3 OCT 2026 — For people whose melanoma has been surgically removed but is at high risk of returning, anti-PD-1 immunotherapy after surgery is a standard way to lower that risk. It works by releasing the brakes on the immune system, which can then attack healthy tissue. A study published on 2 October in JAMA Network Open looked at which side effects outlast treatment and which patients are most likely to have them.
The researchers found that nearly half of patients developed side effects that persisted at least three months after treatment ended, and identified a combination of factors linked to a threefold higher risk.
What the study looked at
The team reviewed 303 patients with stage III or IV melanoma who received anti-PD-1 therapy after surgery at seven hospitals in the United States and Australia between 2015 and 2024, according to the paper. All were followed for more than a year after stopping treatment. The authors include researchers from Vanderbilt University and Melanoma Institute Australia.
During treatment, 220 patients (73%) had immune-related side effects. In 146 patients (48%), at least one side effect became chronic, meaning it was still present three months or more after the drug was stopped.
Which side effects lasted
The most common chronic side effects were an underactive thyroid (31% of chronic cases), arthritis (17%) and skin inflammation (12%). Chronic side effects tended to start later in treatment, at a median of 153 days compared with 99 days for those that resolved, and to need corticosteroids more often.
Who was most at risk
In an exploratory analysis of non-endocrine side effects, those outside the hormone glands, timing and steroid treatment interacted. Patients whose side effect started early and was treated with steroids had three times the odds of it becoming chronic, after accounting for age, sex and whether they continued the drug.
What the blood tests showed
The researchers also measured inflammatory signalling proteins in the blood a year after treatment began. Eight were higher in patients with chronic side effects after correcting for multiple comparisons, including VEGFA and CCL19, and none were higher in patients without them. The authors present these as candidate pathways to test, not as established markers.
How this fits earlier work
The scale of the problem is not new. In 2023 a largely overlapping research group reported in JAMA Network Open that 46.2% of 318 similar patients developed chronic side effects, and that about a third of those had resolved with longer follow-up. What the new study adds is risk factors and blood markers.
Because most of these patients are expected to live for many years, the authors conclude that monitoring and managing chronic side effects must be part of their care. The study is retrospective, so it shows association, not proof that steroids or timing cause side effects to persist.
View author profile → · Editorial policy
