Mutational Profile of Patients with Polycythemia Vera Treated with Ruxolitinib in the Phase III Controlled Response Study

Background. The JAK1/2 inhibitor ruxolitinib (RUX) demonstrated clinical benefit compared with best available therapy (BAT) in patients (pts) with polycythemia vera (PV) in a phase 3 study (RESPONSE TRIAL) (NEJM 2015; 372:426). A reduction of JAK2V617F allele burden of 12.1% from baseline at week (w)32 was observed in pts receiving ruxolitinib compared to BAT (1.2%). Conversely, no information was available about other mutations that may occur in some PV pts.

Aim. To analyze the molecular landscape of PV pts enrolled in the RESPONSE trial specifically as regarded subclonal mutations.

Methods: In the RESPONSE study, PV pts with intolerance or refractoriness to hydroxyurea, showing an enlarged spleen volume (SV) >450 ml and phlebotomy requirement, were randomized (1:1) to receive RUX (n = 110) or BAT (n = 112). After institutional approval and informed written consent, samples for genotyping were available in 150 cases (67.5% of total, 75 each RUX and BAT). Mutations in 22 genes (JAK1, JAK2, JAK3, EZH2, ASXL1, TET2, IDH1, IDH2, CBL, SRSF2, DNMT3A, NFE2, SOCS1, SOCS2, SOCS3, SH2B3, STAT1, STAT3, STAT5A, STAT5B, SF3B1, U2AF1) were analyzed in blood DNA at baseline and at the latest available sample by deep sequencing with Ion Torrent-PGM. CALR mutations were analyzed by capillary electrophoresis. JAK2V617F allele burden was confirmed by RT-qPCR assays.

Results.  149/150 patients had informative sequencing results (RUX, n = 74 and BAT, n = 75). Mutation frequency at baseline in RUX pts was: JAK2V617F 97.3% (mean allele burden=83.7±20.0%); JAK2Exon12 1.3%; other mutations in JAK2 5.4%; JAK3 1.3%; ASXL1 4.0%; TET2 20.3%; EZH2 4.0%; CBL 1.3%; SH2B3 2.0%; SOCS1 2.7%; NFE2 2.7%; STAT5A 1.3%. No somatic variants were detected in CALR, JAK1, SRSF2, IDH1, IDH2, SOCS2, SOCS3, DNMT3, STAT1, STAT5B, SF3B1, U2AF1. These frequencies were comparable in BAT arm. One patient was un-mutated in all assessed genes. 28.4% and 8.1% of RUX and 32.0% and 8.0% in BAT pts had 1 and >2 subclonal mutations, respectively. The proportion of PV pts harboring at least 1 mutation in either ASXL1, EZH2, SRSF2, IDH1/2, was significantly lower (8.1% in RUX and 10.6% in BAT) compared to  reference series of primary myelofibrosis pts (31%) as it was for those having >2 HMR mutations (0.0 in RESPONSE vs 7.4% in PMF) (Leukemia 2014;28:1804)

The median duration of treatment corresponding to the latest available sample for analysis was 82.8w; at that time,  43 pts (58.1%) randomized to RUX achieved a JAK2V617F allele burden reduction ≥10%, of which 15 (20.3%) had >50% reduction. Among the latter pts, the median allele burden was 83.7% at baseline, 84.9%, 55.1% and 44.3% at 1, 2, and 3 years. Three patients attained an allele burden below 5% (from 65.1%, 17.3% and 83.7% at baseline to 3.2%, 0.5% and 1.4%, respectively, at latest follow up).

Of the 27 pts harboring subclonal mutations at baseline, 12 (44.4%) presented a reduction of mutational allele burden ≥10%: 4 in JAK2 (other than JAK2V617F/exon 12 mutations), 4 in TET2, 3 in ASXL1 and 1 each in JAK3, EZH2 and SH2B3. In 10 of the 12 pts, comparable decreases in JAK2V617F allele burden were observed, suggesting reduction of a single clone expressing both mutations. Conversely, in 5 pts (18.5%) the allele burden of a baseline TET2 clone at increased by ≥10% (range: 10-37%); of these, one had concurrent reduction of JAK2V617F burden from 17.3 to 0.5, thereby suggesting two independent clones.   

Eight pts (29.6%) acquired new mutations: 3 in TET2, 3 in U2AF1, 1 in DNMT3A and 1 in IDH1. Among these, 4 pts had achieved a reduction ≥10% of JAK2V617F allele burden (18.4%, 25.4%, 30.5% and 39.4%, respectively). Three pts (4.0%) progressed during treatment (2 myelofibrosis,1 acute leukemia); no novel acquired mutation in the 22 genes was observed in these pts. All 3 pts were homozygous for V617F (92.3%, 72.7% and 59.0%) and did not show appreciable changes of allele burden during treatment.

Conclusions. The current study identifies mutations and mutational combinations at baseline and during follow up in a representative cohort of pts enrolled in RESPONSE trial and treated with Ruxolitinib. We observed progressive reduction of JAK2V617F allele burden that in some cases was associated with concurrent reduction of subclonal mutations. Conversely, emergency of novel clones as observed in some pts, whose significance might be clarified by ongoing analysis of hydroxyurea and phlebotomy treated patients that will be presented at the meeting.