Validation of immunohistochemical tests performed on cytology cell block material: Practical application of the College of American Pathologists’ guidelines

The advent of fiberoptic endoscopy with biopsy has revolutionized procurement of specimens from deep sites. This has translated into more cytologic specimens whereby the material is limited and best handled by cytology laboratory staff. While the diagnosis of the pathologic process is of utmost importance, there is increasing expectation that the diagnosis be specific and accurate as not to require additional biopsy for initiation of treatment. This expectation has driven demand in immunohistochemical (IHC) and molecular studies conducted specifically on material processed as cytology specimens. The Clinical Laboratory Improvement Amendments of 1988 requires laboratories in the United States of America to verify the performance of patient tests. Due to varying laboratory practices with respect to validation of IHC assays, the College of American Pathologists introduced guidelines for analytic validation of IHC tests. These guidelines address how to perform validation by recommending the number of cases in the validation set, comparator concordance, and when to revalidate. The main thrust of the guidelines is based on formalin-fixed paraffin-embedded tissue with only one expert consensus opinion referring to validation of IHC tests on cytology specimens which delegates to the medical director, the determination of number of positive and negative cases to be tested. This article will outline how an academic center approaches validation of IHC studies performed on cytology cell block specimens using the College of American Pathologists guidelines. A stepwise approach from selection of antibodies to validate followed by building the validation panel and evaluating the stain results for concordance against the gold standard of histology tissue specimen will be described. A rationale for dealing with discordant results and future innovations will conclude the report.