How CPCA categorisation sets a nitrosamine acceptable intake limit
How the Carcinogenic Potency Categorisation Approach assigns a nitrosamine to one of five categories, and what acceptable intake limit follows under EMA and FDA.
What the approach is for
Most N-nitrosamines found as impurities have never been tested for carcinogenicity, and testing each one is not practical. CPCA closes that gap: it predicts potency from structure, so a limit can be set for a compound nobody has ever dosed to an animal.
The prediction rests on a single mechanism. An N-nitrosamine becomes genotoxic after α-hydroxylation — oxidation at the carbon adjacent to the N-nitroso group — which ultimately yields a species that alkylates DNA. Structural features that make that pathway easier raise predicted potency; features that hinder it, or that favour clearance by some other route, lower it.
How a category is reached
Each α-carbon is examined for hydrogens available to that oxidation, and the molecule is scored on features that activate or deactivate the pathway. The scores combine into a potency score, and the score maps to a category.
The mapping is fixed, not a matter of judgement: a lower score means higher predicted potency and therefore a lower permitted intake. Where a molecule carries two N-nitroso groups, the more potent group — the lower category — sets the limit for the whole molecule.
The categories and their limits
Categories 4 and 5 both sit at 1500 ng/day, the ICH M7 threshold of toxicological concern, but for different reasons: Category 4 is predicted to be weakly activated, Category 5 not activated by this pathway at all.
| Category | Potency score | AI — EMA | AI — FDA | Structural basis |
|---|---|---|---|---|
| Category 1 | Potency score ≤ 1 | 18 ng/day | 26.5 ng/day | 18 ng/day is the class-specific TTC for N-nitrosamine impurities. FDA instead applies 26.5 ng/day, derived from N-nitrosodiethylamine (NDEA), the most potent robustly tested nitrosamine. |
| Category 2 | Potency score = 2 | 100 ng/day | 100 ng/day | Representative of two potent, robustly tested nitrosamines — NDMA (96 ng/day) and NNK (100 ng/day). Category 2 substances are predicted to be no more potent than these. |
| Category 3 | Potency score = 3 | 400 ng/day | 400 ng/day | Set to reflect a four-fold decrease in carcinogenic potency from Category 2, typically due to a weakly deactivating structural feature. |
| Category 4 | Potency score ≥ 4 | 1500 ng/day | 1500 ng/day | May still be activated by α-hydroxylation, but predicted to be of low potency because the pathway is disfavoured or clearance is favoured. Set at the ICH M7 TTC. |
| Category 5 | Structurally excluded | 1500 ng/day | 1500 ng/day | Not predicted to be activated via α-hydroxylation — no α-hydrogens, steric hindrance, or formation of unstable species that will not react with DNA. Set at the ICH M7 TTC. |
Where EMA and FDA diverge
At Categories 2 through 5 the two agencies agree. At Category 1 they do not, and the difference matters to anyone filing in both jurisdictions.
EMA applies 18 ng/day, the class-specific threshold of toxicological concern for N-nitrosamines. FDA applies 26.5 ng/day, derived from N-nitrosodiethylamine — the most potent nitrosamine for which robust carcinogenicity data exists. Same structure, same score, same category, two limits.
Where the approach stops
CPCA is bounded, and reaching for it outside those bounds produces a number that looks authoritative and is not applicable.
- Applies to N-nitrosamines bearing a carbon atom on both sides of the N-nitroso group.
- Excluded where an α-carbon is directly double-bonded to a heteroatom — N-nitrosamides, N-nitrosoureas and N-nitrosoguanidines are outside the approach.
- Does not apply where the N-nitroso group is attached to a nitrogen within a hetero-aromatic ring.
- Where a molecule carries two N-nitroso groups, the group with the highest predicted potency — the lowest numerical category — sets the AI for the whole molecule.
Common questions
Does CPCA require carcinogenicity data for the specific nitrosamine?
No. The approach predicts potency from structure, which is the reason it exists — most nitrosamine impurities have never been tested, and the limit still has to be set.
Why do EMA and FDA give different limits for the same compound?
Only at Category 1, and because they derive the figure differently. EMA applies 18 ng/day, the class-specific threshold of toxicological concern for N-nitrosamines. FDA applies 26.5 ng/day, derived from N-nitrosodiethylamine, the most potent robustly tested nitrosamine. The score, the features and the category are identical either way.
Can CPCA be applied to any nitrosamine impurity?
No. It applies to N-nitrosamines carrying a carbon atom on both sides of the N-nitroso group. N-nitrosamides, N-nitrosoureas and N-nitrosoguanidines fall outside it, as does any structure where the N-nitroso group sits on a nitrogen within a hetero-aromatic ring.
What happens when a molecule has two N-nitroso groups?
The group with the highest predicted potency — the lowest numerical category — sets the acceptable intake for the whole molecule.
Is a CPCA category enough on its own for a submission?
It establishes the acceptable intake. A submission also needs the risk assessment that identifies where the nitrosamine could form or carry over, the analytical work confirming the level present, and the control strategy holding it below that intake.