Many areas of pathology are by their nature subjective, and IHC is no exception. What is considered positive will depend, to varying degrees, on the intensity of staining, the distribution of stain within the cell, and/or the percentage of cells showing staining. Ideally, the interpretation will be based on criteria set forth in the literature, but this is not always the case, especially because the literature frequently gives divergent data. The variation in published reports can be the result of differences in methodology, the specific antibody used, and, of course, the pathologist interpreting the slides (28). A recent study of estrogen receptor staining in tissue microarrays of breast carcinomas documented significant interlaboratory variability in reporting of results. The variability was least in cases that were strongly positive, but increased dramatically in intermediate and negative cases. The study excluded fixation procedures and interobserver variability as major contributors. Variations in AR methods were cited as a main potential cause for disagreement (29). Even with these attempts to standardize procedures, relatively little concerted effort has been made to standardize interpretation of results.
When evaluating IHC studies, simply knowing the cellular and subcellular localization of the protein being assayed can help eliminate some erroneous interpretations (8). For example, cyclin D1 is a nuclear protein and one would not expect to find it in the cytoplasm (Fig. 1A). Similarly, the presence of HER-2/neu (HER-2) staining in the cytoplasm or nucleus should be regarded as nonspecific because its expression is restricted to the cell membrane (Fig. 1B). Comparison with appropriate internal positive and negative controls is essential and helps confirm that the antibody is marking the right cellular location. In cases where ambiguous results are obtained, validation can be attempted with antibodies against different epitopes of the same molecule or by detection of a related marker.
In addition to the lack of agreement on what constitutes "positive" at a cellular level, there is also the issue of what constitutes positive on a tissue level. In the breast, HER-2 protein overexpression and ER are prime examples. For HER-2 staining, the definition of "positive" rests on strong, uniform staining of the cell membrane, which has been proven by molecular studies to correlate well with amplification of the HER-2 gene. ER studies, on the other hand, have traditionally defined positive based on the percentage of cells that show nuclear reactivity. A cutoff of 10% cells with positive nuclear staining was generally accepted for reporting a positive result, although some argued for higher or lower percentages or for intensity of staining to be used. In 2000, the National Institutes of Health (NIH) issued a statement that any degree of positivity should be considered a positive result and a reason to offer antiestrogen therapy to breast cancer patients.
Finally, a number of artifacts can result in false-positive and false-negative results. In some instances, variable staining over a single slide begs the question of which area to believe. In such cases, either repeating the assay or finding an area in which the internal controls are appropriate might help resolve the issue. The absence of any staining of the tissue, in the presence of internal controls that should be positive, calls into question the immunoreactivity of the tissue or raises the possibility that some step in the process went awry. Testing with a ubiquitous marker such as vimentin can be of value in determining if the tissue has been rendered nonreactive because of processing, and a repeat assay will (hopefully!) solve the issue if the reagents were not appropriately applied.
4.1. CONTROLS Appropriate tissue controls known to react with the antibody being evaluated must be prepared and evaluated at the same time as the test sample. The presence of appropriately reactive positive and nonreactive negative internal controls is probably the best way to determine that an antibody was properly prepared and applied. However, some antigens are not expressed in normal tissues, and internal controls are absent. In these cases, it is necessary to include an external control tissue that is known to express the antigen being evaluated. Standardization of fixation is essential because differences in fixation between the control and test tissue might make a difference in the outcome of the staining procedure. For this reason, use of tissue processed "in house" as the control is most appropriate (8). Because control and test slides are run at the same time, the issues relating to temperature, time, and pH of AR will be virtually identical. One issue that arises is whether the control should be placed on the same or a different slide. Ideally, having them on the same slide as the test material would afford more standardization, but this technique is more time-consuming and technically demanding. The choice of control material also has an impact on the interpretation of results. The control tissue should have the antigen of interest present at a detectable but not overpowering level. The use of low-level positive controls allows one to detect the antigen as well as detect slight alterations in primary antibody sensitivity. Surgical specimens should be used to prepare control slides to be used with routinely processed test specimens, whereas cytospin slides or cell block materials are more appropriate as controls for cytology specimens. In addition to positive controls, negative controls are necessary to evaluate nonspecific background staining or the presence of internal pigments that might interfere with interpretation. This is commonly accomplished by following the regular protocol, but with substitution of antibody diluent for the antibody of interest. Finally, in cases where there is little or no staining with the
antibodies of interest, an antibody to a ubiquitous antigen such as vimentin might be used as a "reporter molecule" to ensure that the tissue was fixed and processed well enough to preserve immunoreactivity. For the present, controls are inherently imperfect because the amount of time between tissue removal and processing, fixation, and intensity can all vary from case to case and even within a given case. Suggestions to use specially cultured cells with more rigid control over these variables might be an improvement, but it will not address all concerns (28).
Was this article helpful?
Prior to planning pregnancy, you should learn more about the things involved in getting pregnant. It involves carrying a baby inside you for nine months, caring for a child for a number of years, and many more. Consider these things, so that you can properly assess if you are ready for pregnancy. Get all these very important tips about pregnancy that you need to know.