Phase 1 Trial Investigating the Safety and Tolerability of Leflunomide in Patients with Steroid-Dependent Chronic Graft-Versus-Host Disease (cGVHD) after Allogeneic Hematopoietic Cell Transplant (alloHCT)

Current FDA approved agents for cGVHD are associated with low complete response rates and high cost. Leflunomide is an FDA-approved drug for treatment of rheumatoid arthritis (RA) and is in clinical trial development to treat multiple myeloma, solid tumors, acute GVHD, and COVID19. Based on its role in promoting lymphocyte apoptosis and its anti-inflammatory properties, we conducted a phase 1 trial (NCT04212416) to study leflunomide's safety in patients with cGVHD (primary objective). The 6-month anti-cGVHD activity, defined by overall response rate (ORR per NIH consensus criteria), was the secondary objective.Patients received a loading dose of leflunomide (100 mg X 3 days) followed by a daily maintenance dose of 20 mg, which is the approved dose in RA (Fig 1). The planned duration of therapy was six 28-day cycles, with responding patients being allowed to continue therapy. The CTCAE v5.0 was used to assess toxicity. Changes in cGVHD severity were assessed by physician reported cGVHD activity assessment form.The trial consisted of two (safety lead-in: n=6, expansion: n=12) with a planned dose de-escalation to 10 mg dose, if >1 patient experienced dose limiting toxicities (DLT) in the safety lead (first 28 days). The median age was 53 years (range 18-72). Most patients received PBSC graft (89%). Donors were matched unrelated (55%) or related (45%). Conditioning was MAC (50%) or RIC (50%). The most common GVHD prophylaxis was tacrolimus-sirolimus (56%) and prior aGVHD was reported in 60% patients. Median prior lines of therapy were 3 (range 2-7). Moderate and severe cGVHD were seen in 9 and 9 patients, respectively. Most common organ involvements were oral (77%), skin (61%), ocular (66%), joint/fascia (44%), lung/BOS (33%) and GI (11%).Adverse events (AE) were GI (diarrhea: 27%, nausea/vomiting: 5%, abdominal pain: 5%), cytopenias (anemia: 16%, leukopenia: 11%), hepatic failure (5%), lab abnormalities (AST/ALT increase: 11%, bilirubin increase: 5%), fatigue (11%), and headache (5%). The AEs (n=29) were mostly grade 1-2 (82%), and grade 3-4 AEs were seen in 18% including one DLT (sepsis). The best ORR at any time in the first 6 cycles was 55% [n=10; partial responses (PR)], 38% (n=6; stable disease) and 11% (n=2; progressed). The mean duration of therapy was 4.6 months for all patients with two patients currently on therapy. Ten patients completed 6 cycles of planned therapy, of whom 80% achieved a PR. Reasons for early discontinuation were response plateau (n=2), AEs (n=2), progressive cGVHD (n=1), and death (n=1; sepsis). There was a mean prednisone dose reduction of 41% from study start to end (9.4 vs 5.5 mg/day)In conclusion, leflunomide is safe in patients with cGVHD, and is associated with a promising ORR in patients with multisystem cGVHD involvement. Correlative immunologic and molecular studies are underway to compare the impact of leflunomide on lymphocytes between responders and non-responders.