Carbonic anhydrase IX (CA IX) as a predictive marker for benefit from post-mastectomy adjuvant locoregional radiation (RT): Results from the British Columbia (BC) Randomized Radiation Trial

Abstract

571 Background: The benefit of adjuvant locoregional RT in post-mastectomy node positive breast cancer is controversial. Three randomized trials (Danish Breast Cancer Cooperative Group 82b/82c Trials and the BC Randomized Radiation Trial) have demonstrated improved survival for the addition of locoregional RT. We have previously shown in a large population based tissue microarray (TMA) series that CA IX, a hypoxia regulated enzyme, is an independent poor prognostic factor. We now investigated whether CA IX would be a predictive marker for locoregional RT in the context of the BC Randomized Radiation Trial. Methods: A TMA series was constructed from 160 available blocks out of the 318 patients randomized on the trial. CA IX expression was evaluated by IHC with a monclonal antibody M75. Any staining for CA IX within tumour cells was scored as positive. The frequency of expression of CA IX and its association with other standard prognostic factors (tumour size, grade, ER, PR, HER-2) as well as other biological correlates (p53, EGFR) was evaluated. Statistical analysis included KM survival analysis and interaction tests. Results: Median follow-up of the trial is greater than 15 years. CA IX was expressed in 23% of cases. In the patients randomized to no RT, CA IX is a poor prognostic marker associated with a worse overall survival (log rank p=0.0002). CA IX expression was predictive of benefit to locoregional RT. In the CA IX positive cohort those patients randomized to RT had significantly improved survival (p<0.0005) with an estimated 15 year overall survival of 73% with RT versus 19% without RT. Conversely in the CA IX negative cohort there was no significant benefit for the addition of RT (p=0.225). The associated 15 year overall survival was 54% and 50% with and without RT respectively. The interaction test between CA IX expression and RT was significant (p<0.01). Conclusions: CA IX appears to be a promising biomarker predictive of benefit for locoregional RT. Further work validating this is warranted. Work is ongoing assessing for the biological rationale (e.g. BNIP3, HIF1α) and other biologic correlates associated with CA IX expression. No significant financial relationships to disclose.