Posterior global malalignment after osteotomy for sagittal plane deformity: it happens and here is why

Multicenter, retrospective analysis of 183 consecutive patients undergoing lumbar osteotomy.

To evaluate cause and impact of posterior postoperative alignment.

Sagittal malalignment in the setting of adult spinal deformity (ASD) has shown significant correlation with pain and disability. Surgical treatment often entails correction of deformity by pedicle subtraction osteotomies (PSO). Key radiographical spinopelvic objectives to reach improvement in clinical outcomes have been previously reported. Although anterior alignment is a cause of poor outcomes, the impact and cause of posterior spinal alignment by PSO has not been reported.

The patient inclusion criteria were age, more than 18 years, with a diagnosis of sagittal plane deformity (C7 plumbline offset >5 cm, a pelvic tilt >20°, or a lumbar lordosis to pelvic incidence mismatch of ≥10°) requiring a surgical procedure involving a lumbar posterior osteotomy and a long fusion. Patients were divided into 3 groups based on postoperative sagittal vertical axis (SVA): neutral alignment (0 < SVA < 50 mm), anterior alignment (SVA > 50 mm), and posterior alignment (SVA < 0 mm). All patients underwent pre- and postoperative full-length sagittal spine radiography. Differences between groups were evaluated using ANOVA and χ² analysis.

Seventy-six patients were postoperatively classified in the anterior group: 59 in the neutral group and 48 in the posterior group. These groups were comparable preoperatively in terms of surgical status (revision vs. primary surgery) and regional alignment (lumbar lordosis and thoracic kyphosis). The patients with posterior alignment were younger and had a significantly lower pelvic incidence (53° vs. 62°), preoperative pelvic tilt (30 vs. 36°), SVA (94 vs. 185 mm) and cervical lordosis (16° vs. 25°) than patients in the anterior alignment group. No significant differences were found in terms surgical procedure. Patients in the posterior alignment group demonstrated a significantly greater change in SVA and pelvic tilt correction (P < 0.05) but with a lower gain in thoracic kyphosis (5 vs. 12°) and reduction of cervical lordosis (4° vs. 22°).

A significantly lower pelvic incidence and lack of restoration of thoracic kyphosis may lead to sagittal overcorrection with a posterior alignment. Although the clinical significance of posterior malalignment is still unclear, this study showed a compensatory loss of cervical lordosis in these patients. Particular attention must be paid to preoperative planning before sagittal realignment procedures. Further study will be necessary to evaluate long-term clinical outcomes of these patients.