Background: The optimal duration of immune checkpoint blockade (ICB) therapy is not well established. Active residual disease is considered prohibitive for treatment discontinuation and its detection by diagnostic CT imaging is limited. Here, we set out to determine the potential added value of 2-[18F]fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET) to identify patients at higher risk of relapse following discontinuation of ICB in advanced melanoma.

Methods: Metastatic melanoma patients who discontinued ICB were identified retrospectively. Eligible patients received FDG-PET and diagnostic CT within four months of ICB discontinuation. We defined morphologic response using RECIST v1.1. Complete metabolic response (CMR) was defined as uptake in tumor lesions below background, whereas any site of residual, FDG-avid disease was rated as non-CMR. The primary endpoint was time to progression (TTP) after therapy discontinuation stratified by morphologic and metabolic imaging response using Kaplan–Meier estimates and log-rank test.

Results: Thiry-eight patients were eligible for this analysis. Median follow-up was 37.3 months since ICB discontinuation. Median TTP in the overall cohort was not reached. A greater proportion of patients were rated as CMR in PET (n = 34, 89.5%) as compared to complete response (CR) in CT (n = 13, 34.2%). Median TTP was reached in patients with non-CMR (12.7 months, 95%CI 4.4-not reached) but not for patients with CMR (log-rank: p < 0.001). All patients with complete response by CT had CMR by PET. In a subset of patients excluding those with complete response by CT, TTP remained significantly different between CMR and non-CMR (log-rank: p < 0.001).

Conclusion: Additional FDG-PET at time of discontinuation of ICB therapy helps identify melanoma patients with a low risk of recurrence and favourable prognosis compared to CT imaging alone. Results may have clinical relevance especially for patients with residual tumor burden.