Role of Imprint Cytology in the Intraoperative Diagnosis of Soft Tissue Tumors

Lakshmy M Ra, Krishna Balachandranb

a. Department of Pathology, Employees’ State Insurance Corporation Medical College Hospital, Parippally, Kerala, India; b. Department of Pathology, Government Medical College, Thiruvananthapuram, Kerala, India*

Corresponding Author: Dr. Lakshmy M R, Assistant Professor, Department of Pathology, Employees’ State Insurance Corporation Medical College Hospital, Parippally, Kerala, India. Phone: +91 98 47 125267, Email: lakshmymr1980@gmail.com

Abstract

Tumors of the soft tissue constitute a heterogeneous group of neoplasms in terms of clinical presentation, morphology and behavior. Intraoperative diagnosis of soft tissue tumors is indicated not only for making a diagnosis, but also for evaluating margin status, determining tumor extent/spread, and obtaining an adequate sample for ancillary diagnostic tests. This study attempted to bring out the values and limitations of imprint cytology in providing a rapid and accurate diagnosis of soft tissue tumors and in particular to explore the potential application of cytology techniques in the intraoperative setting. The accuracy of this method was assessed by comparing the imprint diagnosis with paraffin section diagnosis. It was also aimed to describe the morphology of cells seen in the imprint smears of the different tumor types as well as to correlate the cytologic features with gross pathologic features like tumor size, location, depth, circumscription, encapsulation and presence of necrosis. This study included 87 patients admitted with a clinical diagnosis of soft tissue tumor in Government Medical College, Thiruvananthapuram from February 2007 to December 2008. Imprint and scrape smears were obtained intraoperatively from the tumors, fixed, stained and studied under a microscope. Imprint cytology and histopathology slides were compared. Of the 87 cases, 50 were malignant; 25 benign and 7 were of intermediate grade. Five cases turned out to be other soft tissue lesions. 18 benign tumors and 37 malignant tumors could be accurately categorized by imprint cytology. Weighted kappa statistics showed substantial agreement between imprint cytology and histopathology (0.78). Imprint smears were seen to provide very good cytologic and architectural details. Gross examination of specimens served as an adjunct to enhance the diagnostic accuracy. This study concludes that imprint cytology serves as an invaluable tool in the intraoperative evaluation of soft tissue tumors, the accuracy of which can be enhanced by correlating with clinical and radiological features. Imprints prepared from fresh specimens give excellent cytological details. However, intraoperative diagnosis of benign soft tissue tumors by imprint cytology is preliminary and warrants confirmation.

Key words: Imprint Cytology, Intraoperative Diagnosis, Soft Tissue Tumors

 

Introduction

Soft tissue is defined as the supportive tissue of various organs and the non-epithelial, extra-skeletal structures1 exclusive of lymphohematopoietic tissues. Soft tissue tumors constitute a heterogeneous group of neoplasms that clinically range from totally benign to highly malignant neoplasms.1 Benign soft-tissue tumors are about 100 times more common than malignant soft-tissue tumors.3,16

Although soft tissues constitute much of the human body, the incidence of soft tissue sarcomas is low and accounts for 30 per million per annum. Although fewer than 1% of the malignant tumors in adults are soft tissue sarcomas,16 they are nevertheless responsible for 2% of all deaths caused by cancer.4 Poorly differentiated soft tissue sarcomas in particular carry a poor prognosis that can be attributed to frequent hematogenic metastases (30-40%).

The clinical symptoms accompanying the diagnosis of soft tissue tumors are nonspecific. The most common finding at presentation is painless and gradually enlarging mass.17,22,31 Most soft tissue sarcomas arise de novo, but a small percentage originate in injured tissues such as scars or radiation-exposed areas.

More than 50 subtypes of proliferative soft tissue lesions are presently defined.32 The classification of soft tissue tumors is regularly updated. As per the 2002 WHO classification, the soft tissue tumors are divided into 4 categories2- benign, intermediate (locally aggressive), intermediate (rarely metastasizing) and malignant.

Very few neoplasms challenge the diagnostic skills of pathologists more than soft tissue tumors. An accurate histopathological diagnosis is of utmost importance in the management of soft tissue tumors and is quite a challenging task demanding good clinical and radiological correlation.

Early tissue diagnosis is one of the most important components of multimodality treatment of soft tissue tumors. Imprint cytology is a valuable tool for rapid intraoperative diagnosis19,21,23,30 of soft tissue tumors. The method is rapid, easy, reliable and does not require special equipments.23,27

Smears prepared from fresh specimens give excellent cytological details. Scrape cytology, a modification of imprint cytology, has better diagnostic accuracy than imprint cytology or frozen sections alone.6,13,14,24 Despite its merits like speed, simplicity and excellent cellular detail,18 many centers do not utilize this technique routinely.25

 

Objectives

  1. To assess the diagnostic accuracy of imprint cytology in soft tissue tumors by correlating with histopathology.

  2. To observe whether the correlation of the cytologic features with gross pathologic features such as tumor size, location, depth, circumscription, encapsulation, and presence of necrosis improves diagnostic accuracy.

 

Methodology

Patients admitted with clinical and/or radiological diagnosis of soft tissue tumor in the Department of General Surgery, Government. Medical College, Thiruvananthapuram from February 2007 to December 2008 were included in this study. From this group, 87 cases were selected in whom malignancy was suspected clinically.

Fresh specimens of soft tissue tumors obtained at the time of surgery were cut and direct imprint taken by pressing a glass slide gently on the fresh cut surface. After this, tumor tissue from the cut surface was scraped off using a clean glass slide. This scraped tissue was spread on another clean glass slide.

The imprint and scrape smears were immediately fixed in 95% ethyl alcohol for 5-6 min.18 Rapid hematoxylin and eosin staining was done, and smears were studied under a microscope. A minimum of three imprint smear slides and two scrape smear slides were studied per case.

The tumor tissue was then fixed in 10% formalin, processed, stained with hematoxylin and eosin and studied under a microscope. Whenever possible, histopathological findings were correlated with clinical and radiological features and immunohistochemistry. Imprint cytology and histopathology slides were compared (Table 1).

Table 1. Cytohistological Correlation

thumblarge

Diagnostic accuracy of imprint cytology was determined using sensitivity, specificity, and Kappa value. Ethical Committee approval was obtained from Government Medical College, Thiruvananthapuram before starting the study.

 

Observations

Intraoperative imprint cytology was performed on 87 soft tissue tumors, which were clinically suspected as malignant. Histopathological confirmation was obtained in all cases (Table 2).

Table 2. Tumor types by histopathology

thumblarge

Of the 87 cases, 50 were malignant; 25 benign and 7 were of intermediate grade. 5 cases turned out to be other soft tissue lesions. Soft tissue tumors in this study occurred mostly between 31 and 60 years of age (52.4%). Malignant tumors predominated between fifth and eighth decades (64%) with a peak incidence from 41 to 50 years (22%). Among the 87 cases studied, 52 were males and 35 females.

The tumors ranged in size from 1.5 cm to 33 cm. Most of them (64.4%) were ≤10 cm in size. 71% of the tumors above 10 cm in size, 80% of the tumors above 15 cm in size and all the tumors above 20 cm in size were malignant.

In this study, soft tissue tumors were most frequently encountered in the extremities (57.5%) – lower extremities (40.2%) more than upper extremities (17.2%) – followed by the trunk (21.8%) and the retroperitoneum (18.4%). The head and neck was the least common site (2.3%). Most of the retroperitoneal tumors (81.2%) and the lower extremity tumors (68.6%) were malignant.

Most of the soft tissue tumors were located in the subcutaneous plane (52.4%) followed by the muscular plane (28%). Benign tumors occurred most frequently in the subcutis (84%). Malignant tumors showed more predilection to involve the muscle plane (36%), followed by the subcutis (34%). Of note, most of the tumors located in the muscular plane (72%) and in the body cavity (88%) were malignant.

About 80% of the benign tumors were well-circumscribed. 33 out of the 50 malignant tumors (66%) and 6 out of the 7 tumors of intermediate grade (85.7%) were not circumscribed. 73.3% of the poorly circumscribed tumors were malignant. Encapsulation was noted in 8 sarcomas and 6 benign tumors. 40% of the malignant soft tissue tumors showed necrosis. Necrosis was not observed in any of the benign or intermediate-grade tumors. Only 3 of the 87 cases were recurrent tumors, all of which were high-grade sarcomas.

Based on the imprint and scrape cytology smears, the 87 cases under study were broadly classified into the following 5 groups:

  1. Benign

  2. Atypical

  3. Suspicious of malignancy

  4. Malignant

  5. Inconclusive

Among the 25 benign tumors, 18 could be identified as benign by cytology. Two schwannomas showing cystic change and one cavernous hemangioma could not be diagnosed by cytology as the cytological details were obscured by inflammatory cells, blood and debris. Imprint smears of one granular cell tumor, one traumatic neuroma, and one case of nodular fascitis were also inconclusive due to scanty cellularity. One case of ancient schwannoma26 showed moderate nuclear pleomorphism and was hence categorized as atypical.

Among the 7 soft tissue tumors of intermediate grade, one case of hemangioendothelioma and one case of fibromatosis were categorized as benign tumors by cytology due to the low cellularity and lack of pleomorphism in the imprint smears. One case of hemangiopericytoma and two cases of fibromatosis were categorized as atypical on account of increased cellularity and mild nuclear pleomorphism. The study also included one case of solitary fibrous tumor and one case of inflammatory pseudotumor in which the imprint smears were inadequate for a definite categorization.

Among the 50 soft tissue sarcomas, 37 could be diagnosed as malignant by cytology. Among the remaining 13 cases, five liposarcomas, three rhabdomyosarcomas, two fibromyxoid sarcomas and one fibrosarcoma showed only mild to moderate cellular pleomorphism in smears and were categorized as suspicious of malignancy. Two cases of liposarcomas showed clusters of mature adipocytes only and were incorrectly categorized as benign tumors by cytology.

In the analytical part, for the comparison between imprint cytology and histopathology, weighted kappa was calculated as a measurement of agreement between the two diagnostic measures. Diagnostic accuracy was determined using sensitivity, specificity, positive and negative predictive values, after excluding the 8 soft tissue tumors that showed inconclusive diagnosis in imprint smears, as well as the 5 non-mesenchymal lesions.

 

Discussion

Intraoperative diagnosis is an invaluable aid in the management of soft tissue tumors, especially when clinical and radiological findings are unequivocal. On table, a broad categorization into benign and malignant tumors is possible, thus assisting the surgeon to decide on the extent of resection. This study employed imprint cytology from fresh specimens for intraoperative evaluation of soft tissue tumors and found this to be a simple, rapid and economical method.

An accurate intraoperative diagnosis requires a multimodality interdisciplinary approach correlating clinical, radiological and pathologic information5 including gross examination, cytological and histological examination and ancillary techniques, if facilities are available. The current study correlated the cytologic features with clinical parameters like age and sex of the patient as well as gross pathologic features such as tumor size, location, depth, circumscription, encapsulation and presence of necrosis.

In the present study, though malignant soft tissue tumors were found in all age groups, most of them (64%) occurred above 40 years of age. This is comparable to a study by Fang et al on 1118 cases of primary soft tissue sarcomas, where 63.06% occurred above 40 years of age.7 The male to female ratio for malignant soft tissue tumors was 2:1 in this study as in the study by Bhurgri et al8 on 96 sarcomas, while Fang et al7 found the ratio to be 1.4:1. However, the sex of the patient was not helpful in separating tumors into benign and malignant.

The tumor size was more useful in this respect. In the present study, 71% of the tumors above 10 cm in size and 80% of the tumors above 15 cm in size were malignant. Assessment of the plane of the tumor was found to be a useful parameter in deciding the benign and malignant behavior of the tumors.29 Most of the sarcomas occurred in deeper planes – muscle plane (36%), followed by the subcutis (34%), whereas, 84% of the benign tumors involved the subcutaneous plane. By comparing clinical data for benign tumors and sarcomas, Rydholm A et al9 also found that a tumor 5 cm or larger or a deep tumor is relatively more likely to be a sarcoma.

The lower extremity (48%) was the most favored site for sarcomas. 81.2% of the retroperitoneal soft tissue tumors were malignant. Most of the malignant soft tissue tumors in this study were ill circumscribed (66%) and non-encapsulated (84%). Tumor necrosis was observed only in sarcomas, not in benign tumors. Gustafson P propose tumor necrosis as a strong and reliable factor that can be used to improve prognostic accuracy.10

All the 87 cases selected in the present study were those clinically suspected to be malignant soft tissue tumors. Among them, 25 turned out to be benign tumors. There were 50 malignant soft tissue tumors consisting predominantly of liposarcomas,36 pleomorphic sarcomas,34,35,37 and malignant peripheral nerve sheath tumors.33

Benign spindle cell tumors formed the largest category of the non-malignant tumors (21 cases). Among them, 12 were neural tumors. Correlating with the gross features, a possible neural origin could be suggested in 6 cases, but subtyping of the spindle cell lesions by cytology alone was difficult. This has been well recognized as a limiting factor in other studies such as Rekhi et al.15

All the pleomorphic, round cell and myxoid type of tumors in this study were sarcomas. This is in accordance with the study on fine-needle aspiration cytology of soft tissue tumors by Rekhi et al.15

Of the 82 soft tissue tumors studied, 55 (18 benign and 37 malignant tumors) were correctly categorized by imprint cytology as benign/malignant. A definitive categorization into benign or malignant was not possible in 13 cases. These were classified into 3 groups – atypical, suspicious of malignancy and inconclusive (Table 4).

Table 3. Analysis of imprint cytology

thumblarge

Table 4. Statistical Analysis

thumblarge

Weighted kappa statistics showed substantial agreement between imprint cytology and histopathology (0.78) (Table 3). Imprint cytology showed a high sensitivity (92.7%) and specificity (94.7%) (Table 5). In comparison, imprint method was reported to have an accuracy of 95.5% by Sakai and Lauslahti,11 93.8% by K C Suen et al12 and 98.4% by Shidham VB et al.6

Table 5. Diagnostic accuracy of imprint cytology

thumblarge

The low false-positive rate observed in this study indicates the reliability of imprint smears in diagnosing malignant soft tissue tumors accurately. In instances when a lesion is grossly malignant as in the case of a large ill-circumscribed tumor with areas of necrosis and hemorrhage, imprint cytology confirms the pathologist’s gross impression.

On the other hand, 2 sarcomas and 2 intermediate grade tumors in the present study were wrongly categorized as benign by cytology. A negative imprint, therefore, does not necessarily exclude malignancy. Imprint smears should always be reported after careful correlation with gross pathological findings. If imprint smears of a grossly malignant appearing lesion are negative, situations that can result in a false-negative report should be carefully considered and ruled out before issuing a negative cytological report.

False-negative reports can occur in cytologically well-differentiated tumors, for example in well differentiated liposarcomas as in the present study, where the morphological changes of the neoplastic cells are often subtle. Also, in tumors with a dense fibrous stroma, large areas of necrosis, hemorrhage and/or cystic change, the cell-yield in imprint smears is often inadequate to provide an accurate diagnosis. In such situations, an erroneous interpretation can be avoided by careful correlation with clinical and radiological features as well as gross pathologic features such as size, location, depth and circumscription of the tumor and presence of necrosis.

The present study is in agreement with Marilyn M Bui et al5 in that cytology alone can accurately diagnose certain entities. Marilyn M Bui et al states that imprint smears are better than the frozen section for myxoid and adipocytic tumors. However, imprint cytology may not be as resourceful in diagnosing tumors having decreased cellularity or increased fibrosis.5 Therefore, the accuracy of intraoperative diagnosis can be enhanced by employing imprint cytology and frozen section together.20,28

Imprint cytology can also be utilized to provide samples for ancillary studies5 and also to ensure the adequacy of a biopsy sample.

 

Conclusions

Imprint cytology serves as an invaluable tool in the intraoperative evaluation of soft tissue tumors. There is substantial agreement (weighted kappa=0.78) between imprint cytology and histopathology.

A diagnosis of malignant soft tissue tumor by the imprint smear method is highly reliable. However, intraoperative diagnosis of soft tissue tumors by imprint cytology is preliminary and warrants confirmation.

A negative diagnosis by imprint cytology does not necessarily exclude malignancy.

An interdisciplinary approach by correlating clinical, radiological, and pathological information can enhance the accuracy of diagnosis by imprint cytology.

 

End Note

Author Information

  1. Dr.Lakshmy M R, Assistant Professor, Department of Pathology, Employees’ State Insurance Corporation Medical College Hospital, Parippally, Kerala, India. Email: lakshmymr1980@gmail.com

  2. Dr.Krishna Balachandran, Professor, Department of Pathology, Government Medical College, Thiruvananthapuram, Kerala, India

 

Conflict of Interest

None declared.

References

1. Weiss S, Goldblum J, Enzinger and Weiss’s Soft Tissue Tumors2009; Philadelphia: Elsevier; 1-12.

2. Christopher DM, Fletcher K, Unni K, Mertens F, World Health Organization Pathology and Genetics of Tumours of Soft Tissue and Bone2002; Lyon: IARC Press; 12-8.

3. Shidham VB, Hackbarth MD, FACS. Benign and Malignant Soft Tissue Tumors2009; St. Louis, MO: Jaypee Brothers Medical Publishers

4. Hoeber I, Spillane AJ, Fisher CM, Thomas JM, Accuracy of biopsy techniques for limb and limb girdle soft tissue tumorsAnn Surg Oncol 2001; 8: 80-7.[CrossRef] [PubMed]

5. Bui MM, Smith P, Agresta SV, Cheong D, Letson GD, Practical issues of intraoperative frozen section diagnosis of bone and soft tissue lesionsCancer Control 2008; 15: 7-12.

6. Shidham VB, Dravid NV, Grover S, Kher AV, Role of scrape cytology in rapid intraoperative diagnosis. Value and limitationsActa Cytol 1984; 28: 477-82.[PubMed]

7. Fang ZW, Chen J, Teng S, Chen Y, Xue RF, Analysis of soft tissue sarcomas in 1118 casesChin Med J (Engl) 2009; 122: 51-3.

8. Bhurgri Y, Bhurgri H, Pervez S, Kayani N, Usman A, Bashir I, Epidemiology of soft tissue sarcomas in Karachi South, Pakistan (1995-7)Asian Pac J Cancer Prev 2008; 9: 709-14.[PubMed]

9. Rydholm A, Gustafson P, Rooser B, Willen H, Berg NO. Subcutaneous sarcoma. A population-based study of 129 patientsJ Bone Joint Surg Br 73-B: 662-7.

10. Gustafson P, Soft tissue sarcoma. Epidemiology and prognosis in 508 patientsActa Orthop Scand Suppl 1994; 259: 1-31.[PubMed]

11. Sakai Y, Lauslahti K, Comparison and analysis of the results of cytodiagnosis and frozen sections during operationActa Cytol 1969; 13: 359-68.[PubMed]

12. Suen KC, Wood WS, Syed AA, Quenville NF, Clement PB, Role of imprint cytology in intraoperative diagnosis: value and limitationsJ Clin Pathol 1978; 31: 328-37.[CrossRef] [PubMed] [PMC Free Article]

13. Sadhana D Mahore, Kalpana A. Bothale, Joshi A, Scrape Cytology in Rapid Intraoperative Diagnosis of TumorsIOSR Journal of Dental and Medical Sciences (IOSR-JDMS) e-ISSN: 2279-0853, p-ISSN: 2279-0861 2015; 14: 65-72.

14. Sachin S Kolte, Rahul N Satarkar, Role of scrape cytology in the intraoperative diagnosis of tumor J Cytol2010; 27: 86-90.[CrossRef] [PubMed] [PMC Free Article]

15. Rekhi B, Gorad BD, Kakade AC, Chinoy R, Scope of FNAC in the diagnosis of soft tissue tumors – A study from a tertiary cancer referral center in IndiaCytojournal 2007; 4: 20.[CrossRef] [PubMed] [PMC Free Article]

16. Kransdorf MJ, Mark D, Murphey Radiologic Evaluation of Soft – Tissue Masses: A Current PerspectiveAvailable from: http://www.ajronline.org.

17. Rosenthal TC, Kraybill W, Soft tissue sarcomas: integrating primary care recognition with tertiary care center treatmentAm Fam Phys 1999; 6: 567-72.

18. Kim K, Phillips ER, Paolino M, Intraoperative imprint cytology: Its significance as a diagnostic adjunctDiagn Cytopathol 1990; 6: 304-7.[CrossRef] [PubMed]

19. Scopa CD, Melachrinou M, Apessou D, Bonikos D, Tissue imprints in surgical pathology: A rapid intraoperative diagnostic aidDiagn Cytopathol 1990; 6: 5-8.[CrossRef] [PubMed]

20. Golouh R, Bracko M, Accuracy of frozen section diagnosis in soft tissue tumorsMod Pathol 1990; 3: 729-33.[PubMed]

21. Bamforth J, Osborn GR, Diagnosis from cellsJ Clin Pathol 1958; 11: 473-82.[CrossRef] [PubMed] [PMC Free Article]

22. Kotilingam D, Lev DC, Lazar AJ, Pollock RE, Staging soft tissue sarcoma: Evolution and changeCA Cancer J Clin 2006; 56: 282-91.[CrossRef] [PubMed]

23. Shirley S, Escoffery C, Usefulness of Touch Preparation Cytology in Postmortem Diagnosis: A Study From the University Hospital of the West IndiesThe Internet Journal of Pathology 2004; 3: 2

24. Dudgeon LS, Patrick CV, A new method for the rapid microscopical diagnosis of tumours, with an account of 200 cases so examinedBr J Surg 1927; 15: 250-61.[CrossRef]

25. Godwin JT, Smears from tissue submitted for frozen sectionsActa Cytol 1968; 12: 85[PubMed]

26. Dodd LG, Marom EM, Dash RC, Matthews MR, McLendon RE, Fineneedle aspiration cytology of “ancient”schwannomaDiagn Cytopathol 1999; 20: 307-11.[CrossRef] [CrossRef]

27. Mavec P, Cytologic diagnosis from tumor tissue using the “quick method” during operationActa Cytol 1967; 11: 229-30.[PubMed]

28. Pickren JW, Burke EM, Adjuvant cytology to frozen sectionsActa Cytol 1963; 7: 164-7.[PubMed]

29. Hajdu SI, Shiu MH, Brennan MF, The role of the pathologist in the management of soft tissue sarcomasWorld J Surg 1988; 12: 326-31.[CrossRef] [PubMed]

30. Bulimbašić S, Perić-Balja M, Intraoperative evaluation of head and neck specimens by imprint cytologyActa Clin Croat 2006; 45: 131-2.

31. Pisters PW, Leung DH, Woodruff J, Shi W, Brennan MF, Analysis of prognostic factors in 1,041 patients with localized soft tissue sarcomas of the extremitiesJ Clin Oncol 1996; 14: 1679-89.[PubMed]

32. Singer S, Demetri GD, Baldini EH, Fletcher CD, Management of soft-tissue sarcomas: An overview and updateLancet Oncol 2000; 1: 75-85.[CrossRef]

33. Klijanienko J, Caillaud JM, Lagacé R, Vielh P, Cytohistologic correlations of 24 malignant peripheral nerve sheath tumor (MPNST) in 17 patients: The institut curie experienceDiagn Cytopathol 2002; 27: 103-8.[CrossRef] [CrossRef]

34. Seidal T, Walaas L, Kindblom LG, Angervall L, Cytology of embryonal rhabdomyosarcoma: A cytologic, light microscopic, electron microscopic, and immunohistochemical study of seven cases Diagn Cytopathol1988; 4: 292-9.[CrossRef] [PubMed]

35. Klijanienko J, Caillaud JM, Lagace R, Vielh P, Comparative fine-needle aspiration and pathologic study of malignant fibrous histiocytoma: Cytodiagnostic features of 95 tumors in 71 patientsDiagn Cytopathol 2003; 29: 320-6.[CrossRef] [PubMed]

36. Tomoko S, Ishiiet M, Kawano J, Akamine T, Kojima T, Ttanaka F, Myxoid liposarcoma and round cell liposarcoma-cytological and morphometric studyJ Jpn Soc Clin Cytol 1999; 38: 504-10.[CrossRef]

37. Ng SB, Ahmed Q, Tien SL, Sivaswaren C, Lau LC, Primary pleural synovial sarcoma.A case report and review of the literatureArch Pathol Lab Med 2003; 127: 85-90.[PubMed]

Leave Your Comment

*