PD-1/HDAC Inhibitor Combo Active in Recurrent Melanoma After Progression on Anti-PD-1 Therapy


Entinostat in combination with pembrolizumab induced responses in patients with recurrent or metastatic melanoma who have progressed on or after prior anti–PD-1 therapy, according to results of a phase II trial.

Ryan J. Sullivan, MD

Ryan J. Sullivan, MD

Entinostat in combination with pembrolizumab (Keytruda) induced responses in patients with recurrent or metastatic melanoma who have progressed on or after prior anti—PD-1 therapy, according to results of a phase II trial.1

In findings from the melanoma cohort of the ENCORE-601 trial (NCT02437136) presented at the 2019 AACR Annual Meeting, the PD-1/HDAC inhibitor combination regimen demonstrated an objective response rate (ORR) of 19% (95% CI, 9%-32%) and a clinical benefit rate (CBR) of 36% (95% CI, 23%-50%) in patients with recurrent melanoma.

“In patients with progressing melanoma on prior PD-1 blockade including prior PD-1 and CTLA-4 blockade—clearly a group with limited treatment options—entinostat and pembrolizumab demonstrated significant antitumor activity,” Ryan J. Sullivan, MD, said when presenting the findings. “The combination of entinostat plus pembrolizumab was clearly safe and tolerable.”

Sullivan, an assistant professor of hematology/oncology at Massachusetts General Hospital and an assistant professor of medicine at Harvard Medical School, explained how a growing minority of patients with melanoma have been cured with newer immunotherapy options, such as PD-1 inhibitors, yet most patients still develop resistance and die from metastatic melanoma.

“With a better understanding of the mechanisms of resistance, we may be able to develop more effective therapies,” he said, as well as combination regimens. “Though the frontline setting is a critical place for us to be in terms of clinical trials, I think the most relevant and most unmet need now is what do we do with patients who have received anti—PD-1 antibody therapy and need something else.”

Entinostat is an oral class I-selective histone deacetylase (HDAC) inhibitor that leads to the downregulation of immunosuppressive cell types in the tumor microenvironment. In preclinical models, the agent demonstrated synergy with PD-1 inhibitors.

A phase I study of the combination of pembrolizumab and entinostat in patients with pretreated advanced solid tumors that was presented at last year’s AACR Annual Meeting demonstrated promising antitumor activity and acceptable safety. Sullivan also noted that after a single dose of entinostat, patients showed decreased myeloid-derived suppressor cell (MDSC) frequency at day 14.2

ENCORE-601 is a phase Ib/II trial exploring entinostat in combination with pembrolizumab, which consisted of a dose-escalation and safety confirmation phase followed by a dose-expansion phase. The expansion phase included 4 cohorts: 1) patients with non—small cell lung cancer (NSCLC) who have progressed on anti–PD-1/L1 therapy, 2) patients with NSCLC who are anti–PD-1 naïve, 3) patients with melanoma who have progressed on anti–PD-1, and 4) patients with mismatch repair–proficient colorectal cancer who are anti–PD-1 naïve.

In the melanoma cohort, the focus of Sullivan’s presentation at the 2019 AACR meeting, patients must have had recurrent or metastatic melanoma, prior progression on or after anti—PD-1/L1 therapy, prior BRAF-targeted therapy if indicated, and an ECOG performance status of less than 2.

The cohort enrolled 53 patients as of April 2018 and the participants were administered 5 mg of oral entinostat weekly and 200 mg of intravenous pembrolizumab every 3 weeks.

The primary endpoint for the cohort was ORR by immune-related RECIST (irRECIST) criteria, and secondary endpoints included CBR, progression-free survival (PFS), overall survival (OS), and safety and tolerability.

Median patient age was 62.0 years (range, 20-86), 60% of patients were male, 89% were white, and 55% patients had an ECOG performance status of 0. PD-L1 expression was positive in 53%, negative in 21%, and not evaluable in 19%. Visceral metastases were noted in 64% of patients and 36% had a baseline LDH level above the upper limit of normal.

Prior treatment consisted of a combination of anti—PD-1 and ipilimumab (Yervoy) in 70% and 23% had received prior BRAF/MEK combination treatment. The best recorded response to prior anti–PD-1 therapy was a complete response (CR) in 1 patient, partial response (PR) in 6, and 20 patients had achieved stable disease (SD). Twenty-two patients had progression while on treatment and the response was unknown in 4 patients.

The median duration of prior anti—PD-1 treatment was 5.2 months (range, 0.72-23.1) and the median time between last anti–PD-1 and first dose on this trial was 2.6 months (range, 0.66-37.1).

After treatment with the pembrolizumab and entinostat combination, 1 patient achieved a CR and 9 had PRs. The median duration of response was 13 months (range, 3-20), 4 of which are ongoing. Additionally, 9 patients had SD lasting more than 6 months.

After a median follow-up of 10.6 months, the median PFS was 4.2 months (95% CI, 2.7-7.0) and OS findings were not presented.

“Importantly, at the 1-year mark, 10 patients remained either on therapy or in response,” Sullivan noted.

He pointed out that responses to the combination were seen in patients regardless of the time since the patient last received PD-1 inhibition: “There are certainly some patients who received their PD-1 inhibitor long before being enrolled onto this study and ultimately responded, but there are also a number of these patients who received PD-1 inhibitor right before going onto the study and were defined as progressors. So clearly, it seemed to be helping both those who remotely were treated with PD-1 inhibitors and those who were recently treated with PD-1 inhibitors.”

Common (≥15%) grade 1/2 treatment-related adverse events (TRAEs) included nausea (47.2%); fatigue (32.0%); diarrhea (18.9%); pruritus (17.0%); and neutrophil count decrease, platelet count decrease, and arthralgia (15.1% each). The most common (≥5%) grade 3/4 TRAEs were neutropenia (7.5%), hyponatremia and fatigue (5.7% each).

Grade 3/4 immune-related adverse events (irAEs) were observed in 5 patients, including rash in 2 and colitis, pneumonitis, and autoimmune hepatitis in 1 patient each.

“The dominant toxicity was additive toxicity from entinostat and [there were] no apparent increase in irAEs with the combination,” Sullivan said.

Six patients discontinued treatment due to TRAEs of increased bilirubin, mucosal inflammation, neutropenia, pneumonitis, constipation, and autoimmune hepatitis. However, Sullivan commented that the patient with hepatitis had received a short course of therapy before developing the TRAE, but the patient achieved a response that was maintained off of therapy for over a year.

Biomarker analysis of 49 patients with peripheral blood available for analysis showed that there was no significant change in the T cell population in either responders or non-responders throughout treatment with the combination regimen, but there was a significant reduction in monocytic MDSCs from treatment in responders and a slight reduction in non-responders as well. The reductions in monocytic MDSCs in the responders were maintained throughout treatment for all but 1 patient.

The investigators also compared genetic signatures by RNA sequencing of 4 responders and 4 non-responders pre- and post-treatment. They found that in the responders versus the non-responders, the responders showed an increased inflammatory response as well as more hypoxia, more TNF-alpha signaling, less E2F targets, and less oxidative phosphorylation. In just the 4 responders post-treatment compared with pre-treatment, the investigators found that post-treatment these patients demonstrated more enrichment of immune gene signatures such as IL-6 and JAK-STAT3 signaling, and less MYC and E2F targets. Sullivan said that these findings were potentially interesting and hypothesis generating.

Additionally, post-treatment the responders demonstrated significant upregulation of several immune-related genes, such as CCL21 and CXCL9, and immunoglobulin markers, like IL15RA. The only gene that was significantly downregulated was CXCR1, which has been associated with melanoma progression.

“The preliminary biomarker analysis, and it’s very preliminary on a small number of patients, demonstrates findings consistent with the mechanism of action of entinostat including reduction in circulating myeloid-derived suppressor cells as well as tumor-specific increases in the pre- and on-treatment and enriched in responders versus non-responders’ inflammatory pathways,” Sullivan concluded.


  1. Sullivan RJ, Moschos SJ, Johnson ML, et al. Efficacy and safety of entinostat (ENT) and pembrolizumab (PEMBRO) in patients with melanoma previously treated with anti-PD-1 therapy. Presented at: 2019 AACR Annual Meeting; March 29-April 3, 2019; Atlanta, GA. Abstract CT072.
  2. Tolcher AW, Meyers ML, Gabrilovich D, et al. Safety, efficacy, and immune correlates of alternative doses and schedules of entinostat combined with pembrolizumab in patients with advanced solid tumors - results from SNDX-275-0141 phase I trial. Presented at: 2018 AACR Annual Meeting; April 14-18, 2018; Chicago, IL. Abstract CT179.
Related Videos
Expert on melanoma
Expert on melanoma
Related Content