Abstract PO4-26-07: HER2-low prevalence in early breast cancer (BC) across reference laboratories in Brazil: potential impact of immunohistochemistry (IHC) antibody assay sensitivity

Abstract

Abstract Background: Current definition of HER2-positive and HER2-low BC follows ASCO/CAP guidelines using IHC and/or in situ hybridization (ISH)-based techniques. Although there is no recommendation for a specific antibody assay, the companion diagnostic test for trastuzumab deruxtecan in the USA is Ventana PATHWAY anti-HER2 monoclonal antibody (4B5) run on Benchmark Ultra instrument. In this study, we assessed the prevalence of HER2-low BC in samples stained with Dako anti-HER2 polyclonal antibody (A0485) and analyzed with Autostainer Link 48 Agilent/Dako, commonly used in different large-volume laboratories in Brazil. Methods: Retrospective study of all early-stage BC samples without exposure to neoadjuvant therapy tested for HR plus HER2 status in Oncoclínicas Precision Medicine (Locus Lab) from 2021 to 2023 and analyzed following ASCO/CAP guidelines. Tumors with HER2 IHC score 0 were classified as HER2-zero, whereas tumors with HER2 score 1+ and those with HER2 score 2+ with ISH-negative were classified as HER2-low. To assess the impact of preanalytical factors on HER2-low positivity, we stratified samples as in-house (part of Oncoclínicas network) or external (other institutions’ FFPE preparation conditions not controlled). Results were also validated in independent IRA Lab using the same Dako anti-HER2 polyclonal antibody (A0485) on Autostainer Link 48 Agilent/Dako. Results: Out of 1,638 eligible samples from Locus Lab, 84% were classified as HR-positive/HER2 negative, 12% HER2-positive, and 4% triple-negative. In 1,386 HR-positive/HER2 negative samples, 83% were HER2-zero, and 17% (CI95% 15%-19%) were HER2-low (1+ in 79% and 2+ with ISH negative in 21%). In 70 triple-negative samples, 62% were HER2-zero and 38% (CI95% 27%-51%) were HER2-low (1+ in 44% and 2+ with ISH negative in 56%). HER2-low positivity was 17% in the subset of in-house samples (N=1.090) and 16% in external referrals (P=0.41). In the independent cohort, out of 441 samples from IRA Lab, 80% were HR-positive/HER2 negative, 12% HER2-positive and 8% triple negative. In 352 HR positive/HER2 negative samples, 80% were HER2-zero and 20% (CI95% 16%-25%) were HER2-low. In 36 triple-negative samples, 72% were HER2-zero and 28% (CI95% 14%-45%) were HER2-low. Conclusions: Different from another Brazilian cohort where the prevalence of HER2-low was 50% with monoclonal anti-HER2 antibodies (Reinert et al., SABCS 2020), our data suggests that the reduced proportion of samples with HER2-low status may be related to antibody clone sensitivity and not linked to preanalytical factors or local ancestry/cancer biology. Indeed, a recent Italian cohort also had lower than expected prevalence of HER2-low in surgical specimens using anti-HER2 Dako A0485 assay on Omnis instrument (Rossi et al., Breast Cancer Res Treat 2023). Prior studies have found differential analytical sensitivity of antibodies for HER2-low status, while HER2 score 3+ detection is equivalent (Ruschoff et al., Virchows Archiv 2022). We recommend prospective evaluation of the HER2-low status with companion-diagnostic Ventana 4B5 assay and alternative antibodies and instruments. Citation Format: Leonard M da Silva, Fernanda Orpinelli Rego, Matheus Costa e Silva, Tomas Reinert, Fernanda Koyama, Daniel Bueno da Cunha, Vagner Duarte, Emílio Pereira, Bruno Ferrari, Rodrigo Dienstmann. HER2-low prevalence in early breast cancer (BC) across reference laboratories in Brazil: potential impact of immunohistochemistry (IHC) antibody assay sensitivity [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO4-26-07.