The Use Of Immunocytochemistry In The Diagnosis Of Neoplasia
William Vernau BSc, BVMS, DVSc, PhD, Diplomate ACVP Davis CA
Immunocytochemistry (ICC) is a technique for the detection of antigens in cytologic specimens using antibodies specific for those antigens . Antigen-antibody binding, and hence the detection of the presence of the specific antigen in question, is most often achieved by using a secondary antibody (usually labeled), an enzyme or enzyme complex system and a chromogenic substrate. Localized enzyme results in a color change in positive cells (via the chromagen) which are then visible using a standard light microscope. This technique has been used for diagnosis of tumor types and infectious agents, as well as for detection of immune complex deposition, autoantibodies, and markers of cell growth and differentiation.1 The primary use of ICC is in helping diagnose cancer. However, at this time, there are no immunocytochemical markers that differentiate neoplastic from non-neoplastic cells>. Therefore, immunocytochemical tests should be used only AFTER a diagnosis of malignancy is made using routine morphologic assessment (cytology or histopathology) and when further tumor classification is indicated. The technique has been applied mainly to the classification of tumors into specific cell types or the determination of the origin of metastases. As such, it can be a valuable adjunct (but not a substitute) to cytologic or histologic examination of cells or tissues and may allow a more precise diagnosis of tumor type when the type cannot be determined by routine morphologic assessment. In recent years, with the availability of antibodies to cellular antigens that can predict the potential biological behaviour of various neoplasms, ICC has become even more important in the diagnosis and management of patients with cancer. This technique is now being used to learn more about the behavior, prognosis, and response to therapy of tumors in animals.
Immunocytochemical methods have advantages and disadvantages over other methods used to gain similar diagnostic information.2,3,4 An important advantage in some cases is that cell morphology is preserved , thus the cells and their staining pattern may be observed and assessed simultaneously.5 Additionally, the fixation and processing techniques applied to cytologic specimens are usually “gentler” than those used in histopathologic specimens, resulting in greater antigen preservation and therefore, applicability of a broader range of antibodies for ICC versus immunohistochemistry. As many of the available antibodies can be used on fixed as well as frozen tissue and stored slides, the techniques may be used in retrospective studies or applied to cases in which previous diagnostic tests have not been satisfactory or conclusive. The cost for immunostaining may be higher than other methods with a current range of around $10-$40 per test.
Sensitivity and specificity of immunocytochemistry depend on several factors. Sensitivity of a particular antibody may be quite high, particularly when monoclonal antibodies are used. There are a number of different modifications of immunocytochemical techniques that may be used to optimize sensitivity without increasing background staining. Ultimately, the sensitivity of an antibody and protocol is determined on the basis of the smallest amount of antigen detected as well as the number of cells in a sample that express the antigen. Specificity is similar in that it depends technically on specificity of an antibody for a particular marker (usually higher for monoclonal versus polyclonal antibodies) but is also determined by the specificity of the marker for a particular diagnosis. The majority of currently used cell markers are not perfectly specific and found in more than one cell type. It is therefore usually necessary to select a “panel” (2 or more) of markers and interpret results in concert. Additionally, as stated previously, there are no immunocytochemical markers that differentiate between neoplastic and non-neoplastic cells. This must be kept in mind when ordering immunocytochemical assays and in interpretation of the results of these tests.
Immunocytochemistry may be performed on blood and bone marrow smears and smears of fine needle aspirates, impression smears, scrapings, cavity fluids and effusions. On these samples, cell membrane antigens are most easily detected, and sensitivity for intracytoplasmic and intranuclear antigens may be somewhat reduced. There are, however, techniques that may be used to increase sensitivity.
An important consideration in sampling for immunocytochemistry is that it is usually necessary to assay a sample for the presence of several different markers. This, in addition to the need for positive and negative controls, leads to a requirement for multiple slides from a sample. These multiple slides should preferably be similar in cell numbers and morphologic preservation of cells. Making cytospins of a fluid sample and making multiple impression smears of the cut surface of an excised mass are relatively easy methods of obtaining multiple slides of equivalent quality. It can be more difficult to achieve this with fine needle aspirates, or scrapings, but creation of cell suspensions and resultant cytospinning or synthesis of cell “blocks” (with sectioning) are possible approaches.
Immunocytochemistry may be performed on slides already stained with cytologic stains such as Wright’s stain or Papanicolaou’s stain.6 De-staining of these slides is not necessary, but may be performed.
The actual process of immunocytochemical staining involves “fixation” of the slide then incubation of the cells on the slide with a number of different reagents, interspersed with thorough rinses to remove unbound reagents. This may be done manually by flooding each slide with reagent and dipping or flushing slides for rinses. Automated immunostainers add more equipment cost but have the advantages of increased convenience, standardization and reproducibility. The process may or may not involve pretreatment but should ideally involve a “quenching” step and a “blocking” step before addition of the primary antibody, the secondary antibody and the enzyme complex. The following steps will be discussed in more detail during the talk: Fixation, Quenching, Blocking, Primary antibody, Secondary antibody, Enzyme complex.
The choice of primary antibody used is crucial for the diagnostic utility of an immunocytochemical procedure. Ideally, the primary antibody should be highly specific and sensitive and work well on formalin-fixed, paraffin-embedded tissues as well as frozen or fresh cells. Using a commercially available antibody has the advantage of continual and, usually, easy availability. Each primary antibody must be validated for use in the species it will be testing.
In the diagnosis of neoplasia, it is usually necessary to use a panel of reagents to increase sensitivity and specificity. The choice of markers depends entirely on their purpose. A common purpose is immunocytochemical classification of tumor type. Ideally, this panel will be chosen based on criteria such as clinical findings, cytomorphologic and gross appearance of a lesion, the differential diagnosis (which integrates these findings plus such things as signalment, anatomic location etc), and, most importantly, the availability of markers for the entities within the differential diagnosis.
There are advantages and disadvantages of using monoclonal antibodies as opposed to polyclonal antibodies. Monoclonals have the advantages of being epitope-specific and may be used usually at higher dilutions, thus minimizing cross-reactivity and background staining. However, some monoclonal antibodies are “too” specific and do not work well on all tissues or species; in these cases, polyclonal sera are preferred.
Controls are essential for accurate immunocytochemistry just as they are in every laboratory procedure. Every time an immunocytochemical procedure is performed, a known positive and an antibody negative control should be run. For cytologic samples, cytologic controls should be used. To obtain cytologic positive controls, impression smears may be made of neoplastic and normal tissues expressing the marker. These slides can be fixed and saved. For the antibody negative control, a duplicate slide from the patient is incubated with an irrelevant antibody (preferably isotype matched) or non-immune serum from the same animal species as the primary antibody. If duplicate slides from the patient are not available, a cytologic smear can be divided into two separate areas with a diamond or grease pen. One area is then used for the primary antibody and the other is used as an antibody negative control.
The knowledge and experience of the individual evaluating the slides is an extremely important factor in determining the clinical value of immunocytochemistry. This individual has to be aware of the staining characteristics of true positive reactions as well as technical and biologic factors that can contribute to false positive and false negative staining. The observer should know the normal location of the antigen within the cell, i.e. should cell nucleus, cytoplasm or cell membrane be staining. True positive staining often appears as crisp granules and does not spread or smudge into surrounding areas. If cells within a sample are crushed, ruptured or necrotic, they will often stain falsely positive. Large three-dimensional clusters of cells will trap antibodies within the clusters and result in false positives. False negatives are often due to technical problems that should be apparent by the controls used in the procedure.
There are numerous antibodies available for immunocytochemical staining. These vary widely in their sensitivity and specificity for cell types. They also vary in species-specificity in that there are some species differences in expression of antigens as well as in affinity of some of these antibodies. As stated previously, each primary antibody must be validated for use in the species it will be testing. Some of the most common targets for immunostaining are the “intermediate” filaments. The intermediate filaments are important cytoskeletal proteins and are so called because they are “intermediate” in size between the “thin” actin filaments and “thick” myosin filaments of myocytes. They are present in most mammalian cells, are highly conserved between species and expression of specific intermediate filaments is usually maintained in neoplastic cells. They include cytokeratin, vimentin, desmin, glial fibrillar acid protein, neurofilament protein, lamins and peripherin.
An antibody to cytokeratin is most often used to detect epithelial cells.7,8 The cytokeratins are a complex family of proteins. There are more than 20 distinct keratins in human epithelia that are preferentially expressed in a site specific manner. Consequently, there are also many monoclonal antibodies specific for certain types of epithelial cell membranes such as mammary gland, pancreas, prostate, liver, and lung that have been used in human medicine. The targets of these antibodies are often expressed preferentially in carcinomas of certain types, but they have not been generally shown to be both highly sensitive and specific. Nevertheless, panels of keratin antibodies have been used to identify the site of origin of metastatic carcinomas in people, and to help categorize skin tumors in dogs.8,9 Carcinoembryonic antigen (CEA) is not specific for a certain carcinoma type, but in human medicine, adenocarcinomas have been classified into “usually CEA positive” and “usually CEA negative” types.3 Use of CEA in the veterinary literature is limited though there have been reports of CEA positivity in canine primary liver and pancreatic carcinomas,10,11 primate colon carcinoma,12 and negative results contributing to a diagnosis of mesothelioma in a rhesus monkey.13
Vimentin is an intermediate filament found in mesenchymal cells and is present in sarcomas such as fibrosarcomas, osteosarcomas, hemangiopericytomas, etc. Although it is used in some immunocytochemical panels, it is neither sensitive nor specific as it may or may not be present in undifferentiated tumors of many types and is commonly co-expressed with other intermediate filaments (including cytokeratin in some carcinomas). Desmin is a marker found in both smooth and striated muscle. It may be used for primary diagnosis as a muscle tumor and is considered to be the most sensitive marker of tumors of muscle origin. Confirmatory antibodies for muscle tumor type include muscle actin for smooth muscle and myoglobin for skeletal muscle. These are not always specific, depending on the degree of differentiation of the tumor. An antibody to von Willebrand’s factor is often used to identify endothelial cells in hemangiosarcomas and hemangiomas. This marker is fairly specific to endothelial cells and megakaryocytes. Additional markers used to identify tumors of vascular endothelial origin include Factor VIII, PECAM (CD31) and CD34. Differentiation between melanomas and neurogenic sarcomas may be difficult as both tend to be positive for S-100, but myelin basic protein (MBP) may be used as it should be positive in neurogenic sarcomas and negative in melanomas. Additionally, although less sensitive than S-100, antibodies to Melan A have greater specificity for the diagnosis of melanoma and have been used for this purpose in veterinary species.14,15
Markers for neuroendocrine tumors are generally not specific as to tumor type and are variably-expressed in these tumors, but they may be useful as part of a panel to distinguish, for example, epithelial or round-cell tumors from endocrine tumors. Markers most commonly used (in order of utility) include neuronal-associated enolase, chromogranin A, glial fibrillary acid protein (GFAP) and synaptophysin.
Diagnosis of round cell tumors may be facilitated by use of several markers including those for surface antigens specific to a variety of leukocytes.1,16 These are addressed in greater detail in my next manuscript / talk (Beyond the microscope - New tools for the assessment of hematopoietic neoplasia).
Markers of tumor behavior
There is currently a great deal of interest and activity in determining the presence of certain markers of cell proliferation and differentiation in a variety of tumors, and correlating them with tumor behavior and prognosis. Markers that have been used with immunocytochemical methods include indicators of cell kinetics such as proliferating cell nuclear antigen (PCNA) and Ki-67, both of which identify cycling cells and may be used to evaluate the proliferative activity of a tumor.17,18,19 Over-expression of bcl-2, a gene product that protects cells from apoptosis, has been associated with several tumor types and behaviors.20,21
Detection of increased levels of products of oncogenes and suppressor genes such as p53 and c-erb-2 (HER-2/neu) have been investigated in assessing prognosis of neoplasia in human, and increasingly, in veterinary medicine.22,23,24 Similarly, expression of growth factors and their receptors, such as Epidermal Growth Factor-receptor (EGF-R) and cell adhesion molecules and receptors have been shown to be associated with poor prognosis in some tumor types. Certain steroid hormone receptors, such as estrogen and progesterone, are expressed differentially in tumors that exhibit benign versus malignant behavior. Detection of these receptors is being used prognostically in human breast cancer and may be valuable in veterinary medicine.25
P-glycoprotein is plasma membrane protein that acts as a pump and is associated with multiple drug resistance (MDR). Immunological assessment of P-glycoprotein has been used in order to help predict biological behavior and response to therapy of cases of canine lymphoma.26 Similarly, cyclooxygenase 2 (COX-2) expression has been examined in canine TCC and renal carcinoma in an effort to understand the biology of the tumor and, potentially, to design rational therapeutic strategies.27,28
Variability of expression of many antigens in neoplastic cells can create problems with both sensitivity and specificity of ICC. Additionally, some tumor cells, when they become phagocytic, react with antibodies to antigens that wouldn’t normally be expressed by that cell type. The optimal utilization of ICC results in patient care should include integration of all historical, clinical, morphologic and other (ie. cytochemical, cytogenetic, molecular genetic) information.
1. Caniatti M, Roccabianca P, Scanziani E. Paltrinieri S, Moore PF. Canine lymphoma: immunocytochemical analysis of fine-needle aspiration biopsy. Vet Pathol 33:204-212, 1996.
2. Nadji M, Ganjei P. Immunocytochemistry in diagnostic cytology. Am J Clin Pathol 94:470-475, 1990.
3. Nadji M, Ganjei P, Morales AR. Immunocytochemistry in contemporary cytology. Laboratory Medicine 25:502-508, 1994.
4. Oertel J, Huhn D. Immunocytochemical methods in haematology and oncology. J Cancer Res Clin Oncol 126:425-440, 2000.
5. Wren G. Immunohistochemistry helps many diagnoses. Bovine Veterinarian, Nov-Dec., 2000:4-14, 2000.
6. Dardik J, Epstein JI. Efficacy of restaining prostate needle biopsies with high-molecular weight cytokeratin. Hum Pathol 31:1155-1161, 2000.
7. Sandusky GE, Wightman KA, Carlton WW. Immunocytochemical study of tissues from clinically normal dogs and of neoplasms, using keratin monoclonal antibodies. Am J Vet Res 52:613-618,1991.
8. Walter J. A cytokeratin profile of canine epithelial skin tumours. J Comp Path. 122:278-287,2000.
9. Campbell F, Herrington CS. Application of cytokeratin 7 and 20 immunohistochemistry to diagnostic pathology. Current Diagnostic Pathology 7:113-122, 2001.
10. de las Mulas J, Gomez-Villamandos JC, Perez J, Mozos E, Estrado M, Mendez A. Immunohistochemical evaluation of canine primary liver carcinomas: distribution of alpha-fetoprotein, carcinoembryonic antigen, keratins and vimentin. Res Vet Sci 59:124-127,1995.
11. Rabanal R, Fondevila D, Vargas A, Ramis A, Badiola J, Ferrer L. Immunocytochemical detection of amylase, carboxypeptidase A, carcinoembryonic antigen and alpha 1-antitrypsin in carcinomas of the exocrine pancreas of the dog. Res Vet Sci 52:217-223,1992.
12. Brack M. Lectin histochemistry and carcinoembryonic antigen in spontaneous colonic cancers of cotton-top tamarins (Saguinus oedipus). Vet Pathol 32:668-673,1995.
13. Chandra J, Mansfield KG. Spontaneous pericardial mesothelioma in a rhesus monkey. J Med Primatol 28:142-144, 1999.
14. Ramos-Vara JA, Beissenherz ME, Miller MA, Johnson CG, Pace LW et al. Retrospective study of 338 canine oral melanomas with clinical, histologic, and immunohistochemical review of 129 cases. Vet Pathol 37:597-608,2000.
15. Koenig A, Wojcieszyn J, Weeks BR, Modiano JF. Expression of S100a, Vimentin, NSE, and Melan A/MART-1 in seven canine melanoma cell lines and twenty-nine retrospective cases of canine melanoma. Vet Pathol 38:427-435,2001.
16. Vernau W, Moore PF. An immunophenotypic study of canine leukemias and preliminary assessment of clonality by polymerase chain reaction. Vet Immunol Immunopathol 69(2-4): 145-64,1999
17. Phillips BS, Kass PH, Naydan DK, Winthrop MD, Griffey SM, Madewell BR. Apoptosis and proliferation indexes in canine lymphoma. J Vet Diagn Invest 12:111-117,2000.
18. Kiupel M, Bostock D, Bergmann V. The prognostic significance of AgNOR counts and PCNA-positive cell counts in canine malignant lymphomas. J Comp Path 119:407-418,1998.
19. Funakoshi Y, Nakayama H, Uetsuka K, Nishimura R, Sasaki N, Doi K. Cellular proliferative and telomerase activity in canine mammary gland tumors. Vet Pathol 37:177-183,2000.
20. Bozzetti C, Niazzol R, Naldi N, Guazzi A, Camisa R, et al. Bcl-2 expression on fine-needle aspirates from primary breast carcinoma Cancer 87:224-280.1999.
21. Madewell BR, Gandour-Edwards R, Edwards BF, Walls JE, Griffey SM. Topographic distribution of bcl-2 protein in feline tissues in health and neoplasia. Vet Pathol 36:565-573,1999.
22. Jaffe MH, Gosgood G, Taylor HW, Kerwin SC, Hedlund CS, Lopez MK, Davidson JR, Miller DM, Paranjpe M. Immunohistochemical and clinical evaluation of p53 in canine cutaneous mast cell tumors. Vet Pathol 37:40-46,2000.
23. Ginn PE, Fox LE, Brower JC, Gaskin A, Kurzman ID, Kubilis PS. Immunohistochemical detection of p53 tumor-suppressor protein is a poor indicator of prognosis for canine cutaneous mast cell tumors. Vet Pathol. 37:33-39 2000.
24. Nasir L, Krasner H, Argyle DJ, Williams A. Immunocytochemical analysis of the tumour suppressor protein (p53) in feline neoplasia. Cancer Letters 155:1-7,2000.
25. de las Mulas JM, van Niel M, Millan Y, Blankenstin MA, van Mil F, Misdorp W. Immunohistochemical analysis of estrogen receptors in feline mammary gland benign and malignant lesions: comparison with biochemical assay. Domestic Animal Endocrinoly 18:111-125,2000.
26. Bergman PJ, Ogilvie GK, Powers BE. Monoclonal antibody C219 immunohistochemistry against P-glycoprotein: sequential analysis and predictive ability in dogs with lymphoma. J Vet Intern Med 10(6):354-359,1996.
27. Khan KNM, Knapp DW, DeNicola DB, Harris K. Expression of cyclooxygenase–2 in transitional cell carcinoma of the urinary bladder in dogs. Am J Vet Res 61:478-481,2000.
28. Khan KNM, Stanfield KM, Trajkovic D, Knapp DW. Expression of cyclooxygenase–2 in canine renal cell carcinoma. Vet Pathol 38:116-119,2001.