Abstract P4-10-23: Expression of PD-L1 is independent of PIK3CA/AKT1/PTEN alterations in triple-negative breast cancer (TNBC) and is not associated with response to ipatasertib (IPAT) plus paclitaxel (PAC)

Abstract

Abstract Background: In the FAIRLANE trial (NCT02301988) in early TNBC, adding the oral AKT inhibitor IPAT to neoadjuvant PAC led to numerical increases in rates of pathologic complete response (pCR; primary endpoint) and complete response by MRI in unselected patients, with a numerically greater treatment effect in patients with PIK3CA/AKT1/PTEN-altered tumors [Oliveira et al., Ann Oncol 2019]. The IMpassion130 trial established the efficacy of first-line atezolizumab + nab-PAC in patients with PD-L1-positive advanced TNBC [Schmid et al., NEJM 2018]. Preliminary phase 1b results showed promising activity with a triplet combination of IPAT, atezolizumab, and taxane as first-line therapy for advanced TNBC across PIK3CA/AKT1/PTEN and PD-L1 biomarker subgroups [Schmid et al., AACR 2019]. Using samples from the FAIRLANE trial, PD-L1 expression was characterized to evaluate relationships between IPAT treatment effect and expression of PD-L1 and biomarkers relevant to IPAT. Methods: Pretreatment tumor samples were evaluated for genomic alterations using the FoundationOne® assay (Foundation Medicine). Pretreatment PD-L1 status was determined using the SP142 immunohistochemistry assay (VENTANA Medical Systems), with PD-L1 positivity defined as PD-L1 expression on ≥1% of tumor-infiltrating immune cells. Tumor-infiltrating lymphocytes (TILs) were quantified using the Salgado method [Salgado et al., Ann Oncol 2015]. Gene expression was measured by RNA sequencing. Results: Samples were evaluable for genomic alterations in 144 patients, TILs in 135 patients, gene expression in 111 patients, and PD-L1 expression in 99 patients. The prevalence of PD-L1 positivity was 39%, consistent with the reported prevalence in IMpassion130. Compared with PD-L1-negative tumor samples, PD-L1-positive tumor samples showed significantly higher levels of TILs (mean 37% vs 24%; p=0.004) and significantly higher expression of the genes encoding for PD-L1 (CD274: geometric mean 10.5 vs 3.9 counts per million; p<0.0001) and PD-1 (PDCD1:geometric mean 2.3 vs 0.9 counts per million; p=0.006). The prevalence of PD-L1 positivity was similar in PIK3CA/AKT1/PTEN-altered and -nonaltered tumor samples (36% vs 40%, respectively). For patients receiving IPAT + PAC, no difference was observed according to PD-L1 status for either pCR (16% in PD-L1-positive samples vs 15% in PD-L1-negative samples; p=1) or complete/partial response by MRI (75% vs 64%, respectively; p=0.5). Conclusions: PIK3CA/AKT1/PTEN alterations and PD-L1 expression are independent biomarkers in TNBC. The prevalence of PD-L1 positivity in FAIRLANE (early TNBC) is similar to that in IMpassion130 (advanced TNBC). Pretreatment PD-L1 expression is not associated with pCR or response by MRI to IPAT + PAC. Citation Format: Matthew J Wongchenko, Mafalda Oliveira, Cristina Saura, Paolo Nuciforo, Isabel Calvo, Jay Andersen, José L Passos Coelho, Miguel Gil Gil, Begoña Bermejo, Debra A Patt, Eva Ciruelos, Zhen Shi, Na Xu, Patricia Villagrasa, Aruna Mani, Steven J Isakoff. Expression of PD-L1 is independent of PIK3CA/AKT1/PTEN alterations in triple-negative breast cancer (TNBC) and is not associated with response to ipatasertib (IPAT) plus paclitaxel (PAC) [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-10-23.