Abstract Introduction: Breast cancer is one of the leading malignancies and causes of cancer death for women now in China especially in the developed cities. Early diagnosis and adjuvant therapy have helped to improve the survival of the patients. The technique limitation slowed the pace of development of gene expression profiling. QuantiGene Plex(QGP) assay of bDNA technique without requiring RNA extraction, cDNA synthesis or PCR amplification which is suitable for the FFPE samples analysis. QGP could quantify the expression of RNA in FFPE tissues. Our primary purpose was to verify the value of the QGP platform in conjunction with multi-analysis in paired FFPE and FF tissues. The second purpose is focusing the key genes (ESR1, PGR, ERBB2 and MKI67) expression signature in early breast cancer patients (T1-2N0-1M0), and compare the gene expression profiling with the immunohistochemistry(IHC) corresponding panels by using two prespecified models-power and linear, based on the different combinations of the variables (FF4_P, FF4_L, FFPE4_P, FFPE_L, IHC4_P, IHC4_L). Methods: We retrospectively selected 103 paired archival tumor blocks and fresh frozen samples consecutively from the fresh tissue bank from the year 2006 to now. All the tissues are without neoadjuvant treatment. The tumor specimens with local recurrence or distance metastasis had the priority. The QGP data were harvested from a paralled bDNA assay with multiplex capability of the Luminex platform by coupling xMAP fluorescent beads. We compared the QGP result of FFPE samples with FF and IHC results head-to-head. We used automatic IHC staining instrument (BenchMark XT) to retest all the Ki67 staining. We divide the total sample into training and test groups. According to the data from the training group to create the two models, and through the cross-validation to build a model based on a subset of the provided data, then calculate their average value to create the final model. And then through the test group on the construction of group training model validation. Results: The storage time of the FF and FFPE samples used in our study ranged from 20.5 to 96.6 months. There is no different expression in each storage times intergroup and intra-group. FFPE and FF of QGP assay (single gene markers ESR1, PGR, HER2, and MKI67) correlated with the IHC result. The two models created the different scores. Both the power and linear models scores of FF4, FFPE4 and IHC4 are the prognostic factors in univariate and multivariate analysis of disease free survival with significant statistic(p<0.01), but in the bootstrapping resampling the FFPE4_L was borderline (p=0.068). The AUC of ROC in the test group is from 0.760-0.820 of each scores, and there is no statistic difference between each AUC. Conclusion:The multiplex bDNA assay is a reliable and accuracy technique, and could bridge the gap among the FF tissue to FFPE. The both models of QGP could be complement of the immunohistochemistry panels and provide more reproducible results. The parallel multiplex QGP could leading to a great potential for translational cancer research for future study and resolve the complex pathological interpreting. Citation Format: Xiaoyan Huang, Genhong Di, Xiaojing Xu, Bingyu Sun, Rui Bi, Wentao Yang, Rui Li, Guangyu Liu, Jiong Wu, Zhenzhou Shen, Zhimin Shao. Comparison of the parallel quantitative expression profiling in paired FF and FFPE samples with IHC4 by using two prespecified prognostic models [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P5-10-11.