Portal de Periódicos da CAPES

Artigo Acesso aberto Produção Nacional Revisado por pares

BIBTEX RIS

LBA11 IMpassion031: Results from a phase III study of neoadjuvant (neoadj) atezolizumab + chemotherapy in early triple-negative breast cancer (TNBC)

2020; Elsevier BV; Volume: 31; Linguagem: Inglês

10.1016/j.annonc.2020.08.2239

1569-8041

Nadia Harbeck, H. Zhang, Carlos H. Barrios, Shigehira Saji, Kyung Hae Jung, Roberto Hegg, Andreas Koehler, Joohyuk Sohn, Hiroji Iwata, Melinda L. Telli, Cristiano Ferrario, Kevin Punie, Frédérique Penault‐Llorca, Shilpen Patel, Anh Nguyen Duc, Mario Liste-Hermoso, Vidya Maiya, Luciana Molinero, S.Y. Chui, EA Mittendorf,

Monoclonal and Polyclonal Antibodies Research

Preferred neoadj regimens for patients (pts) with early-stage (e)TNBC include taxane + anthracycline–based therapy. IMpassion031 is a global, phase III, multicentre, double-blind, randomised, placebo-controlled study in pts with high-risk primary invasive eTNBC evaluating the efficacy and safety of neoadj atezolizumab (A) or placebo (P) with nab-paclitaxel (nP) followed by A or P with dose-dense doxorubicin + cyclophosphamide. Here, we report the primary analysis of IMpassion031. Eligible pts were ≥ 18 years old with previously untreated, centrally confirmed, invasive stage II-III eTNBC and tumour size > 2 cm. Pts (n = 333) were randomized 1:1 to receive A 840 mg or P q2w + nP 125 mg/m2 qw for 6 doses of A followed by A 840 mg or P q2w + doxorubicin 60 mg/m2 + cyclophosphamide 600 mg/m2 q2w for 4 doses of A followed by surgery. After surgery, pathologic complete response (pCR; ypT0/isN0) was assessed in all pts and investigators were unblinded to study treatment. Stratification was by diagnosis stage (II vs III) and PD-L1 expression on tumour-infiltrating immune cells (IC; ≥ 1% vs < 1%). Co-primary endpoints were locally assessed pCR in ITT or PD-L1+ (PD-L1 IC ≥ 1%) pts. Event-free survival (EFS) was a secondary endpoint. Safety was assessed. 333 pts were assigned to A-chemo (n = 165) or P-chemo (n = 168). Median follow-up was 20.6 mos in the A-chemo arm and 19.8 mos in the P-chemo arm (data cutoff 3 Apr 2020). pCR was seen in 57.6% (95% CI: 49.7, 65.2) of pts in the A-chemo arm and 41.1% (33.6, 48.9) in the P-chemo arm (Δ16.5%; 5.9, 27.1; 1-sided P = 0.0044 [significance boundary, 0.0184], P = 0.0085 for the intersection hypothesis of ITT and PD-L1+ populations). In PD-L1+ pts (n=152), pCR was seen in 68.8% (57.3, 78.9) vs 49.3% (37.6, 61.1) of pts (Δ19.5%; 4.2, 34.8; 1-sided P = 0.021; not significant). Median EFS was not reached in either arm (HR, 0.76; 95% CI: 0.40, 1.44). In the neoadj phase, Grade 3/4 adverse events (AEs) were balanced, treatment-related serious AEs occurred in 22.6% (A-chemo) and 15.6% (P-chemo), and 1 pt per arm had an unrelated Grade 5 AE. In pts with eTNBC, A + neoadj chemo significantly improved pCR rates regardless of PD-L1 status with an acceptable safety profile.