Premium
Selecting Pharmacogenomic Variables that influence Propofol Pharmacokinetic Mean Residence Times in Surgical Patients using Machine Learning
Author(s) -
Eugene Andy Roger,
Eugene Beata
Publication year - 2018
Publication title -
the faseb journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.709
H-Index - 277
eISSN - 1530-6860
pISSN - 0892-6638
DOI - 10.1096/fasebj.2018.32.1_supplement.lb658
Subject(s) - propofol , medicine , cyp2c9 , bonferroni correction , population , anesthesia , statistics , mathematics , environmental health , cytochrome p450 , metabolism
BACKGROUND Propofol is the most widely used intravenous anesthetic agent for surgery and outpatient procedures. We sought to test the hypothesis that CYP2B6, CYP2C9, and UGT1A9 pharmacogenes predicts prolonged Propofol Mean Residence Times (MRT), a pharmacokinetic parameter, or normal Propofol MRTs. The importance of this study will help further understanding of the pharmacogenomics of Propofol anesthesia to clinically selectively identify patients who would be at risk of delayed emergence or complications from Propofol. METHODS Propofol pharmacokinetic data was obtained from the Mikstacki et al 2017 study that genotyped patients for the following pharmacogenes: CYP2C9 1075 A>C, UGT1A9 98T>C, and CYP2B6 516G>T that included 85 Polish patients (53 males and 32 females) who were rated as a I or II according to the American Society of Anesthesiologists physical status classification system (1). The Random Forest and Adaptive Boost Machine Learning algorithms were used to classify patients as being either having a prolonged MRT or a normal MRT, relative to the population mean. All analyses were conducted using the R statistical programming language (R Foundation for Statistical Computing, Vienna, Austria). RESULTS After assigning the patients into the two dichotomous groups, the final machine learning model included CYP2B6, CYP2C9, UGT1A9, gender, and age (in decades) from a total of 43 patients, from the original 85. Propofol MRTs ranged from 8‐minutes to 504‐minutes, among the two groups. The Random Forest classifier resulted in best predictive performance with an ROC AUC of 0.98 with a variable importance (descending): CYP2B6, gender, CYP2C9, age (in decades), and UGT1A9. CONCLUSION Among the pharmacogenomics and demographic variables tested, the CYP2B6 gene and gender were identified as the most important variables predicting the Propofol pharmacokinetic parameter of Mean Residence Times in surgical patients. Support or Funding Information None.