Abstract
Background: Real-world safety profiles of immune checkpoint inhibitors (ICIs) in older adults remain insufficiently characterized. Although ICIs are widely used across tumor types, older patients, particularly those with frailty, multimorbidity, or polypharmacy, are consistently under-represented in clinical trials, limiting the external validity of trial-derived toxicity estimates. Robust real-world data are therefore essential to clarify the incidence, seriousness, and age-related patterns of immune-related adverse events (irAEs) in routine practice. Methods: This is a nationwide retrospective study of spontaneous ICI-related ADRs reported in INFARMED's Portal RAM (2011-2024). We evaluated the frequency, seriousness, fatality, and organ-specific patterns of ICI-related adverse drug reactions (ADRs) reported to the Portuguese National Pharmacovigilance System. The analytic unit was the ADR case. Endpoints included seriousness (primary), fatality, hospitalization, time-to-onset, and System Organ Class. Multivariable logistic regression adjusted for age, sex, regimen, tumor type, polypharmacy, and calendar period; sensitivity analyses using first ADR per patient were concordant. Results: We identified 2300 eligible ICI-related ADRs (corresponding to 925 patients). Median age at the time of ADR was 65 years (IQR not reported); 33.7% occurred in adults aged ≥70 years, and 62.8% of reports involved male patients. PD-1 inhibitors accounted for 77.5% of ADRs, and monotherapy for 72.9%. Overall, 85.8% of ADRs were classified as serious; 17.9% led to hospitalization and 19.1% were fatal. Serious-event reporting was similar in older and younger adults (≥70 vs. <70 years: 84.5% vs. 86.5%, p = 0.22), and the proportion explicitly labeled immune-related did not differ (9.3% vs. 8.7%, p = 0.56). In contrast, fatal outcomes were significantly more common in older adults (25.3% vs. 16.0%; p < 0.001). Age was associated with distinct organ-specific patterns. Adults ≥ 70 years had higher odds of nervous system disorders (aOR 1.75, 95% CI 1.23-2.48) and immune system disorders (aOR 1.42, 95% CI 1.02-1.98), but lower odds of hepatobiliary (aOR 0.52, 95% CI 0.36-0.76; p = 0.001) and blood/lymphatic disorders (aOR 0.50, 95% CI 0.32-0.79). In multivariable models, age ≥ 70 years did not predict seriousness (aOR 0.98, 95% CI 0.76-1.27), whereas combination therapy remained independently associated with increased seriousness (aOR 1.57, 95% CI 1.13-2.18). Conversely, age ≥ 70 years independently predicted fatal outcomes (aOR 1.66, 95% CI 1.31-2.09). Later calendar periods (2017-2024) were associated with substantially lower fatality (aOR 0.16; 95% CI 0.10-0.27). CTLA-4-containing regimens demonstrated a tendency toward higher fatality (aOR 1.50; 95% CI 0.94-2.37). Conclusions: Chronological age does not seem to increase the likelihood of reporting a serious ICI-related ADR, but, once toxicity occurs, older adults experience higher fatality rates. Age-related phenotypic differences and regimen-specific risks highlight the need for early recognition systems and tailored toxicity management in older populations.