Toxics screening (toxicomics, service)
01Overview and value chain
Markers: [EC: ICH S7A/S7B & OECD Principles of GLP | OECD: Bio-pharmaceuticals & bioinformatics | Regulator: FDA (USA), EMA (EU), NMPA (China)]
Toxics screening (toxicomics) is the contract safety-assessment service that de-risks a drug candidate before it ever reaches a human: a pharma sponsor hands a test article to a CRO, which runs in-silico ADMET prediction, a battery of in-vitro assays (drug-induced-liver-injury, hERG cardiotoxicity, CYP induction, high-content imaging), then GLP in-vivo studies in rodents and non-rodents, and returns the IND-enabling toxicity package the regulator demands. The field is consolidating around a few large GLP CROs — Charles River runs about 15 in-vivo safety sites worldwide, WuXi AppTec’s 2025 revenue grew 15.8% year-on-year (21.4% for continuing operations) on strong CDMO demand into 2026, and Syngene extended its Bristol Myers Squibb research collaboration through to 2035 — while in-silico modelling (Certara’s EMA-qualified Simcyp Simulator Version 25, released for 2026 regulatory submissions) and toxicogenomics (Labcorp/Stemina’s devTOX quickPredict human-iPSC developmental-toxicity assay) move the early tiers onto the computer and into high-content cell models. A full GLP rat study still costs in the hundreds of thousands of USD and runs for months, so the service stacks a cheap in-silico and in-vitro funnel in front of the expensive animal work, each tier gating the next. Toxicomics — omics readouts attached to tox endpoints (transcriptomic, metabolomic signatures of stress) — is the layer that earns this Industry its bioinformatics capability tag, turning a pass/fail animal result into a mechanistic, reproducible molecular signature.
The key directions of toxics screening (toxicomics) are:
- In-vivo GLP toxicology: rodent and non-rodent IND-enabling safety studies (acute, repeat-dose, reproductive, carcinogenicity) run under OECD GLP principles at large CROs (Charles River, Eurofins, WuXi AppTec, Syngene).
- In-vitro ADMET and high-content screening: hepatocyte DILI models, hERG channel cardiotoxicity, CYP enzyme induction (now miniaturised to 96-well), and high-content imaging that flag off-target toxicity before animal dosing (Cyprotex/Labcorp).
- In-silico / model-informed toxicology: EMA-qualified PBPK modelling (Certara Simcyp) and QSAR to predict drug-drug interactions, pharmacokinetics and hepatotoxicity in software, feeding model-informed drug development (MIDD).
- Toxicogenomics and developmental toxicity: human-iPSC assays (devTOX quickPredict) and metabolomic signatures that read developmental toxicity and stress pathways off a cell panel — the toxicomics core that replaces animal endpoints with molecular signatures.
Sectoral value chain
[Test article & GLP setup] ──> [In-silico ADMET triage] ──> [In-vitro assay panel]
│
(gate: kill overt toxicants)
│
▼
[IND-enabling tox report] <─── [Bioanalysis & DMPK] <─── [In-vivo GLP study]Value chain levels
| Level | Description | Key inputs/outputs |
|---|---|---|
| Test-article receipt & GLP setup | Receipt of the test article under chain-of-custody, GLP facility setup, dose-formulation analysis and stability. | In: Test article, GLP facility, formulation lab. Out: Dosed, stability-verified test article. |
| In-silico ADMET triage | PBPK (Simcyp) and QSAR prediction of DDI, pharmacokinetics and hepatotoxicity to rank and deprioritise overt toxicants. | In: Molecular structure, PBPK/QSAR software. Out: Predicted ADMET profile and risk rank. |
| In-vitro assay panel | Hepatocyte DILI, hERG cardiotox, CYP induction (96-well) and high-content imaging across a cell panel. | In: Cell lines, assay reagents, high-content imager. Out: In-vitro IC50/tox-flag dataset. |
| In-vivo GLP study | Rodent and non-rodent acute and repeat-dose toxicity studies under OECD GLP, with toxicokinetic sampling. | In: Dosed animals, GLP study director, vivarium. Out: NOAEL, target-organ toxicity, toxicokinetics. |
| Bioanalysis & DMPK | LC-MS/MS quantitation of plasma drug levels and metabolites, with Phoenix WinNonlin pharmacokinetic modelling. | In: Plasma samples, LC-MS/MS, WinNonlin. Out: PK parameters and metabolic profile. |
| Regulatory tox report | Integrated IND-enabling toxicity package against ICH M3(R2)/S7A/S7B and OECD GLP, ready for FDA/EMA/NMPA submission. | In: All tiers’ data, regulatory scientists. Out: Signed GLP toxicity report. |
Cross-cutting technologies of the sector:
- GLP quality system: OECD Principles of Good Laboratory Practice (and FDA 21 CFR Part 58) govern every animal study — protocol, chain-of-custody, audit trail and a study director accountable for the signed report.
- ICH safety guidelines: ICH S7A/S7B (safety pharmacology, hERG), M3(R2) (nonclinical safety studies needed for human trials) and S5 (reproductive toxicology) fix the assay battery regulators expect.
- 3R and Virtual Control Groups: replacement, reduction and refinement of animal use — Charles River’s peer-reviewed Virtual Control Groups let historical control data stand in for concurrent controls, cutting animal counts without losing statistical power.
02US
The United States hosts the two largest pure-play tox CROs and the in-silico modelling leader, and the FDA sets the IND-enabling toxicity expectations the global industry designs to.
Charles River in-vivo GLP, Certara Simcyp, FDA/ICH guidance
- Charles River Laboratories: the largest preclinical safety CRO, running roughly 15 in-vivo safety sites and publishing peer-reviewed evidence (April 2026) for Virtual Control Groups that reduce animal use in nonclinical toxicology.
- Certara: the biosimulation leader whose EMA-qualified Simcyp Simulator Version 25 (2026) extends PBPK modelling to streamline regulatory submissions and drug-drug-interaction assessment across discovery, DMPK and clinical pharmacology.
- FDA and ICH alignment: the FDA’s IND-enabling framework follows ICH M3(R2) and S7A/S7B, so the US CRO assay battery is the de facto global standard.
03CN
China is the fastest-growing tox CRO market, anchored by a single dominant domestic platform whose 2025 results signal the sector’s 2026 trajectory, under NMPA oversight that has converged on ICH-aligned safety guidelines.
WuXi AppTec DMPK/safety, NMPA-ICH convergence, CDMO pull-through
- WuXi AppTec (603259.SH): the leading Chinese preclinical CRO, reported 2025 revenue up 15.8% year-on-year (21.4% for continuing operations) with strong CDMO demand fueling 2026 growth, offering integrated DMPK and safety assessment to global sponsors.
- NMPA-ICH convergence: since adopting ICH guidelines, China’s NMPA accepts the same nonclinical safety package, letting a Chinese GLP study serve multinational submissions.
- CDMO pull-through: domestic capacity expansion in toxicology and DMPK tracks the broader outsourcing of early-stage drug discovery to Chinese platforms.
04EU
The European Union contributes the original in-vitro ADMET specialist and a large GLP testing network, under EMA oversight that has made PBPK modelling a regulatory-accepted submission route.
Cyprotex in-vitro ADMET, Eurofins GLP network, EMA-qualified PBPK
- Cyprotex (Labcorp): the in-vitro ADMET and high-content-toxicology leader (Macclesfield, UK), now within Labcorp — its devTOX quickPredict human-iPSC developmental-toxicity assay and miniaturised 96-well CYP induction workflows were both presented in 2026, alongside a new in-vitro liver model for drug-induced-liver-injury prediction at the MPS World Summit 2026.
- Eurofins Scientific: the Luxembourg-based testing group runs a BioPharma Product Testing network (Germany, Finland and beyond) delivering GLP toxicology and leachables/extractables assessment (USP <1664.3>) for ophthalmic and parenteral drugs.
- EMA-qualified modelling: EMA qualifies PBPK platforms (Simcyp) so that model-informed DDI and hepatotoxicity prediction now carry weight in European submissions, not just internal triage.
05Leading companies and research institutes
| Company / Institute | Country | Key products / platforms | Tech features | Status 2026 |
|---|---|---|---|---|
| Charles River Laboratories | 🇺🇸 USA | In-vivo GLP toxicology / VCGs | ~15 in-vivo safety sites; AI-assisted workflows; peer-reviewed Virtual Control Groups | commercial |
| Certara | 🇺🇸 USA | Simcyp Simulator v25 | EMA-qualified PBPK/MIDD; in-silico DDI, ADMET and hepatotox prediction | commercial |
| Cyprotex | 🇬🇧 UK | devTOX quickPredict / ADMET panel | Labcorp in-vitro arm; iPSC developmental tox, DILI liver model, 96-well CYP | commercial |
| Eurofins Scientific | 🇱🇺 Luxembourg | GLP toxicology / BioPharma Product Testing | GLP safety studies; USP <1664.3> leachables/extractables for ophthalmics | commercial |
| WuXi AppTec | 🇨🇳 China | DMPK / safety assessment | 603259.SH; 2025 revenue +15.8% YoY (continuing +21.4%); ICH-aligned GLP | commercial |
| Syngene International | 🇮🇳 India | Integrated safety assessment | BMS collaboration extended to 2035; PDE/HBEL toxicology; research CRDMO | commercial |
06Tech stack and innovations
The tox-screening service stack funnels a candidate through a cheap in-silico and in-vitro triage before the expensive GLP animal study, each tier killing toxicants the next tier no longer has to test.
- In-silico ADMET and PBPK modelling:
- Certara’s Simcyp Simulator Version 25 (EMA-qualified, 2026) predicts drug-drug interactions and pharmacokinetics from physiology and in-vitro data, feeding model-informed drug development (MIDD) and regulatory submissions.
- QSAR models flag structural toxicity alerts (hERG, mutagenicity) directly from molecular structure, ranking compounds before any wet-lab assay.
- In-vitro assay panel and high-content imaging:
- Cyprotex/Labcorp runs hepatocyte drug-induced-liver-injury models, hERG cardiotoxicity assays and CYP enzyme induction (now miniaturised from 24-well to 96-well to cut cost and boost throughput).
- High-content imaging and the devTOX quickPredict human-iPSC assay read developmental toxicity (the developmental-toxicity threshold via the ornithine/cystine ratio) and stress signatures off a cell panel.
- GLP in-vivo studies and bioanalysis:
- Rodent and non-rodent acute, repeat-dose, reproductive and safety-pharmacology studies under OECD GLP, with Charles River’s Virtual Control Groups cutting concurrent-control animal numbers using historical data.
- LC-MS/MS bioanalysis plus Phoenix WinNonlin pharmacokinetic modelling deliver the toxicokinetics and NOAEL that anchor the signed GLP report.
07Value chains and production pipelines
Industrial pipeline of an IND-enabling tox screening package (OECD GLP / ICH M3(R2) / FDA 21 CFR Part 58)
┌───────────────────────────┐ ┌───────────────────────────┐
│ 1. Test-article receipt & │ ───> │ 2. In-silico ADMET triage │
│ GLP facility setup │ │ (Simcyp PBPK / QSAR) │
└───────────────────────────┘ └───────────────────────────┘
│
▼
┌───────────────────────────┐ ┌───────────────────────────┐
│ 4. In-vivo GLP study │ <─── │ 3. In-vitro assay panel │
│ (rodent + non-rodent) │ │ (DILI, hERG, CYP) │
└───────────────────────────┘ └───────────────────────────┘
│
▼
┌───────────────────────────┐ ┌───────────────────────────┐
│ 5. Bioanalysis & DMPK │ ───> │ 6. Regulatory tox report │
│ (LC-MS/MS, WinNonlin) │ │ (IND-enabling package) │
└───────────────────────────┘ └───────────────────────────┘Stage 1: Test-article receipt and GLP facility setup
The sponsor’s test article is received under chain-of-custody into a GLP facility, the dose formulation is prepared and its concentration, homogeneity and stability verified, and the study director opens the protocol under the GLP quality system.
Stage 2: In-silico ADMET triage
Certara Simcyp PBPK modelling and QSAR predict the candidate’s drug-drug interactions, pharmacokinetics and hepatotoxicity from its structure and in-vitro data, ranking and deprioritising overt toxicants before any cell is dosed.
Stage 3: In-vitro assay panel
Cyprotex/Labcorp runs the hepatocyte drug-induced-liver-injury model, the hERG cardiotoxicity assay and miniaturised 96-well CYP induction, plus high-content imaging and the devTOX quickPredict human-iPSC developmental-toxicity assay, flagging off-target toxicity cheaply.
Stage 4: In-vivo GLP study
Rodent and non-rodent acute and repeat-dose toxicity studies run under OECD GLP at Charles River, Eurofins, WuXi AppTec or Syngene, with toxicokinetic blood sampling and Charles River’s Virtual Control Groups reducing concurrent-control animal counts using historical data.
Stage 5: Bioanalysis and DMPK
Plasma drug and metabolite concentrations are quantified by LC-MS/MS and modelled in Phoenix WinNonlin to yield the toxicokinetic profile and metabolic pathway — the exposure data that anchors the NOAEL.
Stage 6: Regulatory tox report
All tiers are integrated into a signed IND-enabling toxicity package against ICH M3(R2), S7A/S7B and OECD GLP, with target-organ findings, NOAEL and toxicokinetics ready for FDA, EMA or NMPA submission.
| Supplier | Price | Lead time | Certificates | Risk | Confidence |
|---|---|---|---|---|---|
| Charles River Laboratories | per-study / quote | weeks–months | Commercial Public (NYSE: CRL) GLP / OECD | Low | HIGH |
| Certara | per-license / service | weeks | Commercial Public (NASDAQ: CERT) | Low | HIGH |
| Cyprotex (Labcorp) | per-assay / panel | weeks | Commercial GLP / OECD | Low | HIGH |
| Eurofins Scientific | per-study / quote | weeks–months | Commercial Public (EPA: ERF) GLP / OECD | Low | HIGH |
| WuXi AppTec | per-study / quote | weeks–months | Commercial Public (SSE: 603259) GLP / OECD | Low | HIGH |
| Syngene International | per-study / FTE | weeks–months | Commercial Public (BSE/NSE: SYNGENE) | Low | HIGH |
AI note: toxics-screening-toxicomics-service (EN)
Key directions:
- In-vivo GLP toxicology — rodent and non-rodent IND-enabling safety studies (acute, repeat-dose, reproductive, carcinogenicity) under OECD GLP at large CROs (Charles River, Eurofins, WuXi AppTec, Syngene).
- In-vitro ADMET and high-content screening — hepatocyte DILI, hERG cardiotoxicity, CYP induction (now 96-well) and high-content imaging that flag off-target toxicity before animal dosing (Cyprotex/Labcorp).
- In-silico / model-informed toxicology — EMA-qualified PBPK modelling (Certara Simcyp v25, 2026) and QSAR for drug-drug interactions and hepatotoxicity in software, feeding MIDD.
- Toxicogenomics and developmental toxicity — human-iPSC assays (devTOX quickPredict) and metabolomic signatures reading developmental toxicity off a cell panel — the toxicomics core.
Regulatory:
- US: FDA IND-enabling framework follows ICH M3(R2) and S7A/S7B; GLP studies under 21 CFR Part 58; FDA accepts EMA-qualified PBPK for DDI prediction.
- EU: EMA qualifies PBPK platforms (Simcyp) so model-informed DDI/hepatotoxicity carries regulatory weight; GLP under OECD principles.
- CN: NMPA, post-ICH adoption, accepts the same nonclinical safety package, so a Chinese GLP study serves multinational filings.
Companies not in table: Labcorp (parent of Cyprotex and Stemina — the devTOX quickPredict iPSC assay came in via Labcorp’s Stemina acquisition; Cyprotex is the in-vitro ADMET brand within Labcorp Drug Development, kept as the table’s canonical slug); Charles River and Eurofins also run environmental/ecotox and analytical testing far beyond pharma tox (kept scoped here to drug safety); WuXi Biologics (the biologics arm, separate from WuXi AppTec — not tabled here to avoid scope creep); BioReliance (Merck KGaA’s GLP toxicology arm, a notable omission — it competes with Charles River but was outside the 6-row cap).
Processing note: the funnel stacks a cheap in-silico (Simcyp PBPK, QSAR) and in-vitro tier (DILI, hERG, 96-well CYP induction, devTOX iPSC) in front of the expensive GLP animal study, each tier gating the next; a full GLP rat repeat-dose study still costs in the hundreds of thousands of USD and runs months, so the early tiers exist to kill overt toxicants cheaply.
Relevance: this is the safety bottleneck every drug candidate must clear before human dosing — Charles River’s ~15 in-vivo sites, WuXi AppTec’s 2025 revenue +15.8% YoY (continuing +21.4%) and Syngene’s BMS-to-2035 extension all show a sector still compounding. The honest MECE boundary is with sibling CRO-service articles: crispr-screening-as-a-service (functional genomics, not tox), proteomics-mass-spec-cro (proteomics), strain-engineering-synbio-cro (strain engineering), structural-biology-cryo-em-cro (structure) and process-validation-qualification-cro (CMC) — all different assay domains; this article owns the safety/toxicity slice. The “toxicomics” capability tag (cap:bioinformatics) is earned by the iPSC-metabolomic and transcriptomic readouts replacing animal endpoints.