Tolerability of immune checkpoint inhibitors for cancer treatment in frail, older patients

INTRODUCTION: Immune checkpoint inhibitors (ICIs) are increasingly used in older adults, but tolerability among frail patients remains uncertain. This study examined immune-related adverse events (irAEs) and treatment outcomes in frail and non-frail adults aged ≥65 years. MATERIALS AND METHODS: A retrospective single-centre review was conducted at Princess Alexandra Hospital (Queensland, Australia) of patients aged ≥65 who commenced at least one cycle of ICI between December 2021 and July 2023. Frailty was assessed using the Australian 58-item Frailty Index (FI) (non-frail <0.25, frail ≥0.25). The primary outcome was any irAE as defined by Common Terminology Criteria for Adverse Events (CTCAE v5). Secondary outcomes included treatment discontinuation, cycles, interruptions, and corticosteroid use. Univariable and multivariable logistic regression modelling was performed, adjusting for age, sex, cycles, tumour type, stage, concurrent therapy, and Eastern Cooperative Oncology Group Performance Status (ECOG-PS). RESULTS: Of 122 adults (mean age 75.8 years, 66.4 % male), 51 (41.8 %) were frail. Twenty patients were still receiving ICIs and had not yet completed treatment at last census, leaving 102 records with evaluable outcomes for the final modelling. Any irAE occurred in 53 non-frail adults (74.7 %) and 28 frail adults (54.9 %) (adjusted OR 0.41, 95 % CI 0.19-0.89, p = 0.02). Non-frail adults had more grade 1 events (62.0 % vs 43.1 %, p = 0.045) and skin-related events (48 % vs 27 %, p = 0.04). Grade ≥ 3 events occurred in 11.3 % non-frail vs 11.8 % frail (OR 1.10, 95 % CI 0.38-3.17). Early discontinuation was 60.6 % in non-frail vs 74.5 % in frail (OR 1.90, 95 % CI 0.86-4.19, p = 0.12). Median cycles were 10 (IQR 4-16) in non-frail vs 6 (IQR 4-11) in frail (p = 0.04). DISCUSSION: Frailty (FI ≥0.25) was associated with fewer irAEs, but frail adults received fewer cycles. These findings are exploratory and should be interpreted cautiously as differences in exposure likely confound toxicity comparisons; time-to-event analyses are warranted.