Evaluation of the relevance of touch imprint cytology in the diagnosis of neoplastic lesions of Breast

Assistant Professsor, Kamineni Institute of Medical Sciences, Narketpally, Nalgonda. Head of Department, Department of Pathology, Kamineni Institute of Medical Sciences, Narketpally, Nalgonda. Associate Professor, Kamineni Institute of Medical Sciences, Narketpally, Nalgonda Consultant Pathologist, Yashoda Hospitals, Somajiguda, Hyderabad. Associate Professor, Dr. VRK Women’s Medical College, Aziznagar, Telangana. Associate Professor, Kamineni Institute of Medical Sciences, Narketpally, Nalgonda Head of Department, Department of Blood Transfusion, Kamineni Institute of Medical Sciences, Narketpally, Nalgonda


Introduction
Breast lesions which are more common in the females have gained increasing importance and attained global attention because of increasing mortality and morbidity caused by breast cancer. Clinically, the patients present with a palpable mass, abnormalities in mammography and sometimes with sign of inflammation, nipple discharge 1,2 . Most of the conditions clinically mimic malignancy. Cytology is the study of the morphology of cells that exfoliate naturally and removed by artificial means. Most of the breast lesions including carcinoma are potentially curable, if diagnosed early. For early diagnosis, IC, FNAC and mammography are being used regularly. Some of the diagnostic procedures are preoperative done like mammography, ultrasound, FNA, core needle biopsy. Some diagnostic modalities are used intra operatively like frozen section technique and IC. The diagnosis is to be confirmed by histopathological examination [3][4][5][6][7][8][9][10] . The most important role of diagnostic cytology in the assessment of breast lesions is to make a decision between benign and malignant categories. Observation of normal and abnormal human cells, imprint and exfoliated cells were documented throughout the nineteenth century 11 . Procedures like FNA, core needle biopsy and imprints are useful in sampling of lesions whereas, the use of imaging modalities increases the visualization of lesions 12 .
Breast cytology is generally considered to be a part of initial assessment of breast lesions. It is also used as an adjunct to frozen section in paraffin section histology to assist in reaching a diagnosis 13 . Imprint cytology is a rapid and inexpensive diagnostic procedure which is being used for the diagnosis of lesions at various sites in the body 8,9,10,14 . This procedure has also been applied successfully on breast lesions. It is recommended to use touch preparations to achieve better results 6,[13][14][15][16] . Cytological examination of surgical specimens has proved to be a valuable learning tool. The low percentages of false positive diagnosis suggest that a diagnosis of malignancy by the imprint method is reliable 17 . Intraoperative evaluation of IC can provide the surgeon with information for immediate clinical and surgical management 18 . Cytological examination of imprint or similar preparations is increasing in popularity among pathologists for intra-operative or peri-operative tissue assessment. Several reports in the literature discuss the use of IC in the evaluation of surgical specimens 19 . Although techniques have varied, it seems evident that many authors recognize the usefulness of such cytological study, when employed with appropriate caution, as an adjunct to gross examination and frozen section 20 . A recent report indicates that the accuracy of diagnosis by the IC method is from 94%-97%, almost same as that from the use of frozen section 20 . Imprint cytology has wide applicability in the rapid diagnosis of tumors of various body organs. IJBR (2014) 05 (11) www.ssjournals.com

Material and Methods
This descriptive study on 50 breast neoplastic lesions was conducted at KIMS hospital in collaboration with surgical department to assess the accuracy of IC in breast lesions done in two years. Formalin fixed samples were excluded.
Relevant clinical data regarding age, history and examination was recorded in a proforma. The lesions were identified and cut intra-operatively. Excessive haemorrhagic fluid was washed away. Gross examination of the lesions was recorded. Imprint smears were prepared from the lesion. Imprints were obtained by gently pressing the clean dry glass slides against the cut surface of the lesion and allowed to dry, later fixed in isopropyl alcohol. Touch preparations were stained with rapid H & E. The imprints were examined by pathologists and then the results were informed to the surgeons. After surgery, the specimens were received in the pathology lab, after fixation for 24hrs in formalin grossing were carried out. An automatic tissue processor (Model RH-12 EP Sakura, Fine Technical Co. Ltd., Tokyo, Japan) was used for further processing of tissue blocks. About 17 hours was required for processing by this method. Different concentrations of ethyl alcohol were used for dehydration (70% alcohol for one hour x two changes, 90% alcohol for two hours x two changes and 100% alcohol for two hours x two changes). Clearing in two changes of xylene for one hour each was done. Tissues were impregnated in two changes of paraffin wax with a melting point of 56 o C for a period of 3 hours. Embedding of the tissues was done in paraffin wax using L-shaped metallic moulds. These blocks were put in the refrigerator for a period of 4-6 hours. Each block was cut on a rotatory microtome. About 3-4 micro meter thick tissue sections were obtained and placed in a water bath with a temperature of 5 o C below the melting point of paraffin wax. Cut ribbons of tissues were placed on albuminized glass slides. About 2-3 sections of the tissue were placed on each glass slide. All the sections and touched smears were stained with H & E. The criteria of Yiangou, et al (1996) was for differentiating the benign and malignant breast lesions. The diagnostic classes were as C1inadequate, C2-benign, C3-atypical probably benign, C4-suspicious probably malignant, C5-malignant 7 . Cytodiagnosis was correlated with histopathological diagnosis, using modified Bloom Richardson grading system 22 . Screening tests were applied. Sensitivity, specificity, positive predictive value, negative predictive value and accuracy of IC were calculated considering histopathology as gold standard.

Results
Out of 50 histopathological specimens, only 37 patients underwent FNAC procedure. Histopathological correlation was done in 37cases (there were twenty five benign and twelve malignant lesions). Fibroadenoma cases were from a younger age group. All were mobile, encapsulated and with well defined margins. The color was grey white on cut section and surface was nodular. Histologically, mixed patterns were noted. Imprints revealed cohesive clusters of ductal epithelial cells with finger-like projections. Many naked bipolar nuclei were seen. Mild atypia was present. Fibroadenoma with apocrine change was noted. One case of benign phylloides tumor was noted. This was firm in consistency. Microscopically, leaves-like stromal nodules covered by epithelium projecting into the space were noted. Imprints revealed increased number of spindle-shaped stromal cells.
Twelve malignant cases were studied including cases of infiltrating ductal cell carcinoma. 12 cases were found to be of grade II after applying modified Bloom Richardson grading system. Imprints were well correlated with histological diagnosis. All malignant cases were hard on palpation. No special drug history was noted. On cut section, all the lesions were hard with areas of fibrosis. Cut surface was grey white in all cases. Features like pleomorphism, hyperchromatism and mitosis were clearly discernable on touch imprint as well. Lymphocytes were also appreciated in imprints helping diagnosis of medullary carcinoma. Touch imprints were prepared from twenty five benign and twelve malignant lesions. Cytological diagnosis was divided into five classes from C1 to C5. Smears were adequate in all the cases. In benign cases, it was noted that number of cases with fibroadenoma was predominantly more with touch imprints. In malignant lesions, duct cell carcinoma was easily diagnosed by touch imprints. Cytological diagnosis remained the same with H&E. The cytological diagnosis was classified into three categories i.e. negative, suspicious and positive. For practical purposes, cases with suspicious for malignancy were considered, as positive while calculating the results by applying screening tests. (All the results are shown in Table 1-8and photomicrographs Fig 1-3).   (11) www.ssjournals.com       (11) www.ssjournals.com   (11) www.ssjournals.com

Discussion
Imprint cytology is known since 1927. It was first reported by Dudgeon &Patrick 23 . But recently it has achieved recognition as an adjuvant to frozen section for the intra-operative diagnosis. Intra-operative IC is practically easy, rapid & inexpensive.
The commonly encountered Fibroadenoma, Phylloides tumors, Infiltrating duct cell carcinomas are easy to diagnose by IC. Suen et al 17 has documented 95.8% diagnostic accuracy with 10% false negativity and no false positive cases among 473 breast lesions analysed for IC. Veneti et al 18 studied 351 breast lesions for IC with 98.31% accuracy, 97.1% sensitivity and 99.4% specificity. In our study, on IC no benign lesions were reported as malignant and none of the malignancies were scored as benign. Among the benign lesions, fibroadenoma cases predominated and in malignant lesions, duct cell carcinoma cases were higher in touch imprint cytology. As imprint preparation gives better cellularity, the results of our study were supported by the study conducted by Veneti et al and co workers as they recommended the same.
Rapid Imprint cytological examination of resected tissue has been shown to be a useful adjunct to the use of frozen section as well as an efficient alternate method for reporting fine-needle aspirates 19 . Tung-Kwang Lee 20  In the present study 96% diagnostic accuracy is observed among 50 breast lesions with 2% false negativity and 2% false positivity, 97.3% sensitivity and 92.3% specificity was observed in diagnosing neoplasm of breast. 2% false negativity is due to epitheliosis on imprints which was diagnosed as malignant neoplasm on histopathology. Our analysis is correlated well with all the above mentioned studies.
In cases studied on lumpectomy margins, Klimberg et al 27 supported this method with sensitivity and specificity 100%. Creager et al 27 observed accuracy to be 85%. Cox CE and co-workers also have similar views 29 . In cases of early breast cancer, Saarela et al 30 did not suggest IC as a method of assessing lumpectomy margins. In re-excision specimens, for predicting residual cancer, they had a view that histological margins are the misleading factor. To minimize the risk of residual cancer, they suggested wide excision or mastectomy in the management of multifocal and non-palpable tumor. The present study has not taken lumpectomy margins into consideration.
In benign lesions, counseling and proper treatment plan is suggested to reduce patient anxiety, but this was not recommended by March et al 31 . IC did not provide information about depth of the infiltration but provides information about histological patterns of the tumor as observed by Lee 32 .We observed lymphocytes in case of medullary carcinoma which goes in the favor of conclusion putforth by Lee.
The evaluation of breast lesions in outpatient department by Core Needle Biopsy for IC is recommended as it is a single day procedure. As IC provides the surgeon with information regarding immediate surgical management, this is recommended.

Conclusion
In the present study we observed better Accuracy in IC (96%) than FNAC (89%) in correlation with histopathology. Accordingly, the results show better Sensitivity in IC (97.3%) than FNAC (93.3%) in correlation with histopathology and better Specificity in IC (92.3%) than FNAC (50%) in correlation with histopathology with Positive predictive value -97.29% and Negative predictive value -92.3%. Therefore, IC is a easy, rapid, inexpensive and diagnostic procedure which is being used for the diagnosis of lesions of various sites in the body. IC is an accurate and simple intra-operative method for diagnosing breast lesions which provides prompt information about histological patterns of the tumors. Thus, IC provides initial diagnosis to surgeons for immediate clinical and surgical intervention. Therefore, IC is suggested to be a part of pathologist's repertoire.