Selective thoracolumbar fusion in adult spinal deformity double curves with circumferential minimally invasive surgery: 2-year minimum follow-up

Abstract

OBJECTIVE Selection of the upper instrumented vertebra (UIV) level for adult spinal deformity (ASD) remains controversial. Although selective fusion attempts have been described for fractional curves or adolescent curves, no authors have described selective thoracolumbar fusion performance for ASD with double curves. This study evaluated the clinical impact of selective fusion constructs within the lower thoracic and/or lumbar spine on ASD with double curves. METHODS A retrospective review was performed on an ASD (Cobb angle > 20°, sagittal vertical axis [SVA] > 50 mm, and pelvic incidence minus lumbar lordosis mismatch [PI-LL] > 10°) database consisting of 438 patients who underwent correction with circumferential minimally invasive surgery (CMIS) between 2007 and 2020. The inclusion criteria were ASD double curves (lumbar Cobb angle > 35° and thoracic Cobb angle > 30°), 4 or more levels fused, and minimum 2-year follow-up. Analyses were performed on spinopelvic data and clinical outcome scores. Complications were recorded, specifically the need for revision surgery and hardware-related complications. RESULTS Twenty-one ASD double curve patients underwent selective correction with a mean ± SD (range) follow-up of 91 ± 43 (24–174) months. A total of 141 levels were fused with a mean of 6.7 ± 1.3 (4–8) levels. T10 was the most proximal and most common UIV (10/21 [48%]). Pelvic fixation was performed in 12 patients (57%). Significant improvements in lumbar Cobb angle, thoracic Cobb angle, coronal balance, lumbar lordosis, thoracic kyphosis, SVA, and PI-LL were achieved. The uninstrumented thoracic spine demonstrated 14.5° of mean coronal correction and a mean increase of 9.4° in kyphosis. Significant improvements in visual analog scale (VAS) and Oswestry Disability Index (ODI) scores were observed. Four patients required revision for the following reasons: 1) superficial wound infection requiring irrigation and debridement; 2) bilateral L5 pars fractures requiring L5–S1 anterior lumbar interbody fusion and pelvic fixation; 3) adjacent-segment degeneration at L5–S1 requiring anterior lumbar interbody fusion and pelvic fixation; and 4) proximal junctional kyphosis requiring revision fusion to include the entire thoracic curve. There were no instances of hardware failure such as rod breakage or screw loosening. CONCLUSIONS Selective thoracolumbar fusion with CMIS for ASD double curves can provide significant clinical improvements. Despite limiting fusion constructs to within the lower thoracic and/or lumbar spine, significant correction can be observed in the uninstrumented thoracic curve. The rate of mechanical complications was low, and the 2-year follow-up results suggested that limited fusion constructs are viable options for ASD double curve patients.