1 204 9p21 Deletion in the Diagnosis of Malignant Mesothelioma in Serous Effusions Additional to Immunocytochemistry, DNA-ICM, and AgNOR Analysis Fabiana Botelho de Miranda Onofre, MSc 1 Alexandre Sherlley Casimiro Onofre, MSc 1 Natalia Pomjanski, MD, MIAC 1 Birgit Buckstegge 1 Hans Juergen Grote, MD, MIAC 1,2 Alfred Böcking, MD, FIAC 1 1 Institute of Cytopathology. Heinrich-Heine University, Dusseldorf, Germany. 2 Target Research, Merck KGaA, Darmstadt, Germany. BACKGROUND. The diagnosis of malignant mesothelioma (MM) in serous effusions is difficult but may be achieved by the application of adjuvant methods. METHODS. The authors cytologically diagnosed 33 effusions as suspicious or positive for MM cells by using DNA-image cytometry (DNA-ICM), immunocytochemistry and AgNOR analysis. The authors further detected 9p21 deletions by chromosomal fluorescence in situ hybridization (FISH). In addition, 31 cases of metastatic carcinomas and 39 of tumor cell-negative effusions were investigated. All diagnoses were confirmed by histologic and/or clinical follow-up. RESULTS. DNA aneuploidy was found in 71% of MMs, 100% of metastatic carcinomas, and in none of the negative effusions. Calretinin was positive in 100% of MMs, in none of the metastatic carcinomas, and in 94.9% of negative effusions. BerEP4 showed positivity in 15.6% of MMs, 87.1% of metastatic carcinomas, and in none of the negative effusions. With AgNOR analysis, 89.3% of MMs and 96.7% of metastatic carcinomas showed 2.5 AgNOR dots as satellites and 4.5 as total AgNOR counts. 9p21 deletions were demonstrated in 90.9% of MM cases, 45.2% of metastatic carcinomas, and in none of the negative effusions. By cytology alone, 81.8% of MMs were identified unequivocally. Addition of DNA-ICM improved the prevalence of tumor cell detection to 87.9% and of AgNOR analysis to 97%. The introduction of 9p21 deletions by FISH improved this prevalence to 100%. CONCLUSIONS. Because of these results, the authors propose the sequential application of immunocytochemistry, DNA-ICM, and AgNOR analysis to establish a cytological diagnosis of malignant mesothelioma in serous effusions. In persistent doubtful diagnoses, the authors recommend fluorescence in situ hybridization to analyze the 9p21 deletion. Cancer (Cancer Cytopathol) 2008;114: Ó 2008 American Cancer Society. Coordenaç~ao de Aperfeiçoamento de Pessoal de N ivel Superior grant (F.B.M. Onofre and A.S.C. Onofre). Address for reprints: Alfred Böcking, MD, FIAC, Institute of Cytopathology, Heinrich-Heine University D usseldorf, Moorenstrasse 5, D D usseldorf, Germany; Fax: ; Received October 23, 2007; revision received December 21, 2007; accepted January 25, KEYWORDS: 9p21, CDKN2A, fluorescence in situ hybridization, DNA-ICM, AgNOR analysis, immunocytochemistry, mesothelioma, effusion. Malignant mesothelioma (MM) is a primary aggressive tumor that affects the mesothelial lining of serosal membranes, such as those of the pleural, peritoneal, and pericardial cavities. 1 Exposure to asbestos is considered to contribute to the development of this tumor in the majority of MM patients. 2 5 Studies predict an increasing incidence of this tumor in the next decade. 6 9 The diagnosis of MM is initially based on cytologic examination of serous effusions because of its high accuracy, time effectiveness, and cost effectiveness. However, the diagnosis of MM in serous effusion is difficult and represents one of the classic diagnostic challenges of Ó2008 American Cancer Society DOI /cncr Published online 27 February 2008 in Wiley InterScience (www.interscience.wiley.com).
2 9p21 Deletion and Adjuvant Methods in MM/Onofre et al. 205 cytopathology. 10 The morphological features are not always reliable for differentiation between neoplastic and non-neoplastic mesothelial cells or between mesothelioma cells and those from metastatic adenocarcinomas. 11 Recent advances in therapy for patients with MM can result in an improved outcome if they are applied to stage I disease. 12 Furthermore, from a legal viewpoint, compensation claims of workers who have been occupationally exposed to asbestos demand an accurate and early diagnosis of MM, and unnecessary invasive diagnostic procedures should be avoided. To improve diagnostic sensitivity, various additional approaches have been proposed including DNA-image cytometry (DNA-ICM), immunocytochemistry, and silver staining of nucleolar organizer region-associated proteins (AgNORs). DNA-ICM is a quantitative adjuvant method that identifies DNA stemlines outside the euploid (diploid, tetraploid, or octaploid) regions as abnormal, named DNA aneuploidy, which is used as a sensitive marker for the identification of neoplastic cells. 13,14 A specific immunocytochemical marker for mesotheliomas has not yet been recognized; therefore, panels of markers have been composed of both negative and positive mesothelioma markers. 15 Nucleolar organizer regions (NORs) are large rdna loops that are responsible for the transcription of rrna. 16 AgNOR can be considered as a marker of cell proliferation and could also help distinguish benign from malignant lesions. 16,17 The use of fluorescent in situ hybridization (FISH) for diagnostic purposes has increased significantly in the last few years, primarily because it provides in situ information about genetic or chromosomal changes and because it is applicable to archival and fresh material. 12 Cytogenetic and molecular studies have identified several frequent genetic alterations in mesothelioma, 18 of which one of the most common is homozygous deletion of the 9p21 locus within a cluster of genes (CDKN2B, CDKN2A, and MTAP). 12,19 23 CDKN2A encodes 2 important cell-cycle regulatory proteins, the p16 protein and, in an alternative reading frame, the p14arf protein. p16, a cyclin-dependent kinase inhibitor, acts through CDK4/CDK6 and blocks the phosphorylation of the RB protein, and p14arf binds MDM2, thus preventing the latter from binding p53 and targeting it for degradation. 19,24 The detection of homozygous CDKN2A deletion by FISH would be helpful in confirming a diagnosis of MM instead of reactive mesothelial cells. 12 The aim of this study was to find the prevalence of heterozygous and homozygous deletions at chromosome region 9p21 in abnormal mesothelial cells in effusions of the body cavities by using FISH to differentiate MM from reactive mesothelial cells and these from those of metastatic carcinomas. We, furthermore, wanted to analyze the diagnostic usefulness of cytomorphology, DNA-ICM, immunocytochemistry, AgNOR analysis, and chromosomal FISH for the diagnosis of malignant mesothelioma in serous effusions. MATERIALS AND METHODS Patient Selection Between March 2005 and May 2007, 48 body-cavity effusion specimens from 47 patients were positive or suspicious for MM after routine investigation by conventional cytology, DNA-ICM, immunocytochemistry and AgNOR analysis at the Institute of Cytopathology of the Heinrich-Heine University. The patients were from the University Hospital of Dusseldorf and from hospitals in the surrounding area. A total of 33 effusions (30 pleural and 3 peritoneal) taken from the routine files were analyzed in the current study (28 men and 5 women; mean age, 70 years; range, 47 years to 88 years), in which 6 were cytologically suspicious and 27 positive for mesothelioma. When FISH could not be performed because of a small number of atypical cells, because of the presence of atypical cells in clusters only, or because of unsuitable hybridization, 15 cases were excluded. In addition to MM cases, effusions from 31 metastatic carcinomas (20 pleural and 11 peritoneal; 7 men and 24 women; mean age, 66 years; age range, 43 years to 87 years) and 40 effusions with reactive mesothelial cells (36 pleural and 4 peritoneal; 24 men and 16 women; mean age 67 years; age range, 21 years to 91 years) were opportunistically obtained for subsequent applications of adjuvant methods such as DNA-ICM, immunocytochemistry, AgNOR analysis, and FISH. Cytological Procedure Routinely, 8 slides per native effusion were processed. Three of them were air dried and stained according to May-Grunwald-Giemsa (MGG). Five further slides were immediately fixed in Delaunay fixative and stained according to Papanicolaou (PAP). Two MGG slides were used for Feulgen and silver staining, and PAP slides were used for immunocytochemistry. Slides were evaluated according to generally accepted diagnostic criteria 10,25 by 2 experienced cytopathologists (authors N.P. and A.B.). DNA Image Cytometry After morphologic investigation, the MGG slides were uncovered in xylene and restained according to the
3 206 CANCER (CANCER CYTOPATHOLOGY) June 25, 2008 / Volume 114 / Number 3 method described by Feulgen. 26 Measurements of nuclear DNA contents were performed as previously described. 14 We used a computer-based image analysis system consisting of a Zeiss Axioplan 2 microscope (Zeiss, Jena, Germany) with a 403 objective (numeric aperture, 0.75; Kohler illumination) and a charge-coupled device black-and-white video camera with 572 lines of resolution (VariCam CCIR; PCO Computer Optics, Kehlheim, Germany). The software package used in the current study was the AutoCyte QUIC-DNA-Workstation (AutoCyte, Burlington, NC), which provides shading and glare correction. The latter was performed at a rate of 2.2%. In each case, at least 30 lymphocytes were measured as internal reference cells. The coefficient of variation of reference cells was always below 5%. 27 A minimum of 300 chosen nuclei of interest (reactive or atypical mesothelial or carcinoma cells) were randomly measured per specimen. All technical instruments and all software used met the standard requirements of the consensus reports of the European Society for Analytical Cellular Pathology (ESACP). 13,27,28,29 Several parameters were assessed for diagnostic interpretation. 13,27,28,29 DNA stemline ploidy was defined as the modal value of a DNA stemline in c units (c 5 DNA content). DNA stemline aneuploidy was assumed if the modal value of a stemline was \1.80c or [2.20c and \3.60c or [4.40c. Single-cell aneuploidy was diagnosed when at least 1 cell per slide had DNA content [9c (9cEE [1). 27 Cells between 5c and 8c occur in 6.5% of mesothelial cells in tumor cell-negative effusions. 30 This is the reason why the threshold for the detection of rare aneuploid cells had to be set at 9c and not at 5c. 31,32 We used 2 algorithms for the identification of DNA aneuploidy: abnormal position of any DNA stemline and occurrence of cells [9c. Immunocytochemistry After morphologic investigation, cells of interest were marked on PAP slides by felt-tip pen. Then, coverslips were removed in xylene at room temperature (RT). The coverslips fell off within 24 hours. All steps were performed according to our previous study. 31 Apart from the finding that all of our antibodies were originally tested with tumor-positive and tumor-negative effusions, we, because of scarcity of smears, did not routinely apply positive and negative controls on separate slides. Normal macrophages, lymphocytes, and granulocytes were usually used for internal negative control. Incubations were performed with commercially available monoclonal primary antibodies (Table 1) followed by a biotinylated link antibody and the avidin-biotin complex method of Elite TABLE 1 Antibodies, Clones, Dilutions, Pretreatments and Providers Antibody Clone Dilution Pretreatment Provider BerEP4 BerEP4(1) 1:200 None Dako Calretinin DAK-Calret 1 1:200 None Dako EMA E-29 1:1600 None Dako Mesothelin NCL Meso 1:200 None Novocastra WT-1 6F-H2 1:200 None Dako BerEP4 indicates surface and cytoplasmic glycoprotein; EMA, epithelial membrane antigen; WT-1, Wilms tumor protein 1. Standard (Vector, Burlingame, Calif) for the observation of immunologic reactions. We applied BerEP4 and calretinin in all effusions from MM, metastatic carcinoma, and reactive mesothelial cells. When there were enough slides from morphologically suspicious cases of MM, we also applied epithelial membrane antigen (28 cases), mesothelin (11 cases) and WT-1 (Wilms Tumor protein 1) (15 cases). Antibody reactions were evaluated per case, without knowledge of patient s follow-up in order to avoid any bias. To exclude false-positive results, artificial or unspecific staining cases that stained 10% of cells or showed diffuse weak staining were considered negative. Results were obtained as percentage per slide. AgNOR Analysis Silver staining was performed according to the 1-step method of Ploton et al., 33 Crocker et al., 34 and Rüschoff 35 with some modifications. Routine MGG-slides from effusions were uncovered in xylene. All steps were performed according to our previous study. 36 AgNOR counting was performed on 100 cells for each cytologic slide. These were examined at a magnification under oil immersion. Unequivocal benign cells were not counted in malignant effusions, and only reactive mesothelial cells were counted in benign effusion samples. To standardize the counting, we followed the Crocker method 34 with minor modifications as follows: first, silver-stained dot aggregations or partly disaggregated nucleoli (clusters) treated as 1 structure were counted. Second, individual dots (satellites) outside the clusters of silver-stained structures were counted. Third, clusters and satellites were counted together to obtain the total AgNORs counts. The mean number per nucleus of AgNORs as clusters, as satellites, and as total AgNOR counts (clusters plus satellites) was calculated in each case. For a correct tumor cell identification and to avoid overlap between reactive and malignant cells in effusions, we applied the threshold
4 9p21 Deletion and Adjuvant Methods in MM/Onofre et al. 207 of 2.5 AgNOR dots as satellites and 4.5 as total AgNOR counts as published in our previous study. 36 Values greater than or equal to this threshold were considered to be AgNOR positive. Fluorescence in situ Hybridization For FISH analysis, additional cytospin slides were made from each specimen and stained according to PAP to verify the presence of mesothelial or atypical cells. In each slide, the LSI p16/cep 9 dual-color probe was performed according to the recommendations of the manufacturer (Abbot/Vysis, Downers Grove, Ill) with minor modifications. Briefly, slides were uncovered in xylene, rehydrated through 2 series of 100%, 95%, and 80% ethanol, placed under running tap water for 5 minutes, then put in 0.5% HCl in 70% ethanol for 15 minutes at room temperature. After immersion in 2X saline sodium citrate (SSC) for 5 minutes at 738C, slides were digested by using 0.2mg/mL pepsin in 0.01 mol/l HCl for 15 minutes at 378C in a humidified chamber. The slides were washed in phosphate-buffered saline (PBS) for 5 minutes at room temperature, fixed in 1% neutralbuffered formalin/pbs for 5 minutes, washed in PBS again for 5 minutes, and air dried at room temperature. The FISH probe mix (7 ll LSI/WCP hybridization buffer, 2 ll purified water, and 1 ll 9p21/CEP 9 dual-color probe) was applied, cover slipped, and sealed with rubber cement (Q Biogene, Montreal, Canada). After denaturation at 738C for 10 minutes, slides were incubated at 378C overnight in a humidified chamber. After hybridization, the samples were washed in 0.4X SSC/0.1% NP-40 at 738C and in 0.4X SSC/0.1% NP-40 at room temperature for 2 minutes each. Then, 4.6-diamidine-2-2phenylindole dihydrochloride (DAPI) (Abbot/Vysis, Downers Grove, Ill) was used for counterstaining. The slides were scored on a cell-by-cell basis by using a Zeiss Axio Imager M1 fluorescence microscope (Zeiss, Gottingen, Germany) with a singleband pass filter for DAPI, SpectrumGreen (chromosome 9), and SpectrumOrange (9p21 locus). It contains a charge-coupled device black-and-white video camera with 1.4 Megapixel (AxioCam MRm, München-Hallbergmoos, Germany), and it is equipped with the Axio Vision QuantiFISH software (Zeiss, Hallbergmoos, Germany). Signal counting was performed by an experienced observer with FISH analysis (author F.O.). A minimum of 100 cells per case was counted manually under a 1003 planar objective. Inflammatory cells and macrophages were not counted. Two staining colors, orange for 9p21 locus and green for CEP 9, were simultaneously counted in a given nucleus. In each specimen, polymorphonuclear cells and lymphocytes served as internal controls, and hybridization efficiency was evaluated in these cells. Homozygous deletion was defined as the absence of both 9p21 signals in the presence of at least 1 chromosome 9 centromere signal. Heterozygous deletion was defined as the presence of only 1 9p21 signal or when the number of 9p21 signals was lower than the number of chromosome 9 centromere signals. Follow-up According to patient s histological and/or clinical follow-up, the investigated effusions were classified as either containing malignant cells or not and by what kind of tumor was present when the patient presented with a malignancy. Clinical evidence was considered valid, applying such diagnostic techniques as radiology and/or computed tomography. RESULTS From 33 cases cytologically diagnosed as suspicious or positive for MM, 2 cases were re-evaluated after routine application of immunocytochemistry, DNA- ICM, and AgNOR analysis (Fig. 1). One case, primarily reported as suspicious for MM, was revised as negative for tumor cells after routine application of those adjuvant methods, but histologic and clinical follow-up confirmed MM. The other case was rediagnosed as metastatic nonsmall cell carcinoma after application of those methods, but it was finally confirmed as MM by histologic follow-up. Another case diagnosed as suspicious for MM continued to be suspicious even after the application of adjuvant methods, which was confirmed as MM after histologic follow-up. One case cytologically reported as negative for tumor cells was not corroborated by histologic follow-up, which revealed a pleural involvement of a non-hodgkin lymphoma. In consequence, the additional effusions were resumed as 31 cases with metastatic carcinomas and 39 with reactive mesothelial cells. The most often found etiologies of negative effusions were pulmonary emphysema, pneumonia, congestive heart failure, coronary disease, renal insufficiency, cirrhosis of the liver, rheumatic fever, chronic inflammation, and foreign substances like talc. The primary tumors metastasized to the pleura were localized in the breasts (12 cases), the lungs (3 cases), the ovaries (3 cases), the stomach (1 case), and the endometrium (1 case). In peritoneal carcinosis, the primary tumors came from metastatic carcinomas of the ovary (6 cases), the stomach (2 cases),
5 208 CANCER (CANCER CYTOPATHOLOGY) June 25, 2008 / Volume 114 / Number 3 FIGURE 1. Flow chart illustrating the diagnostic results. MM: malignant mesothelioma; CA: carcinoma. TABLE 2 Prevalence of DNA Aneuploidy in Tumor-Cell Positive and Negative Effusions DNA ploidy status Without tumor cells n 5 39 (%) Follow-up diagnosis Total with tumor cells n 5 62 (%) Carcinoma n 5 31 (%) Mesothelioma n 5 31 (%) Euploidy 39 (100.0) 9 (14.5) 0 (0.0) 9 (29.0) Aneuploidy 0 (0.0) 53 (85.5) 31 (100.0) 22 (71.0) the pancreas (1 case), the endometrium (1 case), and the breast (1 case). DNA Image Cytometry None of the DNA histograms obtained from tumor cell-negative effusions revealed any of the mentioned parameters of DNA aneuploidy. This corresponds to a prevalence of 100% of the DNA euploidy in reactive mesothelial cells (Table 2). The prevalence of DNA aneuploidy was 100% for metastatic carcinomas and 71% for MMs (Table 2). In 2 cases of MM, DNA-ICM was not performed because of scarcity of cells. The prevalence of DNA aneuploidy in effusions depended on the algorithm applied (see table 3). Whereas 61.3% (19 of 31) of MMs showed their greatest DNA stemline within the range of 1.8c and 2.2c and/or 3.6c and 4.4c, only 19.4% (6 of 31) of metastatic carcinomas showed their greatest stemline in this region. Immunocytochemistry The immunocytochemical results are summarized in Table 4. All 32 cases of MM demonstrated positivity with calretinin (Fig. 2a). From 2 effusions from the same patient, immunocytochemistry was performed in only 1 case. BerEp4 was negative in 84.4% cases of MM (Fig. 2b). Only 2 cases demonstrated strong reaction ([80%) with BerEP4, in which calretinin demonstrated strong reaction in tumor cells also. Epithelial membrane antigen and mesothelin showed positivity in the majority of cases of MM and WT-1 in 66.7%. Calretinin was negative in all cases of effusions that contained metastatic carcinomas. In 87.1% of those cases, BerEP4 showed positivity. In 2 cases of metastatic carcinoma of the breast, 1 of the ovary, and 1 of the stomach, calretinin and BerEP4 were negative. The majority of negative cases with reactive mesothelial cells demonstrated positivity with calretinin. BerEP4 was completely negative in all of them. AgNOR Analysis AgNORs were strictly located only within nuclei and were clearly visible as distinct black or brown dots. We could not perform AgNOR counting in 10 cases (4 negatives, 1 metastatic carcinoma, and 5 MMs) because of the small number of cells, overlap of cells, or technical staining problems as hyperstaining or hypostaining with silver nitrate. By using receiver-operator characteristic (ROC) curves (Figs. 3 and 4), we confirmed the statement of our previous study 36 by applying the threshold of 2.5 AgNOR dots as satellites and 4.5 as total AgNOR counts. The AgNOR dots were discrete and smaller in benign effusion cases (Fig. 5a) compared with their coarse and aggregated appearance in malignant effusions (Fig. 5b). The AgNOR analysis results are summarized in Table 5. AgNOR analysis was positive in 89.3% of MM cases. In 1 case, the AgNOR analysis showed 2.16 and 3.16 (satellites/total AgNOR counts); this case was initially reported as suspicious for MM, and, after application of adjuvant methods, it was reevaluated as negative, but histologic and clinical follow-up confirmed MM. Two doubtful cases were found with 2.30 and 4.92 and 3.04 and 4.48 (satellites/total AgNOR counts). In MM cases, the mean
6 9p21 Deletion and Adjuvant Methods in MM/Onofre et al. 209 TABLE 3 Prevalence of DNA Aneuploidy in Effusions According to the Application of Different Algorithms DNA Aneuploidy Cytology No. of cases Single-cell aneuploidy (%) STL Aneuploidy (%) Single-cell and STL aneuploidy (%) Malignant mesothelioma (86.4) 12 (54.5) 9 (40.9) Metastatic carcinomas (83.9) 25 (80.6) 20 (64.5) Total with tumor cell (84.9) 37 (69.8) 29 (54.7) DNA-ICM indicates DNA-image cytometry; STL, stemline; single-cell aneuploidy, at least 1 cell with DNA content [9c. TABLE 4 Immunocytochemical Findings in Malignant Mesotheliomas, Metastatic Carcinomas and Reactive Effusions Grading of positive reactivity Markers No. of cases Negative (%) Positive (%) 11% 39% 40% 79% 80% 100% Malignant mesothelioma BerEP (84.4) 5 (15.6) Calretinin 32 0 (0.0) 32 (100.0) EMA 28 2 (7.1) 26 (92.9) Mesothelin 11 1 (9.1) 10 (90.9) WT (33.3) 10 (66.7) Metastatic carcinoma BerEP (12.9) 27 (87.1) Calretinin (100.0) 0 (0.0) Reactive effusion BerEP (100.0) 0 (0.0) Calretinin 39 2 (5.1) 37 (94.9) BerEP4 indicates surface and cytoplasmic glycoprotein; EMA, epithelial membrane antigen; WT-1, Wilms tumor protein 1. number of satellites was 4.91 (range, ), and the total AgNOR count was 7.04 (range, ). In 96.7% of metastatic carcinoma cases, AgNOR analysis was positive. In only 1 case of metastatic carcinoma of the breast, the AgNOR analysis was doubtful (2.44 satellites/5.16 total AgNOR counts). In metastatic carcinoma cases, the mean number of satellites was 5.31 (range, ), and the total AgNOR count was 7.04 (range, ). Only 1 case of negative effusion demonstrated an increased number of AgNOR dots (2.5 satellites/5.13 total AgNOR counts). This patient was clinically diagnosed as having lung embolism and pneumonia without indication of malignant cells in effusions. Of negative cases, 97.1% were AgNOR-analysis negative. The mean number of satellites was 1.07 (range, ), and the total AgNOR count was 2.94 (range, ) in reactive mesothelial cells. Fluorescence in situ Hybridization We focused on both homozygous and heterozygous deletion of 9p21 as criteria for positivity because of the relatively low frequency of these events in the normal cell population. To avoid overinterpretation of incomplete hybridization, we considered as positive all specimens containing 5 nuclei showing homozygous deletion of 9p21 or containing 15 nuclei showing heterozygous deletion of 9p21. Prevalence of 9p21 deletion in cases of reactive mesothelial cells, MM, and metastatic carcinoma is summarized in Table 6. FISH was positive in 90.9% of MM cases, in which 48.5% showed homozygous deletion (Fig. 6a), 36.4% heterozygous deletion (Fig. 6b), and 6.0% both. From 31 metastatic carcinomas cases, FISH was positive in 45.2%. Cases with homozygous deletion were carcinoma of the breast, the ovary, the pancreas, and the endometrium. The heterozygous deletion cases were carcinoma of the endometrium (1 case), the lungs (1 case), the stomach (2 cases), the breast (2 cases), and the ovary (2 cases). Two cases of carcinoma of the ovary showed homozygous and heterozygous deletion. No 9p21 deletion was detected in any negative effusion. One case cytologically diagnosed as tumor cell-negative effusion demonstrated positivity of 9p21 homozygous deletion, and the histologic follow-up reported a non-hodgkin lymphoma. Another case was first diagnosed cytologically as suspicious for MM and rediagnosed as negative after DNA-ICM, immunocytochemistry, and AgNOR analysis, but histologic and
7 210 CANCER (CANCER CYTOPATHOLOGY) June 25, 2008 / Volume 114 / Number 3 clinical follow-up confirmed MM. In this case, FISH demonstrated 9p21 heterozygous deletion. Application of Adjuvant Methods In the current study, when cytology was used alone for the diagnosis of MM, 81.8% of cases were reported as tumor cell-positive with certainty. The addition of DNA-ICM improved the prevalence to 87.9% and of AgNORs to 97% (Table 7). Application of FISH improved the prevalence of correct MM diagnosis to 100%, especially when 9p21 homozygous and heterozygous deletion were used as parameters. The combination of calretinin positive and BerEP4 negative was found in 84.4% of MM cases and calretinin negative and BerEP4 positive in 87.1% of metastatic carcinoma. FIGURE 2. Malignant mesothelioma cells are a) stained with calretinin (original magnification, 3400) and b) faintly stained with BerEP4 (original magnification, 3400). DISCUSSION Malignant mesothelioma (MM) is an aggressive neoplasm of the serosal membranes with a poor prognosis. Predictive studies have reported an increased incidence of MM for the next decade. 6 8 An early and precise diagnosis of MM in effusions is crucial for patient management and may avoid unnecessary invasive diagnostic procedures. Cytological diagnosis of MM is difficult but may be achieved by the application of adjuvant methods. DNA-ICM is a standardized, quantitative, adjuvant method that measures nuclear DNA content by the cytometric equivalent of nuclear Integrated Optical Density (IOD). 13 None of the DNA histograms measured from negative effusions revealed any of the mentioned parameters of DNA aneuploidy in the FIGURE 3. Receiver-operator characteristic (ROC) curve shows the sensitivity and specificity of AgNOR analysis for tumor cell detection based on the number of total AgNOR counts.
8 9p21 Deletion and Adjuvant Methods in MM/Onofre et al. 211 FIGURE 4. Receiver-operator characteristic (ROC) curve shows the sensitivity and specificity of AgNOR analysis for tumor cell detection based on the number of AgNOR satellites. FIGURE 5. a) Benign reactive mesothelial cells have small and discrete AgNOR dots (clusters) and a low number of satellites (oil immersion; original magnification, 31000), and b) malignant mesothelioma cells show an increased AgNOR number (oil immersion; original magnification, 31000). TABLE 5 Prevalence of AgNOR-Analysis in Tumor Cell-Positive and Negative Effusions Follow-up diagnosis AgNOR-analysis (Satellites/Total AgNOR counts) Without tumor cells n 5 35 (%) Total with tumor cells n 5 58 (%) Carcinoma n 5 30 (%) Mesothelioma n 5 28 (%) Negative (\2.5/\4.5) 34 (97.1) 1 (1.7) 0 (0.0) 1 (3.6) Positive (2.5/4.5) 1 (2.9) 54 (93.1) 29 (96.7) 25 (89.3) Doubtful* 0 (0.0) 3 (5.2) 1 (3.3) 2 (7.1) AgNOR: silver staining of nucleolar organizer regions-associated proteins. * Doubtful represents the cases that satellites or the total AgNOR counts is contradictory (\2.5/[4.5 or [2.5/\4.5). current study. This result is in concordance with the majority of the studies that have analyzed DNA-ICM in effusions. 31,36 39 In contrast, Palaoro et al. 40 found aneuploidy in 3 of 23 (13%) cases of negative effusions, but they did not follow the standard requirements reported by the European Society for Analytical Cellular Pathology (ESACP), which possibly contributed to these false-positive results. 13,14,27,28 In our study, all cases of metastatic carcinomas were found to be DNA aneuploid. The prevalence of
9 212 CANCER (CANCER CYTOPATHOLOGY) June 25, 2008 / Volume 114 / Number 3 TABLE 6 Prevalence of 9p21 Deletion in Cases of Reactive Mesothelial Cells, Malignant Mesothelioma and Metastatic Carcinoma 9p21 Deletion (%) Type of FISH positivity (%) Cytology No. of cases FISH Negative FISH Positive Homo deletion Hetero deletion Homo & Hetero deletion Reactive mesothelial cell (100) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Malignant mesothelioma 33 3 (9.1) 30 (90.9) 16 (48.5) 12 (36.4) 2 (6.0) Metastatic carcinomas (54.8) 14 (45.2) 4 (12.9) 8 (25.8) 2 (6.5) FISH indicates fluorescence in situ hybridization; Homo, positive homozygous deletion of 9p21 5 nuclei; Hetero, positive heterozygous deletion 15 nuclei. FIGURE 6. Fluorescence in situ hybridization in malignant mesothelioma nuclei shows a) (left) 2 signals for chromosome 9 centromeric probe (CEP-9; Spectrum Green) and no signals for the 9p21 locus probe (LSI p16; Spectrum Orange); b) (right) 2 signals for chromosome 9 centromeric probe and only 1 signal for the 9p21 locus. TABLE 7 Prevalence of Cytology Correlated With Adjuvant Methods in Effusions With Malignant Mesothelioma Cells Cytology1FISH Cytology1DNA-ICM1AgNOR1FISH No. of cases Cytology Cytology1DNA-ICM Cytology1AgNOR Homo Homo & Hetero Homo Homo & Hetero Prevalence, % (81.8) 29 (87.9) 32 (97.0) 30 (90.9) 33 (100.0) 32 (97.0) 33 (100.0) DNA-ICM indicates DNA-image cytometry, DNA-aneuploidy; AgNOR, silver staining of nucleolar organizer regions-associated proteins. 2.5/4.5 (satellites/total AgNOR counts); FISH, fluorescence in situ hybridization; Homo, positive homozygous deletion of 9p21 5 nuclei; Hetero, positive heterozygous deletion 15 nuclei. aneuploidy in cells from metastatic carcinomas in effusions varies between 88.5% and 100%. 31,36 39 DNA aneuploidy was found in 71% of MM cases in the current study. This is in agreement with some reports DNA aneuploidy seems to represent a consistent marker of malignancy; however, euploidy cannot exclude it. Our data confirm the higher proportion of euploidy in MMs compared with metastatic carcinomas, which showed no euploid cases. In MMs, 61.3% showed their greatest stemline within the range of 1.8c and 2.2c and/or 3.6c and 4.4c. This parameter is able to contribute to the differential diagnosis between MM and metastatic carcinoma, which showed only 19.4% in that region. Similar findings have been published. 31,36 In recent years, immunocytochemistry has contributed to the differentiation of mesotheliomas from adenocarcinomas As a specific marker for mesotheliomas has not yet been recognized, panels of markers that frequently are expressed in mesotheliomas combined with those that are commonly expressed in adenocarcinomas are used. 15 Calretinin
10 9p21 Deletion and Adjuvant Methods in MM/Onofre et al. 213 and BerEP4 have been used in panels, using calretinin as marker for mesothelial cells and BerEP4 for carcinoma cells. 36,41,42,44,45 In the current study, calretinin revealed positivity in 100% and BerEP4 in 15.6% of MM cases. These results are in concordance with other authors who have found similar results. 36,42,44 Li et al. 41 and Politi et al. 45 reported calretinin positivity and BerEP4 negativity in all of their cases of MM. Our metastatic carcinoma cases showed BerEP4 positivity in 87.1% and calretinin in none of them. Some studies found calretinin positivity in metastatic carcinomas. 36,42,44 It is suggested that a positive background of benign mesothelial cells could be a potential pitfall in the interpretation of calretinin staining patterns. 41 Epithelial membrane antigen revealed positivity in 92.9% of MM cases. Similar results have been found by other studies. 36,44 Only 1 of our MMs was negative for mesothelin, and WT-1 was positive in 66.7%. Ordonez found positivity of WT-1 between 72% and 93% of MMs. 43,46 AgNOR is a proliferation marker useful in the differential diagnosis of benign and malignant cells. Unfortunately, a definitive standard for AgNOR staining and quantification has not yet been achieved. By using a digital-image analysis system, Wolanski et al. 47 reported a significant overlap in nuclei in AgNOR counting. In the current study, we generated ROC curves that confirmed those thresholds suggested by Pomjanski et al. 36 In concordance with previous studies, 40,48 our study showed that malignant cells mostly exhibit a greater AgNOR protein content compared with corresponding benign cells. AgNOR analysis was positive in 89.3% of MM and 96.7% of metastatic carcinomas. Only 1 case of MM showed AgNOR negativity. AgNOR analysis was negative in 97.1% of negative effusions. Only 1 case of negative effusion demonstrated an increased number of AgNOR dots (2.5 as satellites and 5.13 as total AgNOR counts). The AgNOR analysis appears to be a sensitive method to distinguish between benign and malignant cells in effusions. The occurrence of multiple cytogenetic deletions in MM suggests that the loss and/or the inactivation of tumor-suppressor genes may be critical to the development and progression of this tumor. 19,49 The most common cytogenetic abnormality in MM is deletion at the 9p21 locus, 12,19 23 which can be used as a marker for malignancy in serous effusions. 12 In the current study, 90.9% of MM cases demonstrated 9p21 deletion, of which 54.5% were homozygous and 42.4% heterozygous. This occurrence of deletions was reduced in metastatic carcinomas, which were revealed to be homozygous in 19.4% and heterozygous in 32.3%. No 9p21 deletion was detected in any tumor cell-negative effusion. Previous studies have indicated that the 9p21 homozygous deletion is present in approximately 72% to 75% of MM patients 19,22,50 and up to 100% in mesothelioma cell lines. 23,50 Illei et al. 12 confirmed a diagnosis of MM in 12 of 13 patients with positive or suspicious cytology. In concordance with the current study, they also reported that all 19 cytologically negative specimens were negative for 9p21 deletion. 12 Cheng et al. 23 reported homozygous deletion of 9p21-p22 in 43% and homozygous and/or heterozygous deletion in 83% of MM cell lines. To our knowledge, no study has so far reported the 9p21 deletion in metastatic carcinomas by FISH, as well as 9p21 heterozygous deletions as criteria for FISH-positive in serous effusions. The 9p21 heterozygous deletion is a criterion used for FISH positivity also in other tumor sites, such as bladder cancer in bladder-washing samples. 51 Because of the dependence of the type of preparation and staining, volume examined, investigator experience, and number of sufficient specimens investigated, the reported sensitivity of conventional cytology to serous effusions varies between 50% and 84%. 37,52 The sequential use of different adjuvant methods can contribute to establish a diagnosis in cytologically doubtful and suspicious cases. In the current study, 81.8% of MMs were identified unequivocally by cytology alone, as confirmed by follow-up. The addition of DNA-ICM improved the prevalence of tumor-cell detection to 87.9% and to 97% with the addition of AgNOR analysis. Our results confirm the impact of DNA-ICM and AgNOR to diagnose MM. 31,36 The introduction of FISH could improve the prevalence of tumor-cell detection to 100% if 9p21 homozygous and heterozygous deletions were used as diagnostic parameters. To differentiate metastatic carcinoma from MM or reactive mesothelial cells, the application of calretinin and BerEP4 plays an important role. It is generally agreed that no single antibody is sufficiently sensitive or specific. In our study, the combination of calretinin positivity and BerEP4 negativity was found in 84.4% of MM cases and calretinin negativity and BerEP4 positivity in 87.1% of metastatic carcinomas. The documented low sensitivity of conventional cytology to identify MM cells in effusions indicates the necessity of applying adjuvant diagnostic methods to improve diagnostic accuracy. Following our results, we state that immunocytochemistry (calretinin and BerEP4) should be applied to differentiate
11 214 CANCER (CANCER CYTOPATHOLOGY) June 25, 2008 / Volume 114 / Number 3 epithelial cells from mesothelial cells. If BerEP4 is positive and calretinin negative, the diagnosis should be focused on metastatic carcinoma, applying other immunocytochemical panels if the primary tumor is unknown. 53 If BerEP4 is negative and calretinin is positive, then the diagnosis should be focused on differentiating reactive mesothelial cells from MM. The application of DNA-ICM and AgNOR analysis is recommended to establish the detection of tumor cells. If the results of those methods remain doubtful, then the application of FISH (9p21 deletion) is indicated. In conclusion, to establish a cytological diagnosis of malignant mesothelioma in serous effusions, we propose the sequential application of immunocytochemistry, DNA-ICM, and AgNOR analysis. In persistent doubtful diagnoses, we recommend fluorescence in situ hybridization to analyze the 9p21 deletion. Further studies on larger series of patients are needed to evaluate the validity and efficiency of this approach for improving the diagnostic accuracy of effusion cytology. REFERENCES 1. Churg A, Cagle PT, Roggli VL.Tumors of the serosal membranes. In: Atlas of Tumor Pathology. 4th series, fascicle 10. Washington, DC: Armed Forces Institute of Pathology; Yarborough CM. The risk of mesothelioma from exposure to chrysotile asbestos. Curr Opin Pulm Med. 2007;13: Carbone M, Bedrossian CW. The pathogenesis of mesothelioma. Semin Diagn Pathol. 2006;23: Robinson BW, Musk AW, Lake RA. Malignant mesothelioma. Lancet. 2005;366: Wagner JC, Sleggs CA, Marchand P. Diffuse pleural mesothelioma and asbestos exposure in the North Western Cape Province. Br J Ind Med. 1960;17: Hodgson JT, McElvenny DM, Darnton AJ, Price MJ, Peto J. The expected burden of mesothelioma mortality in Great Britain from 2002 to Br J Cancer. 2005;92: Price B, Ware A. Mesothelioma trends in the United States: An update based on Surveillance, Epidemiology, and End Results Program data for 1973 through Am J Epidemiol. 2004;159: Leigh J, Davidson P, Hendrie L, Berry D. Malignant mesothelioma in Australia, Am J Ind Med. 2002; 41: Peto J, Decarli A, La Vecchia C, Levi F, Negri E. The European mesothelioma epidemic. Br J Cancer. 1999;79: Gray W, Mckee GT. Diagnostic Cytopathology. 2nd ed. Edinburgh: Churchill Livingstone; Bedrossian CW. Diagnostic problems in serous effusions. Diagn Cytopathol. 1998;19: Illei PB, Ladanyi M, Rusch VW, Zakowski M. The use of CDKN2A deletion as a diagnostic marker for malignant mesothelioma in body cavity effusions. Cancer (Cancer Cytopathol). 2003;99: Haroske G, Baak JP, Danielsen H, et al. Fourth updated ESACP consensus report on diagnostic DNA image cytometry. Anal Cell Pathol. 2001;23: Böcking A.DNA measurements. When and why? In: Wied GL,Keebler CM, Rosenthal DL, Schenk U, Somrak TM, Vooijs GP, eds. Compendium on Quality Assurance, Proficiency Testing, and Workload Limitations. Chicago: Tutorials of Cytology; 1995: Ordonez NG. What are the current best immunohistochemical markers for the diagnosis of the epithelioid mesothelioma? A review and update. Hum Pathol. 2007;38: Derenzini M. The AgNORs. Micron. 2000;31: Sirri V; Roussel P; Hernandez-Verdun D. The AgNOR proteins: qualitative and quantitative changes during the cell cycle. Micron. 2000;31: Sandberg AA, Bridge JA. Updates on the cytogenetics and molecular genetics of bone and soft tissue tumors. Mesothelioma. Cancer Genet Cytogenet. 2001;127: Illei PB; Rusch VW, Zakowski MF, Ladanyi M. Homozygous deletion of CDKN2A and codeletion of the methylthioadenosine phosphorylase gene in the majority of pleural mesotheliomas. Clin Cancer Res. 2003;9: Hirao T, Bueno R, Chen CJ, Gordon GJ, Heilig E, Kelsey KT. Alterations of the p16 INK4 locus in human malignant mesothelial tumors. Carcinogenesis. 2002;23: Bjorkqvist AM, Tammilehto L, Anttila S, Mattson K, Knuutila S. Recurrent DNA copy number changes in 1q, 4q, 6q, 9p, 13q, 14q and 22q detected by comparative genomic hybridization in malignant mesothelioma. Br J Cancer. 1997;75: Xiao S, Li D, Vijg J, Sugarbaker DJ, Corson JM, Fletcher JA. Codeletion of p15 and p16 in primary malignant mesothelioma. Oncogene. 1995;11: Cheng JQ, Jhanwar SC, Klein WM, et al. p16 alterations and deletion mapping of 9p21-p22 in malignant mesothelioma. Cancer Res. 1994;54: Stahel RA. Malignant pleural mesothelioma: a new standard of care. Lung Cancer. 2006;545:S9 S Koss LG, Melamed MR.Koss diagnostic cytology and its histopathologic bases. Volumes I and II. 5th ed. Philadelphia: Lippincott Williams & Wilkins; Chatelain R, Willms A, Biesterfeld S, Auffermann W, Bocking A. Automated Feulgen staining with a temperature controlled staining machine. Anal Quant Cytol Histol. 1989;11: Haroske G, Giroud F, Reith A, Böcking A ESACP consensus report on diagnostic DNA image cytometry. Part I: basic considerations and recommendations for preparation, measurement and interpretation. European Society for Analytical Cellular Pathology. Anal Cell Pathol. 1998;17: Giroud F, Haroske G, Reith A, Böcking A ESACP consensus report on diagnostic DNA image cytometry. Part II: Specific recommendations for quality assurance. European Society for Analytical Cellular Pathology. Anal Cell Pathol. 1998;17: Böcking A, Giroud F, Reith A. Consensus report of the ESACP task force on standardization of diagnostic DNA image cytometry. European Society for Analytical Cellular Pathology. Anal Cell Pathol. 1995;8: Biesterfeld S, Gerres K, Fischer-Wein G, Böcking A. Polyploidy in non-neoplastic tissues. J Clin Pathol. 1994;47:38 42.
12 9p21 Deletion and Adjuvant Methods in MM/Onofre et al Motherby H, Kube M, Friedrichs N, et al. Immunocytochemistry and DNA-image cytometry in diagnostic effusion cytology. I. Prevalence of markers in tumor cell positive and negative smears. Anal Cell Pathol. 1999;19: Motherby H, Marcy T, Hecker M, et al. Static DNA cytometry as a diagnostic aid in effusion cytology. I. DNA aneuploidy for identification and differentiation of primary and secondary tumor of the serous membranes. Analyt Quant Cytol Histol. 1998;20: Ploton D, Menager M, Jeannesson P, Himber G, Pigeon F, Adnet JJ. Improvement in the staining and in the visualization of the argyrophilic proteins of the nucleolar organizer region at the optical level. Histochem J. 1986;18: Crocker J, Boldy DA, Egan MJ. How should we count AgNORs? Proposals for a standardized approach. J Pathol. 1989;158: Rüschoff J. Nukleolus Organisierende Regionen (NORs) in der Pathomorphologischen Tumordiagnostik. Stuttgart: Gustav Fischer Verlag; Pomjanski N, Motherby H, Buckstegge B, Knops K, Rohn BL, Böcking A. Early diagnosis of mesothelioma in serous effusions using AgNOR analysis. Analyt Quant Cytol Histol. 2001;23: Osterheld MC, Liette C, Anca M. Image Cytometry: an aid for cytological diagnosis of pleural effusions. Diagn Cytopathol. 2005;32: Motherby H, Pomjanski N, Kube M, et al. Diagnostic DNAflow vs. image-cytometry in effusion cytology. Anal Cell Pathol. 2002;24: Kayser K, Blum S, Beyer M, Haroske G, Kunze KD, Meyer W. Routine DNA cytometry of benign and malignant pleural effusions by means of the remote quantification server Euroquant: a prospective study. J Clin Pathol. 2000;53: Palaoro LA, Blanco AM, Gamboni M, Rocher AE, Rotenberg RG. Usefulness of ploidy, AgNOR and immunocytochemistry for differentiating benign and malignant cells in serous effusions. Cytopathology. 2007;18: Li Q, Bavikatty N, Michael CW. The role of immunohistochemistry in distinguishing squamous cell carcinoma from mesothelioma and adenocarcinoma in pleural effusion. Semin Diagn Pathol. 2006;23: Ordonez NG. The diagnostic utility of immunohistochemistry and electron microscopy in distinguishing between peritoneal mesotheliomas and serous carcinomas: a comparative study. Mod Pathol. 2006;19: Ordonez NG. The immunohistochemical diagnosis of mesothelioma. A comparative study of epithelioid mesothelioma and lung adenocarcinoma. Am J Surg Pathol. 2003;27: Comin AE, Novelli L, Boddi V, Paglierani M, Dini S. Calretinin, Thrombomodulin, CEA, and CD15: a useful combination of immunohistochemical markers for differentiating pleural epithelial mesothelioma from peripheral pulmonary adenocarcinoma. Hum Pathol. 2001;32: Politi E, Kandaraki C, Apostolopoulou C, Kyritsi T, Koutselini H. Immunocytochemical panel for distinguishing between carcinoma and reactive mesothelial cells in body cavity fluids. Diagnostic Cytopathol. 2005;32: Ordonez NG. Value of thyroid transcription factor-1, E- Cadherin, BG8, WT1, and CD44s immunostaining in distinguishing epithelial pleural mesothelioma from pulmonary and nonpulmonary adenocarcinoma. Am J Surg Pathol. 2000;24: Wolanski KD, Whitaker D, Shilkin KB, Henderson DW. The use of epithelial membrane antigen and silver-stained nucleolar organizer regions testing in the differential diagnosis of mesothelioma from benign reactive mesothelioses. Cancer. 1998;82: Mohanty SK, Dey P, Rana P. Manual and automated AgNOR count in differentiating reactive mesothelial from metastatic malignant cells in serous effusions. Anal Quant Cytol Histol. 2003;25: Musti M, Kettunen E, Dragonieri S, et al. Cytogenetic and molecular genetic changes in malignant mesothelioma. Cancer Genet Cytogenet. 2006;170: Prins JB, Williamson KA, Kamp MM, et al. The gene for the cyclin-dependent-kinase-4 inhibitor, CDKN2A, is preferentially deleted in malignant mesothelioma Int J Cancer. 1998;75: Zellweger T, Benz G, Cathomas G, et al. Multi-target fluorescence in situ hybridization in bladder washings for prediction of recurrent bladder cancer. Int J Cancer. 2006;119: Metzgeroth G, Kuhn C, Schultheis B, Hehlmann R, Hastka J. Diagnostic accuracy of cytology and immunocytology in carcinomatous effusions. Cytopathology. 2007;DOI: / j x. In press. 53. Pomjanski N, Grote HJ, Doganay P, Schmiemann V, Buckstegge B, Böcking A. Immunocytochemical identification of carcinomas of unknown primary in serous effusions. Diagn Cytopathol. 2005;33: