Expression of Gene Trios as Predictive Markers of Response to Doxorubicin Based Primary Chemotherapy in Breast Cancer Patients.

Some studies have been carried out to identify a gene expression profile predictive of drug response in breast cancer patients and assessed the reproducibility of the model in an independent group of patients. In common, these studies have used the same technique already used to identify the panel of genes in a training set, to verify accuracy of the model in a validation set. We have previously identified through cDNA microarray technology gene trios whose expression was capable of predicting response to neoadjuvant chemotherapy based in doxorubicin and cyclophosphamide in breast cancer patients. We have now evaluated whether expression of these genes analyzed by real time RT-PCR, which represents a more accessible technique, would reproduce cDNA microarray results in separating responsive from non-responsive patients. Twenty eight samples, already studied by cDNA microarray, were analyzed as a technical validation group and subsequently, another 14 samples were evaluated, as a biological validation group. Expression of nine genes (defining five trio combinations, previously identified) was evaluated by RT-PCR in samples from the technical validation group, to define the separating plane, using linear discriminant analysis. Among the five trios, the best separation of tumors was conferred by RPL37A, XLHSRF-1 based trios (with NOTCH1 or NUP210, as third genes), which correctly classified 86% of samples from the technical validation group and 82% in a cross-validation analysis (leave-one-out). Next, samples from the biological validation group were spatially distributed using the pre-established features from these two trios, resulting in 71% correct classification. Additionally, other gene combinations were searched for a higher accuracy in discriminating response and expression of a new trio, RPL37A, SMYD2 and MTSS1, correctly classified 86% and 79% of samples from the technical validation (leave-one-out cross-validation) and the biological validation groups, respectively. Expression values evaluated by cDNA microarray and RT-PCR are reproducible, however they are not exactly the same as these methods are based on different technical principles and normalization approaches. Our data indicate that care should be taken when substituting techniques in an attempt to reproduce a model. In conclusion, expression of the gene trio RPL37A, SMYD2, MTSS1, as evaluated by RT-PCR, is a potential candidate as predictive marker of response to neoadjuvant chemotherapy in breast cancer patients. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 2040.