Role of preoperative Hounsfield units in predicting subsidence after anterior cervical discectomy and fusion in the United States: A retrospective analysis including osteopenia diagnosis

Study Design: A retrospective cohort study. Purpose: This study aimed to investigate the relationship among osteopenia, bone density, and subsidence following anterior cervical discectomy and fusion (ACDF). Overview of Literature: Subsidence following ACDF procedures can lead to worse clinical outcomes. Although studies have linked osteopenia to cage subsidence, no consensus has been established on the relationship between bone density and subsidence. Methods: Patients undergoing ACDF between 2016 and 2021 were included and assigned to the osteopenia cohort based on chart review and dual-energy X-ray absorptiometry scan results. Bone density at each vertebral level of the cervical fusion was assessed by measuring Hounsfield units from preoperative computed tomography (CT) scans, and disk height changes were assessed using immediate postoperative (<6 weeks) and final follow-up (>5 months) radiographs. Subsidence was quantified by the difference between the long-term and immediate postoperative anterior and posterior disk heights. A t-test was performed to evaluate the effect of prior osteopenia diagnosis on segmental subsidence. Multivariable analysis, accounting for age, sex, smoking status, and cage type, further investigated the relationship between Hounsfield units and subsidence. Results: Among the 131 patients (244 levels fused), no significant associations were found between osteopenia diagnosis and anterior (p=0.926) or posterior (p=0.918) subsidence. Preoperative CT measurements for 28 patients (54 fused levels) revealed no correlations between subsidence and Hounsfield units at the vertebral levels above and below the fusion. Of the 54 levels with preoperative CT scans, 22 patients (41%) were diagnosed with osteopenia. Osteopenia did not correlate with Hounsfield units using the endplate (superior, p=0.735; inferior, p=0.693), full vertebrae (superior, p=0.686; inferior, p=0.735), or elliptical (superior, p=0.501; inferior, p=0.465) methods. Conclusions: The results did not reveal the relationship between either the prior diagnosis of osteopenia or Hounsfield units and subsidence. These results highlight the multifactorial nature of postoperative subsidence, and osteopenia or Hounsfield units cannot be used alone to determine the subsidence risk.