Proximal junctional kyphosis in adult scoliosis: comparison of four radiological predictor models.

Objective: The objective of this retrospective study is to identify the best immediate postoperative radiological predictors for the occurrence of proximal junctional kyphosis (PJK). Four proposed methods will be explored.

Methods: A homogeneous database of adult scoliosis from multiple centers was used. Patients with whole spine X-rays at the required follow-up (FU) periods were included. Spinal and pelvic parameters were measured and calculated to compare four predictive methods: Method 1: assessment of the global sagittal alignment (GSA); Method 2: restoration of the theoretical values of lumbar lordosis (LL) and thoracic kyphosis (TK) according to pelvic incidence (PI); Method 3: evaluation of TK + LL, and Method 4: restoration of the apex of sagittal LL to its theoretical values according to various spine shapes in Roussouly Classification. PJK occurrence was assessed at the last FU radiograph.

Results: 250 patients were included; mean age was 56.67 years and mean FU was 2.5 years. PJK occurred in 25.6% of cases. PJK occurred in 19.9% in patients with a GSA <45° and in 29.9% where GSA >45° (p = 0.04, OR = 1.71). Restoring the sagittal apex of the LL to its theoretical values according to PI deceased PJK to 13.5% compared to 38.9% in the other cases (p = 0.01, OR = 4.6). The two other described methods (2 and 3) were not significant predictors.

Discussion: The comparison between the four predictive methods showed that a GSA >45° and restoration of sagittal apex of lordosis according to PI, were the most predictive methods for PJK in ASD. The latter had a higher predictive value. Our findings could prove useful in effective preoperative planning in ASD surgery to reduce PJK rates.

Level Of Evidence: Level IV.