Mil Med 2019 05;184(5-6):e408-e416
Dermatopathology Section, Department of Pathology and Laboratory Medicine, VA Integrated Service Network (VISN-1), West Roxbury, MA.
Introduction: A recent study found that the incidence of melanoma and melanoma-related mortality was decreasing in residents of the New England region. However, it is unknown whether this trend is conserved in Veterans of New England who constitute more than 14% of the national Veteran population. Given this, our goal was to analyze the incidence of melanoma in patients of Veteran Integrated Service Network-1 (VISN-1) (geographically consisting of VA health care facilities in Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont) and to calculate an incidence rate ratio (IRR) of melanoma in VISN-1 compared to the general population. Additional goals were to ascertain the risk/susceptibility of this patient population with a view to improve quality of care and outcomes.
Materials And Methods: Data for 523 cases of melanoma [2000-2011] were obtained from the regional branch of the Veterans Affairs Central Cancer Registry (VACCR) within the geographic area comprising VISN-1. A detailed retrospective chart review was conducted on these cases to gather demographic, risk factor, and clinical practice data. Demographic and incidence data from VISN-1 were compared to the general population via data from Surveillance, Epidemiology and End Results Program (SEER) from the same time period. Person-years (PY) were calculated for both populations to measure IRRs which was further standardized for age and gender.
Results: VISN-1 patients were predominantly older (94.26% >50 years), Caucasian (99.43%) males (96.75%). Compared to the general population, VISN-1 patients experienced more invasive lesions defined as stage T1 or greater (4.33% vs. 57.12%, p < 0.001), but reduced melanoma-associated mortality (40.96% vs. 19.05%, p < 0.001) although all-cause mortality was approximately doubled (52.20% vs. 26.14%, p < 0.001). Metastatic disease-rates were similar in both [approximately 4% in both]. IRR of melanoma in VISN-1 patients was 0.36 (95% CI: 0.20-0.67; p = 0.0063) which persisted in all age groups/genders. 60.92% of VISN-1 patients had recreational sun-exposure history and 72.41% of tobacco use. 95.02% of melanomas were located in continuously/intermittently sun-exposed areas, 93.28% were surgically-treated with a median treatment delay of 31 days [range 18-48]. Median lost to follow-up was 0 day [range 0-681 days].
Conclusions: Compared to the general population, melanoma incidence was lower in the VISN-1 cohort, possibly due to decreased UV index in the New England region, protective effects of past tobacco use, improved access to care through the VA and regional public health educational efforts. Yet melanomas were more often invasive in the VISN-1 cohort due to advanced age and male sex both of which are associated with more advanced disease at diagnosis. A strength of this study is the calculation of IRR using PY as this method enhances accuracy of incidence calculations. The data were limited by the fact that the population was from one geographic region and consisted mainly of elderly Caucasian males. Descriptive variable data such as sun-protective habits and risk factors from military service are limited by potential recall bias given the retrospective study design. Further study is necessary to replicate these results and to compare our data to Veteran populations from different geographic regions within the USA.