Imaging analysis of balanced hemivertebrae in congenital scoliosis in children

English Keywords:Spinal deformity  Children  Hemivertebrae  Imaging

English Abstract:

  【Abstract】 Objectives: To analyze the imaging characteristics of balanced hemivertebrae in children with congenital scoliosis, so as to provide guidance and basis for clinical practice. Methods: The clinical data of 52 congenital scoliosis children with balanced hemivertebrae undergoing surgery between December 2015 and June 2020 in our hospital were retrospectively analyzed. The mean age of the patients was 3.5 years(2.9, 7.3 years). There were 28 males and 24 females. Radiographic parameters were measured on full-spine X-ray before operation, including coronal parameters such as cranial/caudal segmental Cobb angle and compensatory curve, pelvic tilt(PT), and coronal sacral slope(CSS), and sagittal parameters such as thoracic kyphosis(TK), thoracolumbar kyphosis(TLK), sacral slope(SS), and lumbar lordosis(LL). The deformity conditions of anterior hemivertebrae and posterior elements were determined via CT and MRI as unsionous and discordant deformity. Spinal deformity was classified and imaging characteristics were summarized according to the number and position of hemivertebrae and combined spinal/ribs deformity. The deformity was classified into four types: type 1, 2 hemivertebraes; type 2, 3 hemivertebraes; type 3, ≥4 hemivertebraes; type 4, balanced hemivertebrae combined with complicated spine/ribs deformity(Type 1-3 may combine with mild spine/ ribs deformity). Type 1 and 2-4 were classified into non-complicated and complicated deformity, respectively. t test and rank-sum test were applied to compare the main coronal and sagittal parameters, and Chi-square test was used to compare the combined spine/ribs deformity conditions. Results: A total of 130 hemivertebrae with an interval of 3.4 normal vertebrae on average were found in 52 patients including 6(4.6%) cervical, 29(22.3%) upper thoracic, 44(33.8%) lower thoracic, and 51(39.2%) lumbosacral hemivertebrae. Hemivertebraes were separated by mean 3.4 normal vertebrae. Unisonous deformity was found in 110 hemivertebrae(84.6%) in comparison with 20(5.4%) discordant deformity. There were 31 cases classified into type 1, 11 cases into type 2, 4 cases into type 3, and 6 cases into type 4. Cranial and caudal segmental Cobb angle was 32.7°±8.1° and 34.4°±9.0°(t=-0.905, P=0.367), and cranial and caudal compensatory curve was 12.8°±6.6° and 9.7°±6.9°(t=2.308, P=0.023). Data of cranial/caudal segmental Cobb angle and compensatory curve, PT, CSS, TK, SS and LL showed no significant difference(P>0.05) between the complicated and non-complicated types. TLK in complicated and non-complicated deformity was 16.2°±16.4° and 6.4°±22.6°, respectively (t=-1.693, P=0.047). Rib and intraspinal deformity in type 2-4 was higher than that in type 1 (52.4% vs 12.9%, χ2=9.506, P=0.002). The discordant rate of anterior vertebrae and posterior lamina was 25.7% in types 3 and 4 deformity and 11.6% in types 1 and 2 deformity, respectively (χ2=3.926, P=0.048). Conclusions: The deformity of balanced hemivertebrae was complicated and most hemivertebrae distributed in lumbosacral and lower thoracic spine. The influence caused by deformity was less significant in coronal plane than that in sagittal plane, especially in the thoracolumbar region. The bigger the number of hemivertebrae, the more complicated the deformity.