Anti-programmed death ligand 1 (PD-L1) inhibitors in metastatic gastric cancer
Article information
Abstract
Metastatic gastric cancer (GC) with limited therapeutic options has a poor prognosis, and therefore, major therapeutic advances are needed. Comprehensive genomic characterization has improved understanding of GC but a great majority of genomic profiling has not been effectively translated to the clinical benefit yet. The introduction of immunotherapy with anti-programmed death ligand 1 (PD-L1) antibody such as pembrolizumab or nivolumab resulted in a rapid paradigm shift in the field of medical oncology and their clinical indication has been expanding in the past few years. Now, the era of immunotherapy in metastatic GC has arrived with recent trial results in survival benefit from anti-PD-L1 antibody. Herein, we review the mechanisms of immunotherapy, clinical trials that have been conducted or are in progress, and the potential for PD-L1 as a predictive biomarker for GC.
INTRODUCTION
Gastric cancer (GC) is the second leading cause of cancer-related mortality worldwide, with an estimated 723,000 deaths in 2012 [1,2]. Several immune checkpoint inhibitors that block the key immune checkpoints, such as cytotoxic T-lymphocyte antigen 4 (CTLA-4), programmed cell death protein 1 (PD-1), and/or programmed death ligand 1 (PD-L1), had been approved by the U.S. Food and Drug Administration (FDA) in several cancer types including melanoma, bladder, and lung cancer. We extensively reviewed the current ongoing clinical research effort and translational research for GC regarding immune checkpoint inhibitors.
IMMUNE CHECKPOINT INHIBITORS IN GC
Two anti-PD-L1 inhibitors, pembrolizumab and nivolumab, were approved by the U.S. FDA for the treatment of metastatic melanoma in 2014 [3,4]. Subsequently, immune checkpoint inhibitors have been approved by the U.S. FDA for lung cancer, renal cell carcinoma (RCC), Hodgkin’s disease and head and neck cancer since then. Many studies have reported long-term responses to anti-PD-L1 antibodies with acceptable safety profiles in lung cancer [5], RCC [6], bladder cancer [7], and chemoresistant Hodgkin disease [8]. In addition, solid tumors with high mutational load or microsatellite instability (MSI-high) have shown dramatic response to pembrolizumab in recent trials including colon cancer, GC, endometrial cancer, etc. [9,10]. Based on this finding, pembrolizumab has been approved in MSI-high tumors in U.S. Table 1 lists selected clinical trials of immune checkpoint inhibitors for metastatic GC [10-16].
PEMBROLIZUMAB
The phase Ib study KEYNOTE-012 (NCT01848834) evaluated the safety and activity of pembrolizumab in PD-L1 positive GC (n= 36). The primary endpoints were safety and response rate. The adverse events were manageable, and the response rate was 22% in PD-L1 positive GC patients who have failed to several lines of chemotherapy, thus very promising. There was no correlation between PD-L1 expression and clinical responses to pembrolizumab in the KEYNOTE-012 trial but further biomarker analysis was not extensively performed due to limited tissue specimens from the cohort [11]. In KEYNOTE-059 (NCT02335411), in GC patients who received 2 prior lines of therapy (n= 133), the objective response rate (ORR) was 16.4%; for those with PD-L1 positive tumors in the third-line setting (n= 75), the ORR was 22.7%; in patients with PD-L1 negative tumors (n = 58), the ORR was 8.6%. Hence, the ORR was higher in patients with PD-L1 positive tumors when compared with PD-L1 negative GC but importantly, responses were observed in patients with PD-L1 negative tumors [16]. Among the several ongoing trials, the KEYNOTE-061 (NCT02370498) is testing pembrolizumab versus paclitaxel after progression following a first-line platinum-based therapy. The KEYNOTE-062 trial (NCT02494583) is comparing pembrolizumab as monotherapy or in combination with platinum and 5-fluorouracil (5-FU) in a first-line setting (Table 2). The two trials are currently ongoing. Currently, a phase II pembrolizumab trial (NCT02589496) which is a single-center trial at Samsung Medical Center with integration of pre and post-biopsies has been completed patient accrual (n = 60). This trial results, hopefully, may shed light into identifying a subset of GC patients who may benefit the most from pembrolizumab.
NIVOLUMAB
In contrast to pembrolizumab, nivolumab has been investigated in patients regardless to PD-L1 status (both PD-L1 positive and negative GCs) and responses have been seen in both cohorts. Recently, in a randomized phase III trial ONO-4538 (NCT02267343), nivolumab was evaluated as a monotherapy versus placebo (2:1) in metastatic GC after second or later lines (n= 493) [12]. The study reported significant prolongation of overall survival (OS) (5.3 months vs. 4.1 months) and disease-free survival (1.61 months vs. 1.45 months) in nivolumab arm when compared to best supportive care, respectively with statistical significance. This study documented a response rate of 11% and all comer GC patients were enrolled [12]. Based on this trial, nivolumab is expected to receive authority approval in Japan and Korea in late 2017 or early 2018. The GC cohort of CheckMate 032 [10] enrolled patients on nivolumab alone and two different doses of nivolumab in combination with ipilimumab. ORR was the highest with nivolumab 1 mg/kg with ipilimumab 3 mg/kg (26%, 12 of 46 patients), relative to the nivolumab 3 mg/kg (14%, 8 of 59 patients), or nivolumab 3 mg/kg with ipilimumab 1 mg/kg (10%, 5 of 49 patients) cohorts. In the nivolumab alone arm, response rate was higher in PD-L1-positive GC (27%) when compared to PD-L1 negative GC (12%); the highest response rate was observed in the combination arm (nivolumab 1 mg/ kg with ipilimumab 3 mg/kg; 44% in PD-L1 positive and 21% in PD-L1 negative cohort) (Table 3). Given these notable findings, a phase III trial CheckMate-649 (NCT02872116) is currently ongoing to enroll patients in nivolumab plus ipilimumab versus FOLFOX (oxaliplatin plus 5-fluorouracil) or XELOX (oxaliplatin plus capecitabine) as front-line therapy (Table 2).
DURVALUMAB (MEDI-4736)
Durvalumab is currently in a phase I study to evaluate its efficacy and safety in GC patients with GC (NCT01693562). Most frequent treatment-related adverse events were fatigue, nausea, and rash. Evidence of clinical activity has been seen across all histologies [17]. More recently, a phase II PLATFORM study (NCT02678182) is ongoing. The aim of this study is to evaluate the efficacy and safety of durvalumab for maintenance therapy in GC.
AVELUMAB
Another promising immune checkpoint inhibitor is avelumab. A large phase Ib JAVELIN trial (n = 151) of first-line avelumab maintenance treatment reported a response rate of 9% with a median progression-free survival (PFS) of 12 weeks; second-line treatment reported a response rate of 10% with a median PFS of 6 weeks (NCT01772004) [18]. The adverse events reported were most common infusion reactions and fatigue [13]. Given the promising results of this trial, JAVELIN Gastric 100 (NCT#02625610) and JAVELIN Gastric 300 (NCT#02625623) phase III trials are underway (Table 2).
ATEZOLIZUMAB (MPDL-3280A)
A phase I dose-escalation study was conducted to evaluate the pharmacokinetics and safety of atezolizumab. In this trial, atezolizumab monotherapy was administered in patients with non-selected solid tumors, including one GC patient. The ORR was 21% (NCT01375842).
PD-1/PD-L1 EXPRESSION AND PATIENT OUTCOME
Melanoma, RCC, and non-small cell lung cancer express high levels of PD-L1, ranging from 66% to 100% [19-21]. In a study of 102 GC samples, Wu et al. [22] have demonstrated using an immunohistochemical approach that 42.2% of the samples had PD-L1 overexpressing tumor cells. They also showed that PD-L1 is undetectable in normal gastric tissue controls and only weakly detectable in gastric adenomas [22]. Similarly, PD-L1 overexpression was found in 40% of cases of advanced GC from the phase Ib KEYNOTE-012 trial data [11]. Another Chinese study (n = 111) reported PD-L1 positivity in 63% of GC resection specimens [23].
The high PD-L1 expression has been identified as a negative prognostic marker in melanoma [24], RCC [25], and lung cancer [26]. Additionally, PD-L1 upregulation has been detected in gastrointestinal malignancies, such as pancreatic, colorectal, and GC, correlating with poor prognosis [27,28]. Likewise, high PD-L1 expression in GC was associated with nodal metastases, advanced stage, and poor outcome [22,23]. Recently, a small cohort study showed that resected GC patients with a higher CD8+ T-cell density have a higher PD-L1 expression and worse clinical outcome [29]. These biological features in GCs are comparable to those seen in other cancers in which immune checkpoint inhibitors have demonstrated early success. PD-L1 overexpression may also play a role as a predictive response biomarker in GC.
Nevertheless, no biomarker is absolute to predict response to immune checkpoint inhibitors. Hence, an integrated analysis of mutational load, microsatellite instability, needs to be analyzed in responders and non-responders to immune checkpoint inhibitors in GC. The impact of Epstein-Barr Virus status and microsatellite instability status on response to immune checkpoint inhibitors in GC will be very interesting.
CONCLUSION
In spite the development of targeted agents such as trastuzumab and ramucirumab, precision medicine for GC patients has still a long way to go. Based on the promising early trial results in GC with anti-PD-L1 antibodies such as pembrolizumab, nivolumab, and avelumab, the practice guideline and pattern in metastatic GC patients will be rapidly changed in the next few years. Especially, given the high percentage of MSI-high in GC (up to 20%), the impact of immunotherapy in these patients will be substantial. With upcoming KEYNOTE 061, KEYNOTE 062 trials, and CheckMate trials, the positioning of immune checkpoint inhibitors in GC will be refined and tuned in the next few years. In addition, identification of responders versus non-responders to immunotherapy will very likely facilitate and improve survival outcome in GC following immunotherapy.
Notes
No potential conflict of interest relevant to this article was reported.