Purpose:

The aim of this study is to identify risk factors associated with postoperative DJF in long constructs for ASD.

Methods:

A retrospective review was performed at a tertiary referral spine centre from 01/01/2007 to 31/12/2016. Demographic, clinical and radiographic parameters were collated for patients with DJF in the postoperative period and compared to those without DJF. Survival analyses were performed using univariate logistic regression to identify variables with a p value < 0.05 for inclusion in multivariate analysis. Spearman's correlations were performed where applicable.

Results:

One hundred two patients were identified. 41 (40.2%) suffered DJF in the postoperative period, with rod fracture being the most common sign of DJF (13/65; 20.0%). Mean time to failure was 32.4 months. On univariate analysis, pedicle subtraction osteotomy (p = 0.03), transforaminal lumbar interbody fusion (p < 0.001), pre-op LL (p < 0.01), pre-op SVA (p < 0.01), pre-op SS (p = 0.02), postop LL (p = 0.03), postop SVA (p = 0.01), postop PI/LL (p < 0.001), LL correction (p < 0.001), SVA correction (p < 0.001), PT correction (p = 0.03), PI/LL correction (p < 0.001), SS correction (p = 0.03) all proved significant. On multivariate analysis, pedicle subtraction osteotomy (OR 27.3; p = 0.03), postop SVA (p < 0.01) and LL correction (p = 0.02) remained statistically significant as independent risk factors for DJF.

Conclusion:

Recently, DJF has received recognition as its own entity due to a notable postoperative incidence. Few studies to date have evaluated risk factors for DJF. The results of our study highlight that pedicle subtraction osteotomy, poor correction of lumbar lordosis, and sagittal vertical axis are significantly associated with postoperative occurrence of DJF.