Parameters leading to a successful radiographic outcome following surgical treatment for Lenke 2 curves

In Lenke 2 curves, there are conflicting data when to include the PTC into the fusion. Studies focusing on Lenke 2 curves are scant. The number of patients with significant postoperative shoulder height difference (SHD) or trunk shift (TS) is as high as 30 % indicating further research. Therefore, the purpose of the current study was to improve understanding of curve resolution and shoulder balance following surgical correction of Lenke 2 curves as well as the identification of radiographic parameters predicting postoperative curve resolution, shoulder and trunk balance in perspective of inclusion/exclusion of the proximal thoracic curve (PTC).

This is a retrospective study of a 158 Lenke 2 curves. Serial radiographs were analyzed for the main thoracic curve (MTC), PTC, and lumbar curve (LC), SHD, clavicle angle (CA), T-1 tilt, deviation of the central sacral vertical line (CSVL) off the C7 plumb line.Patients were stratified whether the PTC was included in the fusion (+PTC group, n = 60) or not (−PTC group, n = 98). Intergroup results were studied. Compensatory mechanisms for SHD were studied in detail. Adding-on distally was defined as an increase of the lowest instrumented vertebra adjacent disc angle (LIVDA) >3°. Stepwise regression analyses were performed to establish predictive radiographic parameters.

At follow-up averaging 24 months significant differences between the +PTC and –PTC group existed for the PTC (24° vs 28°, p < .01), PTC correction (42 vs 29 %, p < .01), rate of MTC-loss >5° (27 vs 53 %, p < .01), and spontaneous LC correction in patients with a selective thoracic fusion (STF) (80/93 %, p = .04). The number of patients with a new trunk shift (CSVL > 2 cm) was 9 (6 %): 7 in the −PTC vs 2 in the +PTC group (p = .03). Utilization of compensatory mechanisms (99 vs 83 %, p < .01) and adding-on (35 vs 20 %, p < .05) occurred more often in the +PTC vs the –PTC groups. Statistics showed postoperative SHD improvement in both the +PTC and –PTC groups. There were no significant differences regarding SHD, CA and T1-Tilt between groups. However, only in the –PTC group, a significant change between postoperative and follow-up SHD existed (p = .02). Statistics identified a preoperative ‘left shoulder up’ (p < .01) and CSVL (p = .03) predictive for follow-up SHD ≥1.5 cm. A statistical model only for the –PTC group showed 9 parameters highly predictive for a follow-up SHD ≥1.5 cm with highest prediction strength for a PTC >40° (p = .01), a preoperative ‘left shoulder up’ (p < .01) and anterior fusion (p = .02). To account for baseline differences between the +PTC and –PTC groups, 49 matched-pairs were studied. Postoperative differences remained significant between the +PTC and –PTC groups for the PTC (p < .01), MTC (p = .03) and the rate of loss of MTC >5° (p < .01).

Prediction of a successful surgical outcome for Lenke 2 curves depends on multiple variables, in particular a preoperative left shoulder up, preoperative PTC >40°, MTC correction, and surgical approach. Shoulder balance is not significantly different whether the PTC is included in the fusion or not. But, powerful compensation mechanisms utilized to balance shoulder in the –PTC group can impose changes of trunk alignment, main and compensatory lumbar curves.