Incidence and prevalence of surgery at segments adjacent to a previous posterior lumbar arthrodesis

Adjacent segment disease (ASD) after lumbar spinal fusion has been an important reason behind the development of nonfusion stabilization technology. However, the incidence, prevalence, and factors contributing to adjacent segment degeneration in the lumbar spine remain unclear. A range of prevalence rates for ASD have been reported in the lumbar spinal literature, but the annual incidence has not been widely studied in this region. Conflicting reports exist regarding risk factors, especially fusion length.

To determine the annual incidence and prevalence of further surgery for adjacent segment disease (SxASD) after posterior lumbar arthrodesis and examine possible risk factors.

Retrospective cohort study.

Nine hundred twelve patients who underwent 1,000 consecutive posterior lumbar interbody fusion procedures, with mean follow-up duration of 63 months (range, 5 months-16 years).

Further surgery for ASD or surgery-free survival.

A postal and telephone survey. Follow-up rate: 91% of patients. The annual incidence and prevalence of ASD requiring further surgery were determined using Kaplan-Meier survivorship analysis. Cox proportional-hazards (Cox) regression was used for multivariate analysis of possible risk factors. Significance was set at p<.05.

Further surgery for ASD occurred following 130 of 1,000 or 13% of procedures at a mean time of 43 months (range, 2.3-162 months). The mean annual incidence of SxASD over the first 10 years, in all patients, was 2.5% (95% confidence interval [95% CI], 1.9-3.1) with prevalences of 13.6% and 22.2% at 5 and 10 years, respectively. Cox regression modeling found that the number of levels fused (p≤.0003), age of the patient, fusing to L5, and performing an additional laminectomy adjacent to a fusion all independently affect the risk of SxASD. The mean annual incidence figures in the first 10 years after a lumbar fusion were 1.7% (95% CI, 1.3-2.2) after fusion at single levels, 3.6% (2.1-5.2) after two levels, and 5.0% (3.3-6.7) after three and four levels. The 5- and 10-year prevalences were 9% and 16%, 17% and 31%, and 29% and 40% after single-, two-, and three-/four-level fusions, respectively. The risk of SxASD in patients younger than 45 years was one-quarter (95% CI, 10-64) the risk of patients older than 60 years (p=.003). A laminectomy adjacent to a fusion increases the relative risk by 2.4 times (95% CI, 1.1-5.2; p=.03). Stopping a fusion at L5 is associated with a 1.7-fold increased risk (95% CI, 1.2-2.4; p=.007) of SxASD compared with a fusion to S1, for fusions of the same length.

The overall annual incidence and predicted 10-year prevalence of further surgery for ASD after lumbar arthrodesis were 2.5% and 22.2%, respectively. These rates varied widely depending on the identified risk factors. Although young patients who underwent single-level fusions were at low risk, patients who underwent fusion of three or four levels had a threefold increased risk of further surgery, compared with single-level fusions (p<.0001), and a predicted 10-year prevalence of 40%.