Comparison of allograft to autograft in multilevelanterior cervical discectomy and fusion with rigid plate fixation

Abstract

Background context

A relatively high pseudarthrosis rate is associated with multilevel anterior cervical discectomy and fusion (ACDF). Anterior plate fixation increases fusion rate in multilevel ACDF. A debate still exists between the effectiveness of allograft versus autograft in plated multilevel ACDF.

Purpose

To determine the efficacy of allograft versus autograft in fusion rate and clinical outcome in patients undergoing two- and three-level ACDFs with rigid anterior plate fixation.

Study design

A retrospective radiographic and clinical review to assess fusion, risk factors and clinical outcome of 80 consecutive patients who underwent ACDF with rigid anterior plate fixation involving two and three levels with either allograft or autograft.

Patient sample

There were 45 patients (56%) who had autogenous iliac crest tricortical grafts and 35 patients (44%) who received tricortical allograft with an average age of 49 years who were treated by multilevel ACDF with rigid anterior plate fixation at a single institution. Thirty-three Peak polyaxial (Depuy-Acromed, Rayham, MA), 26 Orion (Sofamor-Danek, Memphis, TN), 16 Atlantis (Sofamor-Danek, Memphis, TN) and 5 Synthes (Paoli, PA) anterior cervical plating systems were used. All patients underwent ACDF (61 two-level, 19 three-level) by a Smith Robinson technique. All patients had burring of the end plates, 2-mm distraction of the motion segment and graft countersunk 2 mm from the anterior vertebral border. Anterior cervical plate with unicortical screw purchase was used in all cases. Segmental screw fixation was performed in 46 patients. Soft collars were worn postoperatively for 3 to 4 weeks.

Outcome measures

Follow-up lateral neutral, flexion and extension radiographs were used to assess fusion. The radiographs were reviewed by an independent blinded observer in assessing fusion grades between autograft versus allograft. Clinical outcomes were rated excellent, good, fair and poor based on Odom's criteria.

Methods

Fusion rate and postoperative clinical outcome were assessed in 80 patients who underwent two- or three-level ACDF with rigid anterior plate fixation. Additional risk factors were also analyzed.

Results

Radiographic fusion was assessed in all patients (mean, 16 months). Seventy-eight patients (97.5%) achieved solid arthrodesis. Pseudarthrosis occurred in two patients who had allograft for two-level and three-level fusions. Nonsegmental screws were used in the two-level nonunion case. Postoperative dysphagia developed in one two-level nonunion patient, and revision surgery was performed in the other nonunion three-level patient. Twenty-three patients were smokers, and 26 patients had work-related injuries. Clinical outcome (mean, 20 months) was excellent in 23, good in 48 and fair in 9 patients. No statistical significance was noted between demographics, history of tobacco use, graft-type, end plate preparation technique, intermediate segmental screws, plate-type, clinical outcome of fused and nonfused patients and presence of work-related injuries (p > .05).

Conclusions

A high fusion rate of 97.5% was obtained for multilevel ACDF with rigid plating with either autograft or allograft. In this study, nonunion occurred in patients with allograft but this difference was not statistically significant. Fusion was obtained in 97.8% of patients with segmental screw fixation and 97.1% with nonsegmental screw fixation. Nonsegmental screw fixation may contribute to less than adequate stability and contribute to a higher rate of nonunion, but such effects could not be discerned from this study. Excellent and good clinical outcome was noted in 88.8% of the patients. Proper patient selection and meticulous operative technique is essential to obtain high fusion rates and optimal clinical outcome, which is more important than graft type.

Introduction

Since initially reported in 1955 by Robinson and Smith [1], anterior cervical discectomy and fusion (ACDF) has become an accepted procedure for decompression of neural elements and spinal stabilization in the treatment of degenerative disease, trauma and tumors of the spine. Successful arthrodesis is instrumental in obtaining optimal outcome in ACDF. Selection of the appropriate graft substrate to optimize fusion rate and healing is essential. Various systemic factors, operative technique, plate type and work-related injuries are further components that can affect osseous fusion and clinical outcome.

Single-level ACDF is a highly successful procedure yielding high reported fusion rates, ranging from 83% to 97% and 82% to 94% for autograft [2], [3], [4], [5], [6] and allograft [2], [7], [8], respectively. However, in multilevel ACDF, as the number of grafts increases, the cervical spine is predisposed to decreased fusion rates as contact stress increases between the graft–body interface and further contributes to unacceptable micromotion [9]. Further, in multilevel ACDF, studies suggest that allograft is inferior to autograft in achieving arthrodesis. In 1991, in patients undergoing a two-level ACDF, Zdeblick and Ducker [10] noted nonunion in 63% of those with allograft and in 17% of those with autograft. Comparing the use of allograft plus demineralized bone matrix with autograft in patients with a two-level ACDF, An et al. [11] reported nonunion in 37.5% and 23.5% for allograft and autograft, respectively. Higher nonunion rates at 37% to 70% have been reported for three-level ACDFs [4], [11], [12], [13], [14].

Since the advent of an anterior cervical plating device [15], fusion rates for multilevel instrumented ACDF have increased [14], [16], [17], [18], [19]. Throughout the years, anterior cervical plating devices have evolved to increase internal stability, maintain alignment, facilitate fusion, decrease the incidence of graft extrusion and avoid postoperative halo bracing [20], [21], [22]. As previously discussed, single-level ACDF results in high fusion rates, but the addition of cervical plating may result in similar fusion rates and possibly worse clinical outcomes [23], [24]. Nonetheless, the use of anterior cervical plate fixation in multilevel ACDF is warranted to decrease high pseudarthrosis rates that accompany no internal fixation. A further consideration is that instrumentation failure remains a factor that could contribute to nonunion. Various studies have indicated the failure rate of anterior multilevel instrumentation is as high as 23% [25], [26].

The effectiveness of graft type in two-level ACDF with plate fixation is open to debate, largely because of inadequate allograft sample sizes [16], [18], [19], [24]. The graft efficacy for successful fusion in three-level ACDF with plate fixation is also questionable; previous literature has reported pseudarthrosis rates as high as 53% [27]. Although recent studies have demonstrated increased fusion rates with plate fixation for three-level ACDF, reports are few, and comparison of allograft to autograft in such patients has rarely been addressed [14], [17], [27], [28].

The selection of the appropriate graft substrate is imperative to achieve successful bone fusion and an optimal clinical outcome. However, there are multiple disadvantages associated with each graft type. The harvest site for autogenous iliac crest bone grafts is associated with an amplified risk of morbidity, operative time, hospital stay and postoperative recovery [29], [30], [31], [32], [33], [34]. Furthermore, patient dissatisfaction with the cosmetics of the iliac crest incision also contributes to the shortcomings of such a harvesting method [34]. Disadvantages are also associated with allograft and include histocompatibility differences that may affect proper healing, increased risk of infection, lack of availability and back ordering, size variations and possible structural weakness [10], [11], [35], [36], [37], [38], [39]. The use of allograft does avoid harvest site morbidity, is available in desired quantity, can be configured, decreases hospitalization time and reduces costs.

Although it is known that anterior plate fixation increases the fusion rate in multilevel ACDF patients, a debate still exists between the efficacy of allograft versus autograft and its relationship to clinical outcome in multilevel ACDF procedures. The purpose of this study is to address fusion rates, risk factors and clinical outcome in patients undergoing ACDF with rigid anterior plate fixation involving two or three levels with either allograft or autograft.