Pseudarthrosis after an anterior cervical discectomy and fusion is a well-known complication. Variability exists regarding the rate of pseudarthrosis, with authors reporting rates from 0% to more than 50%. However, not all pseudoarthroses are symptomatic or require a subsequent surgical repair.In the authors' experience, some patients have a stable, stiffer fibrous union, whereas others have a more mobile nonunion. In patients with recurrent or persistent neck pain, however, pseudarthrosis should be considered as a likely etiology. An accurate diagnosis is therefore important in the clinical management of these patients.
The accuracy of a pseudarthrosis diagnosis using radiographic tests is poorly documented in the literature. Numerous studies comment on the occurrence, but fail to elucidate what measures were used to arrive at the diagnosis of a pseudarthrosis. In 1997, Phillips et al listed several radiographic criteria for pseudarthrosis. The only objective measure among these criteria was motion exceeding 2 mm between the spinous processes at the surgically managed segment on flexion-extension lateral radiographs. Other methods of diagnosis included radiographic analysis for the presence of bridging trabeculae, computerized tomographic scans, and the like. To the knowledge of the current authors, however, neither this method nor others have not been evaluated critically as diagnostic techniques.
The current authors used a variety of statistical means to demonstrate why measurement of the tips of the spinous processes is a method that may be used for the diagnosis of pseudarthrosis. Three physicians completed the measurements, and the reliability was higher for the spinous process method. In addition, the ability to detect a pseudarthrosis with more than a 2-mm difference between the tips was more sensitive. This method had high sensitivity, detecting a pseudarthrosis in 10 of the 11 patients who had a known pseudarthrosis. When the Cobb angle method was used, there was less sensitivity, specificity, and correlation between the measurements obtained and the detection of a pseudarthrosis. The authors chose also to calculate the ROC curve. For two screening methods to assess the same outcome (pseudarthrosis), the result with the higher area under the ROC curve is considered the better overall diagnostic test.The area under the curve was much higher using the tips of the spinous process method. In fact, the area under the curve for the Cobb angle method was only slightly better than chance alone (0.66 vs 0.50).
Six patients in this study had anterior cervical plating at the time of the surgery. Two of these patients had a pseudarthrosis. The plate did not interfere with detection of a pseudarthrosis by measuring the tips of the spinous processes. In the authors' experience, over time, a windshield wiper effect is detectable around the screws on follow-up radiographs. Covering the operative disc and vertebral body levels with duct tape rendered this effect invisible during completion of the measurements. The current data regarding the radiographic detection of a nonunion in patients who had a plate were sparce, so no definitive conclusion or possible difficulties regarding this subset of patients could be identified.
Measuring the distances between the tips of the spinous processes of the involved segments provided an objective method for consistently detecting pseudarthrosis radiographically. This method proved to be more accurate than the Cobb method, and there was less measurement variability among multiple observers than with the Cobb measurements. The patients with a known pseudarthrosis consistently had a change in measurements between the tips of the spinous processes, with differences greater than 2 mm using the lateral flexion and extension views. Thus, spinous process measurement on flexion and extension radiographs is an objective, accurate method for the evaluation of anterior cervical fusion, which may help in the clinical management of this patient population.