MRI of lumbosacral spine for low back pain ---- correlation with clinical presentation and plain radiograph

Loading...
Thumbnail Image
Date
2002
Authors
Yong, Pei Yee
Journal Title
Journal ISSN
Volume Title
Publisher
Universiti Sains Malaysia
Abstract
Study design. This was a combined retrospective and prospective study of patients who were referred for MR1 of lumbar spine for low back pain. Objective. To determine the correlation between clinical presentation, plain radiograph signs and MRI of lumbosacral spine in patients with low back pain. Summary of background data. MRl has become the modality of choice in the investigation of causes of low back pain. However, to date, there are not many studies which looked into the correlation between the clinical presentation as wel1 as plain radiograph signs and the MRI findings in low back pain. Methods and materials. Fifty-seven patients who were referred for MRI of the lumbosacral spine for low back pain to look for disc herniation or spinal stenosis were studied. Their clinical data, symptoms and signs were obtained from the medical case record. Standard anteroposterior and lateral lumbosacral radiographs with or without oblique views were done. Plain lumbosacral MR1 according to the department protocol were performed. The images were interpreted by a musculoskeletal radiologist. The plain radiographs and MR images for each patient were read at different time and the radiologist was blinded to patient· s data and clinical presentation. All the data collected were analysed using SPSS for Windows version 9.0 software. Descriptive analysis was performed for demographic data, clinical presentation, frequency of plain radiograph and MRI findings. Kappa statistic was used to study the agreement between findings on plain radiograph and MRI. The correlation between clinical parameters, as well as plain radiograph findings with MRI outcome was also performed. Results. The patients in this study were in the high productivity age group with a mean age of 44.78 years. Fifty-six percent of patients presented with chronic low back pain more than 3 months duration. This study showed that the presence of sensory deficit predicts the level of spinal stenosis. Both sensory and motor deficit are not accurate predictor of nerve root compression. A good linear correlation was observed between age and level of disc degeneration. Univariate analysis showed that there was no significant correlation between clinical presentation with disc herniation, nerve root compression or spinal stenosis. However, from multivariate analysis, age, duration of pain and presence or absence of sciatica were found to be significant predictors of spinal stenosis. The commonest plain radiograph finding was reduction of posterior disc height (less than 6mm). It correlated significantly with nerve root compression and spinal stenosis but not with disc herniation. There was also significant correlation between posterior osteophytes, end plate sclerosis, end plate irregularity, vacuum phenomena. facet arthropathy, spondylolysis and spondylolisthesis with disc herniation, nerve root compression and spinal stenosis. However, due to the small sample size, larger scale study may be necessary to further confirm these observations. The sensitivity of plain radiograph was 92. 7o/o and its positive predictive value was 96.2o/o, but its specificity was poor. Conclusion. Low back pain affects mainly those of high productivity age group. A good linear correlation was observed between age and disc degeneration. There was no significant correlation between age, clinical symptoms and signs with disc herniation, nerve root compression and spinal stenosis on MRI. Nevertheless, age, duration of pain and sciatica are predictors of spinal stenosis. Posterior intervertebral disc height of less than 6mm was significantly related to nerve root compression and spinal stenosis. Posterior osteophytes, end plate changes, vacuum phenomena, degenerated facet joints, spondylolysis and spondylolisthesis also showed significant correlation with disc herniation, nerve root compression and spinal stenosis. Plain lumbosacral radiograph was sensitive but not specific for the investigation of low back pain.
Description
Keywords
Clinical presentation, plain radiograph signs and MRI of lumbosacral spine
Citation