Effectiveness and causes for failure of surveillance of CDKN2A-mutated melanoma families

Abstract

For more than 25 years families with an increased susceptibility to melanoma have been under surveillance at our institution. We sought to investigate the effectiveness of surveillance for CDKN2A-mutated families and causes for failure of the program in patients with more advanced tumors. In a retrospective case-control study, Breslow thickness of melanomas diagnosed in relatives enrolled in the surveillance program were compared with melanomas of unscreened index patients. We investigated the influence of mode of detection and length of surveillance interval on outcome. Surveillance melanomas (n = 226, median thickness: 0.50 mm) had a significantly lower Breslow thickness (multiplication factor: 0.61 [95% confidence interval 0.47-0.80], P < .001) than index melanomas (n = 40, median thickness: 0.98 mm). Index melanomas were more likely diagnosed with a Breslow thickness greater than 1.0 mm (odds ratio: 3.1 [95% confidence interval 1.2-8.1], P = .022). In all, 53% of surveillance melanomas were diagnosed during regular screens, 7% during patients' first screen, 20% between regular screens, and 20% in patients who were noncompliant with the surveillance schedule. The majority of surveillance melanomas (58%) were detected within 6 months after the last screen. There was no correlation between tumor thickness and the length of the screening interval for tumors diagnosed within 24 months since the last screen. The study is retrospective. Surveillance was associated with earlier detection of melanomas. Noncompliance was an important cause for failing surveillance. Shortening surveillance intervals may advance detection of tumors, but may paradoxically have little impact on prognosis.