PD-1 Inhibitors Show Promise in Upper GI Cancers

Brandon G. Smaglo, MD


February 05, 2016

Immunotherapy Inroads Into Upper GI Cancers

Recent advances with immunotherapy in melanoma, lung, and kidney cancers have caused clinicians and patients to eagerly anticipate the results of immunotherapy studies for the treatment of additional cancers. Last year at the 2015 Gastrointestinal Cancers Symposium (GICS), Dr Kei Muro, on behalf of the KEYNOTE-012 investigators, presented results indicating that the PD-1 inhibitor pembrolizumab was promising for the management of advanced gastric cancers.[1] Only one year later, several studies presented at GICS 2016 provide early evidence that immunotherapies will play an important role in the management of gastroesophageal carcinomas.

The phase 2/2 CheckMate-032 study is investigating the role of dual immunotherapy in a variety of solid tumors. Patients are treated with the anti-PD-1 monoclonal antibody nivolumab and are randomized to additional treatment with or without the anti-CTLA4 antibody ipilimumab.[2] Dr Dung Le presented the results of 59 patients with gastric or gastroesophageal junction (GEJ) adenocarcinoma treated with nivolumab only. All patients had to have received at least one prior therapy, and the majority, 83%, had received at least two prior lines of treatment. This is remarkable because the typical patient with advanced esophageal cancer is often a poor candidate for third-line (or higher) therapies on the basis of their reduced performance status, which limits toxicity tolerance. There is also limited availability of third-line therapies to offer them. The side effects that patients experienced were generally tolerable, with only 14% of patients experiencing grade 3 or 4 adverse events. No patient deaths occurred. For this heavily pretreated population, a median overall survival of 6.8 months was achieved, and the 12-month overall survival was 38%.

Importantly, the CheckMate-032 trial also considered the PD-L1 status of tumors. Unlike other antibody therapies, it is not yet clear that tumor expression of PD-L1 is requisite for response. Of the tumors considered in this cohort, 39% of tumor samples were found to express PD-L1. The ORRs of patients with PD-L1-positive tumors vs PD-L1-negative tumors were 18% and 12%, respectively. This suggests that PD-L1 positivity is not a necessary expression for effect of this class of therapy. This biomarker's significance as these immunotherapies are developed seems to differ from the standard black-or-white thinking that surrounds marker expression associated with other such targets.

Dr Toshihiko Doi presented results from the KEYNOTE-028 study of patients with advanced esophageal or GEJ carcinomas treated with pembrolizumab; this included both squamous cell (73.9% of patients) and adenocarcinoma (26.1%) histologies.[3] Importantly, as was the case with the patients presented form the CheckMate-032 trial, the majority of the KEYNOTE-028 patients, 87%, had already received at least two prior systemic therapies for their advanced disease.

In KEYNOTE-028, therapy was well tolerated; only 4 of 23 patients experienced adverse events of grade 3, and no patients died or had to discontinue therapy because of drug toxicity. The PFS rates at 6 and 12 months were 30.4% and 21.7%, respectively, in this heavily pretreated population. Although these results are obviously exciting, total enrollment numbers in this phase 1b trial are low. Consequently, phase 2 and 3 data of patients with advanced esophageal cancers will be much anticipated.

The results of both of these phase 1 studies suggest that immune checkpoint inhibitors are likely to have a role in the management of gastroesophageal carcinomas. However, this benefit needs to be further evaluated and confirmed by larger studies. The combined cohorts of KEYNOTE-28 and CheckMate-032 number only 82 patients, and they include a heterogeneous collection of tumor locations and histologies. For now, use of these therapies for the management of gastroesophageal cancers should remain restricted to the clinical trial setting.

Ramucirumab Needs Some Company in Upper GI Cancers

Regardless of their benefit, immunotherapies are unlikely to provide universal activity against gastroesophageal cancers, and the use of chemotherapies and other biologic therapies remains an important component of the treatment of these cancers.

One critical question for the management of potentially curable esophageal cancer concerns the optimal chemotherapy regimen, to be used as a component of neoadjuvant chemoradiation therapy. The NEOSCOPE trial from the United Kingdom attempted to answer this question by comparing capecitabine/oxaliplatin and carboplatin/paclitaxel in this setting.[4] In both arms of NEOSCOPE, induction chemotherapy was administered prior to initiation of the concurrent therapy. Both approaches were well tolerated, and rates of pathologic complete responses, particularly in the carboplatin/paclitaxel arm, were encouraging. However, as is often the case, total numbers are too low to make a definitive statement. Moreover, because patients respond to both therapies, the larger question may be how to select the regimen that will be more active for an individual's tumor, essentially pairing right drug/right tumor/right patient rather than using some overarching standard.

In a trial entitled MEGA, Dr Peter Enzinger reported on the benefits of adding the anti-VEGF monoclonal antibody ziv-aflibercept to FOLFOX for first-line treatment of advanced gastroesophageal adenocarcinoma.[5] There are currently no biologic therapies available in the treatment of all patients with gastroesophageal cancer in the first-line setting. In general, there is a relative dearth of biologic therapies for these cancers; only 10%-20% of patients with gastroesophageal cancer have HER2-positive tumors and are thus candidates for treatment with the biologic trastuzumab.[6] Consequently, there was interest in seeing whether ziv-aflibercept would prove active in this setting. Unfortunately, the addition of ziv-aflibercept did not significantly improve the benefits of FOLFOX. These results leave ramucirumab as the only drug in class that has an established effect for the treatment of gastroesophageal adenocarcinomas, and then only in the second line of therapy.[7,8] Whether ramucirumab will develop a role in first-line therapy remains to be seen.

Another study looking at the role of biologic therapies in advanced gastric cancer was GATSBY. This trial evaluated the benefit of second-line HER2-targeting therapy—using the drug-antibody conjugate trastuzumab emtansine—in patients with HER2-overexpressing gastroesophageal adenocarcinoma.[9] Patients were randomized to one of three treatment arms: two arms used trastuzumab emtansine as monotherapy, either 3.6 mg/kg dosing on a 21-day cycle or 2.4 mg/kg dosing on a 7-day cycle; the third arm, which served as control, used taxane monotherapy. Recommendations from an independent data monitoring committee selected the weekly dosing for trastuzumab emtansine for further data collection, and thus the results presented considered this dose schedule as it compared to the taxane treatment. Unfortunately, trastuzumab emtansine did not demonstrate better efficacy over the chemotherapy used. Thus, no second-line HER2-directed therapies are available to our patients with HER2-overexpressing gastroesophageal adenocarcinomas, leaving this an area for further trial development and investigation.

The excellent work from these studies presented at the 2016 GICS is slowly moving the bar for the management of these aggressive cancers. Clearly, the greatest promise stems from those studies of immunotherapy, and further results of studies of immune checkpoint inhibitors in the management of gastroesophageal carcinomas are highly anticipated. It is certain that chemotherapy and biologic therapy will retain an important role in the management of these cancers. Understanding the optimal way to employ these therapies, as well as the development of new targets, remain important research goals for the management of these cancers.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.