Background: Treatment for mTNBC is limited, with chemotherapy the mainstay (bevacizumab is approved in > 80 countries). Atezolizumab (atezo; MPDL3280A), a humanized anti-PDL1 antibody, inhibits PD-L1 binding to PD-1 and B7.1 but leaves PD-L2/PD-1 binding intact. mTNBC has high levels of tumor-infiltrating immune cells, increased PD-L1 expression and high mutation rates that may generate immunogenic neoantigens, making it a good candidate for PD-L1–targeted therapy. It is hypothesized that atezo’s single-agent activity in TNBC could be enhanced with chemotherapy by exposing the immune system to high levels of tumor antigens and modulating T-cell and NK cell functions. Nab-paclitaxel (pac) has high antitumor activity that may favorably alter the immune microenvironment, and atezo + nab-pac resulted in promising activity and tolerable safety. A Phase III multicenter, randomized, double-blind, placebo-controlled trial (IMpassion130) has been initiated to evaluate atezo + nab-pac in first-line mTNBC. Methods: Patients (pts) will be randomized 1:1 to receive atezo (840 mg) or placebo on days 1 and 15 plus nab-pac (100 mg/m²) on days 1, 8 and 15 of a 28-day cycle. Stratification factors include presence of liver metastases and prior taxane therapy. Pts will have histologically documented locally advanced TNBC or mTNBC, no prior systemic therapy for advanced TNBC, ECOG PS 0–1 and measurable disease (RECIST v1.1). Pts with known CNS disease (except asymptomatic treated metastases), autoimmune disease or prior immune checkpoint blockade therapy are excluded. Coprimary endpoints include progression-free survival (PFS) and overall survival (OS). Other endpoints include response rate and duration and safety. Tumor biopsies will be obtained at baseline and progression to assess biomarkers of response and immune escape. PD-L1 expression will be centrally evaluated using the SP142 IHC assay. PFS and OS will be compared between study arms using the stratified log-rank test. Hazard ratios for progression or death will be estimated with a stratified Cox proportional hazards model. This trial will enroll 900 pts at > 270 sites globally. Clinical trial information: NCT02425891