Final Response and Survival Results from a Phase II Trial of Pembrolizumab and Entinostat in Relapsed/Refractory Hodgkin Lymphoma

Introduction: Targeting PD-1 is a highly effective strategy in HL and is currently being incorporated into upfront regimens. Strategies for relapsed or refractory (R/R) disease remain an unmet need, especially in those with prior anti-PD1 exposure. Among the potential mechanisms of resistance to anti-PD-1 therapy is downregulation of antigen presentation machinery, including major histocompatibility complex (MHC) I and II, both of which occur at high rates in HL. We hypothesized that histone deacetylase (HDAC) inhibition with the oral class I HDAC inhibitor, entinostat, would restore anti-PD-1 sensitivity, in part by restoring MHC-I and MHC-II expression on Reed-Sternberg cells. Therefore, we performed a phase II trial of entinostat plus pembrolizumab in R/R HL. Herein are the final results from the primary analysis. Methods: Patients with R/R classical HL after ≥ 2 systemic therapies were eligible. Prior therapy with an HDAC inhibitor and/or anti-PD1 therapy was allowed. Treatment consisted of pembrolizumab 200 mg every 21 days plus entinostat 5-7 mg on days 1, 8 and 15 of each 21-day cycle. Treatment was continued until progression of disease (POD), unacceptable toxicity, or death, for a maximum of 35 cycles (two years). If one of the study drugs was discontinued, the other could be continued. The primary endpoint was 12-month progression-free survival (PFS). PFS was measured from treatment initiation to POD or death. For patients who stopped treatment due to an adverse event (AE), progression was dated at the date of subsequent POD or initiation of alternative therapy, whichever occurred first. To ensure conservative estimates of pembrolizumab-entinostat efficacy, consolidation with transplant, radiation, or pembrolizumab maintenance (at the completion of 35 cycles) were considered events. The null hypothesis was a 12-month PFS of 40% versus a 12-month PFS of 60%. Key secondary endpoints included response rates per Lugano criteria and safety assessment. Results: Thirty-nine patients enrolled. The median number of prior therapies was 5 (range: 2-18). Prior therapies included brentuximab vedotin (82%), anti-PD1 (74%), HDAC inhibitor (10%), and/or autoHCT (67%). Twenty-two patients (56%) had prior POD to anti-PD1, including 16 (41%) with POD to anti-PD1 as the last line of therapy prior to enrollment. One patient was inevaluable due to not completing at least one cycle of therapy. Among 38 evaluable patients, 11 had POD, 10 withdrew due to AE, 6 completed the full 35 cycles (with 4 continuing on pembrolizumab off study), 5 were consolidated with transplant/radiation, and 4 remain on treatment without POD. Due to the COVID-19 pandemic, 2 patients withdrew consent. Notably, 4 patients were treated past POD due to clinical benefit. Among the 38 evaluable patients, the complete and objective response rates (CRR/ORR) were 47% and 63%, respectively. Stratifying patients by prior exposure (exposed/naïve) and response (sensitive/resistant) to anti-PD1, CRR/ORR was as follows: (1) anti-PD1 exposed: 36% (10/28)/50% (14/28); (2) anti-PD1 naïve: 80% (8/10)/100% (10/10); (3) anti-PD1 sensitive: 40% (2/5)/40% (2/5); (4) anti-PD1 resistant: 36% (8/22)/55% (12/22), (5) anti-PD1 resistant as last line of therapy: 31% (5/16)/44% (7/16). With a median followup of 33.5 months (range: 3.8-84 months), the 12-month PFS was 52% (95% CI 38-71). The median PFS was 12.1 months (95% CI 7.3-24.5). Adverse events (AE) of ≥ grade 3 occurred in 30 (77%) patients. The most common AEs of ≥ grade 3 were neutropenia (44%) and thrombocytopenia (28%). Primary growth factor prophylaxis was not mandated, but patients responded well to growth factor when administered. Pericardial events (effusion, pericarditis) led to discontinuations in 2 patients (at 2 and 6 months), as did pneumonitis in 2 patients (at 3 and 9 months). Conclusions: As checkpoint blockade is increasingly used in earlier lines of therapy in HL, strategies to recapture immune responses in R/R disease are needed. The combination of pembrolizumab and entinostat showed high response rates and encouraging PFS in R/R HL. Toxicities were primarily hematologic, with two notable pericardial events which led to drug discontinuation. Overall, HDAC inhibition in combination with anti-PD1 blockade is effective in HL, even in patients with prior anti-PD1 exposure and/or resistance.