WORLD CLASS BASIC, TRANSLATIONAL AND CLINICAL RESEARCH

Role of the inflammatory response in cancer

The inflammatory response that takes place at the tumour bed has a dual role in cancer. The most common type of inflammatory profile measured in solid tumours is long known for fostering aggressive tumour growth and for its strong association with dismal prognosis. In contrast, the remarkable clinical responses that some patients experienced following treatment with so-called immune checkpoint inhibitors (ICIs), have highlighted a type of inflammation with profound cancer inhibitory features. The latter inflammatory ‘flavour’ is less often present in clinically apparent tumours and is linked to increased infiltration by select innate and adaptive immune cells with key functions in anti-tumour immunity. Our group investigates the principles and rules that control the establishment of tumour inflammatory environments that either promote or restrain cancer growth spontaneously or following treatment. We argue that improved understanding of the determinants of the intratumoral inflammatory profile will help design better cancer treatments, especially those that exploit the anti-cancer function of the immune system. In this context, we combine fundamental and translational research to inform the design of novel interventions that boost tumour immunity and improve the efficacy of cancer therapy.

Cyclooxygenase (COX)-2/prostaglandin E2

Much of our group current efforts are centred on investigating the role of the cyclooxygenase (COX)-2/prostaglandin E2 (PGE2) pathway in tumour inflammation and immunity. This pathway is very often upregulated across many tumour types, including in lung, colorectal, breast, and pancreatic cancers and has been implicated in promoting several aspects of malignant tumour growth.

Importantly, our past work has identified PGE2 as a powerful instructive signal that stimulates a type of inflammatory flavour that fuels tumour growth and therapy resistance through enabling immune escape. Combining the use of genetically engineered in vitro and in vivo pre-clinical cancer models with the analysis of patient samples, our work specifically pointed to the inhibitory effects of PGE2 on the immune system as the basis for its protumourigenic role. Accordingly, genetic ablation of PGE2 synthesis on cancer cells, or of its receptors EP2 and EP4, on select immune cell subsets led to spontaneous immune-dependent control of tumours that otherwise grow aggressively in wild-type hosts (Zelenay et al. Cell 2015; Bonavita et al. Immunity 2020). Searching for the primary regulators of cancer inhibitory inflammation in these experimental systems, we showed that natural killer (NK) cells are direct targets of PGE2 activity in vivo. When NK cells were rendered insensitive to PGE2 through genetic ablation of EP2 and EP4, NK cells drove an intratumoral inflammatory remodelling characteristic of the so called ‘hot’ inflamed tumours that set the stage for T cell-mediated tumour control.

Pharmacological inhibition of COX-2

Based on this genetic evidence highlighting PGE2 as a primary immunosuppressive factor in the tumour environment, we have explored if and how pharmacological inhibition of COX-2 could equally modulate the inflammatory landscape at the tumour bed and enhance the efficacy of ICIs (Pelly and Moeini et al. Cancer Discovery 2021). To test the value of therapeutically inhibiting the pathway in clinically relevant settings, we first used celecoxib, a selective COX-2 inhibitor widely prescribed for the management of conditions like arthritis. Daily oral administration of celecoxib, at a dose considered safe for human use, alongside treatment with ICI led to greater tumour control compared with the single treatments. Across multiple experiments in independent cancer models, we found that ICIs and selective COX-2 inhibition synergised. About two-thirds of animals fully eradicated their tumours after receiving combination therapy, while around only 25% eradicated their tumours when monotherapy was followed with ICIs, and <5% achieved full tumour eradication following monotherapy with celecoxib. We obtained similar results using forefront antagonist of the PGE2 receptors, EP2 and EP4, establishing a prevalent role of PGE2, among other prostaglandins and COX-2 products, and a potentially safer more targeted approach to limit the immunosuppressive effects of PGE2.

Synergy between immune checkpoint inhibitors and selective COX-2 inhibition

Most surprisingly, the synergy with ICIs was also observed when co-administering corticosteroids, widely considered to be potent immunosuppressants and frequently prescribed to cancer patients for the management of the adverse effects that often patients face following ICI therapy. These findings are of particular clinical relevance as they suggest that corticosteroids, while limiting the toxicities of ICIs, might concomitantly improve the anti-tumour response. Although it remains to be established in which exact settings corticosteroids might boost immune-dependent tumour control, our current work supports the notion that broad non-steroidal and steroidal anti-inflammatory drugs can somewhat paradoxically stimulate a cancer-restraining inflammatory response.

To investigate the underlying mechanisms by which pharmacological inhibition of PGE2 synthesis or signalling improved the efficacy of ICI therapy, we carried out deep cellular and molecular profiling of tumours early following treatment. In doing so, we found that both celecoxib and the EP2/EP4 antagonists can very rapidly stimulate a shift in the inflammatory profile, tilting the balance towards a type of inflammation that is also observed enriched in patients that respond to ICIs. Importantly, these rapid changes were also transient and waned over time unless we re-dosed the tumour-bearing animals with the COX-2/PGE2 inhibitors. This data exposed a notable plasticity in the intratumoural inflammatory landscape and implied that sustained inhibition of the COX-2/PGE2 axis might be required to maximise the efficacy of ICIs.

We demonstrated the relevance of our findings for human cancers by showing that addition of celecoxib also leads to a similar, rapid shift in the inflammatory profile of patient tumours using a recently established experimental platform developed by our collaborators Daniela Thommen and Ton Schumacher from the Netherlands Cancer Institute (NKI). Thanks to a CRUK travel award, Victoria Pelly, a former postdoc in the group spent four months at the NKI where she tested the effect of adding celecoxib to short-term cultures of freshly explanted surgical specimens from a variety of cancer types. She found, similarly as in the mouse cancer models, that celecoxib did not generally dampen inflammation but rather promoted an intratumoural inflammatory switch that was accompanied by enhanced T cell effector function.

Summary

Altogether, our findings establish PGE2 signalling as an independent immune checkpoint and support the rational for using readily available anti-inflammatory drugs to tilt the balance toward cancer-inhibitory inflammation and improve the efficacy of current immune-targeting drugs (Figure). Certainly, these findings have significantly contributed to the design of two investigator-led clinical trials sponsored by The Christie NHS Foundation Trust. The first one, IMpALA (IMmunological effects of AveLumab and Aspirin) is funded by a Breast Cancer Now Catalyst Programme and will test in a neoadjuvant setting the combination of high-dose aspirin, a widely used anti-inflammatory drug with the ICI avelumab. The primary objective is to determine whether co-administering aspirin, a COX-inhibitor with ICI treatment switches the intratumoural inflammatory profile towards one that is more favourable to T cell-mediated tumor control as observed in our preclinical models. The second trial, LION (Lifting Immune CheckpOints with NSAIDs), is a pan-tumour basket trial funded by the J P Moulton Charitable Foundation and The Christie NHS Foundation Trust, in which we will test if addition of a selective COX-2 inhibitor enhances the efficacy of the standard of care immunotherapy in non-small cell lung cancer, triple-negative breast cancer and clear cell renal cell cancer.