Nearly half of patients with resected high-risk melanoma treated with adjuvant anti–PD-1 therapy developed chronic immune-related adverse events, and more than one-third still had ongoing toxicity after a median follow-up of more than 5 years. Roma A. Kankaria, MD, and colleagues reported the findings in JAMA Network Open.
Immune checkpoint inhibitors can produce durable benefit in melanoma, but they can also cause immune-related adverse events (irAEs) that persist after treatment ends. The investigators noted that reliable ways to identify patients who are most likely to develop chronic toxicity remain limited.
Study Details
The retrospective multicenter cohort study included 303 patients with resected stage III to IV melanoma who received adjuvant nivolumab or pembrolizumab at 7 academic centers in the United States and Australia between 2015 and 2024. All patients had at least 12 months of follow-up after discontinuing anti–PD-1 therapy. Chronic irAEs were defined as events that began during treatment and persisted for at least 3 months after therapy ended.
The median age was 64 years, and median follow-up was 67 months. The investigators assessed the incidence and persistence of chronic toxicities, potential clinical risk factors, and circulating inflammatory proteins associated with chronic nonendocrine irAEs.
Key Findings
Overall, 220 patients, or 73%, developed an acute irAE, and 146 patients, or 48%, developed a chronic irAE. Chronic endocrine toxicities occurred in 20% of patients and chronic nonendocrine toxicities were reported in 31%. At the last follow-up, 102 patients, or 34% of the full cohort, still had an ongoing chronic irAE.
“Chronic irAEs were common and persistent, with most events failing to resolve even with prolonged follow-up,” the investigators wrote.
Among patients with chronic irAEs, the most common were hypothyroidism (31%), arthritis or arthralgias (17%), and dermatitis or pruritus (12%). Endocrine toxicities were particularly persistent: hypothyroidism remained ongoing in 42 of 45 affected patients, while all 16 patients with adrenal insufficiency or hypophysitis still had those toxicities at last follow-up.
The investigators noted that the near-universal persistence of endocrine irAEs may reflect permanent tissue damage. Other chronic toxicities, including rheumatologic, ocular, oral, and cutaneous events, were more likely to resolve, although some persisted for years and could affect quality of life.
Another 25 patients, or 8%, experienced chronic toxic effects outside of classical irAEs. Fatigue was the most common, affecting 20 patients, followed by orthostasis and headache.
Factors Associated With Chronic Toxicity
Among nonendocrine irAEs, those that became chronic tended to appear later during treatment than those that resolved, with a median onset of 153 days vs 99 days after anti–PD-1 therapy began (P = .006). Corticosteroid treatment was also more common among patients with chronic vs nonchronic events, at 35% vs 24% (P = .03).
The relationship between corticosteroid exposure and chronic toxicity depended on when the irAE developed. Among events with earlier onset, at the first quartile of 42 days, corticosteroid-treated irAEs had three times the adjusted odds of becoming chronic compared with events not treated with corticosteroids. That association weakened as the time to irAE onset increased. The authors noted that corticosteroid use may simply reflect more severe irAEs, which themselves may be more likely to become chronic.
The investigators also examined blood samples from a small subgroup of patients with ongoing chronic nonendocrine irAEs. Eight inflammatory proteins, including VEGFA and CCL19, were significantly elevated after adjustment for multiple testing compared with patients without chronic irAEs. The authors said the elevated cytokines may point to inflammatory pathways involved in chronic irAEs, although the findings require validation in larger cohorts.
The authors cautioned that the retrospective study relied on electronic medical record documentation, which could lead to underreporting or overreporting of chronic toxicities. Because the study was observational, it could not establish that the identified risk factors caused chronic irAEs, and patients with shorter survival were not captured.
“Living longer increasingly equates to living with, and needing management for, chronic immune checkpoint inhibitor–induced toxic effects,” the investigators wrote. They said further research should focus on identifying patients at greatest risk and developing treatments for long-lasting immune toxicity.
Douglas B. Johnson, MD, of Vanderbilt University Medical Center, is the corresponding author of the study.
DISCLOSURE: For full disclosures of the study authors, visit jamanetwork.com.
