Title : Absolute quantitation of protein biomarkers to improve the accuracy of breast cancer diagnostics
Abstract:
Accurate assessment of protein biomarkers is an integral component of breast cancer diagnostics. For example, overexpression of Her2 is the sole determinant of suitability of Trastuzumab for the patients, and Ki67 levels are critical to determine the necessity of cytotoxic therapy in immunohistochemistry (IHC)-based surrogate assay worldwide. Yet, until now, the prevailing method of protein biomarker assessment is IHC, a method known to be plagued with subjectivity and inconsistency. While extensive efforts have been devoted to improve the consistency of Her2 assessment, Fluorescence in situ hybridization (FISH) remains the golden standard for Her2 assessment. Yet, FISH is not without its own issues. Likewise, it is a challenge for pathologists worldwide to achieve reproducible assessment of Ki67 levels in breast cancer specimens.
We propose that while IHC is irreplaceable for morphological assessment of the protein biomarkers, it is not suitable for their quantitation. Instead, objective quantitation of these protein biomarkers should be adopted in clinical fields to improve the accuracy and consistency of the results. Using Quantitative Dot Blot (QDB) method, we have been able to measure both Her2 and Ki67 levels absolutely and quantitatively. We were able to convert Her2 levels dichotomously using a cutoff at 0.267 nmole/g to achieve concordance with IHC at 93.3% (n=1546), and FISH at 94.2%. This method also eliminated the equivocal cases to improve the efficiency of the assay. For Ki67, using an validated cutoff at 2.31 nmole/g, we were able to significantly improve the performance of surrogate assay based on overall survival (OS) analysis of Luminal-like breast cancer patients in two independent cohorts.
Our results demonstrated the necessity and readiness to assessment protein biomarkers objectively and quantitatively in daily clinical practice. More importantly, the acceptance of this practice provides the basis for “big data” based clinical diagnostics in the near future.
