Diagnostic accuracy of frozen section in detecting malignant brain tumors taking histopathology as gold standard.

摘要

Objectives: To evaluate the diagnostic accuracy of frozen section in detecting malignant gliomas, taking histopathology as gold standard. Study Design: Cross-Sectional study. Setting: Neurosurgery Department, Aga Khan University Hospital, Karachi. Period: From March 2018 to Jan 2019. Material & Methods: Through consecutive sampling technique, total 100 specimens from patients suspected of malignant gliomas on MRI brain with contrast, who were admitted in the neurosurgery department for elective tumor surgery were included. Intraoperatively, the specimen from the lesion identified on imaging through neuro-navigation was taken and sent to histopathology department for frozen section analysis. The histopathology consultant was pre-informed about the case and probable time of receiving the specimen. The specimen for frozen section was sent in a dried container without formalin. Once received in pathology department, the tissue is rapidly cooled through a cryostat which converts tissue water into the ice and makes the tissue rigid for cutting into slices and viewing under the microscope for identifying the neoplastic tissue. The consultant histopathologist was request to provide frozen section report immediately and to keep the remaining specimen for definitive histopathology reporting as well. Data of malignancy on both Frozen Section and formal histopathology report was noted and comparison was done. Results: Mean age of the patients was 36.7±8.76 years. There were 64 (64%) male patients and 36 (36%) female patients. Diagnostic accuracy of frozen section taking histopathology as a gold standard shows that sensitivity was found to be 83.3%, specificity 84.4%, Positive predictive value (PPV) 71.4% and negative predictive value (NPV) 91.5% & Overall diagnosis accuracy was found 96.6%. Conclusion: The Diagnostic accuracy of frozen section (FS) in detection of brain malignant gliomas was found to be satisfactory. It has a diagnostic accuracy of 96.9%, sensitivity 83.3%, specificity 84.4%, Positive predictive value (PPV) 71.4% and negative predictive value (NPV) 91.5%.