Evaluation of the Squash Smear Technique in the Rapid Diagnosis of Central Nervous System Tumors: A Cytomorphological Study
N Pawar, K Deshpande, S Surase, G D’costa, S Balgi, A Goel
Keywords
cns tumor, glioma, squash cytology
Citation
N Pawar, K Deshpande, S Surase, G D’costa, S Balgi, A Goel. Evaluation of the Squash Smear Technique in the Rapid Diagnosis of Central Nervous System Tumors: A Cytomorphological Study. The Internet Journal of Pathology. 2009 Volume 11 Number 1.
Abstract
Introduction
The intraoperative cytology preparation was first introduced by Eisenhardt and Cushing in early 1930 and by Badt in 1937.(1,2) This technique was further championed and documented by Russell
Methodology
The present study was conducted over a period of 6 months. It was conducted in the Department of Neuropathology, Grant Medical College and Sir J.J. Hospital, Mumbai, India. It included 50 cases of CNS neoplasms which were subjected to intraoperative smear cytology. The biopsy samples obtained at the time of surgery were transported immediately to the neuropathology laboratory in isotonic saline for processing. A smear slide was prepared by taking 1-2 millimeters (mm) of the biopsy material with the scalpel blade, placing the material on a slide and crushing with another slide with just enough pressure to spread the tissue into a thin film. It was then fixed in 95% alcohol and stained by Hematoxylin and Eosin (H & E). Further, paraffin sections were prepared by the residual tissue and stained by H & E. Relevant clinical and radiological data were noted. Smear cytology diagnosis was correlated with the histopathological findings. The tumours were classified according to the World Health Organization classification of CNS neoplasm 2007.(6) The observations were then subjected to appropriate statistical analysis methods.
Results
The cases where the intraoperative cytological diagnosis was same as the histological diagnosis including the grade of the malignant tumor were considered as the complete correlation. Of the total number of cases two cases were inconclusive i.e. they were neither diagnosed by histopathological examination (HPE) nor by squash smear cytology. Forty-eight cases gave a definitive finding on HPE. Of these, complete correlation with squash smear cytology was obtained in forty-four cases. Further, the grading of four cases was misinterpreted by smear technique. Thus, on statistical analysis the overall diagnostic accuracy obtained in our study was 88%. The sensitivity obtained was 91.6% while, the specificity and positive predictive value obtained were 100%. The results obtained on subjecting the cases to both histopathological and cytological techniques are shown in table 1.
A significant drop was observed in the diagnostic accuracy of oligodendroglioma, distinguishing between pure astrocytoma, oligodendroglioma and mixed glioma (Table: 2).
Four cases were misinterpreted on cytological examination. But they showed partial correlation, comprising 3 cases of discrepancy of grading and 1 case where one of the components had missed during intraoperative cytological diagnosis. (Table 3)
The most common tumors in our study were Astrocytomas I & II (low grade), which contributed to 50% of the cases. These moderately cellular neoplasms showed minimal anisocytosis with fine to coarse granular chromatin and inconspicuous nucleoli, the cytoplasm was scanty and showed variable processes with fibrillary background (Fig 1).
Figure 4
Astrocytoma III & IV (Glioblastoma Multiforme) contributed to 22% of cases. The evenly spread smears of glioblastomas showed cytological heterogeneity with high cellularity. The cells were closely related to blood vessels in a ‘perivascular gradient pattern’ along with endothelial cell proliferation, glomeruloid bodies, marked cellular pleomorphism, atypia, multinucleation, mitotic figures and necrosis (Fig 2). Anaplastic astrocytoma (Grade III), in addition showed distinct papillary pattern with masses of closely packed cells having hyperchromatic and pleomorphic nuclei, coarse chromatin and prominent nucleoli. Mitotic figures and increased vascular proliferation were also noted.
Figure 5
Oligodendrogliomas (8%) smeared well and showed moderate cellularity comprised of uniform round cells with well-defined cell margins in a non-cohesive pattern. Cytoplasm was clear with fine granular background, which lacked fibrillary material. Few cases showed prominent thin walled vessels and foci of calcification.
Ependymomas (8%) were highly cellular tumors and showed cells in papillary pattern along the blood stroma and many layers of cells lying on it. Perivascular pseudorosettes were prominent. The cells had scanty cytoplasm and round to oval nuclei with fine granular chromatin and conspicuous nucleoli.
Discussion
In our study, the overall diagnostic accuracy obtained on using squash smear technique was 88%. A similar diagnostic accuracy was obtained by a study done in France,(10) while, other studies have reported a diagnostic accuracy varying from 86% to 97.3%. (4, 5, 7, 8, 10, 11,12) Further, in our study a significant reduction in the diagnostic accuracy of oligodendroglioma was seen. This finding correlated with the findings obtained by a study done in Austria.(4) The probable reason for this could be sampling error or the dense fibrillary background of astrocytic component.(13)
The squash smear techniques reported a sensitivity of 91.6%. A similar study done in Brazil reported this finding as 97.9%.(11) As stated by some studies the reason cited for this low value could be significant variance observed in grading different areas of smears of astrocytomas. Some authors have also mentioned that in the undergraded cases there are areas of both less and more aggressive astrocytomas and the cytological sampling fails to show the anaplastic component. Further, literature states that small biopsies are not suitable for grading of gliomas.(7,11,12) A study done in France concludes that multiple biopsies from different area might lessen the false positivity and false negativity.(10) The specificity and positive predictive value obtained was 100%. This correlated with the findings obtained in a study done in Brazil which reported a specificity and positive predictive value of 95% and 99.1% respectively.(11)
In our study, final diagnosis of two cases failed to show concordance with the final diagnosis. This could be due to the causes like biopsy taken from cyst wall, increased fibrous component, lack of architecture on cytology, morphology obscured by necrosis and inflammation, reactive changes and resistance to desegregation.
Conclusion
Intraoperative smear cytological diagnosis is fairly accurate, safe, simple reliable and cost effective tool for rapid diagnosis of CNS gliomas. It is a preferred method because it offers great details of cellular morphology avoiding distortion and ice artifacts often introduced by frozen section technique. Intraoperative smears permit rapid and reliable diagnosis of CNS gliomas, which helps the surgeon to monitor and modify the approach of surgery. The newer investigating modalities and approaches have placed the neuropathologist in a key role in the diagnosis, further management and treatment of patients with CNS lesions. Nevertheless, it is very essential to review the clinical data as well as radiological, CT and MRI findings before intraoperative evaluation of squash imprint of brain lesions.
Correspondence to
Dr. Nilesh H. Pawar,
492-2, Sri Vithal Rakumai Colony, Jyotibanagar, Kalewadi, Pimpri, Pune -411017, MH
Email: