|Abstract / Summary|
Early-onset scoliosis (EOS) is a heterogeneous condition ranging from mild and resolving to severe and progressive deformity. EOS can be classified according to the underlying etiology as congenital/structural, neuromuscular, syndromic or idiopathic. If left untreated, progressive EOS may result in pulmonary compromise and increased mortality. Early fusion of the thoracic spine can lead to a restrictively decreased pulmonary function. Therefore, various distraction-based techniques and implants have been developed to allow continuous growth of the spine, while trying to control the deformity. However, they require repetitive surgical procedures with high complication and infection rates. To avoid this, a magnetically control growing rod (MCGR) was recently developed, which allows non-invasive lengthening procedures while partly correcting the deformity. The technique is relatively new, and generally there is not yet consensus on the surgical indications, timing of surgery and distraction technique. Moreover, the implant-related complication rate remains high and the consequences and long-term results of the treatment are sparsely investigated. The Classification of Early-Onset Scoliosis (C-EOS) is a newly developed classification system trying to standardize research and practice within EOS. However, the content and cut-off values are to some extend based on expert consensus and the reproducibility has not been thoroughly assessed.
The objectives of this thesis were to examine reliability and accuracy of the C-EOS. Secondly, to assess the curve correction and complication rate in the MCGR treatment and the efficacy of a distraction-to-stall principle compared with a targeted distraction principle. Lastly, to investigate the bone mineral density (BMD) in the vertebrae within the instrumentation.
The first study included 60 patients in a reproducibility study of C-EOS. Agreement for etiology was substantial, however, disagreement was found regarding syndromic patients and patients with neural axis abnormalities. Reliability of major curve angle (MCA) was excellent while reliability for kyphosis and annual progression rate (APR) were lower due to larger measurement errors. The limits of agreement for APR exceeded the 10°/year increments suggested in the original paper describing the C-EOS.
The second study included 19 patients treated with MCGR. Major curve angle was reduced from 76° preoperatively to 44° postoperatively. Flexibility predicted curve correction, which was maintained throughout follow-up. The distraction-to-stall principle resulted in satisfactory lengthening with spinal height increase comparable to results published in the literature.
The third study included 39 patients treated with MCGR at two centers using either a distraction-to-stall (group 1) or a targeted distraction principle (group 2). The two groups were comparable besides a mean 18 days longer time interval between lengthening in group 2. We found no difference in achieved distraction. The only variable independently associated with greater achieved distraction within the first year was larger preoperative MCA.
The fourth study included measurements of BMD in 11 patients treated with MCGR. We found a lower BMD in vertebrae spanned by the instrumentation compared with total hip BMD and height-for-age normative values. But the results need to be verified in a longitudinal study.
In conclusion, the C-EOS is a reliable classification system for future research in EOS although minor revisions are required. MCGR effectively corrects the coronal deformity, and distraction-to-stall and targeted distraction are equivalent lengthening principles.