Background
Anterior cervical discectomy and fusion (ACDF), introduced by Cloward [
1], has been accepted as the standard procedure for the treatment of myelopathy and radiculopathy in the cervical spine [
2,
3]. A tricortical iliac crest bone graft is the traditional inter-body fusion material that can show perfect bony fusion and maintain the patency of the neuroforamen. However, donor site complications were reported in fusion with an iliac bone graft, such as subcutaneous hematomas, infections, and chronic wound pain [
4].
To immobilize the unstable motion segment after discectomy, we have to ensure bony fusion and avoid donor site complications at the same time; some fusion devices have been developed for stand-alone use or use in combination with an anterior plate. As described by Bagby [
5], cage fusion technology originated from a surgery by Bagby, some veterinary surgeons and the distraction compression method and was the basic principle for stand-alone intervertebral cage fusion. Although the principle was invented to solve a cervical problem, the carbon fiber fusion cage [
6] and titanium cage [
7] were first used for lumbar inter-body fusions, and they were then applied to treat cervical spinal degenerative lesions by
Hacker [
8] and
Profeta [
9] in 2000. Currently, the titanium cage and polyetheretherketone (PEEK) cage are the two most common cages in clinical practice. The ideal cage has to have a high fusion rate and prevent complications, such as subsidence and loss of correction.
Even though a titanium cage can provide long-term stabilization, increase lordosis, and increase foramina height compared with the iliac bone graft [
10], some inferior clinical outcomes appeared in clinical practice. Loss of correction is a major complication of subsidence that may eventually affect cervical spinal function after the operation. The incidence of subsidence for the titanium cage varied as reported by Gercek who retrospectively reviewed eight patients who received ACDF with a stand-alone titanium cervical cage and found that five of the nine fused levels had radiological signs of cage subsidence [
11]. The subsidence was influenced by many factors, of which an important one is the higher elasticity modulus of the titanium cage. A modulus of elasticity close to cortical bone might contribute to advantages in stress distribution and load sharing, which can contribute to a lower subsidence rate and, thus, better clinical results, making PEEK cages more welcomed by surgeons.
Studies comparing titanium and PEEK cages for the treatment of cervical disc degenerative disease are rarely in the literature. Chou’s team retrospectively compared the results of anterior cervical fusion using titanium cages, PEEK cages and tricortical bone grafts [
12]. They noticed a better fusion rate and less subsidence in the PEEK cages group. However, the study only enrolled a small number of patients and cervical spinal function was not evaluated. In a systematic review by Kersten who compared a PEEK cage with a bone graft, titanium cage, and carbon fiber cage, no difference was found between PEEK and titanium cage [
13]. Therefore, our present study was conducted to critically review and summarize the literature to compare the results of a PEEK cage with a titanium cage for the treatment of cervical degenerative disorders to identify the better choice for the surgeons.
Discussion
Titanium and PEEK cages are commonly used cages for the anterior cervical discectomy and fusion. Our present meta-analysis was conducted to compare the outcomes of both types of cages after ACDF. The results showed a similar fusion rate, loss of correction at the surgery segment, and clinical function by the Odom criteria between the titanium and PEEK cages. Although the incidence of subsidence was higher in the titanium cage group as calculated from the data for all of the studies, it changed after excluding the data from Chen’s study. It seems that the titanium and PEEK cages perform similarly when used for ACDF. However, there are more considerations that should be taken when we are interpreting these results.
No matter whether a type of cage or iliac crest bone is adopted, the important aim was to realize that solid fusion is the basic foundation to maintain good clinical function. A tricortical iliac crest bone graft had been considered as the golden standard for good fusion [
21]. A titanium cage can also maintain fusion well because it was first used for this practice. According to the literature, the fusion rate of the titanium cage was 84 % by Yang et al. [
22], 95 by Moreland et al. [
23], and even 98 % by Schmieder et al. [
24]. It is hard to believe that the fusion rate was only 46.51 % at 12 months in the included study by Chou et al. There was no clear explanation for this in Chou’s study. PEEK is a biocompatible material with many perfect qualities for this application, such as a corrosion resistant ability [
25], the absence of cytotoxicity and mutagenicity [
26] and a close elasticity modulus to bone. Based on these material characteristics, the PEEK cage have been used in ACDF with better fusion rate as reported 94 by Hwang et al. [
27] and 100 % by both Cho et al. [
28] and Niu et al. [
19]. Most of these studies had a relatively short follow-up. In the longest 7-year observation by Chen et al., all of the patients in both groups achieved bony fusion [
18]. Therefore, we hypothesized that the cage material only has an effect on the fusion rate for a short time after ACDF.
Based on the Odom criteria, similar postoperative cervical function was observed in the included studies by Niu and Cabraja. However, Chen reported better results in the PEEK cage group than that of the titanium cage group after their long follow-up observation. The patients enrolled into these three studies were different. The patients received a single level ACDF in the study by Cabraja. Most of patients in the study by Niu received a single level ACDF and some of patients received a two level ACDF. It was significantly different that a three level ACDF was offered to all of the patients in the study by Chen et al. Limited by the few studies that compared titanium and PEEK cages, however, we cannot draw a conclusion that PEEK cages performed better in multiple level ACDF over a long time.
Subsidence was a common follow-up observation that can lead to deterioration of the long-term function. The incidence of cage subsidence for the titanium cages ranged from 13 to 62.5 % for cases in the literature [
11,
12,
18‐
20,
24,
29]. Benefitted by the better material property, the PEEK cage shows less subsidence, varying from 0 to 18 % [
18‐
20,
30‐
32], which is consistent with our result that less cage subsidence was observed in the PEEK cage group. However, some authors assumed that the cage subsidence was not affected by the differences in the modulus of material elasticity [
12,
19]. In addition to the cage material, many risk factors, including suboptimal surgical techniques and parameters of the cage [
33], may result in increased stress of the endplates and, thus, increased risk of subsidence if the bone mineral density of the vertebral body is not strong enough to bear it [
34,
35]. The patients of the PEEK cage group were older than those of the titanium cage group in the included study by Cabraja [
20]. Patients in the PEEK cage group may suffer more from osteoporosis with a higher risk of cage subsidence.
Subsidence will a cause loss of correction of the segmental angle and the Cobb angle from C2-C7. Loss of cervical lordosis is a risk factor that contributes to degeneration in the adjacent segments [
36,
37]. Kyphotic malalignment changes the dynamic kinematics of the cervical spine and accelerates the degenerative process [
38]. It is the opinion of others that subsidence less than 2 mm into the vertebral bodies until fusion is acceptable [
39]. Barsa et al. found that subsidence occurred during early follow-up and no evidence of progression appeared beyond 3 months [
29]. There were no radiographic signs of progressive degenerative changes in any of the adjacent segments during the 2-year follow-up. Wu and his colleagues reported that cage subsidence did not exert a significant impact upon the long-term clinical outcomes [
40]. The study by Chen et al. did not provide evidence of radiological and clinical progression after the long-term follow-up [
18]. Therefore, the actual effect of loss of correction on long-term function is still unclear.
Some limitations must be clarified about our present work. The results provided in this article came from two RCTs and two non-RCTs, and they were analyzed together. The power of our meta-analysis is limited by the small number of high-quality RCTs. The four included studies were different in cages, levels of ACDF, and follow-up time. Although we tried our best to summarize and analyze the data, high heterogeneity existed without a subgroup analysis. Therefore, more high-quality RCTs are needed to improve the methodology, to minimize bias and to confirm the effect of these two types of cages on long-term radiographic performance and clinical function.
Acknowledgement
We would like to acknowledge all authors of the original studies included in this meta-analysis.