Search

Abstract: This article presents an overview of statistical methods used in validation and proficiency testing for qualitative testing in analytical chemistry, paying particular attention to interlaboratory study. A definition of “qualitative analysis” that allows a general treatment of a wide variety of different examples is used. We discuss the characterisation of qualitative method performance, and consider in detail methods for the analysis of false response rates. We review appropriate statistical methods for a variety of circumstances, including the application of logistic regression to interlaboratory and intralaboratory studies of method performance. Finally, we propose provisional recommendations for the design and statistical treatment of validation and proficiency studies. [Copyright &y& Elsevier]

Background: Human epidermal growth factor receptor 2 (HER2) over-expression in breast cancer is associated with aggressive tumor behavior and predicts response to targeted therapy. Accurate HER2 result is paramount for optimal patient management. However, routine HER2 immunohistochemistry (IHC) testing are subjected to intra- and inter-laboratory variability., Objective: This study aims to determine inter-laboratory variation in HER2 IHC testing through a slide-exchange program between five main reference laboratories., Method: A total of 20 breast carcinoma cases with different known HER2 expression and gene status were selected by the central laboratory in five testing rounds. Three unstained tissue sections from each case were sent to participating laboratories, which immunostained and interpreted the HER2 immunohistochemistry result. One of the stained slides was sent to one designated participating laboratory for evaluation. Results were analyzed by the central laboratory., Results: A complete concordance was achieved in six IHC-positive and six IHC-negative cases, its gene status of which was confirmed by in-situ-hybridization (ISH) study. The discordant results were observed in six equivocal cases, one negative case and one positive case with a concordance rate of 50-88.3%. Interestingly, the negative discordant case actually displays tumor heterogeneity. Good inter-observer agreement was achieved for all participating laboratories (k = 0.713-1.0)., Conclusion: Standardization of HER2 testing method is important to achieve optimum inter-laboratory concordance. Discordant results were seen mainly in equivocal cases. Intra-tumoral heterogeneity may impact the final HER2 IHC scoring. The continuous quality evaluation is therefore paramount to achieve reliable HER2 results.