Immune Checkpoint Inhibitor Therapy and Risk of Glomerular Disease: A Large Propensity-Matched Cohort Study

Immune checkpoint inhibitors (ICIs) have transformed cancer treatment but may cause immune-related kidney injury, and their association with glomerular disease remains incompletely understood. We evaluated the risk of incident glomerular disease following ICI therapy using the TriNetX U.S. Collaborative Network. Adult patients with cancer treated between January 2014 and March 2025 who received ICIs were compared with those receiving conventional antineoplastic therapies after 1:1 propensity score matching for demographics, cancer type, comorbidities, medications, and baseline kidney function. The index date was constrained to an eligibility window ending 1 March 2024, and identical exposure requirements were applied to both cohorts. The primary outcome was incident glomerular disease (ICD-10-CM codes N00, N01, N02, N04, N05 and N08), and hazard ratios (HRs) were estimated using Cox proportional hazards models. The proportional hazards assumption was tested for every outcome, death was additionally modelled as a competing event, and subgroup heterogeneity was assessed by formal tests of interaction. Among 71,354 matched patient pairs, glomerular disease occurred in 2.5% of ICI-treated patients and in 2.4% of controls. ICI therapy was associated with a higher hazard of glomerular disease (HR 1.492, 95% CI 1.392-1.599). The excess hazard was concentrated in the first 500 days after initiation (interval HR 1.68); beyond approximately 1000 days, the interval point estimates approached unity. In a supporting analysis in which death was modelled as a competing event, the estimated 5-year cumulative incidence was 4.5% with ICI therapy and 3.2% with comparator therapy, an absolute difference of approximately 1.2 percentage points, and the corresponding difference in restricted mean time free of glomerular disease was about 20 days. These absolute estimates were derived from the survival functions of the matched cohort rather than from patient-level competing-risk regression and should be read as approximations. In exploratory analyses, heterogeneity was demonstrated by race, cancer type and ICI agent. Renal biopsy and urine protein testing were performed at indistinguishable rates in the two cohorts, arguing against differential detection. ICI therapy is associated with a modest absolute increase in the risk of coded incident glomerular disease, concentrated in the first 18 months, which may support targeted rather than indefinite renal surveillance.