Risk factors for radiographic progression of proximal junctional fracture in patients undergoing surgical treatment for adult spinal deformity

OBJECTIVE Proximal junctional fracture (PJFx) at the uppermost instrumented vertebra (UIV) or UIV+1 is the most common mechanism of proximal junctional failure (PJF). Few studies have assessed radiographic progression after PJFx development. Therefore, this study sought to identify the risk factors for radiographic progression of PJFx in the surgical treatment for adult spinal deformity. METHODS In this retrospective study, among 317 patients aged > 60 years who underwent ≥ 5-level fusion from the sacrum, 76 with PJFx development were included. On the basis of the change in the proximal junctional angle (PJA), 2 groups were created: progression group (group P) (change ≥ 10°) and nonprogression group (group NP) (change < 10°). Patient, surgical, and radiographic variables were compared between the groups with univariate and multivariate analyses to demonstrate the risk factors for PJFx progression. The receiver operating characteristic (ROC) curve was used to calculate cutoff values. Clinical outcomes, such as visual analog scale (VAS) scores for back and leg pain, Oswestry Disability Index (ODI) score, the Scoliosis Research Society (SRS)–22 score, and the revision rate were compared between the 2 groups. RESULTS The mean age at index surgery was 71.1 years, and 67 women were enrolled in the study (88.2%). There were 45 patients in group P and 31 in group NP. The mean increase in PJA was 15.6° (from 23.2° to 38.8°) in group P and 3.7° (from 17.2° to 20.9°) in group NP. Clinical outcomes were significantly better in group NP than group P, including VAS-back score, ODI score, and SRS-22 scores for all items. The revision rate was significantly greater in group P than in group NP (53.3% vs 25.8%, p = 0.001). Multivariate analysis revealed that overcorrection relative to the age-adjusted ideal pelvic incidence (PI)–lumbar lordosis (LL) target at index surgery (OR 4.484, p = 0.030), PJA at the time of PJFx identification (OR 1.097, p = 0.009), and fracture at UIV (vs UIV+1) (OR 3.410, p = 0.027) were significant risk factors for PJFx progression. The cutoff value of PJA for PJFx progression was calculated as 21° by using the ROC curve. CONCLUSIONS The risk factors for further progression of PJFx were overcorrection relative to the age-adjusted PI-LL target at index surgery, PJA > 21° at initial presentation, and fracture at the UIV level. Close monitoring is warranted for such patients in order to not miss timely revision surgery.