Further validation of the prognostic model for stage I malignant melanoma based on tumor progression

Prediction of long-term survival for clinical stage I malignant melanoma and guidance of therapy has long relied upon assessment of tumor thickness. This parameter is a strong, but not infallible predictor of prognosis. Clark et al. have developed a prognostic model based upon the concept of tumor progression and the evaluation of six readily assessable clinical and histologic attributes. They report the most accurate prediction of long-term survival of any prognostic method available for melanoma (100% accurate for radial-growth-phase and 84.1% accurate for vertical-growth-phase melanomas). This model was developed and validated via study of a patient population which largely resided in the northeastern portion of the United States. We report additional validation of this model using a data base of 55 patients from a different geographic location (North Carolina) and have observed virtual identity in the accuracy of predicting 8-year survival for radial-growth-phase (100%) and vertical-growth-phase (85.1%) melanomas. Twenty per cent of these cases were randomly selected and subjected to re-evaluation for prognosis with extremely good precision obtained in the prognostic prediction.