Balancing efficacy and toxicity of immune checkpoint inhibitors

By activating the immune system to attack cancer cells, immune checkpoint inhibitors (ICI) may induce long-term control of metastatic cancer in some patients, even after the treatment has stopped. However, ICI can also cause immune-related adverse events (irAEs). These can be severe and may require treatment with corticosteroids, and if symptoms do not improve within 3-5 days, escalation to second-line immunosuppressants like the tumor necrosis factor (TNF) inhibitor infliximab. The studies in this thesis aimed to identify factors associated with (severe) irAEs (part I) and assessed the impact of irAE management on survival (part II).
In chapter 1, we provide an overview of the associations of numerous blood-based and microbial biomarkers with irAEs that were previously reported. None of these biomarkers had sufficiently proven predictive value to be of importance for clinical practice. Using data of 819 patients with melanoma who received ICI in chapter 2, we observed that a more advanced stage of melanoma, characterized by presence of non-lung visceral metastases (stage IV M1c-M1d), was associated with a reduced severe irAE risk. No association of sex with severe irAE incidence was observed. In chapter 3, using data of 251 patients with cancer who completed the SQUASH questionnaire at the start of ICI, we observed that higher physical activity levels were associated with a reduced severe irAE risk and prolonged survival. In an analysis of the gut microbiome in 497 stool samples of 195 patients with cancer receiving ICI in chapter 4, we observed that dysbiosis, characterized by a higher baseline abundance of a pathobionts and a decrease in Ruminococcaceae, was associated with developing a severe irAE.
In part II, we assessed whether immunosuppressive treatment of irAEs impacts ICI-efficacy and survival, starting off with an overview of pre-clinical and clinical studies in chapter 5. In a group of patients with melanoma receiving either ICI-monotherapy or combination therapy in chapter 6, we observed that treatment of irAEs with TNF inhibitors ± corticosteroids was associated with impaired survival compared to corticosteroids only. In an updated cohort of 350 patients with melanoma receiving combined ICI in chapter 7, we observed that this association was not only present for patients receiving TNF inhibitors but applied to second-line immunosuppression in general. We could not conclude if the association between second-line immunosuppression and survival was due to the second-line immunosuppressants themselves or the higher doses of corticosteroids which were likely also administered to these patients. Therefore, we initiated an international cohort of 606 melanoma patients with irAEs in chapter 8, in which we observed that a higher peak dose of corticosteroids and the use of second-line immunosuppression were both independently associated with worse survival. These results were confirmed in an analysis of 834 patients with combined-ICI-induced irAEs in six registrational trials across tumor types in chapter 9. Thus, both corticosteroid peak dose and second-line immunosuppression (including TNF inhibition) to treat irAEs were demonstrated to be associated with impaired survival. These data argue for considerate use of immunosuppression to treat irAEs.