Multivariate analysis of the relationship between survival and the microstage of primary melanoma by Clark level and Breslow thickness.

Abstract

The American Joint Committee on Cancer (AJCC) uses both Breslow thickness and Clark level in its staging system for malignant melanoma. Stage I corresponds to Breslow thicknesses less than 1.5 mm and Clark levels II and III. Stage II corresponds to Breslow thicknesses of at least 1.5 mm and Clark levels IV and V. However, most investigators have found Clark level to be of no prognostic significance once Breslow thickness has been taken into consideration by multivariate analysis. The authors examined the prognostic significance of Clark level by studying patients in the large database of the John Wayne Cancer Institute. Among 5575 patients with melanoma seen during the past 20 years, complete data on microstaging by both Clark and Breslow methods were available for 3323 patients. The 5-year survival rates were as follows: Clark II, 95%; III, 81%; IV, 68%; V, 47%. The Breslow thicknesses were as follows: < 0.75 mm, 95%; 0.75-1.49 mm, 85%; 1.5-3.99 mm, 66%; > or = 4.0 mm, 46%. By univariate analysis, both Clark level and Breslow thickness were highly significant prognostic indicators (P < 0.0001). By multivariate analysis, Breslow thickness remained significant (P < 0.0001). However, even when Breslow thickness was included in the model, Clark level also remained highly significant (P < 0.0015). Decisions regarding therapy for patients with clinical Stage I melanoma should consider both Clark level and Breslow thickness of the primary lesion. When there is a discordance between the two methods of microstaging, the AJCC stage should be amended to reflect the least favorable of the two prognostic indicators.