The potential impact on melanoma mortality of reducing rates of suboptimal excision margins

We estimated the potential benefit of reducing rates of inadequate excision margins in the treatment of localized invasive melanoma. A computer-simulated Markov decision analytic model was created to follow until death a hypothetical cohort of 55 y old Caucasians, newly diagnosed in a community setting with localized invasive melanoma. We considered two scenarios: usual care, and a hypothetical intervention. Markov states included well without local recurrence, local recurrence, cured, and dead. Published population-based data were used for rates of optimal excision margins, local recurrence, and mortality. Two outcome measures were employed: melanoma-related mortality and life expectancy. Major assumptions included: local recurrence occurs within 10 y of diagnosis, and patients revert to general population mortality rates 10 y following melanoma excision or subsequent local recurrence. For usual care, the model estimated 8.17% melanoma-related mortality. Modeling intervention with 100% optimal excision margins reduced this rate to 6.15%, a 25% relative reduction in mortality. This increased average life expectancy by 0.437 y, which equates to approximately 11 additional years in the 4% who would not experience a local recurrence due to improved excision margins. Increasing the percentage of optimal excision margins to 80% would still yield substantial improvement, with 6.83% melanoma-related mortality, saving 0.29 life-years compared with baseline. Results were insensitive to moderate changes in the parameter values. Suboptimal excision margins may account for approximately one-fourth of all melanoma-related mortality for localized invasive melanoma. If intervention can achieve even modest adherence to optimal excision margins, it might substantially reduce mortality.