Objective: Whether cervical disc arthroplasty (CDA) is superior to anterior cervical discectomy and fusion (ACDF) remains controversial, especially in relation to long‐term results. The present study aimed to evaluate the long‐term safety and efficiency of CDA and ACDF for cervical disc disease.Methods: We performed this study according to the Cochrane methodology. An extensive search was undertaken in PubMed, Embase, and Cochrane databases up to 1 June 2019 using the following key words: “anterior cervical fusion,” “arthroplasty,” “replacement” and “artificial disc”. RevMan 5.3 (Cochrane, London, UK) was used to analyze data. Safety and efficiency outcome measures included the success rate, functional outcome measures, adverse events (AE), adjacent segment degeneration (ASD), secondary surgery, and patients’ satisfaction and recommendation rates. The OR and MD with 95% confidence interval (CI) were used to evaluate discontinuous and continuous variables, respectively. The statistically significant level was set at P < 0.05.Results: A total of 11 randomized controlled trials with 3505 patients (CDA/ACDF: 1913/1592) were included in this meta‐analysis. Compared with ACDF, CDA achieved significantly higher overall success (2.10, 95% CI [1.70, 2.59]), neck disability index (NDI) success (1.73, 95% CI [1.37, 2.18]), neurological success (1.65, 95% CI [1.24, 2.20]), patients’ satisfaction (2.14, 95% CI [1.50, 3.05]), and patients’ recommendation rates (3.23, 95% CI [1.79, 5.80]). Functional outcome measures such as visual analog score neck pain (−5.50, 95% CI [−8.49, −2.52]) and arm pain (−3.78, 95% CI [−7.04, −0.53]), the Short Form‐36 physical component score (SF‐36 PCS) (1.93, 95% CI [0.53, 3.32]), and the Short Form‐36 mental component score (SF‐36 MCS) (2.62, 95% CI [0.95, 4.29]), revealed superiority in the CDA group. CDA also achieved a significantly lower rate of symptomatic ASD (0.46, 95% CI [0.34, 0.63]), total secondary surgery (0.50, 95% CI [0.29, 0.87]), secondary surgery at the index level (0.46, 95% CI [0.29, 0.74]), and secondary surgery at the adjacent level (0.37, 95% CI [0.28, 0.49]). However, no significant difference was found in radiological success (1.35, 95% CI [0.88, 2.08]), NDI score (−2.88, 95% CI [−5.93, 0.17]), total reported AE (1.14, 95% CI [0.92, 1.42]), serious AE (0.89, 95% CI [0.71, 1.11]), device/surgery‐related AE (0.90, 95% CI [0.68, 1.18]), radiological superior ASD (0.63, 95% CI [0.28, 1.43]), inferior ASD (0.45, 95% CI [0.19, 1.11]), and work status (1.33, 95% CI [0.78, 2.25]). Furthermore, subgroup analysis showed different results between US and non‐US groups.Conclusion: Our study provided further evidence that compared to ACDF, CDA had a higher long‐term clinical success rate and better functional outcome measurements, and resulted in less symptomatic ASD and fewer secondary surgeries. However, worldwide multicenter RCT with long‐term follow up are still needed for further evaluation in the future.

The optimal treatment for multi-level cervical spondylotic myelopathy (CSM) remains controversial. Posterior approach is most commonly used, but complicated with insufficient decompression and postoperative axial neck pain. The anterior approach is effective in neural decompression with less surgical trauma. However, the profile of the plate or the possible construct failure may cause dysphagia after surgery. Recently, anterior cervical discectomy and fusion (ACDF) with self-anchored cage is reported to have a superior result over ACDF with anterior plates and screws in three-level CSM. The purpose of the study is to compare the clinical and radiological outcomes of ACDF using stand-alone anchored cages to that of laminectomy with fusion (LF) for treating four-level CSM. Twenty-six patients underwent four-level ACDF (Group A) and 32 patients with four-level LF (Group B) were retrospectively reviewed and followed-up for 24 months. Clinical efficacy was evaluated by comparing pre- and post-operative Japanese Orthopedic Association (JOA) and Neck Disability Index (NDI) scores. Operative time, blood loss, fusion, lordosis change and complications were evaluated. There was significantly less blood loss in Group A (163.4 ± 72.1 ml) than Group B (241.0 ± 112.3 ml) (P < 0.05). Both groups demonstrated significant improvements in JOA and NDI scores after surgery with similar operative time. Improvements in cervical lordosis and fused segment lordosis were more pronounced in Group A (11.3 ± 5.9°, 9.7 ± 5.3°) than Group B (5.8 ± 4.6°, 5.5 ± 4.5°) (P < 0.05). Loss of lordosis in the cervical spine and fused segment was more prominent in Group A (11.7 ± 2.2°, 6.7 ± 3.2°) than Group B (7.5 ± 3.8°, 3.7 ± 3.4°) (P < 0.05) at the final follow-up. Complication rate in Group A and Group B was 57.69 and 18.75%, respectively. ACDF using a stand-alone anchored cage showed similar clinical results to LF for the treatment of four-level CSM, with better lordosis correction and less blood loss. However, ACDF was associated with more loss of lordosis after surgery and more non-unions.