A liquid biopsy signature of circulating exosome-derived mRNAs, miRNAs and lncRNAs predict therapeutic efficacy to neoadjuvant chemotherapy in patients with advanced gastric cancer

Abstract

At present, there is no validated marker to identify the subpopulation of patients with advanced gastric cancer (AGC) who might benefit from neoadjuvant chemotherapy (NACT). In view of this clinical challenge, the identification of non-invasive biomarkers for efficacy prediction of NACT in patients with AGC is imperative. Herein, we aimed to develop a non-invasive, liquid-biopsy-based assay by using an exosome-derived RNAs model based on multi-omics characteristics of RNAs. We firstly used a multi-omics strategy to characterize the mRNAs, microRNAs (miRNAs) and long non-coding RNAs (lncRNAs) profiles of circulating exosome enriched fractions in responders to NACT paired with non-responders, using RNA sequencing. Finally, numerous miRNAs, mRNAs and lncRNAs were identified to be associated with the response to NACT in patients with AGC, and it was validated in an independent cohort with promising AUC values. Furthermore, we established a 6-exosome-RNA panel that could robustly identified responders from non-responders treated with fluorouracil-based neoadjuvant chemotherapy.

Keywords: Advanced gastric cancer; Biomarker panel for efficacy prediction; Exosome; Multi-omics characteristics of RNAs; Neoadjuvant chemotherapy.

Fig. 1

Differential expression analyses and gene expression signature for each group. The differentially expressed mRNAs (A), miRNAs (B) and lncRNAs (C) of plasma-derived exosomes enriched fraction between responders and non-responders to NACT. (D) The identified panel of exosomal RNAs including 9 miRNAs, 5mRNAs and 4 lncRNAs, to discriminate potential responders to NACT in AGC patients. (E) The PCA method was used to describe gene expression signatures of candidate exosomal RNAs in responders and non-responders. (G) The expression levels of 8 verified exosomal RNAs with differential expression and consistent expression trend. (H) The ROC analyses of verified 8 exosomal RNAs

Fig. 2

Establishment and validation of a 6-exosome-RNA model. A The linear correlation analysis of 8 validated exosomal biomarkers. B The most robust predictive genes were identified using the least absolute shrinkage and selection operator (LASSO) Cox regression algorithm. C An ensemble of 6 genes remained with nonzero coefficients. D The scores of enrolled patients in the training set which were calculated based on the 6-exosome-RNA model. E-F The boxplot and the receiver operating characteristic (ROC) analyses illustrated the predictive ability of this 6-exosome-RNA model to identify responders in the training set. G The patients’ scores in the validating set calculated according to the 6-exosome-RNA model. H The ROC analysis for the predictive power of the focal 6-exosome-RNA model, and the area under the ROC curve was 0.774