Surgical correction of severe adult lumbar scoliosis (major curves ≥ 75°): Retrospective analysis with minimum 2-year follow-up

Thomas J. Buell, Ching Jen Chen, James H. Nguyen, Peter A. Christiansen, Saikiran G. Murthy, Avery L. Buchholz, Chun Po Yen, Mark E. Shaffrey, Christopher I. Shaffrey, Justin S. Smith

Research output: Contribution to journalArticlepeer-review

17 Scopus citations


OBJECTIVE Prior reports have demonstrated the efficacy of surgical correction for adult lumbar scoliosis. Many of these reports focused on mild to moderate scoliosis. The authors’ objective was to report their experience and to assess outcomes and complications after deformity correction for severe adult scoliosis. METHODS The authors retrospectively analyzed consecutive adult scoliosis patients with major thoracolumbar/lumbar (TL/L) curves ≥ 75° who underwent deformity correction at their institution. Those eligible with a minimum 2 years of follow-up were included. Demographic, surgical, coronal and sagittal plane radiographic measurements, and health-related quality of life (HRQL) scores were analyzed. RESULTS Among 26 potentially eligible patients, 22 (85%) had a minimum 2 years of follow-up (range 24–89 months) and were included in the study (mean age 57 ± 11 years; 91% women). The cohort comprised 16 (73%), 4 (18%), and 2 (9%) patients with adult idiopathic scoliosis, de novo degenerative scoliosis, and iatrogenic scoliosis, respectively. The surgical approach was posterior-only and multistage anterior-posterior in 18 (82%) and 4 (18%) patients, respectively. Three-column osteotomy was performed in 5 (23%) patients. Transforaminal and anterior lumbar interbody fusion were performed in 14 (64%) and 4 (18%) patients, respectively. All patients had sacropelvic fixation with uppermost instrumented vertebra in the lower thoracic spine (46% [10/22]) versus upper thoracic spine (55% [12/22]). The mean fusion length was 14 ± 3 levels. Preoperative major TL/L and lumbosacral fractional (L4–S1) curves were corrected from 83° ± 8° to 28° ± 13° (p < 0.001) and 34° ± 8° to 13° ± 6° (p < 0.001), respectively. Global coronal and sagittal balance significantly improved from 5 ± 4 cm to 1 ± 1 cm (p = 0.001) and 9 ± 8 cm to 2 ± 3 cm (p < 0.001), respectively. Pelvic tilt significantly improved from 33° ± 9° to 23° ± 10° (p < 0.001). Significant improvement in HRQL measures included the following: Scoliosis Research Society (SRS) pain score (p = 0.009), SRS appearance score (p = 0.004), and SF-12/ SF-36 physical component summary (PCS) score (p = 0.026). Transient and persistent neurological deficits occurred in 8 (36%) and 2 (9%) patients, respectively. Rod fracture/pseudarthrosis occurred in 6 (27%) patients (supplemental rods were utilized more recently in 23%). Revisions were performed in 7 (32%) patients. CONCLUSIONS In this single-center surgical series for severe adult scoliosis (major curves ≥ 75°), a posterior-only or multistage anterior-posterior approach provided major curve correction of 66% and significant improvements in global coronal and sagittal spinopelvic alignment. Significant improvements were also demonstrated in HRQL measures (SRS pain, SRS appearance, and SF-12/SF-36 PCS). Complications and revisions were comparable to those of other reports involving less severe scoliosis. The results of this study warrant future prospective multicenter studies to further delineate outcomes and complication risks for severe adult scoliosis correction.

Original languageEnglish (US)
Pages (from-to)548-561
Number of pages14
JournalJournal of Neurosurgery: Spine
Issue number4
StatePublished - Oct 2019
Externally publishedYes


  • Adult spinal deformity
  • Degenerative scoliosis
  • Idiopathic scoliosis
  • Lumbar scoliosis
  • Pseudarthrosis
  • Rod fracture
  • Spine surgery

ASJC Scopus subject areas

  • Surgery
  • Neurology
  • Clinical Neurology


Dive into the research topics of 'Surgical correction of severe adult lumbar scoliosis (major curves ≥ 75°): Retrospective analysis with minimum 2-year follow-up'. Together they form a unique fingerprint.

Cite this