Tommaso D'Anna, Angelo Montana, Giulia La Magna, Marco Piraino, Antonella Pecorella, Maria Giovanna Scioli, Sonia Terriaca, Beatrice Belmonte, Augusto Orlandi, Stefania Zerbo, Francesco Paolo Busardò, Antonina Argo
Interpretation of opioid-related toxicity and death requires integration of exposure, measured toxicology, individual susceptibility, and alternative causes. We systematically reviewed human pharmacogenetic/pharmacogenomic (PGx) and multi-omic studies published from 1 January 2008 to March 2026, identified through MEDLINE/PubMed, Scopus, and Web of Science Core Collection. JBI tools informed risk-of-bias assessment, and findings underwent effect-size-oriented narrative synthesis and descriptive mapping. The 75-study corpus was predominantly PGx (51 studies, 68.0%); only 17 studies addressed transcriptomic/epigenetic, metabolomic, proteomic, or integrated approaches, and seven retained other molecular classifications. These domains therefore differ substantially in evidential maturity. The most coherent PGx findings concerned CYP2D6-dependent codeine and tramadol bioactivation, CYP2B6-dependent methadone disposition, and ABCB1-related tissue distribution. Omics findings were exploratory, with limited external validation and specificity for fatal causation; proteomics rested on a single investigation. Forensic interpretative directness was high or moderate in 36 studies and low or absent in 39. No nitazene-specific study met this review's molecular eligibility criteria. Certainty was very low for the four outcome-focused bodies assessed with GRADE; integration utility and confounder control were appraised narratively. Molecular findings may explain discordance between inherited susceptibility, parent-drug/metabolite patterns, and the observed phenotype. Downstream signatures may also reflect chronic exposure, terminal hypoxia, or postmortem change. Neither the descriptive evidence-map scores nor individual molecular markers provide validated estimates of forensic risk or establish cause of death independently.