Does lumbar spine fusion predispose patients to future total hip replacement?

Abstract

OBJECTIVE Hip-spine syndrome has been well studied since it was first described by Offierski and MacNab in 1983. Today, strong evidence links symptoms of hip and spine pathology to postsurgical outcomes. Recent studies have reported increased rates of hip dislocation in patients previously treated with total hip arthroplasty (THA) who had undergone lumbar fusion procedures. However, the effect of this link on native hip-joint degeneration remains an area of ongoing research. The purpose of this study was to characterize the relationship between use of lumbar fusion procedures and acceleration of hip pathology by analyzing the rate of future THA in patients with preexisting hip osteoarthritis. METHODS This population-level, retrospective cohort study was conducted by using the PearlDiver research program. The initial patient cohort was defined by the presence of diagnosis codes for hip osteoarthritis. Patients were categorized according to use of lumbar fusion after diagnosis of hip pathology. Survival curves with respect to THA were generated by comparison of the no lumbar fusion cohort with the lumbar fusion cohort. To assess the impact of fusion construct length, the lumbar fusion cohort was then stratified according to the number of levels treated (1–2, 3–7, or ≥ 8 levels). Hazard ratios (HRs) were then calculated for the risk factors of number of levels treated, patient age, and sex. RESULTS A total of 2,275,683 patients matched the authors’ inclusion criteria. Log-rank analysis showed no significant difference in the rates of THA over time between the no lumbar fusion cohort (2,239,946 patients) and lumbar fusion cohort (35,737 patients; p = 0.40). When patients were stratified according to number of levels treated, again no differences in the incidence rates of THA over the study period were determined (p = 0.30). Patients aged 70–74 years (HR 0.871, p < 0.001), 75–79 years (HR 0.733, p < 0.001), 80–84 years (HR 0.557, p < 0.001), and ≥ 85 years (HR = 0.275, p < 0.001) were less likely to undergo THA relative to the reference group (patients aged 65–69 years). CONCLUSIONS Although lumbar fusion was initially hypothesized to have a significant effect on rate of THA, lumbar fusion was not associated with increased need for future THA in patients with preexisting hip osteoarthritis. Additionally, there was no relationship between fusion construct length and rate of THA. Although lumbar fusion reportedly increases the risk of hip dislocation in patients with prior THA, these data suggest that lumbar fusion may not clinically accelerate native hip degeneration.