Introduction

The aim of anterior cervical discectomy and fusion (ACDF) in degenerative cervical spine disease is to improve patient symptoms, improve spine stability and restore lordosis, while avoiding complications.

Since the first description of ACDF by Smith and Robinson [1], there have been many advancements in surgical technique, as well as prosthesis options. More recently, there has been a rise in the use of the locking stand-alone cages (LSC), which does not require an anterior plate. LSCs developed with the aim of minimising soft tissue disruption anterior to the vertebrae and reducing the profile of the construct by avoiding an anterior plate [2, 3].

Figures 1 and 2 show examples of the APC and LSC, respectively.

Fig. 1
figure 1

X-ray of anterior plate construct—anteroposterior (a) and lateral (b) views

Full size image
Fig. 2
figure 2

X-ray of locking stand-alone cage construct—anteroposterior (a) and lateral (b) views

Full size image

Studies have focussed on the stability profile, outcomes and complications of LSC in contrast to the more traditional anterior plate construct (APC), which have suggested that LSCs have a lower risk of dysphagia as a complication than APCs, while maintaining similar clinical and radiographic outcomes [4–6]. Subsequently, LSCs have increased in popularity. This meta-analysis compares the clinical and radiographic outcomes as well as complication profile of single-level ACDF for degenerative cervical disease between APC and LSC.

Methods

Literature search strategy

Our study was conducted according to the PRISMA guidelines [7, 8]. Electronic databases were searched from the date of inception till 4 July 2016. The databases included Medline, PubMed, EMBASE and Cochrane Database of Systematic Reviews. In order to increase the sensitivity of our search, we combined the terms “zero-profile”, “integrated”, “self-locking”, “anchored”, “stand-alone” and “cervical” as either keywords or MeSH terms. The identified articles, as well as their references, were reviewed according to the selection criteria for consideration of inclusion in the study.

Selection criteria

The systematic review and meta-analysis included studies which adhered to the following inclusion criteria, of studies which: (1) compared patients which had stand-alone locking cages and those which had an anterior plate construct, (2) included patients that only had an operation on one cervical level, (3) included patients who had surgery for symptomatic degenerative spine disease, (4) reported one of the following outcome measures: operative time, blood loss, Visual Analogue Score (VAS), Japanese Orthopaedic Association (JOA) score, Neck Disability Index (NDI) score, fusion rate, subsidence, cervical lordosis, segmental Cobb angle and prevertebral thickness. Our exclusion criteria included patients that had a non-degenerative indication for ACDF such as trauma or tumour, as well as patients that had a multilevel fusion.

We also excluded conference papers, case reports, letters to the editor and abstracts. Only English language articles were included in the study. The LSC defined in the study included only those that had a profile that did not extend anteriorly to the vertebral end plate and a locking mechanism.

Data extraction and critical appraisal

Articles from the literature search were critically reviewed by two authors (MN, KP) with regard to suitability for inclusion in the study, according to the critical review checklist of the Dutch Cochrane Centre proposed by MOOSE [9].

Data were collected from article text, tables and graphs.

Statistical analysis

The odds ratio (OR) was used as a summary statistic. In the present study, both fixed- and random-effects models were tested. In the fixed-effects model, it was assumed that treatment effect in each study was the same, whereas in a random-effects model, it was assumed that there were variations between studies. χ 2 tests were used to study heterogeneity between trials. I 2 statistic was used to estimate the percentage of total variation across studies, owing to heterogeneity rather than chance, with values greater than 50% considered as substantial heterogeneity. I 2 can be calculated as: I 2 = 100% × (Q − df)/Q, with Q defined as Cochrane’s heterogeneity statistics and df defined as degree of freedom. If there was substantial heterogeneity, the possible clinical and methodological reasons for this were explored qualitatively. In the present meta-analysis, the results using the random-effects model were presented to take into account the possible clinical diversity and methodological variation between studies. Specific analyses considering confounding factors were not possible because raw data were not available. All p values were two-sided. Review Manager (version 5.3, Copenhagen, The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) was used for statistical analysis.

Results

Literature search

Electronic database searches yielded 3665 in PubMed, 2437 in EMBASE and 1940 in Medline. Screening of titles and abstracts yielded 44 articles which were subsequently assessed for suitability for inclusion in the study. Exclusion based on our criteria yielded six studies [4, 10–14] which were subsequently included in our quantitative analysis. Figure 3 shows the PRISMA flowchart for inclusion of articles. Assessment of the quality of articles included in the study according to the MOOSE criteria is described in Table 1. Study characteristics are described in Table 2.

Fig. 3
figure 3

PRISMA flow chart of systematic review

Full size image
Table 1 Study evaluation according to MOOSE guidelines
Full size table
Table 2 Study characteristics
Full size table

Demographics

Our analysis included a total of 325 patients (158 APC and 167 LSC groups), which included a total of 143 females and 182 males. LSC constructs used in the studies included Zero-P (DePuy Synthes), PEEK Prevail (Medronic Sofamor Danek, Memphis, TN), as well as unspecified stand-alone cages involving a screw locking mechanism. A breakdown of mean ages of patients as well as follow-up times in each study is listed in Table 3.

Table 3 Study patient demographics and surgical details
Full size table

Operative time

Four studies noted operative time for APC and LSC groups. Figure 4 shows the forest plot analysis for operative time. Mean operative time was greater for the APC group in three studies [10, 13, 14]. Overall, operative time was significantly greater in the APC group compared to the LSC group (MD 7.22 min, 95% CI 0.33–14.11, p = 0.04).

Fig. 4
figure 4

Operation time

Full size image

Blood loss

Data regarding blood loss during the operation were reported in four studies. Figure 5 describes this information in a forest plot. The APC group was noted to have a higher mean blood loss in all four studies [10, 12–14]. Overall, the APC group had a higher amount of blood loss compared to the LSC group (MD 16.79, 95% CI −1.77 to 35.36, p = 0.08).

Fig. 5
figure 5

Blood loss

Full size image

Clinical outcome scores

Mean difference in post-operative JOA scores between APC and LSC groups was reported in one study [13] and was not significant (MD 0.20, 95% CI −1.0 to 0.60, p = 0.63). Post-operative NDI was reported in one study [13] and noted a non-significant mean difference between the two groups (MD 0.30, 95% CI −0.66 to 1.26, p = 0.54). Similarly, post-operative VAS was reported in one study [14], which found a non-significant mean difference between APC and LSC groups (MD 0.80, 95% CI −1.89 to 0.29, p = 0.15).

Radiological outcome

Forest plots detailing analysis of post-operative segmental Cobb angle and cervical lordosis are presented in Figs. 6 and 7, respectively.

Fig. 6
figure 6

Post-operative segmental Cobb

Full size image
Fig. 7
figure 7

Post-operative cervical lordosis

Full size image

Radiographic fusion was reported in four studies, with results varying from 83 to 100% at the final follow-up, and forest plot analysis is presented in Fig. 8. Meta-analysis showed an odds ratio of 0.45 (95% CI 0.10–2.07, p = 0.31) in favour of a LSC construct; however, this was not significant. There was no significant difference between the two groups in post-operative cervical lordosis (MD 1.40 degrees, 95% CI −0.34 to 3.13, p = 0.12) or post-operative segmental Cobb angle (MD 1.20, 95% CI −0.12 to 2.53, p = 0.07).

Fig. 8
figure 8

Fusion

Full size image

Complications

Subsidence was reported in three studies [4, 10, 11], with no significant difference noted between the two groups (OR 0.70, 95% CI 0.30–1.67, p = 0.42), and Fig. 9 details the forest plot analysis. None of the articles noted the incidence of preoperative dysphagia. Post-operative dysphagia was noted to be higher in the APC group (OR 1.77, 95% CI 0.74–4.25, p = 0.20). In the follow-up period, the APC group still had a higher incidence of dysphagia (OR 6.17, 95% CI 1.04–36.64, p = 0.05). Post-operative and follow-up dysphagia forest plots are described in Figs. 10 and 11, respectively.

Fig. 9
figure 9

Subsidence

Full size image
Fig. 10
figure 10

Post-operative dysphagia

Full size image
Fig. 11
figure 11

Follow-up dysphagia

Full size image

Discussion

The literature includes three recent meta-analyses which compare anterior plate constructs with locking stand-alone cages [15–17]. While these meta-analyses also included patients who underwent multilevel instrumentation, our analysis focuses on patients who only had single-level ACDF.

Surgical measures

Our analysis found that operative time was statistically less in the LSC group, with a mean difference of 7.2 min. Blood loss was also noted to be lower in the LSC group, though the data had high heterogeneity (I 2 = 87%). Dong et al. [17] also found that LSC had a statistically significant shorter operation time and less blood loss. Shao et al. [15] noted no significant difference in operative time, however significant less blood loss LSC group. The decreased surgical time in the LSC group is likely due to the fact that less surgical exposure is required to insert the device. This in turn would contribute to a decreased amount of blood loss, due to less soft tissue structures being involved in dissection and exposure.

Clinical outcomes

All functional outcomes scores (VAS, NDI and JOA) were similar in both LSC and APC groups with no statistically significant difference in scores. This has been noted in previous meta-analyses [16, 17]. Clinical functional outcomes correspond to pain and neurological dysfunction. Both LSC and APC groups require total discectomy, and thus, both procedures would be equivalent in providing cervical decompression of neural structures and restoring intervertebral height. Therefore, it is understandable that both techniques impart a similar functional outcome for patients.

Radiological outcomes

The post-operative segmental Cobb angle was noted to be non-significantly increased in the APC group (MD 1.20°). The post-operative cervical lordosis was also increased in the APC group (MD 1.40), although this was not significant (p = 0.12). The Dong et al. [17] meta-analysis also noted a higher post-operative segmental Cobb in APC (MD −0.98, p < 0.01) and no significant difference in cervical lordosis. A study on interobserver reliability in the measurement of Cobb angle noted a standard deviation of up to 3.2° [18]. The mean differences noted between APC and LSC constructs, therefore, are unlikely to be clinically relevant.

Our analysis did not demonstrate a significant difference in fusion rates between the APC and LSC groups, which was also noted in previous meta-analyses [16, 17]. This finding links to previous biomechanical studies which note the similar ranges of motion and stability profiles in APC and LSC construct types [19–21], which would facilitate a similar degree of motion at the fusion site, and would therefore lead to similar fusion rates.

Complications

The Liu et al. [16] meta-analysis showed a statistically significant difference in subsidence rates, which was higher for LSC (risk difference 0.13, 95% CI 0–0.26). This was largely due to the inclusion of the Shi et al. [22] study, which studies non-contiguous ACDF, and showed a statistically significant higher subsidence rate in LSC patients compared with the APC group. Other studies in the Liu et al. analysis did not demonstrate a statistically significant difference [4, 10]. Our analysis showed no statistically significant difference in subsidence, and this correlates with the similar biomechanical properties of both construct types and similar fusion rates.

Dysphagia is a known complication of ACDF, of which the exact causal mechanism is unknown. Fountas et al. [3] in a review of 1015 patients undergoing ACDF cited an incidence of 9.5%, while Bazaz et al. [23] noted an incidence of 50% at one month after anterior spine surgery. Fountas noted that patients undergoing three-level fusion had a statistically significant higher incidence than those undergoing single- or two-level fusions, suggesting that iatrogenic irritation to soft tissues during surgical exposure is a contributing factor for dysphagia [3]. This is highlighted by the fact that less exposure of the anterior vertebrae is required to insert a LSC. A decreased incidence of dysphagia following instrumentation with small plate profiles [2] implies that plate design may have an effect on soft tissue structures. Previous meta-analyses have noted higher early post-operative and follow-up dysphagia in patients undergoing APC than LSC [15–17]. We also note the subjective nature of describing the symptom of dysphagia, as well as the lack of information regarding the effect that dysphagia has had for a patient. We encourage the use of scoring system for dysphagia in future studies, such as the commonly used Bazaz scoring method [23], which would provide a standardised reporting method for this complication.

Limitations

Our meta-analysis was restricted to the inclusion of only seven articles, which only included one randomised control trial. Furthermore, there were different functional outcome scores used by different articles, which overall resulted in even less articles being used for analysis. This limits the reliability of meta-regression. There was an inherent language bias, where non-English language articles were excluded. While most studies used the Zero-P (DePuy Synthes) prosthesis, there was no uniform homogeneity in prosthesis type or surgical technique. This may have an effect on biomechanical construct stiffness which in turn may influence radiographic and clinical outcomes. We also note the variability in length of follow-up of the studies, which impacts on the reporting of post-operative complications as well as functional outcome scores.

Conclusion

The LSC is equivalent to APC in single-level ACDF with regard to operative time, blood loss, functional outcomes and fusion rates. However, based on the available literature, post-operative dysphagia is noted to be higher in the APC group. Further RCTs are required to confirm our results. We also suggest uniformity in assessing post-operative dysphagia via a structured questionnaire.