Immunohistochemistry: Growing Pains, From a Stain to an Assay.

Abstract

T primary focus of AIMM (Applied Immunohistochemistry and Molecular Morphology) is intrinsic to its title, namely the use of immunohistochemical (IHC) and molecular methods for diagnostic purposes in a morphologic context. Accepted manuscripts fall generally into 2 classes: those that describe the detection of one or more analytes (biomarkers) in tissues (normal and diseased) as applied to diagnosis in anatomic pathology (AP), and those that focus on the underlying methods, rather than the potential value of individual biomarkers. In recent years the policy of AIMM has been to seek out and publish manuscripts from investigators whose work relates to improving the quality, reproducibility, and utility of the methods. The current issue of the Journal contains, by specific invitation, a thoughtful and thought-provoking review by Steven Bogen, bearing the title “A Root Cause Analysis Into The High Error Rate In Clinical Immunohistochemistry.”1 Dr Bogen is a clinical pathologist and the paper borrows from his extensive experience in the clinical laboratory, where error rates are very low. Dr Bogen delivers a novel perspective and analysis of many of the reasons for the observed discrepant error rates in IHC and offers some possible solutions, including novel standard controls. There is also an accompanying invited commentary by Mogens Vyberg, who is the founder of NordiQC, one of the foremost and most comprehensive AP external QC (Quality Control) programs. Dr Vyberg has penned a thoughtful response from the AP perspective.2 Dr Vyberg questions whether the comparison of IHC in the AP laboratory to high performance clinical laboratory assays is “fair.” It is a very good question, one that goes directly to the core of how IHC is used in the AP laboratory, and how its use increasingly must be considered in a modern “fit for purpose” context. Why are we doing this particular IHC test? In short, the question posed by Drs Bogen and Vyberg is “Can we convert an IHC stain to an assay, that purports actually to measure something?” The question is a good one because in seeking an answer, there is much to be learned, not least of which is that having learned, the learning is of no value unless applied in daily practice in AP. In addition, both process and outcome must be closely monitored, issues raised by both Drs Bogen and Vyberg. As an example, AIMM in the past several years has published a “control series” of papers3,4 with recommendations, that if universally adopted, would certainly lead to major overall improvements within and across AP laboratories. But adoption is slow. Then there is another problem, of even greater in import. Not only is there much to be learned, but also there is much to be “unlearned”: the AP laboratory is handicapped by its legacy; it performs stains, not assays, and we in AP have acquired many bad habits. The first use of IHC (immunoperoxidase) method performed on formalin-fixed paraffin embedded tissues for the express purpose of diagnosis was naturally in an AP laboratory, the department of Morbid Anatomy at Oxford, under the directorship of A.H.T. Robb-Smith. At the time, the purpose was to produce a stain that imparted differential color to tissue components, somewhat akin to a PAS or a trichrome stain as an aid to diagnosis. It happened that there was major clinical laboratory input into this new stain, for antibodies were required (these were polyclonal antisera), and these