[Analyses and comparison of 2-D vs. 3-D glenoid versions in atraumatic/traumatic shoulder instability--are 3-D...

Changes in glenoid orientation as a primary cause of shoulder instability have been discussed controversially in the literature. The data of a physiological glenoid version vary widely among different authors and techniques. One reason may be that the previously used 2-D techniques suffer from a limited reproducibility and validity. The objective of this study was therefore to compare the 2-D and 3-D analyses of the glenoid version in patients with shoulder instability.

The shoulders of 28 healthy volunteers and of 14 patients each with atraumatic/traumatic instability were examined in an open MR scanner (0.2 T). The 2-D glenoid version was determined using post-processing techniques according to the technique of Friedman et al. (1992). Afterwards, the 3-D glenoid version was analysed independently of the slice orientation and patient position. The coefficient of correlation (r) between the 2-D and 3-D glenoid versions was calculated using the correlation z test.

The 3-D post-processing technique showed a reproducibility with a coefficient of variation of 8.3 %. Patients with traumatic instability demonstrated no significant difference compared to the healthy control group (4.4 +/- 2.1 degrees vs. healthy: 3.9 +/- 1.3 degrees). In atraumatic shoulder instability the glenoid retroversion was in the mean significantly increased (10.2 +/- 4.9 degrees). The individual values ranged between 2.6 degrees and 16.6 degrees . Also for the contralateral, unaffected side a significantly increased retroversion (6.3 +/- 2.2 degrees) was observed compared to healthy shoulders. There was a significant correlation (r: 0.84) between 2-D and 3-D retroversion.

The presented techniques allow for a reproducible assessment of glenoid version independent of the slice orientation and patient position. Our results demonstrate in the mean only a small difference of +/- 3 degrees between 2-D and 3-D glenoid versions. Therefore under standardised conditions the 2-D CT/MRI should be adequate for measuring the glenoid version except for borderline cases. No significant changes in glenoid version were found in patients with traumatic instability. In atraumatic, posterior instability, in the mean an increased retroversion was observed on both sides. However, the magnitude of these changes varied widely among individuals and should be identified to initiate a causal treatment.