The impact of cage position, radiographic parameters, and hounsfield units on subsidence rate after one to three level anterior cervical discectomy and fusion.

Abstract

Study design: The study design was a retrospective cohort.

Objective: The objective of the study was to validate the relationship between Hounsfield units (HU) and subsidence, including multilevel anterior cervical discectomy and fusion (ACDF). Cage/graft subsidence commonly occurs after ACDF. Prior work on 1-level ACDF found increased subsidence rates in patients with lower HUs.

Materials and methods: Adults who underwent 1-3 level ACDF at a tertiary center and had preoperative computed tomography scans were included (2018-2022). HUs were assessed ~5 mm caudal to the superior endplate. Six-month postoperative radiographs were evaluated for cage/graft positioning, screw loosening, and subsidence. Receiver operating characteristic curves and area under the curve (AUC) assessed the predictive value of segmental/minimum/maximum HU for screw loosening and/or subsidence.

Results: Forty-two patients (82 levels) were included - demographics were similar among patients with versus without subsidence at any level. Average HU, segmental HU, segmental HU above and below 343.7 HU, minimum HU, and maximum HU were similar between patients with versus without subsidence at any level. Among the HU measurements, the maximum AUC was 0.554 (95% confidence interval 0.421-0.687) for screw loosening as predicted by minimum HU with a cutoff of 313 HU. Subsidence was more associated with middle positioning compared to anterior (28.0% vs. 10.5%, P = 0.046).

Conclusions: Contrary to prior literature, this study found no association and minimal predictive ability of segmental, minimum, or maximum HU values and subsidence or screw loosening after 1-3 level ACDF. Middle positioning was associated with cage/graft subsidence. This suggests that central positioning of the cage/graft is a risk factor for subsidence, potentially due to softer cancellous bone centrally.