Abstract P2-02-01: Stage IV stratification using circulating tumor cells (CTCs) enumeration modeling: A retrospective analysis of the MONARCH 2 study

Abstract

Abstract Background: Metastatic breast cancer (MBC) clinical outcome is associated with heterogeneity in site, time of recurrence, treatment response, and disease outcome that is partially subtype dependent. The enumeration of CTCs can independently stratify Stage IV disease. We previously showed the feasibility of predicting CTCs stratification (pred_CTCs) through machine learning. The MONARCH 2 study (M2) showed the efficacy of abemaciclib (Abema), a cyclin dependent kinase 4 & 6 inhibitor (CDK4/6i), and fulvestrant in hormone-receptor positive, HER2 negative (luminal-like) MBC. The aim of this analysis was to assess the potential utility of CTCs stratification in luminal-like MBC treated with CDK4/6i utilizing a dataset from a randomized phase 3 study. Methods: Pred_CTCs were computed in the M2 (NCT02107703) randomized, phase 3 trial through a K nearest neighbor (KNN) algorithm trained on a 2,436 MBC patients pooled dataset from EPAC (European Pooled Analysis of CTCs) and the MD Anderson Cancer Center to identify patients likely to have CTCs ≥ 5/7.5ml blood (pred_Stage IVaggressive vs pred_Stage IVindolent) (Cristofanilli et al. 2019). The association of Pred_CTCs with progression-free survival (PFS) and overall survival (OS) in the M2 trial was explored using a univariate analysis and then a stepwise Cox regression to evaluate its value in the presence of prognostic factors previously identified in the MONARCH2 and 3 cohorts (i.e., ECOG performance status (PS), tumor grade, progesterone receptor status, liver and bone only involvement) (Di Leo A et al. 2018). The differential impact of Abema on OS was then explored through analysis of subgroups. Results: M2 enrolled 669 women with endocrine resistant luminal-like MBC, 644 were eligible for the KNN model. Patients classified as pred_Stage IVaggressive and pred_Stage IVindolent were 183 (28%) and 461 (72%), respectively. After univariable analysis, the prognostic impact of pred_CTCs was observed for both PFS and OS (HR=1.39 95%CI 1.14 - 1.69, P = 0.001 and HR=1.67, 95%CI 1.33 - 2.10 P < 0.001, respectively). Median OS was 48 and 32 months for pred_Stage IVindolent and pred_Stage IVaggressive, respectively. After stepwise Cox regression, variables associated with PFS were pred_CTCs, ECOG PS, liver and bone only disease (see Table 1). Features associated with OS were pred_CTCs, ECOG PS, and bone only disease. Subgroup analysis of Abema treatment effect on OS showed HR=0.68 (95%CI 0.52 - 0.89) and HR=0.89 (95%CI 0.60 - 1.35) in the pred_Stage IVindolent and pred_Stage IVaggressive subgroups, respectively Conclusions: This study represents the first analysis of clinical outcome using predicted modeling of CTCs for stage IV disease stratification. These hypothesis-generating results illustrate the need to expand resistance biomarkers evaluation in combination with CTCs stratification for improved biomarker-driven drug development. PFSHR95%CIP valuepred_CTCStage IVindolent1Stage IVaggressive1.321.02 - 1.710.034ECOG PS0111.441.17 - 1.770.001Bone onlyNo1Yes0.710.54 - 0.930.014LiverNo1Yes1.371.04 - 1.790.024OSpred_CTCStage IVindolent1Stage IVaggressive1.671.2 - 2.18< 0.001ECOG PS0111.781.39 - 2.30< 0.001Bone onlyNo1Yes0.590.43 - 0.800.001 Citation Format: Lorenzo Gerratana, Masha Kocherginsky, Andrew A Davis, Paolo D’Amico, Carolina Reduzzi, Qiang Zhang, Fabio Puglisi, Massimo Cristofanilli. Stage IV stratification using circulating tumor cells (CTCs) enumeration modeling: A retrospective analysis of the MONARCH 2 study [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P2-02-01.