Are Hounsfield Units Better at Predicting Cervical Corpectomy Subsidence than Other Measures of Bone Mineral Density?

Introduction: Subsidence after interbody fusion has been shown to correlate with loss of segmental lordosis, recurrence of preoperative symptoms, and decreased postoperative improvement in pain and disability. While decreased bone strength is known to be a risk factor for subsidence, the optimal bone quality assessment to predict cervical corpectomy subsidence remains unknown. Furthermore, prior cervical corpectomy subsidence studies were limited to indirect and imprecise measurement of subsidence on postoperative lateral x-rays. The present investigation utilized CT-based subsidence assessment to define the incidence of anterior cervical corpectomy and fusion (ACCF) subsidence and determine the relative importance of preoperative and intraoperative subsidence risk factors.

Methods: All adult patients who underwent single-level ACCF at an academic center between 2018-2021 were retrospectively identified. Corpectomy subsidence at the superior and inferior endplate of the ACCF was directly measured on the endplate-facing surface of both coronal and sagittal CT scans obtained at six months- one year postoperatively. Patients were grouped based on the maximum subsidence classified as mild-moderate < 4mm or severe ≥4mm. Preoperative bone quality assessment included: CT Hounsfield units (HUs) of the superior and inferior vertebral bodies, MRI vertebral bone quality (VBQ) of the cervical spine, and dual-XR absorptiometry (DEXA) bone mineral density, T-score, FRAX, and trabecular bone score. Univariate analysis compared patient demographics and medical comorbidities, preoperative bone assessments, and surgical factors between subsidence groups.

Results: A total of 44 patients met the inclusion/exclusion criteria (32 mild-moderate subsidence, 12 severe subsidence). The mean superior and inferior endplate subsidence was 2.6mm and 1.7mm, respectively. The maximum subsidence occurred in superior endplate in 73% of patients and the inferior endplate in 27% of patients. Patient demographics, medical comorbidities, corpectomy level, and corpectomy cage material and placement location did not differ significantly between subsidence groups. Decreased HUs of the inferior vertebral body were significantly associated with severe subsidence (244 vs 294, p=0.020). Other bone quality measures including superior vertebral HUs, cervical VBQ, and all DEXA metrics were not significantly associated with corpectomy subsidence.

Conclusion : Severe subsidence occurred in 27% of patients after ACCF. Subsidence most commonly occurred in the inferior direction with inferior vertebral body HUs as a significant risk factor.