Distinct Mechanisms of Proximal Junctional Kyphosis and Their Clinical Implications

Retrospective cohort study. In patients undergoing adult spinal deformity (ASD) surgery, we sought to: (1) describe mechanisms of proximal junctional kyphosis/failure (PJK/F), and (2) compare time-to-diagnosis, proximal junctional angle (PJA), reoperation, and rate of neurological deficits between PJK/F mechanisms. PJK/F includes several different failure mechanisms. ASD patients (2009-2021) with &#x2265;5-level fusion, sagittal/coronal deformity, and two-year follow-up were included. Primary outcome was mechanism of PJK/F, defined as a PJA&#x2265;10&#xb0; and &#x2265;10&#xb0; change from preoperative. PJK/F mechanisms were: screw pullout, UIV fracture, UIV+1 fracture, screw lucency, fracture dislocation, supradjacent disc degeneration with/without listhesis, and radiographic kyphosis only. Descriptive and bivariate statistics were performed. Among 238 patients, 113 (47.5%) developed PJK/F: screw pullout (7.1%), UIV fracture (15.0%), UIV+1 fracture (8.0%), screw lucency (12.4%), fracture dislocation (11.5%), supradjacent disc degeneration with/without listhesis (31.0%), and radiographic kyphosis only (15.0%). One mechanism was seen in 91 (80.5%) patients, and 2+ mechanisms in 22 (19.5%). Median time-to-PJK/F diagnosis was 5.3 (IQR: 1.4-17.7) months: screw pullout (8.3 m, PJA=19.9&#xb0;), UIV fracture (3.8 m, 25.6&#xb0;), UIV+1 fracture (11.8 m, 28.0&#xb0;), screw lucency (12.8 m, 19.0&#xb0;), fracture dislocation (1.6 m, 27.9&#xb0;), disc degeneration (4.5 m, 25.6&#xb0;), and radiographic kyphosis only (6.1 m, 19.5&#xb0;) ( P =0.986, P <0.001). Reoperation occurred in 45 (39.8%) patients: 6/8 (75.0%) screw pullout, 6/17 (35.2%) UIV fracture, 4/9 (44.4%) UIV+1 fracture, 9/14 (64.3%) screw lucency, 6/13 (46.1%) fracture dislocation, 9/35 (25.2%) disc degeneration, and 5/17 (29.4%) radiographic kyphosis only ( P =0.068). Neurological deficits occurred in 15 (13.3%) patients: UIV fracture 3 (17.6%), UIV+1 fracture 2 (22.2%), screw lucency 5 (35.7%), fracture dislocation 2 (15.4%), disc degeneration 2 (5.7%), kyphosis only 1 (5.9%) ( P =0.093). Supradjacent disc degeneration was the most common PJK mechanism. Fracture dislocation presented earliest and with the greatest kyphosis. Reoperation was most frequent with screw pullout, lucency, and UIV+1 fracture, whereas neurological deficits were most common with lucency and UIV+1 fracture. These results demonstrate that PJK/F occurs in many different forms and ideally should be analyzed independently to further improve our treatment of this vexing complication.