An APIC multinational working group has compiled a new guidance on cleaning validation with the title "APIC Guidance on Aspects of Cleaning Validation in Active Pharmaceutical Ingredients Plants". Publication date is May 2014 and the document can be downloaded from the APIC website. The following is a summary description of the document. The document contains 55 pages and is subdivided into 13 chapters.
The topic cleaning validation gained new importance in the EU with the publication of the EMA Guideline "Guideline on setting health based exposure limits for use in risk identification in the manufacture of different medicinal products in shared facilities" and with the chapter Cleaning Validation in the draft of the revision of Annex 15. The foreword refers to the integration of cleaning validation within a quality system supported by quality risk management processes in order to protect the patients. According to the authors the document is aligned with ISPE Risk-MaPP and it recommends the revised PDA Technical Report 29 as a valuable guidance document. The document is supposed to assist companies in cleaning validation and to serve as a starting point for internal discussions. It should in no way be considered as a technical standard. The document addresses six topics:
Acceptance Criteria (Chapter 4)
The acceptance criteria preferably should be based on the acceptable daily exposure (ADE) calculations whenever this data is available. Alternatively, occupational exposure limits (OEL) are recommended as acceptance criteria. The document then presents examples for the calculation of acceptance criteria. Based upon the ADE the maximum allowable carryover (MACO) is calculated by way of example.
As (further) examples for acceptance criteria and their calculation the document lists:
The same chapter addresses the calculation of acceptance criteria by means of swab and rinse tests. The chapter closes with a rationale for the use of different limits in pharmaceutical and chemical production of active pharmaceutical ingredients. A competent chemist should accompany the finding of a rationale. It is mentioned expressly that in the production of active pharmaceutical ingredients the risk of carry-over of contaminants because of manufacturing processes (such as extraction or filtration ...) can be much lower than in the manufacture of medicinal products. Based on this consideration limits could be set higher. Naturally, these considerations do not apply to physical manufacturing processes (drying, milling). In this case, the methods applied should be those normally used in pharmaceutical production. Annex 1 contains five examples of MACO calculations according to the acceptance criteria mentioned above.
Levels of Cleaning (chapter 5)
Depending on the step of manufacture and/or the use of multi-purpose equipment the cleaning intensity varies. The document recommends that at least three levels of cleaning (level 0-2) are implemented according to the difficulty of cleaning. Depending on the level different activities are required as concerns visual inspection, analytical verification and cleaning validation. An example presents different product changeover scenarios. In that case the number of cleaning validation activities is defined with at least three consecutive runs. In cases in which it takes some time to finalize the cleaning validation cleaning verification has to be performed in the meantime. In the course of cleaning validation the dirty hold time should also be determined.
The whole process is divided into:
Control of Cleaning Process (chapter 6)
The document gives indications concerning the minimum content of cleaning instructions for manual cleaning and for cleaning with automated systems in chapter 6.
Bracketing and Worst Case Rating (chapter 7)
At the very beginning of chapter 7 it is pointed out that a grouping of substances and the use of a worst case product can be used in order to simplify the cleaning validation. It is important in any case that a documented scientific rationale for the chosen worst case exists. The basis for a bracketing procedure should be defined in a company policy, or an SOP. A fictive example then shows a way of implementation as concerns the equipment train, substances and cleaning procedures. Again by way of example, a worst case rating study/risk assessment is presented. Criteria to be taken into consideration:
The single criteria are rated according to their risk. For this the guidance makes proposals of categorization (for instance concerning cleanability) and of classification concerning solubility, ADE concept, therapeutic dose (1-3 oder 1-5). A table summarises the single data and identifies worst case products for the cleaning validation. Explicit reference is again made to change control which might make necessary a new worst case calculation.
Determination of the Amount of Residue (chapter 8)
At the beginning of chapter 8 a practical guidance is provided concerning the validation of analytical methods in the course of a cleaning validation; regulatory requirements (for example FDA) and guidelines are indicated (such as ICH Q2 R1). The second part addresses sampling methods (swab/rinse). A combination of the two methods is described as being generally the most desirable. Interestingly, the subchapter swab sampling mentions the analysis of the first production batch of the following product for impurities as "may" option. In its cleaning validation guidance FDA is critical as concerns this procedure. For swab sampling as well as for rinse or solvent sampling equations are given.
There is also mentioned a third method which is rather unknown - the so-called "stamps" or "coins". These elements are placed on appropriate sampling points in the equipment and remain there during the manufacture of the previous product and during cleaning. After cleaning, the contamination on the coins is analysed and the overall contamination is calculated.
Cleaning Validation Protocol (chapter 9)
In chapter 9 you can find a cleaning validation protocol which is six pages long and divided into 9 chapters:
In Chapter 10 "Validation Questions" frequently asked questions are answered in the style of a Q&A paper. In this document "A" is short for "advice". The questions are answered with references to the single chapters/subchapters in the document.
The document is completed with the chapters 11 (References), 12 (Glossary) and 13 (Copyright and Disclaimer).
Conclusion:
In spite of the many pages the document can be read very well. The many practical examples (such as calculations, cleaning validation protocol,...) may be helpful. The document addresses the ADE concept very intensively. This is a question which confuses industry at the moment. But the document also addresses acceptance criteria already known (dose criterion, amount criterion). It is striking that the document still adheres to the three batches concept. Although the document addresses active pharmaceutical ingredients plants it will surely find readers among manufacturers of medicinal products.