Postmortem metabolomics: Correlating time‐dependent concentration changes of xenobiotic and endogenous compounds

Abstract

Postmortem redistribution (PMR) describes the artificial postmortem concentration changes of xenobiotics that may pose major challenges in forensic toxicology. Only a few studies have systematically investigated time‐dependent postmortem drug concentration changes so far and the a posteriori estimation of the occurrence of PMR is not yet possible. In this context, the general concept that postmortem biochemical changes in blood might parallel drug redistribution mechanisms seems promising. Thus, the current study investigated the possible correlations between time‐dependent postmortem concentration changes of xenobiotic and endogenous compounds; exemplified for authentic morphine (n = 19) and methadone (n = 11) cases. Peripheral blood samples at two time‐points postmortem were analyzed for morphine and methadone concentrations and an (un)targeted postmortem metabolomics approach was utilized to combine targeted quantitative analysis of 56 endogenous analytes and untargeted screening for endogenous compounds (characterizing 1174 features); liquid and gas chromatography–mass spectrometry was used respectively. Individual statistically significant correlations between morphine/methadone and endogenous compounds/features could be determined. Hence, the general applicability of the proposed concept could successfully be confirmed. To verify the reproducibility and robustness of the correlating behavior, a larger dataset must be analyzed next. Once a marker/set of markers is found (e.g. robust correlation with specific xenobiotic or xenobiotic class), these could be used as surrogates to further study the time‐dependent PMR in a broader variety of cases (e.g. independent of a xenobiotic drug present). A crucial next step will also be the attempt to create a statistical model that allows a posteriori estimation of PMR occurrence of xenobiotics to assist forensic toxicologists in postmortem case interpretation.

Postmortem Metabolomics: Strategies to Assess Time-Dependent Postmortem Changes of Diazepam, Nordiazepam, Morphine, Codeine, Mirtazapine and Citalopram

“…Following the proposition of Langford and Pounder on the general possibility that postmortem biochemical changes in blood might parallel drug redistribution mechanisms and a recently published successful proof-of-concept study, the aim was to find and confirm strong positive or negative correlations between time-dependent concentration changes of drugs and endogenous metabolites in a large dataset [ 16 , 17 ]. These could help to further understand the underlying biochemical processes of PMR and might be used for a posteriori estimation on the occurrence of PMR for specific cases, circumventing the problem of generalized models as detailed above.…”

“…For morphine, five endogenous molecules from the targeted workflow were found to significantly correlate with its time-dependent postmortem behavior, namely methionine, phenylalanine, valine, isoleucine and proline ( Table 3 ). None of these amino acids did show a significant correlation with morphine concentration changes in the previously conducted proof-of-concept study [ 17 ]. Previously, significant negative correlations of morphine with creatinine, glutaric acid, hypoxanthine, fructose, pentadecanoic acid (C15:0), palmitoleic acid (C16:1), alanine and linoleic acid (C14:0) were established.…”

“…An endogenous molecule/feature that shows a strong correlation with a drug (of abuse) or drug class could be used as a surrogate marker to further study time-dependent PMR and a posteriori estimation of PMR occurrence could be attempted. A recent proof-of-concept study took up this idea and utilized an untargeted gas chromatography-high resolution mass spectrometry (GC-HRMS) metabolomics workflow to find individual statistically significant correlations between the time-dependent concentration changes of morphine/methadone and some endogenous molecules/features [ 17 ]. As this study was only comprised of a small case set (morphine: n = 19; methadone: n = 11), robustness and reproducibility were still needed to be confirmed, which was one of the aims of the current study, utilizing 477 authentic postmortem cases, where femoral blood samples at two postmortem time points were available.…”

Untargeted metabolomics approaches to improve casework in clinical and forensic toxicology—“Where are we standing and where are we heading?”

“…To overcome confounders due to different storage time, analysis day, and so forth, t1 and t2 samples from one individual were always analyzed within the same analytical batch. Correlation analysis was then only performed for concentration differences between t1 and t2, but not on a global level (Brockbals, Staeheli, et al, 2020). Alternatively, already measured data files can be subjected to re‐evaluation, in case routine data acquisition was performed in an untargeted manner using HRMS methods.…”

“…Their preliminary investigation, analyzing one authentic postmortem case, found a strong positive correlation between amino acids (particularly glycine, leucine, methionine, serine, and valine) and drug concentrations in pulmonary blood samples (Langford & Pounder, 1997). Although the validation in a large case series has never been carried out, a recent proof‐of‐concept study by Brockbals et al expanded on this concept and carried out an untargeted postmortem metabolomics investigation, trying to find possible correlations between time‐dependent postmortem concentration changes of morphine and methadone and endogenous molecules in an authentic case set (Brockbals, Staeheli, et al, 2020).…”