Real-World Evidence for Duvelisib and Romidepsin Combination in Patients with Relapsed/Refractory Peripheral T-Cell Lymphomas

Introduction:

Novel treatments with lineage-specific activity and minimal side-effects are urgently needed for patients with relapsed/refractory (R/R) peripheral T-cell lymphomas (PTCL). A phase I study of duvelisib/romidepsin (duv/romi) demonstrated an overall response rate (ORR) of 58%, a complete response rate (CRR) of 42%, a median PFS of 6.9 m with low rates of Gr 3-4 transaminitis compared to duv single agent (SA; Horwitz SA, et al. Hematological Oncology 2021). Herein, we report on real-world evidence with duv/romi in patients with R/R PTCL in a multicenter cohort.

Methods:

Two US institutions contributed retrospective data from 19 patients with PTCL treated with duv/romi (duv 75 mg BID + romi 10 mg/m2 on days 1, 8, 15 of 28-day cycles) between 2021-2024. Duv dose was reduced to 25 mg BID from cycle 3 onwards. Seven patients received lead-in duv BID SA (median 3 cycles), and one patient received one cycle of lead-in romi SA prior to combination duv/romi. At the data cut-off (07/26/2024), the cohort received a median of three cycles (range, 1-19) of duv/romi. Patients received prophylaxis against VZV and PJP with active pre-emptive monitoring for cytomegalovirus reactivation.

Results:

Histological subtypes of the patients enrolled included 10 with nodal TFH lymphoma, 7 with PTCL-NOS, 1 with ENKTCL, and one with HSTCL. Ten (52.6%) patients identified as male, 14 (73.7%) as White, one (5.3%) as Black, three (15.8%) as Asian, one (5.3%) as Hispanic, and one (5.3%) patient with unknown race and ethnicity. The median age at diagnosis was 61.4 years (range, 30.2-88.8), median IPI at diagnosis was 2 (range, 1-4), median PIT at diagnosis was 1 (range, 0-3), and median ECOG at start of duv/romi was 0 (range, 0-2). Seven patients (36.8%) had primary refractory and 12 (63.2%) had relapsed lymphoma with a median of 1 (range, 1-8) regimen preceding duv/romi. The frontline therapies included anthracycline-based (n=15), azacitidine plus romidepsin (n=2), decitabine-cedazuridine (n=1) and R-ICE (n=1). Three patients had received prior autologous stem cell transplantation in first remission.

After a median follow-up of 7.0 m (range, 3.0-31.5), the ORR and CRR of the 17 evaluable patients were 58.8% (10/17) and 41.2% (7/17), respectively. CRR was 40% in both nTFH (4/10) and PTCL-NOS (2/5), whereas ORR was higher in nTFH at 70% (7/10) compared to PTCL-NOS at 40% (2/5). The median PFS, defined as time from lead-in or combination therapy initiation to radiologic progression, was 8.5 m (95%CI 3.2 – NR; range, 1.7-23.4). Median overall survival and duration of response were not reached. At the data cut off, most (8/10) patients who achieved response (PR or CR) to combination therapy had ongoing response and were actively receiving treatment. Two patients had no response to frontline azacitidine plus romidepsin but achieved PR (n=1) and CR (n=1) on duv/romi. Five of the seven patients with CRR were bridged to allogeneic SCT.

Treatment was well-tolerated in our real-world cohort. Notable Gr 3-4 adverse events included transaminitis (n=3), anemia (n=3), neutropenia (n=5), thrombocytopenia (n=2), and morbilliform rash (n=1) among 18 evaluable patients. Treatment was discontinued due to colitis (n=1) and disrupted due to neutropenia (n=1) and disseminated tuberculosis (n=1). Systematic and serial collection of specimens, including saliva, peripheral blood monocytes, plasma, cell-free DNA, and tumor tissues (pre- and post-treatment), has been performed to define molecular residual disease, biomarkers of response, and innate and acquired resistance.

Conclusion:

This real-world cohort confirms the highly favorable activity and tolerability of duv/romi as previously observed, supporting its clinical relevance in a challenging patient population with few options. We continue to follow current participants, enroll additional study participants, and bank and analyze serially collected specimens.