Cleaning Validation: When PDE Reports and the Worst-Case Selection Do Not Align

Shortened excerpt from the GMP Compliance Adviser, Chapter 21.C.2.5, Insufficient cleaning validation

4 min. reading time | by Lea Joos, former GMP inspector with the Government of Upper Bavaria
Published in LOGFILE 21/2026 

Robust PDE reports form the basis of risk-based cleaning validation. This article highlights common deficiencies identified during GMP inspections and explains how they can be avoided.

The Deficiency

During cleaning validation, the PDE values of the active substances were determined in order to implement the EMA Guideline on setting health-based exposure limits for use in risk identification in the manufacture of different medicinal products in shared facilities. For a combination medicinal product, only the active substance with the higher dose was considered because the risk assessment assumed that the active substance with the lower dose would no longer be detectable once the active substance with the higher dose had been cleaned to an acceptable level. Neither the PDE nor the cleanability or solubility of the lower-dose active substance was included in this informal risk assessment. The previous worst-case substance was retained without further justification. 

Furthermore, the following points could not be determined from the PDE reports:

• The scope and completeness of the literature search, including details of the data used
• The criteria used to select the LOEL for the calculation (Ref.: EU GMP Guide Part I, section 5.21, and Annex 15, section 10.6) 

What was the Problem? 

The deficiency had two parts. 

First: Insufficient integration of PDE values into the cleaning-validation worst-case concept

In this case, a PDE had not been calculated for all active substances because the company assumed that, after active substance A had been cleaned to its acceptance limit, “nothing” of active substance B would remain. The company therefore considered a PDE calculation for active substance B unnecessary. 

The assumption that an active substance is “no longer present” is fundamentally problematic: how many decimal places in which unit mean that “nothing” remains? The answer depends on the potency and toxicity of active substance B, and therefore on its PDE value. Only the PDE can establish whether even a very low residue level after cleaning is sufficiently low. In this case, that PDE was missing. 

The company had also not examined whether active substance B would in fact be cleaned “equally well” and whether this conclusion by analogy was valid. Active substance A might, for example, be highly soluble in water, while active substance B might be poorly soluble. The two substances would then behave differently during cleaning. 

Active substance A also remained classified as the worst-case active substance, irrespective of the PDE values calculated for active substances C and D. The continued validity of this classification had not been assessed on the basis of the PDE values for active substances A, C and D. 

Second: Deficiencies in the PDE report 

The PDE report did not identify the databases used for the literature search. It was therefore impossible to assess whether the literature search was complete in accordance with the EMA PDE guideline (Guideline on setting health-based exposure limits for use in risk identification in the manufacture of different medicinal products in shared facilities).

The PDE report derived the PDE value for active substance A from the LOEL (Lowest Observed Effect Level). However, the EMA guideline first requires identification of the NO(A)EL. If the NO(A)EL is not used, or if several NO(A)ELs exist, the selection of the value used for the subsequent calculation must be justified. This justification was missing. 

The Most Common Pitfall

Existing cleaning-validation concepts frequently used worst-case approaches in which only a small number of active substances were validated as representatives of all active substances. When the EMA guideline was introduced, companies had therefore already established a “worst-case active substance”, although it was based on a different approach. As a result, PDE values were often calculated without a defined concept for integrating them into the existing cleaning-validation programme.

Very few companies have in-house toxicologists who can prepare a PDE report or assess its quality. Employees without toxicological training often do not feel qualified to review such a report.

The report prepared by the external toxicologist is therefore submitted and simply “filed”. In doing so, the contracting company overlooks its own responsibility and accepts the external toxicologist’s assessments and conclusions without critical review. 

How to Avoid This Error 

A defined concept for integrating PDE values into cleaning validation should answer the following questions:

• For which substances must PDE reports be prepared, and for which might they not be required? Which criteria apply?
• Can different active substances be grouped for the purpose of PDE reports? What would be required for this?
• How is the PDE value incorporated into the selection of the worst-case active substance? Which other criteria are considered, and how are they weighted? Are there criteria that must always result in inclusion in the validation
• Can PDEs or ADEs derived from toxicological assessments outside a medicinal-product context be applied to medicinal products? Which minimum criteria must be met? 

An approach for determining the worst-case product is presented in Chapter 8.D Establishing the scope of validation. Selection criteria include active-substance content, solubility and therapeutic dose. Other established criteria include cleanability and toxicity. Toxicity can be substantiated particularly well by the result of the PDE assessment and should therefore be considered when selecting the worst-case product. The different criteria must be weighted on a risk basis.

PDE values may be grouped – for example, a PDE of 1–50 µg/day as “high toxicity”. Such groupings may not adequately reflect the criticality of an individual active substance. In these cases, it must be treated as a worst case on the basis of a single criterion, such as its PDE value or solubility.

It is very difficult to justify not calculating a PDE value for an individual active substance. Any such justification must consider toxicity as well as active-substance content and cleanability. And how can toxicity be substantiated if not by the PDE value? 

Whatever you do, first ask yourself:

Do I really have a justification based on scientific data? 

The company should define its own requirements for external toxicological reports so that they can also be reviewed by non-toxicologists. The objective is to ensure the logical traceability of the toxicologist’s assessment, particularly with regard to:

• The nature and scope of the literature search, including the databases used
• The criteria used to derive the most relevant NO(A)EL, or the justification for using a LO(A)EL
• The criteria used to derive the adjustment factors for the PDE calculation 

Employees without toxicological training can then verify compliance with these requirements and identify major deficiencies in areas such as the literature search and the traceability of the documentation.

Do you have any questions or suggestions? Please contact us at: redaktion@gmp-verlag.de