Justification of Impurity Limits: How to Build Specifications That Are “Defensible” (Not Just Compliant)
In a GMP audit, a specification is not assessed only because it exists. It is assessed based on the question behind it:
“Why is this the right limit for this product?”
The GuideGxP guide highlights a point that many underestimate: simply “staying within the numbers” is formal compliance, but that is not enough. The inspection expectation is substantive compliance: demonstrating that the limits are scientific, up to date, and consistent with risk and with your actual process.
This article gives you a practical, “QA/QP-style” method to:
- define limits that can withstand FDA/EMA scrutiny,
- document the rationale,
- prevent findings related to specifications that are “copied”, incomplete, or not aligned with the dossier.
The Number 1 Problem: “Copied” Limits (Pharmacopoeia/ICH) Without Rationale
The guide gives concrete examples of errors that become inspection findings:
- using the generic limit “≤0.10% unspecified individual impurity” just because it is provided,
- without checking whether your process generates recurring impurities at higher levels,
- and without managing trends close to the limit.
For the inspector, this means: “your specification does not represent the real product” → quality risk.
What Is a “Defensible” Specification? (QA/QP Mindset)
A specification is defensible when you can show a robust logical thread that integrates:
- ICH thresholds (regulatory minimums and dose-based logic)
- Real product data (development, validation, routine, stability)
- Toxicology (when qualification thresholds are exceeded or for critical impurities)
- Compendial constraints (mandatory limits and methods for specific markets)
- QRM (for omissions, alternative choices, testing frequencies, risk-based approach)
7-Step Method: How to Build the “Justification of Specifications” for Impurities
Step 1 — Build the impurity map (before the numbers)
Base document: “Impurity Profile / Impurity Risk Map”
- origin (process, degradation, known impurities),
- probability of formation,
- severity (patient impact),
- detectability (available method / LOQ),
- impacted markets.
If you do not have this map, every limit is just a “floating” number.
Step 2 — Align by impurity type (Q3A/B, Q3C, Q3D, M7)
There is no single “impurity chapter”:
- organic impurities ≠ residual solvents ≠ elemental impurities ≠ mutagenic impurities.
Step 3 — Use historical data to demonstrate control (not only compliance)
This is where the difference arises between:
- specification limit (release/stability, registered),
- internal alert/action limit (statistical control and OOS prevention).
The guide highlights a key concept: being “within specification” but statistically out of control is no longer acceptable as a modern quality management approach.
Step 4 — Link toxicology when needed (and document it)
When:
- qualification thresholds are exceeded,
- potentially genotoxic impurities are present,
- problematic stability trends exist,
you must have a toxicological rationale (or M7 TTC where appropriate).
Step 5 — Justify absences as well (this is the part that saves you in an audit)
The guide explicitly recalls the principle:
“The absence of a specification must be justified.”
Typical examples:
- no residual solvent test → “solvents not used + signed risk assessment”
- no test for a theoretical impurity → “not formed >X% even under forced degradation conditions, report attached”
- no elemental impurity test → “Q3D risk assessment + control over materials/suppliers/packaging + possible periodic testing”
Step 6 — Align with the dossier (the killer finding)
One of the most dangerous findings is a mismatch between:
- registered specification (dossier),
- operational specification (QC),
- and what the QP releases.
Step 7 — Include periodic review (PQR) as a “defense mechanism”
The guide clearly links:
- impurity trends,
- PQR,
- change control,
- CAPA in a lifecycle perspective.
If you cannot show that you periodically review limits and trends, the inspector concludes: “the system does not learn.”
Practical Template: Minimum Structure of the “Justification of Specifications” (Impurities)
You can use this outline as an internal document (or as a QOS section):
- Scope (product, markets, pharmacopoeia versions)
- Impurity profile (table: impurity, origin, maximum observed, trend)
- Applicable ICH thresholds (dose, thresholds, qualification)
- Compendial requirements (Ph. Eur./USP: limits and methods)
- Rationale for each limit (data + justification)
- Rationale for excluded tests (QRM)
- Trending & PQR commitments (frequencies, OOT triggers)
- QA/QP conclusions (approvals, references to reports)
What Really Gets Challenged in Audits (Red Flags)
From the guide, the typical targets are:
- incomplete specifications (e.g. elemental impurities without Q3D risk assessment),
- arbitrary limits (too wide compared to capability),
- lack of toxicological justification where required,
- dossier vs QC mismatch,
- ignored trends until OOS.
Micro-Checklist (QA/QP) Before an Audit
- Do I have a document explaining where every number comes from?
- Do I have a document explaining why some tests are not included?
- Can I show trends and internal action limits?
- Are operational specifications = registered specifications?
- If the inspector asks “why 2.0%?”, do I have a documented answer?
FAQ
1) Can I use pharmacopoeial limits as the “justification”?
No: they are a starting point, not a complete justification. You need to connect them to data and risk.
2) What should I do if the process constantly generates impurities close to the limit?
Manage the trend: internal action limit, OOT investigation, CAPA and/or limit revision with rationale.
3) Is it acceptable not to include an impurity test in the specification?
Yes, but only with documented justification (QRM + data) approved by QA/QP.
4) What is the worst mistake that can “break” an audit?
Different specifications between dossier and laboratory (or QP release).
5) Is toxicology needed for every impurity?
No. But it is needed when qualification thresholds are exceeded or for critical impurities (M7/nitrosamines).
6) Does the PQR really matter for impurities?
Yes: it is the “natural” place where the inspector expects to see trends and decisions.
If you want ready-to-use templates, checklists, and a step-by-step guide to build inspector-proof specifications and rationales for impurities (ICH vs compendia), you can find everything in the full guide on guidegxp.com: “Guide to Pharmacopoeias: Impurities and Critical Tests (Multi-Region, ICH vs Compendia)”.
