Devoted to diagnostic and interventional spine imaging and therapeutics

Library

Analysis of Post-Vertebroplasty Spine MRI According to Origin of Recurrent Back Pain 2004

Dansie, David, M.D., Luetmer, Patrick, M.D., Kallmes, David, M.D., Lane, John, M.D., Thielen, Kent, M.D., Wald, John M.D.

Purpose:
We describe the spectrum of MRI findings of vertebrae treated with vertebroplasty and attempt to demonstrate whether there are MRI imaging features that allow prediction of whether a treated vertebra is a source of post-procedural back pain.

Methods and Materials:
Forty-five patients with osteoporotic compression fractures had vertebroplasty at our institution and had a postprocedural MRI for the evaluation of recurrent back pain prior to October 2002. Each patient’s medical record and imaging studies were reviewed by three physicians, who made a consensus clinical determination of the pain source ( group 1--pain has a definite source other than the treated vertebra, group 2--pain source is indeterminate, group 3--pain source is the treated vertebra). Three neuroradiologists reviewed the sagittal T1 and T2 FSE images of the preprocedural (when available) and the follow-up MRI, and the post-vertebroplasty plain films. By consensus opinion they graded the following features: percent loss of vertebral body height; percent of marrow volume edematous; percent of marrow volume filling with cement; presence of a cement-filled intravertebral cleft, and if so whether there was persistent cleft fluid; percent of non-cemented marrow volume edematous; change in total marrow edema; conversion of marrow from non-edematous to edematous (and vice-versa); and gross loss of height between the vertebroplasty and the follow-up MRI.

Results:
Only one patient had pain clearly attributed to the treated vertebra. This patient had a transverse process fracture of the treate d vertebra. This patient’s imaging was reviewed but not included in the statistical analysis. The remaining 44 patients had a total of 76 treated vertebrae treated prior to the initial follow-up MRI. These were used in the statistical analysis. Thirty-one patients had pain attributed to pathology away from the treated vertebra, the majority of these were new compression fractures. Each imaging feature was compared with the patient’s clinical categorization of pain source. For continuous variables, the two sample t-test or Wilcoxon rank sum test was performed as appropriate. For dichotomous variables, the chi square or Fisher’s exact test was performed as appropriate.

Fifty-two vertebrae were treated in group 1, and 24 in group 2 patients. No MRI finding of treated vertebrae differed significantly between groups 1 and 2, although it is likely some patients in group 2 (indeterminate pain source) had undiagnosed pain originating from the treated leve l(s). In both groups, a significant number of vertebrae (17% of group 1 vs 21% of group 2) lost height between the vertebroplasty and the follow-up MRI. As expected, most vertebrae in both groups demonstrated either moderate (involving >1/3 but <2/3 of marrow volume) or severe (involving >2/3 of marrow volume) vertebral edema on preprocedure MRI (62% of group 1 vs. 47% of group 2). Also, in both groups there was a trend toward interval decreased marrow edema between the preprocedure MRI and the follow-up MRI. However, many vertebrae in both groups (34% of group 1 vs 33% of group 2) continued to have either moderate or severe edema on follow-up. In both groups, there were vertebrae with specific regions of marrow that converted over the imaging interval from edematous to non-edematous, and vice-versa.

Conclusions:
The spectrum of MRI findings in vertebrae clinically successfully treated with vertebroplasty (our group 1 patients) includes persistent edema and interval loss of height. These findings are also found in patients with recurrent, undiagnosed back pain. Although our study lacked statistical power due to low patient numbers, we were unable to identify any statistically significant difference in these MRI features between the two clinical groups. Because significant marrow edema and further loss of height of treated vertebrae can be seen in patients from both clinical groups, interpretation of these findings should be done with caution, and these findings should not be assumed to represent the source of post-vertebroplasty back pain.