Cleaning Validation in Pharma: Complete Guide to HBEL and MACO

In short: Cleaning validation is documented evidence that a cleaning procedure reliably removes the residues of the previous product, cleaning agents and microbes from shared equipment to a level that is safe for the next product. Today that safe level comes from a toxicological health-based exposure limit (HBEL), not from a rule of thumb such as 10 ppm. This guide takes QA, QC, production and validation staff and students through the lifecycle, with a fictional MACO worked example, a printable checklist and FAQs.

What is cleaning validation, and why do we do it?

When one machine makes Product A on Monday and Product B on Thursday, some of A can end up in B. That is cross-contamination. For a mild drug the harm may be small. For a hormone, a cytotoxic, a penicillin or any potent drug, a few micrograms in a patient’s daily dose can matter.

EU GMP Annex 15 (2015) defines cleaning validation as documented evidence that an approved cleaning procedure will reproducibly remove the previous product or cleaning agents from the equipment below the scientifically set maximum allowable carryover level. Section 10 of the annex begins by expecting cleaning validation for all product-contact equipment.

In the US, 21 CFR 211.67(a) requires equipment and utensils to be cleaned, maintained and, where appropriate, sanitised or sterilised at appropriate intervals to prevent malfunctions or contamination that would alter the safety, identity, strength, quality or purity of the drug product beyond established requirements. Section 211.67(b) also requires written cleaning procedures, including protection of clean equipment from contamination before use and inspection for cleanliness immediately before use.

For the basic ideas, read principles of cleaning validation. For a procedure written as an example SOP, see the SOP on cleaning validation. This guide is the wider view: how limits are set today, how the numbers work, and what inspectors expect.

The lifecycle approach

Treat cleaning as a process with a life, not a one-time exercise.

Stage 1: Design

• Know the residues: active, excipients, degradants, cleaning agent, and microbes or endotoxin where relevant.
• Set the limit from toxicology, documented in a risk assessment with references, including a limit for the cleaning agent (Annex 15 §10.6).
• Develop the procedure: agent, concentration, temperature, contact time, rinse volume, drying. Small laboratory cleanability studies help.
• Find the hardest spots: dead legs, gaskets, valves, agitator shafts, sieve meshes, punch tips, filling-line tubing.
• Assess variable factors such as operators and the level of detail in the procedure; if they matter, validate on the worst case (Annex 15 §10.5).

Stage 2: Qualification (the validation runs)

Run the approved protocol on the worst-case product and equipment. Annex 15 §10.13 says the procedure should be performed an appropriate number of times, based on risk assessment, and meet the acceptance criteria. Three consecutive runs is a common choice, but the text fixes no number, so justify yours. For the layout, see the cleaning validation protocol format.

The US FDA inspection guide Validation of Cleaning Processes (7/93) describes the same flow: written general procedures, a specific protocol written beforehand, documented execution and a management-approved report. Annex 15 §10.2 adds that repeated cleaning and retesting until acceptable results are obtained is not acceptable.

Stage 3: Continued verification

Show that the process stays in control: visual inspection after every cleaning, periodic swab or rinse testing at a justified frequency (Annex 15 §10.15 stresses this for manual cleaning), trending against alert limits, and change control review (Annex 15 section 11). The EMA Q&A on HBELs (Q7) expects analytical testing at each changeover in shared facilities unless a documented quality risk assessment justifies otherwise, weighing repeatability (manual is less repeatable than automated), product hazard and whether visual inspection is reliable at the justified limit.

Worst case and grouping

You cannot validate every product on every machine. Annex 15 §10.1 allows grouping of similar equipment with a justification of the unit chosen. §10.10 requires a scientific rationale for the worst-case product and an assessment when a new product arrives; criteria named include solubility, cleanability, toxicity and potency.

1. List all products made on the equipment train.
2. Score each on solubility in the cleaning medium, ease of cleaning, and potency or toxicity.
3. Choose the worst-case previous product (hard to clean, toxic) and the worst-case next product (smallest batch, highest daily dose).
4. Group equipment of the same design and material, and pick the hardest unit.
5. Repeat the matrix whenever a product is added.

For campaign manufacture, §10.9 says the maximum campaign length (in time or batches) should be the basis for validation.

Setting limits: HBEL and PDE versus 10 ppm and 1/1000 dose

The older criteria

The FDA 1993 guide mentions analytical levels such as 10 ppm, biological levels such as 1/1000 of the normal therapeutic dose, and visual cleanliness, and says limits should be practical, achievable and verifiable. WHO TRS 1019, Annex 3, Appendix 3 (2019) lists the same three: nothing visible; no more than 10 ppm of one product in another; and no more than 0.1% of the normal therapeutic dose in the maximum daily dose of the next product. It says the most stringent option should be used.

These older approaches are simple but not tied to actual toxicity. For a harmless substance 10 ppm can be needlessly strict; for a potent drug it can be unsafe.

The HBEL approach

The EMA Guideline on setting health based exposure limits for use in risk identification in the manufacture of different medicinal products in shared facilities (EMA/CHMP/CVMP/SWP/169430/2012) was adopted by CHMP on 20 November 2014 and came into effect on 1 June 2015. It derives a permitted daily exposure (PDE), a substance-specific dose unlikely to cause an adverse effect if a person is exposed at or below it every day for a lifetime. The PDE is the NOAEL of the critical effect multiplied by a body-weight adjustment and divided by adjustment factors F1 to F5, using the method of ICH Q3C(R4) Appendix 3. A qualified toxicologist or pharmacologist must do this work.

Annex 15 §10.6 now says carryover limits should be based on a toxicological evaluation, with the justification documented in a risk assessment. PIC/S transposed the guideline for its members as PI 046-1 (in force 1 July 2018). WHO’s Points to consider when including HBELs in cleaning validation (TRS 1033, Annex 2, 2021) says historical approaches often only showed that a procedure met historical limits, and that manufacturers should show residues are removed to levels considered safe for patients.

The EMA Q&A: the HBEL is a ceiling, not the working limit

The EMA Q&A (EMA/CHMP/CVMP/SWP/246844/2018, 19 April 2018) clarifies:

• HBELs should be established for all medicinal products (Q1).
• The guideline is not intended to set cleaning limits at the level of the calculated HBEL. Limits actually used are treated as alert limits that, given cleaning process capability, give sufficient assurance that excursions above the HBEL are prevented. Results above an alert limit trigger an investigation, and repeated excursions are not acceptable if the cleaning is not in control (Q6).
• For visual inspection, establish the threshold at which residue is readily visible under real lighting and distance, inspect all product-contact surfaces with tools where needed, and train operators, with periodic eyesight tests (Q8).

MACO: the calculation, step by step

MACO is the maximum allowable carryover: the largest mass of the previous product’s residue allowed in the next batch. No guideline above prescribes one formula. The forms below are common industry practice; the HBEL form follows directly from the definition of the PDE.

MBS is the minimum batch size of the next product, MDD its maximum daily dose (as mass of product), and TDD the normal therapeutic daily dose of the previous product. Keep units consistent (mg).

Surface limit (µg/cm²) = MACO (µg) ÷ shared product-contact surface area (cm²). The swab limit is that multiplied by the swab area; the limit in the extract is that divided by the extraction volume.

Worked example (fictional)

All products and numbers are invented for teaching.

Previous product (worst case): “Alphacin 100 mg tablets”, PDE 1.2 mg/day, normal daily dose 100 mg. Next product: “Betazole 500 mg tablets”, smallest batch 300 kg, maximum daily dose 3 tablets = 1,500 mg. Shared surface area: 25 m² = 250,000 cm².

Step 1, HBEL-based MACO. Daily doses in the batch = 300,000,000 mg ÷ 1,500 mg = 200,000. MACO = 1.2 mg × 200,000 = 240,000 mg = 240 g.

Step 2, older criteria. 10 ppm: 10 mg/kg × 300 kg = 3,000 mg = 3 g. 1/1000 dose: 100 × 300,000,000 ÷ (1000 × 1,500) = 20,000 mg = 20 g. Here the older criteria are far stricter than the HBEL, which is common for low-hazard drugs. For a highly potent drug the order reverses.

Step 3, working limit. The HBEL (240 g) is the ceiling. The site sets an example in-house criterion of 10 g as the acceptance limit and 5 g as the alert limit, because the cleaning easily achieves it. A site may choose a limit below the HBEL, never above.

Step 4, surface limit. 10 g = 10,000,000 µg ÷ 250,000 cm² = 40 µg/cm².

Step 5, swab limit. Swab area 5 cm × 5 cm = 25 cm², so 40 × 25 = 1,000 µg per swab. With 20 mL extraction, the limit is 50 µg/mL before recovery correction.

Step 6, recovery. If recovery from stainless steel is 80%, a measured 36 µg/mL is corrected to 36 ÷ 0.80 = 45 µg/mL, within 50 µg/mL. The protocol must state how recovery is applied.

Two cautions. The simple formula assumes even spread; real residue is not even, so sample the worst spots. And Annex 15 §10.6 says acceptance criteria should consider the cumulative effect across the equipment train.

Sampling: swab, rinse and visual inspection

Annex 15 §10.12 says sampling is by swabbing and/or rinsing or other means, the sampling materials and method must not influence the result, and recovery must be shown from all product-contact materials sampled with all methods used. Protocols must state the locations and why they were chosen (§10.11).

For the swabbing procedure, see the SOP for sampling by swab and rinse.

Recovery studies

Spike a known amount of residue at the limit onto a coupon of each product-contact material (stainless steel, glass, PTFE, silicone gasket), swab as in the real procedure, and measure. Example in-house criterion: three levels (50%, 100%, 150% of the limit), three replicates each, two analysts, minimum recovery 70% with low variability. The guidelines fix none of these figures. Improve the technique if recovery is low or erratic, and record the factor used.

Visual inspection done properly

Run a visible residue study: apply the product at falling levels on a coupon under shop-floor lighting and note the level at which it vanishes from sight. If that is above the chemical limit, “visually clean” alone is not enough.

Hold times

Annex 15 §10.8 says the time between manufacture and cleaning, and between cleaning and use, should be taken into account to define dirty and clean hold times.

• Dirty hold time: the longest wait between end of use and start of cleaning. Residues dry and harden. Hold soiled equipment for the maximum proposed time, then clean normally.
• Clean hold time: the longest storage of cleaned equipment before use. Test cleaning agent residue, bioburden and endotoxin at the end of the proposed time, in the real storage condition.

Example in-house criterion: dirty hold 48 hours and clean hold 72 hours for covered equipment in a Grade D room, each proved with three runs. Use your own data. For an injectable-area case, see the hold time study report for dirty equipment in the ampoule section.

Cleaning agents, microbes and endotoxin

Annex 15 §10.6 asks for limits for the removal of cleaning agents. Choose agents that rinse easily and are easy to detect, and test the final rinse for the agent or a marker such as conductivity or pH. §10.7 asks that microbial and endotoxin risk be considered when writing protocols. For sterile and liquid products include bioburden, and for parenterals bacterial endotoxin, in the criteria.

Analytical methods: TOC, HPLC and others

The method must measure at the limit, in the presence of the swab and the cleaning agent.

The FDA guide stresses knowing each method’s specificity and sensitivity, and notes that a negative result alone does not prove absence of residue. Make sure the data stands up on data integrity too: no trial injections, audit trails on, every injection reviewed.

When to revalidate

The FDA 1993 guide expects written procedures to state acceptance criteria and revalidation triggers. Typical triggers, all through change control:

• A new product more potent, less soluble or harder to clean than the current worst case (Annex 15 §10.10).
• Change in formulation, batch size, daily dose or HBEL that alters the limit.
• Equipment change, major repair or change of surface material.
• Change in cleaning agent, concentration, steps, times or temperatures.
• Longer campaign or hold time.
• A failure or alert-limit excursion; handle as a deviation and, if needed, a CAPA (see the deviation management guide).
• Periodic review showing drift, or new toxicology data; the EMA Q&A notes HBEL source data needs periodic re-assessment.

Documentation and record retention

The protocol, raw data, chromatograms, recovery studies, swab maps, report and equipment cleaning logbooks are GMP records. Keep batch-related records at least 1 year after the expiry date of the respective product batch (EU GMP Part I Chapter 4 §4.11, or 5 years after QP certification if longer; 21 CFR 211.180(a)), longer if local law or the marketing authorisation requires. The HBEL assessment, with its data references and the experts involved, should be available on request at inspection (EMA Q&A, Q5); Q4 describes the toxicology or pharmacology expertise expected of the person who derives it. The quality risk management guide covers the assessment tools.

Printable checklist: cleaning validation readiness

This is an example in-house checklist; tick each item before you sign the protocol.

Frequently asked questions

What is the difference between cleaning validation and cleaning verification?

Validation is a planned programme proving the procedure works repeatedly. Verification is routine testing after a batch or campaign. Annex 15 §10.3 accepts verification after each batch for some products, for example investigational medicinal products.

Is 10 ppm still acceptable?

It can remain a practical in-house limit if it is below the HBEL-based limit and justified, but it cannot be the only basis. Annex 15 §10.6 expects a toxicological evaluation, and the EMA Q&A says HBELs should exist for all products.

Do we need validation when the same product is made in consecutive batches?

The FDA 1993 guide says cleaning between batches of the same product needs only a visibly clean standard. Still, control campaign length, degradation and microbial growth (Annex 15 §10.9).

How many validation runs are needed?

An appropriate number, based on risk (Annex 15 §10.13). Three is common but not a stated rule; manual or difficult-to-clean processes may need more.

Can TOC replace a specific HPLC method?

Annex 15 §10.6.2 allows other representative parameters such as TOC and conductivity where specific testing is not feasible. TOC suits water-soluble residues and agents. Use a specific method for very potent drugs or if the agent contains carbon that biases the result.

What if a result is above the limit?

Treat it as a deviation. Do not simply re-clean and re-test. Find the cause (procedure, operator, agent, hold time, method), correct it and repeat the run as justified.

References

1. EudraLex Volume 4, EU GMP Guide, Annex 15: Qualification and Validation (October 2015), sections 10 and 11.
2. EudraLex Volume 4, Part I, Chapter 4: Documentation, §4.11.
3. EMA, Guideline on setting health based exposure limits for use in risk identification in the manufacture of different medicinal products in shared facilities, EMA/CHMP/CVMP/SWP/169430/2012 (effective 1 June 2015).
4. EMA, Questions and answers on implementation of risk-based prevention of cross-contamination in production and the HBEL guideline, EMA/CHMP/CVMP/SWP/246844/2018 (19 April 2018).
5. PIC/S, PI 046-1: Guideline on setting health based exposure limits for use in risk identification in the manufacture of different medicinal products in shared facilities (in force 1 July 2018).
6. US FDA, Guide to Inspections: Validation of Cleaning Processes (7/93).
7. 21 CFR 211.67 and 211.180(a), eCFR (current text checked October 2026).
8. WHO TRS No. 1019 (2019), Annex 3, Appendix 3: Cleaning validation.
9. WHO TRS No. 1033 (2021), Annex 2: Points to consider when including HBELs in cleaning validation.
10. ICH Q3C(R4), Appendix 3 (PDE method).