Immune Checkpoint Inhibitors for Advanced Gastroesophageal Cancers: Key Findings From ASCO GI 2019

Steven Maron, MD

Assistant Attending
Gastrointestinal Medical Oncology Service
Memorial Sloan Kettering Cancer Center
New York, New York

Steven Maron, MD, has disclosed that he has received funds for research support from Roche/Genentech.

View ClinicalThoughts from this Author

Released: February 26, 2019

Below, I highlight key findings presented at the 2019 Gastrointestinal Cancers Symposium with a focus on the role of immune checkpoint inhibitors (ICIs) in the treatment of gastroesophageal cancers.

Combining ICIs With Targeted Agents for HER2-Positive Gastric or GEJ Cancer
The potential for synergy between targeted antibodies and ICIs was one of the most interesting concepts at the 2019 conference. An antibody such as trastuzumab can attract and stimulate multiple immune cell types. Preclinical data suggest that the tumor responds by upregulating PD-L1, a checkpoint molecule intended to dampen the host immune response. We now have ICIs that block this inhibitory signal by binding directly to PD-1 or its ligand, PD-L1. Through this combination of targeted therapy and ICIs, patients may obtain an additive, if not synergistic, benefit. However, we have more to learn about the mechanisms underlying these responses. 

From the phase III ToGA study published nearly a decade ago, we know that the ORR is 47% in patients with HER2‑positive gastroesophageal cancer receiving first-line trastuzumab plus chemotherapy. New data from a phase II study by Janjigian and colleagues build upon this precedent, combining pembrolizumab with trastuzumab plus chemotherapy in the first-line setting. In this small study, the disease control rate was 100% and the ORR was 87%—notably up from the 47% observed in ToGA. After a median follow-up of only 6.6 months, the median PFS was 11.4 months, but OS data were not mature. This study also uniquely integrated a lead-in of pembrolizumab and trastuzumab without chemotherapy, which demonstrated a rapidity of response in these patients even prior to chemotherapy initiation.

Although HER2 positivity was an inclusion criterion for this phase II study, central confirmation revealed that 6 of 35 patients actually had HER2-negative cancer, highlighting the intrapatient molecular heterogeneity found in gastroesophageal cancer. These considerations in determining HER2 status will be important moving into the phase III KEYNOTE‑811 study, for which the investigators are recruiting a similar population to receive trastuzumab plus chemotherapy vs either pembrolizumab or placebo.

In patients with advanced HER2-positive gastric or GEJ cancer and with progressive disease after or resistance to trastuzumab plus chemotherapy, phase II data presented at this meeting by Catenacci and colleagues suggest that the chemotherapy-free combination of the HER2-targeted monoclonal antibody margetuximab plus pembrolizumab may be beneficial (ORR: 22%). This study allowed archival tissue testing for HER2, and a higher ORR was identified in patients with confirmed IHC 3+ HER2 expression and concurrent PD-L1 expression, which was more common in gastric vs GEJ cancer. By assessing circulating tumor DNA (ctDNA), the investigators found that 60% of patients retained ctDNA ERBB2 amplification upon disease progression while receiving trastuzumab.

Second-line ICI Monotherapy for Advanced Esophageal Cancer
The next trial to mention is the phase III KEYNOTE-181 trial, which compared single-agent pembrolizumab vs chemotherapy for second-line treatment of advanced esophageal cancer, including both squamous cell carcinoma and adenocarcinoma. In the intent-to-treat population, there was no difference in median OS with pembrolizumab vs chemotherapy (both 7.1 months; P = .0560). There was a nonsignificant trend toward improved survival with pembrolizumab in the subgroup with squamous cell carcinoma (8.2 vs 7.1 months with chemotherapy). However, among those with PD-L1 CPS ≥ 10, the median OS was significantly improved with pembrolizumab vs chemotherapy (9.3 vs 6.7 months, respectively; P = .0074). 

The question is, do we feel comfortable using this agent in the second-line setting for patients without MSI-high/MMR-deficient disease based on the limited data we have now? Recruitment of a second‑line trial in patients with squamous cell esophageal cancer is incredibly difficult. By the time patients with squamous cell histology progress to second‑line therapy, they are generally quite sick; consequently, chemotherapies with limited response rates and significant toxicities are difficult to justify. In my practice, I find the response observed in KEYNOTE-181 to be enough justification to try this approach in a patient with high PD-L1 expression—preferably as part of a clinical trial.

Future Directions
In the next couple of years, we will see the testing of numerous combinations of targeted agents and ICIs; many ongoing studies will be maturing. It will be important to carefully ascertain which signals warrant movement forward to phase II or III trials, either across tumor types or even within a tumor type, where possible. I hope to see compelling evidence that these combination approaches can convert “cold” tumors excluding T-cells to “hot” tumors.

Your Thoughts
Where do you foresee combinations with ICIs having the greatest impact in your practice? Join the conversation in the comments box or by answering the polling question.

Provided by Clinical Care Options, LLC

Contact Clinical Care Options

For customer support please email:

Mailing Address
Clinical Care Options, LLC
12001 Sunrise Valley Drive
Suite 300
Reston, VA 20191

Supported by an educational grant from
Bristol-Myers Squibb

Leaving the CCO site

You are now leaving the CCO site. The new destination site may have different terms of use and privacy policy.


Cookie Settings