by Alicia Algeciras-Schimnich, PhD, DABCC
Verifying new reagent lot performance is a common task in the clinical laboratory. It is not only considered good laboratory practice, but also laboratory regulations and accreditation standards require the evaluation of each new reagent lot prior to use. Each new reagent lot has the potential to affect quality control (QC) material and/or patient sample performance. In the clinical laboratory, immunoassays have been reported to be more prone to lot-to-lot variability than general chemistry tests. Multiple factors can affect performance of a new reagent lot, including changes in a critical reagent material or in stability of the reagents, reagent damage during transportation or storage, or incorrect calibration.
Assuring lot-to-lot consistency is particularly critical when an analyte is used for long-term follow-up of patients, as in the case with tumor markers, when small changes in concentration might trigger further laboratory testing, imaging, or other clinical interventions. Reagent manufacturers have procedures in place to qualify the release of new reagent lots. The goal of the manufacturer should be to achieve correct recovery of the analyte, meaning that the assay is able to measure the analyte correctly based on a known expected concentration. Manufacturers also seek to minimize lot-to-lot variation when recovering patient samples. Unfortunately, manufacturers’ processes to ensure lot-to-lot consistency vary greatly and their accessibility to patient samples sometimes is limited. Consequently, they might miss changes that appear once a laboratory tests enough patient samples.
Clinical laboratory practices for lot-to-lot evaluation also vary widely, ranging from testing as few as three-to-four samples to as many as 20to 40 samples with each new reagent lot. Regarding the choice of samples that are tested, current approaches include the use of QC material supplied by the reagent vendors, third party QC material, in-house QC material, and patient samples. Caution should be taken if only QC material is used for lot verification. Matrix-related differences between patient samples and QC material could affect how well QC results reflect new reagent lot performance in actual patient samples.
When patient sample comparisons are used, the type of samples tested again varies significantly between laboratories. Some labs might test samples with low, mid and high analyte concentrations; others might test randomly selected samples, while still others might test samples with values that span the analytical measurement range. There are no universally agreed upon acceptance or rejection criteria for new reagent lots. It is up to the laboratory management team to determine what is acceptable. The decision should be based on the historical performance of the assay, the assay impression, and the estimated allowable overall bias for the respective analyte, given its biological variability and clinical use.
To date, there has not been a standardized protocol or guideline to help laboratories deal with lot-to-lot verification. Recognizing the need for one, the Clinical and Laboratory Standards Institute (CLSI) recently published the document EP26-A—User Evaluation of Between-Reagent Lot Variation. This guideline offers a practical protocol that takes into consideration the resource constraints of the clinical laboratory and uses as few patient samples as possible. In practice, the protocol is broken down into two phases.
The first phase –which could be time consuming, but once established, does not need to be repeated– involves gathering data to establish a number of parameters. These include: the maximum difference between the two reagents lots that would be acceptable without having an adverse clinical impact (critical difference); the laboratory-observed method imprecision; and the desired statistical power for detecting significant lot-to-lot changes. This information is then used to determine the number of samples to be tested and the rejection limit necessary to assure the critical difference is detected.
The second phase involves verification of the new reagent lot by testing the determined number of patient samples with both lots of reagents, calculating the average concentration differences between the two lots, and analyzing acceptability of the new lot based on the rejection limit established during the first phase.
Lot-to-lot verification is, without a doubt, necessary to prevent use of suboptimal reagent lots, but it can be an additional burden to the laboratory, especially when it takes multiple attempts to obtain an acceptable reagent lot. Laboratories need to plan ahead and allow time for performing repeat new reagent lot evaluations if acceptability limits are not initially met. Accessibility to the necessary number of patient samples, availability of resources such as instrument time, remaining inventory of the current reagent lot, and technologist time also need to be taken into consideration. With EP26-A, laboratories will be able to adapt a standardized and practical process for lot-to-lot verification, in keeping with resource constraint limits of the current healthcare environment.
Alicia Algeciras-Schimnich, PhD, is an associate professor of Laboratory Medicine and Pathology at the Mayo Clinic in Rochester, Minnesota. She serves as director of Mayo’s Clinical Immunoassay Laboratory and associate medical director of operations for Mayo Medical Laboratories.