15P A Novel RNA Test to Guide Primary Systemic Breast Cancer Chemotherapy

ABSTRACT Disclosure M.E. Trudeau: Holds stock options and is a member of Clinical Advisory Board in Rna Diagnostics. L.B. Pritzker: Study partially funded by Rna Diagnostics. A.M. Parissenti and B. Guo: Holds stock in Rna Diagnostics. K.P.H. Pritzker: Holds stock and is member of board of directors for Rna Diagnostics. All other authors have declared no conflicts of interest. The need for a prognostic biomarker to assess long term breast cancer chemotherapy efficacy is well established.The RNA Disruption Assay (RDA) is a novel prognostic test for women undergoing neoadjuvant chemotherapy that enables assessment of drug efficacy during treatment. RDA was developed a clinical trial for women with locally advanced breast cancer (CAN-NCIC-CTG MA-22). Varying doses of docetaxel and epirubicin were given at two weekly (dose dense) or three weekly intervals. Tumours were biopsied in duplicate at 3 time points; pre-therapy, mid-therapy and post-therapy. Trial endpoints included clinical response, pathological complete response (pCR), disease-free survival and overall survival. RDA is based on analysis of RNA electropherograms generated to assess RNA quality. At mid therapy, patients with tumours responding to treatment exhibited a dramatic reduction in RNA quality (RNA disruption). An algorithm was then developed by combining various features of the electropherogram to discriminate between those patients who subsequently exhibited pCR and those that did not. This algorithm generates an RDA score that indicates the degree of RNA disruption. A high RDA score is associated with subsequent pCR, while a low RDA score indicates that patients are unlikely to receive long term chemotherapy benefit. When applied to patients in the MA-22 study after the 3rd or 4th cycle of chemotherapy, RDA made the following predictions. Of the 85 patients studied, RDA predicted 32% to be non-responders. This prediction had a negative predictive value of 0.99 with a 95% confidence limit of 0.98-1.0 and a false negative rate of 2%. Of patients achieving subsequent pCR, RDA predicted 75% would have an increased chance of responding. This prediction had a positive predictive value of 0.17 reflecting that some patients had a drug effect but did not achieve a pCR. Use of the RDA in clinical practice will enable physicians to reliably identify patients unlikely to respond to breast cancer chemotherapy. Patients with an RDA score that indicates no long term benefit can then avoid the toxic side-effects of chemotherapy and can be switched to an alternate treatment, which may provide a better clinical outcome as well as health care cost savings.