Identification of Preoperative Serum Metabolites Associated with Postoperative Opioid Consumption in Gastric Cancer Patients by Extreme Phenotype Sampling.

BACKGROUND:Postoperative pain increases patients' risk and opioids remain the main analgesics to relieve it. However, improper use of opioids causes many side effects and identification of suitable preoperative biomarkers that predict postoperative opioid consumption may aid clinicians in improving analgesic strategies for patients. The activity of metabolites modulates multiple phenotypes and can function as biomarkers for disease prediction and diagnosis.OBJECTIVES:In this study, we explore whether preoperative serum metabolites are associated with postoperative opioid consumption in gastric cancer patients by extreme phenotype sampling.STUDY DESIGN:This was a case-control, observational study.SETTING:This study was conducted at Beijing Cancer Hospital.METHODS:One hundred and sixty-nine gastric cancer patients participated in this study. After exclusion of 51 patients, postoperative pain intensity and opioid consumption data of 118 patients were collected. Patients were sorted by gender and classified into 2 groups based on opioid consumption during the 24h postoperative period. Patients in the sufentanil high consumption (SHC) group and patients in the sufentanil low consumption (SLC) group were ranked in the top or bottom 30% of sufentanil consumption, respectively. Untargeted metabolomic analysis of preoperative serum samples from both groups was performed by ultra-performance liquid chromatography-mass spectrometry/mass spectrometry (LC-MS/MS) and orthogonal partial least square discriminant analysis. Allele frequencies of DAO rs10156191 and MAOB rs1799836 SNPs in both groups were detected by Sanger sequencing.RESULTS:Thirty-five metabolites in preoperative serum were significantly different between the SLC and SHC groups. Hydrogen phosphate had the highest area under the curve in a ROC analysis (0.98), suggesting that it may serve as a predictive biomarker for postoperative opioid consumption. Differential metabolites unique to the male and female subgroups were also identified. Histidine metabolism was the most altered pathway between the SLC and SHC groups. There were no significant differences in the allele frequencies of 2 SNPs associated with histamine degradation; however, 2 metabolites of histamine degradation, imidazole-4-acetaldehyde, and methylimidazole acetaldehyde, showed different trends in the 2 groups.LIMITATIONS:Our study was restricted to gastric cancer patients with strict exclusion criteria, which may limit the generalizability to other groups.CONCLUSION:Preoperative serum metabolites were associated with postoperative opioid consumption. Different efficiencies of histamine degradation may be one cause of the variable sensitivity of patients to acute pain and warrants further study.