Postmortem metabolomics as a window into the agonal phase of death: a comparison between opioid intoxications and hangings

Abstract Fatal opioid intoxications provide a postmortem diagnostic challenge due to a lack of coherent autopsy findings and the potential of tolerance and individual variations to complicate the evaluation of toxicological results. This study investigates whether postmortem metabolomics could be used to discriminate between opioid intoxications, irrespective of the involved opioid, and hangings, and if the degree of respiratory failure in intoxications could be predicted. Data was acquired from toxicological screenings of femoral blood with high resolution mass spectrometry (HRMS) comprising 338 hangings and 287 opioid intoxications. Cases were divided into a training set and a validation set using a 3:1 ratio. An orthogonal partial least squares discriminant analysis (OPLS-DA) binary classification model was built using 269 features, with an R 2 = 0.474 and Q 2 = 0.377, and classification accuracy of 78.2%. A second model, aiming to classify the degree of respiratory failure did not perform successfully. Metabolite identification revealed differences in acylcarnitine, nicotinamide adenine dinucleotide (NAD) and monoamine metabolism. Analysis of misclassifications revealed a high proportion of misclassified opioid suicides and suggested a potential to differentiate between acute and delayed death in opioid intoxication. Further investigation with larger cohorts is needed to evaluate generalizability and explore opioid-related respiratory failure. However, the current study demonstrates that postmortem metabolomics can assist forensic pathologists in classification of cause of death, and potentially manner of death, in opioid related cases.