Lymphovascular invasion, regression novel prognostic factors in thin melanomas

Mark C. Kelley, MD, MMHC, FACS

The study by Maurichi and colleagues provides a detailed analysis of the prognostic factors for outcome of patients with thin melanoma. More than 76,000 cases of melanoma will be diagnosed in the United States this year, and the vast majority of these will be thin melanomas. Although most of these patients will be cured of their disease with surgical treatment, up to 10% will relapse within 10 years.
Breslow thickness, ulceration, mitotic rate and lymph node status previously have been shown to be strong and independent predictors of recurrence and survival, and they are included in the current American Joint Committee on Cancer (AJCC) staging system. Lymphovascular invasion, regression, tumor infiltrating lymphocytes, and patient factors such as age and anatomic site have been associated with outcome in some studies. The impact of these factors is not consistent, and it is of less magnitude than these other variables. Many analyses do not include a large number of thin melanomas or do not have sufficient long-term follow-up to capture enough events to identify the factors associated with relapse and survival. In those that do, Breslow thickness, ulceration and mitotic rate are identified as key prognostic factors.
This study examined the outcome of 2,243 patients with thin melanomas treated at six regional referral centers in Italy. The large size of the group, long-term follow-up (median, 124 months), routine review of the primary pathology by a small group of experienced dermatopathologists, and the prospective nature of the data collection lend strength to the findings.
The authors confirm age, Breslow thickness, ulceration and mitotic rate to be predictive of regional and distant metastasis and OS. They also identify lymphovascular invasion and extensive (>50%) regression as predictors of regional and distant relapse and OS. These factors also were independent predictors of OS in a Cox regression model that included sentinel node status. The impact of these variables upon sentinel node status also was examined in the subset of 794 patients who underwent sentinel node biopsy. Multivariable analysis confirmed Breslow thickness, ulceration, mitotic rate and lymphovascular invasion, but not regression, to be independent predictors of sentinel node status. The authors used these data to construct a prognostic model for OS of patients with thin melanomas, and propose revision of the AJCC staging system for melanoma to include lymphovascular invasion and extensive regression.
Should these findings alter clinical management of patients with thin melanomas? Although they are intriguing, they must be confirmed in an independent data set before we consider incorporating them into the AJCC staging system. Lymphovascular invasion and extensive regression can be added to the factors that clinicians consider when estimating recurrence risk in thin melanomas. Lymphovascular invasion may be added to the variables used to select patients with thin melanomas for sentinel node biopsy, particularly for lesions between 0.75 and 1 mm in thickness. It also will refine our ability to identify patients at very low risk for metastasis who may safely avoid sentinel node biopsy, thereby reducing morbidity and health care costs.
Although these findings are an important addition to our understanding of melanoma, the applicability in other populations is unclear, and may be influenced by the ability of pathologists with varying levels of experience to correctly identify and report lymphovascular invasion and extensive regression. Accurate prognostic information may help determine the frequency or intensity of surveillance, but it will be of greatest benefit when adjuvant therapies are available that may improve OS.