An ultrarapid CYP2D6 prediction from postmortem DNA did not establish why a particular person died of opioid toxicity. A 2020 veteran study found five predicted ultrarapid metabolizers among 74 usable samples, about 7%, but did not establish that this phenotype was overrepresented in overdose deaths. The relevant toxicology subgroup comparison was also nonsignificant.
How were the decedents selected?
Boyle and Stock identified Utah military veterans who received care at one VA medical center, had a hydrocodone, oxycodone or tramadol prescription in their record, and were identified through overdose death records from January 2009 to January 2016. Banked autopsy blood had to be available.
They obtained 75 samples; one failed genotyping because of poor DNA amplification, leaving 74. Participants were predominantly male, 97%, and Caucasian, 95%, with a mean age of 51 years. These selected records were not a matched case-control study of all opioid users, all veterans or all people who died.
The laboratory predicted function from DNA
Autopsy blood had been placed on FTA filter-paper cards. DNA was extracted from 3 mm punches, then analyzed by an iPLEX genotyping assay and MassARRAY, with copy-number assessment. Each sample was tested twice and the laboratory reported strong concordance, despite variable DNA quantity and quality.
The method tested a defined variant panel. If no targeted variant was detected, the allele was presumed functional *1. That assumption and the panel’s coverage belong with the interpretation; the study did not sequence every possible functional variation.
The authors used historical activity-score categories to predict phenotype: poor at zero, intermediate at 0.5, extensive at 1–2 and ultrarapid above two. This was genotype-to-phenotype translation under the study’s classification scheme, not direct measurement of liver enzyme expression or real-time metabolic activity immediately before death.
The toxicology denominators narrowed further
Toxicology records were available for 68 decedents, of whom 49 had an opioid detected. Twenty of those 49 had hydrocodone, oxycodone or tramadol detected; 29 had other opioids. A prior prescription therefore did not establish that the prescribed drug was present at death.
The paper did not provide one uniform anatomical blood collection site, postmortem interval, quantitative opioid assay range or drug concentration series for this analysis. The genetic results cannot be combined with invented drug concentrations to reconstruct an active-metabolite exposure.
What was statistically significant?
Three of the 20 decedents with the selected prescription opioids detected had an ultrarapid prediction, compared with one of 29 with other opioids: 15% versus 3.4%. Fisher’s exact test gave p = 0.291, which was not statistically significant. The numerical difference should not be presented as a demonstrated association.
The overall five-of-74 finding included an ultrarapid prediction in a person without opioids detected whose death was classified as suicide by asphyxia. There was no directly measured veteran comparison population establishing the expected background frequency. Small numbers precluded a confident overrepresentation claim; a null result did not prove equal risk.
Co-detections and missing evidence limited causation
Twenty-four of the 49 opioid-positive people had non-opioid substances detected, including potentially sedating medications or alcohol. Detection alone did not quantify an interaction, but it complicated attributing death to a single genetic pathway.
The study examined no adherence proxy. Absence of a prescribed opioid could not establish deliberate nonadherence, and the records did not provide a validated dose–genotype–concentration model for individual death reconstruction.
What can the finding contribute?
Postmortem genotyping can add information about inherited drug-metabolism potential. It does not independently prove a toxic active-metabolite concentration, impairment, dosing error or cause of death. Interpretation must retain the exposure evidence, genetic test scope, prediction rules and full case context.
For a separate example of record-based exposure limits, see negative autopsy screens and medication exposure. The genetic study’s conclusions describe its historical data and do not constitute a current prescribing guideline.
Primary source
Boyle J, Stock CJ. CYP2D6 Expression in Veterans Experiencing Opioid Overdose: A Postmortem Review. Pharmacogenomics and Personalized Medicine. 2020;13:289–293. doi:10.2147/PGPM.S261424.
Leave a Reply