Spinopelvic morphology impacts on postoperative proximal junctional kyphosis in congenital scoliosis with thoracolumbar hemivertebrae.

Abstract

Purpose: It aims to investigate the lumbar and pelvic morphology in congenital scoliosis with thoracolumbar hemivertebrae and its impact on proximal junctional kyphosis (PJK) incidence after hemivertebra resection and short fusion.

Methods: 23 congenital scoliosis patients with thoracolumbar hemivertebra aged between 10 and 18 years were enrolled in the retrospective study. Spinopelvic sagittal parameters were analyzed on whole-spine standing lateral radiographs preoperatively, one-week postoperatively and at the final follow-up. Pearson correlations were calculated for local kyphosis (LK), lumbar and pelvic morphology parameters. Binary logistic regression and receiver operating characteristics (ROC) curve analysis were performed to identify the risk factors for PJK.

Results: Thoracolumbar hemivertebra caused LK of 29.2° ± 17.3°, an increased lumbar lordosis (LL) (-64.7° ± 16.3°), lower LL apex (52.2% at L5), and small pelvic incidence (PI) (36.8° ± 6.6°). LK was correlated with lumbar morphology parameters, including LL (r = - 0.837), upper arc of LL (LLUA) (r = - 0.879), thoracolumbar kyphosis (TLK) (r = 0.933), thoracic kyphosis (TK) (r = 0.762) and TK apex (TKA) (r = - 0.749). Surgical treatment improved the lumbar morphology, but not pelvic morphology. At the final follow-up, LL had returned to its preoperative value (p = 0.158). PJK occurred in 30.4% of cases as a compensatory mechanism. Preoperatively, significant differences of parameters between non-PJK and PJK groups were observed in LK and TLK. Binary logistic regression identified three independent risk factors for PJK: preoperative LLA (OR = 0.005, 95%CI = 0.000-0.287, p = 0.011), preoperative TLK (OR = 1.134, 95%CI = 1.001-1.286, p = 0.048), and preoperative lumbar lordosis morphology type (OR = 5.507, 95%CI = 1.202-25.227, p = 0.028). However, residual LK after surgery was not correlated with PJK incidence. ROC curve analysis verified that preoperative TLK > 22.59° was associated with increased PJK incidence after surgery.

Conclusions: Lumbar morphology changes as a compensatory mechanism beneath the thoracolumbar hemivertebra. However, a stable pelvis tends to allow the LL to return to its preoperative value. PJK occurred as a cranial compensatory mechanism for increasing LL and corrected TLK. A larger TLK (> 22.59°) was an independent risk factor for PJK incidence in patients with type 2 and 3A lumbar lordosis morphology.