RESUMO
To study the cyto-pathological appearance of intraoperative crush smears of neurosurgical biopsies after subsequent comparison with paraffin sections. Place and duration of study: The study was conducted at Armed Forces Institute of Pathology, Rawalpindi, for one year from February 2002 to February 2003. One hundred, neurosurgical biopsies were received for intraoperative consultation over a period of one year. Clinical information like age, gender, history, and CT scan or MRI findings were recorded. Crush smears were prepared, fixed in 95% alcohol and stained with rapid haematoxylin and eosin stains. Diagnosis was categorised into inflammatory, benign and malignant tumours considering the cyto-pathological features. Remaining tissue was fixed overnight in 10% formalin for histological sections. Permanent Haematoxylin and Eosin [H and E] sections were used as the "gold standard" for comparison. Out of 100 neurosurgical biopsies, crush smears of 94 were considered suitable for interpretation. The smears were categorised as, 8 inflammatory, 41 benign and 41 malignant tumours. Four cases were labelled as inconclusive. Eighty four of the crush smears diagnosis agreed with the histological diagnosis, cyto-morphology of the lesions on crush smear is discussed. Crush smears are useful in the intraoperative diagnosis of space occupying lesions of central nervous system as majority of the cases have typical diagnostic yield and cyto-morphology. We found crush smear cytology highly reliable, rapid and inexpensive mode of intraoperative diagnosis
RESUMO
To evaluate accuracy of intraoperative crush smears diagnosis of neurosurgical biopsies. Cross-sectional study. The Armed Forces Institute of Pathology, Rawalpindi, from February 2002 to February 2003. One hundred, neurosurgical biopsies were received for intraoperative consultation over a period of one year. Clinical information regarding age, gender, history, and CT scan or MRI findings were recorded. Crush smears were prepared, fixed in 95% alcohol and stained with rapid haemotoxylin and eosin [H and E] stains. Diagnosis was categorized into inflammatory, benign and malignant tumours. Remaining tissue was fixed overnight in 10% formalin for histological sections. Permanent H and E sections were used as the gold standard. The average time required for intraoperative cytological diagnosis was 8 minutes. Out of 100 neurosurgical biopsies, crush smears of 94 were considered suitable for interpretation. There were 8 inflammatory, 41 benign and 41 malignant tumours, including 4 inconclusive. Eightyfour of the crush smears diagnosis agreed with the histological diagnosis. Overall diagnostic accuracy of crush smears was 93.3%. Diagnosis of inflammatory, benign and malignant tumours showed specificity of 98.7%, 96% and 94%, and a sensitivity of 70%, 97.5% and 95% respectively. Crush smears are useful in the intraoperative diagnosis of space occupying lesions of central nervous system. The crush smears cytology was found highly reliable, rapid and inexpensive mode of intraoperative diagnosis