Commenting on the programmed cell death protein 1 (PD-1)/PD-1 ligand (PD-L1) antibodies in tumors with mismatch repair deficiency, Bertram Wiedenmann, MD, PhD, Professor of Internal Medicine and Gastroenterology, Universitätsmedizin Berlin, Germany, suggested, “The efficacy of pembrolizumab (Keytruda) in patients with noncolorectal tumors showing mismatch repair deficiency is impressive but needs to be validated in larger cohorts of patients with various gastroenteropancreatic primary tumors.”

He claimed that the heterogeneity of the patient population, the lack of a control group, and the short follow-up were weaknesses of the study, although he emphasized the strong rationale for studying checkpoint inhibitors in patients with mismatch repair deficiency.

Invited discussant Markus Moehler, MD, PhD, Professor of Gastrointestinal Oncology, Johannes-Gutenberg University in Mainz, Germany, shared his thoughts on the studies of anti–PD-1/PD-L1 agents in esophagogastric cancers. Since these tumors tend to be immunogenic, he said, “Immunotherapy is an important effort to address unmet needs for advanced disease.”

CheckMate-032 vs KEYNOTE-028

Although the two agents were both tested primarily in the third line and had similar toxicity profiles (17% grade 3/4 adverse events), there were differences between the CheckMate-032 study of nivolumab (Opdivo) and the KEYNOTE-028 trial of pembrolizumab that make comparisons difficult:

The populations of the two key trials were somewhat different.

PD-L1 positivity was not mandated in CheckMate-032 but was required for inclusion in KEYNOTE-028.

PD-L1 positivity with nivolumab was based on membranous expression only in tumor cells, not immune cells; for pembrolizumab, both tumor and stroma bands were stained.

KEYNOTE-028, but not CheckMate-032, included Asian patients, who may have a different immune signature than whites.

The requirement of PD-L1 positivity may partly explain the higher response rate seen with pembrolizumab (30%) vs nivolumab (14%), he said, but noted that Checkmate-032 did observe an association between PD-L1 expression and response. Due to small numbers of patients so far, he said, “it’s too early to say whether PD-L1 expression is clearly associated with better responses.”

In developing PD-L1 as a biomarker, Dr. Moehler emphasized the need to clarify the different antibodies used in the trials and cut points used for determining PD-L1 positivity, evaluate regional and time-dependent tumor heterogeneity, and assess the stability of the PD-L1 proteins in tissue blocks and slides.

He also commented on the emerging survival data with these drugs, noting that survival time appears to be somewhat in line with that seen with other new targeted agents in second- or third-line of metastatic esophagogastric cancer. The median overall survival of 5.0 months with nivolumab in esophagogastric cancers is approximately the same as that seen with ramucirumab (Cyramza) in randomized phase III and regorafenib (Stivarga) in randomized phase II trials. Encouragingly, he said, “one-third of a highly pretreated population was alive at 1 year” in CheckMate-032, suggesting that quite a good number of patients benefit with prolonged survival, particularly at the end of the Kaplan-Meier curves by treatment with checkpoint inhibition, which seems to be better than with classical targeted agents or chemotherapy.

Step in the Right Direction

Dr. Moehler was most impressed, however, with the potential of a gene-expression profile, as presented by Dr. Doi within the KEYNOTE trial, that might predict for benefit and suggested that it be developed for all tumor types in which these drugs are active . He also suggested the gene signature be applied to the four main gastric cancer subtypes, as published by The Cancer Genome Atlas Research Network.¹ “It’s clear some of these subtypes express higher PD-L1 levels than others,” he pointed out.

“Molecular and immunologic characterization of the patients is key, and with the gene signature in the pembrolizumab study, we have now seen the first step in the right direction,” Dr. Moehler said. ■

Disclosure: Dr. Wiedenmann reported no potential conflicts of interest. Dr. Moehler has received honoraria from Amgen, Lilly/ImClone, Merck Serono, Roche/Genentech, and Taiho Pharmaceutical.

Reference

1. The Cancer Genome Atlas Research Network: Comprehensive molecular characterization of gastric adenocarcinoma. Nature 513:202-209, 2014.