Background Currently, there are no validated data derived from prospective trials that demonstrate a benefit from anti-HER2 targeted therapy in HER2 negative advanced breast cancer (ABC) patients (pts) bearing HER2 expressing (HER2+) circulating tumor cells (CTCs). Although some retrospective studies suggest that a group of these pts could benefit from a target treatment, a substantial proportion of those will never respond. A better characterization of the HER2+ CTCs could improve the pts selection. We hypothesize that a high correlation between different HER2+ CTCs subgroups and the HER2 status on solid biopsy can predict the likelihood of response of the corresponding subgroup. Here we propose a simple algorithm to identify those HER2+ CTCs subgroups that have demonstrated a higher correlation with HER2+ positivity in the solid biopsy. Methods The IRB-approved study retrospectively analyzed blood samples at two different timepoints (before treatment and at the first evaluation) from 110 ABC pts treated at Northwestern University (Chicago, IL) between 2016 and 2019. CTCs were enumerated through CellSearch™ (Menarini Silicon Biosystems, Bologna, Italy), and characterized for HER2 expression using the CellSearch CXC Kit. HER2 expression in CTCs was defined and categorized in 4 different categories (0,1+,2+,3+) as previously reported (Riethdorf et al., 2010). Four different scores were assessed in their ability to predict a HER2 positive disease (based on metastases HER2 status). Score 0 (negative CTCs), Score 1 (1+ CTCs), Score 2 (2+ CTCs), Score 3 (3+ CTCs) and Score 4 (cHER2 ratio, defined as the sum of 2+ CTCs and 3+ CTCs divided by the total number of CTCs). The performance of these 4 different biomarkers was explored longitudinally through the Wilcoxon test. Results: Out of 110 pts, 49 showed a CTCs count ≥ 5 (stage IV aggressive). Among these pts, Score 0 was associated with a HER2 negative disease with an area under the ROC curve (AUC) of 0.14. No correlation was found between Score 1 and HER2 status (AUC 0.45). A marginally significant association was found in Score 2 and Score 3 (AUC respectively 0.67 and 0.64). A direct correlation was observed between Score 4 and Her2 status (AUC 0.73). The analysis performed on stage IV indolent patients showed a non-significant association among all 4 scores. Significant association was not found among the scores and molecular subtypes, only a numerical increase among HER2+ disease and Score 3 and Score 4 (respectively p=0.06 and p=0.1). The patients were observed longitudinally and a dynamic Score evaluation was performed. No significant changes were found among Score 1 and Score 2 at the two different timepoints (p= 0.35 and p=0.11). Both, Score 0 and Score 4 were significantly decreased among pts at the second timepoint (p=0.0019 and 0.006) Conclusions The data highlighted a strong performance of cHER2 ratio (score4) identifying a HER2 positive disease on tissue biopsy, comparable with the known ability of score0 (presence of only HER2 negative CTCs) to predict a HER2 negative tissue biopsy. Furthermore, the cHER2 ratio has proven to be dependent on the treatment administration, allowing to be considered as a predictive biomarker in future study, to gain a prospective validation.

Citation Format: Paolo D'Amico, Lorenzo Gerratana, Ami Shah, Qiang Zhang, Andrew Davis, Youbin Zhang, Saya Jacob, Elena Vagia, Amir Behdad, Carolina Reduzzi, Giuseppe Curigliano, Massimo Cristofanilli. The HER2 circulating ratio to define HER2 expressing circulating tumor cells in advanced breast cancer [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS2-15.