Correlation of Susceptibility Weighted Imaging Findings of Intracranial Gliomas With Histopathological Findings at Kenyatta National Hospital

Abstract

Background: Gliomas are heterogeneous group of neoplasms and constitute the most common overall malignant brain tumor. Conventional Magnetic Resonance Imaging (MRI) provides excellent anatomic details and soft tissue contrast. Despite this, MRI may not accurately diagnose or grade intracranial tumors in all instances especially for glioma grading. Due to significant morbidity and mortality associated with brain biopsies including complexity of surgery, cost and economic burden on the patients, accurate preoperative diagnosis is important. Several advanced MRI methods and sequences such as Susceptibility Weighted Imaging, Diffusion Weighted Imaging, Perfusion Weighted Imaging and Spectroscopy have therefore been introduced for narrowing down differential diagnosis and grading of gliomas. Internationally, several studies have been done to demonstrate the utility of Susceptibility Weighted Imaging in assessment of gliomas. However, no local study has been done to assess the utility of Susceptibility Weighted Imaging (SWI) in assessment of gliomas. Aim: To determine diagnostic accuracy of Susceptibility weighted imaging in evaluation of brain gliomas using histopathology as gold standard. Methodology: This was a retrospective and prospective cross-sectional study and was conducted in Kenyatta National Hospital Radiology, Neurosurgical, Health Records departments and UON pathology department following approval from KNH-UON ERC. Administrative approval was sort for the retrospective arm to access the medical records of histologically proven brain gliomas which were correlated with SWI findings in patients who met the inclusion criteria. For the prospective part of the study, informed consent was obtained from participants who met the inclusion criteria and the findings were correlated with histopathological results. Study subjects were 47 patients who were diagnosed with a glial tumor based on histopathological examination and who had had MRI done with SWI sequence included and met the inclusion criteria. Relevant data of eligible patients was collected. Data was recorded in data collection tool and cleaned. The collected data was entered and analysed in SPSS version 23.0. The variables analysed included demographics (i.e. age, sex), positive signs in conventional MR and SWI findings. Descriptive data was presented as means with standard deviation, median and mode. SWI findings were summarised and presented using percentages. 2x2 contingency tables were used to calculate diagnostic accuracy data which will be presented as sensitivity, specificity, positive and negative predictive values. Conclusion: SWI has a role in preoperative grading of gliomas in combination with other advanced MRI techniques and conventional MRI studies. The study established, the addition of SWI sequence in reviewing MR images to have overall diagnostic accuracy of 74.5% sensitivity of 84.2%, specificity of 67.9%, positive predictive value of 64.0% negative predictive value of 86.4%. in grading of intracranial gliomas with histopathology being the gold standard. By identifying both haemorrhagic and calcific foci, SWI can help in characterization and grading of gliomas. SWI provides invaluable information about pathophysiology and internal tumour architecture of gliomas. This information is vital in preoperative grading of tumours