A Commentary: Quality Assurance in Immunohistochemistry.

Abstract

A and reliable biomarker testing is essential to provide personalized medicine for cancer patients. Suboptimal immunohistochemical (IHC) assays may result in suboptimal diagnoses and treatment decisions. The rapid development of new methodologies, the introduction of novel predictive markers, and increasing quality expectations, make it of the utmost importance for laboratories continuously to focus on providing optimal staining results and identifying methodological errors. In contrast to most other laboratory methods, IHC is a qualitative, or at best a semiquantitative, method where the interpretation of diagnostic assays in cancer specimens typically is performed in the context of clinical information, tumor morphology, etc. Unexpected staining reactions may indicate a need for reevaluation of the staining quality, readout, and diagnosis, before concluding that the staining reaction is “aberrant.” Often a panel of IHC assays is used in tumor classification, making it easier to discover aberrant staining patterns, but the cause of these aberrations, whether technical errors or biological phenomena, due to unusual tumor expression, may not always be identified, for example, by restaining, especially if the diagnostic issue has been resolved by results of the other assays in the panel. In contrast, predictive assays for critical biomarkers, like HER2 and PD-L1, are standalone tests, where unidentified errors or wrong readouts have direct negative consequences for the patients. In this issue of AIMM, Steven Bogen’s paper, “A root cause analysis into the high error rate in clinical immunohistochemistry,”1 points out that clinical IHC is beset with a higher error rate than other types of clinical laboratory testing. In part, this is due to the inherent limitations in IHC, including the absence of standards, traceable units of measure, and true quantitative monitoring. However, it can be questioned if a comparison between IHC and other, strictly quantitative laboratory testing procedures is fair. Nevertheless, Bogen’s article is relevant for pathology laboratories, which may rely too much on judging their assay quality solely on “routine” and internal quality control, with inconsistent use of external quality assurance (EQA) programs, which often do not challenge the laboratories with sufficient rigor, for example, by failing to utilize EQA test tissues with low level of epitope expression that provide a truer test of assay sensitivity. Some EQA schemes do not utilize a central assessment of staining results in the QA sample (by an expert pathology panel), but leave it to the laboratories to assess their own staining results, introducing a further uncontrolled and unknown variable, namely interpretation by separate pathologists. This approach effectively denies the opportunity for different laboratories to compare their technical results with others in order to discover suboptimal assay performance. EQA or laboratory proficiency testing is, or should be, a service to diagnostic laboratories by which they determine if their testing procedures compare well with their peers and/or a “designated true value.” Nordic Immunohistochemical Quality Control (NordiQC, http://www.nordiqc.org), with ~600 participating laboratories in 50 countries, is primarily aimed at assessing the technical performance of the analytical phase of the laboratory IHC, recognizing that preanalytic and postanalytic (interpretation) elements are not directly evaluated. More than 40,000 IHC slides have been evaluated, revealing that about 20% of the staining results in the breast cancer IHC module and about 30% in the general module