OBJECTIVES: For patients with melanoma, ulceration is an important prognostic marker and interestingly also a predictive marker for the response of adjuvant interferon. A consensual definition and accurate assessment of ulceration are therefore crucial for proper staging and clinical management. We evaluated the prognostic impact of the extent and type of ulceration and the epidermal involvement theoretically preceding it (consumption of epidermis and cleft formation) or seen subsequent to the inflammation (reepithelialization and reactive epidermal hyperplasia), aiming for better prognostic stratification of ulcerated lesions. METHODS: From H&E-stained sections, the status (presence vs absence), extent (percentage of the total tumor length), and type (infiltrative vs attenuative) of ulceration and epidermal involvement were evaluated from 385 patients with cutaneous melanoma. RESULTS: The presence of ulceration (hazard ratio [HR], 1.83), an attenuative type of ulceration (HR, 3.02), and excessive ulceration (HR, 3.57) were independent predictors of poor melanoma-specific survival. Further subdivision of minimal/moderate ulceration showed independent prognostic value only for lesions with epidermal involvement of the surrounding epidermis (HR, 1.78). CONCLUSION: The extent and type of ulceration and involvement of the surrounding epidermis provided more accurate prognostic information than the mere absence or presence and may be useful markers allowing better stratification of ulcerated lesions.
American Journal of Clinical Pathology, 2014, Vol 142, Issue 6, p. 845-856