The major purpose of cytological examination of any serous effusion is to determine whether malignant cells are present or not. This is an extremely important task, as in most cases the presence of malignant cells in effusion indicates an advanced stage of malignancy and is usually associated with poor survival, whereas on other occasions it may be the primary diagnosis (Naylor, 2008). Adenocarcinoma is by far the most common type of neoplastic cells to be found in serous effusion (Nguyen and Thomson, 2010).
Mesothelial cells readily undergo hypertrophy and hyperplasia in response to a wide variety of stimuli, such as inflammation or necrosis of the underlying parenchyma and the presence of foreign substances such as blood or air in the serous cavity (Urrunaga et al., 2013), as well as in some medical conditions, including cirrhosis, renal failure, and AIDS (Cibas, 2014). They often exfoliate in large numbers, either as isolated cells or as cohesive clusters of cells or both. They proliferate and show atypical morphological changes in cytoplasm and nucleus. It should be mentioned that some of these atypical changes can be readily identified in adenocarcinoma cells in serous fluids, which makes the differentiation between reactive mesothelial cells and metastatic adenocarcinoma even more difficult and challenging (Koss and Melamed, 2014).
On the other hand, effusions containing neoplastic cells usually contain benign cells, which may be stimulated to undergo hypertrophy and hyperplasia, and their number may greatly exceed that of the neoplastic cells themselves, which may be difficult to be recognized. In addition, adenocarcinoma cells in effusions exhibit great morphologic variation and commonly exhibit variable forms of differentiation (Butnor, 2006).
In some instances, getting adequate exfoliated cells in serous fluids is not an easy task, and thus because of the scarcity of tumor cells in malignant effusions and morphologic similarities to reactive mesothelial cells, especially those of adenocarcinoma, cytological examination fails to detect malignant cells in ∼40% of effusions (Thapar et al., 2009). Nowadays, the diagnostic accuracy of effusion cytology is enhanced though the utilization of ancillary techniques, and immunocytochemistry is one of the suggested methods, which helps distinguish between reactive mesothelial and adenocarcinoma cells using various cytology specimens such as wet or air-dried, liquid-based cytology smears or paraffin-embedded cell block (Grefte et al., 2008; Murugan et al., 2009; Sayed et al., 2009).
In this current study, we tried to evaluate the effectiveness of two combined markers, E-cadherin (ECAD) and desmin, for discriminating between adenocarcinoma cells and reactive mesothelial cells obtained from serosal cavity fluid.
ECAD is a cell adhesion molecule expressed in epithelial cells that mediates cell–cell adhesion. Abnormalities of the ECAD gene expression lead to a mutant protein that loses its adhesive properties and is expressed in malignant epithelial cells (Haque et al., 2008; Moghaddam et al., 2012). Desmin is one of the intermediate filament proteins that keeps other cytoplasmic proteins in proper relationship to one another, and it works, in addition, to be as a specific marker for myogenic differentiation among soft-tissue tumors (Wick and Hornick, 2010); it has been reported to exhibit positive staining of benign mesothelial cells (reactive mesothelial hyperplasia) in serous fluid. The exact etiology for expression of desmin in mesothelial cells is not known; however, the multipotential role of mesothelial cells with possible muscle differentiation and coexpression of desmin has been proposed by some studies (Afify et al., 2002; Hyun et al., 2012).
Patients and methods
This is a retrospective study that was carried out on 70 cases, in which cytologic examination of peritoneal and pleural fluid was performed by means of Papanicolaou-stained and Diff quick-stained smears, as well as paraffin-embedded cell blocks, at the Pathology Department, Cytology Unit, National Cancer Institute, Cairo University, during the period from August 2013 to December 2014. Forty-three cases were diagnosed as metastatic adenocarcinoma, and 27 cases were diagnosed as reactive mesothelial hyperplasia. The diagnosis was based on typical cytological morphology and/or immunocytochemistry staining, and it was supported by history, clinical and radiological findings. Cases showing metastatic tumors other than adenocarcinoma, as well as suspicious, hypocellular, and unsatisfactory smears, were excluded from the study.
Our study is retrospective carried on paraffin blocks, so according to Declaration of Helsinki, for this type of study, formal consent is not required.
Immunostaining for ECAD and desmin was done for all cases by BenchMark IHC/ISH staining module (Ventana, Roche, Basel, Switzerland). The primary antibodies used were as follows:
- ECAD, clone NCH-38, a ready-to-use monoclonal mouse antibody, by Dako.
- Desmin, clone D33, isotype immunoglobulinG1, κ, a ready-to-use monoclonal mouse antibody, by Dako.
In the evaluation of the ECAD immunostaining, the location (cytoplasmic and/or membranous), the intensity (faint, moderate and strong) of the stain, and the percentage of the immunoreactive cells for ECAD were considered. ECAD expression was scored as positive if at least more than 5% stained cells was present (He et al., 2004; Hasteh et al., 2010).
In the evaluation of desmin immunostaining, the location (cytoplasmic and/or membranous), the intensity (faint, moderate and strong) of the stain, and the percentage of cells immunoreactive for desmin were considered. Desmin protein expression was scored as positive if at least 10% stained cells was present (Minato et al., 2014).
Data were analyzed using integrated best management – statistical package for the social science (IBM SPSS) advanced statistics, version 22 (SPSS Inc., Chicago, Illinois, USA). Numerical data were expressed as mean. Qualitative data were expressed as frequency and percentage. The Fisher’s exact test was used to compare tumor cell immunoexpression between reactive mesothelial hyperplasia and metastatic adenocarcinoma. One-way analysis of variance test was used to determine whether the difference was significant. Significance was established at P value less than 0.05.
The present study included 70 cases of serous effusions: 43 (61%) cases were diagnosed as metastatic adenocarcinoma, whereas 27 (39%) cases were diagnosed as reactive mesothelial hyperplasia.
The age of the all studied patients ranged from 4 to 82 years, with a mean age of 49.5±19 years. The mean age in patients with metastatic adenocarcinoma was 53.3±13.5 years, whereas in cases of reactive mesothelial hyperplasia the mean age was 38.7±18 years.
The studied group included 23 (33%) male and 47 (67%) female patients, with a male to female ratio of about 1 : 2.1. Male patients were mostly diagnosed with reactive mesothelial hyperplasia (62.3%), whereas female patients were mostly diagnosed with metastatic adenocarcinoma (71.4%).
All 43 metastatic adenocarcinoma cases were positive for ECAD, whereas only two cases out of 27 (7%) cases of reactive mesothelial hyperplasia showed positive staining (Table 1). The results showed a statistically significant P value up to 0.05. The two cases of reactive mesothelial hyperplasia that were positive for ECAD showed moderate and strong intensity of staining (Figs 1 and 2). A majority of cases of metastatic adenocarcinoma exhibited moderate (Fig. 3) and strong immunostaining (Fig. 4), 22 (51.2%) and 17 (39.5%), respectively (Table 2).
ECAD had a sensitivity of 100%, specificity of 95%, positive predictive value (PPV) of 96.8%, negative predictive value (NPV) of 100%, and accuracy of 98% in diagnosing metastatic adenocarcinoma.
As for desmin immunoexpression, 25 (92.6%) out of 27 cases of reactive mesothelial hyperplasia were desmin positive compared with only four (9.3%) metastatic adenocarcinoma cases (Table 1). These results were highly statistically significant, with a P value up to 0.001. The primary of those metastatic case were two cases from ovary, one is of breast and one is of lung origin. All 25 positive cases of reactive mesothelial hyperplasia for desmin showed strong immunoreactivity (100%), as shown in Fig. 5, whereas 75% of desmin-positive metastatic adenocarcinoma cases showed faint immunoreactivity (Fig. 6), as shown in Table 3.
Desmin had a sensitivity of 90%, specificity of 90%, PPV of 85.7%, NPV of 93.1, and accuracy of 90% in diagnosing reactive mesothelial hyperplasia.
The combined use of ECAD and desmin as a panel in differentiating reactive mesothelial hyperplasia from metastatic adenocarcinoma is shown in Table 4.
Considering the staining of ECAD and desmin under conditions that the cells were stained with ECAD but not with desmin, the sensitivity was 90%, specificity was 100%, PPV was 100%, NPV was 87%, and accuracy was 94% to identify metastatic adenocarcinoma.
However, under conditions that the cells were stained with desmin but not with ECAD, the sensitivity was 85%, specificity was 100%, PPV was 100%, NPV was 90%, and accuracy was 94% to identify reactive mesothelial cell hyperplasia.
The major purpose of cytologic examination of serous effusions is to determine whether malignant cells are present or not. Cytologic examination of the serous fluid may be the first or the only chance for making the diagnosis of an underlying malignancy (Nguyen and Thomson, 2010).
In this study, a trial for evaluation of effectiveness of two combined markers, ECAD and desmin, for discrimination between adenocarcinoma cells and reactive mesothelial cells obtained from serosal cavity fluid is studied. The results regarding ECAD expression in cases of metastatic adenocarcinoma are close to, but with lower figures, those results recorded by Moghaddam et al. (2012), who used ECAD and fibronectin to differentiate metastatic adenocarcinoma from reactive mesothelial hyperplasia in serous effusions. Our findings also go with, but with higher figures, those recorded by He et al. (2004), who used ECAD and calretinin to differentiate metastatic malignant effusions from benign serous effusions. Nevertheless, the findings regarding expression of ECAD in reactive mesothelial hyperplasia were similar to those reported by He et al. (2004) and Moghaddam et al. (2012). Regarding the intensity of the staining, our cases showed either moderate or strong intensity of staining, but on the contrary the study conducted by He et al. (2004) showed that ECAD expression in carcinoma cells was usually strong.
ECAD expression in all metastatic adenocarcinoma cases was expected, because theoretically only the exfoliated cells originating from epithelial tissues can express ECAD; therefore, detection of ECAD expression is helpful for determining cells from epithelia. Because no epithelial cells are present in benign effusions, the appearance of epithelial cells in effusions means that a metastasis of carcinoma developed from the epithelia (Moghaddam et al., 2012).
In the current study, the findings regarding desmin expression in reactive mesothelial hyperplasia were similar but slightly higher than those reported by Hyun et al., 2012. A slightly lower figure also was reported by Hasteh et al. (2010), who used a different methodology, considering intensity of staining and percentage of the cells positive for desmin. Furthermore, Minato et al. (2014) reported a much lower percentage.
Regarding the cases of metastatic adenocarcinoma in the current work who were positive for desmin, these findings disagree with the study conducted by Hyun et al., 2012, in which all metastatic adenocarcinoma cases were negative for desmin. The occurrence of desmin immunoreactivity in metastatic adenocarcinoma in our current work may be related to the type of carcinoma, especially ovarian adenocarcinoma (two cases out of four). A possible explanation is that the likely origin of ovarian surface epithelial-stromal tumors is the mesothelial surface lining of the ovaries and/or invaginations of this lining into the superficial ovarian cortex that form inclusion cysts (Minato et al., 2014).
In our work, all reactive mesothelial hyperplasia cases positive for desmin showed strong immunoreactivity, whereas most of the metastatic adenocarcinoma cases showed faint immunoreactivity. The use of the scoring system, including the intensity, may reduce the percentage of desmin-positive carcinoma cases (Hyun et al., 2012). Further studies may be needed to clarify this point.
Several studies have been reported to show positive staining of benign mesothelial cells (reactive mesothelial hyperplasia) in serous fluid and tissue sections for desmin. The exact etiology for expression of desmin in mesothelial cells is not known; however, the multipotential role of mesothelial cells with possible muscle differentiation and coexpression of desmin has been proposed by some studies (Afify et al., 2002).
Conclusion and recommendation
ECAD alone has a high sensitivity in detecting carcinoma cells in effusion. The combined use of ECAD and desmin had a lower sensitivity and lower accuracy but higher specificity than the use of ECAD alone in identifying adenocarcinoma cells.
Thus, the combined use of this short panel is helpful for better differentiation of adenocarcinoma from reactive mesothelial hyperplasia in serous effusions and avoids overdiagnosis of metastatic adenocarcinoma if ECAD is used alone.
Caution is advised in the use of these markers in metastases of ovarian adenocarcinoma in effusions, as desmin may be false positive in this type of carcinoma. The use of a scoring system including intensity, as well as percentage of the immunoreactive cells, may help eliminate the positivity for desmin in metastatic adenocarcinoma in effusions, and thus further studies are needed to confirm this point.
Conflicts of interest
There are no conflicts of interest.
Afify AM, Al-Khafaji BM, Paulino AF, Davila RM (2002). Diagnostic use of muscle markers in the cytologic evaluation of serous fluids. Appl Immunohistochem Mol Morphol 10:178–182.
Butnor KJ (2006). My approach to the diagnosis of mesothelial lesions. J Clin Pathol 59:564–574.
Cibas ES (2014). Pleural, pericardial, and peritoneal fluids cytology, diagnostic principles and clinical correlates
, 4th ed. Philadelphia. 132–151.
Grefte JM, de Wilde PC, Salet-van de Pol MR, Tomassen M, Raaymakersvan Geloof WL, Bulten J (2008). Improved identification of malignant cells in serous effusions using a small, robust panel of antibodies on paraffin-embedded cell suspensions. Acta Cytol 52:35–44.
Haque AK, Singhal N, Allen T, Barrios R, Sienko A, Cagle PT, et al (2008). Lack of E-cadherin and Pan-cadherin immunoexpression is associated with metastasis in small cell lung cancer. Mod Pathol 21:34–42.
Hasteh F, Lin GY, Weidner N, Michael CW (2010). The use of immunohistochemistry to distinguish reactive mesothelial cells from malignant mesothelioma in cytologic effusions. Cancer Cytopathol 118:90–96.
He DN, Zhu HS, Zhang KH, Jin WJ, Zhu WM, Li N (2004). E-cadherin and calretinin as immunocytochemical markers to differentiate malignant from benign serous effusions. World J Gastroenterol 10:2406–2408.
Hyun TS, Barnes M, Tabatabai L (2012). The diagnostic utility of D2–40, calretinin, CK5/6, desmin and MOC-31 in the differentiation of mesothelioma from adenocarcinoma in pleural effusion cytology. Acta Cytol 56:527–532.
Koss GL, Melamed MR (2014). Effusions in the absence of cancer and effusions in the presence of cancer In Koss’ diagnostic cytology and its histopathologic bases, 5th ed. Philadelphia. 919–1022.
Minato H, Kurose N, Fukushima M, Nojima T, Usuda K, Sagawa M, et al (2014). Comparative immunohistochemical analysis of IMP3, GLUT1, EMA, CD146, and desmin for distinguishing malignant mesothelioma from reactive mesothelial cells. Am J Clin Pathol 141:85–93.
Moghaddam NA, Tahririan R, Eftekhari M, Tahririan D, Rahmani A (2012). Diagnostic value of E-cadherin and fibronectin in differentiation between reactive mesothelial and adenocarcinoma cells in serous effusions. Adv Biomed Res 1:56.
Murugan P, Siddaraju N, Habeebullah S, Basu D (2009). Immunohistochemical distinction between mesothelial and adenocarcinoma cells in serous effusions: a combination panel-based approach with a brief review of the literature. Indian J Pathol Microbiol 52:175–181.
Naylor BBibbo M (2008). Pleural, peritoneal, and pericardial effusions. Comprehensive cytopathology, 3rd ed. Philadelphia, PA: W.B. Saunders Elsevier. 515–578.
Nguyen GK, Thomson TA (2010). Serous effusions, essential of fluid cytology, 1st ed. London. 27–56.
Sayed DM, el-Attar MM, Hussein AA (2009). Evaluation of flow cytometric immunophenotyping and DNA analysis for detection of malignant cells in serosal cavity fluids. Diagn Cytopathol 37:498–504.
Thapar M, Mishra RK, Sharma A, Goyal V (2009). Critical analysis of cell block versus smear examination in effusions. J Cytol 26:60–64.
Urrunaga NH, Singal AG, Cuthbert JA, Rockey DC (2013). Hemorrhagic ascites. Clinical presentation and outcomes in patients with cirrhosis. J Hepatol 58:11–13.
Wick MR, Hornick JL (2010). Immunohistology of soft tissue and osseous neoplasms David Dabbs diagnostic immuohistochemistry, theranostic and genomic applications, 3rd ed. Philadelphia. 83–173.