Predictors of subsidence after lateral lumbar interbody fusion.

Abstract

Lateral lumbar interbody fusion (LLIF) facilitates the restoration of disc height and the indirect decompression of neural elements. However, these benefits are lost when the graft subsides into the adjacent endplates. The factors leading to subsidence after LLIF are poorly understood. This article presents a case series of patients who underwent LLIF and reports factors correlating with subsidence. A retrospective review of a consecutive, prospectively collected, single-institution database of patients who underwent LLIF over a 29-month period was performed. The degree of subsidence was measured on the basis of postoperative imaging. The timing of postoperative subsidence was determined, and intraoperative fluoroscopic images were reviewed to determine whether subsidence occurred as a result of endplate violation. The association of subsidence with age, sex, cage size and type, bone density, and posterior instrumentation was investigated. One hundred thirty-one patients underwent LLIF at a total of 204 levels. Subsidence was observed at 23 (11.3%) operated levels. True subsidence, attributable to postoperative cage settling, occurred for 12 (5.9%) of the levels; for the remaining 11 (5.4%) levels, subsidence was associated with intraoperative endplate violation noted on fluoroscopy during cage placement. All subsidence occurred within 12 weeks of surgery. Univariate analysis showed that the prevalence of true subsidence was significantly lower among patients with titanium implants (0 of 55; 0%) than among patients with polyetheretherketone cages (12 of 149; 8.1%) (p = 0.04). In addition, the mean ratio of graft area to inferior endplate area was significantly lower among the subsidence levels (0.34) than among the nonsubsidence levels (0.42) (p < 0.01). Finally, subsidence among levels with posterior fixation (4.4% [6/135]) was not significantly different than among those without posterior fixation (8.7% [6/69]) (p = 0.23). Multivariate analysis results showed that the ratio of cage to inferior endplate area was the only significant predictor of subsidence in this study (p < 0.01); increasing ratios were associated with a decreased likelihood of subsidence. Overall, the prevalence of subsidence after LLIF was low in this clinical series. Titanium cages were associated with a lower prevalence of observed subsidence on univariate analysis; however, multivariate analysis demonstrated that this effect may be attributable to the increased surface area of these cages relative to the inferior endplate area.