ACDF has been one of the standard treatments for cervical spondylopathy since it was first described in the 1950s[16, 17]. For multilevel lesions, although many prefer the posterior approach to achieve broader indirect decompression[18], the anterior approach has the characteristic of direct decompression of the neurological structures, less interference with posterior muscles, and hence less surgical trauma[19–22]. As experience with the anterior cervical plate has accumulated, the advantages of an enhanced fusion rate, better cervical spine realignment, and lordosis maintenance have been proven[23]. However, complications of dysphagia and tracheoesophageal lesions, which increase with the number of fusion levels[24, 25], are a concern[26, 27]. The zero-profile anchored cage, which integrates the stand-alone cage and the fixating screws or clips, has been developed to reduce the above complications and shorten the time of operation in multisegment disorder. Several studies have confirmed that these cages effectively avoid the complications of plates with similar surgical efficacy[22, 28]. However, few studies have reported mid- to long-term results.
Fusion
The reconstructed multiaxial CT scan has the best interobserver reliability of predicting pseudoarthrosis[29], so CT scans are used to evaluate for extragraft bone bridging (ExGBB) given its superior diagnostic qualities[13]. However, the evaluation of fusion status is either vague or subjective sometimes, as there are no specific parameters. And it is limited to findings derived from a static moment in time[13]. It fails to assess the dynamic changes during motion, which makes some cases of pseudoarthrosis seen only with movement. Song et al.[30] showed that using ISM ≥ 1 mm as the cutoff for detection of anterior cervical pseudarthrosis on radiographs magnified 150% and made with a superjacent interspinous motion of ≥ 4 mm yielded accuracies comparable with those of CT. In our study, we also found that patients who appeared to be fused on CT did not meet the ISM fusion standard(Fig. 6). In addition, some patients refused to take CT scans due to the high cost and time. To ensure the consistency of evaluation standards we use the ISM criterion to evaluate pseudarthrosis instead of CT.
A systematic review reported the results of ACDF using different kinds of implants[31]. The bone fusion rate was 88.6% in 5738 patients treated with a stand-alone cage (without a screw-plate or integrating fixation system), and the fusion rate was 91.4% in 3971 patients treated with a screw-plate system. Nevertheless, the highest fusion rate was 96.6% in 499 patients treated with a zero-profile anchored cage. The fusion rate of our 5-year follow-up study was 100%, which also proved that the zero-profile anchored cage achieved a satisfactory fusion effect. The zero-profile anchored cage we used was a cage with two anchoring clips. These unique structures offer a fixation mechanism that is similar to the function of a plate and screws[32]. Scholz et al.[33] found that the anchored spacer provided similar biomechanical stability to that of the established anterior fusion technique using an anterior plate and cage. In our study, we also found an excellent fusion rate with good stability. The self-locking clips ensure excellent primary stability of the implant and promote early fusion. Furthermore, the elastic modulus of the anchored cage is similar to that of bone, which theoretically helps to decrease the stress shielding effect and increase bony fusion. The anatomical shape of the anchored cage allows a wide grafting space and close contact between the endplate bone and the implant. In addition, this satisfactory fusion rate may be related to surgical techniques that include optimal preparation of the fusion bed and proper disc space distraction. Although clip fractures have occurred in these cases, they may collapse the intervertebral space, and fracture clips can limit the effect to some extent, resulting in a lack of apparent early instability. Similarly, despite the presence of cases of cage migration, the full clip can still have a limited effect; therefore, they ultimately achieve stable fusion.
The fusion rate at the 1-year follow-up of our series was 85.8%, with the extension of follow-up time, the fusion rate gradually increased to 100%. Our finding was consistent with the previous studies. Sun et al.[34] performed a 5-year 3-level ACDF study and found 81 surgical levels were completely fused at the 5-year follow-up, but six levels remained unfused in the first year after surgery. In a similar three-year study[35], the fusion rates at 1, 2, and 3 years after surgery were 77.9%, 94.1%, and 100%, respectively. In another study[36] involving 78 patients who also used the spinous processes criteria to assess fusion rate, all surgical levels were fused within the first year. Although the researcher claimed that the peak number of new fusion levels was 3–6 months postoperative, it also indicated that the overall fusion rate increased over time. The speed of fusion may relate to age[37], smoking[37], osteoporosis[38], diabetes,[39] drinking[39] and other factors. Indeed, of the 18 levels of pseudarthrosis (14 patients) at 1-year follow-up, 6 patients had been smoking for more than 20 years with at least one pack per day. At the 2-year follow-up, a total of 4 levels of pseudarthrosis (4 patients) were found, and 3 of them were long-term smokers. There were 3 out of the 5 diabetic patients had pseudarthrosis at 1-year follow-up, and two of them still had pseudarthrosis at 2-year follow-up. A total of one patient was both a chronic smoker and a diabetic. Unfortunately, It was not until the fifth year of follow-up that the patient achieved fusion at all surgical levels.
Subsidence
Subsidence is an adverse event after ACDF. Many studies have shown that cage subsidence is the main complication in ACDF with stand-alone cages regardless of the composite material[40–46]. In a systematic review comprising 4784 patients using single- and multilevel cages, the mean subsidence rate was 21.1%, ranging from 0 to 83%[47]. The subsidence of the cage causes loss of intervertebral disc height, which may result in the narrowing of the foramen, nerve root compression, and pseudarthrosis due to cervical instability[47]. Our previous study[9] indicated that subsidence occurs mainly in the early postoperative period. In this study, the subsidence rate was 14.3%. However, there was no new subsidence during 2 to 5 years of follow-up, which indicated that it occurred early after surgery. This was consistent with a prospective study conducted by Igarashi et al.[48] Most subsidence occurred in the first month after surgery, probably due to the great pressure applied to the interior endplate when patients got out of bed in the initial days following surgery [48]. Cho et al.[49] compared Zero-P (zero-profile cage with integrated screws) with a stand-alone cage without integrating a fixation system. In this comparison, the subsidence of the Zero-P was lower, which seems logical since the screws force the cage to be placed in the anterior cortical plane, and the screws themselves are placed cortically.
The endplate may be another factor affecting subsidence. Lowe et al.[50] performed compression tests on vertebral bodies under an intact endplate, partial and complete resection of the endplate, and compared the failure load of each group. They showed that the ultimate compressive strength of intact endplate vertebral bodies was significantly higher than that of vertebral bodies with endplate resection. The injury of the endplate of the vertebral body significantly increased the incidence of the deposition and displacement of the interbody fusion device. Therefore, the endplate should be handled gently to prevent injury to reduce the occurrence of the deposition of the interbody fusion device[51]. In our study, all of the surgeries were performed by the same surgeon with similar endplate preparation. However, the insertion of anchored clips through the cage we used would destroy the endplate to a certain extent, making it one of the factors leading to the inevitable subsidence.
The height and size of the cage are both related to subsidence. Truumees et al.[52] reported that higher distractive and compressive forces were recorded with larger grafts. Yamagata et al.[53] found that a cage height of 6.5 or 7.5 mm had a higher risk of cage subsidence than a height of 4.5 or 5.5 mm. This finding can be explained by the larger amount of stress on the vertebral endplates, which presumably results in subsidence. Yang et al.[54] found that the use of 14 mm anteroposterior diameter cages led to a significantly lower risk of subsidence than using 12 mm diameter cages. This is because the cage with a larger surface area reduces the pressure per unit area of the endplate, thus reducing the risk of subsidence. Combined with our study, we recommend minimizing overdistraction during surgery and using the widest cage to minimize the risk of subsidence.
Alignment
Loss of cervical lordosis after surgery is one of the major disadvantages of using a zero-profile anchored cage. Pereira et al.[55] reported that the use of multiple independent PEEK cages in the treatment of multilevel cervical spondylotic myelopathy achieved good intermediate results. A meta-analysis demonstrated that the cage-plate technique resulted in significantly greater cervical lordosis than the stand-alone cage technique[56]. Many degenerative patients have wedge-shaped vertebrae with a lower anterior wall and a higher posterior wall. Given the zero-profile cage with the same height from the front to back, the cervical lordosis could not be recovered or maintained without additional reduction force by the screw-plate system. In our study, the loss of cervical curvature occurred mainly in the early postoperative period. In our last 2-year follow-up study[9], the maintenance of cervical curvature in the early period after surgery was not ideal, and concerns have been raised to determine the association between the loss of curvature and symptoms. This study proved that the curvature tends to be stable after fusion. The clinical outcomes of all patients were maintained during the 5-year follow-up, indicating that the loss of cervical curvature to a certain distance was unrelated to postoperative symptoms. In our follow-up case, the largest change was from 22.5° immediately after surgery to 3.8° at the 5-year follow-up, but the patient's postoperative symptoms continued to improve. Nevertheless, we failed to determine a definite threshold due to the small sample size of our study. This finding was consistent with the study of Spanos et al.[57], who included single- or consecutive two-level ACDF with a PEEK interbody cage and reported that the slight increase in cervical lordosis after ACDF was lost at the 12-month follow-up but had no correlation with the pain or function of the subjects. The study of Godlewski et al.[58] also demonstrated that the greatest changes in lordosis and disc space height after ACDF were noted immediately after surgery but reduced over time, and this change was not correlated with clinical outcomes. Song et al.[59] compared the PEEK stand-alone cage group with the cage-plate group in 1- and 2-level ACDF. At approximately three years of follow-up, they found that the cage-plate group had a better sagittal alignment, a higher fusion rate, and lower cage subsidence. However, there were no significant differences in clinical outcomes between the two groups. They explained that the clinical sequelae resulting from loss of cervical lordosis may take years to develop and thus may not show up in these relatively short follow-ups. Meng et al.[60] reported a 3-level ACDF and hybrid cervical surgery study with a 5-year follow-up. They also found that there was no correlation between cervical balance and clinical outcomes. This result is also consistent with the results of this study, subsidence and cervical lordosis have no correlation with clinical outcomes.
Hoarseness and dysphagia
Voice hoarseness and dysphagia were prevalent postoperative complaints following ACDF surgery. Rates of immediate postoperative dysphagia after ACDF ranged from 1.7–67%[61], In multiple series, the rates of symptomatic postoperative recurrent laryngeal nerve palsies (RLNP) ranged from 0.9%-8.3%[62–65]. Gowd et al.[61] claim that swallowing dysfunction correlated with advanced age, a prior history of such dysfunction, longer surgical procedures, and a trend toward increased dysphagia with multilevel surgery attributed to greater soft tissue swelling/retraction injury. Tsalimas et al.[8] found that female sex, smoking, surgical approach, use of rhBMP-2, and multilevel surgery were risk factors for dysphagia after ACDF.
In our series, mild post-operative dysphagia was reported by 35.3% of patients at early postoperative follow-up. However, it improved over time and disappeared within six months after surgery (Table 3). It suggests that the onset of dysphagia in multilevel cases could be related more likely to the surgical procedure and soft-tissue swelling. For hoarseness, the primary mechanism of injury is through direct compression or traction of the RLN rather than accidental resection. The endotracheal tube cuff, when inflated, has been reported to exert direct compression forces that can induce nerve ischemia. Gowd et al. put forward that it is likely that there is an additive effect of retractor placement and increased endotracheal cuff pressure that results in voice hoarseness. Furthermore, increased BMI may increase intraoperative pressure and subsequently add to postoperative symptoms. These apparent findings may suggest that attention be paid to each of these contributions during surgery to minimize postoperative hoarseness[61].
Limitations
There were limitations of the study. This was a single-center, retrospective study with a relatively small number of patients. Confounding factors, such as patients' bone mineral density, history of medication, and other risk factors for bone fusion, were not fully excluded. Further prospective randomized multi-center studies with long-term follow-up are needed to investigate the correlation between clinical and radiological findings.