Introduction

Anterior cervical discectomy and fusion (ACDF) has been shown to be a an optimal standard procedure to treat cervical degenerative disc diseases (DDD)1. However, many studies have reported that ACDF results in decreased range of motion (ROM) at the fused segments and compensatory increased motion at the adjacent levels, which leads to the acceleration of adjacent segment degeneration (ASD)2. Sometimes, symptomatic ASD requires a second surgery.

Artificial cervical disc replacement (A-CDR) serves as an alternative to rigid cervical arthrodesis in the treatment of cervical DDD. The concept is that CDR could stabilize the symptomatic level while preserving some motion, which in theory may reduce the risk of ASD3. In-vitro biomechanical study also found that CDR restores the functional biomechanical properties of the motion segment, protects neurovascular structures and re-establishes near-normal kinematics to the functional spinal unit4. Recently, a few high quality clinical studies have shown that, compared with ACDF, CDR is safe, effective and even advantageous5,6,7.

ACDF, anterior approach direct decompression, was associated with better outcomes than posterior approach (indirect decompression) in the treatment of multilevel cervical spondylotic myelopathy8. Traditionally, multilevel ACDF was used to treat cervical spondylotic myelopathy induced by multilevel cervical DDD, resulting in greater loss of mobility in operative levels9. Although multilevel cervical DDD could benefit from multilevel CDR, strict indications are narrow for multilevel CDR10. Hybrid surgery (HS), combined ACDF and CDR, has advantages in both techniques, especially for multilevel cervical DDD in various degeneration grades11,12. At present, several published studies have compared HS with ACDF in the treatment of multilevel cervical DDD13,14,15,16,17. However, controversies still exist between HS and ACDF for treating these patients. Therefore, the current study was conducted to critically review and summarize the literature comparing the results of HS versus ACDF in the treatment of multilevel cervical DDD, in order to identify the better choice.

Methods

Search strategy

Electronic databases including Cochrane Library, Medline (1966–2015.3), PubMed (1966–2015.3), Embase (1980–2015.3), ScienceDirect (1985–2015.3), CNKI (1985–2015.3), WANFANG DATA (1985–2015.3) and CQVIP (1985–2015.3) were searched. Gray studies were identified from the reference of included literature. No language was restricted. The search process was conducted as follows in Fig. 1. The key words “hybrid”, “replacement or arthroplasty”, “fusion” and “cervical” were used in combination with the Boolean operators AND or OR.

Figure 1
figure 1

Flowchart of the study selection process.

Full size image

Inclusion criteria

Studies were considered eligible for inclusion if they met the following criteria:

Study design: Interventional studies (RCTs or non-RCTs).

Population: Patients with multilevel cervical DDD.

Intervention group: HS (ACDF+CDR).

Control group: ACDF.

Outcomes: Reported at least one of the following: blood loss, operative time, subjective pain perception, neck disability index (NDI), Japanese Orthopaedic Association (JOA) scale, C2–C7 range of motion (ROM), adjacent level ROM and complications.

Exclusive criteria

Patients were excluded from the meta-analysis if they had neoplastic etiology (i.e., metastasis or myeloma), infection, traumatic fracture, spondylolisthesis, serious osteoporosis, metal sensitivity or mental diseases.

Selection criteria

Two reviewers independently screened the titles and abstracts for the eligibility criteria. Subsequently, the full text of the studies that potentially met the inclusion criteria were read and the literature was reviewed to determine the final inclusion. Disagreement was resolved by consulting a third reviewer.

Quality assessment

Quality assessment for randomized trial was conducted according to a modification of the generic evaluation tool used by the Cochrane Bone, Joint and Muscle Trauma Group18 and index for non-randomized studies (MINORS) form for non-randomized clinical trials19. The methodological quality of each trial was scored from 0 to 24. Disagreements were resolved by consensus or consultation with the senior reviewer.

Data extraction

Two researchers independently extracted the data from the included literature. In the case of incomplete data, the corresponding author was consulted for details. The following information was extracted: First author name, year of publication, intervening measures, comparable baseline, sample size and outcome measures. Other relevant parameters were also extracted from individual studies.

Data analysis and statistical methods

Pooling of data was analyzed by RevMan 5.1 (The Cochrane Collaboration, Oxford, United Kingdom). Heterogeneity was estimated depending on the value of P and I2 using the standard chi-square test. When I2 > 50%, P < 0.1 was considered to be significant heterogeneity. Therefore, a random-effects model was applied for data analysis. A fixed-effects model was used when no significant heterogeneity was found. In case of significant heterogeneity, subgroup analysis was performed to investigate sources. For continuous outcomes, mean differences (MDs) and 95% confidence intervals (CIs) were presented. Risk difference (RD) and 95% CIs were calculated for dichotomous data.

Results

Literature search

A total of 424 potential studies were identified with the first search strategy. Of these, 419 reports were excluded according to the eligibility criteria. No additional studies were obtained after the reference review. Ultimately, four non-RCTs and one RCT13,14,15,16,17 were eligible for data extraction and meta-analysis, as indicated by the flowchart in Fig. 1. These studies involved a total of 74 patients in the HS group and 86 patients in the ACDF group.

Study characteristics

The main characteristics of the included studies are reported in Table 1. Statistically similar baseline characteristics were observed between the HS and ACDF groups. All studies had small sample sizes, from 14 to 48 patients. The number of operative levels was two or three. The design of CDR varied among different studies. All studies except one were followed up more than 18 months15.

Table 1 Characteristics of included studies.
Full size table

Risk of bias assessment

The quality of the included studies was assessed according to the Cochrane Handbook for Systematic Review of Interventions. Only one RCT17 met the inclusion criteria. Randomization was conducted using the odd or even hospital number, which showed a high risk of selection bias; adequate concealment of allocation was unclear for RCT. For the non-RCTs13,14,15,16, the MINORS score was 15-17 for the retrospective controlled trials. The methodological quality assessment is illustrated in Fig. 2 (RCT) and Table 2 (non-RCT).

Table 2 Quality assessment for non-randomized trials.
Full size table
Figure 2
figure 2

The summary of bias risk of randomized controlled trials.

Full size image

Outcomes for meta-analysis

Operation time

Details regarding operation time were available in five studies13,14,15,16,17. There was significant heterogeneity (χ2 = 34.40, df = 4, I2 = 88%, P < 0.00001); therefore, a random-model was performed. Pooling results demonstrated that operation time made no significant difference between the two groups (MD = 2.21, 95% CI: −11.25 to 15.67, P = 0.75; Fig. 3).

Figure 3
figure 3

Forest plot of operation time between two groups.

Full size image

Blood loss

Four articles reported blood loss13,14,15,16. Significant heterogeneity was found, so a random-model was used (χ2 = 6.61, df = 3, I2 = 55%, P = 0.09). There was significance between HS groups and ACDF groups regarding blood loss (MD = −27.80, 95% CI: −47.08 to −8.52, P = 0.005; Fig. 4).

Figure 4
figure 4

Forest plot of blood loss between two groups.

Full size image

Preoperative and postoperative visual analog scale (VAS)

Two studies showed preoperative and postoperative 24 months VAS14,17. For preoperative VAS, there was no significant heterogeneity, so a fixed-model was performed. No significant difference was found between HS groups and ACDF groups (MD = −0.10, 95% CI: −0.85 to 0.65, P = 0.79; Table 3). For postoperative 24 months VAS, there was significant heterogeneity; thus, a random-model was performed. No significant difference was found between HS groups and ACDF groups (MD = −1.00, 95% CI: −2.46 to 0.47, P = 0.18; Table 3).

Table 3 Meta-analysis results.
Full size table

Preoperative and postoperative C2–C7 ROM

Relevant data regarding the preoperative C2–C7 ROM were documented in four articles14,15,16,17. No significant heterogeneity was found, so a fixed-model was applied. There was significant difference between HS groups and ACDF groups (MD = 0.51, 95% CI: −3.73 to 4.74, P = 0.81; Table 3). Postoperative 24 months C2–C7 ROM were documented in two articles13,17. No significant heterogeneity was found; thus, a fixed-model was applied. There was significant difference between HS groups and ACDF groups (MD = 9.50, 95% CI: 1.44 to 15.40, P = 0.002; Table 3).

Preoperative and postoperative ROM of the superior adjacent segment

Four articles reported the preoperative ROM of the superior adjacent segment14,15,16,17. A fixed-model was used due to low-significance heterogeneity and showed no significant difference between groups (MD = −0.46, 95% CI: −2.15 to 1.23, P = 0.60; Table 3). Two articles reported the postoperative 24 months ROM of the superior adjacent segment13,17. A random-model was used due to high-significance heterogeneity and showed a significant difference between groups (MD = −5.24, 95% CI: −7.57 to −2.92, P < 0.00001; Table 3).

Preoperative and postoperative ROM of the inferior adjacent segment

Preoperative ROM of the inferior adjacent segment was provided in four trials14,15,16,17. No significant heterogeneity was showed between pooling results; thus, a fixed-model was performed. There was no significant difference between groups (MD = −0.45, 95% CI: −1.78 to 0.87, P = 0.50; Table 3). Postoperative 24 months ROM of the inferior adjacent segment was provided in two trials13,17. No significant heterogeneity was shown between pooling results; thus, a fixed-model was performed. There was significant difference between groups (MD = −4.62, 95% CI: −7.28 to −1.96, P = 0.0007; Table 3).

Complications

Data regarding complications were provided in all studies13,14,15,16,17. No significant heterogeneity was shown between pooling results; thus, a fixed-model was performed. There was no significant difference between groups (RD = −0.01, 95% CI: −0.09 to 0.06, P = 0.73; Fig. 5).

Figure 5
figure 5

Forest plot of complications between two groups.

Full size image

Other outcomes

Neck Disability Index (NDI) scores and Japanese orthopaedic association (JOA) scores were used to assess functional recovery in most studies. However, limited data could not be extracted for meta analysis, as shown in Table 4. Only one study reported better recovery of NDI at 24 months postoperatively for the HS group14.

Table 4 Functional assessment of included studies.
Full size table

Discussion

The most important findings of the present study were that HS reduces blood loss, postoperative C2–C7 ROM was close to the physiological level and no decrease in the ROM of the adjacent segment was noted in the HS group. Although most included studies reported consistent results, more consideration should be given when we interpret these results.

For published studies with small samples, we searched and included all the RCT and non-RCT. Four non-RCTs13,14,15,16 and one RCT17 met the inclusion criteria in the meta-analysis. There were biases for randomization and concealment of allocation in the RCT. Quality-assessment scores of non-RCTs ranged from 15 to 17. No prospective calculation of the sample size was described in non-RCTs. In addition, the assessment of the study endpoints was biased. All these shortcomings weaken the level of evidence.

In 2014, Jia et al. performed a systematic review of biomechanical and clinical evidence of HS for multilevel cervical DDD20. Although their results are consistent with our meta-analysis, they did not extract data for further quantitative analysis. To our knowledge, this is the first quantitative meta-analysis to evaluate HS for multilevel cervical DDD by only including studies that had appropriate control and study groups.

Previous meta-analysis21,22,23,24 reported that ACDF was associated with shorter operative times and less blood loss in single level. In our meta-analysis, HS reduced blood loss and did not increase operation time for multilevel cervical DDD. Decreased blood loss and operation time for CDR are attributable to the surgeons experienced with this technology. Therefore, CDR is considered less invasive.

In our meta-analysis, preoperative VAS scores were not significantly different between the two groups. Postoperative 24 months VAS scores significantly improved in each group, without significant differences between the two groups. Although the functional assessment did not perform a quantitative meta-analysis due to limited data, most of the included studies reported that there was no significant difference between the two groups. These results suggest that both procedures effectively reduce patients’ pain and improve function.

HS aims to keep the kinetics of the adjacent segment to the greatest extent avoiding ASD. Present meta-analysis showed that HS improved postoperative 24 months C2–C7 ROM and decreased superior and inferior adjacent segment ROM compared with ACDF. These results are consistent with previous biomechanical studies. Cervical motion after HS is closer to the physiological kinetics than after ACDF.

In all, 7 complications were reported in the two groups, including 1 dysphasia, 1 heterotopic ossification and 1 residual limb symptom for the HS group. For the ACDF group, the corresponding complications were 1, 1 and 2, respectively. Meta-analysis showed that there was no significant difference between the two groups. This result suggests that both the treatments are safe in short-term follow-up. Recently, a meta-analysis of 4 included RCTs with 4–7 years of follow-up showed that CDR results in better clinical and radiographic outcomes7. Complications may increase in patients with middle- or long-term follow-up.

Several potential limitations should be acknowledged in the present meta-analysis: 1) Only one RCT and four non-RCTs were identified and the sample sizes of the included studies were relatively small; 2) There were some methodological weaknesses in the included RCT and non-RCTs; 3) Because some data were incomplete, we failed to conduct meta-analysis such as functional score; 4) Follow-up was relatively short and may lead to underestimating complications.

Conclusion

Hybrid surgery demonstrated excellent clinical efficacy and radiological results. Postoperative C2–C7 ROM was closer to the physiological status. No decrease in the ROM of the adjacent segment was noted in the hybrid surgery group. Large sample sizes, long-term follow up and well designed studies are needed in the future.

Additional Information

How to cite this article: Tian, P. et al. Hybrid surgery versus anterior cervical discectomy and fusion for multilevel cervical degenerative disc diseases: a meta-analysis. Sci. Rep. 5, 13454; doi: 10.1038/srep13454 (2015).