ICH Q3A/BICH Q3A(R2) / Q3B(R2) — Impurities in New Drug Substances / Products

ICH Q3A(R2), Impurities in New Drug Substances, and ICH Q3B(R2), Impurities in New Drug Products, are companion guidelines that apply the same three-threshold logic to two different materials. Q3A covers organic impurities in a new, chemically synthesized drug substance (API): starting materials, by-products, intermediates, reagents, ligands, catalysts, and degradation products that form before or during manufacture and storage of the API itself. Q3B covers degradation products that form in the finished drug product — from the drug substance degrading, or from the drug substance reacting with an excipient or the container-closure system — once it is formulated. Q3B explicitly states it is complementary to Q3A and that impurities already present in the drug substance generally do not need to be separately monitored in the drug product unless they are also degradation products.

Both guidelines were adopted at ICH Step 4 with a later revision to their threshold attachment: Q3A(R2) carries a Step 4 date of 25 October 2006 (the original Step 4 adoption was 7 February 2002, with Attachment 2 revised in 2006), and Q3B(R2) carries a Step 4 date of 2 June 2006 (original Step 4 adoption 5 February 2003, Attachment 2 revised 6 June 2006). Neither guideline applies to drug substances or products used during the clinical research stage of development, and both exclude biological/biotechnological products, peptides, oligonucleotides, radiopharmaceuticals, fermentation products and semi-synthetic derivatives, herbal products, and crude products of animal or plant origin. Both also explicitly exclude polymorphic forms, enantiomeric impurities, and extraneous GMP contaminants from their scope.

The shared mechanism is three dose-indexed thresholds — reporting, identification, and qualification — that scale the required rigor of analytical and safety work to patient exposure. A low-dose product has tighter percentage thresholds than a high-dose product because, at the same percentage impurity level, a higher daily dose delivers more absolute impurity to the patient. Reading Q3A and Q3B as 'the impurities guidelines' obscures this design: they are two applications of one dose-based risk logic, one upstream of formulation and one downstream of it.

Q3A(R2) Attachment 1 sets two dose bands for the drug substance, based on the maximum daily dose of drug substance administered. For a maximum daily dose of 2 g/day or less, the reporting threshold is 0.05%, the identification threshold is 0.10% or 1.0 mg total daily intake (whichever is lower), and the qualification threshold is 0.15% or 1.0 mg total daily intake (whichever is lower). For a maximum daily dose above 2 g/day, all three thresholds drop to a flat percentage: 0.03% reporting, 0.05% identification, and 0.05% qualification.

Footnotes to the table matter as much as the numbers: a higher reporting threshold must be scientifically justified, and lower thresholds can be appropriate for an impurity known or suspected to be unusually toxic. Any impurity above the reporting threshold, plus the sum of all such impurities as 'total impurities,' must be reported with the analytical procedure used. Results below 1.0% are reported to two decimal places and at or above 1.0% to one decimal place, with conventional rounding rules illustrated in the guideline's Attachment 2 worked examples.

The specification then separates impurities into specified (individually listed, each with its own acceptance criterion, which may be identified or unidentified) and a general acceptance criterion for any unspecified impurity set at not more than the identification threshold, plus a total impurities limit. Q3A directs that specified identified impurities and specified unidentified impurities estimated above the identification threshold be listed individually, while an unidentified impurity below that level can remain grouped under the general unspecified cap. Acceptance criteria should be set no higher than levels justified by safety data and consistent with manufacturing and analytical capability — not simply set at the regulatory threshold by default.

Q3B(R2) Attachment 1 indexes its thresholds to the maximum daily dose of the drug substance delivered by the drug product, but it uses more dose bands than Q3A because drug-product doses span a much wider range, from sub-milligram to multi-gram. The reporting threshold has two bands: 0.1% for a maximum daily dose of 1 g or less, and 0.05% above 1 g.

The identification threshold has four bands: below 1 mg, 1.0% or 5 µg total daily intake (TDI), whichever is lower; 1 mg to 10 mg, 0.5% or 20 µg TDI, whichever is lower; above 10 mg to 2 g, 0.2% or 2 mg TDI, whichever is lower; and above 2 g, a flat 0.10%. The qualification threshold also has four bands, using slightly different dose cut-points: below 10 mg, 1.0% or 50 µg TDI, whichever is lower; 10 mg to 100 mg, 0.5% or 200 µg TDI, whichever is lower; above 100 mg to 2 g, 0.2% or 3 mg TDI, whichever is lower; and above 2 g, a flat 0.15%.

As in Q3A, a degradation product above the reporting threshold must be reported numerically (not as 'complies'), with totals summed and rounding conventions applied consistently with the number of decimal places in the applicable threshold. Quantitation limits for the analytical procedure should be no higher than the reporting threshold, and methods must be validated for specificity against samples stressed under light, heat, humidity, acid/base hydrolysis, and oxidation.

Under both guidelines, an impurity or degradation product above the identification threshold should be structurally elucidated; if identification is not feasible, the dossier should include a summary of the laboratory efforts attempted rather than simply omitting the data. Below the identification threshold, identification is generally not required, but Q3B specifically notes that unusually potent, toxic, or pharmacologically active degradants may still need dedicated analytical procedures even below that threshold.

Qualification is the process of acquiring and evaluating data that establishes the biological safety of an impurity or degradation product at the level specified. Both guidelines treat an impurity level already present in batches adequately tested in safety and clinical studies as qualified by that history, and treat an impurity that is also a significant human or animal metabolite as generally qualified without further work. Where no such history exists and the qualification threshold is exceeded, the applicant has three practical paths described in the guidelines' decision trees: reduce the impurity to at or below the threshold (sometimes achieved in Q3B's case through a more protective container closure or storage condition), cite adequate safety data already in the scientific literature, or generate new safety data through additional studies on the drug substance, drug product, or isolated impurity.

Both guidelines allow the proposed qualification threshold itself to be adjusted up or down from the default on a case-by-case basis, with scientific rationale — for example, a lower threshold where a drug class has a known history of adverse reactions linked to a structurally similar impurity, or a higher threshold where the level of safety concern is demonstrably lower. These are not loopholes; Q3A states such proposals 'would be considered on a case-by-case basis,' meaning the rationale has to be built and defended in the dossier, not simply asserted.

Neither Q3A nor Q3B's default qualification threshold is intended to apply to an impurity that is a DNA-reactive (mutagenic) carcinogen. That class of risk is addressed separately by ICH M7(R2), Assessment and Control of DNA Reactive (Mutagenic) Impurities in Pharmaceuticals to Limit Potential Carcinogenic Risk, adopted at Step 4 on 3 April 2023. M7(R2) sets a default Threshold of Toxicological Concern (TTC) acceptable intake of 1.5 µg per person per day for an impurity of unknown mutagenic and carcinogenic potential, derived to be protective across a lifetime of daily exposure; the guideline also provides a less-than-lifetime adjustment formula for shorter treatment durations, calculated as 1.5 µg multiplied by the ratio of a 70-year lifetime (25,550 days) to the actual treatment duration in days.

In practice this means a firm first screens impurities and degradation products (in silico structural alerts plus knowledge of the synthetic route or degradation pathway) to flag any with structural alerts for mutagenicity, then applies the M7 hazard assessment and control framework to those specific impurities instead of the general Q3A/Q3B qualification threshold — because 1.5 µg/day is, for most oral small-molecule doses, a far tighter limit than the percentage-based Q3A/Q3B qualification threshold would produce. Q3A and Q3B remain the controlling framework for every other organic impurity and degradation product in the same specification; M7 is additive scrutiny for the mutagenic subset, not a replacement for the whole impurities program.

Q3A classifies impurities into three categories: organic impurities, inorganic impurities, and residual solvents. Only the organic-impurity category is governed by the Q3A reporting/identification/qualification thresholds described above. Residual solvents are directed to ICH Q3C, which classifies solvents by toxicity class and sets permitted daily exposure limits; Q3A notes that because solvents are generally of known toxicity, appropriate controls are 'easily accomplished' by reference to Q3C rather than by the Q3A threshold logic. Inorganic impurities — heavy or residual metals, inorganic salts, filter aids, and reagents/ligands/catalysts of inorganic origin — are likewise outside the Q3A percentage thresholds and are addressed by ICH Q3D's permitted daily exposure (PDE) limits for elemental impurities, which require a route-specific risk assessment across likely elemental sources.

A specification built on Q3A/Q3B therefore sits alongside, not instead of, separate residual-solvent limits under Q3C and elemental-impurity controls under Q3D, and alongside mutagenic-impurity controls under M7 for the relevant subset. Treating a residual solvent peak or an ICP-MS metal result as if it were a Q3A 'organic impurity' subject to the reporting/identification/qualification percentages is a scope error the guideline itself anticipates by defining the three impurity categories up front.

Both guidelines expect the specification for impurities (Q3A) or degradation products (Q3B) to be built from actual data: stability studies, process-development data, and batch analyses from material representative of the proposed commercial process, not from theoretical worst cases alone. Q3A instructs that the rationale for including or excluding a given impurity from the specification should discuss the impurity profile seen in safety and clinical batches alongside the profile from commercial-process batches, and references ICH Q6A's Decision Tree #1 for setting a specified-impurity acceptance criterion. Q3B's Attachment 3 provides an analogous decision tree specifically for identification and qualification of a degradation product.

A resulting specification typically lists each specified (identified or unidentified) impurity with its own acceptance criterion, a general limit for any unspecified impurity (not more than the identification threshold), and a total-impurities or total-degradation-products limit. Both guidelines caution that acceptance criteria should be set no higher than levels justified by safety data and should be consistent with what the manufacturing process and analytical method can actually and reliably deliver — significant lot-to-lot variation in impurity levels is itself treated as a signal that the manufacturing process may not be adequately controlled or validated, not simply a specification-setting inconvenience to be widened away.

Both guidelines also address later-development changes: an impurity or degradation product newly observed in batches made by the proposed commercial process, even late in development, should still be evaluated against the Attachment 1 thresholds and identified or qualified if it exceeds them, with safety comparisons made against previously qualified material.

Q3B's scope is specifically degradation products — material formed from the drug substance degrading, or from the drug substance reacting with an excipient or the container-closure system — observed during manufacture or on stability. The guideline requires the analytical method to be validated for specificity against the drug substance, synthesis-related impurities, excipients, and excipient-derived impurities, using samples stressed under light, heat, humidity, acid/base hydrolysis, and oxidation, consistent with ICH Q1A stability guidance and ICH Q2 method-validation expectations. Chromatograms from validation studies and from long-term and accelerated stability studies, with peaks labeled, are expected in the registration dossier.

Because degradation products can continue to form across shelf life, the reporting and identification obligations apply at the storage condition and time point where they are actually observed — a degradation product that only appears at an accelerated stability time point, or only near the end of proposed shelf life, still has to be evaluated against the Attachment 1 thresholds at the point it crosses them, not only at release.

Operationally, Q3A/Q3B compliance depends on records that connect four things for any given lot: the raw chromatographic or spectroscopic result, the calculation that converts it into a reported percentage or total daily intake figure, the threshold band that applies (keyed to the correct maximum daily dose for that specific product), and the disposition decision against the specification. Because the identification and qualification thresholds for drug products are expressed in multiple dose bands with a 'whichever is lower' percent-or-TDI comparison, the underlying calculation — and which dose was used — needs to be visible in the record, not just the final pass/fail call, so a reviewer or inspector can re-derive the comparison the guideline's own Attachment 2 worked examples illustrate.

A laboratory information management system or electronic batch record that stores the specification version, the applicable threshold band, the test method and its validated quantitation limit, and the result together — with the reasoning for any specified-impurity classification or unspecified-impurity grouping — makes that traceability auditable. Where V5 is used for this kind of record, the practical role it plays is keeping the specification, the method, the result, and the reviewer's disposition linked in one place with an audit trail, so a question about a specific peak in a specific lot can be traced back to the specification basis and the human quality decision that released it; it does not itself set thresholds, classify impurities, or substitute for the structural identification and toxicological qualification work the guidelines require.

Because Q3A and Q3B both expect rounding conventions to be applied consistently (two decimal places below 1.0%, one decimal place at or above 1.0%) and expect totals to be summed from individually reported values, a records system that recalculates totals from stored raw results, rather than relying on a separately transcribed total, reduces the chance that a transcription error produces a total that disagrees with the specification's total-impurities limit.