A retrospective cohort study with chart review OBJECTIVE.: To compare the reoperation rates for symptomatic nonunions (operative nonunion rates) between posterolateral fusions with pedicle screws (PLFs) and posterior interbody fusion with pedicle screws (PLIFs).
Summary of background data:
Although radiographic nonunions in PLFs and PLIFs are well documented in the literature, there is no consensus on which technique has lower nonunions. Since some radiographic nonunions may be asymptomatic, a more clinically useful measure is operative nonunions, of which there is minimal research.
A retrospective cohort study, using data from the Kaiser Permanente Spine Registry, identified adult patients (≥18 years old) who had elective single and multilevel PLFs and PLIFs. Descriptive statistics and two-year incidence rates for operative nonunions were calculated by fusion-level (1-3), fusion type (PLF vs. PLIF), and levels fused (L3 to S1). Time-dependent multivariable Cox-Proportional Hazards regression was used to evaluate nonunion reoperation rates with adjustment for covariates.
The cohort consisted of 3,065 patients with PLFs (71.6%) and PLIFs (28.4%). Average age was 65.0 ± 11.7, average follow-up time was 4.8 ± 3.1 years, and average time to operative nonunion was 1.6 (±1.3) years. Single and multilevel incidence rates for nonunions after PLF vs PLIF were similar except for 3-level fusions (2.9% (95% CI = 1.0-6.7) vs 7.1% (95% CI = 0.2-33.9)). In adjusted models, there was no difference in risk of operative nonunions in PLIF compared to PLF (HR: 0.8, 95% CI = 0.4-1.6), however patients with L5-S1 constructs with PLFs had 2.8 times the risk of operative nonunion compared to PLIFs (PLF: HR = 2.8, 95% CI = 1.3-6.2; PLIF: HR = 1.5, 95% CI = 0.4-5.1).
In a large cohort of patients with greater than 4 years of follow-up, we found no difference in operative nonunions between PLF and PLIF except for constructs that included L5-S1 in which the risk of nonunion was limited to PLF patients.
Level of evidence: