Prediction of Oncotype Dx recurrence score using clinical parameters: A comparison of available tools and a simple predictor based on grade and progesterone receptor
The Oncotype Dx test is a genomic test currently used in clinical practice to predict the risk of disease recurrence in estrogen receptor (ER)-positive, HER2-negative breast cancer patients with axillary lymph node-negative or micrometastatic disease. The test is one of several similar genomically based tests available. Although it has a good predictive value, it is expensive and thus constitutes a significant financial burden for health systems. Thus, several attempts have been made to devise low-cost tools that could predict the recurrence score derived from the genomic evaluation using easily obtainable clinical parameters.
Methods
Two previously proposed predictive tools were evaluated in a cohort of 201 patients that had undergone the Oncotype Dx test for their efficacy in predicting the Oncotype Dx Recurrence Score (RS). A simple predictor, named GR-PR, based on two available pathologic parameters, grade and progesterone receptor status was devised and also evaluated.
Results
The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of all three tools were compared and found to be similar for all cutoff points of Oncotype Dx RS. The accuracy of GR-PR was comparable to the best performing of the two other prediction tools for all four cutoff points.
Conclusion
The simple GR-PR predictor proposed in this study seems to be at least as accurate as more complex tools and should be the preferred tool for the prediction of Oncotype Dx RS from clinicopathologic parameters when the Oncotype Dx test is not available.
