BACKGROUND

Immune checkpoint inhibitors (ICIs) are associated with severe immune-related adverse events (s-irAEs) that result in hospitalization, emergency department (ED) visits, treatment discontinuation, or death. This study examined the impact of s-irAEs and their earliest management strategies on clinical outcomes in advanced non-small cell lung cancer (NSCLC).

METHODS

Data were derived from ConcertAI Patient360 NSCLC, a US-based electronic medical record database, between January 2012 and May 2021. Eligible patients had advanced NSCLC and received ICI-containing therapy. s-irAEs and management actions were abstracted from unstructured EHR data from ICI initiation through the earliest of 100 days after ICI discontinuation, start of a non-ICI-containing regimen, loss to follow up, end of study period, or death. Multivariable Cox regression analysis was used to evaluate the association between s-irAEs and their earliest management strategies, and real-world progression-free survival (rwPFS) and real-world overall survival (rwOS).

RESULTS

The study included 3211 patients. Median (IQR) age was 67 (60-73) years, and 44.9% were female. Most patients (61.6%) initiated ICIs as first-line therapy; half (50.1%) initiated ICIs as monotherapy, with nivolumab monotherapy (29.5%) as the most common initial ICI-containing regimen in any line. Overall, 8.6% of patients experienced s-irAEs, most often diarrhea (3.5%), pneumonitis (1.4%), and rash (1.3%). Among patients who experienced at least one s-irAEs, over half (57.4%) were hospitalized, and 71.8% were treated with corticosteroids, any time after the occurrence of their first s-irAEs. Median rwPFS was 4.9 (95%CI, 4.6-5.2) months, and median rwOS was 13.6 (12.6-14.7) months from ICI initiation. rwPFS and rwOS were comparable between patients with s-irAEs vs patients without s-irAEs when s-irAEs were first managed with anti-cancer treatment interruptions. Patients with s-irAEs had a 53% (22.3%-91.4%) higher risk of death than patients without s-irAEs when s-irAEs initially required corticosteroids or other immunosuppressants, and a 61% (37.9%-87.9%) higher risk of death when s-irAEs first required hospitalization or ED admission.

CONCLUSION

The impact of s-irAEs on clinical outcomes may depend on the initial intervention required to manage the adverse event. s-irAEs were associated with worse outcomes when they initially required hospital/ED admission, corticosteroids, or other immunosuppression.