Serum microRNA profiles as prognostic or predictive markers in the multimodality treatment of patients with gastric cancer.

Despite the implementation of multimodality treatment strategies, the persistently poor prognosis of gastric cancer patients is predominantly caused by the lack of predictive markers for response assessment in the neoadjuvant setting, preventing individualized therapy. Therefore, the identification of novel predictive and prognostic markers for application in the multimodality treatment of gastric cancer patients is required. The aim of the present study was to characterize the serum microRNA (miRNA/miR) profile of gastric cancer patients undergoing multimodality therapy to identify possible prognostic and predictive markers. The study consisted of 32 patients with gastric cancer who had undergone either primary surgical resection (n=14) or neoadjuvant therapy followed by surgical resection (n=18). Histopathological regression was defined as a major histopathological response when the resected specimens contained <10% vital residual tumor cells. Intratumoral miRNA was isolated from pre-operative or post-neoadjuvant blood serum samples. Initially, microarray analyses were performed in six of the patients that received neoadjuvant treatment (three responders versus three non-responders), to assess the amplification profile of dysregulated miRNAs. Based on these findings, possible predictive or prognostic markers were validated in all study patients by performing single reverse transcription-polymerase chain reaction (RT-PCR) analysis. Depending on the extent of the histopathological regression, a differential miRNA expression profile was identified in the microarray analyses. Based on the amplification profile, miR-21, miR-29a and miR-221 were selected for additional validation. However, the single RT-PCR measurements of the three selected miRNAs did not exhibit any prognostic or predictive value in the patients treated with primary resection or neoadjuvant therapy and resection. Thus, the current pilot study failed to identify a prognostic or predictive value in selected miRNAs using single RT-PCR measurements, however, the microarray results revealed a differential microRNA expression profile depending on the histopathological regression. The findings of the present study may have been affected by the small sample size.