Gene expression signature comparison for solid tumor chemotherapy.

Abstract

e21045 Background: It is unclear which one of signatures of ΔCT or RQ values from gene expression respond better in chemotherapy evaluation. Improved understanding is required for the comparison analysis in FFPE and blood samples. Methods: 370 solid tumor consecutive FFPE samples were tested by quantitative PCR. Among all the samples, 251 cases of solid tumor (108 lung cancer, 72 colon cancer, 44 gastric cancer, 27 breast cancer) with self-blood control were enrolled and thresholds were set at 15-35 cycles, which involved in predictive chemotherapy biomarkers ERCC1, BRCA1, TSand RRM1. For source dataset, a P value was computed for genes ΔCT and genes RQ value between the specimens of solid tumor, tumor subgroups and blood. A summary statistic to solid tumor, subgroup tumors and target gene signatures including comparison algoriths was analyzed with t test, chitest X2 and Pearson’s correlation coefficients. Results: Overall, RNA was isolated for 324 cases (87.6%) among 370 FFPE tissue successfully and over 1.8ng/ul per reaction. Total 251 solid cancers and its subgroups were examined for genes average ΔCT of tumor is 4.698±2.066 (ERCC1), 6.918±2.533 (BRCA1), 5.364±2.452 (RRM1), 6.355±2.459 (TS). Average RQ value of tumor is 38.157±83.086 (ERCC1), 65.030±245.821 (BRCA1), 42.841±166.933 (RRM1), 111.820±545.89 (TS). The data extracted from tumor ΔCT and RQ value (ΔCT/ΔΔCT(X+SD) were categorized and compared, statistics analysis indicated no significant differences of ΔCT value between the solid tumor, 4 subgroups tumors and 4 genes (P>0.05), but relative RQ value were observed in odds repetition and dispersed CV value with more groups difference (p<0.05), without correlation coefficients between ΔCTand RQ value in subgroups of cancer (P>0.05, pearson correlation), analyzed 4 gene synergism expression values based on cutoffs (ERCC1:4.3, TS:8.5, RRM1:8.0, BRCA1:6.5) in cancer subgroup, only half situation of coordinate up or down-regulation occured (p<0.05). Conclusions: ΔCT value- based signature make a stable and accurate resources data from solid tumor FFPE mRNA, which could contribute to the personality chemotherapy evaluation and analysis is more reliable than blood sample.