tazarotene, <60 total applications of imiquimod, <5 applications per week, <12 weeks of treatment, and partial clinical clearance. The sum of a case’s risk factors was considered high-risk (>4 risk factors), intermediate (4 risk factors) or low risk (<3 risk factors). Variables of interest were based on previously published literature1,2 or epidemiological data.3 An invasive melanoma component and previously failed surgical excision were considered proxies for more aggressive pre-existing disease or undetected sites of invasive disease following excision.4 History of prior unsuccessful complete excision included patients who required multiple surgeries or developed recurrence years after an initial excision with negative margins. Since the reliability of histologic and clinical confirmation of clearance of LM has been debated,5 both were considered. Ultimately, histological clearance was not included in the survival analysis due to a lack of recurrences among partial histological responders in our series, as well as biopsy-confirmation in only 64% of cases due to patient preference. RFS was defined as time from end of treatment to the first recorded date of local recurrence. RFS was estimated using Kaplan-Meier plots. Logrank tests were performed for the comprehensive risk profile. Schoenfeld residuals were used to test the proportional-hazards assumption. As fewer than 10 local recurrences were observed, Cox Proportional Hazards regression was not performed.6,7
