Abstract
Introduction
Endplate disruption is a significant risk factor of cement leakage during percutaneous vertebral augmentation. This research is to compare the helpfulness of computerized tomography (CT) and magnetic resonance imaging (MRI) in diagnosing endplate disruption for osteoporotic vertebral compressive fractures (OVCFs).
Materials and Methods
Between January 2013 and January 2014, 60 patients with single-level OVCF were enrolled in this study. Of them, 54 were female and 6 were male patients. Mean age was 69.6 years (54–92 years). Mean disease duration was 44 days (7–120 days). Mean bone mass density was −3.9 (−2.5 to ∼ − 6.7). Endplate disruption was defined so that integrity and continuity of endplate was disrupted on the sagittal and coronary CT images and endplate hypointensity discontinued on sagittal T2-weighted or T2-weighted short tau inversion recovery sequences images. For the first step, two spinal surgeons with knowledge of the standard of endplate disruption in CT and MRI images analyzed both CT and MRI images together at the same time to decide whether endplates of responsible level were disrupted and marked the location of endplate disruption if it existed. The diagnosing result from CT–MRI images combined was referred as the golden standard. Another four spinal surgeons were then divided into two groups to analyze the CT or MRI images blindly to decide whether endplates of responsible level were disrupted and marked the location of endplate disruption if it existed. The two spinal surgeons in each group discussed to reach an agreement in diagnosis if they did not agree at the beginning. Chi-square test and kappa test were used to analyze the difference and consistency in the usefulness in diagnosing the endplate disruption of responsible level between CT–MRI images combined and CT images, between CT–MRI images combined and MRI images, and between CT images and MRI images, respectively.
Results
There was no significant difference in the comparison of the usefulness between CT–MRI combined and CT in diagnosing the endplate disruption (p > 0.05). The concordance rate was 91.7% (55/60) and the discordance rate was 8.3% (5/60). Missed diagnosis rate for CT was 11.9% (5/42). The ability of CT–MRI and CT in diagnosing the endplate disruption had a good consistency (kappa = 0.82, p < 0.05). There was also no significant difference in the comparison of usefulness between CT–MRI combined and MRI in diagnosing the endplate disruption (p > 0.05). The concordance rate of diagnosis was 88.3% (53/60) and the discordance rate was 11.7% (7/60). Missed diagnosis rate for MRI was 7.1% (3/42) and misdiagnosed rate was 22.2% (4/18). The ability of CT–MRI images and MRI images in diagnosing the endplate disruption also had a good consistency (kappa = 0.72, p < 0.05). Furthermore, there was no significant difference in the comparison of usefulness between CT and MRI in diagnosing the endplate disruption (p > 0.05). The concordance rate of diagnosis was 80% (48/60) and the discordance rate was 5% (12/60). The usefulness in diagnosing the endplate disruption between CT and MRI had a moderate consistency (kappa = 0.555, p < 0.05).
Conclusion
Both CT and MRI could diagnose the endplate disruption for OVCFs precisely and they had the same usefulness. Specificity of CT was better than that of MRI and sensitivity of MRI images was better than that of CT.