Assessing the Clinical Utility of Companion Diagnostics - Episode 3

Predictive Value of Diagnostic Testing

Peter Salgo, MD, and John L. Fox, MD, MHA, discuss the definition of predictive value of diagnostic testing in cancer and describe the value of the genomic assays that are currently available in breast cancer.

Dr Fox defines predictive value as “something that helps define optimal therapy for a patient.” As an example, he describes how the detection of certain mutations in colorectal cancer can suggest that a patient will respond to a specific drug. In contrast, the genomic assays in breast cancer predict a general response to chemotherapy but they do not provide information on what specific drug or drug combination the patient would respond to. A test with strong predictive value can help provide more clarity about which treatment is most appropriate for the patient, comments Dr Fox.

Currently, the diagnostic tests in breast cancer only indicate the likelihood that a patient may respond to chemotherapy. Dr Fox is hopeful that in the future, predictive testing will help determine whether or not patients are responding to treatment.

Watch our related Peer Exchange, Oncology Stakeholder Summit 2014: Evidence-Based Decisions to Improve Quality and Regulate Costs