How to rectify the convex coronal imbalance in patients with unstable dystrophic scoliosis secondary to type I neurofibromatosis: experience from a case series

Background There was a paucity of valid information on how to rectify the convex coronal imbalance effectively in dystrophic scoliosis secondary to Type I neurofibromatosis (DS-NF1), while postoperative inadvertent aggravation of CCI occurred regularly resulting in poor patient satisfaction. We aimed to identify the risk factors for persistent postoperative CCI in DS-NF1, and to optimize the coronal rebalancing strategies based on the lessons learned from this rare case series. Methods NF1-related scoliosis database was reviewed and those with significant CCI (> 3 cm) were identified, sorted and the outcomes of surgical coronal rebalance were analyzed to identify the factors being responsible for failure of CCI correction. Results CCI with dystrophic thoracolumbar/lumbar apex was prone to remain uncorrected (7 failure cases in 11) when compared to those with thoracic apex (0 failure cases in 4) (63.6% vs. 0.0%, p = 0.077). Further comparison between those with and without post-op CCI showed a higher correction of main curve Cobb angle (65.9 ± 9.1% vs. 51.5 ± 37.3%, p = 0.040), more tilted instrumentation (10.3 ± 3.6° vs. 3.2 ± 3.1°, p = 0.001) and reverse tilt and translation of upper instrumented vertebra (UIV) to convex side (8.0 ± 2.3° vs. -3.4 ± 5.9°, p < 0.001; 35.4 ± 6.9 mm vs. 12.3 ± 13.1 mm, p = 0.001) in the uncorrected imbalanced group. Multiple linear regression analysis revealed that △UIV translation (pre- to post-operation) (β = 0.832; p = 0.030) was significantly correlated with the correction of CBD. Conclusion Thoracolumbar/lumbar CCI in dystrophic scoliosis was prone to suffer high risk of persistent post-op CCI. Satisfying coronal rebalance should avoid UIV tilt and translation to the convex side, tilted morphology of instrumentation and over correction maneuvers for main curve, the upper hemi-curve region in particular.