Letters to the Editor 1285 We report on our experience with this novel compound maintain ample serum levels in all patients for over 14 days. focusing on its pharmacokinetics and pharmacodynamics after Pegfilgrastim is a reasonable and cost-effective alternative induction polychemotherapy. These analyses were performed in to long-term conventional recombinant human granulocyte the context of a phase II study and were approved by the local colony-stimulating factor in chemotherapy-induced leukopenia ethics committee. Prior to initiation of study medication, all in patients with . However, from these patients gave their written informed consent. Seven days after data, no conclusion can be drawn with respect to its activity in the end of induction chemotherapy comprising high-dose shortening the time to neutrophil recovery. cytarabine and mitoxantron, 6 mg of pegfilgrastim (Neulasta) was administered subcutaneously in patients showing complete M Fiegl, W Hiddemann and J Braess, for the AML Cooperative-Group blast clearance in routine bone marrow aspiration. A second Department of Internal Medicine III, University Hospital injection was scheduled for those patients remaining leukopenic –1 Munich, Campus Grosshadern, Marchioninistr, (that is, leukocytes o1Gl ) 10 days after the first injection. Munich, Germany Blood samples were obtained before the first application and E-mail: [email protected] then daily until leukocyte recovery. Pegfilgrastim plasma levels were measured by commercial granulocyte colony-stimulating factor ELISA (R&D Systems, GmbH, Wiesbaden-Nordenstadt, References Germany). Nineteen patients were included into the analysis. The 1 Dombret H, Chastang C, Fenaux P, Reiffers J, Bordessoule D, median age was 62 years (range: 27–78 years; 7 male/12 Bouabdallah R et al. A controlled study of recombinant human female patients). The median peak plasma level of pegfilgrastim granulocyte colony-stimulating factor in elderly patients after –1 treatment for acute myelogenous leukemia. AML Cooperative Study was 178 ng ml and was reached within 2 days after injection Group. N Engl J Med 1995; 332: 1678–1683. (Figure 1a). Median half-life (T1/2;) was 6 days, which strongly 2 Heil G, Hoelzer D, Sanz MA, Lechner K, Noens L, Szer J et al. correlated to the absolute leukocyte count (Figure 1b). Effective Long-term survival data from a phase 3 study of Filgrastim as an plasma levels of 42ngml–1 as defined by Yang et al.6 were adjunct to chemotherapy in adults with de novo acute myeloid observed in all patients until day 14. Median time to leukocyte leukemia. Leukemia 2006; 20: 404–409. 3 Ohno R, Tomonaga M, Kobayashi T, Kanamaru A, Shirakawa S, recovery was 8 days after application of pegfilgrastim. Masaoka T et al. Effect of granulocyte colony-stimulating factor after Seven patients received a second injection of pegfilgrastim intensive induction therapy in relapsed or refractory acute leukemia. on day þ 10 (median) due to insufficient leukocyte recovery. N Engl J Med 1990; 323: 871–877. A second application led to a median peak plasma level of 4 Kern W, Aul C, Maschmeyer G, Kuse R, Kerkhoff A, Grote-Metke A –1 210 ng ml . T1/2; was dependent on leukocyte count; four of et al. Granulocyte colony-stimulating factor shortens duration these seven patients had persistent neutropeniaFresulting in a of critical neutropenia and prolongs disease-free survival after sequential high-dose cytosine arabinoside and mitoxantrone T1/2; of more than 12 days. Plasma levels were always in the –1 (S-HAM) salvage therapy for refractory and relapsed acute myeloid effective range of 42ngml (Figure 1d). In the three patients leukaemia. German AML Cooperative Group. Ann Hematol 1998; with early leukocyte recovery after the second dose, T1/2; was 77: 115–122. only 2 daysFclearly demonstrating the context-sensitive nature 5 Johnston E, Crawford J, Blackwell S, Bjurstrom T, Lockbaum P, of pegfilgrastim half-life (Figure 1c). Roskos L et al. Randomized, dose-escalation study of SD/01 Pegfilgrastim was safe in our study, as all patients finished compared with daily filgrastim in patients receiving chemotherapy. study medication as scheduled, and no side effects clearly J Clin Oncol 2000; 18: 2522–2528. 6 Yang BB, Kido A, Shibata A. Serum pegfilgrastim concentrations related to the study drug were reported. during recovery of absolute neutrophil count in patients with cancer Concluding from these data, 6 mg of pegfilgrastim is sufficient receiving pegfilgrastim after chemotherapy. Pharmacotherapy 2007; to reach high peak plasma concentrations that are sufficient to 27: 1387–1393.

In , generation of a nuclear export signal through duplication appears unique to nucleophosmin (NPM1) and is restricted to AML

Leukemia (2008) 22, 1285–1289; doi:10.1038/sj.leu.2405045; C-terminus of the NPM mutant.5 Although the transforming published online 6 December 2007 potential of the NPM leukaemic mutants is yet to be proven, the observation that mutants retaining tryptophan 288 always combine with the strongest C-terminus NES motif reveals Nucleophosmin (NPM1) mutations, involving mainly NPM1 mutational selective pressure toward enhanced NPM export exon-121 and occasionally other exons,2,3 occur in about 30% into cytoplasm,6 pointing to this event as critical for of adult acute myeloid leukaemia (AML)1 that show distinct leukaemogenesis.4 biological and clinical features.4 NPM1 mutations characteri- About 80% of NPM1-mutated AML bear ‘ A’,1,4 that stically generate mutants that localize aberrantly in the is, a tandem duplication of a TCTG tetranucleotide at positions cytoplasm of leukaemic cells,1 hence, the term NPMc þ 956–959 of NPM1. Intriguingly, tandem duplications seem to (cytoplasmic-positive) AML. Cytoplasmic NPM accumulation have played a role in evolution of biological diversity,7 as is caused by changes of tryptophan(s) 288 and 290 (or only 290) expansions and contractions of repeats in critical are and creation of a new nuclear export signal (NES) motif at the associated with great morphological diversity (for example,

Leukemia Letters to the Editor 1286 canine limbs and skulls).8 In humans, generation of DNA included sequences containing up to 15 amino acids, so as to repeats is an ongoing process in tissues,9 and duplications of prevent a combinatorial explosion induced by genetic code very short repetitive sequences contribute significantly to redundancy during retrotranslation. Some signals that could not the generation of insertion mutations. Indeed, repeat instability, be clearly identified as contiguous amino acid sequences were a major mutation form, is linked to several neurological excluded, since they were short sequences of a few annotated disorders.9,10 residues interwoven with the non-annotated ones. A total of 28 It is not known whether tandem duplication of short NES sequences were retrotranslated, taking all possible codon nucleotide sequences targeting genes other than NPM1 can combinations into account. Besides these 28 dabatase NES generate NES motifs and ultimately leads to aberrant cytoplas- motifs, we also analysed the C-terminus of 23 known NPM1 mic dislocation and accumulation of unidentified mutated leukaemic mutants. . Therefore, we developed a genome-wide computa- After retrotranslation, each signal was processed in a search tional analysis to assess whether duplication of 3–6 nucleotide for tandem duplications with a period of 3–6 nucleotides as well combinations could lead to generation of a known NES motif in as exact multiple duplications, for example, ‘TCTG TCTG any human . We also used immunohistochemistry to study TCTG’. When a duplication was identified, its last repeat was NPM nucleo-cytoplasmic transport abnormalities in 3017 deleted and its DNA sequences were matched against human samples of different human tumours. Our results strongly suggest genome coding sequences (NCBI repository: ftp://ftp.ncbi.nih. that, in the human genome, duplication-linked generation of a gov/genomes/H_sapiens/RNA/rna.fa.gz, last accessed on 30 July NES motif appears to be unique to the NPM1 mutation, and 2007), in an attempt to find wild-type genes with the potential to confirm that cytoplasmic expression of NPM leukaemic mutants generate NES after the duplication events. The set of genes was is restricted to AML. analysed to exclude artefacts, that is, rediscovery of our original NES. To explore hypothetical duplication events causing known The results of computational analysis are reported in Table 1. NES signals in the human genome, we designed a computa- Notably, in the search starting from the NPM mutational tional approach (Figure 1) based on a customized software variants, we recovered only the NPM1 gene (leukaemic mutant (Supplementary Material and Supplementary Figure 1). We first ‘A’), which was first described by Falini et al.1 This finding accessed NES sequences in the NES database,11 which is served to validate our approach. The other occurrences reported available online at http://www.cbs.dtu.dk/databases/NESbase in Table 1 were filtered out by the last computational step. In (last accessed on 3 July 2007). The current version of the particular, NES-00018, NES-00019, NES-00023, NES-00041.1 database contains 75 sequences carrying experimentally and NES-00041.2, all result in matches with genes that validated NES signals. Our computational analysis was re- physiologically contain the same NES signal. stricted to motifs that contain amino acids both necessary and Our findings so far strongly suggest that, out of the entire sufficient for nuclear export. Moreover, for our study, we only human genome, only the NPM1 gene fulfils the requirements for an ‘ideal’ site for duplication encoding for an NES motif. However, since not all NES motifs in nature (and diseases) are known and since our analysis was restricted to those containing NES motifs (aminoacids) stretches of up to 15 amino acids, we cannot exclude that other (51 sequences) genes may theoretically undergo this process. The unique mechanism underlying generation of a new NES 1. (Retrotranslation) motif, that leads (in concert with C-terminus tryptophans alterations) to aberrant cytoplasmic NPM expression is not the NES-coding nucleotide sequences 9 only distinguishing feature of NPM1 mutations. In fact, (1.10×10 sequences) immunohistochemical studies1,12 support the view that, among human tumours, altered nucleo-cytoplasmic NPM transport due 2. (Tandem repeat identification) to NPM1 mutations is specific for AML. To further confirm these findings, we extended our immunohistochemical investiga- Nucleotide sequences containing tandem repeats tions1,12 in human tumours to a total of 3017 bioptic specimens, (6.86×108 sequences, if L = 3) which included 1123 new de novo AML cases and also new tumour types. The immunohistochemical results on the tumours 3. ( of one repeat unit) so far investigated in this and previous studies are shown in Supplementary Table 1. Cytoplasmic NPM expression (indica- Putative wild-type genes before duplication 8 tive of NPM1 mutations) was restricted to 386/1123 (34.3%) of (6.86×10 sequences, if L =3) de novo AML. A representative example of NPMc þ AML is shown in Figures 2a and b. All other tumours exhibited nucleus- 4. (Exact search in the genome) restricted NPM positivity, which is fully predictive of wild-type NPM1.4 Immunohistochemistry allowed the study of tumour Wild-type genes found in the genome samples even under circumstances that would have presented (6 sequences) difficulties for NPM1 mutation analysis, that is, small percentage of tumour cells (for example, Hodgkin’s disease; Figure 2d), 5. (Result filtering) inability to collect neoplastic cells because of dry-tap (such as in primary myelofibrosis or in hairy cell leukaemia), small NES-generation candidates sized bioptic samples or lack of fresh material. Representative (1 sequence) examples of NPM immunostaining pattern of various types of tumours other than NPMc þ AML are shown in Figure 2(c–f). Figure 1 Flow chart summarizing our computational prediction of putative NES generation through tandem duplication in the human These immunohistochemical studies and recent reports of 13 genome. The number of sequences that will be processed at steps 4 mutational analysis of solid tumours, clearly point to NPM1 and 5 refers to repeat units of length 3. NES: nuclear export signal. mutations and consequent cytoplasmic dislocation of nucleo-

Leukemia Letters to the Editor 1287 Table 1 List of analyzed NES motifs

NES ID Signal Length L ¼ 3L¼ 4L¼ 5L¼ 6

4NES-0001 LALKLAGLDI 10 4NES-0002 EMFRELNEALELKD 14 4NES-0003 SISLSFDESLALCVI 15 4NES-0004 GIDLSGLTLQ 10 4NES-0008 DLCQAFSDVILAV 13 4NES-0009.1 SLPHAILRIDLA 12 4NES-0009.2 DIKEKLCYVALD 12 4NES-0010 LPPLERLTL 9 4NES-0011 KTCTPSGEAPLSACT 15 4NES-0012 LQKKLEELELDEQ 13 4NES-0015 LSEALLQLQF 10 4NES-0018 LLLKKMYLM 9 YES 4NES-0019 LCLSDLSLL 9 YES 4NES-0020 EAALSDALAALQIE 14 4NES-0021 ALPLGKLTL 9 4NES-0023 DIQELSEQIHRLLL 14 YES 4NES-0024 ELKDFLKELNIQVD 14 4NES-0025 LYPELRRILTI 11 4NES-0026 LDRKLLELLW 10 4NES-0030 QEDILDELLGNMVLA 15 4NES-0032 LTKMCTIRM 9 4NES-0041.1 LLERLKELNLDSS 13 YES 4NES-0041.2 LTKRIDSLPL 10 YES 4NES-0055 LTELEISSIFSHCC 14 4NES-0062 IDMLIDLGLDLSD 13 4NES-0063 FDRLETLILL 10 4NES-0064 LLSQDLEPLALSD 13 4NES-0065 LPLTLFPSANL 11 4NES_NPM_1 mutation A/Om/G/H/4/7 DLCLAVEEVSLRK 13 YES 4NES_NPM_2 mutation B/J DLCMAVEEVSLRK 13 4NES_NPM_3 mutation C/D DLCVAVEEVSLRK 13 4NES_NPM_4 mutation E DLWQSLAQVSLRK 13 4NES_NPM_5 mutation F DLWQSLEKVSLRK 13 4NES_NPM_6 mutation G DLWQCFAQVSLRK 13 4NES_NPM_7 mutation H DLWQCFSKVSLRK 13 4NES_NPM_8 mutation J DLWQSLSKVSLRK 13 4NES_NPM_9 mutation L DLSRAVEEVSLRK 13 4NES_NPM_10 mutation K DLWQSLSKVSLRK 13 4NES_NPM_11 mutation M DLCTAVEEVSLRK 13 4NES_NPM_12 mutation N/3 DLCHAVEEVSLRK 13 4NES_NPM_13 mutation O DLWQRFQEVSLRK 13 4NES_NPM_14 mutation P DLCTFLEEVSLRK 13 4NES_NPM_15 mutation Q/I DLWQRMEEVSLRK 13 4NES_NPM_16 mutation Gm/Km/Qm/I DLCRAVEEVSLRK 13 4NES_NPM_17 mutation Lm DLCRGVEEVSLRK 13 4NES_NPM_18 mutation Nm DLCQAVEEVSLRK 13 4NES_NPM_19 mutation 1 DLWQSMEEVSLRK 13 4NES_NPM_20 mutation 6 DLWQDFLNRLFKKIV 15 4NES_NPM_21 mutation 13/I DLCKAVEEVSLRK 13 4NES_NPM_22 mutation 10 DLWQCCSQVSLRK 13 4NES_NPM_23 mutation 12 DLCAAVEEVSLRK 13 Abbreviations: NES, nuclear export signal; NPM, nucleophosmin. Column 1: NES identification name, column 2: NES sequence, column 3: Motifs length, columns 4–7: matching occurrence in human genome.

phosmin as an event specifically restricted to AML. The reason Acknowledgements why NPM1, an ubiquitously expressed gene that encodes for one of the most abundant nucleolar proteins, is selectively This work was supported by the Associazione Italiana per la associated with myeloid but not lymphoid leukaemias,12 Ricerca sul Cancro (AIRC) and by the University of Foggia. Falini B remains a mistery. applied for a patent on the clinical use of NPM mutants. In conclusion, computational analysis and immunohisto- chemistry indicate that NPM1 mutation is a unique genetic A Liso1,6, A Bogliolo2,6, V Freschi2,6, MP Martelli3, SA Pileri4, event that is specific to AML. These findings, taken together with 1 5 5 5 4 M Santodirocco , N Bolli , MF Martelli and B Falini the distinctive biological and clinical features of NPMc þ AML, 1Department of Medical and Occupational Sciences, Institute further point to NPM1 mutations as the genetic marker of a new of Haematology, University of Foggia, Foggia, Italy; leukaemia entity that should be considered for inclusion in the 2Information Science and Technology Institute (ISTI), upcoming WHO Classification. University of Urbino, Urbino, Italy;

Leukemia Letters to the Editor 1288

Figure 2 NPM expression in various human neoplasms. (a) Acute myeloid leukaemia, FAB-M4 (bone marrow biopsy; haematoxylin–eosin; Â 800). (b) The same case as (a) showing aberrant NPM expression in leukaemic cell cytoplasm (arrow); endothelial cells exhibit nucleus-restricted expression of NPM (double arrows) (APAAP, haematoxylin counterstaining; Â 100). (c) Essential thrombocytaemia with clusters of dysplastic megakaryocytes showing nucleus-restricted NPM positivity (arrow); T indicates a bone trabecula (bone marrow biopsy; APAAP, haematoxylin counterstaining; Â 800). (d) Hodgkin and Reed-Sternberg cells of Hodgkin’s disease show restricted NPM nuclear positivity (arrow); double arrows indicate expected cytoplasmic NPM expression in a mitotic figure (lymph node biopsy; haematoxylin counterstaining; Â 800). (e) Breast carcinoma showing the expected NPM nucleus-restricted expression (one representative sample of breast tissue microarray; APAAP technique; haematoxylin counterstaining; Â 800). (f) Melanoma with the expected NPM nucleus-restricted expression. The arrows point to a pigmented tumour cell (skin biopsy; APAAP technique; hematoxylin counterstaining; Â 800). APAAP, alkaline phosphatase anti-alkaline phosphatase staining; NPM, nucleophosmin.

3 Institute of Haematology, University of Bari, Bari, Italy; 3 Pitiot AS, Santamaria I, Garcia-Suarez O, Centeno I, Astudillo A, 4 Institute of Haematology, Haematopathology Unit, University Rayon C et al. A new type of NPM1 gene mutation in AML leading of Bologna, Policlinico S. Orsola, Bologna, Italy and to a C-terminal truncated protein. Leukemia 2007; 21: 1564–1566. 5Institute of Haematology, University of Perugia, Perugia, Italy 4 Falini B, Nicoletti I, Martelli MF, Mecucci C. Acute myeloid E-mail: [email protected] or [email protected] leukemia carrying cytoplasmic/mutated nucleophosmin (NPMc+ 6These authors contributed equally to this work. AML): biologic and clinical features. Blood 2007; 109: 874–885. 5 Falini B, Bolli N, Shan J, Martelli MP, Liso A, Pucciarini A et al. Both carboxy-terminus NES motif and mutated tryptophan(s) are References crucial for aberrant nuclear export of nucleophosmin leukemic mutants in NPMc+ AML. Blood 2006; 107: 4514–4523. 1 Falini B, Mecucci C, Tiacci E, Alcalay M, Rosati R, Pasqualucci L 6 Bolli N, Nicoletti I, De Marco MF, Bigerna B, Pucciarini A, et al. Cytoplasmic nucleophosmin in acute myelogenous leukemia Mannucci R et al. Born to be exported: COOH-terminal nuclear with a normal karyotype. N Engl J Med 2005; 352: 254–266. export signals of different strength ensure cytoplasmic accumula- 2 Albiero E, Madeo D, Bolli N, Giaretta I, Bona ED, Martelli MF et al. tion of nucleophosmin leukemic mutants. Cancer Res 2007; 67: Identification and functional characterization of a cytoplasmic 6230–6237. nucleophosmin leukaemic mutant generated by a novel exon-11 7 Caburet S, Cocquet J, Vaiman D, Veitia RA. Coding repeats and NPM1 mutation. Leukemia 2007; 21: 1099–1103. evolutionary ‘agility’. BioEssays 2005; 27: 581–587.

Leukemia Letters to the Editor 1289 8 Fondon III JW, Garner HR. Molecular origins of rapid and 11 la Cour T, Gupta R, Rapacki K, Skriver K, Poulsen FM, Brunak S. continuous morphological evolution. Proc Natl Acad Sci USA NESbase version 1.0: a database of nuclear export signals. Nucleic 2004; 101: 18058–18063. Acids Res 2003; 31: 393–396. 9 Pearson CE, Nichol Edamura K, Cleary JD. Repeat instability: 12 Martelli MP, Manes N, Pettirossi V, Liso A, Pacini R, Mannucci R mechanisms of dynamic mutations. Nat Rev Genet 2005; 6: et al. Absence of nucleophosmin leukaemic mutants in B and 729–742. T cells from AML with NPM1 mutations: implications for the cell 10 Darvasi A, Kerem B. Deletion and insertion mutations in short of origin of NPMc+ AML. Leukemia 2007; e-pub ahead of print. tandem repeats in the coding regions of human genes. Eur J Hum 13 Jeong EG, Lee SH, Yoo NJ. Absence of nucleophosmin 1 (NPM1) gene Genet 1995; 3: 14–20. mutations in common solid cancers. APMIS 2007; 115: 341–346.

Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)

JAK2 Mutations are present in all cases of polycythemia vera

Leukemia (2008) 22, 1289; doi:10.1038/sj.leu.2405047; 0.1%. To address this issue, in 33 patients, we compared the test published online 13 December 2007 results in parallel with three assays. Of these patients, modified ARMS detected JAK2V617F in all 33 (100%), whereas standard The JAK2V617F point mutation has been described in 65–97% of ARMS detected JAK2V617F in 30 (91%) and pyrosequencing patients with polycythemia vera (PV).1 Previously, 17 phenoty- detected JAK2V617F in 27 (82%). Our results suggest that an pical cases of PV from a large group of patients from three assay with 0.1% sensitivity is required to detect adequately institutions were initially reported as JAK2V617F negative.2 These JAK2V617F in all PV patients. These results also suggest that an cases were reanalyzed in detail to determine the cause, if any, of assay with a sensitivity of 5%, as currently offered by many the JAK2V617F negativity. Reasons for this initial negativity laboratories, might miss a number of PV patients, especially included inaccurate clinical diagnosis, relatively insensitive those who have been treated. laboratory techniques, insufficient material for retesting, and 1 2 3 3 4 possible treatment effect. It was predicted that when tested YL Wang , K Vandris , A Jones , NCP Cross , P Christos , F Adriano2 and RT Silver2 appropriately, JAK2 mutations would be found in all new cases 1 2 Molecular Hematopathology Laboratory, Department of PV. of Pathology and Laboratory Medicine, Weill Medical College To address the assay sensitivity issue, we studied a total of of Cornell University, New York, NY, USA; 105 PV cases diagnosed by Polycythemia Vera Study Group 2Division of Hematology–Oncology, Department of criteria using three different assays of distinct analytical Medicine, Weill Medical College of Cornell University, sensitivity. We defined the sensitivity by mixing DNA from New York, NY, USA; JAK2V617F-containing human erythroid leukemia cells (HEL) 3Wessex Regional Genetics Laboratory, Salisbury District with that from normal peripheral blood mononuclear cells, so Hospital, University of Southampton, Salisbury, Wiltshire, UK and that 1% sensitivity would detect one part of HEL DNA in 100 4 parts of normal DNA. The first method used was pyrosequen- Division of Biostatistics and Epidemiology, Department of cing, which involves synthetic nucleotide extension by DNA Public Health, Weill Medical College of Cornell University, V617F New York, NY, USA polymerase. The method reliably detects JAK2 when the E-mails: [email protected]; [email protected] mutant allele is greater than 5%.1 The second method was a standard amplification refractory mutation system (ARMS) PCR assay, which detects the mutant allele when it occupies more than 1% of total DNA.3 The third method was a modified highly References sensitive ARMS assay.4 The analytic sensitivity of the modified ARMS assay reaches 0.1% and has a false positive rate of less 1 James C, Ugo V, Le Coue´dic J-P, Staerk J, Delhommeau F, Lacout C than 0.01% (we found 1 positive case in 300 individuals without et al. A unique clonal JAK2 mutation leading to constitutive 5 signalling causes polycythaemia vera. Nature 2005; 434: myeloid disorders). Of 105 PV cases, the combination of all 1144–1148. V617F three assays detected the JAK2 allele in 104. In the 2 Verstovsek S, Silver RT, Cross NCP, Tefferi A. JAK2V617F mutational remaining case, the JAK2 exon 12 deletion (E543–D544 del) frequency in polycythemia vera: 100%, 490%, less? Leukemia was detected. Therefore, mutations in JAK2 have been detected 2006; 20: 2067. in 100% of 105 PV patients examined. 3 Jones AV, Kreil S, Zoi K, Waghorn K, Curtis C, Zhang L et al. Of the eight patients who were not detected by the Widespread occurrence of the JAK2 V617F mutation in chronic myeloproliferative disorders. Blood 2005; 106: 2162–2168. pyrosequencing method, all had been treated; four had been 4 Chen Q, Lu P, Jones AV, Cross NCP, Silver RT, Wang YL. on interferon for a median of 13 years, two on imatinib for 3 Amplification refractory mutation system (ARMS), a highly sensitive years and two had been treated by phlebotomy for 7 and 28 and simple PCR assay for the detection of JAK2V617F mutation months, respectively. Therefore, sensitive methods such as in chronic myeloproliferative disorders. J Mol Diagn 2007; 9: ARMS-PCR assays presented here may be of particular value 272–276. in patients treated for some time with different modalities. 5 Wang YL, Lee JW, Kui JS, Chadburn A, Cross NCP, Knowles DM et al. Evaluation of JAK2V617F in B- and T-cell neoplasms: Although it has been realized that a sensitive assay is required identification of JAK2V617F mutation of undetermined significance V617F 2 to identify the JAK2 mutation in PV patients, it has never (JMUS) in the bone marrow of three individuals. Acta Haematolo- been shown which level of sensitivity is adequate, that is 5, 1 or gica 2007; 118: 209–214.

Leukemia