A Novel Transcriptomic Signature for Prediction of Response to Adjuvant Chemotherapy in Patients With Stages II and III Gastric Cancer.

Abstract

Predicting patients who will benefit from postoperative adjuvant chemotherapy is crucial for precision medicine. Therefore, this study comprehensively analyzed messenger RNA (mRNA) expression profiles to identify novel biomarkers and developed a prediction signature for postoperative adjuvant chemotherapy in patients with gastric cancer (GC). Biomarkers were discovered by analyzing two publicly available genome-wide datasets from 343 patients with pathologic stages (pStages) II and III GC. A novel prediction signature was developed based on a quantitative reverse transcription polymerase chain reaction (qRT-PCR) assay using 137 pStages II and III GC frozen tissue specimens. Nine novel mRNAs were identified as candidate biomarkers in biomarker discovery, and a Gene Expression-based ADJuvant chemotherapy Response prediction for stages II and III GC (GEx-ADJ-Res) signature was developed using these candidate biomarkers and key clinicopathologic features by qRT-PCR assay. The GEx-ADJ-Res signature robustly predicted postoperative recurrence in clinical tissue samples (area under the curve [AUC], 0.84). The signature demonstrated sufficient potential for predicting response to postoperative adjuvant chemotherapy (AUC, 0.82) and was shown to be an independent predictor of postoperative recurrence and survival in multivariate analysis. Finally, the GEx-ADJ-Res signature was successfully validated using independent multi-institutional datasets (AUC, 0.91, 0.85, and 0.78, respectively). We identified the novel mRNA biomarkers and developed a novel signature that allowed robust prediction of response to postoperative adjuvant chemotherapy in patients with GC. This signature could become a precision medicine tool in GC treatment.