The importance of hormone receptor status in assigning treatment and the potential use of human
epidermal growth factor receptor 2 (HER2)–targeted therapy have made it beneficial for laboratories to
improve detection techniques. Because interlaboratory variability in immunohistochemistry (IHC) tests may
also affect studies of breast cancer subtypes in different countries, we undertook a Web-based quality
improvement training and a comparative study of accuracy of immunohistochemical tests of breast cancer
biomarkers between a well-established laboratory in the United States (University of Chicago) and afield
laboratory in Ibadan, Nigeria. Two hundred and thirty-two breast tumor blocks were evaluated for estrogen
receptors (ERs), progesterone receptors (PRs), and HER2 status at both laboratories using tissue microarray
technique. Initially, concordance analysis revealed κscores of 0.42 (moderate agreement) for ER, 0.41
(moderate agreement) for PR, and 0.39 (fair agreement) for HER2 between the 2 laboratories. Antigen
retrieval techniques and scoring methods were identified as important reasons for discrepancy. Web-based
conferences using Web conferencing tools such as Skype and WebEx were then held periodically to discuss
IHC staining protocols and standard scoring systems and to resolve discrepant cases. After quality assurance
and training, the agreement improved to 0.64 (substantial agreement) for ER, 0.60 (moderate agreement)
for PR, and 0.75 (substantial agreement) for HER2. We found Web-based conferences and digital
microscopy useful and cost-effective tools for quality assurance of IHC, consultation, and collaboration
between distant laboratories. Quality improvement exercises in testing of tumor biomarkers will reduce
misclassification in epidemiologic studies of breast cancer subtypes and provide much needed capacity
building in resource-poor countries.
© 2013 The Authors. Published by Elsevier Inc. All rights reserved