Abstract

The JAK2V617F mutation is found in 60% of patients (pts) with primary myelofibrosis (PMF) and >90% with post-polycythemia vera MF (PPV-MF); few pts with post-essential thrombocythemia MF (PET-MF) have been reported. There are conflicting results about clinical relevance of V617F mutational status and/or burden in PMF, while no study specifically addressed this point yet in PPV/PET-MF. An association of mutation with poorer survival in PMF has been reported (Campbell P et al, Blood, 2006), while Tefferi A et al (Leukemia, 2008) found a reduced survival only for pts with lowest allele burden. Barosi G et al (Blood, 2007) found that JAK2V617F mutation independently predicted evolution towards large splenomegaly, splenectomy and leukemic transformation. We separately evaluated 2 groups of pts, 240 pts with PMF and 65 pts with secondary forms of MF (PPV-MF n=43, PET-MF n=22); the V617F allele burden was measured in granulocytes by RT-PCR. Diagnosis of PMF satisfied the 2008 WHO criteria, and that of PPV- or PET-MF the IWG-MRT criteria.

PMF: 150 pts (63%) were JAK2V617F-pos, median V617F allele burden 45% (range, 1–100%); percentage of pts in the four allele quartiles was 16%, 50%, 15%, and 19%. Median interval from diagnosis to genotyping was 21 months. In multivariate analysis, JAK2V617F mutated patients had significantly higher leukocyte count (P=.02) and hemoglobin level (P<.0001), that were both positively correlated with V617F burden (P= .03 and P=.01, respectively). Pts with Hb <10g/dL were significantly less among JAK2V617-pos (12% vs 26%; P<.007), accounting for a significantly greater proportion of mutated pts in the Lille low risk category (P=.04). Nor a mutated JAK2 genotype nor the V617F burden had any influence on presence or degree of palpable splenomegaly, thrombosis, hemorrhage, AML evolution, or cytoreductive treatment. 29 pts died (12.0%); survival was independent of JAK2 genotype, and was associated (multivariate analysis) only with age (95% CI, 1.07–1.18) and leukocyte count (95% CI, 1.02–1.10). Since we previously observed a time-dependent accumulation of mutated alleles in PMF, we performed a sub-group analysis in 90 pts in whom quantitation of V617F burden was performed within six months from diagnosis: 56 pts (62%) were mutated, median V617F burden was 36% (range, 10–100%); percentage of pts in the four allele quartiles was 27%, 50%, 16%, and 7%. Similar to the whole pts series, the only parameters significantly associated with allele burden were hemoglobin level, (P=.03), leukocyte count (P=.04) and low Lille score (P=.04). However, we found a statistically significant association (multivariate analysis) of lowest V617F burden quartile with shortened survival compared to patients having greater than 25% V617F allele or to JAK2V671F unmutated patients, corroborating finding from Tefferi et al (Leukemia, 2008). Using Cox analysis, reduced survival was associated with low V617F burden (P=.008), leukocytosis (P=−003) and age (P=.03). These findings indicate that a low V617F burden, measured at the time of diagnosis, has a negative impact on survival, and might represent a novel criterion for risk stratification in PMF.

PPV/PET-MF: 49 pts (75.3%) were JAK2V617F mutated, median V617F allele burden 73% (10–100%). All 43 PPV-MF were mutated compared to 6/22 (27%) of PET-MF (P<.01). Median V617F burden was significantly greater in PPV-MF than in 173 control PV pts (72% vs 52%; P<.01), as well as in PET-MF (57% vs 26% in 200 control ET pts; P<.001), supporting that accumulation of V617F alleles is a mechanism of evolution toward MF in JAK2V617F-positive PV or ET pts. However, since JAK2V617F mutation frequency was significantly lower in PET-MF pts (27.2%) than in ET pts (63.4%; P<.01), genetic mechanisms other than JAK2V617F may have a dominant role in MF transformation of ET. A part for lower platelet count (P=.04) in JAK2V617F-pos PPV/PET-MF pts no other difference with unmutated pts was found. In the analysis of quartiles, we found a correlation of allele burden with age, leukocyte count and circulating CD34+ cell count (all P<.05), while there was no impact on clinical characteristics including spleen size, thrombosis, hemorrhage, AML transformation or overall survival (median follow-up, 40 months). Therefore, both presence and burden of JAK2V617F seem to have little influence on disease phenotype in pts with PPV- or PET-MF, although larger series and longer follow-up might be warranted.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author