Acceptable Daily Intake for Pharmaceutical Impurities: How to Calculate
Acceptable daily intake (ADI) for pharmaceutical impurities is calculated using different methodologies depending on the impurity type. For elemental impurities (ICH Q3D), Permitted Daily Exposure (PDE) is derived from the NOAEL using uncertainty factors. For mutagenic impurities (ICH M7), the Threshold of Toxicological Concern (TTC) of 1.5 mcg/day applies, or compound-specific limits are derived from TD50 carcinogenicity data. For nitrosamines, the Carcinogenic Potency Categorization Approach (CPCA) provides a tiered framework when compound-specific data is unavailable. Less-than-lifetime adjustments permit higher limits for shorter treatment durations.
Key Takeaways
Key Takeaways
- Three primary methodologies apply: PDE (ICH Q3D for elemental), TTC (ICH M7 for mutagenic at 1.5 mcg/day lifetime), and CPCA (tiered framework for nitrosamines).
- Less-than-lifetime adjustments under ICH M7 permit higher acceptable intake limits for treatment durations shorter than a patient's lifetime.
- Incorrect AI calculations are a common root cause of either unnecessarily tight specifications (manufacturing problems) or dangerously loose specifications (patient safety risk).
- The CPCA approach for nitrosamines provides five potency categories when compound-specific TD50 data is unavailable.
- Acceptable daily intake for pharmaceutical impurities is the maximum amount of an impurity that a patient can be exposed to daily without appreciable risk of adverse health effects. The term encompasses several related concepts — Permitted Daily Exposure (PDE), Threshold of Toxicological Concern (TTC), and compound-specific Acceptable Intake (AI) — each calculated differently depending on the type of impurity, available toxicological data, and applicable ICH guideline.
- For CMC teams, calculating acceptable intake limits correctly is a core regulatory competency. Errors in these calculations directly affect specifications, analytical method requirements, and the risk of regulatory deficiency notices. Incorrect AI calculations are also a common root cause of unnecessarily tight specifications that create manufacturing challenges, or dangerously loose specifications that fail to protect patients.
- In this guide, you'll learn:
- The three primary methodologies for calculating acceptable intake limits
- PDE calculation for elemental impurities per ICH Q3D
- TTC application for mutagenic impurities per ICH M7
- Compound-specific AI derivation from carcinogenicity data
- The CPCA approach for nitrosamines
- Less-than-lifetime exposure adjustments
- How to convert AI values to specification limits
- ---
Three Methodologies for Acceptable Intake Calculation
Different ICH guidelines use fundamentally different approaches to establish safe limits, reflecting the distinct toxicological mechanisms involved.
Overview of Approaches
| Approach | Applicable Guideline | Impurity Types | Basis | Key Assumption |
|---|---|---|---|---|
| PDE (Permitted Daily Exposure) | ICH Q3D | Elemental impurities | NOAEL with uncertainty factors | Threshold mechanism; below NOAEL = safe |
| TTC (Threshold of Toxicological Concern) | ICH M7 | Mutagenic impurities (Class 2/3) | Statistical distribution of carcinogenic potency | Linear dose-response; any dose carries some risk |
| Compound-Specific AI | ICH M7, FDA Nitrosamine Guidance | Class 1 mutagenic impurities, specific nitrosamines | TD50 carcinogenicity data | Linear extrapolation to 1-in-100,000 risk |
| CPCA | FDA Guidance (2023) | NDSRIs without compound-specific data | Structure-based potency categorization | Structural analogs predict carcinogenic potency |
PDE Calculation: Elemental Impurities (ICH Q3D)
Methodology
PDE for elemental impurities is calculated from the No-Observed-Adverse-Effect Level (NOAEL) in the most relevant animal study, adjusted by uncertainty factors (also called safety factors or modifying factors).
PDE formula:
Where:
| Factor | Description | Range | Default |
|---|---|---|---|
| NOAEL | No-Observed-Adverse-Effect Level from most relevant study (mg/kg/day) | Study-specific | Study-specific |
| Weight Adjustment | Assumed human body weight | 50 kg (ICH standard) | 50 kg |
| F1 | Interspecies extrapolation factor | See table below | Species-dependent |
| F2 | Interindividual variability | 10 | 10 |
| F3 | Short-term to long-term study extrapolation | 1-10 | Study duration-dependent |
| F4 | Severe toxicity (e.g., teratogenicity, genotoxicity without carcinogenicity) | 1-10 | 1 unless severe endpoint |
| F5 | NOAEL not established (LOAEL used instead) | 1-10 | 1 if NOAEL available; up to 10 for LOAEL |
Interspecies Extrapolation Factors (F1)
| Species | F1 Factor | Basis |
|---|---|---|
| Rat | 5 | Body surface area scaling |
| Mouse | 12 | Body surface area scaling |
| Dog | 2 | Body surface area scaling |
| Rabbit | 2.5 | Body surface area scaling |
| Monkey | 3 | Body surface area scaling |
| Human | 1 | No extrapolation needed |
Worked PDE Calculation Example
Element: Nickel (Ni)
Study: 2-year oral rat study
NOAEL: 5 mg Ni/kg/day
Endpoint: Decreased body weight gain
The published ICH Q3D PDE values reflect the Agency's selection of the most sensitive endpoint, species, and study, plus application of appropriate uncertainty factors. CMC teams should use the published PDE values directly from the ICH Q3D appendix rather than recalculating, unless addressing an element not covered by Q3D.
Route-Specific PDEs
ICH Q3D establishes separate PDEs for different routes of administration because bioavailability varies:
| Element | Oral PDE (mcg/day) | Parenteral PDE (mcg/day) | Inhalation PDE (mcg/day) |
|---|---|---|---|
| As | 15 | 15 | 2 |
| Cd | 5 | 2 | 3 |
| Pb | 5 | 5 | 5 |
| Hg | 30 | 3 | 1 |
| Co | 50 | 5 | 3 |
| Ni | 200 | 20 | 5 |
| V | 100 | 10 | 1 |
| Pd | 100 | 10 | 1 |
| Pt | 100 | 10 | 1 |
For multi-route products (e.g., a product available as both oral and injectable), use the most restrictive PDE across routes for the drug substance specification. This ensures that the drug substance meets requirements regardless of which dosage form it is incorporated into.
TTC Calculation: Mutagenic Impurities (ICH M7)
Methodology
The TTC approach is fundamentally different from PDE calculation. Instead of deriving a safe dose from a single compound's NOAEL, TTC is derived from a statistical analysis of a large database of carcinogenic potency data for hundreds of compounds.
TTC derivation basis:
- Compile TD50 values from the Carcinogenicity Potency Database (CPDB)
- Calculate the theoretical daily dose corresponding to a 1-in-100,000 lifetime cancer risk for each compound
- Determine the distribution of these values
- Set TTC at a level protective against the vast majority of compounds
The result: 1.5 mcg/day for lifetime exposure (> 10 years continuous use)
Applying TTC to Specification Setting
Conversion to concentration limit:
Example calculations:
| Drug Product | MDD | Treatment Duration | Applicable TTC | Specification Limit |
|---|---|---|---|---|
| Statin (chronic) | 40 mg (0.04 g) | Lifetime | 1.5 mcg/day | 37.5 ppm |
| Antibiotic (14-day course) | 500 mg (0.5 g) | ≤ 1 month | 120 mcg/day | 240 ppm |
| Chemotherapy (6-month regimen) | 200 mg (0.2 g) | > 1 to ≤ 12 months | 20 mcg/day | 100 ppm |
| Chronic pain (2-year use) | 100 mg (0.1 g) | > 1 to ≤ 10 years | 10 mcg/day | 100 ppm |
Staged TTC Table
| Treatment Duration | Daily Limit | Total Cumulative Limit | Safety Factor |
|---|---|---|---|
| ≤ 1 month (30 days) | 120 mcg/day | 3,600 mcg | ~80x lifetime TTC |
| > 1 to ≤ 12 months | 20 mcg/day | 7,200 mcg | ~13x lifetime TTC |
| > 1 to ≤ 10 years | 10 mcg/day | 36,500 mcg | ~7x lifetime TTC |
| > 10 years to lifetime | 1.5 mcg/day | ~38,325 mcg | 1x (baseline) |
The staged TTC maintains a constant cumulative lifetime dose cap: as treatment duration shortens, the permitted daily dose increases proportionally.
Compound-Specific AI: Known Carcinogens (ICH M7 Class 1)
Methodology
For Class 1 impurities (known mutagenic carcinogens with available carcinogenicity data), compound-specific AI limits are calculated directly from TD50 data.
AI calculation formula:
Where:
- TD50 = dose producing tumors in 50% of animals over lifetime
- 50 kg = assumed human body weight
- 50,000 = safety factor for 1-in-100,000 lifetime cancer risk
TD50 Data Sources
| Source | Database | Coverage |
|---|---|---|
| CPDB | Carcinogenicity Potency Database (UC Berkeley) | ~1,500 chemicals; gold standard |
| NTP | National Toxicology Program | Long-term carcinogenicity studies |
| IARC Monographs | International Agency for Research on Cancer | Cancer classification data |
| Published literature | PubMed, regulatory reviews | Study-specific data |
Selection of TD50 Values
When multiple TD50 values exist for a compound:
| Scenario | Rule |
|---|---|
| Multiple species (rat, mouse) | Use the most potent (lowest TD50) |
| Multiple sexes | Use the most potent |
| Multiple tumor sites | Use the most potent single-site TD50 |
| Different study designs | Prioritize 2-year bioassay data |
| Different routes | Match to intended human route of exposure |
Worked example:
Compound: NDEA (N-Nitrosodiethylamine)
TD50 (rat, oral, most sensitive site): 0.0265 mg/kg/day
This matches FDA's published AI for NDEA of 26.5 ng/day.
The CPCA Approach: Nitrosamine Drug Substance-Related Impurities
Background
The Carcinogenic Potency Categorization Approach (CPCA) was developed by FDA to address nitrosamine drug substance-related impurities (NDSRIs) — nitrosamines formed from the drug substance itself — where compound-specific carcinogenicity data is typically unavailable.
CPCA Categories
CPCA assigns NDSRIs to potency categories based on structural features known to influence carcinogenic potency:
| Category | AI Limit (ng/day) | Structural Basis | Examples |
|---|---|---|---|
| Category 1 (Highest Potency) | 18 | Alpha-hydrogen on both carbons adjacent to the nitroso group; structures resembling the most potent known nitrosamines | Dialkyl nitrosamines with alpha-H on both sides |
| Category 2 | 100 | Alpha-hydrogen on one carbon adjacent to the nitroso group | Monosubstituted nitrosamines with one alpha-H |
| Category 3 | 400 | No alpha-hydrogen adjacent to the nitroso group; bulky substituents that reduce reactivity | Tertiary structure nitrosamines; sterically hindered |
| Category 4 | 1500 | Structural features associated with low carcinogenic potency or lack of metabolic activation | Nitrosamines with deactivating features |
| Category 5 (Lowest Potency) | 1500 | Sufficient evidence from compound-specific data indicating low potency | Supported by read-across or actual data |
CPCA Rationale
The CPCA is based on the understanding that nitrosamine carcinogenic potency is driven by metabolic activation:
- Cytochrome P450 enzymes (primarily CYP2E1) hydroxylate the carbon alpha to the nitroso group
- The resulting alpha-hydroxy nitrosamine spontaneously decomposes to form a diazonium ion
- The diazonium ion alkylates DNA, forming mutagenic adducts
Therefore:
- More alpha-hydrogens = more metabolic activation = higher potency = lower AI limit
- Fewer alpha-hydrogens or steric hindrance = less activation = lower potency = higher AI limit
Applying CPCA
CPCA categorization is not a simple checklist exercise. FDA expects a documented scientific rationale for the category assignment, including structural analysis, comparison to tested analogs, and consideration of metabolic activation potential. A bare statement of "Category 3 based on no alpha-hydrogen" without supporting analysis will face regulatory scrutiny. Include structural diagrams, analog comparisons, and metabolic pathway analysis in your justification.
Less-Than-Lifetime Exposure Adjustments
When Adjustments Apply
Less-than-lifetime (LTL) adjustments are permitted for both TTC-based limits (ICH M7) and compound-specific AI limits when the drug product is not intended for continuous lifetime use.
Eligibility Criteria
| Criterion | Requirement |
|---|---|
| Labeled indication | Must specify a finite treatment duration |
| Treatment duration | Based on the labeled indication, not actual patient use patterns |
| Chronic use products | No LTL adjustment permitted (antihypertensives, statins, etc.) |
| PRN (as needed) products | Generally assessed as chronic use unless label limits total exposure |
| Intermittent use | May qualify if total cumulative exposure can be estimated |
Adjustment Factors
For ICH M7 (mutagenic impurities):
| Duration | Adjustment Factor (vs. lifetime TTC) | Daily Limit |
|---|---|---|
| ≤ 1 month | 80x | 120 mcg/day |
| > 1-12 months | 13.3x | 20 mcg/day |
| > 1-10 years | 6.7x | 10 mcg/day |
| > 10 years | 1x | 1.5 mcg/day |
For compound-specific AI (nitrosamines):
The same staged approach applies, with adjustment factors applied to the compound-specific AI rather than to the generic TTC:
| Duration | NDMA AI (ng/day) | NDEA AI (ng/day) |
|---|---|---|
| ≤ 1 month | 960 | 265 |
| > 1-12 months | 480 | 132.5 |
| > 1-10 years | 192 | 53 |
| > 10 years (lifetime) | 96 | 26.5 |
Documentation Requirements for LTL Adjustments
When claiming LTL adjustments, document:
- The labeled indication and its expected treatment duration
- Whether the drug is used as monotherapy or in combination (relevant for total nitrosamine exposure)
- Whether repeated courses of treatment are anticipated (e.g., antibiotics used 2-3 times per year)
- Any post-market real-world usage data suggesting longer-than-labeled use patterns
Converting AI to Specification Limits
General Formula
For mcg units:
For ng units:
Setting Specifications at Drug Substance vs. Drug Product
| Level | Calculation | When to Apply |
|---|---|---|
| Drug product | AI / MDD of drug product | When impurity forms in the drug product |
| Drug substance | AI / MDD of drug substance | When impurity originates in drug substance synthesis |
| Both | Split AI proportionally or apply full AI at each level | When impurity could originate from either source |
When the same impurity could be present in both drug substance and drug product (e.g., a nitrosamine that forms during API synthesis and also forms during drug product storage), the AI must be allocated between the two specifications. A common approach is a 30/70 split (30% at drug substance, 70% at drug product) or applying the full AI at each level with the understanding that total exposure from both sources must not exceed the AI. Document your allocation rationale.
Key Takeaways
References
Key Takeaways
- 1. Different impurity types use different calculation methods: PDE (elemental, ICH Q3D) uses NOAEL with uncertainty factors. TTC (mutagenic, ICH M7) uses a statistical population-level approach. Compound-specific AI uses TD50-based linear extrapolation.
- 2. PDE assumes a threshold mechanism: Below the NOAEL, no adverse effect is expected. This is appropriate for non-genotoxic endpoints (organ toxicity, developmental effects).
- 3. TTC and compound-specific AI assume a linear dose-response: Any dose of a mutagenic carcinogen carries some risk. The acceptable level is defined as a 1-in-100,000 lifetime cancer risk.
- 4. CPCA provides a structured framework for NDSRIs: When compound-specific carcinogenicity data is unavailable, structural features predict carcinogenic potency. Categories range from 18 ng/day (highest potency) to 1500 ng/day (lowest potency).
- 5. LTL adjustments require labeled treatment duration: Products labeled for chronic use cannot claim short-duration adjustments. Always base the adjustment on the approved indication, not assumed patient behavior.
- 6. Converting AI to specifications requires MDD: The maximum daily dose determines the concentration limit in ppm or ppb. Lower MDD products have higher concentration limits (more permissive in ppm) for the same AI.
- 7. Use published values when available: ICH Q3D PDEs and FDA-published nitrosamine AIs are the authoritative values. Do not recalculate unless dealing with a compound not covered by published tables.
- ---
- ICH Q3D(R1): Guideline for Elemental Impurities (March 2019)
- ICH M7(R1): Assessment and Control of DNA Reactive (Mutagenic) Impurities in Pharmaceuticals (March 2017)
- FDA Guidance: "Control of Nitrosamine Impurities in Human Drugs" (Version 3.0, March 2023)
- FDA Recommended Acceptable Intake Limits for Nitrosamine Drug Substance-Related Impurities (NDSRIs)
- Carcinogenicity Potency Database (CPDB), UC Berkeley
- ICH Q3A(R2): Impurities in New Drug Substances (October 2006)
- ICH Q3C(R8): Residual Solvents (April 2021)
- Kroes, R. et al. "Structure-based thresholds of toxicological concern (TTC): guidance for application to substances present at low levels in the diet." Food and Chemical Toxicology 42 (2004): 65-83.