ABCB1 has recently been described as a substrate for ixabepilone. ABCB1 was overall expressed at low levels in the tumors examined, while the presence of the T-allele was not significantly associated with ABCB1 mRNA expression, except in the homozygotes. The association of the T variant with tumor PgRpositivity, which was per se a favorable prognostic factor for survival in this study, may have accounted for the observed better patient outcome. Indeed, none of the ABCB1 T LOUREIRIN-B variants remained prognostic for survival upon multivariate adjustment, while PgRpositivity, assessed by IHC, did. In any case, the association of the evolutionarily more recent ABCB1 T-alleles in the SNPs examined with PgR- and ER-positivity in breast carcinomas is a novel finding meriting further investigation. In vitro studies suggest that the tau protein and Danshensu paclitaxel both bind to the same pocket on the inner surface of the microtubules. Conceivably, low expression of the tau protein renders microtubules more sensitive to paclitaxel. These data indicate that tau protein is a potential predictor of sensitivity to all microtubulestabilizing agents, including ixabepilone. Indeed, low ER and MAPT mRNA expression were strongly predictive of sensitivity to ixabepilone in the neo-adjuvant setting, while ixabepilone seems to benefit patients with triple-negative breast cancer. However, MAPT mRNA did not predict benefit from the addition of paclitaxel to epirubicin/CMF dose-dense adjuvant chemotherapy in a different study. On the other hand, tau protein expression assessed by IHC was not related to ixabepilone response in a small phase II study in metastatic breast cancer. In the present series MAPT mRNA and tau protein expression were strongly associated with each other, while high expression of either was associated with favorable outcome. This is in line with ER and PgR being favorable predictors in the present study, since MAPT expression is influenced by ER. It is also in line with the inverse correlation of high MAPT with high Ki67 and TopoIIa-positivity, which are generally associated with poor prognosis. Evidently, all these pro-ER and pro-MAPT related findings appear to be discordant with the reports cited above. Sample size and consistency with respect to breast cancer molecular subtypes, as well as treatment setting may have accounted for this discrepancy. In support to the present findings, however, ER positive breast cancer cell lines were ixabepilone sensitive, while MAPT mRNA expression was not included in gene expression sets predictive of ixabepilone response in a very recent study. Evidently, the issue of MAPT expression levels with respect to ixabepilone efficiency needs further clarification. From a different perspective, this study confirms the wellestablished good prognosis of ER/PgR-positive tumors, the association of MAPT expression with these favorable prognostic factors and the adverse prognostic effect of TopoIIa-positivity. As previously stated, resistance to taxanes may develop via different mechanisms, such as MDR, b-tubulin mutations or overexpression of the bIII-tubulin isoform or microtubule-associated proteins. bIII is one of the eight different isoforms of b-tubulin that have been identified so far and has been linked to paclitaxel resistance in vitro. In contrast to paclitaxel and to different epothilones, ixabepilone binds to bIII-tubulin containing microtubules and stabilizes them but its efficacy does not seem to be affected by bIII-tubulin in vitro. At present, information about a potential link between bIIItubulin expression and clinical activity of ixabepilone is limited.