Introduction: Centromere protein A (CENP-A), an essential centromere protein, has been associated with high grade cancers. This study was undertaken to determine if CENP-A is a prognostic factor for breast cancer patients not receiving systemic therapy or predictive of response to tamoxifen or neoadjuvant chemotherapy.
Methods: mRNA levels of CENP-A and CENP-B, a centromere protein that binds independently of CENP-A, were measured in breast cancer specimens from 484 patients receiving no systemic therapy, 276 patients receiving tamoxifen, and 233 patients treated with neoadjuvant chemotherapy. Associations between CENP-A, CENP-B, Ki-67, relapse, and chemotherapy response were determined.
Results: CENP-A but not CENP-B was higher in estrogen receptor (ER)-negative tumors than ER-positive tumors and positively correlated with Ki-67 expression. Among patients with ER-positive disease who received no systemic therapy or tamoxifen, higher levels of CENP-A were associated with lower rates of 5-year distant relapse free survival (DRFS). On multivariate analyses including Ki-67, high CENP-A expression had a hazard ratio of 10.9 for relapse in patients with ER-positive disease not receiving systemic therapy (95% CI, 2.86 to 41.78; P = 0.00047) and 1.64 for patients with ER-positive disease receiving tamoxifen (95% CI, 0.99 to 2.71; P = 0.054). CENP-A was not an independent prognostic marker in ER-negative tumors. For both ER-positive and ER-negative tumors, CENP-A was not a significant independent predictor of chemotherapy response.
Conclusions: CENP-A was a significant independent prognostic marker for patients with ER-positive breast cancer not treated with systemic therapy but had limited predictive value in tamoxifen treated patients and was not predictive of response to neoadjuvant chemotherapy.