Tara Lin Couch,EAS Consulting Group
The establishment of specifications for raw materials, in-process materials, and finished products is one of the essential regulatory requirements of current good manufacturing practices (cGMPs) for dietary supplements as established by the FDA1-3. These specifications define material and product quality. To ensure that the necessary characteristics—identity, purity, strength, composition, and limits of contaminants—are present, companies must perform thorough material and product testing.
Testing can be conducted either by the manufacturer’s in-house laboratory or by a contract laboratory. In either case, the results will determine whether or not the company can release the material for use and distribute the product. When a test result falls outside of the acceptance criteria for a specified characteristic, the material is considered “out of specification” (OOS). All OOS results must be investigated in accordance to the cGMPs, and quality control personnel must conduct a material review and make a disposition decision4.
OOS investigation
The purpose of an OOS investigation is to determine the cause of the OOS result during testing or manufacture. The investigation is usually, and most practically, conducted as two separate yet linked investigations. First, a laboratory OOS investigation determines if the OOS result is valid. If the result is valid, the next step is to perform a non-conforming material or product investigation. For dietary supplement materials or products, this non-conforming investigation is referred to as a material review1. Even when the particular OOS material or product lot is rejected for use or distribution, the investigations are necessary to determine the cause and whether the issue may impact other lots of materials and/ or products. The FDA Center for Drug Evaluation and Research (CDER) issued a “Guidance for Industry: Out of Specification (OOS) Test Results for Pharmaceutical Production” that describes the agency’s current thinking on the matter. Although this guidance focused on pharmaceutical production, it applies to all FDA-regulated industries that must conduct an OOS investigation, including dietary supplements. It is noted that the “responsibility of a contract testing laboratory in meeting these requirements is equivalent to that of a manufacturing firm” as dictated in the FDA guidance. Laboratory OOS investigations
In addition to determining if the original OOS result is valid, the laboratory OOS investigation should look for “out of trend” (OOT) and unusual test results. OOT test results are those not within the general tendency, pattern, or direction of a series of events, conditions, or previous results. It’s important to note that an OOT test result may or may not be OOS. The most common OOT event is a stability test result that is not consistent with the stability profile generated from previous time intervals or other historical stability studies. An unusual test result is not necessarily OOS but is aberrant, atypical, or otherwise unexpected. Examples of unusual test results that should be investigated using the laboratory OOS investigation process include results that exceed set action or alert levels or do not meet a target level. Sudden occurrences of contaminants not previously observed and the presence of unexpected or extraneous chromatographic peaks should also be considered unusual and investigated.
Upon the occurrence or discovery of an OOS, OOT, or unusual result, the analyst performing the test must notify their supervisor to initiate the laboratory OOS investigation. The laboratory OOS investigation itself is organized into two phases as depicted in Figure 1. As the figure shows, Phase I includes both a documentary review and experimental examination, while Phase II includes only an expanded experimental examination5. This type of thorough OOS process is especially important for contract labs to employ because much of the dietary supplement industry outsources product testing.
Whenever possible, Phase I of the investigation should be done before test preparations are discarded to test hypotheses regarding potential laboratory error. It is highly recommended, and common practice, to segregate analytical and microbiological laboratory OOS investigations due to the strategic differences required during the investigations. Analytes are generally expected to be uniformly distributed throughout the material or product, while bacteria, fungi, and mold begin to grow in a particular region of the material or product and then spread.
Phase I: Documentary
Accuracy assessment
The accuracy assessment is an immediate non-experimental re-examination of the actual solutions, test units, glassware, methodology, and instrumentation used in the original measurements and preparations. It is conducted by the analyst and supervisor to determine if a laboratory error is the cause of the OOS, OOT, or unusual result. Any errors found invalidate the original result, and subsequent non-conforming material or product investigation or dietary supplement material review are not needed.
Historical review
A historical review is performed when a definitive laboratory error is not found during the accuracy assessment. Although not discussed in the FDA guidance document, a historical review of previous testing of the material or product is highly beneficial. A historical review not only includes trending of previous test results and investigations but also an examination of applicable material or product certificates of analysis (COAs) and specifications.
Phase I: Experimental examination
Re-analysis
An experimental examination by re-analysis is performed to assess whether there has been a transient equipment malfunction resulting in an equipment error. This is done by re-injecting or re-introducing the final sample preparation or diluent through the input system. The re-injections should always be performed in replicates of ≥ 3 to obtain a mean and standard deviation and have a meaningful data set to compare to the original result. Any equipment errors found invalidate the original result and, as such, the need for subsequent non-conforming material or product investigation or dietary supplement material review.
Only re-analyze samples using the same standards, reagents, and other chemical preparations that were used in the original chemical analysis. Do not re-analyze sample preparations that are known to be unstable or that have expired, and do not use expired standards, reagents, or other preparations for re-analysis. Insufficient sample preparation amounts must not be diluted and re-analyzed with standard or method modifications because doing so introduces additional variables, making a true assessment of just the equipment impossible.
Re-test
Re-testing of the original analyte sample may be required if no meaningful errors have been found during the documentary reviews or equipment assessment. An experimental examination by re-test is conducted by re-testing the original sample. Ideally, this is accomplished with the same first analyst and a second analyst to provide sufficient data to do an abbreviated accuracy, precision, and ruggedness assessment of the test method performance. No change in the testing methodology is made. Replicates of ≥ 3 are needed from each analyst to obtain a mean and standard deviation for both data sets. The second analyst must be appropriately trained to conduct the test and is preferably more experienced than the original analyst.
All of the data obtained is compared to the original OOS result and to each other to determine if the original result is valid. Consideration must always be given to the expected method variability. Pre-assigned criteria for relative percent standard deviation and percent difference assessment requirements can also be included based on test method validation or verification experiments.
According to the U.S. v Barr Laboratories court decision, retesting “cannot continue ad infinitum” and materials and products cannot be “tested into compliance”6. This means that OOS results cannot be averaged with passing results to obtain a “reported” average value that meets the material or product specification. Furthermore, testing cannot be repeated until a desired result is obtained. It’s critical that the number of tests to be performed is defined and then reviewed and approved by quality unit management prior to the performance of any testing. Best practice is to include these instructions in a laboratory OOS investigation procedure.
Confirmatory testing
Confirmatory testing is conducted when an assignable cause has not been found during the accuracy assessment or historical review of a microbiological OOS investigation. Confirmatory testing only applies to quantitative testing of bacteria, mold, or fungi. No confirmatory tests are performed for absence/ presence microbiological methods of pathogens such as E. Coli, Salmonella, or S. Aureus due to the confirmatory nature of the test methods themselves, potential ramifications of the presence of the bacteria, and the non-homogeneous nature of the microbiological contaminants. A non-conforming material or product investigation, or material review for dietary supplements, is always required in this case. Confirmatory testing is typically performed by re-testing the sample using a larger 25-gram sample size, rather than a 10-gram sample, to either validate or invalidate the original result. Preferably, quantitative confirmation testing should be performed by a different microbiologist. Replicate retests of ≥ 3 additional sub-samples from the original sample should be conducted to examine hypotheses about the cause of the OOS result regarding potential cross-contamination or a transient equipment malfunction for automated microbiological systems.
Phase II: Expanded experimental examination
An expanded experimental examination of the material or product laboratory OOS investigation may be undertaken if Phase I finds no meaningful errors and does not determine a cause for the OOS, OOT, or unusual result. This may include collecting additional samples for testing; sample testing not dictated on the specification, such as different test methodology; testing that encompasses method variations; or the inclusion of additional quality control samples.
A resampling of the material using the same procedure as the initial sampling process (same location and technique) can also help rule out a possible error or contamination concern during the original sampling. Alternatively, an enhanced representative and statistical sampling of the material or product lot such as random sampling of (√n + 1) containers may be beneficial. Spatial sampling of large containers (such as super sacks or bulk bags) to incorporate samples from the top, middle, and bottom or sides and middle using a thief sample device may be needed.
As with the Phase I re-test, all actions or testing to be undertaken for the expanded experimental examination must be defined, reviewed, and approved by quality unit management prior to execution.
Documentation
In a cGMP environment, “if it’s not documented, it didn’t happen,” so each laboratory OOS investigation must be assigned a unique alphanumeric identification and every aspect of the investigation must be documented on an applicable form. OOS investigations must also be reviewed and approved by the quality unit management who will compile and track all OOS investigations for tracking, monitoring, and trending purposes.
Tara Lin Couch, PhD, is the senior director of dietary supplements and tobacco services at EAS Consulting Group, Alexandria, VA (571 447 5500, www.easconsulting.com).
References
1. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/ cfcfr/cfrsearch.cfm?cfrpart=111
2. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/ cfcfr/CFRSearch.cfm?CFRPart=210
3. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/ cfcfr/CFRSearch.cfm?CFRPart=211
4. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/ cfcfr/CFRSearch.cfm?fr=111.113
5. https://www.fda.gov/media/71001/download
6. https://law.justia.com/cases/federal/district-courts/ FSupp/812/458/1762275/