Immune Checkpoint Inhibitors in Advanced Gastroesophageal Cancers

Immune Checkpoint Inhibitors in Advanced Gastroesophageal Cancers


Dr. Mariela Blum Murphy

Dr. Jaffer A. Ajani

By Mariela Blum Murphy, MD, and Jaffer A. Ajani, MD

Article Highlights

  • After aggressive treatment, one-third of patients with locally advanced gastroesophageal cancers will develop metastatic disease with a 5-year survival rate of less than 4%.
  • Among targeted agents, only trastuzumab and ramucirumab have yielded a statistically significant survival advantage, and results from other targeted agents have been discouraging.
  • The merging of immunotherapy with chemotherapy, radiation therapy, and targeted agents opens new possibilities for the treatment of gastroesophageal cancers and other malignancies.

Cure is not possible for most patients with locally advanced gastroesophageal cancers, and much progress remains to be made. Current strategies for localized disease include perioperative chemotherapy and surgery1 or surgery followed by adjunctive therapy with either chemotherapy2 or chemoradiation.3,4 However, after aggressive treatment, one-third of patients with locally advanced gastroesophageal cancers will develop metastatic disease. In patients with advanced disease, the median overall survival is approximately 9 months, and the 5-year survival rate is less than 4%.5 Such patients are empirically treated with a combination of a platinum agent and a fluoropyrimidine,6 with the addition of trastuzumab if the tumor expresses HER2/neu biomarkers.7

After disease progression, options for second-line treatment include taxanes,8 ramucirumab (monotherapy or in combination),9,10 or single-agent irinotecan, which provides a modest survival advantage versus best supportive care.8,11 Among targeted agents, only trastuzumab7 and ramucirumab10 have yielded a statistically significant survival advantage in a few patients (ramucirumab resulted in a marginal advantage as a single agent, and the addition of trastuzumab led to only a modest advantage12), and results from other targeted agents have been discouraging.13,14

Harnessing the Immune System

Fortunately, the paradigm is shifting to harness the power of the immune system. Under ordinary circumstances, most tumors evade the immune system, which is quite equipped to kill tumor cells. Thus, the tumor can thrive in the presence of a competent immune system. Uncovering the mechanisms of T-cell activation and its costimulatory signals15 has provided the opportunity for new treatment avenues. The immune response is regulated by a fine balance between costimulatory and inhibitory signals. For example, co-inhibitory molecules, such as CTLA-4, will inhibit T-cell signaling.16

Likewise, PD-1 will bind to PD-L1 and PD-L2, causing T-cell exhaustion and diminishing T-cell response.17,18 PD-L1 is up-regulated in solid tumors, where it can inhibit cytokine production and the cytolytic activity of PD-1–positive, tumor-infiltrating CD4- and CD8-positive T cells, allowing evasion of the immune response.19 Targeting CTLA-4 and PD-1 or PD-L1 with monoclonal antibodies has produced excellent results in the treatment of melanoma and other solid tumors.20,21 However, the checkpoint inhibition (e.g., PD-1, PD-L1, and CTLA-4) that results in non-specific stimulation of the immune system has shown promising results in the treatment of several solids tumors,22,23 and more recently, in gastroesophageal cancers.24,25

CTLA-4 Inhibition

CTLA-4 inhibition as monotherapy has thus far rendered disappointing results in gastroesophageal adenocarcinoma. Ipilimumab, a fully humanized monoclonal antibody against CTLA-4 was evaluated as sequential or maintenance therapy versus best supportive care following first-line chemotherapy in a multicenter, randomized phase II trial of patients with unresectable, locally advanced, or metastatic gastric cancer (GC) or gastroesophageal junction (GEJ) cancer. The immune-related progression-free survival (PFS) with ipilimumab was not better than best supportive care. The median overall survival (OS) for both arms was approximately 1 year.26

Another CTLA-4 inhibitor, tremelimumab, also a human monoclonal antibody, was investigated in a phase II trial as a second-line treatment for patients with advanced GC and GEJ adenocarcinoma. Eighteen patients received tremelimumab until disease progression. Only one patient achieved a durable partial response. Median time to progression was almost 3 months, and median OS was 4.8 months (95% CI [4.06, 5.59]).27 Based on these studies, further investigation of CTLA-4 inhibition as a monotherapy for gastroesophageal cancer appears unlikely to be fruitful.

PD-L1 Inhibition

Other immunotherapy targets appear promising, including PD-L1, although results for its use as a target in GEJ cancer and GC remain inconclusive. Several agents have already made it to clinical trials, including BMS-936559, a fully human PD-L1–specific IgG4 monoclonal antibody, which was one of the first PD-L1 antibodies used in a phase I study for advanced solid tumors. However, despite the fact that patients with GC were enrolled in this trial, patients with GC were not included in the efficacy population.28

Additionally, durvalumab, a monoclonal antibody directed against PD-L1, is currently under study as both monotherapy and in combination with other agents. Preliminary data from a multi-arm phase I expansion study suggest that the safety of durvalumab is acceptable, even in patients who have been heavily pretreated. In the GEJ/GC cohort (16 patients), clinical activity was seen in 25% of the patients.29

Another PD-L1 inhibitor, atezolizumab, an engineered human monoclonal antibody of IgG1 isotype, was studied in patients with locally advanced or metastatic solid tumors. Unfortunately, only one patient with GC was included in the study; however, this patient achieved partial response with durable results.30

Avelumab, a human IgG1 anti–PD-L1 antibody, has been mainly studied in the Japanese population. A phase 1b study assessed safety and tolerability of avelumab, as well as overall response and PFS in patients with advanced GC or GEJ cancer. Among 55 patients who received avelumab maintenance after first-line chemotherapy in the phase 1b study, one had a complete response and three others had partial responses, for an overall response rate (ORR) of 7.3% with median PFS of 14.1 weeks. In the second-line setting, avelumab demonstrated an ORR of 15% with a median PFS of 11.6 weeks. Stable disease rates were 35.0% in the second-line setting and 47.3% in the maintenance cohort.31 Based on these results, phase III trials with avelumab are ongoing in the first- and third-line setting for GC (JAVELIN Gastric 100 [NCT02625610] and JAVELIN Gastric 300 [NCT02625623]).

To date, the number of patients with GC included in these studies is far too small to draw significant conclusions about use of PD-L1 as a target in GEJ cancer/GC treatment. Further study results and more mature data are needed with PD-L1 inhibitors to establish its role in gastroesophageal tumors.

PD-1 Inhibition

Currently, the most promising data on the efficacy of immunotherapy in GEJ cancer/GC comes from PD-1 inhibition, with preliminary results just becoming available. Nivolumab is a human anti–PD-1 IgG4 monoclonal antibody with a favorable safety profile and efficacy, and it is currently approved by the U.S. Food and Drug Administration for melanoma, non–small cell lung cancer, renal cell carcinoma, and classical Hodgkin lymphoma. The phase I/II, open-label CheckMate 032 study evaluated nivolumab with and without ipilimumab in patients with solid tumors such as triple-negative breast cancer, GC, bladder cancer, and others (NCT01928394). The patients with advanced or metastatic gastric and GEJ cancer, irrespective of PD-L1 status, were assigned to receive nivolumab alone (3 mg/kg IV once every 2 weeks) and treated until disease progression or intolerable toxicity. With a total of 59 patients, the ORR was 12% (seven out of 58 patients; one complete response, six partial responses), and 12 patients (21%) had stable disease. Among responders, median duration of response was 7.1 months (95% CI [3.0, 13.2]). Median OS was 6.8 months (95% CI [3.3, 12.4]); the 12-month OS rate was 38% (95% CI [23.2, 52.7]).25

Similarly, pembrolizumab, a selective, humanized, high-affinity IgG4k monoclonal antibody designed to bind to PD-1, was studied in a multicenter, open-label phase 1b trial (KEYNOTE-012) in patients with PD-L1–positive recurrent or metastatic adenocarcinoma of the stomach or GEJ (NCT01848834). The primary endpoints were safety and tolerability. Thirty-nine patients were enrolled, and all patients were included in the safety analysis. Six grade 3 or 4 treatment-
related adverse events occurred in five (13%) patients. No treatment-related deaths occurred. Thirty-six patients were evaluated for efficacy. ORR was 22% (95% CI [10.1%, 39.2%]) per RECIST version 1.1, and all responses were partial.24 Given these results, pembrolizumab is being studied in combination with chemotherapy and targeted agents.

Preliminary data from KEYNOTE-059 suggest the combination of pembrolizumab, cisplatin, and 5-fluorouracil (5-FU) had a manageable safety profile as first-line therapy in patients with advanced GC. Further evaluation to investigate clinical activity in a randomized, active-controlled phase III study with pembrolizumab as first-line monotherapy and in combination with cisplatin and 5-FU versus placebo, cisplatin, and 5-FU in patients with advanced GC is ongoing (KEYNOTE-062; NCT02494583). Additionally, pembrolizumab is being evaluated in the second line compared to paclitaxel in patients with advanced GC whose disease has progressed after first-line therapy with platinum and fluoropyrimidine (KEYNOTE-061; NCT02370498). PD-1 inhibition offers promise in shaping GC therapies in the coming years.

Targeting the CTLA-4 and PD-1/PD-L1 Axis

Simultaneously targeting the CTLA-4 and PD-1/PD-L1 axis may offer yet another approach and has produced high tumor regression rates in patients with melanoma. Based on this, nivolumab and ipilimumab have been studied as monotherapies and in combination with different doses (nivolumab 3 mg/kg every 2 weeks [N3], nivolumab 1 mg/kg plus ipilimumab 3 mg/kg [N1+I3], or nivolumab 3 mg/kg plus ipilimumab 1 mg/kg [N3+I1]) in advanced GC. A total of 154 patients (96%) were evaluated for efficacy, and the confirmed ORR was 16%: 14% in N3, 26% in N1+I3, and 10% in N3+I1, including two patients with complete response (one in N3; one in N1+I3). The disease control rate was 38%. N1+13 resulted in tumor regression and an encouraging median OS of 6.9 months and a 12-month OS rate of 34% in patients who were chemotherapy refractory with both PD-L1–positive and PD-L1–negative tumors.32 Upcoming research on nivolumab plus ipilimumab is planned to test these against chemotherapy as first-line therapy (NCT02872116).

The merging of immunotherapy with chemotherapy, radiation therapy, and targeted agents opens new possibilities for the treatment of gastroesophageal cancers and other malignancies. The durable responses seen in some patients coupled with the lack of activity in others suggest efforts are needed to identify biomarkers and patients who might benefit the most from these strategies. As we learn more about the underlying oncogenic pathways, tumor microenvironments, and their interactions with the immune system, there is little doubt that we have entered an era where we can envision increasingly effective armamentaria to treat this group of patients.  

About the Authors: Dr. Blum Murphy is an assistant professor in the Department of Gastrointestinal Medical Oncology at The University of Texas MD Anderson Cancer Center. Dr. Ajani is a professor in the Department of Gastrointestinal Medical Oncology and a professor in the Department of Epidemiology at The University of Texas MD Anderson Cancer Center