The results of a retrospective multicohort study, reported in The Lancet Oncology by Wan et al, suggested that the identification of the immune-related adverse event cluster to which a patient belongs may aid in prognosticating response to immunotherapy.

“These insights can be leveraged to counsel patients on the clinical impact of their individual constellation of immune-related adverse events and ultimately develop more personalized surveillance and mitigation strategies,” the investigators commented.

Study Details

The investigators focused on 13,086 patients from the Massachusetts General Hospital, Brigham and Women's Hospital, and Dana-Farber Cancer Institute (MGBD) and 26,172 patients from an independent U.S. population–based network who received immune checkpoint inhibitors for the treatment of cancer. These discovery and validation cohorts were propensity-score matched 1:2 based on demographics, cancer, and immune checkpoint inhibitor characteristics.

The investigators applied immune-related adverse event identification rules to identify patients who did and did not experience such a toxicity. From the MGBD and population-level cohorts, 4,382 and 8,010 patients were excluded, respectively, because of diagnoses of suspected immune-related adverse events within 3 months of receipt of chemotherapy.

Immune-Related Adverse Events

The median duration of follow-up was 317 days in the MGBD cohort and 249 days in the population-level cohort. Immune-related adverse events were reported in 37.7% (n = 3,284 of 8,704) and 30.5% (n = 5,538 of 18,162) of those who were retained from these cohorts, respectively. According to the investigators, positive pairwise correlations of immune-related adverse events were common in both the MGBD and population-level cohorts.

Seven patient clusters demonstrating different development patterns of immune-related adverse events were identified: endocrine, cutaneous, respiratory, gastrointestinal, hepatic, musculoskeletal, and neurologic. At the 6-month landmark timepoint, in the MGBD cohort, the patient clusters that had predominantly endocrine (hazard ratio [HR] = 0.53, P < .0001) and cutaneous (HR = 0.61, P = .0007) immune-related adverse events seemed to experience favorable overall survival; the other clusters were found to have either unfavorable (respiratory: HR = 1.60, P = .0001) or neutral (gastrointestinal: HR = 0.86, P = .23; musculoskeletal: HR = .97, P = .78; hepatic: HR = 1.20, P = .19; neurologic: HR = 1.30, P = .074) outcomes. Similar results were documented in the population-level cohort (endocrine: HR = 0.75, P = .0078; cutaneous: HR = 0.62, P = .0007; respiratory: HR = 1.21, P = .044), except the neurologic cluster demonstrated an unfavorable, rather than a neutral, overall survival outcome (HR = 1.30, P = .013).

The investigators concluded: “We identified seven patient clusters demonstrating different development patterns of immune-related adverse events and found that, in comparison with patients without immune-related adverse events, patient clusters dominated by endocrine and cutaneous immune-related adverse events were associated with improved survival, while those dominated by respiratory and neurologic immune-related adverse events were associated with worse survival outcomes. Our analyses reached similar conclusions across both cohorts, demonstrating their robustness.”

Yevgeniy R. Semenov, MD, MA, of Harvard Medical School, Boston, is the corresponding author of The Lancet Oncology article.

Disclosure: The study was funded by the U.S. National Institutes of Health. For full disclosures of the study authors, visit thelancet.com.