Could CCI or FBCI Fully Eliminate the Impact of Curve Flexibility When Evaluating the Surgery Outcome for Thoracic Curve Idiopathic Scoliosis Patient? A Retrospective Study

Abstract

Purpose: To clarify if CCI or FBCI could fully eliminate the influence of curve flexibility on the coronal correction rate.

Methods: We reviewed medical record of all thoracic curve AIS cases undergoing posterior spinal fusion with all pedicle screw systems from June 2011 to July 2013. Radiographical data was collected and calculated. Student t test, Pearson correlation analysis and linear regression analysis were used to analyze the data.

Results: 60 were included in this study. The mean age was 14.7 y (10-18 y) with 10 males (17%) and 50 females (83%). The average Risser sign was 2.7. The mean thoracic Cobb angle before operation was 51.9°. The mean bending Cobb angle was 27.6° and the mean fulcrum bending Cobb angle was 17.4°. The mean Cobb angle at 2 week after surgery was 16.3°. The Pearson correlation coefficient r between CCI and BFR was -0.856(P<0.001), and between FBCI and FFR was -0.728 (P<0.001). A modified FBCI (M-FBCI) = (CR-0.513)/BFR or a modified CCI (M-CCI) = (CR-0.279)/FFR was generated by curve estimation has no significant correlation with FFR (r=-0.08, p=0.950) or with BFR (r=0.123, p=0.349).

Conclusions: Fulcrum-bending radiographs may better predict the outcome of AIS coronal correction than bending radiographs in thoracic curveAIS patients. Neither CCI nor FBCI can fully eliminate the impact of curve flexibility on the outcome of correction. A modified CCI or FBCI can better evaluating the corrective effects of different surgical techniques or instruments.

Fig 1. Show that FFR(r = 0.811) has a higher predict value than BFR(r = 0.523) in predicting CR.

(A) scatter plot with correction rate (CR) against bending flexibility rate(BFR) and the linear regression equation of estimating CR by BFR. (B) scatter plot with correction rate (CR) against fulcrum flexibility rate (FFR) and the linear regression equation of estimating CR by FFR.

Fig 2. Correlation analysis between BFR (bending flexibility rate), FFR (fulcrum flexibility rate), CCI(correction index), fulcrum bending correction index (FBCI) modified FBCI (M-FBCI), modified CCI (M-CCI).

A: CCI showed significant negative correlation with BFR(r = -0.856, P<0.01). B: FBCI showed significant negative correlation with FFR(r = -0.728, P<0.01). Figs 2A,2B show that both CCI and FBCI could not fully eliminate the influence of BFR and FFR. Means that the preoperative curve flexibility will still affect when using CCI or FBCI to compare the curve corrective ability of different apparatus. C: M- CCI showed no significant correlation with BFR(r = 0.123, P = 0.349). D: M-FBCI showed no significant correlation with FFR(r = -0.008, P = 0.950). Figs 2C,2D show that M-CCI and M-FBCI do not have significant correlation with preoperative curve flexibility. Means that the influence of curve flexibility could be eliminated when use these to compare the curve corrective ability of different apparatus.