Diagnosis of spine pseudoarthrosis based on the biomechanical properties of bone

BACKGROUND CONTEXT

Failure to fuse following anterior cervical discectomy and fusion (ACDF) may result in symptomatic pseudoarthrosis. Traditional diagnosis involves computerized tomography to detect bridging bone and/or flexion-extension radiographs to assess whether segmental motion is above specific thresholds; however, there are currently no well-validated diagnostic tests. We propose a biomechanically rational approach to achieve a reliable diagnostic test for pseudoarthrosis.

PURPOSE

Develop and test a biomechanically based approach to the diagnosis of pseudoarthrosis.

STUDY DESIGN

Literature review, development of theory, reanalysis of a previously published study with surgical exploration as the gold-standard, and retrospective analysis of pooled studies to understand time to fusion.

METHODS

Fully automated methods were used to measure disc space strains (change in disc space height divided by initial height). Measurement error combined with the reported failure strain of trabecular bone led to a proposed strain threshold for diagnosis of pseudoarthrosis following ACDF. We reanalyzed previously reported flexion-extension radiographs for asymptomatic volunteers to assess whether flexion-extension radiographs, in the absence of fusion surgery, can be expected to provide sufficient stress on motion segments to allow for reliable strain-based fusion assessment. The sensitivity and specificity of strain- and rotation-based pseudoarthrosis diagnosis were assessed by reanalysis of previously reported post-ACDF flexion-extension radiographs, where intraoperative fusion assessments were also available. Finally, we assessed changes in strain over time using 9,869 flexion-extension radiographs obtained 6 weeks to 84 months post-ACDF surgery from 1,369 patients.

RESULTS

The estimated error in automated measurement of disc space strain from radiographs was approximately 3%, and the reported failure strain of bridging bone was less than 2.5%. On that basis, we propose a 5% strain threshold for pseudoarthrosis diagnosis. Reanalysis of a study in which intraoperative fusion assessments were available revealed 67% sensitivity and 82% specificity for strain-based diagnosis of pseudoarthrosis, which was comparable to rotation-based diagnosis. Analysis of post-ACDF flexion-extension radiographs revealed rapid strain reduction for up to 24 months, followed by a slower decrease for up to 84 months. When rotation is less than 2°, the strain-based diagnosis differed from the rotation-based diagnosis in approximately 14% of the cases.

CONCLUSIONS

We propose steps for standardizing diagnosis of pseudoarthrosis based on the failure strain of bone, measurement error, and retrospective data. These steps include obtaining high-quality flexion-extension studies, the application of proposed diagnostic thresholds, and the use of image stabilization for conclusive diagnosis, when motion is near thresholds. The necessity for an accurate diagnosis with minimal radiation exposure underscores the need for further optimization and standardization in diagnosing pseudoarthrosis following ACDF surgery.

CLINICAL SIGNIFICANCE

In a symptomatic postspine fusion patient, it is important to diagnose or rule-out pseudoarthrosis. There are currently no well-validated diagnostic tests for this condition. Incorporating strain-based intervertebral motion analysis into the diagnosis could lead to a standardized and validated test for detecting spine pseudoarthrosis.