Verification of the BREASTEST assay in an Australian population: A novel liquid biopsy assay measuring lipid biomarkers for early-stage breast cancer screening.

Abstract

Background: Women with dense breasts are at higher risk of developing breast cancer however there is no agreement on how to screen these women, despite mandatory density reporting becoming more prevalent. Currently available image-based, population level breast screening modalities have poor sensitivity in these women. Likewise, the positive predictive value of imaging is reduced due to a higher false positive rate/recall rate compared to that of women with less dense breasts. Liquid biopsy approaches do not suffer from imaging-dependent challenges with density and provide a promising option in this population. We have discovered a novel liquid biopsy assay using a lipidomic discovery platform by combining liquid chromatography tandem mass spectrometry (LC-MS/MS) and machine learning. The BREASTEST assay is intended to complement standard of care screening and address gaps in the current screening paradigm. Method(s): This study was conducted to verify the performance of the BREASTEST assay in an Australian population (n = 720). This verification study was an observational case-control study that prospectively recruited women with breast cancer (n = 275) or without (n = 446) across 10 clinical sites over a 34-month period. A primary imaging modality was identified for each subject and a binary classification was assigned to the outcome of this imaging (normal/suspicious) to enable comparison to, and combination with, the BREASTEST assay. An assay with a high sensitivity has utility as a rule-out test and would complement population-based imaging (high specificity). Therefore, the assay was designed to achieve a sensitivity of 0.90. The combined specificity of imaging with the assay was calculated to estimate the clinical benefit BREASTEST could bring in ruling-out women without breast cancer. A safe de-escalation rate was also calculated in this study to assess the potential reduction in unnecessary further assessment if this assay was added to standard imaging. Result(s): The utility of the BREASTEST assay was observed when results were combined with primary imaging data in the study cohort. Across all imaging modalities and breast densities, the assay improved the combined specificity in 45-75-year-old women by +6.1% (0.712, 0.773) and had a safe de-escalation rate of 21.0%. Highlighting the potential benefit to women with dense breasts, in women aged 30-49 years with breast density category D the combined specificity was +14.7% (0.585, 0.732) and safe de-escalation rate of 37.5%. The BREASTEST assay alone obtained a sensitivity of 0.90, specificity of 0.369 with an AUC of 0.743. Conclusion(s): The performance and utility of the BREASTEST assay was verified in this study in an Australian population. It has highlighted the potential of this assay in the workup of women with breast conditions, in particular women with dense breasts.