Clavicle length and scapulothoracic angle in young children with internal rotation contracture secondary to brachial plexus birth palsy: a bilateral MRI study.

BACKGROUND

The most common sequela of brachial plexus birth palsy (BPBP) is an internal rotation contracture (IRC) of the shoulder. The contracture is known to result primarily from restricted motion and deformity at the glenohumeral joint, and, to a lesser extent, decreased humeral retroversion (HRV). Limb length discrepancies of the affected arm are also common, raising the question of whether a shortened clavicle alters the orientation of the scapula into more internal rotation to further limit external rotation.

HYPOTHESIS: We hypothesized that on the affected side of young children with IRCs from BPBP, the clavicle length (CL) and scapulothoracic angle (STA) would be reduced, and their reduction would correlate with each other and the severity of the IRC.

METHODS

In this retrospective study, 24 children with a median age of 1.5 years (range: 0.5 to 10.1), as part of treatment for shoulder IRC secondary to BPBP, underwent magnetic resonance imaging inclusive of both humeri, scapulae, clavicles, and thorax. Measures of bilateral STA, CL, glenoid retroversion, and HRV were made. Interobserver and intraobserver reliability were assessed. Paired t-tests were performed to compare STA, CL, glenoid retroversion, and HRV between involved and uninvolved sides, with P value<.05 considered statistically significant. Spearman correlation coefficients were applied to assess associations between age (at magnetic resonance imaging), IRC severity, ratio of CL (involved/uninvolved), and difference in STA.

RESULTS

On the affected side, CL was significantly reduced (67.2 mm, range: 60.2-124.7, vs. 79.2 mm, range: 57.2-129.0; P < .001), corresponding to 94.5% of the uninvolved side (range: 70.5%-107.1%). STA was also significantly reduced on the involved side (37.8°, range: 27°-53°, vs. 42.7°, range: 27.0°-53.0°; P = .001), with a mean difference of 5°. In this cohort, CL and STA did not correlate with each other or with the severity of the IRC.

CONCLUSIONS

Young children with IRC from BPBP sequelae commonly have reduced CL and an altered STA on the involved side. In this skeletally immature cohort, these differences were not correlated with contracture severity, suggesting that they represent only one aspect of the overall pathoanatomy. Quantitively, these findings constitute a small angular component of the contracture severity but are useful in understanding the appearance of the shoulder girdle and counseling families of young children. A study of older children (after the secondary growth spurt) is warranted to assess whether these changes progress further, accentuating upper limb deformity and limiting motion.