Immunohistochemical differentiation between primary adenocarcinomas of the ovary and ovarian metastases of colonic and breast origin. Comparison between a statistical and an intuitive approach.

AIM: To discriminate between adenocarcinomas that are primary to the ovary and metastatic to the ovary, especially of colonic and breast origin, by immunohistochemistry, using stepwise discriminant analysis or a decision tree. METHODS: 312 routinely processed, formalin fixed tissue specimens were used. The tumours were divided into a learning set (n = 159), composed of primary tumours of ovary, breast, and colon, and a test set, comprising 134 metastases from these sites and an additional 19 primary ovarian carcinomas. The immunohistochemical panel was composed of antibodies against cytokeratin 7 (CK7) and 20 (CK20), CA125, vimentin, carcinoembryonic antigen (CEA), gross cystic disease fluid protein-15 (GCDFP-15), and the oestrogen receptor (ER). The staining results of the tumours were expressed as the product of the staining intensity and the percentage of positive tumour cells. Analyses were first performed on the learning set and then evaluated on the test set. RESULTS: Although the immunostaining patterns showed a considerable overlap between the three types of adenocarcinoma, the breast carcinomas were typically positive for GCDFP-15 and often for ER, and negative for vimentin. Ovarian carcinomas were always positive for CK7 and to a lesser extent for CA125. Colonic carcinomas showed prominent positivity for CEA and CK20, while no staining was seen for ER and vimentin. In discriminant analysis, six antibodies (alpha CK7, alpha CK20, alpha CA125, alpha CEA, alpha ER, and alpha GCDFP-15) appeared to be necessary for optimal classification: 89% of the learning set and 82% of the test set were classified correctly. In the decision tree, only four antibodies (alpha CK7, alpha CEA, alpha ER, and alpha GCDFP-15) were used to obtain a correct classification score of 89% for the learning set and 84% for the test set. CONCLUSIONS: Using a semiquantitative assessment of the immunostaining results by a restricted panel of six antibodies with stepwise discriminant analysis, 80-90% of the adenocarcinomas of colon, breast, and ovary can be correctly classified. Discriminant analysis is computer aided and therefore an easy method and for each case a probability value of the classification result is obtained. The intuitive decision tree method provides a slightly better result, requires only four antibodies, and offers a more practical method for the surgical pathologist.

Tracing the origin of adenocarcinomas with unknown primary using immunohistochemistry: differential diagnosis between colonic and ovarian carcinomas as primary sites.

To discriminate adenocarcinoma metastases originating from either colon or ovary, a panel of immunohistochemical markers was evaluated. For this purpose, paraffin sections from 157 primary and metastatic colonic and ovarian carcinomas were immunostained. These cases were divided into a learning group of 46 colonic and 54 ovarian carcinomas and a test group of 29 colonic and 28 ovarian carcinomas, including all metastatic tumors, among which were five with unknown primary site at the time of testing. The sections were immunostained with antibodies against carcinoembryonic antigen (CEA), cytokeratin 7 (CK7), cytokeratin 20 (CK20), CA125, vimentin, and CA19.9. Staining results were expressed as the product of staining intensity and percentage of positive tumor cells. Stepwise discriminant analysis was applied on the learning set to obtain a classification function for both tumors. The validity of the classification function was evaluated using the test set. There was considerable overlap in immunostaining for both tumor types, but colonic carcinomas were typically positive for CEA and CK20 and negative for CK7 and CA125. Ovarian carcinomas were typically positive for CK7 and CA125 and negative for CEA and CK20. In discriminant analysis, the best combination of markers appeared to be CK7 and CEA. Only one sample of the test group (2%) was misclassified. Taking learning and test groups together, 136 of the 157 samples (87%) were correctly classified with high posterior probability (PP > .8). However, from the 28 mucinous ovarian carcinomas, only 19 (68%) could correctly be classified with high PP. When excluding the nonmucinous ovarian carcinomas from the analysis, overall 87 of 103 (84.5%) of the samples were correctly classified (PP > .8) with a combination of CEA, CK7, and also vimentin. From the 28 mucinous ovarian carcinomas, only two (7%) were misclassified, and four could not be classified with sufficient certainty. In neither analysis did CK20, CA125, or CA19.9 emerge as discriminatory parameters. Based on the same data, an intuitive flow chart was constructed with which 129 of 157 cases could be classified (only one falsely) without further statistical analysis. The five metastases with an at first unknown primary could, according to the follow-up, all be classified correctly with high PP. Most ovarian carcinomas, including the mucinous ones, can be discriminated with high probability from colonic carcinomas using a panel of three antibodies directed against CEA, cytokeratin 7, and vimentin.

Decreased expression of cellular markers in Epstein-Barr virus-positive Hodgkin's disease.

Epstein-Barr virus (EBV) has been demonstrated in the Reed-Sternberg cells and their mononuclear variants (Hodgkin cells; H-RS cells) in a substantial number of Hodgkin's disease (HD) cases. Moreover, EBV can modulate both in vivo and in vitro the expression of several cellular genes, including lymphoid differentiation markers. Therefore we investigated, in 64 cases of HD, the relationship between the presence of EBV and the expression of lymphoid (CD45RB), T- (CD3, CD45RO), B- (CD20, MB2 antigen, CDw75), and myeloid-cell lineage markers (CD15), and of activation markers (CD30, EMA, and the 115D8 antigen) on the H-RS cells. EBV-positive cases, as demonstrated by the presence of EBER-1 and -2 RNA and LMP-1 protein expression, showed a significant reduction in the expression on H-RS cells of T-cell lineage (CD3, P < 0.02), B-cell lineage (CD20; P < 0.005), and activation markers (EMA; P < 0.002 and the 115D8 antigen; P < 0.001) as compared with EBV-negative cases. No differences were found in the expression of CD15, CD30, CD45RO, CD45RB, CDw75, or the MB2 antigen on H-RS cells in EBV-positive and EBV-negative HD cases. Interestingly, in 11 cases of EBV-negative HD, B- as well as T-cell lineage markers could be found on some H-RS cells. These data suggest that EBV in H-RS cells is able to down-regulate the expression of T- (CD3) and B- (CD20) cell lineage markers and lymphoid activation markers (EMA and the 115D8 antigen).(ABSTRACT TRUNCATED AT 250 WORDS)

Presence of Epstein-Barr virus harbouring small and intermediate-sized cells in Hodgkin's disease. Is there a relationship with Reed-Sternberg cells?

Forty-four cases of Hodgkin's disease (HD), mostly of the nodular sclerosing type, were investigated for the presence of Epstein-Barr virus (EBV) by polymerase chain reaction (PCR) and DNA and RNA in situ hybridization (DISH, RISH), as well as by immunohistochemistry for the detection of latent membrane protein-1 (LMP-1) of EBV. In situ hybridization (ISH) was combined with immunohistochemistry to correlate the presence and activity of the virus at the cellular level. In 18/34 (53 per cent) cases, EBV-DNA sequences could be detected with the PCR method. In 12/18 positive cases, DISH and RISH were also positive. In the remaining six EBV-PCR positive cases, two were also positive with RISH and LMP-1, whereas no positive signal with DISH could be obtained. All DISH and/or RISH positive cases were also positive for LMP-1. With RISH, not only the Reed-Sternberg cells and their mononuclear variants (RS cells) stained positive, but also small and intermediate cells frequently reacted with the EBV-specific probes (EBER-1 and -2). Double staining with cellular markers (CD3, CD20, CD45, CD45RO, CD68, and the lectin PNA) revealed that most of the smaller EBER-positive cells frequently did not express T, B, or histiocytic markers, but that they, as well as the RS cells, showed cytoplasmic and membranous staining with PNA. These smaller EBER-positive cells were not found in EBV-PCR negative HD. EBER-positive RS cells were almost always LMP-1 positive, as well as a substantial proportion of the intermediate-sized cells, whereas the majority of the small EBER-positive cells remained LMP-1 negative.(ABSTRACT TRUNCATED AT 250 WORDS)

Expression of c-myc and bcl-2 oncogene products in Reed-Sternberg cells independent of presence of Epstein-Barr virus.

AIMS: To evaluate the expression of c-myc and bcl-2 oncogene products in Reed-Sternberg cells in Hodgkin's disease, especially in relation to Epstein-Barr virus infection and expression of EBV encoded latent membrane protein (LMP). METHODS: Tissues from 33 cases of Hodgkin's disease were studied for the presence of EBV DNA by polymerase chain reaction (PCR) and DNA in situ hybridisation (DISH), for the presence of EBER-1 and EBER-2 EBV RNA by RNA in situ hybridisation (RISH); and for the presence of LMP, bcl-2, and c-myc proteins by immunohistochemical staining. RESULTS: A substantial number of Reed-Sternberg cells expressed bcl-2 in 20 of 29 (69%) and c-myc in 30 of 32 (94%) Hodgkin's disease samples. In 18 of the 25 (72%) cases Reed-Sternberg cells expressed both oncogene products. Of these 18 cases, 10 (56%) were EBV-PCR positive; eight (44%) were EBV-PCR negative. CONCLUSIONS: Reed-Sternberg cells in Hodgkin's disease frequently express both bcl-2 and c-myc oncogene products, suggesting that these oncogenes may act in concert in the pathogenesis of the disease. Moreover, the expression of c-myc and bcl-2 proteins in Reed-Sternberg cells is independent of EBV and LMP status.

Epstein-Barr virus DNA in Reed-Sternberg cells of Hodgkin's disease is frequently associated with CR2 (EBV receptor) expression.

We studied 44 cases of Hodgkin's disease for the presence of Epstein-Barr virus (EBV) DNA, its localization and the expression of the EBV receptor on the tumour cells. EBV DNA was found in 52% (16/31) of the Hodgkin's lymphomas using the polymerase chain reaction. With a very sensitive non-radioactive DNA in situ hybridization technique in combination with immunohistochemistry for CD 30 or CD 15 antigens, EBV DNA was localized to Reed-Sternberg cells and its mononuclear variants. The relationship between the presence of EBV DNA and the expression of the EBV-receptor CR2 (CD 21) on Reed-Sternberg cells was studied using the same techniques and two different monoclonal anti-CD 21 antibodies. CR2 could be detected on a substantial number of the Reed-Sternberg cells in EBV DNA positive Hodgkin's lymphomas (9/12; 75%), whereas in EBV negative cases positivity with anti-CD 21 was rare (1/13; 8%). The results indicate that CR2 expression on Reed-Sternberg cells and the presence of EBV DNA sequences are frequently associated in Hodgkin's lymphomas.

Application and comparison of silver intensification methods for the diaminobenzidine and diaminobenzidine-nickel endproduct of the peroxidation reaction in immunohistochemistry and in situ hybridization.

Silver-intensification methods described in the literature for the diaminobenzidine (DAB) and diaminobenzidine-nickel (DAB/Ni) endproduct of the peroxidase reaction were compared in model systems after immunoperoxidase and in situ hybridization. First, these methods were compared in immunohistochemical model systems, using the demonstration of glial fibrillar acidic protein (GFAP) and prostate-specific antigen (PSA) in paraffin sections of human brain and prostate tissue, respectively. When DAB without Ni was used as substrate, tissue argyrophilia caused considerable background staining. Only when this tissue reactivity was quenched with, e.g., CuSO4 with H2O2 or thioglycolic acid, were the results acceptable. A considerable improvement in the signal-to-noise ratio could be obtained when nickel was included in the substrate mixture. The methods that proved to be best for demonstration of GFAP and PSA made use of acid developer solutions. Subsequently, these methods were compared with other sensitive immunostaining methods for demonstration of the gamma-delta T-cell receptor in frozen lymphoid tissue. In this model a considerable increase in the number of positive cells could be obtained using silver intensification. The different methods using DAB/Ni were also compared for use in DNA in situ hybridization (DISH). In this case two model systems were used: human papilloma virus type 11 (HPV-11) DNA in condyloma tissue (abundant target model) and Epstein-Barr virus (EBV) DNA in a mononucleosis lymph node (low target model). For demonstration of HPV-11, all methods gave more or less satisfactory results, which were best with the acid developer solutions. Moreover, for demonstration of EBV DNA, a signal could be obtained only with these developer solutions. Such a method also proved suitable in double immuno-hybrido stainings for the demonstration of EBV DNA in specific antigen-positive Reed-Sternberg cells in paraffin sections of Hodgkin lymph nodes.

Demonstration of changes in cytokeratin expression in condylomata accuminata in relation to the presence of human papilloma virus as shown by a combination of immunohistochemistry and in situ hybridization.

Human papillomavirus (HPV) can be detected in, and is probably involved in the etiology of, the majority of anogenital neoplasias. Infection with the virus induces a number of events in the infected epithelial cells that may lead to the development of benign or malignant tumors. One change that can be detected in the infected cells is in squamous differentiation, which is reflected by the pattern of cytokeratin polypeptide expression. By studying this pattern in relation to the presence of the virus, an indication may be obtained of the influence of the virus on the cellular differentiation in individual cells. By using a combination of DNA in situ hybridization and immunohistochemistry, for HPV and cytokeratin polypeptides, respectively, we studied the presence of HPV6 or HPV11 in condylomata accuminata derived from anogenital skin in relation to the cytokeratin polypeptides K1, 4, 8, 10, 14, and 18. We found that in many samples the presence of the skin-type cytokeratins K1 and K10 was decreased, whereas K13, and to a lesser degree K4, appeared. The cellular localization of these aberrations in cytokeratin expression could be related to the presence of HPV6 or 11 DNA in the tissue.

Macrophage-T suppressor cell interference in the lungs of steroid-treated sarcoidosis patients.

In the bronchoalveolar lavage of sarcoidosis patients the mononuclear cell infiltrate was enumerated on T helper and suppressor lymphocytes as well as macrophages by means of a triple-staining assay on cytospin slides. As was seen on the slides, lymphocytes were often adhered very closely to macrophages. This phenomenon, many times described but not understood, was studied in a group of 13 sarcoidosis patients, of whom 7 received prednisolone treatment. It could be shown that treatment with the corticosteroid was followed by an increase in the percentage of suppressor lymphocytes adhered to macrophages. Second, the number of such alveolar T suppressor lymphocyte-macrophage aggregates was dramatically increased in the prednisolone-treated patients.

Oral hairy leukoplakia in HIV infection: a diagnostic pitfall.

Twenty-nine human immunodeficiency virus (HIV)-infected patients with white, nonremovable lesions on the lateral border of the tongue, clinically suggestive of oral hairy leukoplakia (HL), were studied. In particular, the value of local antifungal therapy in establishing the diagnosis of HL was investigated. In 15 patients (52%) the lesions could be ultimately attributed to a candidal infection of the tongue. In 10 of the remaining 14 patients, a biopsy was obtained from lesions persisting after local antifungal treatment. In all biopsy specimens, the diagnosis of HL was confirmed by histopathologic examination and the demonstration of Epstein-Barr virus DNA by polymerase chain reaction, Southern blot hybridization, and DNA in situ hybridization. The present data confirm that the diagnosis of HL in HIV-infected patients cannot be reliably made on clinical criteria alone, but requires histopathologic confirmation including the demonstration of Epstein-Barr virus DNA, preferably by DNA in situ hybridization. However, with regard to the differential diagnosis of white, nonremovable lesions on the lateral border of the tongue in HIV-infected patients, the present study suggests that persistence of lesions after local antifungal therapy is highly suggestive of HL.

Detection of human papillomavirus and Epstein-Barr virus DNA sequences in oral mucosa of HIV-infected patients by the polymerase chain reaction.

The presence of human papillomavirus (HPV) and Epstein-Barr virus (EBV) was analyzed in 21 oral biopsy specimens of HIV-infected patients using the polymerase chain reaction (PCR) method. Biopsies were categorized as hairy leukoplakia (HL) (n = 12), candidiasis (n = 3), oral warts (n = 2), and clinically normal epithelium (n = 4). For HPV detection a modified general primer-mediated PCR method (GP-PCR), which detects a broad spectrum of HPV genotypes at sub-picogram levels, was used. Human papillomavirus DNA was only found in two oral warts and was identified as HPV type 32. Epstein-Barr virus DNA was detected in 16 biopsy specimens, including the 12 HLs, 2 cases of candidiasis, and 2 samples of normal epithelium. Epstein-Barr virus positivity in HL could be confirmed by Southern blot analysis and DNA in situ hybridization using biotinylated DNA probes (bio-DISH). Epstein-Barr virus bio-DISH was also positive in one sample of normal epithelium from a patient with HL. The results indicate that HL is strongly associated with EBV and not with any of the common HPV types that react with general HPV primers in the PCR. However the detection of EBV in normal oral epithelium by PCR and bio-DISH suggests that the presence of this virus is not exclusively related to HL.

Simultaneous enumeration of T-cell subsets and macrophages in bronchoalveolar lavage fluids by immunoenzyme double staining. Comparison with conventional immunofluorescence.

Bronchoalveolar lavage is a common research and clinical tool for the retrieval of cells from the lower respiratory tract. In addition to conventional morphologic study of these cells, the subtyping of T lymphocytes is often important for reaching a diagnosis of a disease or assessing its activity; subtyping is usually done by a standard immunofluorescence assay on cell suspensions requiring about 5 X 10(5) cells. Since the number of leukocytes in the lavage fluid from many patients is too small to obtain reliable information by this assay, a double immunoenzyme staining of T-lymphocyte subtypes on Cytospin preparations was utilized. This method, which requires a small number of cells, was compared with the standard immunofluorescence assay for the identification and quantitation of lymphocyte subtypes in the lavage fluids of patients with different disorders. Although the immunoenzyme double staining assay is somewhat more laborious, it provides important advantages: (1) simultaneous observation of two lymphocyte subsets and macrophages on the same slide; (2) a considerably smaller number of cells (2 X 10(4) instead of 5 X 10(5] is necessary; (3) the availability of permanent preparations; (4) the possibility of storing the Cytospin slides before staining; and (5) conventional light microscopy can be used. Since the reliability of both techniques appeared to be the same, the double staining assay for routine usage with bronchoalveolar lavage fluids appears to be preferable.

Combined immuno- and non-radioactive hybridocytochemistry on cells and tissue sections: influence of fixation, enzyme pre-treatment, and choice of chromogen on detection of antigen and DNA sequences.

Conditions for combination of DNA in situ hybridization, using biotinylated DNA probes, with immunohistochemistry were investigated on cryostat sections, cytological preparations, and paraffin sections. We found that cryostat sections and cytological preparations are suitable for in situ hybridization of target DNA after fixation in acetone, methanol, ethanol, or Carnoy without further proteinase pretreatment. Acetone is also very suitable for immunostaining of cell surface or cytoskeleton antigens. We therefore performed combined immunoenzyme and in situ hybridization staining using this fixative. The best results were obtained when immunoperoxidase staining with diaminobenzidine/H2O2 was followed directly by in situ hybridization. In addition to immunoperoxidase, alkaline phosphatase-antialkaline phosphatase (APAAP) staining with naphthol ASBI phosphate and New Fuchsin as a substrate could be used. In most instances, detection of the biotinylated hybrid with a streptavidin-biotinylated polyalkaline phosphatase method using nitroblue tetrazolium and 5-bromo-4-chloro-3-indolylphosphate as the substrate was preferable. The double stainings were studied on the following test models: (a) frozen tonsil sections: cell surface antigens (pan T) and ribosomal DNA; (b) frozen genital condyloma sections; cytokeratins and human papillomavirus type 6 + 11 (HPV-6/11) DNA; (c) CaSKi cells: cytokeratins and HPV-16 DNA; (d) infected fetal lung fibroblasts: vimentin and cytomegalovirus (CMV) DNA. An adapted procedure was followed on routinely formaldehye-fixed and paraffin-embedded condyloma tissue. Immunoperoxidase staining for papilloma virus capsid antigen could be combined with DNA in situ hybridization with HPV-6/11 DNA. In this model, however, the accessibility of the target DNA had to be improved by enzyme treatment after the immunostaining and before starting the in situ hybridization.

Immunohistochemistry in bone marrow diagnosis. Value of a panel of monoclonal antibodies on routinely processed bone marrow biopsies.

To see if immunohistochemistry can be used on routinely processed bone marrow biopsies for diagnostic purposes, 73 biopsy specimens, fixed in sublimate-formaldehyde, decalcified in an acetic acid-formaldehyde mixture, and embedded in paraffin, were studied with a panel of antibodies. The specimens included "normal," lymphomatous, and myeloproliferative disorders as well as some biopsies with metastatic carcinoma. The results show that the different cell lines and their localization in the bone marrow can be easily identified and quantitative and qualitative changes can be assessed. Megakaryopoiesis is identified with Factor VIII-related antigen and Ulex europaeus agglutinin (UEA); myelopoiesis stains with MT-1, elastase, Leu M-1, LN-2, LN-3, HECA 452, and 115D8; and in myeloproliferative conditions, myeloblasts and promyelocytes stained with leukocyte common antigen (LCA). Erythroid cells stained with UEA, glycophorin A, and LN-1. Lymphocytes were marked with LCA, MB-2, and LN-2. Plasma cells were stained best with immunoglobulin light chain antisera; only occasional reactivity with LCA and 115D8 was observed. Carcinomas all reacted with MB-2; occasional reactivity with 115D8, HECA 452, LN-1, LN-2, MT-1, and Ber H-2 was seen. A small panel of selected antibodies, such as UEA, Leu M-1, and LCA, and the immunoglobulin light chain antisera can be very helpful in bone marrow diagnosis and would cover most indications.

Two colour DNA in situ hybridization for the detection of two viral genomes using non-radioactive probes.

Methods for the simultaneous detection of two virus types in cytological preparations or tissue sections by non-radioactive in situ hybridization were investigated. As a model system, CaSki cells, which have human papilloma virus type 16 (HPV16) DNA integrated in their cellular genome, were in vitro infected with Herpes simplex virus 2 (HSV2). DNA probes for both viruses were labeled with biotin, acetylaminofluorene (AAF), and transaminated-sulfonated cytosine (TS-modified). Best results were obtained when a mixture of biotinated and haptenized DNA probes (AAF- or TS-modified) was used for hybridization. The biotinated hybrid was demonstrated with a streptavidin-biotinated alkaline phosphatase staining reaction, whereas the haptenized hybrid was visualized by an indirect peroxidase method. Visualisation of both viral DNAs in the same cell was possible by a combination of biotinated HPV16 DNA and haptenized HSV2 DNA.

The expression of intermediate filaments and mam-6 antigen in relation to the degree of morphologic differentiation of carcinoma of the head and neck: diagnostic implications.

In this study, the immunoreactivity of several cytokeratin antibodies; 115 D8, a monoclonal antibody against MAM-6, an epithelial membrane antigen; and two vimentin antibodies, is examined in relation to the degree of morphologic differentiation in carcinomas of the head and neck. The results indicate that a relationship exists between the degree of morphologic differentiation and the expression of cytokeratin, MAM-6, and vimentin, as detected by polyclonal antikeratin, 115 D8 and anti-vimentin. Expression of cytokeratin and MAM-6 is reversely related to vimentin. Polyclonal anti-keratin; CAM 5.2, a monoclonal antibody against cytokeratin 8, 18 and 19; and 115 D8, used in combination, were still able to identify the epithelial nature of undifferentiated/spindle cells. Since these immunohistochemical markers precede light microscopic detectable signs of epithelial differentiation, they can be used for the identification of the epithelial nature of undifferentiated/spindle tumors of the head and neck.

Sensitivity of in situ detection with biotinylated probes of human papilloma virus type 16 DNA in frozen tissue sections of squamous cell carcinomas of the cervix.

The sensitivity of human papilloma virus type 16 (HPV-16) DNA detection by DNA in situ hybridization using biotinylated probes (bio-DISH) was estimated by performing this technique on snap-frozen tissue sections of 10 cervical squamous cell carcinomas containing increasing amounts of HPV-16 as determined by Southern blot hybridization. A protocol using serial sections for bio-DISH and DNA extraction was used. The number of positively stained cells and the detection limit were strongly dependent on the treatment of the sections with proteinase K prior to hybridization. At low proteinase K concentration (0.1 micrograms/ml), the detection limit appeared to be 30-40 HPV-16 DNA copies per carcinoma cell, whereas morphology was preserved. A high proteinase K concentration (1-5 micrograms/ml) often resulted in an increase in the number of positively stained cells but also in a poor morphology. The detection limit was improved to at least 20 HPV-16 DNA copies per carcinoma cell.

Expression of intermediate filament proteins in adrenal cortex and related tumours.

The intermediate filament profile of adrenal cortex and its related tumours has been evaluated. Most adrenocortical cells contained cytokeratin 8 and 18 as demonstrated by monoclonal antibodies CAM 5.2, M20, M9 and RGE53. Cytokeratin immunoreactivity was not confined to a functional zone of the adrenal cortex. Only a small number of the adrenocortical cells showed vimentin immunoreactivity. From normal adrenal cortex through adenomas, to carcinomas, there is a progressive decrease or even loss of cytokeratin immunoreactivity and an increase in vimentin immunoreactivity. Aberrant cytokeratin expression was not found in adrenocortical adenomas and carcinomas with the antibodies used. Awareness of the possible absence of cytokeratin immunoreactivity in adrenocortical carcinomas is important whenever antibodies to cytokeratins and vimentin are used for diagnostic purposes in poorly differentiated neoplasms.