EMA publishes final Q&A Document on the use of PDE (HBEL) Values

In April 2018, the EMA published the final version of the Q&A document on the use of Health-Based Exposure Limits (HBELs) without special notice. The draft version had become necessary in 2016 because the EMA had generated uncertainty within the pharmaceutical industry with their Guideline on setting these limits (PDE values as well). The Permitted Daily Exposure (PDE) values had been defined by the EMA in a new guideline in order to mathematically decide upon the dedicated or multi-purpose use of plants. The new health-based values influence the cleaning validation threshold values as well. In short, it can be determined how much cross contamination may be tolerated from a health risk point of view by using the PDE values. For identifying and using these values, the EMA had to provide a Q&A document which has now been published in its final version.

The document currently available mostly talks of HBELs and the term PDE is used only very occasionally. HBEL seems to be used as a general term and PDE as a part of HBEL. The threshold values described as ADE (Acceptable Daily Exposures) in ISPE Risk-MaPP would also be a sort of HBEL. When calculating the cleaning threshold values, the terms of PDE, ADE and HBEL are used interchangeably.

The differences between the draft and the final version of the Q&A document in brief:
The number of questions listed has been reduced from 14 to 13. The order and to some extent the content of the questions were also changed.
The statement that Health-Based Exposure Limits (HBELs) are required for all products is basically the same (question 1). In addition, these must be checked repeatedly during the product life cycle.

The question of what constitutes a highly hazardous substance (question 2) is no longer included in the document. Instead, question 2 states that every substance is to be classified in a band somewhere between a low up to a very high hazard potential. This means that classifying the substances as "highly hazardous" and "non-highly hazardous" is not considered reasonable.
There is a new question about how a producer should handle HBELs (question 3). Here as well, they refer to risk management processes. In case that these indicate that the specified control measures do not sufficiently protect from contamination, dedicating the production should be considered.

Another new question deals with who should develop resp. define the HBEL values (question 4). According to the current paper, this can be a person with the appropriate toxicological/pharmacological expertise who is familiar with medicinal products and with determining exposure limits (apart from occupational exposure limit, PDE limits are mentioned here as well). This is somewhat confusing, since PDE values are a subset of HBELs.

Question no. 5 is also new: in which way are clients responsible towards contract manufacturers? It is clearly stated that the client is obliged to provide a complete HBEL assessment to the contract manufacturer or else the required data so that the contract manufacturer himself can prepare a HBEL document (or have it prepared).

Question no. 6 remains unchanged: How can threshold values for cleaning be established? While in the draft, the traditional calculations according to the 1/1000 dose or 10 ppm criterion for 'non-highly hazardous' are named as an option, the new document has a different passage. For existing products, the existing limit values for cleaning shall be maintained and regarded as alert values, provided that they are significantly higher than the corresponding HBEL values. So the content stays the same: less strict HBEL values should not lead to a less well-executed cleaning.

The 8th question, which is also new, is interesting as well. It deals with the "visually clean" criterion. It is possible to use this as a criterion for success in cleaning, provided that numerous preconditions are fulfilled. For instance, surface-specific threshold studies are required, all relevant surfaces must be visible (if necessary, by dismantling), light and distance must be defined and the examiners need a specific training etc.
Question 9 is the former question 13. Separating a product class into a dedicated section without further measures to prevent cross contamination within this class is still not allowed.

Question 10 is the former question 5: it is still not possible to use an LD50 value to obtain an HBEL.

Question 11 is the former question 7 (antiparasitic treatment); question 12 is the former question 8 (veterinary medicine), question 13 covers development products. However, it no longer addresses whether HBELs are required for them; the answer to question 1 clearly indicates that these are also required for development products. In fact, it is even pointed out that HBELs have to be reviewed periodically, if the associated data change within the product life cycle - which applies mainly to development products after all.

For more information please see the the new Q&A paper on the use of HBEL resp. PDE values.