Predicting OncoDx Recurrence Score with Immunohistochemical Markers Including Stromelysin
Scott H Bradshaw, Dale Pidutti, Denis H Gravel, EC Marginean, Xinni Song, Susan J Robertson. Ottawa Hospital, Ottawa, ON, Canada; Carleton University, Ottawa, ON, Canada
Background: Recent reports suggest that immunohistochemistry (IHC) markers can be used to give prognostic information in breast cancer that is similar to that contained in the Genomic Health Inc. OncoDxR recurrence score (Onco-RS). Our own previous work suggests that the main drivers in predicting Onco-RS are ER PR and Ki67 proliferation index. One biological factor not included in our previous work is a measure of invasive potential. The goal of this study is to examine the potential predictive value of an IHC based measure of stromelysin-3 (ST-3/MMP-11) expression.
Design: A total ST-3 measure (both tumour and stromal) was selected as the first level of investigation. The patient population consists of 91 women with ER positive, HER2/neu negative breast cancer who completed OncoDxR testing. Total ST-3 expression was quantified on IHC stained slides using colour assisted image analysis and expressed as a ratio of area showing positive staining to total area Also identifiable larger zones of extracellular ST3 were removed from the primary analysis; however extracellular ST-3 was also assessed using a semiquanitative scoring scheme.
Results: There was no correlation between extracellular ST-3 and Onco-RS and no difference in the extracellular ST-3 score in the groups of patients with low (<18), intermediate or high (> 31) Onco-RS. In contrast total ST-3 did show a significant difference between these groups (ANOVA, p= .05) with post hoc analysis showing the difference is accounted for by a significantly higher score in the high risk group as defined by Onco-RS (figure 1). Stepwise multiple regression incorporating Ki67 percentage, total ST-3score, and semi-quantitative ER and PR scores showed no additional increase in ability to predict Onco-RS with inclusion of ST-3.
Conclusions: Although total ST-3 showed a statistically significant separation of total ST-3 values associated with high Onco-RS, it was not found to be an independent factor in multivariate analysis which included ER, PR and Ki67 scores. Future work is planned to correct for tumor cellularity and to address the question of the relative importance of tumour cell versus a tumour stromal/ inflammatory cell source of ST-3 staining.
Wednesday, March 6, 2013 9:30 AM
Poster Session V # 28, Wednesday Morning