What Is the True Response Rate to Ruxolitinib in Persons with Myeloproliferative Neoplasm (MPN)-Associated Myelofibrosis (MF) Needing Therapy for Splenomegaly ?

Data from COMFORT-I and COMFORT-II trials indicate that the JAK1/JAK2 inhibitor ruxolitinib is effective in reducing splenomegaly and reversing symptoms in patients with MPN-associated MF (Verstovsek S. et al, NEJM 2012;366:799; Harrison C. et al. NEJM 2012;366:787). However, not all subjects in these studies needed therapy for splenomegaly according newly-reported criteria (Barosi et al, Leuk Res 2014;38:155). Moreover, splenomegaly response was defined as 35% reduction in spleen volume by magnetic resonance imaging, which does not always correspond to a clinically relevant response as defined by the IWG-MRT/ELN revised response criteria (Tefferi et al. Blood 2012;122:1395).

We evaluated response in 57 subjects with a spleen > 10 cm from the costal margin or with a progressive splenomegaly. Patients had been enrolled in clinical trials, or in individual patient care. The majority of patients were males (32/57, 56.1%), and the median patient age was 57 years (range: 30–81 years). The disease conditions included primary MF in 45 patients (78.9%), post-ET or post- PV MF (12 patients, 21.1%). Genotype was JAK2V617F in 37, 26 with > 50% (45.6%) and 11 (19.3%) with 50% or less allele burden. Fifteen (26.3%) had a CALR mutation, 4 a MPL mutation (7%), and 1 no detectable mutation. Starting dose of ruxolitinib given twice daily was: 20 mg in 32, 15 mg in 22, and 10 mg in 3. Doses were adjusted as appropriate. Median actuarial follow-up was 23.3 months. Thirteen subjects (37%) had a dose-reduction at a median time of 34 days (range 8-510 days). Twenty-eight (49.1%) discontinued therapy at a median of 17. months: 3-month and 1-,2-, and 3-year discontinuation rates were 22%, 40%, 55%, and 67%, respectively. Reasons for therapy-discontinuation were lack of response/progression (N=14), or toxicity/adverse events (N=14).

Nineteen subjects (33%) responded according to the IWG-MRT/ELN response criteria at a median time of 3.9 months (range, 12 days - 21 months). However, 7 subsequently lost their response after a median time of 2.3 months (range, 1-35 months). 3-year continuous response rate was 24%.

We next estimated failure-free survival (FFS), defined as continuing on ruxolitinib, no start of another therapy for splenomegaly, and/or no disease progression. This composite endpoint is more relevant than the individual components, but its validity after ruxolitinib in MF has not been evaluated. Actuarial FFS was 84% at 3 months, 71% at 6 months, 61% at 12 months, 44% at 2 years, and 36% at 3 years. Subjects with a JAK2V617F allele burden >50%, or with a liver <5 cm="" had="" significantly="" longer="" ffs="" log="" rank="" test="" p=".05," and="" respectively="" multivariate="" analysis="" identified="" jak2v617f="" allele="" burden="">50% as independently associated with FFS. Relative failure risk of subjects with >50% JAK2V617F allele burden was 0.28 (95% CI = 0.12 to 0.66), P=0.003 relative to all other genotypes (Figure). There was no significant association between FFS and age, gender, type of MF, hemoglobin concentration, white blood cells, platelet count, reason for starting therapy, spleen size, CD34+ levels, starting dose, or dose reductions.

In conclusion, ruxolitinib produced a 3-year IWG-MRT/ELN response rate of 24% in persons needing therapy for splenomegaly. However, actuarial 3-year FFS was 36%. Patients with JAK2V617F and more than 50% allele burden had significantly better FFS than other genotypes, perhaps reflecting a greater sensitivity to inhibition of JAK1/JAK2 by ruxolitinib. These results may help define which persons with MPN-associated MF and splenomegaly needing therapy are most likely to benefit from ruxolitinib.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal