Risk of renal adverse events (RE-AE) and electrolyte changes with immunotherapy (IT) in a real-world cohort.

Abstract

11149 Background: IT is used across multiple tumor types but is linked to immune-related AE (irAE), including RE-AEs. Acute kidney injury (AKI) is the most common renal irAE (1–2%); however, real-world comparative data on RE-AE & electrolyte trends across tumor types remain limited. We evaluated RE-AE incidence & longitudinal electrolyte & urine trends following IT. Methods: We conducted a retrospective cohort study of 5,556 adult cancer patients treated at Cleveland Clinic Ohio between 2010–2022. Included cancers were advanced-stage (III–IV) bladder, endometrial, esophageal, gastric, head & neck, renal, & lung cancers (NSCLC/SCLC), & metastatic melanoma. We compared IT (± chemo/radiation) vs non-IT regimens. Follow-up began at systemic therapy initiation & continued until RE-AE, death, or last encounter. Outcomes included AKI, AKI progressing to chronic kidney disease (CKD), & clinically significant proteinuria. Poisson regression estimated incidence rates, Kaplan–Meier methods cumulative incidence (CIR), & time-dependent Cox models assessed risk with IT as a time-varying exposure. Longitudinal serum electrolytes & urine parameters were extracted electronically & analyzed via mixed-effects & generalized linear models over a one-year post-initiation period. Results: Among the cohort, 1,291 (23%) received first-line IT & 4,265 (77%) later-line IT. Median age was 66 years; 38% were female. During median follow-up of 14 months (CI 6-34), 195 RE-AEs occurred (IT: 46; non-IT: 149). The incidence rates per 100 person years were 0.18 (0.13, 0.24) in IT group & 0.13 (0.11, 0.15) in non-IT group. 2-year CIR was 4.6% with IT & 3.9% without IT (P = 0.6). Adjusted time-dependent Cox models demonstrated no significantly increased risk for AKI, CKD, or composite RE-AEs associated with IT exposure: [AKI HR 0.99 (0.7–1.5); CKD HR 1.5 (0.97–2.4); RE-AE HR 1.3 (0.9–1.7)]. Most RE-AEs were CTCAE grade II–III. Immunomodulatory therapy was used in 15%, primarily steroids, with a 50% response rate. Longitudinal analyses showed IT was associated with higher mean CO₂ & calcium compared with non-IT (P < 0.05), but not other labs (Table 1). While on therapy, IT was associated with lower urine pH (β = -0.37, CI -0.69 -0.04) but no other UA parameter changes were noted. Conclusions: In this large real-world cohort, IT was not associated with an increased incidence of RE-AE compared with non-IT therapy, suggesting that stringent trial exclusion criteria for baseline renal impairment may be unnecessarily restrictive. IT was associated with statistically significant changes in calcium, CO₂, & urine pH, the clinical significance of which remains uncertain. Laboratory parameter Estimate 95% CI Sodium 0.14 -0.04 – 0.32 Potassium 0.02 -0.001 – 0.04 BUN -0.03 -0.59 – 0.54 Creatinine 0.02 -0.003 – 0.05 CO2/ Bicarbonate 0.16 0.001 – 0.32 Calcium 0.10 0.067 – 0.14 Chloride 0.18 -0.038 – 0.4 GFR -1.59 -4.73 – 1.55 Phosphorus -0.11 -0.25 – 0.04