The New American Joint Committee on Cancer staging system for cutaneous melanoma

Recently, a new staging system for cutaneous melanoma was proposed by the American Joint Committee on Cancer (AJCC) melanoma group. Comments and suggestions regarding this proposed system have been invited.1, 2 We published a first proposal to revise the currently used staging system in 1992 by analyzing data from the German Central Malignant Melanoma Registry (CMMR) suggesting that the classification of primary melanomas should be based solely on Breslow tumor thickness and no longer be based on a combination of Breslow tumor thickness and Clark level of invasion.3 We found that the interobserver reproducibility of Clark level of invasion was rather poor compared with Breslow tumor thickness. In addition, we proposed to change the cutoff values for tumor thickness to 1.0, 2.0, and 4 mm, respectively, and both proposals have been substantiated by an additional evaluation of a larger collective from the CMMR.4-6 We therefore appreciate that these suggestions have been included into the new proposal from the AJCC. In the German-speaking countries, we already proposed a modified staging system according to these principles in 1994.7 The new proposal of the AJCC includes some changes that we believe need to be substantiated further. We performed prognostic evaluations based on the data from the CMMR to analyze the discriminatory power of the different features included into the new staging proposal. The analysis was based on data from 6356 patients with primary melanoma with complete data available with regard to tumor thickness and histopathologic reports of ulceration from 8 departments of dermatology with appropriate follow-up; the median follow-up time was 51 months. Five hundred three of these patients died of cutaneous melanoma. We would like to make some comments. Ulceration has been found to be the second most significant prognostic factor in primary cutaneous melanoma. This finding has been confirmed by the multivariate analysis from our collective (Table 1). However, ulceration does not appear to have prognostic impact for all the different classes of tumor thickness. Ulceration is a rare event in T1 melanomas (Table 2) and demonstrates no prognostic significance in this tumor thickness class (Fig. 1). Conversely, ulceration is a frequent event in thick primary melanomas and it likewise demonstrates no prognostic significance in T4 melanomas (Fig. 1). Therefore, upstaging based on ulceration should only be taken into account for T2 and T3 melanomas. Overall survival curves for tumor thickness class and presence (+) or absence (–)of histopathologically recognizable ulceration. Significant differences in overall survival between patients with and without ulceration were found using the log-rank test for pT2 and pT3 tumors (P < 0.0001) but not for pT1 tumors (P = 0.21) or pT4 tumors (P = 0.14). Less than 10% of patients with T1 melanomas were classified as having a Clark level of invasion of IV or V. For the patients with T1 tumors, we found no significant difference with regard to survival for patients with a Clark level of invasion of II–III or IV–V (Fig. 2). Therefore, we believe that the level of invasion should likewise be removed from the T1 tumor subgroup. In patients with thin primary melanomas (pT1, ≤1.0-mm tumor thickness) no significant differences in overall survival were detectable when patients were subclassified into 2 groups with either a Clark level of invasion of II–III or a Clark level of IV–V. We agree with the creation of a subclassification of the N category based on the number of lymph nodes involved with metastases and on the differentiation between microscopic and macroscopic metastasis. However, the inclusion of ulceration of the primary tumor into the lymph node subclassification is a bit confusing and also could not be reproduced from our data material. Neither ulceration of the primary tumor nor tumor thickness of the primary tumor have been found to be of prognostic significance with regard to the stage of lymph node metastasis in our data set. Therefore, ulceration of the primary tumor should be removed from the N classification. In the new AJCC staging system, satellite and in-transit metastases without metastatic lymph nodes have been subclassified to the N2 category. These patients are assumed to have a less favorable prognosis compared with patients with lymph node metastasis involving one lymph node or microscopically involving several lymph nodes. Our data material does not demonstrate this unfavorable outcome for these patients. In our evaluation of the TNM classification from 1992 we demonstrated that satellite and in-transit metastases displayed a significantly more favorable prognosis compared with lymph node metastasis and we proposed assigning this to Stage IIIA disease in comparison to staging lymph node metastasis as Stage IIIB disease.3, 7 In our current data, we found a median survival of 72 months for patients with satellite and in-transit metastasis and a median survival of 32 months for patients with lymph node metastasis. In the new AJCC staging system, satellite and in-transit metastasis would be better assigned to the N1 category. Subclassification of the M category by site of distant metastasis generally is well accepted. However, it is doubtful as to whether the inclusion of laboratory values may improve the classification. Therefore, although possibly important, we believe it would be better if the lactate dehydrogenase value was withdrawn from the staging system proposal. Clinical staging systems are developed to provide accurate prognostic information for diagnostic therapeutic decision-making and, moreover, to facilitate communication between clinicians. Therefore, clinical staging systems must not be too complicated and should be based on a limited number of features. We suggest simplifying the AJCC staging system proposal by the points mentioned above to alleviate the classification of melanoma patients and simplify clinical communication. Claus Garbe M.D.*, Ulf Ellwanger Ph.D. , Michael Tronnier M.D. , Eva-Bettina Brocker M.D.§, Constantin E. Orfanos M.D.¶, * Section of Dermatologic Oncology, Department of Dermatology, Eberhard Karls University Tuebingen, Tuebingen, Germany, Tuebingen, Germany, Luebeck, Germany, § Wuerzburg, Germany, ¶ Berlin, Germany