Is It Time to Rethink What Thickness Really Means as a Prognostic Indicator in Melanoma? Ten Questions We Should Be Asking Ourselves

February 2019 | Volume 18 | Issue 2 | Original Article | 148 | Copyright © February 2019


Marc L. Frost MD,a Katherine Hrynewycz MD,b C. William Hanke MDc

aSt. Vincent Hospital, Indianapolis, IN bDepartment of Dermatology, Indiana School of Medicine, Indianapolis, IN cLaser and Skin Surgery Center of Indiana, Indianapolis, IN

Abstract
Tumor thickness has been a key tool for prognosis of melanoma. However, with the advent of gene expression profile (GEP) assays for melanoma and the discovery of multiple melanoma subtypes, it is time to reassess how we view tumor thickness as a prognostic indicator.Herein we present ten questions for consideration by the shrewd practitioner when considering prognostic factors of melanoma. J Drugs Dermatol. 2019;18(2):148-151.

INTRODUCTION

1. Why question the value of tumor thickness as a prognostic indicator for melanoma?Since Dr. Clark’s observations in the 1960’s, it has been well-accepted that melanoma thickness is a very useful predictor of metastasis and survival, leading to the general concept that thin tumors behave more favorably than thick tumors.Over the years, however, clinicians have seen too many exceptions to this simplistic statement. How do we account for melanomas as thick as 6 cm, but do not metastasize1? Likewise, how do we reconcile thin tumors (less than 1 mm Breslow thickness), or even in situ disease2 that eventuate into metastatic disease? With such observations in mind, it is prudent to consider if there are alternative indicators for melanoma prognosis.2. When a very thin melanoma behaves aggressively, isn’t it likely due to regression of what was a thick tumor, which is no longer seen on pathology? After all, melanoma is supposed to be unpredictable.Signs of regression can be seen in thin melanomas that metastasize, and it is considered to be a predictor of which thin melanomas will spread.2 But a study by Guitart et al demonstrated nearly half of the thin melanomas that metastasized had no signs of regression microscopically, which included an in situ melanoma. Tas et al demonstrated no association between regression and outcomes.3 Furthermore, regression is again absent on the list of criteria to the recent revision to the AJCC staging.4 To simply write off thin melanomas that metastasize as aconsequence of regression affecting thickness measurements is too simplistic. As the incidence of melanoma continues to increase,5 it becomes all the more imperative to investigate why some thin melanomas spread, and others do not.3. What other tumor factors can be evaluated that may be more accurate than melanoma thickness in predicting outcomes?Gene expression profile (GEP) signatures have been developed for a number of tumors, including uveal melanoma,6 breast cancer,7 glioblastoma,8 and mesothelioma9 to better predict outcomes in these malignancies. In breast cancer, GEP is being used to complement the traditional use of tumor morphology, grade, and immunophenotyping for classification and treatment.10 A GEP test has been developed by Castle Biosciences that looks at 31 genes that are up- or down-regulated in metastatic melanoma. The test, known as DecisionDx-Melanoma, divides melanomas into two classes: Class 1 tumors are low risk melanomas with an 88% 5-year recurrence free survival (RFS) rate. Class 2 tumors are melanomas with a 52% 5-year RFS rate.114. How does GEP testing of melanoma affect the viewpoint of tumor thickness as a prognostic indicator of metastasis and survival?The current thinking of melanoma is as a single disease entity with a single, linear progression leading to risk of metastasis. This progression is based on increasing Breslow thickness, so that for a given tumor that is initially thin, prognosis worsens as it becomes thicker. Our assumption is that a given melanoma becomes “worse,” or “more aggressive”, as it increases in thickness. But with GEP analysis, we now know that there are two subsets of melanoma in