Abstract P5-02-48: Identifying immune-related predictive factors for paclitaxel + bevacizumab therapy in patients with HER2-negative advanced breast cancer- A multicenter retrospective study

Abstract

Abstract Background: In IMpassion 130 trial, programmed death-ligand 1 (PD-L1) expression was seen in about 40% of participated patients with triple negative advanced breast cancer (TN-ABC) [Schmid, 2018, N Engl J Med], however, actual situation of PD-L1 expression in patients HER2-negative ABC patients including with TN-ABC has not been well studied. Furthermore, no biomarker related to chemotherapy for HER2-negative ABC patients except for PD-L1 in TN-ABC has been identified. Recently, several reports regarding the relationship between peripheral immune-related markers; such as absolute lymphocyte count (ALC) or neutrophil-to-lymphocyte ratio (NLR), and efficacy of eribulin therapy [Miyoshi, 2020, Breast Cancer; Watanabe, 2020, Breast Cancer Res Treat] or paclitaxel plus bevacizumab (PB) therapy [Nakamoto, 2021, Sci Rep], however, the relationship between the efficacy of PB therapy and peripheral immune related markers including dynamic change during the therapy or local immune-related markers is unclear. Therefore, we conducted multi-institutional, retrospective study 1) to evaluate the actual situation of PD-L1 expression and other immune-related markers of the primary site by central review, and 2) to explore biomarkers for first-line PB therapy using peripheral and local immune-related markers. Patients and methods: We retrospectively reviewed medical records of HER2-negative ABC patients who received PB therapy as first-line (1L) or second-line chemotherapy for ABC. Clinical data including ALC, NLR and serum albumin (Alb) were extracted from medical records, and the pathology of archived tissues of primary and metastatic site (if available) were centrally reviewed including PD-L1 (VentanaR SP142). Statistical analyses were performed using Kaplan-Meier method, log-rank test, Wilcoxon’s test, and Cox hazard model. Mixed-effects model for repeated measures (MMRM) to evaluate the relationships between dynamic change in immune-related markers and time-to treatment termination (TTT) of PB therapy. Results: We identified 156 HER2-negaive ABC patients who underwent PB therapy, and 114 out of 156 patients were eligible for analyses. Of 114 patients, 63 patients (55.3%) had recurrent disease, and 65 (57.0%) patients had visceral disease. Eighty-seven out of 114 (76.3%) patients received PB therapy as 1L chemotherapy. Eighty-four specimens (73.7%) were diagnosed as estrogen-receptor (ER) positive, and PD-L1 positivity rate were 3.6% (1/84) in ER+ subgroup and 30.0% (6/20) in ER- subgroup, respectively. Paired biopsy specimens were eligible in 14 patients, and significant elevation of Ki67 labeling index was noted. In patients who received 1L PB (n = 87), there was no positive relationship between maintained ALC (>1,500 or >1,000/μL at baseline) or low NLR (< 2.5 or < 3) at the initiation of 1L PB therapy and TTTs. However, low NLR (cut-off at 2.5 and 3) at the initiation of second cycle of PB therapy reduced the risk for treatment-termination as; hazard ratio [HR], 0.427; 95% CI 0.218-0.843; P = 0.0147 and HR, 0.344; 95% CI 0.170-0.731; P = 0.0066, respectively. PD-L1 positive patients (n = 5) showed numerically increased risk of treatment-termination (HR, 2.68; 95% CI, 0.922-6.196; P = 0.0674) for 1L PB therapy, however, there was no significant difference in mortality risk regarding PD-L1 statuses. Multivariate analysis using MMRM disclosed that increase of Alb level was predictive factor for 1L PB therapy (HR, 0.41; 95%CI, 0.18-0.93; P = 0.0338). Conclusion: According to our real-world study, 1) PD-L1 positive rate was lower than that of previous reports, 2) low NLR at the initiation of second cycle of PB therapy and dynamic change in albumin level were identified as predictive factors and 3) PD-L1 overexpression was not a prognostic or a predictive factor for PB therapy in patients with HER2-negative ABC. Funding: Chugai Pharmaceutical CO., LTD. Citation Format: Junichiro Watanabe, Takashi Sugino, Koji Muramatsu, Satoshi Morita, Yu Hidaka, Shogo Nakamoto, Mitsuya Ito, Shoichiro Ohtani, Masahiko Ikeda. Identifying immune-related predictive factors for paclitaxel + bevacizumab therapy in patients with HER2-negative advanced breast cancer- A multicenter retrospective study [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P5-02-48.