JAKAFI (Page 10 of 12)

14.4 Chronic Graft-Versus-Host Disease

Study 5 (REACH-3; NCT03112603) was a randomized, open-label, multicenter study of Jakafi in comparison to best available therapy (BAT) for treatment of corticosteroid-refractory cGVHD after allogeneic stem cell transplantation. Eligible patients were ≥ 12 years old with moderate or severe cGVHD as defined by NIH Consensus Criteria requiring additional therapy after failure of corticosteroid therapy and no more than one additional salvage treatment. Patients were excluded if they had ANC < 1 Gi/L and platelet count < 25 Gi/L, estimated creatinine clearance < 30 ml/min, progressive onset cGVHD, oxygen saturation < 90%, total bilirubin > 2 mg/dL, or diarrhea due to GVHD.

A total of 329 patients were randomized 1:1 to receive either Jakafi 10 mg twice daily (n=165) or BAT (n=164). BAT was selected by the investigator prior to randomization and included the following treatments: extracorporeal photopheresis (ECP), low-dose methotrexate (MTX), mycophenolate mofetil (MMF), mTOR inhibitors (everolimus or sirolimus), infliximab, rituximab, pentostatin, imatinib, or ibrutinib. Randomization was stratified by cGVHD severity (moderate versus severe). On Cycle 7 Day 1 and thereafter, patients randomized to BAT could cross over to Jakafi if they had disease progression, mixed response, unchanged response, cGVHD flare, or toxicity to BAT. All patients also received standard supportive care, including anti-infective medications. GVHD prophylaxis and cGVHD treatment medications initiated before randomization, including systemic corticosteroids, calcineurin inhibitors, and topical or inhaled corticosteroid therapy, were allowed to be continued per institutional guidelines. Table 24 shows the demographics and baseline disease characteristics of the randomized population.

The efficacy of Jakafi was based on overall response rate (ORR) through Cycle 7 Day 1, where overall response included complete response or partial response according to the 2014 NIH Response Criteria and durability of the response. The ORR results are presented in Table 25; the difference in ORR between Jakafi and BAT arms was 13% (95% CI 3%, 23%). The median time to first response in the responders was 3 weeks (range, 2 to 24) for the Jakafi arm and 4 weeks (range, 2 to 25) for the BAT arm. The median duration of response, calculated from first response to progression, death, or new systemic therapies for cGVHD, was 4.2 months (95% CI 3.2, 6.7) for the Jakafi arm and 2.1 months (95% CI 1.6, 3.2) for the BAT arm; and the median time from first response to death or new systemic therapies for cGVHD was 25 months (95% CI 16.8, NE) for the Jakafi arm and 5.6 months (95% CI 4.1, 7.8) for the BAT arm.

ORR results were supported by exploratory analyses of patient-reported symptom severity which showed at least a 7-point decrease in the cGVHD Total Symptom Score at any time through Cycle 7 Day 1 in 66 (40%; 95% CI 32, 48) patients in the Jakafi arm and 47 (29%; 95% CI 22, 36) patients in the BAT arm.