Epidemiology of Eccrine Porocarcinoma in the United States: A Study of 948 Cases

Eccrine porocarcinoma (EPC) is a rare adnexal malignancy that is thought to be characterized by an aggressive course, with metastatic rates of 20%–30% being described in the literature [1]. However, population-based studies have been limited [2] and have not examined recent incidence trends or prognostic factors impacting disease-specific survival (DSS). Here, we use data from the Surveillance, Epidemiology, and End Results (SEER-17) program (2000–2022) to further elucidate the epidemiology of EPC in the United States. Patients with histologically confirmed EPC were identified using the International Classification of Diseases for Oncology, Third Edition (ICD-O-3) code 8409/3. Age-adjusted incidence rates per million person-years were calculated at 95% confidence intervals (CIs) using the Tiwari modification, and annual percent change (APC) was calculated using the weighted least-squares method. Five and 10-year DSS and overall survival (OS) were calculated using the Kaplan–Meier method, and factors impacting DSS were assessed using multivariate Fine–Gray models, treating non–EPC-related deaths as a competing risk. We identified 948 patients with EPC from SEER-17. Patients were diagnosed at an average age of 70.2 ± 15.9 years, and 55.4% of the cohort was male. Tumors primarily occurred on extremities (48.4%), followed by the head/neck (30.4%) and trunk (19.3%). Tumors were diagnosed at an average size of 22.7 ± 23.3 mm; 71.6% of tumors were localized to the skin at the time of diagnosis, whereas 8.6% and 1.3% of tumors presented with regional and distant metastatic spread, respectively. Common surgical approaches included excision with ≤ 1 cm margins (55.5%), excision with > 1 cm (17.0%), and Mohs micrographic surgery (10.2%) (Table 1). We observed an annual age-adjusted incidence of 0.49 persons per million person-years, with a higher incidence among males (0.63) than females (0.38) and a rising incidence trend over the 22-year period (APC 2.8%). We estimated a 5-year and 10-year DSS of 95.8% and 93.6%, respectively, and a 5-year and 10-year OS of 67.8% and 50.2%, respectively. Multivariate analysis identified age > 80 years, regional/distant stage, and poorly-differentiated/undifferentiated grade as significantly associated with worse DSS (Table 2). Tumor size > 20 mm approached statistical significance as a predictor of poorer DSS (sHR 2.36, 95% CI 0.96–5.86, p = 0.06). DSS survival analysis between patients who underwent excision with > 1 cm margins versus Mohs micrographic surgery (MMS) was not conducted, as none of the 97 patients who underwent MMS experienced EPC-related death. Sex, income, metropolitan (vs. non-metropolitan) residence, and marital status were not observed to impact DSS. Our study adds a number of novel insights to the literature on EPC. Although previous studies have suggested EPC to be characterized by high rates of metastatic spread and poor DSS, we observed < 10% of EPCs to present with metastatic spread at time of diagnosis and a > 90% 5-year and 10-year DSS; notably, we observed the 5-year and 10-year OS rates to be far below their respective DSS rates, suggesting that most EPC patients die of non-EPC causes. Additionally, age > 80 years and higher tumor grade were associated with worse DSS. Additionally, we appreciated a trend towards poorer DSS for tumors > 20 mm. Taken together with a recent analysis that observed a 6.8% risk of sentinel lymph node positivity in tumors > 17 mm3, this suggests that heightened vigilance with nodal evaluation may be considered for larger tumors [3]. Finally, the rising incidence of EPC observed in this study may reflect increased ultraviolet exposure in an aging population, consistent with prior molecular studies demonstrating ultraviolet-associated mutational signatures in EPC [4]. Limitations include a lack of data on recurrence and certain tumor-specific variables (e.g., perineural invasion) not captured by SEER registries. Altogether, these findings inform risk stratification for this rare malignancy. The authors have nothing to report. The authors declare no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request.