Combinations of anti-PD-1/PD-L1 monoclonal antibodies and other immune checkpoint inhibitors may lead to more robust antitumour effects in gastrointestinal cancers. Indeed, in advanced melanoma, the combination of PD-1 and CTLA-4 blockade demonstrated a synergistic effect compared to either inhibitor alone [111]. The combination of PD-1 and CTLA-4 inhibition is currently being explored in gastric, colorectal, hepatocellular and pancreatic cancer. Preliminary analyses of the combination of nivolumab and ipilimumab in gastric, GEJ and dMMR colorectal cancer have been reported. In CheckMate-032, 160 patients with metastatic gastric or GEJ cancer were treated with nivolumab alone (3 mg/kg) or in combination with ipilimumab (two cohorts; nivolumab 1 mg/kg + ipilimumab 3 mg/kg or nivolumab 3 mg/kg + ipilimumab 1 mg/kg). The ORR was 14 % for the nivolumab alone group (N3), 26 % for the nivolumab 1 mg/kg + ipilimumab 3 mg/kg group (N1 + I3) and 10 % for the nivolumab 3 mg/kg + ipilimumab 1 mg/kg group (N3 + I1). Median OS was 5.0 months (95 % CI, 3.4–12.4 months) for the N3 group, 6.9 months (95 % CI, 3.6 months to NR) for the N1 + I3 group and 4.8 months (95 % CI, 3.0–9.1 months) for the N3 + I1 group [112]. The interim analysis of CheckMate-132 included 56 patients with dMMR metastatic colorectal cancer, treated with nivolumab 3 mg/kg (N3) or nivolumab 3 mg/kg and ipilimumab 1 mg/kg (N3 + I1). The ORR was 27 % for the N3 group and 15 % for the N3 + I1. Despite the higher response rate in the monotherapy group, the combination was superior with regard to PFS and OS. The 4-month PFS was 55 % for the N3 group and 80 % for the N3 + I1 group, while the 5-month OS was 75 % for the N3 group and 100 % for the N3 + I1 group [113]. These preliminary results demonstrate a promising clinical activity of immune checkpoint inhibitor combinations in gastrointestinal malignancies. Further research is needed to identify patients who are more likely to respond to these combinations compared to PD-1 monotherapy.

As noted above, several other co-inhibitory receptors, such as TIM-3, LAG-3, B7-H3, are expressed in gastrointestinal tumours and are associated with poor prognosis [54, 66, 68, 89, 90]. Preclinical studies in murine models of colon cancer have demonstrated synergy between PD-1 and LAG-3 blockade, leading to complete regression of the tumours in the majority of mice treated with this combination [114]. A large number of early phase trials are currently investigating the safety and efficacy of inhibitors targeting LAG-3 and other immune checkpoints in combination with PD-1 blockade in advanced solid tumours, including gastrointestinal malignancies (Table 6.5).

An alternative strategy to augment the efficacy of PD-1 inhibition is to increase T cell activity. Several co-stimulatory molecules, including members of the TNF receptor family (OX40, CD137 [4-1BB] CD40 and CD27) and members of the B7-CD28 family members (ICOS and B7-H3), lead to enhanced memory and effector T cell responses, increased T cell survival and regression of tumours in murine models of solid tumours [115–119].

Specifically, in preclinical models of colorectal cancer, activation of CD137 with an agonistic monoclonal antibody enhanced NK cell degranulation and cytotoxicity. Combined anti-EGFR and anti-CD137 therapy was synergistic and resulted in complete tumour resolution and prolonged survival [120]. In addition, the clinical impact of a CD40 agonist monoclonal antibody in combination with gemcitabine chemotherapy was evaluated in a phase I clinical trial for patients with chemotherapy-naïve advanced or metastatic pancreatic cancer. Of 21 patients enrolled in the trial, 4 patients (19 %) achieved PR and 11 patients had stable disease. Median PFS was 5.6 months and median OS was 7.4 months. Gemcitabine monotherapy historically achieves an ORR of 5.4 % with median PFS of 2.3 months and median OS of 5.7 months. Therefore, the combination of anti-CD40 and gemcitabine showed promising efficacy in patients with metastatic pancreatic cancer [121].

Based on these encouraging results, combinations of PD-1/PD-L1 inhibitors and T-cell agonists are currently being explored in early phase trials (Table 6.5).

Monoclonal antibodies targeting the human epidermal growth factor (HER) family receptors have been in clinical use in gastrointestinal malignancies for many years. Specifically, the anti-EGFR monoclonal antibodies, cetuximab and panitumumab, are approved for the treatment of KRAS wild-type advanced or metastatic colorectal cancer, while the anti-HER2 monoclonal antibody, trastuzumab, is approved as first-line therapy of metastatic, HER2-positive gastroesophageal cancer. In addition to blocking the EGFR/HER2 signalling pathway, these therapies have a significant impact on the tumour immune microenvironment by promoting antibody-dependent cellular cytotoxicity. Preclinical studies have demonstrated that the therapeutic effect of anti-EGFR and anti-HER2 monoclonal antibodies is dependent on CD8+ and NK cells [122, 123]. In addition, studies in immunocompetent mouse models of HER2-positive breast cancer suggest that PD-1 blockade can significantly improve the therapeutic efficacy of anti-HER2 monoclonal antibody therapy [123]. This potential synergistic effect of PD-1 inhibition and anti-EGFR or anti-HER2 therapy is currently being investigated in KRAS wild-type colorectal cancer and HER2-positive gastric and oesophageal cancer, respectively (Table 6.5).

Another approach to enhance the efficacy of PD-1/PD-L1 inhibitors is to combine them with agents targeting other immunosuppressive components of the tumour microenvironment.

Emerging data show that VEGF plays an important role in tumour immune evasion. In addition to promoting tumour angiogenesis, VEGF blocks the maturation of dendritic cells and therefore impairs antigen presentation, inhibits lymphocyte trafficking through the vasculature into tumours and enhances T cell exhaustion by upregulating immune checkpoint expression [124–128]. Angiogenensis inhibitors, particularly anti-VEGFA monoclonal antibodies, have been shown to augment the anti-tumour immune response and synergise with PD-1 blockade in both preclinical and early clinical studies [128, 129]. In a mouse model of colorectal cancer expressing high levels of VEGF-A, combined therapy with anti-PD-1 and anti-VEGF-A monoclonal antibodies led to a significant decrease in tumour volume, compared to either agent alone. Of note, anti-VEGF-A therapy decreased the proportion of tumour-infiltrating CD8+ T cells expressing immune checkpoints [128]. In patients with metastatic melanoma, combination therapy with bevacizumab and ipilimumab enhanced lymphocyte and macrophage trafficking leading to a more robust immune cell infiltration into the tumour, compared to ipilimumab monotherapy [129]. These studies provide the mechanistic foundation for combining antiangiogenic therapy with immune checkpoint blockade. Clinical trials evaluating the efficacy of combinatorial treatment with PD-1 blockade and monoclonal antibodies targeting VEGF-A or its receptor are currently ongoing in a variety of gastrointestinal malignancies (Table 6.5).

An additional approach to augment anti-PD-1 therapy is the depletion of Tregs. Chemokine receptor 4 (CCR4) is expressed selectively in effector Tregs, which promote immunosuppression. Preclinical in vitro studies demonstrated that CCR4 blockade led to the elimination of Tregs and the induction of T-cell responses specific to human tumour antigens. In addition, administration of the anti-CCR4 monoclonal antibody mogamulizumab to patients with adult T-cell leukaemia-lymphoma selectively depleted effector Tregs, enhanced immune responses and significantly reduced tumour cells [130]. Combination of mogamulizumab with either PD-L1 or CTLA-4 inhibitors is currently being tested in patients with metastatic pancreatic cancer (Table 6.5).

Critical drivers of the immune escape in gastrointestinal malignancies include tumour-associated macrophages (TAMs) and myeloid derived suppressor cells (MDSCs). These cells not only contribute to immune suppression, but also promote cancer cell proliferation, mediate resistance to cytotoxic chemotherapy and augment the metastatic potential of solid tumours [131, 132]. Colony stimulating factor 1 receptor (CSF1R) is expressed on TAMs and MDSCs and can be targeted with monoclonal antibodies. In preclinical models of pancreatic cancer, CSF1/CSF1R blockade was shown to decrease the number of TAMs and reprogram remaining TAMs to support antigen presentation and bolster T cell activation within the tumour microenvironment. This in-turn led to reduced immune suppression and elevated interferon responses, which restrained tumour progression. However, inhibition of CSF1R also increased the levels of PD-L1 and CTLA-4 in these tumours, potentially limiting its long-term efficacy [133]. One rational strategy to overcome immune evasion is to combine PD-1/PD-L1 or CTLA-4 inhibitors with CSF1R blockade. This approach is under investigation in several solid tumours, including pancreatic and gastric cancer (Table 6.5).