Surgical excision, Mohs micrographic surgery, external-beam radiotherapy, or brachytherapy for indolent skin cancer: An international meta-analysis of 58 studies with 21,000 patients.

ABSTRACT

BACKGROUND: The objective of this study was to compare the cosmesis and recurrence rates of conventional excision (CE), Mohs micrographic surgery (MMS), external-beam radiation therapy (EBRT), or brachytherapy (BT), for basal cell carcinoma and squamous cell carcinoma of the skin. METHODS: Population, Intervention, Control, Outcome, Study Design (PICOS), Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA), and Meta-Analyses of Observational Studies in Epidemiology (MOOSE) methods were used to identify studies on PubMed (from 1985 to 2018), including patients with American Joint Committee on Cancer (AJCC) T1-T2N0 basal cell carcinomas and squamous cell carcinomas and ≥10 months follow-up who received CE, MMS, EBRT, or BT. The primary endpoint was cosmesis, classified as "good," "fair," or "poor." The secondary endpoint was 1-year recurrence. Fixed-effects and random-effects meta-analyses were performed to evaluate primary and secondary outcomes with respect to treatment modality. RESULTS: In total, 18,095 studies met initial search criteria. There were 24 CE, 13 MMS, 19 EBRT, and 7 BT studies included with a total of 21,371 patients. The summary effect size for "good" cosmesis was 81% (95% CI, 70.6%-89.6%), 74.6% (95% CI, 63%-84.6%), and 97.6% (95% CI, 91.3%-100%) for CE, EBRT, and BT, respectively. Good cosmesis was 96.0% in the only MMS study that reported cosmesis. BT had improved "good" cosmesis over EBRT (P = .0025) and was similar to CE and MMS. No significant differences were seen for "fair" or "poor" cosmesis. One-year recurrence rates were low throughout at 0.8% (95% CI, 0.3%-1.6%), 0.2% (95% CI, 0%-0.6%), 2% (95% CI, 1.3%-2.7%), and 0% (95% CI, 0%-0.5%) for CE, MMS, EBRT, and BT, respectively. CONCLUSIONS: For T1-T2N0 skin cancers, BT and MMS have improved cosmesis over EBRT and CE. It is unclear whether this is because of treatment superiority or selection and reporting bias. Local control is similar among all modalities at 1 year.