The annual risk of melanoma progression. Implications for the concept of cure.

BACKGROUND: Melanoma may remain clinically dormant for years, and patients may have distant metastatic disease decades after the initial diagnosis is made. Because of this potential for late recurrence, the concept of "cure" for melanoma is not particularly meaningful.
METHODS: To understand better the risks of future disease as a function of time elapsed after diagnosis, the clinical course of melanoma was reviewed in 5838 patients. Using conditional probability methods, the risk of recurrent disease and the risk of death were determined for 1-year and 5-year intervals during the first 15 years of follow-up.
RESULTS: The estimated 5-year risk of recurrence declined from 44% at the time of diagnosis to 21% after 6 years. The 5-year risk of mortality decreased from 26% after 1 year to 16% after 9 years. Among patients with recurrent or metastatic disease, the annual risk of mortality was approximately 20% per year for 3 years; thereafter, the risk declined markedly. Among patients with thick primary lesions, the greatest risk was during the first few years after diagnosis, but in patients with thin lesions, the risk was distributed evenly over 15 years and did not decrease with time.
CONCLUSIONS: Conditional probability methods permit estimation of future risks to address questions frequently asked by patients with cancer who want to know when they can be considered cured of cancer or when the risk of recurrent disease has decreased. These data on the future risk of recurrent disease and mortality can give a patient meaningful information on which to base life decisions.