Immediate and delayed autologous abdominal microvascular flap breast reconstruction in patients receiving adjuvant, neoadjuvant or no radiotherapy: a meta‐analysis of clinical and quality‐of‐life outcomes

BACKGROUND: Effects of postmastectomy radiotherapy (PMRT) on autologous breast reconstruction (BRR) are controversial regarding surgical complications, cosmetic appearance and quality of life (QOL). This systematic review evaluated these outcomes after abdominal free flap reconstruction in patients undergoing postoperative adjuvant radiotherapy (PMRT), preoperative radiotherapy (neoadjuvant radiotherapy) and no radiotherapy, aiming to establish evidence‐based optimal timings for radiotherapy and BRR to guide contemporary management. METHODS: The study was registered on PROSPERO (CRD42017077945). Embase, MEDLINE, Google Scholar, CENTRAL, Science Citation Index and ClinicalTrials.gov were searched (January 2000 to August 2018). Study quality and risk of bias were assessed using GRADE and Cochrane's ROBINS‐I respectively. RESULTS: Some 12 studies were identified, involving 1756 patients (350 PMRT, 683 no radiotherapy and 723 neoadjuvant radiotherapy), with a mean follow‐up of 27·1 (range 12·0–54·0) months for those having PMRT, 16·8 (1·0–50·3) months for neoadjuvant radiotherapy, and 18·3 (1·0–48·7) months for no radiotherapy. Three prospective and nine retrospective cohorts were included. There were no randomized studies. Five comparative radiotherapy studies evaluated PMRT and four assessed neoadjuvant radiotherapy. Studies were of low quality, with moderate to serious risk of bias. Severe complications were similar between the groups: PMRT versus no radiotherapy (92 versus 141 patients respectively; odds ratio (OR) 2·35, 95 per cent c.i. 0·63 to 8·81, P = 0·200); neoadjuvant radiotherapy versus no radiotherapy (180 versus 392 patients; OR 1·24, 0·76 to 2·04, P = 0·390); and combined PMRT plus neoadjuvant radiotherapy versus no radiotherapy (272 versus 453 patients; OR 1·38, 0·83 to 2·32, P = 0·220). QOL and cosmetic studies used inconsistent methodologies. CONCLUSION: Evidence is conflicting and study quality was poor, limiting recommendations for the timing of autologous BRR and radiotherapy. The impact of PMRT and neoadjuvant radiotherapy appeared to be similar.