Variations in screening, reporting patterns, and incidence of postoperative subsidence following spine surgeries using interbody cages

Introduction: Intervertebral cages are placed to sustain foraminal height. Reduction in this vertical space prior to fusion completion/subsidence is the most common complication. Studies report incidence in varying ways including mm measurements, yes/no criteria, and grading scale. We aim to assess patterns of subsidence assessment, subsidence definitions, and differences in incidence of subsidence based on modality of radiographic imaging.

Methods: Literature searches were performed and studies that met inclusion criteria were greater than 30 patients >18 years, underwent interbody fusion with cage insertion, and assessed for subsidence. Variables assessed included sample size, region of the spine, type of surgery, surgical approach, indications, shape of cage, type of cage, measurement mode of linear subsidence (mm scale or yes/no), method of radiology used to assess for subsidence, height of disc space, number of patients with linear subsidence, mean follow up (months), and whether subsidence was described with cage orientation.

Results: A total of 86 studies consisting of 5839 patients were included for meta-analysis: 35 anterior cervical, 2 posterior cervical, 13 posterior thoracic, 2 anterior thoracic, 7 anterior lumbar, 12 lateral lumbar and 6 posterior lumbar, and 9 mixed multilevel approaches. Of the studies included, 28 used X-ray for diagnosis of postoperative linear subsidence, whereas 58 used CT or MRI for this purpose. The mean follow-up time for subsidence assessment was 18.8 months after surgery. 48.1% utilized a yes/no criteria for subsidence grading, relative to the 51.9% that used a millimeter grading scale. The most commonly employed definition of subsidence in the yes/no criteria was cage sinking more than 2 mm into the adjacent vertebral endplate, utilized in 24.4% of all studies. Of the studies that used X-ray alone, the overall incidence subsidence was 8.192% as opposed to studies using CT/MRI, which report significantly higher subsidence rates of 34.373% (p=0.0152).

Conclusion : Screening practices and definitions of postoperative subsidence are heterogeneous. Assessments using X-rays alone severely undercount the incidence of subsidence as opposed to CT/MRI, highlighting the need for reporting specificity for linear subsidence.