Abstract
Locally advanced breast cancer (LABC) cases have a varying five-year survival rate, mainly influenced by the tumor response to chemotherapy. Paclitaxel activity (response rate) varies across populations from 21.5% to 84%. There are some reports on genetic traits and paclitaxel; however, there is still considerable residual unexplained variability. In this study, we aimed to test the association between eleven novel markers and tumor response to paclitaxel and to explore if any of them influenced tumor protein expression. We studied a cohort of 140 women with LABC. At baseline, we collected a blood sample (for genotyping), fine needle aspirates (for Western blot), and tumor measurements by imaging. After follow-up, we ascertained the response to paclitaxel monotherapy by comparing the percent change in the pre-, post- tumor measurements after treatment. To allocate exposure, we genotyped eleven SNPs with TaqMan probes on RT-PCR and regressed them to tumor response using linear modeling. In addition, we compared protein expression, between breast tumors and healthy controls, of those genes whose genetic markers were significantly associated with tumor response. After adjusting for multiple clinical covariates, SNPs on the LPHN2, ROBO1, SNTG1, and GRIK1 genes were significant independent predictors of poor tumor response (tumor growth) despite paclitaxel treatment. Moreover, proteins encoded by those genes are significantly downregulated in breast tumor samples.