What does the ICH Q3E draft require for extractables and leachables, and when will it apply?

1Where ICH Q3E stands, and when it will apply.

ICH Q3E is ICH’s guideline on extractables and leachables (E&L). Leachables migrate into a product from manufacturing equipment, packaging or a delivery device under normal manufacture and storage; extractables are what a laboratory deliberately pulls from those components, the potential leachables.¹ It is a draft: endorsed at Step 2 on 1 August 2025, now at Step 3 while the working group addresses comments, and implemented by no regulator.¹²³

Step 4 is ICH’s adoption of the final text; each regulator then implements it on its own timetable, and FDA will reformat its version to its own guidance style.⁶ ICH’s work plan of 21 July 2026 puts Step 4 in July 2027 and training materials in November 2027. The June 2026 Assembly in Rio de Janeiro reported “good progress”, not adoption; the next meets in Prague on 14 to 18 November 2026.³⁷ So no region can apply a final Q3E before mid-2027.

The numbers may still move: the review of the permitted daily exposures behind the thresholds, “to reach alignment on final QTs for Step 4”, was planned for February to April 2026 and now runs to November 2026.⁸³ Treat every threshold below as a draft value.

2What applies until Q3E is final.

2.1How Q3E sits beside the USP chapters.

US programmes also use two pairs of USP chapters; FDA’s 2026 draft cites <1663> and <1664> beside Q3E.¹²

ELSIE told EMA that the draft’s use of surface-to-volume ratio “stands in contrast to USP 665”; EFPIA that qualifying a manufacturing component on food-contact compliance alone “is not aligned with USP <665>”.²³ Under the draft, a supplier’s <665> report is prior knowledge (4.1), and its peaks are judged against your product’s own AET (4.3).¹

2.2Should you work to the draft now.

For a product filing in 2026 or 2027, build the programme the draft describes and hold its numbers loosely.

Reviewers already ask for its core steps. FDA’s complete response letter of 30 November 2018 on Xaracoll, a bupivacaine collagen implant, said:²⁴

You have not provided an adequate extractables/leachables evaluation to support the safety of your proposed container closure system.

The letter added that many compounds above the safety concern threshold had not been identified. FDA approved the product in August 2020.²⁴²⁵

Investigators look too. FDA’s compliance program for drug manufacturing inspections has them check that “Containers and closures are not additive, reactive, or absorptive to the drug product”.²⁶ After a deviation report at Vitae Enim Vitae Scientific described ink from transfer tubing leaching into a sterile solution, FDA’s warning letter of 17 March 2022 cited the investigation (21 CFR 211.192):²⁷

Your response is inadequate because you failed to evaluate the potential of ink leaching into previously manufactured batches of drug products.

It asked for every batch made with the tubing and “The chemical name, composition, and toxicity of the contaminant that leached into the product”.²⁷ Before a pre-approval inspection, check that you can show which component lots the E&L studies and stability batches used (4.4), and the supplier change clause (3.4).¹

The draft is not in force, and FDA guidance does not bind even when final: FDA says the final version “does not establish any rights for any person and is not binding on FDA or the public”.⁶ The numbers are contested too: EFPIA, Medicines for Europe and PDA asked ICH to define when the 0.5 uncertainty factor applies, and EFPIA asked for an implementation timeline that would not reach products already on the market unless they change or new information arises.²³

Record which threshold version each AET used, so it can be recalculated, and justify the filing against current rules, citing the draft as support.

3What the draft asks, and what to check.

4From safety threshold to AET.

The safety concern threshold (SCT) is the lower of two values: the threshold of toxicological concern (TTC) for mutagenicity, from ICH M7, and a qualification threshold (QT) for other toxicity, derived from about 330 permitted daily exposures (PDEs). The analytical evaluation threshold (AET) turns the SCT into a concentration for one study. It is “not a control threshold” but the level above which a peak should be identified, quantified and reported for safety assessment.¹

• Local thresholds. Topical ophthalmic 20 ppm, subcutaneous and intradermal 50 ppm, dermal and transdermal 500 ppm, inhalation 5 µg/day; intracerebral, intrathecal, epidural and intraocular products need a compound-specific evaluation. Where local toxicity matters, the SCT is the lowest of the TTC, the QT and the local threshold converted to a daily dose.¹
• Class 1, avoid. The ICH M7 cohort of concern, M7 Class 1 mutagens with an acceptable intake below 1.5 µg/day, and non-mutagens the QT would not protect against, each quantified against its own limit, not the SCT. Table A.4.1 gives chronic parenteral PDEs of 0.26 µg/day for benzo(a)pyrene (carbon black) and 4 µg/day for bisphenol A (polycarbonate, epoxy resin).¹
• Class 2, the default. Limited to the TTC and the QT.¹
• Class 3, low potency. Qualified up to 1.0 mg/day, or a compound-specific PDE, without further justification. The draft list includes BHT, erucamide and fatty acids such as stearic and palmitic acid.¹
• Uncertainty factor. Semi-quantitative screens multiply by a factor of no more than 0.5, or one derived from a response-factor database, justified in the report.¹
• Dosing that is not daily. With 30 dosing days or fewer, the TTC is 120 µg, following ICH M7, and the one-month QT can be used, as it can for dosing once a month or less often. A patch or depot is assumed to release every non-mutagenic leachable in one day unless data show otherwise.¹
• Cancer medicines. If the safety assessment follows ICH S9, the TTC does not apply and the SCT is the QT (6.5).¹

5Example: an AET for a prefilled syringe and its filter.

Example, invented for illustration. Product A is a 10 mg/mL subcutaneous injection: maximum approved dose 10 mg once daily, so 1.0 mL from one 1 mL prefilled syringe, labelled for long-term use. The smallest batch is 5 L, filtered through one sterilising filter with a polycarbonate housing. The formulas are the draft’s extractable scenario 1 and leachable scenario 2.¹

Duration drives the workload. Labelled for one month or less, the SCT for Product A would be 26 µg/day, the QT being lower than the TTC of 120, and the syringe AET 13 µg/mL, about 17 times higher.¹

6Checklist: what to ask single-use and packaging suppliers.

Supplier evidence feeds supplier qualification, and change notice belongs in the quality agreement.

7How it runs in Seal.

In Seal, the extractables and leachables blueprint keeps one evidence record from material and component through worst-case exposure, study sample, analytical signal, compound identity, AET and toxicological assessment to a qualification decision for a stated use. Each threshold keeps its formula, inputs, units, version, rounding and approval, so when a dose or presentation changes, recalculation shows which findings and conclusions are affected. Unknowns remain findings in their own right, with their estimated concentration and open work.

For single-use assemblies, supplier, extractables and sterilisation evidence attaches to the component, and a resin, film, formulation, site or sterilisation change resolves to the affected assembly positions, planned batches and studies before it reaches a batch. Supplier quality checks each receipt against the recorded qualification scope; Neil investigates a supplier change and prepares the qualification work for your team to review.