Selective versus nonselective thoracic fusion in Lenke 1C curves: a meta-analysis of baseline characteristics and postoperative outcomes.

Abstract

Lenke 1C curves are challenging to manage surgically due to the structural thoracic deformity and nonstructural lumbar curve. Selective thoracic fusion (STF) is considered the standard of care because it preserves motion of the lumbar segment, yet nonselective STF (NSTF) remains prevalent. This study aims to identify baseline patient characteristics that drive treatment and to compare postoperative outcomes for both procedures. Studies that compared baseline and postoperative demographic data, health-related quality of life (HRQL) questionnaires, and radiographic parameters of patients with Lenke 1C curves undergoing STF or NSTF were identified for meta-analysis. The effect measure is expressed as a mean difference (MD) with 95% CI. A positive MD signifies a greater STF value, or a mean increase within the group. One prospective and 6 retrospective case-control studies with sample size of 488 patients (344 STF and 144 NSTF) were identified. Baseline age, sex, and HRQLs were equivalent, except for better scores in the STF group for the Scoliosis Appearance Questionnaire (SAQ): Unrelated to Deformity item (3.47 vs 3.88, p = 0.01) and the Spine Research Society questionnaire, Item 22: Pain (4.13 vs 3.92, p = 0.04). Radiographic findings were significantly worse in NSTF, as measured by the thoracolumbar/lumbar (TL/L) Cobb angle (MD: -4.29°, p < 0.01) and TL/L apical vertebral translation (AVT) (MD: -6.08, p < 0.01). Radiographic findings significantly improved in STF, as measured in the main thoracic (MT) Cobb angle (MD: -27.78°, p < 0.01), TL/L Cobb angle (MD: -16.24°, p < 0.01), MT:TL/L Cobb ratio (MD: -0.21, p < 0.01), coronal balance (MD: 0.47, p = 0.02), and thoracic kyphosis (MD: 7.87°, p < 0.01); and in NSTF in proximal thoracic (PT) Cobb angle (24° vs 14.1°, p < 0.01), MT Cobb angle (53.5° vs 20.5°, p < 0.01), and TL/L Cobb angle (41.6° vs 16.6°, p < 0.01). Postoperative TL/L Cobb angle (23.1° vs 16.6°, p < 0.01) was significantly higher in STF; but PT Cobb angle, MT Cobb angle, and MT:TL/L Cobb ratio are equivalent. Patients with larger lumbar compensatory curves displaying a larger degree of coronal translation, as measured by the TL/L AVT, are more likely to undergo an NSTF. Contrary to established guidelines, larger MT curve magnitudes and MT:TL/L Cobb angle ratios have not been found to influence the decision to pursue a selective thoracic fusion. Although overall both STF and NSTF groups are found to have effective postoperative coronal balance, the STF group has only modest improvements in the lumbar curve position as determined by a relatively unchanged TL/L AVT. Furthermore, surgeons may prefer NSTF in patients who may have a worse overall perception of their spinal deformity as measured by HRQL measures of pain and desire for appearance change.