Transcriptional expression of 8 genes predicts pathological response to first-line docetaxel + trastuzumab-based neoadjuvant chemotherapy

Esther Schmitt, Frédérique Végran, Sandy Chevrier, Laura Burillier, Muriel Cadouot, Sarab Lizard-Nacol, Bruno Coudert, Pierre Fumoleau, Laurent Arnould, Romain Boidot

Overexpression of HER2 is observed in 20 to 30% of breast carcinomas. The use of trastuzumab has improved the treatment of these patients, especially when it is associated with docetaxel. To optimize the use of this treatment, it seems important to select putative complete responders before treatment administration. In this study, we analyzed by quantitative PCR the expression of 28 genes in HER2-overexpressing tumors treated with trastuzumab + docetaxel-based chemotherapy. We then correlated their expression profile with those of trastuzumab-sensitive and resistant cell lines to classify tumors as having a sensitive (pCR) or resistant (non-pCR) profile. Finally, we used public datasets from the GEO website to validate the reduced gene-expression profile obtained. We identified an 8-gene-expression combination that predicted the response to treatment with an accuracy of 76%. Based on public microarray data, we showed that the expression profile was specific to first-line trastuzumab + docetaxel-based treatment with an accuracy of 85%. Our results showed that by profiling the expression of 8 genes it was possible to predict the response to first-line trastuzumab + docetaxel-based chemotherapy. The use of cancer cell lines as the reference allowed a proper fit with the specificity of different tissues, such as lung or gastric cancers, which could also be eligible to concomitant HER2 inhibition by treatment with trastuzumab or tyrosine kinase inhibitors and docetaxel.