HER2-Ultralow: Prevalence, Characteristics, and Treatment Choices Among Advanced Breast Cancer Patients With Tumors...

Historically, HER2 status in invasive breast cancer has been categorized as HER2-positive (IHC 3+, IHC 2+/ISH+) or HER2-negative (IHC 0, IHC 1+, IHC 2+/ISH-). Patients meeting IHC 0 with membrane staining (HER2-ultralow) criteria may benefit from HER2-targeted therapies such as trastuzumab deruxtecan. This cohort study assessed the prevalence of HER2-ultralow expression by re-scoring HER2 IHC slides using Mayo Clinic electronic health record data. Three hundred patients with advanced breast cancer (Stages III-IV) and documented HER2 IHC 0 status (January 2017-March 2023) were identified. One slide per patient was digitized and independently re-scored by two Mayo Clinic pathologists following the 2023 ASCO-CAP guidelines, including tumor staining percentage to denote HER2-ultralow status. A sensitivity analysis was performed by a third independent pathologist. The re-scored patients had a mean age of 57.7 years (SD = 13.6). Most samples (95%, n = 285) remained scored IHC 0 by at least one pathologist; 60% of these met HER2-ultralow criteria per at least one pathologist. HER2-ultralow prevalence ranged from 43% to 45% per pathologist, with 57% overall interpathologist concordance. Samples with no observable staining comprised most of the concordant cases. Treatment patterns were similar between HER2-ultralow and no-staining groups; however, time to treatment failure (TTF) varied between groups across lines of therapy (LOT). In HR- positive cohorts, median TTF for LOT1 was 7.73 months in patients with HER2-ultralow versus 9.43 months with no observable IHC staining. In HR- negative cohorts, median TTF was 5.00 months for HER2-ultralow versus 3.17 months with no observable IHC staining. Similar TTF trends were observed in LOT2-LOT3. Approximately three in five samples originally classified as IHC 0 met HER2-ultralow criteria, suggesting many patients may benefit from HER2-directed therapy if reclassified. These findings highlight potential challenges of identifying HER2-ultralow expression and suggest the need for enhanced pathologist training, adherence to best practices, and integration of digital pathology and artificial intelligence solutions. Trial Registration: ClinicalTrials.gov_identifier: NCT03734029.