Which Factors Predict Shoulder Asymmetry in Patients with Lenke Type 1 and 3 Curves following Pedicle Screw Instrumentation?

Abstract

Introduction Shoulder asymmetry (SA) has been reported to be a frequent (39%) problem after fusion of the main thoracic (MT) curve in patients with nonstructural proximal thoracic (PT) curves (Lenke Type 1 and 3). The purpose of this study was to analyze factors related to shoulder asymmetry and efficacy of traction X-ray under general anesthesia (TrUGA) in predicting postoperative shoulder asymmetry. Material and Methods A total of 137 (128 female patients, 9 male patients) consecutive patients (average age: 14.2 [11–17]) with Lenke Type 1 and Type 3 curves and preoperative right shoulder elevation (RSE) with (-) T1 tilt were included in this study. Radiographic analysis included preoperative, postoperative, and follow-up AP standing, supine bending, and TrUGA. Magnitudes of curves, T1 tilt, clavicle angle, shoulder height, first rib angle (FRA), and second rib angles (SRA) were measured as shoulder asymmetry parameters. Left or right shoulder elevation > 1 cm is accepted as shoulder imbalance. SRS-22r and shoulder balance questionnaires were also analyzed at follow-up. Results Mean age was 13.96 years (11–17) and average follow-up was 5.21 years (2–12). A total of 107 patients had UIV at T2, 17 patients had UIV at T3, and 13 had UIV at T4. Average correction rates were 69.12% for PT curve, 82.4% for MT curve, and 73.5% for TL/L curve. Postoperatively, 101 patients (73%) had balanced shoulders. A total of 17 patients (12.4%) had LSE and 20 patients (14.6%) had RSE postoperatively. Incidence of LSE was 8.4% (9 patients) for UIV at T2, 29.4% (5 patients) for UIV at T3, and 23.1% (3 patients) UIV at T4. Higher percentage of MT correction was correlated with LSE ( p = 0.043). Shoulder height, clavicula angle, T1 tilt, FRA, and SRA were correlated in TrUGA and at follow-up. Preoperatively, 83% of the patients and at final follow-up 7.8% patients thought they had shoulder asymmetry. Av SRS-22r score was 4.32 (2, 96–5) and was similar in patients who had shoulder asymmetry (4, 36) versus no asymmetry (4, 29). Conclusion Risk factors for shoulder asymmetry in patients with nonstructural PT curves are higher correction percentage of MT and extension of instrumentation up to T3 at PT curve. Extension of fusion up to T2 did not always prevent shoulder asymmetry in patients with nonstructural PT curves. Shoulder asymmetry parameters in TrUGA are helpful in predicting postoperative shoulder asymmetry. HRQL scores are similar in patients with SA versus no SA.