A Phase II Study of Pembrolizumab in Combination with Romidepsin Demonstrates Durable Responses in Relapsed or Refractory T-Cell Lymphoma (TCL)

Background: There are few treatment options for relapsed/refractory (r/r) T cell lymphoma (TCL). We previously reported the safety and efficacy of a single center, single arm, trial of the check-point inhibitor (CPI) pembrolizumab in combination with the Histone deacetylase inhibitor (HDACi) romidepsin in r/r TCL pts who progressed after at least one prior systemic regimen (NCT03278782). The rationale for combining CPI with HDACi was that TCLs often carry multiple gene mutations in epigenetic modifier genes TET2, IDH2, DNMT3A and TCR related genes CD38, RHOA, and FYN that may impair immunogenicity and promote immune escape. Herein, we report the updated results of the Phase II study along with the correlative PDL1 evaluation.

METHODS: In this phase I/II trial, safety was evaluated in the lead-in phase I study using a dose de-escalation strategy, with Pembrolizumab at 200 mg on day 1 and Romidepsin 14 mg/m² on days 1 and 8, the eventual phase II study dosage. Adverse events were monitored and graded according to the NCI Common Terminology Criteria for Adverse Events version 4.0 guidelines. Primary objectives were to determine safety, tolerability, and overall response rate (ORR). Secondary objectives included complete remission rate (CRR), progression free survival (PFS), overall survival (OS) and duration of response (DOR). Outcome assessment were performed utilizing Lugano Revised Response Criteria and correlated with PD-L1 expression by immunohistochemistry and exploratory analysis for cell of origin.

RESULTS: 20 pts were evaluable for the primary endpoint (Table 1); the median age was 67 and a male predominance. The phase I component included 6 pts, with cohort expansion secondary to one DLT of hypotension and renal insufficiency. Prior therapies included induction chemotherapy with CHOP or CHOP like in 90%, Autologous Stem cell transplant (SCT) in 2 pts, radiation in 4 and salvage therapies in 10(Table 1). Ten pts had ≥3 prior therapies. The ongoing phase II enrolled 14 pts with 85% >60 years. In the phase II the ORR was 50% and the median follow-up was 18 months 5 patients achieved CR and 2 PR (AITL, PTCL with TfH and ALCL). In addition, there were 3 CRs (ALCL and PTCL with TfH) in the phase I portion. No responses were seen in the 3 MF pts. Of the 8 pts (Figure 1A) who achieved a CR, 4 have been off treatment for >1 year without relapse, 2 are still on therapy for >2 years, and 1 underwent a haplo-SCT (9 months after the last dose of pembrolizumab, without GVHD). The most common ≥ grade 3 adverse events were nausea, vomiting (n=1) and fatigue (n= 2). Two patients experienced hyperprogression within the first 10 days of treatment. The four patients who stopped therapy due to development of immune related adverse events (iRAE)- Grade 1 cytokine storm, Grade 3 gastritis, Grade 4 colitis and Grade 2 pneumonitis, respectively- all of which were reversible with steroid treatment. 13 pts were available for correlative studies of PD-L1 expression. The median PD-L1 expression in responders was 10 and ranged from < 5% (n=3) to 95% (n=1). PD-L1 in the 2 hyperprogressors were 0% and 10% respectively. The pts who achieved CR had a higher level of PD-L1 compared to those who achieved PR or SD (p= 0.048) (Figure 1B). Of the 2 ALCL pts with CR, 1 pt was ALK+ and 1 was ALK negative; the PD-L1 expression was 95% and 60% respectively.

CONCLUSIONS: The combination of Romidepsin and Pembrolizumab in patients with r/r TCL demonstrates an ORR of 50% which was durable and with acceptable safety in pts over the age of 60. Pts achieving CR included those with diagnosis of ALCL, AITL and PTCL with TfH phenotype and those expressing higher levels of PD-L1. Moreover, they also maintained a longer duration of response than with prior therapies. Additional studies evaluating the cell of origin expression will be reported at the Annual Meeting.