Abstract T-DXd has been approved by the FDA to treat patients with metastatic HER2-low and -positive breast cancer. The utility of current HER2 immunohistochemistry (IHC) assays in evaluating HER2-low tumors is not clear. A simple and objective method to evaluate HER2 expression in breast cancer is urgently needed. RNAScope can detect HER2 RNA levels by in situ hybridization using one regular unstained FFPE slide and processed using the Leica BOND-III autostainer that is readily available in many clinical laboratories. RNA level detected by RNAScope can be quantified by dots/cell using publicly available software. Therefore, RNAScope is a practical assay and could be a promising alternative to IHC to quantify HER2 levels in breast cancer. We evaluated HER2 levels in 605 breast cancer tissue microarray cores using RNAScope and the two most commonly used FDA approved HER2 IHC assays: Ventana PATHWAY (PATHWAY) and Dako HercepTest (HercepTest). Clinical data were available for 505 cores from 347 patients. RNA level (dots/cell) by RNAScope was quantified using publicly available software QuPath. IHC assays were scored as 0, ultralow (UL, >0% but ≤10% incomplete membranous staining), 1+, 2+ and 3+. In addition, HER2 protein levels (AQUA protein level) were quantified from 100 cores through regression analysis, using AQUA score against cell line arrays with pre-calibrated HER2 protein levels determined by mass spectrometry. We used ANOVA to assess the differences in RNAScope results among the five IHC scores, and linear regressions to correlate RNAScope with HER2 AQUA protein levels. We finally evaluated 41 RNAScope whole-slide images (IHC 1+: n=5; 2+: n=26; 3+: n=10) of metastatic tumors from 31 patients treated with T-DXd. No significant differences of RNAScope results were observed among the 0, UL, and 1+ cores in both IHC assays, indicating the current IHC assays cannot differentiate HER levels in HER2-low tumors. However, statistically significant differences (p < 0.0001) were found among the IHC 1+, 2+, and 3+ cores and higher RNAScope dots/cell was associated with higher stage of the tumors. There was a strong correlation (R2 = 0.610) between the RNAScope results and quantitative HER2 AQUA protein level in the 100 cores. There were significantly higher HER2 RNA levels in the 41 metastatic biopsies with higher IHC scores (p < 0.05). When we used RNAScope to measures HER2 levels in metastases right before T-DXd treatment, there was numerically (p=0.881) higher HER2 RNA levels in responders (5.60±8.82 dots/cell) vs non-responders (5.20±5.31). Interestingly, the HER2 RNA levels in bone metastases was statistically higher (p=0.030) in non-responders (5.24±2.87, n=3) than in responders (1.55±0.81, n=5); although number of patients was low. For the non-bone metastases (esophagus, lymph node, liver, brain, lung), HER2 RNAScope values were numerically higher (p=0.261) in responders (9.65±2.87, n=5) than non-responders (5.19±5.74, n=15). In these non-bone metastatic cases, the response rates by IHC scores were 100% in 1+ cases, 24% in 2+ and 33% in 3+. When we used AI assisted categories based on RNAScope results, the response rates were 20% in RNAScope 1+ cases, 20% in 2+, 50% in 3+ cases. Our study shows that current IHC assays are unable to differentiate HER2 levels between IHC 0 and 1+ breast cancer cases, which is a critical issue in properly identifying patients who will benefit from T-DXd treatment. RNAScope results strongly correlate with HER2 protein levels and showed similar RNA levels among IHC 0 and 1+ cases. RNAScope is a simple and objective assay to quantify HER2 levels by dots/cell using publicly available software and may help better identify which patients benefit from T-DXd treatment. Other factors besides HER2 level may also contribute to the response rate in patients treated with T-DXd. RNAScope results in association with immunohistochemistry (IHC) scores by PATHWAY, HercepTest, and in biopsies from patients treated with T-DXd (T-DXd cohort). XX XX No significant (ns) difference of RNAScope dots/cell were seen among 0, UL and 1+ cases by both PATHWAY and HercepTest assays. There were significant differences of dots/cell comparing 1+, 2+ and 3+ cases. Note: ns not significant (p>0.05), * p<0.05, *** p<0.001, **** p<0.0001 Citation Format: Xiaoxian Li, Ji-Hoon Lee, Yuan Gao, Jilun Zhang, Katherine Bates, David Rimm, Huina Zhang, Geoffrey Smith, Diane Lawson, Jane Meisel, Jenny Chang, Lei Huo. RNAScope: a practical approach and promising alternative to immunohistochemistry to quantify HER2 expression in breast cancer [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 PO3-13-03.
Read full abstract