Abstract Background: Emerging data supports PARPi combinations, including PARPi with immune checkpoint blockade (ICB), as effective therapies in TNBC. The Adaptive Multi-Drug Treatment of Evolving Cancers (AMTEC) trial (NCT03801369) is evaluating the PARPi + ICB combination of olaparib (ola) + durvalumab (durva) in mTNBC patients (pts). Deep profiling of paired pre- and on-ola monotherapy TNBC biopsies (Bx) is key to identifying: i) predictive biomarkers to select pts who will benefit from PARPi + ICB, and ii) resistance mechanisms that inform on other rational PARPi combinations. We report on biomarker characterization of paired Bxs from 18 AMTEC pts. Methods: AMTEC pts undergo a pre-ola Bx (Bx1), start one (28-day) cycle of ola monotherapy (300 mg BID), with a repeat on-ola Bx (Bx2) before adding durva (1500 mg Q4W) to ola. Profiling of DNA, RNA and protein signals in Bx1 and Bx2 using WES, RNAseq, RPPA, and spatially resolved single cell proteomics using cycIF and mIHC was correlated with clinical outcomes to identify predictors of ola + durva sensitivity, and adaptive resistance to PARPi therapy. Results: WES/RNAseq - TNBC subtype (Bx1) was a strong predictor of response, with basal immune activated (BLIA), luminal androgen receptor (LAR), and basal immune suppressed (BLIS) subtypes associated with mPFS of 8.7, 2.5, and 1.7 months, respectively (p<0.05). MutSig3 signature in Bx1 (Yes = 7.4 mo vs. No = 2.5 mo; p<0.05), or increases in IFN signaling in Bx2 (Yes = 6.6 mo vs. No = 2.2 mo; p<0.05) were positive predictors of mPFS. RPPA - Change in PD-L1 expression on Bx2 (from Bx1) was a positive predictor (p<0.05). RAS-MAPK pathway activation in Bx1 was predictive of a poor response (p<0.05). mIHC - Two dominant immune cell groups were identified: 1) T cell enriched, and 2) hypoinflammed. On Bx1, most pts in the T cell enriched group achieved a partial response (PR) or stable disease (SD), whereas pts in the hypoinflammed group all had progressive disease (PD, p=0.04). On Bx2, all pts in the T cell enriched group were PR or SD, whereas PD pts comprised the hypoinflammed group (p=0.06). Conclusions: Findings highlight the value of paired Bxs to identify predictive biomarkers of PARPi + ICB sensitivity. Emerging resistance mechanisms justify amending AMTEC to a PARPi biomarker-driven trial evaluating ola in combination with durva, selumetinib (MEKi), or capivasertib (AKTi). Citation Format: Zahi I. Mitri, Evthokia A. Hobbs, Shaun M. Goodyear, Jeong Youn Lim, Joanna Pucilowska, Brett Johnson, Allison L. Creason, Courtney Betts, Lisa M. Coussens, Shannon McWeeney, Christopher L. Corless, Joe W. Gray, Gordon B. Mills. Biomarker-driven selection of polyADP ribose polymerase inhibitors (PARPi)-based combination therapies in patients with metastatic triple negative breast cancer (mTNBC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2149.