Innovative Marker Discovery
Unlike conventional deductive methods that rely on knowledge of the function and pathway for specific genes, Exagen's approach relies on "pure" inductive, data-driven discovery.
The technology mines genome-wide data in a global search to discover the best combinations of genes, without respect to gene function. Combinations are then ranked. Thus, unbiased selection across all genes is assured, and each gene has an equal chance of being in the final combination.
To discover prognostic DNA copy number markers for breast cancer, Exagen concurrently mined gene expression data and DNA copy number data to discover sets of genes that separate patients who have a low risk of recurrence from patients with a higher risk.
Seventeen genes were identified that were consistently found in top combinations, and all seventeen were taken into a retrospective validation study of formalin fixed, paraffin embedded (FFPE) tumor specimens from 229 patients with invasive ductal carcinoma of the breast (Stages I, II, and III).
Seventeen BACs (bacterial artificial chromosomes), each containing one of the seventeen genes, were used as FISH probes to determine the optimal combination of genes to accurately separate low risk from high risk patients.
This study, conducted at the University of New Mexico Cancer Research and Treatment Center, confirmed the validity of the discovery approach and demonstrated that two different sets of three genes each were sufficientone set for hormone receptor1 (HR) positive patients and one set for HR negative patients.
In independent test sets, the HR positive marker set had a 93.8% NPV in lymph node negative patients, and the HR negative set had a 100% NPV in lymph node negative patients. Each marker set had a 91% NPV in patient populations that are both node positive and node negative, exceeding the performance of current criteria used to determine low risk of recurrence.
These two genomic tests, named the Key2 Breast Cancer Prognosis Tests, dovetail with existing patient stratification by HR status and improve the ability to predict those patients who are at low risk of recurrence.
Using these tests, both HR positive and HR negative breast cancer patients can be tested at the point of tumor removal to predict which patients are at low risk of recurrence and who, therefore, may not benefit from additional treatment (after surgical resection and local irradiation).