Christian Rack

Pharmaceutical manufacturers must ensure the quality, safety, and efficacy of their products (1). Examples in the past showed the importance of starting material testing for the quality of the finished product. Unidentified impurities in starting materials, at times, have led to serious harm or even death of patients (2,3,4).

	An essential prerequisite for any analysis is that the sample is representative of the quality of the whole batch (5). A non-representative sample will lead to incorrect assumptions about the quality of the starting material, because it does not truly reflect the characteristics of the lot being sampled (6). The steps from goods receipt to release of the starting material are summarized in 

	Even though quality and other regulatory requirements are mandatory, economic aspects are important as well. When reduced sampling and reduced testing are considered, intelligent approaches fulfill both aspects (7).

	The aim of this article is to give an overview of the viewpoints from different pharmacopoeias and authorities on reduced sampling and testing. The viewpoints are complemented by selected literature. Furthermore, different strategies are discussed to reduce the workload at the steps from sampling to release. The focus of the article is on starting materials or components. 

 provides clarification on terms used herein.

	According to the European Union good manufacturing practice (GMP) guidelines, “the identity of the contents of each container of starting material” has to be assured (8). This requirement is specified by the European Commission (EC) 

 Volume 4, Annex 8, which says, “the identity of a complete batch of starting materials can normally only be ensured if individual samples are taken from all the containers and an identity test performed on each sample” (9).

	According to Annex 8, it is permitted to deviate from this requirement. If it can be assured by a validated procedure that no container is labeled incorrectly, it is permitted to sample only a proportion of the containers. In addition to this, aspects such as the general GMP level, the manufacturing conditions, and the nature of the starting material and the finished drug product should be considered. The Swiss Medic published a technical interpretation that might be helpful for conducting such a validation. The approach involves a scrutinized look at the packaging and labeling process and to assess potential risks at the different process steps (14).

	The reduction via validation is generally possible for starting materials coming from a single product plant or starting materials coming directly from a manufacturer. It is unlikely, however, that a validation could be successful for a starting material that is obtained from brokers or that is for parenteral use (9).

	FDA does not have strict requirements for the number of containers that must be sampled, but the number must be representative, and a scientifically sound sampling plan must be used (15, 16). Aspects like the variability of the material and the quality history of the supplier should be considered (17). ANSI/ASQ Z1.9-2003 for bulk and ANSI/ASQ Z1.4-2003 for discrete material are considered an appropriate standard for a sampling plan (18). An exemption is glycerol (19, 3). Here, FDA and the European Medicines Agency (EMA) share the expectation that some tests have to be performed on each container.

	According to Health Canada, the identity testing of each container is mandatory, but a risk-based reduction is possible. The risk-based reduction is not possible for starting material that is obtained from brokers or for starting material for parenteral use. Starting material coming in different deliveries from the same batch should be considered as different batches (20).

	Detailed requirements for API and excipients used in listed and complementary medicines are given by the Australian Therapeutic Goods Administration (TGA), which distinguishes a pre- and post-qualification phase (21).

	Beside the identity testing, further analysis (e.g., purity, strength, or microbiological testing) is needed to assess the quality of the starting material. It is not required to test those specifications on each container (9). Often, the square root + 1 sampling plan of the World Health Organization (WHO) is used for these purposes (22). But it should be considered that this plan is only one of three options and not WHO’s recommendation for the sampling of starting material. Therefore, the general usage of this plan is seen as critical by the authorities (23, 18).

	The sampling for processing relevant parameters (e.g., water content or particle size distribution for solid oral dosage forms) has its own unique challenges. For each starting material, it must be evaluated if a sampling from different areas within the container is required, based on the possibility of an uneven distribution of the parameters. For the choice of the containers that should be sampled, a computer aided randomization is recommended (18).

) sets clear requirements about the identity testing of monographed substances. The identity tests described in the first identification may be used in all circumstances, whereas the second identification requires full traceability of the starting material to a batch certified to comply with the requirements of the monograph (24).

	The usage of near-infrared spectroscopy (NIR) for pharmaceutical applications has been discussed since the late 1980s (25). The 

. list NIR and Raman spectroscopy as methods for identity testing (26–29). According to EMA, it is permissible to perform identity testing on a statistically representative composite sample when this is supplemented by NIR analysis on every container (30). For the validation of NIR analysis, detailed requirements are given (31, 32).

	It is common practice to assess homogeneity via NIR and ensure identity via compendial methods. No guidance was found related to this practice. However, validation of either the direct identity testing or the homogeneity testing with subsequent identity testing requires similar approaches. Some cases in the past showed that the combination of NIR and the compendial methods was superior to the usage of only compendial methods (33).

	According to the Q&A document of Health Canada, the identity of each container can be assessed by two approaches. First, every container is tested by a discriminating method, which may not necessarily be the monographed method. Second, the identity and the potency are tested on a composite sample (23). 

Use of composite samples for identity testing. 

In general, it is required to perform identity testing on each container. EMA allows the use of composite samples for the identity testing, when the identity of each container is additionally assessed by NIR (30). FDA and Health Canada accept identity testing on composite samples under certain considerations. The main point is that the identity test is specific and selective to detect incorrectly labeled substances in a composite sample (17,20,23).

	For glycerol, FDA and EMA require the testing of each container for identity and the limit test for diethylene glycol (19,3).

Reduction of sampling by choice of container size.

 A pragmatic approach to reduce the efforts for sampling for identity testing could be to increase the container size. As a consequence, fewer tests would be required for the same amount of starting material. However, this approach requires the assessment of potential other problems, such as handling or storage.

	According to Annex 8, it is permitted to use representative composite samples for purposes other than identity testing (9). But no details are given about how many samples of single containers can be combined in a composite sample. An idea can be found in the Q&A document from Health Canada. In the context of identity testing, a maximum of 10 containers can be mixed (23).

	In general, the composite sample should be representative and homogenous. The procedure for the preparation of the composite sample should be described. It should be evaluated to see if the composite sample is homogenous.

 describes the preparation of composite samples of powder starting material. If it is necessary to sample from different places within one container, the samples should not be composited (15,18).

	The term “reduced testing” describes the practice whereby, after receipt of the starting material, not all parameters of the specification are tested. Only a reduced number of parameters are tested, while the other parameters are taken from the certificate of analysis from the manufacturer. This is an established approach in the pharmaceutical industry that requires an established system for the qualification of suppliers (34–37). The DIN ISO 2959-3 from the International Organization for Standardization (ISO) describes the statistical requirements for establishing this approach (38).

	According to ICH Q6, reduced testing should not be established before the market authorization is granted (39). The prerequisite for reduced testing is an audited and qualified manufacturer that has been assessed and classified (34, 35, 36).

	Some requirements in circumstances where reduced testing is permitted or minimum requirements are applicable can be found in different guidelines. According to chapter 5 of the EU-GMP guide, the manufacturer should have appropriate experience with the material (8). ICH Q7 requires full testing on at least three batches (10). Health Canada also recommends testing the first three batches of the starting material and repeating this procedure whenever the manufacturing process has changed (20). Some authors suggest testing the first 30–40 batches in full analysis (37).

	The manufacturer of the medicinal product should establish the frequency for full testing. The accepted differences between the internally generated analytical results and the results of the supplier certificate of analysis should be defined, before the results are compared (35).

	It is industry practice that at each starting material undergoes a full testing on an annual basis. According to Health Canada, it is permitted to do a rotational full testing if one manufacturer supplies more than one starting material. For example, if a manufacturer supplies three starting materials, it is possible to perform a full test on only one of the starting materials each year. At least every five years, a full testing of each starting material should be performed (20).

	If the testing of the drug product manufacturer identifies a discrepancy in comparison with the starting material manufacturer, an investigation should be done first by the drug or medicinal product manufacturer and if he/she does not find the root cause, the investigation should be extended to the supplier/manufacturer of the starting material. Until completion of the measures, the certificate of analysis of the manufacturer should not be accepted (8).

	Skip or periodic testing means that neither the manufacturer of the starting material, nor the pharmaceutical company test a certain parameter (37). If a pharmaceutical company wants to establish skip or periodic testing for a registered specification parameter, this approach must be submitted to the authorities (40, 41). In general, according to FDA, a required criteria cannot be not measured without a surrogate test (42). But for certain parameters (e.g., heavy metals or residual solvents), 

 allows excipient manufacturers to perform skip testing, if validated processes are in place. This approach is also applicable for non-monographed excipients. However, full testing is required annually, and certain parameters, such as identity and strength, must be tested every time. The applied testing frequency of each parameter must be shown on the certificate of analysis (43).

	In the end, it is the responsibility of the pharmaceutical manufacturer to verify if all necessary parameters have been tested. It is recommended to audit the suppliers and to verify the results from the manufacturer with results from in-house testing (43).

	As described in the previous sections, there are different options to reduce the sampling and testing of starting material. The market authorization holder has the ultimate responsibility to ensure the quality of the starting material, but the viewpoints of the different authorities could help to derive a strategy.

	Sampling generally requires identity testing of each container. Reduced sampling is possible, but requires additional work to justify it.  Independently, if the sampling for the full analysis is created from the samples of the identity testing or separately, a general square root +1 sampling plan might be too simple. A risk-based sampling plan is recommended.

	The identity testing of only mixed samples is possible under certain circumstances. The method used must be suitable to identify potential mix-ups even in mixed samples. In combination with a spectroscopic method, such as Raman or NIR, it is possible to perform the monographed or the approved methods on mixed samples.

	The analysis for the full testing requires a representative and homogeneous mixed sample. The preparation of the mixed sample should be described in detail, especially the number of samples that are permitted to be combined in one sample.

	A full testing of all parameters at every goods receipt is not required. Results from the manufacturer’s certificate of analyses can be used. However, it should be defined when it is permitted to move from full testing to reduced testing, at which frequency the full testing is repeated, and under which circumstances it is necessary to move back to full testing.

	The author would like to thank Dr. Jasmin Kohlmann for doing part of the literature research and our valuable discussions about the topic. 

	This article represents a personal view of the author and is not necessarily that of B.Braun.

	1. European Commission, Chapter 1, “Pharmaceutical Quality System,” 

EudraLex Volume 4, EU Guidelines for Good Manufacturing Practice for Medicinal Products for Human and Veterinary Use

, (EC, 31 Jan 2013).
	2. U. Holzgrabe et. al., 

Guidance for Industry, Testing of Glycerin for Diethylene Glycol

 (Rockville, MD, May 2007).
	4. S. Belz, R. Endler, and T. Gumz, 

, No. 38, pp 62 (2011).
	5. European Commission, Chapter 6 “Quality Control”, 

, No. 1, pp 68-75 (2011).
	8. European Commission, Chapter 5, “Production”, 

 (EC, 1 March. 2015).
	9. European Commission, “Sampling of Starting and Packaging Materials,” 

Good Manufacturing Practice Guide for Active Pharmaceutical Ingredients

, Glossary (EC, 31 Jan 2013).
	12. ICH, Q11 

Development and Manufacture of Drug Substances (Chemical Entities and Biotechnological/Biological Entities)

 Title 21, Food and Drugs (U.S. Government Publishing Office, Washington, DC), Part 210.3, pp. 157-158, 1 April. 2019.
	14. Inspektorate–Swissmedic, “Anforderungen an die gebindeweise Garantierung der Identität von Ausgangsstoffen,” Vers. 4.0, 8 August 2018.
	15. US 

, Title 21, Food and Drugs (U.S. Government Publishing Office, Washington, DC), Part 211.84, pp. 164-165, 1 April 2019.
	16. US, 

, Title 21, Food and Drugs (U.S. Government Publishing Office, Washington, DC), Part 211.160(b), p. 172, Apr. 1. 2019.
	17. FDA, “How does FDA interpret the regulations (21 CFR part 211) regarding the establishment of expiry dating for chemicals, reagents, solutions, and solvents?,” Question in 

Questions and Answers on Current Good Manufacturing Practices-Control of Components and Drug Product Containers and Closures

,” (FDA, 28 August 2018.
	18. USP, General Chapter <1097>, “Bulk Powder Sampling Procedures”, 

, pp. 7599-7611 (USP, 3 June 2019).
	19. EMA, 

Guidance on Good Manufacturing Practice and Good Distribution Practice: Questions and Answers

, www.ema.europa.eu/en/human-regulatory/research-development/compliance/good-manufacturing-practice/guidance-good-manufacturing-practice-good-distribution-practice-questions-answers#eu-gmp-guide-annexes:-supplementary-requirements:-annex-8:-sampling-of-starting-and-packaging-materials:-use-of-near-infrared-(nir)-technology-for-container-wise-identity-testing-section.
	20.  Health Canada, “Raw material testing,” 

Good Manufacturing Practices Guide for Drug Products

 (GUI-0001), pp. 32 – 41, 28 February 2018.
	21. Australian Government Department of Health Therapeutic Goods Administration, 

Sampling and Testing for Listed and Complementary Medicines–Technical Guidance on the Interpretation of the PIC/S Guide to GMP,

 Version 2.0 (January 2019).
	22. WHO, WHO Technical Report Series, “Annex 4 WHO Guidelines for Sampling of Pharmaceutical Products and Related Materials,” No. 929, 2005.
	23. Health Canada, 

Good Manufacturing Practices Guide for Drug Products (GUI-0001)

, Appendix B–Questions and Answers “Raw material testing”, pp. 127–131, 28 February 2018.
	24. EDQM, “1. General Notices”, pp. 3 - 10 

 10.0 (EDQM, July 2019).
	25. M. Blanco et al, 

, 123, 135R–150R (1998).
	26. EDQM, “2.2.40. Near-infrared Spectroscopy,” 

 10.0,  pp. 70-75 (EDQM, July 2019).
	27. EDQM, “2.2.48. Raman Spectroscopy,” 

10.0,  pp. 92-94 (EDQM, July 2019).
	28. USP, General Chapter <1119>, “Near-infrared Spectroscopy,” 

,  pp. 7724-7730 (USP, 3 June 2019).
	29. USP, General Chapter <1120>, “Raman Spectroscopy,” 

, pp. 7731-7737 (USP, 3 June 2019).
	30. EMA, 

“EU GMP guide annexes: Supplementary requirements: Annex 8: Sampling of starting and packaging materials: Use of near-infrared (NIR) technology for container-wise identity testing.”
	31. EMEA, “Note for Guidance on the Use of Near Infrared Spectroscopy by the Pharmaceutical Industry and the Data Requirements for New Submissions and Variations,” August 2003.
	32. EMA, 

Guideline on the Use of Near Infrared Spectroscopy by the Pharmaceutical Industry and the Data Requirements for New Submissions and Variations

 (EMA, 27 Jan. 2014).
	33. HK Srivastava, S. Wolfgang, J.D. Rodriguez, 

, 914, pp. 91–99 (31 March 2016).
	34. European Commission, 

Volume 4, “Good Manufacturing Practice, Medicinal Products for Human and Veterinary Use, Part II: Basic Requirements for Active Substances used as Starting Materials” (EC, 13 August 2014).
	35. P. Lienbacher, M. Karner, 

, No. 3, pp. 364-369 (2016).
	36. P. Lienbacher, M. Karner, 

, Volume 2017, Supplement, Issue 1, pp s32-s37 (2017).
	37. UR Mallu et al, 

 5, 2, (2014).
	38. ISO, “ISO 2859-3:2005-Sampling procedures for inspection by attributes-Part 3: Skip-lot sampling procedures,” May 2005.
	39. ICH, 

Q6A Specifications: Test Procedures and Accepatance Criteria For New Drug Substances And New Drug Products: Chemical Substances

, Step 5 version, (ICH, 6 October 1999).
	40. “Guidelines on the details of the various categories of variations”, 

Guidance for Industry Changes to an Approved NDA or ANDA, Rev. 1 

(Rockville, MD, April 2004).
	42. B. Carlin, 

, Volume 31, Issue 9 (2007).
	43. USP, General Chapter <1080>, “Bulk Pharmaceutical Excipients–Certificate of Analysis”, pp. 7527 - 7534, 


	Vol. 43, No. 12
	December 2019
	Pages: 26-30

	When refering to this article, please cite it as C. Rack, "Approaches to Reduced Sampling and Testing for Starting Materials," 

Articles

This article discusses reduced sampling and testing of starting materials or components. Different strategies are presented to reduce the workload at the steps from sampling to release. Viewpoints from the different pharmacopoeias and regulatory authorities, as well as selected literature, are reviewed.