Long Term Follow up of a Phase 2 Study of Ruxolitinib in Patients with Splanchnic Vein Thrombosis Associated with Myeloproliferative Neoplasm

Background: Philadelphia-negative Myeloproliferative Neoplasms (MPN) include Polycythemia Vera (PV), Essential Thrombocythemia (ET) and Myelofibrosis, both Primary (PMF) and secondary to PV or ET (PPV-MF and PET-MF). A MPN is frequently the underlying cause of splanchnic vein thrombosis (SVT). Ruxolitinib, a JAK1/2 inhibitor, efficiently reduced spleen volume and improved symptoms in patients (pts) with MF and PV in the COMFORT-I/II and RESPONSE phase III trials, and in ET pts in a phase II study. We reported (Blood 2014 124:3192) that ruxolitinib was safe in pts with MPN associated to SVT and effective in reducing spleen size at the planned primary endpoint analysis at 24 weeks (w) in a phase II clinical trial. Herein we present follow up data with cut off at 1 year after core period (a total of 72 w of treatment).

Methods: Main enrolment criteria included diagnosis of PV, ET, PMF or PPV-/PET-MF associated with SVT, splenomegaly >5 cm below costal margin (bcm), active anticoagulant or antiaggregant thrombosis prophylaxis, platelet count (plt) >100 x10⁹/L, neutrophils count >1x10⁹/L, normal hepatic and renal function, absence of esophageal varices >grade 2. Pts who completed the 24 w of study treatment and tolerated well the drug and had evidence of clinically-significant improvement were allowed to enter an extension phase aimed at collecting and reviewing safety and efficacy data. The drug was provided free of charge by Novartis, that had no role in trial design nor in data analysis.

Results: Diagnosis of MPN were: PMF 8 (38.1%), PV 5 (23.8%), ET 4 (19.1%), PPV-MF 3 (14.3%), PET-MF 1 (4.8%). Nineteen pts had spleno-porto-mesenteric thrombosis and 3 Budd-Chiari syndrome (BCS); one pt had both sites involved. Initial dose of ruxolitinib was 10 mg BID for PV, 25 mg BID for ET, 15 mg BID for MF pts with baseline (bl) platelet count of 100 to 200x10⁹/L and 20 mg BID for platelet count >200x10⁹/L. Currently 17/21 pts are on active treatment, 14 completed w72; final data for all 17 pts will be available at meeting. One pt with MF discontinued from the study being shifted to commercial ruxolitinib at w60, one ET and one MF pt discontinued for inefficacy at w24 and one MF pt for an unrelated adverse event after w72. Efficacy: 13/21 (61.9%) pts obtained a ≥50% spleen length (sl) reduction by palpation at w24, that was maintained at w72 in 8/14 pts (57.1%). Median sl reduction at w72 was 63% (range 0-100). No significant differences in resistive or pulsatility index of splanchnic artery were noted, nor in esophageal varices status evaluated at w72. 10/11 evaluated pts with echocardiography at w72 showed a median reduction of the cardiac output of 20.1% (range 2.3-42.2) mainly due to a reduction of heart rate and of cardiac index (-21.9%, range 8.8-44.3) due to increase in body surface area. The first effect could be attributed to decrease of proinflammatory cytokines, the second to weight gain associated with ruxolitinib. Symptomatology was evaluated by MPN-SAF up to w24, showing a median total symptom score reduction from 65 to 42. Safety: regardless of drug relationship, the most common adverse events (AE) (% any grade, % grade ≥3) were thrombocythopenia (57.1%; 14.3%) and anemia (33.3%, 19%) that were the main reasons for dose adjustments. Other AE included AST or ALT increase (42.9%, 0%), diarrhea (28.6%, 0%), abdominal pain (23.8%, 0%), ascites (19%, 0%), fever (23.8%, 0%), neutropenia, (9.5%, 9.5%), upper airways infection (19%, 0%), weight gain (14.3%, 4.8%), muscle cramps (14.3%, 0%). Three serious AE occurred: one case of hepatocarcinoma in a pts with BCS, one grade 2 pneumonia and one grade 2 haematemesis not related to esophageal varices. Median ruxolitinib total daily dose at w72, after dose adjustments, was 19.1 mg for MF, 16 mg for PV and 28.3 mg for ET. Median hemoglobin reduced from 12.9 gr/dL (range 9.4-16.7) at bl to 10.7 (8.4-14.4) at w16 and recovered at w72 (12.1, range 10.8-14.7). No pts received transfusions. Median platelet count was 212 x10⁹/L (100-389) at bl, reached to the lowest level at w4 (139, range 48-252) and improved to 160 (69-285) at w72. Median leukocyte count decreased from 7.3 x10⁹/L (1.8-16.4) at bl to 4.08 (1.2-21.7) at w 24, and remained substantially stable through w 72 (4.96; range 2.45-17.3). Median reduction of JAK2 allele burden at w72 was 9% (range 0-38).

Conclusions: At w 72 follow up, ruxolitinib continues to be safe in pts with MPN associated to SVT and maintains efficacy against splenomegaly in 57% of the pts.