How do you calculate MACO and set HBEL- or PDE-based cleaning validation limits?

1What a health-based cleaning limit is.

EU GMP Annex 15 says limits for the carryover of product residues “should be based on a toxicological evaluation”, with the justification documented in a risk assessment, and points to EMA’s guideline on health-based exposure limits.¹ PIC/S describes cleaning validation as evidence that cleaning reproducibly removes residues to below “scientifically set maximum allowable carry-over (MACO) levels”.² WHO calls the same quantity the maximum safe carryover: the most residue that can carry into the next product “without presenting an appreciable health risk to patients”.³

The health-based exposure limit (HBEL) behind it is usually a permitted daily exposure (PDE): “a substance-specific dose that is unlikely to cause an adverse effect if an individual is exposed at or below this dose every day for a lifetime”.⁴ EMA’s questions and answers of April 2018 say HBELs “should be established for all medicinal products”, not only for the potent ones.⁵

In the US, 21 CFR 211.67 requires equipment to be cleaned at appropriate intervals to prevent contamination that would alter the drug’s safety, identity, strength, quality or purity, under written procedures.⁶ FDA’s inspection guide on cleaning validation does not set acceptance limits; it expects the firm’s rationale to be “logical” and the limits “practical, achievable, and verifiable”.⁷

2How a PDE is derived.

EMA’s guideline, in effect since 1 June 2015, derives the PDE in four steps: review all relevant data to identify hazards, identify the critical effects, find the no-observed-adverse-effect level (NOAEL) for those effects, and divide by adjustment factors for uncertainty.⁴

PDE = NOAEL × weight adjustment ÷ (F1 × F2 × F3 × F4 × F5)

The guideline asks for the PDE on a mg/kg body weight basis, with a standard 50 kg body weight where the next product’s dose is stated per patient. If no NOAEL is found, the lowest-observed-adverse-effect level may be used, and a benchmark dose is an accepted alternative. Where several critical effects give several PDEs, the lowest is used by default.⁴ For genotoxic substances with no threshold, a limit of 1.5 µg per person per day may be applied, and a route-to-route correction is needed where bioavailability differs clearly, for example by more than 40%.⁴ EMA says LD50 “is not an adequate point of departure”.⁵

The person deriving the HBEL needs expertise in toxicology or pharmacology and experience of setting exposure limits, and a bought-in assessment needs a recorded assessment of the provider.⁵ WHO adds that the report should be peer-reviewed by another qualified expert.³

3How MACO and the surface limit are calculated.

WHO’s points to consider set out the calculation in steps, for a preceding product A and a following product B.³

1. Maximum safe carryover (g) = PDE of A (µg/day) × batch size of B (kg) ÷ maximum daily dose of B (mg/day).
2. Tabulate the PDE, maximum daily dose, batch size and shared equipment surface for every product.
3. Maximum safe surface residue (mg/m²) = maximum safe carryover (g) × 1,000 ÷ surface shared with B (m²).
4. Tabulate the surface limit for every preceding and following pair, and identify where it cannot be met given the cleaning procedure’s capability or the compound’s visible residue limit.

As an illustration with invented figures: if product A has a PDE of 10 µg/day and product B is made in 100 kg batches with a maximum daily dose of 400 mg, the maximum carryover is 10 × 100 ÷ 400 = 2.5 g. Over 25 m² of shared surface that is 2.5 × 1,000 ÷ 25 = 100 mg/m², or 10 µg/cm².

WHO says the calculation should include the PDE, maximum daily dose, batch size, total shared surface area, sample areas, sample dilution volumes and recovery factors, and that the margin of safety should be identified.³ PIC/S asks for criteria for each item of equipment and for the cumulative carryover across the whole train, consideration of unit doses, and no assumption that a contaminant is spread evenly through a batch.² Annex 15 also asks for limits for cleaning agents.¹

4From a calculated limit to acceptance criteria.

The HBEL justifies the limit; it is not the limit. EMA says its guideline “is not intended to be used to set cleaning limits at the level of the calculated HBEL”. Historically used limits can be kept as alert limits if, given the cleaning process’s capability, they prevent excursions above the HBEL; a result above the alert limit triggers an investigation, and repeated excursions are not acceptable where they show the cleaning method is not in control.⁵ PIC/S says the same and suggests alert limits based on historical limits or a statistical evaluation of past results.² WHO puts it differently: historical limits may be used where they are more stringent than the HBEL-based limit, but alert and action limits should come from statistical analysis of existing data.³

FDA’s guide records limits mentioned in industry literature, such as 10 ppm, 1/1000 of the normal therapeutic dose and no visible residue.⁷ WHO notes that the criteria in historical GMP texts “did not account for HBELs” and that an HBEL-based limit should be calculated and compared against any existing one.³

Visual inspection matters but rarely suffices. Annex 15 says it is “not generally acceptable” for a visual check to be the only criterion, and that cleaning and retesting until results pass “is not considered an acceptable approach”.¹ In shared facilities EMA expects analytical testing at each changeover unless a documented risk assessment justifies otherwise, weighing how repeatable the cleaning is, the product’s hazard and whether visual inspection can detect residue at the HBEL-justified limit.⁵ PIC/S expects verification after every changeover where MACO falls below the limit of visual detection.²

Methods should be validated, with detection and quantitation limits suited to the residue levels. Non-specific methods such as total organic carbon (TOC) may be justified, but the result must then be treated as “entirely due to the target residue”.² For therapeutic proteins and peptides, which cleaning conditions typically degrade, a PDE for the intact product may not be required.⁴¹

5How it runs in Seal.

In Seal, an acceptance limit is a governed calculation rather than a number copied into protocols. The cleaning validation blueprint keeps each limit’s toxicological or pharmacological input, dose basis, batch sizes, surface area, safety factors, units, formula version, rounding, rationale, reviewer and effective dates, and connects it to the product-equipment matrix, the worst-case rationale, cleaning execution, hold times and swab and rinse results. A change to a toxicological input, procedure or surface area identifies the calculations, studies and scheduled campaigns in scope. The manufacturer remains responsible for the toxicological conclusions and the acceptance strategy.

Neil, Seal’s AI agent, can review cleaning validation evidence, mapping results to equipment locations, recovery assumptions and protocol limits before drafting the review. It does not derive a safety limit or release equipment from an incomplete package; validation and quality owners approve the scientific basis and conclusion.