Abstract

Different neoadjuvant chemotherapies are available for triple-negative breast cancer (TNBC). Here, we performed a network meta-analysis to evaluate the pathological complete response (pCR) benefit and safety of treatment regimens. Pairwise and Bayesian network meta-analyses were performed to compare direct and indirect evidence, respectively. Twenty-three studies involving 12 regimens namely standard chemotherapeutic agents, bevacizumab (B)-, platinum salts (P)-, B plus P (BP)-, poly(ADP-ribose) polymerase inhibitors (Pi)-, P plus Pi (PPi)-, capecitabine (Ca)-, gemcitabine (Ge)-, zoledronic acid (Za)-, everolimus (E)-, P plus E (PE)-, and gefitinib (G)-containing regimens. The results showed that P-, B-, PPi-, and Za-containing regimens achieved higher pCR than standard chemotherapeutic agents. BP-containing regimens had a better pCR than B-containing regimens. In indirect comparisons, Za-, BP-, P-, and B-containing regimens were the top four strategies with the highest probability for pCR. Benefit-risk analysis showed that B-containing regimens had the highest acceptability of being the best treatment for better pCR achievement with fewer SAEs. The addition of P, B, BP, PPi, and Za to standard chemotherapeutic agents enhanced the pCR, but a balance between efficacy and safety should be carefully considered. B-containing regimens might be the best choice for neoadjuvant chemotherapy due to its better efficacy and tolerability.