Designing an integrative theoretical framework for predicting endocrine toxicity in a large prospective cohort of cancer patients treated with immune checkpoint inhibitor therapy

Immune checkpoint inhibitors (ICI) are associated with immune-related adverse events (irAEs) involving the endocrine system. The rate of endocrine irAEs can vary significantly depending on the agent used, with an average all-grade occurrence of 23% and a severe-grade occurrence of 19%. Unlike other irAEs, endocrinopathies frequently persist after ICI treatment discontinuation, highlighting the need for risk stratification and predictive monitoring. The current understanding of endocrine irAEs is limited, and there are no reliable methods to identify individuals who are most at risk of developing these complications prospectively or to stratify risk at the initiation of treatment. The current study presents an theoretical framework for systematically investigating underlying mechanisms and risk factors for endocrine irAEs. The primary objective is to generate hypotheses and advance methodological approaches for subsequent predictive modeling and clinical trials, thereby supporting the development of tailored strategies for the early detection and management of these irAEs. Within this framework, we propose that endocrine irAEs are complex and result from multidimensional interactions among various clinical, biological, and lifestyle determinants. We examine endocrine irAE risk through a multifaceted approach encompassing: (1) clinical and treatment factors (ICI agent selection, dosing schedules, treatment sequencing), (2) biomarkers of immune dysregulation (CBC-derived indices and lymphocyte/neutrophil counts), (3) inflammatory mediators (IL-6, IFN-gamma, TNF-alpha, IL-8, TGF-beta), (4) genetic predisposition, and (5) behavioral and environmental exposures (nutrition, physical activity, tobacco use, comorbid diseases). Sex is also hypothesized to be a critical effect modifier, specifically influencing the predictive utility of specific CBC indices for immune dysregulation parameters for susceptibility to irAEs. Using a systems biology framework informed by mechanistic data and current empirical evidence, we develop an integrative approach that combines quantitative biomarkers with patient-reported assessments of behavioral and environmental exposures. Robust predictive models, we propose, require integration of clinical, biological, modifiable risk, and environmental factors, stratified by sex. This approach seeks to move toward a systems-based approach rather than biomarker assessment, providing a structured foundation for subsequent empirical validation and integration into clinical trial methodology. By integrating clinical, biological, and lifestyle factors, this framework supports the development of predictive models and tailored surveillance protocols that can be implemented in future clinical trials and oncology practice.