Immunotherapy a Likely New Standard for GI Cancers

The prognosis for advanced or metastatic upper gastrointestinal (GI) malignancies is poor, with a median overall survival (OS) of less than 1 year. Treatment advances have stagnated for decades, and platinum-based chemotherapy has been the standard treatment for patients newly diagnosed with HER2-negative gastric cancer since the 1990s. However, the results of four phase III trials evaluating the use of immune checkpoint inhibitors presented at the European Society for Medical Oncology Virtual Conference 2020 in September may upend the current paradigm, even though the value of a few of the factors considered should be validated.

The randomized, open-label CheckMate 649 trial tested the PD-1 inhibitor nivolumab (Opdivo; Bristol Myers Squibb) plus chemotherapy versus chemotherapy alone. Researchers enrolled 1,581 patients with treatment-naïve, inoperable advanced or metastatic gastric cancer, gastroesophageal junction cancer, or esophageal adenocarcinoma. Patients with known HER2-positive disease were excluded from the study.

Markus Moehler, MD, PhD, of Johannes Gutenberg Universität Mainz in Germany, reported that patients who received nivolumab plus chemotherapy had a median OS of 13.8 months, whereas those who received chemotherapy alone had a median OS of 11.6 months. The benefit was greater for patients with a PD-L1 combined positive score (CPS) of 5 or higher: 14.4 months versus 11.1 months, respectively.

In apparent contradiction were the results of ATTRACTION-4, another randomized study of nivolumab plus chemotherapy versus chemotherapy alone in 724 patients with inoperable advanced or recurrent HER2-negative, chemotherapy-naïve gastric or gastroesophageal cancer. ATTRACTION-4 did not show an OS benefit with the addition of a PD-1 inhibitor, according to Narikazu Boku, MD, PhD, of the National Cancer Center Hospital in Tokyo, Japan, who presented the findings. Median OS was nearly indistinguishable between the groups—17.5 months with nivolumab plus chemotherapy and 17.2 months with chemotherapy alone. Median progression-free survival was slightly higher (10.5 months versus 8.4 months) in the patients who received nivolumab, but whether that difference affects quality of life or other key factors remains unclear.

One reason for the disparate results between CheckMate 649 and ATTRACTION-4 may be that patients enrolled in the latter study were much more likely to receive second-line treatment after the trial, muddying OS data. This is likely due to the different study populations: Whereas CheckMate 649 included patients in western and Asian countries, ATTRACTION-4 enrolled only patients in Asia. Genetic differences may have contributed, but perhaps more important is that patients in Asian countries tend to present with earlier-stage disease and thus have more treatment options, possibly including second-line immunotherapy, said Josep Tabernero, MD, PhD, of Vall d'Hebron Institute of Oncology in Barcelona, Spain. In fact, 38% of patients had second-line therapy in CheckMate 649, and 8% of patients in the control arm received immunotherapy after the trial, versus 66% and 27%, respectively, in ATTRACTION-4.

The divergent results could also be due to the biomarker-based design of CheckMate 649, in which 82% of patients had tumors with a PD-L1 CPS of 1 or greater, 74% of whom had tumor CPS of 5 or greater. In contrast, ATTRACTION-4 was not designed to assess efficacy in patients based on PD-L1 CPS value, nor was it enriched for patients with high PD-L1 CPS—only 16% of patients had tumors with a PD-L1 CPS of 1 or greater.

The potential predictive ability of PD-L1 CPS was highlighted in KEYNOTE-590. This randomized trial compared the PD-1 inhibitor pembrolizumab (Keytruda; Merck) plus chemotherapy with chemotherapy alone in 749 patients with locally advanced, chemotherapy-naïve, inoperable or metastatic esophageal adenocarcinoma or esophageal squamous-cell carcinoma, or advanced or metastatic esophagogastric junction adenocarcinoma. For all patients who received pembrolizumab, the median OS was 12.4 months, compared with 9.8 months for those who received only chemotherapy, reported Ken Kato, MD, PhD, also of Japan's National Cancer Center Hospital.

Strikingly, the immunotherapy–chemotherapy combination showed even greater benefit in the 51% of patients whose tumors had a PD-L1 CPS of 10 or greater: The median OS was 13.5 months with pembrolizumab plus chemotherapy, compared with only 9.4 months in patients who received placebo plus chemotherapy.

Despite the lack of OS benefit observed in ATTRACTION-4, the results of CheckMate 649 and KEYNOTE-590 will change practice, said Tabernero. In addition, the studies will likely lead to approval of nivolumab and pembrolizumab to treat upper GI cancers in the United States and Europe, as well as some Asian countries, because the latter two trials enrolled enough patients from those countries to be considered by local regulatory authorities.

Equally valuable, Tabernero said, will be the results of the CheckMate 577 trial, which was also presented at the conference. Researchers enrolled 794 patients with esophageal or gastroesophageal junction cancer who were treated with either adjuvant nivolumab or placebo following neoadjuvant chemoradiation therapy and surgery. Researcher Ronan Kelly, MD, MBA, of Baylor University Medical Center in Dallas, TX, noted that disease-free survival (DFS) was 22.4 months with nivolumab versus 11 months with placebo, but unlike in KEYNOTE-590, a higher PD-L1 CPS was not predictive of response. OS data are not yet mature, however, and researchers don't know whether longer DFS will correlate with increased OS.

Although the value of PD-L1 CPS and the utility of DFS as an endpoint in these diseases still require validation, the apparently manageable adverse-event profile with nivolumab and pembrolizumab combined with the efficacy data may mean that the standard of care for upper GI malignancies, static for so long, will finally be amended. –Nicole Haloupek

For more news on cancer research, visit Cancer Discovery online at http://cancerdiscovery.aacrjournals.org/CDNews.