Final Analysis at 5 Years Follow up of Patients with MPN-Associated Splanchnic Vein Thrombosis Treated with Ruxolitinib in a Phase 2 Study

Background: A myeloproliferative neoplasm (MPN) is frequent underlying cause of splanchnic vein thrombosis (SVT). We reported that ruxolitinib, a JAK1/2 inhibitor, was safe in patients (pts) with MPN-associated SVT and effective in reducing spleen size at the planned primary analysis at 24 weeks (w) in an investigator-initiated phase II clinical trial (Pieri L, AJH 2016). Herein we present final long term follow up (FU) data.

Methods. A total of 21 MPN-SVT entered the trial in 2012. Pts who completed the 24w core study and well tolerated the drug and had evidence of clinically-significant improvement were allowed to enter an extension phase up to w72, then entered a long term FU. The drug was provided by Novartis, that had no role in trial design nor in data analysis. Safety data were reported as cumulative incidence of adverse events (CTCAE v4.03). Clinical responses were defined according to ELN and IWG-MRT criteria.

Results.

Patients disposition. Eighteen patients were alive at last follow-up at a median of 5.5y, range 3.7-6.5y. Two pts died before w72 (hepatocarcinoma; unknown cause), 1 pt died at w252 (sepsis). Diagnosis of the 18 pts was: PMF 8 (44%), PV 5 (28%), ET 4 (22%), PET-MF 1 (6%). The DIPSS risk score of the 8 pts with PMF was low in 4 and intermediate-1 in 4. At last FU, 14 patients are still on ruxolitinib (78%), 10 pts still on clinical trial, 4 pts shifted to commercial drug. Four pts (22%) withdrawn treatment after w72 because of inefficacy, withdrawal of consent, unknown reasons, shift to observation only (1 each).

Safety. Thirteen pts (72.2%) had adverse events; the median number of events per patient was 5 (range, 0-37). Five pts had G2 thrombocytopenia plus 1 pt G3 (28%), 9 had G2 anemia (50%), 1 had G2 neutropenia (6%). There were 4 non-hematologic G3 adverse events (22%), one drug-related (CPK increase) that resolved with dose reduction. Six pts developed G1-2 infections (33%), 4 had Herpes Zoster reactivation (22%), 1 developed second cancer (pheochromocytoma at w96; 6%).

Efficacy. Spleen response (defined as percentage reduction of spleen length from left costal margin (LCM) from baseline value as by IWG-MRT criteria) was available in 16 of 18 pts on treatment. Of 8 pts who obtained spleen response at w72, 4 pts had a complete response, while the remaining 4 pts had a spleen reduction >50%. At last FU, 6 of 8 pts (75%) adjudicated as spleen responders at w72 maintained the response. None of the 8 pts who had not obtained a spleen response at w72 acquired it at the last FU. According to ELN/IWG-MRT criteria of disease response, 7 pts had partial response, 11 had stable disease. At abdominal vessel eco scan, thrombosis improved in 3 of 12 evaluated pts (25%) and remained stable in the remaining 9. Of the 16 pts with endoscopic evaluation at last FU, complete resolution of oesophageal varices was documented in 3/12 with baseline varices (25%), worsening in 4 (33%), stable in 5 (42%), while no de-novo formation of F1 varices was documented in the remaining 4 (33%). No pt had gastrointestinal bleeding episodes.

Exploratory endpoints. Included changes in the level of JAK2V617F variant allele burden (VAF) at last FU, available in 13 of 15 JAK2V617F mutated pts. VAF remained stable in 2 pts (13%), increased by 34% (range 25-40) in 3 pts, while 8 pts (62%) had a reduction >10% (range 18-96), of whom 6 (46%) had a reduction >50% (median 82%, range 70-96). By correlative analysis of JAK2 VAF changes and spleen response, 3 of 6 pts with VAF reduction of >50% had complete spleen response, as compared to none of the 3 pts who had an increase of VAF.

Conclusions. After a median follow-up of 5.5 years, ruxolitinib continues to be safe in pts with MPN-associated SVT and maintains efficacy against splenomegaly in 33% of the pts. Of note, the large majority of pts (67%) showed stabilization or improvement of oesophageal varices, supporting the hypothesis that sustained reduction of enlarged spleen might contribute to decrease the upstream venous system pressure. Significant reduction of JAK2V617F VAF>50% was documented in 40% of the pts, although it was not clearly correlated with clinical improvement.