The advent of immune checkpoint inhibitors as a therapeutic strategy to augment the endogenous anticancer immune response has revealed another obesity paradox, as a higher BMI predicts treatment efficacy in several malignancies (6–8). In this issue of the JCI, Gómez-Banoy and colleagues now extend this paradox to endometrial cancer treated with checkpoint inhibitors targeting the interaction between programmed cell death receptor-1 (PD-1) and its ligand (PD-L1) (9). PD-1 is expressed on T cells. When bound to PD-L1, a protein often overexpressed by tumor cells, PD-1 suppresses T cell activation, thereby promoting immune evasion by PD-L1–expressing cells. Disruption of PD-1/PD-L1 interaction by one of several currently available antibody-based therapeutics is the front-line therapy for advanced endometrial cancer. Gómez-Banoy and investigators retrospectively identified 768 patients with endometrial cancer treated with immune checkpoint inhibitors between 2015 and 2022, of whom 524 met criteria of recurrent, advanced, or metastatic endometrial cancer and hence were included in their formal analyses. Most patients in the cohort were treated with anti–PD-1 therapy (85%), with 15% receiving anti–PD-L1 therapy. Recognizing that endometrial cancer is a heterogeneous disease, the authors selected a cohort that generated a sufficiently broad distribution of clinical characteristics to facilitate multivariate analyses. They confirmed that the improved progression-free and overall survival of patients with an elevated BMI in the overweight or obese range was not driven by obvious potential confounders, such as cancer stage, prior therapeutic exposures, or histological or molecular subtypes. Indeed, the relationship between BMI categories and treatment efficacy was strongest in patients with the copy number–high/TP53abnormal (CN-H/TP53abn) molecular subtype, which has been associated with immune cells expressing PD-1/PD-L1 in tumors. These findings provide conceptual support for obesity as having an on-target effect on treatment responsiveness. Despite limitations inherent to the retrospective study design, the reasonable size and diversity of the cohort coupled with appropriate multivariate adjustments for likely confounders provides robust evidence for an obesity paradox in patients with endometrial cancer undergoing treatment with immune checkpoint inhibitors (9).

The Gómez-Banoy study also breaks important ground by moving beyond a simple BMI-centric view of obesity to a consideration of adiposity traits (9). This nuanced viewpoint is important because BMI does not account for differences in lean mass, nor does it reflect the recognition of depot-dependent differences in fat biology that has emerged from mechanistic studies in rodents and humans. The predilection for visceral adipose tissue (VAT), for example, is heritable and its underlying genetic architecture appears distinct from the genetics of obesity itself (10). VAT exhibits greater potential for lipolytic release of fatty acids and poor capacity for metabolically active beige/brown fat phenotypes relative to subcutaneous adipose tissue (SAT) (11, 12). VAT also displays a greater capacity for new fat cell formation and an obesity-related cycle of fat cell death and regeneration, which is associated with increased macrophage influx to resolve lipid-rich dead adipocytes (13–15). Macrophages contained in VAT with obesity appear particularly prone to a proinflammatory cell state, whereas counterbalancing regulatory lymphocyte populations (e.g., T-regulatory cells) are diminished in number and/or function in obese VAT (16, 17). Increased VAT is further associated with a systemic imbalance in potentially toxic proinflammatory cytokines relative to protective antiinflammatory cytokines and adipokines such as adiponectin (11). It is thought that such VAT-specific pathobiology accounts for why VAT volume, as can be measured by clinical imaging, predicts cardiometabolic and cancer risk (18). Most patients (n = 500, 95%) in the Gómez-Banoy cohort (9) had clinical CT scans of the abdomen available for 2D measurement of SAT and VAT quantity at the L3/L4 spine level, which generally correlates with volumetric measures and clinical outcomes (19). This analysis identified a second interesting paradox in that the quantity of putative unhealthy VAT, and not SAT, predicted positive clinical responses to immunotherapy.