Safety and efficacy of sintilimab combined with oxaliplatin/capecitabine as first-line treatment in patients with locally advanced or metastatic gastric/gastroesophageal junction adenocarcinoma in a phase Ib clinical trial

The fifth most commonly diagnosed cancer worldwide is gastric cancer (GC), accounting for about 33% of cancer-related deaths globally and the third most common cancer in China with almost half of worldwide new GC cases occurring in China annually [1, 2]. The standard treatments exhibit regional differences among western countries, Japan/Korea and China, which are considered to be associated with different screening and early detection methods as well as different biological behaviors, disease characteristics and ethnicity [3‐5].

Surgical resection is the only radical therapy for gastric/gastroesophageal junction (G/GEJ) cancer. However, systemic chemotherapy is an alternative main therapy for G/GEJ cancer because of the high relapse rate after post-resection surgery and for the many patients diagnosed at an advanced-stage. For advanced G/GEJ cancer, first-line treatment mainly involves platinum-based chemotherapy using a combination of two or three drugs (trastuzumab is given to patients whose tumor is human epidermal growth factor receptor-2 (HER2) positive), but the overall survival (OS) is disappointing, since the maximum OS time has been reported to be 13.8 months [6‐10]. Any potential novel drug that will increase patient survival times is urgently needed, in particular for first-line treatment.

Immune checkpoint inhibitor treatment is a new approach for tumor immunotherapy [11, 12]. The treatment diminishes the immune system tolerance to tumor cells and improves the effective identification and eradication of tumor cells by blocking T cell inhibition [13]. The programmed death-1 (PD-1) antibody specifically binds to PD-1, thereby inhibiting apoptosis of antigen-specific T cells and thus reducing regulatory T cell (Treg) apoptosis by inhibiting the activation of PD-L1 [14, 15].

The efficacy of anti-PD-1 antibodies monotherapy in patients who had prior chemotherapy for advanced GC has been demonstrated and supported by several trials. In the KEYNOTE-012 and KEYNOTE-059 trials, pembrolizumab monotherapy showed objective response rates (ORR) of 22% (n = 36) [16] and 15.5% (n = 148) [17], respectively in PD-L1 positive advanced GC patients after at least two prior systemic therapies. Based on such results, the Food and Drug Administration approved pembrolizumab for third-line treatment of patients with recurrent or advanced GC. In the ATTRACTION-2 study, nivolumab monotherapy improved OS from 4.1 to 5.3 months (hazard ratio 0.63, 95% CI: 0.51–0.78; P < 0.000), compared with a placebo in advanced GC that was refractory or intolerant to previous treatment regimens [18].

However, between 30 and 60% of patients exhibit no response to PD-1 blockade, which is considered to be associated with T cell exclusion or exhaustion or inadequate T cell trafficking and many immunosuppressive factors accumulate in the tumor microenvironment [19]. New therapy regimens that improve the response and long-term efficacy are desperately needed. The efficacy of anti-PD-1 therapy in combination with chemotherapy has been confirmed in non-small-cell lung cancer [20, 21]. In addition to direct tumor killing, conventional cytotoxic chemotherapy has demonstrated immunoregulatory properties by enhancing tumor antigenicity, disturbing immune suppressive pathways, inducing immunogenic cell death, and increasing effector T-cell reactions [22]. It is safe to hypothesize that anti-PD-1 antibodies in combination with chemotherapy may further improve the clinical outcomes of patients with advanced GC. Sintilimab is a highly selective, monoclonal IgG4 antibody that inhibits interactions between PD-1 and its ligands, with strong anti-tumor response [23]. A phase 1a study for dose escalation, has demonstrated the tolerance and pharmacological activity of sintilimab in patients with advanced-stage solid tumors, but there is limited evidence for the efficacy of antibodies against PD-1 plus chemotherapy in Chinese G/GEJ adenocarcinoma patients. Thus, the present trial was conducted to investigate the safety and efficacy of sintilimab combined with CapeOx as first-line therapy for a cohort of patients with G/GEJ adenocarcinoma.

From 26 Dec, 2017 to 17 Oct 2018, 25 patients were screened and 20 were enrolled in the G/GEJ adenocarcinoma cohort (Fig. 1). The median interval between initial diagnosis and screening was 14 days (range 7–611). Most patients (80.0%) had metastatic disease status and 11 (55.0%) had ECOG scores of 1 (Table 1). The TNM stage summary is shown in Table 1 and the staging of each patient in Supplementary Table 1.

At data cutoff on May 1, 2019, the median follow-up time was 7.8 months (range 6.2–12.3). The median treatment duration was 6.2 months (range 2.1–10.4). All patients received more than 4 cycles of treatment, with the median doses of received sintilimab being 9.5 (range 4–16).

In the present study, the results from the G/GEJ adenocarcinoma cohort in a Phase Ib study demonstrated manageable safety and favorable anti-tumor activity of sintilimab combined with a CapeOx regimen as first-line treatment for unresectable advanced metastatic G/GEJ adenocarcinoma.

In terms of safety, the incidence and severity of TRAEs with sintilimab and CapeOx were generally consistent with those of known toxic effects of conventional chemotherapy [26‐28] and previously reported side effects of other anti-PD-1 antibody combined with chemotherapy regimens [29, 30]. Platelet count, white blood cell count and neutrophil count decreases were most commonly and mostly grade 1 to 2 reported TRAEs and are expected AEs associated with CapeOx [26‐28]. Only 1 patient reported discontinuation of investigational drug application due to a TRAE (abnormal hepatic function). No treatment-related death occurred in this study and in general, the addition of sintilimab to CapeOx showed a manageable safety profile and did not bear extra safety risks.

In the present study, after treatment with sintilimab plus CapeOx, patients with unresectable G/GEJ adenocarcinoma obtained an ORR of 85.0% (95% CI: 62.1–96.8%), which is higher than that of conventional first line chemotherapy. For G/GEJ adenocarcinoma, first-line treatment mainly involves platinum-based chemotherapy and fluoropyrimidine [25]. The ORR of capecitabine-based or oxaliplatin-based therapies for G/GEJ adenocarcinoma was about 30–40% [27, 28]. The ORR for anti-PD-1 antibodies with a chemotherapy regimen were variable. In the KEYNOTE-059 study, the ORR was 60.0% (95% CI: 39.0–79.0%) for pembrolizumab plus cisplatin/5-fluorouracilm (5-FU) as first-line treatment [29]. In the KEYNOTE-062 study, ORRs were 48.6 and 52.5% in patients with a ≥ 1 and ≥ 10 combined positive score (CPS), respectively, after they received pembrolizumab plus cisplatin/5-FU or capecitabine regimen as first-line therapy [31]. In ATTRACTION-04, the ORR for nivolumab with S-1/oxaliplatin was 57.1% (95% CI: 34.0–78.2%) and the ORR for nivolumab with CapeOx was 76.5% (95% CI: 50.1–93.2%) [30]. In another study, the ORR was reported to be 66.7% for an anti-PD1 antibody toripalimab plus CapeOx treatment [32].

Sintilimab plus CapeOx also showed favorable long-term efficacy. Median PFS was 7.5 months (95% CI: 6.2–9.4) and the 6-month PFS rate was 88%. Median OS was not reached and the 6-month and 12- month OS rates were 100.0 and 68.0% respectively, which was higher than for conventional treatments with a median PFS of 5.6 months (95% CI: 5.1–5.7) for capecitabine-cisplatin regimen [27] and a median OS of 11.3 months (95% CI 9.6–13.0) for an epirubicin-oxaliplatin-capecitabine regimen [28]. The median PFS times for anti-PD-1 antibodies with a chemotherapy regimen were variable ranging from 5.7 to 10.6 months, a finding which might be associated with different populations and disease status [30‐32].

Next-generation sequencing (NGS) has researcher enabled to perform target capture sequencing, which has been proposed as a reliable technique to identify mutated driver genes and for the estimation of TMBs. Its use has led to the detection of actionable alterations in various cancer related genes [33]. Regarding high TMB and the efficacy of PD-1 treatments, inconsistent results have been reported in previous studies. Wang et al. (2018) suggested that TMB might be associated with better efficacy for PD-1 monotherapy [32], whereas Mishima et al. [34] did not find a significant relationship between TMB and the response of gastric cancers to PD-1 therapy [34]. The latter data is in accordance with our finding that after treatment with sintilimab in combination with CapeOx, no significant difference in the clinical responses was found between H-TMB and L-TMB patients. However, using the median TMB as a cut-off is difficult to extrapolate to the real world clinic and bias due to the small sample size could not be excluded in the present study. In addition, it has been noted that up to now there is no uniform standard for H-TMB [33] and further investigations are urgently required.

Our results strongly indicate that sintilimab combined with CapeOx is an option for the first-line treatment of patients with advanced or metastatic G/GEJ adenocarcinoma. However, the sample size was small and it was a single-arm study without a comparator. The large scale, double-blinded and randomized Phase III clinical trial ORIENT-16 for previously untreated advanced G/GEJ adenocarcinoma patients is being conducted to evaluate the efficacy and safety of sintilimab combined with CapeOx vs CapeOx alone (ClinicalTrials.​gov Identifier: NCT03745170).