Immunohistochemistry has far surpassed it’s initial expectation as an invaluable tool in the correct recognition of tumours. It is now being increasingly sought after for prognostication of tumours and as a justification for initiaton of expensive targeted therapy in oncology practice. The wider application of IHC has increased the demands from a surgical pathologist who no longer restricts to giving a correct label but also actively participates in the subsequent clinical decision making process.

However many physicians have this notion that IHC is akin to a biochemical test. The IHC result is considered as representing enough “evidence” to neatly categorise disease entities and resolve messy diagnostic dilemmas. In fact clinicians often demand “IHC confirmation” of surgical pathology reports (even where IHC may not be needed and is unlikely to add any value) and ever so often pathologists who lack access to IHC facilities often end their reports stating “IHC confirmation necessary” (almost as a disclaimer) for every possible lesion.

This situation has risen partly due to the belief (more aptly misbelief ) that the “IHC test” can be translated into “a” particular diagnosis. The “IHC test” is looked upon as a tool to usher in the much needed objectivity in routine surgical pathology practice and this is accompanied with an underlying perception that IHC can make-up for lack of diagnostic skill and experience in the complex task of giving a correct histological diagnosis.

The moot point , therefore, is : “ How good is IHC evidence ?” This issue will be addressed under the following headings :

A. Generation of the IHC result
B. Search engines for information
C. Application of IHC result to resolve frequent diagnostic dilemmas
D. Clinical Decision support from IHC result
E. Data warehousing
F. Quality control in IHC

A. Generation of IHC result : Several factors influence the final IHC result and no two tissue specimens will react in the same way even though they could be representing the same anatomic site and could also be matched for stage and grade. The factors responsible for this variation can be Pre-analytical, Analytical, and Post-analytical.

The pre-analytical factors : Optimal preservation of the antigenic epitope is vital. IHC can be affected by the duration of anoxia at surgery, time gap between resection to fixation, the type of fixative, the duration in the fixative, the size of the tissue, the thickness, and whether freezing was done. Finally, the quality of reagents such as the company, the batch, and the shelf-life of antibodies can affect the kind of result obtained. Of these optimal fixation is of special interest because it is a critical yet manageable step.

The analytical factors : include the various techniques used for antigen retrieval namely heat, Microwaving, Pressure cooking, trypsin digestion, autoclaving with different buffers etc. Proper endogenous peroxidase blocking is vital to prevent background staining. Further, whether an autostainer is being used or the staining is done by hand will influence the end result. The biochemical process involved is important, for example the relatively recent “catalyzed signal amplification” method is considered more sensitive than avidin-biotin or extraavidin methods.

The postanalytical (interpretative) factors : Finally, a very much underrated factor is the actual interpretation of the IHC result by the surgical pathologist. Several large studies have addressed the issue of inter and intra-observer errors .The positive or the negative interpretation has several subtle nuances which only a busy practitioner of IHC will realise. Suffice it to say a combination of several observations such as intensity, quantity and localisation of the IHC reaction and visualization of the immunostain in the lesional cells - as opposed to the immuno reaction seen in normal tissues, reactive tissues and other ‘bystanders’ (often referred to as “background” staining) – are features vital to the the final interpretation. Nevertheless the lack of a prescribed threshold level for interpreting a reaction as positive leaves immense scope for inter and intra observer errors. Finally, whatever the interpretation – it is the integration of the information obtained from the IHC test in the histopathologic picture is what matters most in the final interpretion.

B. Search engines : There is burgeoning literature on new antibodies, and newer application of old antibodies. Sharing of epitopes and significant immunoreactivity as an epiphomena rather than a common trait (eg. bcl2, CD34, c-kit), is well-recognized.
Hence, it is necessary to refer to sites which give information of the possible range of reactivity and some of these are as follows :

http://immunoquery.com, http://immunohypermart.net,
http://www.ncbi.nlm.nih.gov/prow

as also the various journals on Immunohistochemistry.

C. Application of IHC to resolve some diagnostic dilemmas in surgical pathology :
Several vexatious diagnostic dilemmas are well-known in surgical pathology practice.
To mention some of the frequent queries include distinguishing mesothelioma vs adenocarcinoma, the subtype of the malignant round cell tumour and poorly differentiated malignancies, the possible primary in case of metastasis from unknown primary (MCUO), neuroendocrine vs neuroectodermal lesions, and many others. Lineage identification by IHC is particularly helpful in lymphomas, melanomas, astrocytomas and pecomas. Algorithms constructed on the basis of previously demonstrated immunoreactivity on a large number of test cases have proved to be accurate in identifying the primary in 67% of MCUO. The IHC approach to find the possible primary is much more cost-effective as opposed to extensive work-up with radiological studies and various endoscopy procedures.

D. Clinical Decision support :
Several immunohistological markers which influence tumour behaviour and help tumour prognostication for microstaging, predicting response to therapy, monitoring drug resistance, detecting growth factors and receptors, evaluating tumour angiogenesis etc. have been described in the literature. Most of these have not as yet been incorporated in a routine diagnostic setting. There are other markers such as CD20, C-kit and HER-2/NEU which are increasingly requested to clinically justify targeted therapy. The application of this information for the Indian patients could be guided by the general guidelines for evidence based medicine with the following considerations: a) will the IHC result make a difference in the management or will the patient benefit from the information given by the IHC result; b) what about the cost to the patient and the delay in the report? c) is the treatment feasible in our socioeconomic setting? Generating IHC results which are not going to affect patient management or are not applicable to an individual patient only burden the IHC laboratory with no value added to an individual patient although they make the typed report look more impressive.

E. Importance of data warehousing :
Unlike blood chemistry results there are no readily available reference ranges for the several IHC antibodies with respect to age, normal tissue, physiological alterations, benign tumours and malignant tumours. In the Indian context IHC is in its infancy and there is a lack of information on Indian co-horts. Geographic variation could influence IHC results such as percentage of ER/PR positivity in breast tumours, ALK1 & CD30 in ALCL, and ALK-1 in B- cell NHL. This highlights the importance of data warehousing to obtain information on Indian patients

F. Quality control / assurance in Immunohistochemistry :
Qualitive assurance is easily applicable to results which are quantifiable and objective. Attempts at quality assurance in IHC comprise establishment of standardized procedures to ensure technical reproducibility, uniformity in interpretation and evaluation and quantification of extent of immunoreaction by the use of a scoring system to bring in objectivity. An interlaboratory trial involving 172 pathologists and 3526 immunostains brought out some unexpected findings. There is a general belief that the staining quality is a major problem in the application of IHC; however a multivariate model in which each step of the diagnostic pathway, beginning with a pre IHC tentative diagnosis, was introduced, revealed that only (i) the correct tentative diagnosis, (ii) the interpretation of the IHC stain and (iii) the conclusions drawn from the IHC stain were independently predictive of the correct final diagnosis. Neither antibody selection nor quality of immunostain correlated independently to the correct final diagnosis. A definitive diagnosis could not be rendered if the morphological tentative diagnosis was incorrect or not included in the differential diagnosis. The results of another large study, evaluating interlaboratory and interobserver agreement for semiquantitative assessment of estrogen receptor (ER) using Tissue array technology wherein 172 laboratories participated, suggested that neither the pH of the formalin buffer or the duration in the fixative greatly influenced the detection of ER. Variability in subsequent IHC practices (such as antigen retrieval) and interpretation of results was a greater source of diagnostic error. The antibody which has been subjected to a lot of scrutiny with respect to overall evaluation of accuracy, specificity, sensitivity and reproducibility is HER 2/neu overexpression to identify the 20%-30% of breast cancer women who could benefit from Herceptin therapy (which is a humanized monoclonal antibody). A comparative study involving results from 94 Laboratories in 21 countries (predominantly European) participating in the National External Quality Assessment scheme for Immunocytochemistry (UK NEQAS-ICC) concluded that the reliability of the Her-2 assay could be greatly improved by stringent quality control and an ongoing quality assurance program using a standard reference obtained from cell lines.

To conclude, standardization in IHC is a daunting task. The multiple variables which affect the final result are not easy to control but standardization of technical steps, and hopefully availability of a reference for most lesions in future (such as in HER-2/neu assays) will bring in uniformity in IHC results regardless of where they are generated.

As of now it will be difficult to find a pathologist who has not, at some point or other, decided to neglect the IHC result.

Recommended reading :
– Berman JJ: Racing to Share Pathology Data. Am J Clin Pathol 121: 169-171, 2004
– Bock BJ, Dolan CT, Miller GC et al: The Data Warehouse as a Foundation for Population-Based Reference Intervals. Am J Clin Pathol 120: 662-670, 2003
– Leong ASY: Immunohistological markers for tumor prognostication. Current Diagnostic Pathology 7: 176-186, 2001
– Lotan Y and Roehrborn CG: Sensitivity and Specificity of Commonly Available Bladder Tumor Markers versus Cytology: Results of a Comprehensive Literature Review and Meta-analyses. Urology 61: 109-118, 2003
– Marchevsky AM and Wick MR: Evidence-Based Medicine, Medical Decision Analysis, and Pathology. Hum Pathol 35: 1179-1188, 2004
– Rhodes A, Jasani B, Anderson E, et al: Evaluation of HER-2/neu Immunohistochemical Assay Sensitivity and Scoring on Formalin- Fixed and Paraffin-Processed Cell Lines and Breast Tumors. A Comparative Study Involving Results From Laboratories in 21 Countries. Am J Clin Pathol 118: 408-417, 2002
– Rüdiger T, Kreipe HH, Nizze H et al: Quality Assurance in Immunohistochemistry. Results of an Interlaboratory Trial Involving 172 Pathologists. Am J Surg Pathol 26(7): 873-882, 2002
– Straus SE and Sackett DL: Review on evidence-based cancer medicine. Ann Oncol 10: 29-32, 1999
– Wasielewski RV, Mengel M, Wiese B et al: Tissue Array Technology for Testing Interlaboratory and Interobserver Reproducibility of Immunohistochemical Estrogen Receptor Analysis in a Large Multicenter Trial. Am J Clin Pathol 118: 675-682, 2002