Redefining Proximal Junctional Kyphosis After Posterior Spinal Fusion for Idiopathic Scoliosis: PJA >12 Is Associated With Increased Pain and Decreased Mobility, But Not Reoperation

Introduction: Proximal junctional kyphosis (PJK) occurs at moderate rates following posterior spinal fusion (PSF) in idiopathic scoliosis (IS). However, the impact of preoperative sagittal alignment and the clinical consequences of PJK, particularly effects on Patient-Reported Outcome Measurement Information System (PROMIS) scores, remain underreported. Methods: A query of an institutional pediatric spinal deformity database identified patients undergoing primary PSF for IS with a minimum 2-year follow-up. Preoperative and latest follow-up radiographs, as well as PROMIS scores, were analyzed. PJK was defined as a final upper instrumented vertebra (UIV)+2 kyphosis angle ≥10 degrees and a change of ≥10 degrees from the preoperative measurement. Results: Eighty patients (mean age: 15 y; range: 11 to 18) met the inclusion criteria; mean follow-up was 3.1 years (range: 2 to 4). No patients underwent revision surgeries. PJK was present in 13% (10/80) of patients. Preoperatively, in the PJK group, there was a significantly lower mean coronal deformity (49.5 vs. 58.2 degrees; P =0.003), pelvic tilt (4 vs. 9 degrees; P =.01), pelvic incidence (43 vs. 52 degrees; P =.02), and higher T1 slope (26 vs. 18 degrees; P =.01). Postoperatively, the PJK group showed significantly lower upper extremity scores (45 vs. 52; P =0.049) and higher mean T1 slope (31 vs. 18 degrees; P =0.0002). Significant changes (preoperative to final) between groups included mobility scores (no-PJK: +4 vs. PJK: −10; P =0.03) and pain interference (no-PJK: 0 vs. PJK: +9; P =0.02). No other PROMIS domains (peer relations, anxiety, etc.), UIV, age, or sex were significantly different between the groups. A logistic regression model from this cohort demonstrated changes in pain interference associated with postoperative PJA ≥12. Conclusion: Thirteen percent of IS patients undergoing PSF had the radiographic diagnosis of PJK, but no patient underwent revision surgery. Preoperative risk factors included smaller major coronal deformity, lower pelvic incidence and tilt, and higher T1 slope. Postoperatively, PJK patients had significant increases in pain and worsening mobility versus non-PJK patients. The current radiographic definition of PJK, though not associated with reoperations, appears clinically valid, as its presence negatively impacts PROMIS mobility and pain interference scores. However, a logistic regression model utilizing our patients’ postoperative proximal junctional angles (PJA) and changes in PROMIS pain interference scores from preoperative to postoperative suggests 12 degrees as an optimal cutoff for differentiating between PJK and non-PJK. Further study is needed to corroborate these findings and better understand how PJK affects function. Level of Evidence: Level II.