Complications and radiographic correction in adult scoliosis following combined transpsoas extreme lateral interbody fusion and posterior pedicle screw instrumentation.

Object: The authors recently used a combined approach of minimally invasive transpsoas extreme lateral interbody fusion (XLIF) and open posterior segmental pedicle screw instrumentation with transforaminal lumbar interbody fusion (TLIF) for the correction of coronal deformity. The complications and radiographic outcomes were compared with a posterior-only approach for scoliosis correction.

Methods: The authors retrospectively reviewed all deformity cases that were surgically corrected at the University of Pittsburgh Medical Center Presbyterian Hospital between June 2007 and August 2009. Eight patients underwent combined transpsoas and posterior approaches for adult degenerative thoracolumbar scoliosis. The comparison group consisted of 4 adult patients who underwent a posterior-only scoliosis correction. Data on intra- and postoperative complications were collected. The pre- and postoperative posterior-anterior and lateral scoliosis series radiographic films were reviewed, and comparisons were made for coronal deformity, apical vertebral translation (AVT), and lumbar lordosis. Clinical outcomes were evaluated by comparing pre- and postoperative visual analog scale scores.

Results: The median preoperative coronal Cobb angle in the combined approach was 38.5 degrees (range 18-80 degrees). Following surgery, the median Cobb angle was 10 degrees (p < 0.0001). The mean preoperative AVT was 3.6 cm, improving to 1.8 cm postoperatively (p = 0.031). The mean preoperative lumbar lordosis in this group was 47.3 degrees, and the mean postoperative lordosis was 40.4 degrees. Compared with posterior-only deformity corrections, the mean values for curve correction were higher for the combined approach than for the posterior-only approach. Conversely, the mean AVT correction was higher in the posterior-only group. One patient in the posterior-only group required revision of the instrumentation. One patient who underwent the transpsoas XLIF approach suffered an intraoperative bowel injury necessitating laparotomy and segmental bowel resection; this patient later underwent an uneventful posterior-only correction of her scoliotic deformity. Two patients (25%) in the XLIF group sustained motor radiculopathies, and 6 of 8 patients (75%) experienced postoperative thigh paresthesias or dysesthesias. Motor radiculopathy resolved in 1 patient, but persisted 3 months postsurgery in the other. Sensory symptoms persisted in 5 of 6 patients at the most recent follow-up evaluation. The mean clinical follow-up time was 10.5 months for the XLIF group and 11.5 months for the posterior-only group. The mean visual analog scale score decreased from 8.8 to 3.5 in the XLIF group, and it decreased from 9.5 to 4 in the posterior-only group.

Conclusions: Radiographic outcomes such as the Cobb angle and AVT were significantly improved in patients who underwent a combined transpsoas and posterior approach. Lumbar lordosis was maintained in all patients undergoing the combined approach. The combination of XLIF and TLIF/posterior segmental instrumentation techniques may lead to less blood loss and to radiographic outcomes that are comparable to traditional posterior-only approaches. However, the surgical technique carries significant risks that require further evaluation and proper informed consent.