10P - Comparison of 28-gene and 70-gene panel in risk-prediction of Chinese women with early-stage HR-positive and HER2-negative breast cancer

Background

Multi-gene expression tests have been widely accepted as risk-prediction tools in early-stage HR-positive and HER2-negative (HR+/HER2-) breast cancer (EBC) after surgery. This study compared the performance of the 28-gene panel that developed from the Asian breast cancer population with the 70-gene panel in Chinese women with EBC.

Methods

Pts with EBC, previously undergone 70-gene testing after primary surgery from Jan 2019 to Oct 2022 were enrolled. FFPE of primary breast tumors were collected for 28-gene testing. Clinicopathological risk factors such as diagnosed age ≤40y, ≥T2, N1, LVI positive, grade III, Ki-67≥ 20% were included in the final analyses. Pts with clinicopathological risk factors ≥2 were clinical high-risk (n=48); others were low-risk (n=51).

Results

A total of 99 Chinese pts with EBC (pT1-2N0-1 or pT3N0M0) were enrolled in this study, with a median diagnosed age of 52 years (29-74 years). The median follow-up period was 38 months, and the last follow-up date was June 20, 2023. Two pts were lost to follow-up, but none recurred or died. The 28-gene and 70-gene panels detected high-risk pts (26% vs. 34%) and low-risk (73% vs. 65%), respectively. The overall agreement of the two tests was 72% (71/99). In addition, the 28-gene panel showed higher consistency with clinical-risk prediction than the 70-gene panel (Kappa: 0.51 vs. 0.39). Among them, the consistency of the 28-gene panel and 70-gene panel with the low clinical-risk prediction was 98% (50/51) and 84% (43/51), respectively; the high-risk prediction was 47% (25/48) and 54% (26/48), respectively. Table: 10P

Risk prediction comparison of 28-gene and 70-gene panels

Conclusions

The performance of the 28-gene panel in risk prediction of EBC was comparable with the 70-gene panel. The 28-gene panel could identify a more low-risk population in Chinese breast cancer pts. An enlarged sample size and longer follow-up for clinical outcomes will further confirm our conclusions.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

National Key Research and Development Program of China (2019YFE0196500).

Disclosure

All authors have declared no conflicts of interest.