[Necessity of intraoperative three-dimensional navigation for low lumbar spine: a randomized controlled study]

To compare the accuracy of lower lumbar pedicle screw placement at different levels with or without navigation system.

A prospective randomized controlled trial was conducted for patients undergoing open lumbar surgery with pedicle screws at both L4 and L5 levels from March 2011 to May 2012. The accuracy of pedicle screws at L4 and L5 level were evaluated by three-dimensional postoperative computed tomography (CT). There was no apparent violation of pedicle or < 2 mm perforation of pedicle in non-violation group while ≥ 2 mm perforation of pedicle in violation group.

A total of 188 patients were recruited and assigned randomly into navigation (n = 47) and control (n = 141) groups. Two groups did not differ in estimated blood loss volume, length of hospital stay, Japanese Orthopedics Association (JOA) score or JOA recovery rate (P > 0.05). Navigation group had a longer operative duration (P < 0.05). At L4 level, 93/94 screws (98.9%) in navigation group and 273/282 screws (96.8%) in control group did not violate pedicle (P > 0.05). At L5 level, 93/94 screws (98.9%) in navigation group and 260/282 screws (92.2%) in control group did not violate pedicle. Thus navigation could achieve a higher accuracy at L5 level (P < 0.05).

The profile of L5 pedicle at coronal plane is tilting oval. Screw may easily violate medioinferior part of pedicle and cause injures of nerve root. For improving the accuracy and safety of pedicle screw at L5 level, we recommend using intraoperative three-dimensional navigation for spondylolisthesis or facet hypertrophy, iliac cohesion or junior operator.