Post-Conference Perspectives: Impact of TAILORx in Choosing Adjuvant Therapy in Early Breast Cancer - Episode 1

Oncotype DX in Early Stage Breast Cancer


Terry P. Mamounas, MD: The Oncotype DX test is a commercially available genomic classifier that uses genes in the breast tumor from the resected specimen to estimate risk of recurrence and also a potential for chemotherapy benefit. It’s used in patients with estrogen receptor [ER]-positive, HER2 [human epidermal growth factor receptor 2]-negative breast cancer. It is comprised of 21 genes—16 are cancer genes and 5 are reference genes. This test has been developed and validated primarily in the population of patients with ER-positive, HER2-negative early stage breast cancer. As I mentioned, it was validated to predict the risk of distant recurrence in patients who are treated with adjuvant endocrine therapy. In those studies it was actually tamoxifen, so 5 years of tamoxifen was the treatment. In that population the 21-gene recurrence score, or Oncotype DX, has been found to be prognostic of risk of distant recurrence. In subsequent studies it was also found to be predictive of chemotherapy benefit, based on some other studies that compared tamoxifen to chemotherapy plus tamoxifen.

Are clinical pathologic factors such as age, clinical tumor size, and grade predictive of chemotherapy benefit? The short answer is most of those are not predictive of chemotherapy benefit. There is some prediction with age as well as grade. However, these factors are overshadowed by the use of the recurrence score in predicting chemotherapy benefit because they’re not as prognostic or predictive of chemotherapy benefit, as is the genomic profile in tests with the recurrence score.

Although the recurrence score correlates with some of the clinical pathologic factors, such as grade or Ki-67 protein, in general the recurrence score cannot entirely be predicted by clinical pathologic factors. This is because sometimes the clinical pathologic factors can be discordant. One factor could be favorable, one unfavorable, and you are not sure where you are. But in the majority of the cases we can be surprised by what we think the recurrence score will be and what we actually get.

In general though, we may have a good idea of what the recurrence score would be. However, we do order the recurrence score because on occasion, and not infrequently, the score could be totally different than we expect.

In the past we used these clinical pathologic factors to make a decision for chemotherapy use. Traditionally we used clinical pathologic factors such as tumor size and nodal status because the risk of the patients was higher where the tumor size is larger or where the nodal status was positive or multiple nodes. But in terms of predicting chemotherapy benefit, those factors are not necessarily the best predictors.

The Ki-67, grade, age, and tumor size are essentially prognostic of outcomes. However, some of these factors also could be predictive of chemotherapy benefit. For example, younger age, high grade, and high Ki-67 do predict chemotherapy benefit. But all these factors predict chemotherapy benefit less than the genomic classifiers, namely the recurrence score.

Transcript Edited for Clarity