Standardizing Estrogen Receptor Measurement
Method Proposed to Narrow Rate of Misclassifications
By Genna Rollins
Immunohistochemistry (IHC) is a common method to assess estrogen receptor (ER) status in breast cancer, but has well-known accuracy issues. Researchers at Yale University recently described a possible method for standardizing ER measurement and reducing the number of specimens misclassified as ER-negative. Their findings are the subject of this issue of Strategies.
Although IHC is widely used to assess ER status in breast cancer, up to 20% of determinations are misclassified. There are several reasons for this, including that pathologists score the amount of ER qualitatively, which introduces the possibility of variability and lack of reproducibility. In addition, different labs use different antibodies, reagents, and protocols to prepare ER slides for interpretation. The challenges with IHC measurement of ER came into sharp focus in 2008 when a 40% misclassification rate was reported between local and central labs in Canada.
All of this has important consequences for patients, as false negatives render the affected individual ineligible for adjuvant endocrine therapy, which has been associated with clinical benefit in women with positive ER status. To improve the accuracy of IHC ER testing in breast cancer, an expert panel of the American Society for Clinical Oncology (ASCO)/College of American Pathologists (CAP) recently recommended lowering the standard of ER positivity from 10% to 1% positive nuclei. Since there is no gold standard IHC assay for ER, the committee also advised that labs choose assays that have both preanalytical and analytical components that conform exactly to assays whose results have been validated against clinical benefit in endocrine therapy.
Concerned about improving the accuracy of ER testing, researchers at Yale University have recently described a potential method for standardizing ER measurement and have compared it to IHC analysis (J Clin Oncol 2011; 29: 2978-84). About 10 years ago, the investigators developed a quantitative immunofluorescence (QIF) method that uses a fluorescence-based algorithm known as automated quantitative analysis (AQUA) to calculate marker expression on a continuous scale based on pixel intensity. Previous research showed that this method correlated well with IHC analysis and predicted response to tamoxifin. The researchers did, however, notice some discrepancies.
“We were looking at the reproducibility of this method in a number of different cohorts, and we noticed that there was a good general trend in agreement with IHC. But we also noticed that there were cases—usually with low AQUA scores—that had been called negative by the pathologist. Since these were right around the threshold, it begged the question of what the threshold should be,” explained senior author, David Rimm, MD, PhD, professor of pathology and director of pathology tissue services at Yale. Rimm also is founder of and a shareholder in HistoRx, which is the exclusive licensee of the Yale-owned AQUA patent.
Rimm’s lab first sought to define an ER cutpoint with biological and clinical relevance for the QIF. They did so by creating an index tissue microarray that would serve as a standard curve for ER expression and included both a panel of cell lines that had been prepared as patient tissue and 40 patient controls. This process enabled them to establish the cutpoint between the highest ER-negative and lowest ER-positive cell lines as 2 pg/µg. The team then went on to evaluate how this standardized threshold using QIF compared with IHC in two archived breast cancer specimen cohorts, one with 252 samples, and the other with 234. Overall 9.1% of the former and 19.7% of the latter were ER-positive by QIF but ER-negative based on IHC.
To assess the impact of the ASCO/CAP expert panel’s recommended change in ER positivity from 10% to 1% positive nuclei, the researchers conducted a retrospective analysis of the Yale Copath database for the preceding 10 years when the 10% standard was used. They compared this to ER-positive results since the lab implemented the 1% standard, and found “almost no difference” between the percentage of cases called ER-positive.
Based upon all of these findings, the researchers suggested that one of the key underlying reasons behind discordant IHC ER results is the intensity of positive staining. “The take home message of the paper is, it’s not the percentage of positive nuclei; it’s the intensity of positive staining, and more importantly, the lack of standardization that’s the problem,” said Rimm. “There’s going to be a small percent of cases where the specimen will be positive for the protein, but the blue hematoxylin stain and the relatively low level of expression of ER combine to make it invisible to pathologists using conventional methods.”
Join us for a discussion of this article, and interact with your peers, on LinkedIn