Risk factors for metastatic cutaneous squamous cell carcinoma: Refinement and replication based on 2 nationwide nested case-control studies

Abstract

Risk factors for cutaneous squamous cell carcinoma (cSCC) metastasis have been investigated only in relatively small data sets. To analyze and replicate risk factors for metastatic cSCC. From English and Dutch nationwide cancer registry cohorts, metastatic cases were selected and 1:1 matched to controls. The variables were extracted from pathology reports from the National Disease Registration Servicein England. In the Netherlands, histopathologic slides from the Dutch Pathology Registrywere revised by a dermatopathologist. Model building was performed in the English data set using backward conditional logistic regression, whereas replication was performed using the Dutch data set. In addition to diameter and thickness, the following variables were significant risk factors for metastatic cSCC in the English data set (n=1774): poor differentiation (odds ratio [OR], 4.56; 95% CI, 2.99-6.94), invasion in (OR, 1.69; 95% CI, 1.05-2.71)/beyond (OR, 4.43; 95% CI, 1.98-9.90) subcutaneous fat, male sex (OR, 2.59; 95% CI, 1.70-3.96), perineural/lymphovascular invasion (OR, 2.12; 95% CI, 1.21-3.71), and facial localization (OR, 1.57; 95% CI, 1.02-2.41). Diameter and thickness showed significant nonlinear relationships with metastasis. Similar ORs were observed in the Dutch data set (n=434 cSCCs). Retrospective use of pathology reportsin the English data set. cSCC staging systems can be improved by including differentiation, clinical characteristics such as sex and tumor location, and nonlinear relationships for diameter and thickness.