Prophylactic TNF blockade uncouples efficacy and toxicity in dual CTLA‑4 and PD‑1 immunotherapy

Combination immunotherapy with the anti‑PD‑1 monoclonal antibody nivolumab and the anti‑CTLA‑4 monoclonal antibody ipilimumab has shown synergistic effects in melanoma, NSCLC and renal-cell carcinoma patients. Mouse models of cancer have also demonstrated improved efficacy using this combination providing enough evidence to conduct a phase I dose-escalation clinical trial in patients with metastatic cutaneous melanoma. Despite strong and sustained clinical responses over 40% of patients developed immune-related adverse events resulting in termination of the trial. A phase III clinical trial has previously determined that a much higher dose of ipilimumab was tolerated than was used in the phase I trial.

Generally adverse events from checkpoint inhibitors can be readily treated with steroids. Frequently however, patients develop more serious immune mediated adverse effects such as colitis and Crohn’s disease and TNF blockade with infliximab is the standard of care. In mice treatment with anti‑PD‑1 and anti‑CTLA‑4 antibodies has also been demonstrated to exacerbate autoimmune colitis.

In this letter to Nature, Perez‑Ruiz et al., demonstrate that, in mouse models, prophylactic blockade of TNF before the start of combined anti‑CTLA‑4 and anti‑PD‑1 therapy prevents autoimmune adverse events and may also enhance the anti-tumour efficacy.

Inflammatory bowel disease (IBD) was induced in mice with dextran sulfate sodium (DSS) in drinking water. IBD-induced mice treated with a combination of anti‑PD‑1 and anti‑CTLA‑4 antibodies had more severe colitis-like symptoms including colonic inflammation, thickening of the wall of the large intestine and weight loss.

Prophylactic treatment with a mouse specific monoclonal anti-TNF or with the TNF inhibitor etanercept (TNFR2‑IgG)16 reduced the severity of combination treatment.

TNF blockade with monoclonal antibody or etanercept had no effect on the anti-tumour efficacy of MC38 or B16-ovalbumin (OVA) – derived tumours. TNF inhibition increased the numbers of mice that had complete rejection of the tumour xenografts compared to those treated with combination immunotherapy.

Prophylactic blockade of Il‑6, which has previously been identified as a potential target for cancer therapy, slightly reduced the anti-tumour effects of anti‑PD‑1 and anti‑CTLA‑4 combined therapies.

In MC38 tumour bearing mice, where colitis had been induced by DSS, anti-TNF blockade in combination with anti‑PD‑1 / anti‑CTLA‑4 improved tumour rejection and survival over those treated with combination immunotherapies alone. The anti‑TNF monoclonal antibody was more effective than etanercept, potentially due to increased anti-drug antibodies or increased penetration of the antibody into the tumour tissues.

Increased tumour infiltration of CD8+ T-cells was observed as a result of combination immunotherapy and this was enhanced with TNF inhibition. A similar increase was also observed for tumour-antigen specific CD8+ T-cells in the tumour microenvironment and draining lymph nodes.

To determine if prophylactic TNF blockade resulted in reversion of a CD8+ exhaustion phenotype, a panel of targets associated with T-cell exhaustion were analysed in the tumour infiltrating CD8+ T-cells, including CTLA‑4, LAG3, BTLA, 2B4, CD160, TIM3 and intracellular Ki67. None of these markers were altered although there was a reduction in the surface expression of PD‑1.

To determine if the enhanced efficacy was due to attenuation of activation induced cell death (AICD), the researchers used cultures of T cell-receptor (TCR)-transgenic mouse CD8+ OT‑I and Pmel‑1 cells and activated them with their respective cognate peptides. In these cells, anti-TNF or etanercept decreased apoptosis in the presence of anti‑PD‑1 and anti‑CTLA‑4 monoclonal antibodies.

Increased viability of tumour-reactive CD8+ T cells in the tumour microenvironment and in tumour draining lymph nodes was also observed in mice bearing MC38 tumours that had colitis as a result of treatment. CD8+ T lymphocytes, from healthy human donors were activated with anti‑CD3 and anti‑CD28 monoclonal antibodies. Less AICD was detected when these cells were treated with ofinfliximab (an anti-TNF monoclonal antibody) or etanercept in culture.

mRNA expression of immune-related genes was examined in patient colon biopsies with checkpoint-induced colitis or from patients diagnosed ulcerative colitis. In all cases of checkpoint-induced colitis TNF mRNA expression was increased, but did not reach the levels observed in the IBD patients. Ingenuity pathway analyses confirmed activation of the TNF pathway in checkpoint induced colitis patients.

A model of xenograft-versus-host disease was developed by infusing human peripheral blood mononuclear cells (PBMCs) into Rag2‑/‑Il2rg‑/‑ mice. This results in inflammation of the colon and treatment with combined ipilimumab and nivolumab caused reduced bodyweight compared to animals that also received etarnercept. Etanercept, but not human IgG control, also reduced inflammation in the large intestine.

Using this humanised model, HT‑29 xenografts were assessed with dual nivolumab and ipilimumab treatment. A reduction in tumour volume was observed over time and co-treatment with etanercept did not affect the efficacy of immunotherapy, however, in this model, etanercept-treated mice had reduced severity of hepatitis and colitis.

These results demonstrate that in the gut and in the liver anti-TNF agents may improve the safety of combined immunotherapies, with no detrimental effects on efficacy, and in some cases may even improve efficacy. Prophylactic TNF blockade may safely allow doses of ipilimumab to be increased in immune checkpoint blockade combined therapies. These results demonstrate that clinically feasible strategies for the use of combined immune checkpoint blockade in cancers can be achieved.

Want to ask a question about

Oncology or Inflammatory Bowel Disease?

Epistem's IBD and Xenograft Models

Epistem offers preclinical in vivo colitis models including off-the-shelf models of adoptive T cell transfer-mediated and DSS-induced acute colitis. Epistem's preclinical models replicate aspects of IBD in humans and can be used to generate target validation and mechanism of action data to confirm efficacy of novel therapeutic agents. Epistem's IBD team has many years of experience, routinely running these models for a range of clients including Biotechnology, Pharmaceutical and Nutraceutical companies.

We offer Xenograft models for a number of cancers and several labelled cell lines for in-life imaging of orthotopic disease. We have syngeneic models also available, in addition, we are happy to develop and run bespoke models in collaboration with Sponsors.

All of our models are supported by our GCLP-accredited gene expression and IHC services. Epistem has recently invested in a next generation sequencing platform to support DNA/RNA/microbiome sequence profiling for our in vivo IBD models.

Perez-Ruiz et al., 2019. Prophylactic TNF blockade uncouples efficacy and toxicity in dual CTLA-4 and PD-1 immunotherapy. Nature, 569, 428–432 DOI: 10.1038/s41586-019-1162-y

About Epistem

Epistem's contract research service is committed to providing reliable, innovative and transferable pre-clinical models and services to support decision making throughout the drug discovery and development pipeline.

Tel: +44 (0)161 850 7600  Email: