Impact of thoracolumbar inflection point on outcomes and complications in adult spinal deformity.

Abstract

Objective: Existing literature on the impact of alignment parameters relative to the thoracolumbar inflection point remains sparse. The authors aimed to investigate the influence of the inflection point, lumbar lordosis apex (LLA), and other alignment parameters on complications, reoperations, and clinical outcomes.

Methods: Patients with adult spinal deformity who underwent fusion of the lower thoracic spine (T7-12) to pelvis, for whom 2-year data were available, were included. Parameters relative to the inflection point, the inflection point from the upper instrumented vertebra (UIV), the LLA, and the theoretical inflection point based on pelvic incidence in the normative populations from Roussouly morphology were assessed. Patients were evaluated based on inflection point changes of at least 1 level from baseline and stratified into caudal, same, and cranial groups. The analysis controlled for invasiveness, baseline deformity, frailty, and PJK prophylaxis.

Results: A total of 435 patients (mean age ± SD 65 ± 8 years) were included in the study. The mean baseline inflection point was at the L1-2 interbody space, which was corrected more cranial to the L1 body after surgery. Development of proximal junctional failure (PJF) or proximal junctional kyphosis (PJK) with reoperation was associated with a more caudal baseline inflection point (L2 inferior endplate vs L1 body, p < 0.001). In the adjusted analysis, patients with a more caudal baseline inflection point had 25% higher odds of developing PJF by 2 years (OR 1.26 [95% CI 1.08-1.46], p = 0.003). Postoperative normalization to the theoretical Roussouly inflection point had decreased rates of rod breakage (4.5% vs 9.3%, p = 0.049) but higher rates of PJF (13.1% vs 7.7%, p = 0.044). Stratifying patients based on inflection point change from baseline, rates of PJK and PJF 2 years after surgery were higher in the cranial group (both p < 0.003), with no difference in meeting the Roussouly target inflection point. Compared with the cranial group, there was 4.4 times lower odds of developing PJF in the caudal group and 2.0 times lower odds in the group with the same inflection point (p < 0.05). Increased distance from the UIV to the inflection point was associated with mechanical complications (MCs) (p < 0.05). Decreased distance between the LLA and inflection point at baseline was associated with MC (p = 0.04).

Conclusions: Although correction of the inflection point to normative values decreased rates of MC, PJK rates remain high suggesting other factors being at play. These correlations between the inflection point and the UIV and LLA underscore the pivotal role of the inflection point in achieving adequate realignment.