Allogeneic Stem Cell Transplantation for Primary or Secondary Myelofibrosis: A Retrospective Intent-to-Treat Analysis and Impact of Mutational Status and JAK1/2 Inhibitor Ruxolitinib Prescription in Patients Who Cannot Proceed to Transplantation

Introduction: Allogenic stem cell transplantation (Allo-SCT) remains currently the only curative treatment for primary myelofibrosis (PMF) or myelofibrosis secondary (SMF) to essential thrombocythemia (ET)/polycythemia vera (PV). Indication of allo-SCT refers to high-risk patients defined by various scores such as Lille score, IPSS, DIPSS or more recently DIPSS+. However, JAK2, CALR, MPL or triple negative mutation profiles have recently been shown to impact on the outcome of PMF patients and thus may challenge the indication of allo-SCT in the future. CALRmut patients seem to have the best survival conversely to triple-negative patients, while JAK2mut and MPLmut status can be considered as an intermediate molecular signature (Tefferi, Leukemia 2014). The outcome of MF patients with a validated allo-SCT indication who do not proceed to transplant is generally unknown although relatively long survival can be documented, especially since the availability of JAK inhibitors such as ruxolitinib. Also, data regarding impact of mutational status on outcome after allo-SCT remain scarce.

Patients and Methods:This single-center retrospective study considered MF patients with a validated indication of allo-SCT between 2000 and 2013. The main objective was to compare the overall survival (OS) between patients who ultimately received allo-SCT or not. Secondary objectives were to analyze the impact of mutational molecular status on OS and the use of ruxolitinib in patients not allografted.

Results: An indication of allo-SCT was validated in 67 patients  (males: 64%; PMF: 59%; SMF post-ET 25%, SMF post-PV 16%). At the time of indication of ASCT, the median age was 59 years (range: 41-69); DIPSS+ score was: int-2 in 40%, high risk in 60%. Mutational status, available for 86% of patients, was as follows: JAK2_V617F (n=37), CALRmut (n=12); MPLmut (n=3), triple-negative (n=6). Thirty-three patients proceeded to allo-SCT (reduced intensity conditioning: 82%) while 34 did not for lack of donor (31%), comorbidity (28%), progressive disease (14%), stable disease on conventional therapy (18%), refusal (9%).  The two groups (allo/non-allo SCT) were comparable for gender, year of indication of allo-SCT, type of MF, number of prior lines of treatment before allo-SCT, DIPSS+ categorization, mutational status and median follow-up. Patients who did not proceed to allo-SCT were significantly older (61 vs 57 years, p<0.005). Two allografted patients received ruxolitinib after transplant while 9 patients received it after invalidation of the transplant procedure. With a median follow-up of 44 months, median OS was 57 months for the whole cohort.  A significantly lower OS was associated to unfavorable karyotype (median: 17 vs 100 months, HR: 2.6, 95%CI: 1.1-6.5, p=0.02), higher percentage of circulating blasts at time of indication of SCT (HR: 1.15, 95%CI: 1.01-1.32, p=0.01) and higher DIPSS+ score (HR: 1.3, 95%CI: 1.01-1.6 p=0.01). There was a trend for a better OS in CALRmut patients compared to other patients (median not reached (NR) vs 35 months, HR: 0.63, 95%CI: 0.25-1.51, p=0,3), especially vs triple-negative patients (NR vs 15 months, HR:0.29, 95%CI: 0.09-0.9 p=0.025). Five-year OS was similar between both groups (allo: 51% vs non-allo 44%, p=ns). However, progression to myelodysplastic syndrome or acute myeloid leukemia was significantly lower for allografted patients (19.2% vs 44.5%, p=0.01). In univariate analysis, allo-SCT benefited to patients with SMF post TE/PV (median OS: allo 115 vs non-allo 18.4 months, HR: 0.18, 95%CI: 0.05-0.5, p=0.004), unfavorable karyotype (median OS: allo: NR vs non-allo 7.9 months, HR: 0.13, 95%CI: 0.04-0.4, p=0.002) or high DIPSS+ score (median OS: allo 120 vs non-allo 18 months, HR:0.41,  95%CI: 0.16-0.9, p=0.04). Also there was a trend for better OS in allografted patients with high JAK2_V617Fburden (>65%) (median OS: NR vs 18 months, HR: 0.18, 95%CI:0.03 -1.11, p=0.065). Interestingly, the survival of patients who did not proceed to allo-SCT was increased by the use of ruxolitinib (median OS: NR vs 20 months, HR: 2.3, 95%CI: 0.08-0.6, p=0.003).

Conclusion: Allo-SCT remains a valid strategy for high-risk MF patients with unfavorable karyotype, high DIPSS+ score and secondary MF while impact of mutational status and JAK2_V617Fburden have to be confirmed in larger studies. MF patients who cannot proceed to transplant likely benefit from JAK2 inhibitor prescription.