Nonunion at the site of attempted fusion has direct implications in regard to the development of implant failure and other subsequent complications. Following an ACDF, if solid fusion fails to occur, the hardware that is placed will eventually fracture or loosen. This is more frequent with longer constructs such as multilevel corpectomies as larger stresses are imparted to the screw/bone interface of the construct (15,16). Wang et al. (12) reported on 80 patients that were treated with a singlelevel ACDF with 6-year follow-up. Forty-four patients had cervical plates, whereas thirty-six had fusions without plates. The pseudarthrosis rates were 4.5% for patients with plating and 8.3% without plating, though this did not reach statistical significance. They did note several significant differences between the groups. The amount of graft collapse for patients with plating was 0.75 mm compared with 1.5 mm for those without a plate, and the amount of kyphotic deformity of the fused segment was 1.2 degrees with plating compared with 1.9 degrees for patients without plating. They concluded that the addition of plate fixation for single-level ACDF is safe and does not increase complication rates, although pseudarthrosis rates were not significantly different when a cervical plate was used.

Wang et al. (17) also evaluated the effectiveness of anterior cervical plating in two-level ACDF procedures. They compared 60 patients, 32 with cervical plates, and 28 treated without plates, followed for an average of 2.7 years. Significant findings included pseudarthrosis rates of 0% for patients with plating versus 25% for those with no plating, as well as increased graft collapse and increased kyphosis in patients experiencing nonunions versus plated patients. They concluded that the addition of plate fixation for two-level ACDF is again a safe procedure with no significant increase in complication rates. Pseudarthrosis rates were significantly lower, and there was significantly less disk space collapse and kyphotic deformity with the plated fusions than with the nonplated fusions.

Finally, Wang et al. (18) reported on three-level ACDF procedures with and without plating. Fifty-nine patients underwent a three-level ACDF, 40 patients with anterior cervical plates and 19 with no plate, followed for an average of 3.2 years. Fourteen patients developed a pseudarthrosis, 18% of patients with plating and 37% of patients with no plate. This difference did not reach statistical significance. Addition of plate fixation for three-level ACDF was safe and did not result in higher complication rates. Although the use of cervical plates may decrease the pseudarthrosis rate, a plated three or more level ACDF is still associated with a high nonunion rate, and other strategies to increase fusion rates should be considered (Fig. 102.1). Similarly, Bolesta et al. (19) reported a 53% pseudoarthrosis rate for three- and four-level ACDFs despite plate fixation. While their patient numbers were low, they suggested that additional dorsal fusion should be considered in such a situation.

More recently, Fountas et al. (20) retrospectively reviewed 1,015 patients who underwent first-time one-, two-, or three-level ACDF and reported upon their complications. Fusion rates for the one-, two-, and three-level ACDF groups were 95.6%, 93.9%, and 90.5%, respectively. This study had several limitations. It was nonrandomized and multiple different techniques were employed, with some patients receiving autograft while others allograft. A small percentage of patients were treated without plates; 6% in the single-level group, 2% in the two-level group, and 1% in the three-level group. They could not find any statistically significant differences in fusion rates comparing the plated versus nonplated groups, though with such a small number of patients in the nonplated group, the study lacked statistical power to show such differences.

Lastly, there has been concern that statically locked plates themselves may lead to pseudoarthrosis in some patients. This is theoretically possible secondary to the effect of bridging an unloaded interbody graft after normal subsidence during healing (21). This concern motivated the development of dynamic plates, as outlined previously. Nunley et al. (22) randomized 66 patients undergoing ACDF to either a dynamic or static anterior cervical plate. Of the single-level fusions, there was no statistically significant difference in fusion rate between plate types. In multilevel constructs, patients with dynamic plates had better clinical outcomes. It should be noted that the style of dynamic plate used was a translational plate.