Polycythemia vera following autologous transplantation for AML: insights on the kinetics of JAK2V617F clonal dominance

To the editor:

A JAK2V617F mutation is harbored by most patients with polycythemia vera (PV) and has pathogenetic¹  and diagnostic²  relevance. We used this molecular marker to trace origin of disease in a case of PV manifested 5 years after autologous transplantation for acute myeloid leukemia (AML).

A 60-year-old man was diagnosed as having AML, M2 FAB subtype with normal karyotype, in January 1999. Complete remission was obtained after induction chemotherapy with fludarabine, cytarabine, and idarubicin, followed by consolidation with idarubicin and etoposide. He was conditioned with oral busulphan and cyclophosphamide and reinfused with 4.3 × 10⁶/kg peripheral blood (PB)–primed C34⁺ cells in December 1999. Clinical and hematologic follow-up, and bone marrow (BM) biopsies performed yearly thereafter, were unremarkable until March 2006, when he reported pruritus after hot showers, visual disturbance, and paresthesia of a few months' duration. Spleen was palpable, white blood cell count (WBC) was 6.1 × 10⁹/L; Hb, 187 g/L (18.7 g/dL); platelet count, 517 × 10⁹/L; LDH, 558 U/L; serum ferritin, 10 ng/mL; and erythropoietin 7.6 U/L; endogenous erythroid colonies were present, and karyotype was normal. BM biopsy showed increased cellularity with panmyelosis, prominent erythroid, and megakaryocytic proliferation; JAK2V617F mutation was found,³  and diagnosis of PV was established.

Archived BM or PB samples were analyzed for V617F mutation with quantitative polymerase chain reaction (PCR; sensitivity, < 1%; interassay, ≤ 3%).⁴  Leukemic blasts at diagnosis, BM cells at remission, and leukapheresis were all V617F negative. A 20% V617F allele burden was first detected in BM aspirate 4 years after transplantation and it was 25% 1 year later (May 2005); ad hoc revision of both biopsies failed to satisfy criteria for PV.⁵  At diagnosis, granulocyte V617F allele burden was 28% (Figure 1).

Although the patient's JAK2 genotype before AML was unknown, normal hematologic parameters 1 year before would exclude a V617F-negative AML transformed from preexisting, unrecognized, V617F-positive PV.⁶  A JAK2V617F-positive essen-tial thrombocythemia after transplantation for AML has also been reported.⁷ 

Availability of sequential samples after AML diagnosis allowed us to trace emergence of mutant clone and to correlate it with disease manifestation. In this patient, the target cell for JAK2V617F mutation was most likely a reinfused hematopoietic stem cell (HSC) that had been exposed to chemotherapy for AML. The time interval between transplantation and appearance of V617F allele was 4 years; of note, this interval is similar to that in a case of PV that developed after chemotherapy for Hodgkin disease⁸  and significantly shorter than the 10 to 20 years reported in nuclear explosion exposure.⁹  Stochastic calculation of the kinetics of clonal dominance in animals determined that if the number of neoplastic HSCs reaches 0.5% of total HSCs, survival and expansion of the clone are assured.¹⁰  In this particular case, expansion and dominance of a putative V617F mutant cell might have been facilitated by the relatively low total number of HSCs after transplantation, and therefore cannot be considered to reproduce clonal evolution kinetics under usual circumstances. On the other hand, we observed that there was only a 2-year lag phase between first detection of mutant cells and manifestation of PV phenotype.

Therefore, although we cannot generalize from a single observation, this case supports the idea that expansion of JAK2V617F mutant clone in PV is strictly linked with appearance of disease-associated abnormalities.