Date of Award

8-6-2009

Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)

First Advisor

Alexa B. Kimball, MD

Second Advisor

Robert Tigelaar, MD

Abstract

TEMPORAL EVOLUTION IN THE HISTOPATHOLOGIC DIAGNOSIS OF BORDERLINE MELANOCYTIC LESIONS. Jason E. Frangos, Lyn Duncan and Alexa B. Kimball. Departments of Dermatology and Dermatopathology, Massachusetts General Hospital, Harvard Medical School, Boston MA. (Sponsored by Dr. Robert Tigelaar, Department of Dermatology, Yale University School of Medicine). While the incidence of cutaneous malignant melanoma has risen steeply over the past half century, increases in the mortality rate have been relatively modest. In an effort to understand this discrepancy, we sought to determine whether a shift toward more malignant diagnoses may have been made by dermatopathologists (DPs) diagnosing severely dysplastic nevi over a time period of 20 years. Forty biopsy slides of dysplastic nevi (28) and thin melanomas (12) from the period 1988-1990 were obtained from the pathology files of the Massachusetts General Hospital (MGH). All DPs that had rendered an original diagnosis for any of the 40 slides as well as the current staff in the MGH Dermatopathology department were invited to re-evaluate the slide-set. Three original DPs and 3 current MGH staff DPs re-read the slide-set. The mean number of melanoma diagnoses by the 6 study participants was 19.7 (median=19.5), an increase of 64% from the original number of melanoma diagnoses in the slide set (12). For lesions originally diagnosed as Melanoma, study participants had a high level of agreement between each other (kappa=0.74) and between each rater and the original diagnosing DP (kappa=0.86). For lesions originally diagnosed as Not Melanoma study participants had a low level of agreement between each other (kappa=0.22) and a low level of agreement between each rater and the original diagnosing DP (mean kappa=0.39). The results of this study indicate that a small set of DPs at a major academic institution tended to read prior non-malignant diagnoses of borderline melanocytic lesions as malignant but not to revise prior diagnoses of malignant melanoma as benign.

Share

COinS