The Impact of Surveillance Imaging Frequency on the Detection of Distant Disease for Patients with Resected Stage III Melanoma

Mbathio Dieng,Sarah J Lord,Robin M Turner,Omgo E Nieweg,Alexander M Menzies,Robyn P M Saw,Andrew J Einstein,Louise Emmett,John F Thompson,Serigne N Lo,Rachael L Morton
DOI: https://doi.org/10.1245/s10434-021-11231-3
Abstract:Background: It is not known whether there is a survival benefit associated with more frequent surveillance imaging in patients with resected American Joint Committee on Cancer stage III melanoma. Objective: The aim of this study was to investigate distant disease-free survival (DDFS), melanoma-specific survival (MSS), post distant recurrence MSS (dMSS), and overall survival for patients with resected stage III melanoma undergoing regular computed tomography (CT) or positron emission tomography (PET)/CT surveillance imaging at different intervals. Patients and methods: A closely followed longitudinal cohort of patients with resected stage IIIA-D disease treated at a tertiary referral center underwent 3- to 4-monthly, 6-monthly, or 12-monthly surveillance imaging between 2000 and 2017. Survival outcomes were estimated using the Kaplan-Meier method, and log-rank tests assessed the significance of survival differences between imaging frequency groups. Results: Of 473 patients (IIIA, 19%; IIIB, 31%; IIIC, 49%; IIID, 1%) 30% underwent 3- to 4-monthly imaging, 10% underwent 6-monthly imaging, and 60% underwent 12-monthly imaging. After a median follow-up of 6.2 years, distant recurrence was recorded in 252 patients (53%), with 40% detected by surveillance CT or PET/CT, 43% detected clinically, and 17% with another imaging modality. Median DDFS was 5.1 years (95% confidence interval 3.9-6.6). Among 139 IIIC patients who developed distant disease, the median dMSS was 4.4 months shorter in those who underwent 3- to 4-monthly imaging than those who underwent 12-monthly imaging. Conclusion: Selecting patients at higher risk of distant recurrence for more frequent surveillance imaging yields a higher proportion of imaging-detected distant recurrences but is not associated with improved survival. A randomized comparison of low versus high frequency imaging is needed.
What problem does this paper attempt to address?