HER2 FISH ratio cut-points and pathologic complete response (pCR), residual tumor (RT), HER2 status, and survival prediction in HER2-positive breast cancer (BC)

e22033 Background: We demonstrated that pCR is correlated with increasing HER2-FISH ratio, while disease-free survival (DFS) with pCR and ER-positivity in HER2-positive breast cancer treated with trastuzumab (SABCS 2008). It is known that quantitative HER2-FISH ratio correlates with ER levels and HER2-positivity imparts a higher grade in ER-positive BC. Collectively, we hypothesized that combined ER (≥10) and a HER2 ratio cut-point may subdivide HER2-positive BC into pCR predictive subtypes.Methods: Of the 80 HER2-positive (IHC3+/FISH+) BC, quantitative HER2 FISH ratio (widely spread over 1–18.3) and ER correlation was noted (r=0.34, p=0.002). Moreover, HER2 ratio (>4) correlated with higher Ki-67 (r= 0.5, p=0.01) and grade (p trend=0.05) in ER-positive subtype, inferring a biologic cut-point. Results: Of patients with stage I-IV BC treated neoadjuvantly (92% trastuzumab-based), pCR was 0% (0/13) in ER-positive-low-HER2 compared to 77% (10/13, p=0.0001), 75% (24/32, p<0.0001) and 37.5% (3/8, p=0.043) in ER-positive-high-HER2, ER-negative-high-HER2 and ER-negative-low-HER2, respectively. DFS was 100, 90, 80% and 60% (logrank-trend p<0.05) in these 4 subtypes (excluding stage IV), respectively, at a median follow-up of 38 months (range 6–72). In ER-negative subtypes, DFS was 97% and 29% (logrank p=0.0001) in patients with or without pCR; of the six with RT, 0% DFS was noted in four with HER2-negative/HER2-reduced (HER2-R) RT, compared to 100% in the two with unchanged HER2 (p=<0.05, logrank test). In ER-positive subtypes, DFS is 95% overall, and 100% in patients with RT; 7 of 10 tested RT were HER2- R. Conclusions: pCR is crucial and high in ER-negative-high-HER2 and is crucial (but low) in ER-negative-low-HER2-positive BC for improved outcome. Improved DFS is associated with high pCR in ER-positive-high-HER2 BC, but is independent of the low pCR in ER-positive-low-HER2-subtype. Thus, combined HER2 and ER offer improved prediction. In hypothesis generating analysis, HER2-R may underlie relapse in ER-negative subtypes (HER2-basal-transitional-residual), while it may be beneficial in ER-positive subtypes (luminal-B- A-transitional) by reducing HER2-pathway mediated endocrine resistance. [Table: see text]