Treatment Options in Early Stages of Hodgkin's Lymphoma, High Cure Rate with Lower Short and Long-Term Toxicity

Hematology, 2005; 10 Supplement 1: 3/5

HODGKIN’S LYMPHOMA

Treatment options in early stages of Hodgkin’s Lymphoma, high cure rate with lower short and long-term toxicity

SANTIAGO PAVLOVSKY

Scientific Director, FUNDALEU, Centro de Internacioń e Investigacioń Clıńica ‘‘Ange´lica Ocampo’’, Buenos Aires, Argentina

Abstract The definition of early stages in HL varied among cooperative groups and clinical trials. Most of them considered early stages; stage I, II, and IIIA without bulky disease. Bulky disease has been defined at the Costwolds Meeting as those tumors with more than 10 cm or a mediastinal involvement of more than one-third of the chest wall diameter. Other factors that have been considered unfavorable within the early stages are old age, high ESR, mixed cellularity or lymphocyte depleted histology, B symptoms or multiple sites of disease.

Keywords: Hodgkin’s lymphoma, early stage, treatment

Brief history of the therapy of HL from 1960s to ABVD and four courses of standard BEACOPP, or 1980s for the comparison of 20 Gy IFRT with 30 Gy IFRT, with a 90% DFS rate and 97% OS rate at 2 years [1]. The standard treatment in pathological stage I/II HL The EORTC and GELA randomized 808 patients in 1960s and 1970s was total nodal irradiation in the H9-U trial in stage I/II with unfavorable (mantle and inverted Y) at 35 Gy to 40 Gy. The clinical features to 6 vs. 4 cycles of ABVD vs. 4 cycles combination of chemotherapy MOPP and MOPP like of BEACOPP baseline followed by 30 Gy IFRT in all regimens (COPP, CVPP, LOPP, etc.) were used in the the arms. The 4 years DFS were 94%, 89%, and 91% 1970s only in advanced stages or in patients relapsing in the 3 arms respectively (P/0.23). The 4 years OS to radiotherapy. ABVD emerged as a second-line rates were 96%, 89%, and 93% respectively (P/ treatment in patients who relapsed or were refractory 0.89) [2]. to MOPP. In the 1980s MOPP was used with ABVD as sequential, alternated or hybrid form in order to increase response rate and DFS. Number of cycles of chemotherapy The GATLA performed a randomized study in stage Best combination chemotherapy I /II without bulky disease of CVPP for 3 vs. 6 cycles without radiotherapy. The DFS at 5 and 10 years Randomized studies done by European and later by were 85% and 85% for these treated with 3 cycles; 93 USA groups demonstrated better results with ABVD and 73% for 6 cycles of CVPP. The OS at 10 years than MOPP, and similar results with ABVD versus were 89% in both arms concluding that 3 cycles of alternated or hybrid MOPP/ABVD with less myelo- CVPP without radiotherapy were equally effective supression, gonadal toxicity and second malignancies. than six cycles [3].

The German Hodgkin’s Disease Study Group The GHSG in a trial HD 10 compared in stage I/II (GHSG) performed the HD11 trial in 1047 inter- without risk factors 4 vs. 2 cycles of ABVD and 30 Gy mediate stage patients. It showed that there was no vs 20 GY of IFRT. The DFS and OS at 2 years in 847 difference for the comparison between four courses of evaluable patients was 97% and 98% respectively,

Correspondence: S. Pavlovsky Scientific Director, FUNDALEU, Centro de Internacio´n e Investigacio´n Clinica ‘‘Ange´lica Ocampo’’, Buenos Aires, Argentina.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389700 4 S. Pavlovsky with no statistical difference between ABVD 2 vs. 4, 0.001) being superior in the two arms with radio- and IFRT 30 Gy vs. 20 GY [4]. therapy.

Extended field vs. involved field radiotheraphy Second malignancies as consolidation after chemotherapy The combination of regimens that included alkylating The GHSG randomized 1064 patients with early agents as MOPP, BEACOPP, has shown to produce stage unfavorable prognosis after 4 cycles of alternate higher rate of AML/MDS, mainly during the first 5 COPP and ABVD to extended-field radiotherapy or years since treatment started. The incidence of second IFRT 30 Gy plus 10 Gy to bulky disease. The CR, solid tumor continued to increase compared to the DFS and OS were 98%, 83% and 91% at 5 years normal population more than 30 years later. However without difference between type of RT arms [5]. most of the long-term reports were in patients treated Bonadonna et al. [6] showed that in 136 patients two or three decades ago with MOPP like regimens with early stage of HL after 4 cycles of ABVD, 36 Gy and wider fields and higher doses of radiation, that is of subtotal nodal plus spleen irradiation were equiva- not the current practice today. lent to IFRT. The overall 12 years DFS and OS were 93% and 95% respectively. Conclusion The current treatment of choice for localized stages Chemotherapy vs. combined modality therapy without ‘‘bulky’’ disease is 2 or 4 cycles of combined The Canada Clinical Trial Group (CCTG) and the treatment with ABVD and low doses of radiotherapy Eastern Cooperative Oncology Group (ECOG) re- (20 to 25 Gy) to involved areas in partial responders ported the results of a randomized trial comparing or bulky disease at diagnosis. The new imaging techniques as positron emission tomography (PET) ABVD 4/6 cycles alone vs. ABVD plus STNRT 35 Gy in limited-stage HL. No difference was observed will help more sensitive discriminate between early in DFS (88% vs. 86%), or OS (94% vs. 96%) responders, late responders, and nonresponders with ABVD chemotherapy and tailored consolidation ra- between both treatment modalities [7]. diation therapy only for these late responders, and Strauss et al. [8] randomized 152 untreated stage nonresponders that probably will be less than 20% of I/IIIA nonbulky HL to 6 cycles of ABVD alone the early stages of HL. versus ABVD followed by 36 Gy of radiation. The 5 years DFS and OS for ABVD and ABVD/RT was 81% vs. 86% and 90% vs. 97% being the differences not significant. References In the current trial of the Argentine Group for [1] Klimm BC, Engert A, Brillant C, Eich H, Mueller-Hermelink Treatment of Acute Leukemia (GATLA), patients H, Herrmann R, Ho A, Hiddemann W, Greil R, Diehl V. Comparison of BEACOPP and ABVD chemotherapy in with stage I/IIIA without bulky disease and who achieved CR after the third cycle of ABVD received intermediate stage Hodgkin’s lymphoma: Results of the fourth interim analysis of the HD 11 trial of the GHSG. Proc ASCO IFRT 25 Gy. A total of 172 patients were evaluable, 23:561s, Abstr 6507, 2005. the DFS and OS at 60 months were 92% and 98% [2] Noordijk EM, Thomas J, Ferme C, van’t Veer MB, Brice P, respectively [9]. Divine´ M, Morschhauser F, Carde P, Eghbali H, Henry-Amar The GHSG compared in a trial, patients who M. First results of the EORTC-GELA H9 randomized trials: the H9-F trial (comparing 3 radiation dose levels) and H9-U received eight cycles of BEACOPP, 30 Gy IFRT or (comparing 3 chemotherapy schemes) in patients with favor- no radiation to bulk or residual disease. CR was able or unfavorable early stage Hodgkin’s Lymphoma. Proc achieved in 93%, the DFS for the total group was ASCO 23:561s, Abstr 6505, 2005. 89% and OS was 95%. There was no difference for [3] Pavlovsky S, Lastiri F. Progress in the prognosis of adult DFS or OS in an attempt to treat analysis between the Hodgkin’s lymphoma in the past 35 years through clinical trials in Argentina: A GATLA experience. Clin Lymphoma RT and no RT arm [10]. 2004;5:/ 102/ /109. The GATLA has compared 6 cycles of CVPP vs. 6 [4] Diehl V, Brillant C, Engert RP, Brillant C, Engert A, Nueller cycles of CVPP plus IFRT 30 GY in early stages of R, Mueller-Hermelink H, Hermann R, Doerken B, Kanz L, et HL without bulky disease. Also, the comparison of the al. H10 investigating reduction of combined modality treatment intensity in early stage Hodgkin’s lymphoma. Interim DFS and OS at 5 and 10 years showed no difference analysis of a randomized trial of the German Hodgkin Study (3). Group (GHSG). Proc ASCO 23:561s, Abstr 6506, 2005. The EORTC and GELA (2) in a H9-F trial [5] Engert A, Schiller P, Josting A, Herrmann R, Koch P, Sieber compared in 783 patients with early and favorable M, Boissevain F, De Wit M, Mezger J, Duhmke E, et al. disease who achieved CR after 6 cycles of EBVP, 36 Involved-ﬁeld radiotherapy is equally effective and less toxic compared with extended-ﬁeld radiotherapy after four cycles of Gy vs. 20 GY of IFRT vs. no radiotherapy. The 4 chemotherapy in patients with early-stage unfavorable Hodg- years DFS was 87%, 84% and 70% respectively (P B/ kin’s lymphoma: Results of the HD (trial of the German Treatment options in early stages of Hodgkin’s Lymphoma 5

Hodgkin’s lymphoma Study Group. J Clin Oncol 2003;21:/ / prospective randomized clinical trial of doxorubicin, bleomy-

3601/3608. cin, vinblastine, and dacarbazine (ABVD) followed by radia- [6] Bonadonna G, Bonfante V, Viviani S, Di Russo A, Villani F, tion therapy versus ABVD alone for stage I, II, and IIIA Valagussa P. ABVD plus subtotal nodal versus involved-ﬁeld nonbulky Hodgkin’s disease. Blood 2004;104:/ 3483/ /3489. radiotherapy in early-stage Hodgkin’s disease: Long-term [9] Pavlovsky S, Corrado CS, Pavlovsky MA, Giunta M, Zop- results. J Clin Oncol 2004;22:/ 2835/ /2841. pegno l, Prates V, Lastiri F, Cerutti I, Narbaitz MC, Rudoy S, [7] Meyer RM, Gospodarowicz MK, Connors JM, Pearcey RG, et al. Risk-adapted therapy in adults previously untreated Bezjak A, Wells WA, Burns BF, Winter JN, Horning SJ, Dar Hodgkin’s lymphoma with ABVD followed by involved ﬁeld AR, et al. Randomized comparison of ABVD chemotherapy radiotherapy. Blood 2004;104:369a, Abstr 1310. with a strategy that includes radiation therapy in patients with [10] Diehl V, Schieler P, Engert A, Wolf J, Josting A, Mueller RP, limited stage Hodgkin’s lymphoma: National Cancer Institute Eich HT, Gossmann A, Mueller-Hermelink HK, Franklin J, et of Canada, Clinical Trials Group and the Eastern Cooperative al. Results of the third interim analysis of the HD12 trial of the

Oncology Group. J Clin Oncol 2005;23:/ 21./ GHSG: 8 courses of BEACOPP with or without additive [8] Strauss DJ, Portlock CS, Qin J, Myers J, Zelenetz AD, radiotherapy for advanced stage Hodgkin’s lymphoma. Blood

Moskowitz C, Noy A, Goy A, Yahalom J. Results of a 2003;102(suppl/ 1):85./ Hematology, 2005; 10 Supplement 1: 6/9

NON-HODGKIN’S LYMPHOMA

Treatment of indolent lymphomas from watch and wait to high dose therapy

GERASSIMOS A. PANGALIS, THEODOROS P. VASSILAKOPOULOS, CHRISTINA KALPADAKIS, MARIA-CHRISTINA KYRTSONIS, STYLIANI I. KOKORIS, MARIA K. ANGELOPOULOU, & PANAYIOTIS PANAYIOTIDIS

First Department of Internal Medicine and Department of Haematology, National and Kapodistrian University of Athens School of Medicine, Laikon University Hospital, Athens, Greece

Indolent B-cell non-Hodgkin’s lymphomas (IBC- nodal MZ lymphomas of MALT type (EMZL), NHL) represent a heterogenous group of chronic focusing mainly on first-line therapies. diseases [1]. Follicular lymphomas grades 1 and 2 (FL1/2) are by far the most common, accounting for approximately 20% of all NHL worldwide. Indolent Follicular lymphomas, grades 1 and 2 IBC-NHL also include small lymphocytic lymphoma Ann Arbor stages I and II (SLL), which is the tissue counterpart of B-chronic lymphocytic leukemia (B-CLL); [2], lymphoplasma- This is the only subgroup of FL1/2, which is cytic lymphoma (LPL; 1%), mantle cell lymphoma considered curable, accounting for approximately (MCL; 6%), and marginal zone lymphomas, either 1/4 of the total patient population. Involved field splenic (SMZL; 1%), nodal (NMZL, 2%) or extra- radiotherapy (IF-RT) may cure approximately half of stage I and one quarter of stage II patients. However nodal MALT (8%). based on recent developments some questions may be Various treatment strategies ranging from ‘‘watch raised. Thus, it is not known whether IF-RT could be and wait’’ policies or oral alkylating agent monother- curative in patients with conventional stage I or II but apy to more aggressive combination chemotherapy with subclinical lesions in PET imaging. Further- (CT), chemoimmunotherapy or even CT followed by more, many of these patients have molecular evidence high dose therapy and autologous stem cell transplan- of disease dissemination as detected by the presence tation (ASCT) have been used in previously untreated of BCL-2 or immunoglobulin heavy chain gene patients with IBC-NHL. However there is no clear rearrangements in the blood or bone marrow DNA. evidence for the superiority of any particular approach It is not known whether such patients may be cured by in terms of overall survival (OS), because: (i) The and which is the place of rituximab in the potential natural history of IBC-NHL is generally prolonged eradication of residual disease. and many of these approaches are ultimately integrated in the overall treatment strategy, thus mini- mizing OS differences, despite significant differences Ann Arbor stages III and IV in progression free survival (PFS); (ii) With the Given that advanced FL1/2 are clearly incurable with exception of FL, the rarity of these disorders raises conventional CT, a ‘‘watch and wait’’ policy is applied further difficulties in the design of randomized trials. in patients with asymptomatic, non-bulky disease. We will review here current treatment approaches This approach is supported by randomized trials [3]. for FL1/2, MCL, SLL, LPL including Waldenstrom’s Chemotherapy based treatment should be insti- macroglobulinemia and NMZL, SMZL and extra- tuted when the patient develops constitutional symp-

Correspondence: Professor Gerassimos A. Pangalis, MD, First Department of Internal Medicine and Department of Haematology, National and Kapodistrian University of Athens School of Medicine, Laikon University Hospital, 17 Agiou Thoma Street, Athens 11527, Greece.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389755 Watch and wait to high dose therapy for indolent lymphomas 7 toms or symptoms related to tumor burden. There is a while the third [11] revealed an OS benefit despite plethora of treatment options (Table I), which differ similar PFS! Firm conclusions on OS cannot be with respect to response rates, PFS, and toxicity, but derived yet, but notably two trials demonstrated a OS differences are not demonstrable so far. significant increase of MDS/ANLL in ASCT-treated Elderly patients can be safely treated with oral patients. Furthermore, rituximab, which prolongs alkylating agents, such as intermittent chlorambucil, PFS compared with conventional CT alone, was not alkylator-based combination chemotherapy (CVP) or used in anyone of these trials. Thus, ASCT is still monotherapy with the anti-CD20 monoclonal anti- experimental in the first-line treatment of FL. In body rituximab. Younger patients can also be treated contrast, ASCT is clearly indicated in relapsed/ with such approaches, but many centers prefer to refractory disease. administer anthracycline-based CT (CHOP or similar Rituximab, when given as first-line monotherapy, regimens, MCP: mitoxantrone, chlorambucil and produces RR of approximately 70% [12]. Mainte- prednisone, etc) or CT based on both anthracyclines nance with 4 bimonthly infusions prolongs the PFS and purine analogues (FND, FCM etc). over rituximab induction alone (median 36 vs. 19 Several randomized trials have now convincingly months) [12]. It is however uncertain whether ritux- demonstrated that the addition of rituximab to con- imab maintenance is superior to induction followed ventional CT produces superior PFS rates, but by retreatment upon progression. Given the favorable follow-up is still short to reveal potential differences toxicity profile of rituximab and the potential of in OS [4/7]. The efficacy of rituximab in the setting durable responses, it can be considered as front- of relapsed/refractory FL [48% response rate (RR) line therapy at least in patients not eligible for with 13 mo median PFS] may in part obviate OS aggressive CT. benefits of chemoimmunotherapy over CT alone. For Recently, radioimmunotherapy with 131I-Tositumo- the time being, it seems reasonable to add rituximab mab (Bexxar) was shown to be very effective, not only to the first-line CT regimen, irrespectively of the for relapsed/refractory disease, but as first-line therapy intensity of the latter. as well. A single one week course of therapy produced A recent meta-analysis suggested that interferon- a RR of 95% with 75% CRs in 76 previously alpha (IFN-a) incorporated in the initial CT regimen untreated stage III/IV FL patients. Among CRs, and/or given as maintenance to responding patients 80% were extended at the molecular level as well, may result in superior PFS rates and a modest while the 5-year PFS was 59% and 5-year OS 89% increase in 10-year OS, in the order of 6/8% [8]. [13]. In the setting of relapsed/refractory disease OS benefits were apparent only in the subgroups of radioimmunotherapy with Bexxar or Y90 Ibritumo- patients receiving more intensive CT or higher doses mab tiuxetan (Zevalin), which can be administered in of IFN-a. However the individual randomized trials an outpatient basis, produces RR of 70/80% with included in this analysis were performed prior to the approximately 30% CR. Radioimmynotherapy even at introduction of rituximab. Thus the benefit of IFN-a myeloablative doses with stem cell support has also in the era of rituximab remains uncertain. produced very promising results in relapsed/refractory The value of autologous stem cell transplantation disease. (ASCT) incorporate in the first-line approach of FL has been tested in 3 randomized trials [9/11].Two of Mantle cell lymphoma them demonstrated significant prolongation of PFS, MCL resembles to the IBC-NHL in that there is a Table I. Treatment Options for Follicular Lymphomas-Grades 1 continuous pattern of relapse and no plateau in and 2 survival curves. However MCL can not be strictly Watch and Wait considered as an IBC-NHL, because median survival Oral alkylating agents (intermittent chlorambucil, etc) is short usually in the range of 2 to 4 years with B/10% Alkylator / based combination chemotherapy (CVP)*§ surviving at ten years after diagnosis. In contrast to Anthracyclin / based combination chemotherapy (CHOP and similar regiments, MCP)*§ FL, a ‘‘watch and wait’’ policy is not advisable in § Fludarabin / based combination chemotherapy (FND, FCM, etc) MCL except perhaps a minority of elderly patients Rituximab monotherapy with poor performance status. However approaches Chemotherapy followed by high dose therapy and autologous stem with very diverse toxicity profiles have been applied. § cell transplantation The preferred treatment approach is higly depended Radioimmunotherapy on patient’s age. *Interferon possibly added during and/or after chemotherapy, CHOP or similar regimens are preferred by most §Rituximab may be added in combination chemotherapy regiments, centers, while others also use fludarabine-based regi- CVP: cyclophosphamide, vincreistine, prednisone, CHOP: cyclo- mens. The addition of rituximab to CHOP moder- phosphamide, vincristine, doxorubicin, prednisone, MCP: mitoxantrone, chlorabucil, prednisone, FND: Fludarabine, ately prolonged PFS but had no effect on OS in a mitoxantrone, desamethasone, FCM: Fludarabine, cyclophospha- recent randomized trial [14]. The MD Anderson mide, mitoxantrone. group has reported impressive preliminary results 8 G. A. Pangalis et al. with the combination of rituximab and hyperCVAD the administration of IFN-a maintenance did not regimen [15]. In the absence of randomized trials, this improve these results [24]. Rituximab produces re- approach is considered experimental given that there sponses similar to that observed in FL and can be is no plateau in PFS curves, the high toxicity of the considered as monotherapy in these patients [25]. As regimen. Studies evaluating the first-line use of ASCT many as 70% of previously untreated patients respond are ongoing, but no definitive OS data have been to induction plus maintenance rituximab, with a published so far. Allogenetic SCT / either myeloa- median PFS of approximately 2.5 years. blative or based on reduced intensity conditioning Combination CT (CVP, CHOP) is usually de- regimens are the only potentially curable approaches served for relapsed disease while purine analogues and deserve further evaluation. may also have a role in the treatment of these patients, In contrast to these high-intensity approaches, although clear data are not yet available. some patients with MCL may achieve relatively durable remissions with chlorambucil monotherapy EMZL [16]. Thus, elderly asymptomatic patients without features of histologic aggressiveness (non-blastoid MALT lymphomas commonly arise in the stomach MCL) may be treated in this way. Rituximab mono- but also can be seen in other extranodal sites such as therapy may produce RR of 25/30% in both un- the skin, salivary glands, lung, ocular adnexa and treated and relapsed/refractory MCL patients, but thyroid. Gastric lymphomas respond well to antibiotic

CRs are very rare (/2%) and the median PFS is in therapy in conjunction with proton pump inhibitor the range of 6/12 months. Maintenance rituximab therapy. A considerable percentage of patients achieve does not appear to improve these results [17]. In the CR [26]. Radiation therapy or alkylating agent rare ‘‘splenic form’’ of MCL, splenectomy may be a therapy alone or in combination with Rituximab is reasonable first-line approach, delaying the adminis- reserved for non-responders. For non-gastric MALT tration of CT [18]. Novel agents, as the proteasome lymphomas the optimal management is not well inhibitor bortezomib and temsirolimus a rapamycin defined. Single alkylating agent or CVP regimen, kinase inhibitor that regultes cyclin-D1 translation are radiotherapy, immunotherapy, surgery alone or in effective in relapsed/refractory patients and require combination have been successfully used [27]. An- further evaluation [19,20]. thracycline based regimens do not seem to improve the response rate [28]. Splenic marginal zone lymphoma Conclusion SMZL is another entity for which the ‘‘watch and wait’’ policy was the preferred approach for patients IBC-NHL may be treated with various strategies, having asymptomatic splenomegaly or non significant without clear superiority of anyone of them. Evi- cytopenias. When treatment is needed, splenectomy is dence-based approaches are likely to emerge for FL still a reasonable option. Various CT regimens, (and probably MCL), but are difficult to be obtained including CHOP or fludarabine-based ones as well for SLL, LPL and marginal zone lymphomas. The as oral alkylating agents have been used, mainly in introduction of rituximab has revolutionized the splenectomy failures or in patients not eligible for treatment of IBC-NHL. ASCT and newer approaches splenectomy [21]. The selection of the CT regimen is including bordezomib and temsirolimus may improve at present arbitrary. The goal of treatment is to their eventual outcome. For the time being, only achieve a good response, but not necessarily a CR. ASCT can cure a fraction of relapsed/refractory Interesting preliminary results have been reported patients at the expense of a high rate of early with rituximab monotherapy, which may delay sple- mortality. nectomy [22]. The subgroup of patients with SMZL and hepatitis C virus infection achieve long-lasting partial remissions of excellent quality with interferon- Acknowledgements a and/or ribavirine without CT [23]. This study was supported in part by a grant provided by IASIS, a non proﬁt organization raising funds for research in leukemias, lymphomas and related dis- SLL, LPL/MW, NMZL orders. Specific data for these subtypes of IBC-NHL, particularly for NMZL, are lacking. We favor treatment of these patients, who present usually with advanced age, References with monthly intermittent chlorambucil for 1/2 years. [1] Jaffe ES, Harris NL, Stein H, Vardiman JW. World Health High-dose chlorambucil produced a RR of 72% with Organization classiﬁcation of tumors. Pathology and genetics 30% CRs in a recent randomized trial, with 5-year of tumors of haematopoietic and lymphoid tissues. Lyon: PFS rates of 20/30%. The addition of epirubicine or IARC Press; 2001. Watch and wait to high dose therapy for indolent lymphomas 9

[2] Pangalis GA, Angelopoulou MN, Vassilakopoulos TP, Sia- tositumomab therapy as initial treatment for follicular lym- kantaris MP, Kittas C. B-chronic lymphocytic leukemia, small phoma. N Engl J Med 2005;352:/ 441/ /449. lymphocytic lymphoma, and lymphoplasmacytic lymphoma, [14] Lenz G, Dreyling M, Hoster E, et al. Immunochemotherapy including waldenstom’s macroglobulinemia: A clinical, mor- with rituximab and cyclophosphamide, doxorubicin, vincris- phologic, and biologic spectrum of similar disorders. Semin tine, and prednisone signiﬁcantly improves response and time Hematol 1999;36:/ 104/ /114. to treatment failure, but not long-term outcome in patients [3] Ardeshna KM, Smith P, Norton A, Hancock BW, Hoskin PG, with previously untreated mantle cell lymphoma: results of a MacLennan KA, et al. Long-term effect of a watch and wait prospective randomized trial of the German Low Grade

policy versus immediate systemic treatment for asymptomatic Lymphoma Study Group (GLSG). J Clin Oncol 2005;23:/ /

advanced-stage non-Hodgkin lymphoma: a randomised con- 1984/1992. trolled study trial. British National Lymphoma Investigation. [15] Khouri IF, Romaguera J, Katarjian H, et al. Hyper-CVAD and Lancet 2003;362:/ 516/ /522. high-dose methotrexate/cytarabine followed by stem-cell [4] Hiddermann W, Dreyling MH, Forstpointner R, Kneba M, transplantation: an active regimen for aggressive mantle-cell Woermann B, Lengfelder E, et al. Combined immunochem- lymphoma. J Clin Oncol 1998;16:/ 3803/ /3809. otherapy (R-CHOP) signiﬁcantly improves time to treatment [16] Vandenberghe E, De Wolf-Peeters C, Vaughan Hudson G, et

Failure in ﬁrst line therapy of follicular lymphoma / results of al. The clinical outcome of 65 cases of mantle cell lymphoma a prospective randomized trial of the German Low Grade initially treated with non-intensive therapy by the British lymphoma study group (GLSG). Blood 2003;102:104a [ab- National Lymphoma Investigation Group. Br J Haematol

stract]. 1997;99:/ 842./ [5] Marcus R, Imrie K, Belch A, et al. CVP chemotherapy plus [17] Ghielmini M, Schmitz SF, Cogliatti S, et al. Effect of single- rituximab compared with CVP as ﬁrst-line treatment for agent rituximab given at the standard schedule or as prolonged advanced stage follicular lymphoma. Blood 2005;105:/ 1417/ / treatment in patients with mantle cell lymphoma: a study of 1423. the Swiss Group for Clinical Cancer Research (SAKK). J Clin [6] Herold M, Pasold R, Srock S, et al. Results of a prosprctive Oncol 2005;23:/ 705/ /711. randomized open label phase III study comparing rituximab [18] Angelopoulou MK, Siakantaris MP, Vassilakopoulos TP, et al. plus mitoxantrone, chlorambucile, prednisolone chemother- The splenic form of mantle cell lymphoma. Eur J Haematol apy (R-MCP) versus MCP alone in untreated advanced 2002;68:/ 12/ /21. indolent non-Hodgkin’s lymphoma (NHL) and mantle-cell- [19] Goy A, Younes A, McLaughlin, et al. Phase II study of lymphoma (MCL). Blood 2004;104:584a [abstract]. proteasome inhibitor bortezomib in relapsed or refractory B- [7] Salles G, Foussard C, Nikolas M. Rituximab added to aIFN / cell non-Hodgkin’s lymphoma. J Clin Oncol 2005;23:/ 667/ / CHVP improves the outcome of follicular lymphoma patients 675. with a high tumor burden: ﬁrst analysis of the GELA- [20] Witzig TE, Geyer SM, Ghobrial I, et al. Phase II trial of GOELAMS FL-2000 randomised trial in 359 patients. Blood single-agent temsirolimus (CCI-779) for relapsed mantle cell 2004;104:160a [abstract]. lymphoma. J Clin Oncol 2005; June 27 [Epub ahead of print] [8] Rohatiner AZ, Gregory WM, Peterson B, Borden E, Solal- [21] Thieblemont C, Felman P, Callet-Bauchu E, et al. Splenic Celigny P, Hagenbeek A, Fisher RI, Unterhalt M, et al. Meta- marginal-zone lymphoma: a distinct clinical and pathological

analysis to evaluate the role of interferon in follicular lym- entity. Lancet Oncol 2003;4:95/103. Review. phoma. J Clin Oncol 2005;23:/ 2215/ /2223. [22] Dimopoulou MN, Kalpadakis C, Dimitriadou EM, et al. [9] Lenz G, Dreyling M, Schiegnitz E, Forstpointner R, Wandt H, Rituximab treatment for splenic marginal zone B-cell lym- Freund M, et al. Myeloablative radiochemotherapy followed phomas (SMZL). Haematologica 2005;90:680 (abstract) by autologous stem cell transplantation in ﬁrst remission [23] Hermine O, Lefrere F, Bronowicki JP, et al. Regression of prolongs progression-free survival in follicular lymphoma: splenic lymphoma with villous lymphocytes after treatment of results of a prospective, randomized trial of the German hepatitis C virus infection. N Engl J Med 2002;347:/ 89/ /94.

Low-Grade Lymphoma Study Group. Blood 2004;104:/ / [24] Baldini L, Brugiatelli M, Luminari S, et al. Treatment of

2667/2674. indolent B-Cell nonfollicular lymphomas: ﬁnal results of the [10] Deconinck E, Foussard C, Milpied N, et al. High-dose LL01randomized trial of the Gruppo Italiano per lo Studio dei therapy followed by autologous purged stem-cell transplanta- Linfomi. J Clin Oncol 2003;21:/ 1459/ /1465. tion and doxorubicin-based chemotherapy in patients with [25] Hainsworth JD, Burris HA 3rd, Morrissey LH, Litchy S, advanced follicular lymphoma: a randomized multicenter Scullin DC Jr, Bearden JD 3rd, Richards P, Greco FA. study by GOELAMS. Blood 2005; prepublished Rituximab monoclonal antibody as initial systemic therapy [11] Sebban C, Belanger C, Brousse N, et al. Comparison of for patients with low-grade non-Hodgkin lymphoma. Blood CHVP /interferon with CHOP followed by autologous stem 2000;95:/ 3052/ /3056. cell blood transplantation with a TBI conditioning regimen in [26] Wotherspoon AC, Doglioni C, Diss TC, et al. Regression of untreated patients with high tumor burden follicular lym- primary low-grade B-cell gastric lymphoma of mucosa-asso- phoma: Results of the randomized GELF94 trial (GELA. ciated lymphoid tissue type after eradication of Helicobacter study group). Blood 2003;102:354a [abstract]. pylori. Lancet 1993;342:/ 575/ /577. [12] Ghielmini M, Schmitz SF, Cogliatti SB, et al. Prolonged [27] Kalpadakis C, Siakantaris MP, Kyrtsonis MC, et al. Non- treatment with rituximab in patients with follicular lympho- gastrointestinal extranodal marginal zone lymphomas masigniﬁcantly increases event-free survival and response (EMZL): A single center experience. Haematologica

duration compared with the standard weekly/4 schedule. 2005;90:682 (abstract). Blood 2004;103:/ 4416/ /4423. [28] Zucca E, Conconi A, Pedrinis E, et al. Nongastric marginal [13] Kaminski MS, Tuck M, Estes J, Kolstad A, Ross CW, Zasadny zone B-cell lymphoma of mucosa-associated lymphoid tissue. K, Regan D, Kison P, Fisher S, Kroll S, Wahl RL. 131I- Blood 2003;101:/ 2489/ /2495. Hematology, 2005; 10 Supplement 1: 10/14

NON-HODGKIN’S LYMPHOMA

Update on lymphoma management: Diffuse large B-Cell NHL

JULIE M. VOSE, NEUMANN M. & MILDRED E. HARRIS

Section of Hematology/Oncology, University of Nebraska Medical Center, Omaha, Nebraska

Keywords: Diffuse large cell, lymphoma, treatment

Introduction designed clinical trials. The development of novel therapies may result in improved outcomes for Diffuse large B-cell lymphoma (DLBCL) is the most patients diagnosed with these common NHL sub- commonly occurring subtype of non-Hodgkin’s lym- types. phoma (NHL) in the Western Hemisphere, compris- ing about one-third of all adult lymphomas [1]. The natural history of this subtype is aggressive, with a Diffuse large B-Cell lymphoma/clinical risk median survival of less than 1 year in untreated stratification patients. Over the past decade, remarkable progress Clinical risk stratification is necessary to define has been made in understanding the biological optimal therapy for patients with ‘‘early stage’’ heterogeneity of DLBCL. There is clear evidence DLBCL. ‘‘Early stage’’ NHL usually refers to disease that, in many cases, the clinical behavior of certain limited to a single side of the diaphragm, including, at DLBCLs can be profiled by the expression of most, stage 1 contiguous extranodal site. It has been molecular markers [2,3]. These markers have not well documented that patients with ‘‘bulky’’ stage 2 only contributed to the development of novel prog- disease (i.e., a mediastinal mass /10 cm or /1/3 of nostic models, allowing clinicians to refine their ability the maximum diameter of the chest) have a prognosis to identify patients at high risk but they have also been indistinguishable from that of patients with advanced- integral in the identification of new therapeutic stage disease; thus, these patients should be treated targets. differently from other patients with early-stage dis- The clinical management of DLBCL has changed ease. dramatically over the past five 5 years. Routine Randomized clinical trials have demonstrated that a incorporation of monoclonal antibody therapy in combined-modality approach incorporating a brief induction treatment regimens has improved OS in duration of chemotherapy followed by involved-field most subgroups of patients with DLBCL. In addition, radiation remains a reasonable standard of care for studies evaluating high-dose chemotherapy and auto- most patients with early stage DLBCL. A SWOG logous stem cell transplantation (SCT) as consolida- study randomized 401 patients with aggressive non- tion treatment during first remission have shown bulky stage 1 or 2 NHL (mainly DLBCL) to 3 cycles promise. Perhaps most exciting is the multitude of of CHOP followed by involved-field radiation (40/50 promising new agents now under development. Gy) or to 8 cycles of CHOP alone [4]. At 5 years, PFS Despite many recent advances, most patients with and OS rates were significantly higher in the com- advanced-stage DLBCL are not cured with conven- bined-modality arm than in the chemotherapy-alone tional therapy. Given this reality, treating physicians arm (77% vs. 64% and 82% vs. 72%, respectively, must recognize the inadequacy of current therapies with less life-threatening toxicity in the combined and urge their eligible patients to participate in well- modality arm (P/0.06).

Correspondence: Julie M. Vose, Section of Hematology/Oncology, University of Nebraska Medical Center, Omaha, Nebraska.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389746 Update on lymphoma management 11

More recently, the Eastern Cooperative Oncology tion therapy. The respective 5-year OS estimates of Group (ECOG) enrolled 210 patients with either OS were 90% and 81%, respectively. Thus, ACBVP diffuse, aggressive stage 1 lymphoma with mediastinal conferred only a modest survival benefit among or retroperitoneal masses greater than 10 cm in patients without bulky disease. diameter (bulky disease), or stage 1E, 2, or 2E The GELA group presented data from an early disease. The 172 patients who had attained CR after analysis of elderly patients with localized aggressive 8 cycles of CHOP were randomized to receive no NHL and an age-adjusted IPI risk of 0 [9]. The report further therapy or involved-field radiation [5]. Dis- suggested that the adding involved-field radiotherapy ease-free survival at 6 years was superior in the to 4 courses of CHOP did not improve CR rates, 5- combined treatment arm (73% vs. 56%; 2-sided P/ year event-free survival, or 5-year OS [23]. Finally, 0.05). However, there was no difference in overall another preliminary report from a SWOG pilot trial survival. Therefore, the benefit of radiation therapy that treated patients with limited stage disease to 3 following a full course of chemotherapy appears to be cycles of R-CHOP, followed by radiation therapy [10] limited to enhanced local control. calculated 2-year progression-free survival and OS at

When any risk factor (age /60 years, high lactate 94% and 95%, respectively, superior to historical dehydrogenase (LDH) level, stage 2 disease, and results with CHOP chemotherapy alone. Of note, performance status ]/2) by the stage-modified this trial required patients to have at least 1 risk factor (‘‘Miller Modification’’) International Prognostic In- in the Miller IPI modification. However, 10-year dex (IPI) is present, outcome is inferior to that of follow-up of SWOG’s original randomized trial [11] patients with no risk factors [6]. For example, in the suggested an increase in late recurrences (/5 years SWOG study, 5- year overall survival was 94%, 71% after completion of therapy) in patients treated with and 50%, respectively, for those with 0 or 1, 2, or 3 combined radiochemotherapy, including a fixed mor- risk factors; and 5-year failure free survival estimates tality rate over the first 10 years, with no evidence of a were 82% for patients with 0 or 1 risk factor, 71% for plateau in the survival curve. Therefore, long-term patients with 2 risk factors, and 48% for patients with follow-up is clearly required before it can be said that 3 risk factors [4] one regimen is superior to another, particularly when These findings have been confirmed by Canadian that regimen is compared with historical controls. researchers who evaluated combined modality therapy Despite the preliminary nature of these follow-up in a similarly defined group of early stage patients [7]. findings, the SWOG investigators currently recom- The overall survival rates at 5 years were 97% for mend 3 cycles of CHOP plus rituximab in addition to patients with no risk factors, 77% for patients with 1/ involved-field radiation for most patients with stage 1 2 risk factors, 58% for patients with 3 risk factors, and and nonbulky stage 2 disease, on the basis of 48% for patients with 4 risk factors, with similar increased survival through the first 9 years and less decrements in PFS reported for increasing numbers associated toxicity. Select elderly patients lacking of risk factors [7]. other risk factors may not require radiation therapy, Similarly, in the ECOG study, the following factors and patients with bulky disease clearly require more were significantly associated with prolonged survival chemotherapy and may benefit from intensified regi- among patients receiving induction CR: age less than mens. By using new approaches such as radioimmu-

60 years (P B/0.001), and fewer than 3 disease sites notherapy and 18F-fluorodeoxyglucose positron

(P/0.01). As in the Canadian trial, these factors were emission tomogram imaging, current clinical trials in also associated with prolonged survival among com- patients with early stage DLBCL and at least 1 Miller plete responders receiving induction therapy [5]. IPI modification risk factor [12] will be useful in The Groupe d’E´ tude des Lymphoˆmes de l’Adulte defining which patients may not require external- (GELA) has reported results from a randomized trial beam radiation. of previously untreated patients younger than 61 years In patients with advanced-stage DLBCL, rituximab with localized, aggressive stage 1 or 2 lymphoma and appears to improve survival when administered in no IPI risk factors. The study compared 3 cycles of combination with standard chemotherapy, but no

CHOP plus involved-field radiotherapy (n/329) or additional benefit is observed with the addition of chemotherapy alone with dose-intensified doxorubi- maintenance rituximab. cin, cyclophosphamide, vindesine, bleomycin, and In a 2002 publication, GELA reported that ritux- prednisone (ACVBP) plus sequential consolidation imab added to standard CHOP conferred a higher OS with high-dose methotrexate, etoposide, ifosfamide, rate for older patients (/60 years) with advanced- and cytosine arabinoside (n/318) [8]. Notably, stage DLBCL [13]. These results truly changed patients with bulky stage 2 disease were not included clinical practice throughout much of the world. Eight in this trial. The 5-year event-free survival estimates cycles of CHOP alone (control arm) or CHOP with were 82% for patients receiving ACVBP chemother- rituximab (treatment arm) produced CR rates of 63% apy alone (followed by intensive consolidation) and and 76%, respectively (P/0.005) and a 2-year OS of

74% for those receiving standard CHOP with radia- 57% and 70% (P/0.007). A recent update of this 12 J. M. Vose et al. trial demonstrated that the survival benefit was but CHOEP-21, and, especially, CHOEP-14 were maintained, and actually continued to improve more toxic than either CHOP regimen. In the parallel through 5 years of follow-up [14]. A subgroup trial (NHL B1) for patients younger than 61 years, analysis of this large study has revealed that 2 groups better complete remission rates were obtained with of patients appear to derive particular benefit from CHOEP than with CHOP (87.6% vs. 79.4%; P/ rituximab: (1) those with an age-adjusted low IPI risk 0.003) as well as improved 5-year event-free survival and (2) those with DLBCL positive for Bcl-2 over- rates (69.2% vs. 57.6%; P/0.004, primary end expression, historically a poor prognostic factor. This point) [19] The benefit of interval reduction was finding suggests that one of the ways in which less clear in the younger than in the older patients. rituximab works is to overcome Bcl-2-associated Although the CHOEP were more myelosuppresive, chemotherapy resistance [15]. they were reasonably well tolerated. Only 3 therapy- Preliminary results from the MabThera Interna- associated deaths occurred, 1 (0.5%) among the tional Trial (MInT), currently evaluating CHOP-like CHOEP-21 and 2 (1.1%) among the CHOEP-14 chemotherapy regimens plus rituximab in patients participants. younger than 60 years, were recently presented [16]. The magnitude of benefit seen with these dose- As in the GELA trial, patients who received rituximab intense regimens is similar to that observed with the plus chemotherapy had a significantly longer 2-year addition of relatively nontoxic rituximab therapy time to treatment failure (81% vs. 58%) than patients reported in other trials. Indeed, a recent retrospective receiving chemotherapy alone. In addition, the 2-year analysis of patients included in the MInT trial OS rates also significantly favored chemotherapy plus suggests that survival differences between different rituximab (95% vs. 85%). CHOP-like regimens, including CHOEP, disappear Similar to the GELA trial, the as-yet unpublished when rituximab is added to standard therapy [20]. larger (N/632) US Intergroup Study randomized a Since trials incorporating monoclonal antibody ther- population of 632 elderly patients to 8 cycles of apy into these dose-intensified regimens are ongoing, CHOP or CHOP plus rituximab given every other the routine use of dose-intense regimens outside of a cycle [17]. Responding patients were then rando- clinical trial is not currently recommended. mized to receive either rituximab ‘‘maintenance’’ (4 Ongoing prospective trials are underway to define doses, every 6 months for 2 years or no maintenance the role of autologous stem cell transplantation therapy. A weighted analysis was used to mathemati- (ASCT) consolidation for patients with high-risk, cally model the groups that had been treated with advanced-stage DLBCL in first remission. CHOP alone or CHOP plus rituximab as induction Several phase 3 trials have evaluated ASCT in therapy, controlling for maintenance exposure. The newly diagnosed patients with DLBCL, either as magnitude of the OS benefit of induction therapy with consolidation therapy after CR or as induction CHOP plus rituximab was similar to that seen in the therapy. In most of these trials, however, high-risk GELA trial, which essentially confirmed the GELA disease was identified by criteria other than the IPI, results. Perhaps the most important contribution of and a variety of schedules incorporating ASCT have the United States Intergroup Study, however, was that been used [21]. it demonstrated a lack of benefit with ‘‘maintenance’’ The Groupe Ouest-Est des Leuce´mies et des Autres rituximab when rituximab was included in the initial Maladies du Sang (GOELAMS) trial randomized 197 chemotherapy regimen. consecutive patients to receive either 8 courses of The role of dose-intense regimens which are standard CHOP chemotherapy, or a complicated rituximab-based therapies is unclear. Recently pub- regimen of ASCT plus chemotherapy, starting with lished results of 2 large German trials (NHL-B1 and cyclophosphamide, vindesine, epirubicin, and predni- NHL-B2) suggest that modifications to the CHOP sone (CEEP), followed by high-dose methotrexate regimen may improve survival. The major limitation and cytarabine, then treated with carmustine, etopo- of these trials is that they did not include rituximab. side, cytarabine, and melphalan (BEAM) for stem cell These 2 trials randomized patients to 6 cycles of conditioning prior to ASCT [22]. Overall, 78% of the CHOP-21 (every 3 weeks) or CHOP-14 (every 2 patients completed the assigned treatment. With a weeks) vs. CHOEP-21 (CHOP plus etoposide 100 median follow-up of 4 years, the estimated event-free 2 mg/m Days 1/3 every 3 weeks) or CHOEP-14 5-year survival rate was significantly higher for pa- 2 (CHOP plus etoposide 100 mg/m Days 1/3 every tients who received ASCT than for those who 2 weeks). Patients in these trials also received radio- received standard CHOP (55% vs. 37%). A retro- therapy (36 Gy) to both extranodal and bulky disease spectively performed subgroup analysis demonstrated sites. One trial (NHL-B2) was limited to patients a survival benefit in patients with age-adjusted high- older than 60 years [18]. Five-year event-free and OS intermediate IPI risk (OS, 74% vs. 44%). rates were respectively 32.5% and 40.6%, for CHOP- These data are reminiscent of the LNH87/2 trial 21 and 43.8% and 53.3% for CHOP-14. Toxicity was results previously published by GELA [23]. In this similar among CHOP-14 and CHOP-21 participants, trial, 1043 patients were initially randomized to Update on lymphoma management 13 treatment with 4 courses of an anthracycline-based Despite the many recent advances summarized in regimen. Patients who achieved CR were randomized this manuscript, most patients with advanced stage to receive additional cycles of sequential chemother- DLBCL are not cured with conventional therapy, apy or ASCT. As in the GOELAMS trial, a retro- Hence, each treating physician must recognize the spective assessment of 451 high-intermediate or high inadequacy of current therapy and urge all eligible IPI risk patients showed that the 8-year OS rate was patients to participate in well-designed clinical trials. higher in the ASCT arm than in the sequential Several of the investigators conducting ongoing clin- chemotherapy arm (64% vs. 49%). Of course, these ical trials have emphasized that providing optimal retrospective subgroups analyses must be interpreted therapy often involves experimenting with both new cautiously because these higher risk patients were not and old agents in novel ways. Further development of initially identified as the target population for these the aforementioned novel therapies should result in trials. Most of the current mature phase 3 trials have improved outcomes for patients suffering from these reported improved disease-free survival (DFS) but common subtypes of NHL. not improved OS with ASCT therapy in patients younger than 60 years with high or high-intermediate IPI risk scores [21]. Moreover, none of these trials References included rituximab therapy, so it is not known [1] Armitage JO, Weisenburger DD. New approach to classifying whether the benefit of is abrogated by the addition non-Hodgkin’s lymphomas: clinical features of the major of rituximab to induction or consolidation therapies. histologic subtypes. Non-Hodgkin’s Lymphoma Classification This is a particularly important question in light of the Project. J Clin Oncol 1998;16:/ 2780/ /2795. [2] Shipp MA, Ross KN, Tamayo P, Weng AP, Kutok JL, Aguiar MInT trial analysis, suggesting that rituximab may RC, Gaasenbeek M, Angelo M, Reich M, Pinkus GS, et al. abrogate the benefit of intensified regimens. Diffuse large B-cell lymphoma outcome prediction by gene- Radioimmunotherapy with iodine-131 (I-131) to- expression profiling and supervised machine learning. Nat situmomab or ibritumomab tiuxetan is quite active in Med 2002;8:/ 68/ /74. the treatment of indolent B-cell lymphoma and is [3] Rosenwald A, Wright G, Chan WC, Connors JM, Campo E, Fischer RI, Gascoyne RD, Muller-Hermelink HK, Smeland worthy of further investigation in other lymphoma EB, Giltnane JM, et al. The use of molecular profiling to subtypes. Zelenetz and colleagues analyzed 71 pa- predict survival after chemotherapy for diffuse large-B-cell tients whose indolent lymphomas underwent Richter’s lymphoma. N Engl J Med 2002;346:/ 1937/ /1947. transformation to more aggressive histologic forms [4] Miller TP, Dahlberg S, Cassady JR, Adelstein DJ, Spier CM, who were treated with I-131 tositumomab in 5 clinical Grogan TM, LeBlanc M, Carlin S, Chase E, Fisher RI. Chemotherapy alone compared with chemotherapy plus radio- trials [24]. The overall response rate for a single therapy for localized intermediate- and high-grade non-Hodg- treatment with I-131 tositumomab was 39%, with a kin’s lymphoma. N Engl J Med 1998;339:/ 21/ /26. median response duration of 20 months. In 24% of [5] Horning SJ, Weller E, Kim K, Earle JD, O’Connell MJ, these patients, response duration was longer than 1 Habermann TM, Glick JH. Chemotherapy with or without year. Given the relatively low toxicity profile of the I- radiotherapy in limited-stage diffuse aggressive non-Hodgkin’s lymphoma: Eastern Cooperative Oncology Group study 1484. 131 tositumomab regimen compared with that of J Clin Oncol 2004;22:/ 3032/ /3038. ASCT, [25] the radioimmunotherapy approach holds [6] Miller TP. The limits of limited stage lymphoma. J Clin Oncol significant promise for patients with transformed 2004;22:/ 2982/ /2984. disease. [7] Shenkier TN, Voss N, Fairey R, Gascoyne RD, Hoskins P, Morschhauser and colleagues have recently com- Klasa R, Klimo P, O’Reilly SE, Sutcliffe S, Connors JM. Brief chemotherapy and involved-region irradiation for limited- pleted a prospective, multicenter phase 2 trial to stage diffuse large-cell lymphoma: an 18-year experience evaluate the efficacy and safety of yttrium-90 ibritu- from the British Columbia Cancer Agency. momab tiuxetan in elderly patients with histologically [8] Reyes F, Lepage E, Ganem G, Molina TJ, Brice P, Coiffier B, confirmed primary refractory or relapsed DLBCL for Morel P, Ferme C, Bosley A, Lederlin P, et al. ACVBP versus whom ASCT is contraindicated [26]. An overall CHOP plus radiotherapy for localized aggressive lymphoma. N Engl J Med 2005;352:/ 1197/ /1205. response rate of 44% was observed for the entire [9] Fillet G, Bonnet C, Mournier N, Thieblemont C, FermeĆ, study population. The median response duration was Quesnez B, Martin C, Blanc M, Conroy Th, Penny AM, et al. 22 months in those patients who had never received Radiotherapy is unnecessary in elderly patients with localized rituximab. By contrast, only 19% of patients treated aggressive non Hodgkin’s lymphoma: results of the GELA with prior chemoimmunotherapy responded to radio- LNH 93 /4 study. Program and abstracts of the 44th Annual Meeting of the American Society of Hematology; December immunotherapy. 6/10, 2002; Philadelphia, Pennsylvania. Abstract 337. Blood These results are encouraging. Both I-131 tositu- 2002;100:/ 92a./ momab and yttrium-90 ibritumomab tiuxetan are [10] Miller TP, Unger JM, Spier C, Stea B, Cantu E, Le Blanc M, currently being evaluated for the treatment of Fisher RI. Effect of adding rituximab to three cycles of CHOP DLBCL in multicenter trials. For example, SWOG plus involved-field radiotherapy for limited-stage aggressive diffuse B-cell lymphoma (SWOG-0014). Program and ab- is currently conducting a trial with I-131 tositumo- stracts of the 46th Annual Meeting of the American Society of mab as consolidation therapy following standard R- Hematology; December 4/7, 2004; San Diego, California. CHOP for patients over 60 with DLBCL. Abstract 158. 14 J. M. Vose et al.

[11] Miller TP, Leblanc M, Spier C, et al. CHOP alone compared Hasenclever D, et al. Two-weekly or 3-weekly CHOP che- to CHOP plus radiotherapy for early stage aggressive non- motherapy with or without etoposide for the treatment of Hodgkin’s lymphomas: update of the Southwest Oncology young patients with good-prognosis (normal LDH) aggressive Group (SWOG) randomized trial. Program and abstracts of lymphomas: results of the NHL-B1 trial of the DSHNHL. the 43rd Annual Meeting of the American Society of Blood 2004;104:/ 626/ /633. Hematology; December 7 /11, 2001; Orlando, Florida. Ab- [20] Pfreundschuh MG, Ho A, Wolf M, Cavallin-Stahl E, Petten- stract 3024. Blood 2001;98:/ 724/ /725a. gell R, Vasova I, Belch A, Walewski J, Zinzani P, Mingrone W, [12] Friedberg JW, Chengazi V. PET scans in the staging of et al. Treatment results on CHOP-21, CHOEP-21, MACOP- lymphoma: current status. Oncologist 2003;8:/ 438/ /447. B and PMitCEOB with and without rituximab in young good- [13] Coifﬁer B, Lepage E, Briere J, Herbrecht R, Tilly H, prognosis patients with aggressive lymphomas: rituximab as Bouabdallah R, Morel P, Van Den Neste E, Salles G, Gaulard ‘‘equalizer’’ in the MinT (TABTHERA International Trial P, et al. CHOP chemotherapy plus rituximab compared with Group) study. Program and abstracts of the 41st Annual CHOP alone in elderly patients with diffuse large-B-cell Meeting of the American Society of Clinical Oncology; May lymphoma. N Engl J Med 2002;346:/ 235/ /242. 13/17, 2005; Orlando, Florida. Abstract 6529. J Clin Oncol [14] Coifﬁer B, Feugier P, Sebban C, Bouabdallah R, Delwail V,

2005;23(suppl):/ 567S./ Tilly H, Gisselbrecht C, Bosly A, Salles G, Reyes F. Long term [21] Fisher RI. Autologous stem-cell transplantation as a compo- results of the GELA Study, R-CHOP vs. CHOP in elderly patients with diffuse large B-Cell lymphoma. Program and nent of initial treatment for poor-risk patients with aggressive abstracts of the 46th Annual Meeting of the American Society non-Hodgkin’s lymphoma: resolved issues versus remaining opportunity. J Clin Oncol 2002;20:/ 4411/ /4412. of Hematology; December 4/7, 2004; San Diego, California. Abstract 1383. Blood 2004;104. [22] Milpied N, Deconinck E, Gaillard F, Delwail V, Foussard C, [15] Mounier N, Briere J, Gisselbrecht C, Emile JF, Lederlin P, Berthou C, Gressin R, Lucas V, Colombat P, Harousseau JL, Sebban C, Berger F, Bosly A, Morel P, Tilly H, et al. et al. Initial treatment of aggressive lymphoma with high-dose Rituximab plus CHOP (R-CHOP) overcomes bcl-2-asso- chemotherapy and autologous stem-cell support. N Engl J ciated resistance to chemotherapy in elderly patients with Med 2004;350:/ 1287/ /1295.

diffuse large B-cell lymphoma (DLBCL). Blood 2003;101:/ / [23] Haioun C, Lepage E, Gisselbrecht C, Salles G, Coifﬁer B,

4279/4284. Brice P, Bosly A, Morel C, Tilly H, Lederlin P, et al. Survival [16] Pfreundschuh M, Truemper L, Gill D, Osterborg A, Petten- benefit of high-dose therapy in poor-risk aggressive non- gell R, Trneny M, Imrie K, Walewski J, Zinzani Pl, Loeffler M. Hodgkin’s lymphoma: final analysis of the prospective First analysis of the completed MabThera International LNH87-2 protocol: a groupe d’E´ tude des lymphomes de (MInT) Trial in young patients with low-risk diffuse large B- l’Adulte study. J Clin Oncol 2000;18:/ 3025/ /3030. cell lymphoma (DLBCL): addition of rituximab to a CHOP- [24] Zelenetz AD, Saleh M, Vose J, Younes A, Kaminski MS. like regimen significantly improves outcome of all patients Patients with transformed low-grade lymphoma attain durable with the identification of a very favorable subgroup with IPI / responses following outpatient radioimmunotherapy with 0 and no bulky disease. Program and abstracts of the 46th tositumomab and iodine I 131 tositumomab (Bexxar). Pro- Annual Meeting of the American Society of Hematology; gram and abstracts of the 44th Annual Meeting of the December 4/7, 2004; San Diego, California. Abstract 157. American Society of Hematology; December 6/10, 2002; [17] Habermann TM, Weller EA, Morrison VA, Cassileth PA,

Philadelphia, Pennsylvania. Abstract 1384. Blood 2002;100:/ / Cohn JB, Dakhil SR, Gascoyne RD, Woda B, Fisher RI, 357a. Peterson BA, et al. Phase III trial of rituximab-CHOP vs. [25] Friedberg JW, Neuberg D, Gribben JG, Mauch P, Anderson CHOP with a second randomization to maintenance ritux- KC, Soiffer RJ, Takvorian T, Fisher DC, Schlossman R, Jallow imab or observation in patients 60 years of age and older with H, et al. Autologous bone marrow transplantation after diffuse large B cell lymphoma. Program and abstracts of the histologic transformation of indolent B cell malignancies. 45th Annual Meeting of the American Society of Hematology; Biol Blood Marrow Transplant 1999;5:/ 262/ /268. December 6/9, 2003; San Diego, California. Abstract 8. [26] Morschhauser F, Huglo D, Martinelli G, Paganelli G, Zinzani

Blood 2003;102:/ 6a./ [18] Pfreundschuh M, Trumper L, Kloess M, Schmits R, Feller PL, Hadjiyiannakis D, Liberati A, Illidge T, Milpied N, Stein AC, Rube C, Rudolph C, Reiser M, Hossfeld DK, Eimerma- H, et al. Yttrium-90 ibritumomab tiuxetan (Zevalin) for cher H, et al. Two-weekly or 3-weekly CHOP chemotherapy patients with relapsed/refractory diffuse large B-cell lymphoma with or without etoposide for the treatment of elderly patients not appropriate for autologous stem cell transplantation: with aggressive lymphomas: results of the NHL-B2 trial of the results of an open-label phase II trial. Program and abstracts DSHNHL. Blood 2004;104:/ 634/ /641. of the 46th Annual Meeting of the American Society of [19] Pfreundschuh M, Trumper L, Kloess M, Schmits R, Feller Hematology; December 4/7, 2004; San Diego, California.

AC, Rudolph C, Reiser M, Hossfeld DK, Metzner B, Abstract 130. Blood 2004;104:/ 41a./ Hematology, 2005; 10 Supplement 1: 15/18

CHRONIC MYELOID LEUKEMIA

Allogeneic transplantation for chronic myelogenous leukemia

PHILIP MCGLAVE

Division of Hematology, Oncology and Transplantation, University of Minnesota Medical School, Minneapolis, MN, USA

Keywords: CML, transplantation

Related donor transplantation predictor of poor outcome, while DNA-based methodology has identified certain single and multiple Over the last quarter century investigators have allele mismatches with adverse effects on outcome in demonstrated that allogeneic hematopoietic cell trans- URD transplant for CML [11]. plantation (HCT) can cure chronic myelogenous leukemia (CML). Early reports suggested the efficacy of related donor transplant after a myeloablative Umbilical cord blood preparative regimen containing total body irradiation Umbilical cord blood (UCB) has been identified as a (TBI) [1/3]. Subsequent studies have identified source of hematopoietic stem cells for transplantation variables which improve outcome such as transplant in early chronic phase, younger recipient age, donor/ [12]. Furthermore, UCB may be particularly useful in recipient compatibility at the major HLA loci and the unrelated donor transplant setting since HLA- male donor gender [4,5]. Myeloablative regimens typed, frozen and stored cells are usually available which do not contain TBI have also proven effective within 1 /2 days through international registries, can in transplant for CML [6]. The use of peripheral be used to repopulate hematopoiesis in adults as well blood progenitor cells as a source of stem cells for as children, and may (arguably) tolerate a greater transplant mobilized with G-CSF is comparable in degree of HLA mismatch with the recipient than most respects to non-mobilized related donor mar- hematopoietic stem cells obtained from adult volun- row, although long-term studies may uncover differ- teer URD [13,14]. The role of UCB transplant in ences in the incidence of chronic GVHD and in the CML is promising, but not yet fully explored. risk of relapse [7]. Donor Leukocyte Infusions Adult unrelated donors Although a clinically relevant ‘‘graft-versus-leukemia’’ Alternative sources of stem cells for CML patients (GVL) effect was first detected over 20 years ago [15], without a suitably HLA-matched related donor have the importance of this effect in transplant therapy for been developed. Adult, volunteer unrelated donors CML has recently been underscored. Early attempts (URD) can be obtained through the National Marrow at ex-vivo T-cell depletion after related donor HCT Donor Program (NMDP) and other registries for for CML resulted in a reduced incidence of acute many, but not all, otherwise eligible patients. Factors GVHD, but an unexpected, extraordinarily high predicting success of transplant are comparable in the incidence of relapse. Analyses of a large International unrelated and related donor setting [8/10]. Large, Bone Marrow Transplant Registry (IBMTR) data set retrospective analyses have identified certain donor- demonstrated a correlation between development of recipient disparities such as mismatch at HLA-C as a GVHD and protection from CML relapse [16].

Correspondence: Philip McGlave, Division of Hematology, Oncology and Transplantation, University of Minnesota Medical School, Minneapolis, MN, USA.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389764 16 P. McGlave

Subsequently, Kolb et al. have demonstrated the transplant compared to the study patients’ immediate efficacy of donor leukocyte infusions (DLI) to sup- pretransplant baseline [31]. press persistent or recurrent malignant clones after allogeneic HCT [17/19]. Predictably, such infusions The imatinib era may provoke or exacerbate GVHD. The development of the selective Bcr-Abl tyrosine kinase inhibitor imatinib (STI-571, Gleevec) has Nonmyeloablative preparative regimens fundamentally changed therapy of CML. Imatinib, These clinical observations, underpinned by consid- given orally on a daily basis as first-line therapy in erable preclinical data [20], prompted investigators to newly diagnosed chronic phase CML patients, results develop less intensive and often non-myeloablative in hematologic, cytogenetic and molecular remissions (NMA) preparative regimens for allogeneic HCT in the majority of cases [32]. Relapses occur and can therapy of CML. These NMA regimens are intended often be attributed to mutations in the BCR-ABL to minimize toxicity while exploiting the GVL effect gene [33]. It is possible that such mutations can be treated successfully with a second generation of [21/25]. Current results suggest that HCT with a NMA preparative regimen is a feasible treatment tyrosine kinase inhibitors possessing higher binding option for individuals not eligible for more standard affinities for the ABL kinases [34 /37]. These exciting preparative regimens. The incidence of nonrelapse developments call into question the historical first-line mortality may be lower; however non-engraftment, role of allogeneic HCT therapy for CML. GVHD, infection and disease persistence or recur- * rence still complicate / transplants [26]. Transplant in the imatinib era Recently, investigators have demonstrated that im- Current clinical results atinib can be used effectively to treat CML relapse Analysis of long-term results using a large data set after allogeneic HCT [38]. By 6 months after imatinib from the Center for International Blood and Marrow therapy for post-transplant relapse, 9 of 10 patients Transplant Research (CIBMTR) reveals an approxi- achieved cytogenetic remission, and the BCR-ABL transcript could not be identified in 7 of these mately 60% incidence of overall survival at 10 years patients. Of note, patients achieving cytogenetic for over 3300 first chronic phase CML patients remission also converted to complete donor chimer- receiving related donor HCT between 1978 and ism. These early results raise the issue of substituting 1997 and 50% for over 1300 similar patients receiving imatinib therapy for DLI in the post-transplant transplants from alternative donors. The incidence of relapse setting. relapse at 10 years for these two groups is approxi- In small, uncontrolled trials, imatinib has also been mately 20%. Of note, the overall survival curve does administered as prophylaxis during the first 100 days not plateau. Important late-occurring causes of death following transplant for poor prognosis CML [39]. include GVHD, infection and relapse [27]. Early results suggest that this approach is feasible, but requires reduced doses of imatinib and tacrolimus and Late effects and quality of life resulted in reversible hematologic suppression. The investigators suggest that an ‘‘adjuvant’’ strategy Late effects and quality of life after allogeneic HCT incorporating imatinib in the early post transplant therapy for CML are important issues in a field where regimen may reduce risk of relapse in high-risk effective alternative therapy is developing rapidly [28]. individuals. A retrospective analysis of late effects in 248 CML Imatinib has also been used to prepare patients for transplant recipients who had survived at least 2 years subsequent NMA HCT therapy of CML. In a was very informative [29]. Compared to siblings, recently reported study, CML patients receiving pre- survivors had a high prevalence of long-term health- transplant imatinib had equivalent time to engraft- related complications including endocrine, ocular, ment and transplant related mortality compared to oral health, gastrointestinal, musculoskeletal, neuro- equivalent patients not pre-treated with imatinib [40]. sensory and neuromotor impairment. In a non-over- Of note, the incidence of molecular remission in the lapping study of 46 CML transplant recipients, imatinib pre-treatment group (83%) was significantly investigators observed a high incidence of late cogni- higher than that in the group not pre-treated with tive deficits and an increase in psychosexual problems imatinib (40%) (P/0.11). Such approaches might be compared to the general population [30]. Of interest, useful to reduce the leukemia load prior to NMA in a third study CML patients and physicians reported HCT. an improved Quality of Life (QOL) Index score, Current results suggest that the majority of chronic decreased signs and symptoms of depression and phase patients receiving imatinib as first-line therapy less alcohol consumption at 12 months following will achieve a complete cytogenetic remission without Allogeneic transplantation for chronic myelogenous leukemia 17 undue toxicity. On the other hand, a recent study [11] Petersdorf EW, Kollman C, Hurley CK, Dupont B, Nadema- suggests that patients achieving a complete cytoge- nee A, Begovich AB, Weisdorf D, McGlave P. Effect of HLA class II gene disparity on clinical outcome in unrelated donor netic remission with imatinib have a lower incidence hematopoietic cell transplantation for CML: the U.S. national of molecular remission, less durable molecular remis- marrow donor program experience. Blood 2001;98:/ 2922/ / sions and a higher level of residual disease in 2929. molecular remission (determined by replicate RT- [12] Gluckman E, Rocha V, Boyer-Chammard A, Locatelli F, PCR testing) than comparable patients treated with Arcese W, Pasquini R, Ortega J, Souillet G, Ferreira E, allogeneic HCT [41]. Furthermore, some CML Laporte JP, et al. Outcome of cord blood transplantation

from related and unrelated donors. N Engl J Med 1997;337:/ / patients will not respond to imatinib or will develop 373/331. imatinib resistance, and the role of newer agents with [13] Laughlin MJ, Eapen M, Rubinstein P, Wagner JE, Zhang MJ, increased binding activity to the ABL-kinase domain Champlin RE, Stevens C, Barker JN, Gale RP, Lazarus HM, in the treatment of imatinib-resistant CML is not yet et al. Outcomes after transplantation of cord blood or bone fully understood. Early transplant in patients unlikely marrow from unrelated donors in adults with leukemia. N to have a durable response to imatinib may be Engl J Med 2004;351:/ 2265/ /2275. [14] Barker JN, Weisdorf DJ, DeFor TE, Blazar BR, McGlave PB, indicated, and methods to predict those who will Miller JS, Verfaillie CM, Wagner JE. Transplantation of two benefit from early transplant are needed. partially HLA matched umbilical cord blood units to enhance engraftment in adult with hematologic malignancy. Blood 2005;105:/ 1343/ /1347. [15] Weiden PL, Sullivan KM, Flournoy N, Storb R, Thomas ED. Antileukemic effect of chronic graft-versus-host disease: con- References tribution to improved survival after allogeneic marrow trans- [1] Doney K, Buckner CD, Sale GE, Ramberg R, Boyd C, plantation. N Engl J Med 1981;304:/ 1529/ /1533. Thomas ED. Treatment of chronic granulocytic leukemia by [16] Goldman JM, Gale RP, Horowitz MM, Biggs JC, Champlin chemotherapy, total body irradiation and allogeneic bone RE, Gluckman E, Hoffman RG, Jacobsen SJ, Marmont AM, marrow transplantation. Exp Hematol 1978;6:/ 738/ /747. McGlave PB, et al. Bone marrow transplantation for chronic [2] McGlave PB, Miller WJ, Hurd DD, Arthur D, Kim TH. myelogenous leukemia in chronic phase. Increased risk for Cytogenetic conversion following allogeneic bone marrow relapse associated with T-cell depletion. Ann Intern Med transplantation for advanced chronic myelogenous leukemia. 1988;108:/ 806/ /814. Blood 1981;58:/ 1050/ /1052. [17] Kolb HJ, Mittermuller J, Clemm C, Holler E, Ledderose G, [3] Champlin R, Ho W, Arenson E, Gale RP. Allogeneic bone Brehm G, Heim M, Wilmanns W. Donor leukocyte transfu- marrow transplantation for chronic myelogenous leukemia in sions for treatment of recurrent chronic myelogenous leuke- chronic or accelerated phase. Blood 1982;60:/ 1038/ /1041. mia in marrow transplant patients. Blood 1990;76:/ 2462/ / [4] Gratwohl A. Prognostic factors in chronic myeloid leukemia: 2465. allografting. Semin Hematol 2003;40:/ 13/ /21. [18] Giralt S, Hester J, Huh Y, Hirsch-Ginsberg C, Rondon G, [5] De Souza CA, Vigorito AC, Ruiz MA, Nucci M, Dulley FL, Seong D, Lee M, Gajewski J, Van Besien K, Khouri I, et al. Funcke V, Tabak D, Azevedo AM, Byington R, Macedo MC, CD8-depleted donor lymphocyte infusion as treatment for et al. Validation of the EBMT risk score in chronic myeloid relapsed chronic myelogenous leukemia after allogeneic bone leukemia in Brazil and allogeneic transplant outcome. Hae- marrow transplantation. Blood 1995;86:/ 4337/ /4343. matologica 2005;90:/ 232/ /237. [19] Slavin S, Naparstek E, Nagler A, Ackerstein A, Kapelushnik J, [6] Socie G, Clift RA, Blaise D, Devergie A, Ringden O, Martin Or R. Allogeneic cell therapy for relapsed leukemia after bone PJ, Remberger M, Deeg HJ, Ruutu T, Michallet M, et al. marrow transplantation with donor peripheral blood lympho- Busulfan plus cyclophosphamide compared with total-body cytes. Exp Hematol 1995;23:/ 1553/ /1562. irradiation plus cyclophosphamide before marrow transplanta- [20] Georges GE, Storb R, Thompson JD, Yu C, Gooley T, Bruno tion for myeloid leukemia: long-term follow-up of 4 rando- B, Nash RA. Adoptive immunotherapy in canine mixed mized studies. Blood 2001;98:/ 3569/ /3574. chimeras after non-myeloablative hematopoietic cell trans- [7] Oehler VG, Radich JP, Storer B, Blume KG, Chauncey T, plantation. Blood 2000;95:/ 3262/ /3269. Clift R, Snyder DS, Forman SJ, Flowers ME, Martin P, et al. [21] Giralt S, Estey E, Albitar M, van Besien K, Rondon G, Randomized trial of allogeneic related bone marrow trans- Anderlini P, O’Brien S, Khouri I, Gajewski J, Mehra R, et al. plantation versus peripheral blood stem cell transplantation for chronic myeloid leukemia. Biol Blood Marrow Transplant Engraftment of allogeneic hematopoietic progenitor cells with purine analog-containing chemotherapy: harnessing graft- 2005;11:/ 85/ /92.

[8] McGlave PB, Xiao OS, Wen W, Anasetti C, Nademanee A, versus-leukemia without myeloablative therapy. Blood 1997;/ Champlin R, Antin JH, Kernan NA, King R, Weisdorf DJ. 89:4531/ /4536. Unrelated donor marrow transplantation for chronic myelo- [22] Bornhauser M, Kiehl M, Siegert W, Schetelig J, Hertenstein genous leukemia: 9 years’ experience of the National Marrow B, Martin H, Schwerdtfeger R, Sayer HG, Runde V, Kroger N, et al. Dose-reduced conditioning for allografting in 44 Donor Program. Blood 2000;95:/ 2219/ /2225. [9] Barker JN, Davies SM, DeFor TE, Burns LJ, McGlave PB, patients with chronic myeloid leukaemia: a retrospective Miller JS, Weisdorf DJ. Determinants of survival after human analysis. Br J Haematol 2001;115:/ 119/ /124. leucocyte antigen-matched unrelated donor bone marrow [23] Or R, Shapira MY, Resnick I, Amar A, Ackerstein A, Samuel transplantation in adults. British Journal of Haematology S, Aker M, Naparstek E, Nagler A, Slavin S. Nonmyelo- 2002;118:/ 101/ /107. ablative allogeneic stem cell transplantation for the treatment [10] Weisdorf DJ, Anasetti C, Antin JH, Kernan NA, Kollman C, of chronic myeloid leukemia in ﬁrst chronic phase. Blood Snyder D, Petersdorf E, Nelson G, McGlave P. Allogeneic 2003;101:/ 441/ /445. bone marrow transplantation for chronic myelogenous leuke- [24] Hallemeier C. Re: High rate of graft failure in 25 patients with mia: Comparative analysis of unrelated versus matched sibling chronic myelogenous leukemia conditioned with a reduced- donor transplants. Blood 2002;99:/ 1971/ /1977. intensity regimen of 550 cGy total body irradiation and 18 P. McGlave

cyclophosphamide for unrelated donor transplantation. Biol inhibitor imatinib (STI571) in chronic phase and blast crisis / / Blood Marrow Transplant 2004;10:726 /727. chronic myeloid leukemia. Cancer Cell 2002;2:/ 117/ /125. [25] Kerbauy FR, Storb R, Hegenbart U, Gooley T, Shizuru J, Al- [34] Deininger MW, Druker BJ. SRCircumventing imatinib resis- Ali HK, Radich JP, Maloney DG, Agura E, Bruno B, et al. tance. Cancer Cell 2004;6:/ 108/ /110. Hematopoietic cell transplantation from HLA-identical sibling [35] Shah NP, Tran C, Lee FY, Chen P, Norris D, Sawyers CL. donors after low-dose radiation-based conditioning for treat- Overriding imatinib resistance with a novel ABL kinase / / / ment of CML. Leukemia 2005;19:990 997. inhibitor. Science 2004;305:/ 399/ /401. [26] Qazilbash MH, Giralt SA, Champlin RE. Nonmyeloablative [36] O’Hare T, Walters DK, Deininger MW, Druker BJ. AMN107: stem cell transplantation for chronic myeloid leukemia. tightening the grip of imatinib. Cancer Cell 2005;7:/ 117/ /119. Hematol Oncol Clin North Am 2004;3:/ 703/ /713, xi. [37] Weisberg E, Manley PW, Breitenstein W, Bruggen J, Cowan- [27] Horowitz MM. Personal Communication. Jacob SW, Ray A, Huntly B, Fabbro D, Fendrich G, Hall- [28] Kiss TL, Abdolell M, Jamal N, Minden MD, Lipton JH, Meyers E, et al. Characterization of AMN107, a selective Messner HA. Long-term medical outcomes and quality-of-life inhibitor of native and mutant Bcr-Abl. Cancer Cell 2005;7:/ / assessment of patients with chronic myeloid leukemia followed 129/141. at least 10 years after allogeneic bone marrow transplantation. [38] DeAngelo DJ, Hochberg EP, Alyea EP, Longtine J, Lee S, J Clin Oncol 2002;20:/ 2334/ /2343. Galinsky I, Parekkedon B, Ritz J, Antin JH, Stone RM, et al. [29] Baker KS, Gurney JG, Ness KK, Bhatia R, Forman SJ, Extended follow-up of patients treated with imatinib mesylate Francisco L, McGlave PB, Robison LL, Snyder DS, Weisdorf (gleevec) for chronic myelogenous leukemia relapse after DJ, et al. Late effects in survivors of chronic myeloid leukemia allogeneic transplantation: durable cytogenetic remission and treated with hematopoietic cell transplantation: results from conversion to complete donor chimerism without graft-versus-

the Bone Marrow Transplant Survivor Study. Blood 2004;/ host disease. Clin Cancer Res 2004;10:/ 5065/ /5071. 104:1898/ /1906. [39] Anderlini P, Sheth S, Hicks K, Ippoliti C, Giralt S, Champlin [30] Hayden PJ, Keogh F, Ni Conghaile M, Carroll M, Crowley RE. Re: Imatinib mesylate administration in the ﬁrst 100 days M, Fitzsimon N, Gardiner N, Vandenberghe E, O’Riordan J, McCann SR. A single-centre assessment of long-term quality- after stem cell transplantation. Biol Blood Marrow Transplant of-life status after sibling allogeneic stem cell transplantation 2004;10:/ 883/ /884. for chronic myeloid leukaemia in ﬁrst chronic phase. Bone [40] Kim DW, Chung YJ, Lee S, Kim YJ, Chung NG, Kim JA, Oh IH, Kim TG, Kim YL, Goh HG, et al. Pretransplant imatinib Marrow Transplant 2004;34:/ 545/ /556. [31] Chang G, Orav EJ, McNamara TK, Tong MY, Antin JH. can improve the outcome of nonmyeloablative stem cell Psychosocial function after hematopoietic stem cell transplan- transplantation without increasing the morbidity in Philadel- phia chromosome-positive chronic myeloid leukemia. Leuke- tation. Psychosomatics 2005;46:/ 34/ /40. [32] Deininger M, Buchdunger E, Druker BJ. The development of mia 2004;18:/ 1907/ /1909. imatinib as a therapeutic agent for chronic myeloid leukemia. [41] Lange T, Bumm T, Mueller M, Otto S, Al-Ali HK, Grom- Blood 2005;105:/ 2640/ /2453. misch L, Musiol S, Franke C, Krahl R, Niederwieser D, et al. [33] Shah NP, Nicoll JM, Nagar B, Gorre ME, Paquette RL, Durability of molecular remission in chronic myeloid leukemia Kuriyan J, Sawyers CL. Multiple BCR-ABL kinase domain patients treated with imatinib vs allogeneic stem cell trans- mutations confer polyclonal resistance to the tyrosine kinase plantation. Leukemia 2005;19:/ 1262/ /1265. Hematology, 2005; 10 Supplement 1: 19/24

MULTIPLE MYELOMA

Myeloma bone disease

ORHAN SEZER

Department of Hematology and Oncology, University Hospital Charite´, Berlin, Germany

Abstract Bone destruction is a hallmark of multiple myeloma, and recent studies demonstrated a strong interdependence between tumor progression and bone resorption. Increased bone resorption as a major characteristic of multiple myeloma is caused by osteoclast activation and osteoblast inhibition (uncoupling). Myeloma cells alter the local regulation of bone metabolism by increasing the receptor activator of NF-kB ligand (RANKL) and decreasing osteoprotegerin (OPG) expression within the bone marrow microenvironment, thereby stimulating the central pathway for osteoclast formation and activation. In addition, they produce the chemokines MIP-1a, MIP-1b and SDF-1a, which also increase osteoclast activity. Furthermore, myeloma cells suppress osteoblast function by the secretion of osteoblast inhibiting factors, e.g. Dickkopf (DKK)-1. The resulting bone destruction releases several cytokines, which in turn promote myeloma cell growth. Therefore, the inhibition of bone resorption could stop this vicious circle and not only decrease myeloma bone disease, but also the tumor progression. Preclinical studies provided strong evidence that the suppression of the osteoclast activity using bisphosphonates, RANKL blockade or inhibition of MIP-1a or MIP-1b is effective both in reducing myeloma bone disease and tumor growth and therefore may offer an important treatment strategy in multiple myeloma.

Keywords: Myeloma, bone, RANKL

Introduction Osteoclast activation Multiple myeloma is a clonal malignancy of ter- Histomorphometric findings minally differentiated plasma cells. Skeletal complica- The main principle of myeloma bone disease is an tions, including bone pain, osteolytic lesions, uncoupled bone remodeling with enhanced osteolytic pathological fractures and hypercalcemia, are a major resorption and decreased bone formation, resulting cause of morbidity and are found in up to 80% in prevailing bone destruction. Bone resorption is of myeloma patients at presentation [1]. In contrast, mediated through osteoclasts, which are derived these symptoms are rarely seen in other B-cell from the granulocyte-macrophage colony-forming malignancies. The increased bone turnover has unit and represent differentiated, multinucleated recently been characterized as an important cells. Histomorphometric analysis of bone biopsies facilitator of proliferation and tumor cell survival from myeloma patients showed osteoclast accumula- in myeloma. Several studies have given growing tion only on bone-resorbing surfaces adjacent to evidence that blocking the osteolytic process may myeloma cells, whereas osteoclasts were not increased have an antimyeloma effect. This review will give in bone not invaded by myeloma [2]. Therefore, it an overview of novel aspects in myeloma bone has been suggested that osteoclast activity is upregu- disease and show possible developments for novel lated by local osteoclast activating factors (OAFs) targets in the therapy of bone destruction in multiple which are produced by either myeloma cells or cells myeloma. of the microenvironment [3] Moreover, ex vivo

Correspondence: Orhan Sezer, Department of Hematology and Oncology, University Hospital Charite´, Berlin, Germany.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389782 20 O. Sezer cocultures of purified preosteoclasts and plasma cells pendent studies performed by several groups showed from myeloma patients showed that myeloma that human bone marrow myeloma cells express cells recruit preosteoclasts and directly induce their RANKL. In a murine model, Oyajobi et al. demon- differentiation into mature osteoclasts, which in strated that interactions between myeloma and stro- turn support survival and proliferation of myeloma mal cells lead to increased RANKL expression in both cells. Several OAFs have been implicated in the cell types. Other authors found RANKL protein and pathogenesis of myeloma bone disease. Recently, mRNA in murine myeloma cells [8]. The expression three major groups of factors have been identified as of RANKL by human bone marrow myeloma cells major osteoclast inducers in multiple myeloma: the was recently demonstrated on protein level by im- receptor activator of NF-kB ligand (RANKL), the munocytochemistry [9] and flow cytometry [10] chemokines macrophage inflammatory protein Expression of RANKL mRNA in plasma cells purified (MIP)-1a and MIP-1b, and stromal derived factor- from bone marrow aspirates of myeloma patients 1a (SDF-1a). could be demonstrated by RT-PCR by several inves-

tigators [11/13]. In a study evaluating the clinical impact of RANKL expression, nonmyelomatous The RANK/RANKL/OPG system plasma cells from controls showed no or only a weak The receptor activator of NF-kB (RANK)/RANKL/ surface expression of RANKL, whereas surface osteoprotegerin (OPG) system has been characterized RANKL could be detected on bone marrow plasma and described as the final common effector system cells from all patients with multiple myeloma. Ac- and the central pathway of osteoclast activation. cording to the bone status determined by conven- RANKL (synonym: tumor necrosis factor-related tional radiography, multiple myeloma patients were activation induced cytokine, TRANCE) is a member divided into a group with osteolytic bone lesions and a of the tumor necrosis factor (TNF) superfamily. It group without osteolysis. Bone marrow plasma cells exists as a cell membrane bound isoform, a secondary derived from patients with osteolytic bone lesions soluble variant that is cleaved from the cellular form showed a significantly higher level of surface RANKL by metalloproteases and TNF-a converting enzyme expression compared to plasma cells from patients

(TACE), and a primary secreted isoform. Under without osteolysis (P B/0.01) [11]. Moreover, mye- normal conditions, RANKL is mainly produced by loma cells have been demonstrated to enhance osteoblastic lineage cells and stromal cells. The RANKL expression by activated T-cells [14] and receptor for RANKL, RANK, is expressed by osteo- endothelial cells. Taken together, myeloma cells en- clast precursors and mature osteoclasts. Activation of hance the local RANKL availability. RANK by RANKL results in differentiation, forma- In addition to the stimulatory effects on RANKL, tion, fusion and survival of preosteoclasts. In addition, myeloma cells decrease the OPG availability within RANKL directly acts on mature osteoclasts, inducing the bone microenvironment. They lead to a down- actin ring formation and activating mature osteoclasts regulation of OPG mRNA and protein secretion in to resorb bone [4]. OPG acts as a soluble neutralizing osteoblasts and stromal cells. Furthermore, myeloma receptor for RANKL and inhibits osteoclast differ- cells produce and shed syndecan-1 (CD 138), a entiation and activation. It is secreted by stromal cells transmembrane proteoglycan, that binds to the he- and other cell types including osteoblast lineage cells. parin-binding domain of OPG and mediates its The biologic effects of RANK, RANKL and OPG internalization and lysosomal degradation [15]. Stu- have been evaluated by several in vitro and in vivo dies showed reduced serum OPG levels in patients studies. In animal models, unbalanced expression of with multiple myeloma, compared to normal controls, these cytokines led to extreme skeletal phenotypes, for and an inverse correlation of the OPG levels with the example severe osteopetrosis in RANKL knockout number of radiographic osteolytic lesions [16]. The mice. In contrast, OPG deficient mice develop combination of these effects results in an increased osteopenia. In humans, an abnormal RANKL/OPG RANKL/OPG ratio that favors the formation and ratio was found both in benign and malignant bone activation of osteoclasts. disease. Recent reports suggested that the RANKL/RANK/ MIP-1a and MIP-1b OPG system is involved in myeloma bone disease. Myeloma cells break the local RANKL/OPG balance MIP-1a has been identified as another important [5,6] by several mechanisms, they increase the avail- factor responsible for osteoclastic bone resorption in ability of RANKL within the bone marrow micro- myeloma. MIP-1a is a low molecular weight chemo- environment. Different studies have demonstrated kine that belongs to the RANTES (regulated on that myeloma cells induce RANKL expression by activation normal T-cell expressed and secreted) stromal cells through direct cell to cell contact [5,7]. family of chemokines. It binds to its receptors Whether myeloma cells directly express RANKL, has CCR1, CCR5 and CCR9 and acts as chemoattractant been discussed controversially. Recent data of inde- of phagocytes and osteoclasts. MIP-1a has been Myeloma bone disease 21 shown to induce late stage of differentiation on SDF-1a positively correlated with the presence of human osteoclast progenitors and osteoclast forma- bone lesions on radiology. In an in vitro osteoclast- tion in bone marrow cultures. Using an RNAse potentiating culture system, SDF-1a increased osteo- protection assay, Choi et al. [17] identified MIP-1a clast motility and bone-resorbing activity. as an osteoclastogenic factor in multiple myeloma. In their study, MIP-1a mRNA expression was signifi- Osteoblast inhibition cantly increased in patients with advanced myeloma as compared to normal controls. Moreover, MIP-1a In contrast to bone metastases in other malignancies, protein levels were elevated in the bone marrow multiple myeloma causes bone destruction without a supernatants of 62% of patients with active myeloma, propositional osteoblastic reaction, resulting in an but only in 17% of patients with stable myeloma and uncoupling of the normal bone remodelling sequence. not in healthy individuals. Recombinant human MIP- The inability of antiresorptive drugs to repair lytic 1a as well as bone marrow plasma from patients with bone lesions suggests that impairment of osteoblast myeloma induced osteoclast formation in human function is an important factor in myeloma bone bone marrow cultures. This effect could be inhibited disease. Serum concentrations of osteocalcin, which by a neutralizing antibody against MIP-1a or trans- reflect the osteoblastic activity, are reduced in patients fection of myeloma cells with an antisense construct with advanced disease. Histomorphometric analysis of to MIP-1a. Abe et al. showed that both MIP-1a and bone biopsies from patients with overt myeloma MIP-1b are produced and secreted by myeloma cells, showed a reduced number and activity of osteoblasts and that the secretion of MIP-1 correlates with the on bone surfaces adjacent to myeloma cells [2]. In ability of myeloma cells to enhance osteoclastic bone vitro studies revealed that myeloma cells affect the resorption [18]. Antibodies against MIP-1a and MIP- growth and function of human osteoblast-like cells. 1b or their receptor CCR5 could block these effects. Both osteoblast growth and function are inhibited The level of MIP-1a expression or secretion was when cultured in medium conditioned by myeloma correlated with the severity of myeloma bone disease cells, suggesting that this effect is due to osteoblast by several authors. inhibiting factors that are secreted by myeloma cells There is evidence that the effects of MIP-1a are [21]. However, the responsible factors remained dependent on the RANKL pathway. Both MIP-1a unclear. and MIP-1 b enhance RANKL expression in stromal A recent study suggested that Dickkopf (DKK)-1 cells.(39) In a murine model of myeloma, injection of could be one of these factors [22]. DKK-1 is an recombinant MIP-1a produced a strong increase in inhibitor of the Wnt signaling pathway, which repre- osteoclast formation in normal mice, but not in sents a major signaling pathway in osteoblasts. Wnt RANK L/OPG animals [19]. glycoproteins bind to the Wnt receptor and its coreceptors LRP5/LRP6 and inhibit the degradation of b-catenin, thus leading to its cytoplasmatic accu- Other factors mulation, translocation into the nucleus and stimula- Before the characterization of RANKL and MIP-1a tion of expression of osteoblastic target genes. In the as osteoclast inducing factors, several cytokines have absence of a Wnt signal, cytoplasmatic b-catenin is been implicated in the pathogenesis of myeloma bone phosphorylated and degraded by the proteasome. disease. These factors, e.g. lymphotoxin, TNFa, IL- Extracellular Wnt antagonists prevent ligand/recep- 1b, IL-6 have been found to be overproduced in some tor interactions and can be divided into two functional myeloma patients and therefore were discussed as classes. Members of the secreted frizzled-related potential OAFs. However, currently they can not be protein (sFRP) class, for example sFRP3 (synonym regarded as main inducers of osteoclast activation in FrzB), bind to Wnt proteins, whereas members of the multiple myeloma. DKK family bind to the LRP5/LRP6 component of A recent study by Zannettino et al. investigated the the Wnt receptor complex. Using gene-expression role of the SDF-1a in the pathogenesis of myeloma profiles of myeloma patients, Tian et al. found an bone disease [20]. SDF-1a, a chemokine highly overexpression of the DKK-1 gene in multiple mye- expressed by bone vascular endothelial and marrow loma patients with focal bone lesions. stromal cells, increases the recruitment and migration In addition to this mechanism, malignant plasma of osteoclast precursors by inducing matrix metallo- cells are able to induce osteoblast apoptosis. Silvestris proteinase-9 (MMP-9) activity. SDF-1a acts through et al. found a significantly increased expression of Fas binding to its receptor CXCR4, which is expressed on ligand (Fas-L) and tumor-necrosis-factor-related leukocytes, mature dendritic cells and osteoclast apoptosis inducing ligand (TRAIL) in myeloma cells precursors. Zannettino et al. found that myeloma and an overexpression of Fas and death receptor (DR) cells produce SDF-1a protein. In their study, mye- 4/5 by osteoblastic lineage cells obtained from patients loma patients exhibited elevated plasma levels of with extensive osteolytic lesions [23]. Further re- SDF-1a as compared to controls, and the level of search on the interaction between myeloma cells and 22 O. Sezer osteoblasts is needed in order to understand the the osteoclast induced bone disease might interrupt mechanism of osteoblast inhibition and identify pos- the feedback loop and have an additional antimye- sible therapeutic targets in the treatment of myeloma loma effect. bone disease. Bisphosphonates Evaluation of bone disease and biochemical Bisphosphonates are currently the most widely used markers class of antiresorptive drugs in the treatment of Evaluation of myeloma bone disease requires conven- myeloma bone disease. They selectively concentrate tional radiography including X-ray scans of the skull, at the boundary between osteoclast and bone resorp- cervical, thoracic and lumbar spine, ribs, proximal tion surface. Their main effect is the suppression of humeri and femora, and pelvis. The more sensitive osteoclast activity and function. Bisphosphonates are technique is magnetic resonance imaging (MRI) and based on a P/C/P structure similar to endogenous can show pathologic results even in patients with pyrophosphate. Two side chains are attached to the normal X-ray scans. However, histomorphometric carbon, influencing the antiresorptive potency. Nitro- studies have demonstrated that increased bone re- gen containing aminobisphosphonates (e.g. pamidro- sorption can be present even in the absence of nate, ibandronate, alendronate and zoledronic acid) radiographic abnormalities. New laboratory para- are more potent than the first generation bispho- meters that reflect bone metabolism may help to sphonates (e.g. clodronate and etidronate). Several overcome the diagnostic problems in estimating the large, randomized, placebo-controlled clinical trials activity of bone disease. have proved the efficacy of bisphosphonates, i.e. Amino-terminal collagen type-I telopeptide (NTx) clodronate p.o., pamidronate i.v. and zoledronic in urine and carboxy-terminal telopeptide of type-I acid, in the therapy of myeloma bone disease [29] collagen (ICTP) in serum reflect the osteoclastic Currently, the most potent bisphosphonate is zole- activity and bone destruction. In myeloma, ICTP dronic acid. Preclinical studies as well as randomized correlates with bone resorption and has a prognostic clinical trials (RCTs) showed that zoledronic acid had value. As shown by Jakob et al. serum ICTP levels a superior effect over pamidronate in the treatment of differ significantly between MGUS and myeloma and hypercalcemia of malignancy [30]. Long-term follow- increase parallel to the myeloma stage according to up data confirm that zoledronic acid was more Durie and Salmon [24]. Another study on myeloma effective than pamidronate in reducing the risk of patients demonstrated that serum levels of ICTP were skeletal complications in patients with bone metas- elevated in patients with abnormal bone MRI but tases from breast carcinoma and was of similar lacking lytic bone lesions in conventional radiography efficacy in patients with multiple myeloma [31]. [25]. Tartrate-resistant acid phosphatase isoform-5b Although the main effect of bisphosphonates is the (TRACP-5b) was described as another novel para- suppression of osteoclast activity, other mechanisms meter reflecting osteoclast activity. Terpos et al. found may add to the effects of aminobisphosphonates in increased TRACP-5b serum levels in myeloma pa- multiple myeloma. Phase III studies evaluating the tients, and the levels were associated with the radio- use of bisphosphonates in smoldering myeloma (pri- graphically assessed severity of bone disease [26]. mary end point: prolongation of time to progression) are ongoing [32]. Treatment approaches RANKL antagonists Several in vitro and in vivo experiments as well as clinical observations have shown the importance of After the identification of the RANK/ RANKL/ OPG the positive feedback between myeloma progression system as the final effector system of osteoclast and bone resorption for sustaining the disease process activation, systemic RANKL blockade has been [27,28]. The close association between bone resorp- evaluated in animal models and first clinical trials, tive activity and myeloma progression has been high- using OPG, OPG-Fc fusion protein or RANK-Fc. lighted in an experiment using C57BL/KaLwRij mice. Treatment of myelomatous SCID-hu hosts with In these animals, ovariectomy induced an accelerated RANK-Fc not only reduced osteoclast formation bone remodelling, resulting in reduced trabecular and myeloma-induced bone resorption, but also bone volume and increased osteoclast number. Non- resulted in a sustained suppression of paraprotein ovariectomized mice were used as control. Injection of levels by more than 80%, reflecting a reduced 5T2MM myeloma cells lead to bone disease in both myeloma cell burden. Moreover, studies in the groups with an increased tumor growth and earlier 5T2MM model showed that RANKL blockade by development of osteolytic lesions in ovariectomized OPG-Fc caused not only inhibition of development of mice, thereby confirming the interdependence of osteolytic bone lesions, but as well a decreased tumor osteoclasts and myeloma cells. Therefore, targeting burden and a significant increase in time to morbidity. Myeloma bone disease 23

The treatment was associated with a decreased nuclear factor-kappaB ligand expression by human myeloma cells mediates osteoclast formation in vitro and correlates with number of osteoclasts but had no effect on apoptosis bone destruction in vivo. Cancer Res 2003;63:/ 5438/ /5445. and proliferation of 5T33MM cells in vitro, indicating [13] Lai FP, Cole-Sinclair M, Cheng WJ, Quinn JM, Gillespie MT, that the antimyeloma effect of RANKL inhibitors is Sentry JW, Schneider HG. Myeloma cells can directly related to inhibition of osteoclast activity. contribute to the pool of RANKL in bone bypassing the In humans, trials with an OPG-Fc fusion protein, classic stromal and osteoblast pathway of osteoclast stimula- with an recombinant osteoprotegerin construct and tion. Br J Haematol 2004;126:/ 192/ /201. [14] Giuliani N, Colla S, Sala R, Moroni M, Lazzaretti M, La with a human monoclonal antibody to RANKL were Monica S, Bonomini S, Hojden M, Sammarelli G, Barille S, et performed. These preliminary observations indicate al. Human myeloma cells stimulate the receptor activator of that targeting the RANK/RANKL/OPG system may nuclear factor-kappa B ligand (RANKL) in T lymphocytes: a inhibit the development of myeloma bone disease and potential role in multiple myeloma bone disease. Blood 2002;/ also decreases myeloma growth and may therefore 100:4615/ /4621. offer a new treatment strategy in multiple myeloma. [15] Standal T, Seidel C, Hjertner O, Plesner T, Sanderson RD, Waage A, Borset M, Sundan A. Osteoprotegerin is bound, internalized, and degraded by multiple myeloma cells. Blood Acknowledgements 2002;100:/ 3002/ /3007. [16] Seidel C, Hjertner O, Abildgaard N, Heickendorff L, Hjorth This work was supported by Deutsche Forschungs- M, Westin J, Nielsen JL, Hjorth-Hansen H, Waage A, Sundan gemeinschaft (DFG). A, Borset M. Nordic Myeloma Study Group. Serum osteoprotegerin levels are reduced in patients with multiple myeloma with lytic bone disease. Blood 2001;98:/ 2269/ /2271. [17] Choi SJ, Cruz JC, Craig F, Chung H, Devlin RD, Roodman References GD, Alsina M. Macrophage inflammatory protein 1-alpha is a [1] Melton LJ 3rd, Kyle RA, Achenbach SJ, Oberg AL, Rajkumar potential osteoclast stimulatory factor in multiple myeloma. SV. Fracture risk with multiple myeloma: a population-based Blood 2000;96:/ 671/ /675. [18] Abe M, Hiura K, Wilde J, Moriyama K, Hashimoto T, Ozaki study. J Bone Miner Res 2005;20:/ 487/ /493. [2] Bataille R, Chappard D, Marcelli C, Dessauw P, Baldet P, S, Wakatsuki S, Kosaka M, Kido S, Inoue D, Matsumoto T. Sany J, Alexandre C. The recruitment of new osteoblasts and Role for macrophage inflammatory protein (MIP)-1alpha and osteoclasts is the earliest critical event in the pathogenesis of MIP-1beta in the development of osteolytic lesions in multiple myeloma. Blood 2002;100:/ 2195/ /2202. human multiple myeloma. J Clin Invest 1991;88:/ 62/ /66. [3] Mundy GR, Raisz LG, Cooper RA, Schechter GP, Salmon [19] Oyajobi BO, Franchin G, Williams PJ, Pulkrabek D, Gupta A, SE. Evidence for the secretion of an osteoclast stimulating Munoz S, Grubbs B, Zhao M, Chen D, Sherry B, Mundy GR. Dual effects of macrophage inflammatory protein-1alpha on factor in myeloma. N Engl J Med 1974;291:/ 1041/ /1046. [4] Lacey DL, Timms E, Tan HL, Kelley MJ, Dunstan CR, osteolysis and tumor burden in the murine 5TGM1 model of Burgess T, Elliott R, Colombero A, Elliott G, Scully S, et al. myeloma bone disease. Blood 2003;102:/ 311/ /319. Osteoprotegerin ligand is a cytokine that regulates osteoclast [20] Zannettino AC, Farrugia AN, Kortesidis A, Manavis J, To LB, Martin SK, Diamond P, Tamamura H, Lapidot T, Fujii N, differentiation and activation. Cell 1998;93:/ 165/ /176. [5] Pearse RN, Sordillo EM, Yaccoby S, Wong BR, Liau DF, Gronthos S. Elevated serum levels of stromal-derived factor- Colman N, Michaeli J, Epstein J, Choi Y. Multiple myeloma 1alpha are associated with increased osteoclast activity and disrupts the TRANCE/ osteoprotegerin cytokine axis to osteolytic bone disease in multiple myeloma patients. Cancer trigger bone destruction and promote tumor progression. Res 2005;65:/ 1700/ /1709. [21] Evans CE, Ward C, Rathour L, Galasko CB. Myeloma affects Proc Natl Acad Sci USA 2001;98:/ 11581/ /11586. [6] Sezer O, Heider U, Zavrski I, Kuhne CA, Hofbauer LC. both the growth and function of human osteoblast-like cells. RANK ligand and osteoprotegerin in myeloma bone disease. Clin Exp Metastasis 1992;10:/ 33/ /38. [22] Tian E, Zhan F, Walker R, Rasmussen E, Ma Y, Barlogie B, Blood 2003;101:/ 2094/ /2098. [7] Giuliani N, Bataille R, Mancini C, Lazzaretti M, Barille S. Shaughnessy JD Jr. The role of the Wnt-signaling antagonist Myeloma cells induce imbalance in the osteoprotegerin/ DKK1 in the development of osteolytic lesions in multiple osteoprotegerin ligand system in the human bone marrow myeloma. N Engl J Med 2003;349:/ 2483/ /2494. [23] Silvestris F, Cafforio P, Tucci M, Grinello D, Dammacco F. environment. Blood 2001;98:/ 3527/ /3533. [8] Croucher PI, Shipman CM, Lippitt J, Perry M, Asosingh K, Upregulation of osteoblast apoptosis by malignant plasma

Hijzen A, Brabbs AC, van Beek EJ, Holen I, Skerry TM, et al. cells: a role in myeloma bone disease. Br J Haematol 2003;/ Osteoprotegerin inhibits the development of osteolytic bone 122:39/ /52. disease in multiple myeloma. Blood 2001;98:/ 3534/ /3540. [24] Jakob C, Zavrski I, Heider U, Brux B, Eucker J, Langelotz C, [9] Sezer O, Heider U, Jakob C, Zavrski I, Eucker J, Possinger K, Sinha P, Possinger K, Sezer O. Bone resorption parameters Sers C, Krenn V. Immunocytochemistry reveals RANKL [carboxy-terminal telopeptide of type-I collagen (ICTP), expression of myeloma cells. Blood 2002;99:/ 4646/ /4647. amino-terminal collagen type-I telopeptide (NTx), and deox- [10] Sezer O, Heider U, Jakob C, Eucker J, Possinger K. Human ypyridinoline (Dpd)] in MGUS and multiple myeloma. Eur J bone marrow myeloma cells express RANKL. J Clin Oncol Haematol 2002;69:/ 37/ /42. 2002;20:/ 353/ /354. [25] Jakob C, Zavrski I, Heider U, Bollow M, Schulz CO, Fleissner [11] Heider U, Langelotz C, Jakob C, Zavrski I, Fleissner C, C, Eucker J, Michael R, Hamm B, Possinger K, Sezer O. Eucker J, Possinger K, Hofbauer LC, Sezer O. Expression of Serum levels of carboxy-terminal telopeptide of type-I col- receptor activator of nuclear factor kappaB ligand on bone lagen are elevated in patients with multiple myeloma showing marrow plasma cells correlates with osteolytic bone disease in skeletal manifestations in magnetic resonance imaging but

patients with multiple myeloma. Clin Cancer Res 2003;9:/ / lacking lytic bone lesions in conventional radiography. Clin 1436/1440. Cancer Res 2003;9:/ 3047/ /3051. [12] Farrugia AN, Atkins GJ, To LB, Pan B, Horvath N, Kostakis [26] Terpos E, de la Fuente J, Szydlo R, Hatjiharissi E, Viniou N, P, Findlay DM, Bardy P, Zannettino AC. Receptor activator of Meletis J, Yataganas X, Goldman JM, Rahemtulla A. Tartrate- 24 O. Sezer

resistant acid phosphatase isoform 5b: a novel serum marker [30] Major P, Lortholary A, Hon J, Abdi E, Mills G, Menssen HD, for monitoring bone disease in multiple myeloma. Int J Cancer Yunus F, Bell R, Body J, Quebe-Fehling E, Seaman J. 2003;106:/ 455/ /457. Zoledronic acid is superior to pamidronate in the treatment [27] Yaccoby S, Pearse RN, Johnson CL, Barlogie B, Choi Y, of hypercalcemia of malignancy: A pooled analysis of two

Epstein J. Myeloma interacts with the bone marrow micro- randomized, controlled clinical trials. J Clin Oncol 2001;19:/ /

environment to induce osteoclastogenesis and is dependent on 558/567. osteoclast activity. Br J Haematol 2002;116:/ 278/ /290. [31] Rosen LS, Gordon D, Kaminski M, Howell A, Belch A, [28] Libouban H, Moreau MF, Basle MF, Bataille R, Chappard D. Mackey J, Apffelstaedt J, Hussein MA, Coleman RE, Reitsma Increased bone remodeling due to ovariectomy dramatically DJ, et al. Long-term efﬁcacy and safety of zoledronic acid increases tumoral growth in the 5T2 multiple myeloma mouse compared with pamidronate disodium in the treatment of model. Bone 2003;33:/ 283/ /292. skeletal complications in patients with advanced multiple [29] Berenson JR, Lichtenstein A, Porter L, Dimopoulos MA, myeloma or breast carcinoma: a randomized, double-blind, Bordoni R, George S, Lipton A, Keller A, Ballester O, Kovacs multicenter, comparative trial. Cancer 2003;98:/ 1735/ /1744. MJ, et al. Efﬁcacy of pamidronate in reducing skeletal events [32] Sezer O, Jakob C, Zavrski I, Heider U, Fleissner C, Freund M. in patients with advanced multiple myeloma. N Engl J Med Bisphosphonates in early multiple myeloma. Eur J Haematol 1996;334:/ 488/ /493. 2003;71:/ 231/ /232. Hematology, 2005; 10 Supplement 1: 25

MULTIPLE MYELOMA

Novel treatments of multiple myeloma

M. A. DIMOPOULOS1, A. ANAGNOSTOPOULOS2, & E. TERPOS3

1University of Athens, Athens, Greece, 2Hematology Department and HCT Unit, G Papanicolaou Hospital, Thessaloniki, Greece, and 3Department of Haematology, 251 General Airforce Hospital, Athens, Greece

Over the last decade significant advances in the of patients with refractory/relapsed MM including treatment of multiple myeloma (MM) have been patients who have failed thalidomide. The combina- achieved. Most prospective randomized trials indicate tion of bortezomib with thalidomide and dexametha- that high dose therapy with autologous stem cell sone in previously untreated patients has been transplantation is associated with improved event- associated with a response rate exceeding 80%. free and overall survival. Thalidomide, an oral More recently the Imid lenalidomide combined with immunomodulatory and anti-angiogenetic agent is dexamethasone has been associated with significantly associated with objective responses in 30% of patients higher response rates and longer event-free survival with advanced and refractory MM. The addition of than dexamethasone alone. Lenalidomide and dexa- dexamethasone to thalidomide is associated with a methasone is also very active in previously untreated 50% response rate in refractory patients and with a patients. The development of genomics and proteo- 70% response rate in previously untreated patients. mics provide the basis for a novel molecular classifica- This oral combination is particularly useful in newly tion of MM, can identify unique targets for diagnosed patients with features of advanced disease combination therapy in individual patients and pro- who are candidates for stem cell collection. Bortezo- vide the framework for clinical protocols to enhance mib is a potent and selective inhibitor of the 26S cytotoxicity and to avoid the emergence of drug proteasome. Bortrezomib has shown activity in 30% resistance.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389791 Hematology, 2005; 10 Supplement 1: 26/28

MULTIPLE MYELOMA

Stem cell transplantation in multiple myeloma

BO BJO¨ RKSTRAND

Institutet at Huddinge Hospital, Stockholm, Sweden

Single autologous transplantation [3/5]. These data have been confirmed in recent randomized trials. In the French IFM94-study [6], High-dose treatment (HDT) with autologous stem 399 previously untreated patients were randomized at cell transplantation (ASCT) has been shown to increase survival and time to progression in multiple diagnosis to receive either single or double ASCT myeloma (IMM) compared with conventional che- following induction treatment. Conditioning in the motherapy. In the French IFM90-study [1], 200 single transplant arm and for the second transplant in patients were randomised to receive prolonged stan- the double transplant arm was melphalan 140 mg/ dard treatment with i.v. combination chemotherapy, sqm plus TBI 8 Gy total dose in 4 fractions. For the or a short phase by induction treatment with the same first transplant in the double transplant group, mel- regimen, followed by high-dose melphalan / TBI phalan 140 mg/sqm was used. The rate of CR or and autologous bone marrow transplantation VGPR was 49% and 63% in the single and double (ABMT). Median survival was signficantly better in transplant groups, respectively (P/0.10). Median OS the ABMT-arm, /60 months, vs. 37 months in the was prolonged with 10 months in the double trans- chemotherapy arm, and also event-free survival was plant group (58 months, vs. 48 months for the single superior, 27 months vs. 18 months. These differences transplant group, P/0.01). The probability of survi- are sustained in a long term follow-up, where actuarial val at 7 years was 21% vs. 42% and of EFS 10% vs. survival at 7 years was 43% and 25%, and EFS was 20% for the single and double transplant groups, 16% and 8% in the ABMT- and chemotherapy arms, respectively. In an Italian randomized study of 386 respectively. However, these latter differences were patients conditioning for the first transplantation was only observed in patients under the age of 60 at the melphalan 200 mg/sqm and for the second melphalan time of inclusion in the study, while there was no 120 mg/sqm plus busulfan 12 mg/kg. Median OS was benefit in older patients [1a]. Similar results have 62 vs. /74 months and median EFS 21 vs. 31 months been demonstrated in a more recently published study for the single and double transplant groups, respec- from the British MRC [2]: 401 patients were rando- tively [7]. In contrast to these results, another French mized to receive either standard chemotherapy randomized study in 230 patients has failed to (cVAMP) or cVAMP followed by high-dose melpha- demonstrate a significant benefit of the double lan and autologous SCT. Median overall survival transplant procedure [8]. Concerning the optimal (OS) was signficantly superior in the ASCT-arm/ timing of the two transplants, an EBMT registry 51 months */ compared with 42 months in the study has recently demonstratred that the second cVAMP-group. The difference was most pronounced transplant should preferably be performed before in patients with a high beta-2-microglobulin (b-2-m) relapse and within 6 to 12 months of the first value at diagnosis, i.e. with a poorer prognosis. transplantation [9]. It can be concluded that tandem autologous Tandem autologous transplantation transplantation seems to improve survival with about During the last 10 years, there have been indications one year compared with single autografting, and that from non randomized trials and treatment programs it is beneficial to perform the second transplant within that repeated HDT with ASCT, i.e. tandem or double a short interval after the first, before disease progres- transplantation, can further improve these results sion.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389809 Stem cell transplantation in multiple myeloma 27

Allogeneic stem cell transplantation with full reducing the treatment-related complications of dose conditioning high-dose conditioning [13,14]. In a retrospective survey of 256 patients reported to the EBMT Allogeneic stem cell transplantation (SCT) with full- myeloma registry, TRM was about 24%, but only dose conditioning has been used in MM for almost 20 13% in low-risk patients, i.e. mainly patients trans- years. The treatment results have been hampered by a planted upfront as a part of first-line treatment [15]. very high transplant-related mortality (TRM) rate of One treatment rationale frequently used is to induce 25/50%, and furthermore, the curative potential can maximal cytoreduction by induction treatment, HDT be questioned since late relapses after several years do (mainly high-dose melphalan) and ASCT, followed by occur [10,11], although molecular remissions can be consolidation with with RIC allogeneic transplanta- achieved in a way not seen after autologous trans- tion and if necessary, also infusion of alloreactive plantation [12]. In 1996, the EBMT Myeloma donor T-lymphocytes (DLI). This strategy has been Registry performed a retrospective analysis comparing used in a number of phase II-studies, of totally 124 189 allogeneic transplant patients with 189 case- patients [16/19]. The conditioning regimens have matched autologous transplant controls [10]. The been variable and of low or intermediate cytotoxic results were consistent with previous surveys, with a intensity, and have included fludarabine, low dow significantly better survival for the autotransplant TBI, melphalan, cyclophosphamide and ATG. Trans- group (median survival after transplantation 34 plnat-related mortality has varied between 4/13% in months, vs. 18 months in the allogeneic group). previously untreated or less treated patients, and Relapse rate was significantly higher in the autotrans- about 25% in relapsed and/or heavily pretreated plant group (70%) vs. 50% for the allogeneic trans- patients. OS and EFS at 2/3 years was 50/100% plant patients. This difference did not however and 30/60%, respectively in less-treated or relapsed * compensate for the large difference in TRM /41% but chemosensitive patients. The incidence of acute in the allogeneic vs. 13% in the autologous transplant and chronic GVHD varied between 10/60% and 40/ * group, respectively /which was the reason for the 65%, respectively. As for controlled prospcctive trials, poorer survival of the allogeneic transplant patients. preliminary data from one study have been presented In spite of extensive analyses of different prognostic [20]. Two hundred and eighty-four previously un- subgroups, no category of patients could be identified treated patients have been included, all receving VAD where allogeneic transplantation would be more induction treatment and high-dose melphalan (200 favorable than autologous, although no survival dif- mg/sqm) with ASCT. For inclusion, the patients ference was seen in female patients. needed to have poor prognosis criteria with respect A more recent EBMT study compared 225 allo- to abnormal karyotype and high b-2-m. Sixty-four geneic bone marrow transplants performed during patients with an HLA-identical sibling donor then 1994 /98 with 339 patients transplanted during received an RIC allogeneic transplantation, while 220 1983 /93 [11] This analysis demonstrated a decrease patients with no available donor were randomized to in TRM, with a reduction of total TRM from 50 to no further treatment or to receive a second autograft 30%, and early mortality from 30 to 20%. This after high-dose melphalan 220 mg/sqm. The allo- resulted in an improvement in outcome, with an geneic transplant conditioning regimen consisted of actuarial survival of about 50% at 4 years after fludarabine, low dose busulfan and ATG. Median transplantation, compared to about 30% in the follow-up time from diagnosis is 23 months (9/44). previously transplanted group of patients. Still, At 4 years, the probability of survival*/on an inten- although improved, TRM is unacceptably high. tion-to-treat basis*/was 40% in the autologous vs.

Based on these and other studies with similar 30% in the allogeneic transplant arms (P/0.36), and results, allogeneic SCT with full dose conditioning EFS 20% vs. 15% (P/0.70). When comparing the can not be recommended, not even for younger patients who actually received their planned trans- patients, until the lethal complications can be handled plants */ 46 in the allogeneic- and 166 in the more effectively. autologous transplant group, the corresponding results for survival was 50% for the autologous and 35%

for the allogeneic transplant patients (P/0.09), i.e. a Allogeneic stem cell transplantation with trend for better outcome after autologous SCT. EFS reduced intensity conditioning was essentially identical in both groups / about 18% Based on the high toxicity of allogeneic SCT with full at 4 years. Transplant-related mortality at 3 months dose conditioning, it is not surprising that MM is one after allogeneic transplantation in the 406 patients obvious target disease for reduced intensity-condi- who completed all treatment was 6%. tioning (RIC) or non-myeloablative allogeneic In a retrospective registry study of the EBMT [21], (NMA) transplantation, also called ‘mini-transplanta- 321 RIC allografts were compared with 196 allo- tion’ where the rationale is to utilize the antitumor transplants with standard conditioning (SC). The effect of the graft-versus-myeloma reaction, while transplants were performed during the same period 28 Bo Bjo¨rkstrand

[8] Fermand J-P, Alberti C, Marolleau J-P. Single versus tandem of time, 1998/2002. No case matching was done, and the groups were not completely comparable with high dose therapy (HDT) supported with autologous blood stem cell (ABSC) transplantation using unselected or CD34- respect to prognostic factors. Median survival was enriched ABSC: results of a two by two designed randomized 23 and 36 months (P/N.S) and PFS was 11 and 17 trial in 230 young patients with multiple myeloma (MM).

months (P/N.S) for the RIC and SC groups, Hematol J 2003;4:/ S59./ respectively. Non-relapse mortality at 2 years was [9] Morris C, et al. Beneﬁt and timing of second transplantations in multiple myeloma: clinical ﬁndings and methodological limita- significantly higher for the SC group / 37% vs. 23% tions in a European Group for Blood and Marrow Transplanta- for the RIC patients. The cumulative relapse inci- tion registry study. J Clin Oncol 2004;22:/ 1674/ /1681. dence at 3 years was higher in the RIC group, 54%, [10] Bjo¨rkstrand B, et al. Allogeneic bone marrow transplantation compared with 26% in the group that received SC. versus autologous stem cell transplantation in multiple mye- Allogeneic SCT with reduced intensity conditioning loma-a retrospective case-matched study from the European

Group for Blood and Marrow Transplantation. Blood 1996;/ is feasible, and TRM is low and acceptable in contrast 88:4711/ /4718. to SC. However, relapse rate seems to be higher, and [11] Gahrton G, et al. Progress in allogeneic bone marrow and long-term outcome does not seem to be improved. peripheral blood stem cell transplantation for multiple mye- High-dose treatment and ASCT followed RIC allo- loma: a comparison between transplants performed 1983/ geneic transplantation is feasible, but preliminary 1993 and 1994 /1998 at EBMT centers. Br J Haematol results from controlled prospective trials still do not 2001;113:/ 209/ /216. [12] Corradini P, et al. Molecular remission after myeloablative demonstrate a benefit in survival or freedom of allogeneic stem cell transplantation predicts a better relapse-

progression compared with single or tandem auto- free survival in patients with multiple myeloma. Blood 2003;/ logous SCT, at least not in high-risk patients. Data 102:1927/ /1929. from other similar studies, like the EBMT study with [13] Badros A, et al. High response rate in refractory and poor-risk a similar design as the French trial but also including multiple myeloma after allotransplantation using a nonmyeloablative conditioning regimen and donor lymphocyte infu- patients without high-risk criteria, will be available in sions. Blood 2001;97:/ 2574/ /2579. the near future. [14] McSweeney PA, et al. Hematopoietic cell transplantation in older patients with hematologic malignancies: replacing high- dose cytotoxic therapy with graft-versus-tumor effects. Blood References 2001;97:/ 3390/ /3400. [15] Crawley C, et al. Results of reduced intensity conditioning [1] Attal M, et al. A prospective randomized trial of autologous allogeneic transplantation in myeloma-a report from the

bone marrow transplantation and chemotherapy in multiple EBMT. Hematol J 2003;4:/ S224./ myeloma. N Engl J Med 1996;335:/ 91/ /97. [16] Badros A, et al. Improved outcome of allogeneic transplanta-

[1a] Harousseau J-L, Attal M. The IFM-90 trial. Hematol J 2003;/ tion in high-risk multiple myeloma patients after nonmyeloa- 4:S55./ blative conditioning. J Clin Oncol 2002;20:/ 1295/ /1303. [2] Child JA, et al. High-dose chemotherapy with hematopoietic [17] Kro¨ger N, et al. Autologous stem cell transplantation followed

stem-cell rescue for multiple myeloma. N Engl J Med 2003;/ by a dose-reduced allograft induces high complete remission 348:1875/ /1883. rate in multiple myeloma. Blood 2002;100:/ 755/ /760. [3] Bjo¨rkstrand B, et al. Double high-dose chemo-radiotherapy [18] Einsele H, et al. Follow-up of patients with progressive with autologous stem cell transplantation can induce mole- multiple myeloma undergoing allografts after reduced-inten- cular remissions in multiple myeloma. Bone Marrow Trans- sity conditioning. Br J Haematol 2003;121:/ 411/ /418. plant 1995;15:/ 367/ /371. [19] Maloney DG, et al. Allografting with nonmyeloablative [4] Barlogie B, et al. Superior outcome after early autotransplan- conditioning following cytoreductive autografts for the treat-

tation (AT) with ‘‘total therapy’’ (TT) compared to standard ment of patients with multiple myeloma. Blood 2003;102:/ /

SWOG treatment (ST) for multiple myeloma (MM). Blood 3447/3454.

1995;86:/ 207a./ [20] Harousseau, J-L Preliminary results of the IFM9903 and [5] Barlogie B, et al. Total therapy with tandem transplants for IFM9904 protocols comparing autologous followed by min- newly diagnosed multiple myeloma. Blood 1999;93:/ 55/ /65. iallogeneic transplantation and double autologous transplant [6] Attal M, et al. Single versus double autologous stem-cell in high-risk de novo multiple myeloma. 10th International

transplantation for multiple myeloma. N Engl J Med 2003;/ Myeloma Workshop, Sydney 2005. 349:2495/ /2502. [21] Crawley, C, et al, Reduced-intensity conditioning does not [7] Cavo M, et al. Single vs. tandem autologous transplants in improve survival compared to standard conditioning for multiple myeloma: Italian experience. Hematol J 2003;4:/ S60/ / patients with myeloma. Bone Marrow Transplant 2005; 35: 61. S48. Hematology, 2005; 10 Supplement 1: 29/32

BLEEDING DISORDERS

Diagnostic approaches to bleeding disorders

NIGEL S. KEY

University of Minnesota Medical School, Minneapolis, Minnesota, USA

Introduction and general approach ing sensitivity, specificity, reproducibility, and inconsistency in interpretation between laboratories. The evaluation of a patient with a possible bleeding Given all these limitations, it is not uncommon to disorder can be one of the most challenging referrals encounter patients with a history that is strongly in hematology practice. In the absence of much suggestive of a bleeding disorder yet no diagnosis is evidence-based outcomes data, the practitioner’s established after an exhaustive laboratory evaluation. experimental judgment will be called upon to for- Because it is generally not feasible to refer every such mulate a working diagnosis and management plan in patient to a research laboratory for advanced investi- many instances. A logical and semi-standardized gation, empiric management plans may need to be approach should include the application of a three- instituted, using some combination of available hemo- part screening process, each of which addresses a static therapies. specific question, as follows:

1. A comprehensive bleeding history should be The bleeding history: The best screening test for obtained with the specific question; ‘what is the a bleeding disorder? pre-test probability that this patient has a bleeding The bleeding history should include the age of onset disorder?’ One possible outcome is that no of symptomatic bleeding, the pattern of bleeding both further laboratory evaluation is deemed to be with respect to anatomical distribution (muco-cuta- necessary, for example in a patient who com- neous vs. deep tissue) and time of onset after plains of minimal muco-cutaneous bleeding hemostatic challenges, and whether hemorrhage has symptoms, and who has had many hemostatic challenges without excess bleeding. been spontaneous and/or provoked after minor and 2. Global screening tests of hemostasis should then major hemostatic challenges. A detailed family pedi- applied with the specific question; ‘what is the gree, and inquiry about systemic disorders and likelihood that there is a hemostatic abnormality medication exposure are also important historical to explain (and are the screening tests compa- aspects. The details of excessive site-specific bleeding tible with) the patient’s history?’ Again, depend- should be carefully documented. For example, histor- ing on the perceived pre-test probability and the ical features of menorrhagia that are predictive of an result of these screening test(s), no further underlying bleeding disorder are an onset at me- evaluation may be judged to be necessary. narche, concomitant iron deficiency anemia, hourly or 3. Specific screening tests of hemostasis may then greater pad/tampon changes, and the passage of clots. be applied in order to address the question; ‘what Until recently, there have been few published studies specific hemostatic disorder does this patient suffer on the positive predictive value of site-specific or from?’ global bleeding symptoms. Indeed, relatively sparse attention has been paid to the development and Unfortunately, the available global and specific la- validation of user-friendly bleeding scales that are boratory assays suffer from many limitations, includ- also highly predictive of an underlying disorder.

Correspondence: Dr Nigel S. Key, Professor of Medicine, Division of Hematology, Oncology, and Transplantation, MMC 480 Mayo Building, 420 Delaware St S.E., Minneapolis, MN 55455, USA. Tel: 612-624-8903. Fax: 612-625-6919. E-mail: [email protected]

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331390050 30 N. S. Key

Possible exceptions are the menstrual blood loss secretion defects. This is an unfortunate weakness as (MBL) scale developed by Kadir et al., although the these disorders are relatively prevalent. definition of ‘abnormal’ (/80 mL of blood loss/ In practice, many hematologists elect to screen for period) is measured by the alkaline hematin method vWD in patients with a compatible muco-cutaneous which is not readily performed in a clinical practice and/or post-surgical bleeding history, and reserve setting. A pictorial chart illustrating blood loss on referral for platelet aggregation studies (to evaluate sanitary napkins with an associated scoring system is for intrinsic platelet disorders) for those subjects in more practical, and has been reasonably well vali- whom vWD has been ruled out. The PFA-100 pattern dated. A pictorial blood assessment score /100 is of abnormality may be particularly useful in this frequently used as cut-off to define excessive menor- regard; when both the collagen/epinephrine and rhagia. collagen/ADP closure times are abnormally pro- In the ISTH criteria for the diagnosis of von longed, vWD is more likely. Conversely, when just

Willebrand disease, a positive muco-cutaneous bleed- the collagen/epinephrine closure time is prolonged / ing history is defined by ]/2 symptoms in the absence and salicylate/NSAID ingestion has been ruled out / of a blood transfusion history, or one symptom platelet aggregrometry is more likely to reveal an requiring blood transfusion, or one symptom recur- abnormality of platelet function. A normal PFA-100 ring on ]/3 distinct occasions. The ongoing European study has a reasonably high negative predictive value Union collaborative study on type 1 von Willebrand for vWD, but probably not for intrinsic platelet disease (MCMDM-1VWD) has recently developed function defects. Therefore, platelet aggregometry and validated a questionnaire for muco-cutaneous should be considered in any patient with a suggestive bleeding that ascribes a severity score based on the history but normal screening studies. clinical impact/outcome of site-specific bleeding; because it is being correlated with laboratory parameters Testing for von Willebrand Disease (vWD) and vWF genotype, it may become the standard assessment tool in the future. Testing for vWD using a combination of the factor VIII activity assay (FVIII:C), vW antigen (VW:Ag) and ristocetin cofactor (VW:RCo) activities is indi- What are the best laboratory global screening cated in most patients presenting with muco-cuta- tests for a bleeding disorder? neous bleeding manifestations. Sub-type classification Global screening tests of hemostasis should evaluate by analysis of vW multimers can be reserved for for both ‘primary’ and ‘secondary’ hemostatic ab- patients with suggestive abnormalities on the above normalities. The Prothrombin and Activated Partial screen. Once again, the historical details will provide Thromboplastin Times (PT & APTT) are reasonably some measure of pre-test probability of vWD. For sensitive and specific assays for factor deficiency example, it has been estimated that 5/20% of women states. Furthermore, these tests are reproducible, with menorrhagia have underlying vWD. Conversely, inexpensive, and easily performed. A platelet count 60/95% of women with vWD suffer from menor- and review of platelet morphology by simple light rhagia. It should be borne in mind that the diagnosis microscopy (which may pick up abnormalities such as of vWD may be masked by the fact that assays for the Gray Platelet and Wiskott-Aldrich Syndromes) is vW factor complex may be erroneously high during mandatory. periods of inflammation and ‘stress’ (e.g., in young More uncertainty revolves around the best labora- children subjected to phlebotomy and in many tory screening strategy for vWD or platelet function hospitalized patients) and perhaps in women on abnormalities in patients presenting with a history of estrogen therapy. In women of menstruating age, abnormal muco-cutaneous bleeding. Formerly, the most studies agree that testing should be performed

Bleeding Time (BT) was most commonly used for on days1/4 of the menstrual cycle, when vW complex this purpose, but it is invasive, relatively insensitive, proteins may be at their nadir. and poorly reproducible. Many laboratories have now Type 1, characterized by a partial quantitative either abandoned the BT completely or added the deficiency of vWF, is the most common variant, TM Platelet Function Analyzer (PFA-100 ) as a global accounting for 70/80% of all vWD. It has pointed screen for platelet disorders. The PFA-100 is sensitive out that the application of commonly used diagnostic to abnormalities that affect high shear-dependent criteria for type 1 vWD may include many false platelet adhesion and aggregation; its utility is sig- positives, since bleeding and low vWF levels often nificantly better than the BT in the detection of vWD associate by chance. Nothwithstanding this proviso, and severe platelet function defects (e.g., Glanz- type 1 vWD is defined by a quantitative deficiency of mann’s Thrombasthenia, Bernard-Soulier Syn- vWF / that is, vW:Ag and vWF:RCo levels /2 S.D. drome), but like the BT, it suffers from a lack of below the (ABO blood type population-adjusted) sensitivity in the detection of mild platelet function mean, with a normal vW multimer pattern / in a disorders, such as storage pool defects and platelet patient with bleeding symptoms and a positive family Diagnostic approaches to bleeding disorders 31 history. Any diagnostic algorithm is further compli- lumi-aggregometry is considered to be a reasonable cated by the fact that the disorder, although classically screen for d-SPD. thought of as autosomal dominant, is variably pene- trant in many families. Among the causes of acquired ‘Third Line’ laboratory investigation vWD, hypothyroidism is probably the most prevalent, and a routine screen of TSH levels as part of a Depending on the suspected defect at this stage, bleeding work-up is appropriate. A FVIII/vW:Ag ratio specialized laboratories may be capable of performing

]/1.50 (compared to 5/1.10 in healthy subjects) is additional sophisticated studies. These studies should helpful in the diagnosis of type 1 vWD, although be considered as confirmatory rather than screening. vWF:RCo levels are the single measure of vWF- Examples include platelet electron microscopy to related activities that best correlate with bleeding examine for the absence of dense granules in sus- severity. Recent data from the on-going MCMDM pected d-SPD. Routine application of this technique

1-VWD study suggest that :/70% of subjects diag- has suggested that ‘forme fruste’ variants of this nosed with type 1 vWD have a detectable candidate disorder, in which there is a partial deficiency of mutation in the vW gene, the majority of which are dense granules, may be relatively prevalent in women missense mutations. with bleeding disorders. Other studies that may be of value in certain scenarios include platelet flow cytometry (e.g., in suspected Glanzmann’s or Bernard- Platelet aggregation tests Soulier syndromes), and specific studies of platelet Although still considered the ‘gold standard’ for the release (e.g., serotonin release), receptor deficiency, diagnosis of intrinsic platelet function defects, platelet or signal transduction. aggregation studies, which were first introduced in the 1960s, suffer from poor accuracy and reproducibility. Tailoring the work-up for bleeding disorders: Aggregometry is usually performed using ADP, epinephrine, arachidonic acid, collagen, and ristocetin As a general rule, the more severe the bleeding agonists. Patients undergoing platelet function tests disorder, the earlier in life the presentation. However, (PFTs) should be instructed to refrain from ingestion recent studies have begun to evaluate the clinical and of all potentially platelet inhibitory drugs for at least laboratory features of mild bleeding disorders in 10 days prior to testing. Dietary factors may also children. These studies, like those in adults, conclude affect aggregation responses. Recent studies have that platelet function defects (abnormal aggregation begun to re-address some of the standardization and/or ATP release) may account for a relatively large challenges of these assays, including pre-analytical proportion of bleeding diatheses in children. In variables. In particular, it has been suggested that the addition, it is clear that the cause of muco-cutaneous addition of platelet-poor plasma (PPP) to platelet-rich bleeding remains enigmatic in many children referred plasma (PRP) to adjust the cell count may inhibit for evaluation of seemingly abnormal muco-cuta- aggregation responses, and it may be preferable to neous bleeding (14). simply use unadjusted PRP; however, this is not yet Racial differences may also be an important con- considered to be standard practice. The variability in sideration in the work-up of bleeding disorders. For clinical laboratory practice in the execution of platelet example, it has been demonstrated that among aggregometry (concentration of agonists, use of re- African-American women presenting with menorrha- ference ranges, mode of reporting, etc) is highlighted gia, von Willebrand’s disease is less common, whereas by a recent survey that included 47 US coagulation intrinsic platelet function defects are more common. laboratories. Some of the variability is explained by a The approach outlined by this review is recom- lack of standardized protocols. However, even when mended primarily for outpatients referred for evalua- provided with a standardized methodology, wide tion. Even more complex is the evaluation of variability still exists among platelet function studies hospitalized in-patients, who are generally referred performed at different sites. It is hoped that the pre- because of post-surgical bleeding. As already men- analytical and assay variables in platelet aggregrome- tioned, it may be difficult or even impossible to try can be better understood and standardized in the establish a diagnosis of vWD in a stressed individual future as a result of recent re-evaluation. in the post-operative period. Once again, a careful Standard platelet aggregrometry is not sensitive to history will provide information on whether the secretion or storage defects, and indeed in classic patient has experienced spontaneous or excessive delta storage pool deficiency (d-SPD), it is well post-surgery/trauma bleeding in the past, or has a recognized that platelet aggregation may be normal. positive family history of such. Understanding Therefore it is recommended that lumi-aggregometry whether bleeding was uni-focal or multi-focal may (which evaluates for simultaneous ATP release in suggest a surgical or coagulopathic cause, respectively. response to agonists such as thrombin and ADP) Other common acquired causes of bleeding in hospi- should be performed when possible. ATP release by talized patients, such as vitamin K deficiency, dilu- 32 N. S. Key tional coagulopathy, or DIC should be considered. must be implemented. Reminders to the patient to However, ultimately, a laboratory evaluation may be avoid all platelet inhibitory agents prior to the indicated; in order to simplify test interpretation, it is procedure, and to the surgeon to take extra care in recommended that this be postponed until the patient securing intra-operative surgical hemostasis are has been discharged if at all possible. worthwhile.

Management of the patient with a positive The future bleeding history and an unremarkable laboratory evaluation Studies that aim to better standardize the clinical history and existing laboratory studies in the diagnosis Even after an extensive tiered laboratory investigation of bleeding disorders will be of particular value. It is of a subject strongly suspected of having a bleeding clear however that assays of platelet function with disorder, no specific abnormality may be discovered. greater sensitivity and specificity are required. To this This outcome may be the case in 40% or more of end, proteomics approaches that evaluate the range of patients believed to have a bleeding disorder on the proteins expressed by normal and abnormal platelets basis of the history. This dilemma can present a are a realistic expectation in the next decade. Simi- therapeutic challenge, since many of these individuals larly, it is reasonable to expect that the pharmaceutical are evaluated prior to elective surgery. The literature armamentarium of hemostatic agents, which is cur- contains very little information on the management or rently quite limited, will be expanded to cope with the outcomes in these subjects. Thus, a careful assessment of the risk/benefit and cost/benefit ratios of shifting paradigm demanding better diagnosis of the empirical use of anti-fibrinolytic agents, DDAVP, and/ causes of bleeding, and avoidance of a simplistic or platelet transfusions, or even recombinant FVIIa, transfusion approach when bleeding does occur. Hematology, 2005; 10 Supplement 1: 33/38

HYPERCOAGULABLE STATES

Antiphospholipid antibody syndrome

REYHAN DIZ-KU¨ C¸ U¨ KKAYA

Department of Internal Medicine, Division of Hematology, Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkey

Antiphospholipid syndrome (APS) is defined as ondary’ when it is associated with other autoimmune recurrent arterial and/or venous thrombosis and disorders especially with systemic lupus erythemato- obstetric complications in the presence of antipho- sus [9]. Besides these autoimmune conditions, aPLA spholipid antibodies (aPLA) [1]. APS may affect any may be present in healthy individuals, in patients with system and organ in the body including heart, brain, hematologic and solid malignancies, in patients with kidney, skin, lung, and placenta (Table I). This certain infections (syphilis, leprosy, HIV, CMV, EBV, syndrome is predominant in females (female to male etc), and in patients being treated with some drugs ratio is 5 to 1), especially during the childbearing (phenothiazines, procainamide, phenytoine, etc). years [2]. It is the most common acquired thrombo- Those antibodies are defined as ‘alloimmune aPLA’, philic disorder in the general population, affects both they are generally transient and not associated with arterial and venous vessels in any diameter [2/7]. the clinical findings of APS [10].

Classification criteria of APS Anti-phospolipid antibodies (aPLA) Preliminary classification criteria for the classification aPLA are heterogenous antibodies directed against of ‘definite’ APS were described at an international phospholipid/protein complexes. Although several meeting at Sapporo, Japan [1]. Clinical criteria aPLA are defined, in two of those (anti-cardiolipin include vascular thrombosis (one or more episode of antibodies (ACLA) and lupus anticoagulant (LA)) arterial, venous, or small vessel thrombosis which clinical studies confirmed an association with the should be documented by radiologically or histologi- clinical complications of APS. Anticardiolipin antibo- cally) and pregnancy morbidity (a). three or more dies (aCLA) are measured by ELISA, and reported by unexplained abortions before 10th week of gestation, GPL ad MPL units for IgG and IgM anti-cardiolipin (b). one or more unexplained fetal death beyond the antibodies, respectively [11]. In patients with APS, 10th week of gestation, (c). one or more premature aCLA are not simply directed to cardiolipin but it births because of pre-eclampsia, eclampsia, or pla- recognizes B2-GPI /cardiolipin complexes in the cental insufficiency. Laboratory criteria are the detec- ELISA microplates. Since low titer aCLA may be tion of lupus anticoagulants, and presence of present in normal population, moderate (20/80 GPL, anticardiolipin IgG and IgM antibodies in medium 20/50 MPL) to high positive (more than 80 GPL or 50 or high titers. Laboratory criteria should be met at two MPL) results are needed for the diagnosis of APS [12]. or more occasions at least 6 weeks apart. According to Lupus anticoagulants are screened by phospholipid- these criteria, ‘definite’ APS is considered if a patient dependent coagulation tests (e.g. activated partial had at least one clinical and one laboratory criteria thromboplastin time, kaolin clotting time, Russell’s [1]. Validation studies showed that Sapporo criteria viper venom time, dilute prothrombin time, or textarin have 71% sensitivity and 98% specificity to diagnose time). A prolonged screening test should not be ‘definite’ APS patients [8]. corrected by mixing of normal plasma (mixing stu- As in the other autoimmune disorders, aPLA and dies), but should be corrected by addition of phospho- APS may accompany the other autoimmune diseases lipid (phospholipid neutralization procedure). For the (most frequently SLE) and certain situations. APS is diagnosis of LA, other coagulopathies including spe- referred as ‘primary’ when it occurs alone, or ‘sec- cific factor inhibitors should be excluded [13].

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389845 34 R. Diz-Ku¨c¸u¨kkaya Table I. Clinical manifestations of the APS

System Findings

Cardiovascular Acute coronary syndromes, angina pectoris, cardiac valve involvement, non-bacterial thrombotic endocarditis, System atherosclerosis, peripheral arterial disease, miyocarditis, claudicatio intermittens Arterial system Thrombosis in the large, medium or small arteries Venous system Thrombosis in the large, medium or small veins Central Nervous Transient ischemic attack, cerebral arterial occlusion, chorea, convulsions, dementia, transverse myelitis System encephalopathy, migrain, Sneddon syndrome, pseudo-tumor cerebri, cerebral venous thrombosis, mononeuritis multiplex Obstetrical Fetal loss, pre-eclampsia, eclampsia, HELLP syndrome, fetal growth retardation, utero-placental insufficiency, infertility Hematological Thrombocytopenia, autoimmune hemolytic anemia, thrombotic thrombocytopenic purpura, hemolytic uremic syndrome Dermatological Livedo reticularis, Raynaud’s phenomenon, leg ulcers, purpura, acrocyanosis, cutanous infarcts, digital gangrenes, Degos’s disease, anetoderma Ophtalmological Retinal arterial and venous thrombosis, amaurosis fugax, transient or persistent visual loss, photophobia Gastrointestinal Budd-Chiari’s syndrome, mesentery embolism, nodular regenerative hyperplasia of the liver, intestinal vaso-occlusive disease, ischemic colitis, pancreatitis, hepatic or splenic infarct Pulmonary Pulmonary embolism, pulmonary hypertension, alveolary hemorrhage, fibrosing alveolitis, respiratory disstres syndrome Renal Renal vein thrombosis, renal arterial thrombosis, acute or chronic renal failure, hypertension, hematuria, nephrotic syndrome Pscychiatric Psychosis, cognitive dysfunction

B2-GPI is a plasma protein with high affinity for Autoimmune mechanisms play a role in the gen- negatively charged phospholipids. It has been showed eration of aPLA, but the exact pathophysiologic that most of the aCLA antibodies are B2 GPI- mechanisms causing thrombotic complications of dependent [14,15]. Although the precise function is APS have not been clarified. Endothelial cell activa- unknown, B2-GPI might neutralize negatively charged tion, inhibition of protein C and protein S activity, phospholipids and inhibits coagulation cascade and the activation of platelets, increased tissue factor expres- presence of anti-B2 GPI antibodies may predispose to sion, and impairment of fibrinolytic activity have all thrombosis. Anti-prothrombin antibodies may present been proposed as possible mechanisms by which the in 50/90% of APS patients [4/6]. Anti-prothrombin aPLA may predispose to thrombosis in patients with antibodies may cause hypoprothrombinemia and may APS [2,3,6]. be associated with bleeding symptoms in patients with In the last decade, more studies have been focused APS. Although it has been showed that both anti- on interaction between aPLA and endothelial cells B2GPI and anti-prothrombin antibodies might be (EC). It has been shown that binding of aPLA associated with clinical features of APS, most of these activates EC, which results in the expression of studies are retrospective, and these antibodies are not P-selectin, intercellular cell adhesion molecule-1 included in the Sapporo criteria [16]. (ICAM-1), and vascular cell adhesion molecule-1,

Although other aPLA (anti-phosphatidylserine, and induces thrombosis in animal models [17/20]. anti-phosphatidylinositol, anti-phosphatidylglycerol, Blank et al. found that synthetic peptides that antibodies directed to zwitterionic phospholipids, neutralize aPLA function inhibit the increase of these etc) may be detected in patients with APS, the clinical adhesion molecules on the surface of EC, inhibit importance of these aPLA are unclear in the absence monocyte adhesion, and prevent the development of of a positive LA and/or aCLA test [1]. experimental APS in mice [17]. Pierangeli et al. showed that aPLA-induced leukocyte adhesion was completely abrogated in ICAM-1 and P-selectin- Thrombosis in the APS deficient (ICAM-1//P-selectin/) mice [19]. Thrombotic complications are the major causes of Combes et al. showed that in vitro generation of morbidity and mortality in patients with APS. endothelial microparticles was increased in patients Although the thrombosis may occur in any site of with lupus anticoagulants [21]. Besides these in vitro the vascular tree; about 2/3 of the thrombotic events studies, we have recently showed that endothelial are venous, mainly deep and superficial veins in the functions determined by brachial artery ultrasound lower extremity; remaining 1/3 is seen in the arterial were impaired in patients with primary APS [22]. system. In patients with APS, the risk of the throm- On the other hand, some researchers have focused botic recurrence is high compared to patients without on the interaction between aPLA and monocytes. APS especially after the discontinuation of oral antic- They have suggested that aPLA increase procoagulant oagulant therapy [2/6]. activity of normal donor monocytes, and the mono- Antiphospholipid antibody syndrome 35 cytes that circulate in patients with APS are in a ial and venous thrombosis, convulsions, and encepha- ‘primed’ state to express tissue factor [23]. lopathy (60%), cardiac manifestations such as valve Since APS is a clinically heterogenous disorder with involvements and myocardial infarctions (53%) are a broad spectrum of presentations, identification of the major clinical findings. Usually severe thrombo- the patients who have high risk of thrombosis is cytopena is present (60%). Hemolysis and dissemi- crucial. Although many epidemiological studies nated intravascular coagulation may occur. showed an increased risk of thrombosis in patients Recommendations for the treatment of CAPS are positive for aPLA, some patients with high titers of intravenous heparin, plasmapheresis, and steroids. aPLA did not develop thrombosis, even in long-term Treatment of CAPS should include the precipitating follow-up. These facts supported the hypothesis that conditions such as the treatment of infections with there are additional inherited or acquired thrombo- appropriate antibiotics, and the treatment of SLE genic factors, which influence the development of flares. The mortality rate is quite high (more than thrombosis in those patients (‘double or multiple hit’ 50%) even in presumably properly treated patients hypothesis) [2]. [35,36]. In the last decades, several genetic factors have been identified in patients with thrombosis, especially with Obstetric complications of APS venous thromboembolism [24]. Natural anticoagulant (protein C, protein S, antithrombin) deficiencies, The association between the aPLA and fetal loss has factor V Leiden A506G mutation and prothrombin been recognised for a long time. Besides the recurrent A20210G mutation were the most common causes of fetal losses, it is now clear that aPLA may cause pre- hereditary thrombophilia. It has been shown that eclampsia, eclampsia, fetal growth retardation, utero- natural anticoagulant deficiencies are quite rare in placental insufficiency, preterm birth, and reproduc- patients with APS. Several studies investigating the tive failure [1,2,8]. role of factor V Leiden A506G mutation and prothrombin A20210G mutation in patients with APS, Thrombocytopenia in APS gave conflicting results [25/33]. It is unclear whether the presence of these mutations increased the throm- Thrombocytopenia is reported in about 20/40% of botic risk of APS patients. patients with APS, is usually mild (70.000/120.000/ Factor XIII Val34Leu polymorphism is a newly mm3), and does not require any clinical intervention. described polymorphism which is located in the three Severe thrombocytopenia (lower than 50.000/mm3)is amino acids away from the thrombin activation site of seen only in 5/10% of the patients [37/39]. Although the factor XIII- A subunit. It has been reported that thrombocytopenia has been defined as a clinical the presence of Leu allele may decrease both arterial criteria in the first classification of APS [40], it was and venous thrombosis risk, and increase bleeding not included in the preliminary classification of tendency. In a cohort of 60 APS patients with definite APS recently proposed in Sapporo [1]. The thrombosis, 22 aPLA-positive patients without patients who had APLA and thrombocytopenia as the thrombosis and 126 healthy controls, we could not only clinical manifestation in the absence of other find any effect of factor XIII Val34Leu polymorphism APS findings were defined as ‘probable’ or ‘possible’ on the thrombosis in patients with APS [34]. APS. In a recent prospective study, however, it has been shown that the ITP patients who presented with thrombocytopenia and had positive tests for APLA Catastrophic APS (CAPS) had increased risk of thrombosis. It has been proposed A minority of APS patients may acutely present with that measurement of APLA, especially LA in patients multiple simultaneous vascular occlusions affecting with initial diagnosis of ITP may identify a subgroup small vessels predominantly, and is termed as ‘cata- of patients with higher risk of developing APS [41]. strophic APS (CAPS)’. Precipitating factors have The pathogenesis of thrombocytopenia in the APS been identified in the majority of the CAPS patients, is not clear. Although there is direct evidence that such as infections, surgery, pregnancy, SLE flares, aPLA may bind platelet membranes and cause plate- withdrawal of anticoagulation therapy, and trauma. let destruction, the relation between aPLA positivity CAPS definition requires thrombotic involvement of and thrombocytopenia is still unclear. Some investi- at least three different organ systems over a period of gators suggest that specific antiplatelet glycoprotein days or weeks. Clinical manifestations of CAPS are (GP)-antibodies rather than aPLA cause thrombocy- related to the extent of organ involvement and topenia in patients with APS, and anti-GP antibodies cytokine release of affected tissues. Renal dysfunction are rare in patients with APS with normal platelet (70%), pulmonary involvement such as ARDS and counts [42,43]. In one study, antibodies directed pulmonary embolism (66%), skin involvement such against the GPIIb-IIIa or GPIb-V-IX complexes as skin necrosis and livedo reticularis (66%), central were found in about 40% of the patients with APS nervous system manifestations such as cerebral arter- who had thrombocytopenia [44]. Anti-platelet GP 36 R. Diz-Ku¨c¸u¨kkaya antibodies in thrombocytopenic patients with APS do a. aPLA-positive individuals with no APS symp- not cross-react with antibodies against phospholipids toms or findings: It is known that aPLA may be or beta2 GP-I [45]. Immunosuppressive treatment of present in 1/5% of healthy individuals [57]. If thrombocytopenia in those patients increases the aPLA-positive individuals have no history or platelet count and reduces the titers of anti-GP findings of APS, treatment should not be con- antibodies, but not the titers of aPLA [46]. These sidered [3,4,58,59]. Although some investigators data may suggest that thrombocytopenia is a second- recommend prophylactic therapy for aPLA-po- ary immune phenomenon that may develop at the sitive individuals if they face to acquired throm- same time with APS. On the other hand, Fabris et al. botic risk (puerperal period, surgery, showed that platelet antigens in thrombocytopenic immobilisation etc.) [7,58], there are no pro- patients with APS were different from those in ITP, spective or controlled studies investigating the and surface glycoproteins were not involved. He also effectiveness of anti-thrombotic prophylaxis in those individuals. found that a 50/70 kDa internal platelet protein had been specifically found in patients with APS and b. Primary prophylaxis of thrombosis in aPLA- thrombocytopenia but not in patients with ITP [47]. positive patients: Although many experts recom- Another issue is the clinical importance of throm- mend to use anti-thrombotic prophylaxis in bocytopenia in the APS. When the APS patients were aPLA-positive patients who fullfill Sapporo divided into three groups according to platelet counts APS criteria and have no history of thrombosis, there are only few studies addressing this issue. as non-thrombocytopenic, moderately (50.000/ 100.000/mm3), and severe thrombocytopenic (below The diffucult point is to define thrombotic risk in 50.000/mm3), the rate of the development of throm- aPLA-positive patients who had no thrombosis. bosis were found as 40%, 32%, and 9%, respectively Erkan et al. showed that APS patients with fetal losses were also at high risk of thrombosis [60], [38]. This data shows that moderate thrombocytope- and they recommended prophylactic aspirin nia does not prevent the development of thrombosis therapy, 325 mg/day. Petri et al. reported that in patients with APS, and anti-thrombotic prophylaxis hydroxychloroquine might have a protective should be considered in those patients [38,39,41]. effect against thrombosis in secondary APS Although thrombocytopenia is a common finding in (SLE-APS) patients [61]. patients with APS, bleeding complications are very c. Treatment of venous thrombosis in APS: Treat- rare, even in severe thrombocytopenia. The presence ment of the first venous thrombotic attack is of bleeding symptoms in an APS patient with mod- similar to those with idiopathic venous throm- erate thrombocytopenia nececcitates a search for the bosis. Heparin and oral anticoagulation therapy presence of anti-prothrombin antibodies [48], and are routinely used for this purpose. Recurrency other diseases which may affect hemostasis such as of thrombosis is higher in patients with APS DIC, uremia etc. Severe thrombocyopenia may re- compared to patients with no aPLA and venous quire therapy. Treatment strategies in those patients thrombosis [62], especially in the first 6 months are similar in those with ITP. Glucocorticoids are of thrombosis and after the cessation of the effective in only 15% of the patients [37]. IVIg and therapy [62/64]. However, the duration and immunosupressive drugs such as cyclophosphamide the intensity of the treatment are not clear. In may be used in patients who have severe bleeding the first studies, it has been suggested that high symptoms and have ‘catastrophic’ APS. Splenectomy intensity (INR/3) oral anticoagulant therapy produces sustained remission in approximately two- should be used in APS [62,64]. Recent studies thirds of the patients [49 /51]. Preoperative vaccina- however, showed that high intensity oral antic- tion procedure is the same as ITP. Antithrombotic oagulant therapy is not superior to conventional prophylaxis should be planned to prevent postopera- dose oral anticoagulant therapy (INR 2/3) for tive thrombosis in those patients. the prevention of thrombosis in those patients

Interestingly there are a few case reports describing [2,3,7,63/66,68]. Ames et al. also showed that the correction of thrombocytopenia with aspirin high intensity oral anticoagulant therapy might [52,53], warfarin [54,55], and anti-malarial drugs increase bleeding complications in patients with [56]. It has been suggested that inhibition of platelet APS [68]. Coexistence of thrombocytopenia is activation, aggregation, and platelet consumption may the major cause for bleeding. The duration of help to increase platelet count in those patients. oral anticoagulant therapy is an another debated issue. Although life-long therapy was recommended in the first studies; recent prospective Management of thrombotic complications in studies yield no firm conclusions. APS d. Treatment of arterial thrombosis in APS: Acute Management of patients with aPLA mainly depends coronary syndromes, transient ischemic attacks, on the presence of clinical symptoms and findings: and cerebral arterial thrombosis are the most Antiphospholipid antibody syndrome 37

common causes of arterial involvement in pa- [9] Asherson RA. A ‘primary’ antiphospholipid syndrome? J tients with APS. Morbidity and mortality are Rheumatol 1998;15:/ 1742/ /46. [10] Nimmo MC, Carter CJ. The antihospholipid antibody syn- high in arterial involvement. In APS patients drome: a riddle wrapped in a mystery inside an enigma. Clin with acute coronary syndromes, high intensity Applied Immunol Rev 2003;4:/ 125/ /140.

oral anticoagulant therapy is recommended. The [11] Harris EN. Antiphospholipid antibodies. Br J Haematol 1990;/ effectiveness of additional aspirin therapy is 74:1/ /9. debated [7]. In APASS (The Antiphospholipid [12] McNeil HP, Chesterman CN, Kirilis SA. Immunology and clinical importance of antiphospholipid antibodies. Adv Im- Antibody in Stroke Study Group) study [68],

munol 1991;49:/ 193./ APS patients with stroke were randomized to [13] Brandt JT, Triplett DA, Alving B, Scharer I. Criteria for the either aspirin (325 mg daily) or warfarin (tar- diagnosis of lupus anticoagulants: An update. Thromb Hae- geted INR 2.2), and were compared for the risk most 1995;74:/ 1185/ /1187. of recurrent stroke. APASS study found no [14] McNeil P, Simpson R, Chesterman C, Kirilis S. Antipho- significant difference between two arms, spholipid antibodies are directed against a complex antigen that includes a lipid-binding inhibitor of coagulation: b2-GPI although the study population was older than (apolipoprotein H). Proc Natl Acad Sci USA 1990;87:/ 4120/ / the APS patients in most of the cohorts. It is 4124. important to determine other disorders to decide [15] Galli M, Comfurius P, Maassen C. Anticardiolipin antibodies for the treatment modalities in APS patients with (ACA) directed not to cardiolipin but a plasma protein stroke. In stroke patients who have atrial fibrilla- cofactor. Lancet 1990;335:/ 1544/ /1545. [16] Galli M. Should we include anti-prothrombin antibodies in tion and cardiac valve disorders are recom- the screening for the antiphospholipid syndrome? J Autoim- mended to use warfarin therapy with moderate mun 2000;15:/ 101/ /105. to high intensity. [17] Blank M, Shoenfeld Y, Cabilly S, Heldman Y, Fridkin M, Katchalski-Katzir E. Prevention of experimental antiphospholipid syndrome and endothelial cell activation by synthetic Conclusion peptides. Proc Natl Acad Sci USA 1999;96:/ 5164/ /68. [18] Pierangeli S, Colden-Stanﬁeld M, Liu X, Barker JH, Ander- APS is an acquired autoimmune disorder with son GL, Harris N. Antiphospholipid antibodies from antipho- thrombotic tendency, obstetric complications and spholipid syndrome patients activate endothelial cells in vitro multi-systemic symptoms. The identification of the and in vivo. Circulation 1999;99:/ 1997/ /2002. APS is extremely important since thrombotic compli- [19] Pierangeli SS, Espinola RG, Liu X, Harris EN. Thrombogenic cations may cause severe morbidity and mortality. effects of antiphospholipid antibodies are mediated by intercellular cell adhesion molecule-1, vascular cell adhesion Although many studies showed the association be- molecule-1, and P-selectin. Circ Res 2001;88:/ 245/ /50. tween aPLA and clinical findings of APS; exact [20] Meroni PL, Raschi E, Camera M, et al. Endothelial activation pathophysiologic mechanisms are still unclear. The by aPL: a potential pathogenetic mechanism for the clinical elucidation of pathogenetic mechanism of APS may manifestation of the syndrome. J Autoimmun 2000;15:/ 237/ / help to identify patients who are at high risk for 40. [21] Combes V, Simon A, Grau G, et al. In vitro generation of thrombosis, and may improve management of the endothelial microparticles and possible prothrombotic activity

patients. in patients with lupus anticoagulant. J Clin Invest 1999;104:/ /

93/102. [22] Mercanoglu F, Erdogan D, Oflaz H, et al. Impaired branchial endothelial function in patients with primary anti-phospho- References lipid syndrome. Int J Clin Pract 2004;58:/ 1003/ /7. [1] Wilson WA, Gharavi AE, Koike T, Lockshin MD, Branch DW, [23] Williams FMK, Jurd K, Hughes GRV, Hunt BJ. Antipho- Piette JC, Brey R, Derksen R, Harris EN, Hughes GR, et al. spholipid syndrome patients’ monocytes are ‘primed’ to International consensus statement on preliminary classifica- express tissue factor. Thromb Haemost 1998;80:/ 864/ /5. tion criteria for definite antiphospholipid syndrome. Arthritis [24] Rosendaal FR. Venous thrombosis: a multicausal disease. Lancet 1999;353:/ 1167/ /1173. Rheum 1999;42:/ 1309/ /1311. [2] Levine JS, Branch DW, Rauch J. The antiphospholipid [25] Ames PRJ, Tommasino C, D’Andrea G, et al. Thrombophilic syndrome. N Eng J Med 2002;346:/ 752/ /763. genotypes in subjects with idiophatic antiphospholipid anti-

[3] Hanly JG. Antiphospholipid syndrome: an overview. CMAJ bodies. Prevalance and signiﬁcance. Thromb Haemost 1998;/

/ 2003;168:/ 1675/ /1982. 79:46. [4] Greaves M. Antiphospholipid antibodies and thrombosis. [26] Montaruli R, Borchiellini A, Tamponi G, et al. Factor V 506 0/ Lancet 1999;353:/ 1348/ /1353. Arg Gln mutation in patients with antiphospholipid

[5] Goodnight SH. Antiphospholipid antibodies and thrombosis. antibodies. Lupus 1996;5:/ 303./ Curr Op Hematol 1994;1:/ 354/ /361. [27] Hansen KE, Kong DF, Moore KD, Ortel TL. Risk factors [6] Santoro SA. Antiphospholipid antibodies and thrombotic associated with thrombosis in patients with antiphospholipid

predisposition: Underlying pathogenetic mechanisms. Blood antibodies. J Rheumatol 2001;28:/ 2018./ 1994;83:/ 2389/ /2391. [28] Chopra N, Koren S, Greer WL, et al. Factor V Leiden, [7] Greaves M, Cohen H, Machin SJ, Mackie I. Guidelines on the prothrombin gene mutation, and thrombosis risk in patients

investigation and management of the antiphospholipid syn- with antiphospholipid antibodies. J Rheumatol 2002;29:/ / drome. Br J Haematol 2000;109:/ 704/ /715. 1683 /1688. [8] Lockshin MD, Sammaritano LR, Schwartzman S. Validation [29] Galli M, Finazzi G, Duca F, Norbis F, Moia M. The G1691-A of the Sapporo criteria for antiphospholipid syndrome. Ar- mutation of factor V, not the G20210-A mutation of factor II thritis Rheum 2000;43:/ 440/ /443. or the C677-T mutation of methylenetetrahydrofolate reduc- 38 R. Diz-Ku¨c¸u¨kkaya

tase genes, is associated with venous thrombosis in patients [48] Bernini JC, Buchanan GR, Ashcraft J. Correction of hypopro- with venous thrombosis in patients with lupus anticoagulants. thrombinemia and severe hemorrhage associated with a lupus Br J Haematol 2000;108:/ 865/ /870. anticoagulant. J Pediatr 1993;123:/ 937./ [30] Forasterio R, Martinuzzo M, Adamczuk Y, Varela MLI, [49] Stasi R, Stipa E, Masi M, et al. Long-term observation of 208 Pombo GV, Carreras LO. The combination of thrombophilic adults with chronic thrombocytopenic purpura. Am J Med

genotypes is associated with deﬁnite antiphospholipid syn- 1995;98:/ 436./

drome. Haematologica 2001;86:/ 735./ [50] Font J, Jimenez S, Cervera R, et al. Splenectomy for refractory [31] Alarcon-Segovia D, Ruiz-Arguelles GJ, Garces-Eisele J, Ruiz- Evans’ syndrome associated with antiphospholipid antibodies:

Arguelles A. Inherited activated protein C resistance in a report of two cases. Ann Rheum Dis 2000;59:/ 920./ patient with familial primary antiphospholipid syndrome. J [51] Hakim AJ, Machin SJ, Isenberg DA. Autoimmune thrombo-

Rheumatol 1996;23:/ 2162./ cytopenia in primary antiphospholipid syndrome and systemic [32] Diz-Kucukkaya R, Demir K, Yenerel MN, Nalc¸aci M, lupus erythematosus: The response to splenectomy. Semin ˙ Kaymakog˘lu S, .Inanc¸ M. Coexistence of homozygous factor Arthritis Rheum 1998;28:/ 20./ V Leiden mutation and antiphospholipid antibodies in two [52] Alarcon-Segovia D, Sanchez-Guerrero J. Correction of throm- patients presented with Budd-Chiari syndrome. Haematologia bocytopenia with small dose aspirin in the primary antipho-

2002;32:/ 2061./ spholipid syndrome. J Rheumatol 1989;16:/ 1359./ [33] Dizon-Tawson D, Hutchinson C, Silver R, Branch W. The [53] Alliot C, Messouak D, Albert F. Correction of thrombocyto- factor V Leiden mutation which predispose to thrombosis is penia with aspirin in the primary antiphospholipid syndrome.

not common in patients with antiphospholipid syndrome. Am J Hematol 2001;68:/ 215./

Thromb Haemost 1995;74:/ 1029./ [54] Wisbey HL, Klestov AC. Thrombocytopenia corrected by ˙ [34] Diz-Kucukkaya R, Hançer VS, .Inanç M, Nalçaci M, Pekçelen warfarin in antiphospholipid syndrome. J Rheumatol 1996;23:/ / Y. Factor XIII Val34Leu polymorphism does not contibute to 769. the prevention of thrombotic complications in patients with [55] Ames PRJ, Orefice G, Brancaccio V. Reversal of thrombocy- antiphospholipid syndrome. Lupus 2004;13:/ 32/ /35. topenia following oral anticoagulation in two patients with

[35] Asherson RA. The catastrophic antiphospholipid syndrome. J primary antiphospholipid syndrome. Lupus 1995;4:/ 491./

Rheumatol 1992;19:/ 508./ [56] Suarez MI, Diaz RA, Aguayo-Canela D, Pujol de la Llave. [36] Asherson RA, Cervera R, Piette JC, et al. Catastrophic Correction of severe thrombocytopenia with chloroquine in

antiphospholipid syndrome :Clues to pathogenesis from series the primary antiphospholipid syndrome. Lupus 1996;5:/ 81./

of 80 patients. Medicine (Baltimore) 2001;80:/ 355./ [57] Petri M. Epidemiology of antiphospholipid antibody syn- [37] Galli M, Finazzi G, Barbui T. Thrombocytopenia in the drome. J Autoimmun 2000;15:/ 145/ /151. antiphospholipid syndrome. Br J Med 1996;93:/ 1/ /6. [58] Gezer S. Antipholipid syndrome. Dis Mon 2003;49:/ 696/ /74. [38] Italian Registry of Antiphospholipid Antibodies (IR-APA). [59] Lockshin MD, Erkan D. Treatment of the antiphospholipid Thrombosis and thrombocytopenia in antiphospholipid syn- syndrome. N Eng J Med 2003;349:/ 1177/ /1179. drome (idiopathic and secondary to systemic lupus erythema- [60] Erkan D, Merrill JT, Yazici Y, Sammaritano L, Buyon JP, tosus): First report from the Italian registry. Haematologica Lockshin MD. High risk of thrombosis rate after fetal loss in

1993;78:/ 313./ antiphospholipid syndrome: effective prophylaxis with aspirin. [39] Cuadrado MJ, Mujic F, Munoz E, Khamashta MA, Hughes Arthritis Rheum 2001;44:/ 1466/ /1467. GRV. Thrombocytopenia in the antiphospholipid syndrome. [61] Petri M. Hydroxychloroquine use in the Baltimore Lupus

Ann Rheum Dis 1997;56:/ 194./ Cohort. Effects on lipids, glucose, and thrombosis. Lupus [40] Harris EN. Annotation: antiphospholipid antibodies. Br J 1996;5(supp/ 1):S16/ /22.

Haematol 1990;74:/ 1./ [62] Khamashta MA, Cuadrado MJ, Mujic F, Taub NA, Hunt BJ, [41] Diz-Kucukkaya R, Hacihanefioglu A, Yenerel M, et al. Hughes GR. The management of thrombosis in the antipho- Antiphospholipid antibodies and antiphospholipid syndrome spholipid-antibody syndrome. N Eng J Med 1995;332:/ 993/ / in patients presenting with immune thrombocytopenic pur- 997. pura: a prospective cohort study. Blood 2001;98:/ 1760/ /1764. [63] Rosove MH, Brewer PM. Antiphospholipid thrombosis: [42] Godeau B, Piette JC, Fromont P, Intrator L, Schaeffer A, clinical course after the first thrombotic event in 70 patients. Bierling A. Specific antiplatelet glycoprotein autoantibodies Ann Intern Med 1992;117:/ 303/ /308. are associated with the thrombocytopenia of primary antipho- [64] Derksen RH, de Groot PG, Kater L, Nieuwenhuis HK.

spholipid syndrome. Br J Haematol 1997;98:/ 873./ Patients with antiphospholipid antibodies and venous throm- [43] Macchi L, Rispal P, Clofent-Sanchez G, et al. Antiplatelet bosis should receive long term anticoagulant treatment. Ann antibodies in patients with systemic lupus erythematosus and Rheum Dis 1993;52:/ 689/ /692. the primary antiphospholipid antibody syndrome: their rela- [65] Prandoni P, Simioni P, Girolami A. Antiphospholipid anti- tionship with the observed thrombocytopenia. Br J Haematol bodies, recurrent thromboembolism, and intensity of warfarin

1997;98:/ 336./ coagulation. Thromb Haemost 1996;75:/ 859./ [44] Galli M, Daldossi M, Barbui T. Anti-glycoprotein Ib/IX and [66] Schulman S, Svenungsson E, Granqvist S. Anticardiolipin IIb/IIIa antibodies in patients with antiphospholipid antibo- antibodies predict early recurrence of thromboembolism and

dies. Thromb Haemost 1994;71:/ 571./ death among patients with venous thromboembolism follow- [45] Lipp E, von Felten A, Sax H, Muller D, Berchtold P. ing anticoagulant therapy. Duration of Anticoagulation Study Antibodies against platelet glycoproteins and antiphospholipid Group. Am J Med 1998;104:/ 332/ /338. antibodies in autoimmune thrombocytopenia. Eur J Haematol [67] Ames PR, Ciampa A, Margaglione M, Scenna G, Iannaccone

1998;60:/ 283./ L, Brancaccio V. Bleeding and re-thrombosis in primary [46] Stasi R, Stipa E, Oliva F, et al. Prevalance and clinical antiphospholipid syndrome on oral anticoagulation: an 8- signiﬁcance of elevated antiphospholipid antibodies in patients year longitudinal comparison with mitral valve replacement

with idiopathic thrombocytopenic purpura. Blood 1994;84:/ / and inherited thrombophilia. Thromb Haemost 2005;93:/ /

4203. 694/9. [47] Fabris F, Steffan A, Cordiano I, et al. Speciﬁc antiplatelet [68] The WARSS, APASS, PICSS, HAS and GENESIS study autoantibodies in patients with antiphospholipid antibodies groups. The feasibility of a collaborative double-blind study and thrombocytopenia. Eur J Haematol 1994;53:/ 232./ using an anticoagulant. Cerebrovasc Dis 1997;7:100 /112. Hematology, 2005; 10 Supplement 1: 39

HYPERCOAGULABLE STATES

Hypercoagulable states

KENNETH BAUER

Harvard Medical School

Abstract The incidence of venous thromboembolism (VTE) in the USA is approximately 1 per 1000 with over 200,000 new cases per year. Despite improved methods for antithrombotic prophylaxis, the incidence of VTE has remained constant for the last 20 years and about 30% of patients developing VTE die within 30 days. Risk factors for VTE include advancing age, obesity, recent surgery or trauma, hospital or nursing home confinement, malignancy, and stroke; among women, pregnancy, oral contraceptives, and hormone replacement therapy play a significant role. Prior to 1993, a cause of thrombophilia was detectable in a relatively small percentage of patients presenting with VTE. Hereditary abnormalities were found in only 5 to 15% of patients and were confined to deficiencies of antithrombin, protein C, and protein S. Discovery of the factor V Leiden and prothrombin G20210A mutations has greatly increased the percentage of Caucasian patients in whom venous thrombosis can, in part, be attributed to hereditary thrombophilia. It has also been demonstrated that elevated plasma homocysteine levels constitute a risk factor for venous as well as arterial thrombosis. These abnormalities, along with the presence of markers of the antiphospholipid antibody syndrome can be identified in a substantial percentage of patients presenting with a first episode of idiopathic venous thromboembolism (i.e. in the absence of a triggering risk factor such as surgery, immobilization, or active malignancy). Testing for the hereditary thrombophilias is warranted in selected subsets of such patients, particularly those with a history of venous thromboembolism in first degree relatives. The selection of patients for testing, the choice of tests to perform, and the timing of the testing are important issues to consider. Routine testing of unselected patients with venous thromboembolism would be warranted if the identification of abnormalities led to an alteration in the type or duration of initial anticoagulant therapy. However with the exception of the antiphospholipid antibody syndrome, the available data do not indicate that the anticoagulation management of the majority of patients with hereditary thrombophilia should be any different from those without identifiable abnormalities. The risk of recurrent venous thromboembolism in patients with an initial episode of venous thromboembolism will be discussed along with the impact of a thrombophilic defect and the likely benefit of prolonged anticoagulation versus the associated bleeding risk associated with prolonged anticoagulation.

Keywords: Hypercoagulability, Thrombophilia, Venous Thromboembolism

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389854 Hematology, 2005; 10 Supplement 1: 40

HYPERCOAGULABLE STATES

Inflammation and coagulation

CHARLES T. ESMON

Oklahoma Medical Research Foundation, 825 N.E. 13th Street, Oklahoma City, Oklahoma 73104, USA

Inflammation is associated with thrombotic predis- sponse. Inflammatory mediators, however, can de- position in patients. The mechanisms by which press these potent natural anticoagulant mechanisms. inflammation contributes to the thrombosis involve Of the natural anticoagulants, the protein C pathway both humoral and cellular changes. Inflammation is one of the systems down regulated by inflammation. impacts all phases of blood coagulation: initiation, Thrombomodulin and the endothelial cell protein C propagation and the inhibition. Inflammatory media- receptor are both required for optimal protein C tors like endotoxin and tumor necrosis factor alpha activation in response to thrombin generation. In induce expression of tissue factor on blood cells, severe inflammatory situations, both proteins are particularly monocytes. Tissue factor then triggers down regulated resulting in a decreased anticoagulant the initiation of coagulation. Normally, negatively response. Furthermore, free protein S levels often charged membrane surfaces are limiting so that even decrease, further impairing the pathway. In immune if some activated coagulation factors are generated, mediated inflammatory disease, the pathway may also they fail to propagate coagulation. C reactive protein, be down regulated. Some anti-phospholipid antibo- an acute phase reactant, levels increase in response to dies severely impair the protein C pathway. Clinically inflammation. The C reactive protein can then induce relevant thrombosis is minimized naturally either by tissue factor formation. Complement membrane attack complex provides a potent stimulus for cells to inhibiting the blood clotting or by rapid lysis of the express negatively charged phospholipid on their out blood clot or both. In addition to shifting the membrane leaflet. Alternatively, exposure to collagen hemostatic balance in favor of clot formation, inflam- in combination with thrombin provides a potent mation elevates plasminogen activator inhibitor levels stimulus to platelets eliciting the formation of micro- resulting in decreased fibrinolytic activity. Procoagu- particles and the subsequent exposure of negatively lant impacts of inflammation are also expressed at the charged phospholipid membrane surfaces that can cellular level. Interleukin 6 can increase platelet propagate coagulation. Even when these two events numbers and their responsiveness to agonists like both occur to augment coagulation, potent natural thrombin. All of these events tend to shift the anticoagulant mechanisms limit the thrombotic re- hemostatic balance in favor of clot formation.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389863 Hematology, 2005; 10 Supplement 1: 41/46

PLATELETS

Giant platelet syndrome

H. SAITO, T. MATSUSHITA, K. YAMAMOTO, T. KOJIMA, & S. KUNISHIMA

Nagoya Medical Center and Nagoya University School of Medicine, Nagoya, Japan

Introduction giant platelet syndromes according to the possible underlying cause: abnormalities in the platelet cytos- Platelets are small, disk-shaped, anuclear cells with a keleton, GPIb/IX/V, and transcription factors. Clin- mean diameter of 2 to 3 mm. They are derived from ical and laboratory features as well as responsible cytoplasmic fragmentation of megakaryocytes, genes and chromosomal localizations are also shown. released into the circulation and survived for There are still many inherited disorders in which the 7/10 days. Megakaryocyte development and platelet formation are regulated by thrombopoietin and underlying genetic abnormality has not yet been other cytokines. It is now generally assumed that elucidated. platelets are released by extension and fragmentation of cytoplasmic processes (proplatelet) of megakaryo- Autosomal dominant macrothrombocytopenias cytes through sinus endothelial cells in the bone with leukocyte inclusions (MYH9 disorders) marrow. It is not known, however, how the number and the size of the platelet are controlled at this May-Hegglin anomaly (MHA), first described a cen- step [1]. tury ago, is the prototype of these disorders (Figure 1) Giant platelet syndrome is a group of unique [2]. Sebastian (SBS), Fechtner (FTNS), and Epstein disorders characterized by the presence of abnormally (EPS) syndromes belong here. All four disorders have large platelets, and is usually accompanied by throm- macrothrombocytopenia however, each of these dis- bocytopenia. Thus, it is also called macrothrombocy- orders is distinguished from others by the presence or topenia. Giant platelets are occasionally observed as absence of granulocyte inclusion bodies and the an incidental finding in routine blood smear examina- presence or absence of a variable combination of tions. Most of them are due to acquired disorders Alport manifestations, including nephritis, deafness such as idiopathic thrombocytopenic purpura (ITP) and cataracts (Table II). Although these four disorders and myelodysplastic syndrome (MDS) (Table I). were previously considered to be separate clinical In contrast, inherited giant platelet disorders are entities, a recent positional cloning approach disclosed rare. The mechanisms of giant platelet formation that these disorders are caused by mutations in the and thrombocytopenia are not fully understood same gene, MYH9, which encodes the nonmuscle in both inherited and acquired disorders. It is myosin heavy chain-A (NMMHCA) [3 /5]. Thus, important from a clinical standpoint that congenital they appear to represent the same entity with different disorders are distinguished from acquired disorders, genetic penetrance and variable phenotypic expres- especially ITP, to avoid unnecessary treatments. We sion. The bleeding tendency is usually mild. will discuss the molecular basis, diagnosis, and The diagnosis of macrothrombocytopenia with management of some major inherited giant platelet leukocyte inclusions has been conventionally made syndromes. on the basis of hematomorphological examinations. It is, however, not always easy to detect granulocyte inclusions on Wright stained smear. Immunofluores- Inherited giant platelet syndrome cence analysis of neutrophil NMMHCA localization In recent years we have seen remarkable progress in has revolutionized the diagnosis of MYH9 disorders the molecular understanding of some giant platelet [6]. Abnormal NMMHCA aggregates and accumu- syndromes. The Table I lists some major inherited lates in the neutrophil cytoplasm, and this abnormal

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389881 42 H. Saito et al.

Table I. Characteristics of Giant Platelet Syndromes

Disease Inheritance Gene Chromosome Clinical and laboratory features

Acquired ITP almost normal RBC and WBC MDS anemia, abnormal WBC Inherited Abnormalities in platelet cytoskeleton Autosomal dominant macrothrombo- AD MYH9 22q12-13 Macrothrombocytopenia, granulocyte inclusions cytopenia with leukocyte inclusions/ with/without Alport manifestations (Table II) MYH9 disorders* Abnormalities in GPIb/IX/V Bernard-Soulier syndrome AR GP1BA 17pter-p12 No ristocetin-induced platelet agglutination GP1BB 22q11 GP9 3q21 Mediterranean macrothrombocytopenia/ AD GP1BA 17pter-p12 No bleeding tendency Bernard-Soulier syndrome carrier GP1BB 22q11 Mild thrombocytopenia with normal ristocetin- induced platelet agglutination GP9 3q21 DiGeorge/Velocardiofacial syndrome AD GP1BB 22q11 Contiguous gene syndrome due to chromosome 22q11 microdeletion Parathyroid and thyroid hypoplasia, cardiac abnormalities, cleft palate, mental retardation Abnormalities in transcription factors X-linked macrothrombocytopenia with XL GATA1 Xp11 Dyserythropoiesis with/without b-thalassemia trait dyserythropoiesis Paris-Trousseau thrombocytopenia/ AD FLI1 11q23 Contiguous gene syndrome due to chromosome Jacobsen syndrome 11q23 microdeletion Growth and psychomotor retardation Dysmegakaryopoiesis associated with giant a-granules Unknown cause Gray platelet syndrome AR, AD unknown Gray or colorless platelets due to absent a-granules

*See Table II for details. AD: autosomal dominant, AR: autosomal recessive, XL: X-linked. subcellular localization of NMMHCA is present in of several cytoplasmic spots with circular to oval every neutrophil from individuals with MYH9 muta- shape. Type III or speckled staining is detected in tion. The localization pattern of neutrophil patients with EPS and isolated macrothrombocytope- NMMHCA in MYH9 disorders can be classified nia, in which Wright or May-Gru¨nwald-Giemsa into three groups according to the number, size, and (MGG)-stained inclusion bodies have never been shape of the fluorescence-labeled NMMHCA gran- identified [6]. ules: type I, II and III (Figure 2). In type I, An MYH9 mutation is strictly associated with the NMMHCA forms one or two large and intensely hematological abnormalities. Although the molecular stained cytoplasmic foci. Type II neutrophils consist mechanism of the production of giant platelets has not been elucidated, it is suggested that abnormal NMMHCA, by interfering with the formation of myosin thick-filament, affects proper proplatelet formation in megakaryocytes. Genetic analysis of many kindred with MYH9 disorders revealed that there is no clear relationship between clinical phenotypes and the sites of the MYH9 mutations. It is likely that the MYH9 mutation alone does not cause associated Alport manifestations and that unknown genetic and/or epigenetic factors might influence the phenotypic consequences of MYH9 mutations [7,8]. We have recently generated and analyzed MYH9 knock-out mice [9]. No homozygous mice were born, Figure 1. May-Hegglin anomaly. suggesting that MYH9 expression is required for Giant platelet syndrome 43

Table II. Macrothrombocytopenia with leukocyte inclusions/MYH9 disorders

Phenotype Macrothrombocytopenia Leukocyte inclusions Alport syndrome*

May-Hegglin anomaly (MHA) / / /

Sebastian syndrome (SBS) / / /

Fechtner syndrome (FTNS) / / /

Epstein syndrome (EPS) / / /

*Nephritis, deafness and cataracts. embryonic development. By an antisense oligonucleo- Bernard-Soulier syndrome tide-directed suppression of transcription, NMMHCA Bernard-Soulier syndrome (BSS) is an autosomal is involved in a rearrangement of the actin cytoskeleton recessive bleeding disorder characterized by giant and loss of cell adhesion [10]. MYH9/// embryos, platelets, thrombocytopenia and prolonged bleeding therefore, appeared to fail to develop a normal time (Figure 3), originally described by Bernard and patterned embryo. Such fetal lethality may partly Soulier in 1948 [15]. BSS is caused by quantitative or explain the absence of naturally occurring homozygous qualitative abnormalities in the glycoprotein (GP) Ib/ mutations in human subjects. In contrast, heterozy- IX/V complex, the platelet receptor for von Will- gous mice (MYH9 // /) were viable and fertile ebrand factor [16,17]. As a result of the absence of without gross anatomical, hematological, and nephro- GPIb/IX/V complexes on the platelet membrane, logical abnormalities. Immunofluorescence analysis platelets cannot stick to the damaged blood showed the normal cytoplasmic distribution of vessel walls, and consequently patients bleed. In the NMMHCA. Interestingly, we found that some but central cytoplasmic domain, GPIb associates with the not all mice have hearing loss. The distribution of actin cross-linking protein, filamin A. Thus, the MYH9 expression in the inner ear has been studied defective linkage between GPIb/IX/V and cytoskele- in the developing fetal, neonate, adult mice [11], ton is the proposed molecular cause of the giant suggesting that MYH9 may have important roles in platelets. the development and maintenance of auditory The classical diagnostic features are a prolonged function. On the other hand, the unconventional bleeding time, moderate to severe thrombocytopenia, myosins, Myosin VIIA is localized in the stereocilia and giant platelets. Especially, giant platelets and the and cell body of hair cells and is critical in differentia- absence of ristocetin-induced platelet agglutination tion, formation, and/or maintenance of sensory are the laboratory hallmarks of BSS (Table I). Flow hair cell structure [12 /14]. Myosin VIIA have been cytometric determination of platelet GPIb/IX expres- linked to nonsyndromic and syndromic hearing loss. sion is a convenient method for diagnosis of BSS in a Therefore, the requirement of NMMHCA in the clinical laboratory. mammalian auditory system might be limited Thus far, 39 mutations in the genes for GPIba, and could be compensated by other conventional GPIbb, and GPIX have been described. Approxi- myosins. Thus heterozygous loss of NMMHCA mately half of the mutations were found in the GPIba might result in the phenotype that does not appear gene (17 mutations), and the remaining were found in uniformly, or the severity of phenotype might be the GPIbb (12 mutations) and GPIX genes (10 unrecognizable. mutations). A majority of the mutations correspond to single-base substitutions or small base deletion and insertion mutations. Recent studies have shown that the phenotypes caused by mutations in the subunits of GPIb/IX span a wide spectrum, from

Figure 2. Subcellular localization of neutrophil NMMHCA. Figure 3. Bernard-Soulier syndrome. 44 H. Saito et al. the normal phenotype to isolated macrothrombocy- responsible for the disease are currently not known. topenia with normal platelet function, to full-blown The most characteristic feature and thus the labora- BSS and platelet-type von Willebrand disease. It is tory hallmark of the syndrome is agranular platelets. important to note that although heterozygous BSS On Wright- or MGG- stained peripheral blood carriers are generally asymptomatic with subnormal smears, platelets appear gray or colorless due to the platelet number and function, they have giant plate- absence of platelet a-granules and their constituents. lets. Several individuals with heterozygous GPIb/IX Because platelet a-granule proteins such as platelet- mutation have been initially identified as having derived growth factor are synthesized but not properly undifferentiated thrombocytopenia or refractory stored in the granules and released from megakaryo- ITP. In Italy, a mutation in GPIba (Ala156Val) was cytes into the bone marrow, myelofibrosis is present in found to be a founder mutation responsible for the most cases [22]. autosomal dominant macrothrombocytopenia previously known as a Mediterranean macrothrombocy- Type 2B von Willebrand disease topenia [18]. In addition, patients with DiGeorge/ velo-cardio-facial syndrome due to a heterozygous Patients have a prolonged bleeding time, decidedly chromosome 22q11.2 microdeletion, which includes low vWF activity measured as ristocetin cofactor the GPIbbgene, have macrothrombocytopenia [19] activity, a mild deficiency of vWF antigen level, and (Table I). enhanced ristocetin-induced platelet aggregation at low concentrations of ristocetin. Although the molecular mechanisms remain to be elucidated, some X-linked macrothrombocytopenia with patients with type 2B vWD have been reported to dyserythropoiesis have giant platelets [23]. Recently, several unrelated families with X-linked macrothrombocytopenia with mild to moderate dys- Approach to patients with erythropoiesis have been found to have mutations in macrothrombocytopenia [24] the GATA-1 gene. GATA-1 is a megakaryocyte- and erythroid-specific transcription factor required for First of all, acquired causes of macrothrombocytope- normal growth and differentiation of both lineages. nia, including ITP and myelodysplastic syndromes, Defective GATA-1 function due to missense muta- should be ruled out. Complete history and physical tions causes reduced transcription and thus protein examination should be carefully performed. In syn- expression of its target genes, including GPIba, dromic forms, patients show complications of physical GPIbb, GPIX, and GPV. The concomitant decrease abnormalities such as facial, cardiac, skeletal anoma- in the expression of GPIb/IX/V and other platelet- lies and/or mental retardation. If the patient pre- specific gene products is related to the macrothrom- viously had normal platelet counts, acquired rather bocytopenia and bleeding tendency in this disorder than congenital conditions are more likely to be the [20]. underlying cause. In inherited macrothrombocytopenias, platelet counts are constantly decreased, ranging 9 9 from as low as 10/10 /l to near normal 150/10 /l. Paris-Trousseau syndrome/Jacobsen syndrome On a peripheral blood smear, the majority of platelets Paris-Trousseau syndrome/Jacobsen syndrome is a are large, being similar to or larger than red blood contiguous gene syndrome characterized by mental cells or small lymphocytes. In contrast, in patients retardation, and facial and cardiac abnormalities due with the much more common ITP, large platelets are to a heterozygous 11q23 deletion. The platelets present but the majority are of normal size. Because contain giant a granules on peripheral blood smears, routine automated blood cell counting systems differ- and in the bone marrow megakaryocytes are increased entiate blood cells by their size and do not recognize with many micro megakaryocytes. Hemizygous dele- giant platelets as platelets, these instruments under- tion of the transcription factor Fli1 contributes to the estimate platelet counts in patients with macrothrom- hematopoietic defects in this disorder [21]. bocytopenia. The mean platelet volume, usually calculated as a parameter of the complete blood count, also does not reflect actual platelet size in the Inherited macrothrombocytopenias of unknown case of giant platelets. Platelet count should therefore cause be determined manually in a calculating chamber or on peripheral blood smears. Careful examination of a Gray platelet syndrome smear also allows morphological assessment of leuko- Gray platelet syndrome (GPS) is characterized by cytes and erythrocytes. If granulocyte inclusion bodies thrombocytopenia and abnormal giant platelets with are obscure or absent, immunofluorescence analysis absent platelet a-granules. Patients with GPS have a for neutrophil NMMHCA localization is helpful to bleeding tendency of variable severity. The gene(s) make a clear distinction. Flow cytometric analysis of Giant platelet syndrome 45 platelet GPIb/IX expression can differentiate BSS inclusions (May-Hegglin anomaly/Sebastian syndrome). heterozygotes from patients with ‘‘true’’ isolated Blood 2001;97:/ 1147/ /1149. [6] Kunishima S, Matsushita T, Kojima T, Sako M, Kimura F, Jo macrothrombocytopenia. EK, Inoue T, Kamiya T, Saito H. Immunofluorescence Patients with congenital macrothrombocytopenia analysis of neutrophil nonmuscle myosin heavy chain-A in generally do not respond to standard ITP treatments, MYH9 disorders: Association of subcellular localization with including corticosteroids, intravenous immunoglobu- MYH9 mutations. Lab Invest 2003;83:/ 115/ /122. lin, and splenectomy. If treatment for bleeding is [7] Kunishima S, Matsushita T, Kojima T, Amemiya N, Choi YM, Hosaka N, Inoue M, Jung Y, Mamiya S, Matsumoto K, clinically indicated, the administration of antifibrino- et al. Identification of six novel MYH9 mutations and lytic agents such as o-aminocaproic acid or tranexamic genotype-phenotype relationships in autosomal dominant acid and recombinant activated factor VII may macrothrombocytopenia with leukocyte inclusions. J Hum transiently improve the episodes [25]. Transfusion of Genet 2001;46:/ 722/ /729. platelets is effective for serious bleeding and as [8] Heath KE, Campos-Barros A, Toren A, Rozenfeld-Granot G, prophylaxis prior to major surgery, but may be Carlsson LE, Savige J, Denison JC, Gregory MC, White JG, Barker DF, et al. Nonmuscle myosin heavy chain IIa muta- complicated by the development of alloantibodies. tions define a spectrum of autosomal dominant macrothrom- In certain instances, hematopoietic stem cell trans- bocytopenias: May-Hegglin anomaly and Fechtner, Sebastian,

plantation (SCT) may be a curative therapeutic Epstein, and Alport-like syndromes. Am J Hum Genet 2001;/ option [26]. It is important to make a proper 69:1033/ /1045. diagnosis to avoid unnecessary treatment. Affected [9] Matsushita T, Hayashi H, Kunishima S, Hayashi M, Ikejiri M, Takeshita K, Yuzawa Y, Adachi T, Hirashima K, Sone M, et al. families should be educated about their diagnosis to Targeted disruption of mouse ortholog of the human MYH9 avoid unnecessary medications and potentially dan- responsible for macrothrombocytopenia with different organ gerous treatments for presumed ITP. When evaluat- involvement: hematological, nephrological, and otological ing patients with refractory ITP or undifferentiated studies of heterozygous KO mice. Biochem Biophys Res thrombocytopenia, congenital macrothrombocytope- Commun 2004;325:/ 1163/ /71. [10] Wylie SR, Chantler PD. Separate but linked functions of nias should be included in the differential diagnosis. conventional myosins modulate adhesion and neurite out- growth. Nat Cell Biol 2001;3:/ 88/ /92. Conclusions [11] Mhatre AN, Li J, Kim Y, Coling DE, Lalwani AK. Cloning and developmental expression of nonmuscle myosin IIA

Inherited giant platelet syndromes are rare conditions, (Myh9) in the mammalian inner ear. J Neurosci Res 2004;/ yet the study of them has been instrumental in 76:296/ /305. [12] Avraham KB, Hasson T, Sobe T, Balsara B, Testa JR, Skvorak elucidating the structure and function of normal AB, Morton CC, Cope land NG, Jenkins NA. Characteriza- platelets as well as the mechanisms of thrombopoiesis. tion of unconventional MYO6, the human homologue of the Further research is needed to understand the patho- gene responsible for deafness in Snell’s waltzer mice. Hum genesis of many congenital disorders with unknown Mol Genet 1997;6:/ 1225/ /1231. causes. [13] Hasson T, Walsh J, Cable J, Mooseker MS, Brown SD, Steel KP. Effects of shaker-1 mutations on myosin-VIIa protein and mRNA expression. Cell Motil Cytoskeleton 1997;37:/ 127/ / Acknowledgements 138. [14] Liang Y, Wang A, Belyantseva AI, Anderson DW, Probst FJ, This work is supported partly by grants-in-aid from Barber TD, Miller W, Touchman JW, Jin L, Sullivan SL, et al. the Ministry of Education, Science, Sports and Characterization of the human and mouse unconventional myosin XV genes responsible for hereditary deafness DFNB3 Culture, from the Ministry of Health and Welfare, and shaker 2. Genomics 1999;61:/ 243/ /258. Japan, and from Sankyou Life Science Foundation, [15] Bernard J, Soulier JP. Sur une nouvelle variete de dystrophie Japan. thrombocytaire-hemorragipare congenitale. Sem Hop Paris 1948;24:/ 3217/ /3223. [16] Lopez JA, Andrews RK, Afshar-Kharghan V, Berndt MC. Bernard-Soulier syndrome. Blood 1998;91:/ 4397/ /4418. References [17] Kunishima S, Kamiya T, Saito H. Genetic abnormalities [1] Hartwig J, Italiano J, Jr. The birth of the platelet. J Thromb of Bernard-Soulier syndrome. Int J Hematol 2002;76:/ 319/ / Haemost 2003;1:/ 1580/ /1586. 327.

[2] May R. Leukozyteneinschlusse. Dtsch Arch Klin Med 1909;/ [18] Savoia A, Balduini CL, Savino M, Noris P, Del Vecchio M, 96:1/ /6. Perrotta S, Belletti S, Poggi, Iolascon A. Autosomal dominant [3] The May-Hegglin/Fechtner Syndrome Consortium. Muta- macrothrombocytopenia in Italy is most frequently a type of

tions in MYH9 result in the May-Hegglin anomaly, and heterozygous Bernard-Soulier syndrome. Blood 2001;97:/ / Fechtner and Sebastian syndromes. Nat Genet 2000;26:/ / 1330 /1335.

103/105. [19] Van Geet C, Devriendt K, Eyskens B, Vermylen J, Hoylaerts [4] Kelley MJ, Jawien W, Ortel TL, Korczak JF. Mutation of MF. Velocardiofacial syndrome patients with a heterozygous MYH9, encoding non-muscle myosin heavy chain A, in May- chromosome 22q11 deletion have giant platelets. Pediatr Res / / Hegglin anomaly. Nat Genet 2000;26:/ 106/ /108. 1998;44:607 /611. [5] Kunishima S, Kojima T, Matsushita T, Tanaka T, Tsurusawa [20] Nichols KE, Crispino JD, Poncz M, White JG, Orkin SH, M, Furukawa Y, Nakamura Y, Okamura T, Amemiya N, Maris JM, Weiss MJ. Familial dyserythropoietic anaemia and Nakayama T, et al. Mutations in the NMMHC-A gene cause thrombocytopenia due to an inherited mutation in GATA1. autosomal dominant macrothrombocytopenia with leukocyte Nat Genet 2000;24:/ 266/ /270. 46 H. Saito et al.

[21] Raslova H, Komura E, Le Couedic JP, Larbret F, Debili N, patients with type 2B von Willebrand disease. Br J Haematol Feunteun J, Danos O, Albagli O, Vainchenker W, Favier R. 2000;110:/ 704/ /714. FLI1 monoallelic expression combined with its hemizygous [24] Kunishima S, Saito H. Congenital macrothrombocytopenias. loss underlies Paris-Trousseau/Jacobsen thrombopenia. J Clin Blood Reviews, in press. [25] Poon MC, d’Oiron R. Recombinant activated factor VII Invest 2004;114:/ 77/ /84. [22] Schmitt A, Jouault H, Guichard J, Wendling F, Drouin A, (NovoSeven) treatment of platelet-related bleeding disorders. Cramer EM. Pathologic interaction between megakaryocytes International Registry on Recombinant Factor VIIa and and polymorphonuclear leukocytes in myeloﬁbrosis. Blood Congenital Platelet Disorders Group. Blood Coagul Fibrino- lysis 2000;11:/ S55/ /68. 2000;96:/ 1342/ /1347. [23] Nurden P, Chretien F, Poujol C, Winckler J, Borel-Derlon A, [26] Locatelli F, Rossi G, Balduini C. Hematopoietic stem-cell Nurden A. Platelet untrastructural abnormalities in three transplantation for the Bernard-Soulier syndrome. Ann Intern

Med 2003;138:/ 79./ Hematology, 2005; 10 Supplement 1: 47/48

PLATELETS

The role of ADAMTS13 in the new pathogenesis of TTP

FLORA PEYVANDI

Angelo Bianchi Bonomi Hemophilia and Thrombosis Center, Via Pace 9, 20122, Milan, Italy

Thrombotic Thrombocytopenic Purpura (TTP) is a Gerritsen [9] and Zheng [10] in 2001. In the same severe microvascular occlusive microangiopathy char- year Levy identified the ADAMTS13 gene and acterized by thrombocytopenia, Coombs-negative described in seven unrelated families 12 distinct hemolytic anemia and ischemic symptoms localized mutations, underlying the molecular mechanism re- mainly but not exclusively in central nervous system sponsible for congenital TTP [11]. At this point and resulting from diffuse platelet thrombi in micro- everything seemed to become clear: an enzyme circulation [1]. To understand the broad clinical deficiency allows ULVWF multimers to circulate; spectrum of TTP it is important to read the original the unusually large multimers cause disseminated case report prior to effective treatment, when the full platelet clumping in the microcirculation; the defi- natural history of this fatal illness was appreciated. In cient protease can be replaced by plasma infusion, 1924 Dr. Moschowitz [2] described the first case in a which is sufficient in patients with congenital defi- previously healthy 16-year-old girl who died with ciency; the inhibitor of ADAMTS13 activity in multiple organ failure and post-mortem examination patients with acquired disease can be removed by showed widespread thrombi in the terminal circula- plasmapheresis, making plasma exchange a more tion of several organs, composed mainly by platelets. effective treatment than plasma infusion. However, These thrombi remain the hallmark of pathologic as clinical studies have extended beyond the initial diagnosis. translational research to community studies [12], the The current era in our knowledge began with the inevitable increase in complexity of these syndromes observation by Moake and his colleagues in 1982 [3] has been appreciated. The main finding of these of unusually large multimers of von Willebrand factor subsequent studies is that only approximately 50% (ULVWF) in the plasma of patients with chronic of patients with overt clinical disease have low plasma relapsing TTP. They are similar to those normally levels of the protease. Simultaneously an enormous contained in vascular endothelial cells and platelets number of in vitro studies have been initiated in order but absent in plasma. These ULVWF multimers to comprehend the pathophysiology of TTP. Between promote platelet-dependent microvascular thrombo- 2002 and 2004 it has been demonstrated that sis, suggesting that patients lacked a VWF protease ULVWF multimers are cleaved by ADAMTS13 able to reduce the size of VWF multimers by cleavage. between Tyr1605-Met1606 residues located within The next key step was the discovery made 14 years the A2 domain as they are secreted in long ‘‘strings’’ later, when Furlan, Tsai et al. [4,5] isolated a VWF- from stimulated endothelial cells [13,14]. The cleaving protease that had properties consistent with ULVWF strings are anchored to the endothelial cell the activity that had been postulated to be absent in membrane via P-selectin molecules that are secreted patients with TTP. These two groups subsequently concurrently with the ULVWF multimers [15]. Spe- published evidence that the consistently very low or cifically, ADAMTS13 enzyme binds under flow con- undetectable plasma levels of this protease were ditions to accessible A3 domains in the monomeric indicative of TTP [6,7]. The VWF-cleaving protease subunit of ULVWF multimers and then cleaves Tyr is now known as ADAMTS13, a member of the 1605-Met 1606 peptide bonds in adjacent A2 do- ADAMTS metalloprotease family (A Disintegrin And main. Majerus et al. [16] have shown that the spacer Metalloprotease with Thrombospondin-1 repeats), and TSP1-1 like domains are required for thanks to the results obtained by Fujikawa [8], ADAMTS13 binding to VWF, that may be modu-

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389890 48 F. Peyvandi lated by CUB and TSP1-2/8 like domains. As a [5] Tsai HM. Physiologic cleavage of von Willebrand factor by a plasma protease is dependent on its conformation and requires consequence of ADAMTS13 deficiency, ULVWF calcium ion. Blood 1996;87:/ 4235/ /4244. multimers are not cleaved after their secretion from [6] Furlan M, Robles R, Galbusera M, Remuzzi G, Kyrle PA, endothelial cells, but remain anchored to the cells. Brenner B, Krause M, Scharrer I, Aumann V, Mittler U, et al. Passing platelets adhere via their GpIb and GpIIb/IIIa von Willebrand factor-cleaving protease in thrombotic throm- to the A1 and the A3 domains of the monomeric bocytopenic purpura and the hemolytic-uremic syndrome. N subunits of ULVWF strings anchored to P-selectin to Engl J Med 1998;339:/ 1578/ /1584. [7] Tsai HM, Lian EC. Antibodies to von Willebrand factor- form large, potentially occlusive, platelet thrombi. cleaving protease in acute thrombotic thrombocytopenic Moreover, considering that severely deficient purpura. N Engl J Med 1998;339:/ 1585/ /1594. ADAMTS13 activity may not always be sufficient to [8] Fujikawa K, Suzuki H, McMullen B, Chung D. Purification of cause TTP, in the last two years some researchers human von Willebrand factor-cleaving protease and its identification as a new member of the metalloproteinase family. have started to investigated other proteins that might Blood 2001;98:/ 1662/ /1666. be involved in the mechanisms underlying TTP. [9] Gerritsen HE, Robles R, Lammle B, Furlan M. Partial amino Pimanda and their colleagues have been demon- acid sequence of purified von Willebrand factor-cleaving strated that the multimeric size of VWF can be protease. Blood 2001;98:/ 1654/ /1661. controlled by disulfide bonds-reduction by thrombos- [10] Zheng X, Chung D, Takayama TK, Majerus EM, Sadler JE, pondin-1 (TSP-1) [17]. VWF reductase activity was Fujikawa K. Structure of von Willebrand factor-cleaving protease (ADAMTS13), a metalloprotease involved in throm- centered around Cys974 in the C-terminal sequence

botic thrombocytopenic purpura. J Biol Chem 2001;276:/ / of TSP-1 but the role of TSP-1 in the aetiology of 41059/41063. TTP is under investigation. Many other aspects of [11] Levy GG, Nichols WC, Lian EC, Foroud T, McClintick JN, ULVWF processing are now rapidly coming into McGee BM, Yang AY, Siemieniak DR, Stark KR, Gruppo R, focus, such as the involvement in ULVWF secretion et al. Mutations in a member of the ADAMTS gene family

cause thrombotic thrombocytopenic purpura. Nature 2001;/ of such cytokines as IL-8, TNF-a and IL-6 in 413:488/ /494. complex with its receptor. This might be explain the [12] George JN, Vesely SK. Thrombotic thrombocytopenic pur- role of inflammation as triggering agent in TTP [18]. pura: from the bench to the bedside, but not yet to the Furthermore different authors are studying the reg- community. Ann Intern Med 2003;138:/ 152/ /153. ulation of ADAMTS13 activity by different factors as [13] Dong JF, Moake JL, Nolasco L, Bernardo A, Arceneaux W, Shrimpton CN, Schade AJ, McIntire LV, Fujikawa K, Lopez thrombin and plasmin, that cleave ADAMTS13 at JA. ADAMTS-13 rapidly cleaves newly secreted ultralarge von specific sites, resulting in the loss of ADAMTS13 Willebrand factor multimers on the endothelial surface under activity [19]; or chloride ions which, binding to VWF ﬂowing conditions. Blood 2002;100:/ 4033/ /4039. and causing its conformational change, have specific [14] Dong JF, Moake JL, Bernardo A, Fujikawa K, Ball C, Nolasco inhibitory effect on ADAMTS13 activity [20]. Finally L, Lopez JA, Cruz MA. ADAMTS-13 metalloprotease interacts with the endothelial cell-derived ultra-large von Will- it has been demonstrated that the VWF A1 domain ebrand factor. J Biol Chem 2003;278:/ 29633/ /29639. inhibits the cleavage of A2 domain by ADAMTS13 [15] Padilla A, Moake JL, Bernardo A, Ball C, Wang Y, Arya M, and this inhibition can be relieved by interaction of A1 Nolasco L, Turner N, Berndt MC, Anvari B, et al. P-selectin domain with platelet GpIb or certain glycosaminogly- anchors newly released ultralarge von Willebrand factor cans (heparin) [21]. multimers to the endothelial cell surface. Blood 2004;103:/ / 2150/2156. Until now it has been paved the way to understand [16] Majerus EM, Anderson PJ, Sadler JE. Binding of

the exact role of ADAMTS13 and the other mod- ADAMTS13 to von Willebrand factor. J Biol Chem 2005;/ ulators in developing TTP but many unresolved 280:21773/ /21778. issues still remain. [17] Pimanda J, Hogg P. Control of von Willebrand factor multimer size and implications for disease. Blood Rev 2002;

16:185/192. Review [18] Bernardo A, Ball C, Nolasco L, Moake JF, Dong JF. Effects of References inﬂammatory cytokines on the release and cleavage of the endothelial cell-derived ultralarge von Willebrand factor [1] Moake JL. Thrombotic microangiopathies. N Engl J Med multimers under ﬂow. Blood 2004;104:/ 100/ /106.

2002;347:589/600. Review [19] Crawley JT, Lam JK, Rance JB, Mollica LR, O’Donnell JS, [2] Moschcowitz E. Hyaline thrombosis of the terminal arterioles Lane DA. Proteolytic inactivation of ADAMTS13 by throm- and capillaries: a hitherto undescribed disease. Proc N Y bin and plasmin. Blood 2005;105:/ 1085/ /1093. Pathol Soc 1924;24:/ 21/ /24. [20] De Cristofaro R, Peyvandi F, Palla R, Lavoretano S, Lombardi [3] Moake JL, Rudy CK, Troll JH, Weinstein MJ, Colannino NM, R, Merati G, Romitelli F, Di Stasio E, Mannucci PM. Role of Azocar J, Seder RH, Hong SL, Deykin D. Unusually large chloride ions in modulation of the interaction between von

plasma factor VIII:von Willebrand factor multimers in chronic Willebrand factor and ADAMTS-13. J Biol Chem 2005;280:/ /

relapsing thrombotic thrombocytopenic purpura. N Engl J 23295/23302. Med 1982;307:/ 1432/ /1435. [21] Nishio K, Anderson PJ, Zheng XL, Sadler JE. Binding of [4] Furlan M, Robles R, Lammle B. Partial puriﬁcation and platelet glycoprotein Ibalpha to von Willebrand factor domain characterization of a protease from human plasma cleaving A1 stimulates the cleavage of the adjacent domain A2 by

von Willebrand factor to fragments produced by in vivo ADAMTS13. Proc Natl Acad Sci USA 2004; 101:10578/ proteolysis. Blood 1996;87:/ 4223/ /4234. 10583 Hematology, 2005; 10 Supplement 1: 49

ACUTE MYELOID LEUKEMIA

Biology adapted treatment of acute myeloid leukemia

WOLFGANG HIDDEMANN on behalf of the German AML Cooperative Group (AMLCG)

Klinikum GroBhadern, III Med. Klinik Marchioninistr. 15, 81377, Mu¨ nchen

Genetic and molecular techniques have provided the double induction therapy with either HAM/HAM means to gain increasing insights into the biology of versus TAD/HAM, priming with G-CSF and the acute myeloid leukemia (AML). These investigations impact of autologous stem cell transplantation versus demonstrated that AML is not a homogeneous long term maintenance as treatment in remission. disease but a heterogeneous group of biologically The new strategy of the AMLCG is directed to a different subentities. These subentities are currently more biology oriented treatment. Hence, besides primarily defined by cytogenetic and molecular ana- AML with the t(15;17) which is been treated in a lyses. Based on cytogenetics three main subgroups of different protocol for a long time already also other AML can be discriminated: AML with balanced subtypes will receive risk adapted regimens. Hence, in translocations, AML with unbalanced aberrations patients with complex karyotype early transplantation and AML without cytogenetically detectable aberra- with those reduced conditioning will be applied. In tions. Within the latter group molecular alterations patients with standard risk AML intensification in way are identified in more than half of all cases such as of the S-HAM regimen is currently explored. In NPM-mutations, FLT3- mutations, MLL-duplica- addition a prospective randomized comparison of tions and mutations of CEBP-a. conventional versus dose-reduced conditioning has These different subtypes reveal a different prog- been initiated. In patients with core binding factor nosis. Based on these insides recent treatment strategies try to adapt the knowledge about the biology to leukemia transplantation will not be performed but develop biology adapted treatment strategies. This diseases will be monitored on a molecular level. process is still under development and is determined Transplantation is only initiated as soon as a mole- by the availability of new agents and their evaluation cular signal is redetected or reveals rising transcript in phase II- studies but also on the more conservative levels. approach to direct currently established protocols to The step towards the biology adapted treatment of biologic risk groups. The AMLCG has thoroughly AML is long and requires the combined efforts of investigated the outcome of biologically different researchers, clinicians and the pharmaceutical indus- subgroups on the basis of a prospective randomized try. The first steps towards this goal, however, have trial asking three main questions: The comparison of been taken.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389908 Hematology, 2005; 10 Supplement 1: 50/53

ACUTE MYELOID LEUKEMIA

The treatment of AML: Current status and novel approaches

ALAN K. BURNETT

Wales School of Medicine, Cardiff University, Heath Park, Cardiff, CF14 4XN

The epidemiology of AML suggests that in Western complex, t(9;22)) have a survival of 15%. All other countries the median age of patients is around 65 lesions and normal have a survival of 50%. The years. Convention in recent years has often resulted in presence of a FLT3 internal tandem repeat (ITD) collaborative groups designing trials for patients mutation which itself significantly increases the re- under or over 60 years. In the former group treatment lapse risk [10], adds refinement to the cytogenetic risk options included consolidation with stem cell trans- group / particularly the intermediate risk group plantation of matched sibling allograft or autologous (Relapse risk ITD/ve 70%; ITD-ve 45%). Muta- cells collected earlier in remission, which was not tions of RAS are present in 12% of patients, but do considered feasible in older patients. not appear to be prognostically important. As a general rule these prognostic factors are independent Treatment in younger patients of treatment, including transplantation. Our recent analysis has been unable to show that there is benefit It is now expected that complete remission (CR) can from transplant in FLT3 positive cases. Other muta- be obtained in 75/90% of patients under 60 years, tions / although relatively rare, are being identified usually with one course of treatment. Treatment which may have prognostic value. mortality should be B/10% with optimal supportive Numerous trials have been conducted over the care. About 50% of patients who achieve complete years with the aim of determining which drug and remission will relapse, mostly over the next two years, dose combination is the most effective. Comparisons so the overall 5 year survival at five years is around of anthracyclines, doses of Ara-C, the presence or not 50%. This represents an improvement of 1/1.5% of a 3rd drug, have all been assessed without any per annum over the last 20 years. No single treat- convincing evidence of overall superiority for one ment modality accounts for this although inten- schedule. Given the intensity of induction and con- sification of treatment made possible by better supportive care seems the most plausible explanation solidation, maintenance treatment is not indicated in [1,2]. younger patients. The precise total number of treat- Stem cell transplantation (both autograft and ment courses required is not fully established but matched sibling allograft) has been carefully assessed MRC studies suggest that the maximum should be 4 [11]. in a number of major collaborative group trials [3/7]. This approach undoubtedly reduces the risk of Once a patient relapses, his/her outcome is again relapse even in patients who have been treated dictated by prognostic factors more than treatment. intensively; however the impact on survival is less These are age, length of first remission and cytoge- clear. The risk of relapse is dominated by a number of netic risk group [12]. Patients who fail a transplant do validated prognostic factors such as cytogenetics, less well, but this may in part be due to a reluctance to FLT3 mutation status, the extent of blast clearance treat such patients intensively. Old age, short first from the bone marrow after treatment course 1, remissions and adverse cytogenetics are virtually presenting WBC and patient age [8,9]. Favourable unsalvageable. Overall the outcome from relapse is risk cytogenetics includes (inv16); t(8;21) and about 20% without transplant, 35% with an autograft t(15;17) and has an overall cure rate of 70/75%. and 45/50% with an allograft (matched sibling or Poor risk cytogenetics (abnormalities of Chs 5,7 3q- volunteer donor) [13].

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389773 The treatment of AML: Current status and novel approaches 51

Potential for improvement in younger patients searching for novel against with a larger number of randomized phase 2 studies. With a remission rate as high as it currently is, it will be difficult to show a benefit on remission rate of a new treatment with adequate statistical power. Transplantation in the older patient However it is well established but induction treatment Until recently it was accepted that standard allogeneic / while not necessarily impacting on remission rate / or autologous SCT in older patients carried a can improve disease free survival by improving the prohibitive mortality. However interest has been ‘‘quality’’ of remission. rekindled by the development of reduced intensity One of the interesting approaches currently in trial transplantation. This is clearly feasible with stable is adding in the immunoconjugate Gemtuzumab establishment of a full chimeric state. There does not TM Ozogamicin (Mylotarg ). Initial studies at full dose yet seem to be a best choice preparative protocol. The 2 (9 mgs/m ) suggested that unexpected liver toxicity slightly less reduced intensity will be strong enough to would be problematic [14]. However a large study achieve full chimerism, but will often be accompanied currently being conducted by the MRC confirms that by acute or chronic GVHD. More reduction in adding Mylotarg in a reduced dose (3 mgs/m2)is preparation intensity will reduce GVHD but may be feasible [15], however any benefit will not be known associated with rejection. for 2/3 years. Initially it was not thought that this transplant Poor risk cases are particularly refractory, and approach would be suitable for AML since several although there is not unreal consistency in the weeks are required to establish full chimerism. How- evidence base such patients are usually offered allo- ever a number of phase 2 experiences now suggest geneic SCT (including MUD transplants) as soon as that durable survivals are seen in AML. However the risk status is known. there is nothing to suggest that this approach has a With the recognition of different molecular ab- survival advantage and it is important that this normalities, the expectation is that an era of molecu- approach develops within the context of a prospective larly target therapy / most likely in combination with clinical trial. The improvements in tissue typing existing treatments is imminent. brought about by molecular matching of unrelated donors, has brought unrelated transplants to be a very practical option. Limited data suggests that this is also Treatment of older patients the case in older patients. These developments mean The problem of treatment of older patients with AML that allogeneic transplant is a viable option for older has a number of facets [16,17]. First there is a patients and is ready to be evaluated in clinical trials. substantial population of older patients who are not considered fit for the usual combination chemother- Novel agents apy approach on offer due to age, frailty or comor- bidity. Second, the success of effective treatment in Given the lack of progress in older patients, and the younger patients is less successful in substantial part probability that the molecular understanding of AML due to aggregation of adverse prognostic factors in will bring new agents to the clinic, it is appropriate to direct new strategies and agents to the older patient. older patients e.g., cytogenetics, prior MDS, and molecularly defined chemo-resistance. Chemoresis- tance is defined by a number of proteins several of Gemtuzumab ozogamicin [Mylotarg] which may be coexpressed in the same patient. The This immunoconjugate links a humanized anti-CD33 most common and predictive is over expression monoclonal antibody to the powerful intercalator of Pglycoprotein. Attempts to modulate the efflux calicheamicin. The key features are that the complex action of the protein using the cyclosporine analogue of CD33 antigen and antibody is rapidly internalized PSC-833 have been unsuccessful [18,19]. Third, the making it a potentially specific targeting vehicle for a survival has not / unlike in younger patients / chemotherapeutic agent. Since 90% of AML cases improved much in the last two decades. Overall the express CD33 and that expression is primarily on current expectation of CR in an older patient is haemopoietic tissue this target is appropriate. A 50 /60%. The relapse rate is high so overall the unique feature is that the clinical link joining the survival is 10 /15% at 5 years. antibody to calicheamicin is only lysed intracellularly. Many randomized trials with large numbers have Free calicheamicin is too toxic to give as a free agent tried to improve outcome, with little clear success. so this is an important property. This raises a further strategic issue in older patients, Although only licences in the US and Japan for which is the question as to whether we should relapsed disease in the elderly. There is much interest continue to focus clinical research and large numbers in either combining it with chemotherapy / either of elderly patients in large phase 3 trials as opposed to simultaneously or sequentially / in or as maintenance 52 A. K. Burnett treatment in older patients. Pilot studies have con- study using a reduced dose as first line treatment firmed the feasibility of this approach [15] but patients who were not considered fit for intensive randomized studies will not be available for 3/5 years. therapy [29]. A remission rate of 56% was observed with an acceptable toxicity profile although the degree of myelosuppression suggests that even a reduced Targeting FLT-3 mutations dose was in fact intensive treatment. Nevertheless

About 30/35% of younger patients and a lower Clofarabine seems worthy of development either as a proportion of older patients will have a FLT-3 component of intensive treatment or at reduced dose mutation. As well as providing valuable prognostic for older patients. information, FLT-3 mutation could provide a target for therapeutic targeting. At least 5 agents with in vitro activity against cells carrying a FLT-3 mutation have had some clinical evaluation (PKC412; References CEP701; MLN518, SU5416) [20/24]. PCK 412 [1] Lowenberg B, Downing JR, Burnett A. Acute myeloid and CEP701 have been assessed in unrandomized leukemia. N Engl J Med 1999;341:/ 1051/ /1062. phase 2 trials in patients with mutations. Although a [2] Burnett AK. Acute myeloid leukemia: Treatment of adults degree of disease control was seen, no CR’s were under 60 years. Rev Clin Exp Hematol 2002;6:/ 26/ /45. obtained with it seems likely that in future these [3] Burnett AK, Wheatley K, Goldstone AH, Stevens RF, Hann agents will be combined with chemotherapy. Second IM, Rees JKH, Harrison G. The value of allogeneic bone generation agents may be more effective but no marrow transplant in patients with acute myeloid leukaemia at differing risk of relapse: results of the UK MRC AML10 Trial. clinical data is yet available. Brit J Haematol 2002;118:/ 385/ /400. [4] Burnett AK, Goldstone AH, Stevens R, Hann IM, Rees JK, Gray RG, Wheatley K. Randomised comparison of addition of Farnesylation autologous bone-marrow transplantation in intensive chemotherapy for acute myeloid leukaemia in first remission: A number of cytoplasmic proteins, including RAS, results of MRC AML 10 Trial. Lancet 1998;351:/ 700/ /708. which are involved in signal transduction, prolifera- [5] Zittoun RA, Mandelli F, Willemze R, de Witte T, Labar B, tion, angiogenesis and apoptotic pathways require to Resegotti L, Leoni F, Damasio E, Visani G, Papa G, et al. be prenylated. This process is mediated by the enzyme Suciu. Autologous or allogeneic bone marrow transplantation Farnesyl Transferase (FTase). FTase adds a 15-mer compared with intensive chemotherapy in acute myelogenous fatty acid chain to the C-terminal residue of substrate leukemia. European Organization for Research and Treatment of Cancer (EORTC) and the Gruppo Italiano Malattie proteins which bear a / CAAX motif (farnesylation). Ematologiche Maligne dell’Adulto (GIMEMA) Leukemia This allows attachment to the inner surface of the Cooperative Groups. New Engl J Med 1995;332:/ 217/ /223. membrane so facilitating binding of guanine nucleo- [6] Harousseau JL, Cahn JY, Pignon B, Witz F, Milpied N, Delain tide. While RAS may be an obvious target in AML, M, Lioure B, Lamy T, Desablens B, Guilhot F, et al. but disruption of other known or unknown proteins, Comparison of autologous bone marrow transplantation and intensive chemotherapy as postremission therapy in adult may also be relevant. Tipifarnib (Zarnestra) is the acute myeloid leukemia. Blood 1997;90:/ 2978/ /2986. agent which is most developed clinically in AML. [7] Cassileth PA, Harrington DP, Appelbaum F, Lazarus HM, Phase 1 in relapsed or refractory disease trials Rowe JM, Paietta E, Willman C, Hurd DD, Bennett JM, established a tolerable daily dose of 600 mgs which Blume KG, et al. Chemotherapy compared with autologous or can be given orally for 21 days a month. An allogeneic bone marrow transplantation in the management of acute myeloid leukemia in first remission. New Eng J Med unrandomized phase 2, involving trial patients, con- 1998;339:/ 1649/ /1656. firmed activity with a remission rate of around 25% in [8] Grimwade D, Walker H, Harrison G, Oliver F, Chatters S, older patients not considered fit for intensive treat- Harrison C, Wheatley K, Burnett AK, Goldstone AH. The ment. Activity was not restricted to patients with RAS predictive value of hierarchical cytogenetic classification in older adults with AML: analysis of 1065 patients entered into mutations [25 /27]. the MRC AML11 Trial. Blood 2001;98:/ 1312/ /1320. In spite of the convenience of an oral agent with [9] Wheatley K, Burnett AK, Goldstone AH, Gray RG, Hann IM, activity as monotherapy, future interest will be in Harrison CJ, Rees JK, Stevens RF, Walker H. A simple, combination therapy. Clofarabine (2-chloro-2?-fluo- robust, validated and highly predictive index for the determi- ro-deoxy-9-b-D arabinofuranosyl adenine is a ration- nation of risk-directed therapy in acute myeloid leukaemia ally designed purine analogue. The design is intended derived from the MRC AML 10 trial. United Kingdom Medical Research Council’s Adult and Childhood Leukaemia to harness the potentially favourable profile of Flu- Working Parties. Brit J Haematol 1999;107:/ 69/ /79. darabine while avoiding the neurotoxicity associated [10] Kottaridis PD, Gale RE, Frew ME, Harrison G, Langabeer S, with the dose of Fludarabine that is effective in AML, Belton AA, Walker H, Wheatley K, Bowen DT, Burnett AK, and resisting deamination. With the potential to et al. The presence of a FLT3 internal tandem duplication in become orally available Clofarabine. activity in ad- patients with acute myeloid leukemia (AML) adds important prognostic information to cytogenetic risk group and response vanced paediatric ALL lead to approval in the USA. to the first cycle of chemotherapy: analysis of 854 patients Activity in AML has been established in a phase 2 trial from the United Kingdom Medical Research Council AML 10 in relapsed disease [28]. An unrandomized phase 2 and 12 Trials. Blood 2001;98:/ 1752/ /1759. The treatment of AML: Current status and novel approaches 53

[11] Burnett AK, Wheatley K, Goldstone AH, Gibson K, Webb D, reversal agent PSC-833 in elderly patients with acute myelo-

Prentice AG, Milligan D. MRC AML12: A comparison of genous leukemia. Blood 2004;104:/ 246a./

ADE vs MAE and S-DAT vs H-DAT9/Retinoic Acid for [20] Kell J. Emerging treatments in acute myeloid leukaemia. induction and four vs ﬁve total courses using chemotherapy or Expert Opin Emerg Drugs 2004;9:/ 55/ /71. stem cell transplant in consolidation in 3459 patients under 60 [21] Giles FJ, Stopeck AT, Silverman LR, et al. SU5416, a small

years with AML. Blood 2002;100:/ 155a./ molecule tyrosine kinase receptor inhibitor, has biologic [12] Breems DA, van Putten WL, Huijgens PC, Ossenkoppele GJ, activity in patients with refractory acute myeloid leukemia or Verhoef GE, Verdonck LF, Vellenga E, de Greef GE, Jacky E, myelodysplastic syndromes. Blood 2003;102:/ 795/ /801. van der Lelie J, Boogaerts MA, Lowenberg B. Prognostic [22] Stone RM, Klimek V, Deangelo DJ, et al. PKC412, an oral index for adult patients with acute myeloid leukemia in ﬁrst FLT3 inhibitor, has activity in mutant FLT3 acute myeloid relapse. J Clin Oncol 2005;23:/ 1969/ /1978. leukemia (AML): A phase II clinical trial. Blood 2002;101:/ /

[13] Burnett AK, Hills R, Goldstone AH, Milligan D, Prentice A, 16/11. Hann I, Webb D, Gibson B, Wheatley K. The impact of [23] Smith BD, Levis M, Beran M, Giles F, Kantarjian H, Berg K, transplant in AML in 2nd CR: A prospective study of 741 in et al. Single-agent CEP-701, a novel FLT3 inhibitor, shows

the MRC AML10 and 12 Trials. Blood 2004;104:/ 179a./ biologic and clinical activity in patients with relapsed or [14] Giles FJ, Kantarjian HM, Kornblau SM, Thomas DA, Garcia- refractory acute myeloid leukemia. Blood 2004;103:/ 3669/ / Manero G, David TA, Waddelow CL, Phan AT, Colburn DE, 3676. Rashid A, Estey AH. Mylotarg (gemtuzumab ozogamicin) [24] Knapper S, Burnett AK, Kell WJ, Clark R, Craig J, Chopra R, therapy is associated with hepatic venoocclusive disease in Culligan D, Levis MJ, Littlewood T, Russell NH, Small D. A patients who have not received stem cell transplantation. Phase II trial of the FLT3 inhibitor CEP701 as ﬁrst line Cancer 2001;15:/ 406/ /413. treatment for older patients with acute myeloid leukaemia not [15] Kell WJ, Burnett AK, Chopra R, Yin JAL, Clark RE, considered ﬁt for intensive chemotherapy. Am Soc Hematol Rohatiner A, Culligan D, Hunter A, Prentice AG, Milligan 2004; Submitted (Abstract). D. A feasibility study of simultaneous administration of [25] Karp JE, Lancet JE, Kaufmann SH, et al. Clinical and biologic gemtuzumab ozogamicin with intensive chemotherapy in activity of the farnesyltransferase inhibitor R115777 in adults induction and consolidation in younger patients with acute with refractory and relapsed acute leukemias: a phase 1 myeloid leukaemia. Blood 2003;102:/ 4277/ /4283. clinical-laboratory correlative trial. Blood 2001;97:/ 3361/ / [16] Hiddeman W, Kern W, Schoch C, Fonatsch C, Heinecke A, 3369. Wormann B, Buchner T. Management of acute myeloid [26] Harousseau JL, Reiffers J, Lowenberg B, et al. ZarnestraTM leukemia in elderly patients. J Clin Oncol 1999;17:/ 3569/ / (R115777) in patients with relapsed and refractory acute 3576. myelogenous leukemia (AML): results of a multicenter phase

[17] Goldstone AH, Burnett AK, Wheatley K, Smith AG, Hutch- 2 study. Blood 2003;102:/ 176a./ inson RM, Clark RE. Attempts to improve treatment out- [27] Lancet JE, Gotlib J, Gojo I, Feldman EJ, Morris L, Thibault comes in acute myeloid leukaemia (AML) in older patients: A, Liesveld JL, Greer J, Dugan K, Raponi M, et al. Tipifarnib the results of the United Kingdom Medical Research Council (ZARNESTRATM) in previously untreated poor risk AML of AML11 trial. Blood 2001;98:/ 1302/ /1311. the elderly: updated results of a multicenter phase 2 trial.

[18] van der Holt B, Lowenberg B, Burnett AK, Knauf KU, Blood 2004;104:/ 249a./ Shepherd J, Piccaluga PP, Ossenkoppele G, Verhoef GE, [28] Kantarjian H, Gandhi V, Cortes J, Verstovsek S, Garcia- Ferrant A, Crump M, et al. The value of the MDR1 reversal Manero G, Giles F, Faderl S, O’Brien S, Jeha S, Davis J, et al. agent PSC-833 in addition to daunorubicin and cytarabine in Phase 2 clincal and pharmacologic study of clofarabine in the treatment of elderly patients with peviously untreated patients with refractory or relapsed acute leukemia. Blood acute myeloid leukemia (AML), in relation to MDR1 status at 2003;102:/ 2379/ /2386. diagnosis. Blood 2005; (Submitted). [29] Burnett AK, Russell N, Kell WJ, Milligan D, and Culligan D. [19] van der Holt B, Sonneveld P, Burnett AK, Vossebeld PJM, A phase 2 evaluation of single agent Clofarabine as ﬁrst Beverloo B, Rosenkranz G, Dugan M, van Putten W, line treatment for older patients with AML who are not Wheatley K, Lowenberg B. Daunorubicin and Cytarabine considered ﬁt for intensive chemotherapy. Blood 2004;104: compared with Daunorubicin, Cytarabine and the MDR1 248a (Abstract) Hematology, 2005; 10 Supplement 1: 54

ACUTE MYELOID LEUKEMIA

Management of APL

RYUZO OHNO

Aichi Cancer Center, 1-1 Kanokoden, Chikusaku, Nagoya 464-8681, Japan

Acute promyelocytic leukemia (APL) has become a cularly for patients with hyperleukocytosis. If patients curable disease by all-trans retinoic acid (ATRA)- relapse hematologically or even molecularly, arsenic based induction therapy followed by two or three trioxide (ATO) will be the treatment of choice under courses of consolidation chemotherapy. Currently careful electrocardiogram monitoring. Am80, liposo- more than 90% of newly diagnosed patients with mal ATRA or gemtuzumab ozogamicin may be used APL achieve complete remission (CR) and over 70% at this stage or later. Whether ATO may be used for of patients are curable. To further increase the CR newly diagnosed APL should wait for the results of and cure rates, detection and diagnosis of this disease on-going randomized trials, and probably for the at its early stage is very important, hopefully before long-range outcome including possible secondary the appearance of APL-associated coagulopathy. malignancy due to ATO. Allogeneic stem cell trans- In induction to ATRA, concomitant chemotherapy plantation will be the treatment of choice after is indispensable, except for patients with low initial patients of age B/50 years have relapsed, provided leukocyte counts. Prophylactic use of intrathecal that they have HLA-identical family donors or DNA- methotrexate and cytarabine should be done, parti- identical unrelated donors.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389926 Hematology, 2005; 10 Supplement 1: 55/62

ACUTE LYMPHOCYTIC LEUKEMIA

A broad and integrated diagnostic work-up for a modern management of Acute Lymphoblastic Leukemia (ALL)

ROBIN FOA` , ANTONELLA VITALE, SABINA CHIARETTI & ANNA GUARINI

Division of Hematology, Department of Cellular Biotechnologies and Hematology, University ‘‘La Sapienza’’ of Rome, Italy

Acute lymphoblastic leukaemia (ALL) represents a clinical management [7]. Detailed studies of indivi- biologically and clinically heterogeneous group of dual patients need to be conducted at specialized diseases characterized by the abnormal proliferation centers, where preservation of viable cells, DNA, and accumulation of immature lymphoid cells within RNA, protein lysates, etc. is possible. An integrated the bone marrow and lymphoid tissues. Malignant biologic approach aimed at identifying prognostic transformation is a consequence of somatic mutations factors implies a coordinated effort through central in a single lymphoid progenitor cell and this mutation handling of all patients’ samples so that all the might occur at different stages of B- or T-cell necessary investigations can be consistently per- development. The diagnosis and classification of formed in each individual case and the patients can ALL is currently a multistep procedure based on be enrolled in the same therapeutic protocols. Be- morphology, cytochemistry, immunophenotype, cyto- sides, an integrated approach, using cytogenetic and genetics, molecular genetics, immunoglobulin (Ig) molecular analysis and leukemia-associated immuno- and T-cell receptor (TCR) gene rearrangements, phenotypes, can allow to identify suitable markers for multidrug resistance (MDR), genomic profiling and monitoring MDR in virtually all childhood and adult relies on the simultaneous application of multiple ALL cases [8,9]. Recent advances in genome tech- techniques [1/4]. Increasing evidence suggests that nologies have opened the way for the analysis of gene chromosomal defects and molecular abnormalities are expression profiles in human leukemias that may lead consistently present in patients with ALL, and pro- to innovative genomic-based classifications of hema- gress in our understanding of the biologic and genetic tologic malignancies, as well as to the design of characteristics of ALL has not only improved our innovative therapeutic strategies. knowledge of leukemogenesis, but has also allowed the identification of prognostic groups with specific Morphology and cytochemistry cellular and molecular features [5,6]. Despite the effort to develop risk classification systems that are ALL has been defined by the presence of more than both reproducible and comparable, a need to refine 30% lymphoblasts in the bone marrow (BM) or the ability to distinguish between higher and lower peripheral blood (PB) according to the French- risk patients still remains. In addition to their biologic American-British (FAB) Co-operative group classifi- relevance and their potential prognostic impact, cation system [10]. In the recently proposed World cytogenetic and molecular abnormalities, as well as Health Organization (WHO) classification scheme, the immunophenotypic combinations, may offer ad- [11] a blast count above 20% is sufficient for a ditional tools for the detection of minimal residual diagnosis of acute leukemia. The morphologic/cyto- disease (MRD) during the clinical course of the chemical examination recognizes three morphologic disease. Evidence suggests that a broad and integrated types: L1, L2 and L3 (Table 1). The prognostic characterization of adult ALL in the context of significance between the L1 and L2 morphologic multicenter protocols is essential for an optimal subtypes of ALL has never been fully proven; simi-

Correspondence: Robin Foa`, Division of Hematology, Department of Cellular Biotechnologies and Hematology, University ‘‘La Sapienza ’’ of Rome, Italy. E-mail: [email protected]

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331390041 56 R. Foa` et al.

Table 1. Blast cell characteristics of ALL subtypes

Pro-B Common Pre-B B T

FREQUENCY

Adults 20/25% 50% 2/5% 20/25%

Children 50/70% 90% 25% 2/5% 15% MORPHOLOGY L1/L2 L1/L2 L1/L2 L3 L1/L2

Nucleat TdT / / / / /

5’Nucleotidase / / / / /

A. phosphatase / / / / /

MPO / / / / / larly, immunophenotypic profiles that may be of is a powerful technique for the characterization of prognostic significance do not correlate with the L1 normal and neoplastic hematopoietic cells, the use of or L2 morphology. Only the L3 type of ALL still holds highly specific MoAb that recognize distinct epitopes as a distinct entity characterized by its morphology of surface and intracellular antigens has improved the and also by its unique immunophenotypic and geno- definition of the origin and level of differentiation of typic features. Unlike acute myeloid leukemia (AML), acute leukemias. There is no consensus on the best no single cytochemical test is specific for ALL; by way to report the analyzed data; it is customary to definition, however, ALL is negative for myeloperox- report percentage of blasts expressing each antigen idase (MPO) in cytochemistry studies and lacks tested and to consider any marker present on more staining with the anti-MPO monoclonal antibody than 20% of blasts as positive: the cut-off level of 20% (MoAb). According to the FAB criteria, cases of is arbitrary, however. Another point to consider in acute leukemia with more than 3% MPO positive leukemia immunophenotyping is the intensity of blasts should be classified as AML. More than 95% of antigen expression; since differences in fluorescent cases of L1 and L2 ALL are positive for terminal intensity may be important in distinguishing leukemic deoxynucleotidyl transferase (TdT) expression and its cells from normal cells and in discriminating among detection is useful in distinguishing reactive lympho- subtypes of leukemia, quantitative flow cytometry cytosis from ALL; nevertheless, TdT expression can (QFCM) may now be used to measure antigen- be found in some cases of AML. Most cases of ALL binding sites on cells more objectively and this have a characteristic localized periodic acid-Schiff approach may be useful both at diagnosis and during (PAS) staining pattern, but this finding is not specific the monitoring of MRD [15]. As shown in Table 2, B- for ALL and’block’ reactivity can be seen in some lineage ALL (70/80% of cases) can be classified into AML. Reactivity for nonspecific esterase can be four groups according to the expression of B-cell detected in a subset of ALL and is usually weaker differentiation antigens and cytoplasmic and surface than that seen in acute monocytic leukemia. Acid immunoglobulins (Ig); also T-ALL (15/25% of cases) phosphatase and alpha-naphtyl-acetate-esterase can be classified into four groups based on the level of (ANAE) reactions give a strong positivity localized thymocyte maturation and antigen expression [16]. T- to the Golgi region in more than 80% of the blast cells ALL can be further classified according to the in cases of T-lineage ALL; nowadays, however, they subtypes of T-cell receptor (TCR) molecules. are no longer routinely utilized. Even though no single Although the affiliation of ALL cases to the B- or T- cytochemical reaction is specific, cytochemistry still cell lineage is relatively easy, about 5% of cases remain represents a relevant component in the integrated difficult to classify as ALL or AML; these cases diagnostic work-up of acute leukemias that helps to coexpress several lymphoid and myeloid antigens, differentiate between ALL and AML. either on the same cells (biphenotypic leukemia) or on two different populations (hybrid leukemia). There is no consensus regarding diagnostic criteria for such Immunophenotype cases. The European Group for the Immunological The immunophenotypic characterization of blast cells Characterization of Leukemia (EGIL) [16] has sug- has several objectives: (a) lineage assignment, (b) gested the use of a scoring system based on different evaluation of cell maturation, and (c) assessment of combinations of B, T and myeloid antigen expression. phenotypic aberrations [12/14], Lineage assignment According to a strict scoring system, four groups can of blast cells by immunophenotype may still represent be identified; the most common group is that in which a major challenge in some acute leukemias; this is the blasts coexpress myeloid and B-lymphoid anti- mainly due to the cross-lineage antigen expression gens, and less commonly myeloid and T-lymphoid and it emphasizes the need to use combinations of antigens. Cases coexpressing T- and B-lymphoid several lineage-associated markers to establish the markers and those with trilineage differentiation are lineage affiliation of the blast cells. Flow cytometry rare. The clinical significance of biphenotypic acute A modern management of Acute Lymphoblastic Leukemia 57

Table 2. Immunologic classiﬁcation of All

IB-lineage ALL:

j CD19/ and/or CD79a/ and/or cyCD22/ pro-B ALL (B-I)

j CD10/ cyIg- common ALL (B-II)

j cyIg/ sIg- pre-B All (B-III)

j sIg/ mature-B ALL (B-IV) IT-lineage ALL:

/ cyCD3/ CD7/ pro-T ALL (T-I)

/ CD2/ and/or CD5/ and/or CD8/ pre-T ALL (T-II)

/ CD1a/ cortical T ALL (T-III)

/ CD1a- mCD3/ mature T ALL (T-IV)

/ Anti-TCR a/b/ a/b/ T ALL (group a) / Anti-TCR g/d/ g/d/ T ALL (group b)

Ø All with expression of one or two myeloid markers (My/ALL) leukemia has not been determined and there has been has also been recorded in a high proportion of Ph a lack of uniformity in treatment; for example, there is chromosome positive ALL. The prognostic signifi- no agreement as to whether induction therapy should cance of CD34 antigen expression in ALL, especially use anti-lymphoid and/or anti-myeloid drugs [17]. To in adult patients, is still not well established, even allow reproducible conclusions to be drawn about the though its presence does not seem to influence clinical optimal treatment of biphenotypic leukemias, in- outcome [25,26]. creased numbers of patients are required for an In conclusion, immunophenotype is an essential objective analysis; due to the rarity of the disease, component of the initial diagnostic evaluation of acute this will only be possible through multicenter studies. leukemias and is also a valuable tool for monitoring Other markers are used to identity the maturation disease after therapy and for the detection of MRD. level of the blast cells and eventually establish atypical The quantification of the level of expression of given or aberrant phenotypes indicative of specific under- antigens on the leukemic population may have ther- lying genetic lesions. A variable proportion of ALL apeutic implications; MoAb have, in fact, reached express apparently nonlineage associated markers, clinical utilization in different lymphoproliferative e.g., myeloid antigens and CD34. Expression of disorders. This applies, in particular, to antibodies myeloid associated markers in ALL is well known, directed against CD20, CD22 and CD52. All three but its clinical significance is controversial. The antigens may be expressed by ALL cells. Thus, the reported incidence of adult ALL showing myeloid percent of positivity and the degree of expression by antigen expression (My/ ALL) ranges from 15% to the leukemic population at diagnosis and at relapse is 50%, while it varies from 4% to 35% in children [18]. important when considering the potential clinical This broad variation may be related to the number of utilization of such antibodies for the management of myeloid antigens studied and their degree of lineage ALL patients. specificity, the sensitivity of the MoAb used, the cut- off level and technical factors (e.g., flow cytometric Cytogenetic and molecular analyses sensitivity and gating strategy). The most frequently expressed myeloid antigens are CD33 (/25%) and Cytogenetic and molecular analyses are important in

CD13 (/20%); CD15 and CD14 can be found in identifying prognostic markers in ALL [5,6,27,28].

/15% of ALL cases, while CD11c is rarely present The study of cytogenetic abnormalities is the basis for on ALL blasts [18,19]. Some studies have shown that unraveling molecular events that may be involved in the expression of myeloid-associated antigens is a the disease, such as the role of fusion transcripts that predictor of poor outcome both in childhood and derive from translocations, tumor suppressor genes adult ALL, while other studies have not confirmed from deletions, or the control of cell cycle regulatory this observation [20/23]. The presence of myeloid genes. There are some limitations associated with antigens can be useful in the immunologic monitoring cytogenetic studies in ALL: the leukemic cells do not of MRD. always produce good metaphases and important CD34 is the most commonly used antigen to define abnormalities can be missed. Reverse-transcriptase immature hematopoietic progenitor cells. CD34 is polymerase chain reaction (RT-PCR), DNA flow present in only 1.5% of BM cells, it is not lineage- cytometry, fluorescence in situ hybridization (FISH) restricted and its expression has been documented and comparative genomic hybridization (CGH), both in AML and ALL [24]. Overall, about 70% of among other techniques, have made it possible to ALL cases are CD34 positive. The incidence of CD34 detect more precisely the molecular and chromosomal expression is more frequent in B-lineage ALL (70/ defects in common subtypes of ALL. However,

80%) than in T-lineage ALL (20/30%); its expression detection of chromosomal abnormalities by classic 58 R. Foa` et al. karyotypic analysis or by molecular techniques has its production of a fusion RNA transcript and a chimeric advantages and inconveniences. Only through a protein (qualitative change), and those which repre- karyotypic analysis can an overall evaluation of the sent Ig/TCR rearrangement errors (quantitative whole genome be carried out and the results obtained change). Qualitative abnormalities are found to pro- can direct further investigations; on the contrary, duce functional fusion genes; one of the most molecular techniques allow detection of specific common is the t(9;22)(q34;q11) which forms the abnormalities in situations where karyotyping is BCR-ABL fusion gene; another is t(1;19)(q23;p13), difficult (e.g., insufficient metaphases or detection of where the E2A gene fuses with PBX1. The rearrange- submicroscopic abnormalities). The chromosomal ment involving the MLL gene on chromosome 11 in abnormalities in ALL can be categorized as numerical the q23 region results in a fusion gene with AF4 on in nature or structural, with or without numerical chromosome 4, band q21. Quantitative abnormalities abnormalities. Hyperdiploidy is the gain of additional result from Ig/TCR rearrangement errors which chromosomes so that the total number of chromo- juxtapose the proto-oncogene to regulatory Ig/TCR somes in a single cell exceeds 46; in ALL, this process sequences, leading to deregulated protein expression, seems to be nonrandom and two forms are usually for example the SIL-TAL1/tald deletions on chromo- distinguished: ALL with 47/51 chromosomes and some 1p32 in T-ALL [29]. A list of the main ALL with 52 or more chromosomes. Hyperdiploidy is molecular genetic abnormalities identified in ALL seen in 5/15% of cases of adult ALL and the and currently used for molecular diagnosis is reported association with a favorable outcome is less obvious in Table 3, even if this list is not exhaustive and than in childhood ALL. Hypodiploidy (chromosomes represents a compromise between the current most

B/46) is found in 2/8% of cases of ALL and is appropriate molecular method to detect or exclude an associated with a poorer outcome. The majority of abnormality and the most widely used technique. chromosomal abnormalities found in ALL are struc- Qualitative fusion transcripts predominate in B-line- tural, usually translocations. More than 30 different age ALL and recombinant errors are rare; in contrast, nonrandom translocations have been identified in they are much more frequent in T-ALL, where they ALL. Since only a relatively limited number of represent the majority of molecular abnormalities. patients have so far been studied and many of these Identification of recurring cytogenetic abnormalities translocations are uncommon, the prognostic impli- and molecular alterations in ALL has had a major cations for most of them have still to be conclusively impact on risk assessment and a number of structural defined. Most of the more common karyotypic and chromosomal changes have been incorporated structural rearrangements have been studied at the into existing classification systems [5,6]. Within our molecular level. In molecular terms, chromosomal multicenter GIMEMA ALL 0496 protocol, a central abnormalities or their submicroscopic equivalents are handling of biologic material at presentation is re- of two general types: those in which the breakpoint quired for all registered cases. This has proven feasible occurs within the involved genes, leading to the and adequate metaphases could be obtained in over

Table 3. Main chromosomal abnormalities characterized at the molecular level in ALL

Disease Gene Involved Abnormality Incidence Molecular detection §

B-ALL BCR ABL $(9;22)(q34;q11) Adults: 30% RT-PCR Childern: 3% c-MYC IgH $(8;14)(q24;q32) 1% FISH E2A PBX1 $(1;19)(q23;p13) 5% RT-PCR

E2A HLF $17;19)(q22;p13) B/1% RT-PCR

IL3 IgH $(5;14)(q31;q32) B/1% DNA-PCR

MLL AF1P $(1;11)(p32;q23) B/1% RT-PCR MLL AF4 $(4;11)(q21;q23) Adults: 5% RT-PCR Infants: 60%

MLL AF9 $(9;11)(q21;q23) B/1% RT-PCR

MLL ENL $(11;19)(q23;p13) B/1% RT-PCR

TEL AML1 $(12;21)(p13;q22) Adults: B/1% RT-PCR Children: 20% T-ALL c-MYC TCRa/d $(8;14)(q24;q11) 2% FISH

HOX11 TCRa/d $(10;14)(q24;q11) 5/10% Southern LMO1 TCRa/d $(11;14)(p15;q11) 1% Southern

LMO2 TCRa/d $(11;14)(p13;q11) 5/10% Southern SIL TAL1 Normal 1p32 Adults: 10% RT-PCR Children: 20%

TAL1 TCRa/d $(1;14)(p32;q11) 1/3% Southern

TCL1 TCRa/d inv(14)(q11;q32) B/1% FISH A modern management of Acute Lymphoblastic Leukemia 59

70% of enrolled cases. The opportunity offered by Minimal residual disease this protocol to combine molecular and cytogenetic One of the most important challenges in leukemia data in the framework of a therapeutic trial, has treatment is to accurately distinguish patients who allowed an integrated molecular-cytogenetic classifi- require more intensive (and potentially more toxic) cation to be proposed [30] which categorizes adult therapy from those in whom high cure rates can be ALL cases into subgroups which are as homogeneous achieved with less intensive therapy. MRD studies can as possible based on defined genetic alterations and provide a direct measurement of leukemic cell re- has also enabled a group of patients to be identified sponses to chemotherapy. This information can be without known cytogenetic or molecular changes. used to improve strategies of risk assessment and The results obtained by this integrated classification treatment selection in the management of ALL have enhanced the importance of a broad genetic patients. Nevertheless, before using MRD data to characterization of patients with ALL and offer a guide therapy, further analysis is required to conclu- further biologic basis for stratified treatment ap- sively establish the predictive value of MRD findings. proaches. Leukemia cells can be potentially distinguished from normal hematopoietic progenitors on the basis of Multidrug resistance morphologic and cytochemical properties, immunophenotype, karyotypic or genetic abnormalities, and Some investigators have studied MDR-1 gene expres- Ig/TCR gene rearrangements. These different char- sion in leukemic cells from ALL patients in an attempt acteristics have been exploited in an attempt to detect to demonstrate a correlation with treatment response small numbers of blasts among normal cells and a and/or patient follow-up. All studies reached the variety of techniques have been developed for the conclusion that the overexpression of the gene is detection of residual disease. The conventional cri- probably partially implicated in the chemoresistance teria for remission in patients with acute leukemia are phenomenon; this resistance can be expressed at based on the morphologic examination of BM sam- diagnosis and can also be acquired after treatment. ples and patients are considered to be in CR when Expression of MDR-1 at diagnosis has no effect on BM aspirates contain less than 5% blasts. At the time the probability of entering CR in pediatric ALL of morphologic CR, however, the extent of MRD patients, but the CR rate in adult ALL appears varies considerably. The methods for MDR analysis significantly lower in MDR positive cases compared include cytogenetics, FISH, Southern blotting, im- with MDR negative cases [31]. The MDR-1 gene munophenotype and PCR techniques. The applic- encodes for a membrane P-glycoprotein p170 (P-gp) ability of these techniques for MRD detection that acts as an adenosine triphosphate (ATP) depen- depends on three parameters: (a) specificity to dis- dent efflux pump. The expression of this gene confers criminate between malignant and normal cells without false positive results), (b) sensitivity limit of at resistance to some chemotherapeutic agents such as 3 vinca alkaloids, anthracyclines, etc. However, the least 10 , and (c) reproducibility and applicability exact prognostic significance of this resistance me- (easy standardization and rapid collection of results for clinic application) [8,9,35,36]. Only a proportion chanism is still unclear [32]. Several studies have of leukemias have specific markers such as chromo- found a correlation between high P-gp expression somal translocations, e.g., t(9;22), t(4;11) or t(1;19), levels and/or P-gp function, and poor response to and conventional karyotypic analysis may be used to chemotherapy in AML [33] and, to a lesser extent, in monitor MRD if an abnormal karyotype is present at ALL [31,34]. One of the main reasons for these diagnosis; however, its low specificity and the risk of contradictory results may be methodologic problems analyzing metaphases from normal cells represent in P-gp detection. The methods currently used are: major obstacles in its routine use. The main advantage (a) measurement of P-gp function by efflux studies, of FISH is that it provides interpretable information (b) P-gp expression levels by MoAb, and (c) MDR-1 based on interphase cells with a low proliferative rate. gene expression encoding for (P-gp) by RT-PCR. Nonetheless, the sensitivity of FISH analysis for MRD However, each of these methods has disadvantages. monitoring is limited. Immunophenotyping techni- MDR-1 detection could thus represent a valuable ques using multicolor-gated flow cytometry are based biologic parameter in the diagnostic screening of ALL on the aberrant expression of antigens by the leukemic patients. The inclusion of this parameter may result in cell population and on the identification of markers the design of biologically based risk adapted thera- that may be found on malignant cells in combinations peutic strategies for the management of adult ALL. that are normally not observed in normal BM and PB The adoption of protocols based on drugs that are not cells. For a productive detection of MRD in ALL, it is P-gp substrates may offer therapeutic advantages for necessary to distinguish leukemic lymphoblasts from CR achievement for ALL patients expressing the their normal counterparts and the intensity of expres- MDR-1 protein. sion may also help in distinguishing leukemic cells 60 R. Foa` et al. from normal progenitors [15,37]. Some immunophe- Gene expression profiling notypic combinations present on leukemic cells are Genomic profiling is becoming a reality that may confined to certain normal tissues and are not found profoundly modify the management of ALL patients. in normal BM or PB cells. Overall, flow cytometry can Several studies have identified unique gene expression be utilized to monitor MRD in about 85 /90% of signatures characteristic of various hematologic and cases. non-hematologic cancers [39,40]. The potential ex- The detection of leukemia-associated clonal genetic ploitation of microchip analysis is manifold: it can changes at the karyotypic and genetic levels has been define the genetic signature of given neoplastic extensively tested by molecular biology techniques, populations, it can help define the lineage affiliation based on PCR analysis. Two types of PCR targets can of the tumor, it can identify sets of genes that be used to detect MRD in ALL patients: leukemia- characterize subsets of patients with distinct responses specific breakpoint fusion regions of chromosome to treatment and, ultimately, have a prognostic rearrangements (translocations, deletions or inver- impact, it may identify new targets for future therapies sions) or junctional regions of leukemia clone-specific based on the under- or overexpression of given genes, rearranged Ig/TCR genes. The presence, at diagnosis, it may allow definition of drug susceptibility or of one of these transcripts allows the monitoring of resistance, etc. These innovative technologies have MRD during the clinical follow-up (e.g., BCR-ABL, been recently utilized in both childhood and adult

ALL-AF4, E2A-PBX1). The Ig heavy chain genes ALL [41/44]. In pediatric ALL, it has been shown

IgH undergo rearrangements in 90/95% of B-lineage that distinct gene expression profiles could be found ALL cases. TCR gene rearrangements occur in 95% in each of the prognostically important leukemia of T-lineage ALL and in 50/70% of B-lineage ALL subtypes, based on immunophenotypic and cytoge- [29]. Because such rearrangements are clonal, analy- netic/ molecular features [41]. In a study dedicated sis of Ig and TCR gene configurations can be used to specifically to childhood T-ALL, microchip analysis track the persistence of malignant clones whose could identify previously unrecognized molecular rearrangements have been determined at diagnosis. subtypes of T-ALL and associate the activation of One of the aims of MRD investigations is to estimate particular oncogenes to defined stages of normal the amount of residual tumor rather than to establish thymocyte development [42]. Hierarchical clustering its presence and current PCR methods do not allow of all adult ALL samples based on gene expression an easy and accurate MRD quantification. More profile identified two well-defined groups which recently, real-time quantitative PCR (RQ-PCR) has correlated precisely with the T- or B-cell immunophe- been used for MRD detection in ALL [38]. This notype of the leukemic cells. Further analysis identi- method exploits the 5’-3’ nuclease activity associated fied gene expression profiles associated with the with Taq polymerase and uses a fluorogenically presence of either ALL1-AF4, BCR-ABL or E2A- labeled target-specific DNA probe; this probe is PBX1 gene rearrangements. Furthermore, an integrated analysis of childhood and adult ALL highlights designed to anneal between the forward and reverse a strong similarity between cases which harbor spe- oligonucleotide primers used for PCR amplification. cific rearrangement regardless of patient’s age [44]. The RQ-PCR technique appears to be a sensitive, With the use of these technologies, it has been shown reproducible and quick method for quantifying that genetically defined subgroups express different MRD. The greatest obstacle to the routine use sets of genes. In individual cases, the genetic lesion of MRD studies in ALL therapy protocols is that could be classified by microarray analysis, while being none of the techniques currently available for MRD negative by RT-PCR [41]. Evidence has also been detection can be applied to all patients. Because PCR provided that the lineage affiliation of rare cases with may detect residual leukemic cells in cases not unique phenotypic features may be clarified on the amenable to flow cytometric investigation, and vice- basis of the genomic profile [45]. Times are mature to versa, it is possible to apply the two techniques in verify whether these innovative technologies will tandem. Correlative studies have demonstrated that change our approach to the characterization of detection of MRD by flow cytometry or by PCR leukemias. Should this be the case, it is foreseeable analysis of leukemia-specific markers is strongly that in the near future all new cases will undergo a associated with subsequent relapse. It should again rapid gene chip analysis that may possibly substitute be stressed that all the above can be attempted only many of the analyses routinely carried out nowadays through a broad, integrated and uniform character- through the efforts of numerous laboratories. As our ization of all cases enrolled in multicenter studies. knowledge increases, it is also likely that ad hoc chips Only through such a broad effort it will be possible to will be designed that contain a number of genes conclusively define the role and impact of biologically sufficient for an adequate classification of given based MRD monitoring for the management of ALL diseases, rather than using the broad chips presently patients. available. It appears realist to hypothesize that we are A modern management of Acute Lymphoblastic Leukemia 61 in the verge of a time when subgroups of patients will [10] Bennett JM, Catovsky D, Daniel MT, et al. Proposals for the be classified on the basis of the genomic profile and classiﬁcation of acute leukaemias. French-American-British (FAB) co-operative group. Br J Haematol 1976;33:/ 451/ /458. that the latter will direct the therapeutic strategy in [11] Harris NL, Jaffe ES, Diebold J, et al. World Health Organiza- terms of both drug decisions and treatment aggres- tion classiﬁcation of neoplastic s of the hematopoietic and siveness. Finally, in the very near future, great lymphoid tissues: Report of the clinical advisory committee

attention will be paid to the identification of new meeting-House, Virginia, November 1997. J Clin Oncol 1999;/ therapies aimed at targeting the specific regulatory 17:3835/ /3849. pathways operational in the different leukemia sub- [12] Huh YO, Ibrahim S. Immunophenotypes in adult acute

lymphocytic leukemia. Hematol/Oncol Clin North Am 2000;/ types. 14:1251/ /1265. [13] Czuczman MS, Dadge RK, Stewatr CC, et al. Value of Conclusions immunophenotype in intensively treated adult acute lympho-

blastic leukemia: Cancer and Leukemia Group B. Blood 1999;/ ALL are classified on the basis of the presumed cell of 93:3931/ /3133. origin and do not represent a single disease but rather [14] Ludwig WD, Raghavachar A, Thiel E. Immunophenotypic classification of acute lymphoblastic leukaemia. Baillieres Clin a heterogeneous collection of diseases with different Haematol 1994;7:/ 235/ /262. genetic profiles and differences in clinical progression, [15] Farahat N, Morilla A, Owsu-Ankomah K, et al. Detection of treatment and outcome. Following a diagnostic work- minimal residual disease in B-lineage acute lymphoblastic up, prognostic data are routinely achieved through leukaemia by quantitative flow cytometry. Br J Haematol physical examination, serum biochemical profiles, 1998;101:/ 158/ /164. peripheral blood count and bone marrow morphol- [16] Bene MC, Castoldi G, Knapp W, et al. Proposals for the immunological classification of acute leukemias. Leukemia ogy. Over the years, information obtained through 1995;9:/ 1783/ /1786. karyotype, molecular genetics, extensive immunophe- [17] Killick S, Matutes E, Powles RL, et al. Outcome of bipheno- notype, multidrug resistance and, more recently, tipyc acute leukemia. Haematologica 199;84:/ 699/ /706. genomic profiling is progressively contributing to a [18] Khalidi HS, Chang KL, Medeiros LJ, et al. Acute lympho- better understanding of the biology of this complex blastic leukemia. Survey of immunophenotype, French-Amer- disease, to the identification of subgroups of patients ican-British classification, frequency of myeloid antigen expression, and karyotypic abnormalities in 210 pediatric with a different clinical outcome, to the more precise and adult cases. Hematopathology 1999;111:/ 467/ /476. monitoring of MRD, to the use of different therapeu- [19] Nakase K, Kenkichi K, Shiku H, et al. Myeloid antigen, tic protocols based on prognostic indicators and, CD13, CD14, and/or CD33 expression in adult acute recently, also to the design of innovative and specific lymphoblastic leukemia patients: Diagnostic and prognostic treatment strategies. implication. Am J Clin Pathol 1996;105:/ 761/ /768. [20] Pui CH, Rubnitz JE, Hancock ML, et al. Reappraisal of the clinical and biologic significance of myeloid-associated antigen expression in childhood acute lymphoblastic leukemia. J Clin References Oncol 1998;16:/ 3768/ /3773. [21] Preti HA, Huh YO, O’Brien SM, et al. Myeloid markers in [1] Pui C-H, Evans WE. Acute lymphoblastic leukemia. N Engl J adult acute lymphocytic leukemia. Cancer 1995;76:/ 1564/ / Med 1998;339:/ 605/ /615. 1570. [2] Bassan R, Gatta G, Tondini C, Willemze R. Adult acute [22] Sobol RE, Mick R, Royston I, et al. Clinical importance of

lymphoblastic leukemia. Crit Rev Oncol/Hematol 2004;50:/ / myeloid expression in adult acute lymphoblastic leukemia. N 223 /261. Engl J Med 1987;316:/ 1111/ /1117. [3] Hoelzer D, Gokbuget N. Recent approaches in acute lympho- [23] Lauria F, Raspadori D, Martinelli G, et al. Increased expres-

blastic leukemia in adults. Crit Rev Oncol/Hematol 2000;36:/ / sion of myeloid antigens markers in adult acute lymphoblastic 49/58. leukaemia patients: Diagnostic and prognostic implications. [4] Cortes JE, Kantarjian HM. Acute lymphoblastic leukemia. A Br J Haematol 1994;87:/ 286/ /292. comprehensive review with emphasis on biology and therapy. [24] Krause DS, Fackler MJ, Civin CI, et al. CD34: structure, Cancer 1995;76:/ 2393/ /2417. [5] Ferrando AF, Look T. Clinical implications of recurring biology and clinical utility. Blood 1996;87:/ 1/ /13. chromosomal and associated molecular abnormalities in acute [25] Thomas X, Archimbaud E, Charrin C, et al. CD34 expression is associated with major adverse prognostic factors in adult lymphoblastic leukemia. Semin Hematol 2000;37:/ 381/ /395. [6] Group Francais de Cytogenetique Hematologique.. Cytoge- acute lymphoblastic leukemia. Leukemia 1995;9:/ 249/ /253. netic abnormalities in adult acute lymphoblastic leukemia: [26] Vanhaeke D, Bene MC, Garand R, et al. Expression and long- Correlations with hematologic findings and outcome. A term prognostic value of CD34 in childhood and adult acute collaborative study of the Group Francais de Cytogenetique lymphoblastic leukemia. Leuk Lymphoma 1995;20:/ 137/ /142. [27] Faderl S, Kantarjian HM, Talpaz M, et al. Clinical signifi- Hematologique. Blood 1996;87:/ 3135/ /3142. [7] Foa` R, et Vitale A. Towards an integrated classification of cance of cytogenetic abnormalities in adult acute lymphoblas- adult acute lymphoblastic leukemia. Rev Clin Exp Hematol tic leukemia. Blood 1998;91:/ 3995/ /4019. [28] Secker-Walker LM, Prentice HG, Durrants J, et al. Cytoge- 2002;6:/ 181/ /199. [8] Campana D, Pui CH. Detection of minimal residual disease in netics add independent prognostic information in adults with acute leukemia: Methodologic advances and clinical signifi- acute lymphoblastic leukaemia on MRC trial UKALL XA. Br / / cance. Blood 1995;85:/ 1416/ /1434. J Haematol 1997;96:601 /610. [9] Stock W, Estrov Z. Studies of minimal residual disease in acute [29] Macintyre EA, Delabesse E. Molecular approaches to the

lymphocytic leukemia. Hematol/Oncol Clin North Am 2000;/ diagnosis and evaluation of lymphoid malignancies. Sem 14:1289/ /1305. Hematol 1999;36:/ 373/ /389. 62 R. Foa` et al.

[30] Mancini M, Scappaticci D, Cimino G, et al. A comprehensive disease in acute lymphoblastic leukemia. J Clin Oncol 1998;/ genetic classiﬁcation of adult acute lymphoblastic leukemia 16:3774/ /3778.

(ALL): analysis of the GIMEMA 0496 protocol. Blood 2005;/ [38] Donovan JW, Ladetto M, Zou G, et al. Immunoglobulin 105:3434/ /3441. heavy-chain consensus probes for real-time PCR quantifica- [31] Goasguen JE, Dossot JM, Fardel O, et al. Expression of the tion of residual disease in acute lymphoblastic leukemia. Blood multidrug resistance-associated P-glycoprotein (P-170) in 59 2000;95:/ 2651/ /2658. cases of de novo acute lymphoblastic leukemia: prognostic [39] Golub TR, Slonim DK, Tamayo P, et al. Molecular classification of cancer: Class discovery and class prediction by gene implications. Blood 1993;81:/ 2394/ /2398. [32] Hegewisch-Becker S, Hossfeld DK. The MDR phenotype in expression monitoring. Science 1999;286:/ 531/ /537. hematologic malignancies: Prognostic relevance and future [40] Alizadeh AA, Eisen MB, Davis RE, et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression perspective. Ann Hematol 1996;72:/ 105/ /117. [33] Legrand O, Simonin G, Perrot JY, et al. Pgp and MRP profiling. Nature 2000;403:/ 503/ /511. [41] Yeoh E, Ross ME, Shurtleff SA, et al. Classification, subtype activities using calcein-AM are prognostic factors in adult discovery, and prediction of outcome in pediatric acute acute myeloid leukemia patients. Blood 1998;91:/ 4480/ /4488. lymphoblastic leukemia by gene expression profiling. Cancer [34] Tafuri A, Gregorj C, Petrucci MT, et al. MDR1 protein Cell 2000;1:/ 133/ /143. expression is an independent predictor of complete remission [42] Ferrando A, Neuberg D, Staunton J, et al. Gene expression in newly diagnosed adult acute lymphoblastic leukemia. Blood signatures define oncogenic pathways in T cell acute lympho- 2002;100:/ 974/ /981. blastic leukemia. Cancer Cell 2002;1:/ 75/ /87. [35] Van Dongen JJM, Macintyre EA, Gabert JA, et al. Standar- [43] Chiaretti S, Li X, Gentleman R, et al. Gene expression profile dized RT-PCR analysis of fusion gene transcripts from of adult T-cell acute lymphocytic leukemia identifies distinct chromosome aberrations in acute leukemia for detection of subsets of patients with different response to therapy and minimal residual disease. Report of the BIOMED-1 Con- survival. Blood 2004;103:/ 2771/ /2778. certed Action: Investigation of minimal residual disease in [44] Chiaretti S, Li X, Gentleman R, et al. Gene expression profiles acute leukemia. Leukemia 1999;13:/ 1901/ /1928. of B-lineage adult acute lymphocytic leukemia reveal genetic [36] Foroni L, Harrison CJ, Hoffbrand AV, et al. Investigation of patterns that identify lineare derivation and distinct mechan- minimal residual disease in childhood and adult acute ism of transformation. Clin Cancer Res 2005; in press. lymphoblastic leukaemia by molecular analysis. Br J Haematol [45] Testa U, Torelli GF, Riccioni R, et al. Human acute stem cell 1999;105:/ 7/ /24. leukemia with multilineage differentiation potential via cas-

[37] Ciudad J, San Miguel JF, Lopez-Berges MC, et al. Prognostic cade activation of growth factors receptors. Blood 2002;99:/ /

value of immunophenotypic detection of minimal residual 4634/4637. Hematology, 2005; 10 Supplement 1: 63

STEM CELL TRANSPLANTATION

Current issues in allogeneic stem cell transplantation

GE´ RARD SOCIE´

Hospital Saint Louis, APHP & University Paris VII

Abstract Hematopoietic stem cell transplantation (SCT) has evolved as a central treatment modality in the management of different hematological malignancies. However, 2 major issues still impact overall mortality and morbidity post transplant.

1. Despite adequate post-transplantation immunosuppressive therapy, acute graft-versus-host disease (GvHD) remains a major cause of morbidity and mortality in the hematopoietic stem cell transplantation setting, even in patients who receive human HLA identical sibling grafts. Up to 30% of the recipients of stem cells or bone marrow transplantation from HLA- identical related donors and most patients who receive cells from other sources (matched-unrelated, non-HLA-identical siblings, cord blood) will develop Grade II or more acute GvHD despite immunosuppressive prophylaxis. Thus, GVHD continues to be a major limitation to successful hematopoietic stem cell transplantation. 2. With improvements in outcome, an increasing number of patients survive free of the disease for which they were treated. Today, about 60% of patients will survive 5 years after diagnosis. Therefore, immediate survival is no longer the sole concern. The aim of the allogeneic SCT now is to cure a patient’s underlying disease and, at the same time, to minimize the incidence of post-treatment complications and ensure the best possible long term quality of life. The long time span between initial therapy and late effects, the multiple factors influencing cancer related health risk and the unknown effect of treatment on aging are common characteristics of late effects. While the treatment strategy of a cancer patient depends widely on the type and extension of the disease, considerations for a long term survivor depend much more on the type of treatment applied, age of the patient, its general health status, as well as his familial and social integration.

I will discuss, based on the most recent knowledge, advances and current issues in GvHD and some typical examples of late effects in survivors, and the practical recommendations that could assist practitioner and patient decision about appropriate healthcare for specific clinical circumstances.

Correspondence: Professor Ge´rard Socie´, Service d’He´matologie Greffe de Moelle & Unite´ INSERM U728/Universite´ Paris VII, Hoˆpital Saint Louis, 1 Avenue Claude Vellefaux, 75475, Paris Cedex 10, France. Tel: 33 1 42 49 98 24. Fax: 33 1 42 41 14 70. E-mail: [email protected]

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389953 Hematology, 2005; 10 Supplement 1: 64/65

STEM CELL TRANSPLANTATION

Current update on reduced intensity regimens for allogeneic stem cell transplantation

DIDIER BLAISE, MOHAMAD MOHTY, CATHERINE FAUCHER, DANIEL OLIVE, & CHRISTIAN CHABANNON

Institut Paoli-Calmettes, 232 Boulevard de Ste. Marguerite, 13273 Marseille, Cedex 9, France

Beside chemotherapy, immunotherapy represents an and identical incidence acute GVHD in comparison approach based on the more or less specific recogni- to bone marrow, despite a higher number of immu- tion of tumoral cells. Allogeneic stem cell transplanta- nocompetent cells infused [1]. However chronic tion (ASCT) has been used now for more than 3 GVHD appears to be more frequent and severe decades in the treatment of malignant hemopathies. [1,2]. The second step considered the preparative or Several lines of evidence could show that donor- so-called conditioning regimen. While it has been derived immuno-competent effectors can exert a postulated for many years that transplant conditioning potent long lasting graft-versus-tumor activity. Today, has to be myeloablative to allow long-term engraft- this approach represents the most efficient and the ment of allogeneic cells, different investigators suc- most widely form of immunotherapy used worldwide. ceeded to perform allogeneic transplantation after a However allogeneic transplantation suffers from non-myeloablative (NMA) or reduced intensity regi- major weakness: Despite a common HLA identity mens (RIC) (Storb, 1997 #545) [3,4]. Early after the between matched donor and recipient, the allogeneic first reports, we started to develop our own strategy recognition of some normal tissues, expressing epithe- using a combination of fludarabine, low dose busulfan lial antigens, is associated with a high rate of and anti-thymocyte globulin (ATG), as initially de- morbidity and mortality. Usually, the indications of scribed by Slavin et al. Our initial work was focused allogeneic transplantation are still limited to the on the optimization aspects [5,6]. We have now in youngest patients (age under 45 or 50). Because of hands a well defined platform allowing rapid engraft- its high toxicity, ASCT has also been limited for years ment, lower supportive care [7], low transplant to diseases for which a potent antitumoral effect had related-infections [8] and mortality while preserving been documented in the early years. Interestingly, the a high anti-tumor effect [5]. Simultaneously, we world experience of this approach in malignant investigated the applicability of this strategy to new diseases relies mainly on patients with leukemia populations of cancer patients. We could confirm the (chronic or acute) and younger than 50 years which feasibility in older populations essentially between 50 represent only a part of the wide spectrum of cancer and 55 years [5], and participated to the demonstra- population. tion of its activity in non-leukemic malignancies such Given the strength and curative potential of allo- as multiple myeloma (MM) [9], ovarian carcinoma geneic transplantation, new strategies aiming to widen [10] and different other metastatic solid tumors [11]. the use of this approach to other cancers, were These clinical achievements leaded us to some im- urgently needed. On these bases, we and others portant biological observations related to post-graft started by the end of the 90s to develop new anti-infectious activity [8], some aspects of the approaches that would allow overcoming the above GVHD reaction [12] and to study post graft immune cited limitations: The first step has been the switch reconstitution [13/15] from bone marrow to peripheral blood stem cells Presently, with the dramatic reduction of trans- (PBSC) as source of graft [1]. We established that plant-related mortality, ASCT has probably entered PBSC transplantation is associated with a quicker the era of global immunotherapy applicable to the hematological recovery, lower transplant-related cost largest number of patients. Despite these tremendous

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389971 Current issues in allogeneic stem cell transplantation 65 advances achieved in less than 5 years, further References investigations are urgently needed in 2 major direc- [1] Blaise D. J Clin Oncol 2000;18:/ 537/ /546. tions: [2] Mohty M. Blood 2002;100:/ 3128/ /3134. [3] Giralt S. Blood 1997;89:/ 4531/ /4536. . Anti-tumor effect exists but tumoral escape is [4] Slavin S. Blood 1998;91:/ 756/ /763. still the major cause of failure of the procedure in [5] Mohty M. Blood 2003;102:/ 470/ /476. [6] Faucher C. Exp Hematol 2003;31:/ 873/ /880. patients with advanced diseases: further work [7] Ivanov V. Transfusion 2004;44:/ 501/ /508. and novel complementary approaches are neces- [8] Mohty M. Leukemia 2003. sary to further enhance the allogeneic anti-tumor [9] Mohty M. Bone Marrow Transplant 2004;34:/ 77/ /84. effect. [10] Bay JO. Bone Marrow Transplant 2002;30:/ 95/ /102. . Procedure-related mortality has decreased nota- [11] Blaise D. Blood 2004;103:/ 435/ /441. [12] Mohty M. Leukemia 2003;17:/ 869/ /875. bly but still exists: optimization of the procedure [13] Mohty M. Leukemia 2002;16:/ 2197/ /2204. is still to be pursued. This is notably the case in [14] Mohty M. Hematology 2002;7:/ 157/ /164. the oldest patients. [15] Mohty M. Leukemia 2005;19:/ 1/ /6. Hematology, 2005; 10 Supplement 1: 66/72

ADVANCES IN MOLECULAR HEMATOLOGY

Split-signal FISH for detection of chromosome aberrations

J. J. M. VAN DONGEN, M. VAN DER BURG, & A. W. LANGERAK

Department of Immunology, Erasmus MC, Rotterdam, The Netherlands

Abstract Chromosome aberrations are frequently observed in hematopoietic malignancies. These aberrations can deregulate expression of an oncogene, resulting in aberrant expression or overexpression, or they can form leukemia-specific chimeric fusion proteins. Detection of chromosome aberrations is an important tool for classification of the malignancy and for the definition of risk groups, which need different treatment protocols. We developed rapid and sensitive split-signal fluorescent in situ hybridization (FISH) assays for frequently occuring chromosome aberrations. The split-signal FISH approach uses two differentially labeled probes, located in one gene at opposite sites of the breakpoint region. In normal karyotypes, two co- localized green/red signals are visible, but a translocation results in a split of one of the co-localized signals. Split-signal FISH has three main advantages over the classical fusion-signal FISH approach, which uses of two labeled probes located in two genes. First, the detection of a chromosome aberration is independent of the involved partner gene. Second, split-signal FISH allows the identification of the partner gene or chromosome region if metaphase spreads are present, and finally it reduces false-positivity.

Keywords: Fluorescence in situ hybridization (FISH), split-signal FISH, chromosome aberrations, fusion gene, peptide nucleic acid, leukemia, lymphoma

Chromosome aberrations in hematopoietic sion of aberrant fusion proteins are particularly found malignancies in precursor-B-ALL, acute myeloid leukemias (AML), and chronic myeloid leukemia (CML). The Chromosome aberrations play an important role in fusion proteins have functional features that differ hematological malignancies [1]. Most of these aberra- from the corresponding wild type proteins and mostly tions concern balanced translocations involving genes play a role in oncogenesis. In addition to the new that play key roles in the development and function of features of the fusion protein, loss of wild type activity hematopoietic cells, such as transcription factors, cell due to the translocation (in some translocations cycle regulators, and signal transduction molecules. enhanced by deletion of the second allele) might Chromosome translocations can result in deregulated contribute to oncogenesis. expression of (onco)genes as a direct consequence of a Several clinical studies have demonstrated that translocation to a regulatory element, e.g., an im- chromosomal translocations are useful markers con- munoglobulin (Ig) or T-cell receptor (TCR) enhancer tributing to classification of the malignancies and to [2,3]. Ig and TCR-gene related chromosome aberra- the definition of risk groups, that need different tions are particularly found in mature B-cell malig- treatment protocols. In precursor-B-ALL MLL gene nancies, such as various types of B-cell non-Hodgkin translocations and t(9;22) with the BCR-ABL fusion lymphomas (B-NHL), and in immature T-cell malig- gene are associated with a poor prognosis, while nancies, mainly T-cell acute lymphoblastic leukemias t(12;21) with the TEL-AML1 fusion gene is asso- (T-ALL). Alternatively, translocations can result in ciated with good outcome. Analogously, in AML fusion of two genes that encode leukemia-specific inv(16) with CBFB-MYH11, t(8;21) with AML1- chimeric (fusion) proteins. Fusion genes with expres- ETO, and t(15;17) with PML-RARA are associated

Correspondence: Prof. J. J. M. van Dongen, , M.D., Ph.D., Dept. of Immunology, Erasmus MC, Dr. Molewaterplein 50 3015 GE Rotterdam, The Netherlands. Tel: /31-10-4088094. Fax: /31-10-4089456. E-mail: [email protected]

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389980 Split-signal FISH for translocations in ALL 67 with good prognosis, while 11q32 (MLL gene) chromosome translocations is limited, because the aberrations show poor outcome in AML as well [1]. breakpoints in many translocations are scattered over

large regions (/25kb). Nevertheless, Southern blot analysis has proven to be useful for detection of MLL Detection of chromosome aberrations translocations. As the MLL gene can have many Several techniques can be used for the detection of translocation partner genes and the breakpoint region chromosome aberrations, each having its inherent is relatively small (6.5 kb), Southern blot analysis is advantages and disadvantages (Table I). An advantage suitable for the detection of MLL rearrangements, of conventional karyotyping is that it is highly informa- independent of the partner gene [5,6]. Southern tive as virtually all abnormalities can be detected. This blotting has also been used for detection of E2A includes not only structural abnormalities, but also (ETV6) gene rearrangements, since the majority of numerical abnormalities such as hypo-, or hyper- E2A (ETV6) breakpoints are located in a breakpoint ploidy. However, the interpretation may be difficult region of 15 kb [7]. if the karyotype is complex. Another disadvantage is PCR analysis on the DNA level is relatively easy for that for some samples no reliable results can be detection of SIL-TAL1 fusion genes [8,9]. but much obtained because of a low mitotic index or poor more complex for other translocations, mainly be- chromosome morphology. In addition, some chromo- cause PCR analysis needs multiple primers, if geno- some abnormalities are cryptic, i.e. they cannot be mic breakpoint regions are larger than 2 to 4 kb [10/ identified via conventional karyotyping, because 12]. changes in chromosome banding patterns are too An alternative approach, which is suitable for marginal to be detected, such as t(12;21), t(5;14) and detection of chromosome translocations resulting in SIL-TAL1 fusions [4]. formation of fusion genes, is detection of fusion genes or Chromosome aberrations can also be identified via fusion gene transcripts via (nested) RT-PCR analysis Southern blot or PCR analysis on genomic DNA. South- [13]. The advantage of this approach is that it reaches ern blotting is considered to be technically demanding sensitivities of 1 cell in 104 to 1 cell in 106 cells, and laborious and the applicability to the detection of enabling detection of minimal residual disease Table I. Techniques for detection of chromosome aberrations

Technique Advantages Disadvantages

Karyotyping Detection of virtually all abnormalities Interpretation may be difficult if the karyotype (structural and numerical abnormalities) is complex Sometimes low mitotic index or poor chromosome morphology Some chromosome abnormalities are cryptic Southern blotting Detection of chromosome aberration independent Technically demanding and laborious of partner gene/chromosome region Detection of chromosome aberrations is limited,

if breakpoints are scattered over large regions (/25 kb) Focused on a specific type of aberration, determined by the probe PCR Can be easy for some chromosome Detection of chromosome translocations is limited, if aberrations e.g. SIL-TAL1 gene fusion breakpoints are scattered over large regions (2 to 4 kb) Sensitive, allowing detection of minimal Variant translocations can easily be missed, if these variants residual disease are not covered by the used primers RT-PCR Suitable for detection of chromosome Variant translocations can easily be missed, if these aberrations with formation of fusion genes variants are not covered by the used primers Sensitive, allowing detection of for minimal residual disease FISH general Dividing cells (metaphase nuclei) and non-dividing Focused on a specific type of aberration, determined by cells (interphase nuclei) can be analyzed the applied probe set Detection of cryptic aberrations fusion-signal Detection of the fusion of two partner genes Translocations with other partner genes are missed involved in the translocation

5/10% false-positivity split-signal Detection of chromosome aberration independent of Partner gene/chromosome region can not be identified if partner gene/chromosome region metaphase spreads are absent Identification of partner gene/chromosome region if metaphase spreads are present Minimization of false-positivity 68 J. J. M. van Dongen et al.

Table II. Examples of classical chromosome aberrations, which are easily detectable by split-signal FISH.

Target gene for Detectable chromosomes Involved Relative frequency per Disease category split-signal FISH aberration genes disease category

Precursor-B-ALL E2A (TCF3) (1;19) (q23;p13) E2A-PBX1 5/6%

t(17;19) (q22;p13) E2A-HLF /1% MLL t(4;11) (q21;q23) MLL-AF4 4% (40% of infant ALL) t(9;11) (p22;q23) MLL-AF9 rare (10% of infant ALL) t(11;19) (q23;p13) MLL-ENL 1% (25% of infant ALL)

TEL (ETV6) t(12;21) (p13;q22) TEL-AML1 25% (childhood); B/2% (adults)

BCR t(9;22) (q34;q11) BCR-ABL 4/7% (childhood); 25/45% (adults)

T-ALL TAL1 del (1) (p32;p32) SIL-TAL1 20/25% (childhood); 10% (adults) t(1;14) (p32;q11) TCRD-TAL1 3%

HOX11L2 (TLX3) t(5;14) (q34;q32) TCRD-HOX11L2 15/20% LMO2 (RBTN2) t(11;14) (p13;q11) TCRD-LMO2 7%

CALM t(10;11) (p13;q14) CALM-AF10 /10%

B-NHL CCND1 (BCL1) t(11;14) (q13;q32) IGH-CCND1 /90% of MCL

BCL2 t(14;18) (q32;q21) IGH-BCL2 /80% of FCL; 25% of DLBCL

MYC t(8;14) (q24;q32) IGH-MYC /90% of Burkitt lymphoma

t(2;8) (p11;q24) IGK-MYC /5% of Burkitt lymphoma

t(8;22) (q24;q11) IGL-MYC /5% of Burkitt lymphoma

MALT1 t(11;18) (q21;q21) MALT1-API2 25/50% of MALT lymphoma

PAX5 t(9;14) (p13;q32) IGK-PAX5 /50% of LPL

BCL6 t(3;14) (q27;q32) IGH-BCL6 10/20% of DLBCL

BCL10 t(1;14) (p22;q32) IGH-BCL10 5/10% of MALT lymphoma

ALK t(2;5) (p13;q35) NPM-ALK /75% of ALCL

t(1;2) (q21;p13) TPM3-ALK /15% of ALCL

Abbreviations: MCL, mantle cell lymphoma; FCL, follicular lymphoma; DLBL, diffuse large B-cell lymphoma; ALCL, anaplastic large cell lymphoma; MALT, mucora associated lympoid tissue; LPL, lymphoplasmacytic lymphoma (immunocytoma).

(MRD) [13/15]. A disadvantage of PCR-based red and green, which flank the breakpoint regions of methods is that variant translocations can more easily the two genes, which are involved in the translocation be missed, if these variants are not covered by the (Figure 1A). In normal karyotypes, i.e. without used primers. chromosome aberration, two red signals and two Detection of chromosome translocations with fu- green signals are detectable. In case of a translocation, sion genes can also be performed at the protein level a red and a green signal will be juxtaposed giving rise via the specific detection of the fusion proteins. This to a co-localized green/red signal, which will generally technique, however, has not yet been implemented in appear as a yellow signal. In addition, separate green routine diagnostics, due to lack of appropriate anti- and red signals of the unaffected chromosomes will be bodies that specifically detect only the fusion protein visible. and not the two wild type proteins of which the fusion The split-signal FISH approach also uses two protein is composed. differentially labeled probes, but these probes are Fluorescence in situ hybridization (FISH) is a mole- located in only one of the two involved genes, here- cular cytogenetic technique, which uses fluorescently after called the target gene, and are positioned at labeled probes for detection of specific chromosome opposite sides of the breakpoint region of the target aberrations. The advantage of this technique is that gene (Figure 1B) [18/20]. In normal karyotypes, two besides dividing cells (metaphase nuclei) also non- co-localized green/red signals usually appearing yel- dividing cells (interphase nuclei) can be analyzed, low will be visible. A translocation will result in a split which allows a rapid screening of a large number of of one of the co-localized signals, resulting in a cells even if the malignant clone did not divide under separate green and red signal together with a fused culture conditions. In addition, also cryptic aberra- signal of the unaffected chromosome [19,20]. tions can be detected [16,17]. A disadvantage of The split-signal FISH approach has several advan- FISH analysis compared to cytogenetics is that this tages over the more traditional fusion-signal FISH. technique is focused on a specific type of aberration, First, the detection of a translocation is independent determined by the applied probe set. of the involved partner gene. This is particularly of great interest for target genes with multiple partner genes such as MLL and E2A (ETV6) (Figure 2) [19]. Fusion-signal FISH versus split-signal FISH Although the detection is independent of the involved There are two main approaches of FISH probe design partner gene or partner chromosome, split-signal for use on (interphase) nuclei, i.e. fusion-signal FISH FISH in principle allows the identification of the and split-signal FISH [18]. The classical fusion-signal partner chromosome, if metaphase spreads are pre- FISH approach uses two differentially labeled probes, sent on the slide. As a result of the translocation, one Split-signal FISH for translocations in ALL 69

Figure 1. Differences between fusion-signal FISH and split-signal FISH. A. Fusion-signal FISH with two probes located in the two genes, which are involved in the chromosome translocation. In normal situations, two green and two red signals will be present. In case of a translocation, a green and a red signal co-localize generally appearing as rise to a yellow signal together with the separate green and red signals of the unaffected genes. B. Spit-signal FISH with two probes positioned at opposite sides of the breakpoint region in one of genes, which are involved in the chromosome translocation. In normal situations, two yellow signals will be present, while in case of a translocation separate green and red signals will be present together with the co-localized signal of the unaffected gene. of the probes moves to the partner chromosome, i.e. cells indistinguishable from normal nuclei. However, der(partner), while the other probe remains on the 5/10% false-negativity (percentage deduced from der(target) chromosome. The split-signal approach fusion-signal FISH) within the leukemic cell popula- therefore also allows the detection of new partner tion will not alter the result in diagnostic material chromosomes or chromosome regions. Further mo- where the percentage of malignant cells is virtually lecular analysis can then be performed to identify the always over 25%. Consequently, 10% reduction from new partner gene, such as panhandle PCR or long 25% to 22.5% has no diagnostic meaning [18]. distance inverse PCR [21,22]. Another advantage of split-signal FISH is absence Reduction of background staining using PNA- of the traditionally high levels of false-positivity as based blocking observed via the fusion-signal FISH approach, which range between 5 and 10%. False-positivity occurs as a The successful use of large genomic probes for FISH result of coincidental co-localization of two signals, is dependent on blocking of the undesired back- which actually represent two separate signals in a ground staining derived from repetitive sequences three-dimensional nucleus, but due to the two-dimen- present throughout the human genome. The finishing sional analysis of the nucleus, are visible as a single co- of the human genome project has shown that a large localized signal. On the other hand, one could argue proportion of the human genome is comprised of that split-signal FISH can give rise to low frequencies tandem repeated sequences (i.e. arranged in blocks) of false-negativity due to the same type of coincidental and interspersed tandem repeated sequences (distrib- co-localization of two separate signals making these uted all around the genome). 70 J. J. M. van Dongen et al.

Figure 2. Split-signal FISH for detection of breaks in the MLL gene. A. MLL gene (11q23) with the position of the breakpoint region and the positions of the centromeric MLL-U probe (239 kb, red) and telomeric MLL-D probe (513 kb, green). B. Metaphase spread of a healthy donor. C. Precursor-B-ALL without MLL gene aberration. D. Precursor-B-ALL with MLL gene translocation. E. Precursor-B-ALL with MLL gene translocation with loss of 3’ part of MLL gene. Previously, heat denaturation and reannealing stu- most frequent repetitive element within and around dies on DNA of higher organisms have distinguished genes. PNA is a DNA analogue in which the three populations of genomic DNA: a slowly rean- deoxyribose phosphodiester backbone is replaced by nealing component (45% of the total DNA) contain- a pseudo-peptide backbone of N-(2-aminoethyl)-gly- ing unique sequences of protein-encoding genes, and cine units to which the nucleobases are attached intermediate and quickly reannealing components through a methylene carbonyl linker (Figure 3) (30% and 25% of the total DNA, respectively) [29,30]. The charge of the pseudo-peptide backbone representing repetitive sequences [23]. The fast com- of PNA is neutral, whereas the charge of the deoxyr- ponent contains small (a few nucleotides long), highly ibose phosphodiester backbone of DNA is negative. repetitive DNA sequences, while the intermediate Because of lower electrostatic repulsion, a PNA/ component contains the interspersed repetitive DNA DNA interaction occurs faster and is stronger than a that can be classified as either SINEs (Short Inter- DNA/DNA interaction [31]. Different PNA oligos spersed Nuclear Elements), LINEs (Long Inter- were selected in such a way that they cover both the spersed Nuclear Elements) or LTRs (Long Terminal upper and lower strand of the repetitive sequences

Repeats) [24/27]. The repetitive units of the inter- and could therefore be used as a blocking reagent mediate reannealing component the major reason that [32]. large genomic nucleic acid probes are not well suited This novel PNA-based method for suppression of for hybridization analysis without blocking the repe- background staining is now included in our FISH titive elements to prevent undesired staining. procedure (DakoCytomation, Glostrup, DK, EU). A Blocking of repetitive sequences can be achieved paraformaldehyde pre-treatment is used to improve using a component of the total DNA, Cot-1 DNA, the brightness of the fluorescence signals. The pre- enriched with repetitive sequences [28]. Recently, a mixed ALL probe sets contain PNA oligos and the novel method has been developed based on selection fluorescently labeled DNA probes, and are denatu- of specific Peptide Nucleic Acid (PNA) oligos, rated together with the target DNA before hybridiza- directed against the Alu sequences, which is the tion in a humified environment overnight. Excess of Split-signal FISH for translocations in ALL 71

HN Base several BAC/PAC clones), which are located in the 2 Base HO O target gene at opposite sides of the breakpoint region N [17/20,33]. Directly labeled FISH probes work O O O smoothly in combination with the newly developed HN Base O P O- PNA-blocking system, which allows combined block- Base O O ing and hybridization in a single step. This single-step N O O hybridization procedure makes split-signal FISH an easy, rapid, and sensitive tool for molecular cytoge- HN Base O O P O- netics (Figure 4). Base The split-signal FISH approach has three major N O O O O advantage over fusion-signal FISH. First, translocations involving the target gene can be detected NHR O independent of the involved partner gene. Second, O P O- OR split-signal FISH allows identification of the partner PNA DNA gene or partner chromosome region, if metaphases are present. The third advantage is the absence of high Figure 3. Chemical structure of a PNA and a DNA backbone molecule. ‘‘Base’’ indicates a purine (adenine, guanine) or a levels of false-positivity due to coincidental co-locali- pyrimidine (cytosine, thymidine). zation, as observed in the traditional fusion-signal probe and PNA oligos is removed by washing under FISH approach. One could argue that split-signal stringent condition, before embedding and examina- FISH can give rise to similar frequencies of false- tion of the hybridization area (Figure 4). negativity due to the same type of coincidental co- localization, but 5/10% false-negativity as deduced from fusion-signal FISH within the leukemic cell Concluding remarks population will not alter the result in diagnostic Split-signal FISH probe sets each consists of two material where the percentage of malignant cells is differentially labeled probes (generally composed of virtually always /10%.

Figure 4. Protocol for FISH with PNA-based suppression of background staining. Slides with tissue or cytology preparation are pre-treated to increase the access of target DNA for the labeled probes. The probe mixture containing PNA oligos and fluorescent labeled probes is applied to the target DNA and co-denaturated, before hybridization. Unspecifically bound probe is removed by washing before the slide is scored with a fluorescent microscope. Normal cells present on the slides serve as control cells. 72 J. J. M. van Dongen et al.

References quantitative reverse transcriptase polymerase chain reaction of fusion gene transcripts for residual disease detection in [1] Rowley JD. The role of chromosome translocations in leukemia-A Europe Against Cancer Program. Leukemia leukemogenesis. Semin Hematol 1999;36:/ 59/ /72. 2003;17:/ 2318/ /2357. [2] Look AT. Oncogenic transcription factors in the human acute [16] Kearney L. The impact of the new FISH technologies on the leukemias. Science 1997;278:/ 1059/ /1064. cytogenetics of haematological malignancies. Br J Haematol [3] Rabbitts TH. Chromosomal translocations in human cancer. 1999;104:/ 648/ /658. Nature 1994;372:/ 143/ /149. [17] Van der Burg M, Smit B, Brinkhof B, Barendregt BH, [4] Harrison CJ. The management of patients with leukaemia: the Verschuren MCM, Dib M, Beverloo HB, Van Dongen JJM,

role of cytogenetics in this molecular era. Br J Haematol 2000;/ Langerak AW. A single split-signal FISH probe set allows 108:19/ /30. detection of TAL1 translocations as well as SIL-TAL1 fusion [5] Lo Coco F, Mandelli F, Breccia M, Annino L, Guglielmi C, genes in a single test. Leukemia 2002;16:/ 755/ /761. Petti MC, Testi AM, Alimena G, Croce CM, Canaani E, [18] Van der Burg M, Poulsen TS, Hunger SP, Beverloo HB, Smit Cimino G. Southern blot analysis of ALL-1 rearrangements at EM, Vang-Nielsen K, Langerak AW, Van Dongen JJM. Split-

chromosome 11q23 in acute leukemia. Cancer Res 1993;53:/ / signal FISH for detection of chromosome aberrations in acute 3800/3803. lymphoblastic leukemia. Leukemia 2004;18:/ 895/ /908. [6] Mathew S, Behm F, Dalton J, Raimondi S. Comparison of [19] Van der Burg M, Beverloo HB, Langerak AW, Wijsman J, van cytogenetics, Southern blotting, and ﬂuorescence in situ Drunen E, Slater R, Van Dongen JJM. Rapid and sensitive hybridization as methods for detecting MLL gene rearrange- detection of all types of MLL gene translocations with a single ments in children with acute leukemia and with 11q23 FISH probe set. Leukemia 1999;13:/ 2107/ /2113. abnormalities. Leukemia 1999;13:/ 1713/ /1720. [20] Van Dongen JJM, Langerak AW. Molecular detection of [7] Rubnitz JE, Downing JR, Pui CH, Shurtleff SA, Raimondi chromosome aberrations (Split-signal FISH probe technol- SC, Evans WE, Head DR, Crist WM, Rivera GK, Hancock ogy). International Patent 1998; PCT/NL98/00270: ML, et al. TEL gene rearrangement in acute lymphoblastic [21] Megonigal MD, Rappaport EF, Jones DH, Kim CS, Nowell leukemia: a new genetic marker with prognostic signiﬁcance. J PC, Lange BJ, Felix CA. Panhandle PCR strategy to amplify Clin Oncol 1997;15:/ 1150/ /1157. MLL genomic breakpoints in treatment-related leukemias. [8] Pongers-Willemse MJ, Seriu T, Stolz F, d’Aniello E, Gameiro Proc Natl Acad Sci U S A 1997;94:/ 11583/ /11588. P, Pisa P, Gonzalez M, Bartram CR, Panzer-Grumayer ER, [22] Felix CA, Jones DH. Panhandle PCR: a technical advance to Biondi A, et al. Primers and protocols for standardized amplify MLL genomic translocation breakpoints. Leukemia detection of minimal residual disease in acute lymphoblastic 1998;12:/ 976/ /981. leukemia using immunoglobulin and T cell receptor gene [23] Britten RJ, Kohne DE. Repeated sequences in DNA. Hun- rearrangements and TAL1 deletions as PCR targets: report of dreds of thousands of copies of DNA sequences have been the BIOMED-1 CONCERTED ACTION: investigation of incorporated into the genomes of higher organisms. Science

/ minimal residual disease in acute leukemia. Leukemia 1999; 1968;161:/ 529/ /540. / / 13:110 118. [24] Brosius J. Retroposons-seeds of evolution. Science 1991;251:/ / [9] Breit TM, Beishuizen A, Ludwig WD, Mol EJ, Adriaansen HJ, 753. van Wering ER, van Dongen JJM. tal-1 deletions in T-cell [25] Novick GE, Batzer MA, Deiniger PL, Herrera RJ. The mobile

acute lymphoblastic leukemia as PCR target for detection of genetic element Alu in the human genome. Bioscience 1996;/ / / / minimal residual disease. Leukemia 1993;7:2004 2011. 46:32/ /41. [10] Wiemels JL, Leonard BC, Wang Y, Segal MR, Hunger SP, [26] Singer MF. SINEs and LINEs: highly repeated short and long

Smith MT, Crouse V, Ma X, Bufﬂer PA, Pine SR. Site-speciﬁc interspersed sequences in mammalian genomes. Cell 1982;28:/ /

translocation and evidence of postnatal origin of the t(1;19) 433/434. E2A-PBX1 fusion in childhood acute lymphoblastic leukemia. [27] Korenberg JR, Rykowski MC. Human genome organization: Proc Natl Acad Sci U S A 2002;99:/ 15101/ /15106. Alu, lines, and the molecular structure of metaphase chromo- [11] Wiemels JL, Alexander FE, Cazzaniga G, Biondi A, Mayer SP, some bands. Cell 1988;53:/ 391/ /400. Greaves M. Microclustering of TEL-AML1 translocation [28] Landegent JE, Jansen in de Wal N, Dirks RW, Baao F, van der breakpoints in childhood acute lymphoblastic leukemia. Ploeg M. Use of whole cosmid cloned genomic sequences for Genes Chromosomes Cancer 2000;29:/ 219/ /228. chromosomal localization by non-radioactive in situ hybridi- [12] Van Dongen JJM, Langerak AW, Bruggemann M, Evans PA, zation. Hum Genet 1987;77:/ 366/ /370. Hummel M, Lavender FL, Delabesse E, Davi F, Schuuring E, [29] Nielsen PE, Egholm M, Berg RH, Buchardt O. Sequence- Garcia-Sanz R, et al. Design and standardization of PCR selective recognition of DNA by strand displacement with a primers and protocols for detection of clonal immunoglobulin thymine-substituted polyamide. Science 1991;254:/ 1497/ / and T-cell receptor gene recombinations in suspect lympho- 1500. proliferations: report of the BIOMED-2 Concerted Action [30] Egholm M, Buchardt O, Christensen L, Behrens C, Freier BMH4-CT98-3936. Leukemia 2003;17:/ 2257/ /2317. SM, Driver DA, Berg RH, Kim SK, Norden B, Nielsen PE. [13] Van Dongen JJM, Macintyre EA, Delabesse E, Gabert JA, G PNA hybridizes to complementary oligonucleotides obeying

C, Saglio G, Gottardi E, Rambaldi A, Dotti P, Griesinger F, the Watson- Crick hydrogen-bonding rules. Nature 1993;365:/ /

Parreira A, et al. Standardized RT-PCR analysis of fusion gene 566/568. transcripts from chromosome aberrations in acute leukemia [31] Nielsen KV, Mu¨ller S, Poulsen TS, Gabs S, Schonau A. for detection of minimal residual disease. Report of the Peptide nuclei acid, protocols and applications. In: Nielsen

BIOMED-1 Concerted Action: Investigation of minimal PE, editor. Horizon Scientiﬁc Press; 2003, pp 1/37. residual disease in acute leukemia. Leukemia 1999;13:/ 1901/ / [32] Van Dongen JJM, Vang Nielsen K, Pluzek KJ, Adelhorst K. 1928. Method and probes for the detection of chromosome aberra- [14] Szczepanski T, Orfao A, van der Velden VH, San Miguel JF, tions (FISH probe/PNA based technology). International van Dongen JJ. Minimal residual disease in leukaemia Patent 1999; PCT/DK99/00245: patients. Lancet Oncol 2001;2:/ 409/ /417. [33] Boomer T, Varella-Garcia M, McGavran L, Meltesen L, [15] Gabert J, Beillard E, Van Der Velden VH, Bi W, Grimwade D, Olsen AS, Hunger SP. Detection of E2A translocations in Pallisgaard N, Barbany G, Cazzaniga G, Cayuela JM, Cave H, leukemias via ﬂuorescence in situ hybridization. Leukemia et al. Standardization and quality control studies of ‘real-time’ 2001;15:/ 95/ /102. Hematology, 2005; 10 Supplement 1: 73/75

ADVANCES IN MOLECULAR HEMATOLOGY

Regulation and dysregulation of hematopoiesis by a cytokine-induced antiapoptotic molecule anamorsin

YUZURU KANAKURA

The Department of Hematology and Oncology, Osaka University Graduate School of Medicine, C9, 2-2, Yamada-oka, Suit, Osaka 565-0871 Japan

Keywords: Apoptosis, hematopoiesis, signaling

Many growth factors and cytokines have a crucuial human homologue of AM in EMBL/GenBank data role in hematopoiesis, and the cytokine-receptor libraries, which revealed 82.6% similarity to murine system is known to involve in normal and abnormal AM at a DNA level and 81.9% similarity at an amino hematopoiesis. Upon binding to receptors, cytokines acid level. Human homologue of AM was originally initially induce activation of cell surface tyrosine found by Lotfus et al. as a molecule with unknown kinases that transmit mitogenic and anti-apoptotic function on chromosome 16. AM encodes an about signals through simultaneous activation of down- 37-kDa protein and the protein sequence data base stream signaling molecules including Ras/MAPK, indicates that AM had generic methyltransferase PI3-K/Akt, and STATs. However, it is not fully motif around amino acids 60/99. understood yet how cytokines mediate anti-apoptotic At first, we examined whether AM was indeed effects. To define molecule(s) that mediates anti- involved in the resistance to factor-deprived apoptosis apoptotic effect of cytokine and confer resistance to of Ba/F3-Ad. For this purpose, we stably expressed apoptosis induced by cytokine deprivation, we have murine AM cDNA in parental Ba/F3 cells, and established a subline named Ba/F3-Ad that can grow investigated their sensitivities to IL-3-deprived apop- and survive under IL-3-deprived conditions from a tosis. In control Ba/F3 cells, the subdiploid fraction murine IL-3-dependent cell line, Ba/F3. In order to formed from apoptotic cells emerged as early as 24 h identify the molecule(s) that conferred the resistance after IL-3 depletion, which was effectively reduced in to factor-deprived apoptosis on Ba/F3, we performed AM-transfected Ba/F3 cells (% of apototic fraction: the expression cloning by constructing a retroviral control Ba/F3, 25% at 24 h, 57% at 36 h, 87% at 48 cDNA library from IL-3-starved Ba/F3-Ad. In short, h; AM-transfected Ba/F3, 3% at 24 h, 4% at 36 h, we infected the retrovirus library into parental Ba/F3 12% at 48 h). In agreement with this finding, the (5/105 clones, the average size of the inserts, 1.3 kb; activation of caspase-3, which is detected as a the infection efficacy was estimated at about 35%), and screened the clones that can survive under IL-3- conversion of left-sided peak to right-sided peak in starved condition. After isolation of several clones that fluorescent intensity, was significantly suppressed in can survive under IL-3-deprived condition, we iso- AM-transfected Ba/F3 cells as compared to that in lated the integrated cDNA from genomic DNA control Ba/F3 cells. However, it should be noted that extracted from one clone by the PCR method. AM alone could not support the growth of Ba/F3 By sequencing the integrated cDNA, we found that cells, since proliferating cells drastically decreased in the coding region of murine AM cDNA consists of AM-transfected Ba/F3 cells under IL-3-deprived 930 bp. Comparison with the DNA data base search condition (% of the cells in S-G2/M phase before revealed that the sequence of AM does not exhibit and after IL-3 depletion: 68% at 0 h vs. 10% at 48 h). homology with any known anti-apoptotic molecules, We also confirmed that AM can confer resistance to including Bcl-2 family proteins, caspase inihibitors, or apoptosis caused by IL-3 depletion in another murine signal transduction molecules. Also, we found a IL-3-dependent cell line, 32D.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389999 74 Y. Kanakura

Next, we examined the expression profile of human tured with 150 Ig/ml of G418. We confirmed the AM homologue in various organs using MTA panels. homologous recombination in the selected ES cell AM was expressed ubiquitously in various tissues, lines with Southern blot and PCR analyses. The wild especially with high expression levels in heart, liver type allele was detected as a 5-kb SacI fragment in and pancreas. As for hematopoietic tissues, it was Southern blot analysis with a 5? flanking probe, while abundantly expressed in fetal liver and spleen. Also, the targeted allele was detected as a 13-kb SacI we found that the expression of AM was detectable fragment. Also, in PCR analyses, the 310-bp fragment from the early stages (7 day) of embryogenesis by the was amplified from the wild type allelic genomic DNA PCR method. with a primer pair I/I, while the 730-bp fragment Since AM was supposed to exert anti-apoptotic was amplified from the targeted allelic genomic DNA effects in Ba/F3-Ad under IL-3-deprived condition, with a primer pair I/I. To generate chimeras, we we examined its expression levels in parental Ba/F3 injected three ES cell lines, in which the homologous and Ba/F3-Ad after IL-3 depletion. As expected, AM recombination was confirmed, into blastocysts of was still expressed in Ba/F3-Ad after 18-h IL-3 C57BL/6J mice. Then, male mice with a high degree deprivation, while its expression was hardly detectable of chimerism were crossed with C57BL/6J females to in parental Ba/F3 cells. However, the addition of IL-3 generate AM/// mice. Genotyping was performed recovered it expression in parental Ba/F3 cells from by Southern blot and PCR analyses using tail- and after 3 h. We next investigated the expression of AM embryo-derived DNA. Finally, we confirmed that the was also regulated by other cytokines in Ba/F3 cells. expression of AM protein in the limb was partially

After IL-3 depletion, clones of Ba/F3 each expressing reduced in AM/// embryos and completely lost in the receptors for EPO, SCF (c-Kit), and TPO (c- AM/// embryos by Western blot analysis using an Mpl) were cultured with the corresponding cytokines anti-AM mAb. for the time indicated. EPO, SCF, and TPO indivi- Genotypic analysis of embryos from AM/// mice dually induced the expression of AM as efficiently as intercrosses revealed that AM/// embryos started IL-3, suggesting that the expression of AM would be to die between E12.5 and E14.5; the rate of dead regulated by common signaling molecule(s) shared by AM/// embryos increased after E14.5 (0% until various cytokines. Based on the fact that Ras is E12.5, 18.8% at E14.5, 36.4% at E16.5, and 44.4% activated by various cytokines and was constitutively at E18.5); and all AM/// mice died at birth, activated in Ba/F3-Ad, we speculated that Ras might suggesting that AM/// mice expire in late gestation. control the expression of AM in Ba/F3 cells. To AM/// embryos were apparently smaller in body examine this possibility, we stably expressed onco- size than AM/// embryos, while AM/// embryos genic H-ras (H-rasG12V) in Ba/F3 cells. Also, we displayed phenotypes similar to AM/// embryos. prepared the clone from Ba/F3, in which dominant Despite no significant difference in the formation of negative H-ras (H-rasS17N) was inducibly expressed blood islands in yolk sac, the fetal liver (FL) and by the IPTG treatment. In Ba/F3 cells transfected spleen of AM/// embryos were remarkably smaller with H-rasG12V, the expression of AM was main- than those of AM/// embryos; the size of FL of tained after IL-3 depletion for up to 36 h, while its AM/// embryos was almost one third of that of expression declined to an undetectable level in the AM/// embryos, and the spleen of AM/// mock clone transfected with an empty vector. embryos appeared scar. Furthermore, AM/// em-

Furthermore, the induced expression of H-rasS17N bryos later than E14.5 looked anemic; AM/// led to the reduction of AM expression even in the embryos at E18.5 showed almost half values of red presence of IL-3. Collectively, these results suggest blood cells (RBC), hemoglobin (Hb) and hematocrit that the expression of AM is completely dependent on (Ht) in the peripheral blood (RBC: 3659/50.8, cytokine stimulation, and, at least partially, regulated 335.59/50.6 and 2109/67.0/104/mm3; Hb: 11.79/ by Ras signaling in Ba/F3 cells. Next, we investigated 1.5, 10.69/1.3 and 6.79/2.3 g/dl; Ht: 39.79/5.8, the intracellular localization of AM by an immuno- 37.89/5.3 and 26.39/7.6% in AM/// (n/10), fluorescence staining with an anti-AM monoclonal AM/// (n/22) and AM/// (n/8) embryos, anitibody (mAb). AM was exclusively localized at respectively). Moreover, RBC of AM/// embryos cytoplasm in Ba/F3 cells irrespective of the stimula- were macrocytic (mean corpuslar volume in AM///, tion with IL-3. AM/// and AM/// embryos were 108.89/4.3,

To assess in vivo roles of AM, we tried to generate 113.19/9.5 and 128.29/16.6, respectively). In addi-

AM null (AM///) mice. We constructed a targeting tion to the defect in hematopoietic organs, heart walls vector, in which the first exon was replaced by the of AM/// embryos were thinner than those in neomycine-resistant cassette (a positive selection AM/// or AM/// embryos, whereas AM/// marker). The vector also included the diphtheria embryos displayed no apparent macroscopic or histo- toxin as a negative selection marker. The targeting logical abnormalities in other organs. vector was introduced into an ES cell line R1 by In order to clarify the mechanism of anemia seen in electroporation, and the transfected cells were cul- AM/// embryos, we analyzed the FL, the main Regulation and dysregulation of hematopoiesis 75 hematopoietic organ at late embryonic days. The combination of cytokines SCF, IL-3, IL-6 and EPO, number of morphologically identifiable small-sized the number of myeloid (granulocyte/macrophage and erythroid cells (erythroblasts) with condensed chro- granulocyte) colonies formed from AM/// FL cells matin nucleus was apparently reduced in FL of was one third to one fourth of those by AM/// FL AM // / embryos compared to that of controls. cells. Notably, AM/// FL cells gave rise to little or Furthermore, erythroblasts were larger in AM // / no mix (myeloid/erythroid) and erythroid colonies,

/ / embryos than in AM / embryos, and more mature while these colonies did develop from AM/// and erythroblasts (i.e., polychromatic and orthochromatic AM/// FL cells. Furthermore, BFU-E colony did erythroblasts) markedly decreased in AM/// em- not develop when E14.5 AM/// FL cells were bryos. In addition, TUNEL assays showed that a cultured with SCF and EPO (Fig. 5E). These results substantial fraction of FL cells led to apoptosis at suggest that AM is indispensable for cytokine-depen- E14.5 in AM // / embryos, while it was hardly dent survival and growth of hematopoietic stem/ detected in AM // / embryos. To determine which progenitor cells in vitro, especially with erythroid type of cells undergo apoptosis, we performed flow lineage. cytometric analysis using AM/// and AM/// FL To characterize the anti-apoptotic function of AM, cells at E14.5. Although Annexin V-positive apoptotic we compared the gene expression profiles between cells were scarcely (only 0.5%) detected in AM/// AM/// and AM/// FL cells at E14.5 using a FL cells, 60.2% of the isolated cells were positive for cDNA microarray. Among 4289 genes, including Bcl- Annexin V in AM/// FL. Furthermore, most 2 family, caspases, cytokines and signal transduction importantly, almost all of these apoptotic cells were Ter-119-positive erythroid cells but not Ter-119- molecules, 184 genes were significantly down-regu- negative cells mainly composed from hepatocytes. lated and 40 were up-regulated in AM // / FL. Since neither the absolute number of hematopoietic Concerning apoptosis related genes, Bcl-xL and Jak2 were down-regulated most significantly. We also stem/progenitor cells (CD34/c-kit/ or CD34 low CD44 high) nor that of very immature proerythro- confirmed that their expression was decreased in AM/// FL cells compared with AM/// FL cells blasts (Ter-119/c-kit/) did not decrease in AM//

/ FL, it was speculated that immature hematopoietic by semiquantative RT-PCR assays. Furthermore, AM cells of AM/// mice may succumb to apoptosis, was found to express in a part of leukemia and and fail to attain terminal differentiation. malignant lymphoma cells at a significantly high level. We next examined whether hematopoietic stem/ These results suggest that AM may play a crucial progenitor cells of AM/// FL cells could survive, role in hematopoiesis through mediating anti-apopto- proliferate and differentiate in response to cytokines in tic effects of various cytokines, and also that AM may vitro. When E14.5 AM///,AM/// or AM/// contribute to abnormal growth of leukemia/lym- FL cells were cultured in methylcellulose with the phoma cells. Hematology, 2005; 10 Supplement 1: 76/78

ADVANCES IN MOLECULAR HEMATOLOGY

WT1 overexpression: A clinically useful marker in acute and chronic myeloid leukemias

GIUSEPPE SAGLIO, SONIA CARTURAN, SARA GRILLO, SARA CAPELLA, FRANCESCA ARRUGA, ILARIA DEFILIPPI, VALENTINA ROSSO, MARIA RAUCO*, ANNA MARINA LIBERATI,* & DANIELA CILLONI

Division of Internal Medicine and Haematology, Department of Clinical and Biological Sciences of the University of Turin, San Luigi Gonzaga Hopsital 10043 Orbassano-Turin, Italy, *Clinica Medica, Universita` di Perugia, Policlinico Monteluce, Italy

Monitoring of acute leukemia patients during and human hematopoiesis still needs to be clarified. The after treatment for the presence of remaining leuke- role of WT1 in the leukemogenesis process appears mic cells minimal residual disease (MRD) have been controversial. The majority of human acute myeloid shown to give major insight into the effectiveness of and lymphoblastic leukemias express high levels of treatment. However, so far applicability of this strat- wild type WT1, [3,4] suggesting that this tumor egy has been limited to those leukaemia subsets suppressor might have paradoxical oncogenic activity characterized by genetic markers amenable to sensi- in the hematopoietic cells. Although the potential tive detection by PCR. Although PCR for immuno- usefulness of WT1 expression as a panleukemic globulin and T-cell receptor gene rearrangement marker was envisaged by Inoue et al. [5] several years represents the gold standard for MRD detection in ago, its introduction in the clinical practice was most cases of acute lymphoblastic leukemias (ALL) limited principally by the background expression level lacking the availability of fusion gene transcripts as normally detected in bone marrow (BM) samples molecular markers, the situation in AML is more from healthy volunteers. More recent data obtained complicated because, at present, more than 50% of using quantitative RT-PCR methods established that them lack any sort of clonality markers suitable for sorted populations of normal progenitors express MRD monitoring. Thus, a number of studies have WT1 at very low levels, sometimes even undetectable been performed in the attempt to identify cytogenetic by very sensitive methods of nested RT-PCR [6/9]. and molecular abnormalities associated with leukemic The introduction of a precise method of quantita- transformation. tive PCR allow to distinguish the levels in normal The Wilms Tumor Gene (WT1) represents a controls and in leukemic samples, so overcoming the molecular marker for the detection of the leukemic obstacle represented by the low amount of WT1 clone useful for monitoring the presence of leukemic transcript present in normal hematopoiesis. Using a cells in all the patients affected by acute and chronic Real Time quantitative PCR, we were able to demon- leukemias as well as myelodysplastic syndromes. The strated the presence of high levels of WT1 in the WT1 gene, cloned in 1990 by Call et al. [1] encodes majority of cases of acute myeloid and lymphoblastic for a protein with the characteristics of a zinc finger leukemias at onset of disease as well as in the different transcription factor. WT1 expression is restricted to a phases of chronic myeloid leukemias (ML) [5]. In small number of tissues [2] including testis, ovaries, addition also the patients affected by myelodysplastic myometrium, stromal cells of the uterus, heart, lung, syndromes express increased amount of WT1 tran- intestine, liver and in the supportive stroma and script with variable level according to the subtypes of splenic capsule of the spleen [2]. In contrast, several MDS [10]. other tissues and cell lines were negative for WT1 By analyzing a large number of normal BM and expression. Although the role of the WT1 gene in the peripheral blood (PB) samples we also established development of malignancies in the kidney appears that the majority of the PB samples score negative and quite well defined, currently its potential function in the median number of WT1 copies detected in the

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331390005 WT1 overexpression 77 positive samples is very low. By contrast. all the could even more sensitive than BM in revealing normal BM samples score positive, but, even if higher impending relapses. Although this point still needs than in PB samples, the median number of WT1 to be demonstrated, replacement of BM with PB copies remains low. To gain a further insight into the sampling could greatly improve patients’ compliance theory which attributes the abnormal levels of WT1 and considerably simplify molecular monitoring. expression to the presence of cells with a high degree The clinical application of WT1 gene as a mole- of immaturity, we tested WT1 expression in regener- cular marker for MRD detection was already sug- ating BM samples obtained from AML patients in gested by Inoue et al. [11] in 1996 and confirmed complete remission during recovery from chemother- later on by different studies. Recently published data apy-induced aplasia. The median value of WT1 was obtained in a pediatric group of AML patients in similar to that detected in normal BM. Similar results which the presence of MRD was assessed by flow were obtained by analyzing several samples of en- cytometric analysis for the blast cell count. They riched CD34-positive cells obtained from normal PB found a strict correlation between the levels of WT1 cell donors. These data confirm that the CD34- transcript and the percentage of blast cells detected by positive cell compartment is not responsible for the flow cytometric analysis. Their results confirm the increased WT1 expression detected in leukemias. finding that clinical remission is constantly associated These results lead to the conclusion that the WT1 with low levels of WT1 while increasing values are overexpression found in leukemic samples is not due associated with relapse. to the degree of immaturity of the cell population, but Finally, similarly to what demonstrated for AML, intrinsically related to the presence of leukemic cells. [6,9] WT1 levels seem to represent a good marker for Conversely, in AML, MDS and CML BM and PB MRD detection also in MDS patients treated with samples, a significant higher level of expression can be intensive therapies aimed at the disease eradication detected. Since the RQ-PCR methods clearly distin- [10]. guish between the level of WT1 transcripts in normal Moreover, it has been demonstrated that even in and leukemic cells, WT1 expression can represent a the transplant setting, as already demonstrated for molecular marker extremely useful in the clinical acute leukemia patients treated with intensive che- setting of MRD assessment. In particular in cases motherapy, the determination of the WT1 amount lacking cytogenetic lesions, it may help to establish the can represent a useful marker to monitor the persis- response to therapy and to monitor the behavior of tence or the reappearance of leukemic cells and that the leukemia clone during follow-up. the finding of increasing amounts of WT1 transcript To validate the role of WT1 as a marker of MRD, during follow-up is predictive of relapse. we studied a number of patients bearing a fusion gene transcript suitable for the quantitative assessment of the MRD amount by RQ-PCR and performed a References simultaneous analysis of the WT1 amount at sequen- [1] Call KM, Glaser T, Ito CY, Buckler A, Pelletier J, Haber DA, tial time intervals during the follow-up [6]. The WT1 Rose EA, Kral A, Yeger H, Lewis WH. Isolation and levels were shown to strictly parallel the behavior of characterization of a zinc ﬁnger polypeptide gene at the

the other molecular markers (fusion gene transcripts) human chromosome 11 Wilms’ Tumor Locus. Cell 1990;60:/ / used for the MRD monitoring. Furthermore, in- 509 /520. creased WT1 expression above the range found in [2] Park S, Schalling M, Bernard A, Maheswaran S, Shipley GC, Roberts D, Fletcher J, Schpman R, RheinWald J, Demetri G. normal BM and/or in normal PB samples during The Wilms tumour gene WT1 is expressed in murine follow-up of AML patients was always found to be mesoderm-derived tissues and mutated in a human mesothe- predictive of an impending hematological relapse even lioma. Nature Genetics 1993;4:/ 415/ /420. in AML patients lacking additional molecular mar- [3] Menssen HD, Renkl HJ, Rodeck U, Maurer J, Nutter M, kers. The increase of the WT1 levels could also Schwartz S, Reinhardt R, Thiel E. Presence of Wilms’ tumor gene (WT1) transcripts and WT1 nuclear protein in the precede the occurrence of the overt hematological majority of human acute leukemias. Leukemia 1995;9:/ 1060/ / relapse of some months, although the kinetics of the 1067. relapse appears highly variable. On the opposite, [4] Miwa H, Beran M, Saunders GF. Expression of the wilms’ normal WT1 values have always been found asso- tumor gene (WT1) in hman leukemias. Leukemia 1992;6:/ / ciated with persisting remissions. Therefore, evalua- 405 /409. [5] Inoue K, Ogawa H, Sonoda Y, Kimura T, Sakabe H, Oka Y, tion of WT1 expression could represent a sort of Miyake S, Tamaki H, Oji Y, Yamagami T, et al. Aberrant universal marker that can allow a rather sensitive overexpression of the Wilms Tumor Gene (WT1) in human evaluation of the MRD in all AML patients with a leukemia. Blood 1997;89:/ 1405/ /1412. degree of sensitivity that can be estimated to reach in [6] Cilloni D, Gottardi E, De Micheli D, Serra A, Volpe G, Messa most cases between 103 and 10 4 [6]. Further- F, Rege-Cambrin G, Guerrasio A, Divona M, Lo Coco F, Saglio G. Quantitative assessment of WT1 expression by real more, the finding of extremely low and often un- time quantitative PCR may be a useful tool for monitoring detectable WT1 levels in the PB of normal individuals minimal residual disease in acute leukemia patients. Leukemia and in leukaemia patients in CCR, suggests that PB 2002;16:/ 2115/ /2121. 78 G. Saglio et al.

[7] Gaiger A, Schmid D, Heinze G, Linnerth B, Greinix H, Halhs cells from acute myeloid leukemia and normal CD34/ P, Tisljar K, Prigljnger S, Mitterbauer M, Novak M, et al. progenitors. Blood 1997;90:/ 4230/ /4232. Detection of the WT1 transcript by RT-PCR in complete [10] Cilloni D, Gottardi E, Messa F, Fava M, Scaravaglio P, Bertini remission has no prognostic relevance in de novo acute M, Girotto M, Marinone C, Ferrero D, Gallamini A, et al. myeloid leukemia. Leukemia 1998;12:/ 1886/ /1894. Very signiﬁcant correlation between the degree of WT1 [8] Inoue K, Tamaki H, Ogawa H, Oka Y, Soma T, Tatekawa expression and the IPSS score in patients affected by T, Oji Y, Tsuboi A, Kim EH, Kawakami M, et al. Wilms’ myelodysplastic syndromes. (submitted)

tumor gene (WT1) competes with differentiation /inducing [11] Inoue K, Ogawa H, Yamagami T, Soma T, Tani Y, Tatekawa signal in hematopoietic progenitor cells. Blood 1998;91:/ 2969/ / T, Oji Y, Tamaki H, Kyo T, Dohy H, et al. Log /term follow- 2976. up of minimal residual disease in leukemia patients by [9] Maurer U, Weidmann E, Karakas T, Hoelzer D, Bergmann L. monitoring WT1 (Wilms tumor gene) expression levels. Blood Wilms Tumor Gene (wt1) mRNA is equally expressed in blast 1996;88:/ 2267/ /2278. Hematology, 2005; 10 Supplement 1: 79/81

TRANSFUSION POLICIES IN ANEMIA

Haemovigilance in developing countries

O¨ NDER ARSLAN

Department of Hematology, Ankara Univeristy Medical School, Cebeci, Ankara, Turkey

The word haemovigilance is derived from the word Haemovigilance and the European Blood pharmacovigilance, which covers activities and sys- Directive tems to collect information in medicinal products, On February 8, 2003 the European Blood Directive especially for adverse drug reactions in humans. 2002/98/EC was published and came into force. In Haemovigilance, initially created in France in the this Directive, Chapter V is dedicated to haemovigi- begining of 90s has Greek and Latin roots: ‘‘haema’’: lance and there are two operating articles dealing with blood and ‘‘vigilans’’: paying special attention. Hae- haemovigilance; ‘‘Article 14: Traceability movigilance concerns blood components. Pharma- covigilance concerns plasma derivatives such as / Member States shall take all necessary measures clotting factor oncentrates, immunoglobulins, albu- in order to ensure that blood and blood min and other fractionated products. Since 1993, in components collected, tested, processed, stored, European legislation plasma derivatives are consid- released and/or distributed on their territory can ered to be pharmaceuticals, and the manufactures be traced from donor to recipient and vice have to comply with the European regulations on versa. To this end, Member states shall ensure pharmacovigilance. that blood establishments implement a system Transfusion reports from US and United King- for identification of each single blood donation dom showed that most of the incidents were caused and each single blood unit and components by clerical errors. Not only identification but also thereof enabling full traceability to the donor as administrative errors also plays an important role. In well as to the transfusion and the recipient France, by a law in 1994, the first definition of thereof. The system must unmistakably identify hemovigilance was introduced. Various number of each unique donation and type of blood com- definitions have been written since then. In the ponent. This system shall be established in European Blood Directive, it is defined as a group accordance with requirements referred to in of organized surveillance procedures relating to Article 29(a). With regard to blood and blood serious, adverse or unexpected events or reactions components imported from third countries, in donors or recipients, and the epidemiological Member states shall ensure that the donor follow-up of donors. The definition given by the identification system to be implemented by European Haemovigilance Network (EHN) is the blood establishments permits an equivalent level one most widely used. ‘‘Haemovigilance is a set of of traceability. surveillance procedures covering the entire transfu- / Member States shall take all necessary measures sion chain (from the donation of blood and its in order to ensure that the system used for the components to the follow-up of recipients of transfu- labelling of blood and blood components col- sions), intended to collect and assess information on lected, tested, processed, stored, released and/or unexpected or undesirable effects resulting from the distributed on their territory complies with the therapeutic use of labile blood products, and to identification system referred to in paragraph 1 prevent the occurrence or recurrence of such in- and the labelling requirements listed in Annex cidents’’ [1]. III.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389728 80 O¨ . Arslan

/ Data needed for full traceability in accordance Switzerland and Norway are associate member, and with this Article shall be kept for at least 30 Brazil, Spain and Romania have expressed their wish years’’. to join as associate members. ‘‘Article 15: Notification of serious adverse events and reactions Haemovigilance systems around the world / Member States shall ensure that any serious adverse events related to the collection, testing, There are significant differences in haemovigilance processing, storage and distribution of blood and around the world, in terms of definition, organiza- blood components which may have an influence tional schemes, state of development and implemen- on their quality and safety, as well as any serious tation [4]. Although mostly developed in EU reactions observed during or after transfusion countries, there are some countries, even basic trace- which may be attributed to the quality and the ability causes a problem. Most of these systems are on safety of blood and blood components are a volunteer basis but there are a few which the report notified to the competent authority and blood of the reactions are mandatory like the one in France. establishments have in place a procedure accu- Basically the existing systems can be classified rately, efficiently and verifiably to withdraw according to their legal status (mandatory vs. volun- fromdistribution blood or blood components tary), their field of application (all events vs. very associated with the notification referred to above. serious reactions), their organization (centralized vs.

/ These serious adverse events and reactions less decentralized), and their financing. These two shall be notified in accordance with the proce- programmes represent two different models for other dure and notification format referred to in Article countries. 29(i)’’. In Germany, the labile blood components are considered as medicinal products and come under the German Drug Law. According to the national European haemovigilance network (EHN) legal provisions covering medicines, side effects have Five countries took the initiative in 1998 to work to be reported according to the rules of pharmacov- together in the field of haemovigilance: Belgium, igilance. Nevertheless it should be mentioned that the France, Luxembourg, Portugal and The Netherlands recent German Transfusion Law also established and the EHN was born. Later, Austria, Denmark, haemovigilance as a separate entity. Finland, Greece, Ireland and the United Kingdom Little is known about the haemovigilance systems of joined to the EHN and Canada, Croatia, Norway and the new EU countries by 2004. Switzerland were adopted as associate members. EHN was established to increase blood safety at a French haemovigilance system European level and has the following objectives [2,3]; (a) favour exchange of valid information between In 1993 by law, in France, haemovigilance became a members, (b) increase rapid alert and/or early warn- national system of surveillance and alert, from blood ing between the members, (c) encourage joint collection to the follow-up of the recipients, gathering activities between the members, (d) undertake educa- and analysing all adverse events of blood transfusion tional activities in relation to haemovigilance, in order to prevent their recurrences. In France, (e) standardization of processes and forms, (f) com- unlike the other European countries, the reporting parison and analysis of data, generated by national of all adverse reactions is mandatory, regardless of systems, (g) assistance in the implementation of the their severity and their transfusion imputability. The European Blood Directive, in relation to legal provi- law states that ‘‘anyone, doctor, chemist, dental sions. surgeon, midwife or nurse, who notices an unex- This European network focuses on two systems; pected or untoward effect due or possibly due to a Rapid Alert/Early Warning system (RAS) and a blood product, must report it at once’’ [5] reporting system for adverse reactions to blood The agencies that take part in haemavigilance component transfusion. Reporting forms has been system in France: The French Control Authority developed in order to standardize the information on Health Products (AFSSaPS), -The French process needed and to take appropriate action to National Blood Service (EFS) and The Institute prevent the occurrence or recurrence of adverse of Sanitary Surveillance (INVS). The National events. French Blood Agency was disbanded and the All EHN activitties can be followed on the EHN control of blood transfusion activity was transferred Web-site (www.ehn-org.net). Ten Members States of to the AFSSaPS, effective for haemovigilance in 1999 the EU / Belgium, France, Denmark, Greece, Ire- [5,6]. land, Portugal, Luxembourg, Finland, Netherlands, The French Haemovigilance Network is built on and United Kingdom-are 10 full members of EHN. three levels [1,7]. Local level with 2000 public and Non-EU Members States as Australia, Canada, private hospital correspondents and one correspon- Haemovigilance in developing countries 81 dent in each of the 18 regional blood centres helped Turkey by a colleague mostly physicians or pharmacists, in In general, the blood system is decentralized with high each distribution site. Every public hospital must also degree of more than 300 blood banks, most of them have an Haemovigilance and Transfusion Safety are associated with a hospital. So it is not surprising Committee, with technical, medical and administra- not to set up a national centralised and efficient tive members. Regional level with 25 haemovigilance surveillance system. Legislations and regulations are regional coordinators that are physicians located in not ready yet. each French administrative region, under the authority of the prefect and regional director of social and sanitary affairs (DRASS). They supervise investiga- China tions and implementation of corrective actions, when The Chinese Government has decided to undertake a required. At national level AFSSaPS defining haemo- major upgrade of the system around blood transfu- vigilances main features, leading and coordinating the sion. In the coming years there will be important actions of the different actors, taking the appropriate changes towards a comprehensive system that will measures to ensure transfusion safety and passing on include haemovigilance. to the Ministry of Health all the epidemiological information necessary to take action. The global traceability of blood products improved References

from 95% in 2000, to 99% in 2003. On average 7000 [1] Faber J-C. Haemovigilance around the world. Vox Sang 2002;/ adverse reactions are reported each year, from shivers 83:71/ /76. [2] Faber J-C. Haemovigilance in Europe: the European haemovi- or pruritus to death. Currently, more than 68,000 gilance network. Transfus Clin Biol 2001;8:/ 285/ /290. adverse reactions are in the national database. In [3] Strengers P. Haemovigilance / Why? EHN website, www. 2003, 6933 adverse reactions were reported, among ehn-org.net which 2911 (42%) had a strong association with [4] Faber JC. Worldwide overview of existing haemovigilance transfusion. There were 1898 acute (65%) and 1013 systems. Transfus Apheresis Sci 2004;31:/ 99/ /110. [5] Rebibo D, Hauser L, Slimani A, Herve P, Andreu G. The (35%) delayed adverse reactions. More than 50% of French Haemovigilance System: organization and results for acute reactions were allergic and 22% were FNHTR. 2003. Transfus Apheresis Sci 2004;31:/ 145/ /153. Almost all the 1013 delayed adverse reactions [6] Engelfriet CP, Reesink HW. Haemovigilance systems. Vox Sang (98.5%) were red cell immunization. Seven viral 1999;77:/ 110/ /120. [7] Council of Europe. Guide on the preparation, use and quality infections (6 HCV and 1 CMV) with strong transfu- assurance of blood components. Recommendation No. sion imputability was recorded [5]. R(95)15, 9th ed. January 2003, ISBN 92-871-5075-3. Hematology, 2005; 10 Supplement 1: 82/85

TRANSFUSION POLICIES IN ANEMIA

Blood transfusion and alternatives in elderly, malignancy and chronic disease

FATEN MOFTAH

Medical director of NBTS, MOH&P, Egypt

Introduction When anemia develops it leads to various physiologic changes, the most important of which is the Blood transfusion is an essential part of modern compensatory response which aims at preserving the medicine and health care, however it should be oxygen supply to the tissues. When anemia is severe prescribed only to treat conditions associated with enough that the patient is de-compensated, it is very significant morbidity or mortality that cannot be important that red cells transfusion is considered to prevented or managed by other means [1]. increase the oxygen carrying capacity needed for the Both blood transfusion service and clinical users tissues. De-compensation might happen in case of should share efforts to establish and implement significant cardiovascular disease, fever or infections policies and strategies aiming at reducing the need which increase the demand for oxygen. for transfusion, minimize unnecessary transfusions and ensure a safe, rational and appropriate use of blood and blood components. Blood should not be General common features associated with transfused unless it has been obtained from appro- anemia related to elderly, malignancy, and priately selected donors, has been tested for transfu- chronic disease sion transmissible infections and tested for The rate at which anemia develops determines the compatibility between donor’s red cells and patient’s severity of the symptoms and accordingly the con- plasma [1]. sequences. Subsequently anemia can be classified as:- Anemia of chronic disease is the second prevalent anemia after anemia caused by iron deficiency, occurs / Mild in patients with acute or chronic activation [3]. / Moderate

/ Severe Definition Moderate anemia may cause no symptoms; never- Anemia is defined as a hemoglobin concentration in theless it reduces the patient’s reserves to adjust to any blood that is below the expected value, when age, event such as infection. gender, pregnancy and certain environmental factors, Severe anemia is an important factor in reducing such as altitude, are taken into consideration [1]. the patient’s tissue supply to critical levels. In this However anemia in general results as a conse- situation, urgent treatment is required and the need quence of one or more of the following generic for transfusion should be assessed. Severe anemia is causes:- usually a sign of an advanced underlying disease. The clinical picture of anemia in the above men-

/ Increased loss of red blood cells. tioned three conditions are due to either anemia itself

/ Decreased production of red blood cells. or the underlying cause. However chronic anemia in

/ Increased destruction of red blood cells. these situations may have some symptoms which will

/ Increased demand of red blood cells. increase greatly if the patient develops sudden hemo-

/ Increased production of abnormal red blood lysis, bleeding or during other physiological condi- cells. tions such as pregnancy or labor.

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389917 Blood transfusion 83

Anemia has been associated with relatively poor Red cells transfusion prognosis among patients with various conditions, In cases such as elderly, malignancy and chronic including cancer, chronic kidney disease and conges- diseases, red cell concentrate is the component of tive heart failure [3]. choice, as whole blood carries the risk of volume Symptoms such as fatigue and shortness of breath overload. Red cells are prepared by centrifugation and are subjective, but are still useful in determining the separation from whole blood. It should be stored at need for red cell transfusion in patients with chronic 48C and has to be administered within 60 minutes of anemia [4]. issuance from controlled storage temperature. Red Anemia of chronic disease is not itself usually a cell concentrate should be administered through cause of symptoms. The anemia is mild and well standard [170 micron] giving set. tolerated unless it is superimposed on other threaten- One red cell concentrate unit should increase the ing conditions. The importance of recognizing anemia hemoglobin level in an adult by 1 g/dl. of chronic disease is in identifying its underlying cause Red cell concentrate [packed red cells, or plasma [5]. reduced red cells] is the simplest red cell component. This component can further be leukodepleted Special considerations when transfusing an using leukocyte filters, irradiated by gamma irradia- elderly, malignancy and chronic disease tion, washed by normal saline to remove the plasma proteins or suspended in additive nutritive Blood transfusion in general should be prescribed, solution to extend its shelf life and reduce its administered and monitored very carefully by a hematocrite. It is advisable to leukodeplete all red competent medical person. This is due to the fact cell concentrate units used for patients who are that it carries potential risks and hazards. The patient anticipated to receive multiple transfusions. This is may have several causes of anemia, such as malnutri- to avoid all the untoward consequences of leukocytes. tion deficiency, infection, malignancy or hemoglobi- Immune-compromised patients such as malignancy nopathy. There are few considerations which should under chemotherapy should receive irradiated cellular be considered in the transfusion management of the components to avoid the possibility of developing following three conditions:- GVHD. Compatibility of red blood cells with the intended

/ Elderly: the rate of red cell transfusion must be recipient must be verified by suitable pre-transfusion slower than in younger age patients. Elderly testing [2]. Cross-matching [compatibility testing] is patients are subjected to heart failure due done between the donor’s red cells and patient’s to blood volume overload if the rate of transfu- serum for every unit for transfusion. It is recom- sion is fast. Diuretics should be considered in mended to transfuse fully ABO, Rh and K compatible the appropriate dose when transfusion is inevi- red cells. Red cell antibody screening should be table. Moderate anemia warrants correction in performed in every transfusion; this is to avoid any patients older than 65 years old, especially those hemolytic reaction which in these conditions might with additional risk factors [such as coronary lead to very serious consequences. artery disease, pulmonary disease, or chronic The medical person who gives the transfusion to a kidney disease] or combination of these factors patient is responsible for the control of identity and [3]. other safety measures. Verification of identity shall be carried out either by asking the patient to tell his/her / Malignancy: usually patients with malignant disorders who need red cell transfusion are likely name or other identification details on a wrist band to need other blood components as well such as which has been attached to the patient according to platelet concentrates. Due to the frequency of well-specified rules [2]. Close monitoring of patients during transfusion is this therapy, one should consider at the begin- essential in order to manage any adverse effects in due ning of transfusion protocol several issues as time. The dose and frequency of red cell transfusion leukodepletion, irradiation and erythropoietic has to be appropriate to the patient’s age, weight, and agents. clinical condition. / Chronic disease: same considerations of malig- Because of the risk of damage to the blood nancy are applied in cases of anemia due components, no medical products or infusion solu- to chronic disease. In patients with renal failure tions may be added to blood units [2]. who are receiving dialysis and in patients with cancer who are undergoing chemotherapy, correction of anemia up to levels of 12 g/dl is Alternatives to red cell concentrate transfusion associated with an improvement of quality of Transfusion is usually not needed for chronic anemia, life [3]. and the unnecessary transfusions in these situations 84 F. Moftah may lead to unavoidable complications. Simple In order that blood transfusion becomes effective, it preventive measures and the use of oral iron replace- is very important to treat the underlying cause of ment can greatly reduce the prevalence of iron anemia whenever it is possible. The need for transfu- deficiency anemia and reduce the need for blood sion can often be avoided by either preventing or early transfusion [1]. diagnosis or treatment of anemia and conditions that Erythropoietic agents for patients with anemia of lead to anemia, or correction of anemia and the chronic disease are currently approved for use by replacement of depleted iron stores by prescribing patients with cancer who are undergoing chemother- oral iron in the relevant dose and frequency. apy, patients with chronic kidney disease, and patients Transfusion are particularly helpful in the context with HIV infection who are undergoing myelosup- of either severe anemia [B/ 8.0 g/dl] or life threatening pressive therapy [3]. anemia [B/ 6.5 g/dl] particularly when the condition is aggravated by complications that involve bleeding [3]. Transfusion guidelines Clinicians who are managing anemia in the above mentioned situations should have good knowledge The philosophy of the appropriate clinical use of about red cell components specially indications, risks, blood [ACUB] is a rational move to be adopted by storage and benefits. They should as well inform their clinicians. In order to achieve this practice, health care patients about the potential hazards and benefits of providers and systems should implement effective and red cell transfusion and alternative if relevant. sustainable health programs and services. This leads to the control and standardization of the blood transfusion practice. Recommendations Establishing national guidelines for blood transfu- . Red cell transfusion should be considered only sion protocols is a very important tool for proper when anemia has caused inadequate oxygen management of all conditions that need blood trans- supply to the needs of the patient. Otherwise fusion or alternatives, including anemias. Govern- patients should be managed early enough to ment support is very useful in this policy. avoid this situation. National guidelines should elaborate issues such as . Treatment of the underlying cause of anemia information on different blood components, blood usually prevents further decrease of oxygen ordering policies, compatibility, handling, monitoring carrying capacity, hence making management and management of adverse effects to blood. Many easier and avoids blood transfusion that carries countries have considered this and produced their potential hazards. national guidelines. WHO has issued a group of . National training programs should be in place publications regarding the use of blood components which aims at standardizing the practice of blood which are very reasonable and can be applied in many transfusion among all clinicians. health systems. . Decisions to transfuse blood should always be based on careful assessment of clinical and laboratory indicators that transfusion is necessary Management to save life or prevent significant morbidity [1]. One should always bear in mind that blood transfu- . Do not transfuse more than necessary, in other sion is an adjuvant therapy, and it is always accom- words do not restore laboratory hemoglobin panied by other types of treatment measures. These level. It should be raised to relieve clinical other treatment measures have to be taken into condition. consideration while ordering, administering and monitoring blood transfusion. It is very rational that every Conclusions patient is managed individually and golden rules to be avoided. Transfusion can be a life saving intervention, however Blood transfusion management in the above men- like all other treatments; it may result in acute or tioned cases depends on various factors such as age, delayed complications and carries the risk of transfu- underlying illness, duration & rate at which anemia sion transmissible infections [1]. developed and the degree of compensation. Red There are no reliable parameters to guide the need cell transfusion is considered only when anemia is for red cell transfusion. The decision to transfuse red severe enough to cause reduced oxygen supply to cell is a complex one and depends on factors such tissues leading to de-compensated situation. The as the cause of the anemia, its severity and chronicity, aim of red cell transfusion in this case is to supply the patient’s ability to compensate for anemia, the the patient with sufficient hemoglobin to improve likelihood of further blood loss and the need to the hypoxia and not to increase the lab hemoglobin provide some reserve before the onset of tissue level. hypoxia [4]. Blood transfusion 85

Clinical experience of blood prescribers is crucial to [2] Guide to the preparation, use and quality assurance of blood the management of anemia as there is no reliable components. 8th edition, council of Europe. [3] Guenter Weiss MD, Lawrence T, Goodnough MD. Anemia trigger or golden rule to be followed. of chronic disease. The New Engl J med March 2005;1011/ 1023. [4] Guidelines for the clinical use of red cell transfusion, Br J

References Haematol, 2001;113:24/31. [1] The clinical use of blood, WHO/BTS/99.2. [5] Cecil text book of medicine, 22nd edition. Hematology, 2005; 10 Supplement 1: 86/88

TRANSFUSION POLICIES IN ANEMIA

In search of the transfusion threshold

JEFFREY L. CARSON & RICHARD C. REYNOLDS

Division of General Internal Medicine, UMDNJ-Robert Wood Johnson Medical School, New Brunswick, New Jersey, USA

Introduction finding suggests animals with coronary artery disease are less tolerant of anemia than animals with normal Every medical decision involves weighting the risk hearts. versus benefit. The risks related to red blood cell transfusion include the adverse effects of allogeneic blood and the risks of anemia. The potential benefits Human data of blood transfusion are reduction of mortality and morbidity and improvement in functional recovery. In Studies in patients who decline blood transfusion for this lecture, I will focus on risks related to anemia and religious reasons provide critical insights into risks what is know about the benefits of allogeneic transfu- associated with anemia. In a small study of 125 sion. I will complete the presentation with a descrip- patients, [4] mortality rose with lower preoperative tion of a new clinical trial called FOCUS. hemoglobin levels and greater operative blood loss. The fatality rate was 61.5% in patients with preoperative hemoglobin concentrations below 6 gdl1, Risk from anemia but only 7.1% in patients with preoperative hemoglobin concentrations greater than 10 gdl1. In this Animal data small case series, patients with a hemoglobin level 1 The heart is the most vulnerable to the effects of above 8 gdl and operative blood loss below 500 ml, anemia. The heart extracts a high percentage of none of the patients died. oxygen. When anemia develops coronary blood flow The largest study in patients who refused blood must increase to maintain oxygen delivery. Blood flow transfusion confirmed animal data that patients with increases because of decrease blood viscosity that cardiovascular disease are more susceptible to anemia results from anemia. then as patients without cardiovascular disease. In a A series of experiments have been performed in cohort study in 1,958 adult surgical patients who canines to evaluate the effect of anemia (Table I). underwent a surgical procedure in an operating room, After hemodilution, healthy animals survive hemoglo- [5] mortality rose as preoperative hemoglobin levels 1 bin concentrations between 3 and 5 gdl [1/3]. fell. However, patients with cardiovascular disease However, myocardial ischemia is detectable by elec- had a much higher risk of death as the hemoglobin trocardiograph changes at hemoglobin concentrations concentration fell below 10 gdl 1 than patients with- below 5 gdl1. At hemoglobin levels around 3 gdl1, out cardiovascular disease (Figure 1). severe physiologic derangement develops. This is A recent study combined data from two cohorts of manifested by anerobic metabolism (lactate produc- Jehovah’s Witness patients and examined the morbid- tion) myocardial dysfunction, and death. Some ani- ity and mortality associated with postoperative hemo- mals survive with hemoglobin levels as low as 1 to 2 globin concentrations below 8 gdl 1[6]. Of 2083 gdl1. consecutive patients, 300 (15%) had postoperative 1 Animals with experimentally induced coronary Hgb levels B/8 gdl . None of the patients with artery disease poorly tolerate anemia. In the presence postoperative hemoglobin levels between 7.1/8.0 of coronary stenosis from 50% to 80%, myocardial died although 9.4% had significant morbidity. The ischemia and/or reduced cardiac function develop at 30 day mortality was 34% when the postoperative 1 1 hemoglobin levels between 7 to 10 gdl . This hemoglobin level fell to 4.1/5.0 gdl .

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389719 In search of the transfusion threshold 87

Table I. Effects of anemia in animals undergoing hemodilution. the trial, patients were randomized to a transfusion Hemoglobin strategy in which blood was administered when the 1 Group Event Concentration (gdl 1) hemoglobin concentration fell below 7.0 gdl (and maintained between 7.0 to 9.0 gdl1). In the liberal B Normal ST segment changes /5 transfusion group, patients were transfused to main- Lactate production B/3 tain hemoglobin concentration between (10.0 gdl1 Decreased ventricular function B/3 1 Death B/3 and 12.0 gdl ). Surprisingly, the 30-day mortality was slightly lower (not significant) in the restrictive Coronary ST segment changes 7/10 Artery transfusion group than the liberal group (18.7% vs. Disease 23.3%). Overall, these finding were not statistically significant, although in patients less than 55 years of The effect of anemia was assessed in normal age and with Apache scores less than 20 (less ill volunteers undergoing isovolemic anemia to 5 patients), mortality was significantly better in patients gdl1. Transient electrocardiogram changes occurred randomized to the restrictive transfusion group. The 1 results were different in patients with cardiovascular with hemoglobin levels between 5/7 gdl in 5.7% of volunteers [7,8]. Subtle changes in cognition were disease. In patients with ischemic heart disease found in young volunteers with hemoglobin levels (defined as myocardial infarct, angina, congestive 1 heart failure, and cardiogenic shock), the liberal between 5/7 gdl [9]. Fatigue developed when the hemoglobin level fell to 7 gdl1 and increased as the transfusion group had a small and non-significant hemoglobin level dropped to 5 gdl1 [10]. Heart rate improved outcome. is linearly related to hemoglobin concentration [11]. These studies suggest that important clinical effects Observational studies can be measured in young, normal humans with 1 hemoglobin levels between 5/7 gdl . It is unclear The seven large observational studies that evaluated how these results relate to sick patients undergoing the transfusion thresholds came to different conclu- surgery but it seems possible such patients would be sions and must be interpreted very cautiously. Studies even less tolerant of anemia. in critically ill patients [15,16], patients undergoing coronary artery bypass surgery, [17] and surgical patients undergoing hip fracture repair [18] came to Efficacy of transfusion different conclusions. A very large study in acute myocardial infarction suggested [19] mortality was Clinical trial data reduced by transfusion in patients with hematocrit Clinical trials are essential to establish efficacy of any levels less than 33%. However, a more recent analysis treatment including blood transfusion. Unfortunately, of clinical trials in acute coronary syndrome found there is limited evidence. Of the 10 randomized that blood transfusion increased mortality [20]. Three clinical trials, all but one was too small to evaluate small studies found higher rates of cardiac events in clinical outcomes [12]. A meta-analysis that com- anemic patients. Thus, there is great heterogeneity of bined the results of these trials, [12] found the the study populations and results. Furthermore, the Transfusion Requirements in Critical Care (TRICC) results of these observational studies should be inter- trial contributed 83% of the information for the preted very cautiously because statistical approaches analysis of mortality in the meta-analysis [13]. The to adjusting for differences between patients receiving other trials were too small to reliably evaluate the transfusion versus those not transfused may not be effect of transfusion thresholds on clinical events. successful. This point is emphasized by the results The TRICC trial evaluated two transfusion triggers of recent clinical trials evaluating hormone replace- in 838 volume resuscitated patients admitted to ment therapy where the results on cardiovascular risk intensive care unit [13,14]. In the restrictive arm of were opposite of observational studies. Similarly, the

13 No Cardiovascular Disease 10 Yes-Cardiovascular Disease 7 Odds Ratio Odds

1 6 7 8 9 10 11 12+ Preoperative Hg (g/L) Figure 1. Odds of death in patients who decline blood transfusion stratiﬁed by presence of cardiovascular disease [5]. 88 J. L Carson & R. C. Reynolds

TRICC investigators found difference in their clinical [3] Anderson HT, Kessinger JM, McFarland WJ, Jr., et al. trial than their observational study [15]. Response of the hypertrophied heart to acute anemia and coronary stenosis. Surgery 1978;84:/ 8/ /15. [4] Carson JL, Poses RM, Spence RK, et al. Severity of anaemia FOCUS: A new clinical trial and operative mortality and morbidity. Lancet 1988;1:/ 727/ / 9. Given the limited high quality evidence, we have [5] Carson JL, Duff A, Poses RM, et al. Effect of anaemia and initiated a new NIH funded trial called FOCUS. The cardiovascular disease on surgical mortality and morbidity. transfusion trigger trial for Functional Outcomes in Lancet 1996;348:/ 1055/ /60. [6] Carson JL, Noveck H, Berlin JA, et al. Mortality and Cardiovascular Patients Undergoing Surgical Hip morbidity in patients with very low postoperative Hb Fracture Repair (FOCUS) is a randomized clinical levels who decline blood transfusion. Transfusion 2002;42:/ / trial designed to test the hypothesis that higher blood 812/8. transfusion threshold improves functional recovery [7] Leung JM, Weiskopf RB, Feiner J, et al. Electrocardiographic and reduces morbidity and mortality. Patients who ST-segment changes during acute, severe isovolemic hemodi- undergo surgery for hip fracture, have a history of lution in humans [In Process Citation]. Anesthesiology 2000;/ cardiovascular disease, and have a postoperative 93:1004/ /10. 1 [8] Weiskopf RB, Viele MK, Feiner J, et al. Human cardiovascular hemoglobin level less than 10 gdl within 3 days of and metabolic response to acute, severe isovolemic anemia surgery are eligible. Patients will be randomized to [see comments]. JAMA 1998;279:/ 217/ /21. receive enough blood to raise the hemoglobin level [9] Weiskopf RB, Kramer JH, D P, et al. Acute severe isovolemic above 10 gdl1 any time the hemoglobin level is anemia impairs cognitive function and memory in humans. detected to be below 10gdl 1 during the hospitaliza- Anesthesiology 2000;92:/ 1646/ /52. [10] Toy P, Feiner J, Viele MK, et al. Fatigue during acute tion or to receive transfusion if symptoms of anemia isovolemic anemiain healthy, resting humans. Transfusion develop. Transfusion is permitted but not required if 2000;40:/ 457/ /60. 1 hemoglobin level is less than 8 gdl . The primary [11] Weiskopf RB, Feiner J, Hopf H, et al. Heart rate increases outcome is ability to walk 10 feet (or across a room) linearly in response to acute isovolemic anemia. Transfusion without human assistance at 60 days. The most 2003;43:/ 235/ /40. important secondary outcome is postoperative un- [12] Carson JL, Hill S, Carless P, et al. Transfusion triggers: A systematic review of the literature. Arch of Int Med 2001. stable angina, myocardial infarction or death. Medical [13] Hebert PC, Wells G, Blajchman MA, et al. A multicenter, records will be reviewed while the patient is in the randomized, controlled clinical trial of transfusion require- hospital. Patients will be telephoned at 30 and 60 days ments in critical care. Transfusion Requirements in Critical after entry into the study to determine functional Care Investigators, Canadian Critical Care Trials Group [see capacity and vital status. Long term mortality will be comments]. N Engl J Med 1999;340:/ 409/ /17. determined by searching vital statistic registries in US [14] Hebert PC, Wells G, Marshall J et al. Transfusion requirements in critical care. A pilot study. Canadian Critical Care and Canada. The pilot study for the trial has been Trials Group [published erratum appears in JAMA 1995 Sep published [21]. 27;274(12):944]. JAMA 1995;273:1439/44. [15] Hebert PC, Wells G, Tweeddale M, et al. Does transfusion practice affect mortality in critically ill patients? Transfusion Summary Requirements in Critical Care (TRICC) Investigators and the We have very limited data to guide the transfusion Canadian Critical Care Trials Group. Am J Resp & Crit Care decision. The best evidence suggests that mortality Med 1997;155:/ 1618/ /23. [16] Vincent JL, Baron JF, Reinhart K, et al. Anemia and blood and morbidity rises as the blood count falls but that in transfusion in critically ill patients. JAMA 2002;288:/ 1499/ / 1 patients without cardiovascular disease a 7 gdl level 507. 1 is tolerated in most patients. Below 5/6 gdl , the [17] Spiess BD, Ley C, Body SC, et al. Hematocrit value on mortality and morbidity rises rapidly. In patients with intensive care unit entry inﬂuences the frequency of Q-wave cardiovascular disease, animal and human studies myocardial infarction after coronary artery bypass grafting. suggest that a higher hemoglobin concentration may The Institutions of the Multicenter Study of Perioperative Ischemia (McSPI) Research Group. J Thorac Cardiovasc Surg be necessary but there is limited evidence from 1998;116:/ 460/ /7. clinical trials to inform this question. We expect that [18] Carson JL, Duff A, Berlin JA, et al. Perioperative blood

results from the new clinical trial called FOCUS will transfusion and postoperative mortality. JAMA 1998;279:/ / provide important new information to help guide 199 /205. transfusion decisions. [19] Wu WC, Rathore SS, Wang Y, et al. Blood transfusion in elderly patients with acute myocardial infarction. N Engl J Med 2001;345:/ 1230/ /6. [20] Rao SV, Jollis JG, Harrington RA, et al. Relationship of blood References transfusion and clinical outcomes in patients with acute [1] Wilkerson DK, Rosen AL, Sehgal LR, et al. Limits of cardiac coronary syndromes. JAMA 2004;292:/ 1555/ /62. compensation in anemic baboons. Surgery 1988;103:/ 665/ /70. [21] Carson JL, Terrin ML, Barton FB, et al. A pilot randomized [2] Hagl S, Heimisch W, Meisner H, et al. The effect of trial comparing symptomatic vs. hemoglobin- level-driven red hemodilution on regional myocardial function in the presence blood cell transfusions following hip fracture. Transfusion of coronary stenosis. Basic Res Cardiol 1977;72:/ 344/ /64. 1998;38:/ 522/ /9. Hematology, 2005; 10 Supplement 1: 89/91

CURRENT THERAPIES IN HEMOGLOBINOPATHIES

Combined use of oral chelators and desferrioxamine in thalassemia

A. PIGA, S. ROGGERO, F. MARLETTO, L. SACCHETTI, & F. LONGO

The standard iron chelation therapy is based on the patients an additive effect may be reached combining use of deferoxamine (DFO) [1]. A subcutaneous the two chelators. In some subject a synergistic effect infusion of 20/50 mg/kg/day over 8/12 hours 6/ has been observed. A ‘‘shuttle’’ hypothesis has been 7 days a week promotes a total iron excretion of formulated to explain the described synergistic effect:

0.15/0.5 mg/kg/day [2]. This may counterbalance the DFP, crossing cell membranes more easily, could bind mean iron input from standard transfusional regimens excess intracellular iron and mobilize into plasma, of 0.25/0.5 mg/kg/day [3]. The effectiveness of this where DFO, with its higher affinity, may take it and treatment is mainly determined by the compliance, speed up its excretion. that may vary significantly [4]. During these last few years the oral chelator deferiprone (DFP) has been approved in many countries and gained room at least Clinical studies as a second line option for patients where DFO is not Several studies have been published or presented at tolerated or inadequate [5,6]. The relative lower meetings on long-term use of the two drugs [10/18]. efficacy is partially counterbalanced by the advantages Unfortunately none of these is a randomized in compliance due to the administration route. Recent controlled trial and most do not satisfy high quality results from independent studies suggest that defer- standards, nor the safety has been investigated iprone may be more cardio protective than deferox- systematically. Furthermore many different treatment amine. Patients on long-term treatment with schemes have been used, making difficult to summar- deferiprone have a better myocardial MRI pattern ize findings. Anyway the results seem to indicate [7], and less chance to develop a new cardiac disease a higher efficacy of combination therapy, in terms or to worsen an existing one [8]. of urinary iron excretion, lowering serum ferritin levels, trend to normalize in iron-related MRI abnormalities both in the liver and in the heart, Rationale and in systolic function parameters. These observa- The efficacy and tolerability profiles of DFO and DFP tions, as the hypothesis that deferiprone could are very different, due mostly to the different physio- be faster or more efficient in removing excess iron chemical and pharmacological properties: molecular from the heart have raised a lot of attention from weight, Fe:chelator molar ratio, Fe affinity, Fe binding clinicians, but should be verified with formal con- stability, and excretion. Some in vitro data suggested trolled studies. the possibility of an additive effect of combining the two chelators. A recent study from Link showed that Definition the combination of DFO and DFP improved significantly the iron depletion rate from rat heart cells [9]. A problem of terminology exists: in papers and Clinical data are scarce. Grady performed the first presentations sometimes the prescription and dose exhaustive study on iron balance in thalassemia, timing is not fully described and under the umbrella assessing in each patient the fecal and urinary iron ‘combination’ are associated very different treatments excretion with each chelator and finally in combina- such as the daily taking of both drugs at full doses and tion. The preliminary results show that in most simultaneously, or alternating the two drugs in some

Correspondence: A. Piga, Thalassemia Center, Department of Pediatric Hematology-Oncology, University of Torino, Italy. E-mail: [email protected]

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331389737 90 A. Piga et al. way during the week. An approach to a common Side effects terminology could be the following: This page is still to be written, as the experience with combination treatment is relatively recent. From the / Mono-therapy: a single chelator is prescribed published studies it does not come out a trend to and taken for more than three months increase of the known side effects of both drugs nor / Alternate therapy: in a single day a single the evidence of new ones. A raised prevalence of chelator is taken; the two chelators take turn on agranulocytosis has not been confirmed. a weekly, monthly or quarterly basis (e.g., DFP five days a week and DFO two days a week)

/ Combination therapy: prescription of more New iron chelators than one chelator, to be taken in the same day The development of new iron chelators may enhance at least for a significant part of the period greatly the possibilities of combination treatment. In . Sequential: in a single day two chelators are particular ICL670, a new tridentate oral chelator has taken in sequence; no substantial overlapping good chances to be approved in the near future, due to of the two drugs in the plasma (e.g., DFP the positive results of extensive clinical studies thrice a day and DFO night time) [19,20]. Its pharmacokinetics and pharmacodynamics . Simultaneous or concomitant: in a single offer interesting perspectives of potential combination day two chelators are taken at the same time; with DFO or DFP. Well designed randomized con- substantial overlapping of the two drugs in the trolled trials may answer important questions, but the plasma (e.g., DFO infusion starts at 7 PM and sponsorship of combination therapy trials from in- ends at 7 AM; DFP taken at 8PM, 11 PM and dustry is unlikely. 7 AM).

Some other forms of combination exist, as a course of Conclusions intravenous DFO at each transfusion to reinforce the For the clinical practice: there is growing evidence iron chelation done at home. that DFO and DFP combination treatment is more effective than s.c. DFO or DFP alone and may give Potential applications benefit in certain clinical conditions such as heart disease. Little is known as regards the safety. Combination treatment should be potentially consid- For the research: the results of many observations ered every time there is a need to search for an are large randomized controlled trials on combination additive or synergistic effect. This include the reversal therapy compared to mono-therapy. Clinically rele- of severe iron-related complications, the most impor- vant outcomes should be studied, as liver iron tant of which is heart disease. At a new onset of heart concentration on short-term and heart disease and disease it is important to provide a full protection survival on long-term. from cardio-toxicity with a continuous treatment, to minimize the presence of non-transferrin bound iron in the plasma. Intensive DFO chelation with contin- References uous intravenous infusion has been demonstrated to [1] Propper RD, Cooper B, Rufo RR, et al. Continuous sub- be effective. The addition of deferiprone seems to cutaenous administration of deferoxamine in patients with iron enhance the efficacy and to reduce the duration of overload. N Engl J Med 1977;297:/ 418/ /423. intensive treatment. [2] Hussain MA, Green N, Flynn DM, Hoffbrand AV. Effect of dose, time, and ascorbate on iron excretion after subcutaneous Also for the prevention of iron-related complica- desferrioxamine. Lancet 1977;1:/ 977/ /979. tions may be an important field of application. The [3] Porter JB. Practical management of iron overload. Br J recent development of MRI techniques to quantify Haematol 2001;115:/ 239/ /252. the heart iron and the relationship of MRI signal to [4] Gabutti V, Piga A. Results of long-term iron-chelating therapy. Acta Haematol 1996;95:/ 26/ /36. systolic function give the rationale for a more effective [5] Olivieri NF, Brittenham GM, Matsui D, et al. Iron-chelation prevention of this complication. therapy with oral deferipronein patients with thalassemia Another indication may be the achievement of safe major. N Engl J Med 1995;332:/ 918/ /922. tissue levels, in any patient where the iron overload is [6] Cohen AR, Galanello R, Piga A, De SV, Tricta F. Safety and effectiveness of long-term therapy with the oral iron chelator severe or in special conditions, such as the preparation deferiprone. Blood 2003;102:/ 1583/ /1587. to pregnancy or to a stem cell transplant or to antiviral [7] Anderson LJ, Holden S, Davis B, et al. Cardiovascular T2-star treatment for hepatitis. (T2*) magnetic resonance for the early diagnosis of myocar- Another potential application regards patients with dial iron overload. Eur Heart J 2001;22:/ 2171/ /2179. dose-related side effects to DFO or DFP. Individually [8] Piga A, Gaglioti C, Fogliacco E, Tricta F. Comparative effects of deferiprone and deferoxamine on survival and cardiac tailored combination of the two drugs may minimize disease in patients with thalassemia major: a retrospective side effects, maintaining efficacy. analysis. Haematologica 2003;88:/ 489/ /496. Combined use of oral chelators and desferrioxamine in thalassemia 91

[9] Link G, Konijn AM, Breuer W, Cabantchik ZI, Hershko C. chelation in thalassemia. J Pediatr Hematol Oncol 2004;26:/ /

Exploring the ‘‘iron shuttle’’ hypothesis in chelation therapy: 451/453. effects of combined deferoxamine and deferiprone treatment [16] Gomber S, Saxena R, Madan N. Comparative efﬁcacy of in hypertransfused rats with labeled iron stores and in iron- desferrioxamine, deferiprone and in combination on iron

/ / loaded rat heart cells in culture. J Lab Clin Med 2001;138: chelation in thalassemic children. Indian Pediatr 2004;41:/ / 130 /138. 21/27. [10] Wonke B, Wright C, Hoffbrand AV. Combined therapy with [17] Wu KH, Chang JS, Tsai CH, Peng CT. Combined therapy

deferiprone and desferrioxamine. Br J Haematol 1998;103:/ / with deferiprone and desferrioxamine successfully regresses 361 /364. severe heart failure in patients with beta-thalassemia major. [11] Balveer K, Pyar K, Wonke B. Combined oral and parenteral Ann Hematol 2004;83:/ 471/ /473. iron chelation in beta thalassaemia major. Med J Malaysia [18] Tsironi M, Deftereos S, Andriopoulos P, et al. Reversal of 2000;55:/ 493/ /497. heart failure in thalassemia major by combined chelation [12] Hoffbrand VA, Wonke B. Long term deferiprone chelation therapy: a case report. Eur J Haematol 2005;74:/ 84/ /85. therapy. Adv Exp Med Biol 2002;509:/ 127/ /139. [19] Galanello R, Piga A, Alberti D, et al. Safety, tolerability, and [13] Mourad FH, Hoffbrand AV, Sheikh-Taha M, et al. Compar- pharmacokinetics of ICL670, a new orally active iron- ison between desferrioxamine and combined therapy with chelating agent in patients with transfusion-dependent iron desferrioxamine and deferiprone in iron overloaded thalassae-

overload due to beta-thalassemia. J Clin Pharmacol 2003;43:/ / mia patients. Br J Haematol 2003;121:/ 187/ /189. [14] Alymara V, Bourantas D, Chaidos A, et al. Effectiveness and 565 /572. safety of combined iron-chelation therapy with deferoxamine [20] Nisbet-Brown E, Olivieri NF, Giardina PJ, et al. Effectiveness and safety of ICL670 in iron-loaded patients with thalassae- and deferiprone. Hematol J 2004;5:/ 475/ /479. [15] D’Angelo E, Mirra N, Rocca A, Carnelli V. Combined therapy mia: a randomised, double-blind, placebo-controlled, dose- with desferrioxamine and deferiprone: a new protocol for iron escalation trial. Lancet 2003;361:/ 1597/ /1602. Hematology, 2005; 10 Supplement 1: 92/99

CURRENT THERAPIES IN HEMOGLOBINOPATHIES

Sickle cell disease: A multigenic perspective of a single gene disorder

ABDULLAH KUTLAR

Professor of Medicine, Medical College of Georgia, Director, Sickle Cell Center, Augusta, GA

Half a century has elapsed since the elucidation of the role in the pathogenesis of these syndromes. In the molecular basis of the sickle hemoglobin mutation in past two decades, contributions by Hebbel et al. and the mid 50s. Although significant progress has been others, has established that SCD is an ‘‘inflammatory made in our understanding of the disease and in the state’’ and that the endothelium is activated [2]. In development of new therapies, many questions are more recent years, evidence has accumulated that still unanswered, and a cure remains elusive. This is decreased nitric oxide (NO) bioavailability, contrib- particularly evident in the clinical heterogeneity of uted by scavenging of NO by cell free hemoglobin, a sickle cell disease (SAD), which ranges from a mild product of the hemolytic process, plays a significant clinical course with survival into the 6th and 7th role in the vascular pathobiology of SCD [3]. Thus, decades to a very severe presentation with significant the emerging view of the pathogenesis of many organ damage and death at relatively young ages [1]. complications of SCD involves complex interactions Most of the advances in unraveling the phenotypic between sickle reticulocytes, neutrophils, monocytes, heterogeneity of SCD have been made in the past 30 and the endothelium. It therefore follows that many years, after the dawn of the molecular era. Studies in factors that are important in the pathways leading to this period have clearly shown the importance of high inflammation, cell adhesion, NO metabolism, vascu- hemoglobin F (Hb F) determinants and a-thalassemia lar reactivity and coagulation will be important in the as modifiers of disease severity. The impact of these pathophysiology of the disease and that variations in modifiers has been well established and will not be the expression of a number of molecules in these discussed in detail here [1]. We will rather try to pathways will contribute to the heterogeneity of SCD. summarize recent advances on the effect of non- The approach to the study of this complexity in the globin genetic modifiers on the phenotype of SCD. post-genome era should involve several different This process has been greatly facilitated by the methodologies including: (1) analysis of polymorph- completion of the Human Genome Project, which isms in ‘‘candidate genes’’ by high throughput meth- has identified a large number of single nucleotide ods and the association of these polymorphisms with polymorphisms (SNPs) in and around many genes distinct phenotypic features of the disease, (2) the that can potentially impact on different aspects of study of differential gene expression in various tissues SCD pathophysiology. The realization that a-thalas- using cDNA microarrays, and (3) the application of semia and variation in Hb F levels alone do not proteomics. The tissues and cells that are particularly account for the vast clinical diversity of the disease has important in this regard include blood cells (neutro- provided further impetus to such an endeavor. phils, monocytes, reticulocytes), bone marrow, en- The well-known consequences of the sickle muta- dothelial cells, and liver tissues. A candidate gene tion and its downstream effects are: (1) a chronic approach should include a large number of patients hemolytic anemia, (2) episodic vasoocclusion with with and without a certain complication (e.g., stroke) resultant characteristic painful episodes, and (3) to determine the role of these polymorphisms as risk chronic organ damage. The clinical phenotype of factors. Pooling studies to discover the role of genome SCD is highly variable; complications such as acute wide SNPs as risk factors for certain complications of chest syndrome, stroke, leg ulcers, and avascular the disease is also emerging as a viable approach. In all necrosis do not occur in all of the patients suggesting of these studies, careful characterization of the phe- that factors beyond the b6 Glu0/Val mutation play a notype remains of critical importance. A study of

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331390069 Sickle cell disease 93

Table I. Frequency of various complications among patients with responsible for the variation seen in Hb F levels SCD. among SCD patients. This has prompted the search Phenotype Frequency for other trans-acting regulatory elements controlling Hb F levels. One of these putative elements is a Stroke/CVA 11% Retinopathy 3% quantitative trait locus (QTL) on the X-chromosome

Acute Chest Syndrome 30/50% (Xp22), which presumably controls the production of

Gallstones 30/36% F-cells. This F-cell production locus (FCP) and the Splenic sequestration 5 /6% /158 C0/T polymorphism in the Gg promoter are Nephropathy/microalbuminuria 30% estimated to account for /50% of the variation in Hb Priapism 6/42% Avascular Necrosis 50% F in SCD [5]. Other QTLs that influence Hb F levels / / Leg ulcers 4/6% have been located at 6q22.3 23.2, and 8q [6 8]. In a study of 180 single nucleotide polymorphisms (SNPs) in 38 candidate genes in 280 SCD patients, the sickle cell patients from different populations is likely strongest association with Hb F levels was found to yield important information because of the differ- with SNPs in the 6q 22.3/23.2 region. Detailed ences of the genetic backgrounds in these populations analyses of this region, identified 12 SNPs in the and its potential implications on the disease pheno- introns of four genes, associated with a 20/30% type. The important disease related ‘‘phenotypes’’ variation in Hb F [9]. The genes in question were should include features such as stroke and CNS phosphodiesterase 7 (PDE7), microtubule-associated disease, frequency of vasoocclusive episodes (VOEs), protein 7 (MAP7), peroxisomal biogenesis factor 7 frequency of the acute chest syndrome, avascular (PEX7) and mitogen-activated protein kinase 5 necrosis of the hips and shoulders, gallbladder disease (MAP3K5). Although the precise mechanisms of and cholecystectomy, leg ulcers, renal involvement, this effect have not yet been identified, there is some priapism, and retinopathy. Table I lists these impor- data to suggest that the product of these genes may be tant ‘‘sub-phenotypes’’ of SCD and their frequency in involved in the regulation of g-globin gene expression the SCD patient population in the U.S. in various cellular systems (K562 cells). For the past several years, our center and many In a collaborative study between Boston University other centers have focused on the study of non-globin and the Medical College of Georgia Sickle Cell genetic modifiers of sickle cell disease. Some of the Centers, genetic determinants of Hb F response to findings of these early studies will be summarized hydroxyurea was studied in 214 African/American below. SCD patients. Two hundred twenty six SNPs in 46 candidate genes associated with Hb F regulation and Hb F and Hb F response to hydroxyurea drug metabolism were analyzed. SNPs in two genes, in CYP2C9, a member of the cytochrome P 450 It is well established that fetal hemoglobin has an family and in Aquaporin 9 (AQP9), a membrane ameliorative effect on SCD because the gamma chains channel that stimulates urea transport and transports of Hb F are excluded from the deoxyhemoglobin S uncharged solutes, were associated with a good polymer. Because of this potent anti-sickling effect, response to hydroxyurea [10]. enhancing Hb F production has been a major therapeutic goal in SCD. In fact, this goal has been achieved to a large extent by the use of the S-phase Genetic polymorphisms as risk factors for stroke specific chemotherapeutic agent, hydroxyurea [4]. Stroke occurs in 11% of US sickle cell patients by 20 Several genetic determinants are known to contri- years of age [11]. Increased TCD (Transcranial bute to the heterogeneity of baseline Hb F levels in Doppler) velocities of blood flow (/200 cm/sec) in SCD. Haplotypes of the bS chromosomes are among major intracranial arteries has been shown to be a the most extensively studied. Senegalese and Asian/ predictor of stroke risk in children with SCD. The Indian haplotypes are associated with higher Hb F value of TCD as a surrogate marker of cerebrovas- levels, and therefore a milder clinical and hematologic cular disease and a predictor of stroke risk has been phenotype compared to the other African haplotypes validated in the randomized STOP trial, and a (Benin, Bantu, and Cameroon). The common feature therapeutic intervention based upon this risk stratifi- of the Senegal and Asian haplotypes is the presence of cation (prophylactic transfusions in children with high aC0/T polymorphism in the promoter of the Gg gene TCD velocities) has been proven to reduce the stroke

(/158 C0/T, detectable by the restriction enzyme risk [12]. Despite the validation of the utility of TCD Xmn I). The search for other cis-acting regulatory as a clinical tool for predicting stroke risk, biologic elements in the b-globin cluster has failed to yield factors underlying the development of cerebrovascular consistent results. Thus, it has become clear that bS disease and stroke in the SCD population are poorly haplotypes and variations in cis-acting elements understood. The presence of a-thalassemia has been associated with different haplotypes are only partially shown to be protective of cerebrovascular disease and 94 A. Kutlar stroke [11]. Several recent studies have sought an Table II. Association of SNPs with stroke in SCD. association between certain genetic polymorphisms SNP Effect Reference and stroke risk in SCD. Driscoll et al. [13] found a VCAM1 G1238C Protective Taylor et al. 2002 higher rate of stroke in siblings with SCD (P/ VCAM1 T-1594C Permissive Hoppe et al. 2004 0.0012), suggesting the role of genetic factors. A IL4R S503P Permissive Hoppe et al. 2004 recent study has found an association with HLA TNF-a -308 Protective Hoppe et al. 2004 genotypes and stroke (HLA DRB1*0301 and HLA LDLR/Nco1 Protective Hoppe et al. 2004 DRB1*0302 genotypes were associated with in- ADRB2 Q27E Protective Hoppe et al. 2004 creased stroke risk); of particular interest is the HLA genes Protective and Hoppe et al. 2002 permissive association of certain genotypes with distinct subtypes BMP6 Permissive Sebastiani et al. 2005 (large vessel vs. small vessel stroke) in a retrospective TGFBR2 Permissive Sebastiani et al. study of the CSSCD population [14,15]. This latter TGFBR3 Permissive Sebastiani et al. study found an association between an Interleukin-4 P-selectin Permissive Sebastiani et al. receptor polymorphism (IL-4R S503P), TNF-a - 308G, and b-adrenergic receptor-2 (ADRb2 Q27E) TM polymorphisms and large vessel stroke. In the small based Massarray system (Sequenom, San Diego, vessel stroke group, VCAM T -1594C polymorphism CA). Data analyses are currently ongoing. A summary and low density lipoprotein receptor LDLR NcoI of SNPs associated with stroke risk is summarized in polymorphism emerged as risk factors. The combina- Table II. tion of TNF-a -308 GG homozygosity and the IL- 4RS503P polymorphism conferred a particularly Avascular necrosis strong risk for large vessel stroke (Odds ratio/5.5) in this study. Another recent study [16] reported a Studies of large patient populations such as the protective role for VCAM G1238C polymorphism CSSCD (Cooperative Study of Sickle Cell Disease) against symptomatic stroke (odds ratio/0.35, P/ have shown that a-thalassemia, age, high hematocrit, 0.04). Steinberg et al. [17] studied 113 SCD patients and frequent VOEs are risk factors for the develop- with a history of stroke and 493 controls. They found ment of avascular necrosis (AVN) of the femoral head that polymorphisms in four genes were associated in SCD [19]. We undertook a survey of the frequency with stroke risk: Klotho (KL), TGF-beta receptor of the thermolabile MTHFR (methylenetetrahydro- (TGFbR3), Annexin 2 (ANXA2), and bone morpho- folate reductase) C677T mutation in the sickle cell genic protein 6 (BMP6). The same group of investi- population, because of its association with elevated gators utilized a Bayesian network approach to serum homocysteine levels and resultant vascular analyze gene-gene interactions to develop a risk model complications and in particular, its relationship to for stroke in SCD. They analyzed 235 SNPs in 80 avascular necrosis. Overall, MTHFR was present in candidate genes in1398 unrelated subjects with SCD. 16% of the sickle cell patients in our center (1.8% In addition to four clinical variables including a- homozygote). There was a strong association of the thalassemia and Hb F, they found that SNPs in 11 presence of MTHFR with AVN with 35.6% of the genes interacted in a complex manner to modulate AVN patients having the MTHFR mutation as stroke risk. These included SNPs in BMP6, Trans- opposed to only 12.9% of sickle cell patients without forming Growth Factor beta-receptors (TGFbR2 and AVN (P/0.006). Furthermore, the presence of con- TGFbR3), and P-Selectin (SELP). Using this model comitant a-thalassemia and MTHFR was found to be to predict stroke in a different group of SCD patients, additive in terms of the risk of developing AVN [20]. they reported an overall predictive accuracy of 98.2% This association however was not confirmed in the [18]. A recently completed study at our center is high Hb F population of Kuwaiti sickle cell patients, looking at the association of high TCD (as a risk suggesting that different genetic factors may be factor for ischemic stroke) with 28 genetic polymorphisms in 25 candidate genes in 630 patients operative in different populations [21]. Some other (230 high TCD, 400 normal TCD) in STOP and smaller studies in different patient populations have STOP II trials. The candidate genes include those also failed to show an association between this associated with coagulation and thrombophilia (Fac- MTHFR polymorphism and the risk of avascular tor V, Factor VII, Factor XIII, prothrombin, throm- necrosis in SCD [22 /24]. A recently published study bomodulin, fibrinogen, PAI-1, MTHFR), endothelial [25] analyzed 442 patients with AVN and 455 SCD cell function and inflammation (VCAM-1, ICAM-1, controls from the CSSCD cohort. They studied SNPs selectins, TNF-a, Apo A and Apo E), platelet func- in 66 candidate genes and found significant association and activation (GpIIb/IIIa, GpIb IX-V, GpIa/ tion of AVN with 7 SNPs in 7 genes. These are: IIIa), and vascular reactivity (ACE). The analysis of BMP6, TGFBR2, TGFBR3, EDN1 (Endothelin-1), polymorphisms was accomplished by a high through- ERG (v-ets erythroblastosis virus E26 oncogene like), put SNP genotyping method using the MALDI-TOF KL (Klotho), and ECE1 (Endothelin converting Sickle cell disease 95

Table III. R485K Polymorphism and catheter induced thrombosis. Table IV. UGT1A1 Polymorphism and bilirubin levels in sickle cell disease. R485K Thrombosis n/10 Control n/10 7/7 (n/22) 6/7 (n/35) 6/6 (n/10)

/// 62

/// 32Retics 10.9 11.4 11.3

/// 16Hb F 6.6 11. 8.4 Total bilirubin 5.6 3.8 3.8 P/0.006; Odds Ration/4.9. Cholecystectomy 59% 51% 60% enzyme 1). The precise mechanism(s) whereby variation in these genes causes AVN is not yet understood. xyurea therapy was recently reported by Heeney et al. [35]. These authors reported a normalization of the bilirubin levels in sickle cell patients with the 6/6 Catheter induced thrombosis and factor V genotype upon treatment with hydroxyurea in con- R485K polymorphism trast to those with 7/7 and 6/7 genotypes whose First described as a neutral polymorphism in the bilirubins failed to fall to normal levels. An interesting Factor V gene in nucleotide 1628 in exon 10 by aspect of the effect of this polymorphism was reported

Gandrille et al. the Arg0/Val substitution at residue by Haverfield et al. in the Jamaican study. These 485 was found at high frequency (32.4%) in sub- investigators found an association between the (TA)7/ Saharan Africans [26]. This polymorphism was later (TA)7 genotype and symptomatic gallstones except associated with a risk of thrombosis and coronary for the younger cohort. Similarly in our study, in B artery disease in Asian populations and reported to multivariate analyses, in the younger age group ( /10 lead to a mild activated protein C (APC) resistance years of age), Hb F levels were found to be a more [27]. A study of the frequency of this polymorphism important determinant of bilirubin levels then the in our sickle cell population showed a similar gene UGT1A1 genotype. These observations indicate that frequency of 0.29 (45.1% heterozygous, 6.6% homo- the influence of genetic modifiers on a particular zygous). Although no association of this polymorph- phenotype may vary with age in patients with SCD. ism was established with some complications of SCD (frequent VOEs, acute chest syndrome, leg ulcers, Other phenotypes AVN, and priapism), interestingly the presence of this Few studies exist on genetic risk factors that may be mutation was associated with an increased risk of associated with complications such as acute chest central venous catheter associated thrombosis (P/ syndrome, pulmonary hypertension, leg ulcers, and 0.006, odds ratio 4.9, Table III) [28]. priapism. Sharan et al. and reported an association of acute chest syndrome in females with a T-786C UGT1A1 polymorphism and bilirubin levels polymorphism in the endothelial NO synthase gene [36]. In one study in 45 patients with pulmonary The gene UGT1A1 encodes the enzyme UDP Glu- hypertension as judged by a tricuspid regurgitant jet curonosyl Transferase-1 that mediates the glucuroni- of /2.5 m/sec, an association was found with SNPs dation of bilirubin. A common polymorphism in the in BMPR2 (bone morphogenic protein receptor 2) TATA box of this gene was found to be associated and ADCY6 (adenylate cyclase 6) [37]. In a study of with a decreased enzyme activity and indirect hyper- 148 patients with SCD and history of priapism and bilirubinemia (Gilbert’s syndrome). The sequence 529 controls without priapism, it was shown that a A(TA)6TAA is presumed the wild type while polymorphism in the klotho (KL) gene was associated A(TA)7TAA is associated with Gilbert’s syndrome. with an odds ratio of 2.6. [38]. In a study by Ashley- The co-inheritance of this (TA)7 polymorphism with Kotch et al. 2005, longevity (survival /50 years of hereditary hemolytic anemias, such as hereditary age) was associated with three SNPs in the klotho spherocytosis, b-thalassemia, and G-6-PD deficiency gene and SNPs in the nitric oxide synthase 2 results in marked hyperbilirubinemia and increased (NOS2A) and (TGFbR2) genes [39]. No studies incidence of gallstones [29]. We, and others, have have yet been reported showing an association of studied the frequency of this polymorphism in the polymorphisms with a risk of sickle nephropathy. sickle cell population and its impact on bilirubin levels and incidence of gallstones and cholecystectomy [30/ 34]. In all of these studies, an association was found Gene expression profiling between the (TA)7 genotype and bilirubin levels; Gene expression profiling utilizing cDNA microarrays those with 7/7 had the highest mean bilirubin levels is a relatively novel method, particularly in terms of its compared to the genotypes with 6/6 and 6/7. Our applications to the study of SCD. The quality of the results are summarized in Table III. Furthermore, an data obtained from microarray studies depends upon impact of the 7/7 genotype on the results of hydro- several factors: (1) a meticulous study design, parti- 96 A. Kutlar

Table V. Demographic and hematologic data in mild vs. severe SCD patients (mean values).

Mild (n/8) Severe (n/4) P-value

Gender 6 M, 2 F 2 M, 2 F Age 35.5 25 0.03 Crisis/year 1.1 5.8 0.0002 Hb g dL1 8.9 8.6 0.6 HCT% 27.4 24.9 0.4 MCV fl 91 96 0.4 Retics% 10.7 10.8 0.8 9 1 WBC/10 l 11.6 11.2 0.8 ANC (per ml) 6005 5850 0.9 9 1 Platelets/10 l 376 341 0.6 Total Bilirubin mg dL1 3.8 6.2 0.2 Hb F% 6.3 10.7 0.2

amenable to gene expression profiling studies include peripheral blood (PB) cells, endothelial cells (EC) (circulating EC, as well as EC from various tissues), bone marrow (BM), and liver tissue. Few published reports exist in this area. These include the study of gene expression profiling from the kidney tissue in transgenic sickle cell mice [40], profiling of cultured human pulmonary artery endothelial cells exposed to plasma from patients with SCD and acute chest syndrome and patients with SCD in steady state [41,42], a study of differential gene expression in PB cells from SCD patients with different crisis frequency [43], and a study of gene expression profiles form mononuclear cells in SCD patients on and off hydroxyurea (HU) therapy [44]. Jison et al. purified mononuclear cells from PB in 27 patients with SCD in steady state, 10 of which were on HU and 13 normal controls. They demonstrated differential gene expression of 112 genes involved in heme metabolism, cell cycle regulation, antioxidant and stress responses, inflammation, and angiogenesis. HU did not significantly alter mononuclear cell gene expression. Klings et al. exposed cultured human pulmonary artery endothelial cells to plasma from normal volunteers as well as to plasma from SCD patients at steady state and SCD patients with acute chest syndrome. They found that 50 genes were differentially expressed in endothelial cells upon exposure to plasma from SCD patients in steady state compared to normal plasma. These genes involved Figure 1. Granulocyte gene expression proﬁling: Severe and mild cholesterol biosynthesis, lipid transfer, cellular stress patients vs. controls. response, and extracellular matrix proteins. Upon exposure to plasma from SCD patients with acute cularly a clear and unambiguous definition of the chest, another 58 genes were differentially expressed. ‘‘phenotype(s)’’ to be studied, (2) elimination of We have conducted preliminary studies of gene technical/methodological pitfalls (quantity/quality of expression with commercially available cDNA micro- the RNA, methods of specimen collection and pro- arrays (Affymetrix U95, Affymetrix, Santa Clara, CA) cessing, conditions and timing of hybridization, elim- from PB neutrophils and reticulocytes in SCD pa- ination of intra-patient variation by repeated tients with discordant rates of painful episodes in an sampling, etc.), and (3) expertise in bioinformatics effort to elucidate the factors contributing to different in the interpretation of vast amounts of data gener- crisis frequency. We analyzed the gene expression ated. In patients with SCD, the tissues that will be profiles of neutrophils from four patients with ‘‘se- Sickle cell disease 97

vere’’ disease (/3 vaso-occlusive episodes [VOE] per and 2-fold), MAP2K3 (increased 3.5-fold and 2- year), eight patients with ‘‘mild’’ disease (B/3 VOE/ fold). These data show that neutrophils in SCD year) and compared these to each other and to the patients are activated with higher expression of genes gene expression profiles of neutrophils from five age in the TNF, MAPK, and NFkB pathways consistent and sex matched, healthy, non-sickle cell, African/ with an inflammatory state. Since neutrophil apopto- American individuals. A summary of demographic sis is considered critical for the resolution of inflam- features, mean hematologic values, and Hb F in mation, delayed or inhibited apoptosis of neutrophils ‘‘mild’’ and ‘‘severe’’ SCD patients is shown in Table would further maintain this inflammatory state, even V. Figure 1 depicts differences in gene expression during the so-called ‘‘steady state: of the disease. patterns between non-sickle cell controls and mild Further analyses and identification of differentially and severe SCD patients. In general, a larger number expressed genes and pathways between ‘‘mild’’ pa- of genes were differentially expressed between ‘‘mild’’ tients vs. controls and ‘‘severe’’ vs. ‘‘mild’’ patients is patients vs. controls, compared to that between in progress. A list of some of the differentially ‘‘severe’’ vs. ‘‘mild’’ patients. Out of the differentially expressed genes in neutrophils from ‘‘severe’’ and expressed genes (314 genes for severe vs. control, 718 ‘‘mild’’ patients is shown in Table VI. We conclude genes for mild vs. control), those with greater than that the analyses of gene expression in neutrophils can two fold expression were analyzed with the gene be a useful tool in identifying pathways and genes that MAPP software for localization into biological path- distinguish SCD patients from controls and in differ- ways. Genes related to cellular proliferation, growth entiating mild and severe phenotypes. and maintenance, DNA repair, DNA replication, and The application of the microarray technology to the cell cycle progression were expressed at significantly study of SCD is still in its infancy. Although the higher levels in SCD patients compared to controls. preliminary data look interesting, these results need to The most significant finding was the significantly be confirmed and validated through a large number of higher expression of genes leading to NFkB activation experiments in a large number of patients. The and inhibition of apoptosis: IAP-1 (increased 6.7 fold available data appear promising and show the feasi- and 4.7 fold in mild and severe patients respectively), bility of the application of this technology to SCD. IkB (decreased 0.14 fold and 0.3 fold), Apaf-1 (decreased 0.4-fold in mild), and c-jun (decreased Conclusion 0.4-fold in severe); Traf-2 (TNF receptor associated factor-2; increased 3.5-fold and 2-fold); genes in the The development of high throughput genotyping MAPK signaling pathway: ERK-2 (increased 3.5-fold methods, the microarray technology, and the comple-

Table VI. Differentially expressed genes in severe vs. mild SCD patients.

Fold Difference Gene ID Gene Function P value

11.6 NAD Kinase Signal Tx, Energy Metabolism, Detox Rx 0.005 9.5 GPR43 7 Transmem domain rec for signal tx for a neuroendocrine pp, Galanin 0.004 6.4 BTG family Cell cycle regulation, antiproliferative 0.003 5.4 HLA-G2.2 0.002 4.7 FLJ13052 0.005 4.6 Similar to tumor necrosis factor Inhibits TRAIL (1.2) mediated apoptosis 0.007 4.4 HLA-G1 Role in inhibiting natural killer cell function, tumors can escape from 0.004 immsurv. 3.8 HLA-G2.1 0.006 3.2 IFI30 Catalyzes disulfide bond reduction for proteolysis of the internalized 0.005 proteins. Const. Expressed in Ag pre cells and induced by IFNg in other Imp in Ag presentation 3.1 HLA-DRB1 0.005 3.0 KIAA0991 Rearranged with MLL in therapy induced AML 0.009 2.6 Helix-loop-helix zipper protein NFkB activation by inhibiting the I-kB proteins 0.004

2.4 Aryl Hydrocarbon receptor- Aryl hydrocarbon receptor (AHR) / HSP90 complex translocates to 0.002 interacting protein nucleus, HSP90 dissociates and activated AHR dimerizes with ARNT w then bind enhancer elements to regulate transcription of xenobiotic metabolic enzymes 2.4 RPL28 Structural mammalian ribosomal protein 0.002 2.3 A20 Inhibits NFkB activity and TNF mediated apoptosis. Critical for 0.004 limiting inflammation by terminating TNF induced NFkB responses. TNF dramatically increases A20 expression in all ts. 2.1 DUT Hydrolyzes d UTP to d UMP. Limits d UMP incorporation into DNA 0.008 during replication and repair and protects from fragmentation and death. 98 A. Kutlar tion of the Human Genome Project have already for Fetal Hemoglobin Expression in Adults. BMC Gneomics begun to revolutionize medicine and biology. The 2004;5:/ 33./ [7] Garner CP, Tatu T, Best S, Creary L, Thein SL. Evidence of application of these techniques to the study of SCD is genetic interaction between the beta-globin complex and in its very early stages and is likely to yield novel chromosome 8q in the expression of fetal hemoglobin. Am J information. Human Genetics 2002;70:/ 793/ /799. A candidate gene approach and study of genome- [8] Garner C, Silver N, Best S, Menzel S, Martin C, Spector TD, wide SNPs and their association with certain compli- Thein SL. A quantitative trait locus on chromosome 8q inﬂuences the switch from fetal to adult hemoglobin. Blood cations of SCD is expected to further clarify the basis 2004;104:/ 2184/ /2186. of clinical/phenotypic heterogeneity of this single gene [9] Wyszynski DF, Baldwin CT, Cleves MA, Amirault Y, Nolan disorder. The pros and cons of a candidate gene VG, Farrell JJ, Bisbee A, Kutlar A, Farrer LA, Steinberg MH. approach vs. genome wide association studies need to Polymorphisms near a chromosome 6q QTL area are asso- be balanced carefully. A candidate gene approach is ciated with modulation of fetal hemoglobin levels in sickle cell limited by our current knowledge of the pathophysiol- anemia. Cell Mol Biol 2004;50:/ 23/ /33. [10] Wyszynski DF, Baldwin CT, Cleves MA, Farrell JJ, Bisbee A, ogy and by necessity may omit some of the genetic Kutlar A, Farrer LA, Steinberg MH. Genetic polymorphisms influences that may contribute to the genotype in associated with fetal hemoglobin response to Hydroxyurea in

question. It is also of utmost importance to carefully patients with sickle cell anemia. Blood 2004;104:/ 34a./ and unambiguously define the phenotype. Genome- [11] Ohene-Frempong K, Weiner SJ, Sleeper LA, Miller ST, Embury S, Moohr JW, et al. Cerebrovascular accidents in wide studies are likely to provide some information on sickle cell disease: Rates and risk factors. Blood 1998;91:/ 288/ / novel genetic influences; however, there limitation is 294. the unrealistically large sample size that may be [12] Adams RJ, McKie VC, Hsu L, Files B, Vichinsky E, Pegelow required. Application of methods such as the Bayesian C, et al. Prevention of a ﬁrst stroke by transfusion in children networks may be important in providing information with sickle cell anemia and abnormal results on transcranial doppler ultrasonography. N Engl J Med 1998;339:/ 5/ /11. on gene/gene interactions. [13] Driscoll MC, Hurlet A, Styles L, McKie V, Files B, Olivieri N, Although the association studies and the results of et al. Stroke risk in siblings with sickle cell anemia. Blood microarray data reviewed above are far from conclu- 2003;101:/ 2401/ /2404. sive, some interesting points and recurring themes [14] Hoppe C, Klitz W, Noble J, Vigil L, Vichinsky E, Styles L. have emerged. Many complications of SCD can be Distinct HLA associations by stroke subtype in children with related to variation in genes in the TGF-beta super sickle cell anemia. Blood 2003;101:/ 2865/ /2869. family and genes that play a role in vascular reactivity, [15] Hoppe C, Klitz W, Cheng S, Apple R, Steiner L, Robles L, Girard T, Vichinsky E, Styles L. Gene interactions and stroke inflammation, and apoptosis. These are emerging also risk in children with sickle cell anemia. Blood 2004;103:/ / from the microarray data. A study of gene expression 2391/2396. profiling coupled with proteomics will shed further [16] Taylor VI JG, Tang DG, Savage SA, Leitman SF, Heller SI, light into the functional aspects and differences Serjeant GR, et al. Variants in the VCAM1 Gene and risk for between patients. This information will not only symptomatic stroke in sickle cell disease. Blood 2002;100:/ / 4303/4309. lead to a better understanding of the pathogenesis [17] Steinberg MH, Baldwin CT, Wyszynski DF, Nolan VG, and pathophysiology of many complications of the Amirault Y, Farrell JJ, Bisbee A, Gavras H, Farrer LA, Adams disease, but will also likely result in the identification R, et al. Stroke in sickle cell anemia: Association with single of novel therapeutic targets and discovery of new nucleotide polymorphisms in genes affecting vascular func- genes with prognostic and therapeutic implications. tion. Blood 2003;102:/ 406A./ [18] Sebastinani P, Ramoni MF, Nolan V, Baldwin CT, Steinberg MH. Genetic dissection and prognostic modeling of overt stroke in sickle cell anemia. Nature Genetics 2005;37:/ 435/ / 440. References [19] Milner PF, Kraus AP, Sebes JI, Sleeper LA, Dukes KA, [1] Steinberg MH, Rodgers GP. Pathophysiology of sickle cell Embury SH, et al. Sickle cell disease as a cause of osteone- disease: Role of cellular and genetic modiﬁers. Sem in Hemat crosis of the femoral head. NEJM 1991;325:/ 1476/ /1481. [20] Kutlar A, Kutlar F, Turker I, Tural C. The methylenetetrahy- 2001;38:/ 299/ /306. [2] Hebbel RP, Vercellotti GM. The endothelial biology of sickle drofolate reductase (C677T) mutation as a potential risk cell disease. J Lab Clin Med 1997;129:/ 288/ /293. factor for avascular necrosis in sickle cell disease. Hemoglobin [3] Rother RP, Bell L, Hillmen P, Gladwin MT. The clinical 2001;25:/ 213/ /217. sequelae of intravascular hemolysis and extracellular plasma [21] Adekile AD, Kutlar F, Haider MZ, Kutlar A. Frequency of the hemoglobin: A novel mechanism of human disease. JAMA 677C0/T mutation of the methylenetetrahydrofolate reduc- 2005;6:293:/ 1653/ /1662. tase gene among Kuwaiti sickle cell disease patients. Am J [4] Charache S, Terrin ML, Moore RD, Dover GJ, Barton FB, Hematol 2001;66:/ 263/ /266. Eckert SV, McMahon RP, Bonds DR, MSH Investigators. [22] Zimmerman SA, Ware RE. Inherited DNA mutations con- Effect of Hydroxyurea on the frequency of painful crises in tributing to thrombotic complications in patients with sickle / / sickle cell anemia. NEJM 1995;332:/ 1317/ /1322. cell disease. Am J Hematol 1998;59:267 /272. [5] Steinberg MH. Predicting clinical severity in sickle cell [23] Andrade FL, Annichino-Bizzacchi JM, Saad STO, Costa FF, anemia. Brit J Haematol 2005;129:/ 465/ /481. Arruda VR. Prothrombin mutant, factor V leiden, and [6] Close J, Game L, Clark B, Bergounioux J, Gerovassili A, thermolabile variant of methylenetetrahydrofolate reductase

Thein SL. Genome annotation of a 1/5 Mb region of human among patients with sickle cell disease in Brazil. Am J Hematol chromosome 6q23 encompassing a Quantitative Trait Locus 1998;59:/ 46/ /50. Sickle cell disease 99

[24] De Castro L, Rinder HM, Howe JG, Smith BR. Thrombo- glucuronosyl transferase 1A promoter polymorphisms, bs- philic genotypes do not adversely affect the course of sickle cell globin gene haplotype, co-inherited a-thalassemia trait and

disease (SCD). Blood 1998;92:/ 161./ Hb F on steady-state serum bilirubin levels in sickle cell [25] Baldwin C, Nolan VG, Wyszynski DF, Ma QL, Sebastiani P, anemia. Accepted for publication in Eur J Haematol, 2005. Embury SH, Bisbee A, Farrell J, Farrer L, Steinberg MH. [35] Henney MW, Howard TA, Zimmerman SA, Ware RE. Association of Klotho, bone morphogenic protein 5, and UGT1A Promoter polymorphism inﬂuence bilirubin response annexin A2 polymorphisms with sickle cell osteonecrosis. to hydroxyurea therapy in sickle cell anemia. J Lab Clin Med Blood 2005;106:/ 372/ /375. 2003;141:/ 279/ /282. [26] Gandrille S, Greengard JS, Alhenc-Gelas M, Juhan-Vague I, [36] Sharan K, Surrey S, Ballas S, Borowski M, Devoto M, Wang Abgrall JF, Jude B, et al. Incidence of activated protein C KF, Sandler E, Keller M. Association of T-786C eNOS gene resistance caused by the ARG 5065 GLN mutation in factor V polymorphism with increased susceptibility to acute chest in 113 unrelated symptomatic protein C- deﬁcient patients. syndrome in females with sickle cell disease. Br J Haematol Blood 1995;86:/ 219/ /224. 2004;124:/ 240/ /243. [27] Le W, Yu JD, Tao R, et al. Association of the R485K [37] Ashley-Koch A, DeCastro L, Lennon-Graham F, Jonassaint J, polymorphism of the factor V gene with poor response to Jackson TL, Price J, Galloway J, Ataga K, Orringer EP, Vance activated protein C and increased risk of coronary artery JM, et al. Genetic polymorphisms associated with the risk for disease in the Chinese population. Clin Genet 2000;57:/ 296/ / pulmonary hypertension and proteinuria in sickle cell disease. 303.

Blood 2004;104:/ 464./ [28] Ustun C, Adams GT, Kutlar F, Elam D, Clair B, Daitch L, et [38] Nolan VG, Baldwin CT, Ma QL, Wyszynski DF, Amirault Y, al. Factor V R485K polymorphism is a risk factor for catheter Farrell JJ, Bisbee A, Embury SH, Farrer LA, et al. Association induced thrombosis in sickle cell disease. 30th Anniversary of of single nucleotide polymorphisms in KLOTHO with priap- the National Sickle Cell Disease Program and the SCDAA, ism in sickle cell anemia. Br J Haematol 2004;128:/ 266/ /272. Washington, DC, September 17/21, 2002. [39] Ashley-Koch AE, De Castro L, Lennon-Graham F, Jonassaint [29] Sampietro M, Lipica L, Perrero L, et al. The expression of J, Jakcson TL, Price J, Galloway J, Jones S, Randall E, Eckman uridine diphosphate glucuronosyltransferase gene is a major J, et al. Clinical and genetic proﬁles of the aging sickle cell determinant of bilirubin level in heterozygous b-thalassemia patient. Abstract Presented at the 28th Annual Meeting of the and in Glucose-6-phosphate dehydrogenase deﬁciency. Br J National Sickle Cell Disease Program, April 9 /13, 2005. Haematol 1997;99:/ 437/ /439. [40] Nagel RL, Fabry ME, Kaul DK, Rybicki AC. Up-regulation [30] McKie K, Kutlar F, Sromek E, Litaker M, Woods KF, Kutlar of gene expression in the chronically damaged kidney in sickle A. Uridine diphosphate glucuronosyl transferase-1 (UGT1A1) promoter polymorphism and bilirubin levels in transgenic mice. Blood (Supplement) 2002;100:/ 2607,/ 663a. [41] Adwoye A, Safaya S, Framtpon G, Lenburg M, Klings EX,

patients with sickle cell disease. Blood 1999;94/ Supplement 1 Odhiambo A, et al. Pulmonary artery endothelial cells exposed

(Part 1 of 2):197a(861)./ [31] Passon RG, Howard TA, Zimmerman SA, Schultz WH, Ware to acute chest syndrome plasma express a novel repertoire of

RE. Influence of bilirubin uridine diphosphate-glucuronosyl- genes. Blood (Supplement) 2002;100:/ 1756,/ 453a. transferase 1A promoter polymorphisms on serum bilirubin [42] Klings ES, Safaya S, Adewoye AH, Odhiambo A, Framtpon levels and cholelithiasis in children with sickle cell anemia. J G, Lenburg M, Gerry N, Sebastiani P, Steinberg MH, Farber HW. Differential gene expression in pulmonary artery en- Pediatr Hematol/Oncol 2001;23:/ 448/ /451. [32] Haverfield EV, McKenzie CA, Forrester T, Bouzekri N, dothelial cells exposed to sickle cell plasma. Physical Geno- Harding R, Serjeant G, Walker T, Peto TE, Ward R, Weath- mics 2005;21:/ 293/ /298. erall DJ. UGT1A1 variation and gallstone formation in sickle [43] Kutlar A, Kutlar F, Clair B, Nechtman J, Joshi RM, Daitch L. Gene expression profiling of peripheral blood cells from sickle cell disease. Blood 2004;105:/ 968/ /972. [33] Fertrin KY, Melo MB, Assist AM, Saad ST, Costa FF. UDP- cell patients with differing crisis rates. Blood (Supplement)

Glucuronosyltransferase 1 gene promoter polymorphism is 2002;100:/ 3549,/ 24b. associated with increased serum bilirubin levels and cholecys- [44] Jison ML, Munson PJ, Barb JJ, Suffredini AF, Talwar S, tectomy in patients with sickle cell anemia. Clinical Genetics Logun C, Raghavachari N, Beigel JH, Shelhamer JH, Danner 2003;64:/ 160/ /162. RL, et al. Blood mononuclear cell gene expression profiles [34] Adekile A, Kutlar F, McKie K, Addington A, Elam D, Holley characterize the oxidant, hemolytic, and inflammatory stress L, Clair B, Kutlar A. The influence of uridine diphosphate of sickle cell disease. Blood 2004;104:/ 270/ /280. Hematology, 2005; 10 Supplement 1: 100/103

PEDIATRIC HEMATOLOGY

Juvenile myelomonocytic leukemia

CHRISTIAN P. KRATZ & CHARLOTTE M. NIEMEYER

Department of Pediatrics and Adolescent Medicine, University of Freiburg, Freiburg, Germany

Juvenile myelomonocytic leukemia (JMML) is a rare active GTP-bound (Ras-GTP) conformations. So- clonal myeloproliferative disorder (MPD) of early matic NRAS or KRAS2 mutations are identified in childhood [1]. The median age at diagnosis is 2 years 15/25% of JMML cases [8,9]. These lesions lead to [1]. There is a male predominance with a male:female substitutions at codons G12, G13, or Q61 and impair ratio of 2:1. Pallor, fever, infection, skin bleeding and the intrinsic Ras GTPase and confer resistance to cough are the most common presenting symptoms. GTPase-activating proteins. In 11% of patients with Typically, there is marked hepatosplenomegaly. JMML the clinical diagnosis of neurofibromatosis JMML rarely involves the central nervous system. type 1 (NF1) can be made [10]. However, it has been shown that some infants with JMML carry germ line neurofibromatosis type 1 gene (NF1) mutations Laboratory findings although the diagnosis of NF1 had not been made Laboratory findings include leukocytosis, anemia and clinically [11]. Neurofibromin, the gene product of thrombocytopenia. The median white count is 33/ the NF1 tumor suppressor gene, negatively regulates 109/L [1]. An absolute monocyte count greater than Ras by functioning as GTPase activating protein. 9 1/10 /L is required for the diagnosis [2]. The blast Thus, loss of neurofibromin leads to aberrant Ras cell percentage in peripheral blood (PB) rarely ex- signaling. Indeed, analysis of JMML cells from ceeds 20% and is most often below 2% [1]. Cytoge- patients with NF1 showed an elevated percentage of netic studies of JMML cells show a normal karyotype Ras in the active GTP-bound conformation [12]. in 65% of patients, monosomy 7 in 25%, and other Taken together, approximately 85% of all patients abnormalities in 10% [1]. Patients with monosomy 7 with JMML carry mutations in PTPN11, RAS, or present with a normal or moderately elevated HbF. In NF1 (Figure I). Of note, these mutations are largely patients with normal karyotype HbF is often elevated mutually exclusive suggesting that they disrupt iden- [1]. tical pathways in myeloid cells [4,5].

Aberrant Ras signaling Murine models The Ras signaling transduction pathway is deregu- Mouse models support the idea that in JMML lated in JMML. JMML myeloid progenitors are mutations in NF1, RAS, or PTPN11 are initiating hypersensitive to granulocyte-macrophage colony sti- rather than later cooperating events. By generating mulating factor (GM-CSF) in vitro [3]. This observa- mice whose hematopoietic system was reconstituted tion is linked to mutations that lead to aberrant Ras with NF1 deficient hematopoietic stem cells it has signaling. Approximately 35% of the cases harbor been shown that NF1 loss, by itself, is sufficient to somatic mutations in exons 3 or 13 of the PTPN11 produce a JMML-like disorder in mice [13]. In gene [4/6]. The PTPN11 gene product, SHP-2, is a addition, somatic inactivation of NF1 results in a non-receptor protein tyrosine phosphatase that relays fatal myeloproliferative disease with shift in hemato- signals from growth factor receptors to Ras and other poiesis from the BM to spleen, an increased number signaling molecules [7]. RAS genes encode 21-kDa of T-and B-cells and resistance to apoptosis [14]. signal switch molecules that regulate cell fates by Similarly, somatic activation of oncogenic Ras in cycling between inactive GDP-bound (Ras-GDP) and hematopoietic cells initiates a rapidly fatal myelopro-

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331390078 Juvenile myelomonocytic leukemia 101

35% 25% 25%

GM-CSF receptor GAB2 RAS NF1 SHC GDP SOS RAS

GRB2 GTP

SHP-2 RAF

MAPK

Figure I. Model outlining the roles of SHP-2, RAS, and NF1 in the granulocyte-macrophage colony-stimulating factor (GM-CSF) signal transduction pathway. In juvenile myelomonocytic leukemia (JMML), molecular alterations have been demonstrated in PTPN11 (the gene encoding SHP-2), RAS and NF1 in approximately 35%, 25% and 25% of patients, respectively. liferative disorder [15]. Finally, in a murine model of However, molecular studies have greatly facilitated Noonan syndrome (see below), mice carrying an the diagnostic approach and mutational studies have activated PTPN11 allele develop a mild myeloproli- become an important tool in the diagnostic process of ferative disease [16]. JMML. JMML can be mimicked by a variety of infectious agents such as cytomegalovirus, Epstein- Barr virus, human herpes virus 6 and parvovirus B19. Noonan syndrome and juvenile myelomoncytic Positive results of these viruses do not exclude the leukemia diagnosis of JMML. Noonan syndrome is an autosomal dominant developmental disorder characterized by dysmorphic facial Natural course if the disease features, growth retardation and variable congenital heart defects. In one half of the cases the disorder is JMML is a rapidly fatal disorder for most children if caused by heterozygous germ line PTPN11 mutations left untreated. The median survival time without [7]. A small number of infants with Noonan syn- hematopoietic stem cell transplantation (HSCT) is drome develop a JMML-like disease [17]. In these young children with Noonan syndrome and JMML- Table I. Diagnostic guidelines for juvenile myelomoncytic leukemia like disease, the spontaneous remission of the disorder (JMML) [adopted from [2]] is well documented [17]. PTPN11 mutations ob- I. Suggestive clinical features served in Noonan Syndrome are predicted to have Hepatosplenomegaly generally mild gain-of-function effects. In contrast, Lymphadenopathy PTPN11 mutations associated with JMML lead to Pallor Fever stronger activation of SHP-2. Mutations in patients Skin rash with Noonan syndrome who develop a JMML-like II. Laboratory criteria (all 3 are mandatory) disease are predicted to have intermediate effects. Absence of Philadelphia chromosome (BCR/ABL Consistent with this model, in vitro and in vivo rearrangement) 9 experiments on primary hematopoietic cells and cell Peripheral blood monocyte count /1/10 /L B lines show that somatic mutants confer more pro- Blast percentage in BM /20% nounced effects on cell growth than common mutants III. Criteria for definite diagnosis (at least 2 must be fulfilled) only found in Noonan syndrome [18,19]. Hemoglobin F increased for age Myeloid precursors on peripheral blood smear 9 White blood count /10/10 /L Diagnosis Clonal abnormality Granulocyte-macrophage colony-stimulating factor (GM-CSF) Diagnostic criteria have been proposed by the Inter- hypersensitivity of myeloid progenitors in vitro national JMML Working Group in 1998 (Table I) [2]. 102 C. P. Kratz & C. M. Niemeyer about 1 year [1]. Low platelet count, age above 2 phylaxis may significantly contribute to successful years at diagnosis and high HbF at diagnosis are main leukemia control. However, donor lymphocyte infu- predictors of short survival [1]. Blastic transformation sion in JMML relapse has largely been unsuccessful is infrequent in JMML. Most untreated patients die [22]. Serial quantitative chimerism studies can iden- from respiratory failure caused by infiltration of the tify patients with increasing mixed chimersim who are lungs with JMML cells. at high risk for relapse of JMML. Immediate withdrawal of immunosuppressive therapy is advised in these patients [23]. Despite aggressive re-emergence Antileukemic therapy of the malignant clone and short interval between the The role of antileukemic therapy prior to HSCT is first and second HSCT, a substantial proportion of currently uncertain. The current JMML study of children can be cured after a second HSCT [24]. the Children’s Oncology Group (COG) prescribes cytoreductive therapy consisting of fludarabine and high-dose cytarabine concomitantly with 13-cis reti- References noic acid prior to HSCT, while most patients in [1] Niemeyer CM, Arico M, Basso G, Biondi A, Cantu Rajnoldi Europe traditionally receive mercaptopurine or no A, Creutzig U, et al. Chronic myelomonocytic leukemia in therapy. Therapeutic strategies targeting individual childhood: a retrospective analysis of 110 cases. European components of the Ras signaling pathway include the Working Group on Myelodysplastic Syndromes in Childhood administration of the GM-CSF analog E21R and the (EWOG-MDS). Blood 1997;89:/ 3534/ /43. farnesyltransferase inhibitor R115777 (Zarnestra†) [2] Niemeyer CM, Fenu S, Hasle H, Mann G, Stary J, van Wering E. Differentiating juvenile myelomonocytic leukemia from [20]. infectious disease. Blood 1998;91:/ 365/ /367. [3] Emanuel PD, Bates LJ, Castleberry RP, Gualtieri RJ, Zucker- man KS. Selective hypersensitivity to granulocyte-macrophage Stem cell transplantation colony-stimulating factor by juvenile chronic myeloid leukemia Allogeneic HSCT is currently the only curative hematopoietic progenitors. Blood 1991;77:/ 925/ /9. [4] Tartaglia M, Niemeyer CM, Fragale A, Song X, Buechner J, treatment for JMML. The analysis of the EWOG- Jung A, et al. Somatic mutations in PTPN11 in juvenile MDS/EBMT trial of 100 patients with JMML trans- myelomonocytic leukemia, myelodysplastic syndromes and planted with a preparative regiment of busulfan, acute myeloid leukemia. Nat Genet 2003;34:/ 148/ /50. cyclophosphamide and melphalan shows a 5-year [5] Loh ML, Vattikuti S, Schubbert S, Reynolds MG, Carlson E, probability of EFS of 52% [21]. The EFS of patients Lieuw KH, et al. Mutations in PTPN11 implicate the SHP-2 phosphatase in leukemogenesis. Blood 2004;103:/ 2325/ /31. transplanted from a matched family donor (MFD) [6] Kratz CP, Niemeyer CM, Castleberry RP, Cetin M, Berg- and unrelated donor (URD) are not significantly strasser E, Emanuel PD, et al. The mutational spectrum of different. In the current EWOG-MDS/EBMT trial, PTPN11 in juvenile myelomonocytic leukemia and Noonan the 5-year cumulative incidence of relapse was 35% Syndrome/myeloproliferative disease. Blood 2005 May 31, [21]. Relapse occurs early, at a median of 2 to 6 [Epub ahead of print]. [7] Tartaglia M, Mehler EL, Goldberg R, Zampino G, Brunner months from transplantation [21] and generally HG, Kremer H, et al. Mutations in PTPN11, encoding the within the first year. HSCT shortly after diagnosis is protein tyrosine phosphatase SHP-2, cause Noonan syn- recommended, because younger age at HSCT pre- drome. Nat Genet 2001;29:/ 465/ /8. dicts improved survival [21]. In the prospective [8] Kalra R, Paderanga DC, Olson K, Shannon KM. Genetic HSCT study from EWOG-MDS, multivariate analy- analysis is consistent with the hypothesis that NF1 limits myeloid cell growth through p21ras. Blood 1994;84:/ 3435/ /9. sis shows age greater than 4 years and female sex [9] Flotho C, Valcamonica S, Mach-Pascual S, Schmahl G, predicted poorer outcome [21]. Cytogenetic abnorm- Corral L, Ritterbach J, et al. RAS mutations and clonality alities do not confer a worse prognosis [21]. In the analysis in children with juvenile myelomonocytic leukemia current HSCT study of the EWOG-MDS, splenect- (JMML). Leukemia 1999;13:/ 32/ /7. omy did not improve the survival after HSCT [21]. [10] Castro-Malaspina H, Schaison G, Passe S, Pasquier A, Berger R, Bayle-Weisgerber C, et al. Subacute and chronic myelomonocytic leukemia in children (juvenile CML). Clinical and Relapse hematologic observations, and identification of prognostic factors. Cancer 1984;54:/ 675/ /86. Disease recurrence remains the major cause of treat- [11] Side LE, Emanuel PD, Taylor B, Franklin J, Thompson P, ment failure. Median time from HSCT to relapse is Castleberry RP, et al. Mutations of the NF1 gene in children with juvenile myelomonocytic leukemia without clinical evi- 4 /6 months with only few patients relapsing more dence of neurofibromatosis, type 1. Blood 1998;92:/ 267/ /72. than 1 year after transplantation [21]. Graft-versus- [12] Bollag G, Clapp DW, Shih S, Adler F, Zhang YY, Thompson leukemia (GVL) effect plays an important role in P, et al. Loss of NF1 results in activation of the Ras signaling curing children with JMML with HSCT. Re-emerging pathway and leads to aberrant growth in haematopoietic cells. donor cells and frank hematologic relapse have been Nat Genet 1996;12:/ 144/ /8. [13] Largaespada DA, Brannan CI, Jenkins NA, Copeland NG. successfully eradicated by reduction of ongoing im- Nf1 deficiency causes Ras-mediated granulocyte/macrophage munosuppressive therapy. Reducing the intensity and colony stimulating factor hypersensitivity and chronic myeloid duration of graft-versus-host disease (GVHD) pro- leukaemia. Nat Genet 1996;12:/ 137/ /43. Juvenile myelomonocytic leukemia 103

[14] Le DT, Kong N, Zhu Y, Lauchle JO, Aiyigari A, Braun BS, et mutations in primary hematopoietic cells. Blood 2005;106:/ /

al. Somatic inactivation of Nf1 in hematopoietic cells results in 311/7.

a progressive myeloproliferative disorder. Blood 2004;103:/ / [20] Niemeyer CM, Kratz C. Juvenile myelomonocytic leukemia. 4243 /50. Curr Treat Options Oncol 2003;4:/ 203/ /10. [15] Braun BS, Tuveson DA, Kong N, Le DT, Kogan SC, Rozmus [21] Locatelli F, Nollke P, Zecca M, Korthof E, Lanino E, Peters J, et al. Somatic activation of oncogenic Kras in hematopoietic C, et al. Hematopoietic stem cell transplantation (HSCT) in cells initiates a rapidly fatal myeloproliferative disorder. Proc children with juvenile myelomonocytic leukemia (JMML): Natl Acad Sci U S A 2004;101:/ 597/ /602. results of the EWOG-MDS/EBMT trial. Blood 2005;105:/ /

[16] Araki T, Mohi MG, Ismat FA, Bronson RT, Williams IR, 410/9. Kutok JL, et al. Mouse model of Noonan syndrome reveals [22] Yoshimi A, Bader P, Matthes-Martin S, Stary J, Sedlacek P, cell type- and gene dosage-dependent effects of Ptpn11 Duffner U, et al. Donor leukocyte infusion after hematopoietic mutation. Nat Med 2004;10:/ 849/ /57. stem cell transplantation in patients with juvenile myelomo- [17] Bader-Meunier B, Tchernia G, Mielot F, Fontaine JL, nocytic leukemia. Leukemia 2005;19:/ 971/ /7. Thomas C, Lyonnet S, et al. Occurrence of myeloproliferative [23] Yoshimi A, Niemeyer CM, Bohmer V, Duffner U, Strahm B,

disorder in patients with Noonan syndrome. J Pediatr 1997;/ Kreyenberg H, et al. Chimaerism analyses and subsequent 130:885/ /9. immunological intervention after stem cell transplantation in [18] Mohi MG, Williams IR, Dearolf CR, Chan G, Kutok JL, patients with juvenile myelomonocytic leukaemia. Br J Hae- Cohen S, et al. Prognostic, therapeutic, and mechanistic matol 2005;129:/ 542/ /9. implications of a mouse model of leukemia evoked by Shp2 [24] Chang YH, Jou ST, Lin DT, Lu MY, Lin KH. Second (PTPN11) mutations. Cancer Cell 2005;7:/ 179/ /91. allogeneic hematopoietic stem cell transplantation for juvenile [19] Schubbert S, Lieuw K, Rowe SL, Lee CM, Li X, Loh ML, et myelomonocytic leukemia: case report and literature review. J al. Functional analysis of leukemia-associated PTPN11 Pediatr Hematol Oncol 2004;26:/ 190/ /3. Hematology, 2005; 10 Supplement 1: 104/107

PEDIATRIC HEMATOLOGY

Hemophagocytic lymphohistiocytosis

G. E. JANKA

Universita¨ts-Kinderklinik, Ha¨matol./Onkologie Martinistr., Hamburg

Hemophagocytic lymphohistiocytosis (HLH) is a life- ism, followed by cytomegalovirus and other herpes threatening disease characterized by unremitting viruses. A fairly frequent cause for HLH is infection fever, hepatosplenomegaly, cytopenias, as well as by leishmania, an organism for which very effective changes in coagulation and lipids. The symptoms treatment exists. are due to high levels of cytokines secreted by immune IAHS occurs in children and adults and is probably effector cells which display functional defects. more frequent than the familial form. In contrast to the first publication by Risdall most patients with IAHS reported subsequently had no known under- History and classification (Table I) lying immune deficiency. HLH was first described by Farquhar and Claireaux It has to be emphasized that the identification of an in 1952 as a familial disease [1]. Interestingly in their infectious organism does not help to discriminate cases hemophagocytosis, which has given the disease between FHLH and IAHS since the former is also its name, could not be found during lifetime but was triggered by an infection in most cases. Age is the only prominent on autopsy. Familial hemophagocytic lym- parameter which may be helpful to distinguish both phohistiocytosis (FHLH) is an autosomal recessive forms: 70% of FHLH cases occur within the first year disorder with an estimated frequency of 0.12/100 000 of life whereas IAHS patients are usually older. children per year [2]. Several genetic defects have However, in the author’s experience about 10% of been described for FHLH (see below). In addition babies with HLH have a transient and therefore not well-characterized immune deficiency syndromes familial form; on the other hand FHLH has occa- such as Che´diak-Higashi syndrome (CHS), Griscelli sionally been described in older children. syndrome (GS), and x-linked proliferative syndrome The picture of HLH in patients with rheumatic (XLP) may present initially with HLH or develop diseases, especially systemic onset juvenile rheuma- HLH later [3]. Whereas in FHLH the symptoms of toid arthritis, is commonly named macrophage-acti- HLH are the primary and only manifestation, the vation syndrome (MAHS) [7]. It has recently been occurrence of HLH in these immune defects is suggested that this condition be included as a separate optional. entity in the category of acquired HLH [8]. HLH can In 1979, Risdall et al. described a picture indis- also be a complication in patients with malignant tinguishable from FHLH in adults who received diseases especially lymphomas and some inborn immunosuppressive treatment after organ transplan- errors of metabolism [9]. tation and experienced a viral infection [4]. A few children were included and not in all patients a virus Genetics and pathophysiology could be identified. The disease was named virus- associated hemophagocytic syndrome (VAHS). Sub- HLH arises on the basis of various genetic or acquired sequently it became evident that any infectious agent immune deficiencies. Only recently several genetic including bacteria, protozoa and fungi could trigger defects have been described in FHLH: Mutations in HLH [5,6]; thus the term infection-associated hemo- the perforin gene [10], the UNC13D gene [11], and phagocytic syndrome (IAHS) replaced VAHS. the syntaxin 11 gene [12]. Whereas in our experience Viruses, however, remain the most frequent triggering mutations in one of these three genes can be found in agents with Epstein-Barr virus as the leading organ- 80% of the Turkish patients, only 30% of the German

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331390087 Hemophagocytic lymphohistiocytosis 105

Table I. Classification of hemophagocytic lymphohistiocytosis Table II. Symptoms and laboratory findings at first presentation and at diagnosis* Genetic forms

/Familial HLH (M.Farquhar) At first presentation At diagnosis

/Immune deficiency syndromes % of patients % of patients Chediak-Higashi syndrome Griselli syndrome Fever 78 100 X-linked lymphoproliferative syndrome Splenomegaly 78 98 Bicytopenia 56 98 Acquired forms 1 FibrinogenB/1.5 g l 23 65 /Infection-associated hemophagocytic syndrome 1 Triglycerides/3 mmol l 52 76 /Macrophage activation syndrome Hemophagocytosis 36 88 /Malignancy-associated hemophagocytic syndrome NK cell activity negative or 100 100 decreased 1 sCD25/ /2400 U ml 100 90 1 patients harbored a perforin or UNC13D mutation. Ferritin/ /500 ng nl 57 70 1 Mutations in syntaxin 11 so far have only been found LDH/ /500 Ul 55 51 1 in patients from Turkey. The genes implicated in the ALT / /100 Ul 30 38 1 Bilirubin/ /34 mmoll 30 32 pathogenesis of Che´diak-Higashi syndrome (the 1 CSF cells/ /5 ml 32 39 LYST gene), Griscelli syndrome (The RAB27A 1 CSF protein/0.5 gl 32 46 gene) and x-linked proliferative syndrome (the SH2D1A gene) have also been identified. In a *(Janka et al. 2005) sCD25 /a chain of the soluble interleukin 2 substantial number of patients thought to have genetic receptor; LDH /lactate dehydrogenase; ALT /alanine amino- transferase; CSF/cerebrospinal fluid. In some instances (sCD25, disease based on a positive family history or a LDH) percentage of patients with high values was lower at diagnosis relapsing course, the gene defect is as yet unknown. due to unspecific therapies. The clinical picture of HLH is due to hyperinflammation caused by hypersecretion of inflammatory Typical laboratory values are thrombocytopenia cytokines by activated T-lymphocytes and histiocytes and anemia, and less frequently neutropenia. Pancy- infiltrating all organs. In spite of the excessive activa- topenia is progressive in untreated patients. Charac- tion and expansion of cytotoxic cells, patients with teristic biochemical findings are high triglycerides, a HLH have severe impairment of cytotoxic function of low fibrinogen or a global coagulation disorder, and natural killer cells and cytotoxic T cells [13]. All high ferritin. Less frequent are high transaminases, known defects in HLH are involved in the function of bilirubin and LDH, low total protein and hyponatre- cytolytic granules: granule trafficking (LYST), dock- mia [9]. A bone marrow examination is mandatory ing at the membrane (RAB27A), granule priming but only a minority of patients have hemophagocy- (UNC13D), deficient granule content (PFR), im- tosis at presentation. A bone marrow biopsy is usually paired lymphocyte activation (SH2D1A) and prob- even less sensitive. Characteristically the bone marrow ably impaired interaction between dendritic and killer is cellular in spite of profound peripheral pancytope- cells (syntaxin 11). Defective cytolytic activity seems nia. Erythropoiesis is usually increased and may to be the common denominator which predisposes to exhibit dysplastic changes. The cerebrospinal fluid HLH. shows a moderately increased cell count and/or The mechanisms leading to impaired cytotoxic protein content in more than 50% of the cases even function in apparently immune competent patients in the absence of neurological symptoms. with acquired HLH are less clear. Interference by A valuable disease marker is a high level of the a cytokines or virus-encoded proteins are possible chain of the soluble interleukin 2 receptor (sCD25). mechanisms. The prevalence of EBV associated Impaired natural killer (NK) cell activity is a hallmark HLH in Asia suggests a specific genetic susceptibility. of the disease. It is found in FHLH, CHS, GS and XLP as well as in acquired HLH. In the latter, Clinical symptoms and laboratory findings impaired NK cell activity may be due in part to (Table II) reduced numbers of NK cells and is usually reversible. The disease typically starts after a free interval weeks to months after birth. In rare cases already newborns Diagnostic criteria and problems become symptomatic. The initial symptoms at first The revised diagnostic criteria of the Histiocyte are compatible with a normal infection. The first sign Society are shown in Table III. is high fever frequently associated with signs of an At initial presentation the diagnosis may be difficult upper respiratory or gastrointestinal infection. Nearly because of several reasons: all patients have an enlarged liver and/or spleen.

Enlarged lymph nodes, transient rashes and neurolo- / an infectious organism is identified, a normal gical symptoms such as an opisthotonic posture, infection suspected and the severe symptoms of seizures or cerebral nerve palsies are less frequent [14]. HLH are overlooked 106 G. E. Janka

/ not all necessary diagnostic criteria may be ease. Since many patients do not have a family history present or a proven genetic defect a surrogate marker for

/ fever may subside spontaneously and recur after genetic disease is persistent disease activity or relapses days or weeks on or off treatment. In patients without family

/ unspecific measures as transfusions may lead to history and complete resolution of all symptoms transient improvements elective cessation of therapy is recommended to

/ organomegaly, disturbed liver function and high prevent an unnecessary SCT for transient, acquired triglycerides may suggest a HLH. This is not without risk since a relapse may be

/ metabolic disorder accompanied by severe symptoms. Thus these pa-

/ liver failure and severe coagulopathy may be the tients have to be closely monitored to restart therapy leading symptom in time.

/ neurological symptoms suggesting encephalitis Standard treatment for HLH is a combination of may precede systemic HLH for corticosteroids, cyclosporin A and etoposide. All

/ several months or be the only manifestation patients with known familial disease, suspected genetic disease due to age below 1 year and patients with All symptoms of HLH, however, to a lesser extent life-threatening symptoms such as coagulopathy, pro- may also be found in a normal immunological found cytopenia or neurological disease should re- response. It is the severity of the clinical picture and ceive therapy according to the present HLH 2004 laboratory findings and the progression of symptoms protocol (available at website: www.histio.org/society/ which should alert the physician that this is an protocols). Etoposide may be life-saving especially in impaired immune response leading to HLH. patients with EBV-associated HLH [17] There are a Whereas hemophagocytosis is very characteristic few patients with acquired HLH and mild symptoms and, if present, a valuable diagnostic criterium it in whom corticosteroids and immunoglobulins may should be emphasized that the lack of hemophagocy- be sufficient. However, disease may progress rapidly tosis does not exclude the diagnosis of HLH. and this risk outweighs the possible side effects of The diagnostic criteria for HLH which are based on etoposide. For children with MAS treatment with familial and infection-associated acquired cases may high doses of corticosteroids and/or cyclosporin A is not be relevant for patients with MAS. Accordingly recommended. these modified diagnostic criteria have recently been Reactivation during treatment is a frequent pro- advocated [15]. blem, either systemically or in the CNS. Immuno- suppressive treatment should be reinforced and SCT be performed as early as possible. Intrathecal therapy Therapy and prognosis is recommended in CNS relapses in view of the Untreated familial disease is fatal in all cases. Also deleterious long-term effects of uncontrolled CNS acquired IAHS has a high fatality rate of 50% in disease. children [16]. If a treatable organism is found appro- Most children respond to treatment within 1/4 priate therapy should be given but with the possible weeks. There is no established salvage therapy for exception of leishmaniasis antiinfectious therapy is non-responders. Antithymocyte globulin which was usually not sufficient to control HLH. The immediate shown to be effective as first line treatment [18] is aim of treatment is to suppress hypercytokinemia rarely effective when the HLH protocol fails. Single that is responsible for the life-threatening symptoms. patients have responded to daclizumab or alemtuzu- In familial cases this has to be followed by stem mab. cell transplantation (SCT) as the only curative dis- The results of the HLH 94 study have recently been published [19]. The overall survival rate in 113 Table III. Diagnostic criteria for HLH* children was 55% at 3 years. Patients with a transplant Familial disease/known genetic defect from a matched sibling donor (MSD) or matched Clinical and laboratory criteria (5/8 criteria) unrelated donor had the same results.

/Fever

/Splenomegaly

/Cytopenia/ /2 cell lines Conclusion

/Hypertriglyceridemia and/or hypofibrinogenemia 1 HLH is a well described clinical entity characterized /Ferritin/ /500 ng ml 1 /sCD25/ /2400 U ml by fever, hepatosplenomegaly, cytopenias and various /Decreased or absent NK-cell activity laboratory changes. Hyperinflammation is due to /Hemophagocytosis in bone marrow, CSF or lymphnodes genetic or acquired cytotoxic defects of immune effector cells. HLH is still frequently overlooked. Supportive evidence are cerebral symptoms with moderate pleocy- tosis and/or elevated protein, elevated transaminases and bilirubin, Treatment with chemoimmunotherapy in due time 1 LDH/1000 Ul controls the disease in most patients and facilitates * Janka and Schneider 2004; 1For method see Schneider et al. 2002. SCT for genetic cases with a high cure rate. Hemophagocytic lymphohistiocytosis 107

References [12] Zur Stadt U, Schmidt S, Kasper B, Beutel K, Diler AS, Henter JI, Kabisch HI, Schneppenheim R, Nurnberg P, Janka [1] Farquhar JW, Claireaux AE. Familial haemophagocytic reti- G, Hennies HC. Linkage of familial hemophagocytic lympho- culosis. Arch Disease in Childhood 1952;27:/ 519/ /525. histiocytosis (FHL) type-4 to chromosome 6q24 and identi- [2] Henter JI, Elinder G, Soder O, Ost A. Incidence in Sweden ﬁcation of mutations in syntaxin 11 gene. Hum Mol Genet and clinical features of familial hemophagocytic lymphohistio- 2005;14:/ 827/ /834. cytosis. Acta Paediatrica Scandinavica 1991;80:/ 428/ /435. [13] Schneider EM, Lorenz I, Muller-Rosenberger M, Steinbach [3] Dufourcq-Lagelouse R, Pastural E, Barrat FJ, Feldmann J, Le G, Kron M, Janka-Schaub GE. Hemophagocytic lymphohis- Deist F, Fischer A, De Saint Basile G. Genetic basis of hemophagocytic lymphohistiocytosis syndrome (Review). Int J tiocytosisis associated with deﬁciencies of cellular cytolysis but normal expression of transcripts relevant to killer-cell-induced Mol Med 1999;4:/ 127/ /133. [4] Risdall RJ, McKenna RW, Nesbit ME, Krivit W, Balfour HH, apoptosis. Blood 2002;100:/ 2891/ /2898. Jr, Simmons RL, Brunning RD. Virus-associated hemopha- [14] Janka GE. Familial hemophagocytic lymphohistiocytosis. Eur gocytic syndrome: a benign histiocytic proliferation distinct J Ped 1983;140:/ 221/ /230. [15] Ravelli A, Magni-Manzoni S, Pistorio A, Besana C, Foti T, from malignant histiocytosis. Cancer 1979;44:/ 993/ /1002. [5] Reiner AP, Spivak JL. Hematophagic histiocytosis. A report of Ruperto N, Viola S, Martini A. Preliminary diagnostic 23 new patients and a review of the literature. Medicine guidelines for macrophage activation syndrome complicating / / / (Baltimore) 1988;67:/ 369/ /388. systemic juvenile idiopathic arthritis. J Ped 2005;146:598 [6] Fisman DN. Hemophagocytic syndromes and infection. 604. Emerging Infectious Diseases 2000;6:/ 601/ /608. [16] Janka G, Imashuku S, Elinder G, Schneider M, Henter JI. [7] Stephan JL, Zeller J, Hubert P. Macrophage activation Infection- and malignancy-associated hemophagocytic syn- syndrome and rheumatic disease in childhood: a report of dromes. Secondary hemophagocytic lymphohistiocytosis. He- / / / four new cases. Clin Exp Rheumatol 1993b;11:451 456. matol Oncol Clin North Am 1998;12:/ 435/ /444. [8] Ramanan AV, Baildam EM. Macrophage activation syndrome [17] Imashuku S, Kuriyama K, Teramura T, Ishii E, Kinugawa N, is hemophagocytic lymphohistiocytosis/need for the right Kato M, Sako M, Hibi S. Requirement for etoposide in the terminology. J Rheumatol 2002;29: 1105; author reply 1105. treatment of Epstein-Barr virus-associated hemophagocytic [9] Janka G, Henter JI, Imashuku S. Clinical aspects and therapy lymphohistiocytosis. J Clin Oncol 2001;19:/ 2665/ /2673. of hemophagocytic lymphohistiocytosis. In: Weitzman S, [18] Stephan JL, Donadieu J, Ledeist F, Blanche S, Griscelli C, Egeler RM, editors. Histiocytic Disorders of Children and Fischer A. Treatment of familial hemophagocytic lymphohis- Adults. Cambridge University Press; 2005. pp 353 /379. tiocytosis with antithymocyte globulins, steroids, and cyclos- [10] Stepp SE, Dufourcq-Lagelouse R, Le Deist F, Bhawan S, porin A. Blood 1993;82:/ 2319/ /2323. Certain S, Mathews PA, Henter JI, Bennett M, Fischer A, de [19] Henter JI, Samuelsson-Horne A, Arico M, Egeler RM, Saint Basile G, Kumar V. Perforin gene defects in familial Elinder G, Filipovich AH, Gadner H, Imashuku S, Komp

hemophagocytic lymphohistiocytosis. Science 1999;286:/ / D, Ladisch S, et al. Treatment of hemophagocytic lympho- 1957/1959. histiocytosis with HLH-94 immunochemotherapy and Bone [11] Feldman J, Callebaut I, Raposo G, Certain S, Bacq D, Dumont C, Lambert N, Quachee-Chardin M, Chedeville G, Marrow Transplantation. Blood 2002;100:/ 2367/ /2373. [20] Janka GE, Schneider EM. Modern management of children Tamary H, et al. Munc13/4 is essential for cytolytic granules fusion and is mutated in a form of familial hemophagocytic with haemophagocytic lymphohistiocytosis. Brit J Haematol 2004;124:/ 4/ /14. lymphohistiocytosis (FHL3). Cell 2003;115:/ 461/ /473. Hematology, 2005; 10 Supplement 1: 108/110

PEDIATRIC HEMATOLOGY

Fanconi anemia: Current management

HOON KOOK

Department of Pediatrics, Blood & Marrow Transplantation Center, Chonnam National University Hwasun Hospital, Hwasun, Korea

Abstract Fanconi anemia (FA) is an autosomal recessive chromosomal instability disorder, characterized by congenital anomalies, defective hematopoiesis and a high risk of developing acute myeloid leukemia and certain solid tumors. All racial and ethnic groups are at risk, and at least 11 complementation groups have been identified and the genes defective in eight of these have been identified (FANCA, C, D2, E, F, G, L and BRCA2). FA-A is the most common complementation group, accounting for approximately 65% of all affected individuals. The gold-standard screening test for FA is based on the characteristic hypersensitivity of FA cells to the crosslinking agents, such as mitomicin C or diepoxybutane. Recent progress has been made in identifying the genes bearing pathogenetically relevant mutations, but slower progress has been made in defining the precise functions of the proteins in normal cells, in part because that the proteins are multifunctional. Molecular studies have established that a common pathway exist, both between the FA proteins and other proteins involved in DNA repair such as NBS1, ATM, BRCA1 and BRCA2. Stem cell transplantation (SCT) is the only option for establishing normal hematopoiesis. To reduce undue toxicities due to inherent hypersensitivity, nonmyeloablative conditioning for transplants has been advocated. This review summarizes the general clinical and hematologic features and the current management of FA. Fanconi anemia (FA) is the commonest type of inherited bone marrow failure syndrome with the birth incidence of around three per million. The inheritance pattern is autosomal recessive with the estimated heterozygote frequency being one in 300 in Europe and the US.

Clinical features (AML). The relative risk of AML is 800-fold, and the median age in reported cases is 14 years. The most frequent characteristic birth defects in FA Also, patients with FA are at a particularly high risk include skin hyperpigmentation and/or cafe au lait of developing specific solid tumors at unusually young spots (55%), short stature (51%), abnormal thumbs ages, including head, neck, esophagus, and gynecolo- and radii (43%), abnormal head (26%), eyes (23%), gical squamous cell carcinomas, as well as liver kidneys (21%) and ears (21%). Low birth weight tumors. The relative risk of developing total cancers 5 ( /2,500 g) and developmental disability are found in is 48 in FA patients in comparison with general 11% of patients, respectively. However, 25% or more population, with the highest relative risk of 4,300 for of known FA patients have few or none of these vulva and/or anus tumor. features. Therefore, FA should be considered in any young The most important clinical features of FA are adult with any of the following subtle but character- hematological and these are responsible for the great- istic physical anomalies, hematologic cytopenias, un- est morbidity and mortality in homozygotes. At birth, explained macrocytosis, MDS/AML, or squamous the blood count is usually normal and macrocytosis is cell cancer even in the absence of severe pancytopenia often the first detected abnormalities. This is followed or a positive family history. by thrombocytopenia and anemia, and pancytopenia typically presents between the ages of 5 and 10 years, FA diagnosis with the median age of onset being 7 years. Moreover, patients with FA may even present with myelodys- In the proper clinical context the gold-standard plastic syndrome (MDS), or acute myeloid leukemia screening test for FA is based on the characteristic

ISSN 1024-5332 print/ISSN 1607-8454 online # 2005 Taylor & Francis DOI: 10.1080/10245330512331390096 Fanconi anemia: Current management 109 hypersensitivity of FA cells to the crosslinking agents Fanconi proteins A, C, D2, E, F, and G function in a (mitomicin C [MMC], diepoxybutane [DEB], cispla- common cellular pathway was substantiated by data tin). Culture of replicative cells (usually phytohe- showing that proteins A, C, E, F, and G form a magglutin [PHA]-stimulated peripheral blood constitutive nuclear protein complex. Activation of lymphocytes or skin fibroblasts) in the presence of this protein complex by DNA damage or cell cycle low doses of either MMC, or DEB followed by progression results in the conversion of the down- examination of metaphase spreads for evidence of stream FANCD2 protein from an unubiquitinated chromosomal breaks and radial chromosomes can isoform for from to a monoubiquitinated isoform. establish the diagnosis of FA. Data are reported as Monoubiquitination does not occur if the protein aberration per cell, as well as percentage of cells with complex A to G is not intact, and therefore Fanconi aberrations, usually for 20 to 100 cells. The percen- cells from A, C, E, F, and G patients do not show tage of cells with aberration may be useful, because monoubiquitinated FANCD2. In normal cells after patients with hematopoietic somatic mosaicism may monoubiquitination, FANCD2 localizes to nuclear have only a few cells with breaks. foci where it co-localizes with other DNA-repair Flow cytometry can detect the proportion of cells proteins such as BRCA1. Although BRCA2 may that are arrested at G2/M cell cycle after culture with also be genetically linked to the Fanconi pathway, its a clastogen such as nitrogen mustard. It has advantage exact role is unclear. of examining thousands of cells and is less labor- A genotype-phenotype study examined the intensive and subjective, but it is usually done in a consequences of mutations of FANCC in patients. specialized laboratory. Kaplan-Meier analysis showed that IVS4 or exon 14 Complementation analysis requires patient lym- mutations define poor-risk subgroups clinically, they phocytes, EBV-lymphoblasts, or fibroblasts to culture are associated with earlier onset of hematologic with cells or retroviruses which introduce known abnormalities and poorer survival compared with normal FANC genes into the patient’s cells. This patients with other exon I mutation and with the test is limited to the availability of cells or cloned non-FANCC population. This was confirmed in a DNA from known FA genotypes, and is performed in 20-year follow-up perspective by the International a very limited number of research laboratories. Fanconi Anemia Registry (IFAR). Another report Mutated genes can also be identified by denaturing describing the association of complementation group high performance liquid chromatography (DHCLP) and mutation type with clinical outcome showed that heterodulplex analysis. Fanconi patients with mutations in the FANCG gene and those homozygous for null mutations in FANCA are also high risk groups with a poor hematologic Genetics and molecular pathogenesis outcome. At least eleven complementation groups are known to date. Genes for 8 groups have been characterized Management (FANCA, C, D2, E, F, G, L and BRCA2). FANCA is the most common complementation group, in which The projected median survival of patients with FA is more than 200 mutations have been documented. approximately 30 years but survival is extraordinary Identification of the 8 subtypes facilitated the cloning variable. The most life-threatening early event is bone of the FA genes. The first gene, FANCC on chromo- marrow failure. Matched sibling donor (MSD) SCT some 9q22.3, was discovered in 1992. The breast is now accepted as the best therapy available to cure cancer susceptibility gene, BRCA2, has been identi- the FA patient of marrow aplasia and to prevent fied as a FA gene (formerly known as FANCD1); progression to myelodysplasia or leukemia. An other- biallelic mutations in BRCA2 have been observed in wise healthy patient with FA and significant pancyto- 3 FA subtype B and D1 cells, suggesting that BRCA2 is penia (ANCB/1,000/mm , hemoglobin 5/8 g/dL or a 3 the FANC gene corresponding to both of these platelet count 5/40/50,000/mm ) with an available complementation groups. HLA-MSD is an excellent candidate for hematopoie- Knock-out mouse models of FA provide insight tic SCT. Initial efforts to transplant FA patients using into the role of individual mutations. Knock-out of standard preparative regimens and graft-versus-host FANCA and FANCG and two different knock-outs of disease (GVHD) prophylaxis were plagued by two FANCC have been generated in mice. The phenotype serious and often lethal problems: severe toxicity from of these mutant mice is identical and consists of chemotherapy and exaggerated GVHD. Bone marrow cellular sensitivity to DNA cross-linking agents, transplantation (BMT) protocols were subsequently abnormal G2-M progression of the cell cycle similar modified for FA, and the outcomes improved sub- to Fanconi patients, and hypoplasia of gonads. stantially. Cells and cell lines from Fanconi patients are The data on HLA-identical MSD SCT performed phenotypically similar, regardless of the complemen- on 151 FA patients from 42 institutions were sum- tation group that they represent. The hypothesis that marized in 1995 with the 2-year survival rate of 66%. 110 H. Kook

Using cyclophosphamide (total dose, 20 mg/kg) and blood) is better than another. At this time, most thoracoabdominal irradiation (5 Gy) for conditioning centers have reserved the use of umbilical cord blood and cyclosporine A for GVHD prophylaxis, 50 FA for those patients who cannot identify a suitable patients transplanted form MSD had a 5-year disease- marrow donor. free survival of about 75%. Another protocol for 16 For patients who are not candidates for transplan- FA patients with MSD used cyclophosphamide alone tation, androgen therapy sometimes induces mean- (total dose 100 mg/kg). The actuarial survival rate at ingful responses in pancytopenic patients. This 37 months was 88%. Fludarabine, a purine antime- treatment is known to affect liver function adversely. tabolite with potent immunosuppressive properties, Thus, several transplant centers recommend andro- was successfully incorporated into a conditioning gens not be given to any FA patients unless no suitable protocol without radiation. The University of Minne- donor is available. Use of cytokines such as G-CSF or sota is currently using T-cell depleted bone marrow erythropoietin has been beneficial in some patients, (with Isolex CD34 positive selection) or cord blood provided the marrow shows no evidence of a clone or and a preparative therapy of ATG 150 mg/m2, cyclo- dysplasia. phosphamide 20 mg/kg, and fludarabine 175 mg/m2. Knowledge of the complementation group or mu- GVHD prophylaxis is cyclosporine and short course tation may permit the potential use of gene therapy methylprednisolone. The survival on 11 patients was and pre-implantation genetic diagnosis and embryo 100% with no GVHD and no significant organ selection both to rule out FA and rule in an HLA toxicity. match. SCT from an alternate donor (i.e., HLA- For patients with stable disease, annual surveillance mismatched related or unrelated) is complex and exams and bone marrow aspiration (with cytogenetic challenging, and is associated with poor survival. studies) and biopsy are suggested. For patients with The world-wide experience using unrelated donors complex cytogenetic abnormalities or MDS, closer in the treatment of FA patients prior to the year 2000 follow-up is warranted. revealed that survival at 3 years after unrelated donor SCT was approximately 30%. Regimen-related toxicity, graft rejection, and infection were the primary reasons for treatment failure. The identification of risk Suggested Readings factors, including: age (/10 years), prior use of ] [1] Fanconi Anemia: Standards for Clinical Care. Owen J, Frohn- androgen, number ( /3) of malformations, and mayer L, Eiler ME editors. Fanconi Anmeia Research Fund, absence of fludarabine in the preparative therapy, Inc. 2nd ed. 2003. provides a rationale for tailoring therapy to the [2] Freedman MH. Inherited forms of bone marrow failure. In: individual patient. The preparative therapy most often Hoffman R, Benz Jr. EJ, Shattil SJ, Furie B, Cohen HJ, used currently consisted of fludarabine, cyclopho- Silberstein LE, McGlave P, editors. Hematology. Basic Princi- ples and Practice, 4th ed. Philadelphia: Elsvier Inc.; 2005. pp sphamide and total body irradiation (TBI). Most 339/379. recent data reported the 5-year survival of 58% with [3] Tischkowitz M, Dokal I. Fanconi anaemia and leukaemia */ matched unrelated SCT. Alternate donor SCT is clinical and molecular aspects. Br J Haematol 2004;126:/ 176/ / indicated in patient age B/35 years, severe cytopenia 191. 3 [4] Bagby GC, Lipton JM, Sloand EM, Schiffer CA. Marrow (HgbB/8 g/dL, ANCB/500/mm ,PLTB/20,000/ Failure. Hematology, Am Soc Hematol Educ Program 2004; mm3), MDS/leukemia, high risk mutations (i.e., 318/336 FANCC IVS4, FANCC exon 14, or BRCA2) in the [5] Young NS, Alter BP. Clinical features of Fanconi anemia. In: absence of HLA-MSD. Young NS, Alter BP, editors. Aplastic Anemia, Acquired and Since the first transplantation in 1998 for a Fanconi Inherited. Philadelphia: WB Sanders; 1994. pp 275 /309. patient, umbilical cord blood is being increasingly [6] Joenje H, Oostra AB, Wijker M, di Summa FM, van Berkel CG, Rooimans MA, et al. Evidence for at least eight Fanconi used as a donor source. Engraftment and survival anemia genes. Am J Hum Genet 1997;61:/ 940/ /944. after cord blood transplantation has been comparable [7] Bagby GC Jr. Genetic basis of Fanconi anemia. Curr Opin to BMT, although engraftment was slower with cord Hematol 2003;10:/ 68/ /76. blood. The incidence of acute and chronic GVHD [8] Rosenberg PS, Greene MH, Alter BP. Cancer incidence in was reduced with cord blood grafts, even in cases of persons with Fanconi’s anemia. Blood 2003;101:/ 882/ /886. [9] Vlachos A, Lipton JM. Hematopoietic stem cell transplant for HLA-mismatched transplants. At this time, however, inherited bone marrow failure syndroems. In: Mehta P, et al. there are no data to indicate whether one stem cell editors. Pediatric Stem Cell Transplantation. Sudbury, MA: source (marrow vs peripheral blood vs umbilical cord Jones and Barlett; 2004. pp 281/311.