|Year : 2012 | Volume
| Issue : 4 | Page : 602-609
Cytomorphological profile of neoplastic effusions: An audit of 10 years with emphasis on uncommonly encountered malignancies
Sanjay Gupta1, Pushpa Sodhani1, Shyama Jain2
1 Division of Cytopathology, Institute of Cytology and Preventive Oncology (ICMR), Noida, India
2 Department of Pathology, Maulana Azad Medical College, New Delhi, India
|Date of Web Publication||29-Jan-2013|
Scientist E, Division of Cytopathology, Institute of Cytology and Preventive Oncology (ICMR), I-7, Sector 39, Noida - 201 301
Source of Support: None, Conflict of Interest: None
Context: The diagnosis of metastatic cancer in fluids is of capital importance as, in most such instances, a rapid fatal outcome of the disease is anticipated.
Aim: To determine the spectrum and cytomorphological features of the common and unusual malignancies presenting with effusions.
Methods and Materials : A total of 11,562 effusion samples received for cytopathological examination over a 10-year period were analyzed retrospectively. Cytomorphological features of neoplastic effusions were studied. Special stains and immunocytochemistry (ICC) were performed to aid the diagnosis in difficult cases.
Observations : The effusion samples comprised of pleural (5018), peritoneal (6340) and pericardial (204) fluids. A definitive diagnosis of classifiable malignancy could be given in 836 (7.3%) of these cases (5.7% adenocarcinomas and 1.6% uncommon malignancies). Adenocarcinoma was the most frequent cause of malignant pleural (70%) and peritoneal effusions (86.9%). The most common primary site for pleural metastasis was lung (35.7%), while for peritoneal metastasis, it was the ovary (54.3%). Among the uncommon neoplastic effusions, hematopoeitic malignancies were the most frequent, followed by squamous cell carcinomas. Primary malignant mesotheliomas were the most challenging to diagnose on effusion cytology. ICC was useful to arrive at a definitive diagnosis in difficult cases.
Conclusions: Cytology is a useful tool to detect malignant effusions. However, in uncommon malignancies presenting as effusions, a detailed clinical history and ancillary investigations are often required to make a correct diagnosis.
Keywords: Effusion, fluid, cytology, pleural, peritoneal, pericardial, ascitis
|How to cite this article:|
Gupta S, Sodhani P, Jain S. Cytomorphological profile of neoplastic effusions: An audit of 10 years with emphasis on uncommonly encountered malignancies. J Can Res Ther 2012;8:602-9
|How to cite this URL:|
Gupta S, Sodhani P, Jain S. Cytomorphological profile of neoplastic effusions: An audit of 10 years with emphasis on uncommonly encountered malignancies. J Can Res Ther [serial online] 2012 [cited 2021 Jan 25];8:602-9. Available from: https://www.cancerjournal.net/text.asp?2012/8/4/602/106574
| > Introduction|| |
The first line of investigation of a suspected neoplastic effusion is often the cytological examination of fluid tapped from pleural, peritoneal and pericardial cavities. The responsibility of the pathologist is two-folds: (1) to identify cancer cells accurately and (2) to identify the tumor type and, if possible, the site of primary origin.
Although the tumors often shed abundant malignant cells, singly and in clusters, the interpretation of malignancy is much more difficult in body fluid than in any other cytologic media because of the exuberant proliferation of cells within the fluids.  Adenocarcinomas, which are the most common tumors producing effusions, are often relatively easy to diagnose. However, there are occasions when the cytopathologist encounters some unusual malignancies in the fluids. The diagnosis in such cases requires a constellation of cytomorphological criteria and correlation with the clinical history of the patient. The present study was undertaken to analyze the spectrum of usual and uncommon malignancies presenting as effusions in our setup and also to determine the subtle cytomorphological features that might aid their diagnosis.
| > Materials and Methods|| |
A total of 11,562 effusion samples, where primary diagnosis was known, were received from a major tertiary care hospital of North Delhi, over a 10-year period (January 1998 to December 2007). These were retrospectively analyzed to determine the cytomorphological spectrum of malignancies presenting with effusions. Special emphasis was given to the subtle cytological features that might aid the diagnosis of uncommon malignancies. Two smears were available from each effusion sample, one of which was air dried and stained by Giemsa while the other was wet fixed and stained by standard Papanicolaou's method. Special stains, like Periodic acid Schiff, mucicarmine and fat stain, and immunocytochemistry (ICC) were performed to support the diagnoses as and when required. For immunocytochemical staining, a repeat fresh sample was obtained and smears made from the centrifuge deposit. They were fixed in 1:1 mixture of cold acetone and methanol, wrapped in aluminum foil and stored in refrigerator till further processing. Immunostaining was carried out using standard Peroxidase-anti-Peroxidase (PAP) protocol. The antibodies used were S-100 (for neuroendocrine tumors and melanoma), HMB45 (for melanoma), panel of CEA, MOC 31 and CK7/CK20 for adenocarcinoma, and CK5/6 (for squamous malignancies), Calretinin and WT 1 for mesotheliomas.
| > Results|| |
The effusion samples (11,562) comprised of pleural (5018), peritoneal (6340) and pericardial (204) fluids. Of these, 10,488 (90.7%) were reported as negative for malignancy; in 86 cases (0.74%), a diagnosis of "malignant cells - typing not possible" was given and in 152 (1.3%), "Atypical cells - suspicious of malignancy" were found. A definitive diagnosis of classifiable malignancy could be offered in 836 cases (7.2%); these included 654 (5.7%) adenocarcinomas and 182 (1.6%) uncommon malignancies [Table 1]. The age of patients ranged from 6 years to 85 years (mean age 51.5 years).
In 86 cases where a diagnosis "malignant cells - typing not possible" was given and where primary site was not known at the time of tapping the effusion, follow-up was available in 46 patients. The primary site for malignancy was detected to be lung (16), breast (6), bone (4), ovary (2), gall bladder (4) and pancreas (2). No primary site could be ascertained in 12 cases.
In 152 cases with diagnosis "atypical cells - suspicious of malignancy," repeat cytology was available in 84 cases. Twenty-four cases turned out to be negative, while in 60 cases the diagnosis was still "suspicious of malignancy"; further diagnostic workup of these cases revealed malignancy in lung (8), gall bladder (6), liver (2), breast (2), esophagus (2), kidney (2), bone (2), nasopharynx (2), adrenal (2) and lymph node (2). However, no primary site could be detected in 30 cases.
Adenocarcinoma was the most frequent cause of malignant pleural and peritoneal effusions. Among 400 neoplastic pleural effusions, 280 (70%) were due to adenocarcinoma while 120 were due to other uncommon malignancies. Similarly, of 428 cases of malignant ascitis, 372 (86.9%) were due to adenocarcinomas while 56 were attributed to other malignancies. Of the eight neoplastic pericardial effusions, only two (25%) were due to adenocarcinoma. On further analysis, among the pleural effusions due to adenocarcinomas, lung was the most frequent primary site (35.7%), followed by breast (30%), ovary (19.5%) and gastrointestinal tract (GIT) (2.9%) [Figure 1]a and b. In 50 effusions (17.8%), the primary site was not known. Among malignant peritoneal effusions with adenocarinoma cells (372), ovary was the predominant primary site (54.3%), followed by GIT (22.6%) [Figure 1]c. Primary cancers of the breast (5.3%) and lung (2.7%) infrequently metastasized to peritoneal cavity. In 56 cases (15.1%), the primary site could not be ascertained [Table 2].
|Table 2: Distribution of adenocarcinomas in effusions correlated with primary malignancy|
Click here to view
Uncommonly encountered malignancies in Effusions [Table 3]
Pleural effusions (n=120)
Hematopoietic malignancies were encountered in 58 cases [Figure 2]a and b; these comprised of non-Hodgkin's lymphomas (NHL) (50), chronic lymphocytic leukemia (2), Hodgkin's lymphoma (HL) (4) and multiple myeloma (2). Squamous cell carcinoma (SCC) [Figure 3]a was diagnosed in 32 cases [primary sites: lung (24), head and neck region (6), cervix (2)], small cell anaplastic carcinoma of lung in 10 cases, malignant melanoma [Figure 3]b in four [primary sites : Buccal mucosa (2) and skin (2)] small round cell tumor (4) [Figure 3]c. Two cases each of metastatic germ cell tumor and granulosa cell tumor of ovary [Figure 3]d and eight cases of primary malignant mesothelioma were detected.
Ascitic fluid (n=56)
Hematopoietic malignancies were encountered in 32 cases [Figure 2]c and d [NHL 30, chronic myeloid leukemia (CML) 2] and SCC was encountered in 14 cases (primary : Cervix 8, anus 6). Two cases each of germ cell tumor of testis, malignant melanoma of rectum and transitional cell carcinoma (TCC) bladder [Figure 3]e were detected. Four cases of primary mesothelioma presented with ascitis.
Pericardial effusions (n=6)
Two cases each of NHL, CML and mesothelioma [Figure 4] were picked up. Among all three sites for effusion, a total of 94 hematopoietic malignancies were detected. Of these, correct morphological diagnosis could be offered in 73 cases (77.7%), including 66/82 NHL, 1/4 HL, 2/4 CML, 2/2 CLL and 2/2 multiple myeloma. In the remainder of the cases, ICC, immunophenotyping and histochemistry had been resorted to in order to arrive at the conclusive diagnosis. Of a total of 46 SCCs diagnosed in effusions, diagnosis could be given on cytology alone in 26 (56.5%) cases. In the remaining, differential diagnoses of adenocarcinoma or poorly differentiated carcinoma were considered and immunomarkers were used to establish the diagnosis of SCC.
|Figure 4: Primary mesothelioma presenting as pericardial effusion in a 55-year-old female. Note cells exhibiting surface microvilli and the morular cluster (inset). Giemsa ×400|
Click here to view
Small cell anaplastic carcinoma of lung could be accurately typed on cytology in 5/10 (50%) cases; ICC was performed to differentiate from other small cell tumors in the rest. Malignant melanomas could be correctly diagnosed in 4/6 (66.6%) effusions, based on characteristic morphology. In two cases where the differential diagnosis of poorly differentiated adenocarcinoma or florid reactive mesothelial cells was kept, ICC using S-100 and HMB-45 was contributory to final diagnosis. Two/four (50%) germ cell tumors could be typed on cytology alone while special stains/ICC were resorted to in the other two cases. None of the two TCC could be correctly identified on cytology, and were labeled as papillary tumors.
Primary mesotheliomas presented diagnostic challenge on cytology even when diagnosis was known. Only in 2/14 (14.3%) cases, was the morphology characteristic enough to allow a definite diagnosis on cytology. In all others, differential diagnosis of adenocarcinoma was kept. In these cases, special stains (PAS, mucicarmine and fat stain) and immunocytochemical panel consisting of Calretinin, MOC 31, CK 5/6, CK 7/20 and CEA was put up to arrive at the conclusion. The vacuoles in malignant mesothelioma cells stained for fat but not for mucin, contrary to the vacuoles in metastatic adenocarcinoma. Mesothelioma cells expressed Calretinin and CK 5/6, while adenocarcinomas expressed CEA, MOC 31 and CK7/20 in different combinations [Figure 5].
|Figure 5: Immunoexpression of Calretinin (a) in mesothelioma and of CEA (b) in an adenocarcinoma in pleural fluid specimens. Peroxidase– antiperoxidase ×400|
Click here to view
| > Discussion|| |
Serous effusions are a frequently encountered clinical manifestation of metastatic disease. The cancer cells of different origins that shed in effusions differ in their biology and have unique phenotypic and genotypic characteristics. Cytological examination of the fluids is often the first line of investigation to detect and, if possible, type these neoplastic cells based on their subtle morphological features. In the present study, the neoplastic effusions comprised around 8% of the total effusion samples, while in 1.3% "atypical cells - suspicious of malignancy" were reported.
The most common type of tumor to produce metastasis in the serous cavities is the broad group of adenocarcinomas, most often from breast, lung, ovary and GIT. ,,,, In the present study also, adenocarcinomas comprised the most frequent cause of malignant pleural (70%) and peritoneal effusions (86.9%). The most frequent primary site for pleural metastasis was lung, while it was ovary for peritoneal metastasis of adenocarcinomas. The tumor cells showed classical features of adenocarcinoma, i.e., the tendency to form smoothly contoured cohesive groups composed of large cells with eccentric, malignant-appearing nuclei, prominent nucleoli and vacuolated cytoplasm. However, at times, adenocarcinoma cells in effusions were seen to exhibit great morphologic variation. In well-differentiated bronchioloalveolar-type adenocarcinoma, the clusters of tumor cells were composed of cells with a low nuclear/cytoplasmic ratio and mucin-containing cytoplasm. In giant cell carcinoma, the cells tended to cluster and nuclei were eccentric; cytoplasmic vacuolations were absent.
Metastatic breast cancer cells are most often found in pleural fluids, although, occasionally, they may be found in peritoneal and pericardial fluids. ,,, Cancer cells from the usual infiltrating duct carcinoma were small to medium sized and arranged in compact, smoothly round dense clusters (proliferation spherules). Lobular carcinomas typically manifested as small, isolated tumor cells that resembled macrophages or small mesothelial cells. However, individual tumor cells showed irregularity of nuclear shape, heavy nuclear membrane and prominent nucleoli. In addition, the cytoplasm displayed vacuoles containing mucus droplets in some cases.  Cells of lobular carcinoma also had a tendency to form Indian file chains.
Adenocarcinoma of the ovary is a frequent cause of ascitis and, to a lesser extent, of pleural effusion. , The neoplastic cells in the effusion sample presented as large acinar or papillary clusters admixed with numerous single large and hypervacuolated cells. A few adenocarcinomas, however, presented with relatively small cells with little vacuolation in cytoplasm, and these were difficult to differentiate from mesothelial cells. With experience, it may be possible to suggest with accuracy the primary site of adenocarcinoma, especially if given information about the sex of the patient and the site from which the fluid was obtained. Yet, there are always cases in which the deduction turns out to be wrong. Clinical background of the patient usually provides the best clues about the origin of the cells in such cases. ,
Serous effusions may be a complication of hematopoietic malignancies like lymphomas and leukemias. ,,,,
In the present review, hematopoietic malignancies were detected in 58/120 (48.3%) pleural, 32/56 (57.1%) ascitic and 4/6 (66.6%) pericardial effusions. NHL were the most common cause of effusions due to hematopoietic malignancies. Most effusions due to lymphoma appeared clear, slightly blood-tinged stained or, occasionally, chylous. One of the most important aspects of cytology of NHL and leukemia was that the neoplastic cells did not exhibit genuine attachment to each other. Another striking feature of these neoplasms in serous fluids was massive necrosis, a feature rarely observed with carcinomas. The cytological picture in HL was nonspecific, consisting mainly of lymphocytes, plasma cells, macrophages and mesothelial cells. Pleural eosinophilia was rare and the pathognomic Reed Sternberg cells were only occasionally observed, rendering the diagnosis of HL difficult on effusion cytology.
Pleural effusions were more commonly encountered in lymphoid leukemia than in myeloid leukemia, as reported by various other authors. ,, In patients with leukemia, the possibility of alternative causes of effusion, such as bacterial or viral infections or complications of chemotherapy, should be excluded. Serous effusions in myeloproliferative disorders like CML were occasionally observed in our series. The predominant cells in fluid were mature and immature granulocytes with a variable number of blasts.
Pleural effusions have been reported to occur in approximately 6% of patients with myeloma. , In our cases, the myeloma cells revealed the usual morphologic features of plasma cells, but were larger and often multinucleated with prominent nucleoli. These features helped distinguish them from plasma cells found in mixed inflammatory reactions.
SCC presented with pleural effusions in 32 (26.7%) and ascitis in seven (12.5%) cases. Among pleural fluids with metastatic SCC, the most common primary was lung, followed by head and neck region. In ascitic effusions due to SCC, the most common primary was uterine cervix, followed by anus. In a review of 7389 patients whose serous effusions had been examined cytologically, malignant squamous cells were found in only 46 patients, mostly originating from primary neoplasms of the lung, female genital tract or larynx.  The helpful diagnostic features of these cancers in fluids were presence of keratinizing squamous cells, anucleated squames of odd shapes and sizes and pearl formation. In the absence of clear evidence of keratinization, it is easy to overlook their squamous origin in effusions.
Other rare neoplasms that caused pleural effusions included small cell anaplastic carcinoma of lung, malignant melanoma, germ cell tumor, Ewing's sarcoma and granulosa cell tumor of ovary. In the peritoneal cavity, rare causes of effusions included germ cell tumor, melanoma and transitional cell carcinoma of urinary bladder. In all these cases, the primary site as well as the clinical diagnosis, were known at the time of tapping the effusion and thus it could be typed accurately on cytology.
Primary malignant mesotheliomas rarely present with effusions and are the most challenging tumors to diagnose on effusion cytology. ,, The presence of numerous mulberry-shaped papillary clusters of cells in an effusion of a patient without a known primary cancer provided clues to the possibility of a primary carcinomatous mesothelioma. The malignant mesothelial cells may closely resemble benign mesothelial cells because of abundant, occasionally faintly vacuolated cytoplasm, often with a distinct, clear cell border and denser perinuclear area, a comparatively small, somewhat hyperchromatic nuclei. Bizarre cell forms, nuclear irregularities and large nucleoli, abnormal mitoses, gland-like structures and peculiar multinucleated cells were found to favor a malignant diagnosis. As reported by Boon et al., we also noted that the cytoplasm of pleural mesothelioma cells was smaller and contained few vacuoles, while the cells in the peritoneal mesothelioma were larger and were characterized by numerous perinuclear vacuoles.  Despite the knowledge of the biopsy diagnosis and careful evaluation of the smears, the diagnosis of mesothelioma on cytology was difficult and could be made in only two of the 14 cases. In the rest, histochemical staining and/or immunomarkers were required to differentiate them from adenocarcinomas. ,
To conclude, effusion cytology is a useful tool to detect malignant effusions. It is possible to type the tumor in a substantial number of cases, especially if the metastasis is from an adenocarcinoma, and, in such cases, it may be possible to suggest the possible primary site based on cytomorphological features. However, it is rather difficult to accurately type the unusual malignancies presenting as effusions, unless the primary site is known. Ancillary tests like ICC may be judiciously used to aid the diagnosis in morphologically challenging cases.
| > References|| |
|1.||Koss LG, Melamed MR. Effusions in the presence of cancer. In : Koss LG, editor. Koss' Diagnostic Cytology and its Histopathologic Bases. Vol II, 5 th ed. Phildelphia, Pennsylvania, USA : Lippincott Williams and Wilkins; 2006. p. 950. |
|2.||Pradhan SB, Pradhan B, Dali S. Cytology of body fluids from different sites : An approach for early diagnosis of malignancy. J Nepal Med Assoc 2006;45:353-56. |
|3.||Sears D, Hajdu SI. The cytologic diagnosis of malignant neoplasms in pleural and peritoneal effusions. Acta Cytol 1987;31:85-97. |
|4.||Ong KC, Indumathi V, Poh WT, Ong YY. The diagnostic yield of pleural fluid cytology in malignant pleural effusions. Singapore Med J 2000;41:19-23. |
|5.||Johnston WW. The malignant pleural effusion. A review of cytopathologic diagnoses of 584 specimens from 472 consecutive patients. Cancer 1985;56:905-9. |
|6.||Awasthi A, Gupta N, Srinivasan R, Nijhawan R, Rajwanshi A. Cytopathological spectrum of unusual malignant pleural effusions at a tertiary care centre in north India. Cytopathology 2007;18:28-32. |
|7.||Butler EB, Stanbridge CM. Cytology of body cavity fluids : A color Atlas. London : Chapman and Hall; 1986. |
|8.||Monte SA, Ehya H, Lang WR. Positive effusion cytology as the initial presentation of malignancy. Acta Cytol 1987;31:448-52. |
|9.||Foot NC. Identification of types and primary sites of metastatic tumors from exfoliated cells in serous fluids. Am J Pathol 1954;30: 661-77. |
|10.||Spriggs AI, Jerrome DW. Intracellular mucous inclusions. A feature of malignant cells in effusions in the serous cavities, particularly due to carcinoma of the breast. J Clin Pathol 1975;28:929-36. |
|11.||Johnson TL, Kumar NB, Hopkins M, Hughes JD. Cytologic features of ovarian tumors of low malignant potential in peritoneal fluids. Acta Cytol 1988;32:513-8. |
|12.||Davidson B. Ovarian carcinoma and serous effusions. Changing views regarding tumor progression and review of current literature. Anal Cell Pathol 2001;23:107-28. |
|13.||Spriggs AI, Boddington MM. Atlas of serous fluid cytopathology. A guide to the cells of pleural, pericardial, peritoneal and hydrocele fluids. In : Gresham GA, editor. Current Histopathology series. Vol. 14. Dordrecht, Kluwer Academic Publishers, 1989. |
|14.||Spieler P, Gloor F. Identification of types and primary sites of malignant tumors by examination of exfoliated tumor cells in serous fluids. Comparison with the diagnostic accuracy on small histologic biopsies. Acta Cytol 1985;29:753-67. |
|15.||Das DK. Serous effusions in malignant lymphomas : A review. Diagn Cytopathol 2006;34:335-47. |
|16.||Melamed MR. The cytological presentation of malignant lymphomas and related diseases in effusions. Cancer 1963;16:413-31. |
|17.||Wilson MS, Theil KS, Goodwin RA, Brandt JT. Comparison of Papanicolaou's and Wright-Giemsa stains in the examination of body fluids for Hodgkin's disease. Arch Pathol Lab Med 1988;112: 612-5. |
|18.||Peterson IM, Raible M. Malignant pleural effusion in Hodgkin's lymphoma. Report of a case with immunoperoxidase studies. Acta Cytol 1991;35:300-5. |
|19.||Alexandrakis MG, Passam FH, Kyriakou DS, Bouros D. Pleural effusions in hematologic malignancies. Chest 2004; 1546-55. |
|20.||Mohapatra MK, Das SP, Mohanty NC, Dash PC, Bastia BK. Haemopericardium with cardiac tamponade and pleural effusion in chronic myeloid leukemia. Indian Heart J 2000;52:209-11. |
|21.||Bourantas KL, Repousis P, Tsiara S, Christou L, Konstantinidou P, Bai M. Chronic myelogenous leukemia terminating in acute megakaryoblastic leukemia. Case report. J Exp Clin Cancer Res 1998;17:243-5. |
|22.||Geisinger KR, Buss DH, Kawamoto EH, Ahl ET Jr. Multiple myeloma. The diagnostic role and prognostic significance of exfoliative cytology. Acta Cytol 1986;30:334-40. |
|23.||Sasser RL, Yam LT, Li CY. Myeloma with involvement of the serous cavities. Cytologic and immunochemical diagnosis and literature review. Acta Cytol 1990;34:479-85. |
|24.||Smith-Purslow MJ, Kini SR, Naylor B. Cells of squamous cell carcinoma in pleural, peritoneal and pericardial fluids. Origin and morphology. Acta Cytol 1989;33:245-53. |
|25.||Klempman S. The exfoliative cytology of diffuse pleural mesothelioma. Cancer 1962;15:691-704. |
|26.||Naylor B. The exfoliative cytology of diffuse malignant mesothelioma. J Pathol Bacteriol 1963;86:293-8. |
|27.||Boon ME, Veldhuizen RW, Ruinaard C, Snieders MW, Kwee WS. Qualitative distinctive differences between the vacuoles of mesothelioma cells and of cells from metastatic carcinoma exfoliated in pleural fluid. Acta Cytol 1984;28:443-9. |
|28.||Delahaye M, van der Ham F, van der Kwast TH. Complementary value of five carcinoma markers for the diagnosis of malignant mesothelioma, adenocarcinoma metastasis, and reactive mesothelium in serous effusions. Diagn Cytopathol 1997;17:115-20. |
|29.||Yaziji H, Battifora H, Barry TS, Hwang HC, Bacchi CE, McIntosh MW, et al. Evaluation of 12 antibodies for distinguishing epithelioid mesothelioma from adenocarcinoma : Identification of a three-antibody immunohistochemical panel with maximal sensitivity and specificity. Mod Pathol 2006;19:514-23. |
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]
[Table 1], [Table 2], [Table 3]
|This article has been cited by|
||Squamous cells in effusions: A study of 24 cases
| ||Adhikari, L.J., Elkins, C.T., Wakely Jr., P.E. |
| ||Journal of the American Society of Cytopathology. 2014; 3(1): 21-28 |
||Hücre bloǧu yöntemi akciǧer kanserine eşlik eden eksudatif plevral sivilarda taniyi artirmaktadir | [The cell block method increases the diagnostic yield in exudative pleural effusions accompanying lung cancer]
| ||Köksal, D., Demiraǧ, F., Bayiz, H., (...), Berktaş, B., Berkoǧlu, M. |
| ||Turk Patoloji Dergisi/Turkish Journal of Pathology. 2013; 29(3): 165-170 |