Respiratory failure in patients treated with immune checkpoint inhibitors.

Abstract

e24169 Background: Immune checkpoint inhibitors (ICIs) are increasingly used across cancer types, yet the causes and outcomes of acute respiratory failure (ARF) in ICI-treated patients remain poorly defined. Although checkpoint inhibitor pneumonitis (CIP) is a recognized immune-related adverse event, its contribution to severe ARF and short-term outcomes is uncertain. Methods: We conducted a retrospective cohort study of adults treated with ICIs who developed ARF requiring ≥24 hours of non-invasive or invasive ventilatory support across five hospitals in a single health system (2017-2024). ARF etiology was determined by manual chart review with multidisciplinary adjudication to identify CIP. Clinical characteristics, cancer features, and outcomes were compared between CIP and non-CIP ARF. Cox proportional hazards models evaluated associations between CIP and mortality. Results: Among 197 patients with ICI-associated ARF, 29 (15%) were adjudicated as CIP-related and 168 (85%) had non-CIP etiologies. Median age was 66 years (IQR 59–74), and demographics were similar between groups. Lung cancer was more common in CIP than non-CIP ARF (62% vs 35%, p = 0.01). Time from first ICI exposure to ARF did not differ (median 112 vs 138 days, p = 0.44). CIP patients more often received non-invasive ventilation alone (69% vs 45%) and were less likely to require isolated invasive mechanical ventilation (3% vs 44%, p < 0.01). CIP was associated with more severe hypoxemia (median SpO₂/FiO₂ 120 vs 148, p = 0.01) but less non-respiratory organ dysfunction (median SOFA 2 vs 5, p < 0.01). Ninety-day mortality was 65.5% in CIP and 70.8% in non-CIP ARF (p = 0.56). CIP was not associated with 90-day mortality in adjusted analysis (HR 0.66, 95% CI 0.40-1.08). CIP was associated with lower in-hospital mortality (adjusted HR 0.43, 95% CI 0.21-0.88) and shorter time to hospital discharge (adjusted SHR 1.81, 95% CI 1.07-3.05). Conclusions: CIP accounted for a minority of ARF events after ICI therapy and was not associated with higher short-term mortality despite more severe hypoxemia. Outcomes following ARF appear driven more by underlying illness severity and cancer burden than by immune-mediated lung injury. Improved diagnostic precision may help guide immunosuppression and prognostication in this high-risk population. Causes of Acute Respiratory Failure (ARF) in ICI-treated patients. ARF Etiology n (%) Pneumonia 54 (27%) Progressive cancer 37 (19%) Volume overload 29 (15%) CIP 29 (15%) Aspiration 26 (13%) Sepsis / ARDS 9 (5%) Pulmonary embolism 9 (5%) Neuromuscular 3 (2%) Total 197 (100%)