[110] Can Conventional Histopathologic Prognostic Parameters of Invasive Breast Carcinoma Predict the Oncotype DX™ Recurrence Score.

Shikha Bose, Abdul ASR Mannan, Tobias Rathgeb. Cedars-Sinai Medical Center, Los Angeles

Background: The Oncotype DXTM (ODX) is a 21-gene RT-PCR based commercial assay that is being increasingly used in the management of ER positive (+), lymph node negative (-) breast cancers. The assay provides prognostic and predictive information in the form of a recurrence risk score (RS) that separates patients into low, intermediate, or high risk. This study is designed to determine if histologic and conventional immunophenotypic features of breast carcinoma are able to predict the ODX RS.
Design: Morphologic and immunophenotypic features of 224 invasive breast carcinomas excised between 2006 and 2010 were compared to the ODX RS. Features examined included components of the Modified Bloom-Richardson score i.e. nuclear and histologic grade, and mitosis; and expression of ER, PR, Her2/neu receptor and Ki67. Percent positivity was recorded for ER, PR and Ki67. Her2/neu was evaluated according to CAP guidelines. We modeled the continuous ODX score using linear regression, fitted with STATA 10 software.
Results: 100 (45%) of the carcinomas had a low RS, 91 (41%) an intermediate RS, and 31 (14%) had a high RS. The linear combination of histopathologic variables, 1.82 nuclear grade + 0.64 histologic grade + 3 mitoses – 0.13 ER – 0.08 PR + 0.17 Ki67 + 4.1 Her2/neu result, was statistically significantly related to the ODX score, F(7, 210) = 39.29, p < .0001. The sample multiple correlation coefficient was 0.57, indicating that approximately 75% of the variance of the ODX RS in the sample can be accounted for by the linear combination of pathology measures. In terms of categories of risk, in 58% of cases in our sample, our classification agreed with the ODX classification of risk. Of the 91 cases with the intermediate ODX RS, 34 (15.2%) had a low risk assessment while 5 (2.2%) had a high score with our method. See table.

Correlation of the ODX RS with our pathology-based score
Percentage (n)Oncotype DX Recurrence ScoreTotal cases
Our pathology-based score1 (low)2 (intermediate)3 (high) 
1 (low)29 (65)15.2 (34)0.9 (2)45.1 (101)
2 (intermediate)15.6 (35)23.2 (52)5.8 (13)44.6 (100)
3 (high)0 (0)2.2 (5)8 (18)10.3 (23)
Total cases44.6 (100)40.6 (91)14.7 (33)100 (224)

Conclusions: 1. Linear combination of histopathologic variables shows good correlation with Oncotype DX recurrence score. 2. For cases assigned an intermediate Oncotype DX recurrence risk, which leaves physicians with indeterminate course of action, our risk assessment may augment the decision for treatment selection.
Category: Breast

Tuesday, March 1, 2011 9:30 AM

Poster Session III # 50, Tuesday Morning


Close Window