C77-05 Incidence and Risk Factors for Pneumonitis After Check Point Inhibitor Immunotherapy in Patients With Genitourinary Cancers

Abstract Rationale Genitourinary cancers (GU) contribute significantly to global cancer mortality. Immune checkpoint inhibitors (ICI) have improved outcomes for patients with GU cancers. However, immune-related adverse events, including pneumonitis, can increase morbidity and mortality. Although clinical trials measure ICI-pneumonitis (ICI-P) incidence at 2-3% in GU cancers, real-world studies of incidence and ICI-P risk factors are lacking. We evaluated whether pre-existing interstitial (ILD) or obstructive lung disease (COPD), smoking history, combination CTLA-4 blockade, pre-ICI thoracic radiation, and elevated baseline absolute monocyte count were associated with ICI-P in GU cancer patients. Methods We conducted a retrospective study of 510 patients with GU cancers treated with ICIs. Incidence and risk factors for ICI-P were evaluated using a semi-competing risk model. Covariates included age, sex, BMI, smoking history, pre-existing lung or autoimmune disease, histology, prior therapies, and prior chest radiation. The primary outcome was time to pneumonitis, with death before pneumonitis as a competing risk. We used extended Cox proportional hazards models with ICI-P as a time-dependent variable to examine whether ICI-P increased mortality. Results The mean age was 65±11 years (72% male) and the most common cancer was renal cell carcinoma (RCC). The mean duration of ICI treatment was 370 days. ICI-P occurred in 9% of patients. Treatment duration was marginally associated with ICI-P (HR 1.01, 95% CI 0.99-1.03). COPD, ILD, asthma, pre-ICI chest radiation, monocyte count, histology, prior lines of therapy, smoking status, age, and sex were not associated with ICI-P risk, though ILD was rare. In extended Cox models, ICI-P did not increase mortality (HR = 1.3, 95% CI 0.7-2.5). Pre-ICI chest radiation increased mortality (HR = 1.8, 95% CI 1.1-3.2), possibly due to lung metastasis, while longer treatment duration was protective (HR = 0.97 per month, 95% 0.97-0.98), possibly indicating selection bias. Conclusion In GU cancer patients treated with ICI-therapies, ICI-P occurred at a higher incidence than in clinical trials (9%), but risk factors for ICI-P in patients with lung cancer (e.g., chest radiation, ILD, monocyte count) were not associated with ICI-P in GU cancers. In addition, unlike in lung cancer, ICI-P in patients with GU cancers did not increase mortality. Further work will examine whether quantitative assessments of ILAs are associated with ICI-P, and whether stratified analyses in RCC alone may aid in the detection of ICI-P risk factors. Our findings highlight that ICI-P risk factors should be examined in unique cohorts individually and not in aggregate. This abstract is funded by: None