All Cause Mortality in Patients With Basal and Cutaneous Squamous Cell Carcinoma: A Systematic Review and Meta-Analysis

Abstract

Purpose: The mortality risk associated with a history of KC is unclear. We conducted a systematic review and meta-analysis to assess the relative mortality in patients with a history of KC. Methods: PubMed, Web of Science, and Embase databases were searched to March 2015. A random-effects model meta-analysis was employed to calculate summary estimate of the relative mortality risk in patients with a history of squamous cell carcinoma and basal cell carcinoma. Results: The summary estimates from 3 studies and our own Skin Cancer cohort showed a 1.00 relative mortality rate among those with a history of basal cell carcinoma (95%CI 0.91-1.10) and a relative mortality of 1.35 (95%CI 1.21-1.49) among those with a history of SCC. A χ2 test for heterogeneity was significant for the BCC estimate (p=0.001) and non-significant for the SCC estimates (0.094). I2 statistics were 53.1% and 80.8% for SCC and BCC, respectively. A sensitivity analyses suggest that there is no statistically significant difference in the relative mortality of patients with a history of BCC versus controls whereas patients with a history of SCC have an increase risk of all-cause mortality compared to controls. Limitations: Studies that reported relative survival could not be included in the pooled analysis and as such the small number of studies reporting estimates that could be included in the quantitative analysis is the main limitation of this project. Another limitation was that the overall estimates showed moderate to high heterogeneity. Conclusion: The relative mortality patters between SCC and BCC are quite different. We interpret our meta-analysis to show an increased risk of mortality in individuals with SCC and no difference in relative mortality in patients with a history of BCC versus controls, despite the moderately high heterogeneity measures given that studies consistently show the same direction of effect in both the overall pooled analysis and the analyses by subgroups. Better cohort and population-based studies are required to confirm these findings.