Magnetic Resonance Imaging Evaluation of Common Spinal Intramedullary Tumours: Ependymoma and Astrocytoma

N Sultana,S Jabeen,S Rima,U K Nag,S K Sarkar
Abstract:This cross-sectional study was carried out in the Department of Radiology and Imaging, Bangabandhu Sheikh Mujib Medical University (BSMMU), Dhaka, Bangladesh, from January 2015 to June 2016 in collaboration with Department of Neurosurgery and Department of Pathology of same hospital, to assess MRI findings of common IMSCTs and to find out the validity of MRI in the diagnosis of spinal ependymoma and astrocytoma, thereby differentiating these two common intramedullary entities. For this purpose, 40 patients aged from 15 to 60 years, who were diagnosed or suspected cases of having intramedullary spinal cord tumours were enrolled. These patients underwent preoperative MRI for evaluation of the spinal cord tumours during the study period in department of Radiology and Imaging. Also, patients were diagnosed by MRI as having IMSCTs incidentally was included. All of them were followed by histopathological examination of the same lesions after surgery. Among these 40 patients after exclusion for valid reason 28 cases were included as study population. MR images were obtained on a 1.5 Tesla (Avanto Magnatom, Siemens) unit with a spine surface coil. MRI findings were compared with histopathology after surgery keeping histopathology as gold standard. Out of 28 clinically and MRI diagnosed IMSCTs cases, 19 cases of ependymoma, 8 cases of astrocytoma and one diagnosed as hemangioblastoma by MRI. The mean age was 34.11±9.55 years, ranging from 15 to 56 years for ependymoma, 26.88±8.08 years ranging from 16 to 44 years for astrocytoma. Highest incidence (47.4%) was in between 31 to 40 years for ependymoma, 50.0% were in between 21-30 for Astrocytoma. On MRI majority 12(63.2%) of the cord ependymomas and majority 5(62.5%) of astrocytomas were in cervical region. While considering axial location ependymomas are mostly 17(89.5%) central and astrocytomas 5(62.5%) are eccentric in location. It was observed that out of 19 cases of ependymoma more than half 10(52.6%) had elongated shape, 12(63.1%) had well defined margin. Associated syringohydromyelia was present in 16(84.2%) cases. On T1WI 11(57.9%) and 8(42.1%) cases were iso and hypo respectively. On T2WI 14(73.7%) cases were hyper intense. After Gd-DTPA in most cases, 13(68.4%) cases showed diffuse enhancement. Noticeable and sizeable solid component observed among 13(68.4%) of the cases. Hemorrhage with cap sign was found in more than one third 7(36.8%) cases. Out of 8 cases of astrocytomas 4(50.0%) had lobulated shape, ill-defined margin 5(62.5%). T1WI: Iso 5(62.5%), hypo 3(37.5%), T2WI: hyper 5(62.5%), After Gd-DTPA: focal and heterogenous enhancement 3(37.5%) and rim enhancement 4(50.0%). Component: mixed 4(50.0%), cystic 3(37.5%) and solid 1(12.5%). Hemorrhage without cap sign 2(25.0%), associated syringohydromyelia 1(12.5%). In the case of evaluation of intramedullary ependymoma sensitivity of MRI in the present series is 94.44%, specificity 80.0%, Positive predictive value (PPV) 89.5%, Negative predictive value (NPV) 88.9% and accuracy 89.28%. In case of MRI evaluation of intramedullary astrocytoma sensitivity of MRI in the present study is 85.71%, specificity 90.47%, PPV 75%, NPV 95% and accuracy 89.2%. Thus present study shows MRI is a sensitive and effective noninvasive imaging modality in diagnosing common intramedullary spinal cord tumours.
What problem does this paper attempt to address?