Association Between Serum 25-Hydroxyvitamin D Levels and Immune-Related Adverse Events During PD-1/PD-L1 Inhibitor Therapy: A Retrospective Study.
Zhang Yue Y, Zeng Peng P, Fan Weibin W, Lin Bin B et al.
Vitamin D exerts pleiotropic immunomodulatory effects, yet its relationship with immune-related adverse events (irAEs) during PD-1/PD-L1 inhibitor therapy remains unclear. We examined the association between serum 25-hydroxyvitamin D [25(OH)D] concentrations and documented irAEs in a real-world cohort. In this single-center retrospective study, we analyzed 46 serum 25(OH)D test records from 35 patients treated with PD-1/PD-L1 inhibitors between October 2023 and April 2026; lung cancer was the predominant tumor type. Documented irAEs were adverse events recorded in the dataset, adjudicated as immunotherapy-related by the treating clinicians and graded according to CTCAE version 5.0. Vitamin D status was classified as deficient (<30 nmol/L), insufficient (30 to <50 nmol/L), or sufficient (≥50 nmol/L). Because repeated records from the same patient are not statistically independent, record-level analyses were treated as exploratory, whereas the primary analysis was performed at the patient level using each patient's first available 25(OH)D measurement to ensure independence of observations. Eleven record-level irAEs were documented in 9 patients; after exclusion of one event with unverifiable immune-mediated attribution (isolated hypokalemia), 8 patients had adjudicated irAEs in the primary analysis. At the exploratry record level, irAE proportions decreased stepwise across vitamin D categories: 60.0% (deficient), 20.0% (insufficient), and 6.3% (sufficient) (Cochran-Armitage trend, P=0.003). In the primary patient-level analysis, proportions were 62.5%, 13.3%, and 8.3%, respectively (trend P=0.008); first-test 25(OH)D concentrations were lower in patients with irAEs than in those without (median 26.70 vs 45.50 nmol/L; P=0.007), although the binary comparison (deficient/insufficient vs sufficient) was not statistically significant (P=0.216; OR=4.813, 95% CI: 0.517-44.824). In 10 of the 11 event-positive records, the documented event preceded the linked vitamin D measurement. Lower serum 25(OH)D concentrations were associated with documented irAEs in this small exploratory dataset. Given the temporal ambiguity and residual confounding, the findings are hypothesis-generating and cannot establish causality or prediction; whether vitamin D status can help risk-stratify patients for irAEs-rather than merely reflecting immune activation or nutritional status-requires prospective studies with protocol-defined pretreatment sampling.