Belgian Recommendations for Analytical Verification and Validation of Immunohistochemical Tests in Laboratories of Anatomic Pathology

Abstract

Analytical verification and validation of immunohistochemical (IHC) tests and their equipment are common practices for today's anatomic pathology laboratories. Few references or guidelines are available on how this should be performed. The study of Sciensano (the Belgian national competent authority regarding licensing of medical laboratories) performed in 2016, demonstrated a significant interlaboratory variation in validation procedures of IHC tests among Belgian laboratories. These results suggest the unavailability of practical information on the approach to the verification and validation of these tests. The existing Belgian Practice Guideline for the implementation of a quality management system in anatomic pathology laboratories has been reviewed to meet this demand and, in addition, to prepare the laboratories for the EU-IVD revised regulations (IVDR). This paper describes Belgian recommendations for the verification and validation of IHC tests before implementation, for ongoing validation, and for revalidation. For each type of test (according to the IVDR classification and the origin) and its intended use (purpose), it addresses how to perform analytical verification/validation by recommending: (1) the number of cases in the validation set, (2) the performance characteristics to be evaluated, (3) the objective acceptance criteria, (4) the evaluation method for the obtained results, and (5) how and when to revalidate. A literature study and a risk analysis taking into account the majority of variables regarding verification/validation of methods have been performed, resulting in an expert consensus recommendation that is a compromise among achievability, affordability, and patient safety. This new consensus recommendation has been incorporated in the aforementioned ISO 15189:2012-based Practice Guideline.

FIGURE 1

Verification of CE-IVD IHC tests according to IFU or validation of modified CE-IVD IHC tests with reference. (1) Type 1 (diagnostic): laboratories should retrospectively and objectively demonstrate their experience for at least 8 of the 15 Ab biomarkers within the verification panel [2 Abs with cytoplasmic staining pattern (C), 2 Abs with nuclear staining pattern (N), 2 Abs with membranous staining pattern (M), and 2 Abs with a more complex staining protocol]. Criteria to demonstrate experience with the type 1 IHC method are, for example, at least good or optimal EQA results for the most recent EQA program within the past 7 years without any change in staining pattern (Ab-antigen interaction) and interpretation of the results (eg, scoring algorithm), traceable IQC results over a certain period (at least 6 mo) or no registration of nonconformities with regard to the method/technique used within the past year. Type 2 (prognostic/pharmaco-predictive): laboratories should demonstrate experience with the test through at least good or optimal EQA results for at least 2 consecutive EQA programs and through concordant inter/intraobserver results. (2) Method verification: a panel of 15 Ab biomarkers is selected for each method (same detection system and staining platform). This panel of 15 Abs, with different pretreatment buffers, consists of 3 Abs with a membranous staining pattern (M), 3 Abs with a cytoplasmic staining pattern (C), 3 Abs with a nuclear staining pattern (N), and 6 Abs with a technical more complex or challenging staining protocol (eg, concentrated Ab). The number of type 1 Abs can be reduced when type 2 Abs using the same detection method and platform have been verified according to the specific type 2 verification requirements (eg, using 3 type 2 and 12 type 1 Abs instead of 15 type 1 Abs). The verification of the accuracy of the method with 5⁺/5⁻ or 10⁺/10⁻ tissues (negative internal control cells can be present in the tissue sections with antigen-positive cells) is applicable for each Ab 1 to Ab 15. The repeatability/reproducibility is demonstrated by staining at least 3t in triplicate (3s). The repeatability can be demonstrated by staining 3t on at least 3 different positions in the staining platform. Depending on the staining platform, a minimum of 3 slides (3 specimens on 1 slide) and a maximum of 9 slides are distributed within one run. The reproducibility among different runs can be demonstrated by 2 additional stains in 2 other, additional runs. From the 16th Ab marker onwards, a verification of the accuracy with 2⁺/2⁻ tissues is recommended after optimization. (3) For the definitions of the performance characteristics, see the section “Definitions” in the main text. (4) The obtained results should be evaluated and compared with the reference or comparator to check whether the predetermined acceptance criteria are met. These criteria are based upon expert consensus results as published in (inter)national publications and differ according to the evaluated performance characteristics and the sample size of the validation set (Tables 2 and 3). Examples are: comparison with (1) the expected results of the reference staining, (2) results of a validated IHC test in the same or in another laboratory, (3) results of the EQA provider, (4) comparator method (eg, ISH test and PCR) and comparison between the results obtained from the within run, between run and readout precision examinations. (5) For a new type 1Ab that does not use the same detection method/staining platform/interpretation, the laboratory checks whether it has experience or not. Subsequently, for each Ab, after optimization, a verification is performed on 5⁺/5 or 10⁺/10 tissues to demonstrate the accuracy and 3t×3s to demonstrate the repeatability/reproducibility. Ab indicates antibody; EQA, external quality assessment; IFU, Instructions For Use; IHC, immunohistochemical; IQ, installation qualification; IQC, internal quality control; ISH, in situ hybridization; OQ, operational qualification; PCR, polymerase chain reaction; PQ, performance qualification; Repeat, repeatability; Reprod, reproducibility; 3s, 3 sections; 3t, 3 tissues.

FIGURE 2

Validation of modified CE-IVD IHC tests without reference. (1) Method validation: a panel of 15 Ab biomarkers is selected for each method (same detection system and staining platform). This panel of 15 Abs, with different pretreatment buffers, consists of 3 Abs with a membranous staining pattern (M), 3 Abs with a cytoplasmic staining pattern (C), 3 Abs with a nuclear staining pattern (N), and 6 Abs with a technical more complex or challenging staining protocol (eg, concentrated Ab). The validation of the method with 15⁺/15⁻ tissues (negative internal control cells can be present in the tissue sections with antigen-positive cells) is applicable for each Ab 1 to Ab 15. From the 16th Ab marker onwards, a verification with 3 to 5 positive and 3 to 5 negative tissues (depending on the sample availability) is recommended after optimization. (2) For the definitions of the performance characteristics, see the section “Definitions” in the main text. (3) The obtained results should be evaluated and compared with the reference or comparator to check whether the predetermined acceptance criteria are met. These criteria are based upon expert consensus results as published in (inter)national publications and differ according to the evaluated performance characteristics and the sample size of the validation set (Tables 2 and 3). Examples are: comparison with (1) the expected results of the reference staining, (2) results of a validated IHC test in the same or in another laboratory, (3) results of the EQA provider, (4) comparator method (eg, ISH test, PCR) and comparison between the results obtained from the within run, between run and inter/intraobserver examinations. (4) For a new type 1Ab that does not use the same detection method/staining platform/interpretation, after optimization, a verification is performed on 15⁺/15⁻ tissues to demonstrate the accuracy and 3t×3s to demonstrate the repeatability/reproducibility. Ab indicates antibody; EQA, external quality assessment; IHC, immunohistochemical; IQ, installation qualification; ISH, in situ hybridization; OQ, operational qualification; PCR, polymerase chain reaction; PQ, performance qualification; Repeat, repeatability; Reprod, reproducibility; 3s, 3 sections; 3t, 3 tissues.

FIGURE 3

Validation of non–CE-IVD IHC tests. (1) For non–CE-IVD type 1 tests without reference and for which limited (control) materials are available, but used within the context of a verified/validated CE-IVD detection method, we recommend starting the initial validation with a set of 2 positive and 2 negative tissues and progressively increase the validation set over time until the target of 40 positive and 40 negative tissues is reached. (2) For the definitions of the performance characteristics, see the section “Definitions” in the main text. (3) The obtained results should be evaluated and compared with the reference or comparator to check whether the predetermined acceptance criteria are met. These criteria are based upon expert consensus results as published in (inter)national publications and differ according to the evaluated performance characteristics and the sample size of the validation set (Tables 2 and 3). Examples are: comparison with (1) the expected results of the reference staining, (2) results of a validated IHC test in the same or in another laboratory, (3) results of the EQA provider, (4) comparator method (eg, ISH test and PCR) and comparison between the results obtained from the within run, between run and inter/intraobserver examinations. Ab indicates antibody; EQA, external quality assessment; IHC, immunohistochemical; IQ, installation qualification; ISH, in situ hybridization; OQ, operational qualification; PCR, polymerase chain reaction; PQ, performance qualification; Repeat, repeatability; Reprod, reproducibility; RUO, Research Use Only; 3s, 3 sections; 3t, 3 tissues.