Could screw/hook insertion at the apical vertebrae with rib head dislocation effectively retract the corresponding rib head from spinal canal in dystrophic scoliosis secondary to type 1 neurofibromatosis?

BACKGROUND: Rib head dislocation (RHD) in dystrophic scoliosis of type 1 neurofibromatosis (DS-NF1) is a unique disorder caused by skeletal dystrophy and scoliotic instability. No particular surgical manipulation is mentioned in the literature to instruct the spine surgeons to effectively obtain more migration of the dislocated rib head without resection. The present study aimed to investigate the effectiveness of screw/hook insertion at vertebrae with RHDs on the retraction of penetrated rib head from spinal canal. METHODS: 37 neurologically intact patients with DS-NF1 and concomitant 53 RHDs undergoing scoliosis surgery without rib head excision were retrospectively reviewed. We used pre and postoperative whole-spine radiographs to determine the Cobb angle and the vertebral translation (VT), and the CT scans to evaluate the intraspinal rib length (IRL) and rib-vertebral angle (RVA). The dislocated ribs were assigned into two groups according to the presence of screw/hook insertion at vertebrae with RHD: screw/hook group and non-screw/hook group. RESULTS: 37 dislocated ribs with screws/hooks insertion at corresponding vertebrae were assigned into the screw/hook group and the remaining 16 dislocated ribs consisted of the non-screw/hook group. In the screw/hook group, the correction rates of Cobb angle and VT were significantly higher than the non-screw/hook group after surgery (58.7 ± 16.0% vs. 30.9 ± 12.4%, p = 0.003; 61.8 ± 18.8% vs. 35.1 ± 16.6%, p = 0.001; respectively). Similarly, more correction rates of IRL and RVA were found in the screw/hook group than the non-screw/hook group (63.1 ± 31.3% vs. 30.1 ± 20.7%, p = 0.008; 17.6 ± 9.7% vs. 7.2 ± 3.6%, p = 0.006; respectively). Multiple linear regression analysis revealed that the correction rates of Cobb angle, VT and RVA contributed significantly to correction of IRL (β = 0.389, 0.939 and 1.869, respectively; p = 0.019, 0.001 and 0.002, respectively). CONCLUSION: Screw/hook insertion at dystrophic vertebrae with RHDs contributed significantly to the degree of retraction of penetrated rib head from spinal canal. This effectiveness is mediated by more corrections of VT and RVA.