The knowns and the unknowns in HER2 testing in breast cancer.

The 2007 American Society of Clinical Oncology (ASCO)/ College of American Pathologists (CAP) guidelines for HER2 immunohistochemical and fluorescence in situ hybridization (FISH) testing in breast cancer address the potential impact of several " knowns " (eg, effect of fixative composition and fixation time). 1,2 However, data continue to be generated that shed further light on the impact of " unknowns, " ie, selected preanalytic, analytic, and postanalytic (ie, interpretive) factors, on HER2 immunohistochemical testing. The study by Manion and colleagues 3 in the June 2011 issue of the Journal focuses on 1 analytic factor that was not addressed by the ASCO/CAP guidelines, namely, the impact of the choice of antibody clone used for HER2 immunohis-tochemical testing. Manion et al 3 reported that by switching from the rabbit polyclonal antibody A0485 (the same antibody that is part of the HeceptTest kit, DAKO, Carpinteria, CA) to the rabbit monoclonal antibody SP3, they could essentially halve the number of 2+ cases, at the price of a probably statistically insignificant increase in the false-negative rate (ie, FISH-positive, immunohistochemically negative cases). Can their results be extrapolated to other laboratories wanting to optimize HER2 immunohistochemical testing? As noted by the authors themselves, the apparent differences in performance of the A0485 rabbit polyclonal vs the SP3 rabbit monoclonal noted in their study have not necessarily been corroborated by other investigators. 4-7 And in a study from our laboratory published in abstract form, 8 involving a series of 416 breast cancer cases, we also found that the use of SP3 reduced the number of 2+ cases, albeit not by the magnitude reported by Manion et al. Because all of our cases had also been subjected to FISH analysis, we were able to determine that concordance rates between immunohistochemical and FISH testing were independent of the use of the SP3 or the A0485 antibody. However, in subsequent but unpublished SP3 validation studies performed in our laboratory on 551 breast cancers, we found nearly identical proportions of cases showing 2+ immu-nostaining with the A0485 or the SP3 antibody. Among the cases showing amplification by FISH, there was a significant number of cases that had been classified as 2+ using A0485 that were scored as 3+ using SP3. But among the cases that were not amplified by FISH, there was a significant number of cases that had been classified as 0/1+ using A0485 but as 2+ using SP3. These 2 …