BACKGROUND

The prevention of locoregional recurrence (LRR) is crucial in breast cancer, as it translates directly into reduced breast cancer-related death. Breast cancer is subclassified into distinct intrinsic biological subtypes with varying clinical outcomes.

AIMS

To perform a systematic review and network meta-analysis (NMA) to determine the rate of LRR by breast cancer molecular subtype.

METHODS

A NMA was performed as per PRISMA-NMA guidelines. Molecular subtypes were classified by St Gallen expert consensus statement (2013). Analysis was performed using R and Shiny.

RESULTS

Five studies were included including 6731 patients whose molecular subtypes were available. Overall, 47.3% (3182/6731) were Luminal A (LABC: estrogen receptor (ER) + /human epidermal growth factor receptor-2 (HER2) - /progesterone receptor (PR) + or Ki-67 < 20%), 25.5% (1719/6731) were Luminal B (LBBC: ER + /HER2 - /PR - or Ki-67 ≥ 20%), 11.2% (753/6731) were Luminal B-HER2 + (LBBC-HER2: ER + /HER2 +), 6.9% (466/6731) were HER2 + (HER2 ER - /HER2 +), and finally 9.1% (611/6731) were triple-negative breast cancer (TNBC: ER - /HER2 -). The median follow-up was 74.0 months and the overall LRR rate was 4.0% (271/6731). The LRR was 1.7% for LABC (55/3182), 5.1% for LBBC (88/1719), 6.0% for LBBC-HER2 (45/753), 6.0% for HER2 (28/466), and 7.9% for TNBC (48/611). At NMA, patients with TNBC (odds ratio (OR) 3.73, 95% confidence interval (CI) 1.80-7.74), HER2 (OR 3.24, 95% CI 1.50-6.99), LBBC-HER2 (OR 2.38, 95% CI 1.09-5.20), and LBBC (OR 2.20, 95% CI 1.07-4.50) were significantly more likely to develop LRR compared to LABC.

CONCLUSION

TNBC and HER2 subtypes are associated with the highest risk of LRR. Multidisciplinary team discussions should consider these findings to optimize locoregional control following breast cancer surgery.