Predicting survival and recurrence in localized melanoma: a multivariate approach

Area of Science:

• Oncology
• Dermatology
• Biostatistics

Background:

• Prognostic factors for melanoma recurrence and survival are known, but predictive models are scarce.
• Key clinical questions involve predicting long-term survival and post-remission recurrence risk.
• Existing research often lacks comprehensive predictive modeling for melanoma outcomes.

Purpose of the Study:

• To present a generalized multivariate prognostic model for melanoma survival and recurrence.
• To address clinical needs in predicting patient outcomes after melanoma diagnosis.
• To provide tools for estimating individual patient prognosis.

Main Methods:

• Developed a predictive model using a large database of 4,568 localized melanoma patients.
• Utilized detailed clinical and pathological information with long-term follow-up data.
• Analyzed prognostic factors including tumor thickness, ulceration, Clark's level, location, and sex.

Main Results:

• Tumor thickness at diagnosis emerged as the single most significant prognostic factor for all outcomes.
• Tumor ulceration, Clark's level, lesion location, and sex influenced overall survival in specific subgroups.
• Tumor thickness accurately predicted melanoma recurrence and death, even after disease-free intervals of 2, 5, or 10 years.

Conclusions:

• The developed prognostic model effectively predicts melanoma recurrence and survival.
• Tumor thickness is a robust predictor of melanoma prognosis across various time points.
• The model aids in personalized risk assessment and clinical management of melanoma patients.