Abstract Introduction: Outcomes for patients (pts) with mTNBC remain poor, with chemotherapy as a historic mainstay treatment. Atezo (anti-PD-L1) demonstrated single-agent activity in mTNBC, and initial data suggested the addition of chemotherapy was safe with encouraging activity (Adams, 2016). Here we report long-term outcomes, with OS, for atezo + nab-pac in mTNBC. Methods: Arm F of multi-cohort Ph Ib study (NCT01633970) enrolled pts with locally advanced unresectable/stage IV mTNBC and 0 to 2 prior regimens for advanced disease. Atezo 800 mg IV q2w and nab-pac 125 mg/m2 IV qw (3 weeks on/1 week off) were given concurrently (with nab-pac run-in for biopsy cohort); maintenance atezo was permitted until loss of clinical benefit. In-tumor PD-L1 expression (VENTANA SP142 assay), CD8 and stromal TILs were assessed by IHC; blood-based CD8+Ki67+ via FACS, and CXCL10 cytokine RNA by NanoString. Results: Of 33 evaluable pts, 82% had ECOG PS 1, 58% had visceral mets, and median age was 55 years (data cutoff, April 30, 2017). Grade 3-4 treatment-related AEs (TRAE) occurred in 73%, with no Grade 5 TRAEs. Atezo-related AEs ≥ 20% included fatigue, pyrexia and decreased neutrophil count. Five pts discontinued nab-pac due to typical chemotherapy-related toxicities, and 3 withdrew from atezo due to TRAEs. Responses occurred across all subgroups (Table), with numerically higher ORR and longer OS/PFS in pts with less pre-treatment and higher PD-L1 status. Serial biopsy samples (n = 14) had elevated PD-L1+ TILs after combination treatment (but not after nab-pac alone). Peripheral blood analyses found transient on-treatment increases in CD8+Ki67+ and CXCL10, which were independent of efficacy. Conclusions: In this single-arm study cohort, atezo + nab-pac was well tolerated with promising efficacy, especially for treatment-naive pts. Ongoing randomized Ph III trial IMpassion130 (NCT02425891) is investigating this regimen in untreated mTNBC. Table. EfficacySubgroup/OutcomeCurrent Line of TherapyPD-L1 IC StatusaAll Ptsb (N = 33)1L (n = 13)2L+c (n = 20)IC1/2/3 (n = 12)IC0 (n = 12)Confirmed ORR (95% CI)54% (25, 81)30% (12, 54)42% (15, 72)33% (10, 65)39% (23, 58)mDOR (range)7.8 mo (2.9-11.5+)10.9 mo (3.7-20.9+)9.1 mo (2.9-16.2)10.2 mo (3.7-20.9+)9.1 mo (2.9-20.9+)mPFS (95% CI)8.6 mo (5.2, 11.5)5.1 mo (3.3, 7.3)6.9 mo (5.2, 11.0)5.1 mo (3.5, 6.8)5.5 mo (5.1, 7.7)mOS (95% CI)24.2 mo (11.5, NE)12.4 mo (7.5, 21.9)21.9 mo (13.1, NE)11.4 mo (7.5, NE)14.7 mo (10.1, NE)IC, tumor-infiltrating immune cells; mDOR, median duration of response; mPFS, median progression-free survival; mOS, median overall survival; mTNBC, metastatic triple-negative breast cancer; NE, not estimable; ORR, objective response rate; PD-L1, programmed death-ligand 1.a Excludes 9 pts with unknown PD-L1 status.b Includes 25 pts enrolled in the serial biopsy cohort and 8 in the safety cohort. Median durations of atezolizumab exposure, safety follow-up and efficacy follow-up were 5.6 mo (range, 0-30), 6.9 mo (range, 1.7-30.3) and 24.4 mo (95% CI: 22.1, 28.8), respectively.cIncludes 1 pt who received 3 prior therapies for mTNBC.Citation Format: Paula R. Pohlmann, Jennifer R. Diamond, Erika Hamilton, Sara M. Tolaney, Wei Zhang, Koho Iizuka, Paul Foster, Luciana Molinero, Roel Funke, Sylvia Adams. Atezolizumab (atezo) + nab-paclitaxel (nab-pac) in metastatic triple-negative breast cancer (mTNBC): 2-year update from a ph Ib trial [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr CT028.