Audit and internal quality control in immunohistochemistry
- 1Quantitative Biomarkers Group, Cancer Research Centre, The Queens University of Belfast and Institute of Pathology, Royal Group of Hospitals Trust, Belfast BT12 6BA, Northern Ireland, UK
- Dr Maxwell p.maxwell{at}qub.ac.uk
- Accepted 7 March 2000
Abstract
Aims—Although positive and negative controls are performed and checked in surgical pathology cases undergoing immunohistochemistry, internal quality control procedures for immunohistochemistry are not well described. This study, comprising a retrospective audit, aims to describe a method of internal quality control for immunohistochemistry. A scoring system that allows comparison between cases is described.
Methods—Two positive tissue controls for each month over a three year period (1996–1998) of the 10 antibodies used most frequently were evaluated. All test cases undergoing immunohistochemistry in the months of April in this three year period were also studied. When the test case was completely negative for a given antibody, the corresponding positive tissue control from that day was examined. A marking system was devised whereby each immunohistochemical slide was assessed out of a possible score of 8 to take account of staining intensity, uniformity, specificity, background, and counterstaining. Using this scoring system, cases were classified as showing optimal (7–8), borderline (5–6), or unacceptable (0–4) staining.
Results—Most positive tissue controls showed either optimal or borderline staining with the exception of neurone specific enolase (NSE), where most slides were unacceptable or borderline as a result of a combination of low intensity, poor specificity, and excessive background staining. All test cases showed either optimal or borderline staining with the exception of a single case stained for NSE, which was unacceptable.
Conclusions—This retrospective audit shows that immunohistochemically stained slides can be assessed using this scoring system. With most antibodies, acceptable staining was achieved in most cases. However, there were problems with staining for NSE, which needs to be reviewed. Laboratories should use a system such as this to evaluate which antibodies regularly result in poor staining so that they can be excluded from panels. Routine evaluation of immunohistochemical staining should become part of everyday internal quality control procedures.
