Lumbo-pelvic related indexes: impact on adult spinal deformity surgery.

Purpose: Lumbo-pelvic indexes appeared recently in the literature taking advantage from the relationship between pelvic incidence (PI) and lumbar lordosis (LL). Schwab proposed to subtract LL from PI (PI-LL) as Boissière proposed the lumbar lordosis index (LLI), which is the ratio between LL and PI (LL/PI). Both indexes have been described to weight LL by a constant parameter not affected by degenerative processes, the PI. The aim of this study is to evaluate these parameters in adult spinal deformity (ASD) by analyzing their relationship with spinal malalignment and vertebral osteotomies.

Methods: Two groups of patients with an ASD were realized; an adult scoliosis group (n = 78) and a postoperative flat-back syndrome group (n = 20). In the adult scoliosis group, 28 patients underwent an osteotomy [pedicle subtraction osteotomy (PSO) or Smith Petersen osteotomy] and 50 patients were corrected by posterior fusion without osteotomy. In the postoperative flat-back syndrome group all patients underwent a PSO. All patients had preoperative and postoperative full spine EOS radiographies to measure spino-pelvic parameters. The lack of lordosis was calculated, after prediction of theoretical LL from Legaye's formula, by subtracting measured LL to theoretical LL. Correlation analysis between the different parameters was performed.

Results: Both lumbo-pelvic parameters highly correlated with spinal malalignment (r = 0.97 for PI-LL and r = -0.97 for LLI for total patients) and were highly predictive of a spinal osteotomy performance (r = 0.88 for PI-LL >28° and r = 0.94 for LLI <0.5). Sagittal vertical axis (r = 0.67) and pelvic tilt (r = 0.64) correlated moderately with spinal malalignment for total patients. The LLI was more correlated with spinal osteotomies in the adult scoliosis group (r = 0.86 for PI-LL >28° and r = 0.94 for LLI <0.5), as Schwab's index was more precise to predict osteotomies in the postoperative flat-back syndrome group (Youden index = 0.95 for PI-LL >28° vs 0.90 for LLI <0.5).

Conclusions: This study highlights the necessity to considerer spinal malalignment with lumbo-pelvic indexes as they appear to be highly correlated with lack of LL. They can be used as mathematical tools to detect spinal malalignment in ASD and guide the surgeon's decision of realizing a vertebral osteotomy for ASD sagittal correction. They can be used as well for the interpretation of clinical series in ASD.