Effect of mismatch between pelvic incidence and lumbar lordosis on sagittal imbalance of ankylosing spondylitis following kyphosis correction

English Keywords:Ankylosing spondylitis  Kyphosis  Pelvic incidence and lumbar lordosis mismatch  Sagittal imbalance  Osteotomy

English Abstract:

  【Abstract】 Objectives: The current study aimed to investigate the effect of pelvic incidence and lumbar lordosis mismatch(PI-LL) on sagittal imbalance of ankylosing spondylitis(AS) following corrective osteotomy. Methods: 85 AS patients(75 men, 10 women) who underwent modified pedicle subtraction osteotomy(PSO) between January 2010 and October 2019 were enrolled in this study, with an average age of 37.1±8.8 years(20-67 years). 63 underwent one-level osteotomy, and 22 underwent two-level osteotomy. The preoperative, postoperative(3-4 weeks after surgery) and final follow-up radiographic parameters were measured, including pelvic incidence(PI), pelvic tilt(PT), sacral slope(SS), lumbar lordosis(LL), PI-LL, osteotomized vertebral angle(OVA), and sagittal vertical axis(SVA). The function of limbs and health-related quality of life before operation and at the final follow-up were evaluated using Scoliosis Research Society-22(SRS-22) patient questionnaire and Oswestry disability index(ODI). At the final follow-up, SVA>5cm was regarded as spinal imbalance, and PT>25° was considered pelvic imbalance. Correlations between postoperative PI-LL and final follow-up SVA and PT were analyzed using Pearson correlation coefficient. Threshold of PI-LL was figured out using receiver operating characteristic(ROC) curve and maximum Youden index to obtain the optimal postoperative PI-LL. Patients were divided into two groups by whether meeting the optimal postoperative PI-LL, and effects of different postoperative PI-LL on spino-pelvis sagittal alignment at the final follow-up were compared and evaluated. Subsequently, linear regression was used to evaluate the relationship between lumbar OVA and PI-LL correction to figure out the regression equation. Results: 85 AS patients were followed up for 30.8±6.3 months(24-84 months). The final follow-up LL(-31.6°), PT(31.8°), SS(15.5°), PI-LL(16.7°) and SVA(8.6cm) were improved significantly(P<0.05); and the final follow-up ODI(23.45%) and SRS-22 score(3.91) were better than those before surgery(P<0.05). Postoperative PI-LL positively correlated with the final follow-up SVA and PT(r=0.525 and 0.659 respectively, P<0.01). Calculated with SVA using ROC and maximum Youden index, the optimal threshold of postoperative PI-LL was 12.8°; while, calculated with PT, it was 10.5°. Postoperative PI-LL of ≤10.5° met the needs for preventing both spinal and pelvic imbalance. Patients with postoperative PI-LL≤10.5° had better PT(25.4° vs 36.6°), LL(-40.8° vs -24.1°), PI-LL(4.0° vs 26.2°) and SVA(5.6cm vs 10.9cm) than those with postoperative PI-LL>10.5°, and significantly deceased the rate of spinal (36.1% vs 75.5%) and pelvic(38.9% vs 85.7%) imbalance at the final follow-up(P<0.05). Lumbar OVA highly correlated with PI-LL correction(r=0.707, P<0.001), and the equation was Lumbar OVA=17.12+0.62×(PI-LL correction), adjusted R2=50.1%. Conclusions: Postoperative PI-LL strongly correlated with final follow-up SVA and PT in AS patients after kyphosis correction. Reconstructing postoperative PI-LL ≤10.5° could achieve better spino-pelvis sagittal alignment for AS kyphosis, and decrease the risk of spinal and pelvic imbalance in mid- to long-term follow-up.