Patterns of Oncotype DX Recurrence Scores – Analysis Based on Levels of ER & PR Expression and Proliferation Markers
Meredith Burge, Steven Frame, Patrick McGrath, Edward Romond, Marie-Louise Fjallskog, Cecilia Ahlin, Michael Cibull, Yolanda Brill, Luis M Samayoa. University of Kentucky, Lexington, KY; Uppsala University, Uppsala, Sweden
Background: In 2010 we reported that in 20 – 30% of patients undergoing Oncotype DX, the test was probably of no clinical significance, since Low (L) and High (H) Recurrence Scores (RS) for Grade (G) I & GIII tumors could be accurately predicted based on tumor differentiation, Progesterone (PR) status (+/-) and high Estrogen (ER) expression (95% of tumor cells). We currently test the possibility of identifying other groups of ER (+) patients with predictable RS by analyzing the role of proliferation markers (PM) and levels of PR expression in the same patient population.
Design: Histopathologic material from 72 patients with known Oncotype DX RS was analyzed for the following: tumor grade (Elston Modification of Bloom-Richardson Score), levels (%) of ER and PR expression, proliferating index (PI) according to Cyclin A and Ki-67 expression (%), and number of mitoses in 10 high power fields (HPF). All analyses were performed in sections from definitive surgical specimens. Levels of ER, PR, Ki-67 and Cyclin A were performed by Immunohistochemistry (IHC) and scored by a consensus of 3 pathologists. Each variable was then correlated with the corresponding RS.
Results: See Figures I and 2
Conclusions: 1) The combination of tumor grade and levels of PR expression in patients with tumors showing >80% (+) ER are predictors of Oncotype RS. 2) Although average values of Cyclin A, Ki-67 and mitotic counts correlated with average RS accordingly, regression analysis failed to show an acceptable correlation in predicting RS scores individually. 3) Depending on clinical judgement, these observations may result in stricter indications for Oncotype DX, i.e.: only in patients with GII < 75% PR (+) tumors and GIII > 30 % PR (+) tumors, ultimately resulting in cost containment equivalent to > 50% (28/72) of resources currently disbursed without clinical validation.
Tuesday, March 20, 2012 1:00 PM
Poster Session IV # 15, Tuesday Afternoon