Evidence suggesting the presence of a stem cell clone anteceding the acquisition of the JAK2-V617F mutation

The JAK2-V617F mutation is mostly found in classic Philadelphia-chromosome negative myeloproliferative diseases, such as polycythemia vera, essential thrombocytosis (ET) and myeloid metaplasia/myelofibrosis. However, it remains unclear whether the JAK2-V617F mutation is the essential and sufficient molecular event leading to myeloproliferative disease (MPD). Several observations tend to support the existence of a pre-JAK2 mutated clone: (i) the diversity of hematological phenotypes associated with the mutation, which contrasts with the stereotyped phenotype of Philadelphia-positive chronic myelogenous leukemia, (ii) familial cases of MPDs where the JAK2-V617F mutation is thought to be an event occurring on a background of already clonal hematopoiesis or on a germline mutated cell predisposed to acquire new somatic mutations^1,2 and (iii) clonality studies in cohorts of JAK2-V617F-positive MPDs.^3,4 Furthermore, the fact that the leukemic blasts of acute myeloid leukemia (AML) derived from JAK2-V617F-positive MPDs are often negative for the mutation^5,6 raises the question whether MPD and AML arise from two different progenitor cells or are derived from a common ancestor in which the JAK2-V617F mutation is a late event leading to MPD.

Here, we describe the unusual case of a patient diagnosed with de novo AML in which the allogeneic transplantation setting allowed for the unique observation of the sequence of events leading to the acquisition of the JAK2-V617F mutation, and the possible multistep pathogenesis underlying the development of secondary AML derived from a preexisting MPD.

A 50-year-old man was diagnosed in March 2000 with AML FAB-M2 CD7+ characterized by an unfavorable hyperploid complex karyotype (53XY +2, +6, +8, +8, +13, +14, +18). Standard cytarabine-daunorubicin^3–7 induction led to a complete remission and a normal karyotype. This was followed by a high-dose cytarabine and daunorubicine consolidation.^3,4 The patient received a 6/6 HLA-matched allogeneic stem cell transplant from his sister. Immediately prior to transplant, karyotype and bone marrow studies were normal, while the complete blood count showed normocytic anemia. The conditioning regimen consisted of a busulphan–cyclophosphamide long course protocol with cyclosporine and methotrexate short course as prophylaxis of graft-versus-host disease (GVHD). The patient was infused with a peripheral stem cell graft of 9.9 × 10⁶ CD34 per kg and 19.7 × 10⁷ CD3 per kg. Neutrophil engraftment took place on day 16 and the patient was discharged on day 24. At 5 months after transplant, the patient presented with extensive chronic GVHD (cGVHD), which was persistent and required triple immunosuppressive therapy. At 35 months after transplant, the patient started to exhibit leucocytosis, followed by thrombocytosis. Progression of both the leucocytosis and thrombocytosis, attaining 85.3 × 10⁹ l⁻¹ and 856 × 10⁹ l⁻¹, respectively, led to the initiation of hydroxyurea 500mg daily at 42 months after transplant, which normalized his blood counts. The patient refused a second allogeneic transplantation. At 54 months after transplant, AML relapse was diagnosed on the basis of an 80 and 19% blastosis in the peripheral blood and bone marrow aspiration, respectively. The karyotype was normal. The patient died at 55 months post-transplantation from a septic choc occurring during his induction therapy.

This case of a patient transplanted for de novo AML and developing post-myeloablative transplantation, an MPD and ultimately leukemia relapse was further analyzed. Samples were obtained with informed consent from the patient and Institutional Review Board approval. Chimerisms were performed by PCR amplification of short-tandem repeat polymorphic markers on DNA of polymorphonuclear cells, T-cells and mononuclear fraction at different time points after transplantation. Cytogenetic studies were performed on the marrow pre-transplantation, and at 26, 37 and 54 months post-transplantation using standard procedures. The chimerism analysis shown in Figure 1 revealed that complete donor hematopoiesis was present 1 month after transplant in the peripheral blood polymorphonuclear cells, and remained so for 24 months, where routine chimerism analysis showed a mixed donor-host population indicating resurgence of the host hematopoiesis. At 26 months after transplant, chimerism studies on marrow polymorphonuclear cells revealed complete host hematopoiesis, and an extensive work-up showed no evidence of leukemic relapse. Hemoglobin was 112 g l⁻¹, WBC 12.7 × 10⁹ l⁻¹, platelets 329 × 10⁹ l⁻¹, neutrophils 9.4 × 10⁹ l⁻¹and lymphocytes 1.8 × 10⁹ l⁻¹. The karyotype was 46, XY [25]. At 35 months after transplant, when the complete blood count was compatible with a MPD, the marrow aspiration and biopsy showed no relapse of leukemia with a normal 46, XY [29] karyotype. The retrospective analysis of available samples for the JAK2-V617F mutation, using a real-time quantitative PCR assay,⁷ clarified the evolution of the patient (Figure 1). The initial pre-transplant AML was negative for the mutation. The mutation was also absent right before transplantation and remained negative for 7 months after the loss of donor hematopoiesis. At 32 months post-transplantation, a positive result for the mutation appeared coinciding with the development of leucocytosis and thrombocytosis at 35 months post-transplantation. The percentage of mutant JAK2 alleles progressively increased to reach 40% (possibly 80% cells heterozygous for the mutation) in both marrow and peripheral blood polymorphonuclear cells. At 54 months post-transplantation, relapse of AML was diagnosed and the karyotype on the mononuclear fraction of the peripheral blood (80% of blasts) was 46, XY [20]. The JAK2-V617F mutation analysis of the same mononuclear fraction showed 58% positivity, indicating that a significant subset of cells were homozygous for the mutation.

Figure 1.

Temporal evolution between donor chimerism and appearance of the JAK2-V617F mutation after allogeneic stem cell transplantation. A: Chronic graft-versus-host disease, B: phenotypically normal host hematopoiesis, C: time to transformation and relapse into JAK2-V617F positive AML, D: progressive thrombocytosis and leucocytosis, E: start of Hydrea, F: AML relapse with 58% positivity for JAK2 in peripheral blood. AML, acute myeloid leukemia.

The bone marrow transplant setting of this case allows for the unique observation of the sequence of events leading to both JAK2-V617F-positive MPD and its transformation into AML. The resurgence of a phenotypically normal host hematopoiesis at 2 years after myeloablative stem cell transplantation strongly suggests the presence of a mutated stem cell resistant to both the intensity of the preparative transplant regimen and the immunological challenge of the extensive cGVHD the patient went through. The recipient hematopoiesis was phenotypically normal but failed to reconstitute the T-cell compartment, which remained of donor origin. Ten months after the loss of donor myeloid chimerism, the appearance of a MPD associated with the detection of the JAK2-V617F mutation further suggests that the cells that reconstituted the host hematopoiesis had some level of genomic instability rendering them susceptible to the acquisition of the JAK2-V617F mutation. The subsequent evolution into AML confirms the genomic instability of this pre-JAK2 clone, as rapid evolution into AML is unlikely to occur in classic JAK2-V617F-positive MPDs. This sequence of events revealed by the transplantation setting is unique as the observation of a phenotypically normal but presumably mutated hematopoiesis preceding the development of de novo MPD in the general population would not be possible. Therefore, this case supports the hypothesis that there is a preexisting clone prior to the acquisition of the JAK2-V617F mutation. We also have to consider the alternative hypothesis that the resurgence of host hematopoiesis and loss of donor chimerism were not associated with a clonal growth advantage and resistant phenotype, but rather the result of a late graft failure with autologous reconstitution. This has been reported very rarely in the context of myeloablative, non-T-cell depleted transplantation, and has been described particularly in a small number of patients with Philadelphia-positive chronic myelogenous leukemia.⁸ These cases were characterized by the early occurrence of autologous reconstitution, by T-cell depletion in half the cases, and by absence of cGVHD in almost all the cases.⁸ In the present case, resurgence of host hematopoiesis occurred late and in the presence of a persistent donor-derived T-cell effector population (data not shown).

Finally, the fact that the patient was initially diagnosed with a JAK2-V617F-negative AML and that he ultimately died with a JAK2-V617F-positive AML argues in favor of a common pre-leukemic clone that leads to AML independent of the acquisition of the JAK2-V617F mutation. This is in line with the observation that only a proportion of AMLs secondary to JAK2-V617F-positive MPDs are negative for the mutation.^5,6 The preexisting JAK2-V617F-negative clone must therefore remain present even in the face of a full-blown JAK2-V617F-positive MPD and conserve the ability to acquire a transforming mutation leading to AML (Figure 2).

Figure 2.

Sequence of events leading to both JAK2-V617F positive myeloproliferative disease (MPD) and its transformation into JAK2-V617F positive acute myeloid leukemia (AML) in a patient initially diagnosed with a JAK2-V617F-negative AML who underwent allogeneic stem cell transplantation. Shown in dashed lines: Potential pathway for transformation into JAK2-V617F-negative AML.