Low versus null HER2 tumour expression in "HER2-negative" breast cancer: long-term outcomes based on phenotypes

Two new categories of breast cancer (BC) have been proposed among HER2-negative patients: HER2 0 + and HER2-low breast cancer. We combined these two categories with Perou's classification. We aimed to identify potential differences in clinicopathological features and prognosis using a new, unofficial classification: Luminal A HER2 0 + , Luminal A HER2-Low, Luminal B HER2 0 + , Luminal B HER2- Low, Triple Negative HER2 0 + , and Triple Negative HER2-Low.

We conducted a retrospective analysis of our database from January 1, 2005, to December 31, 2018. Cox Regression served as the basis for our study.

We identified 1704 BC tumor cases from 1,639 HER2-negative patients (65 had bilateral BC). Among these, 608 cases were HER2 0 + , and 1096 were HER2-Low (aggregate). The median follow-up period was 120 months after surgery. None of the patients received anti-HER2 therapy. Case distribution was as follows: Luminal A HER2 0 + : 259. Luminal A HER2-low: 501. Luminal B HER2 0 + : 219. Luminal B HER2-low: 499. Triple-Negative HER2 0 + : 130. Triple-Negative HER2-low: 96. There was a 12.2% excess in the Triple Negative rate in the HER2 0 + group, compared with the HER2 Low group (aggregate), which was highly significant (Chisquare, p < 0.01). Although the HER2 0 + versus HER2 Low Hazard Ratio (HR) for Specific Mortality was of borderline significance:1.39 (IC 1.00-1.92, p = 0.049), the TN imbalance complicated a direct comparison between the two groups. After stratification using the noncanonical classification, the HR was highly significant, but only for the Luminal A subtype: 2.28 (IC 1.09-4.36, p = 0.028).

In the noncanonical classification, the effect of the unbalanced Triple-Negative proportions disappeared, and a significant finding emerged: HER2 0 + status had a negative prognostic influence exclusively in Luminal A patients.