Can a Forensic Scientist Identify a Poison From Its Metabolites Even When the Original Poison Is Gone?
How metabolites, degradation products, and modern mass spectrometry can reveal a hidden poisoning — and where that evidence stops short of proof.
A person is found dead at home. There is no obvious injury, no forced entry, nothing dramatic at the scene. The family mentions a history of depression. The police file gets marked as a probable suicide, pending toxicology.
Routine screening comes back clean. No common sedatives, no opioids, no alcohol worth mentioning. Several days have already passed between death and autopsy, and by the time the blood reaches the laboratory, whatever caused this death does not appear to be sitting in the sample anymore.
At this point, most investigations would stall. No parent compound, no obvious poison, no straightforward answer for the family or the court.
But the laboratory does not stop at "the compound wasn't found." It looks at what else is in the sample — and finds a second molecule. That molecule is not the poison itself. It is something the body built in response to the poison: an oxidised fragment, a conjugate, a breakdown product that only exists because something upstream of it once did too.
The difficult question is not whether the laboratory found a chemical. It is what that chemical actually means.
Could that second molecule be the forensic clue that survives after the original poison has already disappeared? Sometimes, yes. Sometimes, no. And the gap between those two answers is where forensic toxicology actually happens.
Yes, but only under specific conditions. When a poison is chemically unstable, rapidly metabolised, or present at very low concentration, its metabolites or degradation products can sometimes be detected even after the parent compound itself is gone. However, finding a metabolite documents biochemical transformation — it does not automatically prove exposure caused death. Interpretation depends on the specimen, the compound's metabolic pathway, postmortem changes, and corroborating evidence.
1. What Happens to a Poison After It Enters the Body?
Before asking where a poison "went," it helps to trace where it goes in the first place. Once a toxic substance enters the body — by ingestion, inhalation, injection, or absorption through skin — it typically moves through four broad stages: absorption into the bloodstream, distribution to tissues and organs, metabolism (biotransformation), and elimination.
The liver does most of the chemical heavy lifting. Enzyme systems there process foreign compounds (xenobiotics) in two broad phases. Phase I reactions — oxidation, reduction, hydrolysis — usually performed by cytochrome P450 enzymes, tend to make a molecule more reactive or more water-soluble as a first step. Phase II reactions then attach a water-soluble group onto the molecule or its Phase I product: glucuronidation, sulfation, acetylation, and related conjugation reactions. The end result is usually a metabolite that the kidneys can excrete into urine, or that leaves the body via bile and the gut.
This matters forensically because the parent poison is often only the starting point of a chemical trail. Aconitine, for instance, is metabolised by cytochrome P450 enzymes (notably CYP3A and CYP1A1/2) into a series of less toxic derivatives in liver microsomes, and forensic detection of aconitine and its early metabolites in blood and urine has become the practical basis for confirming Aconitum poisoning cases.[41] Methanol follows an even simpler two-step route: alcohol dehydrogenase first oxidises it to formaldehyde, which aldehyde dehydrogenase then converts to formic acid — the metabolite actually responsible for most of methanol poisoning's toxic effects, including the characteristic optic nerve and basal ganglia damage.[35]
The forensic question at every one of these steps is the same one: where does the poison go, and what does it leave behind when it gets there?
2. What Exactly Is a Metabolite? (And What It Isn't)
Forensic writing about toxicology often uses "metabolite," "biomarker," and "degradation product" as if they were interchangeable. They are not, and the article's core argument depends on keeping them separate.
| Term | What it means | Example |
|---|---|---|
| Parent compound | The original substance as ingested or administered, unchanged | Heroin (diacetylmorphine) |
| Metabolite | A product formed by the body's own enzymatic processing of the parent compound | 6-monoacetylmorphine (6-MAM), morphine |
| Conjugated metabolite | A Phase II metabolite chemically bonded to glucuronic acid, sulfate, or similar, for excretion | Morphine-3-glucuronide, morphine-6-glucuronide |
| Degradation product | A breakdown product formed by chemistry (heat, hydrolysis, microbial action) rather than living metabolism | Ethanol formed by postmortem microbial fermentation |
| Biomarker of exposure | Any measurable substance (parent, metabolite, or physiological change) that indicates contact occurred | Urinary dialkyl phosphates after organophosphate exposure[36] |
| Biomarker of effect | A measurable biological change caused by the exposure, not the chemical itself | Reduced cholinesterase activity after organophosphate exposure |
| Postmortem artifact | A compound that appears or changes concentration only because of processes that occur after death | Ethanol generated by putrefactive bacteria and yeasts[58,59] |
These distinctions are not academic hair-splitting. "The substance was present" is a chemistry statement. "The biological system shows evidence consistent with exposure" is a slightly stronger interpretive statement, still short of causation. "The evidence supports that the substance contributed to death" is a medicolegal conclusion that requires far more than a single analytical hit — it requires dose, timing, corroborating pathology, and exclusion of alternative explanations.
3. Why Might the Original Poison Disappear?
A poison can become undetectable for reasons that have nothing to do with each other, and forensic toxicologists have to reason through which one (or combination) applies in a given case.
- Rapid metabolism. Some compounds simply convert too fast. Heroin's half-life in blood is measured in minutes: enzymatic deacetylation strips one acetyl group within roughly two to four minutes, and the resulting 6-MAM is itself hydrolysed to morphine over the following minutes to an hour.[35,42]
- Chemical instability. Some parent compounds or even their immediate metabolites are unstable outside living tissue. A 2024 case report described 5-aminometonitazene and 5-acetamidometonitazene appearing in postmortem casework and raised the specific question of whether some nitro-substituted nitazene opioids degrade or convert further after death, complicating the distinction between an ante-mortem metabolite and a postmortem artifact.[56]
- Postmortem redistribution (PMR). After death, cell membranes break down and compounds diffuse passively from drug-reservoir organs (liver, lungs) into surrounding blood, so a peripheral blood concentration taken at autopsy may not reflect the concentration at the moment of death.[1,7,8]
- Specimen storage and degradation. Temperature, time in storage, and even the container matter. Aconitine, for example, has been shown to degrade substantially in postmortem specimens stored at room temperature for 30 days, while refrigeration or freezing largely prevents this loss.[64]
- Time between exposure and sampling. The longer the interval, the more of the original compound has already been metabolised, excreted, or degraded before the sample is even collected.
Detectability, in other words, depends on both chemistry and biology working together — and on how much time the investigation had before the specimen was even collected.
4. Can a Metabolite Be More Useful Than the Poison Itself?
Counter-intuitively, yes — in specific, well-documented circumstances. The clearest textbook example in forensic toxicology is heroin.
Heroin (diacetylmorphine) itself is essentially undetectable in blood within minutes of use. But 6-MAM, its first metabolite, is unique: it is not produced by morphine or codeine use through any other route, so its presence is treated as definitive proof that heroin, specifically, was used.[35] The catch is that 6-MAM itself has a short half-life — commonly cited as roughly 6 to 25 minutes before further hydrolysis to morphine — so it, too, may be gone by the time a postmortem sample is drawn.[40] Because of this, researchers have studied alternative specimens: 6-MAM detected in vitreous humour or urine, in the absence of blood 6-MAM, has been used to argue that more time passed between heroin use and death, since vitreous humour lacks the esterases that rapidly clear 6-MAM from blood.[34]
A parallel case exists in a 2026 Belgian case report on protonitazepyne, a nitazene-class synthetic opioid. Investigators documented pronounced accumulation of the parent compound in bile, with the parent and its metabolites entirely absent from urine — a hepatobiliary elimination pattern not previously described for this compound. The report also provided the first experimental confirmation of protonitazepyne acid as a genuine metabolite, illustrating how a case can hinge on choosing the right specimen once the "expected" one comes back empty.[98]
But here is the problem: a metabolite being more persistent does not automatically make it more informative. It only helps if the toxicologist knows what it means.
5. How Do Forensic Toxicologists Actually Find a Metabolite?
The general analytical pipeline runs from specimen to interpretation: extraction and sample preparation, chromatographic separation, ionisation, mass analysis, fragmentation, spectral interpretation, comparison against reference data, confirmation, and reporting.
| Method | Best suited for | Key strength | Key limitation |
|---|---|---|---|
| GC-MS | Volatile and semi-volatile compounds (alcohols, some pesticides) | Well-established, strong spectral libraries (e.g., NIST26)[13] | Requires volatility or derivatisation; less suited to large, polar molecules |
| LC-MS/MS | Targeted quantitation of known drugs and metabolites | High sensitivity and selectivity via multiple reaction monitoring[36] | Must know what to look for in advance |
| LC-HRMS | Suspect and non-targeted (untargeted) screening | Full-scan accurate-mass data enables retrospective re-analysis[22,24] | Complex data processing; identification is often tentative without a reference standard |
| GC-HRMS | Volatile unknowns and confirmatory structural work | Combines chromatographic separation with accurate mass | Less widely deployed than LC-HRMS in routine casework |
6. What Does a Mass Spectrometer Actually See?
At its simplest, a mass spectrometer measures the mass-to-charge ratio (m/z) of ionised molecules and molecular fragments. An intact ionised molecule (the precursor ion) can be selected and deliberately broken apart (fragmented) inside the instrument; the resulting fragments (product ions) form a spectrum that acts like a structural fingerprint.
Because a given molecule tends to fragment in characteristic, reproducible ways, the pattern of product ions provides real structural clues — but not automatically a unique identification. Structural isomers (molecules with the same formula, arranged differently) can produce very similar or even indistinguishable fragmentation spectra, which is one reason chromatographic retention time and other corroborating data matter as much as the spectrum itself.[73]
7. What Is Targeted Toxicology?
In targeted analysis, the laboratory already has a hypothesis: it is looking for a specific pesticide, a specific drug, or a specific known metabolite, and it calibrates the instrument (typically LC-MS/MS in multiple reaction monitoring mode) around that hypothesis. This gives excellent sensitivity and legally defensible quantitation, but by design it will miss anything outside the target list.[24]
8. What Is Suspect Screening?
Suspect screening sits between targeted and fully open-ended analysis. The laboratory does not know the exact compound but suspects a class or family — a batch of novel psychoactive substances, an emerging pesticide, a nitazene analogue — and searches full-scan HRMS data against a curated list of plausible candidates, even ones without confirmed reference standards on site.[21]
9. What Is Non-Targeted (Untargeted) Screening?
Untargeted screening reframes the question entirely. Instead of "is poison X present?", the laboratory asks: "what unusual chemical signals are present in this sample at all?" Full-scan, high-resolution accurate-mass data is collected without a predetermined target list, and the resulting features are compared against spectral libraries, with structurally plausible candidates generated computationally when no direct library match exists.[19,20]
One validated workflow evaluated untargeted HRMS screening across 132 toxicologically relevant compounds and reported a mean limit of identification around 8.8 ng/mL, later applying the method successfully to 31 routine casework samples.[20] Another, larger evaluation of a data-independent HRMS screening method was validated against 166 authentic impaired-driving and postmortem specimens, and explicitly cautioned that any laboratory adopting untargeted LC-HRMS screening needs to understand its strengths and limitations across different drug categories before relying on it in casework.[24,25]
Crucially, untargeted screening generates leads and candidates. It does not, by itself, produce a legally definitive identification — that still requires confirmation, ideally against a certified reference standard.
10. How Can Scientists Identify Something That Isn't in the Database?
This is where 2026's spectral libraries earn their reputation. NIST26, the current release of the NIST Mass Spectral Library, expanded its electron-ionisation library by roughly 35,000 compounds to more than 382,000 total, and grew its tandem (MS/MS) library to over 68,000 compounds represented by 3.2 million spectra — used across food, drugs, cosmetics, environmental samples, body fluids, and forensic casework.[13,14,16]
The library's most forensically relevant new feature is Hybrid Search, which is specifically built to help identify compounds that are not in the library outright — by finding library entries structurally similar to an unknown, based on shared fragment patterns and a known precursor mass, and then estimating what chemical modification might explain the difference.[11,12] Wiley's parallel 2026 release of its Mass Spectra of Designer Drugs database added over 700 new spectra and 400 new compounds spanning fentanyl variants, xylazine, nitazene opioids, and synthetic cannabinoids, reflecting how fast the novel psychoactive substance landscape is moving.[18]
11. Which Specimens Are Most Useful?
| Specimen | What it may reveal | Advantages | Limitations |
|---|---|---|---|
| Peripheral blood | Recent exposure; parent compound and short-lived metabolites | Preferred to minimise postmortem redistribution artefacts[6] | Parent compound may already be gone; subject to PMR |
| Urine | Cumulative metabolite excretion over hours to days | Longer detection window than blood for many metabolites[41] | Cannot establish timing, dose, or impairment alone |
| Vitreous humour | More stable, less redistribution-prone reference matrix | Anatomically isolated; resists bacterial contamination[112] | Smaller sample volume; some compounds diffuse poorly into it |
| Liver / kidney | Tissue-bound parent compound and metabolites | High concentrations for some compounds; useful reservoir | Prone to postmortem redistribution artefacts[6] |
| Hair | Weeks to months of chronic or repeated exposure | Long retrospective window via segmental analysis[65,68] | External contamination can mimic ingestion; cannot pinpoint acute dose[67] |
| Bile | Compounds eliminated via hepatobiliary rather than renal route | Can be positive when urine is negative for some drugs[98] | Less standardised interpretation data than blood or urine |
| Gastric contents | Undigested/unabsorbed material, route and timing clues | Useful for confirming oral ingestion and approximate timing | Does not confirm systemic absorption or toxic effect |
12. Can Urine Reveal a Poison That Blood No Longer Shows?
Sometimes. Because urinary metabolites accumulate and are excreted over a longer window than many parent compounds circulate in blood, urine can remain positive after blood has cleared. A comparative LC-MS/MS study of heroin metabolites in urine found that morphine-3-glucuronide offered a notably longer detection window than 6-MAM itself, which has a short urinary half-life of roughly 36 minutes.[41,42] Similarly, urinary dialkyl phosphate metabolites are the standard biomarker used to document organophosphate pesticide exposure, including in biomonitoring studies of Indian farmworkers.[51,52,53]
But a positive urine metabolite alone cannot establish impairment at a specific time, the exact dose taken, or precise timing relative to death. It documents that metabolism occurred — nothing more, on its own.
13. Can Hair Reveal Past Exposure?
Hair incorporates circulating compounds as it grows, creating a rough timeline that can be read back through segmental analysis — cutting the hair shaft into sections corresponding to different growth periods.[65] A 2025 case report presented what its authors describe as the first detailed toxicological study of fatal aconitine poisoning to include both hair and root plant-material analysis via LC-HRMS, suggesting the approach could help reconstruct repeated or chronic exposure scenarios in the future.[63]
The major caveat is external contamination. Pascal Kintz's forensic case series demonstrated that hair testing can produce false positives from environmental contamination or from postmortem incorporation via body fluids like putrefactive fluid, and concluded that despite decontamination washing procedures, this kind of artifact cannot always be excluded in postmortem cases — making interpretation of hair results a genuine analytical challenge rather than a simple yes/no readout.[67]
14. Can Metabolites Help Detect Pesticide Poisoning?
Organophosphate pesticides are metabolised relatively quickly — over hours to a few days — into a family of six common dialkyl phosphate (DAP) metabolites, which are the standard biomarker for documenting exposure in both occupational and forensic contexts.[54,55] A validated UFLC-MS/MS method developed in India specifically targeted these six DAP metabolites in urine, and a separate biomonitoring study of 120 Indian farmworkers found dimethylphosphate detectable in every single sample tested, with diethylthiophosphate and diethyl-dithiophosphate present in the large majority of workers who reported unsafe handling practices.[51,52,53] Personal protective equipment use was associated with significantly lower metabolite concentrations in that study.[51]
India-specific forensic method development also exists for individual organophosphate compounds: a 2023 method published through the Directorate of Forensic Science validated a microextraction and GC-MS approach for detecting dichlorvos (DDVP) directly in autopsy blood, stomach content, and liver samples, reflecting how routinely Indian forensic laboratories encounter fatal organophosphate poisoning.[82]
15. Can Metabolites Help Detect Drug Poisoning?
The heroin/6-MAM/morphine pathway discussed earlier remains the clearest published example of metabolite-based proof of a specific parent drug. The novel synthetic opioid literature extends the same logic to newer, more dangerous compounds. Nitazene-class opioids — a structurally distinct family from fentanyl analogues that has expanded rapidly in the illicit drug supply since 2019 — have required forensic toxicologists to characterise entirely new metabolic pathways essentially in real time, sometimes identifying a compound's metabolite in casework before its full metabolism was even mapped in the literature.[97,102,103] A 2026 retrospective study of 57 nitazene-positive Swedish autopsy cases over four years found metonitazene and protonitazene were the most frequently encountered analogues, and identified probable metabolites (N-desethyl, acetamido, and 5-amino forms) for metonitazene specifically.[100]
16. Can Metabolites Prove That a Person Died From a Poison?
This is the section where most misunderstandings happen, and where the article's central caution matters most.
| Metabolite evidence CAN support | Metabolite evidence CANNOT establish on its own |
|---|---|
| That biochemical transformation of a specific compound occurred | That the compound was the cause of death |
| Exposure to a specific parent compound (in some cases, e.g. 6-MAM proving heroin use)[35] | The exact dose or timing of exposure |
| A plausible route of elimination (renal vs. hepatobiliary)[98] | Whether the exposure was therapeutic, accidental, chronic, or acute |
| A relative timeline when compared across specimens with different clearance rates[34] | Impairment or toxic effect at a specific moment |
Finding a metabolite is not automatic proof of fatal poisoning. Exposure, absorption, metabolism, toxic effect, and contribution to death are five distinct concepts, and forensic interpretation has to work through all five rather than stopping at the first one that returns a positive result. A person on long-term therapeutic medication, someone with incidental environmental or dietary exposure, and someone who died from an acute fatal dose can all show the same metabolite on a screen — the concentration, the specimen pattern, the clinical history, and the autopsy findings are what separate those scenarios, not the metabolite's mere presence.
17. What Is Postmortem Redistribution?
Postmortem redistribution refers to changes in drug and metabolite concentrations that occur after death, driven by passive diffusion from reservoir organs, cell breakdown, blood pooling (hypostasis), and putrefaction.[1,7,8] A 2023 review found that the physicochemical properties of a compound — lipophilicity, protein binding, and volume of distribution — along with cadaveric changes and postmortem interval, all influence how much a measured concentration can differ from the concentration present at the moment of death.[6] This is precisely why peripheral blood is generally preferred over central or cardiac blood for interpretation, and why some laboratories use ratios like the liver-to-peripheral-blood (L/P) concentration ratio as a marker to flag when redistribution may have significantly affected a result.[8] NIST's Organization of Scientific Area Committees identified postmortem redistribution and distribution research as an ongoing, medium-priority research need in forensic toxicology as recently as 2025, underlining that this remains an active and unresolved area of the science rather than a fully solved problem.[10]
18. What Happens to Metabolites After Death?
Metabolite concentrations do not freeze at the moment of death. Some compounds continue to change chemically in the specimen itself, and some new compounds can form after death that were never present during life.
Postmortem ethanol production by bacteria and yeasts is the best-documented example of this problem. Multiple studies describe blood-alcohol concentrations forming after death purely from microbial fermentation, sometimes reaching levels high enough to be mistaken for genuine antemortem drinking; one case report documented a postmortem blood-alcohol concentration that rose from 0.18 g/L to 0.85 g/L over three days of storage, traced to contamination by E. coli, Enterococcus faecalis, and Candida parapsilosis.[113] Because of this well-known artifact, toxicologists increasingly rely on direct ethanol metabolites such as ethyl glucuronide and ethyl sulfate — compounds only produced through living metabolism, never by postmortem fermentation — specifically to validate that a measured ethanol result reflects genuine antemortem drinking rather than postmortem microbial production.[112]
The 2024 case report on 5-aminometonitazene and 5-acetamidometonitazene raised a parallel, more troubling possibility for nitazene opioids: that some nitro-substituted analogues might be chemically unstable enough to transform further after death, which would complicate any attempt to distinguish a genuine ante-mortem metabolite from a purely postmortem chemical artifact.[56]
19. What Is the Difference Between a Toxic Metabolite and an Exposure Biomarker?
A metabolite is not automatically dangerous, and a dangerous compound does not automatically produce a dangerous metabolite. Formic acid, methanol's principal metabolite, is directly responsible for methanol poisoning's characteristic toxicity — it is a toxic metabolite in the fullest sense.[35,50] By contrast, urinary dialkyl phosphates are themselves relatively low-toxicity compounds; their forensic value lies purely in what their presence indicates about exposure to the parent organophosphate, not in any toxicity of their own — making them exposure biomarkers rather than toxic metabolites.[54]
20. Real Forensic Cases: Metabolites in Published Casework
| Case / Study | Toxicant | Parent detected? | Metabolite detected? | Specimen | Method | Key finding / limitation |
|---|---|---|---|---|---|---|
| Bicker et al., Austria, 2013[40] | Aconitine (monkshood) | Yes, in most matrices | Mesaconitine used as internal standard; degradation tracked | 8 postmortem matrices incl. blood, urine, bile, vitreous | LC-MS/MS | Aconitine degraded substantially at room temperature over 30 days; refrigeration prevented loss |
| Case report, 2025[63] | Aconitine | Yes, in hair and root material | Study focused on parent alkaloid in novel matrices | Hair, plant root, blood | LC-HRMS, LC-MS/MS | First reported combined hair+root LC-HRMS study; hair method proposed for reconstructing chronic exposure |
| Rosano et al., USA, 2009[50] | Ethylene glycol | Yes, in 12 fatal ME cases | Glycolic acid | Postmortem blood | GC-MS (ion trap) | Parent and metabolite correlated poorly (R²=0.15) in two cases, showing metabolite alone is not a reliable stand-in for parent concentration |
| Case report[93] | Ethylene glycol | Yes, lethal levels confirmed | Calcium oxalate crystals in kidney/brain (metabolic marker) | Antemortem and postmortem samples, tissue histology | Comprehensive toxicologic testing + histopathology | Homicidal poisoning; toxic metabolite (oxalate) crystal deposition corroborated toxicology |
| Turkey autopsy series, 2026[30] | Methanol | Yes, 138 cases | Formic acid | Blood and vitreous humour | Quantitative biochemical + pathological correlation | Formic acid detected in ~89% of blood samples; organ pathology (cerebral edema, putaminal necrosis) correlated with metabolite levels |
| India (Uttarakhand) autopsy study[35] | Methanol | Case-dependent | Formic acid (via alcohol/aldehyde dehydrogenase pathway) | Postmortem blood, clinical records | Autopsy-based biochemical correlation | Delay between ingestion and metabolite-driven symptom onset can obscure diagnosis |
| Belgian case report, 2026[98] | Protonitazepyne (nitazene opioid) | Yes, notably in bile | Protonitazepyne acid (first experimental confirmation) | Blood, bile, urine | LC-MS/MS | Parent and metabolites entirely absent from urine despite fatal exposure; bile was the informative specimen |
| JAT case report, 2024[56] | Metonitazene-related nitro-nitazenes | Case-dependent | 5-aminometonitazene, 5-acetamidometonitazene | Postmortem blood | LC-MS/MS | Raised open question of whether these are genuine metabolites or postmortem degradation artefacts |
| Synthetic cannabinoid fatalities (cited in JAT review)[4] | 5F-ADB, 5F-PB-22, AB-CHMINACA | Case-dependent, often low/absent | Compound-specific metabolites used for interpretation | Postmortem blood | LC-MS/MS / LC-HRMS | Cited as an example requiring an evaluative-opinion framework due to interpretive uncertainty around synthetic cannabinoid metabolite significance |
| Milan retrospective review, 2018–2019[9] | Mixed (136 cases reviewed) | Case-dependent | Case-dependent | Multiple postmortem specimens | Mixed analytical methods | Toxicological analysis requested in only 10.3% of 1,323 total autopsies, illustrating real-world case selection constraints |
21. The Role of LC-MS/MS
LC-MS/MS remains the workhorse of confirmatory forensic toxicology because it combines liquid chromatography's separation power with tandem mass spectrometry's sensitivity and selectivity, typically operated in multiple reaction monitoring mode for definitive, quantitative confirmation of a known target.[43] Its central limitation is the same as any targeted method: it confirms what you already suspected, and it will not flag something entirely outside its target list.
22. The Role of High-Resolution Mass Spectrometry
HRMS platforms add accurate-mass, full-scan acquisition on top of chromatographic separation, which is what makes suspect screening, non-targeted analysis, and retrospective re-analysis of old data possible in the first place — a sample analysed today can, in principle, be re-interrogated months later once a new compound's reference spectrum becomes available.[22,23] Recent methodological work has focused heavily on improving in-silico structural annotation tools that help bridge the gap between "we detected an unusual mass" and "we can propose a plausible structure for it."[23]
23. Can AI Help Find a Poison Nobody Was Looking For?
Machine learning is being actively applied to HRMS interpretation in forensic toxicology — a 2026 review in the Journal of Analytical Toxicology surveys how AI models process molecular representations, spectral data, and embeddings to support general unknown screening, while explicitly framing itself as an overview of "past implementations and future directions" rather than a description of mature, universally deployed casework tools.[73] Related 2025 work applies AI methods specifically to novel psychoactive substance prediction and identification, including predicting mass spectra directly from proposed molecular structures.[75]
24. Can Metabolomics Reveal That Something Went Wrong Before the Poison Is Identified?
Metabolomics looks at the broader pattern of small-molecule changes in a biological sample, rather than searching for one specific compound. A 2024 study from Sweden's National Board of Forensic Medicine explored postmortem metabolomics as a high-throughput cause-of-death screening tool, motivated partly by declining global autopsy rates and their impact on cause-of-death accuracy.[33] Broader reviews describe metabolomics contributing to postmortem interval estimation and to flagging biochemical perturbations consistent with drug-induced multi-organ injury — hepatic injury patterns linked to methamphetamine, heroin, and other substances have all been studied through this lens.[26]
The consistent caveat across this literature is that metabolomics remains, for cause-of-death applications specifically, a promising but still-developing research area rather than routine casework infrastructure; reviewers repeatedly note that more validation studies are needed before broader forensic deployment.[27,32]
25. What Can Fool a Forensic Toxicologist?
| Source of uncertainty | Why it matters |
|---|---|
| Postmortem redistribution | Measured concentration may not reflect the concentration at time of death[1,6] |
| Postmortem microbial production | Ethanol (and related volatiles) can be generated after death by bacteria/yeast, mimicking antemortem intake[110,111,113] |
| Matrix effects / ion suppression | Co-eluting compounds can suppress or enhance ionisation, distorting quantitation |
| Compound instability | Some parent compounds and metabolites degrade during storage even under good conditions[64] |
| External contamination (esp. hair) | Environmental or postmortem fluid contact can produce false positives despite decontamination washing[67] |
| Structural isomers | Different compounds can produce near-identical fragmentation spectra, requiring retention time and other corroboration[73] |
| Insufficient reference standards | Novel or rare compounds may lack certified standards needed for confirmatory identification[62] |
| Therapeutic / dietary / environmental exposure | A positive result may reflect legitimate medication, food, or incidental contact rather than poisoning |
26. Why Can Two Laboratories Get Different Toxicology Results?
Differences in extraction technique, instrumentation, calibration, limits of detection, available reference standards, method validation status, and reporting thresholds can all produce genuinely different results between two competent laboratories examining related specimens from the same case. This is precisely why standardised validation frameworks exist — the ABS Standard 036 "Standard Practices for Method Validation in Forensic Toxicology" is cited repeatedly across the untargeted-screening literature as the benchmark laboratories are expected to meet for qualitative methods, covering interference, carryover, and limit of detection performance specifically.[24] Quality assurance frameworks exist precisely to make results across laboratories more comparable, not to eliminate all variation.
27. Can a Poisoning Investigation Be Solved Without Finding the Original Poison?
Sometimes. When a compound is unstable or rapidly metabolised but its metabolic pathway is well characterised — as with heroin/6-MAM/morphine, or methanol/formic acid — a metabolite-only result combined with concentration, specimen pattern, and clinical or scene evidence can support a confident conclusion.[35,50] When the pathway is poorly characterised, as with several newly emerging nitazene analogues, the same absence of parent compound leaves genuine ambiguity that published case reports openly acknowledge rather than resolve.[56,98] Indirect evidence can contribute meaningfully to an investigation; it cannot always substitute for the parent compound, and honest forensic reporting says so explicitly when that is the case.
28. Indian Forensic Context
India's forensic toxicology infrastructure runs through Central Forensic Science Laboratories (CFSL) and State Forensic Science Laboratories (SFSL), which handle viscera, biological fluids, and suspected poisons using HPLC, HPTLC, and GC-MS as standard equipment.[86,87] Method-development papers published through India's Directorate of Forensic Science — such as the 2023 dichlorvos microextraction-GC-MS method — show Indian laboratories actively building validated, casework-ready methods for the organophosphate poisonings that remain a major cause of fatal intoxication regionally.[82] Indian research groups (including ICMR's National Institute of Nutrition) have also independently developed and validated UFLC-MS/MS methods for the standard six dialkyl phosphate metabolites used to document organophosphate exposure, and Indian biomonitoring studies of farmworker populations have applied LC-MS/MS DAP metabolite testing at meaningful scale.[51,54,55]
Based on the available literature, metabolite-based biomonitoring for organophosphate exposure appears to be an active area of validated Indian method development and occupational-health research, while broader untargeted or non-targeted metabolite-screening workflows — of the kind increasingly used in European and North American forensic toxicology — are not yet documented as routine casework practice in the sources reviewed here. This distinction matters: it would be inaccurate to claim routine operational use of non-targeted metabolomic screening in Indian forensic laboratories without stronger direct evidence, and no such evidence appeared in this research.
Recognised challenges echoed across the international literature apply with particular force in resource-constrained settings: the availability of certified reference standards for novel and regional compounds, instrumentation access at the district and state level, laboratory capacity relative to national caseload, ongoing method validation, and specialised training in interpreting metabolite evidence rather than simply generating it.
29. What Does 2026 Look Like for Poison Detection?
| Status | Technology | Evidence basis |
|---|---|---|
| Established | Targeted LC-MS/MS confirmation | Standard confirmatory method across forensic toxicology casework[36,43] |
| Established | GC-MS with NIST spectral library matching | Decades of use; NIST26 released 2026 with expanded EI/MS2 coverage[13,14] |
| Emerging | LC-HRMS non-targeted / suspect screening | Validated in multiple recent method papers; adoption growing in casework labs[20,22,24] |
| Emerging | NIST26 Hybrid Search for near-match identification | Newly released 2026 feature; early-stage forensic adoption[11,12] |
| Emerging | Postmortem metabolomics for cause-of-death screening | Active research programmes; not yet routine casework tool[33] |
| Experimental | Machine-learning-assisted HRMS unknown annotation | 2025–2026 reviews describe method development, not deployed casework standard[73,75] |
| Experimental | Multi-omics (proteomics, microbiomics) for PMI/cause-of-death | Reviewed as a research direction with acknowledged validation gaps[26] |
30. The Future
Could forensic toxicology eventually identify a poison from its biological "footprint" even when the original molecule is completely absent? The trajectory of the field — toward metabolomics, exposomics, toxicogenomics, and AI-assisted interpretation of increasingly large spectral libraries — points that direction. But every source reviewed for this article that discusses these emerging tools frames them as promising and actively developing, not as settled forensic capability. Predictions about where this goes belong in the "future" category, clearly separated from what a laboratory can defensibly testify to today.
31. The Final Question, Answered Honestly
Can a forensic scientist identify a poison from its metabolites even when the original poison is gone? Yes, in some circumstances — when the metabolic pathway is well characterised, the metabolite is stable enough to survive to sampling, the right specimen is chosen, and the analytical method has the sensitivity and reference data to confirm it. Heroin's conversion to 6-MAM and morphine, and methanol's conversion to formic acid, are two of the clearest, most literature-supported examples of exactly this.[35,50]
No, not universally. Some compounds and their metabolites degrade beyond recovery, some metabolic pathways are still being mapped in real time as new substances emerge, some specimens are unavailable or compromised, and even a successfully identified metabolite documents transformation — not automatically cause of death. The honest answer sits between those two positions, and which side of it a given case falls on depends entirely on the chemistry, the specimens available, and the quality of the analytical and interpretive work behind the result.
Key Takeaways
- A metabolite proves biochemical transformation occurred — not automatically that a poison caused death.
- Heroin/6-MAM/morphine and methanol/formic acid are the field's clearest examples of forensically decisive metabolite evidence.
- Postmortem redistribution and postmortem artifact formation (e.g., microbial ethanol production) can both mimic and obscure real toxicological findings.
- Untargeted LC-HRMS screening and NIST26's Hybrid Search extend what can be identified, but generate candidates requiring confirmation, not automatic answers.
- AI-assisted mass spectrometry interpretation is an active 2025–2026 research area, not yet a substitute for confirmed, reference-standard-based identification.
- Specimen choice matters enormously: bile, vitreous humour, or hair can sometimes succeed where blood or urine come back negative.
- India's documented metabolite-based forensic toxicology strength is strongest in organophosphate pesticide biomonitoring; broader non-targeted screening is not established as routine casework here.
Frequently Asked Questions
Can a poison disappear from the body?
Yes. Rapid metabolism, chemical instability, elimination, and postmortem degradation can all reduce a parent compound below detectable levels, sometimes within minutes of exposure for compounds like heroin.[35]
Can toxicologists detect metabolites after the poison itself is gone?
Often, yes, if the metabolite is more stable or persists longer than the parent compound, and if the right specimen and analytical method are used.[34,98]
What is the difference between a poison and its metabolite?
The poison (parent compound) is the substance as originally taken; the metabolite is what the body's enzymes convert it into. They can differ significantly in toxicity, stability, and detection window.
Can urine reveal poisoning after blood tests are negative?
Sometimes, because urinary metabolites can accumulate and persist longer than a rapidly cleared parent compound in blood.[41]
Can mass spectrometry detect unknown poisons?
It can detect unusual signals and generate structural candidates through non-targeted screening and hybrid searching, but confirmed identification of a truly unknown compound typically still requires a reference standard.[19,20]
What is untargeted toxicology?
An analytical approach that scans a sample broadly for any unusual chemical signals, rather than searching only for a predetermined list of suspected substances.[21]
What is high-resolution mass spectrometry?
A mass spectrometry technique that measures molecular mass with very high precision, enabling full-scan data acquisition, suspect screening, and retrospective re-analysis of stored data.[22,23]
Can metabolites prove poisoning?
Not automatically. A metabolite documents exposure and biochemical transformation; proving fatal poisoning requires dose, timing, corroborating pathology, and exclusion of alternative explanations.
Can a metabolite identify the original poison?
In some documented cases, yes — 6-MAM uniquely identifies heroin use, for example, because no other route produces it.[35]
Can postmortem redistribution affect toxicology results?
Yes, significantly. Concentrations measured at autopsy can differ substantially from concentrations at the time of death due to passive diffusion and cadaveric changes.[1,6]
Can decomposition change toxicology findings?
Yes. Microbial and chemical processes during decomposition can produce new compounds (like postmortem ethanol) or degrade existing ones, complicating interpretation.[111,113]
Can hair reveal previous poison exposure?
Often, through segmental analysis reflecting weeks to months of exposure, though external contamination can produce misleading positives.[65,67]
Can AI detect unknown poisons?
AI and machine learning tools are being actively developed to assist HRMS interpretation, but current literature frames them as candidate-generation aids, not confirmed-identification replacements.[73,75]
What is toxicometabolomics?
A branch of metabolomics focused specifically on detecting toxin- or drug-induced changes in the body's broader metabolic profile, used in research toward cause-of-death screening.[26,33]
Can a forensic laboratory identify a poison that isn't in its database?
Sometimes, using techniques like hybrid searching that find structurally similar library matches, though this generates candidates requiring further confirmation rather than an automatic identification.[11,12]
Why can two toxicology laboratories produce different results?
Differences in extraction methods, instrumentation, calibration, reference standards, and validated detection limits can all produce genuinely different results between competent laboratories.[24]
Is metabolite evidence accepted in forensic investigations?
Yes, extensively, particularly where the metabolic pathway is well characterised, but it is weighed alongside specimen quality, concentration, and corroborating evidence rather than treated as automatically conclusive.
Is metabolite-based poison detection used in India?
Yes, particularly for organophosphate pesticide exposure via validated dialkyl phosphate metabolite testing; broader non-targeted metabolomic screening is not documented as routine Indian forensic casework in the literature reviewed here.[51,54,82]
References
Peer-Reviewed Research
- Drummer OH. Post-mortem redistribution of drugs and other factors affecting interpretation: A review. WIREs Forensic Science. 2023;5(3):e1480. DOI: 10.1002/wfs2.1480.
- Abdelaal GMM, Hegazy NI. Postmortem redistribution of drugs: a literature review. Forensic Sci Med Pathol. 2023;20(4):1483–1490. DOI: 10.1007/s12024-023-00709-z.
- Di Candia D, Giordano G, Boracchi M, Zoja R. Postmortem forensic toxicology cases: A retrospective review from Milan, Italy. J Forensic Sci. 2022. DOI: 10.1111/1556-4029.15050.
- Hupatz H, Rahu I, Wang W-C, Peets P, Palm EH, Kruve A. Critical review on in silico methods for structural annotation of chemicals detected with LC/HRMS non-targeted screening. 2024. DOI: 10.1007/s00216-024-05471-x.
- Screening for Forensically Relevant Drugs Using Data-Independent High-Resolution Mass Spectrometry. Chem Res Toxicol. DOI: 10.1021/acs.chemrestox.3c00379.
- Validation of a non-targeted method devoted to identification and quantitation of toxicologically relevant compounds in plasma with HRMS. 2023. Via ScienceDirect.
- Non-targeted analysis (NTA) and suspect screening analysis (SSA): a review of examining the chemical exposome. PMC10403360.
- Swan S, Sarkisian M, Pasin D, Rodda LN. Applications of machine learning for the general unknown screening of HRMS data within forensic toxicology. J Anal Toxicol. 2026;50(3):bkaf109. DOI: 10.1093/jat/bkaf109.
- AI Methods for New Psychoactive Substance (NPS) Design and Analysis. Psychoactives (MDPI). 2025;6(2):17.
- Zhang J, et al. The Role of Protein Degradation in Estimation Postmortem Interval and Confirmation of Cause of Death in Forensic Pathology: A Literature Review. PMC10855206.
- Applications of Metabolomics in Forensic Science with Focus on Forensic Toxicology and Time-of-Death Estimation. Metabolites. 2021;11(12):801. DOI: 10.3390/metabo11120801. PMID: 34940558.
- Research Progress of Metabolomics in Forensic Pathology. PMID: 32705848.
- Applications of Metabolomics in Forensic Toxicology and Forensic Medicine. Int J Mol Sci. 2021;22:3010. DOI: 10.3390/ijms22063010. PMC8002074.
- Ward LJ, Kling S, Engvall G, Söderberg C, Kugelberg FC, Green H, Elmsjö A. Postmortem metabolomics as a high-throughput cause-of-death screening tool for human death investigations. iScience. 2024. DOI: 10.1016/j.isci.2024.109794.
- Postmortem distribution of heroin metabolites in femoral blood, liver, cerebrospinal fluid, and vitreous humor. PMID: 15189677.
- Postmortem tissue distribution of morphine and its metabolites in a series of heroin-related deaths. PMID: 30156385.
- A Quantitative and Comparative Study of Heroin-Related Metabolites in Different Postmortem Fluids and Tissues. PMC11945869.
- Post-Mortem Analysis of Heroin Biomarkers, Morphine and Codeine in Stomach Wall Tissue in Heroin-Related Deaths. Toxics. 2022;10:473. DOI: 10.3390/toxics10080473.
- Identifying cases of heroin toxicity where 6-acetylmorphine (6-AM) is not detected by toxicological analyses. PMC4967084.
- Comparison of the Detection Windows of Heroin Metabolites in Human Urine Using Online SPE and LC-MS/MS. PMID: 31095707.
- Heroin-Related Compounds and Metabolic Ratios in Postmortem Samples Using LC–MS-MS. J Anal Toxicol. 2021;45(3):215.
- Postmortem Methanol and Formic Acid Levels and Their Pathological Correlates: Diagnostic Implications from an Autopsy Study. Çevik FE, Bugra A, Sahin HC, et al. Diagnostics. 2026;16(2):233. DOI: 10.3390/diagnostics16020233.
- Epiglottis Cartilage, Costal Cartilage, and Intervertebral Disc Cartilage as Alternative Materials in the Postmortem Diagnosis of Methanol Poisoning. PMC9964933.
- Fatal methanol poisoning with different clinical and autopsy findings: Case report and literature review. 2021. Via ScienceDirect.
- Ferrari LA, Arado MG, Nardo CA, Giannuzzi L. Post-mortem analysis of formic acid disposition in acute methanol intoxication. Forensic Sci Int. 2003;133(1–2):152–158. PMID: 12742704.
- The relationship of methanol and formate concentrations in fatalities where methanol is detected. PMID: 17868264 (review context via Kugelberg & Jones, 2007).
- Methanol Poisoning: An Autopsy-Based Study at the Tertiary Care Center of Uttarakhand, India. PMC9236883.
- Lari S, Pandiyan A, Vanka J, Jee B, Yamagani P, Kumar SB, Naidu M, Jonnalagadda P. Residual Dialkyl Phosphate Metabolite Concentrations of Organophosphate Pesticides Among Indian Farmworkers: Implication of Exposure and Hazard Assessment. PMID: 37696790.
- Kumar D, Sinha SN, Vasudev K. Development and Validation of a New UFLC–MS/MS Method for the Detection of Organophosphate Pesticide Metabolites in Urine. Molecules. 2023;28:5800. DOI: 10.3390/molecules28155800.
- DDVP by VA-DLLME-GC-MS in autopsy blood, stomach content, and liver. J Chromatogr B. 2023;1215:123582. Via Directorate of Forensic Science, India (dfs.nic.in).
- Distribution of Aconitum alkaloids in autopsy cases of aconite poisoning. Via ResearchGate.
- Bicker W, Monticelli F, Bauer A, Roider G, Keller T. Quantification of aconitine in post-mortem specimens by validated liquid chromatography-tandem mass spectrometry method: Three case reports on fatal ‘monkshood’ poisoning. Drug Test Anal. 2013;5:753–762. DOI: 10.1002/dta.1501.
- Research Progress of Aconitine Toxicity and Forensic Analysis of Aconitine Poisoning. Forensic Sciences Research. 2018;5(1):25.
- Herbal Medicine Must Be Treated with Care—A Case Report on Aconitine. Forensic Sciences (MDPI). 2021;1(1):5.
- A deadly root and the science behind it: LC-HRMS and LC-MS/MS analysis in an aconite-induced suicide. PMID: 41032181 (2025).
- Usman et al. Forensic toxicological analysis of hair: a review. Egyptian Journal of Forensic Sciences. 2019;9:17. DOI: 10.1186/s41935-019-0119-5.
- Kintz P. Segmental hair analysis can demonstrate external contamination in postmortem cases. Forensic Sci Int. 2012;215(1–3):73–76. PMID: 21354729.
- Kintz P. Hair analysis in forensic toxicology. WIREs Forensic Science. 2019. DOI: 10.1002/wfs2.1196.
- Rosano TG, Swift TA, Kranick CJ, Sikirica M. Ethylene Glycol and Glycolic Acid in Postmortem Blood from Fatal Poisonings. J Anal Toxicol. 2009;33(8):508–513. DOI: 10.1093/jat/33.8.508. PMID: 19874660.
- Novel Technique for Simultaneous Ethylene Glycol and Its Metabolites Determination in Human Whole Blood and Urine Samples Using GC–QqQ–MS/MS. J Xenobiot. 2024;14:65. DOI: 10.3390/jox14030065. PMC11417905.
- Homicidal ethylene glycol intoxication: a report of a case. PMID: 16738434.
- Fatal mono-intoxication with novel synthetic opioid protonitazepyne (N-pyrrolidino protonitazene). 2026. Via ScienceDirect.
- A case of fatal intoxication with the novel synthetic opioid N-pyrrolidino protonitazene. Int J Legal Med. 2025. DOI: 10.1007/s00414-025-03618-8.
- Fatal intoxications with nitazenes: postmortem findings, toxicology, and in vitro characterization at the μ-opioid receptor of six nitazene analogues. Arch Toxicol. 2026. DOI: 10.1007/s00204-026-04407-x.
- Identification of 5-aminometonitazene and 5-acetamidometonitazene in a postmortem case: are nitro-nitazenes unstable? J Anal Toxicol. 2024;48(9):691.
- Kriikku P, et al. Post-Mortem Identification and Toxicological Findings of Fluetonitazepyne and Isotonitazepyne. Drug Test Anal. 2025. PMC12580158.
- The emergence of nitazenes: a new chapter in the synthetic opioid crisis. Arch Toxicol. 2025. DOI: 10.1007/s00204-025-04102-3.
- Postmortem in vitro ethanol production — It could be more common than we think! 2017. Via ScienceDirect.
- Preanalytical factors influencing the results of ethanol analysis in postmortem specimens. J Anal Toxicol. 2024;48(1):9.
- Kugelberg FC, Jones AW. Interpreting results of ethanol analysis in postmortem specimens: a review of the literature. Forensic Sci Int. 2007;165(1):10–29. PMID: 16782292.
- Vitreous Humor Versus Peripheral Blood for Accurate Estimation of Antemortem Alcohol Levels: A Systematic Review of Postmortem Human Studies. PMC12793923.
Reviews
- Compound identification in forensic toxicological analysis with untargeted LC-MS-based techniques. Via ResearchGate.
- Forensically Relevant Drug Screening Using LC–HRMS (expert interview). Chromatography Online. 2024.
- Artificial Intelligence in Forensic Toxic Science: Emerging Trends and Analytical Techniques. DOI: 10.22214/ijraset.2025.74270.
- AI redefines mass spectrometry chemicals identification: retention time prediction in metabolomics and for a Human Exposome Project. Front Public Health. 2025;13:1687056. PMC12647117.
Government / Institutional Sources
- NIST. NIST26: Updates to the NIST Electron Ionization and Tandem Mass Spectral Libraries. nist.gov, 2026.
- NIST. Electron Ionization Library Component of the NIST/EPA/NIH Mass Spectral Library and NIST GC Retention Index Database. nist.gov.
- NIST expands its library of ‘chemical fingerprints’ to identify unknown substances. EurekAlert!, June 2026.
- From crime labs to space rocks, NIST helps identify mystery chemicals. Nextgov/FCW, 2026.
- Forensic Tox R&D Need: Postmortem Distribution and Redistribution (2025). NIST / OSAC Forensic Toxicology Subcommittee, nist.gov.
- Central Forensic Science Laboratory, Chandigarh — official functions and accreditation. cfslchandigarh.gov.in.
- State Forensic Science Laboratory (SFSL), Jharkhand — Toxicology Division duties. Department of Police, Government of Jharkhand.
