H E M a T O L O G Y R E V I E W S — V O L U M E 1 , 2 0

hematology reviews — volume 1, 2009 HEMATOLOGY REVIEWS

is published by PAGEPress Publications. The journal is completely free online at www.pagepress.org/hr Publishing costs are offset by a publication fee charged to authors.

For more information and manuscript submission: www.pagepress.org/hr

Copyright Information All works published in PAGEPress journals are subject to the terms of the Creative Commons Attribution License (http:⁄⁄creativecommons.org/licenses/by-nc/3.0) unless otherwise noted. Copyright is retained by the authors. Any non-commercial reuse is permitted if the original author and source are credited.

Correspondence Our publishing ofﬁces are located in via Giuseppe Belli 4, 27100 Pavia, Italy. Our telephone number is +39.0382.1751762 and our fax number is +39.0382.1750481. E-mail: [email protected]

All PAGEPress journals are Open Access. PAGEPress articles are freely available online and deposited in a public archive immediately upon publication. HEMATOLOGY REVIEWS ISSN 1970-6804 - eISSN 1970-6790

Editor-in-Chief

Giovanni Martinelli, Bologna, Italy

Deputy Editor

Francesca Palandri, Italy

Editorial Board

Giuliana Alimena, Italy David Dingli, USA Muller Fabbri, USA Mario Federico, USA Francesca Gualandi, Italy Jean-Luc Harousseau, France Karl-Anton Kreuzer, Germany Delong Liu, USA Hans E. Johnsen, Denmark Taira Maekawa, Japan Anne F. McGettrick, Ireland Ruben Mesa, USA Markus Raderer, Austria Manuela Schmidinger, Austria Evangelos Terpos, Greece Elisabeth Walsby, UK

Editorial Staff

Nicole Pezzolo, Managing Editor Cristiana Poggi, Production Editor Anne Freckleton, Copy Editor Jeanette Mitchell, Copy Editor Filippo Lossani, Technical Support HEMATOLOGY REVIEWS 2009; Volume 1

Aurora kinase inhibitors: which role in the treatment of chronic myelogenous leukemia patients resistant to imatinib? Giovanni Martinelli, Cristina Papayannidis, Ilaria Iacobucci, Simona Soverini, Daniela Cilloni, Michele Baccarani...... 1

Double versus single high-dose melphalan 200 mg/m2 and autologous stem cell transplantation for multiple myeloma: a region-based study in 484 patients from the Nordic area Bo Björkstrand, Tobias W. Klausen,Kari Remes,3 Astrid Gruber, Lene M. Knudsen, Olav J. Bergmann, Stig Lenhoff, Hans E. Johnsen...... 9

Tuberculosis associated thrombocytopenic purpura: effectiveness of antituberculous therapy Raphael Borie, Claire Fieschi, Eric Oksenhendler, Lionel Galicier...... 14

Cardiotoxicity of tyrosine kinase inhibitors in chronic myelogenous leukemia therapy Zhenshu Xu, Shundong Cang, Ting Yang, Delong Liu...... 17

Bacterial contamination of platelet concentrates: pathogen detection and inactivation methods Dana Védy, Daniel Robert, Danielle Gasparini, Giorgia Canellini, ...... 22

Human heart-type fatty acid-binding protein as an early diagnostic marker of doxorubicin cardiac toxicity Ashraf H. ElGhandour, Manal El Sorady, Sahar Azab, Mohammed ElRahman ...... 29

Palliative splenic irradiation in primary and post pv/et myelofibrosis: outcomes and toxicity of three radiation schedules Mario Federico, Guido Pagnucco, Antonio Russo, Giovanni Cardinale, Patrizia Guerrieri, Francesco Sciumè, Catherine Symonds, Letizia Cito, Sergio Siragusa, Nicola Gebbia, Roberto Lagalla, Massimo Midiri, Antonio Giordano, Paolo Montemaggi ...... 33

MicroRNAs: tiny players with a big role in the pathogenesis of leukemias and lymphomas Francesca Fanini, Ivan Vannini, Muller Fabbri...... 40

Histone deacetylase inhibitors in multiple myeloma Sarah Deleu, Eline Menu, Els Van Valckenborgh, Ben Van Camp, Joanna Fraczek, Isabelle Vande Broek, Vera Rogiers, Karin Vanderkerken ...... 46

The role of JAK2 abnormalities in hematologic neoplasms Mohammed K. Alabdulaali ...... 56

Up-front fludarabine impairs stem cell harvest in multiple myeloma: report from an interim analysis of the NMSG 13/03 randomized placebo controlled phase II trial Hans E. Johnsen, Lene M. Knudsen, Anne K Mylin, Peter Gimsing, Henrik Gregersen, Niels Abildgaard, Nils Frost Andersen, Torben Plesner, Annette Vangsted, Torben Mourits-Andersen, on behalf of the Nordic Myeloma Study Group ...... 62

Maintenance therapy in multiple myeloma Jean-Luc Harousseau ...... 65

[Hematology Reviews 2009; volume 1] Transcriptional regulation of the human ALDH1A1 promoter by the oncogenic homeoprotein TLX1/HOX11 Kim L. Rice, Mansour Heidari, Ross H. Taplin, Ursula R. Kees, Wayne K. Greene ...... 70

Communication between bone marrow niches in normal bone marrow function and during hemopathies progression Sara Lamorte, Leonor Remédio, Sergio Dias ...... 80

A sensitivity comparison of the Quick and Owren prothrombin time methods in oral anticoagulant therapy Juha Horsti ...... 87

Clear cell renal cell carcinoma with vaginal and brainmetastases: a case report and literature review Tobe Momah, Etwaru Dhanan, Phillip Xiao, Vasantha Kondamudi ...... 92

A case of follicular lymphoma complicated with mesenteric panniculitis Yotaro Tamai, Osamu Imataki, Ichiro Ito, Kimihiro Kawakami ...... 94

Massive pleural effusion due to IgG multiple myeloma Kathryn J. Lang, Surjit Lidder, Robin Aitchison ...... 97

The progress of prothrombin time measurement Juha Horsti ...... 99

Confirmation of the validity of using birth MCV for the diagnosis of alpha thalassemia trait Akram M. Al-Hilali, Aisha M. Al-Jallaf, Sajida Chunkasseril...... 103

Best practices for transfusion for patients with sickle cell disease Ted Wun, Kathryn Hassell ...... 106

[Hematology Reviews 2009; volume 1] Spreading information in science

PAGEPress aims to be a truly international organization by providing access to the scientiﬁc literature to anyone, anywhere, by publishing works from every nation, and by engaging a geographically diverse group of scientists in the editorial process.

PAGEPress aims to create a new chapter of the so-called Library of Science using Open Access

PAGEPress is a brand of MeditGroup, an Italian media PAGEPress OpenAccess journals are an ideal platform company which has been developing services for the for the publication of your research enabling you to reach international scientific community since 1992. the widest available audience of professionals in your PAGEPress is now launching a range of peer-reviewed, field of expertise. Publication in our journals means that open access scholarly journals covering several scientific your research articles will be available for immediate fields. free access online. PAGEPress shortens the time needed All material published by PAGEPress, whether submit- before publication, offers a high quality peer review sys- ted to or created by PAGEPress, is published under an tem and submission to many online OpenAccess reposi- open access license that lets others remix, and build tories. upon your work non-commercially, and although their Many researchers are choosing to publish in our new works must also acknowledge you and be non-com- OpenAccess journals thanks to the promise of a quick mercial, they don’t have to license their derivative works and discerning peer review, immediate free access to on the same terms. your research, the wide range and prestige of our jour- PAGEPress strives to set the highest standards of nals, and much much more. Find out more about why excellence in all aspects of its activities, whether this be many authors choose to publish with us and consider the its journal image, its style of presentation, the quality of advantages to be gained. the editorial process at every level, the transparency of All research articles published by PAGEPress are sub- its operations and procedures, its accessibility to the sci- ject to a rigorous peer review system. The Editorial entific community and the public, and its educational Board of each journal is responsible for the form the value. peer review process will take. Find out more about the PAGEPress is committed to a fair and rigorous editori- peer review policy for each of our journals secure in the al process. The decision to accept an article for publica- knowledge that every article you submit to tion will be based purely on scientific quality and impor- a PAGEPress journal will enjoy a speedy and efficient tance. appraisal. PAGEPress welcomes and actively seeks opportunities PAGEPress allows and encourages authors to deposit to work together with any group (scientific/scholarly both their pre- and post-prints in open-access institu- societies, physicians, patient advocacy groups, educational archives or repositories. The primary benefit of tional organizations) and any publisher who shares our self-archiving is reaching a larger audience which commitment to Open Access and to making scientific enhances the visibility and impact of your research. information available for the benefit of science and the public good.

[Hematology Reviews 2009; volume 1] Submit your paper to www.pagepress.org/hr Hematology Reviews 2009; volume 1:e1

Aurora kinase inhibitors: blastic phase (BP). Imatinib (Gleevec, Glivec; STI571), a relatively selective tyrosine kinase Correspondence: Giovanni Martinelli, which role in the treatment of inhibitor that blocks the catalytic activity of Institute of Hematology and Medical Oncology chronic myelogenous leukemia Bcr-Abl, is now the first-line treatment for all "Seràgnoli", University of Bologna, newly diagnosed CML patients. Despite excel- via Massarenti, 9 - 40138 Bologna, Italy patients resistant to imatinib? E-mail: [email protected] lent clinical results, there is still a need to Giovanni Martinelli,1 improve therapy for patients with CML and Key words: chronic myeloid leukemia, Bcr-Abl, + Cristina Papayannidis,1 Ilaria Iacobucci,1 Ph ALL. More than 80% of newly diagnosed imatinib, resistance, mutations, dasatinib, CML patients treated with imatinib in CP Simona Soverini,1 Daniela Cilloni,2 nilotinib, inhibitors. achieve a complete cytogenetic remission, as Michele Baccarani1 typified by the absence of the Philadelphia Funding: this study was supported by European 1Institute of Hematology and Medical chromosome at the examination of 20 bone LeukemiaNet, COFIN 2003 (Molecular therapy Oncology “L. and A. Seràgnoli”, marrow meta-phases.3 However, residual Bcr- of Ph-positive leukemias), FIRB 2001, A.I.L., University of Bologna, Italy Abl transcripts persist in the majority of these A.I.R.C., Fondazione del Monte di Bologna e 2Division of Hematology and Internal patients, as assessed by sensitive assays such Ravenna, Italy. Medicine, Department of Clinical and as nested reverse transcription-polymerase Received for publication: 16 December 2008. chain reaction, and represent the potential Biological Science, University of Turin, Revision received: 13 January 2009. Italy pool from which disease recurrence may orig- Accepted for publication: 2 March 2009. inate. While responses in CML in CP patients have been shown to last more than five years,3 This work is licensed under a Creative Commons most responding patients with AP- and BP- Attribution 3.0 License (by-nc 3.0). CML, as well as those with Ph+ ALL, relapse Abstract ©Copyright Giovanni Martinelli et al., 2009 early despite continued therapy. Resistance to Licensee PAGEPress, Italy At present, there are no compounds in clin- imatinib is most commonly mediated by Abl Hematology Reviews 2009; 1:e1 4 ical development in the field of chronic kinase domain mutations. We and other doi:10.4081/hr.2009.e1 myeloid leukemia (CML) or Philadelphia-posi- authors have reported that approximately half tive (Ph+) acute lymphoblastic leukemia (ALL) CML patients have evidence of point mutations within the Abl kinase domain at the time that have been documented to harbor signifi- not only to imatinib therapy but also to all of cant activity against the imatinib-resistant of resistance to imatinib. Mutations target critical contact points between imatinib and the newly developed tyrosine kinase T315I mutation. Recent reports on the pre- inhibitors entered in clinical trials. Co-crystal clinical activity of some emerging tyrosine Bcr-Abl or, more often, induce a conformation 5 structure analysis indicates that, on binding, kinase inhibitors such as ON012380, VX-680 to which imatinib is unable to bind. In the remaining patients, the reasons for imatinib the hydroxyl group of threonine 315 forms a and PHA-739358 promise possible clinical effi- 10 resistance have to be traced to Bcr-Abl gene crucial hydrogen bond with imatinib. cacy against this specific Bcr-Abl mutant form. Moreover, the side chain of threonine also Here, we focus on the role of aurora kinase amplification or overexpression, clonal cytoge- sterically controls the binding of the inhibitor inhibitor VX-680 and PHA-739358 in blocking netic evolution, or altered levels of transport to hydrophobic regions adjacent to the ATP- the leukemogenic pathways driven by wild- molecules responsible for imatinib influx and 4 binding site.11 In 10-15% of imatinib-resistant type and T315I-Bcr-Abl in CML or Ph+ ALL by efflux (ABC transporters, hOCT1). patients, especially those in more advanced reviewing recent research evidence. We also Abl mutations are at present the most phases of disease, a threonine to isoleucine discuss the possibility of employing aurora extensively investigated and best character- amino acid substitution may be observed. kinase inhibitors as a promising new thera- ized mechanism of resistance to imatinib. So The T315I abrogates imatinib binding peutic approach in the treatment of CML and far, at least 90 different point mutations have because it disrupts the above mentioned Ph+ ALL patients resistant to first and second been isolated from relapsed CML patients who 6-7 generation TK inhibitors. are resistant to imatinib. The clinical and hydrogen bond and introduces a bulkier pathogenetic impact of mutations varies isoleucine side chain into the gatekeeper according to their different degree of residual position.12 However, this explanation is not sensitivity to imatinib. Indeed, while certain the most up-to-date. In fact, as recently Introduction Bcr-Abl mutations retain in vitro sensitivity to demonstrated, the T315I resistance to ima- imatinib at physiologically relevant concen- tinib mainly results from the breakdown of trations and therefore may not be clinically The molecular signature of chronic myeloid interactions between imatinib and both E286 meaningful, others require increased doses of 13 leukemia (CML) and Philadelphia-positive and M290. As a result, biochemical and cel- imatinib, and some confer a highly resistant (Ph+) acute lymphoblastic leukemia (ALL) is lular IC50 values of imatinib for the T315I-Bcr- phenotype (Table 1).9 the Bcr-Abl hybrid gene, originating from a Abl have been shown to be >6400 times high- reciprocal t(9;22) chromosomal translocation er than those of wild-type Bcr-Abl (Table 1).9 on the 22q- derivative, commonly referred to Some authors have suggested that the T315I as the Philadelphia chromosome.1 The result- The T315I mutation is highly is associated with highly aggressive disease ing fusion protein, Bcr-Abl, displays deregulat- phenotype and poor outcome if no timely ed tyrosine kinase activity and drives CML.2 resistant to imatinib therapeutic reassessment is made.14,15 The disease begins with an indolent chronic However, the effects of the T315I mutation on phase (CP) marked by the expansion of An amino acid substitution occurring at kinase activity in vitro and transforming effi- myeloid cells with normal differentiation, and the so-called “gatekeeper” residue, i.e. ciency of Bcr-Abl in vitro and in vivo have then inexorably proceeds to advanced phases, threonine 315, has attracted particular inter- been very recently investigated, suggesting i.e., accelerated phase (AP) and the terminal est since it confers a high level of resistance that in the absence of imatinib, there is nei-

[Hematology Reviews 2009; 1:e1] [page 1] Review ther increased kinase activity nor any growth F317L, M351T, E355G, F359V, H396R, F486S).19 ty against imatinib-resistant Bcr-Abl mutants, advantage for cells carrying T315I-Bcr-Abl as The only imatinib-resistant Bcr-Abl isoform including the T315I, in biochemical and cellu- compared to wild-type Bcr-Abl.5 that was clearly insensitive to dasatinib was lar assays.33 the T315I mutant, which retained kinase activity even in the presence of micromolar concentrations of the compound (Table 1).19 The two second-generation Accordingly, imatinib-resistant patients har- Aurora kinases as targets for inhibitors in clinical boring the T315I mutation have been shown cancer development, dasatinib and not to benefit from dasatinib in the recent nilotinib, are ineffective phase I trial.34 Between these new promising drugs, VX- against the T315I mutant Nilotinib is a close relative of imatinib with 680 and PHA-739358, two aurora kinase A, B and C inhibitors, have a leading place. The more than 20-fold improved affinity for wild- To counteract the problem of resistance due aurora kinases are a family of serine/threo- type Bcr-Abl.16 It is highly efficacious in to point mutations, several second-generation nine kinases involved in many cellular func- patients with imatinib-resistant Ph+ CML. In inhibitors have been synthesized and tested tions, including progression through mitosis, vitro experiment with cell lines transformed in pre-clinical assays: nilotinib (AMN107),8,16-18 by regulating spindle formation, chromosome with mutated forms of Bcr-Abl showed IC50 pro- dasatinib (BMS-354825),8,19-23 bosutinib,24 VX- segregation and cytokinesis.35-37 The overex- liferation inhibition for most mutations with 680,21,25 AP23464,26,27 bafetinib,28,29 PD166326, pression of aurora kinases has been reported the exception of the T315I, which remains PD180970 and PD173955,10,30-32 and in many human solid tumors, leading to refractory to nilotinib8 (Table 1). Accordingly, ON012380.33 Two of them are currently being defects in centrosome function, aberrant spin- evaluated in phase II clinical trials – the dual- clinical responses have been observed in dle assembly, misalignment of chromosomes, specificity Src/Abl inhibitor dasatinib and the patients with various imatinib-resistant Bcr- abnormal cytokinesis and genetic instability, imatinib derivative nilotinib. Dasatinib is a Abl mutations but not in patients positive for determining the activation of oncogenic path- 35 novel, dual Src and Abl inhibitor entered in the T315I in the recent phase I trial. ways.38-40 Many authors reported an aberrant clinical trials. It has been shown to be ~300 Despite the pressing need for a clinically expression of the aurora A and B kinases also times more potent than imatinib in Bcr-Abl effective T315I-Bcr-Abl inhibitor, relatively few in leukemia cells, suggesting a potential role of inhibition assays. Excellent results in terms of pre-clinical candidates have been reported. A these molecular targets in the treatment of hematologic and cytogenetic response in CML potential pitfall might be the tendency to CML and ALL.41-42 Aurora kinase function is and Ph+ ALL patients resistant to imatinib screen initially for Abl kinase inhibition mediated by the phosphorylation of several have been reported after dasatinib adminis- rather than for T315I-specific inhibition. A substrates that have important roles in cell tration.34 Pre-clinical studies have demonstrat- promising approach is to design inhibitors division, such as proteins survivin, CENP-A ed that dasatinib is active against at least targeting other regions of Bcr-Abl. For exam- and serine 10 on histone H3.37 The aurora fourteen imatinib-resistant Bcr-Abl mutants ple, ON012380, a putative substrate-competi- kinases range in size from 309 to 403 amino (M244V, G250E, Q252H/R, Y253F/H, E255K/V, tive inhibitor, exhibited low nanomolar activi- acids. They have a C terminal domain that is

Table 1. Comparison between imatinib, dasatinib and nilotinib IC50 values obtained in Ba/F3 cellular proliferation assays. Adapted from [8]. Cellular proliferation Abl variant Imatinib Nilotinib Dasatinib IC50 (nM) Fold-change IC50 (nM) Fold-change IC50 (nM) fold-change Wild-type 260 1 13 1 0.8 1 M244V 2,000 8 38 3 1.3 2 G250E 1,350 5 48 4 1.8 2 Q252H 1,325 5 70 5 3.4 4 Y253F 3,475 13 125 10 1.4 2 Y253H >6,400 >25 450 35 1.3 2 E255K 5,200 20 200 15 5.6 7 E255V >6,400 >25 430 33 11 14 F311L 480 2 23 2 1.3 2 T315I >6,400 >25 >2000 >154 >200 >250 F317L 1,050 4 50 4 7.4 9 M351T 880 3 15 1.2 1.1 1.4 F359V 1,825 7 175 13 2.2 3 L387M 1,000 4 49 423 H396P 850 3 41 3 0.6 0.8 H396R 1,750 7 41 3 1.3 2

[page 2] [Hematology Reviews 2009; 1:e1] Review responsible for regulation of the protein levels KEN motif via proteasomal degradation; a highly con- D-box activating motif served catalytic domain; and a short N-terminal domain that varies in length between the 1 51 131 383 402 kinases and contributes to the differing loca- H2N COOH Aurora A tions of the kinases within cells.43 (Figure 1). The aurora A isotype (also known as aurora, Aurora-2, AIK, AIR-1, AIRK1, AYK1, BTAK, Eg2, 1 76 251 343 H2N COOH Aurora B MmIAK1 and STK15) is widely expressed in proliferating normal tissues, with expression being cell-cycle-dependent and peaking at the 1 8 249 75 G2/M point of the cell cycle. During mitosis, H2N COOH Aurora C the kinase is virtually confined to the spindle poles, where it is needed for centrosome separation and maturation.44 An overexpression of aurora A causes an increase in centrosome Figure 1. Schematic representation of domain organization of aurora kinases. Aurora numbers and aneuploidy,45 leading to the trans- kinases have three domains: the N-terminal and C-terminal domains which contain most of the aurora’s regulatory motifs and the catalytic domain in the central region. The align- formation of mammalian cells and also causes ment of auroras A and B allows the identification of one distantly conserved ‘KEN’ motif, resistance to apoptosis induced by taxol in spanning 11-18 residues. The ‘KEN’ motif acts as a Cdh– dependant anaphase-promoting human cancer cell lines. Moreover, this kinase complex recognition signal. is a key regulatory component of the p53 pathway as its overexpression leads to increased p53 degradation, which facilitates oncogenic transformation.46 Human aurora A has been Table 2. Novel compound aurora-kinase inhibitors in clinical trial development. proposed as a “drugable” target in several TK inhibitor Company Phase Target(s) tumors including pancreatic,46,47 hepatocellular,48 breast,49 nonendometriod,50 and ovarian MK-0457 Merck I-II Aurora, FLT-3, JAK-2 carcinomas,51 gliomas52 and aggressive non- PHA-739358 Nerviano II Aurora 53 Hodgkin’s lymphoma. Aurora B (also known KW-2449 Kyowa I Aurora as Aurora-1, AIM-1, AIK2, AIR-2, AIRK-2, ARK2, Deciphera Decifera I Abl IAL-1 and STK12) activity is required for bipo- lar chromosome orientation and condensa- AS703569 Merck Serono I-II Aurora, Abl, JAK-2 tion.54 Aurora B kinases are chromosomal pas- AZD1152 Astra-Zeneca I-II Aurora senger proteins, which are found in cells in a complex with inner centromere protein (INCENP) and survivin. The overexpression of an aurora B kinase-dead mutant (K-R) causes Table 3. Comparison between the binding affinity of imatinib and of the aurora kinase multiple defects in the mitotic machinery, inhibitors BIRB-796 and MK-0457 for wild-type and drug-resistant Abl variants. including the loss of kinetochore attachment Adapted from [21]. Kd, nanomolar to microtubules and the exit from mitosis without anaphase or cytokinesis.55 Increased Abl variant Imatinib BIRB-796 MK-0457 Aurora B expression correlates with increased Wild-type 2 2,000 20 grade in glioma and colon cancer56-57 and with Q252H 20 4,000 10 anaplasia in thyroid carcinoma.58 Aurora C (also known as AIK3) expression plays a role Y253F 40 2,000 20 in spermatogenesis at the time when cells E255K 100 >10,000 50 assemble the two meiotic spindles and also co- M351T 10 2,000 8 operates with aurora B to regulate mitotic F359V 20 8,000 20 chromosome dynamics in mammalian cells.59 Aurora C overexpression has been detected in T315I 6,000 40 5 tumor cell lines in vitro60 and in biopsy samples H396P 60 >10,000 7 from colorectal carcinoma.61

Table 4. Comparison between the binding affinity of imatinib and of the aurora kinase Novel aurora kinase inhibitors inhibitors PHA-739358 and MK-0457 for wild-type and drug-resistant Abl variants. effective against the T315I- Adapted from [63]. Bcr-Abl WT E255V T315I M351T (P loop) Several compounds have been pre-clinically Imatinib 0.230 0.610 >20.000 0.100 screened for their inhibitory activity against PHA-739358 0.021 0.014 0.005 0.015 aurora kinases (VX680, MLN8054, AZD1152, MK-0457 0.083 0.205 0.085 0.045 R766, R763, PHA-739358, AT9283) and many of

[Hematology Reviews 2009; 1:e1] [page 3] Review them are being tested in clinical phase I/II tri- Tyr412, which is located in the kinase activa- als (Table 2). MK-0457 (VX-680) is a pan-auro- tion loop of Abl and is also active against the Binding mode of VX-680 and ra kinase inhibitor with demonstrated in vitro T315I mutant of Abl, which is resistant to other PHA-739358 to Abl activity against wild-type and mutated Bcr-Abl, ATP competitive inhibitors in the clinic, such including the T315I form, as well as FLT3 and as gleevec, and second generation TK The compound VX-680, developed by Vertex JAK-2.21 Fascinatingly, Carter et al. have found inhibitors. A multicentric phase I/II study, Pharmaceuticals as an inhibitor of the aurora that the aurora kinase inhibitor VX-680, aimed to test PHA-739358 in patients with kinases, is a Y-shaped molecule, with a already in phase I trials, and the p38 inhibitor chronic, accelerated or blast phase CML relaps- N-methyl-piperazine group forming the base BIRB-796, in clinical trials for inflammatory ing on gleevec or c-Abl therapy and preferably or leg of the “Y”, a pyrimidine group at the disease, inhibit the imatinib- and dasatinib- with T315I mutation in Bcr-Abl kinase is ongo- fork, and a methylpyrazole group at one arm resistant T315I-Bcr-Abl with high affinity ing. and a substituted phenyl group at the other (Tables 3 and 4). In fact, contrasting results related to this compound have been published. In particular, BIRB-796 binds with good affinity to T315I-Bcr-Abl (Kd = 40 nM), but has significantly weaker affinity for wild-type and other imatinib-resistant forms of Abl, with Kd values >1 μM.21 In contrast, as reported by other authors, the compound fails to inhibit methylpyrazole the proliferation of cells expressing T315I, sug- N gesting a lack of clinical benefit for patients HN N harboring such a mutation.22 H H pyrimidine N In a recent phase I-II study, MK-0457 was N cyclopropyl shown to be active in patients with T315I phe- N O notype-refractory CML or Ph-positive ALL, with N S phenyl no significant extramedullary toxicity.62 Because of a potential heart safety issue revealed in one patient who experienced QTc VX-680 prolongation, the enrolment on phase II proto- (cyclopropaneN carboxylic acid {4-[4-(4-methyl-piperazin-1- N-methylpiperaziney1)-pyrimidin-2ylsulphanyl]-phenyl}-amide col was halted in November 2007. Further- more, an innovative phase I clinical study of sequential and concomitant treatment with Figure 2. Chemical structure of VX-680 aurora kinase inhibitor. [Reprinted and adapted with permission from http://kinasepro.wordpress.com]. dasatinib and MK-0457 has been conducted, based on the suggestion that such a combina- tory approach would suppress the emergence of T315I and other resistant clones, improving upon the response rate for dasatinib and the durability of response. To date, 3 patients with wild-type chronic myeloid leukemia (CML) or Ph-positive acute lymphoblastic leukemia (ALL) have been enrolled, and this innovative therapeutic combination showed a relevant hematologic activity and a good safety profile. Thr 315 - gatekeeper Thr 315 - gatekeeper PHA-739358 is a small molecule that selectively inhibits the ATP site of Aurora-A (IC50=13 nM) and Aurora-B (IC50=79 nM) kinases.63 DFG motif Starting from the rationale that aurora kinases DFG motif play an important role in mitosis and that the interruption of their function has significant potential in the treatment of cancer, the drug, Pro 396 formulated for intravenous infusion, is being developed for therapeutic use in solid tumors and in patients with Philadelphia positive His 396 leukemias. Interestingly, PHA-739358, when tested against a panel of more than 30 kinases, has shown a strong cross-reactivity with c-Abl Abl: Imatinib Abl: VX-680 (IC50=25 nM). Its inhibitory activity on ABL in cells was confirmed in K562 leukemia cells Figure 3. Structure of Abl domain kinase bound to imatinib (left) and to VX-680 (right). which bear the Philadelphia chromosome [Reprinted and adapted with permission of AACR from: Young MA, et al. Structure of related translocation Bcr-Abl. Furthermore the kinase domain of an imatinib-resistant Abl mutant in complex with the Aurora kinase inhibitor VX-680. Cancer Res 2006 Jan 15;66(2):1007-14]. PHA-739358 inhibits phosphorylation of

[page 4] [Hematology Reviews 2009; 1:e1] Review arm (Figure 2). A recent study25 showed that seem to cause any widespread conformational D), although the conformation of the proteins VX-680 forms a hydrogen bond with the strict- changes but creates a steric hindrance that around the ATP-binding site shows some dif- ly conserved Asp381 of the Asp-Phe-Gly (DFG) would interfere with the binding of inhibitors, ferences because in the aurora A structure the motif in the Abl kinase domain and maintains such as imatinib, nilotinib, and dasatinib, DFG motif is more similar to the ‘‘out’’ confor- it in an orientation close to one that is normal- which make use of the hydrophobic pocket. mation. However, all of the essential contacts ly seen in active kinases, although the activa- The binding mode of PHA-739358 is very sim- between PHA-739358 and Abl T315I involve tion loop of Abl is not phosphorylated in this ilar to that reported for the complex of the highly conserved elements. The molecule structure (Figure 3). Furthermore, VX-680 same compound with aurora A (Figure 5B and makes three hydrogen bonds with the protein does not deeply penetrate into the kinase domain as imatinib does and it is anchored to it by four hydrogen bonds. Three of these are formed between two carbonyl groups (Glu316 and Met318) and an amide nitrogen (Met318) in the “hinge region” of the kinase and three nitrogen atoms, one in the linker between the pyrimidine group and the methylpyrazole Leu 210 group, and the other two in the methylpyrazole group. These bonds are a common feature of kinase inhibitors and are independent of the sequence of the kinase.59 Likewise, the fourth hydrogen bond, made by VX-680 to the side Aurora: VX-680 (model) T315I-Bcr-Abl: VX-680 (model) chain of Asp381 of the DFG motif, is to a strictly invariant catalytic residue. Using these four Figure 4. Binding mode of VX-680 to aurora A and T315I-Bcr-Abl. Both leucine210 and 315 anchors, the inhibitor makes contact with 14 isoleucine side chains can be accommodated between the two arms of the “Y” of VX- 680. [Reprinted and adapted with permission of AACR from: Young MA, et al. Structure side chains within the kinase domain, eight of of the kinase domain of an imatinib-resistant Abl mutant in complex with the Aurora which are identical between Abl and aurora. kinase inhibitor VX-680. Cancer Res 2006 Jan 15;66(2):1007-14]. One of the non-conservative substitutions is at the gatekeeper position, where Thr315 in Bcr-Abl is replaced by Leu210 in aurora A kinase (Figure 4). The side chains of AB isoleucine (at position 315 of Bcr-Abl) and I315 leucine (at position 210 of aurora A kinase) K27 can be accommodated readily between the two E266 sides of the “Y” of VX-680. For this reason, VX-

680, in contrast to imatinib, is able to inhibit I315 D381 the kinase activity of both wild-type Bcr-Abl F317 and T315I-Bcr-Abl. To understand the structur- L370 al basis of the capability of PHA-739358 to Y320 bind and inhibit the T315I mutant, the crystal Y395 structure of the inhibitor-protein complex was 63 determined (Figure 5). The protein is in the D typical conformation of active kinases, with F382 I315/L21D the activation loop in the extended DFG ‘‘in’’ DFG motif conformation. Indeed, Asp381 points into the F275 D274 2+ active site and interacts with Mg ion that D381 occupies a position similar to the one usually seen in the structures of kinases in complex H O with ATP. The glycine loop adopts an extended C N N O conformation, in contrast to the other publicly T319/P214 HN available Abl structures where the loop is O more distorted, which could be due to the specific binding mode of our inhibitor. The puri- N fied T315I Abl kinase domain used for crystal- N lization experiments is predominantly phosphorylated on the activation loop at Tyr393, Figure 5. Structure of Abl-T315I-PHA 739358 complex. (A) Ribbon representation of the whereas Tyr253, Tyr257, and Tyr264 are phos- structure of T315I Abl mutant with PHA-739358. The mutated gatekeeper residue phorylated at lower levels. These interactions Ile315 and the activation loop with the phosphorylated residue Tyr393 are highlighted probably stabilize the active conformation of in green. (B) Close-up view of the binding site of PHA-739358 showing the final 2 Fo- Fc electron density map, contoured at 1 j, associated with the ligand. (C) Chemical for- the activation loop, which is, however, very mula of PHA 739358. (D) Comparison of PHA-739358 complexes with the aurora A similar to the structures reported for dasatinib structure. Details of the binding of PHA-739358 to Abl (green carbon atoms) and to in complex with the WT Abl kinase domain64 aurora A (yellow carbon atoms) showing the residues of the hinge region and of the DFG and of MK-0457 in complex with the Abl motif of both proteins. [Reprinted and adapted with permission of AACR from: mutant H396P.25 The mutation of the threo- Modugno M, et al. Crystal structure of the T315I Abl mutant in complex with the aurora kinases inhibitor PHA-739358. Cancer Res 2007 Sep 1;67(17):7987-90]. nine to the more bulky isoleucine does not

[Hematology Reviews 2009; 1:e1] [page 5] Review

A B C

Ile315 Thr 315

Phe317 Phe317 Phe317

Figure 6. Comparison of PHA-739358 complex with imatinib and MK-0457 complexes. (A) and (B) PHA-739358 complex (green carbon atoms) superimposed with the structure of imatinib (magenta carbon atoms). The gatekeeper residues Ile315 of PHA-739358 complex (A) and Thr315 of the imatinib complex (B) are shown with van der Waals spheres. In the T315I mutant, the isoleucine side chain causes a steric clash with imatinib; in addition, the hydrogen bond between imatinib and the side chain oxygen of threonine is lost. (C) Structure of Abl-PHA-739358 complex (green carbon atoms) superimposed on the structure of MK-0457. [Reprinted and adapted with permission of AACR from: Modugno M, et al. Crystal structure of the T315I Abl mutant in complex with the aurora kinases inhibitor PHA-739358. Cancer Res 2007 Sep 1;67(17):7987-90].

backbone of the hinge region: the two nitro- date the isoleucine substitution. Further - Bcr-Abl other than the ATP binding pocket. An gen atoms of the pyrrolopyrazole core interact more, the pyrrolopyrazole scaffold of PHA- intriguing alternative is to explore the possi- with the carbonyl oxygen of Glu316 and with 739358 is situated within van der Waals dis- bility of whether molecules that have been the amide nitrogen of Met318, whereas the tance of the side chain of Ile315 mimicking developed as inhibitors for other protein nitrogen of the amide group hydrogen bonds the interaction between the inhibitor and kinases and are already undergoing clinical to the carbonyl oxygen of Met318. In addition, Leu210 in aurora A. It is possible that this trials might include the T315I-Bcr-Abl mutant the side chain nitrogen of the conserved favorable hydrophobic packing interaction among their off-targets. Although off-target Lys271 is within hydrogen bonding distance may explain why PHA-739358 is more active activity may lead to undesirable side effects, it of the oxygen of the carbonyl group and the against the mutant than the WT protein. PHA- has to be recognized that focusing on com- oxygen of the methoxy group. As in the auro- 739358 could represent a valuable novel agent pounds that are already being tested in clinical ra structure, the benzyl group packs against to target the T315I Bcr-Abl mutation, and pre- practice may speed up the development of suc- Leu370 (Leu263 in aurora), whereas the N- clinical and clinical data are coming through cessful therapeutic strategies. Recent studies methyl-piperazine points toward the solvent to support this concept. have shown that MK-0457 (VX-680) and PHA- accessible area of the kinase pocket. The 739358, two small-molecule aurora kinase gatekeeper residue in the aurora kinases is inhibitors, have in vitro activity against the Leu210, a large and hydrophobic residue very T315I-Bcr-Abl. Moreover, preliminary data similar to isoleucine, and we have observed Conclusions showed promising clinical efficacy in patients that PHA-739358 binds in the ATP-binding affected by Philadelphia positive leukemias, pocket of aurora A without any steric hin- The T315I is responsible for approximately relapsing or resistant to first and second gen- drance with the gatekeeper residue. Indeed, 15% of the cases of relapse in CML and Ph+ eration TK inhibitors. Such a remarkable effi- the co-crystal structure reported here reveals ALL patients on imatinib therapy. The clinical cacy raises the question of whether aurora that the compound is bound to the Abl T315I relevance of this mutant is likely to increase kinases may also harbor some pathogenetic kinase domain in a way that accommodates considerably as to date it seems to represent significance in CML and/or Ph+ ALL or may be the substitution of isoleucine for threonine. the main mechanism of resistance to dasa- selectively deregulated by the T315I-Bcr-Abl, Figure 6 shows the structure of the Abl T315I tinib and nilotinib, the second-generation and whether auroras may be a suitable sec- complex with PHA-739358 superimposed on inhibitors already being developed clinically. ondary target for inhibition. those of the Abl WT with imatinib and Abl Structural analyses indicate that the substitu- H396P with MK-0457. In the T315I mutant, tion of threonine with isoleucine at residue the isoleucine side chain causes a steric clash 315 eliminates a crucial hydrogen-bonding References with imatinib and the hydrogen bond between interaction and introduces a steric clash imatinib and the side chain oxygen of threo- which abrogates binding and effective inhibi- 1. Bartram CR, de Klein A, Hagemeijer A, van nine is lost (Figure 6A and B). On the con- tion of Bcr-Abl by imatinib as well as by sever- Agthoven T, Geurts van Kessel A, Bootsma D, trary, both PHA-739358 and MK-0457 bind in al novel inhibitors. A possible approach to the et al. Translocation of c-ab1 oncogene corre- such a way to avoid the gatekeeper residue development of second-line strategies over- lates with the presence of a Philadelphia chromosome in chronic myelocytic leu - (Figure 6C) and this provides an explanation coming resistance induced by the T315I muta- kaemia. Nature 1983;306:277-80. for the ability of both compounds to accommo- tion is to design inhibitors binding regions of 2. Groffen J, Stephenson JR, Heisterkamp N,

[page 6] [Hematology Reviews 2009; 1:e1] Review

Bartram C, de Klein A, Grosveld G. The human Leukemia 2006;20:650-7. abl mutants. Clin Cancer Res 2003;9:1267-73. c-abl oncogene in the Philadelphia transloca- 17. Weisberg E, et al. Characterization of 32. La Rosee P, Corbin AS, Stoffregen EP, tion. J Cell Physiol Suppl 1984;3:179-91. AMN107, a selective inhibitor of native and Deininger MW, Druker BJ. Activity of the Bcr- 3. Druker B, Guilhot F, O'Brien SG, Larson R. mutant Bcr-Abl. Cancer Cell 2005;7:129-41. Abl kinase inhibitor PD180970 against clinical- Long-term benefits of imatinib (IM) for 18. Weisberg E, Manley PW, Breitenstein W, ly relevant Bcr-Abl isoforms that cause resist- patients newly diagnosed with chronic myel- Bruggen J, Cowan-Jacob SW, Ray A, et al. ance to imatinib mesylate (Gleevec, STI571). ogenous leukemia in chronic phase (CML- AMN107 (nilotinib): a novel and selective Cancer Res 2002;62:7149-53. CP): the 5-year update from the IRIS study. inhibitor of BCR-ABL. Br J Cancer 2006;94: 33. Gumireddy K, Baker SJ, Cosenza SC, John P, J Clin Oncol 2006 ASCO Annual Meeting 1765-9. Kang AD, Robell KA, et al. A non-ATP-compet- Proceedings 2006;24:6506. 19. Shah NP, Tran C, Lee FY, Chen P, Norris D, itive inhibitor of BCR-ABL overrides imatinib 4. Hochhaus A, Hughes T. Clinical resistance to Sawyers CL. Overriding imatinib resistance resistance. Proc Natl Acad Sci USA imatinib: mechanisms and implications. with a novel ABL kinase inhibitor. Science 2005;102:1992-7. Hematol Oncol Clin North Am 2004;18:641-56, 2004;305:399-401. 34. Talpaz M, Shah NP, Kantarjian H, Donato N, ix. 20. Burgess MR, Skaggs BJ, Shah NP, Lee FY, Nicoll J, Paquette R, et al. Dasatinib in ima- 5. Soverini S, Colarossi S, Gnani A, et al. Sawyers CL. Comparative analysis of two clin- tinib-resistant Philadelphia chromosome- Frequency, distribution and prognostic value ically active BCR-ABL kinase inhibitors positive leukemias. N Engl J Med 2006;354: of ABL kinase domain (KD) mutations in dif- reveals the role of conformation-specific 2531-41. ferent subsets of Philadelphia-positive binding in resistance. Proc Natl Acad Sci USA, 35. Keen N, Taylor S. Aurora-kinase inhibitors as (Ph+) patients (pts) resistant to Imatinib 2005;102:3395-400. anticancer agents. Nat Rev Cancer 2004;4: (IM) by the GIMEMA Working Party on CML. 21. Carter TA, Wodicka LM, Shah NP, Velasco AM, 927-36. Blood 2005;106:435. Fabian MA, Treiber DK, et al. Inhibition of 36. Marumoto T, Zhang D, Saya H. Aurora-A, a 6. Melo JV, Chuah C. Resistance to imatinib drug-resistant mutants of ABL, KIT, and EGF guardian of poles. Nat Rev Cancer 2005;5:42- mesylate in chronic myeloid leukaemia. receptor kinases. Proc Natl Acad Sci USA 50. Cancer Lett 2007;249:121-32. 2005;102:11011-6. 37. Carvajal RD, Tse A, Schwartz GK. Aurora 7. Branford S. Chronic myeloid leukemia: 22. O'Hare T, Druker BJ. BIRB-796 is not an effec- kinases: new targets for cancer therapy. Clin molecular monitoring in clinical practice. tive ABL(T315I) inhibitor. Nat Biotechnol Cancer Res 2006;12:6869-75. Hematology Am Soc Hematol Educ Program 2005;23:1209-10. 38. Matthews N, Visintin C, Hartzoulakis B, Jarvis 2007;2007:376-83. 23. Tokarski JS, Newitt JA, Chang CY, Cheng JD, A, Selwood DL. Aurora A and B kinases as tar- 8. Golemovic M, Verstovsek S, Giles F, Cortes J, Wittekind M, Kiefer SE, et al. The structure of gets for cancer: will they be selective for Manshouri T, Manley PW, et al. AMN107, a Dasatinib (BMS-354825) bound to activated tumors? Exppert Rev Anticancer Ther novel aminopyrimidine inhibitor of Bcr-Abl, ABL kinase domain elucidates its inhibitory 2006;6:109-20. has in vitro activity against imatinib-resistant activity against imatinib-resistant ABL 39. Wang X, Zhou YX, Qiao W, et al. Over- chronic myeloid leukemia. Clin Cancer Res mutants. Cancer Res 2006;66:5790-7. expression of Aurora kinase A in mouse 2005;11:4941-7. 24. Golas JM, Arndt K, Etienne C, Lucas J, Nardin mammary epithelium induces genetic insta- 9. O' Hare T, Walters DK, Stoffregen EP, Jia T, D, Gibbons J, et al. SKI-606, a 4-anilino-3- bility preceding mammary tumor formation. Manley PW, Mestan J, et al. In vitro activity of quinolinecarbonitrile dual inhibitor of Src Oncogene 2006;25:7148-58. Bcr-Abl inhibitors AMN107 and BMS-354825 and Abl kinases, is a potent antiproliferative 40. Zhou H, Kuang J, Zhong L et al. Tumour ampil- against clinically relevant imatinib-resistant agent against chronic myelogenous leukemia ified kinase STK15/BTAK induces centrosome Abl kinase domain mutants. Cancer Res cells in culture and causes regression of ampilification, aneuploidy and transforma- 2005;65:4500-5. K562 xenografts in nude mice. Cancer Res tion. Nat Genet 1998;20:189-93. 10. Corbin AS, La Rosee P, Stoffregen EP, Druker 2003;63:375-81. 41. Yang J, Ikezoe T, Nishioka C, Tasaka T, BJ, Deininger MW. Several Bcr-Abl kinase 25. Young MA, Shah NP, Chao LH, Seeliger M, Taniguchi A, Kuwayama Y, et al. AZD1152, a domain mutants associated with imatinib Milanov ZV, Biggs WH 3rd, et al. Structure of novel and selective aurora B kinase inhibitor, mesylate resistance remain sensitive to ima- the kinase domain of an imatinib-resistant induces growth arrest, apoptosis, and sensi- tinib. Blood 2003;101:4611-4. Abl mutant in complex with the Aurora kinase tization for tubulin depolymerizing agent or 11. Nagar B, Bornmann WG, Pellicena P, Schindler inhibitor VX-680. Cancer Res 2006;66:1007-14. topoisomerase II inhibitor in human acute T, Veach DR, Miller WT, et al. Crystal struc- 26. O'Hare T,s Pollock R, Stoffregen EP, Keats JA, leukemia cells in vitro and in vivo. Blood tures of the kinase domain of c-Abl in com- Abdullah OM, Moseson EM, et al. Inhibition of 2007;110:2034-40. plex with the small molecule inhibitors wild-type and mutant Bcr-Abl by AP23464, a 42. Ikezoe T, Yang J, Nishioka C, Tasaka T, PD173955 and imatinib (STI-571). Cancer Res potent ATP-based oncogenic protein kinase Taniguchi A, Kuwayama Y, et al. A novel treat- 2002;62:4236-43. inhibitor: implications for CML. Blood 2004; ment strategy targeting Aurora kinases in 12. Gorre ME, Mohammed M, Ellwood K, Hsu N, 104:2532-9. acute myelogenous leukemia. Mol Canc Ther Paquette R, Rao PN, et al. Clinical resistance 27. Azam M, Nardi V, Shakespeare WC, Metcalf CA 2007;6:1851-7. to STI-571 cancer therapy caused by BCR-ABL 3rd, Bohacek RS, Wang Y, et al. Activity of dual 43. Hu HM, Chuang CK, Lee MJ, Tseng TC, Tang gene mutation or amplification. Science SRC-ABL inhibitors highlights the role of TK. Genomic organization, expression, and 2001;293:876-80. BCR/ABL kinase dynamics in drug resistance. chromosome localization of a third Aurora- 13. Lee TS, Potts SJ, Kantarjian H, et al. Proc Natl Acad Sci USA 2006;103:9244-9. related kinase gene, Aie1. DNA Cell Biol Molecular basis explanation for imatinib 28. Kimura S, Naito H, Segawa H, Kuroda J, Yuasa 2000;19:679-88. resistance of BCR-ABL due to T315I and P- T, Sato K, et al. New tyrosine kinase inhibitors 44. Kufer TA, Nigg EA, Sillje HN. Regulation of loop mutations from molecular dynamics in the treatment of chronic myeloid leukemia. Aurora-A kinase on the mitotic spindle. simulations. Cancer 2008;112:1744-53. Curr Pharm Biotechnol 2006;7:371-9. Chromo-soma 2003;112:159-63. 14. Pricl S, Fermeglia M, Ferrone M, Tamborini E. 29. Naito H, Kimura S, Nakaya Y, Naruoka H, 45. Meraldi P, Honda R, Nigg EA. Aurora-A overex- T315I-mutated Bcr-Abl in chronic myeloid Kimura S, Ito S, et al. In vivo antiproliferative pression reveals tetraploidization as a major leukemia and imatinib: insights from a com- effect of NS-187, a dual Bcr-Abl/Lyn tyrosine route to centrososme amplification in p53-/- putational study. Mol Cancer Ther 2005; kinase inhibitor, on leukemic cells harbouring cells. EMBO J 2002;21:483-92. 4:1167-74. Abl kinase domain mutations. Leuk Res 2006; 46. Rojanala S, Han H, Munoz RM, Browne W, 15. Nicolini FE, Corm S, Le QH, Sorel N, Hayette 30:1443-6. Nagle R, Von Hoff DD, et al. The mitotic ser- S, Bories D, et al. Mutation status and clinical 30. Dorsey JF, Jove R, Kraker AJ, Wu J. The pyri- ine threonine kinase, Aurora-2, is a potential outcome of 89 imatinib mesylate-resistant do[2,3-d] pyrimidine derivative PD180970 target for drug development in human pan- chronic myelogenous leukemia patients: a inhibits p210Bcr-Abl tyrosine kinase and creatic cancer. Mol Cancer Ther 2004;3:451-7. retrospective analysis from the French inter- induces apoptosis of K562 leukemic cells. 47. Li D, Zhu J, Firozi PF, Abbruzzese JL, Evans group of CML (Fi(phi)-LMC GROUP). Leu- Cancer Res 2000;60:3127-31. DB, Cleary K, et al. Overexpression of onco- kemia 2006;20:1061-6. 31. Huron DR, Gorre ME, Kraker AJ, Sawyers CL, genic STK15/BTAK/Aurora A kinase in human 16. Miething C, Feihl S, Mugler C, Grundler R, von Rosen N, Moasser MM, et al. A novel pyri- pancreatic cancer. Clin Cancer Res 2003;9: Bubnoff N, Lordick F, et al. The Bcr-Abl muta- dopyrimidine inhibitor of abl kinase is a pico- 991-7. tions T315I and Y253H do not confer a growth molar inhibitor of Bcr-abl-driven K562 cells 48. Jeng YM, Peng SY, Lin CY, Hsu HC. Over- advantage in the absence of imatinib. and is effective against STI571-resistant Bcr- expression and amplification of Aurora-A in

[Hematology Reviews 2009; 1:e1] [page 7] Review

hepatocellular carcinoma. Clin. Cancer Res non-Hodgkin’s lymphoma. Br J Haematol mitotic cells. Cell Motility and the Cyto - 2004;10:2065-71. 2003;121:439-47. skeleton 2004;59:249-63. 49. Hoque A, Carter J, Xia W, Hung MC, Sahin AA, 54. Dewar H, Tanaka K, Nasmythm K, Tanaka TU. 60. Kimura M, Matsuda Y, Yoshioka T, Okano Y. Sen S, et al. Loss of Aurora A/STK15/BTAK Tension between two kinetochores suffices Cell cycle-dependent expression and centro- overexpression correlates with transition of for their bi-orientation on the mitotic spin- some localization of a third human in situ to invasive ductal carcinoma of the dle. Nature 2004;428:93-7. aurora/Ipl1-related protein kinase, AIK3. breast. Cancer Epidemiol. Biomarkers Prev 55. Murata-Hori M, Wang Y. The kinase activity of J Biol Chem 1999;274:7334-40. 2003;12:1518-22. aurora B is required for kinetochore-micro- 50. Moreno-Bueno G, Sanchez-Estevez C, Cassia tubule interactions during mitosis. Current 61. Takahashi T, Futamura M, Yoshimi N, Sano J, R, Rodriguez-Perales S, Diaz-Uriarte R, Biology 2002;12:894-99. Katada M, Takagi Y, et al. Centrosomal kinas- Dominguez O, et al. Differential gene expres- 56. Araki K, Nozaki K, Ueba T, Tatsuka M, Hashi - es, HsAIRK1 and HsAIRK3, are overexpressed sion profile in endometrioid and nonen- moto N. High expression of Aurora-B/Aurora in primary colorectal cancers. Jpn J Cancer dometrioid endometrial carcinoma: STK15 is and Ipll-like midbody-associated protein Res 2000;91:1007-14. frequently overexpressed and amplified in (AIM-1) in astrocytomas. J Neuro Oncol 62. Giles FJ, Cortes J, Jones D, Bergstrom D, nonendometrioid carcinomas. Cancer Res 2004;67:53-64. Kantarjian H, Freedman SJ. MK-0457, a novel 2003;63:5697-702. 57. Katayama H, Ota T, Jisaki F, Ueda Y, Tanaka T, kinase inhibitor, is active in patients with 51. Gritsko TM, Coppola D, Paciga JE, Yang L, Sun Odashima S, et al. Mitotic kinase expression chronic myeloid leukemia or acute lympho- M, Shelley SA, et al. Activation and overex- and colorectal cancer progression. J Natl cytic leukemia with the T315I BCR-ABL muta- pression of centrosome kinase BTAK/Aurora- Cancer Inst 1999;91:1160-2. tion. Blood 2007;109:500-2. A in human ovarian cancer. Clin. Cancer Res 58. Sorrentino R, Libertini S, Pallante PL, 63. Modugno M, Casale E, Soncini C, Rosettani P, 2003;9:1420-6. Troncone G, Palombini L, Bavetsias V, et al. Colombo R, et al. Crystal structure of the 52. Reichardt W, Jung V, Brunner C, Klein A, Aurora B overexpression associates with the T315I Abl mutant in complex with the aurora Wemmert S, Romeike BF, et al. The putative thyroid carcinoma undifferentiated pheno- serine/threonine kinase gene STK15 on chro- type and is required for thyroid carcinoma kinases inhibitor PHA-739358. Cancer Res mosome 20q13.2 is amplified in human cell proliferation. J Clin Endocrinol Metab 2007;67:7987-90. gliomas. Oncol Rep 2003;10:1275-9. 2005; 90:928-35. 64. Tokarski JS, Newitt JA, Chang CYJ, et al. The 53. Hamada M, Yakushijin Y, Ohtsuka M, Kaki- 59. Sasai K, Katayama H, Stenoien DL, Fujii S, structure of dasatinib (BMS-354825) bound moto M, Yasukawa M, Fujita S. Auro - Honda R, Kimura M, et al. Aurora-C kinase is a to activated ABL kinase domain elucidates its ra2/BTAK/STK15 is involved in cell cycle novel chromosomal passenger protein that inhibitory activity against imatinibresistant checkpoint and cell survival of aggressive can complement Aurora B kinase function in ABL mutants. Cancer Res 2006;66:5790-7.

[page 8] [Hematology Reviews 2009; 1:e1] Hematology Reviews 2009; volume 1:e2

Double versus single high-dose group and 64% in the doubletransplant 2 (p=0.9). In a multivariate analysis of variables, Correspondence: Hans E. Johnsen, melphalan 200 mg/m including single versus double transplantation, Professor, Clinical Haematology, Ålborg Hospital and autologous stem cell β2 microglobulin level, age, sex and disease Science and Innovation Center (AHSIC), Århus stage, only β2 microglobulin level was predic- University Hospital, Ålborg, Denmark transplantation for multiple E-mail: [email protected] myeloma: a region-based tive for overall survival (p>0.0001) and progression free survival (p=0.001). In accor- Key words: high-dose melphalan, double trans- study in 484 patients from dance with these results, a 1:1 case-control plantation. the Nordic area matched comparison between double and single transplantation did not identify significant Acknowledgments: the scientific program was sup- Bo Björkstrand,1 Tobias W. Klausen,2 differences in overall and progression free sur- ported by grants from the Danish Cancer Society (grant no DP 07014, DR 07017), the Danish Kari Remes,3 Astrid Gruber,4 vival. In this retrospective analysis up front double Research Council (grant no. 99 00 771, no. 271-05- Lene M. Knudsen,2 Olav J. Bergmann,2 0286, no. 271-05-0537, 22-00-0314), the Multiple transplantation with melphalan (200 mg/m2) Stig Lenhoff,5 Hans E. Johnsen2,6 Myeloma Research Foundation (senior grant as compared to single transplantation did not 2003-4, contract no. 14), EU 6th FP (grant no Departments of Hematology at seem to improve the final outcome among LSHC-CT-2006-037602 to MSCNET coordinator HE 1Huddinge University Hospital, Sweden, patients in the Nordic area. These data are in Johnsen) and by unrestricted research grants 2Copenhagen University Hospital, accordance with recent publications from the from the Nordic Cancer Union, Amgen, Roche, and Herlev, Denmark Bologna 96 trial indicating that a second trans- Schering-Plough in Denmark, Norway, and Sweden. The Nordic Myeloma Study Group 3Department of Medicine, University plant should not be recommended up front as standard care. (NMSG) is the umbrella organization in which Central Hospital, Turku, Finland forum discussions and decisions about trial design 4Karolinska Hospital, Karolinska Institute, were approved. The participants in the Nordic Stockholm, Sweden Myeloma Study Group are listed at http://www. 5Lund University Hospital, Lund, Sweden nordic-myeloma.org/. The Myeloma Stem Cell Introduction Network is a consortium of 9 partners within the 6Ålborg Hospital, University of Århus European Myeloma Network (EMN) collaborating Denmark for the Nordic Myeloma Study Multiple myeloma (MM) is the second most on studies of cancer stem cells in Multiple Group, Denmark Myeloma. The partners in EMN and MDCNET are common hematologic cancer after non- listed at http://www.myeloma-europe.org/ Hodgkin's lymphoma. More than 50,000 patients in Europe alone have MM and about Contributions: all authors listed below have Abstract half of these patients are diagnosed when approved submission for publication. Further they are younger than 65 years of age, and information about the contributions of each part- increasingly detected under the age of 40. ner who participated in this collaborative project Autologous stem cell transplantation is still Today, MM is the most common indication is given below. BB assisted in study design and considered the standard of care in young for high-dose therapy supported by hemato- patient data; TWK assisted in data collection, analysis and interpretation; KR, AG, LMK, OJB patients with multiple myeloma (MM). This poietic stem cell transplantation in the world, disease is the most common indication for and SL assisted in study design and patient data; and more data support the benefit of this pro- HEJ is corresponding author/guarantor and co- high-dose therapy (HDT) supported by cedure. These remarkable results radically ordinator, assisted in study design, provision of hematopoietic stem cell transplantation and altered the disease management in patients biomaterial and patient data, data collection, much data support the benefit of this proce- below 65 years of age. Thus, stem cell trans- analysis and interpretation, and manuscript dure. Results of randomized studies are in plantation has been recommended for these preparation. favor of tandem autologous transplantation patients as part of the initial therapy or at the Received for publication: 10 January 2009 although the effect on overall survival is time of disease progression. However, the unclear. Based on sequential registration trials Revision received: 10 February 2009 median duration of response is short and Accepted for publication: 9 March 2009 in the Nordic area, we aimed to evaluate the almost all patients ultimately relapse.1-9 outcome of conventional single or double HDT. The InterGroupe Francophone du Myélome This work is licensed under a Creative Commons During 1994-2000 we registered a total of 484 (IFM) took the next logical step and asked if Attribution 3.0 License (by-nc 3.0). previously untreated patients under the age of the combination of two cycles of high-dose 60 years at diagnosis who on a regional basis therapy and hematopoietic stem cell rescue ©Copyright Hans E. Johnsen et al., 2009 initially were treated with single [Trial NMSG might improve survival. The group assigned Licensee PAGEPress, Italy #5/94 and #7/98 (N=383)] or double [Trial Hematology Reviews 2009; 1:e2 399 patients with untreated MM below 60 doi:10.4081/hr.2009.e2 Huddinge Karolinska Turku Herlev (N=101)] years of age to receive VAD (vincristine, adri- high-dose melphalan (200 mg/m2) therapy sup- amycin and dexamethasone) as induction ported by autologous stem cell transplantation. therapy and afterwards assigned these A complete or very good partial response was patients randomly to single or double trans- double transplant benefit that had not been achieved by 40% of patients in the single transplantation conditioned by melphalan 140 evident in early analyses.10 This important plant group and 60% of patients in the double mg/m2 and TBI (standard single transplanta- study demonstrates that double transplanta- transplant group (p=0.0006). The probability of tion) or melphalan 140 mg/m2 followed by 140 tion is one of the options for treating myelo- surviving progression free for five years after mg/m2 and TBI (experimental double trans- ma, particularly those younger than 60 years the diagnosis was 25% (95% CL 18-32%) in the plantation).10 Following a median observation of age who have a suboptimal response to a singletransplant group and 46% (95% CL 33- of approximately six years, response rates in single transplant. 55%) in the double transplant group the two groups were not significantly differ- However, the IFM study has raised a num- (p=0.0014). The estimated overall five-year ent, but the probabilities of event free survival ber of questions. First and most important, do survival rate was 60% in the single transplant and overall survival were prolonged with a the beneficial responses in the double trans-

[Hematology Reviews 2009; 1:e2] [page 9] Article plant group reflect the higher total dose of therapy supported by autologous stem cell ered ineligible for the induction therapy. The melphalan? In other words, is a single trans- transplantation. The aim was to evaluate the treatment was divided into 4 phases: (I) induc- plant with the use of maximally tolerated outcome of conventional single or double HDT. tion therapy; (II) peripheral blood stem cell doses of melphalan (200 mg/m2) as effective harvest; (III) high-dose therapy with single or Double transplant study population: HKTH as a double transplantation strategy with the double high-dose melphalan 200 mg/m2 given From June 1994-June 2000, 101 patients high dose of melphalan administrated twice? as a single dose intravenously, followed by with newly diagnosed myeloma <60 years Second, what should be recommended, as stem cell transplantation and (IV) follow-up were entered into a phase II trial evaluating only one of several phase II-III studies has (described in details in 1). Patients with pro- double high-dose melphalan (200 mg/m2) documented an effect on overall survival?11,12 gressive disease or with emerging contraindi- Without doubt the IFM study has to be con- therapy with autologous stem cell support. This included patients from Huddinge and cations to phases II to III were taken off the sidered the Proof of Principle but in the light treatment protocol. of the study design, as well as the overall Karolinska Hospitals in Stockholm, Sweden, results from other studies, it is still unknown Turku University Hospital in Finland, and Diagnostic criteria if a second transplant should be recommend- from June 1997 Herlev University Hospital, The diagnosis of MM was accepted if crite- ed in all cases, even if the response to the first Copenhagen in Denmark. This trial covered a ria A+C, A+D, or B+C+D of the following were transplant has been inferior. population of 3 million. The number of new fulfilled: (A) serum monoclonal component In this unclear situation we now see alter- cases of myeloma <60 years in this population (M-protein) concentration of immunoglobulin native progress in the treatment of MM by during the study period was estimated to be (Ig)G > 30 g/L, IgA > 20 g/L, the presence of an new drugs currently being analyzed in ran- 200 patients. M-protein of IgD or IgE regardless of concen- domized trials. In the near future, ongoing Single transplant population NMSG tration, or Bence-Jones proteinuria > 1 g/24 h; studies will clarify the role of these novel #5/941 and NMSG #7/9814 NMSG #5/94 (B) M-protein in serum or urine at a lower agents, including thalidomide and its analogs, From March 1994 until June 1997, 122 concentration than described under A; (C) at and bortezomib etc., in the context of autolo- Swedish patients with newly diagnosed myelo- least 10% plasma cells in bone marrow aspirate gous stem cell transplantation. However, trial ma <60 years were entered into NMSG #5/94 or biopsy-verified plasmacytoma of bone or soft designs including consolidation therapy such trial evaluating one cycle of high-dose melpha- tissue; and (D) osteolytic bone lesions. Only as that planned by the NMSG may be ham- lan therapy with autologous stem cell support. patients with symptomatic disease were regis- pered by a double autologous transplantation One hundred and seven of these were treated tered. strategy, not yet documented to have an effect according to the specified treatment protocol Criteria for response on survival. Here the Nordic group reports the and received single transplantation. This trial Complete response was defined as the dis- data analysis of a total of 484 MM patients covered a population of 15 million. A total of appearance of M-protein from serum and urine transplanted from 1994-2000 including double 348 Nordic patients were reported to the study in agarose gel electrophoresis and < 5% plas- transplantation of 101 patients. The conclu- secretariat. The expected number of new cases ma cells in a bone marrow aspirate. Very good sions are based on results from two sequen- of myeloma <60 years in this population dur- partial remission (VGPR) was 90-99% reduc- tial phase II trials evaluating double trans- ing the study period was estimated to be 450. plantation in 4 selected centers (Huddinge, In this trial, a highly significant survival tion of M-protein. Partial response was defined Karolinska, Turku and Herlev) by comparing advantage was found for high-dose melphalan by at least a 50% reduction of the initial serum the outcome with data from 383 single trans- therapy, with a prolongation of the median sur- M-protein concentration and a reduction of planted patients included in trial NMSG #5/94 vival from 44 to 62 months.1 Bence-Jones proteinuria to <0.2 g/24 h. Minor and #7/98 from the other centers.1,13,14 response was defined by a 25-50% reduction of NMSG #7/98 the initial serum M-protein concentration and From January 1998 until June 2000, 452 a reduction in Bence-Jones proteinuria by at patients <65 years were registered in a similar least 50% but exceeding 0.2 g/24 h. Design and Methods trial evaluating high-dose melphalan with No statistically significant differences were autologous stem cell support, and using a observed between the groups at any stage of Approval and patient eligibility matched historical patient group as control. Of the treatment with regard to these compar- The scientific protocols were reviewed and these, 276 Swedish, Norwegian and Danish isons. To fulfill the criteria for complete, par- approved by the regional ethics committees in patients aged <65 years were treated accord- tial, or minor response, the patients were not Denmark, Sweden, Finland and Norway, and ing to the specified treatment protocol and allowed to have any other signs of myeloma all patients gave written informed consent received single transplantation. This trial also progression, such as persisting hypercalcemia before study entry. Patients less than 60 years covered a population of 15 million. A total of or progressive renal insufficiency, skeletal of age who had Durie-Salmon stage I with at 452 patients were reported to the study secre- disease, or bone marrow insufficiency due to least one bone lesion, II, or III myeloma were tariat. The number of expected new cases of plasma cell infiltration. Progression was eligible. The criteria for exclusion were prior myeloma <65 years in this population during defined by a confirmed increase in the serum treatment for myeloma, another cancer, the study period was estimated to be 580. The abnormal cardiac function, chronic respirato- M-protein concentration by more than 25% main purpose of this study was to evaluate the from the level at the time of best response, an ry disease, abnormal liver function or psychi- impact of high-dose melphalan therapy in increase of Bence-Jones proteinuria to more atric disease. patients aged 60-64 years in a population- than 1.0 g/24 h, or other unequivocal signs of based study. Age was found to influence out- disease progression, such as hypercalcemia, Design and aims of the program come after intensive therapy, which, however, progressive skeletal disease, or soft-tissue This study was planned to include previous- prolonged survival but with less superiority ly untreated patients under the age of 60 years plasmacytoma. Progression, death without than in younger patients.14 at diagnosis who on a regional basis initially progression, and occurrence of a secondary were treated with single [Trial NMSG #5/94 Treatment plan malignancy were all considered as events. and #7/98 (N=383)] or double [Trial HKTH All patients could be treated according to the Event free and overall survival was calculated (N=101)] high-dose melphalan (200 mg/m2) protocol provided that they were not consid- from the start of therapy.

[page 10] [Hematology Reviews 2009; 1:e2] Article

Follow-up evaluation plantation procedure. In the double transplant (Figure 1b), respectively. Of the 33 deaths in All registered patients were followed until group, the median follow-up was 48 months this group, 61% were attributed to myeloma, death. Patients were evaluated before the start (range, 10-108) from the time of transplanta- while 3% were related to the transplantation of phase II and phase III, and thereafter every tion. The median durations of event free, pro- procedure. As compared with single transplan- sixth week. gression free, and overall survival were 46, 46, tation, double transplantation improved pro- and 76 months, respectively. The estimated gression free survival (p=0.001) (Figure 1b) Statistical analysis probabilities of event-free, progression free, whereas overall survival was similar (p=0.9) All analyses were performed on treatment and overall survival five years after the inclu- (Figure 1a). received basis and are not an intention-to- sion were 44% (Figure 1a), 45%, and 64% treat study. The proportions of patients with a given characteristic were compared using Fisher's exact test for variables with frequency scale and Wilcoxon rank-sum test for the remaining variables. Event free and overall survival rates were calculated according to the Kaplan-Meier method, and survival comparisons between groups were made by the log- a 1.0 b 1.0 rank test. The Cox proportional hazards DTx DTx STx STx regression model was used to estimate the 0.8 p = 0.9 0.8 p = 0.001 prognostic importance of different variables. Age, bone marrow plasma cells, blood hemoglobin, serum calcium, serum creatinine, blood 0.6 0.6 platelets, and serum albumin were included as continuous variables. The following variables 0.4 0.4 were dichotomized: sex (male vs. female), stage according to Durie and Salmon (I or II vs. Proportion surviving III), M-protein class (IgG vs. other; IgA vs. 0.2 Proportion progression-free 0.2 other; light chains only vs. other), and osteolytic bone lesions (none vs. limited or advanced). 0.0 0.0 In the multivariate analyses, forward stepwise 0 24 48 72 96 0 24 48 72 96 variable selection was used. Survival from inclusion Time to progression from ASCT

N. at risk N. at risk DTx 101 85 40 91DTx 101 75 29 71 Stx 383 289 78 26 2 Stx 383 200 37 12 1 Results Figure 1. Overall survival (a) and progression free survival (b) following double transplantation in trial HKTH (N=101), compared to single transplantation in NMSG #5/94 Baseline characteristics at diagnosis plus #7/98 (N=384). The estimated probabilities are shown for double (DTx) and single Table 1 shows the base-line clinical and (STx) transplantation. Tables below the graph indicate patients at risk for the estimate. demographic characteristics of the 484 patients entering this analysis. There were no differences between the treatment groups but for disease stage. Table 1. Base-line characteristics of the patients according to treatment group. Variable Double transplantation Single transplantation p-value Response rates in the collaborative trials N 101 383 Following the final preparative treatment Age* 55 years 54 years 0.4# with high-dose melphalan and autologous stem Sex (female, male) 44/57; (44 and 56%) 151/232; (39 and 61%) 0.5§ cell transplantation, the overall rates of com- β2 microglobulin* 2.8 mg/L 3.4 mg/L 0.1# plete or very good partial response for patients Stage I/II/III 12/23/64; (12, 23 and 65%) 16/108/259; (4, 28 and 67%) 0.015§ who actually received a single or a double trans- *Median values; #Mann-Whitney test; §Fisher’s exact test. plant (Table 2) were in the collaborative trials 60% and 40%, respectively (p=0.0006).

The survival outcome Table 2. Comparison between double and single transplantation. In the single transplant group, the median Variable Double transplantation Single transplantation p-value follow-up was 34 months (range, 12-115) from the time of transplantation. The median dura- N (data/missing) 101 (98/3) 383 tions of event free, progression free, and over- CR 56 (57%) 139 (36%) 0.003§ all survival were 33, 33, and 78 months, respec- CR + VGPR 59 (60%) 155 (40%) 0.0006§ tively. The estimated probabilities of event- OS % censored See Figure 1a See Figure 1a 0.9* free, progression free, and overall survival five OS median (years) 6.3 6.5 years after inclusion were 25% (Figure 1a), PFS % censored See Figure 1b See Figure 1b 0.0014* 26%, and 60% (Figure 1b), respectively. Of the PFS median (years) 3.8 2.7 96 deaths in this group, 85% were attributed to progressive myeloma, 2% related to the trans- #Mann-Whitney test; §Fisher’s exact test; *Log rank test (See Figure 1a and b)

[Hematology Reviews 2009; 1:e2] [page 11] Article

Analysis for prognostic variables and have not been included in this analysis. due to double transplantation was a major In a multivariate analysis of all 484 patients This number is in accordance with the litera- concern. However, the hematopoietic recon- (Table 3), overall survival was significantly relat- ture.10 The most common reasons were a deci- stitution was similar after one or two trans- ed to baseline serum levels of β2 microglobulin sion for allogeneic transplantation, poor per- plantations13 but the rates of death caused by (p<0.0001) but not to the maximal response to formance status, and poor stem cell collection toxic effects were increased in the IFM study. treatment (p=0.2), not to age (p=0.5), disease owing to an insufficient response after the Analysis of the present phase II trials after stage or treatment assignment (p=0.4). initial VAD treatment. a median follow-up period of 3-4 years docu- The risk of life-threatening toxic effects mented significant improved response rates Case control study comparing double and single transplantation To illustrate the impact of β2 microglobulin, each double transplanted patient from the study group was matched with one case of a Table 3. Statistical analysis of variables on overall (OS) and progression free survival. single transplanted patient (N=101) from the OS Univariat Multivariat corresponding NMSG database, according to Variables RR (95% CI) p-value RR (95% CI) p-value β2 microglobulin, age, sex and disease stage in this order. The results are shown in Figure Single vs. double transplantation 1.0 (0.7-1.5) 0.9 0.8 (0.5-1.3) 0.4 2 and document no survival benefit. (Log) β2 microglobulin 1.8 (1.5-2.2) < 0.0001 1.8 (1.4-2.2) < 0.0001 Age 1.02 (0.99-1.05) 0.3 1.02 (0.99-1.05) 0.3 Treatment-related toxicity Sex (Male vs. female) 1.3 (0.9-1.8) 0.1 1.2 (0.8-1.8) 0.4 The hematopoietic reconstitution was recognized to be similar in the two groups as Stage I, II or III - 0.3 NI NI expected.13 There were 2 (0.5%) treatment- Stage (II, III vs. I) 2.1 (0.8-5.6) 0.1 1.5 (0.5-5.0) 0.4 related deaths in the single transplant group of Response non-CR vs. CR 1.3 (0.9-1.8) 0.2 NI NI 383 patients and 3 (3.0%) in the double transplant group of 101 patients (p=0.06). PFS Univariat Multivariat (N=380) Minimal residual disease Variables RR (95% CI) p-value RR (95% CI) p-value In the double transplant group, the number Single vs. double transplantation 1.7 (1.2-2.3) 0.0009 1.3 (0.9-1.9) 0.1 of bone marrow malignant plasma cells 2-3 (Log) β2 microglobulin 1.3 (1.1-1.6) 0.0006 1.3 (1.1-1.6) 0.001 months post high-dose therapy was estimated Age 1.00 (0.98-1.02) 0.9 1.01 (0.98-1.03) 0.6 by conventional recommended flow cytometry Sex (Male vs. female) 1.2 (0.9-1.6) 0.1 1.2 (0.9-1.6) 0.2 and revealed no significant differences between the levels following the first and the Stage I, II or III - 0.06 NI NI second transplant. The median level of plasma Stage (II, III vs. I) 2.0 (1.0-3.9) 0.02 1.3(0.6-2.8) 0.5 cells was 0.22% and 0.16%, respectively (N=17; Response non-CR vs. CR 1.9 (1.4-2.4) < 0.0001 NI NI p=0.3). NI, Variable not included in multivariate analysis

Discussion a 1.0 b 1.0 DTx DTx Double autologous transplantation regi- STx STx p = 0.5 p = 0.7 mens have been used to treat myeloma for the 0.8 0.8 past decade, although the advantage over single transplantation is unclear. Despite the favorable results reported by IFM,10 it remains 0.6 0.6 the case that progressive myeloma will develop in almost 80% of patients within seven years 0.4 0.4 after they have undergone double transplantation. This unclear situation is further extend- Proportion surviving ed by progress in the number of new drugs cur- 0.2 Proportion progression-free 0.2 rently being analyzed in randomized trials. In the near future, ongoing studies will clarify the 0.0 0.0 role of inflammatory mediators and protea- 0 24 48 72 96 0 24 48 72 96 some inhibitors in the context of autologous Survival from inclusion Time to progression from ASCT stem cell transplantation. However, trial design including consolidation therapy with N. at risk N. at risk new drugs may be hampered by the biased DTx 101 85 40 91 DTx 101 75 29 71 Stx 101 87 44 18 2 Stx 101 59 22 81 double autologous transplantation up front strategy which has not yet been documented to Figure 2. Case control analysis comparing double and single transplantation. Overall sur- have an effect on overall survival.11-12 Among vival (a) and progression free survival (b) following double transplantation in trial the patients enrolled in the double transplant HKTH was compared to 101 case controlled single transplantations in NMSG #5/94 group in the present study, 23% could not plus #7/98. The estimated probabilities are shown for double (DTx) and single (STx) receive their assigned second transplantation transplantation. Tables below the graph indicate patients at risk for the estimate.

[page 12] [Hematology Reviews 2009; 1:e2] Article comparing the two groups, but the probability future. In this unclear situation we now see a including 914 patients. Hematol J 2001;2:272-8. estimates of overall survival were unchanged range of randomized trials evaluating new 8. Palumbo A, Triolo S, Argentino C, et al. Dose- following double transplantation, a lack of drugs. In the near future, several ongoing intensive melphalan with stem cell support benefits which allow us to recommend that up studies will clarify the role of these novel (MEL100) is superior to standard treatment front double transplantation is not to be agents, including thalidomide and its analogs, in elderly myeloma patients. Blood 1999;94: offered to all MM patients. bortezomib etc., in the context of autologous 1248-53. Preliminary accounts of other comparisons stem cell transplantation. However, consoli- 9. Fermand JP, Ravaud P, Chevret S, et al. High- of single and double transplantation have also dating trials such as that designed and ongo- dose therapy and autologous peripheral stem failed to show a difference in outcome, but the ing within NMSG, may be hampered by double cell transplantation in multiple myeloma: up- follow-up periods may be too short to show the autologous transplantation strategies not yet front or rescue treatment? Results of a mul- survival benefit as found in the IFM trial, or documented to have an effect on survival. In ticenter sequential randomized clinical trial. the studies may have been too small to pro- conclusion, by combining our analysis and Blood 1998;92:3131-6. 10. Attal M, Harousseau J-L, Facon T, et al. Single vide a basis for definitive conclusions. One recent results in the literature18 it is likely that versus double autologous stem-cell trans- major concern in the IMF trial is that the double transplantation as initial strategy will plantation for multiple myeloma. N Engl J power of the study was calculated based on not improve the final outcome among most Med 2003;349:2495-502. evaluation of response to therapy and not on patients with MM. Therefore, we recommend 11. Vesole DH, Barlogie B, Jagannath S, et al. survival estimation. Response rates analyzed consolidation trials studying the effect of new High-dose therapy for refractory multiple on an intention-to-treat basis in the two drugs to be based on a single high-dose mel- 10 myeloma: improved prognosis with better groups did not differ significantly. This is 2 phalan 200 mg/m conditioning supported by supportive care and double transplants. unexpected and points to the risk of bias in autologous transplantation. handling the treated patients during the pro- Blood 1994;84:950-6. tocol therapy. From a clinical point of view, the 12. Barlogie B, Jagannath S, Desikan KR, et al. Total therapy with tandem transplants for different conditioning regimen used during newly diagnosed multiple myeloma. Blood the first transplantation may have biased the References 1999;93:55-65. selection of patients with a good or a poor 13. Roer O, Hammerstrøm J, Lenhoff S, et al. response for later intense salvage and may 1. Lenhoff S, Hjorth M, Holmberg E, et al. Quality assessment of autografting by proba- have influenced the improved overall survival. Impact on survival of high-dose therapy with autologous stem cell support in patients bility evaluation: model estimation by clinical Without doubt selection and enthusiasm for younger than 60 years with newly diagnosed end-points in newly diagnosed multiple salvage treatments represent an important multiple myeloma: a population-based study. myeloma patients. Cytotherapy 2006;8:79-88. aspect of the global therapeutic strategy in Blood 2000;95:7-11. 14. Lenhoff S, Martin Hjorth, Jan Westin, et al. 2. Gore ME, Selby PJ, Viner C, et al. Intensive MM, particularly when relapse occurs after Impact of age on survival after high-dose 15 treatment of multiple myeloma and criteria autologous transplantation. Such a bias therapy with autologous stem cell support in might explain the very late significant sur- for complete remission. Lancet 1989;2:879-82. 3. Barlogie B, Gahrton G. Bone marrow trans- newly diagnosed multiple myeloma: a popula- vival differences, first identified 4-5 years fol- plantation in multiple myeloma. Bone Marrow tion-based study. Br J Haematol 2006;133:389- 10 lowing inclusion into the IFM study. Transplant 1991;7:71-9. 96. During the past decade, the advances in 4. Attal M, Harousseau J-L, Stoppa A-M, et al. A 15. Morris C, Iacobelli S, Brand R, et al. Benefit prospective, randomized trial of autologous treatment have been fast and furious. MM and timing of second transplantations in mul- patients have become a fertile testing ground bone marrow transplantation and chemotherapy in multiple myeloma. N Engl J Med tiple myeloma: clinical findings and method- for a number of new antineoplastic agents. 1996;335:91-7. ological limitations in a European Group for Thalidomide, once infamous, has shown itself 5. Child JA, Morgan GJ, Davies FE, et al. High- Blood and Marrow Transplantation registry as a potent immunomodulatory drug that dose chemotherapy with hematopoietic study. J Clin Oncol 2004;22:1674-81. induces responses even in patients with ther- stem-cell rescue for multiple myeloma. 16. Singhal S, Mehta J, Desikan R, et al. Antitumor apy-resistant disease. Thalidomide therapy is N Engl J Med 2003;348:1875-83. activity of thalidomide in refractory multiple 6. Moreau P, Facon T, Attal M, et al. Comparison a viable initial alternative to chemotherapy 2 myeloma. N Engl J Med 1999;341:1565-71. 16 of 200 mg/m melphalan and 8 Gy total body when combined with dexamethasone. irradiation plus 140 mg/m2 melphalan as con- 17. Richardson P, Barlogie B, Berenson J, et al. A Bortezomib, a proteasome inhibitor, has sub- ditioning regimens for peripheral blood stem phase 2 study of bortezomib in relapsed, stantial single-agent activity against myelo- cell transplantation in patients with newly refractory myeloma. N Engl J Med 2003;348: ma. Studies are under way that incorporate diagnosed multiple myeloma: final analysis of 2609-17. streatment with these drugs early in the the InterGroupe Francophone du Myélome 18. Cavo M, Tosi P, Zamagni E, et al. Prospective, 9502 randomized trial. Blood 2002;99:731-5. course of the disease, as are trials with borte- randomized study of single compared with 17 7. Blade J, San Miguel JF, Fontanillas M, et al. zomib in combination with other agents. Increased conventional chemotherapy does double autologous stem-cell transplantation How such developments will alter therapy for not improve survival in multiple myeloma: for multiple myeloma: Bologna 96 clinical MM patients remains to be seen in the near long-term results of two PETHEMA trials study. J Clin Oncol 2007 Jun 10;25:2434-41.

[Hematology Reviews 2009; 1:e2] [page 13] Hematology Reviews 2009; volume 1:e3

Tuberculosis associated and ITP were concomitant. In one case (patient 3), chest radiography abnormalities Correspondence: Raphael Borie, thrombocytopenic purpura: were detected four months after the diagnosis Service de Pneumologie A, Hopital Bichat, effectiveness of of thrombocytopenia. For each patient, ITP 46 rue Henri Huchard, diagnosis was based on a platelet count below 75877 Paris CEDEX 18, France antituberculous therapy E-mail: [email protected] 30×109/L and normal bone marrow examina- Raphael Borie,1,2 Claire Fieschi,2 tion. No patient had splenomegaly. Other sec- Key words: immune thrombocytopenic purpura, ondary forms of ITP were excluded. All patients Eric Oksenhendler,2 Lionel Galicier2 immune, peripheral, thrombocytopenia responded to high-dose intravenous immuno - 1 Department of Pneumology A, Hospital globulin (HD-IVIg) as in most cases of ITP;2 Acknowledgments: potential conflicts of interest: Bichat, Assistance Publique, however, none had complete and sustained none. Financial support: none. For this retrospec- Hôpitaux de Paris, Paris, France response. In 3 patients, platelet counts tive, observational, non-interventional analysis of 2Department of Clinical Immunology, remained below 30×109/L despite corticosteroid medical records, French law does not require the specific approval of an internal review board or and HD-IVIg. These patients had partial Hospital St Louis, Assistance Publique, the consent of patients. Hôpitaux de Paris, Paris, France response to danazol or vincristine. Myco - bacterium tuberculosis was identified in culture Received for publication: 14 January 2009 in 3 patients. Histological examination of an Revision received: 25 February 2009 adenopathy confirmed the diagnosis in the 2 Accepted for publication: 9 March 2009 Abstract others. Each strain of mycobacterium tuberculosis was sensitive to conventional therapy. All This work is licensed under a Creative Commons patients completed treatment and were consid- Attribution 3.0 License (by-nc 3.0). Association of immune thrombocytopenic ered to be cured for tuberculosis. Within two purpura and tuberculosis is a rare condition. In ©Copyright Raphael Borie et al., 2009 months of anti-tuberculous therapy, all 5 Licensee PAGEPress, Italy 5 patients presenting with this association, patients were in ITP complete response and anti-tuberculous therapy was effective on both Hematology Reviews 2009; 1:e3 were off specific therapy for thrombocytopenia. doi:10.4081/hr.2009.e3 tuberculosis and thrombocytopenia suggesting Patient 2 required corticosteroid for three a causal relationship between tuberculosis and months for associated hemolytic anemia. immune thrombocytopenic purpura well known cause of ITP and tuberculosis.1 Autoimmune disorders and lymphoproliferative Discussion diseases, that may cause ITP, are by themselves Introduction or when treated, immunosupressive conditions.1 None of our patients had such an etiolo- ITP may require treatment with oral pred- Immune thrombocytopenic purpura (ITP) is nisone or HD-IVIg. Response rates are almost gy. characterized by a low platelet count associated 75%; however, only 20% of adult patients with Disseminated tuberculous infection may with the presence of platelet autoantibodies. ITP achieve persistent remission.1 In the pres- cause peripheric or central thrombocytopenia. The diagnosis of ITP remains a diagnosis of ent series, all patients had persistent or refrac- Tuberculous splenic abcess is rare but a possi- exclusion, and a bone marrow examination tory ITP. No patient was cured at the time of ble cause of hypersplenism and peripheric should be performed in patients with atypical tuberculosis diagnosis or within the month fol- thrombocytopenia. None of our patients had features. In 5-10% of cases, ITP is associated lowing the onset of ATT. However, within three splenomegaly. All patients had bone marrow with chronic infection (HIV, HCV), systemic months of ATT, ITP was in complete response aspiration to confirm peripheric thrombocy- autoimmune disorders, lymphoproliferative dis- in all patients. None required further ITP treat- topenia, while bone marrow involvement of orders, and primary immunodeficiency. In ment or splenectomy. tuberculosis may provoke central thrombocy- adults, ITP is usually chronic (i.e. more than six Persistence of thrombocytopenia during the topenia. Hemophagocytic lymphohistiocytosis months in duration).1 The response to treat- first month of ATT may be explained by the fact is an uncontrolled activation of T cells and ment is defined as an increase in platelet count that Mycobacterium tuberculosis can remain macrophages, and an overproduction of inflam- above 50×109/L, with at least a 2-fold increase in alive in the first two months of anti-tubercu- matory cytokines responsible for thrombocy- the initial value and remission is considered if lous therapy.3 These results suggest a direct topenia and anemia. It has been previously platelet count reaches 150×109/L.2 relationship between tuberculosis and ITP. decribed secondary to tuberculosis.7 However, We describe 5 cases of the association of ITP is a rare manifestation of tuberculosis. bone marrow aspirate excluded this diagnosis. tuberculosis and ITP and the effectiveness of To our knowledge only 18 reports describing 27 Appearance of thrombocytopenia after treat- anti-tuberculous therapy (ATT) on both tuber- cases of ITP associated with tuberculosis have ment of tuberculosis is classically induced by culosis and thrombocytopenia (Figure 1). been reported.4-6 All patients were treated with rifampicin and recovered after stopping ATT in combination with corticosteroid, HD- rifampicin.8 On the contrary, development of IVIg or vincristine. In all cases platelet count tuberculosis after long-term steroid treatment recovered with ATT. The correction of thrombo- is not uncommon while steroids are immuno- Case Reports cytopenia in all patients within two months of suppresive. One of our patients (case 3) had ATT suggests a causal relationship between received corticoid before the diagnosis of Patients were 3 men and 2 women (Table 1). tuberculosis and thrombocytopenia, although tuberculosis was made. At first she was asymp- Median age was 38 years. Major symptoms the mechanism remains unknown. tomatic and chest radiography was considered were pulmonary symptoms in 3 cases and Few causatives conditions may bind throm- as normal. But four months later, she was bleeding in 2 cases. However, thrombocytope- bocytopenia to tuberculosis. The presence of an coughing and dyspneic; chest radiography nia was symptomatic in 4 patients. In 4 of the underlying disease causative of thrombocytope- revealed abnormalities and Ziehl coloration of 5 cases, clinical manifestations of tuberculosis nia may favor tuberculosis. HIV infection is a the sputum was positive for BAAR. The short

[page 14] [Hematology Reviews 2009; 1:e3] Article

Table 1. Characteristics of the 5 patients with immune thrombocytopenic purpura and tuberculosis. Patient Age/Sex Delay between Tuberculosis ITP and Symptoms CT scan Diagnosis Duration of tuberculosis treatment 1 56/M None Pulmonary Tree in bud pattern Sputum culture 6 months bronchiolitis of the upper lobes 2 16/F None Asthenia Retroperitoneal Biopsy of the 9 months adenopathy adenopathy 3 40/F 4 months Pulmonary Cavernous Sputum culture 6 months destruction of the left upper lung 4 38/M None Fever Pulmonary Biopsy of an 6 months micronodules and adenopathy mediastinal adenopathies 5 26/M None Pulmonary Cavernous destruction Sputum culture 6 months of both upper lobes

F, female; M, male.

Figure 1. Platelet counts in 5 patients presenting idiopathic thrombocytopenic purpura concommi- tant to tuberculosis. Thrombocytopenia resolved within three months of anti- tuberculous therapy.

course of steroids (less than a month) and the rare condition, estimated to occur in less than ed positive for antinuclear antibodies. brief period between steroid treatment and the 1% of cases of tuberculosis. Mycoba- Tuberculosis is a rare but curative cause of discovery of radiographic abnormalities sug- cter ium tuberculosis may share antigene with thrombocytopenia. Thrombocytopenic patients, gest that active tuberculosis was already pres- platelet leading to antiplatelet antibody forma- and particularly chronic and hard to treat ITP, ent when steroids were prescribed and that the tion. Specific HLA presentation of tuberculosis should benefit from tuberculosis depistage. treatment only favored progression of tubercu- could also lead to antiplatelet immunity In cases of tuberculosis associated ITP, losis. In all other cases, active tuberculosis and response in some patients.1 However, only one recurrence of thrombocytopenia is frequent in thrombocytopenia were concomitant. patient developed an auto-antibody. In this case, the first two months, and patients may benefit The pathophysiology of thrombopenia in these were anti-hematia antibodies revealed from close observation leading to a continua- tuberculosis remains unanswered. This is a with Coombs testing. None of our patients test- tion of ITP specific therapy.

[Hematology Reviews 2009; 1:e3] [page 15] Brief Report

ent presence of acid-fast bacilli in sputum bocytopenic purpura associated with pul- References smears in pulmonary tuberculosis. Chest monary tuberculosis. Intern Med 2006;45: 1999;116:726-31. 739-42. 4. Madkaikar M, Ghosh K, Jijina F, Gupta M, 1. Cines DB, Blanchette VS. Immune thrombo- 7. Brastanios PK, Swanson JW, Torbeson M, Rajpurkar M, Mohanty D. Tuberculosis and Sperati J, Karakousis PC. Tuberculosis-asso- cytopenic purpura. N Engl J Med 2002;346: immune thrombocytopenia. Haematologica ciated haemophagocytic syndrome. Lancet 995-1008. 2002;87: ELT38. 2. Cines DB, Bussel JB. How I treat idiopathic 5. Ghobrial MW, Albornoz MA. Immune throm- Infectious Disease 2006;6:447-54. thrombocytopenic purpura (ITP). Blood bocytopenia: a rare presenting manifestation 8. Kant S, Natu NK, Mahajan V. Rifampicin, 2005;106:2244-51. of tuberculosis. Am J Hematol 2001;67:139-43. ethambutol and pyrazinamide-induced 3. Al-Moamary MS, Black W, Bessuille E, Elwood 6. Tsuro K, Kojima H, Mitoro A, Yoshiji H, throm bocytopenia. Int J Clin Pharmacol Ther RK, Vedal S. The significance of the persist- Fujimoto M, Uemura M, et al. Immune throm- 2008;46: 440-2.

[page 16] [Hematology Reviews 2009; 1:e3] Hematology Reviews 2009; volume 1:e4

Cardiotoxicity of tyrosine patients, and secondary resistance presented as relapse in 17% of patients and disease progres- Correspondence: Delong Liu, Associate Professor kinase inhibitors in chronic sion in 7% after 4.5 years.2 After six years of fol- of Medicine, Division of Hematology and myelogenous leukemia therapy low-up, 34% of patients had discontinued ima- Oncology, New York Medical College, Valhalla, tinib treatment, mostly (14%) because of an NY 10595, USA. E-mail: [email protected] Zhenshu Xu,1,2 Shundong Cang,1 unsatisfactory therapeutic effect (defined as Key words: tyrosine kinase inhibitors, imatinib, 2 1 Ting Yang, Delong Liu lack of efficacy/progression), but a number of dasatinib, nilotinib, cardiotoxicity. 1Division of Hematology and Oncology, patients (5%) also stopped receiving the drug as New York Medical College, a result of adverse events (AEs) or abnormal lab- Acknowledgments: ZX and SC are CAHON 3 Research Scholars (Chinese American Hemato- Valhalla, NY, USA oratory values. Further treatment options therefore continue to be developed. logists Oncologist Network, www.cahon.org). 2Department of Hematology, Institute of Second-line choices for treatment include SC is a recipient of a fellowship grant from the Hematology, Union Hospital, Fujian International Scholar Exchange Foundation. This increasing the dose of imatinib, or changing work was partly supported by New York Medical Medical University, Fuzhou, China ® therapy to dasatinib [Sprycel ; Bristol-Myers College Blood Diseases Fund. Writing and edito- Squibb Co. (BMS), New York, USA] or nilo- rial support were provided by Zoila Mora and Josh tinib (Tasigna®; Novartis Pharmaceuticals Collis, and funded by Bristol-Myers Squibb. Corp., New Jersey, USA). Newer agents for Abstract imatinib failures are still under clinical devel- Received for publication: 13 January 2009 Revision received: 6 March 2009 opment.4 Dasatinib potently inhibits BCR-ABL.5 Accepted for publication: 11 March 2009 Emerging evidence suggests that the three Compared with imatinib, dasatinib has 325- tyrosine kinase inhibitors currently approved fold greater activity against native BCR-ABL.6 This work is licensed under a Creative Commons for the treatment of patients with chronic Furthermore, dasatinib inhibits all imatinib- Attribution 3.0 License (by-nc 3.0). myelogenous leukemia (CML) – imatinib, resistant mutations of this molecule (key dasatinib, and nilotinib – have potential car- mediators of imatinib resistance) except the ©Copyright Zhenshu Xu et al., 2009 diotoxic effects. The mechanisms behind T315l mutant, which is resistant to all current- Licensee PAGEPress, Italy these events, and the relations between them, 6-8 ly available TKIs. It is, however, relatively Hematology Reviews 2009; 1:e4 are largely unclear. For example, relative to 9 insensitive to the F317L mutation. doi:10.4081/hr.2009.e4 dasatinib and nilotinib, severe congestive Dasatinib was originally approved at the heart failure and left ventricular dysfunction dosage of 70 mg twice daily, following data are rare but prominent with imatinib treat- from the START (the SRC/ABL Tyrosine Kinase ment, particularly in patients receiving higher Inhibition Activity: Research Trials of doses (>600 mg/day). In comparison with ima- Dasatinib) program of clinical studies.10-12 In Common adverse events tinib, prolongation of the QT interval is rela- November 2007, the label for dasatinib was associated with tyrosine kinase tively common in patients treated with either changed to include updated dosing informa- dasatinib or nilotinib. In contrast to nilotinib, inhibitors tion, safety information from more than 2,100 pericardial effusions are observed with both patients, and data from a randomized compar- imatinib and dasatinib. It is suggested that A number of AEs are associated with TKIs. ison with high-dose imatinib.13 The recom- these data, an evaluation of cardiac status, use The most common ones are hematologic. In of concomitant medications, and potential risk mended starting dose for patients with chron- the IRIS study, anemia, neutropenia, and factors should be considered in the manage- ic phase (CP) CML is now 100 mg once daily. thrombocytopenia (all grades) were each ment of CML. The starting dose for advanced disease reported by 45-61% of patients receiving ima- remains 70 mg twice daily. tinib.1 Grade 3-4 anemia, neutropenia, and Nilotinib (an analog of imatinib) is also thrombocytopenia were observed in 3%, 14%, active against all imatinib-resistant BCR-ABL and 8% of patients, respectively. The updated 14 Introduction mutations except T315l. However, nilotinib is dasatinib label includes data from the START-R relatively inactive against common mutations trial, a comparison of dasatinib 70 mg twice 4 BCR-ABL tyrosine kinase is a key molecule in the ATP-binding P-loop domain, and those daily with high-dose imatinib (800 mg/day) in responsible for the pathophysiology of chronic at the F359 residue in the catalytic domain, of patients with CP CML resistant to standard- 6 myelogenous leukemia (CML). Tyrosine BCR-ABL. Patients harboring such mutations dose imatinib.13,16 In this study, the incidences kinase inhibitors (TKIs) directed against BCR- at baseline do not respond to nilotinib and of grade 3-4 thrombocytopenia and neutrope- ABL are currently the cornerstone of treatment progress quickly during treatment. nia for high-dose imatinib were 14% and 39%, for patients with CML. Imatinib (Gleevec®; Furthermore, patients who do not have these respectively. The START-R study also showed Novartis Pharmaceuticals Corporation, New mutations at baseline may eventually develop that dasatinib has significant response and Jersey, USA) was the first TKI approved for them when resistance to imatinib occurs.15 progression-free survival benefits compared CML. Imatinib has marked response and sur- Nilotinib has recently been approved by the with high-dose imatinib, and that significantly vival benefits over interferon-α plus low-dose FDA for treating imatinib-resistant and -intol- more patients discontinue high-dose imatinib cytarabine and is presently the only TKI erant patients with CP or accelerated phase than dasatinib.16 licensed for first-line treatment.1,2 However, CML, but not those suffering from blast crisis. The 100 mg once daily dose for dasatinib in despite the remarkable success of imatinib, TKIs share a number of common adverse patients with CP CML was approved as a result many patients discontinue treatment because effects (AE), including neutropenia and of data from a phase III dose modification of either resistance or intolerance to this drug. thrombocytopenia. Cardiotoxicities of the TKIs study.17 Shah and colleagues compared dasa- In the pivotal phase III IRIS (International have become a clinical concern. This review tinib 100 mg once daily, 50 mg twice daily, 140 Randomized Study of Interferon and STI571) provides an overview of common AEs and car- mg once daily, and 70 mg twice daily doses in trial, primary resistance was observed in 24% of diotoxicities associated with these TKIs. CP CML. In this study, response rates were

[Hematology Reviews 2009; 1:e4] [page 17] Article similar between doses, and overall efficacy was cardiac failure is a rare feature of imatinib expanded access study. The early occurrences equivalent between the 70 mg twice daily and therapy.25,26 of some of these deaths relative to the start of 100 mg once daily regimens. In the 100 mg The mechanism underlying imatinib- nilotinib treatment suggest that ventricular once daily dose there was significantly less induced cardiac failure is currently unclear. In repolarization may have contributed to their grade 3-4 thrombocytopenia (22%; p=0.004) an in vitro study, physiological concentrations occurrence.18 compared with the 70 mg twice daily dose of imatinib significantly and adversely affected Another common cardiac AE associated with (37%). Grade 3-4 anemia, leukopenia, and mitochondrial membrane potential, apoptosis, nilotinib therapy is myocardial ischemia. In neutropenia were observed in 10%, 16% and cell viability, and cellular ultrastructure.27 This the pivotal phase II study, it was reported in 7% 33% of patients, respectively. The 100 mg once cardiotoxic effect may be linked to inhibition (21/321) of patients with CP CML receiving daily dose was also associated with the lowest of BCR-ABL. Imatinib was reported to cause nilotinib as second-line treatment.31 frequencies of treatment discontinuation stress-induced and dose-dependent mitochon- Uncommon cardiac events (0.1-1% of all (overall: 16% vs. 23%; toxicity alone: 4% vs. drial changes in murine ventricular myocytes, patients in clinical trials) include cardiac fail- 11%, respectively). which was reduced by re-engineering the ima- ure, angina pectoris, atrial fibrillation, pericar- For nilotinib, the incidence rates of grade 3- tinib molecule such that BCR-ABL inhibition dial effusion, coronary artery disease, car- 4 thrombocytopenia (28%), neutropenia was hampered.28,29 Nonetheless, the re-engi- diomegaly, cardiac murmur, and bradycardia. (28%), and anemia (8%) at the recommended neered molecule may have had altered activi- Rare events (of uncertain frequency) include dose (800 mg/day) in patients with CP CML ties besides reduced BCR-ABL inhibition. myocardial infarction, ventricular dysfunction, appear to be similar to those for dasatinib 100 A second cardiac AE associated with ima- pericarditis, cardiac flutter, and extrasystoles.18 mg once daily.17,18 tinib therapy is fluid retention manifesting as The mechanisms underlying these AEs are The incidences of non-hematologic AEs are pericardial effusion. Grade 3-4 fluid retention still unclear. In a manner similar to imatinib, much lower than those for hematologic events reactions, which included pericardial effu- nilotinib was found to decrease the cellular for all TKIs, and are broadly similar between sions, were reported in 2% of patients in the viability of rat cardiomyocytes cultured in TKIs at their current recommended doses. IRIS study and in 6% of all other CML clinical vitro, although the integrity of the mitochondr- Cutaneous toxicity is more common for TKIs 20 studies. Frank pericarditis has been observed ial membrane potential was unaffected.27 against receptor tyrosine kinases.19 in <0.1% of patients receiving imatinib (all However, nilotinib has also been found to 20 indications). Other cardiac AEs include inhibit human ether-a-go-go related gene tachycardia, hypertension, hypotension, flush- (hERG) potassium currents with an IC50 of ing, and peripheral coldness, were each 0.66 μM. This concentration is approximately Cardiotoxicities observed in reported in 0.1-1.0% of patients.20 one-tenth the expected Cmax for this com- chronic myeloid leukemia Precautions and general guidelines for dose pound, well within therapeutic levels. This adjustment for cardiac AEs associated with mechanism is likely to underlie nilotinib- Current evidence suggests that TKIs have imatinib treatment are included in the pre- induced QT prolongation. Inhibition of hERG potential cardiotoxic effects. Cardiac AEs scribing information and have been summa- channels is established as a cause of QT pro- reported include palpitations, arrhythmia, QT rized in Table 2. CML patients with existing longation for a number of compounds, and is a prolongation, pericardial effusions, myocardial cardiac disease or cardiac risk factors should 20,23 significant barrier in the development of new ischemia, myocardial infarction, and conges- be monitored and treated accordingly. 32 tive heart failure (CHF). All of the clinically Patients should also be weighed regularly and drugs. Indeed, the phase II development of available BCR-ABL inhibitors report the poten- monitored for signs and symptoms of fluid the aurora kinase inhibitor MK-0457 (VX-680) tial for cardiotoxicity in their respective pack- retention. Unexpected weight gain should be was recently suspended, pending a full analy- age inserts (Table 1).13,18,20 investigated carefully, and treated appropri- sis of all efficacy and safety data. The decision ately.20,30 Significant fluid retention (local or was based on preliminary safety data, in which Imatinib general) can usually be managed by interrupt- QT prolongation was observed in one patient.33 Although rare, severe CHF and left ventric- ing imatinib treatment and using diuretics or The potential for QT prolongation and sud- ular dysfunction have been reported during other supportive care.20 In severe cases of fluid den death associated with nilotinib, although imatinib treatment, especially in patients with retention, imatinib should be withheld until rare, necessitates vigilant monitoring. In par- risk factors or comorbidities. In the IRIS trial, this is resolved. ticular, ECGs should be performed at baseline, 1,2,20 this was observed in 0.7% of patients. seven days after initiation of treatment, peri- Cardiotoxicity associated with high-dose ima- Nilotinib odically throughout therapy, and following Nilotinib can cause QT prolongation and tinib was not reported during the START-R dose adjustments. Electrolyte levels should be 16 sudden death and carries a black box warning study. In smaller studies of high-dose ima- monitored periodically throughout therapy. tinib, existing CHF was exacerbated in one for these side effects.18 Nilotinib prolongs the Nilotinib is contraindicated for patients with study (6% of patients), and was related to QT interval in a concentration-dependent hypokalemia, hypomagenesmia or long QT 21,22 manner, and the common AEs occurred in 1- mortality in another (4% of patients). 18 18 syndrome. The nilotinib prescribing informa- In a large review of patients enrolled in 10% of all patients in clinical trials. In a study imatinib clinical studies (n=1,276), 1.8% of in healthy volunteers, nilotinib was associated tion recommends dose adjustments for QT pro- patients had symptoms suggestive of systolic with a maximum mean QT interval increase of longation, presented in Table 2. heart failure. Most had risk factors for cardiac 18 ms (adjusted for placebo).18 In imatinib- failure. In total, 0.6% of patients had cardiac resistant patients with CML, nilotinib caused a Dasatinib events attributed to imatinib treatment. The maximum mean QT change of 10 ms from The events of dasatinib-induced QT prolon- authors concluded that CHF is rare during baseline; QT increase >60 ms and QT >500 ms gation are rare although a warning for such a imatinib treatment.23 A similar incidence of associated with nilotinib were reported in 2.1% possible event is given. In single-arm studies CHF (1%) was reported in a nine-year retro- and <1% of patients, respectively. Sudden of dasatinib, nine patients (1%) had QT prospective review at a single institution.24 deaths were reported in 0.6% of patients in a longation reported as an AE.13 The mean QT Reports of smaller studies also confirm that clinical study, and at a similar frequency in an interval increased by 3-6 ms (Fridericia’s

[page 18] [Hematology Reviews 2009; 1:e4] Article

Table 1. Reported incidence of cardiotoxicity during tyrosine kinase inhibitors treatment. Toxicity n Incidence (%) Reference Imatinib Pericardial effusions NR 6 Novartis20 Systolic heart failure 1276 1.8 Atallah et al23 Congestive heart failure 553 0.7 Novartis20 Left ventricular dysfunction 553 0.7 Novartis20 Cardiac failure NR 0.1-1.0 Novartis20 Flushing NR 0.1-1.0 Novartis20 Hypertension NR 0.1-1.0 Novartis20 Hypotension NR 0.1-1.0 Novartis20 Peripheral coldness NR 0.1-1.0 Novartis20 Tachycardia NR 0.1-1.0 Novartis20 Pericarditis NR <0.1 Novartis20 Dasatinib Severe pericardial effusions NR 1 BMS13 Congestive heart failure 911 4 Brave et al.34 Arrhythmia NR 1-<10 BMS13 Palpitations NR 1-<10 BMS13 QT prolongation > 500 ms -300 <1 BMS13 Angina pectoris NR 0.1-<1 BMS13 Cardiomegaly NR 0.1-<1 BMS13 Myocardial infarction NR 0.1-<1 BMS13 Pericarditis NR 0.1-<1 BMS13 Ventricular arrhythmia NR 0.1-<1 BMS13 Acute coronary syndrome NR <0.1 BMS13 Myocarditis NR <0.1 BMS13 Nilotinib Myocardial ischemia 321 7 Kantarjian et al.31 Increase in QTcF >60 ms from BL 232 2.1 Novartis18 Palpitations NR 1-10 Novartis18 Angina pectoris NR 0.1-1 Novartis18 Atrial fibrillation NR 0.1-1 Novartis18 Bradycardia NR 0.1-1 Novartis18 Cardiac failure NR 0.1-1 Novartis18 Cardiac murmur NR 0.1-1 Novartis18 Cardiomegaly NR 0.1-1 Novartis18 Coronary artery disease NR 0.1-1 Novartis18 Pericardial effusion NR 0.1-1 Novartis18 Deatha 867 0.6 Novartis18 BL, baseline; NR, not reported, representing all patients within each agent’s clinical trials. aPossibly resulting from ventricular repolarization. method); this increase was not clinically rele- patients in all clinical studies, and the pre- Both pericardial effusions and cardiac fail- vant.34 In total, <1% of patients had a QT scribing information for dasatinib includes a ure associated with dasatinib therapy may be increase to >500 ms. In contrast with nilotinib, warning for this toxicity.13 Severe CHF has also caused by similar mechanisms to those asso- the IC50 for dasatinib for the inhibition of hERG been reported in 1% of all patients.13 In single- ciated with imatinib treatment, although, in currents (14.3 μM) is 100 times the expected arm studies, CHF or ventricular dysfunction contrast with imatinib, dasatinib has not been Cmax for this drug.27 This may explain why QT occurred in 4% (20/911) of patients.34 found to significantly affect mitochondrial prolongation is more clinically prominent for However, in the dose optimization study, dasa- membrane potential, apoptosis, cell viability, nilotinib than it is for dasatinib. tinib 100 mg once daily was not associated or cellular ultrastructure at physiological concentrations.27 Common cardiac AEs (observed in 1 – <10% with any incidence of severe CHF or pericar- Uncommon cardiac AEs associated with of all patients in clinical trials) include dial effusion in any patient; both AEs were dasatinib include angina pectoris, car- arrhythmia and palpitations. Severe pericar- reported in patients receiving the 70 mg twice diomegaly, pericarditis, ventricular arrhyth- dial effusions have been reported in 1% of all daily dose (all grades, 4%; grade 3-4, 3%).17 mia, and myocardial infarction (reported in

[Hematology Reviews 2009; 1:e4] [page 19] Article

Table 2. Precautions and dose modifications on the emergence of cardiac events during tyrosine kinase inhibitors treatment. Agent Dose modification Precautions Imatinib20 On the emergence of a severe event, withhold until Carefully monitor any patient with or at risk of cardiac failure. the event has resolved. Resume at an appropriate dose, All patients with signs or symptoms depending on initial severity of event. of cardiac failure should be evaluated and treated. Dasatinib13 On the emergence of a severe event, withhold until the Administer with caution to patients with or at risk of QTc prolongation: event has resolved or improved. Resume at an those with hypokalemia, hypomagnesemia, or congenital long QT appropriate dose, depending on initial severity of event. syndrome; or taking medicines known to prolong QT, including anti-arrhythmic drugs, and cumulative high-dose anthracycline therapy Correct hypokalemia or hypomagnesemia prior to administration. Nilotinib18 QTc >480 ms: 1) withhold therapy, correct serum Do not administer to patients with long QT syndrome. potassium and magnesium levels if below normal, Do not administer drugs known to prolong QT, and review concomitant medication; 2) resume at prior dose in and strong CYP3A4 inhibitors. <2 weeks if QTc returns to <450 ms and <20 ms of baseline; Correct hypokalemia or hypomagnesemia prior to administration, 3) reduce dose to 400 mg/day if QTc 450-480 ms after 2 weeks; and periodically monitor serum electrolyte levels during therapy. 4) discontinue if QTc returns to >480 ms after dose reduction; Perform ECGs at baseline, 7 days after treatment starts, 5) repeat ECG assessment approx. 7 days after any dose adjustment. periodically as indicated clinically, and after any dose adjustment.

0.1 – <1% of patients). Rare events (reported 8. Tokarski JS, Newitt JA, Chang CY, et al. in <0.1% of patients) include myocarditis and References The structure of dasatinib (BMS-354825) acute coronary syndrome.13 bound to activated ABL kinase domain elu- Dasatinib should be administered with cau- 1. O'Brien SG, Guilhot F, Larson RA, et al. cidates its inhibitory activity against ima- tion to any patient at risk of cardiac problems Imatinib compared with interferon and tinib-resistant ABL mutants. Cancer Res (especially prolongation of the QT interval). low-dose cytarabine for newly diagnosed 2006;66:5790-7. Patients with pre-existing congenital long QT chronic-phase chronic myeloid leukemia. 9. Soverini S, Colarossi S, Gnani A, et al. syndrome or those receiving anti-arrhythmic N Engl J Med 2003;348:994-1004. Resistance to dasatinib in Philadelphia- medicine should be monitored closely.13 2. Druker BJ, Guilhot F, O'Brien SG, et al. positive leukemia patients and the pres- Patients with hypokalemia or hypomagne- Five-year follow-up of patients receiving ence or the selection of mutations at semia should have these conditions corrected imatinib for chronic myeloid leukemia. N residues 315 and 317 in the BCR-ABL prior to receiving dasatinib. As several anti- Engl J Med 2006;355: 2408-17. kinase domain. Haematologica 2007;92: neoplastic agents, including the anthracy- 3. Hochhaus A, Druker BJ, Larson RA, 401-4. clines, are associated with cardiotoxicity, the O'Brien SG, Gathmann I, Guilhot F. IRIS 6- 10. Cortes J, Rousselot P, Kim DW, et al. treatment history of a patient should also be year follow-up: sustained survival and Dasatinib induces complete hematologic taken into consideration before commencing declining annual rate of transformation in and cytogenetic responses in patients with any therapy. Fluid retention events can typi- patients with newly diagnosed chronic imatinib-resistant or -intolerant chronic cally be managed by supportive measures. myeloid leukemia in chronic phase (CML- myeloid leukemia in blast crisis. Blood Dose interruption is also indicated (Table 2).13 CP) treated with imatinib. Blood 2007;110: 2007;109:3207-13. Abstract 25. 11. Guilhot F, Apperley J, Kim DW, et al. 4. Cang S, Liu D. P-Loop mutations and novel Dasatinib induces significant hematologic therapeutic approaches for imatinib fail- and cytogenetic responses in patients with Conclusions ures in chronic myeloid leukemia. imatinib-resistant or -intolerant chronic J Hematol Oncol 2008;1:15. myeloid leukemia in accelerated phase. Although rare, severe CHF and left ventricu- 5. Lombardo LJ, Lee FY, Chen P, et al. Blood 2007;109: 4143-50. lar dysfunction have occurred with imatinib Discovery of N-(2-chloro-6-methyl- phe - 12. Hochhaus A, Kantarjian HM, Baccarani M, treatment, especially in patients receiving nyl)-2-(6-(4-(2-hydroxyethyl)- piperazin- et al. Dasatinib induces notable hemato- higher doses. Cardiac failure is also an uncom- 1-yl)-2-methylpyrimidin-4- ylamino) thia- logic and cytogenetic responses in chronic mon feature of nilotinib and dasatinib therapy. zole-5-carboxamide (BMS-354825), a dual phase chronic myeloid leukemia after fail- However, CHF has not yet been reported in Src/Abl kinase inhibitor with potent anti- ure of imatinib therapy. Blood 2007;109: patients with CP CML receiving dasatinib with tumor activity in preclinical assays. J Med 2303-9. a starting dose of 100 mg once daily. Compared Chem 2004;47:6658-61. 13. Bristol-Myers Squibb Company. Dasatinib with imatinib, palpitations and prolongation of 6. O'Hare T, Walters DK, Stoffregen EP, et al. prescribing information. Bristol-Myers the QT interval are relatively common with In vitro activity of Bcr-Abl inhibitors Squibb Company, 2007. Available at: both dasatinib and nilotinib. Nilotinib carries a AMN107 and BMS-354825 against clinical- http://www.bms.com/products/data/ black box warning for such AE and sudden ly relevant imatinib-resistant Abl kinase 14. Weisberg E, Manley PW, Breitenstein W, et death has been observed. These AEs may be domain mutants. Cancer Res 2005;65: al. Characterization of AMN107, a selec- related to BCR-ABL inhibition and therefore be 4500-5. tive inhibitor of native and mutant Bcr-Abl. genuine class effects. Further research is 7. Shah NP, Tran C, Lee FY, Chen P, Norris D, Cancer Cell 2005;7:129-41. needed to clarify the mechanisms underlying Sawyers CL. Overriding imatinib resist- 15. Branford S, Shou Y, Lawrence R, Rudzki Z, these effects. ance with a novel ABL kinase inhibitor. Hughes T. The P-loop mutations Y253H Science 2004;305:399-401. and E255K/V may develop more frequently

[page 20] [Hematology Reviews 2009; 1:e4] Article

than T3151 during nilotinib therapy after therapy in patients with chronic myeloge- agent imatinib mesylate. Nat Med 2006;12: imatinib failure and are associated with nous leukemia. Blood 2003;101:473-5. 908-16. progression in patients with Ph-positive 22. Zonder JA, Pemberton P, Brandt H, 29. Fernandez A, Sanguino A, Peng Z, et al. An leukemia. Haematologica 2007;92(S1): Mohamed AN, Schiffer CA. The effect of anticancer C-Kit kinase inhibitor is 337. dose increase of imatinib mesylate in reengineered to make it more active and 16. Kantarjian H, Pasquini R, Hamerschlak N, patients with chronic or accelerated phase less cardiotoxic. J Clin Invest 2007;117: et al. Dasatinib or high-dose imatinib for chronic myelogenous leukemia with inad- 4044-54. chronic-phase chronic myeloid leukemia equate hematologic or cytogenetic 30. National Comprehensive Cancer Network: after failure of first-line imatinib: a ran- response to initial treatment. Clin Cancer NCCN Clinical Practice Guidelines in domized phase 2 trial. Blood 2007;109: Res 2003;9:2092-7. Oncology. Chronic myelogenous leukemia 5143-50. 23. Atallah E, Durand JB, Kantarjian H, Cortes V.3.2008. Available at: http://www.nccn.org/ 17. Shah NP, Kantarjian HM, Kim DW, et al. J. Congestive heart failure is a rare event Intermittent target inhibition with dasa- in patients receiving imatinib therapy. professionals/physician_gls/PDF/cml.pdf. tinib 100 mg once daily preserves efficacy Blood 2007;110:1233-7. 31. Kantajarian HM, Hochhaus A, Cortes J, et and improves tolerability in imatinib- 24. Breccia M, Cannella L, Frustaci A, al. Nilotinib is highly active and safe in resistant and -intolerant chronic-phase Stefanizzi C, Levi A, Alimena G. Cardiac chronic phase chronic myelogenous chronic myeloid leukemia. J Clin Oncol events in imatinib mesylate-treated chron- leukemia (CML-CP) patients with ima- 2008;26:3204-12. ic myeloid leukemia patients: A single tinib-resistance or intolerance. Blood 18. Novartis Pharmaceuticals Corporation: institution experience. Leuk Res 2008;32: 2007;110: Abstract 735. Nilotinib prescribing information. 835-6. 32. Pollard CE, Valentin JP, Hammond TG. Novartis Pharmaceuticals Corporation, 25. Perik PJ, Rikhof B, de Jong FA, Verweij J, Strategies to reduce the risk of drug- 2007. Available at: http://www.pharma.us. Gietema JA, van der Graaf WT. Results of induced QT interval prolongation: a phar- novartis.com/product/pi/pdf/tasigna.pdf. plasma N-terminal pro B-type natriuretic maceutical company perspective. Br J 19. Huang X, Seiter K, Patel S, Ahmed N, Liu peptide and cardiac troponin monitoring Pharmacol 2008;154:1538-43. D. Severe toxicity of skin rash, fever and in GIST patients do not support the exis- 33. Vertex Pharmaceuticals Incorporated: diarrhea associated with imatinib: case tence of imatinib-induced cardiotoxicity. Vertex's collaborator Merck suspends report and review of skin toxicities associ- Ann Oncol 2008;19:359-61. patient enrollment in clinical trials of MK- ated with tyrosine kinase inhibitors. Drug 26. Ribeiro AL, Marcolino MS, Bittencourt HN, 0457 (VX-680) pending full analysis of Design, Development and Therapy 2008 et al. An evaluation of the cardiotoxicity of clinical data. Vertex Pharmaceuticals http://www.dovepress.com/getfile.php?fileI imatinib mesylate. Leuk Res 2008;32: Incorporated, 2007. Available at: D=3682. 1809-14. 20. Novartis Pharmaceuticals Corporation: 27. Freebern WJ, Fang HS, Slade MD, et al. In http://investors.vrtx.com/releasedetail.cfm Imatinib prescribing information. Novartis vitro cardiotoxicity potential comparative ?ReleaseID=276543 Pharmaceuticals Corporation, 2007. assessments of chronic myelogenous 34. Brave M, Goodman V, Kaminskas E, et al. Available at: http://www.pharma.us.novar- leukemia tyrosine kinase inhibitor thera- Sprycel for chronic myeloid leukemia and tis.com/product/pi/pdf/gleevec_tabs.pdf pies: dasatinib, imatinib and nilotinib. Philadelphia chromosome-positive acute 21. Kantarjian HM, Talpaz M, O'Brien S, et al. Blood 2007;110: Abstract 4582. lymphoblastic leukemia resistant to or Dose escalation of imatinib mesylate can 28. Kerkelä R, Grazette L, Yacobi R, et al. intolerant of imatinib mesylate. Clin overcome resistance to standard-dose Cardiotoxicity of the cancer therapeutic Cancer Res 2008;14:352-9.

[Hematology Reviews 2009; 1:e4] [page 21] Hematology Reviews 2009; volume 1:e5

Bacterial contamination of times higher than the combined risk of HIV, HBV, HCV and HTLV-1/2.9 Transfusion related Correspondence: Dana Védy, platelet concentrates: sepsis is often not recognized, and the real Service Régional Vaudois de Transfusion pathogen detection and clinical and fatal prevalence are probably Sanguine, Route de la Corniche 2, CH-1066 inactivation methods underestimated.10 Epalinges, Switzerland E-mail: [email protected]

Conflict of interest: none Dana Védy, Daniel Robert, Danielle Gasparini, Giorgia Canellini, Platelet transfusion-related Received for publication: 13 January 2009 Sophie Waldvogel, Jean-Daniel Tissot bacterial contamination: Revision received: 19 March 2009 Accepted for publication: 24 March 2009 Service Régional Vaudois de Transfusion some data Sanguine, Lausanne, Switzerland This work is licensed under a Creative Commons In the United States, bacterial contamination Attribution 3.0 License (by-nc 3.0) is considered, after transfusion errors, as the ©Copyright Dana Védy et al., 2009 second most common cause of death related to Licensee PAGEPress, Italy. Abstract transfusion.1 As mentioned by Holme et al., the Hematology Reviews 2009; 1:e5 estimated number of patients who received bac- doi:10.4081/hr.2009.e5 Whereas the reduction of transfusion related terial contaminated platelets per year ranges viral transmission has been a priority during the from 2,000-4,000, resulting in 200-600 cases of last decade, bacterial infection transmitted by clinical sepsis, and an estimated 40-533 fatali- 7%, and other organisms in 4%. The clinician 2 transfusion still remains associated to a high ties. The death risk in the USA has been esti- should be aware that uncommon pathogens morbidity and mortality, and constitutes the mated at 1/500,000 platelet concentrates. might be encountered in platelet prepara- Between 1995 and 2001, the English hemovigi- most frequent infectious risk of transfusion. tions.4,15 lance system reported 21 transfusion-related This problem especially concerns platelet con- Infrequently, donor bacteremia may be pres- bacterial contamination, leading to 6 deaths, in centrates because of their favorable bacterial ent during blood collection, also leading to growth conditions. This review gives an which 5 were attributed to platelet contamina- platelet contamination. In a large study evalu- overview of platelet transfusion-related bacterial tion.13 In France, Fournier-Wirth et al. reported ating the risk of septic platelet transfusion contamination as well as on the different strate- that the incidence of transfusion related reac- reactions, 23 septic transfusion reactions were gies to reduce this problem by using either bac- tions due to bacterial contaminated platelets observed; 15 out of 23 reactions (62.5%) were terial detection or inactivation methods. was about 1/25,000 units that was associated with severe morbidity or even mortality in about attributed to skin flora contamination, and 8 3 cases each year in the period 1999-2003.14 The out of 23 (34.8%) resulted from bacteria, more death rate was estimated to be one death per likely associated to transient donor bac- 16 Introduction 200,000 distributed platelet concentrates.13 teremia. Furthermore, the authors also showed that pooled platelet concentrates from buffy coats have a higher bacterial contamina- Reduction of transfusion related viral trans- tion rate compared to apheresis platelets. A mission such as HIV, HBV or HCV has been a total of 32,333 pooled-platelet concentrates and priority for blood transfusion services. Platelet contamination: of 134,159 single-donor platelet concentrates Implementation of antibody screening followed source and species of bacteria were transfused with a rate of septic transfu- by nucleic acid amplification techniques sion reaction of 13/32,333 and 10/134,159, (NAT) introduced in blood donor screening Several mechanisms may lead to platelet respectively. Thus, they observed a 5.4 fold during the last decade, have largely con- contamination, the major being contamination increased rate with pooled-donor platelets.16 In tributed to this reduction.1-5 Nevertheless, all by the skin flora at the site of punction.3,7 an AABB Educational Session in Transfusion blood products remain under the threat of Pathogens are essentially gram positive bacte- Medicine, Hillyer et al. reported that the preva- emerging blood-transmitted infections (virus, ria like Staphylococcus aureus, coagulase neg- 6 lence of the bacterial contamination of whole protozoans, helminths, bacteria, prions) and ative Staphylococci, viridans group Strepto- blood derived platelets was 33.9 per 100,000 bacterial transfusion related transmission is cocci, Bacillus spp., Corynebacteria as well as still associated to a high morbidity and mortal- anaerobic diphteroid gram positive bacilli such units compared to apheresis platelets whose 1,5,17 ity.1-3,5,7-10 Platelets are especially affected by this as Propionibacterium acnes.3,8,11 In 2004, in a prevalence was 51.0 per 100,000 units. risk, and septic infections are most commonly review dealing with transfusion-transmitted The effect of bacterial screening of aphere- associated with platelet transfusion because of bacterial infections, Wagner reported that sis platelets has been evaluated. Eder et al. the favorable bacterial growth conditions that about 56% of bacteria detected in platelet units pointed out that while septic reactions associ- are: i) room temperature storage allowing the implicated in clinical situations of transfusion ated with platelet transfusion were estimated growth of even small bacterial inoculums and, associated sepsis were aerobic gram-positive to occur in about 1:25,000 transfusions, the ii) the biological composition of platelet con- bacteria.10 However, in the presence of gram rate of contaminated platelets determined centrates.3-5,9,11,12 negative organisms, the outcome was more after implementation of detection techniques The estimated rate of bacterial platelet con- frequently fatal (60%) when compared with was about 1:2,000 to 1:3,000. The authors tamination is about 1/2,000-3,000 units (whole- gram positive ones (40%). The distribution of hypothesized that low-level contamination was blood and apheresis-derived platelets)1-3,5 and a bacteria found in cases of transfusion related not necessarily associated with a clear clinical severe sepsis may be associated with one out sepsis showed Staphylococcus spp. in 42% of response, and that more particularly, neu- of 6 of contaminated platelet units transfused.3 cases, Escherichia coli in 9%, Bacillus spp. in tropenic and febrile patients were often under The risk of bacterial contamination of platelet 9%, Salmonella spp. in 9%, Streptococcus spp. antibiotic therapy that may change the clinical concentrates has been estimated to be 50-250 in 12%, Serratia spp. in 8%, Enterobacter spp. in picture.18

[page 22] [Hematology Reviews 2009; 1:e4] Article

latency phase). Whereas in some cases fever teria in leukocyte reduced apheresis platelets. Platelet transfusion-related and chills are present during the transfusion, However, in the absence of sufficient data, a bacterial infection in many cases the patient remains asympto- comparison between this method and both matic. In the symptomatic recipient, fever and BacT/ALERT® and Pall eBDS® is not possible. chills may be observed within two hours after BacT/ALERT® is an automated colorimetric The occurrence of a severe bacterial infec- the start of transfusions. Hypotension, nausea, blood culture method allowing the detection of tion related to platelet transfusion has been vomiting, diarrhea, oliguria, respiratory symp- both aerobic and anaerobic bacteria as well as correlated to bacterial proliferation in platelet toms and shock may be observed. The severity yeasts and fungal microorganisms, in two cul- concentrates, and a bacterial contamination of a septic transfusion reaction also depends ture bottles, one for aerobic and one for anaer- 5 >10 CFU/mL has been considered as a serious on the patient’s characteristics and may be obic pathogens. The bottom of the bottles con- infectious risk.13 In 2006, Yomtovian et al. more severe in immunocompromised tains a pH sensitive liquid sensor which reported data from a surveillance program for 1,5,8 patients. changes its color according to the amounts of detection of bacterial contamination of In order to reduce the risk of post-transfu- CO2 released. Whereas bacteria produce CO2, platelets in a university hospital from 1991 to sion sepsis from bacterial contaminated yeasts and fungi are high acid-producing 2004.5 These data concern passive surveillance platelets, various procedures have been imple- organisms. Inoculation of approximately 7.5 (transfusion reaction-triggered) and active mented: improved donor selection, single- mL of the platelet sample is necessary for each surveillance (prospective methods). During donor apheresis platelets, reduction of the con- bottle, and is performed 24 hours after collec- the surveillance period, 216,283 platelet units tamination risk of the phlebotomy process by tion. The bottles are then incubated at 37°C. were transfused. Only one type of bacterial optimal skin disinfection procedures and the The CO2 production is correlated with the alter- contamination was detected by the passive removal of the first 10-30 mL of the initial col- surveillance method (Pseudomonas aerugi- ation of the reflection of the light on the sen- lected blood, storage time limitation, bacterial sor. When the color changes, an alarm is set nosa, 3-day old random-donor platelets). detection and inactivation methods.1,3,7 Indeed, Thirty-eight contaminated platelet units or off. The bottles are continuously analyzed and as the major cause of platelet contamination is tested for up to 5-7 days. This system has been pools were detected by active surveillance. Six the normal skin flora at the site of punction3,7 used for routine detection of platelet bacterial of them were not transfused. When transfused, optimal skin disinfection may significantly contamination in the Netherlands, Belgium the same contaminant was isolated from blood reduce contamination. However, when the col- and Wales for several years.4,8,10 cultures obtained from the recipient in 7 cases. lection needle enters the skin, bacteria may be In 2002, McDonald et al. reported data Coagulase-negative staphylococci was the introduced into blood. Different studies have resulting from the evaluation of the most frequently isolated bacteria species fol- suggested that the removal of the first aliquot BacT/ALERT® system after inoculation, in 2- lowed by Staphyloccocus aureus. Thirteen of of the initial collected blood may reduce the day old apheresis platelets, of different bacter- the patients who received the 32 contaminated risk of bacterial contamination in blood dona- ial species (9 gram positive, 5 gram negative) platelets (31 detected by active surveillance tion.1,20,21 and one detected by passive surveillance) and one fungal microorganism at a concentra- developed transfusion reactions. Transfusion tion of 10-100 CFU/mL. The results were com- reaction rate was significantly higher for pared to thioglycollate broth cultures. The pooled random-donor platelets. Severe transfu- Methods for detection of mean time for the bacterial detection ranged sion reactions occurred in 9 cases, and 3 between 9.1-48.1 hours, with the exception of bacterial contamination of 22 patients died of complications, likely to be Propionibacterium acnes (89-177.6 hours). associated with the transfusion of contaminat- platelets: culture methods Analogous results were obtained by Brecher et ed platelet units (Pseudomonas aeruginosa, al.,19 who also confirmed the large delay Staphylococcus aureus and Serratia marce- When compared with bacterial detection required for detection of Propionibacterium scens). Nine of the 13 transfusion-reactions methods, surrogate tests such as pH measure- acnes. However, it is important to note that were associated with bacterial counts >106 ment, glucose level determination or Gram’s anaerobic Propionibacterium acnes has a poor CFU/mL. The virulence of the bacteria stain coloration have been shown to have low growth in the aerobic platelet conditions, and appeared to be more important then the bacte- sensitivity and thus, can not be recommended.1 that its clinical relevance is all but clear.2,7,19,22 rial load. However, transfusion reactions with Two methods have been approved by the FDA As bacterial load may be very low in freshly col- fever, rigors and hypotension were observed for the detection of bacterial contamination in lected platelets (<10 CFU/mL sometimes even with coagulase-negative staphylococci at bac- platelets. The first one is BacT/ALERT® <1 CFU/mL)10 a storage period ≥24 hours is terial counts as low as 102 CFU/mL.5 (BioMérieux Inc.), an automated colorimetric required before inoculation,10,19,22 in order to As bacterial growth increased over time, blood culture method, based on the detection allow bacterial growth and to avoid false nega- platelet units, were older, the risk of high bac- of carbon dioxide produced by proliferating tive results. terial amounts was higher, and consequently microorganisms, which allows the detection of To better assess the prevalence and nature there was a higher risk of sepsis. In 1984, the both aerobic and anaerobic bacteria as well as of bacterial contamination of pooled and FDA allowed the extension of platelet storage yeasts and fungus.4,8,10 The second one is apheresis platelets, the GERMS (German from five to seven days. This decision led to an eBDS® (Pall Corporation), an enhanced bacte- Evaluation of Regular Monitoring Study) increased rate of platelet transfusion-related rial detection method based on the measure- Group of the Red Cross Transfusion Services infections, directly associated with the oldest ments of oxygen consumption by bacteria in initiated a prospective multimember study transfused units. Therefore, in 1986, the stor- the milieu.4,14 A comparison between the including 9 different centers. The results of age was reduced from seven to five days.5,19 advantages and disadvantages of both methods this study have been reported by Schrezen- As mentioned above, severity of the clinical is presented in Table 1 [adapted from (4)]. meier et al. in 2007. The BacT/ALERT® method manifestations related to contaminated The Verax Biomedical Platelet PGD® Test for was used to analyze 52,243 platelet concen- platelet units are determined by quantitative/ Bacterial detection (Verax Biomedical, Inc.) is trates (15,198 collected by apheresis and qualitative parameters (number of CFU a new rapid and qualitative immunoassay test 37,045 pooled platelets). Of the 282 platelet infused, type of bacteria, rate of proliferation, for the detection of aerobic and anaerobic bac- concentrates with a first positive culture, a

[Hematology Reviews 2009; 1:e5] [page 23] Article

Table 1. A comparison of the advantages and disadvantages of BacT/ALERT® and Pall eBDS®, according to the data reported by Schmidt et al.4 The sensitivity, efficiency and manageability of BacT/ALERT® and Pall eBDS® were compared in a multicenter-study carried out in 4 German centers. It concerned 6,307 pooled platelets and 4,730 apheresis.

Pall eBDS® BacT/ALERT® Advantages Sensitivity* Sensitivity* Good specificity** Detection of aerobic and anaerobic bacteria, yeasts, and fungal microorganisms Closed system The bottles are continuously analyzed and tested, for up to 5-7 days Easy and quick handling Easy and quick handling Disadvantages Detection of aerobic and only facultative Specificity* anaerobic bacteria A punctual measure in time whereas platelets Open system may be transfused until 5 days after collection Platelets may be released before the time of detection

*This evaluation revealed a better sensitivity for the BacT/ALERT® system, identifying 4 positive samples that were missed with the Pall eBDS®. **The specificity of BacT/ALERT® (0.25%, 28 out of 11,037 tested samples) was significantly lower compared to Pall eBDS® (0.03%, 3 out of 11,037 tested samples).

bacteria was identified in 135. Among these center for routine bacterial screening of inoculated into a sampling bag. The pouch con- 135 concentrates, 37 were confirmed positive platelets in October 2001. Over a period of two tains two tablets: i) a sodium polyanethol by second cultures. There was no significant years, 28,104 pooled platelets were screened: sulphonate dissolvable tablet that reduces the difference in the rate of confirmed positive 203 of them were detected as being contami- natural inhibitors of bacterial growth and also platelet concentrates between apheresis and nated and 125 out of 203 had already been acts as a platelet aggregating agent in order to pooled platelets. Among these 37 confirmed released at the time of the detection. Ninety lower platelet competitive consumption of O2 positive platelet concentrates, most bacteria percent of them were already transfused. No and, ii) a trypticase soy medium-containing isolated were skin flora bacteria. Propioni- adverse reactions such as fever, hypotension, tablet that enhances the sensitivity by provid- bacterium acnes was found in 54% of them, or other unexplained clinical deterioration, ing nutrients for bacterial growth. The sample and staphylococci species in 43% of them, with were observed after a contaminated transfu- is incubated for 24 hours at 35°C, and the level Staphylococcus epidermidis in most of the sion. Early detection (≤48 hours) of bacterial of O2 is measured.2,4,10,14 cases. Serratia marcescens and Staphylococcus contamination with microbiological confirma- Like BacT/ALERT®, the Pall eBDS® in vitro aureus, which are potentially pathogenic bac- tion was possible in 59 concentrates, and 48 of sensitivity was determined to be between 1-10 teria, were isolated respectively in one plasma them could be recalled. For late detection (>48 CFU/mL. Moreover, in order to reduce the risk pooled platelet concentrate and in one aphere- hours), only 33 of 125 contaminated concen- of false negative results, a period of storage sis platelet concentrate. This study demon- trates could be recalled. The authors also 10 ≥24 hours has been advocated. The perform- strated that a large scale routine screening of observed that 68% of the BacT/ALERT® screen- ance of the Pall eBDS® was evaluated in four platelets prepared by different blood centers ing cultures became positive after an incuba- test sites after inoculation 1-15 CFU/mL of 10 was feasible. However, the authors observed tion of 48 hours.15 They noted that this low rate different bacterial species known to be associ- that such a screening did not allow the preven- was in contrast with several studies that ated with fatal platelet transfusion related out- tion of the transfusion of contaminated units showed higher rates of bacterial detection come. The inoculated samples were trans- because of the interval of time needed after after an incubation of 48 hours. For instance, ferred in Pall eBDS® bags immediately after inoculation and the detection of the contami- Wagner and Robinette showed that deliberate inoculation or after a 24 hour storage at 22°C. nant. In addition, the risk of false negative bacterial inoculation in platelet concentrates The results were reported by Holme et al. who results was not eliminated by this approach.7 could be detected after 48 hours in all of the Two cases of life-threatening sepsis due to cases.15,23 Brecher et al. reported detection of showed that all samples incubated 24 hours Bacillus cereus contamination despite the inoculated microorganisms in apheresis after inoculation were detected as being con- BacT/ALERT® detection method were platelet units (with the exception of taminated (100% sensitivity) and no false pos- described by te Boekhorst et al. in the Propionibacterium acnes) in a mean time of itives were obtained with 713 uninoculated 2 Netherlands.15 9.3-18.9 hours (10 CFU/mL) or 8.7-18.2 hours platelets. In another important multi-center German (100 CFU/mL).15,24 Te Boekhorst et al. pointed Fournier-Wirth et al. reported the results study, comparing three bacterial detection out the fact that this difference could be from a study performed in 3 transfusion cen- methods under routine conditions, Schmidt et explained by a presumably very low bacterial ters in France on pooled and apheresis al. reported 2 severe transfusion reactions load in their platelet concentrates compared to platelets. The aim of this study was to evaluate after transfusion of 2 split apheresis platelet the experimental inoculation load.15 the ability of the Pall eBDS® system to detect concentrates contaminated with Klebsiella As previously mentioned, Pall eBDS® is a bacterial contamination after a reduction of pneumoniae, despite a negative screening culture method system based on the bacterial the incubation time. The results were com- with both BacT/ALERT® and Pall eBDS® detec- consumption of oxygen and only allows detec- pared with the BacT/ALERT® system, consid- tion methods.4 tion of aerobic and facultative anaerobic bacte- ered in this study as a reference. Low levels (5- In the Netherlands, te Boekhorst et al. ria. Twenty-four hours after collection, 3 mL of 50 CFU/mL) of 5 different strains of bacteria reported their experience with the the platelet concentrate are filtered in order to were inoculated in the platelet units. The BacT/ALERT® system implemented in their remove platelets and leukocytes, and then are platelets were stored at 22°C, 24 hours before

[page 24] [Hematology Reviews 2009; 1:e5] Article sampling. Time to detection ranged between 8- nated sample containing 1-10 CFU/mL, have 17 hours. Sixty-three contaminated bags were also been evaluated.10,25 Pathogen inactivation incubated for 18 and 24 hours at 35°C. Sixty- In 1999, Chaney et al. described a new one out of 63 (96.82%) were detected as being process allowing the targeting of bacterial ribo- Pathogen reduction technologies allow inac- positive after 18 hours and all at 24 hours. The somal RNA in five steps: i) cell lyses leading to tivation of viruses and bacteria in contaminat- O2 level of the 2 samples negative at 18 hours the release of ribosomal RNA, ii) hybridization ed platelet concentrates. Two main different (contaminated with Bacillus cereus) was near of bacterial ribosomal RNA with biotin and approaches have been described. the detection threshold. However, the negative ruthenium-labeled oligonucleotide probe pairs, samples detected an O2 level >17%. There were iii) capture of the labeled ribosomal RNA with Psoralen based method ® no false positive results (100% specificity). streptavidin-coated magnetic beads, and then The INTERCEPT blood system (Cerus They conclude that the Pall eBDS® method iv) setting of the RNA on an electrode surface Corporation) uses amotosalen which is a syn- allows testing of platelet concentrates 42 hours and detection of ruthenium-labeled ribosomal thetic psoralen, an organic compound found in 29 after collection similarly to the Bac/T ALERT® RNA by application of voltage, v) followed by fruit and vegetables like limes and celery. system.14 the generation of an electrochemiluminescent Amotosalen compound has the potential to The limitation of the Pall eBDS® detection signal. By this approach, the authors were able penetrate into cells, to cross the nuclear mem- system concerns the failure to detect anaero- to obtain a linear relationship between the brane, and to reversibly intercalate into helical bic bacteria. However, anaerobic organisms electrochemiluminescent signal representing regions of nucleic acid. Exposure to a long- are only rarely associated with fatal infection the ribosomal RNA level with the number of wave-length ultraviolet light (UVA, 320-400 after platelet transfusions.2,10,14 CFU/mL.26,27 However, even if this method nm) leads to covalent crosslinks between amo- The sensitivity, efficiency and manageabili- appeared to be suitable for routine application, tosalen molecules and pyrimidine bases, block- ty of BacT/ALERT® and Pall eBDS® were com- its sensitivity was not sufficient. It only allowed ing DNA and RNA replication (cells and pared in a multi-center study carried out in 4 the detection of approximately 105 CFU/mL.27 pathogens with nucleic acid genomes). After German centers.4 It concerned 6,307 pooled Störmer et al. recently published the results light treatment, the residual amotosalen as platelets and 4,730 apheresis platelets. A of a study evaluating bacterial spreading in the well as its metabolites are removed by a com- microbiological reference laboratory evaluated different blood components infected by inocu- pound adsorbing device during a prolonged all initially positive results. This evaluation lation of Klebsiella pneumoniae and Staphylo- incubation. Platelets do not have nuclei and 29-35 revealed a better sensitivity for the coccus epidermidis. Using RT-PCR, Klebsiella are not affected by psoralens. In 2005, Lin et BacT/ALERT® system, identifying 4 positive pneumoniae was detected in platelet concen- al. reported results that advocate the efficien- samples that were missed with the Pall eBDS®. trates immediately after its preparation. cy of such an approach for virus inactivation. However, the enhanced sensitivity of the Staphylococcus epidermidis, which has a slow- The authors studied 10 different families of BacT/ALERT® system was offset by a reduced er growth, was only detected 24 hours after the viruses including: i) the most relevant for specificity which was defined in this study as whole separation process. The authors con- blood transfusion like HIV-1, HIV-2, HBV, the number of false-positive test samples. The cluded that a 24 hour storage was necessary HTLV-1, HTLV-2 or CMV, ii) viruses of emerg- ® ing interest like parvovirus B19, West Nile specificity of BacT/ALERT (0.25%, 28 out of before processing to RT-PCR.17 virus, severe acute respiratory syndrome- 11,037 tested samples) was significantly lower Even if molecular based technologies repre- ® human coronavirus, and vaccinia virus, and, compared to Pall eBDS (0.03%, 3 out of 11,037 sent a high potential for bacterial detection, iii) model viruses like duck hepatitis virus (an tested samples). However, the authors conclud- their applicability in the context of detection of HBV model), bovine viral diarrhea virus (an ed that this reduced specificity might be bacteria in platelet concentrates has not yet ® HCV model), bluetongue virus, feline conjunc- acceptable. BacT/ALERT detected 32 positive been demonstrated. tivitis virus, simian adenovirus 15 and porcine samples with microbiological confirmation out Furthermore, their use is limited by the cost, ® parvovirus. According to the FDA, the process of 11,037 and Pall eBDS detected one positive complexity of use and, more critically, by the was defined as being effective if the pathogen sample. These samples were considered as ini- availability of bacterial-derived nucleic acid tially positive. However, these initially positive load was reduced by 6-10 logs. The results of amplification reagents.25 results were not confirmed by the analysis of a this study showed a significant log reduction In a review of the literature published in second sample, either from the satellite bag, or of enveloped viruses that were uniformly sen- 2004 about how to improve the bacteriological from the original platelet bag, or from a related sitive to inactivation. The non-enveloped safety of platelet transfusions, Blajchman et al, erythrocyte bag (in case of pooled platelet). A viruses showed a variable sensitivity to inacti- pointed out that all bacterial contaminants can- possible explanation may be an exogenous vation. Parvovirus B19 and human adenovirus not currently be detected by molecular methods contamination during the test procedure. For 5 were inactivated to the limit of detection, ® and that it is not clear if bacterial DNA or rRNA whereas the other non-enveloped viruses were that reason, Pall eBDS which is a closed sys- 28 is the most appropriate test marker. 31 tem may be preferable to BacT/ALERT® which resistant to inactivation. In 2006, Singh et al. One of the most important practical prob- is an open one. reported that the photochemical treatment of lems with the use of broad-range PCR is the plasma with amotosalen inactivates high lev- contamination of the assay by exogenous bac- els of gram positive (Streptococcus epider- terial DNA of the nucleotide amplification midis) and gram negative (Klebisella pneumo- Detection of bacterial reagents, particularly of the bacterial derived niae, Yersinia enterocolitica) bacterias. They enzymes and the bacterial DNA sequences contamination with molecular determined that the mean log reductions commonly found in human blood. Moreover, achieved were >7.3 for Streptococcus epider- based methods the detection of a minor amount of bacterial midis and Yersinia enterocolitica and >7.4 for DNA among a high quantity of human DNA Klebisella pneumoniae. Photochemical treat- In order to provide more rapid, sensitive and may also constitute a problem.25 To our knowl- ment was also efficient at high initial titers on highly specific results, molecular technologies, edge, methods based on bacterial amplification spirochetes with mean-reductions >5.9 for based on the detection of ribosomal RNA of a techniques are not currently employed in rou- Treponema pallidum and >10.6 for Borrelia wide variety of bacteria in a platelet contami- tine screening of platelet concentrates. burgdorferi and on protozoa with mean-reduc-

[Hematology Reviews 2009; 1:e5] [page 25] Article tion >6.9 for Plasmodium falciparum, >5 for tained <3×1011 platelets. Thus, patients who studies, the same changes in treated and non- Trypanosomia cruzi and >5.3 for Bancrofti received photochemical treated platelets treated platelets, without modification of FvW, microti. They also showed the maintenance of received more platelet transfusions and had a fibrinogen and FVa levels after five days of clotting time and plasma coagulation factor shorter interval between transfusion than storage. They also showed that treated activity after photochemical treatment.35,36 patients who received conventional platelets.41 platelets have adhesive and cohesive functions Genomic DNA in leukocytes is modified by pso- Studies performed in animals and humans similar to non-treated platelets.45 Ruane et al. ralens that are able to inactivate more than 5.4 reported no evidence for a toxicity of psoralen have shown that the Mirasol® PRT system is logs of T lymphocytes and to disrupt more base treatment.34,40,42 Webert et al. related acute toxi- able to inactivate viruses and bacteria in pairs (1:83) than gamma-irradiation city with amotosalen after UVA activation in platelet concentrates. They observed signifi- (1:37,000).35 Grass et al. reported that a dose of rats with a lethal threshold dose of 150 mg/kg. cant log reductions for cell-associated and cell- 0.05 μmol/L of amotosalen, which is 3,000 fold In dogs, central nervous system alterations free HIV, West Nile virus, porcine parvovirus, lower than the dose used for viruses and bacte- were observed after a threshold dose of and for gram positive and gram negative bacteria inactivation, was sufficient to inactivate T 30 mg/kg. These doses were respectively ria such as Staphylococcus epidermidis or lymphocytes after 1 J/cm2 of UVA illumina- 150,000 and 30,000 fold higher than the dose Escherichia coli. The authors noted that the tion.33 The extreme sensitivity of the T lympho- used for pathogen inactivation. Reproductive platelet pH and the lactate production rate cytes to psoralens suggests that this treatment toxicity, determined by histological rather than were, as suggested by the literature data, pre- has the potential to reduce the incidence of functional evaluation, was observed at a dictive of recoveries in 50.8-59.8%. Therefore, leukocyte mediated adverse immune reactions threshold of 0.35 mg/kg, which was 350 times they used platelet pH and lactate production associated with platelet transfusion like trans- higher than the dose used for pathogen inacti- rate as platelet quality indicators and conclud- fusion associated graft-versus-host disease vation. Moreover, all substances used for ed that platelet cell quality was maintained and platelet related febrile non-hemolytic pathogen inactivation are water soluble and after treatment and during storage.46 Goodrich 12,33,37 transfusion reaction. Platelet concentrates rapidly excreted avoiding the bioaccumulation et al. reviewed Mirasol® PRT performances and treated with psoralens seem to have a compa- of trace amounts in treated blood products.35 In confirmed that this system was able to inacti- rable in vitro function compared with non- addition, transfusion of platelets or plasma vate viruses and various bacteria species such treated platelet concentrates and their viability treated with this method did not appear to as Staphylococcus epidermidis, Staphylococcus 32,38-40 seem to be conserved. induce adverse immunological response when aureus, Escherichia coli, Pseudomonas aerugi- In 2004, McCullough et al. reported the evaluated by searching the presence of nosa, Bacillus cereus, Serratia marcescens. results of the SPRINT trial which was a 29,35,41 neoantigens. These data suggested a significant reduction prospective, randomized, controlled, double- In 2005, Lin et al. evaluated the potential of of the risk of platelet related bacteria transmis- blind parallel group phase III study carried out photochemical treatment with amotosalen to sion.16,44,47 Kumar et al. showed an increased to evaluate the efficacy and the safety of photo- create neoantigens. They quantified the genomic DNA degradation in leukocytes and chemical treated with amotosalen platelets amounts of residual amotosalen and photo- bacteria after riboflavin and UV light treat- compared with non-treated control platelets products in photochemical treated platelets ment,48 also suggesting that Mirasol® PRT tech- collected by apheresis. The efficacy was and plasma. Patients’ serum samples from 7 nology may be an alternative to gamma irradi- defined by prevention and treatment of signif- Phase III clinical trials including 523 patients ation to prevent transfusion associated graft- icant bleeding: proportion of patients with who received more than 8,000 units of treated versus-host disease.43 WHO grade 2 bleeding as primary efficacy end platelets or plasma, were assayed by enzyme In December 2007, Klein et al. reported in a point, the proportion with WHO grade 3-4 linked immunsorbent assay (ELISA) for anti- bleeding, number of days of WHO grade 2 consensus conference about pathogen inactiva- bodies to amotosalen neoantigens. The results tion (Toronto, March 2007) that pathogen inac- bleeding, one hour and 24 hour platelet count indicated that no neonatigens were detected increments, corrected count increments, num- tivation methods should be implemented as by ELISA after photochemical treatment with soon as feasible and safe methods to inactivate ber of days to next platelet transfusion, num- amotosalen.29 ber of platelet transfusion incidence of platelet a broad spectrum of infectious agents are avail- 12 refractoriness and number of erythrocyte con- Riboflavin based method able. The authors established a list of existing centrate transfusions as secondary efficacy The Mirasol® PRT (Navigant Biotechnolo- criteria and procedures that should be changed end points. Safety end points were defined by gies Inc.) system is similar to the psoralen- in case of implementation of pathogen inactiva- the number of platelet transfusion reactions, based method, but uses a different photo-sen- tion such as the suppression of screening tests development of antibody to potential amotos- sitizer, riboflavin (vitamin B2) instead of pso- for: i) Treponema pallidum; ii) agents of low alen neoantigens and overall safety. The study ralen. Riboflavin is a natural component found infectious titer and destroyed early by pathogen included 671 thrombocytopenic patients need- in food (milk, beer, eggs, yeasts, leafy vegeta- inactivation like West Nile virus, which is actu- ing platelet transfusion support: 318 received bles), and is classified as a “Generally- ally systematically tested in the US and Canada; photochemical treated platelets, 327 control Regarded-As-Safe” compound by the FDA.30,43,44 iii) agents that are sensitive to pathogen inacti- platelets and 26 of them did not require Riboflavin interacts with nucleic acids after vation for which redundant safety measures are platelet transfusion. This trial showed that exposure to UV light (280-360 nm) and causes taken (CMV, HTLV, HbsAg and those for which whereas the hemostatic effect of both photo- irreversible damage to DNA/RNA (direct elec- the methods of detection available nowadays chemical treated and control platelets were tron transfer, production of singlet oxygen and lack specificity and sensitivity like those used to comparable, the transfusion of treated production of hydrogen peroxide leading to the detect bacteria). Moreover, gamma irradiation platelets was associated with a lower platelet formation of hydroxyl radicals). A compound of blood components performed to eliminate the count increment after transfusion. This lower adsorbing device removal process for residual risk of transfusion associated graft-versus-host platelet count increment could partly be riboflavine metabolites may not be neces- disease could be eliminated. The authors con- explained by the lower mean platelet dose in sary.30,43 Perez-Pujol et al. have studied the clude that an agent that is known to be ade- the photochemical treated group (3.7×1011 vs. impact of this method on the functional and quately inactivated by these technologies 4×1011 in the control group; p<0.01) and by a biochemical characteristics of platelet concen- should not require screening tests unless of an greater proportion of treated platelet that con- trates. They observed, with flow cytometry unusually high infectious titer.12

[page 26] [Hematology Reviews 2009; 1:e5] Article

the authors of this review), the choice [Transfusion-transmitted bacterial infec- Towards a change of paradigm: between all these different approaches is very tion: residual risk and perspectives of pre- inactivation versus detection difficult. vention]. Transfus Clin Biol 2003; 10:192- 200. The ability of pathogen reduction technolo- 14. Fournier-Wirth C, Deschaseaux M, Defer gies to inactivate a broad spectrum of organ- C, et al. Evaluation of the enhanced bacte- isms (virus, fungus, bacteria, parasites) is one References rial detection system for screening of con- of the most convenient answers to face the taminated platelets. Transfusion 2006; 46: rapidly evolving epidemiological environment 1. Hillyer CD, Josephson CD, Blajchman MA, 220-4. as well as the continuous appearance of new et al. Bacterial contamination of blood 15. te Boekhorst PA, Beckers EA, Vos MC. pathogens. Multiple different factors effective- components: risks, strategies, and regula- Clinical significance of bacteriologic ly contribute to the occurrence of emerging or tion: joint ASH and AABB educational ses- screening in platelet concentrates. re-emerging infectious agents: migration, sion in transfusion medicine. Hematology Transfusion 2005;45:514-9. travel, conflicts, climatic changes, demogra- Am Soc Hematol Educ Program 2003;575- 16. Ness P, Braine H, King K, A et al. Single- phy, and numerous less trivial factors like for 89. donor platelets reduce the risk of septic instance pet trade through e-business.6,49-52 The 2. Holme S, McAlister MB, Ortolano GA, et al. platelet transfusion reactions. Transfusion occurrence of pathogens with a strong epidem- Enhancement of a culture-based bacterial 2001;41:857-61. ic potential, and/or with high prevalence, as detection system (eBDS) for platelet prod- 17. Stormer M, Cassens U, Kleesiek K, Dreier well as the diversity of existing pathogens that ucts based on measurement of oxygen J. Detection of bacteria in platelet concen- are not systematically detected using standard consumption. Transfusion 2005;45:984-93. trates prepared from spiked single dona- screening approaches, strongly argue for the 3. Palavecino E, Yomtovian R. Risk and pre- tions using cultural and molecular genetic introduction of inactivation procedures rather vention of transfusion-related sepsis. Curr methods. Transfus Med 2007;17:61-70. than continuously introducing new biological Opin Hematol 2003;10:434-9. 18. Eder AF, Kennedy JM, Dy BA, et al. tests, each being characterized by its own sen- 4. Schmidt M, Karakassopoulos A, Burkhart Bacterial screening of apheresis platelets sitivity and specificity. J, et al. Comparison of three bacterial and the residual risk of septic transfusion detection methods under routine condi- reactions: the American Red Cross experi- tions. Vox Sang 2007;92:15-21. ence (2004-2006). Transfusion 2007; 47: 5. Yomtovian RA, Palavecino EL, Dysktra AH, 1134-42. Conclusions et al. Evolution of surveillance methods for 19. Brecher ME, Means N, Jere CS, et al. detection of bacterial contamination of Evaluation of an automated culture system The bacterial transmission still constitutes platelets in a university hospital, 1991 for detecting bacterial contamination of a significant problem in transfusion medicine. through 2004. Transfusion 2006; 46:719- platelets: an analysis with 15 contaminat- Because of their favorable bacterial growth 30. ing organisms. Transfusion 2001; 41:477- conditions (storage at room temperature, bio- 6. Alter HJ, Stramer SL, Dodd RY. Emerging 82. logical composition) platelets are of special infectious diseases that threaten the blood 20. Bruneau C, Perez P, Chassaigne M, et al. concern. As shown in this review, data avail- supply. Semin Hematol 2007;44:32-41. Efficacy of a new collection procedure for able suggest that bacterial detection with the 7. Schrezenmeier H, Walther-Wenke G, preventing bacterial contamination of technology available nowadays may present a Muller TH, et al. Bacterial contamination whole-blood donations. Transfusion 2001; false sense of security. To our knowledge, there of platelet concentrates: results of a 41:74-81. is no recently approved “revolutionary” bacter- prospective multicenter study comparing 21. Wagner SJ, Robinette D, Friedman LI, ial detection technology that can dramatically pooled whole blood-derived platelets and Miripol J. Diversion of initial blood flow to and definitively make platelet transfusion safe. apheresis platelets. Transfusion 2007; prevent whole-blood contamination by By contrast, pathogen inactivation has demon- 47:644-52. skin surface bacteria: an in vitro model. strated its efficiency. However, long-term stud- 8. Besson Faure I. Rapid screening for bacte- Transfusion 2000;40:335-8. ies are still needed to demonstrate the safety rial contamination of blood products. J Lab 22. McDonald CP, Rogers A, Cox M, et al. of this approach. Newer pathogen inactivation Med 2006;30:91-100. Evaluation of the 3D BacT/ALERT automat- technologies are currently under development, 9. Ribault S, Harper K, Grave L, et al. Rapid ed culture system for the detection of like CryoFacet red blood cell and platelet tech- screening method for detection of bacteria microbial contamination of platelet con- nology using counterflow elutriation to remove in platelet concentrates. J Clin Microbiol centrates. Transfus Med 2002;12:303-9. infectious agents in plasma, followed by ozona- 2004;42:1903-8. 23. Wagner SJ, Robinette D. Evaluation of an tion and treatment with germicidale UV before 10. Wagner SJ. Transfusion-transmitted bac- automated microbiologic blood culture releasing.35 Other approaches such as “glyco- terial infection: risks, sources and inter- device for detection of bacteria in platelet engineering” the platelets in order to allow ventions. Vox Sang 2004;86:157-63. components. Transfusion 1998;38:674-9. their cold storage and rapid bacterial detection 11. Easley S, Yomtovian RA, Sullivan P, Jacobs 24. Brecher ME, Hay SN, Rothenberg SJ. (at the moment of release) are under develop- MR. Development of proficiency testing for Evaluation of a new generation of plastic ment.35,53 detection of bacterial contamination of culture bottles with an automated micro- In summary, this review describes recent platelet products. Transfusion 2007;47: bial detection system for nine common technologies that may be used to make platelet 251-5. contaminating organisms found in PLT transfusion safe, with a particular emphasis 12. Klein HG, Anderson D, Bernardi MJ, et al. components. Transfusion 2004;44:359-63. on the prevention of bacterial contamination. Pathogen inactivation: making decisions 25. Dreier J, Stormer M, Kleesiek K. Real-time Based on the literature review, and more par- about new technologies. Report of a con- polymerase chain reaction in transfusion ticularly for people who are not directly sensus conference. Transfusion 2007;47: medicine: applications for detection of involved in the development and/or implemen- 2338-47. bacterial contamination in blood products. tation of these particular technologies (like 13. Morel P, Deschaseaux M, Bertrand X, et al. Transfus Med Rev 2007;21:237-54.

[Hematology Reviews 2009; 1:e5 [page 27] Article

26. Chaney R, Rider J, Pamphilon D. Direct Proceedings of a consensus conference: 44. Goodrich RP, Edrich RA, Li J, Seghatchian detection of bacteria in cellular blood prod- pathogen inactivation-making decisions J. The Mirasol PRT system for pathogen ucts using bacterial ribosomal RNA-direct- about new technologies. Transfus Med Rev reduction of platelets and plasma: an ed probes coupled to electrochemilumi- 2008;22:1-34. overview of current status and future nescence. Transfus Med 1999;9:177-88. 36. Singh Y, Sawyer LS, Pinkoski LS, et al. trends. Transfus Apher Sci 2006;35:5-17. 27. Rider J, Newton A. Electrochemilumine- Photochemical treatment of plasma with 45. Perez-Pujol S, Tonda R, Lozano M, et al. scent detection of bacteria in blood compo- amotosalen and long-wavelength ultravio- Effects of a new pathogen-reduction tech- nents. Transfus Med 2002;12:115-23. let light inactivates pathogens while nology (Mirasol PRT) on functional 28. Blajchman MA, Goldman M, Baeza F. retaining coagulation function. Trans- aspects of platelet concentrates. Trans- Improving the bacteriological safety of fusion 2006;46:1168-77. fusion 2005; 45:911-9. platelet transfusions. Transfus Med Rev 37. McCullough J. Pathogen inactivation: a 46. Ruane PH, Edrich R, Gampp D, et al. 2004;18:11-24. new paradigm for blood safety. Trans- Photochemical inactivation of selected 29. Lin L, Conlan MG, Tessman J, et al. fusion 2007;47:2180-4. viruses and bacteria in platelet concen- Amotosalen interactions with platelet and 38. Lin L, Cook DN, Wiesehahn GP, et al. trates using riboflavin and light. Trans- plasma components: absence of neoanti- Photochemical inactivation of viruses and fusion 2004;44:877-85. gen formation after photochemical treat- bacteria in platelet concentrates by use of 47. Engelfriet CP, Reesink HW, Blajchman MA, ment. Transfusion 2005;45:1610-20. a novel psoralen and long-wavelength R et al. Bacterial contamination of blood 30. Solheim BG, Seghatchian J. Update on ultraviolet light. Transfusion 1997; 37:423- components. Vox Sang 2000;78:59-67. pathogen reduction technology for thera- 35. 48. Kumar V, Lockerbie O, Keil SD, et al. peutic plasma: an overview. Transfus 39. van Rhenen DJ, Vermeij J, Mayaudon V, et Riboflavin and UV-light based pathogen Apher Sci 2006;35:83-90. al. Functional characteristics of S-59 pho- reduction: extent and consequence of DNA 31. Lin L, Hanson CV, Alter HJ, et al. tochemically treated platelet concentrates damage at the molecular level. Photochem Inactivation of viruses in platelet concen- derived from buffy coats. Vox Sang 2000; Photobiol 2004;80:15-21. trates by photochemical treatment with 79:206-14. 49. Morris JG, Potter M. Emergence of new amotosalen and long-wavelength ultravio- 40. Pineda A, McCullough J, Benjamin RJ, et pathogens as a function of changes in host let light. Transfusion 2005;45:580-90. al. Pathogen inactivation of platelets with susceptibility. Emerg Infect Dis 1997; 32. Janetzko K, Lin L, Eichler H, et al. a photochemical treatment with amotos- 3:435-41. Implementation of the INTERCEPT Blood alen HCl and ultraviolet light: process used 50. Chomel BB, Belotto A, Meslin FX. Wildlife, System for Platelets into routine blood in the SPRINT trial. Transfusion 2006; exotic pets, and emerging zoonoses. bank manufacturing procedures: evalua- 46:562-71. Emerg Infect Dis 2007;13:6-11. tion of apheresis platelets. Vox Sang 2004; 41. McCullough J, Vesole DH, Benjamin RJ, et 51. Marano N, Arguin PM, Pappaioanou M. 86:239-45. al. Therapeutic efficacy and safety of Impact of globalization and animal trade 33. Grass JA, Hei DJ, Metchette K, et al. platelets treated with a photochemical on infectious disease ecology. Emerg Inactivation of leukocytes in platelet con- process for pathogen inactivation: the Infect Dis 2007;13:1807-9. centrates by photochemical treatment SPRINT Trial. Blood 2004;104:1534-41. 52. Gayer M, Legros D, Formenty P, Connolly with psoralen plus UVA. Blood 1998; 42. McCullough J. Pathogen inactivation of MA. Conflict and emerging infectious dis- 91:2180-8. platelets. Transfusion alternatives in eases. Emerg Infect Dis 2007;13:1625-31. 34. Allain JP, Bianco C, Blajchman MA, et al. transfusion medicine 2006;2:121-6. 53. Sorensen AL, Hoffmeister KM, Wandall Protecting the blood supply from emerging 43. Seghatchian J, de Sousa G. Pathogen- HH. Glycans and glycosylation of platelets: pathogens: the role of pathogen inactiva- reduction systems for blood components: current concepts and implications for tion. Transfus Med Rev 2005;19:110-26. the current position and future trends. transfusion. Curr Opin Hematol 2008; 35. Webert KE, Cserti CM, Hannon J, et al. Transfus Apher Sci 2006; 35:189-96. 15:606-11.

[Hematology Reviews 2009; 1:e5] [page 28] Hematology Reviews 2009; volume 1:e6

Human heart-type fatty bicin.1 Anthracyclines cause a dose-dependent cardiomyopathy that often leads to congestive Correspondence: Ashraf ElGhandour, acid-binding protein as an heart failure. Late onset cardiomyopathy can Hematology Unit, Internal Medicine Department, early diagnostic marker of appear months to years after treatment is com- Alexandria, Egypt. doxorubicin cardiac toxicity pleted.2 The mechanism underlying cardiotox- E-mail: [email protected] ic effects of anthracyclines is generally accept- Received for publication: 13 January 2009. 1 1 ed to be via formation of free radicals generat- Ashraf H. ElGhandour, Manal El Sorady, Revision received: 9 April 2009. Sahar Azab,2 Mohammed ElRahman3 ed by iron-doxorubicin complexes that damage Accepted for publication: 17 April 2009. cardiac cellular membranes.3 Cardiac damage 1 2 Hematology Unit, Cardiology caused by anthracyclines is cumulative. With This work is licensed under a Creative Commons Department, Faculty of Medicine, total doses of doxorubicin less than 500 mg/m2, Attribution 3.0 License (by-nc 3.0) Alexandria University heart failure is seen in less than 7% of cases.4 ©Copyright A. ElGhandour et al., 2009 3 Human heart fatty acid-binding protein (H- Military Medical Academy, Alexandria, Licensee PAGEPress, Italy Alexandria, Egypt FABP) is a small protein abundant in the Hematology Reviews 2009; 1:e6 cytosol, which is readily released into the cir- doi:10.4081/hr.2009.e6 culation following myocardial damage. Recent studies in laboratories and the emergency Abstract department have shown that heart-type fatty Patients’ age ranged from 14 to 56 years acid-binding protein (H-FABP), a more recent- with a mean of 36.2±10.02 years. Twenty-two Progressive cardiotoxicity following treat- ly developed cardiac biomarker, is able to of them were males, while 18 were females. ment with doxorubicin-based chemotherapy in detect myocardial damage as soon as one hour Ten age and sex-matched healthy controls with patients with non-Hodgkin’s lymphoma (NHL) after onset of ischemia and, therefore, is no history of cardiac problems were included may lead to late onset cardiomyopathy. So, regarded the earliest plasma marker avail- in our study. None of our patients had symp- early prediction of toxicity can lead to preven- able.5,6 A bedside test for H-FABP, providing toms or signs of cardiac disease at presentation of heart failure in these patients. The aim results within 15 min,7 could potentially reduce tion, nor had they undergone previous cardiac of this work was to investigate the role of H- diagnostic uncertainty for patients suspected surgery. Patients with severe hepatic or renal FABP as an early diagnostic marker of anthra- of ACS in primary care. Recently, Setsuta and disease or other medical conditions in which cycline-induced cardiac toxicity together with colleagues reported that elevated levels of H- combination chemotherapy may be contraindi- brain natriuretic peptide (BNP) as an indica- FABP were associated with subsequent cardiac cated were excluded from our study. Also, tion of ventricular dysfunction in such events in patients with chronic heart failure patients with early stage disease (stage I and patients. Our study was conducted on 40 NHL due to a variety of causes.8 On the other hand, non-bulky stage II) who are candidates for patients who received 6 cycles of a doxorubicin brain natriuretic peptide (BNP) is an amino involved-field irradiation in addition to containing chemotherapy protocol (CHOP), acid peptide chiefly secreted by the ventricular chemotherapy were excluded. not exceeding the total allowed dose of doxoru- myocardium in response to strain. Thus, it At presentation, full history and clinical bicin (500 mg/m2). Ten healthy controls were may be viewed as a marker of myocardial load examination, complete blood picture, bone included in our study. Human heart-type fatty and the plasma measurement of BNP is being marrow examination and lymph node biopsy acid-binding protein (H-FABP) was assessed used increasingly in the diagnosis, prognosis were carried out in all patients. Routine labo- 24 hours after the first cycle of CHOP. Plasma and monitoring of patients with congestive ratory investigations, ECG and echocardiogra- levels of BNP were estimated both before start- heart failure.9,10 Since heart failure is a com- phy were also performed. ing chemotherapy and after the last cycle of plex clinical syndrome, a single biochemical Blood samples were obtained for estimation CHOP. Resting echocardiography was also per- marker may not reflect all of its characteristics. of BNP before starting the first cycle of chemo - formed before and at the end of chemotherapy Thus, the serial and combined measurements therapy and after completion of the sixth cycle. cycles. The ejection fraction (EF) of 8 of our of markers of myocyte injury may open new Blood was withdrawn for assessment of H- patients decreased below 50% at the end of the perspectives in heart failure. 8 The aim of our FABP within 24 hours of administration of dox- sixth cycle. Elevated levels of both H-FABP and work was to investigate the value of H-FABP as orubicin in the first cycle of CHOP12 H-FABP BNP were found in all patients wth EF below an early diagnostic marker of anthracycline- was measured utilizing human cardiac fatty 50% and both markers showed a positive corre- induced cardiotoxicity together with BNP as an acid-binding protein ELISA test kit, Oxis lation with each other. We concluded that H- indicator of ventricular dysfunction in non- research catalog number: 11230, Oxis Intern - FABP may serve as a reliable early marker for Hodgkin’s lymphoma patients receiving doxo - ational Inc., Foster City, CA, USA.13 prediction of cardiomyopathy induced by dox- rubicin-based chemotherapy. Re-evaluation of ECHO findings regarding orubicin. Thus, in patients with elevated H- the left ventricular ejection fraction and FABP, alternative treatment modalities with no parameters of diastolic function14 was carried cardiac toxicity may be considered in order to out after the end of the sixth cycle of chemo- prevent subsequent heart failure in these Materials and Methods therapy. patients. Our study was conducted on 40 patients with non-Hodgkin’s lymphoma who received 6 cycles of an anthracycline containing regimen Results Introduction (CHOP) with the following doses:11 Cyclophosphamide: 750 mg/m2 IV Day 1 Patients’ data are summarized in Table 1. Anthracyclines are highly efficacious anti- Doxorubicin: 50 mg/m2 IV Day 1 The mean age of our patients at presentation neoplastic agents but their utility is limited by Vincristine: 1.4 mg/m2 IV Day 1 was 36.2±10.02 years. Fifty-five percent were progressive cardiotoxicity. They include doxo- Prednisone: 100 mg/d PO Days 1-5 males, while 45% were females. All patients rubicin, daunorubicin, epirubicin and idaru- This cycle was repeated every 21 days. had normal ECG findings with no clinical evi-

[Hematology Reviews 2009; 1:e6] [page 29] Article dence of heart failure prior to chemotherapy. Table 1. Biochemical markers and ECHO findings in our patients. After completing the first cycle of CHOP, N. Minimum Maximum Mean plasma levels of H-FABP were found to be elevated in 10 patients. After completion of the Age 40 14 56 36.2± 10.02 sixth cycle of chemotherapy, 8 of the 10 BNP before chemotherapy (pg/mL) 40 50 89 71.35±10.04 patients with elevated H-FABP showed a sig- BNP after chemotherapy (pg/mL) 40 50 280 108.32±77.66 nificant reduction in the left ventricular ejec- H-FABP (ng/mL) 40 8 30 17.25±6.44 tion fraction to levels below 50% and the EF before chemotherapy (%) 40 58 72 65.57±4.04 reduction in EF showed a significant correla- EF after chemotherapy (%) 40 40 70 61.4±9.79 tion with H-FABP levels (Figure 1). Six of them also showed evidence of diastolic dysfunction in the form of impaired relaxation and an E/A ratio below 1. Two patients of those with post- Table 2. Comparison between group I and II patients as regards BNP and H-FABP. chemotherapy ejection fraction values less N. Mean T p than 50% had clinical manifestations of heart failure. Fifteen of our patients (37.5%) showed Age GI 32 38.84±10.02 -1.79 0.081 evidence of diastolic dysfunction after 6 cycles GII 8 41.75±8.44 of chemotherapy but 3 of them already had E/A BNP before chemotherapy (pg/mL) GI 32 70.28±9.45 -1.185 0.265 ratios below 1 when evaluated before starting GII 8 75.62±11.8 chemotherapy. We divided our patients into BNP after chemotherapy (pg/mL) two groups according to echocardiographic GI 32 70.4±9.52 -28.048 0.000* evidence of left ventricular systolic dysfunc- GII 8 260±18.5 tion after the sixth cycle of chemotherapy H-FABP (ng/mL) (Table 2): Group I; patients who had left ven- GI 32 13.84±2.68 -13.103 0.000* tricular ejection fraction levels below 50%; GII 8 27.15±2.53 Group II; patients who had no echocardio- EF before chemotherapy (%) graphic evidence of heart failure. GI 32 65.84±4.14 0.903 0.384 On comparing the mean plasma BNP val- GII 8 64.5±3.66 ues in both groups, we found that prior to EF after chemotherapy (%) chemotherapy, they did not show a significant GI 32 65.65±5.02 15.47 0.000* difference (mean value group I 70.28±9.45, GII 8 44.37±2.97 group II 75.62±11.84). After six cycles of che - * p Significant . GI: Patients with post chemotherapy EF above 50% . GII: Patients with post- chemotherapy EF below 50%. motherapy, both groups differed significantly as regards plasma BNP, with the mean value in group I being 70.4±9.527 and in group II being 260±18.51 (p<0.001). There was also a signifi- HFABP cant elevation of BNP values in group II R = - 0.836 p<0.001 patients after chemotherapy and this elevation correlated significantly with the reduced ejection fraction in these patients (Figure 2). There were no patients in whom clinical cardiac failure occurred without an associated rise in the BNP level above the threshold value. From the 10 patients who exhibited elevated H- FABP levels after the first chemotherapy cycle, 8 also showed high BNP levels at the end of the sixth cycle and both biochemical markers cor- Figure 1. Correlation related positively with each other (Figure 3). between ejection fraction after 6 cycles of CHOP and HFABP.

Discussion

Cardiotoxicity leading to congestive heart trations of these biomarkers and the function- increase and concentrations over 100 ng/mL.16 failure is a well-known complication of anthra- al alterations associated with doxorubicin-in- On the other hand, Daugaard and colleagues cyclines. Biochemical methods to assess and duced myocardial damage. A significant corre- concluded that in spite of correlations between monitor cardiac function after anthracycline lation between the left ventricular ejection peptide concentrations and reduced ejection administration, if informative would be of fraction after 6 cycles of CHOP and plasma lev- fraction values, neither baseline levels nor 15 utmost value. We examined the diagnostic els of both HFABP and BNP was found in our serial measurements can safely substitute EF role of human heart-type fatty acid binding patients. Pichon and colleagues stated that an monitoring in patients undergoing anthracy- protein and brain natriuretic peptide to predict infra-clinical cardiotoxicity of anthracyclines cline therapy.10 Brain natriuretic peptide the impairment of left ventricular function in as defined by BNP elevation is frequent but (BNP) was originally discovered in the porcine NHL patients treated by CHOP. We also studied reversible and that patients who developed brain but was subsequently found to be pre- the correlations between the plasma concen- heart failure showed a continuous BNP dominantly a cardiac hormone.17 Unlike atrial

[page 30] [Hematology Reviews 2009; 1:e6] Article natriuretic peptide (ANP), which is secreted by BNP the atria in response to increased atrial pres- R = - 0.892 sure, BNP is derived chiefly from the cardiac p<0.001 ventricles in response to ventricular stresses.18 The plasma levels of both peptides are inversely correlated with measures of cardiac function and recent studies have shown BNP to be a more sensitive marker of cardiac impairment than ANP.19, 20 Raised plasma BNP levels have previously been shown to herald the clinical picture of cardiac failure by days to weeks.21 Various guidelines based on changes in systolic Figure 2. Correlation and diastolic left ventricular function deter- between ejection frac- mined either by ECHO or by radionuclideven- tion after 6 cycles of triculography (RVG) have been proposed for CHOP and BNP levels. monitoring patients receiving anthracycline therapy. The advantages of ECHO over RVG are better availability, lower costs and lack of exposure to ionizing radiation.16 Although H-FABP is viewed as a marker of myocyte injury and BNP is considered indicative of ventricular strain, BNP R = - 0.892 the finding that both markers correlated with p<0.001 each other in our patients shows that some degree of myocyte injury is associated with increased ventricular load. It is well-known that myocardial structure is also altered in congestive heart failure (CHF). Non-contiguous areas of myocardial cell death and foci of replacement fibrosis are typical morphological changes in advanced CHF. Therefore, cytosolic proteins may be released into the circulation through leakage due to increased permeability of the membranes of injured myocytes.22 Thus, the Figure 3. Correlation present study highlights the practical impor- between HFABP& BNP in our patients. tance of measuring some biochemical markers such as plasma H-FABP and BNP to monitor left ventricular dysfunction in patients receiving anthracycline therapy as an early indication of subclinical cardiotoxicity.

References al. Fatty acid-binding protein and the early mo poietic stem cell transplantation with 1. Johnson SA, Richardson DS. Anthra- detection of acute myocardial infarction. plasma brain natriuretic peptide. Bone cyclines in haematology: pharmacokinet- Clin Chim Acta 1998;272:87-92. Marrow Transplant 2000;26:309-13. ics and clinical studies. Blood Rev 1998; 6. van Nieuwenhoven FA, Kleine AH, Wodzig 10. Daugaard G, Lassen U, Bie P, et al. Natri- 12:52-71. WH, et al. Discrimination between myo- uretic peptides in the monitoring of 2. Kremer LC, van Dalen EC, Offringa M, et cardial and skeletal muscle injury by anthracycline-induced reduction in left al. Anthracycline-induced clinical heart assessment of the plasma ratio of myoglo- ventricular ejection fraction. Eur J Heart failure in a cohort of 607 children: long- bin over fatty acid-binding protein. Fail 2005;7:87-93. term follow-up study. J Clin Oncol 2001; Circulation 1995;92:2848-54. 11. McKelvey EM, Gottlieb JA, Wilson HE, et al. 19:191-6. 7. Chan CP, Sum KW, Cheung KY, et al. Hydroxydaunomycin (adriamycin) combi- 3. Shadle SE, Bammell BP, Cusack BJ, et al. Development of a quantitative lateral-flow nation chemotherapy in malignant lym- Daunorubicin cardiotoxicity: evidence for assay for rapid detection of fatty acid-bind- phoma. Cancer 1976;38:1484-93. the importance of the quinone moiety in a ing protein. J Immunol Methods 2003; 279: 12. Hermann-Arnhof KM, Hanusch-Enserer U, free-radical-independent mechanism. Bio- 91-100. Kaestenbauer T, et al. N-terminal pro-B- chem Pharmacol 2000;60:1435-44. 8. Setsuta K, Seino Y, Ogawa T, et al. Ongoing type natriuretic peptide as an indicator of 4. Coukell AJ, Faulds D. Epirubicin. An updat- myocardial damage in chronic heart fail- possible cardiovascular disease in severely ed review of its pharmacodynamic and ure is related to activated tumor necrosis obese individuals: comparison in patients pharmacokinetic properties and therapeu- factor and Fas/Fas ligand system. Circ J with different stages of heart failure. Clin tic efficacy in the management of breast 2004;68:747-50. Chem 2005;51:138-43. cancer. Drugs 1997;53:453-82. 9. Snowden JA, Hill GR, Hunt P, et al. 13. Wodzig KW, Pelsers MM, van der Vusse GJ, 5. Glatz JF, van der Vusse GJ, Simoons ML, et Assessment of cardiotoxicity during hae - et al. One-step enzyme-linked immunosor-

[Hematology Reviews 2009; 1:e6] [page 31] Article

bent assay (ELISA) for plasma fatty acid- assays and radionuclide ventriculography. 20. Davidson NC, Naas AA, Hanson JK, et al. binding protein. Ann Clin Biochem 1997; In vivo 2005;19:567-76. Comparison of atrial natriuretic peptide, 34:263-8. 17. Mukoyama M, Nakao K, Hosoda K, et al. B-type natriuretic peptide, and N-terminal 14. Choy AJ, Darber D, Lang C, et al. Detection Brain natriuretic peptide as a novel car- proatrial natriuretic peptide as indicators of left ventricular dysfunction after acute diac hormone in humans. J Clin Invest of left ventricular systolic dysfunction. Am myocardial infarction: comparison of clini- 1991;87:1402-12. J Cardiol 1996;77:828-31. cal, echocardiographic and neurohormon- 18. Yasue H, Yoshimora M, Sumida H, et al. 21. Hunt PJ, Richards AM, Nicholis MG. et al. al methods. Br Heart J 1994;72:16-22. Localisation and mechanism of secretion Immunoreactive amino-terminal pro-brain 15. Germanakis I, Kalmanti M, Parthenakis F, of B-type natriuretic peptide in normal natriuretic peptide (NT-proBNP): a new et al. Correlation of plasma N-terminal subjects and patients with heart failure. marker of cardiac disease. Clin Endocrinol pro-brain natriuretic peptide levels with Circulation 1994;90:195-200. 1997;47:287-96. left ventricle mass in children treated with 19. Richards AM, Nicholis MG, Yandle TG, et 22. Schaper J, Froede R, Hein St, et al. Im - anthracyclines. Int J Cardiol 2006;108:212- al. Plasma N-terminal pro-brain natriuret- pairment of the myocardial ultrastructure 5. ic and adrenomedulin. New neurohormon- and changes of the cytoskeleton in dilated 16. Pichon MF, Cvitkovic F, Hacene K, et al. al predictors of left ventricular function cardiomyopathy. Circulation 1991; 83:504- Drug-induced cardiotoxicity studied by and prognosis after myocardial infarction. 14. longitudinal B-type natriuretic peptide Circulation 1998;97:1921-9.

[page 32] [Hematology Reviews 2009; 1:e6] Hematology Reviews 2009; volume 1:7

Palliative splenic irradiation in ences in terms of treatment efficacy were seen among dose groups, hematologic toxicity rates Correspondence: Mario Federico, Policlinico primary and post PV/ET distributed differently. Severe cytopenias Universitario Paolo Giaccone, U.O. Radioterapia, myelofibrosis: outcomes occurred in 50% of courses in the HDG, and in DiBiMeL, Sezione Scienze Radiologiche, and toxicity of three radiation the 14.3% and in 0% of the IDG and LDG, via del Vespro 127, 90127 Palermo, Italy respectively. Spleen reduction and pain relief E-mail: [email protected] schedules lasted for a median of 5.5 months in all groups. Due to the efficacy and tolerability of the low- Key words: myelofibrosis, splenomegaly, pallia- 1,5,6 3 tion, radiotherapy, low-dose irradiation. Mario Federico, Guido Pagnucco, dose irradiation 4 patients from the LDG and Antonio Russo,5,6 Giovanni Cardinale,3 IDG were retreated and received on the whole Acknowledgments: our grateful thanks go to Prof. Patrizia Guerrieri,2 Francesco Sciumè,2 12 RT courses. Multiple retreatments did not Giovanni Barosi who kindly reviewed the manu- 6 6,7 Catherine Symonds, Letizia Cito, show decremental trends in terms of rates of script. Sergio Siragusa,4 Nicola Gebbia,5 response to radiation nor in terms of duration Roberto Lagalla,1 Massimo Midiri,1 of clinical response. Moreover, retreatment Received for publication: 16 January 2009. Antonio Giordano,6 Paolo Montemaggi2 courses did not cause an increased rate of Revision received: 2 April 2009. Accepted for publication: 17 April 2009. 1U.O. Radioterapia, DiBiMeL, Università adverse effects and none of the retreated patients experienced severe hematologic toxi- degli Studi di Palermo, Italy This work is licensed under a Creative Commons cities. The average time of clinical benefit in 2U.O. Radioterapia, ARNAS Civico, Attribution 3.0 License (by-nc 3.0) retreated patients was much longer (21 Dipartimento Oncologico M. Ascoli, months, range 44-10) than patients who were ©Copyright Mario Federico et al., 2009 Palermo, Italy not retreated (5.75 months, range 3-6). Licensee PAGEPress, Italy 3U.O. Ematologia, ARNAS Civico, Dipartimento Oncologico M. Ascoli, Hematology Reviews 2009; 1:e7 Palermo, Italy doi:10.4081/hr.2009.e7 4U.O. Ematologia, Università degli Studi Introduction di Palermo, Italy patient’s discomfort. Also even though a new 5U.O. Oncologia Medica, Dipartimento di Primary myelofibrosis1 (PM) is a Phila- generation of “target drugs” are currently Chirurgia e Oncologia, Università degli delphia negative chronic myeloid disorder under intense investigation with some encour- Studi di Palermo, Italy (CMD) currently classified with polycythemia aging results, splenomegaly control still remains a crucial step for patients’ quality of 6Sbarro Health Research Organization, vera (PV) and essential thrombocythemia (ET) 2 life improvement. Temple University, Philadelphia, USA as a chronic myeloproliferative disease To date, splenectomy or splenic irradiation 7CROM, Centro Ricerche Oncologiche, (MPDs). PM is a rare disease mainly affecting 3 (SI) are the sole treatment modalities to con- Mercogliano, Italy older people with a median survival of 3.5-5 years.4 The pathogenetic mechanism is not trol drug resistant splenomegaly in MF clearly understood but probably relates to a patients. When technically achievable splenec- clonal stem-cell disorder that leads to ineffec- tomy is currently the preferred treatment tive erythropoiesis, dysplastic megakaryocyte modality for MF based upon good, long-lasting Abstract hyperplasia and an increased ratio of imma- outcome in term of organomegaly-related ture to total granulocytes.5 These findings are symptom palliation.9,10 Unfortunately, it is con- Splenectomy and splenic irradiation (SI) characteristically accompanied by reactive sistently associated with a significant rate of are the sole treatment modalities to control bone marrow (BM) fibrosis that develops and mortality as well as intra- and peri-operative drug resistant splenomegaly in patients with is mediated by megakaryocyte-derived fibro- complications.9,10 SI, instead, has been general- myelofibrosis (MF). SI has been used in poor genic cytokines.6 ly preferred in patients not undergoing surgery surgical candidates but optimal total dose and Collagen fibrosis, presumably along with due to a poor general status or decline and fractionation are unclear. We retrospectively many other factors, interferes with normal allows for a good but transitory splenomegaly reviewed 14 MF patients with symptomatic hematopoietic processes, ultimately leading to palliation. In fact, the major shortcoming of splenomegaly. Patients received a median of erythroid hypoplasia.7,8 Due to BM fibrosis, in radiation is that its palliative effect on 10 fractions in two weeks. Fraction size ranged MF patients as well as those with post ET/PV splenomegaly generally does not last longer from 0.2-1.4 Gy, and total dose varied from 2- MF, an extramedullary hematopoietic process than six months. 10.8 Gy per RT course. Overall results indicate starts in the spleen or in multiple organs as an There is a general agreement that emerges that 81.8% of radiation courses achieved a sig- attempt to override BM failure, often leading to from the literature to use RT at dose levels nificant spleen reduction. Splenic pain relief the development of splenomegaly or hepato - lower than in other hematologic malignancies; and gastrointestinal symptoms reduction were splenomegaly. Moreover, splenomegaly exacer- however, few studies have taken a retrospec- obtained in 94% and 91% of courses, respec- bates cytopenias through the sequestration tive look at SI.11 The indication for SI is still tively. Severe cytopenias occurred in 13% of and destruction of hematopoietic elements.9 controversial12 and there is not a precise radiation courses. Furthermore patients were Progressive high-grade splenomegaly unequivocal definition of the optimal total divided in three groups according to the radia- occurs in the majority of MF patients. dose and fractionation, mainly due to the lim- tion dose they received: 6 patients in the low- Unfortunately the standard current pharmaco- ited number of patients included in existing dose group (LDG) received a normalized dose logical therapeutic options, due to their short studies and the wide range of radiation sched- of 1.67 Gy; 4 patients in the intermediate-dose periods of response, fail to control organo- ules adopted. Moreover, it is unclear if re-irra- group (IDG) received a normalized dose 4.37 megaly and organomegaly-associated symp- diation of MF patients is a safe strategy to Gy; the remaining 4 patients in the high-dose toms (abdominal pain and early satiety, weight extend the overall time of clinical benefit that group (HDG) received a normalized dose of 9.2 loss, portal hypertension and profound a single SI course allows. Here we aim to Gy. Subgroup analysis showed that if no differ- fatigue), which account for much of the assess outcomes and complication rates of

[Hematology Reviews 2009; 1:e7] [page 33] Article splenic irradiation in three cohorts of patients prognostic parameters13 (Hb levels <10 g/dL years, patients received different total doses treated with different “low-dose” irradiation Land WBC <4 or >30×109/L) in three cate- and dose per fraction. To compare the various schedules. gories: high- , intermediate- and low-risk. Four RT treatments we used the Normalized Tumor patients belonged to the high-risk, 4 to the low- Dose14 (NTD10), defined as the total dose deliv- risk and 6 to the intermediate-risk groups, ered in 2 Gy fractions that corresponds to a respectively. All patients had already under- particular biologically effective dose level and Patients and Methods gone a cytoreductive pharmacological treat- is calculated according to the formula: ment: 8 received hydroxyurea as a single After approval from our institutional modality treatment, one received hydroxyurea research review committee, we retrospectively plus Ara-C, 2 patients received hydroxyurea reviewed data concerning 15 patients (10 plus 6-mercaptopurine, one received hydrox- male, 5 female, median age at diagnosis 61 yurea and melphalan, one received busulphan years, median age at first irradiation 67 years), and 3 patients were given thalidomide in asso- 11 with a histologically proven diagnosis of PM ciation with conventional cytoreductive treat- and 4 with a post ET-MF complaining of a high- ments. Radiation treatment was delivered by a grade symptomatic splenomegaly that were Siemens 15 MV Linac with multi leaf collima- where n is the number of RT fractions and d consecutively referred to our institution from tor; all patients had a CT scan simulation (slice the fraction size in Gy. The value of the 1997-2007 (Table 1). All patients had a drug thickness 10 mm) in the supine position. The α/β Linear Quadratic Model15 was empirically fixed resistant splenomegaly and lacked any further treatment planning system (Plato system v to 10 as for early responding tissues. By stan- treatment options. Before being admitted to 2.6.3.) was used and no patient immobilization dardizing the delivered dose of all 22 adminis- radiation, patients had previously been judged devices were adopted during the simulation unfit for surgery due to their general status or and treatment. tered treatments into a 2 Gy isoeffective treat- had refused splenectomy. Fourteen out of 15 Two portal arrangements were alternatively ment, we were able to make a correct radiobi- underwent splenic irradiation and one was used to encompass the entire spleen volume: ological comparison among different RT excluded due to pre-existing advanced heart antero-posterior (AP-PA), opposed parallel or schedules. On the basis of the NTD values, failure (patient 14). In the 14 irradiated opposed tangential in the attempt to reduce patients were divided into three different patients the first course of radiation occurred the dose to the left kidney. If the left kidney groups but it should be underlined that RT at a median of 58 months from the diagnosis of was displaced posteriorly, a tangential schedules were not chosen on the basis of MF. All 14 treated patients had a severe arrangement was provided; if the kidney was patients’ clinical parameters but rather were splenomegaly with splenic pain, abdominal displaced medially an antero-posterior dependent on the progressive modification of discomfort, and weight loss; 11 patients (84%) approach was arranged. In the planned evalua- our institutional treatment philosophy. had in concurrence constitutional symptoms tion process between target coverage and kid- The initial patients, who had received a total such as night sweats, low-grade fever and an ney sparing we assigned priority to left kidney dose of 10 Gy with a dose per fraction in the initial state of cachexya. All except 3 required sparing in order to reduce the total dose to the order of 1 Gy, were designated as our high- red blood cell transfusions (≥2 units per organ in case of multiple courses of splenic dose group (HDG). Patients who had received month). irradiation. our current standard of treatment (0.2 Gy frac- Patients were scored (at the time of their Since our institutional standards of radiation up to a total dose of 2 Gy in 10 fractions) first irradiation) on the basis of Dupriez’s tion for MF have changed during the past ten were designated as our low-dose group (LDG).

Table 1. Patients’ characteristics at the time of first irradiation. Patient N. Age at Interval Dupriez Previous Symptoms RBC Pre- Pre- irradiation diagnosis score treatments at time transfusion irradiation WBC irradiation PLT and sex irradiation of radiation U/month ×109/L ×109/L (intent to treat) 1* 53 F 14 y. HR HU S , P 2U 39.9 232 2* 65 F 16 y. IR HU S , P, CS 4U 6.48 381 3 62 F 3 y. LR HU; MPH S , P 2U 9.55 21 4* 75 F 8 y. LR HU S , P, CS 2U 8.36 44 5 67 M 4 y. IR HU S , P, CS 2U 5.7 210 6 67 M 8 y. IR HU; Ara-C S , P NT 38.1 190 7 77 M 8 y. LR HU; 6-MP S , P, CS 2U 10.3 423 8 87 M 1 y. HR B; Th S , P, CS 2U 2.63 119 9 46 M 4 y. IR HU S , P, CS NT 29.6 307 10 67 M 1 y. IR HU S , P, CS 2U 4.9 121 11 70 M 2 y. HR HU S , P, CS 2U 89 143 12 58 M 7 y. LR HU; Th S , P, night sweats NT 10.89 329 13* 65 M 2 y. IR HU; 6-MP; Th S , P, CS 2U 8 673 14 76 M 4 y. HR HU S , P, CS 4U 1.04 67 15 55 F 4 y. HR HU; Th S , P, CS 2U 42.51 143 S: Splenomegaly; P: splenic pain; CS: constitutional symptom; HU: Hydroxyurea; Th: Thalidomide; 6-MP: 6-Mercaptopurine; ARA-C: Arabinosylcytosine; MPH: Melphalan; B: Busulphan; HR: High-risk; IR: Intermediate-risk; LR: Low-risk; NT:Not transfused.*Post ET - MF.

[page 34] [Hematology Reviews 2009; 1:e7] Article

Whereas the intermediate-dose group (IDG) ity in patients’ characteristic due to the accru- patients had an improvement in their body reflected the transition or better our “dose al criteria, we did not perform any statistical weight while SI was less effective in reducing finding effort” toward lower doses with the aim data analysis, as it would not be statistically patients’ transfusion requirements. In only to reduce treatment related toxicities representative or pertinent. 35.3% (6/17) of courses there was a slight (Supplementary Tables). The IDG encompass- improvement of anemic state, but this was es patients who had received a wide range of transient and shorter than spleen size reduc- treatment with radiobiological characteristics, tion and pain relief. which, in some cases, may partially overlap Results Within the entire study population, grade 4 with the LDG. However, we decided to aggre- RTOG life-threatening cytopenias occurred in gate our patients in this way in order to obtain Total delivered dose per RT course ranged 21.5 % of patients (3/14) or 13.6% of RT cours- homogeneity in the low- and in the high-dose from 2 to 10.8 Gy, the dose per fraction varied es. In all cases it developed in the first week groups. from 0.2-1.4 Gy. RT courses were generally after completing radiation and required hospi- Patients in which three or more of the fol- administered over a two week period (median talization. Interestingly, RT complications dis- lowing criteria were present were considered number of fraction per RT course was 10), tributed differently among groups. In the LDG, responsive to treatment: subjective absence of patients received RT five days per week contin- no grade 4 RTOG adverse effects occurred. MF-related gastrointestinal symptoms (bulky uously; 4 patients had multiple courses of RT, Patients in the IDG experienced 14.3% of RT effect), absence of splenic pain, consistent and one patient received 4 courses. In the first courses followed by severe cytopenias (1/ 7 reduction of the spleen volume (not less than group (low-dose group, LDG), 6 patients courses), while in the HDG, 50% of RT treat- 50% of the initial size) assessed by clinical received a median NTD of 1.67 Gy (0.6 stan- ments were too toxic (Supplementary Figure examination (according to the formula: spleen dard deviation). In the second (intermediate- 1). Both non-responding patients (patients #9 volume = 4/3π 1° diameter ° 2° diameter ° 3° dose group, IDG), 4 patients had a median and 8) experienced severe acute complica- diameter) and, finally, reduction of the RBC NTD of 4.37 Gy (1.89 standard deviation). The tions. One (patient #8) appeared to have been transfusion units required per month. third group (high-dose group, HDG) contained rescued from cytopenia but three months later To evaluate toxicity and response to treat- 4 patients who received a normalized median developed a leukemic transformation that led ment, patients had undergone clinical exami- dose of 9.2 Gy (0.46 standard deviation). to death. The second patient (patient #9), com- nation and blood cell count twice a week dur- According to the above-defined criteria, 12 plained of a massive splenomegaly, did not ing the period of irradiation and for the follow- patients were considered responsive. Overall respond to SI and underwent splenectomy 12 ing two weeks. If no toxicity occurred, blood response rates after all 22 RT treatments indi- months after RT. One month after splenectomy tests were scheduled weekly for the following cate that 81.8% of courses achieved a signifi- the patient died as a result of sepsis. month. The evaluation of the spleen reduction cant spleen size reduction; however, better The median time of symptom relief after a was carried out 20 days after patients had com- results were achieved on splenic pain relief single RT course was 5.5 months and no differ- pleted radiation. (94.45% of RT courses) and reduction of gas- ences were found among dosage groups. Treatment related toxicity was limited to trointestinal symptoms (91% of courses). No According to the patients’ general conditions, myelosuppression and was measured on the significant difference in terms of spleen size the cumulative RT dose delivered and the rate basis of RTOG acute toxicity scoring criteria. reduction and splenic pain relief emerged after of spleen shrinkage in response to previous An RT course after which a post-radiation subgroup analysis. Patients in the LDG had irradiation, retreatment after splenic relapse grade 4 (WBC count ≤1×109/L and/or PLTs spleen size reduction and splenic pain relief in was considered in 4 patients. The 4 retreated count ≤20×109/L) acute cytopenia had devel- 91% and 100% of courses, respectively, while patients received on the whole 12 RT courses oped was considered too toxic. RBC count was in the IDG and in the HDG, 76.5% and 75% of and one patient received 4 courses without any excluded from toxicity scoring because almost courses obtained a spleen size reduction. Pain acute toxicity. Two of the retreated patients all patients were already heavily transfused relief was achieved in 86% and 100%, respec- belonged to the LDG and the remaining 2 to long before receiving RT. Due to the small size tively (Table 2). the IDG. However, it is important to note that of the study cohort and the lack of homogene- After completing radiation all responsive of the patients retreated from the IDG, one

Table 2. Splenic irradiation results (by NTD group). Number of patients Median Number Median Dose NTD10 Response PM Post dose of courses per fraction m sd % of courses % of courses Median duration Hematologic ET MF delivered with reduction with pain relief of response toxicity in spleen size (in months) Grade 4 RTOG

LDR 33200 cGy 11 20 cGy 1.67 Gy 0.60 91% 100% 6 months 0% (2-4 Gy range) (range 3-12) HDR 401000 cGy 4 110 cGy 9.20 Gy 0.46 75% 100% 4 months 50% range (range 6-0) ( 2/4 ) (980-1080cGy)

IDR 31500 cGy 7 50 cGy 4.37 Gy 1.89 76.5% 86% 5 months 14.3% Range (range 6-0) (1/7 ) (300-800 cGy)

Low-dose group (LDG); high-dose group (HDG); intermediate-dose group (IDG); median (m) standard deviation (sd).

[Hematology Reviews 2009; 1:e7] [page 35] Article received treatments of 0.3 up to 3 Gy in 10 frac- with the latter increasing to 26% when the ent number of fractions, as well as the differ- tions total, which could be considered radiobi- three-month post-splenectomy period was con- ent schedule of irradiation reported, makes it ologically partially overlapping with the treat- sidered, and a morbidity rate of 39.3% and difficult to define a standard of treatment.33 ment dosages of the LDG. 31%, respectively. After splenectomy, up to 25% In order to be able to make dose-effect com- In comparison to the first irradiation, multi- of patients may experience accelerated parisons, the major drawback of some of the ple retreatments did not show decremental hepatomegaly and extreme thrombocytosis.21 published series is that the total dose and the trends in terms of rates of response to radia- Moreover, splenectomy has also been correlat- fractionation scheme seem not to be decided tion nor in terms of duration of clinical ed to a significantly higher incidence of blast up-front from the treatment but modified dur- response. Even in the case of one patient, who transformation. ing the irradiation on the basis of the single received 4 RT courses, there was no change in A large Italian study demonstrated a crude patient response with a consequent high vari- the duration of symptoms’ palliation. More- transformation rateinsplenectomized patients ability in the total dose, fractionation and over- over, after retreatment courses we did not of 26.4% in comparison to 11.9% in non- all treatment time. Some authors31 used the observe an increased rate of adverse effects splenectomized patients with the cumulative common 5 daily fractions per week schedule and none of the retreated patients experienced actuarial transformation rate of 55% in but increased the fraction size during the radi- severe hematologic toxicities. The average splenectomized vs. 27% in non-splenectomized ation course (from 0.4-0.5 Gy/fraction in the time of clinical benefit (Supplementary Figure patients at 12 years after diagnosis. The over- first week of treatment, up to 0.8-1 Gy/fraction 2) in retreated patients was very much longer all relative risk of blast transformation was during the following weeks) until the palliative (21 months, range 44-10) than patients who therefore 2.61 times higher among splenec- effect or toxicity is reached. Other authors30 were not retreated (5.75 months, range 3-6). tomized patients.22 In conclusion, despite the give radiation 2-3 times per week with an impact on symptoms, no overall survival bene- altered time factor. Both such approaches can fit has been demonstrated after splenectomy9,23 be empirically effective but generate data that on the contrary, this procedure is associated are difficult to compare with the common Discussion with a substantial risk of operative mortality, radiobiology algorithms that are based on larg- early and late morbidity and is contraindicated er daily fraction sizes (around 2 Gy) and with Splenomegaly rapidly occurs in all MF in patients with thrombocytosis. Furthermore, a time of inter-course sub-lethal DNA damage patients and is one of the causes of major dis- splenectomy has been shown to be a predictor repair of 24 hours between fractions. Given comfort. Curative treatments are to-date still of treatment failure in case of allo-BMT.16 that it is hard to make radiobiological compar- limited in MF. Allogenic bone marrow trans- Alternative treatments to manage spleno - isons among some published series, it is clear plantation (allo-BMT) has shown promising megaly, with lower morbidity and mortality that, still now, the most critical issue regard- results in younger patients but its role in eld- rates, would offer a significant improvement ing a rational use of RT is the definition of an erly patients is controversial. In particular, sev- in the clinical management of MF patients. optimal total dose and fractionation. eral studies suggest that in individuals older Radiotherapy has been used in selected sit- The leading idea of our approach to SI has than 45 the treatment’s risk-related mortality uations to control extramedullary hemato- been to adopt a relative long fractionation outweighs the benefits.16 On the contrary, poiesis, as in spinal localizations,24 in pul- schedule of 10 fractions in two weeks inde- other studies more recently explored the use monary hypertension25 or in symptomatic pendent of the total dose delivered with the of allo-BMT also in patients older than 60 with hepatomegaly26 with promising results. intent to generate comparable results, also in some interesting results.17,18 Currently, BMT in However, its role in splenic palliation remains case of treatments differing in total dose and the elderly is still a matter of debate since the controversial because of the lack of robust data dose per fraction. This approach should also number of patients accrued in clinical trials is (Table 3). It has been shown that splenic irra- minimize the incidence rate of post-attinic limited and the follow-up time short. Since MF diation can be very effective in reducing severe cytopenias and favor a rapid recovery of remains a disease of the elderly, standard and spleen size and splenic pain with response early blood precursors from RT. In fact, since a palliative treatments to manage cytopenias rates comparable to splenectomy.33 The major strong dose-sparing effect of fractionation on and massive organomegaly still retain a rele- shortcoming of radiotherapy is the reliance on bone marrow precursors34,35 has been proven, vant role in a consistent proportion of patients. its transient effect that normally does not we believed that it would be meaningful to also Splenomegaly can be effectively controlled exceed six months. apply this concept to extramedullary hemato- by conventional cytotoxic chemotherapy19 until As an alternative to splenectomy, SI has poiesis sites. Therefore, we decided to utilize a patients become drug resistant. More recently been considered in poor surgical candidates or long RT schedule (median 10 fractions) even atiangiogenic drugs and target drugs are in patients who declined surgery. In these when it could appear unjustified to do so due expected to offer a new chance of treatment patients, that generally are in a worse condi- to the minimal total dose delivered. for all patients. In particular a new class of tion compared to those that undergo surgery, Regarding the total dose, at the beginning molecules designed to inhibit Jak have been palliative splenic irradiation has shown mor- of our experience, we adopted an aggressive tested in different phase II trials with positive tality rates that are comparable to splenecto- RT regimen (1 Gy per fraction up to a total results.20 Jak inhibitors have shown a signifi- my.27 On the other hand, a high rate of severe dose of 10 Gy) but we observed a high inci- cant activity on splenomegaly but there is no life-threatening cytopenias has been reported dence of severe side effects. This raised the reason to think that, along with their use, also in patients that underwent splenic irradiation, concern that the same stem clonal disorder resistant patients will be selected. ranging from 32% (16/50 courses) of the Mayo that underlies MF could make hematopoietic After massive splenomegaly is established, Clinic series27 where lower doses of RT were precursors more sensitive to radiation. In splenectomy is considered the principle pallia- used (median dose per course 2.775 Gy) to order to reduce the incidence of acute cytope- tive measure because it offers a lengthy relief 35% (6/17 courses) of a French series31 where nias we progressively reduced total RT doses of symptoms. Unfortunately splenectomy is a more aggressive treatment was delivered until we established our actual standard of weighted by significant morbidity and mortali- (median dose per course 9.8 Gy). care (0.2 Gy per fraction up to 2 Gy total dose). ty rates. The two largest single institution Although a general trend in favor of low Our findings show that extremely low-dose series from Barosi10 and Tefferi9 reported a doses is emerging in the literature, the wide treatments are isoeffective as compared to mortality rate of 8.4% and 9%, respectively, variability of total radiation doses, the differ- higher dose regimens in effectively reducing

[page 36] [Hematology Reviews 2009; 1:e7] Article

Table 3. Synoptic table of published data on palliative SI in myelofibrosis. Author Number of patients Median # Median Estimated NTD10 Response MF Post dose of RT dose Median Dev % of courses % of courses Median duration PV/ET delivered courses per standard with reduction with pain of response fraction in spleen size relief (In months)

Elliot27 18 5 277.5cGy 50 50 cGy 3,162 Gy 2,784 94% 96% 6 range range (30-1365 Gy) (1-41) Greenberger28 13 1 650 cGy range 21 57.14 cGy 5,807 Gy 3,204 95% 100% NV (40-1728 cGy) range (1-73)

Parmentier29 54690 cGy range 12 25cGy 5,845 Gy 6,451 92% NA NA (180-2900 cGy) range (12,5-75) Wagner30 60NA NA NA NA NA 80% 63% NA From 200-450 cGy in 25-50 cGy fraction 3 times per week

Bouabdallah31 15 0 980cGy 17 Daily fr. NA NA 81% 90% Spleen size reduction (60-3050 cGy) 40-100 cGy 6 months range median (1-24 months) duration Splenic pain 7 months 22 days Range (1-19 months) Mc Farland 32 42range 13 Irradiation NA NA 92% NA MF: 1-16 months 300-600 cGy twice wk: Post PV/MF: 2-12 months 1stw50cGy 2ndw75cGy 3rdw100cGy

Present study LDG 33200 cGy 11 20 cGy 1.67 Gy 0.603 91% 100% 6 months range (range 3-12) (200-400 cGy) HDG 401000 cGy 4 110 cGy 9.205 Gy 0.465 75% 100% 4 months range (range 6-0) (980- 1080 cGy) IDG 31500 cGy 7 50 cGy 4.375 Gy 1.892 76,5% 86% 5 months range (range 6-0) (300-800 cGy)

NA: not assessable.

splenomegaly. Unfortunately, we cannot tion number, and fraction size, to be able to tion could be that in the definition of toxicity explain the functionality of low-dose treatment correctly compare different treatments we criteria, differing from the Mayo report, we did regimens in being so effective as compared to used the NTD formula, a radiobiological tool not consider hemoglobin levels since the high-dose treatments; however, these findings commonly used in the clinic to evaluate the majority of our patients were transfused from biological effectiveness of modified RT frac- a long time before receiving radiotherapy. are in concordance with the hypothesis of low- 10 36 tionations. The overall NTD of all 22 RT Another possible explanation could rely on the dose hypersensitivity. The suggestive issue of courses in our series is 2.59 Gy, a value compa- medical treatment that patients received radiobiology has been intensely investigated rable with the median NTD10 estimated from before undergoing radiation: in fact, it is inter- in vitro37 and postulates a hypersensivity state the Mayo series (3.16 Gy). Interestingly our esting to note that the only 2 patients in both of cells when irradiated at low doses (<0.4-0.5 patients seem to have a lower overall incidence series that received melphalan as medical Gy). Recently, there have been several indirect rate of grade 4 RTOG (13.6% of courses vs. treatment before radiation later experienced confirmations of this theory in clinical studies, 32%). This discrepancy is somehow difficult to severe post-attinic cytopenias. linking low-dose hypersensivity to tumor be explained since there are just slight differ- To compare outcomes after different radia- ences in the normalized radiation dose that regression38 as well as to the occurrence of tion doses we stratified our patients into three patients of the two groups received. Even a 39 groups according to the NTD10 value they adverse effects, at dose levels under the slighter difference in terms of patient charac- threshold generally accepted for toxicity or teristics can be found between the Mayo Clinic received. We found that, if no differences in tumor control. series and ours (median age at the time of the terms of spleen shrinkage or pain relief Since in our series, as well as in others first irradiation 65 vs. 67 years; time intercur- emerged among patients who underwent dif- reported,27-29,31 there is an inherent discrepancy ring between diagnosis and irradiation 44 vs. ferent RT regimens, daily fractions of 0.2 Gy up due to variability in total dose delivered, frac- 58 months, respectively). A possible explana- to 2 Gy is significantly the safer fractionation

[Hematology Reviews 2009; 1:e7] [page 37] Article scheme since it is not associated to grade 4 rence of toxicity. We conclude that our actual 10. Barosi G, Ambrosetti A, Buratti A, et al. hematologic toxicities. In our patients, inde- standard of 2 Gy delivered in 10 fractions over Splenectomy for patients with myelofibro- pendently from the dose, radiotherapy was two weeks has a NTD of 1.67 Gy, a value two- sis with myeloid metaplasia: retreatment very effective in reducing massive spleno - to three-fold lower than other published series. variables and outcome predictions. megaly, but did not resolve completely the This schedule of treatment has been shown to Leukemia 1993;7:200-6. spleen enlargement (Online Supplementary be extremely well tolerated and to date in our 11. Mesa RA, Tefferi A. Surgical and radiother- Tables). It is possible to argue that, since we experience is not associated with severe apeutic approaches for myelofibrosis with found a safe RT schedule, it would be mean- hematologic toxicities. Such optimal treat- myeloid metaplasia. Semin Oncol 2005;32: ingful to prolong the radiation treatment until ment compliance encouraged repeating irradi- 403-13. a complete splenomegaly remission. On the ation in responsive patients and this favored a 12. Coughlin C, Papac R, Roberts K. Leu- contrary, we decided to maintain a conserva- drastic increase in the average time of clinical kemia. In: Perez C, Brady L. Halperin EC tive approach and to stop the treatment once benefit. eds. Principles and practice of radiation the planned final dose was achieved. Two main oncology. 4th ed. 2004. Lippincott Williams considerations led to our decision: first of all & Wilkins, Philadelphia pp. 2142-3. the fact that the palliative effect of radiothera- 13. Dupriez B, Morel P, Demory JL et al. py seems to last no longer than six months References Prognostic factors in agnogenic myeloid independently from the dose delivered. We metaplasia. A report on 195 cases with a were concerned that reducing the spleen size 1. Mesa RA, Verstovsek S, Cervantes F, new scoring system. Blood 1996;88:1013-8. until normalization could result in a small Barosi G et al. Primary Myelofibrosis 14. Fowler JF, Tomè WA, Fenwick JD et al. A increase of the time free from symptoms at the (PMF) post polycythemia vera myelofibro- challenge to traditional radiation oncolo- cost of a probably higher incidence of severe sis (post-PV MF), post essential thrombo- gy. Int J Radiat Oncol Biol Phys 2004;60: cytopenias. Secondly, since the aim of our cythemia myelofibrosis (post-ET MF), 1241-56. treatment was strictly palliative, we consid- blast phase PMF (PMF-BP): Consensus on 15. Fowler JF. The Linear Quadratic formula and progress in fractionated radiotherapy. ered it meaningful, once symptom relief was terminology by the international working Br J Radiol 1989;62:679-94. achieved, to stop the treatment with the intent group for myelofibrosis research and 16. Cervantes F. Myelofibrosis: biology and to minimize the patient’s absorbed dose per treatment (IWG-MRT). Leuk Res 2007; treatment options. Eur J Haematol 2007; course of RT in order to potentially be able to 737-40. 79:13-7. repeat the treatment in the future. 2. Tefferi A. Classification, diagnosis and management of myeloproliferative disor- 17. Samuelson SJ, Sandmaier BM, Heslop HE In fact, because of the low incidence of mild ders in the JAK2V617F Era. Hematology et al. Allogenic hematopoietic cell trans- adverse effects in the LDG (and in the lower Am Soc Hematol Educ Program 2006;240- plants in patients with myelofibrosis age dose burden of the IDG) we were able to repeat 5. 60 and older. Blood 2008;112:Abstr. 2798. the irradiation several times thus prolonging 3. Visani G, Finelli C, Castelli U, et al. Myelo - 18. Popat UR, Rondon G, Alousi AM et al. the clinical benefit much more than expected. fibrosis with myeloid metaplasia: clinical Myelofibrosis and reduced intensity allo- Four patients safely underwent 12 RT cours- and haematological parameters predicting genic hematopoietic stem cell transplanta- es with no occurrence of grade 4 RTOG hema- survival in a series of 133 patients. Br J tion (RISCT). Blood 2007;110:Abstr. 3062. tologic toxicity. All the retreated patients Haematol 1990;75:4-9. 19. Mesa R. How I treat symptomatic spleno - belong to the low-or intermediate-risk group. 4. Guardiola P et al. Allogenic stem cell megaly in patients with myelofibrosis. In these patients the intensity and the persist- trasplantation for agnogenic myeloid Blood 2009. Epub ahead of print. ence of splenic response to irradiation did not meta plasia: a European Group for Blood 20. Verstovsek S, Kantarjian HM, Pardanani A change under multiple retreatment courses. and Marrow Trasplantation, Società et al. A phase I/II study of INCB018424, an However, retreatment increased the average Francaise de Greffe de Moelle, Gruppo oral, selective JAK inhibitor, in patients time of symptoms relief four fold longer than Italiano per il trapianto di Midollo Osseo e with primary myelofibrosis (PMF) and in un-retreated patients (21 vs. 5.75 months). Fred Hutchinson Cancer Research Center post polycythemia vera/essential thrombo- It could be argued that a possible bias in our Collaborative Study. Blood 1999;93:2831-8. cythemia myelofibrosis (Post PV/ET MF). J work is that all the post ET-MF patients were 5. Tefferi A. Myelofibrosis with myeloid Clin Oncol 2008;26 Abstr. 7004. allocated in the LDG or in the inferior burden meta plasia. N Engl J Med 2000;342:1255- 21. Mesa RA, Nagorney DS, Schwager S. et al. of other IDG but this fact could not modify the 65. Palliative goals, patient selection, and consistency of the presented data; especially 6. Reilly JT. Idiopathic myelofibrosis: patho- perioperative platelet management: out- because just one patient with post ET-MF has genesis, natural history and management. comes and lessons from 3 decades of been retreated so far. We propose that the Blood Rev 1997;11:233. splen ectomy for myelofibrosis with results regarding the average time of clinical 7. Thiele HM, Werden J, et al. Prognostic fac- myeloid metaplasia at the Mayo Clinic benefit in retreated patients can be considered tors in idiopathic (primary) osteomyelofi- Cancer 2006;107:361-70. substantially valid for primary MF patients. brosis. Cancer 1997;80:708. 22. Barosi G, Ambrosetti A, Centra A et al. Furthermore, it deserves to be mentioned that, 8. Thiele J, Windecker R, Kvasnicka HM et al. Splenectomy and risk of blast transforma- in comparison to MF patients, a shorter inter- Erythropoiesis in primary (idiopathic. tion in myelofibrosis with myeloid meta- val free from symptoms has been reported osteomyelofibrosis: quantification, PCNA- plasia. Blood 1998;91:3630-6. after SI32 in post-ET MF patients. reativity and prognostic impact. Am J 23. Bembassat J, Gilon D, Penchas S. The With all the limitations inherent in the Hematol 1994;46:36. choice between splenectomy and medical small number of patients examined, we found 9. Tefferi A., Mesa RA, Nagorney DM. Splen - treatment in patients with advanced agno- that in our series Dupriez’s score (calculated ectomy in myelofibrosis with myeloid genic myeloid metaplasia. Am J Hematol at the time of patient’s referral to the radio- metaplasia: a single institution experi- 1990;33:128-35. therapy department) is not predictive of ence with 223 patients. Blood 2000; 95: 24. Koch CA, Li CY, Mesa RA et al. Non hepato - response to palliative radiotherapy or occur- 2226-33. splenic extramedullary hemato-poiesis:

[page 38] [Hematology Reviews 2009; 1:e7] Article

associated disease, pathology, clinical Radiat Oncol Biol Phys 1977;2:1075-81. rate relationships for long term repopulat- course and treatment. Mayo Clin Proc 30. Wegner H, McKeough P, Desforges J et al. ing hemopoietic stem cells in a murine 2003;78:1223-33. Splenic irradiation in the treatment of bone marrow transplant model. Radiat Res 25. Steensma DP et al. Low dose single frac- patients with myelogenous leukaemia or 1993;136:118-25 tion whole lung radiotherapy for pul- myelofibrosis with myeloid metaplasia. 36. Joiner MC, Marples B, Lambin P et al. Low monary hypertension associated with Cancer 1986;58:1204-7. dose hypersensivity: current status and myelofibrosis with myeloid metaplasia. Br 31. Bouabdallah R Coso D, Gonzague- possible mechanism. Int J Radiat Oncol J Haematol 2002;118:813-6. Casabianca L, et al. Safety and efficacy of 26. Tefferi A et al. Radiation therapy for symp- splenic irradiation in the treatment of Biol Phys 2001;49:379-89. tomatic hepatomegaly in myelofibrosis patients with idiopathic myelofibrosis: a 37. Marples B, Wounters BG, Collis SJ et al. with myeloid metaplasia. Eur J Haematol report on 15 patients. Leuk Res 2000; 24: Low-dose hypersensivity: a consequence 2001;66:37-42. 491-5. of ineffective cell cycle arrest of radiation- 27. Elliott MA, Chen MG, Silverstein MN et al. 32. McFarland J, Kouzma C, Millard F et al. damaged G2-phase cells. Radiat Res 2004; Splenic irradiation for symptomatic Palliative irradiation of the spleen. Am J 161:247-55. splenomegaly associated with myelofibro- Clin Oncol 2003;26:178-83. 38. Harney J, Short S, Shah N et al. Low dose sis with myeloid metaplasia. Br J 33. Elliott MA, Tefferi A. Splenic irradiation in hyper-radiosensivity in metastatic tumors. Haematol 1998;103:505-11. myelofibrosis with myeloid metaplasia: a Int J Radiat Oncol Biol Phys 2004;59:1190- 28. Greenberger J, Chaffey J, Rosenthal D et review. Blood Rev 1999;13:163-70. 5. al. Irradiation for control of hypersplenism 34. Down JD, Boudewijn A, van Os R et al. 39. Yorke E, Jackson A, Rosenzweig KE et al. and painful splenomegaly in myeloid Variations in radiation and repair among metaplasia. Int J Radiat Oncol Biol Phys different hematopoietic stem cell subset Correlation of dosimetric factors and radi- 1977:2:1082-9. following fractionated irradiation. Blood ation pneumonitis for non-small cell lung 29. Parmentier C, Charbord P, Tibi M et al. 1995;86:122-7. cancer patients in a recently completed Splenic irradiation in myelofibrosis. 35. Van Os R, Thames HD, Konings AW et al. dose escalation study. Int J Radiat Oncol Clinical findings and ferrokinetics. Int J Radiation dose-fractionation and dose- Biol Phys 2005;63:672-82.

[Hematology Reviews 2009; 1:e7] [page 39] Hematology Reviews 2009; volume 1:e8

MicroRNAs: tiny players with a mature miRNA, which guides the RISC complex mainly (but not exclusively) to the 3’- Correspondence: Muller Fabbri, Department of big role in the pathogenesis of UTR (3’-untranslated region) of the target Molecular Virology, Immunology, and Medical leukemias and lymphomas mRNAs. The final outcome of the target Genetics and Comprehensive Cancer Center, mRNA is based upon the miRNA:mRNA Ohio State University, Columbus, OH, USA Francesca Fanini,1 Ivan Vannini,1 degree of base pair complementarity. A per- E-mail: [email protected] Muller Fabbri1,2 fect base pair match (occurring mainly in plants), leads to mRNA cleavage, whereas an F.F. and I.V. equally contributed to this work. 1Istituto Scientifico Romagnolo per lo Acknowledgments: Dr. Fabbri is recipient of a imperfect complementarity (predominant in Studio e la Cura dei Tumori, Meldola, 2009 Kimmel Scholar Award. nematodes and mammals) results in transla- Italy; tional silencing of the target, albeit also in 2Department of Molecular Virology, Received for publication: 29 April 2009. case of imperfect base pairing a reduction of Accepted for publication: 5 May 2009. Immunology, and Medical Genetics and the target mRNA has been described.3 A fur- Comprehensive Cancer Center, Ohio State ther level of complexity in miRNA gene regu- This work is licensed under a Creative Commons University, Columbus, OH, USA lation has been recently provided by showing Attribution 3.0 License (by-nc 3.0) that miRNAs can also directly up-regulate the ©Copyright F. Fanini et al., 2009 expression of a target gene by binding to its Licensee PAGEPress, Italy Abstract 3’-UTR.10 The development of high throughput meth- Hematology Reviews 2009; 1:e8 MicroRNAs (miRNAs) are small non-coding ods to detect miRNA expression in human doi:10.4081/hr.2009.e8 RNAs (ncRNAs) with important regulatory samples,11,12 has provided invaluable tools to functions. After an initial phase, aimed at investigate the role of these ncRNAs in sever- identifying whether a deregulation in miRNA al human pathologies. It is now generally expression occurred between hematologic accepted that miRNAs are aberrantly miRNAs in chronic leukemias malignancies and their normal counterparts, expressed in almost all human cancers (both currently an increasing number of studies are solid and hematologic malignancies), with Chronic lymphocytic leukemia focusing on the functional significance of respect to the normal counterpart.13-15 This Chronic lymphocytic leukemia (CLL) is the these aberrancies. The identification of review will focus on the role of miRNAs in the most common type of leukemia, characterized miRNA targeted genes has cast a new light on pathogenesis of human leukemias and lym- by the slow accumulation in blood, bone mar- the role of these tiny ncRNAs in human can- phomas, highlighting diagnostic, prognostic, row and lymphatic tissue of small, non-prolif- cerogenesis, providing a new rationale to the and therapeutic implications of this class of erating, mature B lymphocytes, which display observed diagnostic, prognostic and therapeu- ncRNAs. Table 1 summarizes the most fre- typical surface markers such as CD19 and tic implications of miRNA aberrant expression quently de-regulated miRNAs in hematologic CD20 in addition to CD5.18 in human hematologic malignancies. malignancies. More than 50% of CLLs are characterized by hemizygous and/or homozygous deletion of the genomic region 13q14.19,20 Calin et al. showed that a cluster of miRNAs, namely miR-15a/16- Introduction miRNAs and leukemias 1, is located at chromosome 13q14 and that these miRNAs, situated within a 30-kb region MicroRNAs (miRNAs) are a class of non- Leukemia is a malignant disorder of the of loss in CLL, are both deleted or down-regu- coding RNAs (ncRNAs) with regulatory func- blood or bone marrow characterized by an lated in approximately 68% of CLL cases.21 tions, involved in a variety of biological abnormal proliferation of blood cells. Based Among the targets of miR-15a/16-1, was identi- processes.1-6 Initially transcribed by RNA poly- on the stage of differentiation of the malig- fied BCL2,22 an anti-apoptotic protein which is merase II as long, capped and polyadenylated nant cells, leukemias can be divided into over-expressed in the majority of CLL B cells,23 precursors (pri-miRNAs),7,8 miRNAs undergo acute and chronic. Acute leukemias are char- and it is believed to mediate the anti-tumoral a complex processing mechanism. First, a acterized by the rapid increase of immature effect of these miRNAs. This finding is sup- double-stranded RNA-specific ribonuclease blood cells (blood cell progenitors or primi- ported by the fact that restoration of called Drosha, in conjunction with its bind- tive stem cells with multilineage potential) miR15a/16-1 induces apoptosis in MEG-01 cell ing partner DGCR8 (DiGeorge syndrome crit- and are the most common forms of leukemia line, derived from acute megakaryocytic ical region gene 8, or Pasha) processes pri- in children, while chronic leukemias arise leukemia, pointing out a role for miR15a/16-1 miRNAs into hairpin RNAs of 70-100 from more mature cells during differentia- as tumor-suppressor genes (TSGs) in CLL, and nucleotides (nt) known as pre-miRNAs.9 By tion and occur more often among adults. maybe in other malignancies in which this means of Exportin 5, pre-miRNA is translo- Additionally, the diseases are subdivided, cluster is down-regulated or lost.22,24 Croce’s cated from the nucleus to the cytoplasm, according to the origin of the neoplastic cells, group first identified a unique microRNA sig- where a ribonucleic complex, composed of a into lymphoid leukemias, concerning the nature associated with prognostic factors and ribonuclease III (Dicer), and TRBP (HIV-1 lymphoid pathway (B and T cells) and disease progression in CLL,25 and described a transactivating response RNA binding pro- myeloid leukemias, which involve the cells germ-line mutation in pre-miR-16 sequence, tein) cleaves it in a 18-24 nt duplex. Finally, that normally go on to form red blood cells, which causes low levels of microRNA expres- the duplex interacts with a large protein com- granulocytes, monocytes and platelets. It was sion both in vitro and in vivo. This mutation plex called RISC (RNA-induced silencing recently reported that the pattern of miRNA was associated with deletion of the normal complex), which includes proteins of the expression varies strongly during the differ- allele in leukemic cells of 2 CLL patients one of Argonaute family (Ago1-4 in humans). One entiation of hematopoietic stem/progenitor which with a family history of CLL and breast strand of the miRNA duplex remains stably cells, playing a critical role in events that cancer.25 Interestingly, Raveche et al. described associated with RISC and becomes the result in hematologic disorders.16,17 a point mutation in the 3’-DNA adjacent to

[page 40] [Hematology Reviews 2009; 1:e8] Type ofArticle paper

Table 1. Main aberrantly expressed miRNAs in hematologic malignancies. miRNA Location Function Targets Malignancy miR15a/16-1 cluster 13q14.3 TSG/OG BCL2 CLL, HL miR-29b 7q32.3 (miR-29b-1) TSG TCL1 CLL 1q32.2 (miR-29b-2) miR-181b 1q31.3 (miR-181b-1) TSG TCL1, TLR4, CARD8, CASP1, CLL, AML 9q33.3 (miR-181b-2) IL1B, SLC11A1, MSR1, CD64 miR-181a 1q31.3 (miR-181a-1) TSG BCL2, TLR4, CARD8, CASP1, CLL, AML 9q33.3 (miR-181a-2) IL1B, SLC11A1, MSR1, CD64 miR-155 21q21.3 OG AGTR1, FGF7, ZNF537, CLL, AML (FLT-IDT+), ZIC3, IKBKE HL, NHL (BL, ABC-DLBCL) miR-17-92 cluster 13q31.3 OG/TSG E2F1, PTEN, BIM CML, ALL, HL, NHL (B cell lymphomas) miR-203 14q32.33 TSG ABL1 CML miR-128a (miR-128-1) 2q21.3 OG/TSG N/A High in ALL vs. AML; low in HL EBV+ miR-128b (miR-128-2) 3p22.3 OG/TSG N/A High in ALL vs. AML; low in HL EBV+ let-7b 22q13.31 OG/TSG RAS High in AML vs. ALL miR-223 Xq12 OG N/A High in AML vs. ALL miR-204 9q21.11 TSG HOXA10, MEIS1 AML (NPM1 mut) miR-34b 11q23.1 TSG CREB AML miR-9 1q22 (miR-9-1) OG PRDM1/Blimp-1 HL 5q14.3 (miR-9-2) 15q26.1 (miR-9-3) let-7a 9q22.32 (let-7a-1) OG/TSG RAS, PRDM1/Blimp-1 HL 11q24.1 (let-7a-2) 22q13.31 (let-7a-3) miR-96 7q32.2 TSG N/A Low in HL EBV+ miR-21 17q23.1 OG PTEN, PDCD4 HL, NHL (ABC-DLBCL) miR-24 9q22.32 (miR-24-1) OG N/A HL 19p13.12 (miR-24.2) miR-150 19q13.33 TSG N/A HL miR-221 Xp11.3 OG N/A NHL (ABC-DLBCL) miR-143 5q33.1 TSG ERK5 NHL (B-cell lymphomas) miR-145 5q33.1 TSG ERK5 NHL (B-cell lymphomas)

OG: oncogene; TSG: tumor suppressor gene; BCL2: B-cell CLL/lymphoma 2; TCL1: T-cell leukemia/lymphoma 1A; TLR4: toll-like receptor 4; CARD8: caspase recruitment domain family, member 8; CASP1: caspase 1, apoptosis-related cysteine peptidase (interleukin 1, beta, convertase); IL1B: interleukin 1, beta; SLC11A1: solute carrier family 11 (proton-coupled divalent metal ion transporters), member 1; MSR1: macrophage scavenger receptor 1; CD64: Fc fragment of IgG, high affinity 1a, receptor; AGTR1: angiotensin II receptor, type 1; FGF7: fibroblast growth factor 7 (ker- atinocyte growth factor); ZNF537: teashirt zinc finger homeobox 3; ZIC3: Zic family member 3 (odd-paired homolog, Drosophila); IKBKE: inhibitor of kappa light polypeptide gene enhancer in B- cells, kinase epsilon; E2F1: E2F transcription factor 1; PTEN: phosphatase and tensin homolog; BIM: BCL2-like 11 (apoptosis facilitator); ABL1: c-abl oncogene 1, receptor tyrosine kinase; RAS: rat sarcoma virus; HOXA10: homeobox A10; MEIS1: myeloid ecotropic viral integration site 1 homolog; CREB: cAMP responsive element binding protein 1; PRDM1/Blimp1: PR domain containing 1, with ZNF domain; PDCD4: programmed cell death 4 (neoplastic transformation inhibitor); ERK5: mitogen-activated protein kinase 7; N/A: not available; CLL: chronic lymphocytic leukemia; ALL: acute lymphocytic leukemia; CML: chronic myeloid leukemia; AML: acute myeloid leukemia; HL: Hodgkin’s lymphoma; NHL: Non-Hodgkin’s lymphoma; BL: Burkitt’s lymphoma; DLBCL: diffuse large B-cell lymphoma; ABC: activated B-cell phenotype; EBV+: Epstein-Barr virus positive; FLT-IDT+: fms-related tyrosine kinase 3 gene internal tandem dupication; NPM1 mut: nuclephosmin-1 mutation.

miR-16-1 region, which lets down the expres- ly or indirectly affected by miR15a/16-1 clus- ranted to verify this hypothesis. sion of this miRNA in the CLL-prone New ter.27 Among the genes which are down-regu- Another 2 miRNAs with a TGS function in Zealand black mouse strain model, indicating lated by this cluster, there are both oncogenes CLL are miR-29b and miR-181b, which direct- that this miRNA functions as an oncosuppres- (OGs) and TSGs, suggesting that even though ly target TCL1 (T-cell leukemia/lymphoma sor in CLL.26 By analyzing the result of the overall effect of miR15a/16-1 overexpres- 1A), an OG which co-activates the protein miR15a/16-1 hyperexpression in MEG-01 cells sion is anti-tumoral in CLL, it cannot be kinase AKT (v-akt murine thymoma viral and in CLL patients both at the transcriptome excluded that other pathways may be impacted oncogene homolog 1) and takes part in the and at the proteome level, Calin et al. found and lead to an overall oncogenic outcome in regulation of many pathways involved in cell that about 14% of all human genome is direct- other diseases. Further investigations are war- survival, proliferation and death.28 As evi-

[Hematology Reviews 2009; 1:e8] [page 41] Article denced by transgenic mice models, TCL1 is a be intriguing to investigate whether the Acute myeloid leukemia marker of aggressive CLL and its high levels silencing effect of miR-203 on ABL1 can have Recently, several works have analyzed which are associated with high levels of ZAP-70 (70 any indirect or modulatory consequence on aberrancies in the miRNome (defined as the kD zeta chain-associated protein kinase) and the miR-17-92 cluster. full spectrum of miRNAs expressed in a partic- unmutated IgVH status which are both mark- ular cell type) occur in acute myeloid leukemia ers of poor prognosis in CLL.29,30 In a recent (AML) patients. Garzon et al. investigated 240 study, a correlation between low levels of miR- AML patient samples to determine whether 29c and poor prognosis was observed in CLL miRNAs in acute leukemias miRNAs are associated with cytogenetic patients. In a 110 patient cohort with a medi- abnormalities and clinical features in acute an follow-up of 72 months, the authors Acute lymphocytic leukemia myeloid leukemia. For this purpose, they eval- showed for the first time that a specific Acute lymphocytic leukemia (ALL) is one of uated miRNA expression of CD34+ cells and threshold of expression for miR-29c and miR- the most common malignancies observed in 122 untreated adult AML cases using a 223 can predict treatment-free survival (TFS) the pediatric age and it represents about the microarray platform. They identified some 31 and overall survival (OS). While miR-181b 80% of cases of acute leukemia in children. It miRNAs differentially expressed between targets TCL1, miR-181a directly targets BCL2, arises from the clonal proliferation of lym- CD34+ normal cells and the AML samples 22,32 together with miR-15a/16-1 cluster, sug- phoid progenitors in the bone marrow whose whose expression was also closely associated gesting a central role of the miR-181 family consequence is the occurrence of several dif- to selected cytogenetic and molecular abnor- members in the pathogenesis of CLL. A study ferent cytopenias in the peripheral blood asso- malities, such as t(11q23), isolated trisomy 8, aimed at investigating a possible correlation ciated with the appearance of peripheral blast and FLT3-ITD (fms like tyrosine kinase 3 in between the two groups of miRNAs is war- cells. In 2007, Zanette et al. analyzed miRNA tandem duplication) mutations. Furthermore, ranted. expression profiles in 7 ALL samples and 6 they found that patients with high expression Finally, miR-155 was showed to be up-regu- normal CD19+ samples in order to compare of miR-191 and miR-199a were associated with + 33 lated in CLL versus normal CD19 B cells, and them. They first reported the involvement of worse overall and event-free survival than AML the role of this miRNA will be analyzed more miR-128b, miR-204, miR-218, miR-331 and patients with low expression, suggesting a 40 extensively in the paragraph of lymphomas. miR-181b-1 in hematologic malignancies, possible prognostic role for these 2 miRNAs. with miR-128b as the most represented In normal karyotype, AML nucleophosmin-1 (NPM1) and FLT3-ITD mutations are frequent- Chronic myeloid leukemia miRNA in ALL. In their analysis also the miR- ly found. In NPM1 mutated cases up-regulation Using microarray analysis (miCHIP) and 17-92 cluster resulted up-regulated in all 7 of miR-10a, -10b, several let-7 and miR-29 fam- miRNA-specific quantitative real-time samples.36 The nucleotide sequences and ily members, as well as down-regulation of reverse transcriptase-polymerase chain reac- organization of this cluster is highly con- miR-204 were observed. Intriguingly, miR-204 tion (miR-qRT-PCR), Venturini et al. reported served in vertebrates,37 and its expression is targets the HOX genes HOXA10 and MEIS1, BCR-ABL1- and c-MYC-dependent transacti- tightly regulated during B-cell development. A vation of miR-17-92 cluster, a polycistronic suggesting that the HOX up-regulation role of miR-17-92 cluster to promote survival gene located in human chromosome 13 ORF observed in NPMc+ AML (cytoplasmic nucle- of early B-cell progenitors in a cell- 25 (C13orf25) at 13q31-q32 and composed of ophosmin) may be due at least in part to loss autonomous manner, and control cell survival 7 mature miRNAs namely miR-17–5p, miR- of HOX regulator-miRNAs.41 FLT3-ITD+ samples during the pro-B to pre-B transition was 17–3p, miR-18a, miR-19a, miR-20a, miR-19b showed up-regulation of miR-155 which was recently suggested by Ventura et al. They pro- and miR-92–1, in chronic myeloid leukemia demonstrated to be strongly but independently posed that in human B-cell lymphomas miR- cell lines.34 BCR-ABL1-MYC pathway can associated with FLT3-ITD mutations.41 In a ret- 17-92 overexpression is induced at the pro-B induce this transactivation in early chronic rospective study conducted on samples of to pre-B transition where it acts to silence the phase but not in blast crisis CML CD34+ cells, leukemia cells from patients who had cytoge- suggesting a role for miR-17-92 cluster in the expression of the pro-apoptotic protein Bim, netically normal AML and high-risk molecular pathogenesis of chronic myeloid leukemia leading to an abnormal survival of pro-B lym- features such as FLT3–ITD+, a wild-type NPM1 38 (CML).34 phocytes. In another study, Mi et al. tried to or both, Marcucci et al. found that expression Recently, Bueno et al. identified a fragile understand the distinct mechanisms in leuke- levels of microRNA-181 family were inversely chromosomal region lost in specific mogenesis between ALL and AML and to iden- correlated with those of predicted target genes hematopoietic malignancies, which encodes tify markers for diagnosis and treatment, per- encoding proteins which take part in pathways about 12% of all genomic miRNAs, including forming a large-scale genomewide microRNA controlled by toll-like receptors and inter- miR-203 which directly targets ABL1.35 In a T- expression profiling assay. They identified 27 leukin-1β42 (Table 1). In another study, Fazi et cell lymphoma model, both genetic and epige- miRNAs that are differentially expressed al. analyzed patient’s primary leukemia blasts, netic mechanisms inactivated this miRNA. between ALL and AML. Among them, miR-128a and demonstrated that those carrying the Specifically the miR-203 promoter resulted and miR-128b which are up-regulated in ALL t(8;21), which generates the most common hypermethylated in chromosome Philadelphia vs. AML, and let-7b, miR-223 which are down- acute myeloid leukemia-associated fusion pro- positive (Ph+) tumors, including B-cell ALLs, regulated in ALL vs. AML, can lead to a differ- tein AML1/ETO, exhibited low levels of miRNA- primary CMLs and cultured CML cell lines ential diagnosis between the acute leukemias, 223, a regulator of myelopoiesis. They showed while no methylation was observed in with an accuracy of 98%. Furthermore, they that miR-223 is a direct transcriptional target hematopoietic tumors that do not carry ABL1 found that overexpression of miR-128 in ALL of AML1/ETO which causes heterochromatic alterations. Re-expression of miR-203 was at least partly associated with its promot- silencing of this miRNA by recruiting chro- reduces ABL1 and BCR-ABL1 fusion protein er hypomethylation and not with an amplifica- matin remodeling enzymes at an AML1-bind- levels and inhibits tumor cell proliferation in tion of its genomic locus, suggesting an ing site on the pre-miR-223 gene. Ectopic miR- an ABL1-dependent manner, putting forward important role of epigenetics in the regulation 223 expression, RNA interference against the role of miR-203 as a TSG whose re-expres- of the expression of miRNAs in acute AML1/ETO, or demethylation enhance miR-223 sion might have therapeutic benefits in spe- leukemias.39 The leukemogenic mechanism of levels and restore cell differentiation. These cific hematopoietic malignancies.35 It would miR-128b remains still poorly understood. findings confirm a central role of miR-223 in

[page 42] [Hematology Reviews 2009; 1:e8] Type ofArticle paper myeloproliferative disorders and reflect the ter members, miR-16, miR-21, miR-24, and nate DLBCL, follicular lymphomas and reactive essential function of this miRNA in normal miR-155. A significant down-regulation in HL lymph nodes with an accuracy of 98%.58 It is myeloid ontogeny.43 Recently, a possible mech- was observed for miR-150 whereas an impor- compelling that among the de-regulated anism for CREB (cyclic AMP-responsive ele- tant up-regulation in HL was shown for miR- miRNAs of Roehle’s work, there are miR-17-5p ment binding protein) overexpression in 155.47 Moreover, the Authors identified AGTR1, and miR-106a, members of two paralogous leukemia mediated by mir-34b has been pro- FGF7, ZNF537, ZIC3, and IKBKE as true miR- clusters with a documented role as oncogenes vided. Using real-time quantitative PCR, 155 target genes in HL.47 Interestingly, high in several human malignancies.37 Pigazzi et al. discovered that miR-34b was sig- levels of miR-155 have also been observed in Located at 13q31-32, miR-17-92 cluster is nificantly less expressed in myeloid cell lines the germinal center during normal lym- frequently amplified in malignant B-cell lym- which are known to have high CREB protein phopoiesis.48 Since HL has its origin in the phomas, and is over-expressed in 65% of B-cell level. When exogenous miR-34b expression lymph-nodal germinal center, the hypothesis lymphoma patients.59 Overexpression of miR- was induced in vitro the CREB levels that overexpression of miR-155 is a result of 17-92 in murine multipotent progenitor cells decreased as a consequence of a direct interac- an aberrant lymphocytic block of differentia- (MPPs) of MYC-transgenic mice caused an tion between miR-34b and the CREB 3’-UTR. tion in the germinal center is compelling. increased lymphomagenesis.59 Xiao et al. gen- Moreover, mir-34b restored expression caused erated mice with high miR-17-92 lymphocytic alteration in CREB target gene expression and Non-Hodgkin lymphomas expression and described a higher rate of lym- cell cycle abnormalities suggesting a role as Up-regulation of miR-155 has also been phoproliferative disorders, autoimmunity and potential TSG. In leukemia cell lines, they described in Non Hodgkin lymphomas (NHLs), premature death.60 These pathological charac- showed that miR-34b/miR-34c promoter was as well as in several other solid and hematolog- teristics were induced by down-regulation of methylated and this epigenetic regulation ic malignancies. The first hint of miR-155 the oncosuppressor genes PTEN and BIM.60 should control the observed mir-34b expres- involvement in lymphomagenesis derived from O’Donnell et al. showed that c-MYC activates 2 sion levels to preserve Creb protein overex- the observation that the final part of the B-cell members of the cluster (namely miR-17-5p and pression. The inverse relationship between integration cluster (BIC) non-coding RNA miR-20a), which directly target E2F1, a c-MYC miR-34b and CREB was also supported in vivo (ncRNA), where miR-155 is positioned, transactivated transcription factor promoting in a cohort of 78 pediatric patients.44 increases MYC-mediated lymphomagenesis in cell-cycle progression.61 In summary, by direct- a chicken experimental model.49 Kluiver et al. ly transactivating E2F1 and indirectly (through showed that the expression of both miR-155 miR-17-92 cluster) repressing its translation, and its host gene is induced by protein kinase c-MYC functions as a key-regulator of cell-cycle miRNAs and lymphomas C and NF-kB in several cell lines, except progression. Recently, Li et al. investigated Ramos, a Burkitt’s lymphoma cell line, in genome-wide miRNA expression and copy MiRNAs are involved in the pathogenesis of which, for yet unclear reasons, the overexpres- number in 86 DLBCLs (59 primary tumors and both Hodgkin and non-Hodgkin lymphomas. sion of BIC is not accompanied by up-regula- 27 cell lines), and identified a collection of There are many studies that describe the role tion of miR-155.50 Transgenic and knockout miRNAs that robustly segregated DLBCLs into of miRNAs in these two groups of lymphomas. (KO) miR-155 mice models have significantly three subgroups not related to the cell-of-ori- contributed to clarify the function of this gin classification, extent of T-cell infiltrate and Hodgkin lymphoma ncRNA. In Costinean’s transgenic mouse tumor site.62 In particular, the three newly Nie et al. showed that miR-9 and let-7a tar- model, miR-155 overexpression caused a poly- identified subsets of DLBCLs had different get PRDM1/Blimp-1 in Hodgkin lymphomas clonal pre-leukemic pre-B-cell proliferation fol- prognosis and showed a markedly different cell lines. In fact, the levels of miR-9 and let-7a lowed by full blown B-cell malignancy, provid- MYC transcriptional activity, which was corre- inversely correlated with PRDM1/Blimp-1 ing in vivo evidence of this miRNA involve- lated with the dominance of MYC-regulated expression in Hodgkin Lymphoma (HL) cells. ment in the pathogenesis of hematologic miRNAs in their expression signatures.62 Similar to their in vitro counterparts, the malignancies.51 Conversely, KO mice models Finally, also miR-143 and miR-145 expression majority of cells in primary HL cases showed showed impairment of cytokine production is reduced in B-cell malignancies.63 In Burkitt’s weak or no PRDM1/Blimp-1 expression.45 toward TH2 lymphocyte differentiation, and lymphoma Raji cell line, restoration of these Navarro et al., analyzed the expression of compromised function of dendritic cells,52,53 two miRNAs caused a dose-dependent growth miRNAs in 49 HL patients and 10 reactive confirming the data of Tili et al., who showed a inhibitory effect linked to a down-regulation of lymph nodes. They identified 25 miRNAs dis- central role of miR-155 in regulating the Erk5, a MAP kinase directly targeted by miR- criminating HL from reactive lymph nodes and immune response.54 143 and miR145.63 36 miRNAs differentially expressed the nodu- In a subgroup of NHLs, the diffuse large B- lar sclerosis and mixed cellularity subtype.46 cell lymphomas (DLBCL), miR155 up-regula- The obtained results, validated in a set of dif- tion occurs in the activated B-cell phenotype ferent cell lines, showed that miR-96, miR- (ABC-DLBCL) with respect to the germinal Conclusions 128a, and miR-128b were selectively down-reg- center B-cell-like phenotype (GCB-DLBCL).55,56 ulated in HL with EBV. Moreover, only one of Since ABC-DLBCL and GCB-DLBCL have 5- There is increasing evidence in literature to the miRNAs differentially expressed in EBV+ year survival rates of 30% and 59%, respective- support a central role for miRNAs in the patho- cases was enclosed in the 25 miRNA that dis- ly,57 miR-155 expression harbors prognostic genesis of hematologic malignancies. tinguish HL from reactive lymph nodes, lead- implications. In ABC-DLBCL also high levels of Aberrancies in miRNA expression lead to ing to the intriguing conclusion that EBV miR-21 and miR-221 were described, suggest- abnormal gene regulation and, ultimately, to a might not be a relevant event in HL pathogen- ing a 3 miRNA-signature specific of the histo- severe alteration in proteic effectors with esis.46 Gibcus el al. described a specific miRNA type of DLBCL with severe prognosis.56 Roehle oncogenetic and/or oncosuppressor function. expression profile in HL cell lines. By compar- et al. compared the miRNome of DLBCL, follic- From the initial contribution of high through- ing HL with a panel of B-cell non-Hodgkin lym- ular lymphomas and reactive lymph nodes. In put techniques, which allowed the identifica- phomas, they identified a signature of HL-spe- particular, they found that 4 miRNAs (namely tion of the de-regulated miRNAs, currently cific miRNAs, which included miR-17-92 clus- miR-330, -17-5p, -106a and -210) can discrimi- many efforts are being made to understand the

[Hematology Reviews 2009; 1:e8] [page 43] Article functional implications of miRNA aberrant 12. Lu J, Getz G, Miska EA, et al. MicroRNA 2008;105:5166-71. expression. For several hematologic malignan- expression profiles classify human can- 28. Pekarsky Y, Santanam U, Cimmino A, et al. cies, a clear diagnostic and prognostic role of cers. Nature 2005;435:834-8. Tcl1 expression in CLL is regulated by miRNA de-regulation has been demonstrated. 13. Fabbri M, Croce CM, Calin GA. MicroRNAs. miR-29 and miR-181. Cancer Res 2006;66: The identification of miRNA targeted mRNAs Cancer J 2008;14:1-6. 11590-3. is clarifying which oncogenic pathways are 14. Fabbri M, Garzon R, Andreeff M, et al. 29. Yan XJ, Albesiano E, Zanesi N, et al. B cell affected by the leukemic/lymphomatous MicroRNAs and noncoding RNAs in hema- receptors in TCL1 transgenic mice resem- miRNome. This knowledge represents the tological malignancies: molecular, clinical ble those of aggressive, treatment-resist- essential background to the development of and therapeutic implications. Leukemia ant human chronic lymphocytic leukemia. new miRNA-based therapeutic approaches, 2008;22:1095-105. Proc Natl Acad Sci USA 2006;103:11713-8. which combine the advantage of a physiologic 15. Garzon R, Croce CM. MicroRNAs in normal 30. Herling M, Patel KA, Khalili J, et al. TCL1 molecule (such as an miRNA-derived drug), to and malignant hematopoiesis. Curr Opin shows a regulated expression pattern in the possibility of targeting with a single mole- Hematol 2008;15:352-8. chronic lymphocytic leukemia that corre- cule several members of the same and/or of 16. Chen CZ, Li L, Lodish HF, Bartel DP; lates with molecular subtypes and prolifer- several aberrantly activated pathways in the MicroRNAs modulate hematopoietic line- ative state. Leukemia 2006;20:280-5. malignancy. More studies are warranted to age differentiation. Science 2004;303:83-6. 31. Stamatopoulos B, Meuleman N, Haibe- define how big is the impact of these tiny RNA 17. Chen CZ, Lodish HF. MicroRNAs as regula- Kains B, et al. MicroRNA-29c and micro molecules in human cancerogenesis. However, tors of mammalian hematopoiesis. Semin RNA-223 downregulation has in vivo sig- there is no doubt that these ncRNAs play a role Immunol 2005;17:155-5. nificance in chronic lymphocytic leuke mia in hematologic malignancies. Decoding this 18. Cheson BD, Bennett JM, Grever M, et al. and improves disease risk stratification. role has diagnostic, prognostic and therapeutic National Cancer Institute-sponsored Blood 2009 Jan 14 [Epub ahead of print]. implications of the utmost importance. Working Group guidelines for chronic lym- 32. Neilson JR, Zheng GX, Burge CB, Sharp phocytic leukemia: revised guidelines for PA. Dynamic regulation of miRNA expres- diagnosis and treatment. Blood 1996;15: sion in ordered stages of cellular develop- 4990-7. ment. Genes Dev 2007;21:578-89. References 19. Bullrich F, Fujii H, Calin G, et al. Character- 33. Marton S, Garcia MR, Robello C, et al. ization of the 13q14 tumor suppressor locus Small RNAs analysis in CLL reveals a 1. Ambros V, Lee RC. Identification of in CLL: identification of ALT1, an alterna- deregulation of miRNA expression and microRNAs and other tiny noncoding tive splice variant of the LEU2 gene. Cancer novel miRNA candidates of putative rele- RNAs by cDNA cloning. Methods Mol Biol Res 2001;61:6640-8. vance in CLL pathogenesis. Leukemia 2004;265:131-58. 20. Bullinger L, Dohner K, Bair E, et al. Use of 2008;22:330-8. 2. Bartel DP. MicroRNAs: genomics, biogene- gene-expression profiling to identify prog- 34. Venturini L, Battmer K, Castoldi M, et al. sis, mechanism, and function. Cell nostic subclasses in adult acute myeloid Expression of the miR-17-92 polycistron in 2004;116:281-97. leukemia. N Engl J Med 2004;350:1605-16. chronic myeloid leukemia (CML) CD34+ 3. He L, Hannon GJ. MicroRNAs: small RNAs 21. Calin GA, Dumitru CD, Shimizu M, et al. cells. Blood 2007;109:4399-405. with a big role in gene regulation. Nat Rev Frequent deletions and down-regulation of 35. Bueno MJ, Perez de Castro I, Gomez de Genet 2004;5:522-31. micro- RNA genes miR15 and miR16 at Cedron M, et al. Genetic and epigenetic 4. Plasterk RH. Micro RNAs in animal devel- 13q14 in chronic lymphocytic leukemia. silencing of microRNA-203 enhances ABL1 opment. Cell 2006;124:877-8. Proc Natl Acad Sci USA 2002;99:15524-9. and BCR-ABL1 oncogene expression. 5. Pasquinelli AE, Hunter S, Bracht J. 22. Cimmino A, Calin GA, Fabbri M, et al. miR- Cancer Cell 2008;13:496-506. MicroRNAs: a developing story. Curr Opin 15 and miR-16 induce apoptosis by target- 36. Zanette DL, Rivadavia F, Molfetta GA, et al. Genet Develop 2005;15:200-5. ing BCL2. Proc Natl Acad Sci USA 2005; miRNA expression profiles in chronic lym- 6. Carleton M, Cleary MA, Linsley PS. 102:13944-9. phocytic and acute lymphocytic leukemia. MicroRNAs and cell cycle regulation. Cell 23. Kitada S, Andersen J, Akar S, et al. Braz J Med Biol Res 2007;40:1435-40. Cycle 2007;6:2127-32. Expression of apoptosis-regulating pro- 37. Mendell JT. miRiad roles for the miR-17-92 7. Lee Y, Kim M, Han J, et al. MicroRNA teins in chronic lymphocytic leukemia: cluster in development and disease. Cell genes are transcribed by RNA polymerase correlations with In vitro and In vivo 2008;133:217-22. II. Embo J 2004;23:4051-60. chemoresponses. Blood 1998;91:3379-89. 38. Ventura A, Young AG, Winslow MM, et al. 8. Cai X, Hagedorn CH, Cullen BR. Human 24. Bonci D, Coppola V, Musumeci M, et al. Targeted deletion reveals essential and microRNAs are processed from capped, The miR-15a-miR-16-1 cluster controls overlapping functions of the miR-17 polyadenylated transcripts that can also prostate cancer by targeting multiple onco- through 92 family of miRNA clusters. Cell function as mRNAs. Rna 2004;10:1957-66. genic activities. Nat Med 2008;14:1271-77. 2008;132:875-86. 9. Cullen BR. Transcription and processing of 25. Calin GA, Ferracin M, Cimmino A, et al. A 39. Mi S, Lu J, Sun M, et al. MicroRNA expres- human microRNA precursors. Mol Cell MicroRNA signature associated with prog- sion signatures accurately discriminate 2004;16:861-5. nosis and progression in chronic lymphocyt- acute lymphoblastic leukemia from acute 10. Vasudevan S, Tong Y, Steitz JA. Switching ic leukemia. N Engl J Med 2005;353:1793- myeloid leukemia. Proc Natl Acad Sci USA from repression to activation: microRNAs 801. 2007;104:19971-6. can up-regulate translation. Science 2007; 26. Raveche ES, Salerno E, Scaglione BJ, et al. 40. Garzon R, Volinia S, Liu CG, et al. 318:1931-4. Abnormal microRNA-16 locus with synteny MicroRNA signatures associated with 11. Liu CG, Calin GA, Meloon B, et al. An oligo- to human 13q14 linked to CLL in NZB cytogenetics and prognosis in acute nucleotide microchip for genome-wide mice. Blood 2007;109:5079-86. myeloid leukemia. Blood 2008;111:3183-9. microRNA profiling in human and mouse 27. Calin GA, Cimmino A, Fabbri M, et al. MiR- 41. Garzon R, Garofalo M, Martelli MP, et al. tissues. Proc Natl Acad Sci USA 2004; - 15a and miR-16-1 cluster functions in Distinctive microRNA signature of acute 101:9740-4. human leukemia. Proc Natl Acad Sci USA myeloid leukemia bearing cytoplasmic

[page 44] [Hematology Reviews 2009; 1:e8] Type ofArticle paper

mutated nucleophosmin. Proc Natl Acad coding RNA, cooperates with c-myc in lym- lymphoma. Int J Cancer 2007;121:1156-61. Sci USA 2008;105:3945-50. phomagenesis and erythroleukemogene- 57. Rosenwald A, Wright G, Leroy K, et al. 42. Marcucci G, Radmacher MD, Maharry K, et sis. J Virol 2002;76:4275-86. Molecular diagnosis of primary mediasti- al. MicroRNA expression in cytogenetical- 50. Kluiver J, Haralambieva E, de Jong D, et al. nal B cell lymphoma identifies a clinically ly normal acute myeloid leukemia. N Engl Lack of BIC and microRNA miR-155 favorable subgroup of diffuse large B cell J Med 2008;358:1919-28. expression in primary cases of Burkitt lymphoma related to Hodgkin lymphoma. J 43. Fazi F, Racanicchi S, Zardo G, et al. Epi - lymphoma. Genes Chromosomes Cancer Exp Med 2003;198:851-62. genetic silencing of the myelopoiesis reg- 2006;45:147-53. 58. Roehle A, Hoefig KP, Repsilber D, et al. ulator microRNA-223 by the AML1/ETO 51. Costinean S, Zanesi N, Pekarsky Y, et al. MicroRNA signatures characterize diffuse oncoprotein. Cancer Cell 2007;12:457-66. Pre-B cell proliferation and lymphoblastic large B-cell lymphomas and follicular lym- 44. Pigazzi M, Manara E, Baron E, Basso G. leukemia/high-grade lymphoma in E(mu)- miR-34b targets cyclic AMP-responsive miR155 transgenic mice. Proc Natl Acad phomas. Br J Hematol 2008;142:732-44. element binding protein in acute myeloid Sci USA 2006;103:7024-9. 59. Ota A, Tagawa H, Karnan S, et al. leukemia. Cancer Res 2009;69:2471-8. 52. Thai TH, Calado DP, Casola S, et al. Identification and characterization of a 45. Nie K, Gomez M, Landgraf P, et al. Regulation of the germinal center novel gene, C13orf25, as a target for MicroRNA-mediated down-regulation of response by microRNA-155. Science 2007; 13q31-q32 amplification in malignant lym- PRDM1/Blimp-1 in Hodgkin/Reed- 316:604-8. phoma. Cancer Res 2004;64:3087-95. Sternberg cells: a potential pathogenetic 53. Rodriguez A, Vigorito E, Clare S, et al. 60. Xiao C, Srinivasan L, Calado DP, et al. lesion in Hodgkin lymphomas. Am J Pathol Requirement of bic/microRNA-155 for nor- Lymphoproliferative disease and autoim- 2008;173:242-52. mal immune function. Science 2007;316: munity in mice with increased miR-17-92 46. Navarro A, Gaya A, Martinez A, et al. 608-11. expression in lymphocytes. Nat Immunol MicroRNA expression profiling in classic 54. Tili E, Michaille JJ, Cimino A, et al. 2008;9:405-14. Hodgkin lymphoma. Blood 2008;111:2825- Modulation of miR-155 and miR-125b lev- 61. O'Donnell KA, Wentzel EA, Zeller KI, et al. 32. els following lipopolysaccharide/TNF-alpha c-Myc-regulated microRNAs modulate 47. Gibcus JH, Tan LP, Harms G, et al. Hodgkin stimulation and their possible roles in reg- E2F1 expression. Nature 2005;435:839-43. lymphoma cell lines are characterized by a ulating the response to endotoxin shock. J 62. Li C, Kim SW, Rai D, et al. Copy number specific miRNA expression profile. Neo - Immunol 2007;179:5082-9. abnormalities, MYC activity and the genet- plasia 2009;11:167-76. 55. Eis PS, Tam W, Sun L, et al. Accumulation 48. Kuppers R. Mechanisms of B-cell lym- of miR-155 and BIC RNA in human B cell ic fingerprint of normal B-cells mechanisti- phoma pathogenesis. Nat Rev Cancer lymphomas. Proc Natl Acad Sci USA 2005; cally define the microRNA profile of DLBCL. 2005;5:251-62. 102:3627-32. Blood 2009 Mar 10 [Epub ahead of print]. 49. Tam W, Hughes SH, Hayward WS, Besmer 56. Lawrie CH, Soneji S, Marafioti T, et al. 63. Akao Y, Nakagawa Y, Kitade Y, et al. P. Avian bic, a gene isolated from a com- MicroRNA expression distinguishes be - Downregulation of microRNAs-143 and -145 mon retroviral site in avian leukosis virus- tween germinal center B cell-like and activat- in B-cell malignancies. Cancer Sci 2007; induced lymphomas that encodes a non- ed B cell-like subtypes of diffuse large B cell 98:1914-20.

[Hematology Reviews 2009; 1:e8] [page 45] Hematology Reviews 2009; volume 1:e9

Histone deacetylase inhibitors (MEK)/extracellular signal regulated kinase (ERK)-pathway, the phosphatidylinositol-3 Correspondence: Prof. Dr. K. Vanderkerken, in multiple myeloma kinase (PI3K)/ protein kinase B (Akt)-pathway, Department of Hematology and Immunology, the nuclear factor κ B (NFκB)-pathway and the Vrije Universiteit Brussel (VUB), Laarbeeklaan 1 1 Sarah Deleu, Eline Menu, Wnt-pathway.5,6 Moreover, cells interacting with 103, B-1090 Brussels, Belgium. Els Van Valckenborgh,1 Ben Van Camp,1 the MM cells in the BM, such as stromal cells, E-mail: [email protected] Joanna Fraczek,2 Isabelle Vande Broek,1 endothelial cells, osteoblasts, osteoclasts and Vera Rogiers,2 Karin Vanderkerken1 mesenchymal stem cells are also potential tar- Received for publication: 9 February 2009. Revision received: 20 April 2009. 1 gets to overcome the drug resistance against Myeloma Research Center-Brussels, Accepted for publication: 27 April 2009. 7,8 Department Hematology and conventional chemotherapy. Acknowledgment: we acknowledge grants of VUB Immunology, Vrije Universiteit Brussel MM represents 1% of all cancers and it is the second most commonly diagnosed hematologic (OZR-GOA), FWO-Vlaanderen and Stichting (VUB), Brussels malignancy. The incidence is higher with tegen Kanker. 2Department of Toxicology, Vrije increasing age and is 4-5 per 100,000 individu- Universiteit Brussel - VUB, Brussels, This work is licensed under a Creative Commons als each year worldwide. The median age at Attribution 3.0 License (by-nc 3.0) Belgium diagnosis is 67 years.9 The most common clinical characteristics in MM are bone pain, ane- ©Copyright S. Deleu et al., 2009 mia, recurrent infections and renal failure.10 Licensee PAGEPress, Italy The standard induction therapy for elderly Abstract patients with symptomatic myeloma, and who Hematology Reviews 2009; 1:e9 are not candidates for stem cell transplantion, doi:10.4081/hr.2009.e9 Novel drugs such as bortezomib and high- used to be melphalan (M) and prednisone (P). dose chemotherapy combined with stem cell Recently, improved effects on survival have transplantation improved the outcome of multi- been seen in patients receiving MP combined Modifications of these nucleosomes on the ple myeloma patients in the past decade. with lenalidomide (Revlimid®) (MPR), borte- histone level, as well as the DNA level, can However, multiple myeloma often remains zomib (Velcade®) (MPV) or thalidomide alter the chromatin state which can be open or incurable due to the development of drug resist- (MPT).11 Only the latter has been accepted as closed. The post-translational modifications on ance governed by the bone marrow micro - standard therapy. High-dose therapy plus the core histones are most common on the environment. Therefore targeting new path- autologous stem cell transplantation is consid- amino-terminal lysine rich tail which passes ways to overcome this resistance is needed. ered the standard therapy for front-line treat- through and around the enveloping DNA dou- Histone deacetylase (HDAC) inhibitors repre- ment of MM patients aged <65 years.12,13 The ble helix.18 These modifications, such as acety- sent a new class of anti-myeloma agents. most common pre-transplantation induction lation, methylation, ubiquitinylation, sumoyla- Inhibiting HDACs results in histone hyperacety- therapies used today are thalidomide-dexam- tion, phosphorylation and glycosylation are lation and alterations in chromatine structure, ethasone, bortezomib-based regimes, and crucial in modulating gene expression, as they which, in turn, cause growth arrest differentia- lenalidomide-dexamethasone.14,15 New agents affect the accessibility and interaction of DNA tion and/or apoptosis in several tumor cells. such as bortezomib, thalidomide and lenalido- with other non-histone protein complexes that Here we summarize the molecular actions of mide in the treatment of MM do not only tar- could contain transcriptionally co-activating or HDACi as a single agent or in combination with get the MM cells directly, but also influence co-repressing elements.19,20 More o ver, methyla- other drugs in different in vitro and in vivo the interactions of the MM cells with the BM tion of DNA, maintained by the epigenetic myeloma models and in (pre-)clinical trials. microenvironment. Combining these new enzymes, methyltransferases and demethy- agents with conventional chemotherapy and lases, also affects the chromatin structure high-dose chemotherapy with autologous indirectly by recruiting protein complexes con- stem cell transplantation increases the out- taining enzymes such as histone deacetylases Introduction come of MM patients, although eventually all (HDAC).21 HDAC and the opposite enzyme his- MM patients relapse. Therefore, identification tone acetyltransferases (HAT) are the most Multiple myeloma (MM) is a plasma cell of new key molecules in MM cells and in the analyzed enzymes involved in the post-transla- malignancy, characterized by an accumulation BM microenvironment is crucial for the devel- tional modifications of histones. Both enzymes of monoclonal plasma cells in the bone marrow opment of new therapeutic strategies. maintain the acetylation status of histones and (BM) and high levels of monoclonal immu - There is growing evidence that not only non-histone proteins. HAT acetylates histones noglobulines or paraprotein in blood and/or gene defects such as deletions, mutations and resulting in neutralizing the positive charge of urine. Complex interactions between MM cells chromosomal abnormalities are responsible histones and a more relaxed, transcriptionally and the BM microenvironment are required for for the onset and progression of cancer. active chromatin, while HDAC remove the the growth and progression of MM and result in Several studies have shown that epigenetic acetyl group resulting in a more compact, tran- the development of drug resistance, angiogene- changes, i.e. heritable traits mediated by scriptionally inactive chromatin structure.22 sis and induction of bone disease.1-4 Enhanced changes in DNA other than nucleotide Inhibiting HDAC leads to hyperacetylation of understanding of the interactions between MM sequences, play a key role in the downregula- histones and results in gene expression alter- and the BM microenvironment has led to the tion of tumor suppressor genes and/or upregu- ation. In tumor cells, several HDAC inhibitors identification of new molecular targets. Novel lation of oncogenes and, therefore, are also (HDACi) have shown promising anti-cancer therapeutic approaches target growth factors involved in the onset and progression of sever- activities with anti-proliferative, pro-apoptotic [e.g. insulin-like growth factor-1 (IGF-1), inter- al malignancies.16,17 Chromatin remodeling is and anti-angiogeneic properties.23-28 leukin-6 (IL-6) and vascular endothelial growth one of the main processes in epigenetic regu- This review provides an overview of the factor (VEGF)], adhesion molecules and signal- lations. Nucleosomes are the repeating units anti-myeloma activity of different HDACi in ing cascades in the MM cells such as the mito- of chromatin which contain 146 bp DNA pre-clinical settings and the latest clinical tri- gen-activated protein kinase kinase wrapped around a core histone octamer. als with HDACi ongoing in MM patients.

[page 46] [Hematology Reviews 2009; 1:e9] Type ofArticle paper

anti-epileptic drug valproic acid (VPA) and aliphatic chain for a spacer, and a functional The histone deacetylase family sodium butyrate are only effective at mM con- group (except depsipeptide). The cap group centrations and thereby form the less potent may be necessary for packing the inhibitor at Eighteen HDACs have been identified in class of HDACi.41 Clinical evaluations have the rim of the tube-like active site pocket, humans and are subdivided into four classes been performed with these compounds either while the aliphatic group forms interactions based on their homology to yeast HDACs and alone or in combination and are well tolerated with the residues of the lining pocket. For TSA, their enzymatic activities.29,30 Class I HDACs (1, in patients. However due to the short plasma the hydroxamic acid group (the functional 2, 3 and 8) are homologs to the yeast Rpd3 and half-life, high doses are needed to obtain a group) coordinates the zinc through its car- can generally be detected in the nucleus. They therapeutic effect.40 The first natural hydroxa- bonyl and hydroxyl groups, resulting in the for- are ubiquitously expressed in several human mate was TSA and is now considered as the mation of a penta-coordinated zinc and there- cell lines and tissues. Based on phylogenetic reference compound of hydroxamate based by altering the activity of the enzyme.53,54 analysis, Gregoretti et al. divided class I into Ia inhibitors. Most of the synthetic hydroxamate (HDAC1 and 2), Ib (HDAC3) and Ic (HDAC8).31 based HDACi target class I and class II with Class II HDACs (4, 5, 6, 7, 9 and 10) are relat- high potency. SAHA has a potency at µM range ed to yeast Hda1 (histone deacetylase 1) and and has recently been approved for the treat- Pre-clinical observations of can shuttle between the nucleus and cyto- ment of cutaneous T-cell lymphoma. M-car- HDACi in multiple myeloma plasm. This class is divided into class IIa boxycinamic acid bishydroxamide (CBHA) is (HDACs 4, 5, 7 and 9) and class IIb (HDAC6 another potent second generation inhibitor Anti-myeloma activity of HDACi as and 10) which contain two deacetylase which is the structural basis for example a single agent in vitro domains.30 Since HDAC6 contains a unique LAQ824 and PXD101, both effective at nM alpha-tubulin deacetylase (TCAD) domain, it range towards classes I and II. Two of the HDACi modulate the gene expression can specifically deacetylate alpha-tubulin.32 newest hydroxamate based HDACi are LBH589 profile of multiple myeloma cells The third class HDACs are the sirtuins (SIRT and ITF2357 with very low IC50 values at nM Microarray based studies showed that 1, 2, 3, 4, 5, 6 and 7) which are homologs to the concentrations.43,48 Benzamides include MS- HDACi induce transcriptional modulations of yeast Sir2 (silent information regulator 2) 275 and CI-994 and are generally less potent 7-10 % of the genes in malignant cell lines by family. These enzymes require nicotine ade- than the hydroxamates and cyclic tetrapep- acetylation of histones and non-histone pro- nine dinucleotide (NAD)+ for their deacetylase tides. Cyclic tetrapeptides, include the natural teins.55-57 The patterns of the HDACi induced activity in contrast to the zinc-catalyzed mech- product depsipeptide (FK 228 or FR 901228) gene expression alterations are quite similar anism used in class I and II HDACs.29 The sir- and apicidin. Depsipeptide is a prodrug and for different HDAC inhibitors. Definite differ- tuins appear to deacetylate non-histone pro- has to be metabolically activated via reduction ences, however, could be observed by different teins and transcription factors including p53. of the disulfite binding.45 Recently KD5170, a agents in different cancer cell lines.58,59 They can not be inhibited by HDACi such as novel mercaptoketone-based histone deacety- In MM, the first cDNA array using SAHA in suberoylanilide hydroxamic acid (SAHA) or lase inhibitor, has been developed. KD5170 the human MM1S cell line was performed by Trichostatin A (TSA).33 HDAC11 represents showed significant anti-proliferative activity Mitsiades et al. SAHA caused selective gene class IV and contains residues in the catalytic against a variety of human tumor cell lines, expression alterations of oncogenes, prolifera- core regions similar to both class I and II including human MM cell lines.44 tive/anti-apoptotic transcription factors, cell enzymes but does not have strong enough cycle regulators and members of the IGF-1R Isoenzyme-selectivity of pan-HDACi and identity to be placed in either class.34 and IL-6R signaling cascades.55 Recently, gene mechanism of HDAC inhibition expression profiling of MM1S cells exposed to In general, none of these inhibitors, except HDAC inhibitors VPA have also been performed and showed that tubacin, exhibit specificity towards one isoen- VPA also targeted genes involved in the cellu- Structural classification of HDAC inhibitors zyme. However, they inhibit the enzyme activ- lar pathways crucial for the survival of the MM Butyrate and TSA were among the first ity of HDACs with varying efficiency (Table 1). cells as seen for SAHA. Furthermore, they chemicals to be identified as HDAC inhibitors. For example, depsipeptide preferentially could demonstrate modulation of genes that Dimethylsulfoxide was used to aid superinfec- inhibits HDAC1 and 2 compared to HDAC4 and contribute to RNA splicing/transcription and tion of murine erythroleukemia cells with the 6, whereas the potency of MS-275 to inhibit DNA replication, indicating that HDACi could Friend virus, whereas TSA was originally iso- HDAC1 is 26 times higher compared to HDAC3 affect cell growth differently from apoptosis or lated as an antifungal agent from culture medi- and appears to lack the ability to inhibit the cell cycle regulation.56 um of Streptomyces hygroscopicus. Later on, it HDAC6 and 8.45,49 Tubacin, the HDAC6 selective was discovered that these compounds could inhibitor, induces hyperacetylation of α-tubu- HDACi inhibit the proliferation of multiple induce cell differentiation and a correlation lin and has no effect on the histone acetylation myeloma cells with histone hyperacetylation, which was status, while other hydroxamate inhibitors like Before investigating the molecular effect of maintained by inhibiting HDACs, could be TSA, SAHA and LBH589 induce histone – and HDACi in certain human MM cell lines, assays shown.35-39 It subsequently opened a new field α-tubulin hyperacetylation.42,50-52 such as 3-(4,5-dimethylthiazol-2-yl)-2,5-di - of research. Since then, a large number of nat- X-ray crystallographic analyses clarify the phenyl tetrazolium bromide (MTT)- or 3H-thy - ural and synthetic HDACi have been developed structure of an HDAC enzyme using an HDAC- midine incorporation assays were performed to by several companies and used as anti-tumor like protein (HDLP) isolated from an anaero- study the anti-proliferative effect of the HDACi. agents in pre-clinical and clinical settings bic bacterium, on the one hand and on the Table 2 shows an overview of different HDACi (Table 1). On the basis of their chemical struc- other hand how inhibitors such as SAHA and and their concentration range needed to inhib- ture, major HDACi can be divided into four cat- TSA mediate HDAC inhibition. The HDAC cat- it the proliferation of the human MM cell lines egories: short-chain fatty acids, hydroxamates, alytic domain consists of a tube like pocket and/or primary human MM cells. benzamides and cyclic tetrapeptides.26,46,47 whereby a Zn2+ cation is positioned near the HDACi such as VPA, FK228 and ITF2357 Among the various classes of HDACi, short bottom of this narrow pocket. The basic struc- affected the viability of IL-6 dependent as well chain fatty acids such as phenylbutyrate, the ture of the HDACi contains a cap group, an as IL-6 independent MM cell lines, indicating

[Hematology Reviews 2009; 1:e9] [page 47] Article

Table 1. Isoenzyme-selectivity of pan-HDACi. posed to occur selectively in tumor cells.77 The U266 human MM cell line, although express- Class Compound HDAC specificity Company/Sponsor Ref. ing significant levels of DR4 and caspase-8, is Short-chain fatty acid Butyrate Class I, IIa Merck 40 resistant to Apo2L/TRAIL and this resistance Valproic acid Class I, IIa NCI 41 could be overcome with VPA. This sensitizing Hydroxamate SAHA Class I, II Merck 40 effect of VPA is mediated by the redistribution PXD101 Class I, II CuraGen 40 of DR4 to lipid rafts followed by an improved LAQ824 Class I, II Novartis 40 DR4 signaling.62 However, opposite results LBH589 Class I, II Novartis 40 have been obtained by Schwartz et al. who Tubacin Class IIb BI and MIT 42 ITF2357 Class I, II Italfarmaco 43 have demonstrated that VPA activated caspase- 3 but not caspase-9 and caspase-8 in the U266, Mercaptoketone KD5170 Class I, II Kalypsys 44 OPM2 and RPMI human MM cell lines.63 In the Cyclic tetrapeptide Depsipeptide Class I Gloucester 45 MM1S line, treated with LBH589, no upregula- NCI: National Cancer Institute; BI: Broad Institute; MIT: Massachusetts Institute of Technology tion of death receptors and their ligands could be observed. Caspase-8, however, was activated and the gene expression of the TOSA gene, negative regulator of the Fas ligand (FasL) or Table 2. Potency of HDACi used in different in vitro MM models. TRAIL induced apoptosis was downregulated.69 HDACi Range MM cells Ref. SAHA sensitized MM1S cells to a Fas-activat- NaB mM U266, RPMI 8226, ARH-77, OPM2 60 ing monoclonal antibody CH-11 and to recom- VPA mM OPM1, MM1S, DOX-40, INA-6, OPM2, 56, 61, 62, 63 binant TRAIL. This sensitizing effect was asso- NCI-H929, LP-1, RPMI 8226, U266 ciated with decreased expression of the anti- SAHA µM MM1S 55, 64 apoptotic protein FLICE-like inhibitory protein LAQ824 Sub-µM primary human MM cells, MM1S, 65 (FLIP) and members of the inhibitors of apop- MM1R, RPMI 8226, -LR5, -MR20, -Dox40 tosis (IAP) family such as X-linked IAP 64 KD5170 R MM1S (XIAP). Sub-µM H929, U266, primary human MM cells 66 Despite these results showing that HDACi FK228 nM U266, RPMI 8266 67 affect the extrinsic pathway, in MM and other ITF2357 nM CMA-03 68 malignant cells it is still not clear how important the death-receptor pathway is for the ther- LBH589 nM primary human MM cells, MM1S, 69 MM1R, U266, -LR7, -Dox40 apeutic effects of HDACi. MM1S: dexamethasone S, IL-6 independent; MM1R: dexamethasone R; RPMI 8226, OPM1, CMA-03, DOX-40: IL-6 independent; LR5: melphalan The intrinsic apoptotic pathway is mediated R; MR20: mitoxantrone R; Dox40: doxorubicin R; U266: autocrine secretion of IL-6; INA-6, CMA-03: IL-6 dependent; OPM2: IL-6 dependent, by the mitochondria whereby proapoptotic sig- dexamethasone R when IL-6 is added; ARH-77: Epstein-Barr virus (EBV) positive cell line and thereby not considered as a genuine MM cell nals result in the release of mitochondrial line.70 S: sensitive; R: resistant intermembrane proteins, such as cytochrome c (cyto-c), apoptosis inducing factors (AIF) and second mitochondria-derived activator of caspase (Smac). Cytosolic cyto-c binds to apoptot- that the anti-myeloma activity of the HDACi is HDACi induce cell death in multiple myeloma ic protease activating factor (Apaf-1), resulting not influenced by one of the key growth factors Besides inhibition of proliferation, HDACi in Apaf-1 oligomerization and subsequent caspase-9 activation while cytosolic Smac binds to in MM.61,67,68 Furthermore, co-culturing the MM induced cell death is one of the major mecha- XIAP and thereby eliminates its inhibitory cells with bone marrow stromal cells (BMSC) nisms to inhibit the survival of the myeloma effect on caspase-9. Cytosolic AIF induces cas- cells. Extrinsic and intrinsic apoptotic path- does not protect the cells from cell death pase-independent apoptosis.78 Members from ways as well as non-apoptotic cell death such as induced by the HDACi LAQ824, ITF2357, the BCL2 family partially regulating this path- 65,66,68,71 autophagy have been reported in myeloma cells LBH589 or KD5170. These data suggest way, contain the pro-apoptotic (e.g. Bax, Bak, treated with an HDACi. Figure 1 demonstrates that HDACi could overcome the protective Bid and Bim) and anti-apoptotic (e.g. Bcl2, effects of the HDACi on the compounds of the effect of the BM micro-environment. The BclxL and Mcl1) proteins. The BCL2 protein MM1S cells were resistant to KD5170 and intrinsic and extrinsic apoptotic pathway. Bid, can be cleaved by caspase-8 after death- The extrinsic apoptotic pathway is activated showed no increase in histone acetylation, receptor ligation, and truncated Bid (tBid) by ligand binding to death receptors such as whereas KD5170 sensitive cell lines exhibited localizes to the mitochondria to initiate the Fas (Apo-1 or CD95), tumor necrosis factor 79 histone hyperacetylation after KD5170 treat- intrinsic apoptotic pathway. receptor-1 (TNFR-1) and TNF-related apopto- ment.66 This finding indicates that inhibition of MM cells contain higher levels of the anti- sis-inducing ligand (TRAIL or Apo2L) recep- the HDAC enzymes is necessary for the anti- apoptotic proteins Bcl2 and Mcl1 and lower lev- tors (DR4 and -5), resulting in activation of els of the pro-apoptotic protein Bax compared tumor effects of the HDACi. caspase-8 and caspase-10. Apo2L/TRAIL inter- to normal plasma cells.76,78 These findings could JNJ-26481585, a recently developed novel acts with two death receptors (DR4 and DR5) play a role in the survival of the MM cells and hydroxamate based HDACi with prolonged and potently induces apoptosis in various the resistance to chemotherapeutic agents. pharmacodynamic properties, has anti-prolif- tumors, including primary MM cells and MM How HDACi activate the intrinsic apoptotic erative effect at nM concentrations in the cell lines, while exerting minimal or no toxici- cascade is cell context dependent and is still murine 5T33MM model.72 This murine MM cell ty in normal cells.75,76 not completely understood. Treatment of the line is derived from the 5TMM mouse model Several studies have demonstrated that U266 MM human cell line and primary MM which mimics the human disease closely at the HDACi can upregulate the expression of both human cells with depsipeptide resulted in a molecular, cellular and clinical level.73,74 death receptors and their ligands and are pro- decrease of the anti-apoptotic proteins Mcl1,

[page 48] [Hematology Reviews 2009; 1:e9] Type ofArticle paper

involving the degradation of long-lived proteins or cytoplasmatic organelles through the lysosomal machinery.83 Schwartz et al. demonstrated for the first time that autophagy might be involved in VPA induced cytotoxicity in human myeloma cell lines. Only cleavage of caspase-3 and autophagic vacuoles in the cytoplasm could be observed in the myeloma cells treated with VPA, indicating that autophagic cell death might be involved.63 HDACi induce cell cycle arrest HDACi, except tubacin, induce cell cycle arrest at G1/S phase. The events in the G1 phase are coordinated by the three early G1 D cyclins (1, 2, 3) and their associated cyclin- dependent kinases (CDKs) 4/6 (G1 progression) and CDK 2 (G1/S transition). The transcriptional regulation of the genes, necessary for G1 progression and G1/S transition, depends on the phosphorylation state of the retinoblastoma (Rb) protein. Phosphorylation of the Rb protein by G1 D cyclin/ cyclin-dependent kinase (CDK) results in the release of E2F, allowing transcription activation and further progression through G1 and initation of S Figure 1. Induction of the extrinsic and intrinsic apoptotic pathway by HDACi in myelo- phase. The CDK inhibitors, including the INK4 ma cells. The extrinsic apoptotic pathway is triggered by ligand binding and leads to acti- family (p16) and the Cip/Kip family (p21, p27 vation of caspase-8, which, in turn, activates caspase-3. Activation of the intrinsic apop- and p57), are proteins that negatively regulate totic pathway results in the release of three compounds: (a) cytochrome-c (cyto-c) which binds to apoptotic protease activating factor (Apaf-1) to activate caspase-9, (b) apoptosis the cell cycle by competing with the cyclin D - inducing factors (AIF) and (c) second mitochondria-derived activator of caspase (Smac). CDK binding and therefore inhibiting the CDK FLICE-like inhibitory protein (FLIP) and members to the inhibitors of apoptosis (IAP) complex kinase activity. In MM, constitutive are able to prevent apoptosis induced by death receptors or intrinsic pathway respective- phosphorylation of the Rb protein may be fun- ly. Symbols denote compounds that are up-regulated (*), down-regulated (**), activated damental to the growth and development of the (∏) or translocated to cytosol (r) by HDACi in myeloma cells. tumor.52 The mRNA level of the three G1 D cyclins are elevated in virtually all MM tumors compared to healthy plasma cells and could be Bcl2, BclxL and an increase in Bax; this latter caspase-dependent, while the SAHA induced due to an Ig translocation or an unknown could only be observed in primary human MM cell death is calpain-dependent and caspase- mechanism. The elevated levels of the D cells.67 ITF2357 induced apoptosis through the independent.64,69 Furthermore, SAHA promotes cyclins are not sufficient to promote a cell cycle intrinsic pathway rather than through the cleavage of Bid to tBid while overexpression of and need a corresponding increase of CDK4 or extrinsic pathway in the KMS18 MM cell line the anti-apoptotic protein Bcl2 inhibited CDK6.84 Furthermore, several reports have since no cleavage of caspase-8 nor upregula- SAHA-induced apoptotic signaling.64 Recent demonstrated that p16 is frequently hyperme- tion of DR-4 have been found, whereas cleav- data indicate that KD5170 mediates cell death thylated in primary human MM cells. However, age of caspase-3 and -9 and downregulation of through mitochondrial perturbation in the no decreased mRNA could be found.85,86 Bcl2 and Mcl1 could be demonstrated.68 MM1S U266 cells. KD5170 provoked Bax activation HDACi induce cell cycle arrest in the G1/S cells treated with LBH589 underwent translo- and cleavage of caspase-9 and caspase-3, caus- phase which is mostly associated with induc- cation of cyto-c and AIF from the mitochondria ing loss of mitochondrial membrane potential tion of p21. This has been observed in the MM to the cytosol, upregulation of Apaf-1 and cleav- and subsequent pro-apoptotic factor release. cell lines treated with VPA, NVP-LAQ824, age of Bid, caspase-9 and caspase-3. Further- The fact that AIF was released, and that the LBH589, NaB, SAHA and ITF2357.60,61,64,65,68,69 MM more, gene expression profiling revealed a nuclear condensation was partially blocked in cells treated with VPA or LBH589 also showed novel apoptosis and caspase activation cells pre-treated with z-VAD-fmk before expo- a reduction of cyclin D1 and/or cyclin D2, indi- inhibitor, AVEN, which was downregulated by sure to the HDACi, suggest that KD5170 cating that induction of p21 is not solely treatment with LBH589.69 These data represent induced apoptosis through both caspase- responsible for cycle arrest.56,61 clear evidence that LBH589 caused cell death dependent and caspase-independent path- through mitochondrial perturbations. Both ways. Furthermore, KD5170 induced oxidative HDACi inhibit the aggresomal pathway LBH589 and SAHA induced poly (ADP-ribose) stress and oxidative DNA damage in myeloma in multiple myeloma polymerase (PARP) cleavage in MM cells by cells as evidenced by the upregulation of heme The aggresomal protein degradation system two different enzymes, caspase-3 and calpain, oxygenase-1 and H2A.X phosphorylation, represents an alternative system to the protea- respectively. Using the calpain inhibitor, which is a marker of DNA double strand some for degradation of polyubquitinated mis- calpeptin, and the caspase-3 inhibitor, benzy- breaks.66,80 folded/unfolded proteins (Figure 2).87 When loxycarbonyl-Val-Ala-Asp methylester-fluoro - Autophagy, an alternative model for apopto- degradation of misfolded proteins exceeds the me thyl ketone (z-VAD-fmk), they could demon- sis, has been reported to contribute to the proteasomal degradation through e.g. protea- strate in MM1S cells that the LBH589 induced HDACi induced cell death in several tumor cell some inhibitors, proteins interact with other cell death is calpain-independent and partially lines.81,82 Autophagy is a catabolic process unfolded or partially folded proteins, resulting

[Hematology Reviews 2009; 1:e9] [page 49] Article in accumulation of ubiquitinylated proteins, peripheral blood mononuclear cells (PBMCs) both transfected and untransfected cell lines, organized into perinuclear stuctures termed could be observed at µM levels.92 This indicates indicating that downregulation of the IL-6R is “aggresomes”.88,89 Aggresomes are formed by that tubacin sensitivity is independent of drug not required for the induction of p21 or apopto- the retrograde transport of the aggregated pro- resistance and that tubacin selectively targets sis.60 This observation again confirms that teins on microtubules (MT) and travel to the malignant cells. HDACi act on multiple cellular pathways. MT organizing center (MTOC) region, where Several studies provide evidence that HDACi affect cytokines and proteins they are sequestered as a single structure sus- HDACi suppress angiogenesis through a direct implicated in multiple myeloma survival, ceptible for lysosomal degradation. Movement effect on the growth and differentiation of progression and immune escape of aggresomes requires intact microtubules endothelial cells on one hand and by down-reg- Mitsiades et al. showed that SAHA suppress- and association with motor dynein. ulating the expression of pro-angiogenic es the expression of receptor genes involved in 93-95 HDAC6 deacetylates alfa-tubulin and plays a genes in tumor cells on the other hand. The MM cell proliferation, survival and/or migra- anti-angiogenic effect of HDACi in myeloma key role in the aggresomal pathway since it tion such as IGF-1R, IL-6R and its key signal has been demonstrated using OPM-2 and KM3 can bind poly ubiquitinated proteins and transducer gp130, TNF-R, CD138 (syndecan-1) cells treated with VPA. VPA decreases VEGF dynein, facilitating the transport of aggre- 55 and CXCR-4. Furthermore, in MM1S cells they secretion and VEGF receptor expression, somes along the MTs.32,90 Targeting HDAC6 could demonstrate that SAHA suppressed resulting in inhibition of the vascular tubule with tubacin or pan HDAC inhibitors such as autocrine IGF-1 production and paracrine IL-6 formation of endothelial cells in co-cultures SAHA or LBH589, results in hyperacetylation of secretion of BMSC by triggering MM cell bind- with myeloma cells. These data confirm the alfa-tubulin, accumulation of polyubiquitinat- ing. This suggests that SAHA can overcome anti-angiogenic effect of HDACi on myeloma 71,91 ed proteins and apoptosis. It has been cell adhesion-mediated drug resistance.55,64 which is important to suppress spread of the shown that tubacin inhibits MM cell growth in OPM-2 cells treated with NaB decreased IL-6R MM cells.61,96,97 drug-sensitive (MM1S, U266, INA-6 and but when cells were transfected with an Recently, De Bruyne et al. showed that the RPMI8226) and drug-resistant cell lines expression vector of IL-6R no decrease of the tetraspanin CD9 which shows an inverse cor- (RPMI-LR5 and RPMI-Dox40) with an IC50 receptor could be observed. Increased p21 relation between its expression level and between 5-20 µM, whereas no cytotoxicity in expression and apoptosis could be observed in tumor metastasis in solid tumors, is epigenet- ically down-regulated in MM and could be up- regulated by treating myeloma cells with LBH589. Myeloma cells expressing CD9 become more susceptible for natural killer mediated cytolysis and the expression correlates with non-active MM disease. These observations suggest that the immune escape of the tumor cells and molecules, correlating with the MM disease status, can be affected by HDACi.98

Anti-myeloma activity of HDACi in combination therapy in vitro Bortezomib Bortezomib, a first-in-class, potent and reversible proteasome inhibitor, has been successfully introduced in clinical practice and represents the standard of care in symptomatic MM patients.99 The anti-myeloma activity of bortezomib is a result of NF-κB inhibition, upregulation of various apoptotic pathways, and effects on the tumor micro-environment.100-103 Pei et al. were the first to demonstrate in vitro that HDACi in combination with bortezomib resulted in an improved cytotoxic effect compared to their effect as single agent. Sequential exposure of U266 and MM1S cells to bortezomib and SAHA or NaB potently induced caspase-3, -8 and -9 activation and release of the pro-apoptotic mitochondrial proteins cyto-c and Smac, resulting in a synergistic induction of apoptosis. This effect was Figure 2. The aggresome pathway prevents accumulation of misfolded proteins. Unfolded associated with a reduction in NF- B DNA or misfolded proteins, that exceed proteasomal degradation, form aggregates and are κ transported to the microtubule organizing center (MTOC) for degradation. This trans- binding activity, modulation of JNK activation port requires HDAC6 which deacetylates alfa-tubulin and binds both polyubiquitinated and a reactive oxygen species (ROS)-depend- proteins and dynein. Inhibiting HDAC6 with tubacin, whether or not combined with the ent downregulation of Cyclin D1, Mcl-1 and proteasome inhibitor bortezomib, accumulates misfolded or unfolded proteins and leads XIAP. Combining bortezomib with PXD101 to apoptosis. caused oxidative stress accompanied by an

[page 50] [Hematology Reviews 2009; 1:e9] Type ofArticle paper enhanced effect on Bim expression, DNA dam- proteins, by simultaneously activating the Furthermore, the enhanced anti-myeloma age, MAPK p38 activation and p53 phosphory- intrinsic and extrinsic pathways.40,106 activity of LBH589 with bortezomib could be lation. These observations indicate that there demonstrated in vivo by Atadja et al. using a DNA methyltransferase inhibitors are several molecular mechanisms that may disseminated luciferized MM1S MM xeno - 5-azacitidine is a DNA methyltransferase contribute to the synergy between bortezomib graft mouse model.117 One of the major limita- inhibitor and shows activity against MM.107 5- and HDACi.104 Specific inhibition of the aggre- tions of these in vivo experiments is the lack azacitidine and analogs such as 5-azacytidine somal pathway by tubacin together with proof the interaction of MM cells with a human (decitabine) are interesting tools to investi- teasome inhibition by bortezomib also result- micro-environment and therefore a protective gate hypermethylation in tumorigenesis and ed in an accumulation of ubiquitinated pro- effect of the BM micro-environment against the clinical efficacy is currently being teins followed by a synergistic anti MM-activi- the anti-myeloma activity of the HDACi in assessed in phase II trials.108,109 Several investi- ty, mediated by stress-induced JNK activation, vivo cannot be excluded. gations have already shown that hypermethy- followed by caspase/PARP cleavage.92 In addi- Recently, the syngeneic murine 5T33 and lated tumor suppresor genes can be most effi- tion, further investigations on cytoskeletal 5T2MM models, which mimic the human ciently reactivated by combining DNA events showed that bortezomib alone lead to myeloma disease closely, have been used to demethylating agents with HDACi, this could aggresome formation and, combining it with investigate the anti-myeloma activity of JNJ- thereby result in an enhanced reduction of LBH589 or SAHA, both inhibiting HDAC6, 26481585.74 Injecting C57Bl/KaLwRij mice with tumor cell growth.110-114 resulted in a disruption of aggresome forma- 5T2 or 5T33MM cells results in a migration of Treatment of the human myeloma cell line, tion leading to apoptosis.71,91 Nawrocki et al. the MM cells to the BM followed by tumor U266 with NaB and decitabine resulted in a G1 demonstrated that the oncogen Myc regulates growth, induction of angiogenesis and induc- arrest, whereas no cell cycle arrest could be the sensitivity of MM cells to bortezomib in observed when the compounds were used as tion of a MM bone disease (only in the 5T2MM combination with SAHA. Oncogenic activation single agents. Also, the expression level of the model). 5T2 and 5T33MM mice treated with of Myc is a hallmark of nearly all rapidly divid- p16 gene on RNA and protein level was signif- JNJ-26481585 resulted in a significant decrease ing malignant cells. In MM, the Myc expres- icantly increased when both epigenetic agents in tumor load and a reduction in the MM bone 72 sion is directly correlated with intracellular were applied simultaneously.115 Our group disease. Moreover, when a very low dose of endoplasmatic reticulum (ER) content and could also show in the human myeloma JNJ-26481585 was combined with bortezomib, protein synthesis rate. Bortezomib in combi- Karpas707 cell line that the upregulation of the MM bone disease was more reduced than seen nation with SAHA resulted in an induction of pro-apoptotic protein Bim by LBH589 could be with bortezomib alone (Deleu et al., personal the pro-apoptotic BH3-only protein Noxa and enhanced by decitabine, while decitabine observations, 2009). These in vivo studies ER stress indicated by a disruption of calcium alone had no effect on Bim expression.116 demonstrated that the antimyeloma activity of homeostasis and activation of caspase-4. the HDACi as single agents or in combination Further knock-down studies demonstrated Conventional therapeutic agents with bortezomib could not be overcome by the that caspase-4 and Noxa play significant roles LAQ824, depsipeptide and LBH589 showed BM micro-environment. in Myc-driven sensitivity to the combination of an enhanced decrease in survival of human bortezomib and SAHA.91 MM cell lines with the conventional therapeu- Enhanced anti-MM activity of the combina- tic agents such as dexamethasone and mel- tion therapy could not only be observed in pri- phalan.65,67,69 Targeting different pathways could Clinical observations of HDACi mary human MM cells but also in co-culture contribute partially to the enhanced anti-MM in multiple myeloma conditions and conditions with exogenous effect; namely caspase-8 is activated by growth factors IL-6 or IGF-1. Taken together, LAQ824 and not by dexamethasone whereby Several clinical trials with HDACi alone or bortezomib in combination with HDACi may combining both agents provides an additional in combination with other antimyeloma agents represent a promising therapeutic strategy apoptotic signal to those already induced by are ongoing (Table 3).118-125 Phase I clinical tri- that can overcome drug-resistance.71,69,91,92 dexamethasone. Further investigations are needed to clarify the molecular mechanism of als showed that HDACi, such as SAHA, LBH589 Death receptor ligands the synergism between chemotherapeutic and depsipeptide are well tolerated in myeloma In several tumor cells, an enhanced apoptot- agents and HDACi. patients. In phase II clinical trials, the activity ic effect can be observed using HDACi and of the HDACi as single agent was limited. activators of the TRAIL and Fas pathway. Anti-myeloma activity of HDACi in However, combining HDACi with dexametha- However, the molecular mechanism underly- vivo sone and/or bortezomib resulted in a more ing this synergism is still unclear and is cell- To study the pathogenesis of MM and to find promising therapeutic setting in the treatment type specific. new treatment strategies, different animal of MM, even in patients with refractory and Fandy et al. demonstrated that TSA, as well models have been developed, each with their relapsed MM. as SAHA, in combination with TRAIL have own advantages and disadvantages.73 potent synergistic effect in the ARO-1 MM To determine whether in vivo the anti- cells.105 Similar apoptotic effects have been myeloma effects of LAQ824, VPA and KD5170 observed in MM1S, U266 and H929 cell lines correlate with their in vitro activity, human Future directions treated with KD5170 and TRAIL.66 The fact that MM xenografts in immunodeficient mice SAHA and TSA could up-regulate the two death were used. Xenograft murine models were It has become clear that pan-HDACi have receptors DR4 and DR5 in the MM cells, cou- subcutaneously injected with RPMI8226, anti-neoplastic activities by affecting multiple pled with a downregulation of anti-apoptotic OPM1 or H929 and daily treatment with pathways involved in cell growth, survival, proteins (Bcl-2 and XIAP) could explain the LAQ824, VPA or KD5170, respectively, started immune response and tumor vasculature. synergistic effect of combination therapy.105 when tumors were measurable. These in vivo However, the precise underlying mechanism of However, it has been shown that HDACi could studies resulted in a significant decrease in the inhibition of the different HDACs by pan- also achieve synergy with TRAIL without tumor growth and a significant increase in HDACi and their biological role in MM patho- changing the TRAIL receptors or anti-apoptotic survival of mice treated with the HDACi.65-67 genesis remain to be clarified. A greater

[Hematology Reviews 2009; 1:e9] [page 51] Article

Table 3. Ongoing clinical trials with HDACi as single agent or in combination therapy in tors and potential elements of a heritable MM patients. epigenetic code. Curr Opin Genet Dev Regimen Clinical trial Ref. 2006;16:125-36. 20. Strahl BD, Allis CD. The language of cova- Depsipeptide+bortezomib+dexamethasone Phase I/II 118 lent histone modifications. Nature 2000; LBH589 Phase I/II 119, 120 403:41-5. 21. Fuks F. DNA methylation and histone LBH589+bortezomib Phase IB 121 modifications: teaming up to silence SAHA Phase I 122 genes. Curr Opin Genet Dev 2005;15:490- 5. SAHA+bortezomib Phase I 123, 124 22. Lehrmann H, Pritchard LL, Harel-Bellan ITF2357 Phase II 125 A. Histone acetyltransferases and de - acetylases in the control of cell proliferation and differentiation. Adv Cancer Res 2002;86: 41-65. 23. Dokmanovic M, Marks PA. Prospects: his- understanding of the molecular effects of the ment of multiple myeloma. Ann Hematol tone deacetylase inhibitors. J Cell Bio ch- HDACi and the role of HDACs is essential in 2007;86:159-72. em 2005;96:293-304. selecting patients who are potential candi- 7. Anderson KC. Targeted therapy of multi- 24. Minucci S, Pelicci PG. Histone deacety- dates for HDACi therapy and in designing com- ple myeloma based upon tumor-microen- lase inhibitors and the promise of epige- bination studies. The development of isoform- vironmental interactions. Exp Hematol netic (and more) treatments for cancer. specific inhibitors would be a valuable tool to 2007;35: 155-62. Nat Rev Cancer 2006;6:38-51. investigate the biological role of specific 8. Podar K, Chauhan D, Anderson KC. Bone 25. Timmermann S, Lehrmann H, Polesskaya HDACs. However, it is still not clear whether marrow microenvironment and the iden- A, Harel-Bellan A. Histone acetylation and selective inhibition of HDACs has therapeutic tification of new targets for myeloma disease. Cell Mol Life Sci 2001;58:728-36. advantages over a pan-HDACi. Clinical trials therapy. Leukemia 2009;23:10-24. 26. Vanhaecke T, Papeleu P, Elaut G, Rogiers demonstrated promising anti-tumor responses 9. Parkin DM, Bray F, Ferlay J, Pisani P. V. Trichostatin A-like hydroxamate his- to HDACi, mainly in combination with other Global cancer statistics, 2002. CA Cancer tone de acetylase inhibitors as therapeutic agents such as bortezomib or dexamethasone J Clin 2005;55:74-108. agents: toxicological point of view. Curr which are already in clinical use. Therefore, 10. Jagannath S. Pathophysiological under- Med Chem 2004;11:1629-43. the development of new and improved HDACi pinnings of multiple myeloma progres- 27. Witt O, Deubzer HE, Milde T, Oehme I. should be encouraged together with their use sion. J Manag Care Pharm 2008;14:7-11. HDAC family: What are the cancer rele- in combination therapy to improve the out- 11. Palumbo A, Falco P, Corradini P, et al. vant targets? Cancer Lett 2009;277:8-21. come for MM patients. Melphalan, prednisone, and lenalidomide 28. Marks PA, Jiang X. Histone deacetylase treatment for newly diagnosed myeloma: inhibitors in programmed cell death and a report from the GIMEMA--Italian cancer therapy. Cell Cycle 2005;4:549-51. Multiple Myeloma Network. J Clin Oncol 29. Blander G, Guarente L. The Sir2 family of References 2007;25: 4459-65. protein deacetylases. Annu Rev Biochem 12. Kumar SK, Dingli D, Dispenzieri A, et al. 2004;73:417-35. 1. Atadja P, Shao W, Growney J, et al. Impact of pretransplant therapy in 30. Marks PA, Dokmanovic M. Histone deace - LBH589: efficacy and protection of bone patients with newly diagnosed myeloma tylase inhibitors: discovery and develop- integrity in multiple myeloma as demon- undergoing autologous SCT. Bone ment as anticancer agents. Expert Opin strated in cell lines and in vivo mouse Marrow Transpl ant 2008;41:1013-9. Investig Drugs 2005;14:1497-511. model. Haemato logica 2007;92(s1):151. 13. Child JA, Morgan GJ, Davies FE, et al. 31. Gregoretti IV, Lee YM, Goodson HV. 2. Caers J, Van Valckenborgh E, Menu E, et High-dose chemotherapy with hemato - Molecular evolution of the histone de - al. Unraveling the biology of multiple poietic stem-cell rescue for multiple acetyl ase family: functional implications myeloma disease: cancer stem cells, myeloma. N Engl J Med 2003;348:1875-83. of phylogenetic analysis. J Mol Biol 2004; acquired intracellular changes and inter- 14. Rajkumar SV, Palumbo A. Management of 338:17-31. actions with the surrounding micro-envi- newly diagnosed myeloma. Hematol Oncol 32. Hubbert C, Guardiola A, Shao R, et al. ronment. Bull Cancer 2008;95:301-13. Clin North Am 2007;21:1141-56, ix-x. HDAC6 is a microtubule-associated de - 3. Chauhan D, Anderson KC. Mechanisms of 15. Harousseau JL. Induction therapy in mul- ace tylase. Nature 2002;417:455-8. cell death and survival in multiple myelo- tiple myeloma. Hematology Am Soc 33. Marmorstein R. Structure and chemistry ma (MM): therapeutic implications. Hematol Educ Program 2008;2008:306-12. of the Sir2 family of NAD+-dependent Apoptosis 2003;8:337-43. 16. Baylin SB, Ohm JE. Epigenetic gene histone/protein deactylases. Biochem Soc 4. Roodman GD. Role of the bone marrow silencing in cancer – a mechanism for Trans 2004;32:904-9. microenvironment in multiple myeloma. J early oncogenic pathway addiction? Nat 34. Gao L, Cueto MA, Asselbergs F, Atadja P. Bone Miner Res 2002;17:1921-5. Rev Cancer 2006;6:107-16. Cloning and functional characterization 5. Bruno B, Giaccone L, Rotta M, Anderson K, 17. Lund AH, van Lohuizen M. Epigenetics of HDAC11, a novel member of the human Boccadoro M. Novel targeted drugs for the and cancer. Genes Dev 2004;18:2315-35. histone deacetylase family. J Biol Chem treatment of multiple myeloma: from 18. Davie JR. Covalent modifications of his- 2002;277:25748-55. bench to bedside. Leukemia 2005;19:1729- tones: expression from chromatin tem- 35. Candido EP, Reeves R, Davie JR. Sodium 38. plates. Curr Opin Genet Dev 1998;8:173-8. butyrate inhibits histone deacetylation in 6. Piazza FA, Gurrieri C, Trentin L, Semen - 19. Nightingale KP, O'Neill LP, Turner BM. cultured cells. Cell 1978;14:105-13. zato G. Towards a new age in the treat- Histone modifications: signalling recep- 36. Friend C, Scher W, Holland JG, Sato T.

[page 52] [Hematology Reviews 2009; 1:e9] Type ofArticle paper

Hemoglobin synthesis in murine virus- model. J Cancer Molecules 2008;4:117-21. 63. Schwartz C, Palissot V, Aouali N, et al. induced leukemic cells in vitro: stimula- 51. Fraczek J, Deleu S, Lukaszuk A, et al. Valproic acid induces non-apoptotic cell tion of erythroid differentiation by di - Screening of amide analogues of death mechanisms in multiple myeloma methyl sulfoxide. Proc Natl Acad Sci USA Trichostatin A in cultures of primary rat cell lines. Int J Oncol 2007;30:573-82. 1971;68:378-82. hepatocytes: search for potent and safe 64. Mitsiades N, Mitsiades CS, Richardson 37. Yoshida M, Hoshikawa Y, Koseki K, et al. HDAC inhibitors. Invest New Drugs 2008; PG, et al. Molecular sequelae of histone Structural specificity for biological activi- Sep 30. [Epub ahead of print]. deacetylase inhibition in human malig- ty of trichostatin A, a specific inhibitor of 52. Urashima M, Ogata A, Chauhan D, et al. nant B cells. Blood 2003;101:4055-62. mammalian cell cycle with potent differ- Inter leukin-6 promotes multiple myeloma 65. Catley L, Weisberg E, Tai YT, et al. NVP- entiation-inducing activity in Friend leuk- cell growth via phosphorylation of retino - LAQ824 is a potent novel histone deacety- emia cells. J Antibiot (Tokyo) 1990;43: blastoma protein. Blood 1996;88:2219-27. lase inhibitor with significant activity 1101-6. 53. Finnin MS, Donigian JR, Cohen A, et al. against multiple myeloma. Blood 2003; 38. Yoshida M, Nomura S, Beppu T. Effects of Structures of a histone deacetylase homo- 102:2615-22. trichostatins on differentiation of murine logue bound to the TSA and SAHA 66. Feng R, Ma H, Hassig CA, et al. KD5170, a erythroleukemia cells. Cancer Res 1987; inhibitors. Nature 1999;401:188-93. novel mercaptoketone-based histone de- 47:3688-91. 54. Van Ommeslaeghe K, Elaut G, Brecx V, et acetylase inhibitor, exerts antimyeloma 39. Richon VM, Emiliani S, Verdin E, et al. A al. Amide analogues of TSA: synthesis, effects by DNA damage and mitochondri- class of hybrid polar inducers of trans- binding mode analysis and HDAC inhibi- al signaling. Mol Cancer Ther 2008;7: formed cell differentiation inhibits his- tion. Bioorg Med Chem Lett 2003;13:1861- 1494-505. tone deacetylases. Proc Natl Acad Sci USA 4. 67. Khan SB, Maududi T, Barton K, et al. 1998;95:3003-7. 55. Mitsiades CS, Mitsiades NS, McMullan Analysis of histone deacetylase inhibitor, 40. Bolden JE, Peart MJ, Johnstone RW. CJ, et al. Transcriptional signature of his- depsipeptide (FR901228), effect on multi- Anticancer activities of histone deacety- tone deacetylase inhibition in multiple ple myeloma. Br J Haematol 2004;125:156- lase inhibitors. Nat Rev Drug Discov myeloma: biological and clinical implica- 61. 2006; 5:769-84. tions. Proc Natl Acad Sci USA 2004;101: 68. Golay J, Cuppini L, Leoni F, et al. The his- 41. Kramer OH, Zhu P, Ostendorff HP, et al. 540-5. tone de acetylase inhibitor ITF2357 has The histone deacetylase inhibitor valproic 56. Neri P, Tagliaferri P, Di Martino MT, et al. anti-leukemic activity in vitro and in vivo acid selectively induces proteasomal In vivo anti-myeloma activity and modula- and inhibits IL-6 and VEGF production by degradation of HDAC2. EMBO J 2003;22: tion of gene expression profile induced by stromal cells. Leukemia 2007;21:1892- 3411-20. valproic acid, a histone deacetylase 900. 42. Haggarty SJ, Koeller KM, Wong JC, et al. inhibitor. Br J Haematol 2008;143:520-31. 69. Maiso P, Carvajal-Vergara X, Ocio EM, et Domain-selective small-molecule 57. Van Lint C, Emiliani S, Verdin E. The al. The histone deacetylase inhibitor inhibitor of histone deacetylase 6 expression of a small fraction of cellular LBH589 is a potent antimyeloma agent (HDAC6)-mediated tubulin deacetylation. genes is changed in response to histone that overcomes drug resistance. Cancer Proc Natl Acad Sci USA 2003;100:4389-94. hyperacetylation. Gene Expr 1996;5:245- Res 2006; 66:5781-9. 43. Garber K. HDAC inhibitors overcome first 53. 70. Pellat-Deceunynk C, Amiot M, Bataille R, hurdle. Nat Biotechnol 2007;25:17-9. 58. Gray SG, Qian CN, Furge K, et al. Micro- et al. Human myeloma cell lines as a tool 44. Hassig CA, Symons KT, Guo X, Nguyen array profiling of the effects of histone for studying the biology of multiple myelo- PM, Annable T, Wash PL, et al. KD5170, a deacetylase inhibitors on gene expr e s - ma: a reappraisal 18 years after. Blood novel mercaptoketone-based histone sion in cancer cell lines. Int J Oncol 2004; 1995;86:4001-2. deacetylase inhibitor that exhibits broad 24:773-95. 71. Catley L, Weisberg E, Kiziltepe T, T et al. spectrum antitumor activity in vitro and 59. Peart MJ, Smyth GK, van Laar RK, et al. Aggresome induction by proteasome inh - in vivo. Mol Cancer Ther 2008;7:1054-65. Identifica tion and functional significance i b itor bortezomib and alpha-tubulin 45. Furumai R, Matsuyama A, Kobashi N, et of genes regulated by structurally differ- hyperacetylation by tubulin deacetylase al. FK228 (depsipeptide) as a natural pro- ent histone deacetylase inhibitors. Proc (TDAC) inhibitor LBH589 are synergistic drug that inhibits class I histone deacety- Natl Acad Sci USA 2005;102:3697-702. in myeloma cells. Blood 2006;108:3441-9. lases. Cancer Res 2002;62:4916-21. 60. Lavelle D, Chen YH, Hankewych M, De - 72. Deleu S, Lemaire M, Arts J, et al. The 46. Miller TA, Witter DJ, Belvedere S. Histone Simone J. Histone deacetylase inhibitors effect of JNJ-26481585, a novel hydroxa- deacetylase inhibitors. J Med Chem 2003; increase p21(WAF1) and induce apopto- mate-based histone deacetylase inhibitor, 46:5097-116. sis of human myeloma cell lines inde- on the development of multiple myeloma 47. Marks PA, Richon VM, Miller T, Kelly WK. pendent of decreased IL-6 receptor in the 5T2MM and 5T33MM murine mod- Histone deacetylase inhibitors. Adv Can - expression. Am J Hematol 2001;68:170-8. els. Leukemia 2009 in press. cer Res 2004;91:137-68. 61. Kaiser M, Zavrski I, Sterz J, et al. The 73. Caers J, Asosingh K, Van Riet I, et al. Of 48. Bhalla KN. Epigenetic and chromatin effects of the histone deacetylase mice and men: disease models of multiple modifiers as targeted therapy of hemato- inhibitor valproic acid on cell cycle, myeloma. Drug Discovery Today: Disease logic malignancies. J Clin Oncol 2005;23: growth suppression and apoptosis in mul- Models 2004;1:373-80. 3971-93. tiple myeloma. Haematologica 2006;91: 74. Vanderkerken K, Asosingh K, Croucher P, 49. Hu E, Dul E, Sung CM, et al. Identification 248-51. Van Camp B. Multiple myeloma biology: of novel isoform-selective inhibitors with- 62. Gomez-Benito M, Martinez-Lorenzo MJ, lessons from the 5TMM models. Immunol in class I histone deacetylases. J Pharma - Anel A, et al. Membrane expression of Rev 2003;194:196-206. col Exp Ther 2003;307:720-8. DR4, DR5 and caspase-8 levels, but not 75. Ashkenazi A. Targeting death and decoy 50. Deleu S, Fraczek J, Lukaszuk A, et al. Mcl-1, determine sensitivity of human receptors of the tumour-necrosis factor Screening of Trichostatin analogues myeloma cells to Apo2L/TRAIL. Exp Cell superfamily. Nat Rev Cancer 2002;2:420- based on cellular potency in the 5T33MM Res 2007;313:2378-88. 30.

[Hematology Reviews 2009; 1:e9] [page 53] Article

76. Oancea M, Mani A, Hussein MA, Almasan aggresome formation and cell viability in 14374-9. A. Apoptosis of multiple myeloma. Int J response to misfolded protein stress. Cell 104. Feng R, Oton A, Mapara MY, et al. The his- He m atol 2004;80:224-31. 2003;115:727-38. tone deacetylase inhibitor, PXD101, 77. Nakata S, Yoshida T, Horinaka M, et al. 91. Nawrocki ST, Carew JS, Maclean KH, et potentiates bortezomib-induced anti-mul- Histone de acetylase inhibitors upregulate al. Myc regulates aggresome formation, tiple myeloma effect by induction of death receptor 5/TRAIL-R2 and sensitize the induction of Noxa, and apoptosis in oxidative stress and DNA damage. Br J apoptosis induced by TRAIL/APO2-L in response to the combination of borte- Haematol 2007;139:385-97. human malignant tumor cells. Oncogene zomib and SAHA. Blood 2008;112:2917-26. 105. Fandy TE, Shankar S, Ross DD, et al. 2004;23:6261-71. 92. Hideshima T, Bradner JE, Wong J, et al. Interactive effects of HDAC inhibitors and 78. van de Donk NW, Lokhorst HM, Bloem AC. Small-molecule inhibition of proteasome TRAIL on apoptosis are associated with Growth factors and antiapoptotic signal- and aggresome function induces syner- cha nges in mitochondrial functions and ing pathways in multiple myeloma. gistic antitumor activity in multiple mye - exp re ssions of cell cycle regulatory genes Leukemia 2005;19:2177-85. loma. Proc Natl Acad Sci USA 2005;102: in multiple myeloma. Neoplasia 2005;7: 79. Green DR. Apoptotic pathways: paper 8567-72. 646-57. wraps stone blunts scissors. Cell 2000; 93. Deroanne CF, Bonjean K, Servotte S, et al. 106. Rosato RR, Grant S. Histone deacetylase 102:1-4. Histone deacetylases inhibitors as anti- inhibitors in clinical development. Expert 80. Burma S, Chen BP, Murphy M, et al. ATM angiogenic agents altering vascular Opin Investig Drugs 2004;13:21-38. phosphorylates histone H2AX in response endothelial growth factor signaling. 107. Kiziltepe T, Hideshima T, Catley L, et al. 5- to DNA double-strand breaks. J Biol Chem Oncogene 2002;21:427-36. Azacytidine, a DNA methyltransferase 2001;276:42462-7. 94. Michaelis M, Michaelis UR, Fleming I, et inhibitor, induces ATR-mediated DNA 81. Shao Y, Gao Z, Marks PA, Jiang X. al. Valproic acid inhibits angiogenesis in double-strand break responses, apoptosis, Apoptotic and autophagic cell death vitro and in vivo. Mol Pharmacol and synergistic cytotoxicity with doxoru- induced by histone deacetylase in - 2004;65:520-7. bicin and bortezomib against multiple hibitors. Proc Natl Acad Sci USA 2004; 95. Qian DZ, Kato Y, Shabbeer S, et al. Targ- myeloma cells. Mol Cancer Ther 2007;6: 101:18030-5. eting tumor angiogenesis with histone 1718-27. 82. Yamamoto S, Tanaka K, Sakimura R, et al. de a c etylase inhibitors: the hydroxamic 108. Jones PA, Baylin SB. The fundamental Suberoylanilide hydroxamic acid (SAHA) acid d e r i v a tive LBH589. Clin Cancer Res role of epigenetic events in cancer. Nat induces apoptosis or autophagy-associat- 2006; 12:634-42. Rev Genet 2002;3:415-28. ed cell death in chondrosarcoma cell 96. Dong XF, Song Q, Li LZ, et al. Histone 109. Ocio EM, Mateos MV, Maiso P, et al. New lines. Anticancer Res 2008;28:1585-91. deacetylase inhibitor valproic acid drugs in multiple myeloma: mechanisms 83. Gozuacik D, Kimchi A. Autophagy as a cell inhibits proliferation and induces apopto- of action and phase I/II clinical findings. death and tumor suppressor mechanism. sis in KM3 cells via downregulating VEGF Lancet Oncol 2008;9:1157-65. Oncogene 2004;23:2891-906. receptor. Neuro Endocrinol Lett 2007;28: 110. Cameron EE, Bachman KE, Myohanen S, 84. Ely S, Di Liberto M, Niesvizky R, et al. 775-80. et al. Synergy of demethylation and his- Mutually exclusive cyclin-dependent 97. Kitazoe K, Abe M, Hiasa M, et al. Valproic tone deacetylase inhibition in the re- kinase 4/cyclin D1 and cyclin-dependent acid exerts anti-tumor as well as anti- expression of genes silenced in cancer. kinase 6/cyclin D2 pairing inactivates angiogenic effects on myeloma. Int J Nat Genet 1999;21:103-7. retinoblastoma protein and promotes cell Hematol 2009;89:45-57. 111. Keen JC, Yan L, Mack KM, et al. A novel cycle dysregulation in multiple myeloma. 98. De Bruyne E, Bos TJ, Asosingh K, et al. histone deacetylase inhibitor, scriptaid, Cancer Res 2005;65:11345-53. Epigenetic silencing of the tetraspanin enhances expression of functional estro- 85. Dib A, Barlogie B, Shaughnessy JD Jr, CD9 during disease progression in multi- gen receptor alpha (ER) in ER negative Kuehl WM. Methylation and expression of ple myeloma cells and correlation with human breast cancer cells in combination the p16INK4A tumor suppressor gene in survival. Clin Cancer Res 2008;14:2918- with 5-aza 2'-deoxycytidine. Breast multiple myeloma. Blood 2007;109:1337- 26. Cancer Res Treat 2003;81:177-86. 8. 99. Dmoszynska A. Diagnosis and the current 112. Klisovic MI, Maghraby EA, Parthun MR, et 86. Gonzalez-Paz N, Chng WJ, McClure RF, et trends in multiple myeloma therapy. Pol al. Depsipeptide (FR 901228) promotes al. Tumor suppressor p16 methylation in Arch Med Wewn 2008;118:563-6. histone acetylation, gene transcription, multiple myeloma: biological and clinical 100. Adams J. The proteasome: structure, apoptosis and its activity is enhanced by implications. Blood 2007;109:1228-32. function, and role in the cell. Cancer DNA methyltransferase inhibitors in 87. Wickner S, Maurizi MR, Gottesman S. Treat Rev 2003;29:3-9. AML1/ETO-positive leukemic cells. Leuk e - Posttranslational quality control: folding, 101. Hideshima T, Mitsiades C, Akiyama M, et m ia 2003;17:350-8. refolding, and degrading proteins. al. Molecular mechanisms mediating 113. Shaker S, Bernstein M, Momparler LF, Science 1999;286:1888-93. anti myeloma activity of proteasome Momparler RL. Preclinical evaluation of 88. Bennett EJ, Bence NF, Jayakumar R, inhibitor PS-341. Blood 2003;101:1530-4. antineoplastic activity of inhibitors of Kopito RR. Global impairment of the ubiq- 102. Hideshima T, Richardson P, Chauhan D, DNA methylation (5-aza-2'-deoxycyti- uitin-proteasome system by nuclear or et al. The proteasome inhibitor PS-341 dine) and histone deacetylation (tricho- cytoplasmic protein aggregates precedes inhibits growth, induces apoptosis, and statin A, depsipeptide) in combination inclusion body formation. Mol Cell 2005; overcomes drug resistance in human against myeloid leukemic cells. Leuk Res 17:351-65. multiple myeloma cells. Cancer Res 2003;27:437-44. 89. Kopito RR. Aggresomes, inclusion bodies 2001;61: 3071-6. 114. Zhu WG, Otterson GA. The interaction of and protein aggregation. Trends Cell Biol 103. Mitsiades N, Mitsiades CS, Poulaki V, et histone deacetylase inhibitors and DNA 2000;10:524-30. al. Molecular sequelae of proteasome methyltransferase inhibitors in the treat- 90. Kawaguchi Y, Kovacs JJ, McLaurin A, et inhibition in human multiple myeloma ment of human cancer cells. Curr Med al. The deacetylase HDAC6 regulates cells. Proc Natl Acad Sci USA 2002;99: Chem Anticancer Agents 2003;3:187-99.

[page 54] [Hematology Reviews 2009; 1:e9] Type ofArticle paper

115. Du HL, Ren LM, Chen H, et al. Re-expres- 2008;112: 3698. patients with advanced multiple myeloma. sion of p16 gene in the myeloma cell line 119. Ottmann OG, Spencer A, Prince HM, et al. Leuk Lymphoma 2008;49:502-7. U266 induced by synergy of sodium Phase IA/II study of oral panobinostat 123.Weber D, Badros AZ, Jagannath S, et al. butyrate and 5-Aza-2'-deoxycytidine. Di Yi (LBH589), a novel pan-deacetylase Vorinostat Plus Bortezomib for the Jun Yi Da Xue Xue Bao 2002;22:981-4. inhibitor (DACi) demonstrating efficacy Treatment of Relapsed/Refractory Multiple 116. De Bruyne E, Bos T, Deleu S, et al. in patients with advanced Hematologic Myeloma: Early Clinical Experience. ASH Regulation of Bim expression by IGF-1 in Malignancies. ASH Annual Meeting the 5T33MM murine model for multiple Abstracts 2008;112:958. Annual Meeting Abstracts 2008;112:871. myeloma. ASH Annual Meeting Abstracts 120.Wolf JL, Siegel D, Matous J, et al. A Phase 124.Badros A, Philip S, Niesvizky R, et al. Phase 2007;110:3512. II Study of Oral Panobinostat (LBH589) in I Trial of Suberoylanilide Hydroxamic Acid 117. Atadja P, Shao W, Growney J, et al. Adult Patients with Advanced Refractory (SAHA) + Bortezomib (Bo rt) in Relapsed LBH589: efficacy and protection of bone Multiple Myeloma. ASH Annual Meeting Multiple Myeloma (MM) Patients (pts). integrity in multiple myeloma as demon- Abstracts 2008;112:2774. ASH Annual Meeting Abstracts 2007;110: strated in cell lines and in vivo mouse 121.Siegel Dd, Sezer O, San Miguel JF, et al. A 1168. model. Haematologica 2007;92:151 Phase IB, Multicenter, Open-Label, Dose- 125.Galli M, Salmoiraghi S, Golay J, et al. A 118. Harrison SJ, Quach H, Yuen K, et al. High Escalation Study of Oral Panobinostat Phase II Multiple Dose Clinical Trial of response rates with the combination of (LBH589) and I.V. Bortezomib in Patients Histone Deacetylase Inhibitor ITF2357 in bortezomib, dexamethasone and the pan- with Relapsed Multiple Myeloma. ASH histone deacetylase inhibitor romidepsin Annual Meeting Abstracts 2008;112:2781. Patients with Relapsed or Progressive in patients with relapsed or refractory 122.Richardson P, Mitsiades C, Colson K, et al. Multiple Myeloma: Preliminary Results. multiple myeloma in a phase I/II clinical Phase I trial of oral vorinostat (suberoy- ASH Annual Meeting Abstracts 2007;110: trial. ASH Annual Meeting Abstracts lanilide hydroxamic acid, SAHA) in 1175.

[Hematology Reviews 2009; 1:e9] [page 55] Hematology Reviews 2009; volume 1:e10

The role of JAK2 abnormalities In this article, our aim is to review the role of JAK2 abnormalities in the pathogenesis, Correspondence: Mohammed K. Alabdulaali, in hematologic neoplasms diagnosis, classification, severity and manage- Blood Diseases Center, King Fahad Hofuf ment of hematologic neoplasms. Hospital, P.O. Box: 1 31982, Hassa, Saudi Arabia Mohammed K. Alabdulaali E-mail: [email protected] Blood Diseases Center, King Fahad Hofuf Hospital, Hassa, Saudi Arabia Key words: JAK2, myeloproliferative neoplasms, Classification of myeloid neo- leukemia. plasms Acknowledgment: I would like to thank Dr. Khalid M. Alayed, King Khalid University Hospital, for Abstract As JAK2 abnormalities are mainly identified his helpful discussion and enlightening views. in myeloid neoplasms, we will summarize the Received for publication: 3 May 2009. In 2005, an activating mutation in the Janus current classification scheme for these disor- kinase 2 (JAK2) was identified in a significant Revision received: 3 June 2009. ders. In general they are divided into 3 major Accepted for publication: 24 June 2009. proportion of patients with myeloproliferative groups: acute leukemia, chronic leukemia and neoplasms, mainly polycythemia vera, essen- myelodysplastic syndrome. Chronic leukemia This work is licensed under a Creative Commons tial thrombocythemia and primary myelofibro- can be sub-divided into BCR/ABL positive or Attribution 3.0 License (by-nc 3.0) sis. Many types of mutations in the JAK-STAT negative, those BCR/ABL positive will be pathway have been identified, the majority are ©Copyright M.K. Alabdulaali et al., 2009 labeled as CML regardless of their clinico- related to JAK2. Currently JAK2 mutations are Licensee PAGEPress, Italy pathological features unless it is presenting as important in the area of diagnosis of myeloid acute leukemia, while those BCR/ABL negative neoplasms, but its role beyond the confirma- Hematology Reviews 2009; 1:e10 will be divided further into classical MPNs, doi:10.4081/hr.2009.e10 tion of clonality is growing and widening our non-classical MPNs with or without dysplasia, knowledge about these disorders. In addition and myelodysplastic syndrome (MDS) (Figure to that, clinical trials to target JAK2-STAT path- 1). The diagnosis and classification of these way will widen our knowledge and hopefully disorders are based on peripheral blood signal transmission downstream and have will offer more therapeutic options. In this counts, blast percentage, type of myelosis, seven domains, two of which are structurally review, we will discuss the role of JAK2 abnor- presence of significant dysplasia, extent of similar. One of these (JH1) is an activating malities in the pathogenesis, diagnosis, classi- fibrosis, clinical features, biochemistry and domain while the other (JH2) seems to exert fication, severity and management of hemato- most importantly genetics.2-4 an inhibitory effect (Figure 3). logic neoplasms. JAK2 abnormalities are not only associated Upon ligand binding to its specific receptor, with most of the classical myeloid neoplasms JAK protein will be activated and it will then but they are also seen in association with other phosphorylate the downstream signaling mole- Introduction myeloid neoplasms except BCR/ABL positive cules like STATs, which will be actively trans- CML and acute lymphoid leukemia (ALL) where ported to the nucleus where it will activate it is only rarely reported as we will see later on. transcription factors (Figure 4). For decades the diagnosis, classification One of the interesting things regarding the Abnormalities in JAK1 have been reported and management of hematologic neoplasms clinical features of myeloid neoplasms is their in ALL of mainly T cell type where it is found in was based on the clinico-pathological features tendency to transform to acute leukemia; the nearly 20% of the cases,8 JAK2 is found in of these disorders. Major progress in both the myeloid neoplasms and rarely rearranged in therapeutics and our understanding of these classical MPNs, beside their pre-leukemic ALL,6 JAK3 have been reported in more than diseases did not occur until 1960 when behavior, progress and regress to each other 3-5 Philadelphia chromosome was found in bone (Figure 2). So, why does an abnormal gene 50% of transient abnormal myelopoiesis in marrow cells of patients with chronic myeloge- give rise to different disorders? Why do classi- Down syndrome patients, acute myeloid nous leukemia (CML), followed by the identifi- cal MPNs progress and regress? What are the leukemia (AML) megakaryoblastic type (M7) cation of the molecular defect by the fusion of roles of JAK2 abnormalities in the pathogene- and in some cases of severe combined immune 9-11 BCR and ABL genes in 1982. Fifteen years later sis of hematologic neoplasms? deficiency (SCID), and TYK2 might have a imatinib, an ABL tyrosine kinase (TYK) role in lymphoid neoplasms and natural killer 12 inhibitor, was developed in the late 1990s. cell functional defects. Nowadays the diagnosis of CML is dependent on the identification of t(9;22) (q34;q11) or JAK family BCR/ABL fusion gene and imatinib is one of its first line therapeutics. Other hematologic neo- Janus kinase is a family of intracellular non- JAK2 plasms are tracking the footprints of CML. receptor tyrosine kinases that transduce BCR/ABL negative myeloproliferative neo- cytokine-mediated signals. At present, it con- JAK2 was mapped on the short arm of chro- plasms (MPNs), particularly the classical poly- sists of 4 members: JAK1, JAK2, JAK3 and mosome 9p24 in 1992 by Pritchard and his cythemia vera (PV), essential thrombocythemia TYK2.6 colleagues,13 It has 140 kb spanning 25 exons (ET) and primary myelofibrosis (PMF), are cur- Janus kinases was named after Janus or to form 1132 aminoacid JAK2 protein.14 It rently drawing the attention of many scientists Ianus who in Roman mythology was believed to works as a signaling molecule for many in the fields of hematology, oncology, pathology, be the God of gates, beginnings and endings. cytokines including: INF-γ,15 erythropoietin genetics and pharmacology, after the identifica- He was imagined as having two faces or heads (EPO),16 prolactin,17 thrombopoietin (TMP), tion of Janus kinase 2 (JAK2) mutation in a sig- facing in opposite directions.7 G-CSF, GM-CSF18 and IL-319 via activating nificant number of patients diagnosed for these Indeed, Janus kinases are located just many signaling pathways like: MAPK, PI3,16 disorders in 2005.1 beneath the cellular receptors to control the ERK20 and STATs.14 PRV1 (CD177) and NF-E2

[page 56] [Hematology Reviews 2009; 1:e10] Review also appear to be activated and over- Figure 1. The cur- expressed by JAK2.6 One of the most impor- AML Chronic neoplasms MDS rent classification tant pathways that are activated by JAK2 is of myeloid neo- STAT5 followed by activation of BCL-XL and plasm. CNL: finally upregulation of BCL2 where the cell Rec. cytogen. abnormalities BCR/ABL +ve BCR/ABL -ve chronic neutro- will gain a survival advantage.6 20% blasts philic leukemia, ≥ Classical CMML: chronic ET CML myelomonocytic PV l e u k e m i a , PMF JCMML: juvenile Types of JAK2 abnormalities Classical CNL myelomonocytic Fibrosis CMML leukemia, aCML: and pathogenesis Neutrophilia JCMML atypical chronic Monocytosis aCML myelogenous In general we can divide these into 4 cate- Increased megaka Eosinophilia leukemia. gories:4,21-25 Mastocytosis A. Rearrangements JAK2 can be rearranged to: - TEL/ETV6: t(9;12) (p24;p13) reported in Non classical MPN 2-95% Figure 2. The rate some CML like MPNs and T-cell ALL; MPN/MDS of progression of - BCR: t(9;22) (p24;q11.2) reported in some MDS chronic myeloid CML like MPNs; neoplasms to - PCM1: t(8;9) (p22;p24) reported in some 2-5% leukemic phase. MPNs, AML and ALL; Pre: pre-fibrotic 1-2% 3-15% stage of PMF and - NF-E2: der(9) t(9;12) (p24;q13) reported ET PV unexplained in some cases of MDS. AML thrombotic events B. Point mutations 15-20% prior to the devel- ALL opment of MPN. - V617F G >T at nucleotide 1849 on exon14, 2-15% 5-30% reported in mainly the classical MPNs; PMF - T875N reported in AML (M7); Pre - R683G and less frequently other R683 point mutations reported in 18-28% of ALL in Down syndrome patients and 10% of a high-risk cohort of childhood ALL in patients without Down syndrome. C. Deletions/Insertions - Exon12: there are more than eight reported mutations including deletions Figure 3. The structure of Janus kinases and insertions in 538 to 543 codons report ed in 4% of PV cases; - IREED del which is a five amino acids deletion in JH2 pseudokinase domain reported 1 23 in B-cell ALL in Down syndrome patients. Ligand Ligand D. Numerical Ligand - It can present as trisomy (+9) or be overexpressed due to amplification. The majority of these abnormalities are affecting the JH2 domain leading to loss of inhibitory effects on JH1 domain, hence the later will be auto-activated. Suppressors of cytokine signaling (SOCS) 1 and 3 are negative regulators for JAK2 kinase, these suppres- Activated JAK sors will also be phosphorylated and stabilized JAK JAK by the hyperactive tyrosine kinase.26 SOCS3 Dimerized promoter methylation is another mechanism P STATs that is found in a group of patients.27 P It is reported that 5-10% of MPN patients have at least one relative who is affected by P P this disorder, and in familial MPNs the risk of STAT developing these disorders increases 6-fold.6,28 Phosphorylated STAT Recent studies suggest that particular single nucleotide polymorphisms (SNP) are associat- Activation of ed with a higher risk of developing JAK2 V617F transcription factors mutation. Out of about 659 SNPs, rs10974944 and rs12343867 are reported in 77% and 85% in association with JAK2 V617F mutation Figure 4. Steps of JAK-STAT pathway.

[Hematology Reviews 2009; 1:e10] [page 57] Review respectively, with a significant difference if without overt neoplasms.37 Not only, V617F compared to wild-type JAK2.29 The roles of JAK2 abnormali- mutation alone is seen in a wide range of ties in hematologic neoplasms myeloid neoplasms. This fact has created much hypothesis regarding the pathophysiology of these neoplastic conditions to the extent Prevalence of JAK2 abnormali- Confirmation of clonality The long list of congenital, secondary or that some authors suggest that some disorders ties in hematologic neoplasms reactive causes for cytosis, cytopenia and dys- are in fact phases of a single disease. Many plasia is creating difficulties in labeling cases factors possibly play a role (Figure 5). JAK2 V617F mutation is reported only in for neoplastic conditions without ruling them The type of abnormality myeloid neoplasms with a high frequency in PV, out, a process that requires extensive investi- Some abnormalities are only reported in ET, PMF and refractory anemia with ring sider- gation steps not to mention the time needed to myeloid neoplasms. These are JAK2 oblast and thrombocytosis (RARS-t). It is rarely confirm persistence. Confirming clonality can reported in CML,30 but not in ALL or molecularly bypass all these issues and can solve the prob- rearrangements to BCR: t(9;22) and NF-E2: characterized eosinophilic neoplasms and mas- lematic cases in which a neoplastic condition der(9)t(9;12), V617F and T875N point muta- tocytosis, i.e. those with abnormalities in is co-existing with a secondary or congenital tions and exon 12 mutations. On the other PDGFRA, PDGFRB, FGFR1 or KIT. JAK2 cause. Many methods are used to detect clonal- hand, R683G and IREED del are reported in rearrangements are seen in some cases of AML, ity but the most dependable is by performing lymphoid disorders. While rearrangements to atypical CML and ALL, while exon 12 mutations cytogenetic, fluorescence in situ hybridization TEL/ETV6: t(9;12) and PCM1: t(8;9) are 3,4,31,32 are only reported in PV (Table 1). (FISH) and/or molecular tests. Abnormal reported in both lineages. Some of these immunophenotype and loss of X-linked poly- abnormalities are only reported in a single morphism have many limitations in MPNs. It entity (Table 1). is currently widely accepted that detecting Screening and quantification The targeted cell or receptor JAK2 mutation, particularly V617F mutation, is tests a major criterion in diagnosing myeloid neo- It is suggested that the phenotype is deter- plasms, especially MPNs. It is now incorporat- mined by the ability of the affected stem cells Detecting JAK2 mutations, determination of ed in the WHO 2008, Nordic 2007, and BCSH to differentiate into different lineages, and on hetero-or homozygosity, and the allele burden, 2007 classifications.3,35,36 the receptor affected by the mutation. For i.e. the ratio of the mutant allele to the wild- example, EPO, TMP, G-CSF or GM-CSF recep- type, are the aims of carrying out molecular Phenotype and lineage determina- tors are targeted to give PV, ET, PMF or chron- studies. Variable screening techniques are tion ic myelomonocytic leukemia, respectively.38 available and these are basically utilizing JAK2 abnormalities are seen in many types Another example is the identification of direct sequencing, allele specific polymerase of hematologic neoplasms as mentioned earli- homozygous JAK2 V617F mutation in the chain reaction (PCR) analysis or ultra sensi- er, and also are reported in 70-80% of cases endothelial cells in the vessels of PV patients tive PCR techniques with variable sensitivities with unexplained hepatic venous thrombosis with Budd-Chiari syndrome.39 that can detect mutant gene at the levels of 20%, 3% and 0.01%, respectively. However, quantification techniques are preferred in Table 1. Prevalence of JAK2 abnormalities in hematologic neoplasms. order to estimate the mutant allele burden, to Neoplasms JAK2 V617F mutation Other JAK2 abnormalities monitor the disease course and effect of therapeutics. Peripheral blood and bone marrow Polycythemia vera >95% 4% JAK2 exon12 mutation samples, frozen plasma and paraffin-embed- Essential thrombocythemia 50-60% ded trephine bone marrow biopsies can all be Idiopathic myelofibrosis 40-50% 2,33 used. Eric Lippert and his colleagues studied Chronic neutrophilic leukemia 20% the concordance of assays designed for the Chronic myelomonocytic leukemia 3-13% quantification of JAK2 V617F and reported the Juvenile chronic myelomonocytic leukemia 20% highest sensitivity by using Taqman allele specific PCR with reverse and forward primers Atypical chronic myeloid leukemia 20% JAK2 rearrangement, t(8;9), t(9;12), t(9;22) with detection sensitivity up to 0.2% and MPN/MDS-U 12-20% 0.15%, respectively. But they found these tech- MPN/MDS (RARSt) 50-70% niques laborious, having false positive results Myelodysplastic syndrome (RAEB,-5q) 1-7% JAK2 rearrangement, t(9;12) and they were, as expected, capable of detect- Secondary AML 4% JAK2 rearrangement, t(8;9) ing only the mutation of interest. On the other AML (M6) Few reported cases hand, they found pyrosequencing and direct AML (M7) 15% JAK2 T875N DNA sequencing to be the least sensitive, lim- Mastocytosis 0-25% JAK2 rearrangement, t(8;9), t(9;12) ited to the level of 2% and 5%, respectively, but still having the advantage of detecting new Eosinophilic neoplasms 0-2% mutations.34 It has recently been suggested Chronic myelogenous leukemia Rare reported cases that immunoprecipitations and Western blot- ALL Non JAK2 rearrangement, t(8;9), ting to test for SOCS3 tyrosine phosphorylation t(9;12), JAK2 R683G and less may be a novel bio-marker of MPNs resulting frequently other R683 point mutations from a JAK2 mutation and a potential reporter ALL in Down syndrome patients Non JAK2 R683G, IREED del, of effective JAK2 inhibitor therapy currently in other point mutations, insertions and deletions clinical development.26

[page 58] [Hematology Reviews 2009; 1:e10] Review

Genetic background of the patient abnormality TypeJAK2 of Gender is one of the possible determinants, Figure 5. Factors that might determine the phenotype of as ET tend to occur in females more than JAK2 mutation positive males, while the opposite is true of PV. Iron disorders. metabolism is another factor, as iron depletion will lead to ET phenotype rather than PV. The Genetic modifiers opposite picture will be seen in relation to EPO bioavailability. SNPs are another possible par- ticipant as rs10974944 is significantly seen in Mutation dosage patients with PV if compared to ET.4,29,38 Dosage effect This appears to be the most important fac- EPO level Phenotype Iron level tor, and it can be explained by different thresh- olds of JAK2 kinase activity at which variable receptors, cytokines and proteasomes need to lineage of cell react. This is supported by the reported results in which mutation occurs of homozygous JAK2 V617F mutation occurring in 25-30% PV compared to 2-4% in ET. Few cases are reported that demonstrate dosage 100 dependent phenotype, like those transforming 90 from ET to PV with increasing JAK2 V617F 80 mutation burden, or from PMF to PV with bur- 70 den reduction after treatment with hydrox- 60 PMF yurea (HU).40 We have also encountered a case JAK 50 of PMF that transformed to ET after treatment mutation 40 ET PV with HU (K Alkhairi, KM Alayed, MK burden 30 Alabdulaali, unpublished data, 2007) By moni- 20 toring allele burden it is found that JAK2 V617F 10 Blasts mutation burden is increasing significantly 0 from ET, PV to PMF.6,38,40-44 The levels at which Thrombocytosis Erythrocytosis MF Trasformation AL the burden is found seem to correspond to a specific phenotype of 25%, 55%, 50% and 60% Figure 6. JAK2 mutation burden during progression of MPNs. AL: acute leukemia, ET: for ET, PV, PMF and post ET or PV myelofibro- essential thrombocythemia, MF: myelofibrosis, PMF: primary myelofibrosis, PV: poly- sis, respectively.43 Hypermethylation of the cythemia vera. SOCS3 promoter was identified in nearly 1/3 of patients with myelofibrosis and in both JAK2 mutation positive and negative cases,27 but not with wild-type JAK2. In a systematic literature increased risk of evolution to acute leukemia. in other types of MPNs; this again will review, Panayiotis D. Ziakas analyzed 17 stud- However, there is good evidence of its associa- strengthen the role of dosage effect. ies with 2,905 ET patients, of whom 1,646 tion with progression from ET to PV to PMF. If Pre-JAK2 mutation event (65.7%) patients were JAK2 V617F mutation evolution to leukemia then occurs, in the Several findings raise the possibility of a positive. Thrombotic events occurred in 523 majority of cases the leukemic cells will be pre- JAK2 mutation event leading or participat- (31.8%) of mutation positive cases and in 255 JAK2 mutation negative as mentioned earlier. ing in the development of neoplasia: the pres- (20.3%) of patients with wild-type JAK2. The overwhelming blasts will, therefore, ence of JAK2 mutation negative MPNs with Finally, he concluded that JAK2 V617F patients reduce the allele mutation burden which might 25% abnormal cytogenetics, the finding that have a 2-fold risk of developing thrombosis be considered a sign of evolution if it is not 50% of JAK2 mutation positive cases are seen (OR 1.84, 95%CI 1.40-2.43).46 Presence of related to chemical cytoreductive therapy with recurrent cytogenetic abnormalities, and inherited thrombophilia can increase the real- (Figure 6).41,44 more interestingly, that the majority of JAK2 tive risk of the development of thrombosis in mutation positive cases develop JAK2 mutation patients <60 years of age with ET and JAK2 negative leukemic blasts when they progress to V617F mutation from 2.23 (95%CI 1.57-3.18) to AML.4,6,38,45 Some of the cytogenetic abnormali- 7.66 (95%CI 2.66-22.03).47 High mutation JAK2 as a therapeutic target ties are suspected to present more in associa- dosage are significantly associated with larger tion with JAK2 mutation to develop a specific spleen size, higher WBC count, higher lactate For decades, the treatment options for phenotype like del 20q in cases with PV.4 dehydogenase (LDH) level and higher risk of patients with MPNs was limited to palliation developing venous thrombosis or cardiovascu- and preventive measures against the develop- Clinical severity and progression lar events in patients with ET or PV. However, ment of thrombosis using aspirin, phlebotomy, these findings are not sufficient for chemical splenectomy, splenic radiation or the use of Thrombosis, bleeding, splenomegaly, bone cytoreduction intended risk stratification steroids, androgens or EPO in patients with marrow failure and evolution to acute which is still dependent on age, prior throm- PMF.1 Since the middle of the 20th century, aims leukemia are the main complications of MPNs. botic event and cardiovascular event-related are widening and touching the areas of pre- In ET, many studies suggested that JAK2 muta- risk factors.2,4,41,48 vention of leukemic transformation and cure tion positive cases have a higher risk of devel- It is still debatable whether an increased with the use of cytoreductive medications like oping venous thrombosis compared to patients JAK2 mutation burden is associated with busulfan, HU and interferon-alpha (INF-α),

[Hematology Reviews 2009; 1:e10] [page 59] Review together with the trials of stem cell transplan- of hematologic neoplasms particularly MPNs. myeloproliferative disorder and acute tation. But the use of these options always car- Its mutations, rearrangements or del are megakaryoblastic leukaemia accompany- ries the risk of leukemogenesis. So, given the important to confirm clonality, and their type ing Down syndrome. Br J Haematol 2008; fact that MPNs are very slow progressing disor- and dosage can help in classifying hematolog- 141:681-8. ders and most of the patients are kept under ic neoplasms. Currently, JAK2 abnormalities 12. Stoiber D, Kovacic B, Schuster C, et al. strict control with non-chemical approaches, are mainly utilized in confirming clonality in TYK2 is a key regulator of the surveillance the use of these options is limited to high-risk diagnosing classical BCR/ABL negative myelo- of B lymphoid tumors. J Clin Invest 2004; patients. proliferative neoplasms; however, we believe 114:1650-8. In the last decade, the use of tyrosine kinase that their role will expand to be added in more 13. Pritchard MA, Baker E, Callen DF, et al. inhibitors is expanding in the management of hematologic neoplasms, to include the type of Two members of the JAK family of protein hematologic neoplasms, especially after the JAK2 abnormality in the sub-classifications tyrosine kinases map to chromosomes successful results of imatinib in CML patients. criteria and, possibly, to consider JAK2 muta- 1p31.3 and 9p24. Mammalian Genome However, the use of imatinib in MPNs is only tion burden in the process of classification and 1992;3:36-8. achieving limited benefits. It is reported that in the assessment of transformation. 14. Saltzman A, Stone M, Franks C, et al. the use of oral imatinib in PV can reduce the The JAK2 V617F positive MPNs are more Cloning and characterization of human need for phlebotomy and spleen size, while closely related and appear with variable clini- Jak-2 kinase: high mRNA expression in parenteral administration can achieve remis- co-pathological phenotypes in response to dif- immune cells and muscle tissue. Biochem. sion in 22% of cases.1,4,49 Currently, there are ferent modifiers. They seem to have a more Biophys. Res. Commun 1998;246:627-33. many trials of new agents on patients with severe clinical course but there is still not suf- 15. Argetsinger LS, Campbell GS, Yang X, et al. PMF, including farnesyl transferase, the auro- ficient available data to include this mutation Identification of JAK2 as a growth hor- ra family of serine/threonine kinases, vascular in the risk stratification. mone receptor-associated tyrosine kinase. endothelial growth factor (VEGF) tyrosine JAK2/STAT pathway is an important target Cell 1993; 74:237–44. kinase, proteasome and fibrogenesis for new therapeutics, but more studies are 16. Witthuhn BA, Quelle FW, Silvennoinen O, inhibitors, VEGF neutralizing antibodies and needed to minimize their toxicity. et al. JAK2 associates with the erythropoi- GX15-070MS, an antagonist of the BH3-bind- etin receptor and is tyrosine phosphorylat- ing groove of the Bcl-2 family.1,4 ed and activated following stimulation Many specific JAK2 inhibitors, like with erythropoietin. Cell 1993;74:227-36. INCB018424, TG101209, TG101348, XL019 and References 17. Campbell GS, Argetsinger LS, Ihle JN, et al. TG10134841, are currently in phase I/II trials Activation of JAK2 tyrosine kinase by pro- on advanced stages of MPNs l such as PMF, 1. Mughal TI, Goldman JM. Chronic myelo- lactin receptors in Nb2 cells and mouse post-ET or PV myelofibrosis. Recent reports proliferative Disorders. 2008; Informa, UK. mammary gland explants. Proc Nat Acad from the trials with INCB018424 are showing 2. Tefferi A. Classification, diagnosis and Sci USA 1994;91:5232-6 significant improvement in splenomegaly, con- management of myeloproliferative disor- 18. Parganas E, Wang D, Stravopodis D, et al. stitutional symptoms, control of myeloprolifer- ders in the JAK2V617F Era. Hematology Jak2 is essential for signaling through a ation and reduction in allele burden. Non-spe- 2006:240-5. variety of cytokine receptors. Cell 1998; cific JAK2 inhibitors are also in clinical trials. 3. Steven HS, Campo E, Harris NL, et al. WHO 93:385-95. These include: CEP-701 (an FLT3 inhibitor), Classification of Tumors of Haematopo- 19. Kralovics R, Passamonti F, Buser AS, et al. tipifarnib (a farnesyltransferase inhibitor), ietic and Lymphoid Tissues. 2008; IARC, A gain-of-function mutation of JAK2 in ITF2357 (an HDAC inhibitor) and hypomethy- Lyon. myeloproliferative disorders. N. Engl J lating agents. The main drawbacks of JAK2 4. Silver RT, Tefferi A. Myeloprolife ra tive dis- Med 2005;352:1779-90. inhibitors are hematologic toxicity like neu- orders: biology and management. 2008; 20. James C, Ugo V, Le Couedic JP, et al. A tropenia and thrombocytopenia, non-hemato- Informa Healthcare, USA. unique clonal JAK2 mutation leading to logic side effects which are mainly immuno- 5. Harmening DM. Clinical Hematology and constitutive signalling causes poly- logical and endocrinological. Beside that, cur- fundamentals of hemostasis. 4th edition, cythaemia vera. Nature 2005;434:1144-8. rently used therapeutic options are well toler- 2002; F. A. Davis. 21. Lacronique V, Boureux A, Della Valle V, et ated, and pegylated INF-α and parenteral ima- 6. Levine RL, Gilliland G. Myeloproliferative al. A TEL-JAK2 fusion protein with consti- tinib are reported to achieve 18-22% remission disorders. Blood 2008; 112:2190-8. tutive kinase activity in human leukemia. rates, respectively. Another unpromising issue 7. Janus, Shorter Oxford English Dictionary. Science 1997;278:1309-12. is the limitation in curing the disease or pre- Vol 1, 3rd edition. Oxford University Press. 22. Peeters P, Raynaud SD, Cools J, et al. venting evolution in the presence of a pre- 1973. Fusion of TEL, the ETS-variant gene 6 JAK2 mutation event and JAK2 mutation nega- 8. Porcu M, Gielen O, Cools J, et al. JAK1 (ETV6), to the receptor-associated kinase tive leukemic transformation. The develop- mutation analysis in T-ALL cell lines. JAK2 as a result of t(9;12) in a lymphoid ment of resistance in vitro is another worrying Haematologica 2009;94:435-7. and t(9;15;12) in a myeloid leukemia. issue.1,4,6,48-50 For the anticipated toxic effects of 9. De Vita S, Mulligan C, Mcelwaine S, et al. Blood 1997;90:2535-40. JAK2 inhibitors, few agents targeting compo- Loss-of-function JAK3 mutations in TMD 23. Malinge S, Ben-Abdelali R, Settegrana C, nent downstream of JAK2 are currently inves- and AMKL of Down syndrome. Br J et al. Novel activating JAK2 mutation in a tigated like those inhibiting BCL-XL or SOCS1 Haematol 2007;137:337-41. patient with Down syndrome and B-cell mimetics.1 10. Kiyoi H, Yamaji S, Kojima S, Naoe T. JAK3 precursor acute lymphoblastic leukemia. mutations occur in acute megakaryoblas- Blood 2007;109:2202-4. tic leukemia both in Down syndrome chil- 24. Kearney L, Gonzalez De Castro D, Yeung J, Conclusions dren and non-Down syndrome adults. et al. Specific JAK2 mutation (JAK2R683) Leukemia 2007; 21:574-6. and multiple gene deletions in Down syn- JAK2 abnormalities are important contribu- 11. Sato T, Toki T, Kanezaki R, et al. Functional drome acute lymphoblastic leukemia. tors but not the sole events in the development analysis of JAK3 mutations in transient Blood 2009;113:646-8

[page 60] [Hematology Reviews 2009; 1:e10] Review

25. Mullighan CG, Zhang J, Harvey RC, et al. bone marrow biopsies of patients with cythemia. Blood 2007;110:840-6. JAK mutations in high-risk childhood chronic myeloproliferative diseases. J Mol 42. Stauffer Larsen T, Pallisgaard N, Boe acute lymphoblastic leukemia. Proc Natl Diagn 2006;8:299-304. Møller M, et al. The JAK2 V617F allele bur- Acad Sci USA 2009;106:9414-8. 34. Lippert E, Girodon F, Hammond E, et al. den in essential thrombocythemia, poly- 26. Elliott J, Suessmuth Y, Scott LM, et al. Concordance of assays designed for the cythemia vera and primary myelofibrosis – SOCS3 tyrosine phosphorylation as a quantification of JAK2V617F: a multicen- impact on disease phenotype. Eur J of potential bio-marker for myeloproliferative ter study. Haematologica 2009;94:38-45. Haematol 2007;79:508-15. neoplasms associated with mutant JAK2 35. Odense HH, et al. Guidelines for the diag- 43. Antonioli E, Guglielmelli P, Poli G, et al. kinases. Haematologica 2009;94:576-80. nosis and treatment of patients with poly- Influence of JAK2V617F allele burden on 27. Fourouclas N, Li J, Gilby DC, et al. cythemia vera, essential thrombocythemia phenotype in essential thrombocythemia. Methylation of the suppressor of cytokine and idiopathic myelofibrosis. Nordic MPD Haematologica 2008; 93:41-8. signaling 3 gene (SOCS3) in myeloprolif- Study Group, 2007. 44. Passamonti F, Rumi E. Clinical relevance erative disorders. Haematologica 2008; 36. McMullin MF, Bareford D, Campbell P, et of JAK2 (V617F) mutant allele burden. 93:1635-44. al. Guidelines for the diagnosis, investiga- Haematologica 2009;94:7-10 28. Landgren O, Goldin LR, Kristinsson SY, et tion and management of polycythaemia 45. Schaub FX, Jäger R, Looser R, et al. Clonal al. Increased risks of polycythemia vera, /erythrocytosis. Br J Haematol 2005;130: analysis of deletions on chromosome 20q essential thrombocythemia, and myelofi- 174-95. and JAK2-V617F in MPD suggests that brosis among 24 577 first-degree relatives 37. Goulding C, Uttenthal B, Foroni L, et al. del20q acts independently and is not one of 11 039 patients with myeloproliferative The JAK2(V617F) tyrosine kinase muta- of the predisposing mutations for JAK2- neoplasms in Sweden. Blood 2008; 112: tion identifies clinically latent myeloprolif- V617F. Blood 2009;113:2022-7. 2199-204. erative disorders in patients presenting 46. Ziakas PD. Effect of JAK2 V617F on throm- 29. Olcaydu D, Harutyunyan A, Jäger R, et al. A with hepatic or portal vein thrombosis. Int botic risk in patients with essential throm- common JAK2 haplotype confers suscepti- J Lab Hematol 2008;30:415-9. bocythemia: measuring the uncertain. bility to myeloproliferative neoplasms. 38. Chloé J. The JAK2V617F mutation in poly- Nature Genetics 2009; 41:450-4. cythemia Vera and other myeloprolifera- Haematologica 2008;93:1412-4. 30. Conchon MR, Costa JL, Novaes MM, et al. tive disorders: one mutation for three dis- 47. De Stefano V, Za T, Rossi E, et al. Influence Simultaneous detection of JAK2 V617F eases? Hematology 2008;69-75. of the JAK2 V617F mutation and inherited mutation and Bcr-Abl translocation in a 39. Sozer S, Fiel MI, Schiano T, et al. The pres- thrombophilia on the thrombotic risk patient with chronic myelogenous ence of JAK2V617F mutation in the liver among patients with essential thrombo- leukemia. Int J Hematol 2008;88:243-5. endothelial cells of patients with Budd- cythemia. Haematologica 2009;94:733-7. 31. McLornan D, Percy M, McMullin MF. JAK2 Chiari syndrome. Blood 2009;113:5246-9. 48. Vannucchi AM, Guglielmelli P, Pieri L, et V617F: a single mutation in the myelopro- 40. Spivak JL, Silver RT. The revised World al. Treatment options for essential throm- liferative group of disorders. Ulster Med J Health Organization diagnostic criteria for bocythemia and polycythemia vera. Expert 2006;75:112-9. polycythemia vera, essential thrombocyto- Rev Hematol 2009;2:41-55. 32. Hellström-Lindberg E, Cazzola M. The role sis, and primary myelofibrosis: an alterna- 49. Rambaldi A, Barbui T, Barosi G. From pal- of JAK2 mutations in RARS and other tive proposal. Blood 2008;112:231-9. liation to epigenetic therapy in myelofibro- MDS. Hematology 2008;52-9. 41. Vannucchi AM, Antonioli E, Guglielmelli P, sis. Hematology 2008: 83-91 33. Horn T, Kremer M, Dechow T, et al. et al. Clinical profile of homozygous JAK2 50. Levine RL, Heaney M. New advances in the Detection of the activating JAK2 V617F 617V>F mutation in patients with poly- pathogenesis and therapy of essential mutation in paraffin-embedded trephine cythemia vera or essential thrombo- thrombocythemia. Hematology 2008:76-82.

[Hematology Reviews 2009; 1:e10] [page 61] Hematology Reviews 2009; volume 1:e11

Up-front fludarabine impairs and “feeding” post-treatment relapse and disease progression.4-5 The clinical impact of Correspondence: Hans E Johnsen, stem cell harvest in multiple these cells needs to be evaluated by in vivo Department of Haematology, Medical Center, myeloma: report from an targeted therapy. Therefore, we initiated a Aalborg Hospital Science and Innovation Center, interim analysis of the NMSG randomized phase II multicenter trial compar- AHSIC - Sdr Skovvej 15, Aarhus University 13/03 randomized placebo ing induction therapy by cyclophosphamide Hospital, 9000 Aalborg, Denmark plus dexamethasone with and without fludara- E-mail: [email protected] controlled phase II trial bine, a DNA repair inhibitor. Fludarabine, 9- - β Key words: multiple myeloma, clinical trial, flu- D-arabinofuranosyl-2-fluoroadenine, is an darabine. Hans E. Johnsen,1 Lene M. Knudsen,2 analog to adenosine cytotoxic against dividing 3 3 6-7 Anne K. Mylin, Peter Gimsing, and resting cells. In vivo, the combination of Acknowledgments: first, we would like to Henrik Gregersen,1 Niels Abildgaard,2 a DNA damaging agent, e.g. adriamycine or acknowledge the Pharmacy Copenhagen County Nils Frost Andersen,4 cyclophosphamide combined with fludarabine for assistance in preparing the placebo and Torben Plesner,5 Annette Vangsted,6 is clinically active against B cells in CLL and blinding procedures. Second the GCP Unit at Torben Mourits-Andersen7 on behalf of low-grade follicular lymphomas.8-10 Although it Aarhus University Hospital for monitoring the the Nordic Myeloma Study Group has been documented active against leukemia study. This study was supported by unrestricted and lymphoma, only recent data has suggested research grants from Schering Nordic AB and 1Aalborg Hospital University of Aarhus, efficacy in MM. In a previous open phase II Amgen AB to the Nordic Myeloma Study Group 2Department of Hematology, Odense and in part by individual grants to HEJ from the pilot study, we have documented that addition 3 Nordic Cancer Union (grant no 56-9350/56-9351 University Hospital; Rigshospitalet, of fludarabine to induction therapy is clinical- 4Aarhus Hospital, University of Aarhus; and 56 100 02-9101/9102 95), the Danish Cancer ly feasible with only minor toxicity. A benefi- Society (grant n. 945 100-15) the Danish 5Vejle Hospital; 6Herlev University cial clinical outcome was suggested including Research Agency (grant n. 99 00 771) and EU 6th 7 Hospital, Esbjerg Hospital, Denmark a reduction of minimal residual disease FP (grant n. LSHC-CT-2006-037602 to MSCNET (MRD) following the addition of fludarabine.11 coordinator HEJ). However, one concern during the trial design discussion was the adverse impact of fludara- Received for publication: 12 March 2009. Abstract bine on stem cell harvest experienced in Revision received: 5 July 2009. advanced CLL,13 which, however, was not con- Accepted for publication: 21 July 2009. The impact of chemotherapy resistant B sidered in untreated patient treatment. This work is licensed under a Creative Commons cells in multiple myeloma (MM) needs to be The main objective of the subsequent Attribution 3.0 License (by-nc 3.0) evaluated by in vivo targeted therapy. Here we NMSG n.13/03 phase II trial was to generate report the conclusions from a phase II ran- data on toxicity, safety and efficacy by adding ©Copyright H.E. Johnsen et al., 2009 domized, placebo controlled trial adding flu- fludarabine to standard induction therapy.12 In Licensee PAGEPress, Italy darabine to the induction with cyclophos- the close follow-up of the patient cohort it was phamide-dexamethasone. Based on an inter- Hematology Reviews 2009; 1:e11 decided to perform an interim analysis, which doi:10.4081/hr.2009.e11 im toxicity and safety analysis, the trial was concluded that fludarabine in the experimen- stopped following inclusion of 34 of a planned tal arm inhibits stem cell mobilization capaci- 80 patients due to a reduced number of ty and reduced the number of patients reach- patients (4/17) actually harvested in the ing high-dose therapy and the trial was Trial design experimental arm compared to the control arm stopped. Consequently, fludarabine in combi- This was a randomized, placebo controlled, (11/17; p<0.05). In conclusion, the scheduled nation with alkylating agents should not be single blinded, phase II study evaluating toxi- fludarabine dosage in 2 cycles combined with administrated as up-front therapy, if high- city and safety of fludarabine added to alkylating therapy impairs stem cell mobiliza- dose therapy supported by autologous trans- cyclophosphamide and dexamethasone tion and standard therapy in young MM plantation is standard care. patients and should not be administrated up- (CyDex) as induction therapy in younger front. newly diagnosed symptomatic multiple myeloma requiring therapy. The treatment regimen Materials and Methods CyDex as standard induction therapy was documented in NMSG trial n.11/01.12 Patients Introduction Approval and patient eligibility were randomized at diagnosis either to The scientific protocols were reviewed and CyDex + placebo (control Arm A) or CyDex + Due to a range of new drugs there has been approved by the regional ethics committees in fludarabine (experimental Arm B). a continuous progress in the treatment of Denmark and the Danish Drug Agency multiple myeloma (MM). However, only a few (Sagsnr. KA 03103 ms) and all patients gave Treatment procedure patients are considered cured so far, most written informed consent before study entry. Fludarabine was considered as the only likely due to the nature of the disease.1-3 All patients were over 18 years of age and were investigational drug in this study adminis- Recent data have indicated that the myeloma referred to the departments for diagnostic trated in induction phase I. cell hierarchy includes resistant circulating evaluation. Patients under 60 years of age who clonal memory B cells, which differ consider- had Durie-Salmon stage I with at least one Phase I ably from the classical end stage plasma cells, bone lesion, II, or III myeloma were eligible. Arm A (conventional arm): CyDex + place- infiltrating the bone marrow. The pathophysi- The criteria for exclusion were prior treatment bo, two (three) cycles in Phase I: two courses ological significance of these cells is for myeloma, another cancer, abnormal cardiac of CyDex: cyclophosphamide 1000 mg/m2 IV unknown, but hypothetically they may serve as function, chronic respiratory disease, abnor- day 1 and dexamethasone 40 mg/day PO on “sleeping” myeloma stem cells responsible for mal liver function or psychiatric disease. day 1-4, and 9-12 + placebo PO; repeated once

[page 62] [Hematology Reviews 2009; 1:e11] Article day 21. The third cycle of CyDex (without circulating myeloma compartments identified placebo) was only given if the phase II treat- Results and Discussion a significant reduction of CD19+ B cells and ment could not be initiated within six weeks myeloma plasma cells in the fludarabine-arm, after the start of CyDex II. Other steroids in Treatment cycles given during concluding that fludarabine therapy in MM equipotent dose could be used instead of dex- induction phase I was feasible with a potential clinical efficacy. amethasone.12 All patients in the conventional arm However, in the current trial, unexpected side Arm B (experimental arm): CyDex plus flu- received the scheduled cycles of therapy. effects were initially observed which may darabine, two (three) cycles in Phase I: two However, in the experimental arm this was explain the drop-out of the 6 patients (Table 1) courses of CyDex: cyclophosphamide 1000 only the case for 11/17 patients as 6 patients before the end of induction therapy. mg/m2 day 1 IV and dexamethasone 40 mg/d were stopped before or following the first cycle (or other steroids in equipotent dose) PO on of therapy (Table 1). Three of these patients Toxicity and adverse events follow- days 1-4, and 9-12, combined with fludarabine did not start therapy, suggesting bias from the ing induction phase I 40 mg/m2 PO day 1-3 each cycle; repeated clinician, for whom the therapy arm was not In accordance with the CTC criteria, no dif- once day 21. The third cycle of CyDex (with- blinded. Such a bias may be the result from the ference in severe toxicity was found. However, out fludarabine) was only given if the phase II relative intensive dosage of fludarabine analysis of laboratory quantities following the 2 treatment could not be initiated within six administrated in this trial of 40 mg/m PO for second treatment showed a borderline reduc- weeks after the start of the second CyDex plus three days in each of two cycles, compared to tion of blood lymphocytes from mean 1.12 (SD 2 fludarabine course. our previous trial where we administered a 0.4) to 0.73 (SD 0.6; p=0.055) and an dose of 25 mg/m2 intravenously for five days. increased plasma creatinine level from mean Common trunk (phases II–IV) Following discussions within NMSG the proto- 57.8 (SD 14.2) to mean 124.2 (SD 28.8; This was as described in previous reports col group selected oral administration over p=0.035). All other variables registered from NMSG.3,12 In brief, the priming and three days attempting to reach an equivalent showed no difference including performance apheresis phase II included cyclophos- total dose of fludarabine as used previously. score. Many clinical trials in CLL have shown phamide 2 g/m2 given as a single dose intra- Our previous experience2 administrating a the combination of fludarabine and cyclophos- venously during 60 minutes. Uroprotection total dosage of 125 mg/m2 fludarabine over five phamide to have tolerable toxicity,13-15 however, with Mesna 160% of the cyclophosphamide days was moderate neutropenia, no thrombo- the observed significant reduction in renal dose divided in four doses (before 3, 6 and 9 cytopenia or severe infectious episodes. We function may be due to latent myeloma specif- h after start of cyclophosphamide) and diure- observed all 9 of the fludarabine-treated ic kidney impairment. All serious adverse sis of at least 2.5 L/m2 the following 24 hours. patients responding to treatment with 2 com- events were reported to The Trial Secretariat Granulocyte colony-stimulating factor (G- plete remissions and 7 partial remissions, within one working day of discovery or notifi- CSF) was initiated day 4 after cyclophos- compared to 5 responders (all PR) in the con- cation of the event. Initial serious adverse phamide as Neupo-gen® 5-10 ug/kg daily trol-arm. Furthermore, the effect on the blood event information and all amendments or addi- adjusted to appropriate vial size. Peripheral blood stem cell leukapheresis were performed during mobilization, guided by CD34 blood Table 1. Number of treatment cycles during induction therapy (phase I). 6 levels, by harvest of a minimum of 2×10 Variable Placebo (n) % Intervention (n) % p CD34+ cells per kilogram body weight. Following harvest of a sufficient graft, the No treatment (0) 0.0 (3) 17.6 0.054 patients passed to phase III: high-dose thera- (Mann Whitney test) py with melphalan 200 mg/m2 given as a sin- 1 treatment (0) 0.0 (3) 17.6 gle dose intravenously, followed by stem cell 2 treatments (15) 88.2 (9) 52.9 infusion 48 hours later, and G-CSF (Neupo- 3 treatments (2) 11.8 (2) 11.8 gen® 5 µg/kg daily or Neulasta® 12 mg) one injection from day 4 after graft reinfusion, until the absolute neutrophil count is more than 1.0×109/L for three consecutive days. Table 2. Fraction of patients subjected to harvest of an autograft (phase II). The patients were followed as outpatients Variable Placebo (n) % Intervention (n) % p during phase IV. Number of patients subjected (11) 64.7 % (4) 23.5 % 0.037 Statistical analysis to at least one harvest + The proportions of patients with a given Mean total harvest of CD34 (11) 736±465 (4) 416±248 0.12 6 characteristic were compared using Fisher’s cells ×10 ± SD + exact test for binary data. The distributions of Mean average number of CD34 (11) 584±501 (4) 251.2 (152.6) 0.071 cells 106 per harvest ± SD continuous quantities were examined to con- × trol that they followed Gaussian distributions Number of apheresis to collect Placebo (n) Intervention (n) p with good approximation using Kolmogorow- >5×106 CD 34+ cells/kg Smirnow's test supplemented by Q-Q plots. If they did, either directly or following No harvest session/mobilization failure (6) 35.3% (13) 76.5% 0.11 transformation (square root or logarithmic), 1 harvest session (7) 41.2% (02) 11.8% the mean values of the two groups were com- 2 harvest sessions (2) 11.8% (01) 05.9% pared using a t-test. If not the two groups 3 harvest sessions (2) 11.8% (01) 05.9% were compared using Mann Whitney's test.

[Hematology Reviews 2009; 1:e11] [page 63] Article tions were recorded on the Adverse Event received a full dose of fludarabine. This was 4. Rasmussen T, Jensen L, Johnsen HE. Levels Form. This was reviewed and documented that reported and reviewed by the Danish Drug of circulating CD19+ cells in patients with CMV-reactivating was seen in one patient in Agency. Follow-up in December 2008 revealed multiple myeloma. Blood 2000; 95:4020-1. the standard arm and 3 patients in the experi- that 12/13 and 9/13 patients had responded to 5. Rasmussen T. The presence of circulating mental arm. therapy with 4/13 and 3/13 obtaining CR at any clonal CD19+ cells in multiple myeloma. time during follow-up. The number of patients Leuk Lymphoma 2001; 42:1359-66. Priming for stem cell mobilization dying from complications or progressive dis- 6. Gandhi V, Plunkett W. Cellular and clinical and harvest during phase II ease was 2/17 and 5/16, respectively, with no pharmacology of Fludarabine. Clin Fifteen of 17 patients and 12/17 were primed significant difference between the two arms. Pharmacokinet 2002; 41:93-103. with standard care cyclophosphamide and rhG- 7. Johnson SA, Thomas W. Therapeutic poten- CSF in arms A and B, respectively. Successful tial of purine analogue combinations in the mobilization to reach the level of >10 CD34+ treatment of lymphoid malignancies. cells per microliter blood triggered leukaphere- Conclusions Hematol Oncol 2000; 18:141-53. sis and was obtained in 11/17 patients in arm 8. Flinn IW, Byrd JC, Morrison C, et al. A but only in 4/17 patients in arm B (Table 2). In conclusion, the scheduled fludarabine Fludarabine and cyclophosphamide with fil- This difference was significant at the interim dosage in two cycles combined with alkylating grastim support in patients with previously analysis performed by an independent group of therapy impairs stem cell mobilization and untreated indolent lymphoid malignancies. experts and the decision was taken to stop the standard therapy in young MM patients and Blood 2000;96:71-5 trial. Comparison of the total and average should not be administrated up-front. This 9. Eucker J, Schille C, Schmid P, et al. The number of CD34+ cells harvested did not reveal observation is in accordance with recent data combination of fludarabine and cyclophos- differences in patients actually undergoing from up-front therapy in CLL. phamide results in a high remission rate apheresis. The situation concerning published We are now left with the challenge of under- with moderate toxicity in low-grade non- data about the adverse impact of fludarabine standing the mechanisms of fludarabine Hodgkin's lymphomas. Anticancer Drugs on stem cell harvest is now clearer. In a survey responsible for the negative side effect on 2002;13:907-13. of advanced CLL from 122 centers of the mobilization of normal hematopoietic progeni- 10. Kano Y, Akutsu M, Tsunoda S, et al. In vitro European Group of Blood and Marrow tors, as well as the potential therapeutic effect cytotoxic effects of Fludarabine (2-F-ara-A) Transplantation (EBMT), it was concluded that on marrow and blood B cells in MM and other in combination with commonly used attention should be given to the timing of B-cell malignancies. This is of special interest antileukemic agents by isobologram analy- mobilization with respect to the last dose of as the myeloma cell hierarchy includes resist- sis. Leukemia 2000;14:379-88. fludarabine.13 This has been further supported ant circulating clonal memory B cells, which 11. Bjorkstrand B, Rasmussen T, Remes K, et by a study of B-CLL after front-line treatment differ considerably from the classical end stage al. Feasibility of Fludarabine added to VAD with fludarabine (30 mg/m2 per day) and plasma cells, infiltrating the bone marrow. The during induction therapy in multiple myelo- cyclophosphamide (200 mg/m2 per day) both pathophysiological significance of these cells ma: a randomised phase II-study. Eur J given orally for five consecutive days in six is at present unknown, but hypothetically they Haematol 2003;70:379-83. monthly courses. After evaluation performed may serve as “sleeping” myeloma stem cells 12. Mellqvist UH, Lenhoff S, Johnsen HE, et al. two months after the last course, responding responsible for and “feeding” post-treatment for the Nordic Myeloma Study Group. patients were considered for PBPC collection. relapse and disease progression,4-5 as studied Cyclophosphamide plus dexamethasone is Following conventional rhG-CSF, priming until by the Myeloma Stem Cell Network supported an efficient initial treatment before high- adequate blood CD34 circulation was achieved by the 6th FP from the EU. dose melphalan and autologous stem cell and a harvest procedure was initiated success- transplantation in patients with newly diag- fully in only 12 of the 32 CLL patients.14 nosed multiple myeloma: results of a ran- The present report supports this observation domized comparison with vincristine, dox- and further adds to our knowledge that as few References orubicin, and dexamethasone. Cancer as one to two series of fludarabine may result 2008;112:129-35. in poor mobilization and impair standard ther- 1. Attal M, Harousseau JL, Stoppa AM, et al. 13. Michallet M, Thiébaut A, Dreger P, et al. apy. Recently, the stem cell toxicity has been A prospective, randomized trial of autolo- Peripheral blood stem cell (PBSC) mobi- supported by the observation that the risk for gous bone marrow transplantation and lization and transplantation after fludara- sMDS/AML was correlated to the use of flu- chemotherapy in multiple myeloma. bine therapy in chronic lymphocytic darabine based on an unknown mechanism.15 Intergroupe Francais du Myelome. N Engl J leukaemia (CLL): a report of the European Med 1996;335:91-7. Blood and Marrow Transplantation (EBMT) Response evaluation following 2. Bjorkstrand BB, Ljungman P, Svensson H, CLL subcommittee on behalf of the EBMT phase II et al. Allogeneic bone marrow transplanta- Chronic Leukaemias Working Party Response evaluation performed following tion versus autologous stem cell transplan- (CLWP). Br J Haematol 2000;108: 595-601. phase I-II prior to high-dose therapy revealed tation in multiple myeloma: a retrospective 14. Tournilhac O, Cazin B, Leprètre S, et al. 7/13 and 5/7 patients obtained a partial remis- case-matched study from the European Impact of frontline fludarabine and sion. The trend towards more patients achiev- Group for Blood and Marrow Transplant- cyclophosphamide combined treatment on ing CR in the intervention arm observed in our ation Blood 1996;88:4711-8. peripheral blood stem cell mobilization in previous phase II study could not be confirmed 3. Lenhoff S, Hjorth M, Holmberg E, et al. B-cell chronic lymphocytic leukemia. Blood in this trial (results not shown). There was no Impact on survival of high-dose therapy 2004;103:363-5. difference in graft quality, evaluated by time to with autologous stem cell support in 15. Laurenti L, Tarnani M, Chiusolo P, et al. Low neutrophil and platelet recovery. One patient patients younger than 60 years with newly incidence of secondary neoplasia after died from treatment complication in the exper- diagnosed multiple myeloma: a population- autotransplantation for lymphoproliferative imental arm due to protocol violation: the based study. Nordic Myeloma Study Group. disease: the role of pre-transplant therapy. patient had impaired renal function and Blood 2000; 95:7-11. Clin Transplant 2008;22:191-9.

[page 64] [Hematology Reviews 2009; 1:e11] HematologyHematology Reviews Reviews 2009; 2009; volume volume 1:exxxx 1:e12

Maintenance therapy ing prolonged remissions with α-interferon maintenance in patients responding to con- Correspondence: Jean-Luc Harousseau, Rene in multiple myeloma ventional induction therapy,5 a number of ran- Gauducheau Cancer Center, Bd Jacques Monod domized trials were performed but their 44805, Nantes/St Herblain Cedex, France Jean-Luc Harousseau results were controversial.6 Two meta-analyses E-mail: [email protected] Rene Gauducheau Cancer Center, of randomized trials showed that with interfer- Received for publication: 23 July 2009. Nantes/St. Herblain Cedex, France on maintenance, time to PFS and OS was Accepted for publication: 1 August 2009. increased by four to seven months.7,8 However, most investigators considered that the benefit This work is licensed under a Creative Commons was small and needed balancing against cost Attribution 3.0 License (by-nc 3.0) Introduction and potential toxicity of prolonged treatment with α-interferon. ©Copyright J.-L. Harousseau 2009 Licensee PAGEPress, Italy The treatment of multiple myeloma (MM) In addition, α-interferon has been used has changed dramatically in the past twenty after ASCT, with the hypothesis that it might Hematology Reviews 2009; 1:e12 years with the introduction of high-dose thera- be more effective in patients with minimal doi:10.4081/hr.2009.e12 py plus autologous stem-cell transplantation residual disease. In a retrospective analysis of (ASCT) in younger patients and, more recent- the European Bone Marrow and Blood ly, of three novel agents (thalidomide, borte- Transplant Registry, interferon maintenance zomib, and lenalidomide). was associated with improved PFS and OS in Spencer et al. (2009), respectively,12,15 thalido- 9 When conventional chemotherapy was the patients responding to high-dose therapy. mide was not used during induction treatment. only available possibility, complete responses However, two randomized trials failed to con- In the British study by Morgan et al. (2008), 10,11 (CR) were very rare and the objective of main- firm this result. there was an initial randomization at diagno- tenance was to prolong remission duration by sis (thalidomide versus no thalidomide),16 and continuing the same type of treatment that in the Tunisian study by Abdelkefi et al. induced the initial response. With recent ther- (2008), all patients received thalidomide plus apeutic improvements, CR achievement Thalidomide as maintenance dexamethasone as induction treatment prior becomes a realistic goal that, in most cases, is therapy after autologous to ASCT.14 significantly correlated with the outcome.1 stem-cell transplantation In four of these six trials, thalidomide was Therefore, both the nature and the impact of prescribed until relapse or until severe adverse maintenance therapy have changed. Mainten - event, while in the studies by Abdelkefi et al. When thalidomide maintenance therapy ance therapy is based currently on novel (2008) and Spencer et al. (2009), the duration was first introduced in the field of ASCT for agents, and its objective is not only to control of thalidomide was fixed (six months in the MM, the median duration of response did not the clone but also to further decrease the Tunisian study and 12 months in the exceed three years and almost all patients did Australian study).14,15 The daily dose of thalido- tumor burden and improve the quality of 12 relapse ultimately. Six randomized trials have 14 response. mide varied from 100 mg/day to an initial evaluated the impact of thalidomide mainte- dose of 400 mg/day in the first two of the six A number of randomized studies show a 12-17 nance after ASCT in MM. However, these 12,13 benefit from maintenance therapy with novel studies. Despite these disparities, all six trials differed in their design and in the dose agents (until now, mostly thalidomide), at studies showed a benefit in favor of thalido- and duration of thalidomide maintenance least in terms of response rate and progres- mide in terms of response rate (CR, or greater (Table 1). sion-free survival (PFS). However, there is still than or equal to very good partial remission In two of these studies, patients were ran- a debate as concerns the impact on overall sur- and PFS) (Table 2). domized initially to receive thalidomide vival (OS) and the optimal administration of Results are not that clear-cut as regards OS. throughout their treatment (both before and maintenance therapy. While in the initial publication of the French after ASCT).13,17 Therefore, the putative impact and Tunisian studies12,14 OS was significantly of thalidomide was not expected to be a result longer in the thalidomide trials, with longer of the maintenance effect only. follow-up this survival advantage disappeared, Maintenance therapy in the In the other four studies, patients were ran- and Aldelkefi et al. (2009) did publish a retrac- domized after ASCT to receive either thalido- tion recently.18 pre-thalidomide era mide or not. While in two of these trials the On the contrary, the first publication from control group was to receive no further treat- the Arkansas group showed no significant dif- In patients initially treated with convention- ment,12,16 in the Tunisian study by Abdelkefi et ference in OS between the two groups, al chemotherapy, maintenance chemotherapy al. (2008), thalidomide maintenance was com- because of a shorter OS after relapse in has failed to demonstrate any significant ben- pared to a second ASCT,14 and in the Australian patients initially treated with thalidomide.11 efit.2,3 Therefore, the recommendation was to study by Spencer et al. (2009) the combination However, with a longer follow-up, the OS stop chemotherapy once stable response had of thalidomide plus alternate-day prednisone curves diverge after five years.19 A second been achieved (the so-called plateau phase). was compared to the administration of pred- report of the same trial, showed a trend in One study showed a survival advantage for nisone alone.15 favor of thalidomide and a significant benefit patients receiving alternate-day prednisone In two of these trials, patients were random- in the subgroup of patients with karyotypic maintenance therapy at a dose of 50 mg when ized only if they were not progressing after abnormalities.13 In the Dutch (Lokhorst et al., compared with 10 mg after standard induction induction treatment plus ASCT,12,15 while in the 2008) and British ) and British (Morgan et al., chemotherapy.4 other two studies, all patients were random- 2008) studies, the PFS benefit did not translate In the nineteen-eighties α-interferon main- ized, whatever the degree of post-ASCT into a significant OS benefit, again because of tenance therapy represented a great hope. response.14,16 Finally, in the French and a shorter OS after relapse in the thalidomide Following the initial randomized study show- Australian studies by Attal et al. (2006) and group.16,17

[Hematology Reviews 2009; 1:e12] [page 65] Review

Table 1. Post-ASCT maintenance with thalidomide: randomized studies. Author Induction ASCT Thalidomide administration Design Attal et al.12 No Thal Double Starting dose 400 mg/d until relapse After ASCT (if no progression) No treatment vs. Thal + Pamidronate Barlogie et al.13 50% Thal Double Starting dose 400 mg/d until relapse Initial randomization Thal vs. no Thal Abdelkefi et al.14 Thal Single + Thal vs. double 100 mg/d for 6 months After ASCT Spencer et al.15 No Thal Single 20 mg/d for 1 year After ASCT (if no progression) Morgan et al.16 50% Thal Single 100 mg/d until relapse After ASCT Lokhorst et al.17 50% Thal Double 200 mg/d until relapse Initial randomization TAD → Thal VAD → IFN ASCT, autologous stem cell transplantation; Thal, thalidomide; d, day; TAD, thalidomide adriamycin dexametasone; VAD, vincristine adriamycin dexamethasone; IFN, interferon.

Table 2. Results of randomized studies on post-ASCT maintenance with thalidomide. Author N. of Median Response PFS or EFS OS initial OS updated patients follow-up rate publications results Barlogie et al.13 668 42 m 62% vs. 43% 5-year EFS 5-year OS p=0.09 p<0.001 56% vs. 44% 65% vs. 65% p=0.01 Attal et al12 597 40 m *67% vs. 55% or 57% 3-year EFS 4-year NS p=0.03 52% vs. 37% 87% vs. 75% p=0.04 p=0.002 Abdelkefi et al.14 195 33 m *68% vs. 54% 3-year PFS 3-year NS p=0.04 85% vs. 57% 85% vs. 65% p=0.04 p=0.02 Spencer et al.15 269 3 y 65% vs. 44%* 3-year PFS 3-year ND p<0.001 42% vs. 23% 86% vs. 75% p<0.001 p<0.004 Morgan et al.16 820** 32 m NA Better in patients NS ND with < VGPR Median OS p<0.007 Lokhorst et al.17 556 52 m *66% vs. 54% Median EFS 73 m vs. 60 m ND p=0.005 34 m vs. 22 m p=0.77 p<0.001 *, complete response; +, very good partial response (VGPR); **, including patients receiving non-intensive induction; ASCT, autologous stem cell transplantation; PFS, progression-free survival; EFS, event-free survival; OS, overall survival; NS, not significant; m, months; y, years; ND, not done; NA, not available.

How can we analyze these and conventional chemotherapy, ASCT was ical trials. Therefore there were more thera- differences? superior in terms of PFS but not always in peutic possibilities (including thalidomide) at The first message is that OS data should not terms of OS, partly because patients in the relapse in the no-thalidomide group. be analyzed and published too early. In the past conventional chemotherapy group could These considerations clearly raise the issue the only possibility at relapse was convention- receive ASCT at relapse.20 A randomized study of the optimal duration of maintenance thera- al chemotherapy. Currently we have more pos- comparing early versus late ASCT showed no py. Should it be fixed as in the Tunisian or sibilities (e.g., ASCT, thalidomide and lenalido- significant difference in terms of OS.21 Australian studies,14,15 or unlimited as in the mide, bortezomib), and survival after relapse Moreover, the risk of thalidomide-induced other four studies?12,13,16,17 The theoretical inter- may be longer than duration of first response. peripheral neuropathy is clearly related to the est of prolonged maintenance is to further Since it is possible to achieve median OS of cumulative dose, and prolonged exposure car- improve the level of tumor burden reduction, more than or equal to five years now, OS ries the risk of severe neuropathy, precluding hence prolonging PFS but, on the other hand, should not be analyzed before at least five-year or limiting the use of bortezomib. Prolonged this benefit might be hampered by reduced sal- median follow-up time. exposure to thalidomide might select clones vage possibilities, hence a shorter OS after Secondly, OS obviously depends on salvage resistant not only to thalidomide but also to relapse. A randomized study addressing this treatments. If thalidomide is prescribed until other agents. We already know that lenalido- question might be extremely useful. relapse or severe toxicity, one can imagine that mide is less effective in patients resistant to thalidomide should not be used at relapse. thalidomide than in thalidomide-naïve Additional questions Therefore, more patients receive thalidomide patients.22 at relapse in the no-thalidomide trials, and the Finally, salvage treatment depends on the What is the optimal schedule of design of these studies is rather early (up availability of novel agents. When the British administration? front) thalidomide versus late (at relapse) trial by Morgan et al. (2008) was performed, In the first two studies by Attal et al. (2006) thalidomide. This was actually the case with there was a limited access to bortezomib, and and Barlogie et al. (2006), the initial daily ASCT: in randomized trials comparing ASCT lenalidomide was not available except for clin- dosage was high (400 mg) and the duration of

[page 66] [Hematology Reviews 2009; 1:e12] Type ofReview paper treatment was unlimited, which explains the This important question should be proposed. However, relapses remain frequent, high incidence of severe neuropathy (Grade addressed in future trials. To date, Paiva et al. especially in the absence of chronic-graft ver- ≥3), as well as the percentage of patients who (2008) have shown that immunophenotypic sus host disease.26 Therefore, post-transplant discontinued treatment because of toxicity.12,13 CR, as assessed by multi-parameter flow immunotherapy with donor-lymphocyte infu- Although no randomized study has compared cytometry, is associated with a better outcome sions and/or novel agents has been tested with different schedules of administration, doses of than CR as defined only by immunofixation.24 the objective of upgrading the level of 100 mg or 200 mg/day during 6-12 months response. In a preliminary experience, Kroger appear to be effective and better tolerated.13,14 et al. (2009) have proposed novel agents (thalidomide, bortezomib, or lenalidomide) to Is thalidomide maintenance useful for all Other novel agents as mainte- patients who were not in CR after allo-SCT and patients, and are we able to predict donor lymphocyte infusions.28 They could con- patients who will benefit from maintenance nance therapy after autologous vert partial remission to CR in 59% of patients therapy? stem-cell transplantation and to molecular remissions in 50% of Unfortunately there is no clear response to patients. this question. In the French study by Barlogie Lenalidomide, which is better tolerated than et al. (2006), patients with del(13) apparently thalidomide and can be prescribed safely for did not benefit from thalidomide maintenance, long periods of time, appears to be an ideal but at the time of this trial other abnormalities candidate for maintenance therapy. However, Thalidomide maintenance that are frequently associated with del(13), this agent is more myelotoxic than thalido- therapy after non-intensive such as t(4;14) or 17p del, were not routinely mide and the optimal dose of lenalidomide 12 studied. We know now that the negative prog- after high-dose therapy is not known. Two induction treatment nostic impact of del(13) is mostly a result of large randomized trials from the Intergroupe 23 these two additional abnormalities. We have Francophone du Myelome (IFM ) and the In a recent trial Ludwig et al. (2009) evalu- no published data on the impact of thalidomide Cancer and Acute Leukemia Group B (CALGB) ated thalidomide plus interferon compared in this subgroup of patients with these poor- groups have tested lenalidomide as mainte- with interferon alone in 135 elderly patients risk abnormalities, although in a preliminary nance after ASCT, but results of these studies with at least stable disease after induction report (Lokhorst et al., 2008), thalidomide are not available. treatment with either thalidomide dexametha- appeared to perform poorly in patients with In addition, bortezomib has been evaluated sone or melphalan-prednisone (MP). Athough del(17p).17 in this setting by Morgan et al. (2008) and PFS was significantly longer in the thalido- In the updated analysis of the Arkansas Paiva et al. (2008).16,24 Since bortezomib is mide group (24 months versus 13 months, study by Adelkefi et al. (2009), thalidomide sig- 29 associated with a high incidence of peripheral p>0.024), OS was similar in both groups. nificantly improved OS of patients with cytoge- neuropathy when used on a bi-weekly schedule Five randomized trials have compared MP netic abnormalities as defined by conventional 2 and MP plus thalidomide (MPT) as the primary 18 at a dose of 1.3 mg/m , the issue of the toxici- karyotyping. This heterogeneous subgroup of 30-35 ty/efficacy ratio is crucial. treatment in elderly patients. Although the patients generally is considered as poor-risk, If the objective of post-ASCT therapy is to design of these studies and the inclusion crite- since the possibility of studying mitoses is increase the level of response with a consolida- ria were slightly different, all five studies have associated with a more proliferative disease. tion effect further, short-term treatment with shown a benefit of MPT in terms of response Finally, in the studies by Attal et al. (2006) combinations of novel agents might be attrac- rate, and PFS was significantly longer in the and Morgan et al. (2008), only patients who did tive as well. Ladetto et al. (2008) recently MPT groups of four out of five studies (Table 3). not achieve at least a very good partial showed encouraging results with four courses However, in only two studies this benefit response (VGPR) after transplant benefitted of consolidation treatment with bortezomib- translated into a significantly longer sur- from thalidomide maintenance,12,16 but this was thalidomide-dexamethasone.25 In this study, vival,32,33 and in these two studies there was no not confirmed by Spencer et al. (2009).15 six out of 24 patients, who were at least in maintenance while in the other three trials VGPR after ASCT, achieved molecular remis- there was a maintenance with thalidomide Does thalidomide act as a maintenance or sions and none of them had a relapse with a alone in the MPT groups. In the Italian study by consolidation therapy? median follow-up of 26 months. Palumbo et al. (2008), the shorter survival In all the studies reviewed, the PFS prolon- after relapse in the MPT group might have gation was associated with a CR or CR/VGPR been explained by a lower percentage of increase. Moreover, the fact that patients who patients receiving thalidomide as salvage showed CR after ASCT did not benefit from Novel agents as maintenance treatment at relapse.31 Since the three studies thalidomide in at least two of the studies therapy after allogeneic were not designed to address the question of could mean that thalidomide might act more maintenance, it is not possible to consider that by increasing the post-ASCT CR rate than by stem-cell transplantation the lack of survival benefit in the MPT group is controlling the residual clone. In other words, related to maintenance thalidomide. However, post-ASCT thalidomide might be considered Currently, allogeneic stem-cell transplanta- one can conclude from these studies that there as a consolidation therapy, and might be tion (allo-SCT) following a myeloablative con- is no evidence that maintenance thalidomide administered with the objective of further ditioning regimen has almost been abandoned is useful in elderly patients initially treated decreasing the tumor burden. If this is true, in MM because of excessive toxicity.26 with MPT. Until now, in elderly patients, avail- we still have to determine the optimal level of Reduced-intensity conditioning is associated able data do not show a benefit from mainte- response. Is CR with negative immunofixa- with reduced transplant-related mortality but nance treatment with thalidomide, at least in tion the requested level or should we try to with increased relapse rate compared to stan- terms of OS. At least two randomized studies obtain higher levels of response (stringent dard allo-SCT.27 In order to decrease the relapse addressing the question of maintenance treat- response, immunophenotypic response, or rate, a strategy with tandem ASCT-reduced ment with lenalidomide in elderly patients are even molecular response)? intensity conditioning allo-SCT is currently ongoing.

[Hematology Reviews 2009; 1:e12] [page 67] Review

Table 3. Melphalan-prednisone versus melphalan-prednisone-thalidomide studies: absence of impact of maintenance with thalidomide on overall survival. Author Palumbo et al.31 Facon et al.32 Hullin et al.33 Golbrandsen et al.34 Wijermans et al.35 Number of patients (MPT) 331 (167) 447 (125) 232 (113) 362 (182) 301 (152) Age (yrs) 72 69 78.5 74.5 72 Median range 60-85 65-75 76-91 49-92 - WHO 3/4 (%) 58 7 30 4 MPT regimen Number of cycles 6 12 12 Until plateau Until plateau M dosing 4 mg/m2 0.25 mg/kg 0.2 mg/kg 0.25 mg/kg 0.2 mg/kg d1-7 d1-4 d1-4 d1-4 d1-5 T dosing 100 mg/d Up to 400 mg/d 100 mg/d Up to 400 mg/d 200 mg/d Maintenance YES NO NO YES YES Outcome (MPT vs. MP) Response rate (%) 76 vs. 48* 76 vs. 35* 62 vs. 30* 49 vs. 28* 66 vs. 47* PFS (m) 22 vs. 14.5* 27.5 vs. 18* 24 vs. 19* 20 vs. 18 14 vs. 10* OS (m) 45 vs. 47.5 51.5 vs. 33* 45 vs. 27.5* 29 vs. 33 37 vs. 30

Significant difference, (p<0.005); d, day; M, melphalan; P, prednisone; T, thalidomide; MP, melphalan-prednisone; WHO, World Health Organization; OS, overall survival; PFS, progression-free survival.

1430-4. gous transplantation in multiple myeloma: Conclusions 6. Bataille R, Harousseau J-L. Multiple results of a multicenter randomized clini- Myeloma. N Engl J Med 1997;336:1657-64. cal trial. Blood 2008;111:1805-10. In younger patients, post-ASCT mainte- 7. Fritz E, Ludwig H. Interferon-α treatment 15. Spencer A, Prince HM, Roberts A, et al. nance therapy with thalidomide appears to in multiple myeloma: meta-analysis of 30 Consolidation therapy with thalidomide increase tumor burden reduction further, randomized trials among 3948 patients and prednisolone prolongs the survival of which translates in prolonged PFS. However, Ann Oncol 2000;11:1427-36. multiple myeloma patients undergoing a the benefit in terms of OS is not clear and 8. Myeloma Trialists’ Collaborative Group. single ASCT procedure. J Clin Oncol 2009; many questions remain regarding the respec- Interferon as therapy for multiple myelo- 27:1788-93. tive role of consolidation versus maintenance, ma: an individual patient data overview of 16. Morgan GJ, Jackson GH, Davies FE, et al. the optimal drug and the optimal schedule of 24 randomized trials and 4012 patients. Br Maintenance thalidomide may improve administration and duration of treatment, as J Haematol 2001;113:1020-34. progression free but not overall survival: well as the characteristics of patients who may 9. Bjorkstrand B, Svensson H, Goldschmidt results from the Myeloma IX maintenance benefit from this approach. In elderly patients H, et al. α-interferon maintenance treat- randomisation. Blood 2008;112:656. there is currently no evidence that mainte- ment is associated with improved survival 17. Lokhorst HM, van der Holt B, Zweegmann nance treatment improves OS. after high-dose treatment and autologous S, et al. Final analysis of Hovon 50 random- stem cell transplantation in patients with ized Phase III study on the effect of multiple myeloma: a retrospective registry thalidomide combined withadriamycin, study from the EBMT. Bone marrow trans- dexamethasone and high-dose melphalan References plant 2001;27:511-5. in patients with multiple myeloma Blood 10. Cunningham D, Powles R, Malpas J, et al. 2008;112:461. 1. Harousseau J-L, Attal M, Avet-Loiseau H. A randomized trial of maintenance inter- 18. Abdelkefi A, Ladeb S, Torjman L, Retract - The role of complete response in multiple feron following high-dose chemotherapy ation of reference 14. Blood 2009; 113: myeloma. Blood 2009 Jul 28. [Epub ahead in multiple myeloma:long-term results Br 6265. of print]. J Haematol 1998;102:195-202. 19. Barlogie B, Pinedda-Roman M, van Rhee F, 2. Alexanian R, Balzerzac S, Haut A, et al. 11. Barlogie B, Kyle RA, Anderson KC, et al. et al. Thalidomide arm of Total therapy 2 Remission maintenance therapy for multi- Standard chemotherapy compared with improves complete remission and survival ple myeloma. Arch Intern Med 1975; 135: high-dose chemoradiotherapy for multiple in myeloma patients with metaphase cyto- 147-52. myeloma: final results of phase III US genetic abnormalities. Blood 2008; 112: 3. Belch A, Shelley W, Bergsagel D, et al. A Intergroup trial S9321. J Clin Oncol 2006; 3115-21. randomized trial of maintenance versus 24:929-36. 20. Harousseau J-L, Moreau P. Autologous no maintenance melphalan and pred- 12. Attal M, Harousseau J-L, Leyvraz S, et al. stem-cell transplantation in multiple nisone in responding multiple myeloma Maintenance therapy with thalidomide myeloma. N Engl J Med 2009;360:2645-54. patients. Br J Cancer 1988;57:94-9. improves survival in multiple myeloma 21. Fermand JP, Ravaud P, Chevret S, et al. 4. Berenson JR, Crowley JJ, Grogan TM, et al. patients. Blood 2006; 15:3289-94. High-dose therapy and autologous periph- Maintenace therapy with alternate-day 13. Barlogie B, Tricot G, Anaissie E, et al. eral blood stem cell transplantation in mul- prednisone improves survival in multiple Thalidomide and hematopoietic stem cell tiple myeloma: upfront or rescue treat- myeloma patients Blood 2002;99:3163-8. transplantation for multiple myeloma. N ment? Results of a multicenter sequential 5. Mandelli F, Avvisati G, Amadori S, et al. Engl J Med 2006;354:1021-30. randomized trial. Blood 1998;92:3131-6. Maintenance therapy with recombinant 14. Abdelkefi A, Ladeb S, Torjman L, et al. 22. Wang M, Dimopoulos MA, Chen C, et al. alfa-2b in patients with multiple myeloma Single autologous stem cell transplanta- Lenalidomide plus dexamethasone is responding to conventional induction tion followed by maintenance therapy with more effective than dexamethasone alone chemotherapy. N Engl J Med 1990; 322: thalidomide is superior to double autolo- in patients with relapsed or refractory mul-

[page 68] [Hematology Reviews 2009; 1:e12] Review

tiple myeloma regardless of prior thalido- ma: lower non-relapse mortality but high- controlled trial. Blood 2008;112: 3107-14. mide exposure. Blood 2008;112:4445-51. er relapse rates compared with myloabla- 32. Facon T, Mary JY, Hulin C, et al. Melphalan 23. Avet-Loiseau H, Attal M, Moreau P, et al. tive conditioning. Blood 2007; 109:3588- and prednisone plus thalidomide versus Genetic abnormalities and survival in mul- 94. melphalan and prednisone alone or tiple myeloma: the experience of the 28. Kroger N, Badbaran A, Lioznov M, et al. reduced-intensity autologous stem cell Intergroupe Francophone du Myélome. Post-transplant immunotherapy with transplantation in elderly patients with Blood 2007;109:3489-95. donor-lymphocyte infusions and novel multiple myeloma (IFM 99-06): a ran- 24. Paiva B, Vidriales MB, Cervero J, et al. agents to upgrade partial into complete domised trial. Lancet 2007;370:1209-18. Multiparameter flow cytometry is the most and molecular remission in allografted 33. Hulin C, Facon T, Rodon P, et al. Efficacy of relevant prognostic factor relevant prog- patients with multiple myeloma. Exp melphalan and prednisone plus thalido- nostic factor for multiple myeloma Hematol 2009;37:791-8. mide in patients older than 75 years with patients who undergo autologous stem cell 29. Ludwig H, Zdenek A, Tothova E, et al. newly diagnosed multiple myeloma: IFM transplantation. Blood 2008;112:4017-23. Thalidomide/Interferon versus interferon 01/01 Trial. J Clin Oncol 2009; May 18 25. Ladetto M, Pagliano G, Ferrero S, et al. maintenance after thal/dex versus MP. [Epub ahead of print]. Major shrinking of residual tumor cell bur- Clin Lymphoma Myeloma 2009;147. 34. Gulbrandsen N, Waage A, Gimsing P, et al. den and achievement of molecular remis- 30. Palumbo A, Bringhen S, Caravita T, et al. A randomised placebo controlled study sions in myeloma patients undergoing Oral melphalan and prednisone chemo - with melphalan/prednisone versus mel- post-transplant consolidation with borte- therapy plus thalidomide compared with phalan/prednisone/thalidomide: quality of zomib, thalidomide and dexamethasone: melphalan alone in elderly patients with life and toxicity. Haematologica 2008; aqualitative and quantitative PCR study. MM: randomized controlled trial. Lancet 93(S1):84(abstract 209). Blood 2008;112:3683. 2006; 367:825-31. 35. Wijermans PW, Zweegman S, van Marwijk 26. Harousseau J-L. The allogeneic dilemma. 31. Palumbo A, Bringhen S, Liberati AM, et al. Kooy M, et al. MP versus MPT in elderly Bone Marrow Transplant 2207;40:1123-8. Oral melphalan, prednisone, and thalido- myeloma patients: the final outcome of the 27. Crawley C, Iacobelli S, Bjorkstrand B, et al. mide in elderly patients with multiple HOVON-49 study. Clin Lymphoma Reduced-intensity conditioning for myelo- myeloma: updated results of a randomized, Myeloma 2009; (S14) (abstract 116).

[Hematology Reviews 2009; 1:e12] [page 69] Hematology Reviews 2009; volume 1:e13

Transcriptional regulation of Introduction Correspondence: Wayne K. Greene, the human ALDH1A1 promoter School of Veterinary and Biomedical Sciences, by the oncogenic homeoprotein TLX1 (HOX11), TLX2 and TLX3 belong to the Faculty of Health Sciences Murdoch University, TLX1/HOX11 ancient NKL family of homeobox genes that Perth, 6150 Western Australia. includes HEX, LBX1/2, MSX1/2, NKX3-2/BAPX1 E-mail: [email protected] 1 1 1 and NKX2-5/CSX. TLX1 encodes a transcrip- Kim L. Rice, Mansour Heidari, Key words: T-ALL, TLX1, homeoprotein, aldehyde tion factor that, although required during nor- Ross H. Taplin,2 Ursula R. Kees,3 dehydrogenase, transcriptional regulation. mal embryogenesis,2,3 was actually discovered Wayne K. Greene1 as a consequence of its aberrant expression in Acknowledgments: we thank Jette Ford and 1 4-7 School of Veterinary and Biomedical T-cell acute lymphoblastic leukemia (TALL). Rachael Brake for their technical advice and Sciences, Faculty of Health Sciences, Two distinct TLX1-expression categories have assistance. This work was supported by grants Murdoch University, Australia; been identified in T-ALL.8,9 High level TLX1 from the National Health and Medical Research 2School of Chemical and Mathematical expression (13%) is typically associated with Council and Murdoch University. KLR was sup- Sciences, Faculty of Minerals and Energy, 10q24 chromosomal abnormalities and confers ported by an Australian Postgraduate Award and MH by a scholarship from the Iranian Ministry of Murdoch University, Australia; a favorable prognosis whereas TLX1-low T-ALLs (22%) have an intact 10q24 locus and Health and Medical Education. 3Division of Children’s Leukaemia and expression does not impact prognosis.9 Cancer Research, Telethon Institute for Contributions: KLR and MH, performed the Aberrant expression of related TLX3 is addi- Child Health Research, Australia experiments and contributed to data analysis and tionally found in another 22% of T-ALLs follow- interpretation; WKG and KLR, wrote the manu- 10 ing 5q35 chromosomal rearrangements, script; WKG, RHT and URK, designed the study, which highlights the significant role of TLX interpreted the data, and critically revised the family members in T-cell oncogenesis. manuscript. Abstract Confirmation that the gene product of TLX1 is an oncoprotein has come from mouse mod- Present addresses: KLR, Northwestern University The homeoprotein TLX1, which is essential els, which have shown that enforced expres- Feinberg School of Medicine, Robert H. Lurie to spleen organogenesis and oncogenic when sion of TLX1 impairs cell differentiation and Comprehensive Cancer Center, Division of aberrantly expressed in immature T cells, leads to malignancy.11-15 Current models for the Hematology/Oncology, 303 East Superior St, Lurie functions as a bifunctional transcriptional mechanism by which TLX1 promotes leukemia 5-250, Chicago, IL 60611, USA. MH, Department of Medical Genetics, Tehran University of regulator, being capable of activation or are based on its ability to act indirectly, either Medical Sciences, Tehran, Iran. repression depending on cell type and/or pro- by enhancing chromosome instability16,17 or by moter context. However, the detailed mecha- regulating gene expression through specific Conflict of interest: the authors reported no poten- nisms by which it regulates the transcription protein-protein interactions with key cellular tial conflict of interests. of target genes such as ALDH1A1 remains to regulatory molecules such as the protein ser- be elucidated. We therefore functionally ine/threonine phosphatases PP1 and PP2A, Received for publication: 18 May 2009. Revision received: 30 July 2009. assessed the ability of TLX1 to regulate and the transcriptional coactivator/acetyltrans- 18-20 Accepted for publication: 31 July 2009. ALDH1A1 expression in two hematopoietic ferase, CREB-binding protein (CBP). Thus, cell lines, PER-117 T-leukemic cells and TLX1 may mediate its transforming function by This work is licensed under a Creative Commons human erythroleukemic (HEL) cells, by use of simultaneously inhibiting the phosphatase Attribution 3.0 License (by-nc 3.0) luciferase reporter and mobility shift assays. activity of PP1/PP2A to promote cell cycle pro- ©Copyright K.L. Rice et al., 2009 We showed that TLX1 physically interacts gression via upregulation of pathways such as 19 Licensee PAGEPress, Italy. with the general transcription factor TFIIB via those downstream of E2F and MYC, and sequestering CBP at heterochromatin to Hematology Reviews 2009; 1:e13 its homeodomain, and identified two activi- doi:10.4081/hr.2009.e13 accomplish a differentiation block.20 ties in respect to TLX1-mediated regulation of TLX1 has also long been suspected to act as the CCAAT box-containing ALDH1A1 promot- a sequence-specific transcription factor21,22 that er. The first involved CCAAT-dependent tran- preferentially binds to the core sequence version of the aldehyde form of vitamin A (reti- scriptional repression via perturbation of TAAT/GTG in vitro.22,23 However, with the excep- nal) to its biologically active form, retinoic GATA factor-containing protein complexes 31 tion of its associations with heterochromatic acid. ALDH1A1 apparently has normal roles in assembled at a non-canonical TATA (GATA) 32 satellite 2 DNA24 and its own promoter,25 no embryonic development, and in the box. A structurally intact homeodomain was direct target genes for TLX1 have been con- renewal/differentiation of hematopoietic stem essential for repression by TLX1 although vincingly identified. Individual genes suspect- cells (HSCs),33 where it is known to be highly direct DNA binding was not required. The ed to be regulated by TLX1 have been described expressed.34 ALDH1A1 is further implicated in second activity, which involved CCAAT-inde- in various settings, including spleen develop- regulating the polarity of HSC differentiation pendent transcriptional activation did not ment (Aldh1a1, Wt1),26,27 erythroid differentia- by favoring the development of a myeloid require an intact homeodomain, indicating tion (ALDH1A1, c-Kit, Vegfc),15,28 and T-cell rather than a lymphoid cell fate.30,35 In agree- that the activation and repression functions leukemia (ALDH1A1, FHL1, NR4A3).26,29,30 ment with this, ALDH1A1 expression can dis- of TLX1 are distinct. These findings confirm Nevertheless, the regulatory role that TLX1 criminate between acute myeloid (AML) and ALDH1A1 gene regulation by TLX1 and sup- plays in such cases still remains to be deter- acute lymphoid leukemia,36 and while we have port an indirect model for TLX1 function, in mined. demonstrated aberrant ALDH1A1 expression which protein-protein interactions, rather The best characterized TLX1 target gene, in T-ALL,30 ALDH1A1 is reportedly down-regu- than DNA binding at specific sites, are crucial ALDH1A1 (aldehyde dehydrogenase 1A1) lated in AML.37 Thus, ALDH1A1, which is regu- for its transcriptional activity. belongs to a subfamily (class 1A) of ALDH lated by TLX1 in its normal chromosomal con- genes whose main biological role is the con- text26,30 is of interest due to its associations

[page 70] [Hematology Reviews 2009; 1:e13] Article with both normal development and leukemoge- Hilden, Germany), were transiently transfect- Electrophoretic mobility shift assay nesis. Here, we explored the molecular mecha- ed as previously described.39 In brief, PER-117 (EMSA) nism(s) by which TLX1 regulates the ALDH1A1 or HEL cells (1×107) were co-transfected by Double-stranded EMSA probes were pre- gene and find that it occurs in a non-DNA bind- electroporation (300V, 960 µF) with 15 µg of pared from the following oligonucleotides: ing fashion through protein-protein interac- luciferase reporter plasmid (or as negative ALDH CAT (5’-AGTTTGTTCATCCAATCGTATCC- tions. We further show that TLX1 interacts control, pGL3-Basic) and 5 µg of pSV-β-Gal GAG-3’), ALDH CATMut (5’-AGTTTGTTCAT directly with the general factor TFIIB via its control plasmid. Cells were harvested 24 h gactgCGTATCCGAG-3’), ALDH GATAB (5’-GCC- homeodomain, indicating a role for TLX1 in later followed by measurement of luciferase CGTGCAGATAAAAAAGGAACA-3’), ALDH gene regulation via the basal transcriptional and β-galactosidase activities using the GATABMut (5’-GCCCGTGCActcagcAAAGGA machinery. Tropix Dual-Light luminescent reporter gene ACA-3’). Probes (20 pmol) were labeled by assay system (Applied Biosystems, Foster City, incubation (37°C for 30 min) with T4 polynu- CA, USA). Transcriptional activity was defined cleotide kinase (15 U, Gibco BRL) and 40 µCi as the ratio of luciferase activity (in relative [γ32P]-ATP (3000 Ci/mmol) in a volume of 20 Materials and Methods light units; RLU) from pGL3-Basic derivatives µL. The radiolabeled probes were purified relative to β-galactosidase activity from pSV- using Microspin G-50 columns (Amersham Cell culture and expression plasmids β-Gal, which reflected the efficiency of trans- Biosciences) and made double-stranded by The PER-117 and ALL-SIL T-cell lines, and fection. All experiments were repeated a total annealing with an equimolar amount of com- erythroleukemic cell line HEL, were cultured of three times on different days. plementary oligonucleotide in 1 x annealing as previously described.30 The coding regions For measuring the effect of TLX1 on buffer (10 mM Tris-HCl, pH 7.4, 50 mM NaCl, 1 of human TLX1 and TFIIB were amplified by ALDH1A1 promoter activity, transfections mM EDTA). Probes were incubated in 1 x bind- RT-PCR from ALL-SIL cDNA generated by were similarly performed with the additional ing buffer (20 mM HEPES, pH 7.6, 50 mM KCl, Thermoscript RT (Invitrogen, Carlsbad, CA, inclusion of 15 µg of expression plasmid, 1 mM EDTA, 1 mM DTT, 5% glycerol) with 0.5 USA) using PfuTurbo DNA Polymerase either pEF-BOS/TLX1, pEF-BOS/TLX1ΔH3 or µg poly(dI-dC) (ICN, Costa Mesa, CA, USA) (Stratagene, La Jolla, CA, USA) and primers pEF-BOS as control.40 In this case, transcrip- and 6 µg of nuclear extract in a final volume of containing an Nhe I restriction site. The tional activity was defined as the log (base 2) 15 µL. The samples were incubated at 4°C for resulting products were cloned into the Nhe I transformation of the ratio of luciferase activ- 30 min and then analyzed by electrophoresis site of the pCINeo mammalian expression vec- ity (in relative light units; RLU) from pGL3- on 4% native polyacrylamide gels in 0.5 x TBE tor (Promega, Madison, WI, USA) and Basic derivatives relative to β-galactosidase at 10 V/cm. For competition experiments, an sequence verified. activity from pSV-β-Gal. Statistical analysis excess of unlabeled competitor oligonucleotide was performed in SPLUS 2000 using a mixed was added to reaction mixtures. Additional Luciferase reporter constructs effects model with day of experiment as a ran- competitor oligonucleotides used were those The human ALDH1A1 proximal promoter dom effect and luciferase reporter plasmid containing TLX1 (5’-TTCCATTCGATAATTC- region was amplified by high-fidelity PCR from and expression plasmid as fixed effects. CATTCGA-3’) or GATA (5’-GAAACCTGT- genomic DNA as previously described.29 For the Interactions between contrasts comparing GATAAGTGTATGCAG-3’) binding sites. construction of the -978/+42 construct, a 1020 pEF-BOS/TLX1 with pEF-BOS, and contrasts Bandshifts were performed in the presence of bp fragment of the ALDH1A1 promoter was comparing ALDH1A1 constructs with pGL3- anti-TLX1 by adding 2 uL of polyclonal rabbit amplified using the forward primer 5’- Basic, were examined. This revealed the antiserum raised against the C-terminus (sc- GCGAGCTCCACAATCAGAGCATCCAGAGTA- extent to which the effect of adding TLX1 was 880, Santa Cruz Biotechnology, Santa Cruz, 3’and reverse primer 5’-GCGCTAGCCTCCTG- different for the various promoter constructs CA). Normal rabbit serum was used for the no GAACACAGGTGACTGGCT-3’ (introduced Sac compared to pGL3-Basic. antibody control. Following electrophoresis, I/Nhe I restriction sites in italics). To create the gels were transferred to 3MM paper the -303/+42, -201/+42, -146/+42 and -91/+42 Preparation of nuclear extracts (Whatman, Maidstone, UK), dried and autora- constructs, various lengths of the ALDH1A1 Cells (1×107) were washed twice with 10 diographed at -80°C. promoter were amplified using the same mL of cold PBS, resuspended in 400 µl Buffer reverse primer together with the forward A (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM Western blotting and immunopre- primers 5’-GCGAGCTCAAAAAAATAATAACTG- EDTA, 0.1 mM EGTA) with protease inhibitors cipitation of TLX1 complexes GCCTTAGTG-3’, 5’-GCGAGCTCCAGGTACAAAT- (100 µg/mL aprotinin, 5 µg/ml leupeptin, 1 Western blotting was performed with 50 ug TCGATGCTGGAGCACTGG-3’, 5’-GCGAGCTCAA- µg/ml pepstatin A, 0.5 mM PMSF) and incubat- of nuclear extract electrophoresed through a AGGCTTCCTGCCCTAGGTG-3’ and 5’-GCG- ed on ice for 30 min. 12% SDS-PAGE gel and transferred to a Hybond AGCT CT GAGTTTGTTCATCCAATCG-3’, respec- NP-40 was added to a final concentration of ECL membrane (AP Biotech, Little Chalfont, tively. The -50/+42 and +1/+42 constructs were 0.5% and the cells vortexed for 10 s. The nuclei UK). The membrane was blocked overnight in generated directly by oligonucleotide synthe- were pelleted by centrifugation at 6500 g for 1 TBS-Tween buffer (1 mM Tris, 7.5 mM NaCl, sis. All promoter fragments were directionally min and resuspended in 100 µL Buffer C (20 0.05% v/v Tween-20) containing 3.5% (w/v) cloned into the Sac I/Nhe I sites of the mM HEPES, pH 7.9, 420 mM NaCl, 1 mM gelatin and incubated with 1:1200 diluted rab- luciferase reporter vector pGL3-Basic EDTA, 1 mM EGTA, 20% glycerol) with pro- bit anti-TLX1 antiserum (sc-880, Santa Cruz (Promega, Madison, WI, USA). Insert identi- tease inhibitors. The nuclear suspension was Biotechnology) (1 h, RT). The membrane was ties were confirmed by automated DNA stirred vigorously on ice for 30 min. The sam- incubated with HRP-conjugated secondary sequencing. The FHL1 -821 promoter construct ple was centrifuged at 13 000 g for 10 min, and antibody (1 h, RT) prior to visualization of has been previously described.38 aliquots of the nuclear extract were frozen bands by enhanced chemiluminescence (AP immediately in liquid nitrogen and stored at Biotech). For immunoprecipitation, nuclear Luciferase and β-galactosidase -80°C until required. The protein concentra- extract from ALL-SIL cells (5×106) was pre- reporter gene assays tion of nuclear extracts was determined by the cleared with the addition of protein A/G Plasmid DNAs (1 µg/µL), prepared using a Bradford assay (Bio-Rad Laboratories, agarose (Santa Cruz Biotechnology) for 15 purification kit (Plasmid Maxi, QIAGEN, Hercules, CA, USA). min at 4°C. Pre-cleared extracts were incubat-

[Hematology Reviews 2009; 1:e13] [page 71] Article ed with 5 µg of affinity-purified rabbit anti- sequences between -50 and -91 and between - TLX1 polyclonal antibody (sc-880; Santa Cruz Results 50 and -146 were required for maximal promot- Biotechnology), or no antibody (as control), for er activity in PER-117 and HEL cells, respec- 4 h at 4°C with constant gentle rocking. Functional analysis of the human tively (Figure 2A). Consistent with previous Immune complexes were bound to protein A/G ALDH1A1 promoter results in Hep3B liver cells,44 deletion of the agarose beads, centrifuged and washed with Ectopic expression of TLX1 was previously CCAAT box at -74 caused a dramatic reduction 1.2 mL of cold IP wash buffer (100 mM Tris-HCl shown to modulate endogenous ALDH1A1 in promoter activity, suggesting that it is a key pH 7.4, 1% NP40 and 1% deoxycholic acid) con- expression.26,30,38 To pursue the molecular element for ALDH1A1 expression in multiple taining 500 mM LiCl (once) and 1.2 mL of IP mechanism(s) underlying this regulatory cell types. wash buffer (four times). effect, we cloned 1020 bp of the human ALDH1A1 5’-flanking region into the pGL3- Regulation of the ALDH1A1 pro- MALDI-TOF mass spectrometric Basic luciferase reporter vector. This sequence moter by TLX1 analyses comprised nucleotides -978 to +42 relative to To determine the effect of TLX1 on the tran- Immunoprecipitated proteins were resolved the transcriptional start site (Figure 1). Among scriptional activity of the ALDH1A1 promoter, by SDS-PAGE and visualized by silver stain- the conserved promoter elements identified transactivation assays were carried out in ing.41 Bands were excised directly from gels were a TATA-like sequence (GATA box) at -33 PER-117 and HEL cells co-transfected with an into 96-well microtiter plates (Titertek, and a single CCAAT box at -74, which has pre- expression plasmid containing TLX1 or empty Huntsville, AL, USA), destained,42 and in-gel viously been shown to be functionally impor- vector (pEF-BOS) as a negative control. The 44 trypsin digestions performed according to tant for ALDH1A1 expression. Transient results were expressed as a ratio of the normalized transcriptional activity of each of the Shevchenko et al.41 Peptides were extracted transfection of a series of deletion constructs (-978/+42, -303/+42, -201/+42, -146/+42, -91/ two promoter constructs with and without with 10 mg/mL α-cyano-4-hydroxycinnamic TLX1 (Figure 2B). Western blot analysis con- acid in 50% acetonitrile containing 0.1% triflu- +42, -50/+42 and +1/+42; Figure 1) into PER- firmed that TLX1 protein was expressed in the oroacetic acid, and aliquots of 0.5 uL applied 117 and HEL cells (which both lack TLX1 expression) showed that the ALDH1A1 promot- transfected cells (Figure 2C). Consistent with directly onto a target plate and allowed to air er was functional in both cell lines and that the our previous findings,15,29,38 the effect of TLX1 dry. Tryptic peptide masses were then obtained using a Voyager-DE STR MALDI-TOF mass spectrometer (Applied Biosystems, Foster City, CA, USA). Known trypsin autocleavage peptide masses (842.51, 1045.56 and 2211.1 Da) were used for internal calibration of each spectrum. The peptide masses were used to search the Swiss-Prot and NCBI nr protein databases using the MS-FIT database tool. A protein was considered to be identified if a minimum of five measured peptide masses matched calculated tryptic peptide masses and if the peptides identified by these matches provided at least 20% sequence coverage of the identified protein.

GST-pulldown assay TLX1 and TFIIB cDNAs in pCIneo were transcribed in vitro with T7 RNA polymerase. The products were labeled with [35S]-methionine (Amersham Biosciences) using the TNT coupled transcription-translation system (Prome- ga). GST-TLX1 fusion proteins were expressed from pGEX-6P-1 as described previously.43 The GST-pulldown assay was performed by incubat- ing 15 µg of GST, GST-TLX1 or GST-TLX1ΔH3 immobilized on glutathione sepharose beads (20 µL) with 5 µl of proteins translated in vitro and labeled with [35S]-methionine in 500 µL of binding buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5 mM EDTA, 0.5% NP-40, 1 mM DTT, 0.5 mM PMSF, 0.05% BSA) for 18-20 h at 4°C with continuous rotation. Bound proteins Figure 1. Nucleotide sequence alignment of the human and mouse ALDH1A1 promoters. Conserved nucleotides are marked with an asterisk. Numbers at right indicate nucleotide were washed three times with 500 µL of cold positions of the human sequence relative to the transcriptional start site (bent arrow). binding buffer and eluted in SDS sample Potential regulatory elements, including a conserved TLX1 site at -257, are boxed and buffer. The eluted proteins were resolved on labeled. Solid triangles with numbers demarcate the 5’ ends of the promoter deletion con- 12% SDS-PAGE gels and visualized by autora- structs used in the luciferase reporter assays. The two regions used for gel shift analysis are underlined at -85 to -60 and -42 to -19. BRE, TFIIB recognition element. diography.

[page 72] [Hematology Reviews 2009; 1:e13] Article on ALDH1A1 was reversed in the transcription- a al assay as compared to regulation of the Figure 2. Activity of the human 25 ALDH1A1 promoter and its differential endogenous gene. In PER-117 cells, TLX1 regulation by TLX1. (A) ALDH1A1 pro- repressed ALDH1A1 promoter activity (Figure 20 moter activity in the hematopoietic cell 2B), and this required the sequence between lines PER-117 and HEL. Luciferase activ- -50 and -91, which contains the CCAAT box. 15 ity expressed as fold-change over the promoter-less pGL3-Basic vector after nor- Deletion of the CCAAT box (-91ΔCAT) com- 10 malizing the luciferase signals to a β- pletely abolished TLX1-mediated repression

Normalized fold activation galactosidase transfection control. The (Figure 2B), suggesting that downregulation 5 indicated ALDH1A1 promoter deletion may possibly occur via this site. Interestingly, constructs were transiently transfected the shortest constructs (-50/+42 and +1/+42) into PER-117 cells (black bars) and HEL +1 -50 -91 -91ΔCAT -146 -201 -303 -978 cells (gray bars). Deletion of the CCAAT were mildly stimulated by TLX1 in both PER- pGL3 Constructs box (-91ΔCAT construct) abrogates pro- 117 and HEL cells, (Figure 2B) indicating the moter activity. The values represent the presence of a separate, positive-acting TLX1- average of duplicate data points from response sequence localized to the region b three independent experiments. (B) The between +1 and +42. No effect by TLX1 was 1.5 ALDH1A1 promoter is differentially regulated by TLX1 in PER-117 and HEL observed on the region between -201 and -303, 1 cells. The indicated ALDH1A1 luciferase which contains a conserved TLX1 in vitro bind- 0.5 reporter constructs were co-electroporat- ing site at -257 (Figure 1), suggesting that ed with a TLX1 expression plasmid and TLX1 binds to a distinct recognition sequence 0 β-galactosidase expression plasmid into the cell lines PER-117 (black bars) and in the ALDH1A1 promoter in vivo, or alterna- -0.5 HEL cells (gray bars). Transactivation/ tively, that it acts via a non-DNA binding mech- Log2 fold change -1 repression is shown as the fold-change anism. In either case, these data indicate that (on a log2 scale) compared to back- TLX1 operates via at least two mechanisms in -1.5 ground levels obtained when an empty respect to the ALDH1A1 promoter; a general -2 expression vector (no TLX1) was co-elec- +1 -50 -91 -91ΔCAT -146 -201 -303 -978 troporated. Asterisks denote fold changes transactivating activity via an element located pGL3 Constructs that were statistically different to zero between +1 and +42 and a strong, cell line- (p<0.05). The -91ΔCAT construct har- specific repressive activity via an element bors a deletion of the CCAAT box. (C) Western blot confirming TLX1 protein located between -91 and -50. C PER117 HEL expression in the transiently transfected U TUT PER-117 and HEL cells. U, untransfect- The homeodomain is required for ed controls; T, TLX1 transfected. The TLX1-mediated repression faint band observed in the controls is To assess whether the homeodomain of non-specific. TLX1 is required for the transcriptional activities of TLX1, a mutant TLX1 expression vector (TLX1ΔH3) was employed, which lacks the DNA recognition helix (helix 3) of the homeodomain. Whereas, TLX1 repressed the activity 2.5 of the basal ALDH1A1 promoter (-91/+42) 4- fold in PER-117 cells, TLX1ΔH3 not only lacked 2.0 the ability to negatively regulate ALDH1A1, but Figure 3. TLX1 requires an intact homeodomain for transcriptional repression switched to become an activator, stimulating 1. but not activation in vitro. Luciferase transcription by approximately 5-fold (Figure activity of the ALDH1A1 -91/+42 and 3). A similar result was obtained when the pro- FHL1 -821/+181 promoter constructs 1.0 moter of another target gene, FHL1,38 was used cotransfected into the indicated cell lines

Relative luciferase activity with either an empty expression vector as (Figure 3). By contrast, the positive regulation 0.5 negative control (striped bars), TLX1 of the -91/+42 construct observed in HEL cells (black bars) or TLX1ΔH3, a mutant lack- was comparable when either TLX1 or TLX1ΔH3 ing helix 3 of the homeodomain (gray was used (5.5-and 8-fold induction, respective- bars). The values represent the average of ly). Thus, the homeodomain is crucial for ALDH1A1 -91 FHL1 -821 ALDH1A1 -91 duplicate data points from three independent experiments. TLX1-mediated repression but is not required PER117 HEL for TLX1-mediated activation, indicating that these two activities are distinct.

TLX1 does not affect formation of cells, either expressing empty pEF-BOS (as to the two cell types, while additional complex- transcriptional complexes at the -74 control) or pEF-BOS/TLX1, were incubated es P and H were unique to PER-117 and HEL CCAAT site with a radiolabeled probe (ALDH CAT) span- cells, respectively. These complexes were We opted to focus on the mechanism by ning the CCAAT motif (-85 to -60; Figure 1). As inhibited by the addition of a 35-fold molar which TLX1 mediates CCAAT-dependent shown in Figure 4A, incubation of nuclear excess of unlabeled self-competitor (ALDH repression. Mobility shift assays were there- extract from PER-117 or HEL cells with ALDH CAT) but not by a 35-fold molar excess of fore performed to identify whether TLX1 CAT resulted in the appearance of at least two CCAAT mutant competitor (ALDH CATMut; directly binds to the CCAAT box or interferes specific DNA-protein complexes, together with CCAAT to GACTG). Contrary to expectation, in with the binding of other factors at this site. a third non-specific complex (C2). The both PER-117 and HEL cells the mobility shift Nuclear extracts derived from PER-117 or HEL strongest, complex C1, appeared to be common pattern was identical regardless of TLX1

[Hematology Reviews 2009; 1:e13] [page 73] Article expression status (or addition of TLX1 anti- expression with ALDH GATAB resulted in the beled consensus GATA oligonucleotide (GATA; body; data not shown), suggesting that TLX1 appearance of four specific DNA-protein com- containing WGATAR with different flanking does not affect the formation of DNA-protein plexes, C1, C2, P1 and P2. HEL nuclear extracts sequences to ALDH GATAB) was used (Figure complexes at the CCAAT box. only produced three specific complexes (C1, 5A). This indicated that GATA factors are pres- C2 and H), the strongest of which (C2) ent in all complexes formed in this assay. TLX1 alters DNA-protein complex appeared to migrate similarly in both cell However, TLX1 does not appear to bind DNA at formation at the -33 GATA box types. These complexes were inhibited by the this site, since a 35-fold molar excess of an TLX1 may directly or indirectly inhibit addition of a 35-fold molar excess of unlabeled unlabeled consensus satellite 2 oligonu- CCAAT-dependent transcription via the basal self-competitor (ALDH GATAB) but not by a 35- cleotide (TLX1), capable of being bound by apparatus. The ALDH1A1 promoter lacks a fold molar excess of GATA box mutant competi- TLX1,24 could not compete for complex forma- canonical TATA box but does possess a related tor (ALDH GATABMut; GATAAA to CTCAGC). tion (Figure 5A). In agreement with this con- GATA box (GATAAA) utilized by a number of Strikingly, in PER-117 but not HEL cells, clusion, no supershift or inhibition of complex genes whose transcriptional initiation involves expression of TLX1 resulted in a significant formation was observed following addition of interplay between TFIID and GATA factors.45-48 alteration in complex formation. Of the four TLX1 antibody (Figure 5B). Notably, however, EMSAs were thus performed to identify specific PER-117 complexes, the formation of PER-117 complex C1, which is abolished by whether TLX1 directly binds to the GATA box or two (C1 and P1) was completely inhibited by TLX1, re-formed in the presence of TLX1 anti- interferes with the binding of other factors at TLX1, while the intensity of the remaining two body, indicating that TLX1 directly contributes this site. Nuclear extracts derived from PER- (C2 and P2) was enhanced. Thus, TLX1-medi- to the disruption of this low mobility complex. 117 or HEL cells, either expressing empty pEF- ated regulation of ALDH1A1 in T cells, but not BOS (as control) or pEF-BOS/TLX1, were incu- erythroid cells, is associated with an alteration TLX1 interacts with TFIIB bated with a radiolabeled probe (ALDH GATAB) of transcription factor binding at the GATA box. To help understand the transcriptional regu- spanning the GATA box (-42 to -19; Figure 1). Complex formation at the GATA box in both latory function of TLX1, we searched for bind- As shown in Figure 4B, incubation of nuclear PER-117 and HEL cells was also strongly inhib- ing partners using an immunoprecipitation extract from PER-117 cells lacking TLX1 ited when a 35-fold molar excess of an unla- strategy in leukemic T cells (ALL-SIL) that

a ALDH ALDH ALDH ALDH b ALDH ALDH ALDH ALDH Competitor: CAT CATMut CAT CATMut Competitor: GATAB GATABMut GATAB GATABMut 117 117 117 HEL HEL HEL 117 117 117 HEL HEL HEL Extract: 117 TLX 117 TLX 117 TLX1 HEL TLX1 HEL TLX1 HEL TLX1 Extract: - 117 TLX1 117 TLX1 117 TLX1 HEL TLX1 HEL TLX1 HEL TLX1

*C1 P C1 *P1

*C2 H

C3 C2

*P2 H

Free Free probe probe

Figure 4. Effect of TLX1 on ALDH1A1 promoter DNA-binding complexes. (A) TLX1 does not affect DNA-protein complex formation at the -74 CCAAT box. EMSA using a 32P-labeled double-stranded oligonucleotide containing the CCAAT site at -74 of the human ALDH1A1 promoter (ALDH CAT). The assay was performed with nuclear extracts prepared from PER-117 (left lanes) or HEL cells (right lanes) with and without overexpression of TLX1. The migration of the free probe is indicated along with the positions of specific DNA-protein complexes (C1 and P in PER¬117 cells; C1 and H in HEL cells). Overbars denote the addition of a 35-fold molar excess of unlabeled double-stranded self (ALDH CAT) or mutant (ALDH CATMut) probe as competitor. (B) TLX1 alters DNA-protein complex formation at the -33 GATA box. EMSA performed as above using a labeled oligonucleotide containing the GATAAA site at -33 of the human ALDH1A1 promoter (ALDH GATAB). The migration of the free probe is indicated along with the positions of specific DNA-protein complexes (C1, C2, P1 and P2 in PER-117 cells; C1, C2 and H in HEL cells). Overbars denote the addition of a 35-fold molar excess of unlabeled double-stranded self (ALDH GATAB) or mutant (ALDH GATABMut) probe as competitor. Asterisks denote complexes altered by TLX1.

[page 74] [Hematology Reviews 2009; 1:e13] Article aberrantly express TLX1 as a consequence of a 35S-labeled TFIIB or TLX1 protein. The latter vator and repressor,26,30,38 however the mecha- 10q24 chromosomal translocation. A TLX1 was included as a positive control since TLX1 nism(s) responsible for these divergent roles antibody was employed to isolate naturally has previously been shown to homodimerize.24 is poorly understood. In this study, we further occurring nuclear protein complexes, which Following extensive washing, retention of investigated the function of TLX1 by analyzing were separated by SDS-PAGE and silver TFIIB was observed with GST-TLX1 but not its ability to transcriptionally regulate stained (Figure 6A). Excised gel slices repre- with control GST beads (Figure 6B), demon- ALDH1A1, a gene previously identified as senting discrete molecular mass intervals strating that TLX1 has the capacity to interact being TLX1-dependent in developing mouse were digested with trypsin and analyzed by with TFIIB. Interestingly, retention of TFIIB, spleen as well as in several cell lines (PER-117, MALDI-TOF mass spectrometry in order to but not TLX1 control, was greatly diminished HEL, NIH-3T3),26,30,49 and which contains a pre- determine the identity of the protein bands. when GST-TLX1ΔH3 beads were used, indicat- dicted TLX1 binding site, conserved between Table 1 summarizes the eight proteins detecting that helix 3 of the TLX1 homeodomain con- human and mouse, at -257 upstream of the ed. Among these were the centromeric pro- tributes strongly to the TFIIB interaction. transcriptional start site. In the first instance, teins CENP-E and CENP-F, an intriguing find- our data confirmed ALDH1A1 as a regulatory ing given that TLX1 has previously been local- target of TLX1, with the polarity of effect ized to centromeric regions.24 Of particular observed in terms of activation/repression interest was the detection of the general tran- Discussion being heavily dependent upon cell type. In scription factor TFIIB, which was identified by PER-117 T cells, transient luciferase reporter the presence of 15 peptides with sequence cov- TLX1 has previously been characterized as a assays with nested deletions of the ALDH1A1 erage of 49% (Table 1). To confirm that TLX1 DNA-binding factor that preferentially associ- promoter revealed that TLX1-mediated repres- could physically interact with TFIIB in a specif- ates with the core sequence TAAT/GTG in sion occurred in a CCAAT box-dependent man- ic manner, we performed a GST pulldown assay vitro22,23 and the similar sequence TA/GATTC ner involving an element located between -91 (Figure 6B). Glutathione-Sepharose beads present in satellite 2 DNA.24 In addition, TLX1 and -50. By contrast, TLX1 stimulated tran- containing GST-TLX1, GST-TLX1ΔH3, or GST can switch, in a cell type- and promoter con- scription in a CCAAT-independent manner alone were incubated with in vitro translated text-dependent manner, between roles as acti- from a proximal location (-91/+42) in human

a ALDH TLX1 GATA ALDH TLX1 GATA b GATAB GATAB 117 Competitor: HEL Extract: TLX1 117 117 117 117 HEL HEL HEL HEL 117 HEL TLX1 Extract: - 117 TLX1 117 TLX1 117 TLX1 117 TLX1 - HEL TLX1 HEL TLX1HEL TLX1 HEL TLX1 anti-TLX1:

*C1 *C1

*P1 *P1

*C2 *C2

C3 C3

*P2 *P2 H H

Free Free probe probe

Figure 5. Perturbation of GATA-containing complexes by TLX1. (A) GATA factor(s) but not TLX1 binds to the ALDH1A1 promoter GATA box. EMSA using a 32P-labeled double-stranded oligonucleotide containing the GATAAA site at -33 of the human ALDH1A1 promoter (ALDH GATAB). The assay was performed with nuclear extracts prepared from PER-117 (left panel) or HEL cells (right panel) with and without overexpression of TLX1. The migration of the free probe is indicated along with the positions of specific DNA- protein complexes (C1, C2, P1 and P2 in PER-117 cells; C1, C2 and H in HEL cells). Overbars denote the addition of a 35-fold molar excess of unlabeled double-stranded self (ALDH GATAB), satellite 2 DNA (TLX1) or GATA consensus (GATA) probe as competitor. Asterisks denote complexes altered by TLX1. (B) TLX1 directly disrupts the low mobility ALDH1A1 promoter GATA box complex C1. EMSA performed as above using the -33 GATAAA site oligonucleotide (ALDH GATAB) in the absence (-) or presence (+) of TLX1 antibody. The migration of the free probe is indicated along with the positions of specific DNA-protein complexes (C1, C2, P1 and P2 in PER-117 cells; C1, C2 and H in HEL cells). Complex C1 reappears in PER-117 cells expressing TLX1 in the presence of anti-TLX1 antibody (asterisk).

[Hematology Reviews 2009; 1:e13] [page 75] Article

Input (50%) TLX1 TFIIB

Figure 6. TLX1 interacts with TFIIB. kDa (A) Isolation of TLX1-associated pro-

r Ab teins. Nuclear extracts from ALL-SIL

M No Ab leukemic T cells were immunoprecip- 200 itated using TLX1-specific antibody (Ab), or no antibody (No Ab) as neg- 66.2 TLX1 TFIIB GST GST-TLX1 Δ H3 GST-TLX1 GST GST-TLX1 Δ H3 GST-TLX1 ative control. Coprecipitated proteins 45 were resolved by SDS-PAGE and silver stained. (B) GST pulldown assay. GST, GST-TLX1 and GST-TLX1ΔH3 31 immobilized on glutathione-Sepha - rose beads were incubated with in vitro-translated, 35S-labeled TLX1 as 21.5 positive control, and TFIIB. Bound proteins were washed, eluted, resolved by 12% SDS-PAGE and 6.5 detected by autoradiography: 50% of the labeled input proteins TLX1 and TFIIB are shown on the left.

Table 1. Proteins identified by immunoprecipitation and peptide mass fingerprinting. Identified protein Matched Mass UniProtKB/Swiss-Prot peptide cover (kDa) accession number (%) General transcription factor IIB (TFIIB) 49 34.83 Q00403 Centromere protein F (CENPF) 39 367.76 P49454 Centromere protein E (CENPE) 35 316.42 Q02224 RAD50 homolog (RAD50) 31 153.89 Q92878 Retinoblastoma binding protein 9 (RBBP9) 28 21.00 O75884 NK homeobox 6-1 (NKX6-1) 28 37.85 P78426 Lymphoblastic leukemia 1 (LYL1) 23 28.63 P12980 Zinc finger protein 20 (ZNF20) 20 61.57 P17024

erythroleukemic (HEL) cells. Transactivation, assays was inversely related to its previously transcription in PER-117 cells but switched to which was also observed in the T-cell back- observed effect on endogenous ALDH1A1 lev- become an activator of transcription. This indi- ground when a minimal promoter sequence els in PER-117 and HEL cells.30 This phenome- cated that the repression/activation activities was used (-50/+42), was abolished in HEL cells non has been observed previously in respect to of TLX1 are separable with a structurally intact with the addition of extra DNA sequence (- other putative gene targets of TLX1, namely homeodomain being absolutely required for 146/-92) suggesting that a specific factor NR4A3, KIT and FHL1.15,29,38 Although puzzling, transcriptional repression, but not activation. bound to this region is capable of abrogating the finding that a homeoprotein can act as a Given these findings, a reasonable assumption the stimulatory potential of TLX1. Thus, TLX1 repressor or activator of transcription depend- was that TLX1 would repress ALDH1A1 tran- possessed two activities in respect to the ing on promoter context is a common one. In scription by directly binding promoter DNA at ALDH1A1 promoter, namely CCAAT-dependent many cases the activity that predominates has or near the CCAAT box. This crucial element is repression, which was cell type-specific and been found to be highly dependent on the likely generally required for ALDH1A1 expres- cryptic CCAAT¬independent activation, which nature of the cis-regulatory DNA sequence sion, since it was also identified as the major was unmasked by deleting upstream regulato- since this can, in turn, affect the interaction of positive element in the Hep3B cell line and in ry sequences. Neither of these activities homeoproteins with co-regulatory molecules Hepa-1 mouse hepatoma cells where it was involved the TLX1 recognition site at -257, such as TALE (three amino acid loop exten- bound by NF-Y and CCAAT/enhancer binding which initially suggested that TLX1 either sion) homeoproteins, CBP (CREB-binding pro- protein β (C/EBPβ), respectively.44,55 Moreover, bound a distinct sequence in vivo, as is the tein) or Groucho.50-54 the CCAAT box is capable of being bound by case with the regulation of its own promoter,25 CCAAT-dependent repression by TLX1 was the CCAAT binding transcription factor or that it acted via a non-DNA binding mecha- found to require the DNA recognition helix (CTF1/NFIC) with which TLX1 is known to nism. (helix 3) of the homeodomain. Intriguingly, interact in a functional manner.56 However, Remarkably, the effect of TLX1 on ALDH1A1 the TLX1ΔH3 mutant lacking this helix was EMSA assays using PER-117 (or HEL) nuclear promoter activity in the transient reporter not only incapable of repressing ALDH1A1 extracts revealed that TLX1 did not directly

[page 76] [Hematology Reviews 2009; 1:e13] Article bind the CCAAT box, nor did it affect protein the co-activator/histone acetyltransferase mol- and Vnd/Nkx2-251,53,54,65 is capable of acting as a complex formation at this site. We therefore ecule CBP.20 This interaction leads to seques- bi-functional transcriptional regulator, whose hypothesized that TLX1 may abrogate CCAAT- tration of CBP at repressive chromatin activation and repression activities operate in a dependent transcription of ALDH1A1 by acting domains, which is consistent with our previ- cell-type specific manner and via two distinct through the basal transcriptional apparatus. ous finding that TLX1 can localize to hete- mechanisms. The first involves the ability of Supporting this notion, a previous study rochromatin via binding at satellite 2 DNA.24 TLX1 to repress transcription, possibly by per- demonstrated that TLX1 is capable of repress- The TLX1-CBP interaction was also shown to turbing interactions between CCAAT box-binding transcription via the RNA polymerase II require helix 3 of the TLX1 homeodomain, ing transcriptional activators and proteins holoenzyme in a manner that is DNA-inde- thereby suggesting an alternative, although (GATA factors/basal transcriptional machinery) pendent, yet requires helix 3 of the home- not mutually exclusive, explanation for TLX1- assembled at a non-canonical TATA (GATA) odomain.49 Indeed, we found that TLX1 altered mediated repression and its reversal in the box. This activity does not appear to involve the formation of GATA-containing protein com- TLX1ΔH3 mutant. Riz et al.20 further showed a direct DNA binding, although a structurally plexes at the non-canonical TATA (GATA) box correlation between the presence of TLX1 and intact homeodomain is required, presumably in located at -33 on the ALDH1A1 promoter and a lack of CBP-associated acetylation of GATA1. order for TLX1 to interact with TFIIB and/or that this occurred specifically in PER-117 cells As is the case for GATA2 and GATA3,59,60 the CBP. Chromatin immunoprecipitation assays to where repression by TLX1 was observed. The DNA binding and transcriptional activity of identify the specific transcription factor(s) GATA box is a dual regulatory site capable of GATA1 is heavily dependent on its acetylation bound at this composite CCAAT-box/TLX1 binding both GATA factors and members of the status.61 Taken together, we suggest a model in responsive element and at the GATA box before basal transcriptional machinery, which is over- which interactions between TLX1 and both and after TLX1 expression are required to sub- represented among erythroid-specific gene TFIIB and CBP lead to the transcriptional stantiate this hypothesis. The second activity promoters. Studies to date have indicated that repression observed in our transient luciferase involves the ability of TLX1 to stimulate tran- the GATA box is a specialized/weakened TATA assays. Whereas contacts with TFIIB may scription through as yet unidentified regulato- box57 whose activity depends on interplay inhibit the formation of a functional PIC, bind- ry elements in the proximal ALDH1A1 promoter between binding of GATA factors and general ing to CBP may prevent it functioning as a co- and does not require an intact homeodomain. transcription factors such as TFIID.45-48 Based activator and acting to acetylate GATA factors. Our understanding of the role of TLX1 in nor- on our data, we therefore postulate that TLX1 This would explain the substantially altered mal development and in T-cell leukemogenesis can operate as a transcriptional repressor by formation of GATA-containing protein com- is crucially dependent on deciphering the altering the balance between specific factors plexes observed at the ALDH1A1 promoter mechanisms by which TLX1 is capable of regu- binding at the GATA box without actually bind- GATA box, despite the lack of DNA binding by lating gene expression. Future work to charac- ing DNA itself. This, at least in part, may TLX1. It is conceivable that a general ability of terize TLX1 target genes and to fully define the involve direct protein contacts by TLX1 since TLX1 to indirectly regulate target genes by protein participants involved in TLX1-mediated inclusion of anti-TLX1 antibody into EMSA altering GATA factor activity, whether via CBP gene regulation will represent important steps binding reactions reduced TLX1-associated or an alternative mechanism, may be conse- towards this goal. inhibition of one of the GATA box complexes quential for its role as an oncoprotein. This is identified. Intriguingly, the related HEX pro- particularly in view of evidence linking both a tein has been reported to interact with GATA2 blockage in T-cell development and leukemo- to inhibit its binding to the flk-1/KDR gene,58 genesis to incapacitated GATA function.62,63 References suggesting that direct antagonism of GATA fac- Interestingly, a non-DNA binding mode of tors may be a feature of NKL homeoprotein action for TLX1 that involves activity regula- 1. Holland PW, Takahashi T. The evolution of function more generally. tion of other transcription factors is reminis- homeobox genes: Implications for the We further showed that TLX1 specifically cent of another important T-ALL oncoprotein, study of brain development. Brain Res Bull interacts with TFIIB, a member of the general SCL/TAL1, which induces leukemia by recruit- 2005;66:484-90. transcriptional machinery. TFIIB, which can ing the co-repressor/histone deacetylase 2. Roberts CW, Shutter JR, Korsmeyer SJ. bind promoter DNA in a sequence-specific man- mSin3A to inhibit the transcriptional activity Hox11 controls the genesis of the spleen. ner at a TFIIB recognition element (BRE), plays of E47/HEB.64 Nature 1994;368:747-9. a central role in pre-initiation complex (PIC) There is little evidence to date showing that 3. Kanzler B, Dear TN. Hox11 acts cell assembly, providing a bridge between promoter- TLX1 can bind directly to natural DNA autonomously in spleen development and bound TFIID and RNA polymerase II. This points sequences to regulate target gene expression. its absence results in altered cell fate of to a mechanism to account for how TLX1 can Instead, TLX1 has been shown to operate indi- mesenchymal spleen precursors. Dev Biol inhibit basal transcription in a non-DNA bind- rectly by interacting with other proteins, most 2001;234:231-43. ing manner, both in this study and that of notably the phosphatases PP1 and PP2A to reg- 4. Hatano M, Roberts CW, Minden M, et al. Owens et al.49 Specifically, TLX1 may directly ulate gene cascades in various pathways such Deregulation of a homeobox gene, HOX11, affect repression of the ALDH1A1 promoter, as RB/E2F and p107/MYC.18,19 Consistent with by the t(10;14) in T cell leukemia. Science which contains a BRE adjacent to the GATA box, this paradigm, our data confirm that TLX1 is 1991;253:79-82. by inhibiting the rate/extent of PIC formation capable of regulating ALDH1A1 expression in a 5. Kennedy MA, Gonzalez-Sarmiento R, Kees via contacts with TFIIB. Notably, the TLX1-TFIIB non-DNA binding manner by affecting tran- UR, et al. HOX11, a homeobox-containing interaction was reduced in the absence of scriptional complexes at the proximal ALDH1A1 T-cell oncogene on human chromosome homeodomain helix 3. Thus our results provide promoter, although clearly, additional cis-regu- 10q24. Proc Natl Acad Sci USA 1991;88: one potential mechanism to explain the switch latory elements (that may influence TLX1 pro- 8900-4. in activity from repression to activation tein interactions) are required to recapitulate 6. Dubé ID, Raimondi SC, Pi D, Kalousek DK. observed with the TLX1ΔH3 mutant. the effect of TLX1 on endogenous ALDH1A1 A new translocation, t(10;14)(q24;q11), in Significantly, it has been recently reported gene expression. We showed that TLX1, like T cell neoplasia. Blood 1986;67:1181-4. that TLX1, like many other homeoproteins,52 other homeodomain transcription factors 7. Dubé ID, Kamel-Reid S, Yuan CC, Lu M, can form a mutually inhibitory complex with including tinman/Nkx2-5, abdominal-A, Nkx6.1 Wu X, Corpus G, et al. A novel human

[Hematology Reviews 2009; 1:e13] [page 77] Article

homeobox gene lies at the chromosome 10 20. Riz I, Akimov SS, Eaker SS, et al. 33. Chute JP, Muramoto GG, Whitesides J, et breakpoint in lymphoid neoplasias with TLX1/HOX11-induced hematopoietic dif- al. Inhibition of aldehyde dehydrogenase chromosomal translocation t(10;14). Blood ferentiation blockade. Oncogene 2007;26: and retinoid signaling induces the expan- 1991;78:2996-3003. 4115-23. sion of human hematopoietic stem cells. 8. Kees UR, Heerema NA, Kumar R, et al. 21. Dear TN, Sanchez-Garcia I, Rabbitts TH. Proc Natl Acad Sci USA 2006;103:11707-12. Expression of HOX11 in childhood T-line- The HOX11 gene encodes a DNA-binding 34. Kastan MB, Schlaffer E, Russo JE, et al. age acute lymphoblastic leukemia can nuclear transcription factor belonging to a Direct demonstration of elevated aldehyde occur in the absence of cytogenetic aberra- distinct family of homeobox genes. Proc dehydrogenase in human hematopoietic tion at 10q24: a study from the Children's Natl Acad Sci USA 1993;90:4431-5. progenitor cells. Blood 1990;75:1947-50. Cancer Group (CCG). Leukemia 2003;17: 22. Shimizu H, Kang M, Iitsuka Y, et al. 35. Storms RW, Green PD, Safford KM, et al. 887-93. Identification of an optimal Ncx binding Distinct hematopoietic progenitor com- 9. Bergeron J, Clappier E, Radford I, et al. sequence required for transcriptional acti- partments are delineated by the expres- Prognostic and oncogenic relevance of vation. FEBS Letters 2000;475:170-4. sion of aldehyde dehydrogenase and CD34. TLX1/HOX11 expression level in T-ALLs. 23. Tang S, Breitman ML. The optimal binding Blood 2005;106:95-102. Blood 2007;110:2324-30. sequence of the Hox11 protein contains a 36. Zhou X, Mao KZ. LS Bound based gene 10. Berger R, Dastugue N, Busson M, et al. predicted recognition core motif. Nucleic selection for DNA microarray data. Bio - t(5;14)/HOX11L2-positive T-cell acute lym- Acids Res 1995;23:1928-35. informatics 2005;21:1559-64. phoblastic leukemia. A collaborative study 24. Heidari M, Rice KL, Phillips JK, et al. The 37. Stirewalt DL, Meshinchi S, Kopecky KJ, et of the Groupe Français de Cytogénétique nuclear oncoprotein TLX1/HOX11 associ- al. Identification of genes with abnormal Hématologique (GFCH). Leukemia 2003; ates with pericentromeric satellite 2 DNA expression changes in acute myeloid 17:1851-7. in leukemic T-cells. Leukemia 2006;20: leukemia. Genes Chromosomes Cancer 11. Hawley RG, Fong AZ, Lu M, Hawley TS. The 304-12. 2008;47:8-20. HOX11 homeobox-containing gene of 25. Brendolan A, Ferretti E, Salsi V, et al. A 38. Rice KL, Kees UR, Greene WK. Tran- human leukemia immortalizes murine Pbx1-dependent genetic and transcription- scriptional regulation of FHL1 by hematopoietic precursors. Oncogene 1994; al network regulates spleen ontogeny. TLX1/HOX11 is dosage, cell-type and pro- 9:1-12. Development 2005;132:3113-26. moter context-dependent. Biochem 12. Hawley RG, Fong AZ, Reis MD, et al. Trans - 26. Greene WK, Bahn S, Masson N, Rabbitts Biophys Res Commun 2008;367:707-13. forming function of the HOX11/TCL3 TH. The T-cell oncogenic protein HOX11 39. Brake RL, Kees UR, Watt PM. Multiple neg- homeo box gene. Cancer Res 1997;57: 337- activates Aldh1 expression in NIH 3T3 ative elements contribute to repression of 45. cells but represses its expression in mouse the HOX11 proto-oncogene. Oncogene 13. Hough MR, Reis MD, Singaraja R, et al. A spleen development. Mol Cell Biol 1998;18: 1998;17:1787-95. model for spontaneous B-lineage lym- 7030-7. 40. Masson N, Greene WK, Rabbitts TH. phomas in IgHµ-HOX11 transgenic mice. 27. Koehler K, Franz T, Dear TN. Hox11 is Optimal activation of an endogenous gene Proc Natl Acad Sci USA 1998;95:13853-8. required to maintain normal Wt1 mRNA by HOX11 requires the NH2-terminal 50 14. Owens BM, Hawley TS, Spain LM, et al. levels in the developing spleen. Dev Dyn amino acids. Mol Cell Biol 1998;18:3502-8. TLX1/HOX11-mediated disruption of pri- 2000;218:201-6. 41. Shevchenko A, Wilm M, Vorm O, Mann M. mary thymocyte differentiation prior to the 28. Greene WK, Ford J, Dixon D, et al. Mass spectrometric sequencing of pro- CD4CD8 double-positive stage. Br J Enforced expression of HOX11 is associat- teins silver-stained polyacrylamide gels. Haematol 2006;132:216-29. ed with an immature phenotype in J2E Anal Chem 1996;68:850-8. 15. Dixon DN, Izon DJ, Dagger S, et al. erythroid cells. Br J Haematol 2002;118: 42. Gharahdaghi F, Weinberg CR, Meagher TLX1/HOX11 transcription factor inhibits 909-17. DA, et al. Mass spectrometric identifica- differentiation and promotes a non- 29. Hoffmann K, Dixon DN, Greene WK, et al. tion of proteins from silver-stained poly- hematopoietic phenotype in murine bone A microarray model system identifies acrylamide gel: a method for the removal marrow cells. Br J Haematol 2007;138:54- potential new target genes of the proto- of silver ions to enhance sensitivity. 67. oncogene HOX11. Genes Chromosomes Electrophoresis 1999;20:601-5. 16. Chen E, Kwon YT, Lim MS, et al. Loss of Cancer 2004;41:309-20. 43. Heidari M, Rice KL, Kees UR, Greene WK. UBR1 promotes aneuploidy and acceler- 30. Rice KL, Izon DJ, Ford J, et al. Over-expres- Expression and purification of the human ates B cell lymphomagenesis in HOX11- sion of stem cell associated ALDH1A1, a homeodomain oncoprotein HOX11. transgenic mice. Oncogene 2006;25:5752- target of the leukemogenic transcription Protein Exp Purif 2002;25:313-8. 63. factor TLX1/HOX11, inhibits lympho - 44. Yanagawa Y, Chen JC, Hsu LC, Yoshida A. 17. Chen E, Lim MS, Rosic-Kablar S, et al. poiesis and promotes myelopoiesis in The transcriptional regulation of human Dysregulated expression of mitotic regula- murine hematopoietic progenitors. Leuk aldehyde dehydrogenase I gene. The struc- tors is associated with B-cell lymphomage- Res 2008;32:873-83. tural and functional analysis of the pro- nesis in HOX11-transgenic mice. Onco- 31. Molotkov A, Duester G. Genetic evidence moter. J Biol Chem 1995;270:17521-7. gene 2006;25:2575-87. that retinaldehyde dehydrogenase Raldh1 45. Fong TC, Emerson BM. The erythroid-spe- 18. Kawabe T, Muslin AJ, Korsmeyer SJ. (Aldh1a1) functions downstream of alco- cific protein cGATA-1 mediates distal HOX11 interacts with protein phos- hol dehydrogenase Adh1 in metabolism of enhancer activity through a specialized phatases PP2A and PP1 and disrupts a retinol to retinoic acid. J Biol Chem 2003; beta-globin TATA box. Genes Dev 1992; G2/M cell-cycle checkpoint. Nature 1997; 278:36085-90. 6:521-32. 385:454-8. 32. Haselbeck RJ, Hoffmann I, Duester G. 46. Aird WC, Parvin JD, Sharp PA, Rosenberg 19. Riz I, Hawley RG. G1/S transcriptional net- Distinct functions for Aldh1 and Raldh2 in RD. The interaction of GATA-binding pro- works modulated by the HOX11/TLX1 onco- the control of ligand production for embry- teins and basal transcription factors with gene of T-cell acute lymphoblastic onic retinoid signaling pathways. Dev GATA box-containing core promoters. A leukemia. Oncogene 2005;24:5561-75. Genet 1999;25:353-64. model of tissue-specific gene expression. J

[page 78] [Hematology Reviews 2009; 1:e13] Article

Biol Chem 1994;269:883-9. 53. Iype T, Taylor DG, Ziesmann SM, et al. The cells inhibits flk-1/KDR-mediated vascular 47. Surinya KH, Cox TC, May BK. Trans- transcriptional repressor Nkx6.1 also func- endothelial growth factor signaling. J Biol criptional regulation of the human ery- tions as a deoxyribonucleic acid context- Chem 2004;279: 20626-35. throid 5-aminolevulinate synthase gene. dependent transcriptional activator during 59. Yamagata T, Mitani K, Oda H, et al. Identification of promoter elements and pancreatic beta-cell differentiation: evi- Acetylation of GATA-3 affects T-cell sur- role of regulatory proteins. J Biol Chem dence for feedback activation of the nkx6.1 vival and homing to secondary lymphoid 1997;272:26585-94. gene by Nkx6.1. Mol Endocrinol 2004;18: organs. EMBO J 2000;19:4676-87. 48. Tsuchiya T, Okada M, Ueda M, Yasukochi 1363-75. 60. Hayakawa F, Towatari M, Ozawa Y, et al. Y. Activation of the erythropoietin promot- 54. Yu Z, Syu LJ, Mellerick DM. Contextual Functional regulation of GATA-2 by acety- er by a point mutation from GATA to TATA interactions determine whether the Dro - lation. J Leuk Biol 2004;75:529-40. in the -30 region. J Biochem 1997;121:193- so phila homeodomain protein, Vnd, acts 61. Boyes J, Byfield P, Nakatani Y, Ogryzko V. 6. as a repressor or activator. Nucleic Acids Regulation of activity of the transcription 49. Owens BM, Zhu YX, Suen TC, et al. Res 2005;33:1-12. factor GATA-1 by acetylation. Nature 1998; Specific homeodomain-DNA interactions 55. Elizondo G, Corchero J, Sterneck E, 396:594-8. are required for HOX11-mediated transfor- Gonzalez FJ. Feedback inhibition of the 62. Ting C, Olson MC, Barton KP, Leiden JM. mation. Blood 2003;101:4966-74. retinaldehyde dehydrogenase gene ALDH1 Transcription factor GATA-3 is required for 50. Pinsonneault J, Florence B, Vaessin H, by retinoic acid through retinoic acid development of the T-cell lineage. Nature McGinnis W. A model for extradenticle receptor alpha and CCAAT/enhancer-bind- 1996;384:474-8. function as a switch that changes HOX ing protein beta. J Biol Chem 2000;275: 63. Shimizu R, Kuroha T, Ohneda O, et al. proteins from repressors to activators. 39747-53. Leukemogenesis caused by incapacitated EMBO J 1997;16:2032-42. 56. Zhang N, Shen W, Hawley RG, Lu M. GATA-1 function. Mol Cell Biol 2004;24: 51. Choi CY, Lee YM, Kim YH, et al. The home- HOX11 interacts with CTF1 and mediates 10814-25. odomain transcription factor NK-4 acts as hematopoietic precursor cell immortaliza- 64. O'Neil J, Shank J, Cusson N, et al. either a transcriptional activator or repres- tion. Oncogene 1999;18:2273-9. TAL1/SCL induces leukemia by inhibiting sor and interacts with the p300 coactivator 57. Stewart JJ, Fischbeck JA, Chen X, Stargell the transcriptional activity of E47/HEB. and the Groucho corepressor. J Biol Chem LA. Non-optimal TATA elements exhibit Cancer Cell 2004;5:587-96. 1999;274:31543-52. diverse mechanistic consequences. J Biol 65. Merabet S, Kambris Z, Capovilla M, et al. 52. Shen WF, Krishnan K, Lawrence HJ, Chem 2006;281:22665-73. The hexapeptide and linker regions of the Largman C. The HOX homeodomain pro- 58. Minami T, Murakami T, Horiuchi K, et al. AbdA Hox protein regulate its activating teins block CBP histone acetyltransferase Interaction between Hex and GATA tran- and repressive functions. Dev Cell 2003; activity. Mol Cell Biol 2001;21:7509-22. scription factors in vascular endothelial 4:761-8.

[Hematology Reviews 2009; 1:e13] [page 79] Hematology Reviews 2009; volume 1:e14

Communication between bone tions, hemopoiesis can also occur in extramedullary sites like thymus, spleen and liver. Correspondence: Sergio Dias, CIPM, IPOLFG, marrow niches in normal bone HSCs are localized in specialized microenvi- R. Professor Lima Basto, 1099-023 Lisboa, marrow function and during ronments within hematopoietic tissues called Portugal. hemopathies progression niches.2-6 Within the BM, two anatomical and E-mail: [email protected] functional niches have been proposed, the 7-10 11 Key words: bone marrow niches, hematopoietic Sara Lamorte,1 Leonor Remédio,1 endosteal niche and the perivascular niche. It has been suggested that about 60% of bone- stem cell properties, molecular signals. Sergio Dias1,2,3 marrow HSCs are adjacent to perivascular 1Centro de Investigação de Patobiologia Acknowledgments: the authors would like to niches and up to 20% of HSCs localize in the thank other members of the Angiogenesis Lab for Molecular, CIPM, Instituto Português de endosteal niches; the remaining HSCs are their suggestions. Associação Portuguesa contra 11,12 Oncologia de Lisboa Francisco Gentil, believed to be scattered throughout the BM. a Leucemia (APCL) and Liga Portuguesa Contra o E.P.E.; 2CEDOC, Faculdade de Ciências Endosteal niches, located at the inner bone Cancro (LPCC) fund our projects focusing on the Médicas, Universidade Nova de Lisboa, surface, contain quiescent HSCs, character- bone marrow microenvironment. Sara Lamorte is Lisboa, Portugal; 3Instituto Gulbenkian ized by a low proliferative rate; whereas acti- funded by a fellowship from the University of Ciência, Oeiras, Portugal vated HSCs, which undergo differentiation and Torino, Italy. Leonor Remédio is funded by an FCT ultimately mobilization to the peripheral circu- (Portuguese Government) fellowship (SFRH/BD/ lation, are in close contact to sinusoids of the 43531/2008). BM microvasculature in the perivascular Received for publication: 26 May 2009. niche10,13-18 Endosteal niches may thus repre- Abstract Accepted for publication: 27 July 2009. sent a reserve of HSCs, while perivascular Hematopoietic stem cell (HSC) chemotaxis, niches connect HSCs to the blood stream. This work is licensed under a Creative Commons adhesion, proliferation, quiescence and differ- The endosteal niche mainly comprises Attribution 3.0 License (by-nc 3.0) entiation are regulated by interactions with endosteal cells, osteoblasts and osteoclasts, while the perivascular niche contains mainly ©Copyright S. Lamorte et al., 2009 bone marrow (BM) niches. Two niches have Licensee PAGEPress, Italy. endothelial cells. Stromal cells, including retic- been identified in the adult BM: the endosteal Hematology Reviews 2009; 1:e14 (close to the bone) and the perivascular niche ular and mesenchymal cells, are common com- doi:10.4081/hr.2009.e14 (close to blood vessels). A vast body of litera- ponents of both niches. They are scattered ture has revealed the molecular basis for the throughout the trabecular space of the BM and interaction of HSCs with the two niches. surround the endothelial cells. As these cells However, the signals that regulate the commu- are a component of both endosteal and vascu- Migration of HSCs from the endosteal to the nication between the two niches have not been lar niches, they may serve as a cellular link perivascular niche is regulated by c-kit/Stem 15 well defined. Taking in consideration several between them. The cellular components of Cell Factor (SCF); CXC chemokine receptor 4 clinical and experimental arguments this the niches interact with each other to support (CXCR4)/stromal-cell derived factor-1 (SDF-1) review highlights the molecular cues, involved HSC adhesion, quiescence, chemotaxis and, in and granulocyte colony-stimulating factor (G- in the communication between the BM niches, the case of the vascular niche, differentia- CSF), pathways.7,22-24,42 Endothelial cells and 10,14,16,17,19-21 which regulate the basic properties of HSCs in tion. Thus, the HSC properties and reticular cells have been shown to produce physiological and malignant conditions. As functional responses depend on specific inter- SDF-1, generating a gradient from the perivas- such, it aims at clarifying the most important action with BM niches (Table 1). cular to the endosteal niche, which may thus advances in basic and clinical research focus- promote HSCs migration, since CXCR4 is ing on the role of different factors in the regu- Chemotaxis expressed on HSC.22-24 Mobilization of HSCs lation of the BM microenvironment. Bone marrow niches recruit hematopoietic from the endosteal to the vascular niche is stem cells essential for hemopoietic recovery following myeloablation. In this case, the soluble form of HSC chemotaxis towards the endosteal membrane stem cell factor (sSCF), released Introduction niche has been suggested to be mediated by from osteoblasts after cleavage by SDF-1- osteoponin (Opn) and calcium ion concentra- induced matrix metalloproteinase-9, promotes 2+ 25 tion ([Ca ]o). Opn, a glycoprotein expressed HSC homing to the perivascular niche by inter- Hematopoietic stem cells reside in bone on endosteal bone surface by osteoblasts, pro- acting with its receptor c-Kit.7,22 G-CSF, pro- marrow niches, which regulate their fate motes HSC migration, as shown in vivo stud- duced by osteoblasts, promotes the mobiliza- -/- Hematopoietic stem cells (HSCs) are self- ies with Opn mice. In these mice, there is a tion of HSCs into the peripheral blood by up- renewing cells which give rise to all types of long-term engraftment defect after transplan- regulating CXCR4 expression on HSCs and – – + + mature blood cells. HSCs can be subdivided tation with wild-type Lineage Sca 1 c-Kit cells decreasing SDF-1 expression in the BM. G- -/- into long-term HSCs (LT-HSCs) and in short- and a compromised ability of the Opn BM CSF, in fact, induces the expression of prote- term (ST-HSCs). LT-HSCs can give rise to all microenvironment to sustain hematopoiesis. olytic enzymes such as elastase, cathepsin G, blood lineages and have unlimited self-renew- These effects seem to be indirect, since there MMP-2, and MMP-9, which cleave SDF-1.42,43 al capacity. LT-HSCs produce ST-HSCs which is no evidence, in vitro, of a chemotactic role are still multipotent but with limited self- for Opn on HSCs. Moreover, the high extracel- Adhesion renewal capacity. ST-HSCs differentiate fur- lular [Ca2+]o, maintained by the osteoclasts ther into lineage-committed progenitor cells activity, promotes HSC localization to the Bone marrow niche promotes hematopoietic which are responsible for the large-scale pro- endosteal niche, through calcium-sensing stem cells adhesion duction of mature blood cells.1 receptor (CaR): CaR-/- HSCs show a defect in HSC adhesion to the endosteal niche is reg- The bone marrow (BM) is the major site of the binding to collagenase I present at the ulated by different molecular interactions adult hemopoiesis, but, in pathological condi- bone endosteal surface. including N-cadherin/β-catenin; Tie-2/Angio-

[page 80] [Hematology Reviews 2009; 1:e14] Article poietin-1 (Ang-1); Osteopontin (Opn)/ß1 integrin; Annexin II (Anxa2)/Anxa2 receptor Table 1. HSC properties are regulated by molecular cues conveyed by the bone marrow endosteal and vascular niches. (Anxa2r) and CaR-collagen I pathways.10,25,28,33 The asymmetrical distribuition of N- HSC properties Molecular interactions cadherin/β-catenin on the cell surface of HSCs HSC-NICHE and osteoblasts, respectively, and, in particu- Endosteal niche Perivascular niche lar, the localization of these molecules at the Chemotaxis CXCR4/SDF-122-24 site of interaction of LT-HSC with spindle- 25 22 + – Opn G-CSF shaped N-cadherin CD45 osteoblastic (SNO) CaR/Ca2 cells, suggested a role for N-cadherin/ß- c-Kit/SCF22 catenin in HSCs adhesion on the endosteal niches.10 Studies performed by Kiel and collab- 10 26 orators failed to show significant numbers of Adhesion N-cadherin/ß-catenin α4ß1 integrin/VCAM1 Tie2/Ang110 4 integrins/E-selectin27 N-cadherin expressing HSCs, questioning α ß1 integrin/Opn25 whether HSC adhesion to osteoblasts is medi- Anxa2/Anxa2r 28 ated by N-cadherin.44 CaR/Collagen I25 Tie2, a receptor tyrosine kinase expressed Quiescence / proliferation MPL/THPO 29-32 by a small fraction of BM cells highly enriched Tie2/Ang133,34 for HSC activity in adult murine BM, binds its Notch1/Jagged135,36 wnt/β catenin47,48 ligand, Ang-1, expressed by osteoblasts at the Opn25,29 14,37 surface of trabecular bone.33 Regarding Opn, PTH/PTHr its expression is restricted to the endosteal Differentiation FGF-413 38 bone surface and contributes to HSCs adhe- SDF-1/CXCR4 α4ß1 integrin/VCAM13 sion to the endosteal region via ß1 integrin 13 25 VE-Cadherin expressed by HSC. Osteoblasts also express Notch1/Delta39-41 high levels of Anxa2, a calcium-dependent phospholipid-binding protein, and it has been shown, both in vitro and in vivo, that Anxa2 of positive regulators, such as c-myc.46 This ulates c-myc mRNA expression through a regulates HSCs homing and binding to the pathway is also involved in promoting HSCs PI3K- and MAPK-dependent pathway, thereby endosteal niche, through the binding to its lig- proliferation in the perivascular niche.30-32 promoting HSC proliferation.30-32 Wnt proteins and Anxa2r.28 Thus, THPO/MPL pathway exerts distinct func- are expressed by BM stromal cell and exposure Adhesion of HSCs to the perivascular niche tions on HSC, depending on cell localization. to Wnt was shown to stimulate proliferation is mediated by 4 1 integrin/vascular cell adhe- α β Opn/OpnR, instead, contributes to the mainte- and self-renewal of HSCs in vitro.47,48 sion molecule1 (VCAM1) and α4/E-selectin nance of HSC quiescence either by inhibiting, interaction.26,27,45 4 1 integrins, expressed by α β in a dose-dependent manner, the entry into cell Differentiation HSCs, interact with VCAM-1, constitutively cycle and/or by reducing cell apoptosis.25,29 A 26 expressed on BM endothelial cells. Since mouse genetic model, in which the gene PTHr Perivascular niches mediate hematopoietic inactivation of E-selectin and α4 integrin is constitutively active in osteoblasts, showed stem cell differentiation reduces drastically hematopoietic progenitor an increase in HSCs along with osteoblasts. Differentiation of HSCs occurs only in the and stem cell (HPSC) homing into lethally Moreover, there was high expression of Notch perivascular niches and is mediated by FGF-4; irradiated mice, it has been proposed that E- I ligand, Jagged, on osteoblasts, suggesting SDF-1; VCAM-1/α4ß ; VE-cadherin and Notch1 selectin ligands and α4 integrin cooperate in that the PTH/PTHr pathway can promote HSC pathway.13,38,39,49 SDF-1 is necessary for HSC adhesion to perivascular niches.27 proliferation through activation of Notch.14,37 myelopoiesis and B-lymphopoiesis, as shown Several gain- and loss-of-function experiments by the severe reduction of B-lymphopoiesis Proliferation versus quiescence of Notch target genes and ligands have sug- and lack of BM myelopoiesis in CXCR4- and Endosteal niches promote HSC quiescence gested a role for Notch in HSC quiescence and SDF-1 deficient mice.49 SDF-1 and FGF-4 pro- 35 The balance between HSC proliferation and self-renewal. However, recently Maillard et al. mote megakaryocyte maturation and platelet quiescence is likewise regulated by several have demonstrated rather conclusively that production: FGF-4 supports the adhesion of pathways. In the endosteal niche several inter- inactivation of the Notch pathway in HSCs megakaryocytes to sinusoidal BM endothelial actions, involved in the maintenance of HSC does not interfere with their self-renewal; cells (BMECs), thereby enhancing their sur- quiescence, have been identified: Tie- transplantation of hematopoietic progenitors vival and maturation, while SDF-1 augments 2/Angiopoietin-1 (Ang-1); thrombopoietin with inhibited Notch signaling induced stable platelet production by promoting their migra- long-term reconstitution of irradiated hosts 13,38 (THPO)/MPL; Opn/OpnR; parathyroid hormone tion across BMECs. VCAM-1 enhances the and a normal frequency of progenitor fractions + (PTH)/PTH receptor (PTHr) and interaction of α4ß1 integrin megakaryocytes enriched for LT-HSCs.36 Notch1/Jagged1.14,25,29,33-35,37,46 Tie2, which is with BMECs. VE-cadherin is essential for expressed by SP-HSCs, binds Ang-1+ expressed Perivascular niches promote hematopoietic VCAM-1 expression in BMECs, which in turn is on osteoblasts and induces HSC quiescence.33,34 stem cells proliferation and self-renewal required for FGF-4 mediated adhesion and LT-HSCs expressing MPL, the THPO receptor, In the vascular niche, HSC proliferation is SDF-1-induced transendothelial migration of are closely associated with THPO-producing associated with (THPO)/c-mpl and Wnt/β megakaryocytes. Neutralizing antibodies to osteoblasts. The THPO/MPL pathway is catenin pathway. THPO is expressed on BM VE-cadherin decrease the localization of involved in HSC quiescence through activation stromal cells and acts synergistically with ery- megakaryocytes to the vascular niche and dis- of genes coding for negative regulators of cell thropoietin to promote erythroid progenitors rupt megakaryocyte maturation and throm- cycle, such as p12Cip1 and p57Kip2, and inhibition and megakaryocytes proliferation. THPO stim- bopoiesis.13 Notch1 seems to provide a key reg-

[Hematology Reviews 2009; 1:e14] [page 81] Article ulatory signal in determining T- versus B-lym- sion of hemopathies. In chronic myeloid ALL cells beneath BM fibroblast layers: both phoid lineage commitment. Mice transplanted leukemia (CML), myelodysplastic syndromes cell types produce SDF-1, thereby enhancing with BM, transfected for retroviruses encoding (MDS) and multiple myeloma (MM) circulat- the adhesion molecules involved in the migra- a constitutively active form of Notch1, three ing endothelial cells (CECs), mobilized from tion and homing of these cells to the BM.72,73 weeks after transplantation showed immature the BM, share chromosomal aberrations with CD4+CD8+ T cells in the BM and a block in the malignant hematopoietic cells.57-59 These Adhesion 39 malignant CECs suggest the presence of aber- early B-cell lymphopoiesis. Notch1 activation Aberrant niches mediate cell-adhesion- rant niches in the BMM. Moreover, in B-cell seems to be driven by Delta-1-expressing stro- mediated drug resistance (CAM-DR) 40,41 lymphomas, identical genetic aberration could mal cells. It has been demonstrated, in vitro and in be found both in malignant cells and in the vivo, that cell-cell adhesion between hemato- Hemopathies require the support of aber- microvascular BM endothelial cells.60,61 poietic cells and components of the BM niches, rant bone marrow niches Irradiation and chemotherapy can change such as stromal cells, is involved in drug resist- Several hemopathies are characterized by a the BMM inducing hematopoietic and ance in AML.74,75 AML resistance to chemother- pre-malignant phase that progresses to a endothelial injury and allowing cells, proteins apy seems to be promoted by the adhesion- malignant phase. The molecular basis of this and cytokines to move between the vascular dependent secretion of WNT antagonists by progression remains poorly understood. The and endosteal niches.62 Radiation-induced osteoblasts.76 CAM-DR is mediated by integrins data in the literature suggest the likelihood of injury can also contribute to cell damage in the α4 and β1, as shown in MM, CML and AML cell such progression is very low and a malignant microenvironment in an indirect way, as a con- lines.77-79 Direct correlation has been found clone can remain “stable” for years. Moreover, sequence of an inflammatory-type response.63 between the expression of integrins that medi- in several diseases both phases are character- Moreover, it has been shown that ionizing irra- 50,51 ate adhesion to FN and drug resistance. ized by virtually the same genetic changes. diation results in altered osteoblast differenti- Coculture of ALL cells lines with BM stroma Taking in consideration these aspects, it is ation ability of BM mesenchymal stem cells, cells (BMSCs) resulted in reduced apoptosis legitimate to speculate that the genetic destruction of the endosteal niche and conse- induced by etoposide. In this stroma model, changes are necessary for the immortalization quently hematopoietic injury.64 Another possi- drug resistance required direct cell-cell con- of a malignant clone but insufficient to pro- bility is that malignant cells through direct and tact, since it could not be conferred by the addi- mote the progression to a malignant phase. So indirect signaling can modify the features of tion of stromal conditioned media.80 Moreover, other factors must take part in the progression. the vascular niche. For example, factors pro- The role of the hematopoietic BM microen- the presence of BMSCs during treatment of duced by acute lymphoblastic leukemia (ALL) myeloma cell lines significantly decreases the vironment in malignant progression has been cells can induce proliferation, migration and studied extensively and its importance was apoptosis during exposition of mitoxantrone, morphogenesis of human BM vascular an inhibitor of topoisomerase II.81 Notch-1 sig- well illustrated in recent, in vivo, studies. endothelial cells.65-67 The tumor-derived factor Widespread inactivation of retinoblastoma pro- naling seems to be involved in protection of VEGF and tumor necrosis factor-α (TNF-α) tein (Rb) resulted in myeloproliferative dis- MM from drug-induced apoptosis: overexpres- produced in the tumor microenvironment have sion of Notch-1 in Notch-1(-) myeloma cells ease, characterized by extramedullary hemo- been shown to modify the phenotype of poiesis and increased mobilization and differ- up-regulated p21 and resulted in protection endothelial cells inhibiting ICAM-1 and VCAM- 82 entiation of HSCs from the BM. The phenotype from drug-induced apoptosis. BM niches may 1 clustering on endothelial surfaces with provide a survival advantage for malignant was not recapitulated upon inactivation of Rb implications for immune-cell trafficking.68 in HSCs maintained in wild-type environ- cells following initial drug exposure and facili- Moreover, our own data suggests TNF-α is cru- 52,53 + Δ Δ tate the acquisition of acquired drug resist- ment. Moreover, Mx-Cre Pten / ∆mice cial for the onset and also for the progression develop rapid and aggressive myeloprolifera- ance, determining disease relapse following of BM dysfunction, such as in MDS (Cachaco chemotherapy. tion that progressed to leukemia in 4-5 weeks et al., 2009, unpublished data). post deletion. When Pten deletion was active in The possible mechanisms by which aberrant Aberrant niches show impaired adhesive the context of a wild-type BM microenviron- BM niches modify HSCs properties are dis- capacity, leading to a loss of quiescence ment, phenotypic and functional HSCs were cussed. and consequently to expansion of lost without evidence of myeloproliferation or malignant hematopoietic stem cells transformation.54,55 Finally, BM from wild-type Migration/chemotaxis It has been hypothesized that HSC mobiliza- mice transplanted into mice with a deficient tion results from impaired adhesion to BM retinoic acid receptor γ (RARγ) microenviron- Aberrant niches may promote recruitment niches, allowing their migration into the ment rapidly develop myeloproliferative syn- of malignant hematopoietic stem cells peripheral blood, spleen and other extra- dromes (MPS).56 These results strongly sup- The perivascular niche expresses unique medullary sites. This could explain the port the notion that the progression of the combinations of cell adhesion molecules increase in circulating CD34+ cells reported in hemopathies is not entirely cell autonomous and/or chemokines capable of attracting malig- primary myelofibrosis (PMF) patients.83,84 but depends on interactions between malignant HSCs. For example, it has been shown, in The impaired adhesion could be explained nant cells and the BM microenvironment vitro and in vivo, that E-selectin and SDF-1 are by several mechanisms. (BMM). As described above, BM niches sup- expressed in vascular “hot spots” correspon- Altered expression of membrane adhesion port HSC properties such as adhesion, quies- ding to the regions that attract leukemic molecules and integrins. For example, HSCs of cence, chemotaxis and differentiation, and cells.69-71 Distruption of the interaction between CML patients have reduced adhesion mole- regulate the balance between self-renewal and SDF-1 and its receptor CXCR4 inhibits the cules expression including L-selectin, CD44 differentiation. The idea outlined in this homing of Nalm-6 cells, an acute lymphoblastic and N-cadherin. This decrease correlates with, review is that alteration of the two BM niches, leukemia cell line, to the vascular niche.71 in vitro, reduced adhesive capacity of HSCs triggered by the aberrant expression of key These observations raise the possibility that E- from CML patients.85 molecules or cellular cues between the selectin and /or SDF-1 can regulate malignant A disruption of CXCR4/SDF-1 axis. In idio- endosteal and the perivascular niche, impairs cell homing. Moreover, BM endothelial and pathic myelofibrosis (IM) the constitutive HSC responses, contributing to the progres- stromal cells seem involved in the migration of mobilization of CD34+ cells could be the conse-

[page 82] [Hematology Reviews 2009; 1:e14] Article quence of the creation of a proteolytic sensitive to anti-VEGF and VEGF receptor the receptivity of the endosteal niche.112 PTH, microenvironment within the BMM. It has treatments.102-105 The expanded BM endotheli- acting also on the perivascular niche, can be been shown that malignant cells and the BMM um may support malignant HSC growth by pro- used for the treatment of ischemic vascular produce metalloproteinase.86-88 Thus, the tecting them from chemotherapy-induced disease.113 Moreover, it has recently been increased production of metalloproteinase-9 apoptosis and/or promoting their proliferation shown that pharmacological use of PTH might disrupt adhesive interaction between in a paracrine way through the release of fac- increases the number of HSCs mobilized into CD34+ HSCs and BM niches through degrada- tors such as G-CSF, IL-10, IL-6 and vascular the peripheral blood for stem cell harvests, tion of SDF-1 or cleavage of its receptor endothelial growth factor-C (VEGF-C).106,107 protects stem cells from repeated exposure to CXCR4, leading to the release of the HSCs into Aberrant vascular niches can induce cytotoxic chemotherapy and expands stem the peripheral blood.22,89 114 quiescence in malignant cells playing a role cells in transplant recipients. in tumor maintenance The second strategy is to abrogate the Proliferation vs. quiescence Adhesion of malignant HSCs to BMSCs may interaction between malignant HSCs and BM niches, by blocking their physical binding or Aberrant niches determine an imbalance induce quiescence by inhibiting cell prolifera- the growth factors secreted by the BMM. As between proliferation and quiescence, tion. For example, Notch-1 activation in MM described before, the chemokine axis SDF- accelerating the onset and progression of cells, after incubation on BMSCs, results in the 1/CXCR4 is involved in the retention of HSCs malignancy accumulation of the cells in G0/G1 phase of cell 82,108 within the BM. Thus, destruction of this inter- BM cells display a different set of adhesion cycle. Aberrant niches may thus contribute action allows the mobilization of HSCs from molecules, extracellular matrix elements, to the maintenance of a malignant pool of the BM to the peripheral blood. This approach growth factors and chemokines. Spleen fibrob- HSCs. has been established clinically using G-CSF or lasts isolated from PMF patients, in contrast to antibody against CXCR4.115-117 The combination primary fibroblasts purified from the spleen of Differentiation of both result in an enhancement of HSC healthy subjects, are able to support the prolif- Aberrant niches can induce malignant mobilization from the BM.118 The proteasome eration of autologous patient CD34+ cells, but transformation of normal hematopoietic inhibitor PS-341, currently used in MM thera- not that of their normal counterparts.90 stem cells py, blocks the growth of MM cells by decreas- Moreover, it has been shown that somatic The donor cell leukemia (DCL), a hemopa- ing their adherence to BMSCs and the related mutations that occur in BM stromal cells, such thy following hemopoietic cell transplation, is 119 as p53 mutations, render these cells supportive protection against drug-induced apoptosis. apparently the result of malignant transforma- Another strategy is the inhibition of TNF- of ALL growth.81 Finally, aberrant vascular nich- α tion of normal donor hematopoietic cells in the production by BM cells, with a monoclonal es produce several factors, such as VEGF; IL-6; 109 transplant recipient. One of the hypotheses antibody against the extracellular domain of granulocyte-macrophage and granulocyte is that the host microenvironment in which TNF-α, called infliximab. Two studies have colony-stimulating factors, that are able to sup- the original malignancy developed may trigger port malignant hemopoiesis.91-93 For example, it investigated the use of infliximab in patients malignant transformations in donor cells, with low-risk MDS. In both reports, the drug has been shown that coculture of AML cells favored by the immunocompromised status showed a limited but significant activity and with microvascular endothelial cells increases after transplantation and by perturbation of no particular side-effects.120 proliferation and inhibits apoptosis of AML the host BMM following multiple rounds of 93 Another recently approved therapeutic cells. chemotherapy. approach involves inhibiting angiogenesis; Providing self-renewing and proliferative Studies in Drosophila Melanogaster, by Kai several inhibitors of VEGF are currently used cues to malignant HSCs. ALL stromal cells reg- et al., suggest that a vacant niche can engage in the treatment of different hemopathies.104, 121 ulate self-renewal and proliferation of a ectopic cells, normal hematopoietic and non- + The concept behind most of these thera- Philadelphia-chromosome positive (Ph )/VE- hematopoietic cells, with a resultant change in + peutic approaches implies that to increase cadherin subpopulation of leukemia cells by phenotype. Depending on the specific system, promoting the expression of VE-cadherin, sta- it seems that non-stem cells can acquire either therapeutic efficacy it is necessary to use a bilizing β catenin and up-regulating BCR-abl a more proliferative phenotype or revert to a strategy in which the seed (malignant HSCs) transcripts.94 This way, due to the stromal sup- stem cell-like condition. These findings and the soil (altered BMM) must be targeted port, malignant cells circumvent the require- strongly support the possibility that BM niches simultaneously. ment of exogenous Wnt signaling for self- can contribute to hemopathies, inducing aber- renewal. Human MM cells also become inde- rant transformation of normal cells, including pendent of the IL-6/gp130/STAT3 survival path- HSCs.110,111 way when cocultured in the presence of Conclusions BMSCs.95 This evidence confirms the idea that Bone marrow niches as therapeu- BMSCs can provide alternative survival and tic target This paper highlights the key data demon- proliferative signals to BM malignant cells. Based on the idea that the BMM has a rele- strating that changes in the signals delivered Angiogenesis, the branching of new vant role in the progression of hemopathies, by BM endosteal and/or perivascular niches microvessels from pre-existent blood vessels, novel therapeutic approaches are being devel- may lead to an impairment of survival, differ- is kept at set point in which there is a balance oped to revert the malignant phenotype by tar- entiation and proliferation of HSCs. Thus, between pro- and anti-angiogenic molecules. geting environmental cues. The strategies aberrant BM niches participating in HSC regu- The angiogenic switch, unbalanced set point used until now can be summarized into three lation contribute in a crucial way to the pro- in favor of pro-angiogenic molecules, favors categories. gression of hemopathies. Therefore, the the production of new microvessels.96 The first strategy is to modify the niche molecular cues that contribute towards BM Increased angiogenesis has been described in itself. For example, Ballen and colleagues niches alteration during the onset and devel- a number of hemopathies.97-101 The extent of have tested the hypothesis to use parathyroid opment of hemopathies represent a new chal- BM neo-vessel formation correlates also with hormone (PTH) to augment the engraftment lenging therapeutic target. patient prognosis and these hemopathies are efficiency of cord blood transplant, modifying

[Hematology Reviews 2009; 1:e14] [page 83] Article

etic stem cell niche. Nature 2003;425: 841-6. 30. Chanprasert S, Geddis AE, Barroga C, Fox References 15. Kiel MJ, Morrison SJ. Uncertainty in the NE, Kaushansky K. Thrombopoietin (TPO) niches that maintain haematopoietic stem induces c-myc expression through a PI3K- 1. Zhong JF, Zhao Y, Sutton S, et al. Gene cells. Nat Rev Immunol 2008;8:290-301. and MAPK-dependent pathway that is not expression profile of murine long-term 16. Kopp HG, Avecilla ST, Hooper AT, Rafii S. mediated by Akt, PKCzeta or mTOR in TPO- reconstituting vs. short-term reconstituting The bone marrow vascular niche: home of dependent cell lines and primary megakary- hematopoietic stem cells. Proc Natl Acad Sci HSC differentiation and mobilization. ocytes. Cell Signal 2006;18:1212-8. USA 2005;102:2448-53. Physiology (Bethesda, Md.) 2005;20:349-56. 31. Kaushansky K, Broudy VC, Grossmann A, et 2. Dexter TM, Wright EG, Krizsa F, Lajtha LG. 17. Rafii S, Mohle R, Shapiro F, et al. Regulation al. Thrombopoietin expands erythroid pro- Regulation of haemopoietic stem cell prolif- of hematopoiesis by microvascular endothe- genitors, increases red cell production, and eration in long term bone marrow cultures. lium. Leuk Lymphoma 1997; 27:375-86. enhances erythroid recovery after myelo- Biomedicine/[publiee pour l'AAICIG] 1977; 18. Wilson A, Oser GM, Jaworski M, et al. suppressive therapy. J Clin Invest 1995;96: 27:344-9. Dormant and self-renewing hematopoietic 1683-7. 3. Islam A, Catovsky D, Goldman JM, Galton stem cells and their niches. Ann N Y Acad 32. Nagahisa H, Nagata Y, Ohnuki T, et al. Bone DA. Histomorphological study of cellular Sci 2007;1106:64-75. marrow stromal cells produce thrombopoi- interactions between stromal and haemo- 19. Adams GB, Scadden DT. The hematopoietic etin and stimulate megakaryocyte growth poietic stem cells in normal and leukaemic stem cell in its place. Nat Immunol 2006; and maturation but suppress proplatelet for- bone marrow. Histopathology 1984;8:293- 7:333-7. mation. Blood 1996;87:1309-16. 313. 20. Arai F, Hirao A, Suda T. Regulation of hema - 33. Arai F, Hirao A, Ohmura M, et al. Tie2/an gio - 4. Lord BI, Testa NG, Hendry JH. The relative topoietic stem cells by the niche. Trends poietin-1 signaling regulates hematopoietic spatial distributions of CFUs and CFUc in Cardiovasc Med 2005;15:75-9. stem cell quiescence in the bone marrow the normal mouse femur. Blood 1975;46: 65- 21. Yin T, Li L. The stem cell niches in bone. J niche. Cell 2004;118:149-61. 72. Clin Invest 2006;116:1195-201. 34. Puri MC, Bernstein A. Requirement for the 5. Nilsson SK, Johnston HM, Coverdale JA. 22. Heissig B, Hattori K, Dias S, et al. Recruit - TIE family of receptor tyrosine kinases in Spatial localization of transplanted hemo- ment of stem and progenitor cells from the adult but not fetal hematopoiesis. Proc Natl poietic stem cells: inferences for the local- bone marrow niche requires MMP-9 medi- Acad Sci USA 2003;100:12753-8. ization of stem cell niches. Blood 2001;97: ated release of kit-ligand. Cell 2002;109: 35. Duncan AW, Rattis FM, DiMascio LN, et al. 2293-9. 625-37. Integration of Notch and Wnt signaling in 6. Schofield R. The relationship between the 23. Mohle R, Bautz F, Rafii S, et al. The chemo - hematopoietic stem cell maintenance. Nat spleen colony-forming cell and the haemo- kine receptor CXCR-4 is expressed on Immunol 2005;6:314-22. poietic stem cell. Blood Cells 1978; 4:7-25. CD34+ hematopoietic progenitors and 36. Maillard I, Koch U, Dumortier A, et al. 7. Kollet O, Dar A, Shivtiel S, et al. Osteoclasts leukemic cells and mediates transendothe- Canonical notch signaling is dispensable for degrade endosteal components and promote lial migration induced by stromal cell- the maintenance of adult hematopoietic mobilization of hemato-poietic progenitor derived factor-1. Blood 1998;91:4523-30. stem cells. Cell Stem Cell 2008;2:356-66. cells. Nat Med 2006;12:657-64. 24. Sugiyama T, Kohara H, Noda M, Nagasawa 37. Stier S, Cheng T, Dombkowski D, et al. 8. Kubota Y, Takubo K, Suda T. Bone marrow T. Maintenance of the hematopoietic stem Notch1 activation increases hematopoietic long label-retaining cells reside in the sinu- cell pool by CXCL12-CXCR4 chemokine sig- stem cell self-renewal in vivo and favors soidal hypoxic niche. Biochem Biophys Res naling in bone marrow stromal cell niches. lymphoid over myeloid lineage outcome. Commun 2008;366:335-9. Immunity 2006;25:977-88. Blood 2002;99:2369-78. 9. Taichman RS, Reilly MJ, Emerson SG. 25. Nilsson SK, Johnston HM, Whitty GA, et al. 38. Lane WJ, Dias S, Hattori K, et al. Stromal- Human osteoblasts support human hema- Osteopontin, a key component of the derived factor 1-induced megakaryocyte topoietic progenitor cells in vitro bone mar- hematopoietic stem cell niche and regulator migration and platelet production is row cultures. Blood 1996;87:518-24. of primitive hematopoietic progenitor cells. dependent on matrix metalloproteinases. 10. Zhang J, Niu C, Ye L, Huang H, He X, Tong Blood 2005;106:1232-9. Blood 2000;96: 4152-9. WG, et al. Identification of the haemato-poi- 26. Papayannopoulou T, Priestley GV, Naka- 39. Pui JC, Allman D, Xu L, et al. Notch1 expres- etic stem cell niche and control of the niche moto B, et al. Molecular pathways in bone sion in early lymphopoiesis influences B size. Nature 2003;425:836-41. marrow homing: dominant role of alpha(4) versus T lineage determination. Immunity 11. Kiel MJ, Yilmaz OH, Iwashita T, Yilmaz OH, beta(1) over beta(2)-integrins and select - 1999;11:299-308. Terhorst C, Morrison SJ. SLAM family recep- ins. Blood 2001;98:2403-11. 40. de La Coste A, Six E, Fazilleau N, et al. In tors distinguish hematopoietic stem and 27. Katayama Y, Hidalgo A, Furie BC, et al. vivo and in absence of a thymus, the progenitor cells and reveal endothelial nich- PSGL-1 participates in E-selectin-mediated enforced expression of the Notch ligands es for stem cells. Cell 2005;121:1109-21. progenitor homing to bone marrow: evi- delta-1 or delta-4 promotes T cell develop- 12. Watchman CJ, Bourke VA, Lyon JR, et al. dence for cooperation between E-selectin ment with specific unique effects. J Spatial distribution of blood vessels and ligands and alpha4 integrin. Blood 2003; Immunol 2005;174:2730-7. CD34+ hematopoietic stem and progenitor 102:2060-7. 41. Jaleco AC, Neves H, Hooijberg E, et al. cells within the marrow cavities of human 28. Jung Y, Wang J, Song J, et al. Annexin II Differential effects of Notch ligands Delta-1 cancellous bone. J Nucl Med 2007;48:645-54. expressed by osteoblasts and endothelial and Jagged-1 in human lymphoid differenti- 13. Avecilla ST, Hattori K, Heissig B, et al. cells regulates stem cell adhesion, homing, ation. J Exp Med 2001;194:991-1002. Chemokine-mediated interaction of hema - and engraftment following transplantation. 42. Semerad CL, Christopher MJ, Liu F, et al. G- to poietic progenitors with the bone marrow Blood 2007;110:82-90. CSF potently inhibits osteoblast activity and vascular niche is required for thrombo - 29. Stier S, Ko Y, Forkert R, et al. Osteopontin is CXCL12 mRNA expression in the bone mar- poiesis. Nat Med 2004;10:64-71. a hematopoietic stem cell niche component row. Blood 2005;106:3020-7. 14. Calvi LM, Adams GB, Weibrecht KW, et al. that negatively regulates stem cell pool size. 43. Petit I, Szyper-Kravitz M, Nagler A, et al. G- Osteoblastic cells regulate the haemato-poi- J Exp Med 2005;201:1781-91. CSF induces stem cell mobilization by

[page 84] [Hematology Reviews 2009; 1:e14] Article

decreasing bone marrow SDF-1 and up-reg- 58. Fang B, Zheng C, Liao L, et al. Identi- 72. Bradstock KF, Makrynikola V, Bianchi A, et ulating CXCR4. Nat Immunol 2002;3: 687- fication of human chronic myelogenous al. Effects of the chemokine stromal cell- 94. leukemia progenitor cells with heman- derived factor-1 on the migration and local- 44. Kiel MJ, Radice GL, Morrison SJ. Lack of evi- gioblastic characteristics. Blood 2005;105: ization of precursor-B acute lymphoblastic dence that hematopoietic stem cells depend 2733-40. leukemia cells within bone marrow stromal on N-cadherin-mediated adhesion to 59. Zhang H, Vakil V, Braunstein M, et al. layers. Leukemia 2000; 14:882-8. osteoblasts for their maintenance. Cell stem Circulating endothelial progenitor cells in 73. Shen W, Bendall LJ, Gottlieb DJ, Bradstock cell 2007;1:204-17. multiple myeloma: implications and signifi- KF. The chemokine receptor CXCR4 45. Scott LM, Priestley GV, Papayannopoulou T. cance. Blood 2005;105:3286-94. enhances integrin-mediated in vitro adhe- Deletion of alpha4 integrins from adult 60. Rigolin GM, Fraulini C, Ciccone M, et al. sion and facilitates engraftment of leukemic hematopoietic cells reveals roles in home- Neoplastic circulating endothelial cells in precursor-B cells in the bone marrow. Exp ostasis, regeneration, and homing. Mol Cell multiple myeloma with 13q14 deletion. Hematol 2001;29:1439-47. Biol 2003;23: 9349-60. Blood 2006;107:2531-5. 74. Hazlehurst LA, Landowski TH, Dalton WS. 46. Yoshihara H, Arai F, Hosokawa K, et al. 61. Streubel B, Chott A, Huber D, et al. Role of the tumor microenvironment in Thrombopoietin/MPL signaling regulates Lymphoma-specific genetic aberrations in mediating de novo resistance to drugs and hematopoietic stem cell quiescence and microvascular endothelial cells in B-cell physiological mediators of cell death. interaction with the osteoblastic niche. Cell lymphomas. N Engl J Med 2004;351:250-9. Oncogene 2003;22:7396-402. stem cell 2007;1:685-97. 62. Wright EG. Ionizing radiation and leu - 75. Matsunaga T, Takemoto N, Sato T, et al. 47. Van Den Berg DJ, Sharma AK, Bruno E, kaemia: more questions than answers. Interaction between leukemic-cell VLA-4 Hoffman R. Role of members of the Wnt Hematol Oncol 2005;23:119-26. and stromal fibronectin is a decisive factor gene family in human hematopoiesis. Blood 63. Barcellos-Hoff MH, Park C, Wright EG. for minimal residual disease of acute myel- 1998;92:3189-202. Radiation and the microenvironment - ogenous leukemia. Nat Med 2003;9: 1158- 48. Willert K, Brown JD, Danenberg E, et al. Wnt tumorigenesis and therapy. Nature Rev 65. proteins are lipid-modified and can act as 2005;5:867-75. 76. De Toni F, Racaud-Sultan C, Chicanne G, et stem cell growth factors. Nature 2003;423: 64. Ma J, Shi M, Li J, et al. Senescence-unrelat- al. A crosstalk between the Wnt and the 448-52. ed impediment of osteogenesis from Flk1+ adhesion-dependent signaling pathways 49. Ma Q, Jones D, Borghesani PR, et al. bone marrow mesenchymal stem cells governs the chemosensitivity of acute Impaired B-lymphopoiesis, myelopoiesis, induced by total body irradiation and its con- myeloid leukemia. Oncogene 2006;25: 3113- and derailed cerebellar neuron migration in tribution to long-term bone and hematopoi- 22. CXCR4- and SDF-1-deficient mice. Proc etic injury. Haematologica 2007; 92:889-96. 77. Damiano JS, Cress AE, Hazlehurst LA, Shtil Natl Acad Sci USA 1998;95:9448-53. 65. Dilly SA, Jagger CJ. Bone marrow stromal AA, Dalton WS. Cell adhesion mediated drug 50. Kyle RA, Therneau TM, Rajkumar SV, et al. A cell changes in haematological malignan- resistance (CAM-DR): role of integrins and long-term study of prognosis in monoclonal cies. J Clin Pathol 1990;43:942-6. resistance to apoptosis in human myeloma gammopathy of undetermined significance. 66. Sudhoff T, Wehmeier A, Kliche KO, et al. cell lines. Blood 1999; 93:1658-67. N Engl J Med 2002;346:564-9. Levels of circulating endothelial adhesion 78. Damiano JS, Hazlehurst LA, Dalton WS. Cell 51. Ma X, Does M, Raza A, Mayne ST. Myelo-dys- molecules (sE-selectin and sVCAM-1) in adhesion-mediated drug resistance (CAM- plastic syndromes: incidence and survival in adult patients with acute leukemia. Leu - DR) protects the K562 chronic myelogenous the United States. Cancer 2007;109:1536-42. kemia 1996;10:682-6. leukemia cell line from apoptosis induced by 52. Walkley CR, Orkin SH. Rb is dispensable for 67. Veiga JP, Costa LF, Sallan SE, et al. Leu- BCR/ABL inhibition, cytotoxic drugs, and self-renewal and multilineage differentia- kemia-stimulated bone marrow endotheli- gamma-irradiation. Leukemia 2001;15: tion of adult hematopoietic stem cells. Proc um promotes leukemia cell survival. Exp 1232-9. Natl Acad Sci USA 2006;103:9057-62. Hematol 2006;34: 610-21. 79. Hazlehurst LA, Dalton WS. Mechanisms 53. Walkley CR, Shea JM, Sims NA, et al. Rb reg- 68. Bouzin C, Brouet A, De Vriese J, et al. associated with cell adhesion mediated drug ulates interactions between hemato-poietic Effects of vascular endothelial growth factor resistance (CAM-DR) in hematopoietic stem cells and their bone marrow microen- on the lymphocyte-endothelium interac- malignancies. Cancer Metastasis Rev vironment. Cell 2007;129: 1081-95. tions: identification of caveolin-1 and nitric 2001;20:43-50. 54. Yilmaz OH, Valdez R, Theisen BK, et al. Pten oxide as control points of endothelial cell 80. Mudry RE, Fortney JE, York T, et al. Stromal dependence distinguishes haemato-poietic anergy. J Immunol 2007;178:1505-11. cells regulate survival of B-lineage leukemic stem cells from leukaemia-initiating cells. 69. Jo DY, Hwang JH, Kim JM, et al. Human cells during chemotherapy. Blood 2000;96: Nature 2006; 441:475-82. bone marrow endothelial cells elaborate 1926-32. 55. Zhang J, Grindley JC, Yin T, et al. PTEN non-stromal-cell-derived factor-1 (SDF-1)- 81. Narendran A, Ganjavi H, Morson N, et al. maintains haematopoietic stem cells and dependent chemoattraction and SDF-1- Mutant p53 in bone marrow stromal cells acts in lineage choice and leukaemia pre- dependent transmigration of haematopoiet- increases VEGF expression and supports vention. Nature 2006;441:518-22. ic progenitors. Br J Haematol 2003;121: 649- leukemia cell growth. Exp Hematol 2003; 56. Walkley CR, Olsen GH, Dworkin S, et al. A 52. 31:693-701. microenvironment-induced myeloprolifera- 70. Peled A, Grabovsky V, Habler L, et al. The 82. Nefedova Y, Cheng P, Alsina M, et al. tive syndrome caused by retinoic acid recep- chemokine SDF-1 stimulates integrin-medi- Involvement of Notch-1 signaling in bone tor gamma deficiency. Cell 2007;129: 1097- ated arrest of CD34+ cells on vascular marrow stroma-mediated de novo drug 110. endothelium under shear flow. J Clin Invest resistance of myeloma and other malignant 57. Cortelezzi A, Fracchiolla NS, Mazzeo LM, et 1999; 104:1199-211. lymphoid cell lines. Blood 2004;103: 3503-10. al. Endothelial precursors and mature 71. Sipkins DA, Wei X, Wu JW, et al. In vivo 83. Andreasson B, Swolin B, Kutti J. Patients endothelial cells are increased in the imaging of specialized bone marrow with idiopathic myelofibrosis show peripheral blood of myelodysplastic syn- endothelial microdomains for tumour increased CD34+ cell concentrations in dromes. Leuk Lymphoma 2005;46: 1345-51. engraftment. Nature 2005;435:969-73. peripheral blood compared to patients with

[Hematology Reviews 2009; 1:e14] [page 85] Article

polycythaemia vera and essential thrombo- 2007;110:3334-44. al. Development of leukemia in donor cells cythaemia. Eur J Haematol 2002;68: 189-93. 95. Chatterjee M, Honemann D, Lentzsch S, et after allogeneic stem cell transplantation--a 84. Barosi G, Viarengo G, Pecci A, et al. al. In the presence of bone marrow stromal survey of the European Group for Blood and Diagnostic and clinical relevance of the cells human multiple myeloma cells become Marrow Transplantation (EBMT). Haemato - number of circulating CD34+ cells in myelo - independent of the IL-6/gp130/ STAT3 path- logica 2005;90:969-75. fibrosis with myeloid metaplasia. Blood way. Blood 2002;100:3311-8. 110.Kai T, Spradling A. An empty Drosophila 2001;98:3249-55. 96. Folkman J. Role of angiogenesis in tumor stem cell niche reactivates the proliferation 85. Bruns I, Czibere A, Fischer JC, et al. The growth and metastasis. Semin Oncol 2002; of ectopic cells. Proc Natl Acad Sci USA 2003; hematopoietic stem cell in chronic phase 29:15-8. 100:4633-8. CML is characterized by a transcriptional 97. Hussong JW, Rodgers GM, Shami PJ. 111.Kai T, Spradling A. Differentiating germ profile resembling normal myeloid progeni- Evidence of increased angiogenesis in cells can revert into functional stem cells in tor cells and reflecting loss of quiescence. patients with acute myeloid leukemia. Blood Drosophila melanogaster ovaries. Nature Leukemia 2009; Jan 22 [Epub Ahead of 2000;95:309-13. 2004;428:564-9. Print]. 98. AR, Kay NE, Peterson LC. Increased bone 112.Ballen KK, Shpall EJ, Avigan D, et al. Phase 86. Ries C, Loher F, Zang C, et al. Matrix metal- marrow angiogenesis in B cell chronic lym- I trial of parathyroid hormone to facilitate loproteinase production by bone marrow phocytic leukemia. Leukemia 2000;14:1414- stem cell mobilization. Biol Blood Marrow mononuclear cells from normal individuals 8. Transplant 2007;13:838-43. and patients with acute and chronic myeloid 99. Lundberg LG, Lerner R, Sundelin P, et al. 113.Napoli C, William-Ignarro S, Byrns R, et al. leukemia or myelodysplastic syndromes. Bone marrow in polycythemia vera, chronic Clin Cancer Res 1999;5:1115-24. myelocytic leukemia, and myelofibrosis has Therapeutic targeting of the stem cell niche 87. Barille S, Akhoundi C, Collette M, et al. an increased vascularity. Am J Pathol 2000; in experimental hindlimb ischemia. Nat Metalloproteinases in multiple myeloma: 157:15-9. Clin Pract 2008;5:571-9. production of matrix metalloproteinase-9 100.Padro T, Ruiz S, Bieker R, Burger H, Steins 114.Adams GB, Martin RP, Alley IR, et al. (MMP-9), activation of proMMP-2, and M, Kienast J, et al. Increased angiogenesis Therapeutic targeting of a stem cell niche. induction of MMP-1 by myeloma cells. Blood in the bone marrow of patients with acute Nature Biotechnol 2007;25:238-43. 1997;90:1649-55. myeloid leukemia. Blood 2000;95:2637-44. 115.Broxmeyer HE, Orschell CM, Clapp DW, et 88. Janowska-Wieczorek A, Marquez LA, 101.Vacca A, Ribatti D, Roncali L, et al. Bone al. Rapid mobilization of murine and human Matsuzaki A, et al. Expression of matrix marrow angiogenesis and progression in hematopoietic stem and progenitor cells metalloproteinases (MMP-2 and -9) and tis- multiple myeloma. Br J Haematol 1994;87: with AMD3100, a CXCR4 antagonist. J Exp sue inhibitors of metalloproteinases (TIMP- 503-8. Med 2005;201:1307-18. 1 and -2) in acute myelogenous leukaemia 102.Fragoso R, Elias AP, Dias S. Autocrine VEGF 116.Cashen AF, Link D, Devine S, DiPersio J. blasts: comparison with normal bone mar- loops, signaling pathways, and acute Cytokines and stem cell mobilization for row cells. Br J Haematol 1999; 105:402-11. leukemia regulation. Leuk Lymphoma 2007; autologous and allogeneic transplantation. 89. Xu M, Bruno E, Chao J, et al. Constitutive 48:481-8. Curr Hematol Rep 2004;3:406-12. mobilization of CD34+ cells into the periph- 103.Kelaidi C, Eclache V, Fenaux P. The role of 117.Levesque JP, Hendy J, Takamatsu Y, et al. eral blood in idiopathic myelofibrosis may lenalidomide in the management of myelo - Disruption of the CXCR4/CXCL12 chemotac- be due to the action of a number of proteas- dysplasia with del 5q. Br J Haematol 2008; tic interaction during hematopoietic stem es. Blood 2005;105:4508-15. 140:267-78. cell mobilization induced by GCSF or 90. Brouty-Boye D, Briard D, Azzarone B, et al. 104.Kumar S, Witzig TE, Dispenzieri A, et al. cyclophosphamide. J Clin Invest 2003;111: Effects of human fibroblasts from myelo- Effect of thalidomide therapy on bone mar- 187-96. metaplasic and non-myelometaplasic hema - row angiogenesis in multiple myeloma. 118.Calandra G, McCarty J, McGuirk J, et al. + to poietic tissues on CD34 stem cells. Int J Leukemia 2004;18:624-7. AMD3100 plus G-CSF can successfully mobi- Cancer 2001;92:484-8. 105.Ribatti D, Vacca A, Dammacco F, English D. lize CD34+ cells from non-Hodgkin's lym- 91. Cervi D, Shaked Y, Haeri M, et al. Enhanced Angiogenesis and anti-angiogenesis in phoma, Hodgkin's disease and multiple natural-killer cell and erythropoietic activi- hematological malignancies. J. Hematother myeloma patients previously failing mobi- ties in VEGF-A-overexpressing mice delay F- Stem Cell Res 200312:11-22. lization with chemotherapy and/or cytokine MuLV-induced erythroleukemia. Blood 2007; 106.Dias S, Choy M, Alitalo K, Rafii S. Vascular treatment: compassionate use data. Bone 109:2139-46. endothelial growth factor (VEGF)-C signal- Marrow Transplant 2008;41: 331-8. 92. Koistinen P, Siitonen T, Mantymaa P, et al. ing through FLT-4 (VEGFR-3) mediates 119.Hideshima T, Richardson P, Chauhan D, et Regulation of the acute myeloid leukemia leukemic cell proliferation, survival, and al. The proteasome inhibitor PS-341 in - cell line OCI/AML-2 by endothelial nitric resistance to chemotherapy. Blood 2002; hibits growth, induces apoptosis, and over- oxide synthase under the control of a vascu- 99:2179-84. lar endothelial growth factor signaling sys- 107.Rafii S, Shapiro F, Pettengell R, et al. Human comes drug resistance in human multiple tem. Leukemia 2001;15:1433-41. bone marrow microvascular endothelial myeloma cells. Cancer Res 2001;61: 3071-6. 93. Hatfield K, Ryningen A, Corbascio M, cells support long-term proliferation and dif- 120.Maciejewski JP, Risitano AM, Sloand EM, et Bruserud O. Microvascular endothelial cells ferentiation of myeloid and megakaryocytic al. A pilot study of the recombinant soluble increase proliferation and inhibit apoptosis progenitors. Blood 1995;86:3353-63. human tumour necrosis factor receptor of native human acute myelogenous 108.Nefedova Y, Landowski TH, Dalton WS. (p75)-Fc fusion protein in patients with leukemia blasts. Int J Canc 2006;119: 2313- Bone marrow stromal-derived soluble fac- myelodysplastic syndrome. Br J Haematol 21. tors and direct cell contact contribute to de 2002;117:119-26. 94. Wang L, O'Leary H, Fortney J, Gibson LF. novo drug resistance of myeloma cells by 121.Klasa RJ, List AF, Cheson BD. Rational Ph+/VE-cadherin+ identifies a stem cell like distinct mechanisms. Leukemia 2003;17: approaches to design of therapeutics tar- population of acute lymphoblastic leukemia 1175-82. geting molecular markers. Hematology Am sustained by bone marrow niche cells. Blood 109.Hertenstein B, Hambach L, Bacigalupo A, et Soc Hematol Educ Program 2001:443-62.

[page 86] [Hematology Reviews 2009; 1:e14] Hematology Reviews 2009; volume 1:e15

A sensitivity comparison of the min K antagonists coumarin and warfarin are inexpensive and the most widely used medi- Correspondence: Juha Horsti, Tampere Quick and Owren prothrombin cines in the prevention and treatment of University Hospital, Centre for Laboratory time methods in oral thromboembolism in various clinical situa- Medicine, P.O. Box 2000, FIN-33521 Tampere, anticoagulant therapy tions, and the benefit of OAT has been proved.5-12 Finland. E-mail: [email protected] The major drawback with warfarin is a narrow Key words: International Normalized Ratio, oral Juha Horsti therapeutic window and individually variable anticoagulant therapy, prothrombin time, responses to the treatment. Thus frequent pro- warfarin. Department of Clinical Chemistry, Centre thrombin time (PT) checks are required to for Laboratory Medicine, Tampere ensure that anticoagulation remains within Acknowledgments: The author thanks Kari University Hospital, Tampere, Finland the therapeutic range, which is 2.0-3.0 Borchers and Henrik Biström from ILS Scandi - International Normalized Ratio (INR). navia for their cooperation and invaluable help. New medications for oral anticoagulation therapy have been developed over a number of Received for publication: 23 June 2009. Abstract years and anticipated without laboratory test Revision received: 3 September 2009. Accepted for publication: 3 September 2009. control in an effort to replace warfarin. The Prothrombin time (PT) is the leading test new medicines, however, have not proved This work is licensed under a Creative Commons for monitoring oral anticoagulation therapy superior to warfarin. The new molecules are Attribution 3.0 License (by-nc 3.0). (OAT). According to the World Health expensive and involve serious side effects or Organization recommendation, International only narrow indications for OAT.13 ©Copyright J. Horsti 2009 Licensee PAGEPress, Italy. Normalized Ratio (INR) results obtained from According to a recent review,14,15 prothrom- Hematology Reviews 2009; 1:e15 the same patient samples with the major PT bin time has served as a basis for OAT moni- doi:10.4081/hr.2009.e15 methods (Quick and Owren) should be the toring since its first description by Quick more same when the therapeutic range is the same. than 70 years ago. PT measures vitamin K- In our study blood samples were obtained from 3,4 dependent coagulation factors II, VII, and X. ures both of these coagulation factors, which 207 OAT patients. We analyzed the samples Warfarin (VKAs) inhibits the synthesis of is a drawback in OAT control. In the Owren using two Quick and two Owren PT (combined coagulation factors II, VII, IX, and X in the liver, method the sample volume in the reaction thromboplastin) reagents for INR and which remain partially inactive unless nine to mixture is 5% only, in the Quick method 33%. assessed the sensitivity and true coagulation 13 of the amino-terminal glutamate (Glu) On account of this difference in sample dilu- activity using a new-generation PT method. residues are carboxylated to form Ca2+-binding tion the Owren PT is considered to be a more The INR values with the Quick PT and Owren -carboxyglutamate (Gla) residues.3,4 Thus it is γ sensitive method than the Quick PT. PT methods were very similar around the nor- inevitable that the more medication is admin- In the present study, we sought to compare mal range, while unacceptable differences istered, the more inactive coagulation factors INR results for warfarin therapy from the same were seen within the therapeutic range and at will result, hampering and inhibiting the PT OAT patient samples using both the Owren and higher INR values. The Quick PT results as measurement. the Quick PT methods. The aim was to study INR are clearly lower than those given by Today, two major PT methods are accepted the method and reagent sensitivity of Owren Owren PT and the difference increases toward for anticoagulant medication control and they and Quick PT using the new-generation PT higher INR. The new PT method functions well are used globally for chronic anticoagulation method developed by Horsti et al., which can with both Owren PT reagents, and we can cal- checks: either the Quick PT, which is based on measure active coagulation without inhibition culate the true active INR. The Quick PT meth- the technique described by Quick and his col- from inactive coagulation factors.13,27,28 ods show no sensitivity to coagulation inhibi- leagues in 1935,14,15 or the Owren PT16 (com- tion measurement. The harmonization of the bined thromboplastin reagent). The Owren PT INR is an important goal for the safety of OAT is the predominant approach used in the patients. More accurate INR results reduce Nordic countries, the Benelux, and Japan. The morbidity and mortality, and the therapeutic accuracy and comparability of PT results are Materials and Methods ranges should be similar worldwide. In this essential to the safety of the individual in anti- study we found unacceptable differences in coagulant therapy and to improve the applica- Patients and blood sampling INR results produced by the two PT methods. bility of anticoagulation guidelines. The World Venous blood samples were obtained from The new method showed a lack of sensitivity to Health Organization (WHO) recommendation 207 hospital and health-center patients Quick PT. For the global harmonization of OAT for the use of INR given in the mid-1980s was referred to the PT test for the monitoring of therapy and for INR accuracy only the more aimed to harmonize PT results for OAT, regard- oral anticoagulant therapy. In our region a “P- sensitive Owren PT method should be used. less of the laboratory, reagent, instrument, or INR” test code is used for this purpose. Hence, method used.17,18 the patient samples represented all possible Unfortunately, an increasing body of evi- phases of anticoagulation: pretreatment, dose- dence indicates that this goal has not been adjusting phase, and steady-state phase. All Introduction achieved.19-26 the procedures were approved by the responsi- Different reagents are known to vary with ble committee of our institution in accordance Oral anticoagulation therapy (OAT) is one of respect to the source of thromboplastins and with the Helsinki Declaration of 1975. The the most commonly used medications world- other reagent components,26 and these have blood (3.15 mL) was drawn into citrate coagu- wide. The purpose of the treatment is to bal- not been characterized in detail. The Owren lation tubes (Greiner Labortechnik GmbH, ance the risks of hemorrhage and thrombosis method is reagarded as the “reference Vacuette cat. No. 454332, 9NC) containing 0.35 in the patient. Arterial and venous complica- method” in which fibrinogen and factor V are mL 0.109 mol/L (3.2%) citrate solution. The tions commonly are involved in morbidity and added to the reagent, and these do not depend sample needle (Terumo, Venoject needle, mortality in OAT patients globally.1-4 The vita- on K-vitamin antagonists. The Quick PT meas- Quick Fit, cat. No. MN-2138MQ) was 0.8x40

[Hematology Reviews 2009; 1:e15] [page 87] Article mm. The sample tubes were centrifuged at 1850 g for 10 min at 20°C to separate the plas- Table 1. Differences in INR results between four PT methods at variable INR levels as INR units and percentages. The correlation equations between methods were used for the ma. All the measurements were commenced calculations. within eight hours from blood collection. Nycotest Owren's Diff Diff PT-Fib Diff Diff Reco Diff Diff Prothrombin time determination PT INR PT INR % INR INR % INR 2G INR % INR The calculation formula is for the 1,00 1,00 0,05 0,00 1,08 8,03 - 1,01 0,74 - International Normalized Ratio: INR = (sam- 0,08 0,01 plesec/normalsec)ISI, where ISI is the International 2,00 1,90 5,25 0,11 1,88 6,23 0,12 1,68 16,16 0,32 Sensitivity Index. When ISI is close to 1.0, the reagent is sensitive and ISI is insignificant in 3,00 2,79 6,98 0,21 2,67 10,99 0,33 2,35 21,79 0,65 the INR calculation. The PT coagulation times 4,00 3,69 7,85 0,31 3,47 13,37 0,53 3,02 24,61 0,98 were measured using a fully automated ACL 5,00 4,58 8,37 0,42 4,26 14,79 0,74 3,69 26,30 1,32 TOP CTS coagulation analyzer (Instrument- 6,00 5,48 8,72 0,52 5,06 15,75 0,94 4,35 27,43 1,65 ation Laboratory, IL, Lexington, MA, USA). For the Owren PT (combined thromboplastin reagent) the coagulation reaction contained 10 µL of citrated sample plasma, 60 µL of dilu- ference in intercepts also in INR units and Microsoft Excel from Analyse-it Software Ltd ent, and 140 µL of reagent. The volumes for subtracted this from total INRTot (more expla- programs (Table 1). “dilution measurement” were 5 µL + 65 µL + nation about the new method in reference 28 The difference between the Owren PT meth- 140 µL. The test reagents were Nycotest PT, [patent pending for PT method]),27,28 ods in INR terms is below 10% over the whole cat. No. 1002488 (rabbit brain thromboplastin) INRAct = INRTot – INRInh measuring range. In the therapeutic range, the and diluent (Nycotest PT, dilution liquid, INRs were calculated using the formula INR = average difference is 0.16 INR; 6.10%. The dif- cat.No. 999002) from Axis-Shield as (samplesec / normalsec)ISI ference between Nycotest PT and Pt-Fib HS is lot.10132661, ISI=1.12; Owren's PT, cat.no. GHI The dilution factor for the Owren PT and 8.03-15.75% and in the therapeutic range an 131-10 (rabbit thromboplastin), and diluent Quick PT was 2.0. average of 0.23 INR; 8.6%. The difference between Nycotest PT and Recomb 2G is 0.74- (Owren's buffer, cat.No. GHI 150) from Analytical imprecision and statistics 27.43% and in the therapeutic range an aver- Medirox as lot 75051, ISI=1.16. For the one- The within-run precision of the PT tests was age of 0.49 INR; 19.0%. stage prothrombin time with Quick, 100 µL of measured using normal plasma (n = 10 deter- The Owren methods showed excellent corre- coagulation reagent was added to 50 µL of cit- minations). The respective CVs were: for lation (y = 0.96x + 0.052; intercept 95% CL rated plasma. Sample volumes for the dilution Nycotest PT 1.81%; Owren's PT 0.90%; 0.034 to 0.071; slope 0.948 to 0.973), while the were 100 µL + 25 µL + 25 µL (Factor Diluent, RecombiPlasTin 2 G 0.55%; PT-Fibrinogen HS correlations between Quick and Owren PT HemosIL, cat.no. 0009757600 from IL). The PLUS 2.15%, and pooled plasma (about 2 INR): were not good: Nyco PT vs. Pt-FibHS (y = test reagents were: RecombiPlasTin 2 G cat. for Nycotest PT 1.26%; Owren's PT 1.14%; 0.7955x + 0.2802; intercept 95% CL 0.242 to No. 0020003050 (recombinant human tissue RecombiPlasTin 2 G 1.07%; PT-Fibrinogen HS 0.319; slope 0.769 to 0.823), and Nyco PT vs. thromboplastin) from Instrumentation PLUS 1.94%. The Microsoft Excel 5.0 and Rec 2G (y = 0.669x + 0.3357; intercept 95% CL Laboratory, Lexington USA as lot. N0486316, Analyse-it for Microsoft Excel from Analyse-it 0.290 to 0.381; slope 0.642 to 0.700). The corre- ISI=0.97; PT-Fibrinogen HS PLUS cat. no. Software Ltd programs were used to obtain the lation was not good either between the Quick 0008469810 (rabbit brain thromboplastin) correlation functions and INR results. methods: Pt-FibHS vs. Rec 2G (y = 0.8267x + from Instrumentation Laboratory, Lexington 0.1055, non-linear relationship; intercept 95% USA as lot. N0185573 ISI=1.11. CL 0.063 to 0.150; slope 0.800 to 0.857). Additional reagents The new-generation PT method functions Results well with the two Owren PT methods and Control Plasma Normal cat.No. 1002387 makes it possible to calculate the true active from Axis-Shield as lot. 1D71AOA, NKP Normal We compared the results of four different INR without inhibition (Figure 2). The correc- Control Plasma cat.No. GHI 162 from MediRox commercial INR determination methods (two tion lowers the INR values closer to Quick PT AB as lot. 72011, and Normal Control Assayed Quick PT and two Owren PT, ISI values 0.97; values. Marked individual differences were cat.No. 0020003110 from IL as lot. N0386069. 1.11 and 1.12; 1.16) from 207 blood samples seen in the therapeutic range 2-3 INR and higher values between Rec 2 G INR (Quick PT) The new-generation prothrombin from patients in imminent or ongoing OAT. All the ISI values used for reagents were manufac- and Owren's PT INRACT (Owren PT) (Figure time method turer values. The INR values obtained with the 3). The correlation between active INR results We constructed PT (sec) vs. C (C = plasma different methods were similar around the is good (y = 0.8955x + 0.104; intercept 95% CL or calibrator dilution factor) plots for normal normal range or INR 1 (Figure 1). In contrast, 0.087 to 0.122; slope 0.885 to 0.907). The Quick and OAT plasmas. This is consistent with an marked differences were seen in the therapeu- PT methods show no sensitivity to the coagula- uncompetitive inhibition principle with oral tic range (2-3 INR) and higher INR values tion inhibition measurement. anticoagulants. From the line equation the y- between the Quick PT and Owren PT methods. axis intercept is calculated. This is the so- The INRs produced by the Quick PT are clearly called minimal clotting time (tmin) with an infi- lower than those given by the Owren PT, and nite number of clotting factors. The difference the difference increases toward higher INRs. Discussion in intercepts (y-axis) between normal plasma Using correlation equations between the four and OAT plasma indicates the action of uncom- methods, the differences were calculated The global harmonization of INR and thera- petitive inhibition in seconds without a cali- applying the Microsoft Excel 5.0 and Passing & peutic ranges for different clinical indications bration effect. We went on to calculate the dif- Bablok method comparison, Analyse-it for is an important goal for the benefit of patients

[page 88] [Hematology Reviews 2009; 1:e15] Article

Figure 1. INR values in increasing order from 207 analyzed patient samples on anticoagulation therapy. Nycotest PT and Owren's PT are Owren methods (combined thromboplastin reagent), and PT- FibHS and Rec 2G are Quick methods.

Figure 2. INRTOT and INRACT values in increasing order from 207 analyzed patient samples on anticoagulation therapy. Owren's PT reagent (combined thromboplastin reagent) and the new-generation PT method are used for INR measurement.

Figure 3. Rec 2G, Quick PT, INR, and Owren's PT (combined thromboplastin reagent) INRACT values in increasing order from 207 analyzed patient samples on anticoagulation therapy. The new-generation PT method is used for INRACT measurement.

[Hematology Reviews 2009; 1:e15] [page 89] Article as well as clinicians. WHO originally envisaged inhibition depends on the medication and the 8. Hirsh J, Dalen JE, Anderson DR, et al. Oral INR as a reliable and safe method for obtaining patient's individual metabolism, and thus each anticoagulants’ mechanism of action, clin- results from anticoagulation trials and as a patient’s sample must be corrected individual- ical effectiveness, and optimal therapeutic basis for expert guidelines for treatment. ly.27,28 The Owren PT methods are sensitive, but range. Chest 1998;114:445-69. However, the results of earlier studies and this more susceptible to the inactive coagulation 9. Schulman S. Care of patients receiving present study demonstrate that the goal has factors and inhibition in measurement. In an long-term anticoagulant therapy. N Engl J not been achieved. The narrow therapeutic earlier study we measured only the active Med 2003;349:675-83. range and the increase in complications and coagulation factors, INRAct (F II, F VII, FX) with- 10. Gage BF, Fihn SD, White RH. Management mortality outside this range call for the devel- out inhibition, which provided a new possibili- and dosing of warfarin therapy. Am J Med opment of more accurate laboratory analytics. ty to develop anticoagulant therapy and more 2000;109:481-8. A computer simulation study of serial INR appropriate care for OAT patients. The active 11. Wilson SE, Watson HG, Crowther MA. Low measurements has been conducted within the coagulation factors are responsible for throm- dose oral vitamin K for management of most widely used therapeutic range (INR 2.0 - bosis or bleeding in vivo and medication asymptomatic patients with an elevated 3.0); the authors concluded that analytical should be based on this principle.13,27,28 The INR: a brief review. CMAJ 2004;170:821-4. imprecision should be <5% and analytical bias inhibition interferes with the calibration pro- 12. Dunn AS, Turpie AG. Perioperative man- < ± 0.2 INR units.29 cedure and measurement. agement of patients receiving oral antico- In this study, agreement between the Quick This study was conducted from the point of agulants. Arch Int Med 2003;163:901-8. PT and Owren PT methods was particularly view of a clinical laboratory without clinical 13. Horsti J. An improved method for PT meas- poor and failed to meet the qualification data and outcomes are missing between meth- urement in OAT patients. Clin Lab Int requirements. For this study we selected a ods (Quick PT, Owren PTTOT, Owren PTACT), 2008;32:16-8. newly developed recombinant PT reagent, but which would give the final answer as to the 14. Quick AJ, Stanley-Brown M, Bancroft FW. A the result was not acceptable. We may wonder superiority of PT methods. This study revealed study of the coagulation defect in hemo- whether the difference in sample volume in the lack of sensitivity in both Quick PT meth- philia and in jaundice. Am J Med Sci 1935; the reaction mixture (5% and 33%) accounts ods but good sensitivity and correlation in the 190:501-11. for the measurement sensitivity. Both Owren Owren PT activity (INRAct) measurements. The 15. Quick AJ. The prothrombin time in PT reagents are sensitive to inactive coagula- editorials of Clinical Chemistry have posed the haemophilia and in obstructive jaundice. tion factor measurement (inhibition), while critical question and sought answers: “Has the J Biol Chem 1935;109:73-4. the Quick PT reagents are not. Time Arrived to Replace the Quick Pro - 16. Owren PA. Thrombotest. A new method for In our earlier study we used DadeBehring thrombin Time Test for Monitoring Oral Anti - controlling anticoagulant therapy. Lancet Innovin with human placenta thromboplastin coagulant Therapy?”19 On the basis of recent 1959;2:754-8. (Quick PT reagent), which is the most sensi- studies and our opinion we can answer: “Yes, 17. WHO Expert Committee on Biological tive reagent to inactive coagulation factors and with the Owren PT without inhibition.” Standardisation. Thirty-third Report. inhibition. The sensitivity thus does not Technical Report Series 687. WHO Geneva depend only on differences in principle 1983:81-105. between the Quick PT or Owren PT methods. 18. International Committee for Standard- The thromboplastin and other components References isation in Haematology. International used in the reagent affect the reagent sensitiv- Committee on Thrombosis and Haemo - ity. Horsti24 has compared the Quick PT and 1. Palareti G, Leali N, Coccheri S, et al. stastis. ICSH/ICTH recommendations for Owren PT methods for the harmonization of Bleeding complications of oral anticoagu- reporting prothrombin time in oral antico- INR results and concluded that Quick PT yields lant treatment: an inception cohort, agulant control. Thromb Haemost 1985; clinically divergent and Owren PT clinically prospective collaborative study (ISCOAT). 53:155-6. acceptable INR results. It would be interesting Lancet 1996;348:423-8. 19. Jackson CM, Esnouf MP. Has the time to study the effect of lupus anticoagulants on 2. Odén A, Fahlén M. Oral anticoagulation arrived to replace the quick prothrombin Quick PT and Owren PT methods. The results and risk of death: a medical record linkage time test for monitoring oral anticoagulant of recent studies, unfortunately, demonstrate study. BMJ 2002;325:1073-5. yherapy? Clin Chem 2005;51:483-5. that attempts at harmonization have to 3. Ansell J, Hirsh J, Poller L, et al. The phar- 20. Cunningham MT, Johnson GF, Pennell BJ, improve.25 The unacceptable situation for oral macology and management of the vitamin et al. The reliability of manufacturer-deter- anticoagulation was confirmed further by this K antagonists: the seventh ACCP Conf - mined, instrument-specific international present investigation. erence on Antithrombotic and Throm- sensitivity index values for calculating the How can the harmonization of INR results bolytic Therapy. Chest 2004;126: 204-33. international normalised ratio. Am J Clin be improved worldwide? For global use, WHO 4. Hirsh J, Fuster V, Ansell J, et al. American Pathol 1994;102:128-33. should recommend only the Owren PT, which Heart Association/American College of 21. van Rijn JL, Schmidt NA, Rutten WP. is the superior method. The advantages of Cardiology Foundation Guide to Warfarin Correction of instrument- and reagent- Owren PT are the fact that it measures only Therapy. Circulatione 2003;107:1692-734. based differences in determination of the vitamin K-dependent coagulation factors F II, F 5. Turpie AG. Safer anticoagulant therapy International Normalised Ratio (INR) for VII, and F X (not fibrinogen and F V), and pro- after heart valve replacement. Recom- monitoring anticoagulant therapy. Clin vides greater sensitivity and extensive dilution mendations for less intense regimens. Chem 1989; 5:840-3. of the interfering matrix substances in the Postgrad Med 1997;101:85-6, 89-90, 93-4. 22. Ng VL, Levin J, Corash L, et al. Failure of final reaction mixture.19,25 The patient samples 6. Hirsh J. Optimal intensity and monitoring the International Normalised Ratio to gen- also contain inactive coagulation factors, the warfarin. Am J Cardiol 1995;75:39-42. erate consistent results within a local amount of which increases concomitant with 7. Stein PD, Alpert JS, Bussey HI, et al. medical community. Am J Clin Pathol increased anticoagulant medication and high- Antithrombotic therapy in patients with 1993;99:689-94. er INR levels, interfering with and causing mechanical and biological prosthetic heart 23. Horsti J. Agreement of owren and quick error in the coagulation measurement. The valves. Chest 2001;119:220-7. prothrombin times: effects of citrate and

[page 90] [Hematology Reviews 2009; 1:e15] Article

calcium concentrations and international methods: comparison of seven commercial 28. Horsti J, Uppa H, Vilpo JA. A new genera- sensitivity index correction. Clin Chem reagents. Clin Chem 2005;51:553-60. tion method for quick and owren PT. The 2001;47:940-3. 26. van den Besselaar AM. International stan- Open Hematol J 2008;2:81-5. 24. Horsti J. Comparison of quick and owren dardisation of laboratory control of oral 29. Kjeldsen J, Lassen JF, Petersen PH, et al. prothrombin time with regard to the har- anticoagulant therapy: a survey of throm- Biological variation of International monisation of the international nor- boplastin reagents used for prothrombin Normalised Ratio for prothrombin times, malised ration (INR) system. Clin Chem time testing. J Heart Valve Dis 1993;2:42- and consequences in monitoring oral anti- Lab Med 2002;40:399-403. 52. coagulant therapy: computer simulation of 25. Horsti J, Uppa H, Vilpo J. Poor agreement 27. Horsti J, Uppa H, Vilpo JA. A New genera- serial measurements with goal-setting for between different prothrombin time tion prothrombin time method for INR. analytical quality. Clin Chem 1997;43: International Normalized Ratio (INR) The Open Medicinal Chem J 2008;2:11-15. 2175-82.

[Hematology Reviews 2009; 1:e15] [page 91] Hematology Reviews 2009; volume 1:e16

Clear cell renal cell carcinoma area of enhancement measuring 7 cm in the inferior aspect of the left kidney (Figure 2), Correspondence: Tobe Momah, PGY-II Depart - with vaginal and brain and a 6x5.6 cm focus of soft tissue enhancement of Family Medicine, Brooklyn Hospital metastases: a case report ment in the region of the vagina (Figure 3). Center, 121 Dekalb Avenue, Brooklyn, New York, and literature review These findings were suggestive of cystic renal NY 11201, USA. E-mail: [email protected] cell carcinoma and metastatic clear cell carci- Key words: renal cell carcinoma, clear cell, 1 2 noma of the vagina, respectively. Tobe Momah, Etwaru Dhanan, metastases. Phillip Xiao,3 Vasantha Kondamudi1 Treatment 1PGY-II Department of Family Medicine; Conflict of interest: the authors report no con- The patient was started on Lovenox (enoxa- flicts of interest. 2Department of Urology; 3Department of parin) and coumadin, but she started bleeding Pathology; Brooklyn Hospital Center, PV a week later and was admitted to our inpa- Received for publication: 18 June 2009. Brooklyn, New York, NY, USA tient service with shortness of breath, and for Revision received: 24 September 2009. anticoagulation treatment and inferior vena Accepted for publication: 1 October 2009. cava filter placement. She later underwent a This work is licensed under a Creative Commons laparoscopic hand-assisted left nephrectomy Attribution 3.0 License (by-nc 3.0). Abstract with histopathological examination of the 6.6x6x4.5 cm left lower kidney pole specimen, ©Copyright T. Momah et al., 2009 which was consistent with renal clear cell car- There are very few cases of clear cell renal Licensee PAGEPress, Italy. cinoma (Figure 4). The tumor extended into Hematology Reviews 2009; 1:e16 cell carcinoma with metastases to the vagina the left perirenal adipose tissue, was a doi:10.4081/hr.2009.e16 and brain reported in the literature. Our case Fuhrman nuclear grade G3 tumor, involved the study highlights this rare clinical occurrence left renal vein, and was staged as T3bN0Mx and its associated complications including pul- secondary to the vaginal metastases. monary embolism. In addition we discuss cur- of life and is three times more common in men than in women.2 rent management guidelines for treating and Clinical course and outcome diagnosing the disease, and how this manage- The patient tolerated surgery well, and after ment improves prognosis. Clinical presentation discharge received palliative radiation therapy The most common symptom on presentation (4860 cGy in 27 fractions of 180 cGy) to the in RCC cases with vaginal metastases is vagi- Case report pelvis. However, she was re-admitted three nal bleeding,1,2 and its most common sites for months after discharge for persistent head - metastases include extra-renal regions like aches and right-sided upper extremity numb- 1 Presentation and history lung, lymph nodes, liver, and bones. Our case ness and weakness, which on CT of the head was unique for its metastatic sites, clinical A 55-year-old African American female with was found to be because of a secondary left pos- presentation, and associated complications. a past medical history of hypertension, nico- terior parietal mass. She underwent an Clear cell RCC with metastases to the vagina tine dependence (stopped smoking eleven intracranial incision biopsy of the left parietal and brain is very uncommon with one case of years ago), and Para 2022 (status post-right brain tumor that showed metastatic clear cell the former1 and none of the latter reported in salpingectomy for ectopic pregnancy) present- carcinoma histologically, consistent with a renal the literature. In postmenopausal women only ed with six weeks of persistent postmeno - primary tumor. She received postoperative ther- one case of primary clear cell carcinoma of the pausal bleeding per vagina (PV). Prior to pres- apeutic irradiation of the cranium for metastat- vagina has been reported3 because these entation the patient had been menopausal for ic brain lesions and recently was started on tumors tend to occur more in young patients ten years and used only amlodipine and avalide chemotherapy (Sutent 50 mg once daily orally). exposed to DES in utero.4 Therefore, a solitary tablets as medications. She has a positive fam- She continues taking her antihypertensive lesion of clear cell carcinoma in the vagina in ily history (sister) of breast cancer and denied medication but had to discontinue her anticoag- a postmenopausal patient is more likely to be a intra-partum exposure to diethyl stilbesterol ulation drugs because of recurrent PV bleeding. metastatic secondary than a primary lesion, (DES). The physical examination revealed a She currently has good performance status and and a consideration of clear cell tumor primary loose necrotic mass with a pedunculated polyp otherwise is stable medically. sites should be performed immediately. This attached to the posterior wall of the vagina and was done in our case and revealed a primary measuring 3.5x1.5x1 cm. Endocervical and site in the left kidney, which is typically the pri- endometrial curetting were negative for malig- Discussion mary site for clear cell RCC as the retrograde nancy while polypectomy of the vaginal mass filling of the left renal vein from the left ovari- revealed foci of necrosis, surface ulceration, Epidemiology an vein and uterovaginal plexus eventually and a polypoid tumor consistent with clear cell Clear cell renal cell carcinoma (RCC) is the involves the left kidney.5 carcinoma of the vagina (Figure 1). most common primary site for clear cell tumors, yet clear cell RCC metastases to the Complications Initial management vagina or brain are rare with only one case of The vascular effects of clear cell RCC A computerized tomography (CT) scan of the former reported in the literature to date.1 include the intravascular growth of the tumor the abdomen with contrast had been done four Eighty cases of RCC with vaginal metastases into the left renal vein, and may have caused months previously and showed a 4.4 cm left have been reported in the literature,2 and the tumor emboli to the left lower lung pul- kidney mass with peripheral enhancement. A because clear cell carcinomas constitute about monary arteries that resulted in our patient’s CT scan of the chest, abdomen, and pelvis done 80% of all renal cell carcinomas,2 most of these pulmonary emboli. The vascular involvements after the diagnosis of clear cell malignancy cases may be clear cell carcinomas in fact, but of clear cell carcinoma are well documented in revealed a pulmonary embolism (PE) in the are under-reported as such. Clear cell RCC the literature6 but this is the first reported case left lower lobe pulmonary arteries, an enlarged usually occurs in the sixth or seventh decade with an associated pulmonary embolism.

[page 92] [Hematology Reviews 2009; 1:e16] Case Report

deficits is suggestive of brain metastases. A high indicative index for a clear cell primary tumor is a requirement for the diagnosis of a secondary clear cell vaginal malignancy and for avoiding complications like pulmonary embolism, further metastatic spread, and the consequent deterioration of the patient.

References Figure 1. Vaginal metastases from renal cell carcinoma, clear cell type. Microscopic 1. Queiroz C, Bacchi C, Oliveira C, et al. examination of the tumor from the vagina Cytologic diagnosis of vaginal metastases reveals the compact alveolar architecture of cells with clear cytoplasm and distinct cell from renal cell carcinoma. A case report. boundaries in a background of delicate Acta Cytol 1999;43:1098-100. vasculature. (H&E stain; magnification Figure 3. Computerized tomography scan of 2. Boggess M, Hester O, Srcher S. Renal clear 100X.) the pelvis showing soft tissue enhancement cell carcinoma appearing as a left neck measuring 6x5.6 cm in the left wall of the vagina (arrow) consistent with metastatic mass. Cytopathology 1991;2:47-8. clear cell carcinoma of the vagina. 3. Hughes J, Jensen C, Donnelly A, et al. The role of fine-needle aspiration cytology in the evaluation of metastatic clear cell tumors. Cancer Cytopathol 1999;87:380-9. 4. Watanabe Y, Ueda H, Nozaki K, et al. Advanced primary clear cell carcinoma of the vagina not associated with diethyl- stilbestrol. Acta Cytol 2002;46:577-81. 5. Matias-Guiu X, Lerma E, Prat J. Clear cell tumors of the female genital tract. Semin Diagn Pathol 1997;14:233-9. 6. Papac R, Poo-Hwu W. Renal cell carcinoma: a paradigm of lanthanic disease. Am J Clin Oncol 1999;22:223-31. 7. Allard J, McBroom J, Zahn C, et al. Vaginal Figure 2. Computerized tomography scan Figure 4. Renal cell carcinoma, clear cell metastases and thrombocytopenia from of the abdomen showing a 7 cm area of type. Microscopic examination of the kid- renal cell carcinoma. Gynecol Oncol 2004; enhancement in the left lower pole of the ney tumor shows alveolar architecture of 92:970-3. kidney (arrow) consistent with renal clear the cells with clear cytoplasm and distinct 8. Greenberg R, Charles R, Samana A, et al. cell carcinoma. cell boundaries in a background of delicate Surgical management of recurrent geni- vasculature. (H&E stain; magnification 400X.) tourinary malignancies. Semin Oncol 1993;20:473-92. 9. Walsh P, Retik A, Vaughan E, et al., editors. Campbell’s Urology. 8th edn. Philadelphia: Diagnostic methods or severe pain is present, however, the use of Saunders, 2002, pp 2714-9. 9 Fine needle aspiration cytology to establish palliative nephrectomy is recommended. 10. McNichols D, Segura J, DeWeerd J. Renal the anatomic site of origin in metastatic clear cell carcinoma: long term survival and late cell RCC cases, when combined with clinical Prognosis recurrence. J Urol 1981;126:17-23. features,7 is 95% accurate and is easier to Multiple metastases, recurrent metastases, 11. Dimarco D, Lohse C, Zincke H, et al. Long administer in outpatient settings and follow- and renal vein involvements are poor prognos- term survival of patients with unilateral up cases than are endometrial biopsies. tic indicators for survival10 and, as seen in our sporadic multifocal renal cell carcinoma However, the use of magnetic resonance imag- patient, are prone to develop complications according to histologic subtype compared ing (MRI) and CT scans still provide the main- like pulmonary embolism. Patients with multi- to patients with solitary tumors after radi- stay of primary clear cell RCC diagnoses (and focal clear cell RCC (like our patient with two cal nephrectomy. Urology 2004;64:462-7. subsequent diagnoses of secondary metas- metastatic sites) are more likely to have a con- tases) owing to their greater utility and accu- tralateral recurrence than are patients with racy. solitary clear cell RCC (risk ratio: 2.91, p=0.142).11 Treatment The use of whole body radiation therapy in cases involving a solitary metastatic site clear Conclusions cell carcinoma currently is not recommended as adjuvant loco-regional radiation therapy is The development of recurrent shortness of the preferred treatment for clear cell RCC.8 In breadth in a clear cell RCC patient with vaginal diffuse or non-resectable disease, or when metastases is indicative of pulmonary paraneoplastic syndrome, severe hemorrhage, embolism, and the presence of neurological

[Hematology Reviews 2009; 1:e16] [page 93] Hematology Reviews 2009; volume 1:e17

A case of follicular lymphoma based on guesswork, computed tomography (CT) findings are helpful for cases in which Correspondence: Osamu Imataki, Division of complicated with mesenteric mesentery biopsy is problematic.4 The clinical Hematology, Department of Internal Medicine, panniculitis course of MP is usually benign, and sponta- Faculty of Medicine, Kagawa University,1750-1 Ikenobe, Miki-cho, Kita-gun, Kagawa, 761-0793, neous remission is quite common without spe- Japan. E-mail: [email protected] Yotaro Tamai,1 Osamu Imataki,2 cific treatment.5 Ichiro Ito,3 Kimihiro Kawakami1 The pathogenesis of MP remains obscure Key words: follicular lymphoma, mesenteric pan- and many associated diseases are known, niculitis, autologous peripheral blood stem cell 1Division of Hematology & Stem Cell including mechanical stimulation such as sur- transplantation. Transplantation, Shizuoka Cancer Center, gery or trauma, infections, vascular diseases, Shizuoka, Japan; 2 Division of Received for publication: 11 July 2009. and malignant and non-malignant neoplasms.5 Hematology, Department of Internal Revision received: 3 October 2009. The occurrence rate of lymphoma as an under- Accepted for publication: 5 October 2009. Medicine, Faculty of Medicine, Kagawa lying disease is 15% or more, and no malignant University, Kagawa, Japan; 3Division of involvement of the mesentery was diagnosed This work is licensed under a Creative Commons Attribution 3.0 License (by-nc 3.0). Pathology, Shizuoka Cancer Center, in almost all cases.5 In a case series of MP Shizuoka, Japan accompanied with malignant lymphoma, ©Copyright Y. Tamai et al., 2009 Kipfer et al. (8 cases) suggested that mesen- Licensee PAGEPress, Italy. teritis was a non-specific response to underly- Hematology Reviews 2009; 1:e17 ing lymphoma.5 Other reports introduced MP doi:10.4081/hr.2009.e17 Abstract incidence during the clinical follow-up course

Mesenteric panniculitis (MP) is a rare disease occasionally complicated with lymphoma. A 55-year old female presented with MP accompanied by malignant lymphoma. This patient was first treated for follicular lymphoma and subsequently for panniculitis. After 6 courses of R-CHOP chemotherapy, the treatment response was partial. An additional course of salvage chemotherapy led to a complete response. Since the mesenteric mass progressed simultaneously with the regression of other lymphoma lesions, we performed a biopsy of the mesenteric mass and pathologically confirmed an MP lesion without lymphoma. Subsequent high-dose chemotherapy led to CR and the MP lesion remained stable. In the present case, MP progressed with chemotherapy. We concluded that mesenteric lesions suspected of progressing or recurring should be diagnosed pathologically even if asymptomatic.

Introduction

An extensive involvement of mesenteric fat tissue of the small bowel with chronic non-specific inflammatory disease, first reported by Jura et al. in 19241 was described, detailed, and named as mesenteric panniculitis (MP) by Ogden et al. in 1960.2 MP is always discussed with pathological findings concerning its etiology and 3 components of histology, fibrosis, Figure 1. Imaging and pathologic evaluation of mesenteric panniculitis. During treatment and before SCT, we evaluated the clinical course of the mesenteric mass by CT (A) chronic inflammation, and fat necrosis, sug- and 18F-FDG-PET (B). (A) Clinical course of mesenteric lesion evaluated by CT imag- gesting various mechanisms of disease (called ing. Mesenteric hypertrophy before treatment reduced during chemotherapy (i), after sal- sclerotic mesenteritis, mesenteric panniculi- vage chemotherapy (ii), and before PBSCT (iii). Arrows indicate hypertrophic mesenteritis, and mesenteric lipodystrophy, respective- um. (B) Clinical course of mesenteric lesion evaluated by FDG-PET. An accumulation positioned at the mesenterium detected by FDG-PET (i) was more intensified after sal- ly).3 Histologically, MP lesion is characterized vage chemotherapy (ii) and did not improve before PBSCT (iii). Arrows indicate an by the infiltration of mesenteric fat tissue by increased accumulation in accordance with the enlarged mesenterium. (C) Pathological inflammatory cells, mainly consisting of lym- findings of biopsied mesenteric mass. Fatty necrosis surrounded by foamy macrophages phocytes and plasmacytes, along with a mix- and reactive lymph adenopathy was characteristic of the mesenterium. ture of fat necrosis and fibrosis.3 Although the Immunohistological staining failed to detect CD20, CD10, CD79a, and bcl-2, with no indications of malignant lymphoma: (i) H-E stain (x 400), (ii) H-E stain (x 1000). diagnosis of MP before biopsy is primarily

[page 94] [Hematology Reviews 2009; 1:e17] Case Report of lymphoma. Ogden (2 cases)6 and Perez- 7 Table 1. Laboratory data on admission. The profile of the peripheral blood cell count was Ferriols (1 case) both reported on the associa- within the normal range. Biochemical data showed a normal LDH level and no organ tion of these 2 diseases. disorders. Cytogram of bone marrow revealed no obvious morphological changes and the involvement of lymphoma cells was not detected.

Component of blood cell Case Report White blood cells 9100 /μL (3500~9100) Granulocytes 69% (32~79) Eosinophils 1% (0~6) A 55-year old woman presented with right Basaophils 2% (0~2) cervical lymph node swelling in July, 2004, fol- Monocytes 3% (0~8) lowed by inguinal lymph node swelling a few Lymphocytes 25% (18~59) weeks later. She was admitted to our hospital in Red blood cells 415x104 /μL (376~500) August, 2004. The results of the blood and bone Hemoglobin 12.6 g/dL 11.3~15.2) marrow examinations performed at that time Hematocrit 37.4% (33.4~44.9) Platelets 23.1x104 /μL(13.0~36.9) are shown in Table 1. A lymph node biopsy of the inguinal area revealed grade 2 follicular Biochemistry Total protein 7.7 g/dL (6.7~8.3) lymphoma. CT images of her neck, thorax and Albumin 4.6 g/dL (4.0~5.0) abdomen, and FDG-PET revealed stage IV dis- Total billirubin 0.4 mg/dL (0.3~1.2) ease. She was diagnosed as having stage IV Glutamic pyruvic transaminase 19 IU/L (5~40) lymphoma with femoral bone involvement, as Lactate dehydrogenase 195 IU/L (115~245) confirmed by femoral MRI. Her follicular lym- Alkaline phosphatase 259 IU/L (115~359) phoma international prognostic index (FLIPI) Blood urea nitrogen 15.9 mg/dL (8.0~22.0) was intermediate (stage and lymph-node Creatinine 0.6 mg/dL (0.47~0.79) Urine acid 3.8 mg/dL (2.5~7.0) lesions). IgH-BCL rearrangement (FISH) was Natrium 142 mEq/L (136~147) detected in her lymph nodes and bone marrow Potassium 4.1 mEq/L (3.6~5.0) cells, although neither morphological abnor- Chloride 106 mEq/L (98~109) mality nor apparent involvement of lymphoma Coagulation cells were revealed. We did not examine her for Prothrombin time 97% (70~130) IgH-BCL rearrangement of peripheral blood. B Activated partial thromboplastin time 28.5 sec (24.3~36.0) symptoms were absent; however, since she International normalized ratio 1.02 (0.85~1.15) complained of bone pain and it was progres- Fibrinogen 274 mg/dL (150~400) sive, we decided to treat her with chemothera- Bone marrow 4 py. R-CHOP chemotherapy consisted of ritux- Number of nuclear cells 9.4 x10 /μL (10~20) Megakaryocytes 43.8x104 / L (50~150) imab (375 mg/m2, day 0), cyclophosphamide μ 2 2 Myeloid/erythroid ratio 2.36 (2~4) (750 mg/m , day 1), doxorubicin (50 mg/m , day Blasts 0.2% (1.0~1.5) 1), vincristine (1.4 mg/m2, day 1), and pred- Promyelocytes 0.6% (2.0~6.5) nisolone (100 mg/kg body weight, day 1-5), injected tri-weekly until 6 courses were completed. At this point only a partial response (PR) was observed through objective radiologi- cal evaluations (for nodal masses FDG-PET was Past retrospective analysis reported that, in negative and more than 50% regression was Discussion general, MP occurs in patients in their 60s observed on CT). After R-CHOP, 2 courses at a 3- (median age of occurrence), has slightly high- week interval of the ACES regimen (cytarabine Mesenteric panniculitis is a lipodystrophy er incidence in men, is accompanied by 2,000 mg/m2 day 5, carboplatin 100 mg/m2 day 1- characterized by non-specific inflammatory abdominal pain, and is often associated with 4, etoposide 80 mg/m2 days 1-4, methylpred- disease first reported by Ogden et al.2 Emory et malignant lymphoma.8,9 Generally, isolated MP nisolone 500 mg/body days 1-5) were used as al. reviewed 84 cases of MP pathologically and does not need to be treated.4 Kipfer et al. salvage chemotherapy, after which her disease documented that a mixed histology of 3 compo- reported that 8 among 53 reviewed cases of MP was considered to be in CR (IgH-BCL rearrange- nents (fibrosis, chronic inflammation, and fat with malignant lymphoma led to complications ment in her lymph nodes and bone marrow necrosis) existed in the lesions, which could related to MP.5 They could find no pathological became negative). At the same time, a contrast- be classified according to the dominant patho- involvement of lymphoma in the panniculitis enhanced peritoneal mass appeared on CT logical changes, i.e. mesenteric lipodystrophy lesion in any case. In 3 of the 8 cases, MP imaging (Figure 1A) and FDG-PET (Figure 1B). (ML) type, mesenteric panniculitis (MP) type, existed with malignant lymphoma concomi- We identified the pathological etiology of the and sclerosing mesenteritis (SM) type. tantly. In 2 other patients, the lymphoma lesion peritoneal lesion proceeding to the peripheral However, not all cases can be definitively cate- was found two or three months later. Another 3 blood stem cell transplantation (PBSCT) by CT- gorized in these 3 subtypes because the histo- patients were diagnosed as having malignant guided biopsy, which revealed mesenteric pan- logically dominant subtype is often too varied lymphoma occurring in a different area from niculitis (MP) not concomitant with the lym- to define as ML, MP, or SM, respectively.3 Due the MP lesion. These findings suggest that MP phoma lesion (Figure 1C). Her MP did not pro- to this pathological variation and the various can precede or proceed to lymphoma. However, duce symptoms through her treatment course. underlying diseases of MP, the etiology of this no panniculitis has been revealed in lym- However, the bone lesion recurred in June, disease is thought to be a single pathology phoma lesions by biopsy, and no obvious asso- 2005, when she received PBSCT. She has main- with various causes including immunological ciation has been reported until now. In our tained CR with no progression of the MP mass reactions, malignancies, infection, physical case, the mesenteric lesion progressed and for more than three years. stimulation, or traumatic disorders. was detected by FDG-PET imaging during

[Hematology Reviews 2009; 1:e17] [page 95] Case Report chemotherapy, which reduced to remission the PET-negative subjects and 88% sensitivity (7 Ann Surg 1965; 161:864-75. lymphoma lesions (lymph nodes and femoral out of 8 subjects). We concluded that mesen- 7. Perez-Ferriols, Feber-Bosch. Primary soft bone head). This discrepant course made it teric lesions suspected of progressing or recur- tissue lymphoma associated with mesen- problematic for physicians to decide on a treat- ring with positive uptake with FDG-PET should teric lipodystrophy. Am J Dermato-pathol ment plan. If we had regarded the mesenteric be diagnosed pathologically even if asympto- 1993;15:363-7. mass in our patient as refractory lymphoma, matic. 8. Delgado Plasencia L, Rodríguez Ballester we would have had to administer a salvage reg- L, López-Tomassetti Fernández EM, et al. imen. If the mesenteric mass had been benign, Mesenteric panniculitis: experience in our we could have continued the existing regimen. center. Rev Esp Enferm Dig 2007;99:291-7. A biopsy confirmed the remission of her lym- References 9. Béchade D, Durand X, Desramé J, et al. phoma and resolved this problem. Autologous Etiologic spectrum of mesenteric panni- PBSCT brought about a durable CR in the 1. Jura V. Sulla mesenterite retrattile e scle- culitis: report of 7 cases. Rev Med Interne patient. The MP lesion regressed after the rosante. Policlinico (sez. prat) 1924;31: 2007;28:289-95. planned treatments, with no recurrence for 575-8 10. Juweid ME, Stroobants S, Hoekstra OS, et three years. Therefore, we recognized that her 2. Ogden WW 2nd, Bradburn DM, Rrives JD. al. Imaging Subcommittee of International MP was chemotherapy-related panniculitis, Panniculitis of the mesentery. Ann Surg Harmonization Project in Lymphoma. Use since panniculitis is thought to be brought 1960;151:659-68. of positron emission tomography for about by physical or chemical stimulation.5 3. Emory TS, Monihan JM, Car NJ, Sobin LH. response assessment of lymphoma: con- Although FDG-PET was an effective and Sclerosing mesenteritis, mesenteric pan- sensus of the Imaging Subcommittee of powerful tool for the staging, localization, and niculitis and mesenteric lipodystrophy: a International Harmonization Project in response assessment of malignant lymphoma, single entity? Am J Surg Pathol 1997;21: Lymphoma. J Clin Oncol 2007;25:571-8. the differential diagnosis for lymphoma 392-8. 11. Janikova A, Bolcak K, Pavlik T, et al. Value lesions with physiological accumulation or 4. Horton KM, Lawler LP, Fishman EK. CT of [18F]fluorodeoxyglucose positron emis- inflammatory change is sometimes difficult.10,11 Findings in sclerosing mesenteritis (pan- sion tomography in the management of Though FDG-PET/CT fusion imaging is more niculitis): spectrum of disease. Radio - follicular lymphoma: the end of a dilem- efficient for diagnosis of the active lymphoma graphics 2003;23:1561-7. ma? Clin Lymphoma Myeloma 2008;8:287- lesion,11 a pathological diagnosis via biopsy is 5. Kipfer RE, Moertel CG, Dahlin DC. 93. essential for cases in which a false positive is Mesenteric lypodystrophy. Ann Intern Med 12. Zissin R, Metser U, Hain D, et al. suspected. Zissin et al. reviewed 19 oncological 1974;80: 582-8. Mesenter ic panniculitis in oncologic patients with MP12 and reported that FDG-PET 6. Ogden WW, Bradburn DM, Rives JD. patients: PET-CT findings. Br J Radiol had high specificity (100%) among 11 FDG- Mesenteri Panniculitis: review of 27 cases. 2006;79:37-43.

[page 96] [Hematology Reviews 2009; 1:e17] Hematology Reviews 2009; volume 1:e18

Massive pleural effusion due to The effusion persisted, necessitating therapeutic drainage which produced four liters of Correspondence: Kathryn J. Lang, Department of IgG multiple myeloma bloodstained fluid. Computed tomography Haematology, Wycombe General Hospital, UK. (CT) of the chest was radiologically suggestive E-mail: [email protected] Kathryn J. Lang,1 Surjit Lidder,2 of mesothelioma (Figure 2) but cytological 1 Key words: hematology, multiple myeloma, Robin Aitchison analysis of the pleural fluid showed plasma pleural effusion. 1Department of Haematology, Wycombe cells consistent with myeloma (Figure 3). General Hospital, UK; 2Department of Subsequent immunohistochemistry Received for publication: 1 October 2009. Revision received: 21 October 2009. Trauma and Orthopaedics, Royal London revealed κ light chain restriction and VS38 positivity. A diagnosis of pleural effusion sec- Accepted for publication: 22 October 2009. Hospital, Whitechapel, London, UK ondary to multiple myeloma was made and the This work is licensed under a Creative Commons patient was commenced on bortezomib and Attribution 3.0 License (by-nc 3.0). dexamethasone therapy seven days after first ©Copyright K.J. Lang et al., 2009 Abstract presenting. No further pleural aspiration was necessary from this point and repeat chest Licensee PAGEPress, Italy. Hematology Reviews 2009; 1:e18 radiography (Figure 4) taken seven days after doi:10.4081/hr.2009.e18 Pleural effusion directly attributable to mul- starting treatment showed resolving pleural tiple myeloma is exceedingly uncommon and is disease and the patient reported improved said to occur in only 1-2% of cases. Of these symptomatology. around 80% occur in IgA disease. We report a tion but to related pathologies such as pul- case of myelomatous pleural effusion (proven monary embolus, heart failure and nephrotic on cytological and immunohistochemical analy- syndrome. Effusion directly due to myeloma is sis) in a patient with the IgG-κ subtype. We Discussion said to occur in less than 1% of cases with describe the diagnosis, pathogenesis and man- approximately 80% of these being in IgA dis- agement of this condition and show the radio- Pleural effusion is uncommon in multiple ease.3 logical and cytological evidence of the case. myeloma; it has been suggested that it occurs Involvement of the cavities is an ominous in approximately 6% of cases. Most of these feature in multiple myeloma and thought to are not directly due to myelomatous infiltra- represent either a late manifestation in the Introduction

Multiple myeloma is a clonal B-cell malignancy characterized by proliferation of plasma cells accumulating mainly in the bone marrow and secreting paraprotein.1 Myeloma accounts for around 1% of all malignancies and UK national statistics for 2006 report an incidence of 6.4 cases per 100,000 annually.2 Mean age at diagnosis is 65 years. Pleural effusion directly attributable to myeloma is exceedingly rare (1-2% of cases).3 We present a case of pleural effusion arising rapidly in a patient with long-standing multiple myeloma and found to be a myelomatous effu- Figure 1. Chest radiographs showed a large Figure 3. Cytological analysis of the pleu- sion. left sided pleural effusion. ral fluid showed plasma cells consistent with myeloma.

Case Report

A 74-year old man with longstanding IgG kappa myeloma presented with a two week history of increasing dyspnoea on exertion. Blood tests showed a normocytic anemia (Hb: 7.4 g/dL, MCV: 90.1 fl), and renal impairment (urea: 9.2 mmol/L, creatinine: 248 µmol/L). Serum electrophoresis suggested rapid disease progression with an increase in paraprotein from 14 g/L to 50 g/L over a six week period. Chest radiographs showed a large left sided pleural effusion (Figure 1). Diagnostic tap of the effusion produced heavily blood stained Figure 2. Computed tomography of the Figure 4. Repeat chest radiography taken fluid with features suggestive of an exudate chest was radiologically suggestive of me - seven days after starting treatment showed (pH: 7.62, total protein: 93 g/L, LDH: 506 U/L). so thelioma. resolving pleural disease.

[Hematology Reviews 2009; 1:e18] [page 97] Case Report natural history of the disease or a feature of therapeutic and diagnostic thoracocentesis. the aggressive behavior of myeloma. Once myelomatous pleural effusion is con- References Diagnosis is based on cytological analysis and firmed on cytology, and/or immunohistochem- immunohistochemical staining, pleural biopsy istry, systemic chemotherapy is required to 1. Provan D, Singer CRJ, Baglin T, Lilleyman may also yield a diagnosis of myeloma. slow the advance of the disease. There is as J. Oxford Handbook of Clinical Haemato- Malignant plasma cells within the cytological yet no consensus as to how best to manage logy, 2nd ed. Oxford University Press, preparation show a typical basophilic cyto- this rapid and aggressive phase of disease but 2004. plasm with large, eccentric nuclei and promi- we anecdotally show evidence of a good 2. UK National Cancer Statistics 2006, nent nucleoli.4 response to bortezomib and dexamethasone Cancer Research UK. http://info.cancerre- Several possible mechanisms are postulat- therapy. searchuk.org/cancerstats/types/multiple- ed for myelomatous pleural effusion: invasion We suggest that this case highlights the myeloma/incidence/ Accessed 01/10/2009. from adjacent skeletal lesions; extension from vital importance of diagnostic clarity, where 3. Shirdel A, Attaran D, Ghobadi H, Ghiasi T. chest wall plasmacytomas; tumor infiltration possible, in finding the underlying etiology of Myelomatous pleural effusion. Tanaffos, of the pleura; and mediastinal lymph node a pleural effusion. Myelomatous pleural effu- 2007;6:68-72. involvement causing lymphatic obstruction.4,5 sion carries a poorer prognosis and more work 4. Gogia A, Agarwal PK, Jain S, Jain KP. Currently, the exact mechanism of pathogene- is needed to understand the exact pathogene- Myelomatous pleural effusion. J Assoc sis remains unknown. sis of this condition. Phys India, 2005;5:734-6. The management is two-stage. First, the 5. Yu FC, Chang JY, Hwang SH, et al. Kappa- effects of the effusion must be alleviated with light chain multiple myeloma with myelomatous pleural effusion: Case report and review of the literature. J Med Sci 1994;15: 49-56.

[page 98] [Hematology Reviews 2009; 1:e18] Hematology Reviews 2009; volume 1:e19

The progress of prothrombin more than 150 anticoagulant compounds and found one particularly active molecule, which Correspondence: Juha Horsti, Centre for time measurement was named warfarin (a 4-hydroxy compound) Laboratory Medicine, Tampere University after the patent holder, the Wisconsin Alumni Hospital, P.O. Box 2000, FIN-33521 Tampere, Juha Horsti Research Foundation. Professor Paul Owren Finland. E-mail: [email protected] Centre for Laboratory Medicine, Tampere found factor V and developed a new PT method8 University Hospital, Tampere, Finland subsequent to the “Quick method” to over- Key words: anticoagulant therapy, INR, oral anti- come its drawbacks. The Owren reagent (com- coagulation, vitamin K antagonist, warfarin. bined thromboplastin reagent, Thrombotest) is Received for publication: 8 June 2009. used mainly in the Nordic countries, Benelux, Revision received: 8 September 2009. Abstract and Japan. By reason of its different reagent Accepted for publication: 21 October 2009. composition, the Quick technique is sensitive Warfarin is the most widely used medicine to the coagulation factors fibrinogen, II, V, VII, This work is licensed under a Creative Commons Attribution 3.0 License (by-nc 3.0). for oral anticoagulant therapy (OAT). It and X, while the Owren technique is affected by deficiencies in factors II, VII, and X. The Quick inhibits the synthesis of coagulation factors II, ©Copyright J. Horsti 2009 VII, IX, and X in the liver and results in the pro- method measures factor V and fibrinogen, Licensee PAGEPress, Italy duction of inactive or partially active versions which are not dependent on warfarin therapy, Hematology Reviews 2009; 1:e19 of these factors. Inactive coagulation factors and this constitutes a drawback for OAT. doi:10.4081/hr.2009.e19 interfere with prothrombin time measurement (Quick and Owren PT) measuring the sum of decreases from 13 to 9, only 70% of the activi- coagulation activity and inhibition. The nar- ty of the clotting factor remains; when one row therapeutic range here involves a danger Warfarin of serious complications and the risk of bleed- molecule contains six carboxylated residues, 12 ing or thrombosis. The new-generation PT The coumarins or VKAs have been the main- only 2% activity is present. The liver excretes method can measure coagulation activity and stay of OAT for over 50 years. Warfarin is the both active and inactive coagulation factors to inhibition separately. This new technique pro- most widely used drug for OAT world-wide, and the plasma and both affect International 13 motes patient care and anticoagulant medica- it has a predictable onset and duration of Normalized Ratio (INR) measurement, which tion (warfarin, dicoumarol) based on coagula- action and excellent bioavailability. The thera- possibly has escaped notice in the World tion activity in vivo. Both therapy and laborato- py is effective for a variety of clinical indica- Health Organization (WHO) recommendation ry controls should be unquestionably accurate tions of anticoagulation. The dosage and the for the prothrombin time methodology. and based solely on in vivo coagulation activi- clinical response still vary markedly among ty. Inactive coagulation factors (inhibition) patients, depending on genetic inheritance, render measurement, calibration, and harmo- age, and metabolism. nization. The use of the new-generation PT Warfarin is a racemic mixture of two opti- International Normalized Ratio method based on measurement of coagulation cally active isomers, the R and S forms. It is activity in vivo could develop vitamin K antag- rapidly absorbed from the gastrointestinal The calculation formula is for: INR = (sam- ISI onist (VKA) therapy for the marked benefit of tract (within 90 min) and its half-life is 36-42 plesec/normalsec) , where ISI is the patients. h. In the circulation warfarin is bound to plas- International Sensitivity Index. When the ISI is ma proteins (mainly albumin). It accumulates near 1.0, the reagent is sensitive and ISI has rapidly in the liver, where it is metabolized.9 little meaning in the INR calculation. Warfarin interferes with the cyclic intercon- The Quick and Owren PT methods are the History version of vitamin K and its 2,3 epoxide (vita- most common and generally accepted means of min K epoxide) and thus is a VKA. Vitamin K is monitoring VKA therapy. A comparison Discovery of vitamin K antagonists (VKAs) a cofactor in the carboxylation of vitamin K- between the PT methods has been published in in anticoagulation occurred in the 1920s, when dependent coagulation factors. VKA inhibits a recent article14 and the Owren PT was superi- the veterinarian Frank Schofield studied the the synthesis of coagulation factors II, VII, IX, or. WHO recommends the use of the INR to hemorrhagic disease affecting cattle consum- and X in the liver, and they remain partially harmonize PT results and therapeutic ranges ing sweet clover. He ascribed the bleeding to a inactive unless 9 to 13 of the amino terminal globally both for patient care in clinical prac- toxin in the clover.1 Karl Link and his team glutamate (Glu) residues are carboxylated to tice and in the scientific literature, as the studied spoiled sweet clover and, in 1939, form the Ca2+-binding γ-carboxyglutamate units used formerly proved inadequate for extracted dicumarol and identified the hemor- (Gla) residues (Figure 1).9-11 This carboxyla- international communication.15,16 The chal- rhagic agent, which was the substance in the tion step renders the coagulation factors func- lenge is for global clinical laboratories to har- sweet clover affecting coagulation.2 The dis- tionally active: binding to Ca2+ and the phos- monize INR testing further to a level where covery of dicoumarol made it possible to inhib- pholipid surface.10 Therapeutic dosages of war- INR results are consistent regardless of the it thrombosis and to study anticoagulation farin decrease, by 30-50%, the total amount of methods used. therapy for humans in the early 1940s.3,4 In the each vitamin K-dependent coagulation factor The use of the INR system still involves diffi- decades 1930 and 1940 Professor Armand synthesized by the liver. The secreted mole- culties with sample citrate concentration,17,18 dif- Quick developed a routine prothrombin time cules from the liver are under-carboxylated, ferent reagents and thromboplastins,19 and (PT) coagulation test, which preceded the use resulting in diminished biological activity (10- instruments,20,21 and with ISI and “local ISI” cal- of VKAs.5,6 This has served as a basis for oral 40% of normal).11 ibration22-28 to harmonize results. Horsti and col- anticoagulant therapy (OAT) monitoring from Warfarin treatment reduces the number of leagues measured 150 samples from patients on the onset. However this first drug for OAT sub- Gla residues (normal, 9 to 13) per clotting fac- oral anticoagulation using seven commercial sequently was found to have drawbacks by rea- tor molecule, with a concomitant fall in coagu- reagents and four different calibrator kits. son of its long half-life. Karl Link7 synthesized lant activity. When the number of residues Agreement between INR results was poor.29

[Hematology Reviews 2009; 1:e19] [page 99] Article

dilutions42 without inactive coagulation fac- this method two measurements for one sample ISI calibration tors. In principle, calibrators should not con- are needed and the cost per sample is twice as tain inhibitory coagulation factors, which much. The average correction does not suffice The reagent manufacturer informs the ISI cause error in calibration.40 for accurate INR results. In addition, the num- value or the laboratory can measure the ISI ber of inactive coagulation factors varies with an ISI calibration kit (local calibration) markedly in different calibrator kits (manufac- for a PT reagent lot. The meaning of ISI as tured or local). This means difficulties in cali- power in the INR calculation equation is dis- The new-generation prothrom- bration and harmonization of different played in the previous section. bin time method reagents.29,41 The harmonization of INR results First, the aim in the original recommenda- using different PT methods and reagents is 16,30 problematic if attempted according to the WHO tion was to harmonize INR results by cali- Today the Quick and Owren PT methods recommendation.29,43 Measuring only the active brating reagent ISI values with Human measure the sum of active coagulation factors coagulation factors without inhibition harmo- Combined, which was the primary reference (FII, FVII, FX) and inhibition by their inactive nization between different reagents succeeds preparation (IRP code 67/40). Thromboplastins coagulation factor counterparts. The new-gen- well.41 from different sources (human brain, rabbit eration PT method can measure separately brain, rabbit lung, and ox brain) yield quite dif- both active coagulation factors (FII, FVII, FX) ferent levels of PT. The hierarchy of reference and the inhibition caused by inactive or par- thromboplastin preparations was presented by tially inactive coagulation factors.40 Inhibition 31 van den Besselaar and associates. As the varies markedly between individual patient Conclusions WHO calibration procedure was complex and samples, depending on the medication and demanding in the second stage, the recom- metabolism. Correction of inhibition must be New medications for OAT have been devel- mendation for ISI calibration was local calibra- made individually for each patient sample.40 In oped over a number of years and anticipated tion using certified lyophilized plasmas.32-37 A third possible approach for calibration was presented by van den Besselaar and group. This alternative means of determining the ISI involves the use of freshly pooled plasmas from 20 normal individuals and 60 patients receiving coumarin (OAT). Such numbers of samples are necessary to obtain a precise calibration line for ISI calculation. Freshly pooled plasma can be used to determine reagent or instrument ISI with acceptable precision, or as good a result as with the WHO calibration model.38 Poller and colleagues (European Concerted Action on Anticoagulation, ECAA) have compared local ISI calibration and “Direct INR” in the correction for locally reported INRs. For local INR correction (harmonization) they used seven normal plasmas and 20 from patients on warfarin. The results were variable.39 It is clear that plasmas from patients on warfarin therapy always involve varying amounts of inactive coagulation factors, depending on the level of anticoagulation and individual patient characteristics (see section on Warfarin). All these former calibration models involve the principle that the calibrator contains an average amount of inactive coagulation factors (inhibition), which means an average correction for patient INR results. Is the average correction of inactive coagulation factors appropriate for individual patient samples in a measuring range from 1 to 5 INR? Is it a fact Figure 1. The enzyme reactions involved in the metabolic function of vitamin K. Vitamin that calibrators, according to biochemical prin- K-dependent carboxylase catalyzes the transformation of peptide- (Factors II, VII, IX, and ciples, contain inactive coagulation factors X) bound glutamate residues (Glu) to γ-carboxyglutamate (Gla) residues in the presence (inhibition)? Do different reagents behave of vitamin K hydroquinone, carbon dioxide, and molecular oxygen (I). Vitamin K hydroquinone is oxidized in the reaction to vitamin K 2,3 epoxide. The reduction of the latter identically to active and inactive coagulation to vitamin K quinone is catalyzed by vitamin K epoxide reductase, which can use certain factors?29 In our earlier studies we measured dithiols as the reductants (II). Vitamin K quinone can be reduced to vitamin K hydro- inactive coagulation factors (inhibition) in quinone in the reactions catalyzed by either a dithiol-dependent (III) or an NAD(H)P- four different kits and noted conspicuous vari- dependent (IV) entzyme. Warfarin (WARF) blocks reactions II and III. (Uotila L. Recent findings on the functions and requirements of vitamin K in humans. Klinlab 1998;3:97- ation in inhibition.40,41 ISI calibration should be 101.) based on normal plasma and normal plasma

[page 100] [Hematology Reviews 2009; 1:e19] Article without laboratory test control in an effort to 3. Butt HR, Allen EU, Bollman JL. A prepara- of 3.2% vs. 3.8 % sodium citrate concentra- displace warfarin medication. The new medi- tion from spoiled sweet clover C, 3 methyl- tion on routine coagulation testing. Am J cines, however, have not proved their superior- ene-bis-(4 hydroxy-coumarin) which pro- Clin Pathol 1997;107:105-110. ity over warfarin. The new molecules are too longs coagulation and prothrombin time of 19. van den Besselaar AM. International stan- expensive for global use or involve serious blood: preliminary report of experiments dardisation of laboratory control of oral side-effects and possibly will never be as popu- and clinical studies. Mayo Clin Proc 1941; anticoagulant therapy: a survey of throm- lar as warfarin, which is an old and cheap 16:388. boplastin reagents used for prothrombin means widely used and accepted globally. The 4. Townsend SR, Mills ES. Effect of the syn- time testing. J Heart Valve Dis 1993;2:42- only drawback with warfarin medication is reg- thetic haemorrhragic agent 4-hydroxy- 52. ular laboratory control. It would be important to coumarin in prolonging the coagulation 20. Chantarangkul V, Tripoldi A, Clerici M, et develop warfarin therapy with greater atten- and prothrombin time. Can Med Assoc J al. Assessment of the influence of citrate tion to laboratory control, which helps patient 1942;46:214. concentration on the International Norm- care. Different reagents and thromboplastins 5. Quick AJ, Stanley-Brown M, Bancroft FW. A alised Ratio (INR) determined with twelve react variably with inactive coagulation factors study of the coagulation defect in hemo- reagent-instrument combinations. and cause difficulties in calibration using philia and in jaundice. Am J Med Sci 1935; Thromb Haemost 1998;80:258-62. patient plasmas that contain inhibiting coagu- 190:501-11. 21. Cunningham MT, Johnson GF, Pennell BJ, lation factors. Data on inactive coagulation 6. Quick AJ. The prothrombin time in hemo- et al. The reliability of manufacturer-deter- factors are individual and should be corrected philia and in obstructive jaundice. J Biol mined, instrument-specific international individually. Every new anticoagulant therapy Chem 1935;109:73-4. sensitivity index values for calculating the patient should be tested for inhibition at the 7. Link KP. The discovery of dicoumarol and international normalised ratio. Am J Clin commencement of therapy. its sequels. Circulation 1959;9:97-107. Pathol 1994;102:128-33. The accuracy and harmonization of patient 8. Owren PA. Thrombotest. A new method for 22. van Rijn JL, Schmidt NA, Rutten WP. INR results for different reagents is of the controlling anticoagulant therapy. Lancet Correction of instrument- and reagent- utmost importance for scientific publications, 1959;2:754-8. based differences in determination of the medication, and patient care. Today’s measur- 9. Ansell J, Hirsh J, Poller L, et al. The phar- International Normalised Ratio (INR) for ing principle is taking the sum of active coag- macology and management of the vitamin monitoring anticoagulant therapy. Clin ulation factors (FII, FVII, FX), and inhibition of K antagonists: the seventh ACCP Conf- Chem 1989;35:840-3. inactive coagulation factors is not a satisfacto- erence on Antithrombotic and Thrombo- 23. Ng VL, Levin J, Corash L, et al. Correction ry approach from the standpoint of accurate lytic Therapy. Chest 2004;126: 204-233. of instrument- and reagent-based differ- patient care. Active coagulation factors in vivo 10. Hirsh J, Fuster V, Ansell J, et al. American ences in determination of the Interna- are responsible for thrombosis and bleeding. Heart Association/American College of tional Normalised Ratio (INR) for moni- The errors in INRs are too great, which in Cardiology Foundation Guide to Warfarin toring anticoagulant therapy. Am J Clin many ways affects the success of medication Therapy. Circulation 2003;107:1692-734. Pathol 1993;99:689. and patient well-being, also using the same 11. Majerus PW, Tollefsen DM. In: Goodman 24. Talstad I. Why Is the standardisation of calibration (local calibration).29 The new-gen- and Gilman’s Pharmacological Basis of prothrombin time a problem? Haemostasis eration PT offers possibilities of more accurate Therapeutics. 11th ed. Brunton LL, Lazo 2000;30:258-67. INR results and patient care based on control JS, Parker KL, ed. New York: McGraw-Hill, 25. Kirkwood TB. Calibration of reference of active coagulation factors. The inhibition 2006, pp 1475-86. thromboplastins and standardisation of renders measurement, calibration, and harmo- 12. Goodnight SH Jr, Hathaway WE. Disorders the prothrombin time ratio. Thromb nization.29,40,41 The therapeutic INR ranges of Hemostasis and Thrombosis: a clinical Haemost 1983;49:238-44. guiding anticoagulant therapy using “old guide. 2nd ed. New York: McGraw-Hill, 26. Hermans J, van den Besselaar AM, methods” are based on the principle that both 2001, p 554. Loeliger EA, et al. A collaborative calibra- calibrators and patient samples have inactive 13. Suttie JW. Synthesis of vitamin K-depend- tion study of reference materials for coagulation factors (inhibition) on average, ent proteins. FASEB J 1993;7:445-52. thromboplastins. Thromb Haemost 1983; which compensate for each other in the final 14. Horsti J. Has the Quick or the Owren pro- 50:712-7. INR result. Thus the therapeutic ranges are thrombin time method the advantage in 27. Craig S, Stevenson KJ, Dufty JM, et al. available for the new-generation PT method. harmonization for the International Local INR correction: justification for a The therapeutic range lies between thrombo- Normalized Ratio system? Blood Coag simplified approach. J Clin Pathol 1997;50: sis and bleeding, and complications are seri- Fibrinolysis 2002; 13:641-6. 783-9. ous and general.44 The medication and care of 15. International Committee for Standard - 28. Stevenson KJ, Craig S, Dufty JM, et al. OAT patients must be rendered better and safe isation in Haematology. International System ISI calibration: a universally appli- using a more sensitive PT test. Committee on Thrombosis and Haemo- cable scheme is possible only when stasis. ICSH/ICTH recommendations for coumarin plasma calibrants are used. Br J reporting prothrombin time in oral antico- Haematol 1997;96:435-41. agulant control. Thromb Haemost 1985; 29. Horsti J, Uppa H, Vilpo J. Poor agreement References 53:155-6. between different prothrombin time 16. WHO Expert Committee on Biological International Normalized Ratio (INR) 1. Schofield FW. Damaged sweet clover: the Standardisation. Thirty-third Report. methods: comparison of seven commercial cause of a new disease in cattle simulating Technical Report Series 687. WHO Geneva reagents. Clin Chem 2005;51:553-60. hemorrhagic septicemia and blackleg. J 1983;81-105. 30. Hermans J, van den Besselaar AMHP, Am Vet Med Assoc 1924;64:553-75. 17. Horsti J. Preanalytical aspects of routine Loeliger EA, et al. A collaborative calibra- 2. Link KP. The anticoagulant from spoiled coagulation measurements. Scand J Clin tion study of reference materials for sweet clover hay. Harvey Lect 1943-4; Lab Invest 2001;61:167-8. thromboplastins. Thromb Haemost 1983; 39:162-216. 18. Adcock DM, Kressin DC, Marlar RA. Effect 50:712-7.

[Hematology Reviews 2009; 1:e19] [page 101] Article

31. van den Besselaar AMHP, Lewis SM, Action on Anticoagulation (ECAA). Minim - “Direct INR” methods on correction of Mannucci PM. Status of present and can- um lyophilised plasma requirement for ISI locally reported International Normalized didate international reference prepara- calibration. Am J Clin Pathol 1998; 109: Ratios: an international study. J Thromb tions (IRP) of thromboplastin for the pro- 196-204. Haemost 2007;5:1002-9. thrombin time: a report of the Sub- 36. Poller L, van den Besselaar AMHP, 40. Horsti J, Uppa H, Vilpo JA. A new genera- committee for Control of Anticoagulation Jespersen J, et al. A comparison of linear tion prothrombin time method for INR. (Letter). Thromb Haemost 1993;69:85. and orthogonal regression analysis for Open Med Chem J 2008;2:11-5. 32. Clarke K, Taberner DA, Thomson JM, et al. local INR determination in ECAA coagu- Assessment of value of calibrated lometer studies. Br J Haematol 1998;102: 41. Horsti J, Uppa H, Vilpo JA. A new-genera- lyophilised plasmas to determine Inter - 910-7. tion method for Quick and owren PT. Open national Sensitivity Index for coagulome- 37. Plesch W, van den Besselaar AM. Valid - Hematol J 2008;2:81-5. ters. J Clin J Clin Pathol 1992;45:58-60. ation of the International Normalized 42. Lindahl TL, Egberg N, Hillarp A, et al. INR 33. Poller L, Triplett DA, Hirsh J, et al. Ratio (INR) in a new point-of-care system calibration of Owren-type prothrombin Assessment of value of calibrated lyo- designed for home monitoring of oral anti- time based on the relationship between philised plasmas to determine Inter - coagulation therapy. Int J Lab Hematol PT% and INR utilizing normal plasma national Sensitivity Index for coagulome- 2009;31:20-5. samples. Thromb Haemost 2004;91:1223- ters. Am J Clin J Clin Pathol 1995; 103:358- 38. van den Besselaar AM, Witteveen E, 31. 65. Schaefer-van Mansfeld H, et al. Effect of 43. Horsti J. Comparison of Quick and Owren 34. Poller L, Barrowcliffe TW, van den plasma pooling on the International prothrombin time with regard to the har- Besselaar AMHP, et al. European Concert- Sensitivity Index of prothrombin time sys- monisation of the international nor- ed Action on Anticoagulation. A simplified tems. Blood Coagul Fibrinolysis 1998; statistical method for local INR using lin- 9:645-51. malised ration (INR) system. Clin Chem ear regression. Br J Haematol 1997; 98: 39. Poller L, Keown M, Ibrahim S, et al. Lab Med 2002;40:399-403. 640-7. European Concerted Action on Anti - 44. Odén A, Fahlén M. Oral anticoagulation 35 Poller L, Barrowcliffe TW, van den coagulation. Comparison of local Inter - and risk of death: a medical record linkage Besselaar AM, et al. European Concerted national Sensitivity Index calibration and study. Br Med J 2002;325:1073-5.

[page 102] [Hematology Reviews 2009; 1:e19] Hematology Reviews 2009; volume 1:e20

Confirmation of the validity of presence of Hb Barts would put the diagnosis on solid grounds. Hb Barts disappears a few Correspondence: Akram Al-Hilali, Consultant using birth MCV for the weeks after birth.1 Hematopathologist, Dubai Hospital, Dubai, UAE. diagnosis of alpha It is too expensive to perform HPLC or HBEP E-mail: [email protected] for all neonates, even in areas where frequen- thalassemia trait Key words: MCV, MCH, Hb Barts, high perform- cy of the α-thalassemia gene is high. So, if we ance liquid chromatography. Akram M. Al-Hilali, Aisha M. Al-Jallaf, can predict, with cheap tests, the presence of -thalassemia then Hb separation can be per- This work was presented in part at the 32nd Sajida Chunkasseril α formed for the highly probable cases only. This Congress of the International Society of Hematology Unit, Pathology Department, issue was addressed by work carried out in Haematology held in Bangkok, Thailand in October 2008. Dubai Hospital, Dubai, UAE Thailand,9 and in Jeddah, Saudi Arabia a few years ago.6 MCV below 90 and MCH below 30 Received for publication: 18 June 2009. were found to be of very high predictive value Revision received: 17 September 2009. in the Jeddah study. Accepted for publication: 21 October 2009. Abstract There was a need to repeat the study in This work is licensed under a Creative Commons another setting, where thalassemics are of Attribution 3.0 License (by-nc 3.0). Thirty-four blood samples of neonates in variable ethnic groups, like in Dubai. α-tha- Dubai, UAE, with an MCV below 90 fL were lassemia is present in the UAE nationals and ©Copyright A.M. Al-Hilali et al., 2009 Licensee PAGEPress, Italy. checked by high performance liquid chro- residents of Iranian, Arab, Pakistani and Hematology Reviews 2009; 1:e20 matography (HPLC) for hemoglobin variants Filipino origins in Dubai and it was expected doi:10.4081/hr.2009.e20 to confirm a previous study carried out in that neonates in such a population would make Western Province of Saudi Arabia which a good sample for repetition of the study and showed a very high predictive index of such confirmation of the validity of using the MCV MCV for alpha (α-) thalassemia minor (ATM). and MCH as predictive parameters in individu- tologists for various reasons. MCH below 30 pg was an additional factor als of various ethnic groups. As a control group, 26 neonates, who had which supported such a prediction. The Dubai MCV between 90 and 95 fL were also checked study confirmed the original finding with 100% by HPLC. Again, these were not selected and all of such neonates showing Hb Barts band. A consecutive cases with such an MCV level control group of 26 neonates with an MCV Materials and Methods were included into the control group. It took between 90 and 95 fl showed Hb Barts in only eight months to collect the entire test and con- 11 cases (42.3%). Of these, 6 (23.1%) were All full blood counts (FBC) performed for trol cases because only a minority of neonates preterm babies, expected to have higher MCV. neonates were checked. Counts were per- in Dubai have blood counts checked. Five cases (19.2%) had an MCH below 30 pg, formed using a Coulter machine model LH7D1 though MCV was 90 or higher. Three of the (Beckman Coulter, Brea, Ca, USA). Thirty-four preterm babies also had MCH below 30. The consecutive neonates, with MCV below 90 fL study confirmed the Saudi results in neonates. were checked by HPLC, using Waters 2695 Results It seems very highly probable that a term machine (Milford, Ma, USA), and short β-tha- neonate with MCV below 90 and MCH below 30 lassemia program, to look for Hb Barts band. All individuals in the test group had MCV has ATM. Blood films of all babies included in the below 90 fL (mean 85.35, range 77-88.9). MCH study were checked by the hematopathologist was below 30 in 33/34 patients (97%). One for red cell morphology. FBCs had not been patient (3%) had an MCH of 31.1 pg. Mean ordered for these babies because of any suspi- MCH was 27.82 and range 25.1-31.1. Introduction cion of thalassemia. They were random and Hb Barts was demonstrated in all the 34 judged to be necessary by the attending neona- cases (100%). It ranged from a trace (less than The presence of Hb Barts in blood at birth is a finding in α-thalassemia trait, intermedia (HbH disease) and major (hydrops fetalis).1,2 Table 1. Results on cord blood of test group. Diagnosis of α-thalassemia major is not a Hb Barts MCV MCH problem and it is incompatible with life3 as <90 ≥90 <30 ≥30 Total cases most of the hemoglobin will be Barts. Neonates with α-thalassemia intermedia have HbF, HbA Positive 34 0 33 1 34 and Hb Barts, usually more than 15%.4 This % 100 0 97 3 condition is easy to diagnose later in life, even Negative 0 NA 00 0 if not investigated at birth, due to the presence 000 of HbH.6 The problem lies with α-thalassemia trait, Table 2. Results on cord blood of test group. single or double gene deletion types. Most of these cases have normal HbA2 levels if Hb Barts MCV MCH checked after the age of six months,6,7,8 though <90 ≥90 <30 ≥30 Total cases red cell indices are thalassemic. This group, Positive NA 11 7 (3 preterm) 4 (3 preterm) 11 which makes up the bulk of gene carriers, will % 42.3 26.9 15.3 42.3 be difficult to confirm by Hb electrophoresis Negative NA 15 0 15 15 (HBEP) or by HPLC in childhood and adult life. % 57.6 0 57.6 57.6 If one can check HBEP or HPLC at birth then

[Hematology Reviews 2009; 1:e20] [page 103] Article

1%) to 10.6% (mean 4.65%). The 10.6% figure Figure 1. Investigational scheme was obtained in the baby with the lowest MCV for α-thalassemia screening of (77) and MCH (25.1) (Table 1), who had Hb neonates. level of 12.5 and later proved to have Hb H disease. MCV in the control group ranged from 90.1 to 94.8 fL (mean 92.27) and MCH from 28.7 to 31.7 (mean 30.33). Hb Barts was demonstrated in 11 cases (42.3%). It ranged from 2-6% in those cases. Fifteen cases (57.6%) were negative for Hb Barts. Analyzing the 11 Barts-positive cases further showed that 7 of them had MCH below 30 and 6 were preterm (3 of them had MCH below 30). Only one baby (3.8%) out of the 26 controls and positive Barts had MCH above 30 and was full term (Table 2). Blood films of all babies included in the test group showed numerous target cells and microcytosis, but no signs of active hemol- MCH is below 30 pg. It is noteworthy that Hb ysis. Barts is not only a diagnostic finding for α-tha- References lassemia, but it tends to disappear a few weeks after birth.10 Trying to confirm β-thalassemia 1. Weatherall DJ, Clegg JB. The thalassemia Discussion trait at a later age, when suspicion is raised by syndrome 4th edition. Blackwell Science, low red cell indices leading to erythrocytosis, Oxford, UK 2004. hematologists first resort to HBEP or HPLC, 2. Weatherall DJ. The molecular basis for After the work carried out in Jeddah, Saudi looking for low HbA2. However, low HbA2 is not 6 phenotypic variability of the common tha- Arabia, where the population is basically local 6 a common finding in this condition. Also, low lassemias. Mol Med Today 1995;1:15-20. with a low expatriate mix, as well as some vari- Hb A2 is found in other conditions, like βtha- 3. Henthorn J, Anionwu E, Brozovic M. ation in the origin of the local population itself, lassemia. Iron deficiency anemia may also Screening cord blood for sickle haemoglo- it was thought useful to the medical literature lower HbA2 to below normal level.12-14 Failing to binopathies. BMJ 1984; 289:479-80. to repeat the study at another site to confirm confirm the condition by this routine test, as in 4. Higgs DR. Alpha thlassemias. Baillieres the remarkable results of the original study. the majority of cases, one can try molecular Clin Haematol 1993;6:117-50. Medical literature does mention that α-tha- studies on the α-polypeptide gene to look for 5. Higgs DR, Bowden DK. Clinical and lassemia trait individuals are born with micro- deletions or mutations known to lead to α-tha- Laboratory Features of alpha-thalassemia 5,10,11 cytosis. However, there is no clear demar- lassemia trait. These deletions and mutations Syndromes. In: Steinberd MH, Forget PG, cation of the level of MCV and MCH where one are so numerous5,15,16 that, in practice, laborato- Higgs DR, Nagel RL, eds. Disorders of can expect, with high confidence, to find α- ries only try the common local deletions or Hemo globin: Genetic, Pathophysiology 3,5,11 thalassemia. The work carried out in mutations, using the proper probes for them, if and Clinical Management. Cambridge 9 Thailand took the MCV cut-off value of 95. In these are known from previous studies. In a University Press, Cambridge, UK. 2001: our experience, in both works performed in mixed population the task will be difficult, as p.431-69. Jeddah and in Dubai a MCV cut-off value of 95 one can imagine. The cost of these molecular 6. Al-Hilali A, Abu Saud K, Soufi S, La Cock will give a lot of negative cases and make the tests is certainly higher than Hb separation C. Birth MCV & MCH are very reliable pre- need for HPLC/HBEP much higher, thus studies. In parts of the world where α-tha- dictors of the presence or absence of increasing the cost while one of the objectives lassemia gene is of high frequency, the cost alpha thalassaemia in the neonate. Clin of this scheme is to lower costs. will be enormous if one wants to diagnose all Med Blood Dis 2009;2:1-4. The advantage of using MCV and MCH lev- carriers. It should be remembered that most 7. Kyriacou K, Kyrri A, Kalogirou E, et al. Hb els is that these values are easily available countries harboring a high frequency for the Bart's level in cord blood and alpha tha- from a simple automated blood count (FBC). If gene are also poor in resources.17 The simple lassemia mutations in Cyprus. Hemo- one can prove that there are limit levels for strategy we are recommending, proven by the globin 2000;24:171-80. these two parameters which carry a high prob- two studies in two different locations, is valid 8. Rugless MJ, Fisher CA, Stephens AD, et al. ability of α-thalassemia trait and HbH disease, and is recommended by the authors for WHO Hb Bart's in cord blood: an accurate indi- then there will be a small proportion of and local public health schemes to establish cator of alpha thalassemia. Hemoglobin patients for whom one needs to perform Hb the diagnosis and gene frequency. Further 2006;30:57-62. separation procedure, by alkaline electro- studies are being considered to try to split pop- 9. Tritipsombat J, Sanchaisuriya K, Fucha- phoresis or HPLC, to demonstrate Hb Barts, ulations with one, two and three gene dele- roen G, et al. Hb profiles and hematologic and thus prove α-thalassemia presence. It is tions/mutations from levels of MCV and MCH features of thalassemic newborns: appli- confirmed from this study that if MCV is below caused by them and the HPLC performed as a cation to screening of α-thalassemia and 90 and/or MCH is below 30 there is a high result of obtaining the low MCV and MCH fig- Hb E. Arch Pathol Lab Med 2008;132:1739- probability of over 95% that the case is α-tha- ures. The simple investigational scheme that 45. lassemia trait. MCV may be above 90 if the we suggest to be followed for the α-tha- 10. El-Hazmi MAF. Haemoglobinopathies, tha- baby is born premature. In such cases one can lassemia screening of neonates is shown in lassaemias and enzymopathies in Saudi extend the MCV level to 95, whether or not Figure 1. Arabia: the present status. Acta Haematol

[page 104] [Hematology Reviews 2009; 1:e20] Article

1987;78:130-4. Hemoglobin A2 concentration in iron defi- and Clinical Management. Cambridge 11. Sanchaisuriya K, Fucharoen S, Fucharoen ciency anemia. Acta Haematol 1971;45: University Press, Cambrige, UK. 2001: G, et al. A reliable screening protocol for 77-81. 941-57. thalassemia and hemoglobinopathies in 14. Kuczynski A. The relationship between 16. Vulliamy T, Kaeda J. Molecular Tech ni - pregnancy. Am J Clin Pathol 2005;123:113- the serum iron concentration and erythro- ques. In: Dacie & Lewis Practical Haema - 8. cyte haemoglobin A2 level. J Med 1971;2: tology, London: Churchill Livingstone, 9th 12. Wasi P, Disthasongchan P, Na-Nakron S. 136-42. ed. 2002.p.493-525. The effect of iron deficiency on level of 15. Old JM. DNA-based diagnosis of hemoglo- 17. Weatherall DJ, Clegg JB. Inherited hemo- haemoglobins A2 and E. J. Lab Clin Med bin disorders. In: Steinberg MH, Forget globin disorders: an increasing global 1968;71:85-91. PG, Higgs DR, Nagel RL, eds. Disorders of health problem. Bull World Health Organ 13. Steiner J, Marti HR, Dean D. Decreased Hemoglobin: Genetic, Pathophysiology 2001;79:704-12.

[Hematology Reviews 2009; 1:e20] [page 105] Hematology Reviews 2009; volume 1:e21

Best practices for transfusion complications of sickle cell disease.2 Some transfusion practices, such as perioperative Correspondence: Ted Wun, Division of Hematol - for patients with sickle cell transfusion,3 transfusion during pregnancy,4 ogy and Oncology, UC Davis Cancer Center, 4501 disease and chronic transfusion to prevent stroke in X Street, Sacramento, CA 95817, USA. children found to be high-risk from transcranial E-mail: [email protected] Ted Wun,1 Kathryn Hassell2 doppler (TCD) screening and/or because of pre- 5-7 Key words: sickle cell, transfusion, guidelines. 1UC Davis School of Medicine, vious stroke, are based on prospective randomized studies. Other practices, such as trans- Sacramento; 2University of Colorado Funding: supported by an R13 (HL090075) grant fusion for prolonged painful episodes, exchange at Denver, USA from the NHLBI and the Office of Rare Diseases transfusion for acute chest syndrome, and RBC (TW) and by the UC Davis Clinical and Trans - transfusion for those with pulmonary hyperten- lational Science Center and National Center for sion, are based on less firm evidence.2 Barriers Research Resources (UL1 RR024146). Abstract to chronic transfusion therapy include patient and family reluctance, inconvenience, intra- Received for publication: 7 September 2009. venous access issues, alloimmunization, red Accepted for publication: 27 November 2009. The β-globin gene mutation in sickle cell cell availability, infectious disease complica- anemia results in anemia and repeated bouts 8 This work is licensed under a Creative Commons tions, immune modulation, and iron overload. Attribution 3.0 License (by-nc 3.0). of vascular occlusion. The cumulative effect of Based on surveys from the Sickle Cell Adult these vasocclusive events is progressive dam- Provider Network (SCAPN), whereas transfu- ©Copyright T. Wun and K. Hassel, 2009 age to many organs including the kidneys, sion practice for patients with pregnancies is Licensee PAGEPress, Italy lungs, and brain. The transfusion of red blood fairly standard, it varies considerably in adults Hematology Reviews 2009; 1:e21 cells (RBC) can ameliorate many of these com- with ischemic stroke (K. Hassell, unpublished doi:10.4081/hr.2009.e21 plications, but can be associated with both communication). acute and chronic complications, including It was recognized that many practices and iron overload. The objective of the Best recommendations regarding transfusion of of Strong, Moderate, or Uncertain (Table 2) Practices in Transfusion Medicine for Patients RBC for patients with SCD were not evidenced that paralleled ones used in other guideline 13 with Sickle Cell Disease (SCD) Conference based. Nonetheless, it was felt important to recommendations. At the end of each presen- was to review the available published evidence highlight those that did have a basis in tation, comment was invited from the audience and clinical experience surrounding the use of prospective, randomized trials, and to catalog and other panel members. The final recom- RBC transfusions for sickle cell disease by a practices that were in common use but not evi- mendations were developed by consensus at panel of experts. The expert panel developed denced based. At a minimum, such a catalog the end of the conference with audience input. explicit clinical guidelines for the use of RBC in could serve as a basis for critically examining Panel members adhered to several princi- SCD patients. The panel also made recommen- current practices and setting research priori- ples while developing their recommendations. dations for further research. A set of guidelines ties for future studies. The overall purpose of First, in keeping with the theme, best prac- were produced for dissemination to pertinent the conference was to provide succinct best tices, as opposed to standard or practical ones, stakeholders. If implemented, these clinical practices guidelines for transfusion that had were to be recommended. The bar was to be set high to promote what was felt to be optimal pathways have the potential to optimize the use the potential to improve care.9 of red blood cell transfusions in SCD. medical care. Second, recommendations were to be evidence-based if possible, but could also be based on expert opinion. This was in recog- Materials and Methods nition that very little available data was truly Introduction evidenced-based. In the absence of data specific to adults, discussants were asked to extrap- This conference was conceived to be part of olate from data in children. Finally, panel mem- Sickle cell anemia was the first recognized a series of five best practices guidelines focus- bers were asked to be as prescriptive as possi- “molecular disease”. A single DNA base muta- ing on the care of adult patients with SCD. ble, without being dogmatic. tion, leading to a single amino acid substitu- This conference was not a “consensus” confer- tion (Glu to Val), results in hemoglobin S ence, nor was it an exhaustive, meta-analyses (HbS), the major hemoglobin defect, and a or Cochrane-type review of the literature. 1 myriad of clinical manifestations. Acute clini- Rather, it was a focused literature review with Results cal manifestations include painful crisis, acute key citations as determined by experts in the chest syndrome, stroke, and priapism. field. For a more comprehensive overview, the Irreversible major organ damage occurs to the reader is directed to several excellent reviews A set of guidelines were produced for dis- brain, heart, lungs, kidneys, eyes, and femoral on this topic.2,10-12 Several months prior to the semination to pertinent stakeholders. heads. Although anemia accounts for some of meeting, the conference coordinator (TW) the morbidity of sickle cell disease, vascular asked each invited expert (Table 1) to forward Recommendation #1: indications occlusion underlies most of the acute and succinct recommendations for each assigned for acute red blood cell transfusions chronic complications.1 The SCD phenotype is topic and provide key citations. These served as Transfusion is indicated for the therapy of predominantly expressed in those with the basis for the agenda. the acute clinical situations detailed in Table 3. homozygous S (HbSS), or compound heterozy- The conference took place September, 2007 This recommendation was strong for acute gotes of HbS and HbC (HbSC) or HbS and β- immediately preceding the annual Sickle Cell chest syndrome (ACS) with hypoxemia or wors- thalassemia (HbS/B0-thal or HbS/β+-thal). Clinical Research Meetings and was open to ening respiratory distress,14,16 ischemic stroke There is no doubt that both acute and chron- the public. Each expert was asked to present without or without post-infarct bleeding,8,17 ic RBC transfusions, both in the acute and her or his recommendations and assign a splenic and hepatic sequestration with worsen- chronic setting, can ameliorate many of the strength based on a simplified scoring system ing anemia, aplastic crisis with worsening ane-

[page 106] [Hematology Reviews 2009; 1:e21] Article mia, symptomatic acute blood loss, and multior- Table 1. Speakers and topics. 18 gan failure. It was acknowledged that there is Marilyn J. Telen, M.D., Wellcome Professor of Medicine and Chief of Hematology, Duke University neither a universal definition of acute chest Medical Center syndrome, nor a system for grading the severity Indications for acute and chronic red cell transfusions of ACS.14-16 Nonetheless, all panel members Naomi L.C. Luban, M.D., Professor of Pediatrics and Pathology, Chief of Laboratory Medicine and agreed that hypoxemia requiring supplemental Pathology, Children’s National Medical Center oxygen and/or worsening tachypnea were indi- Selection of the appropriate red cell product cations. It was also recognized that the term Paul V. Holland, M.D., Clinical Professor of Medicine and Pathology, UC Davis School of Medicine “significant decrease from baseline” in regards Infectious and immunologic complications of transfusion in patients with SCD to splenic and hepatic sequestration is subjec- Paul Swerdlow, M.D., Professor of Medicine, Wayne State University School of Medicine tive. This determination is best made by clini- Exchange versus simple transfusion, Hb and HbS target cians under specific clinical circumstances. For Elliot P. Vichinsky, M.D., Director, Hematology and Oncology, Children’s Hospital Oakland Research isolated intracerebral hemorrhage, transfu- Institute sions are of uncertain benefit and, anecdotally, Perioperative transfusions have exacerbated central nervous system Gary Brittenham, M.D., Professor of Medicine and Pediatrics, Columbia University College of Physicians (CNS) bleeding. Nonetheless, in cases where and Surgeons Assessment and Management of iron overload in patients with SCD there was an infarctive component to the bleed, some panel members felt that transfusion might be useful. Panel members were in more agreement about transfusion for transient Table 2. Definition of grading. ischemic attacks (TIA), but there are some data Strong suggesting that transfusions are not effective at At least one, prospective randomized study in adults or a randomized study where a non-transfusion prevention:19 however, in the absence of clinical arm was considered unethical by the panel trials, this recommendation was graded as only Moderate moderate. Silent cerebral infarcts (SCI) were Prospective, randomized data from children not addressed at this conference, but there are Prospective, non-randomized or observational studies in adults or children some data to suggest that transfusion may Definitive study is thought not feasible by the panel lower the incidence of SCI as defined by MRI.20 Uncertain However, as these data were generated in chil- Retrospective studies or case reports in adults or children dren who would be considered candidates for transfusion therapy regardless (because of high risk TCD findings), a separate recommendation fied as high risk by TCD examination.5,7 In may not be necessary. The indications for red Recommendation #3: acute red cell patients requiring general anesthesia with blood cell transfusion for severe infections and transfusion is not indicated for a evidence of cardiac, pulmonary, renal, or persistent priapism are based on less convinc- decrease in hemoglobin from hepatic dysfunction, many panelists and ing anecdotes and are thus uncertain.21 The baseline not associated with attendees also felt red cell transfusions were association of sickle cell disease, priapism, symptoms of anemia indicated. The panel discussed general anes- exchange transfusion, and neurological events In a certain sense, this recommendation is thesia is likely not the issue, but rather proce- in concert with #2. Patients with SCD have (ASPEN syndrome) was also of concern regard- dure times of 45-60 minutes or longer place 21,22 variable severity of chronic anemia. Physio- ing transfusion therapy for priapism. patients at higher risk for complications, pre- logical compensation includes high cardiac Because the ASPEN syndrome is based largely dominantly pulmonary. There was disagree- output and increased plasma volume. The on case reports, there was insufficient evidence ment as to what constituted a high risk surgi- practice of routine transfusion for an arbitrary to consider priapism a contraindication to cal procedure. There was agreement that open level of hemoglobin should be discouraged in transfusion. A recent report suggested that thoracic and vascular procedures were high SCD patients and all patients. The panel exchange transfusion was safe and efficacious risk, but hip and knee replacement might be 22 emphasized that patients with chronic anemia if the resulting hemoglobin level was <11 g/dL, considered moderate risk (similar to cholecys- may be more sensitive to volume challenge and suggests that ASPEN syndrome results from tectomy). Abdominal procedures may also be prone to develop symptomatic volume over- increased whole blood viscosity. associated with increased risk, particularly in load. Whether a patient tolerates a specific patients with HbSC disease.26 Anecdotal expe- level of hemoglobin depends on the clinical cir- Recommendation #2: red cell rience suggests that prophylactic transfusions cumstances (e.g., infection, rapidity of decline, transfusion is not indicated for can reduce the frequency of painful episodes co-morbid conditions) and the determination the therapy of uncomplicated acute in patients refractory to hydroxyurea, but only of whether the anemia is “symptomatic” is painful episodes regardless of as part of a comprehensive pain program. duration somewhat subjective. This recommendation is dependent on sound clinical judgment. The panel discussed that RBC transfusion Recommendation #5: chronic during otherwise typical painful episodes was Recommendation #4: prophylactic transfusion may be useful in common in both community and academic set- red cell transfusions stabilizing or reversing certain tings but there was little to no evidence to sup- end-organ dysfunction port this practice.23 The term uncomplicated in Red cell transfusions can be used in the this context means without the indications for prevention of sickle cell complications. The These include pulmonary hypertension, con- acute transfusion stated in recommendation panel felt that these should be strongly recom- gestive heart failure, and renal dysfunction. 3,24,25 #1. The panel felt that this practice should be mended for high risk surgical procedures, However, this recommendation is almost discouraged, therefore warranted a separate secondary prevention of stroke, and primary entirely based on personal experience and/or recommendation to emphasize the point. prevention of stroke in those children identi- anecdotal reports. Many participants observed

[Hematology Reviews 2009; 1:e21] [page 107] Article that aggressive local therapy was the most Table 3. Recommendations. effective treatment for non-healing leg ulcers. #1 As the sickle cell population ages and develops Acute complications requiring acute red cell transfusion more chronic end-organ dysfunction, this will Acute Chest Syndrome with hypoxemia, or increasing respiratory distress (Strong) be an important area for research. Without hypoxemia or increasing respiratory distress (Moderate) Acute stroke (ischemic with or without bleed) (Strong) Recommendation #6: type of red Isolated intracranial hemorrhage (Uncertain with caveat) cell product TIA (Moderate) Retinal Artery Occlusion (Moderate) The red blood cell products used for transfu- Acute splenic sequestration with a significant drop from baseline Hb (Strong) sion in patients with SCD should all be leuko- Acute hepatic sequestration (Strong) cyte-reduced, sickle hemoglobin negative, Aplastic crisis (Strong) matched for at least the red cell antigens Cc, Significant, symptomatic acute blood loss (Strong) Ee, and K and any antigens for which the Persistent priapism (Uncertain) patients has developed alloantibodies (past or Multiorgan failure syndrome (Strong) present). Alloimmunization to red cell anti- Acute life-threatening infection (Uncertain) gens is a common occurrence in transfused #2 SCD patients.27,28 Limited phenotype-matching Acute red cell transfusion is not indicated for the therapy of uncomplicated painful episodes regard- has been shown to decrease significantly the less of duration (Strong) rate of alloimmunization amongst non-alloim- #3 munized sickle cell patients.29 Recent studies Acute red cell transfusion is not indicated for a decrease in hemoglobin from baseline not associated showed that limited phenotype-matching is with symptoms of anemia (Strong) practiced by a minority of transfusion medi- #4 cine services surveyed in North America.30 Prophylactic red cell transfusions are indicated for Discussion centered on the fact that not all cel- High risk surgical procedures (Strong) Surgical procedure requiring anesthesia in a patient with evidence of significant end-organ dam- lular blood products in the US are pre-storage age (Moderate) leukocyte reduced. While this is the standard Cardiac, pulmonary, renal, hepatic insufficiency (Uncertain) in the European Union and the United Secondary prevention of stroke (Strong) Kingdom, only about 70% of blood banks in the For patients who had strokes as children and are now adults (Moderate) United States routinely stock pre-storage Primary prevention of stroke for patients who were identified as children as being high risk (Strong) leukocyte reduced products. Not all transfusion Refractory (to hydroxyurea), frequent painful episodes same evidence as a component medicine services verify that blood being of a comprehensive pain protocol (Moderate) transfused to a SCD patient is sickle hemoglo- Exacerbation of SCD during pregnancy (Moderate) bin negative. In areas where there is a large Consider if acute Hb < 6.0 g/dL (Moderate) African American blood donor population, this #5 issue becomes more relevant. Chronic transfusion may be useful in stabilizing or reversing Progressive congestive heart failure (Moderate) Recommendation #7: target Progressive chronic renal insufficiency (Uncertain) Progressive pulmonary hypertension (Uncertain) hemogoblin level Elevated tricuspid regurgitant jet (Uncertain) When transfusing red blood cells for acute Non-healing leg ulcer (Uncertain) indications, the target hemoglobin and hemat- #6 ocrit level should not exceed 10 g/dL and 30%, The red blood cells given to patients with sickle cell disease should be respectively. The panel and audience believed Leukocyte-reduced (pre-storage strongly preferred) (Strong) that there were sufficient rheological data and Antigen-matched for Cc, Ee, K and any antigens for which there is a clinically significant alloanti- rationale to avoid higher levels to prevent body currently present or in the past (Strong) increases in whole blood viscosity.31-33 In addi- Hemoglobin S negative (Strong) tion, there is some clinical evidence suggest- #7 ing that acute clinical deterioration occurs For acute sickle cell complications, the dose of red cells should be calculated to result in a Hb when whole blood viscosity is suddenly ≤10 g/dL or ≤Hct 30% (Strong) increased (ASPEN Syndrome noted above). #8 The panel acknowledged that much of the data The goal of exchange transfusion for acute indications is to achieve a HbA > 70% on post-transfusion were generated in vitro; nonetheless, this rec- hemoglobin fractionation (Strong) ommendation was strong. The preferred method is erythrocytapheresis with residual HbS < 30% (Strong) #9 Recommendation #8: target HbA Exchange transfusion is preferred over simple transfusion in adults for level Acute stroke (Strong) Severe acute chest syndrome (Strong) For acute exchange transfusions, a level of Multi-organ failure (Strong) HbA% >70% should be targeted. It was acknowl- Preoperative transfusion with baseline Hb > 10 g/dL (Moderate) edged that this has never been rigorously stud- Transfusion of volume sensitive patients (Moderate) ied. Establishing a target in terms of the per- Hyperosmolar contrast agents (Uncertain) centage of HbA rather than HbS was more sen- Refractory priapism (Uncertain) sible for compound heterozygous patients. For #10 example, it is simpler to target raising the HbA Every patient with sickle cell disease should be vaccinated for hepatitis A/B (Strong) to 70% in a patient with HbSC than to target a continued on next page

[page 108] [Hematology Reviews 2009; 1:e21] Article

Table 3. Continued. 2. Wanko SO, Telen MJ. Transfusion management in sickle cell disease. Hematol #11 Patients with sickle cell disease who are intermittently or repeatedly transfused should be periodi- Oncol Clin North Am 2005;19:803-8. cally evaluated for iron overload. Iron-chelating therapy with deferasirox or deferoxamine should be 3. Vichinsky EP, Neumayr LD, Haberkern C et used to maintain the body iron burden in an optimal range that reduces the risk of iron toxicity while al. The perioperative complication rate of avoiding adverse effects of excessive chelator administration (Strong) orthopedic surgery in sickle cell disease: #12 report of the National Sickle Cell Surgery Each patient with sickle cell disease should have a comprehensive transfusion record, including phe Study Group. Am J Hematol 1999;62:129- type, type and number of transfusions, and adverse events 38. #13 4. Koshy M, Burd L, Wallace D et al. Pro- There should be a system for sharing information on phenotype and alloantibody information amongst phylactic red-cell transfusions in pregnant regional blood providers patients with sickle cell disease. A randomized cooperative study. N Engl J Med HbS of 15% and HbC of 15%.11 The preferred sis and a target range of 3 to 7 mg Fe/g liver, 1988;319:1447-52. method for exchange is automated erythrocyta- dry weight, is recommended as prudent. Liver 5. Adams RJ, Brambilla DJ, Granger S et al. pheresis. In the absence of an apheresis serv- biopsy with its inherent risks is still the stan- Stroke and conversion to high risk in chil- ice or because of patient preference, manual dard, but MRI T2* can also be used and its dren screened with transcranial Doppler exchange is a suitable alternative.17 availability is only a function of proper calibra- ultrasound during the STOP study. Blood tion. Centers taking care of chronic transfu- 2004;103:3689-94. Recommendation #9: simple vs sion patients should develop the ability to offer 6. Adams RJ, McKie VC, Brambilla D et al. exchange transfusion this mode of iron assessment. Stroke prevention trial in sickle cell anemia. Control Clin Trials 1998;19:110-29. There are few data comparing simple to Recommendation #12 7. Adams RJ, McKie VC, Hsu L, et al. exchange transfusion. Exchange was preferred Prevention of a first stroke by transfusions strongly over simple transfusion for the treat- Every patient with sickle cell disease should in children with sickle cell anemia and ment of severe acute chest syndrome, stroke, have a portable comprehensive transfusion abnormal results on transcranial Doppler and multi-organ failure. A multi-center, ran- record, including phenotype, type and number ultrasonography. N Engl J Med 1998;339:5- domized study showed exchange transfusion of transfusions, alloantibody status and 11. to be non-superior to simple transfusion in the adverse events. Strongly advocated by the 8. Telen MJ. Principles and problems of perioperative setting.24 However, the panel patients present, this again would ensure transfusion in sickle cell disease. Semin endorsed the use of exchange as preferable in timely provision of appropriate and safer red Hematol 2001;38:315-23. advance of high risk operations. There was no cell transfusions. 9. Olson V, Coyne P, Smith V, Hudson C. clarity as to which modality was preferable in Critical pathway improves outcomes for the setting of acute priapism and use of hyper- Recommendation #13: information patients with sickle-cell disease. Oncol osmolar contrast agents, with the latter just to sharing Nurs Forum 1997;24:1682. be avoided with the availability of iso-osomot- There should be a system for sharing 10. Josephson CD, Su LL, Hillyer KL, Hillyer ic contrast agents. patient information on RBC phenotype, and CD. Transfusion in the patient with sickle alloantibody and autoantibody status amongst cell disease: a critical review of the litera- Recommendation #10: vaccination regional blood providers. While allowing for ture and transfusion guidelines. Transfus Every sickle cell patient should be vaccinat- privacy concerns, this was felt to be necessary Med Rev 2007;21:118-33. ed for hepatitis A and B as early as possible to to avoid further alloimmunization, hemolytic 11. Swerdlow PS. Red cell exchange in sickle prevent post-transfusion infection. This was transfusion reactions due to anamnestic cell disease. Hematology Am Soc Hematol non-controversial and strong. alloantibody responses (which may simulate Educ Program 2006:48-53. acute aplasia and crisis38), and help ensure 12. Minniti CP, Kratovil T, Luban NL. Recommendation #11: iron assess- availability of appropriately selected RBC prod- Children's National Medical Center's ment ucts. This recommendation was stressed by transfusion protocol for sickle hemoglo- patients and family members in attendance. binopathies. Immunohematology 2006;22: There was consensus that chronically trans- 117-20. fused sickle cell disease patients should be 13. Proceedings of the Seventh ACCP Confe - regularly assessed for iron overload and rence on Antithrombotic and Thrombolytic offered chelation therapy to decrease the risk Conclusions Therapy: evidence-based guidelines. Chest of iron related toxicity. There was considerable 2004;126:172S-696S. discussion regarding the number of lifetime 14. Golden C, Styles L, Vichinsky E. Acute transfusions that should trigger an assess- If implemented, these clinical pathways chest syndrome and sickle cell disease. ment of iron stores, and the optimal methods have the potential to optimize the use of red Curr Opin Hematol 1998;5:89-92. and extent of this assessment.34 The panel blood cell transfusions in SCD. 15. Morris C, Vichinsky E, Styles L. Clinician noted that some published recommendations assessment for acute chest syndrome in are an extrapolation from the experience with febrile patients with sickle cell disease: is β-thalassemia patients whose iron burden is it accurate enough? Ann Emerg Med 1999; significantly higher based on pathophysiology References 34:64-9. of their disease.35-37 Despite limited evidence, 16. Vichinsky EP, Neumayr LD, Earles AN et al. the consensus of the panel is that the hepatic 1. Francis RB, Johnson CS. Vascular occlu- Causes and outcomes of the acute chest storage iron concentration should be main- sion in sickle cell disease: current con- syndrome in sickle cell disease. National tained below a threshold of 15 mg Fe/g liver, cepts and unanswered questions. Blood Acute Chest Syndrome Study Group. N dry weight, to avoid progression to liver fibro- 1991;77:1405-14. Engl J Med 2000;342:1855-65.

[Hematology Reviews 2009; 1:e21] [page 109] Article

17. Wayne AS, Kevy SV, Nathan DG. Trans - Med 1995;333:206-13. megaly on the relationship between fusion management of sickle-cell disease. 25. Vichinsky EP. Current issues with blood venous haematocrit and red-cell volume. Blood 1993;81:1109-23. transfusions in sickle cell disease. Semin Br J Haematol 1964;10:567-75. 18. Hassell KL, Eckman JR, Lane PA. Acute Hematol 2001;38:14-22. 33. Schmalzer EA, Lee JO, Brown AK et al. multiorgan failure syndrome: a potentially 26. Neumayr L, Koshy M, Haberkern C et al. Viscosity of mixtures of sickle and normal catastrophic complication of severe sickle Surgery in patients with hemoglobin SC red cells at varying hematocrit levels. cell pain episodes. Am J Med 1994;96:155- disease. Preoperative transfusion in Implications for transfusion. Transfusion 62. Sickle Cell Disease Study Group. Am J 1987;27:228-33. 19. Pegelow CH, Adams RJ, McKie V et al. Risk Hematol 1998;57:101-8. 34. Harmatz P, Butensky E, Quirolo K et al. of recurrent stroke in patients with sickle 27. Vichinsky EP, Earles A, Johnson RA et al. cell disease treated with erythrocyte trans- Alloimmunization in sickle cell anemia Severity of iron overload in patients with fusions. J Pediatr 1995;126:896-9. and transfusion of racially unmatched sickle cell disease receiving chronic red 20. Pegelow CH, Wang W, Granger S et al. blood. N Engl J Med 1990;322:1617-21. blood cell transfusion therapy. Blood 2000; Silent infarcts in children with sickle cell 28. Orlina AR, Unger PJ, Koshy M. Post-trans- 96:76-9. anemia and abnormal cerebral artery fusion alloimmunization in patients with 35. Brittenham GM, Badman DG. Noninvasive velocity. Arch Neurol 2001;58:2017-21. sickle cell disease. Am J Hematol 1978; measurement of iron: report of an NIDDK 21. Merritt AL, Haiman C, Henderson SO. 5:101-6. workshop. Blood 2003;101:15-9. Myth: blood transfusion is effective for 29. Vichinsky EP, Luban NL, Wright E et al. 36. Brittenham GM, Cohen AR, McLaren CE et sickle cell anemia-associated priapism. Prospective RBC phenotype matching in a al. Hepatic iron stores and plasma ferritin CJEM 2006;8:119-22. stroke-prevention trial in sickle cell ane- concentration in patients with sickle cell 22. Ballas SK, Lyon D, Hall N, Kent T. Safety of mia: a multicenter transfusion trial. anemia and thalassemia major. Am J blood exchange transfusion for priapism Transfusion 2001;41:1086-92. Hematol 1993;42:81-5. complicating sickle cell disease. J Clin 30. Osby M, Shulman IA. Phenotype matching 37. Brittenham GM, Sheth S, Allen CJ, Farrell Apheresis 2006;21:16. of donor red blood cell units for nonalloim- DE. Noninvasive methods for quantitative 23. Ballas SK. Management of sickle pain. munized sickle cell disease patients: a sur- Curr Opin Hematol 1997;4:104-11. vey of 1182 North American laboratories. assessment of transfusional iron overload 24. Vichinsky EP, Haberkern CM, Neumayr L Arch Pathol Lab Med 2005;129:190-3. in sickle cell disease. Semin Hematol et al. A comparison of conservative and 31. Dintenfass L. Rheology of packed red blood 2001;38:37-56. aggressive transfusion regimens in the cells containing hemoglobins a-a, S-a, and 38. Diamond WJ, Brown FL, Bitterman P et al. perioperative management of sickle cell S-S. J Lab Clin Med 1964;64:594-600. Delayed hemolytic transfusion reaction disease. The Preoperative Transfusion in 32. Huber H, Lewis SM, Szur L. The influence presenting as sickle-cell crisis. Ann Intern Sickle Cell Disease Study Group. N Engl J of anaemia, polycythaemia and spleno - Med 1980;93:231-3.

[page 110] [Hematology Reviews 2009; 1:e21]