Do Combined Histopathological Features of ER Positive Breast Carcinoma Correlate with OncotypeDx Recurrence Score?
Shahrzad Ehdaivand, Rochelle A Simon, Cunxian Zhang, M Ruhul Quddus, Joyce J Ou, JinJun Xiong, Katrine Hansen, Michele M Lomme, WanLin Shen, Margaret M Steinhoff, W Dwayne Lawrence, C James Sung. Brown University/Women & Infants Hospital, Providence, RI
Background: Gene expression assays, such as OncotypeDx (ODx), show promise to predict recurrence and help guide treatment of the heterogeneous group of estrogen receptor (ER) positive breast carcinomas, including cases with metastatic disease in lymph nodes. However, this technology is relatively costly and patient selection criteria are subjective. Histopathological features, along with ER, progesterone receptor (PR), and HER2 status, are routinely reported and remain the current gold standard for predicting response to treatment and prognosis. This study aims to correlate routine histopathological features with ODx Recurrence Score (ODxRS) and seeks to determine if a subgroup of cases may not benefit from the added cost of ODx.
Design: The slides and charts of 206 patients who had ODx performed between July 2004 and July 2011 were examined. Ten histopathological parameters were evaluated and assigned value including tumor size, tubule formation, nuclear grade, mitotic count, lymphovascular invasion, lymph node status, quantitative ER and PR (using current CAP reporting standards), HER2 status, and extent of adjacent DCIS. A formula was developed to calculate a Histopathological Score (HS) combining nine weighted parameters (DCIS was excluded). The weight of each parameter was based on the strength of correlation with the reported ODxRS. The total HS for each case was compared to the ODxRS.
Results: Histopathological grade along with ER, PR, and HER2 status significantly correlated with ODxRS. Additionally, cases with a combined HS of ≤50 correlated with an ODx score of ≤30 (n=141, 69% of total cases). All resultant scores correlated with ODx scores with an R=0.565 (Figure 1). The extent of DCIS did not correlate with ODxRS. Nodal status, lymphovascular invasion, and tumor size did not independently correlate with the ODxRS.
Conclusions: Routinely obtained histopathological parameters can predict an ODxRS of ≤30 when the calculated HS is ≤50 (69% of our cases); therefore, it may not be necessary to perform ODx testing in these cases. Further studies are needed to validate the utility of combined HS in selecting cases for ODx testing.
Tuesday, March 20, 2012 1:00 PM
Poster Session IV # 11, Tuesday Afternoon