Stand-Alone Cage Versus Anterior Plating for 1-Level and 2-Level Anterior Cervical Discectomy and Fusion

A Randomized Controlled Trial

Abstract

Study Design: 

Prospective, randomized controlled trial.

Objective: 

The aim was to compare perioperative and radiographic outcomes between stand-alone and anterior plated 1 and 2-level anterior cervical discectomy and fusion (ACDF).

Summary of Background Data: 

ACDF with interbody spacer and separate plate/screw construct (PLATE) may be associated with a higher incidence of postoperative dysphagia, increased operative time, and other complications. Therefore, some have opted to utilize an interbody cage with integrated screws and no plate (CAGE) with good results.

Materials and Methods: 

Patients with 1-level to 2-level degenerative disease were prospectively enrolled and randomized into 1 of 2 treatment arms consisting of either PLATE or CAGE reconstruction. Patients were followed for a minimum of 1 year postoperatively. Primary endpoints included improvement on patient-reported outcome metrics, construct integrity, cervical alignment, successful arthrodesis, and subsequent revision surgeries.

Results: 

Forty-six patients were included: 12 with 1-level PLATE, 12 with 1-level CAGE, 12 with 2-level PLATE, and 10 with 2-level CAGE. For 1-level ACDF, PLATE patients reported worse swallow function on Swallowing Quality of Life Questionnaire at 6 weeks (P=0.050) and 6 months (P=0.042). Pseudarthrosis requiring revision was observed in one PLATE patient. For 2-level ACDF CAGE patients reported worse disability on neck disability index (P=0.037) at 6 weeks, as well as worse neck disability index (P=0.017) and visual analog scale neck (P=0.010) at 6 months. However, PLATE patients reported worse swallow function on Swallowing Quality of Life Questionnaire at 6 weeks (P=0.038). There were no differences in the rates of fusion, loss of disc height correction, subsidence, or in sagittal parameters between cohorts for both 1-level and 2-level ACDF.

Conclusion: 

There was greater incidence of transient postoperative dysphagia in both single and 2-level PLATE cohorts. However, early postoperative outcomes were worse for 2-level CAGE in certain patient-reported metrics. This suggests that although anterior instrumentation may be associated with a higher likelihood of dysphagia, it may also lead to higher short-term stability and improved patient-reported outcomes for 2-level fusion.

Anterior cervical discectomy and fusion (ACDF) is the gold-standard treatment for cervical radiculopathy and myelopathy when conservative management has failed.^1,2 Although the procedure has evolved since it was initially described by Smith and Robinson in 1955,³ modern goals include achieving bony union in order to provide stability, restoration of normal cervical lordosis, and decompression of the spinal cord and entrapped nerve roots. Furthermore, the anterior approach minimizes postsurgical pain and morbidity associated with posterior fusion techniques, while also lending itself to improved restoration of cervical lordosis and lower rates of infection and revision.⁴ In order to provide immediate rigid fixation to the fusion segment and decrease micromotion, surgeons commonly opt for the addition of a separate anterior cervical plate and screw construct, as past studies have demonstrated that this method may result in superior arthrodesis and lordotic correction, while decreasing the risk for interbody subsidence, extrusion, and pseudarthrosis.^5,6 Biomechanical load-sharing properties of anterior plating have been well described.

Despite the advantage of immediate stability, the addition of an anterior plate does have theoretical disadvantages such as plate or screw failure, loosening, malposition, and/or a higher incidence of adjacent segment degeneration (ASD).^7–12 Furthermore, plate fixation requires additional manipulation and retraction of the tissue of the anterior neck and creates additional anterior “bulk,” thereby potentially increasing the risk of transient or long-term dysphagia, or dysphonia. This is particularly true for multilevel ACDF surgery.^13–17

To address the concerns associated with anterior plating, stand-alone interbody cage devices have been developed including low-profile or zero-profile constructs with integrated screws. This design has been shown to provide a similar degree of biomechanical stability conferred by anterior plating, while avoiding increased retraction and anterior bulk associated with plating.^18–20 However, current published data is limited and currently there is no consensus among spine surgeons regarding the best technique to achieve both fusion and reliable improvements in clinical outcome measures in patients undergoing ACDF.^21–25

This prospective, randomized controlled trial sought to examine the radiographic, patient-reported, and perioperative outcomes of patients undergoing ACDF for 1 and 2-level degenerative disease, causing symptoms of radiculopathy and/or myelopathy. Patients were randomized into 2 distinct treatment arms with either an interbody device and a separate plate and screw construct (PLATE), or an integrated, low profile stand-alone interbody cage (CAGE) in order to determine clinical and radiographic differences between the treatment groups for both 1-level and 2-level procedures.

MATERIALS AND METHODS

Patient Selection

Institutional review board approval was obtained for this prospective, randomized, controlled trial at a single tertiary academic medical center. All procedures were performed between July of 2017 and February 2020 by the senior surgeon and patients were followed for a minimum of 1 year following surgical intervention. The study was divided into 2 subgroups with 2 treatment arms that were conducted simultaneously; the first comprised of patients receiving treatment with 1-level ACDF and the second with 2-level ACDF. Criteria for enrollment included diagnosis of radiculopathy, myelopathy, myeloradiculopathy, herniated nucleus pulposus, degenerative disc disease, spondylosis, osteophytic complexes, or foraminal stenosis at 1 or 2 contiguous levels of the cervical spine between C3 and C7. All patients underwent a period of conservative treatment before being considered for surgery and had confirmatory imaging with computed tomography or magnetic resonance imaging. Criteria for exclusion from the study included patients under the age of 18, and patients receiving treatment for indications of revision, trauma, tumor, or infection.

After obtaining informed consent, patients in each arm were randomized using a random number generator in a 1:1 ratio to receive treatment with either an interbody device and anterior plating (PLATE), or interbody cage with an integrated 3-screw construct (CAGE). Patients were followed radiographically and clinically 6 weeks, 6 months, and 1 year after surgery. Neither the surgeon nor the patient was blinded to the treatment group. A summary of the enrollment strategy is summarized by Figure 1.

FIGURE 1: CONSORT flow diagram of the enrollment process. CONSORT indicates Consolidating Standards of Reporting Trials.

Surgical Technique/Devices

A standard surgical approach to ACDF was performed in all patients using the technique described by Smith and Robinson.³ After removing anterior osteophytes, the intervertebral discs and cartilaginous endplates were resected and the endplates were prepared down to bleeding cortical bone. The posterior longitudinal ligament was resected with an extended bilateral direct foraminotomy performed in all cases. In PLATE patients, a synthetic interbody spacer was impacted at the appropriate level(s) and a separate static plate and screw construct with variable angle screws was added to bridge the fusion segment. Patients randomized to CAGE similarly had a synthetic interbody cage placed with an integrated titanium 3-screw construct. Specific manufacturers of the devices were heterogenous but all synthetic spacers in both the CAGE and PLATE groups were nonporous polyetheretherketone (PEEK) without surface modifications in order to homogenize endplate interaction and biology. All cases in both treatment arms were treated with a similar biological mixture consisting of autograft harvested from a suction trap, and 1cc/level i-FACTOR peptide biological (Cerapedics Inc., Westminster, CO).

Outcome Measures

Patient electronic medical records were reviewed for baseline demographics and medical comorbidities, which included age, sex, body mass index, smoking status, hypertension, diabetes mellitus, American Society of Anesthesiologists (ASA) Physical Status Class, and the Charlson comorbidity index. Intraoperative characteristics including operative time and estimated blood loss were used as metrics to compare surgical procedure efficiency and complications. Patient-reported outcomes surveys were assessed preoperatively and postoperatively for differences in primary patient outcomes, including improvement in pain and disability. Surveys included the neck disability index (NDI), visual analog scale (VAS) arm, VAS neck, and Veteran’s RAND 12-Item Health Survey (VR12). These surveys were used to describe a patient’s functional status preoperatively and postoperatively. The Swallowing Quality of Life Questionnaire (SWAL-QOL), a validated and sensitive 44-item tool, was used to identify patients with oropharyngeal dysphagia and quantify severity.²⁶ Given that dysphagia is a subjective sensation, tools measuring patient-reported outcomes such as the SWAL-QOL are considered more clinically relevant than more invasive objective instruments such as dynamic swallowing studies.²⁷

Radiographic outcome measures included functional intervertebral level height, subsidence of the interbody device into the superior and inferior endplates, and parameters of cervical sagittal alignment including T1 slope, cervical sagittal vertical axis, cervical lordosis, and fusion segment lordosis. Disc space height was measured as the distance between endplates on sagittal radiographs at the anterior, middle, and posterior disc spaces. Subsidence was measured separately as the distance the interbody cage settled into the superior and inferior endplates of the motion segment. To correct for magnification on X-ray and improve the accuracy of all measurements, the magnification factor was acquired by measuring the width of the C5 vertebral body on sagittal view magnetic resonance imaging and all preoperative and postoperative X-rays. This magnification factor was then applied to all measurements in order to obtain the “indexed” values. The presence of fusion was evaluated using the Bridwell grading system using radiographic images from the 6-month and 1 year follow-up time points.²⁸ The development of ASD was determined clinically by the development of symptoms of radiculopathy or myelopathy and was supported by radiographic evidence.

Statistical Analysis

Categorical and continuous variables were analyzed through appropriate descriptive statistical testing using SPSS version 26.0 (IBM Corporation, Armonk, NY). Baseline patient demographics, comorbid medical conditions, perioperative characteristics, and primary clinical and radiographic outcomes were compared across 1-level and 2-level PLATE and CAGE cohorts using χ² with the Fisher exact test for categorical variables and the Mann-Whitney U test or independent t test for continuous variables. Magnitude of improvement on patient-reported outcomes (PROs) was assessed if there were differences in baseline scores. Categorical variables were reported as proportions and percentages of total counts. Means and SDs were used to describe continuous variables. The threshold for statistical significance was set a priori to P<0.05. A priori power analysis called for 64 patients in each group (128 total) in order to detect a 5% difference in arthrodesis rate. However, the trial was stopped at the reported timepoint because of concerns regarding differences in early postoperative neck pain and disability, in addition to the cost of continuing the trial.

RESULTS

Demographics and Operative Data

At the end of the study’s enrollment period, 46 patients had been randomized into treatment groups, 24 of whom underwent single-level ACDF and 22 who underwent a 2-level procedure. Patient demographics are demonstrated by Table 1.

TABLE 1 - Patient Demographics Data for Single-Level and 2-Level PLATE and CAGE Groups

	Single-Level				Two-Level
Variable	PLATE	CAGE	P		PLATE	CAGE	P
Age	57.25±10.56	57.18±12.97	0.989		62.08±9.17	53.50±10.82	0.058
Sex			0.640				0.999
Male	83.3% (10/12)	66.7% (8/12)			50% (6/12)	50% (6/12)
Female	16.7% (2/12)	33.3% (4/12)			50% (6/12)	50% (6/12)
BMI	34.48±7.53	30.16±6.32	0.154		31.42±7.60	29.71±3.84	0.527
ASA ≥3	60% (3/5)	22.2% (2/9)	0.266		87.5% (7/8)	25% (1/4)	0.067
CCI	0.70±0.82	0.36±0.67	0.317		0.82±1.78	0.22±0.67	0.356
Smoking history	50% (5/10)	50% (5/10)	0.999		63.6% (7/11)	28.6% (2/7)	0.335
Opioid usage	18.2% (2/11)	18.2% (2/11)	0.999		18.2% (2/11)	0% (0/10)	0.476
Hypertension	50% (6/12)	27.3% (3/11)	0.400		75% (9/12)	40% (4/10)	0.192
Diabetes	9.1% (1/11)	18.2% (2/11)	0.534		18.2% (2/11)	10% (1/10)	0.593
Fusion Segment			0.406				0.561
C3-C4	33.3% (4/12)	16.7% (2/12)		C3-C5	8.3% (1/12)	0% (0/10)
C4-C5	0% (0/12)	8.3% (1/12)		C4-C6	25% (3/12)	40% (4/10)
C5-C6	58.3% (7/12)	41.7% (5/12)		C5-C7	58.3% (7/12)	60% (6/10)
C6-C7	8.3% (1/12)	25% (3/12)		C6-T1	8.3% (1/12)	0% (0/10)
C7-T1	0% (0/12)	8.3% (1/12)
Operative time (min)	64.76±31.16	87.33±21.51	0.163		118.25±15.66	111.25±31.67	0.610
EBL	16.25±12.08	15.00±13.82	0.816		19.17±12.76	15.00±7.5	0.395

ASA indicates American Society of Anesthesiologists Physical Status; BMI, body mass index; CCI, Charlson comorbidity index; EBL, estimated blood loss.

Of patients who had a single-level fusion, 12 were randomized to PLATE and 12 to CAGE, with a mean age of 57.25±10.56 and 57.18±12.97 years, respectively (P=0.989). Twelve 2-level patients were randomized into PLATE and 10 into CAGE, with an average age of 62.08±9.17 and 53.50±10.82 years, respectively (P=0.058). There were no differences in baseline demographics, comorbidities, operative levels, operative time, or estimated blood loss between single-level or 2-level PLATE and CAGE groups.

Combined Single-Level and 2-Level Cohort

Patient-Reported Outcomes

There were no differences in baseline PROs between PLATE and CAGE cohorts across general health, disability, and pain metrics, nor were there differences in magnitude of improvement at 6 weeks, 6 months, or 1 year (Fig. 2). Evaluation of swallowing dysfunction through the SWAL-QOL assessment revealed similar baseline function between cohorts; however, PLATE patients transiently reported significantly worse swallow function 6 weeks after surgery (74.21±10.44 vs. 88.81±10.14, P=0.004). There were no differences in swallowing function between the PLATE and CAGE groups at 6 months or 1 year postoperatively.

FIGURE 2: Results from patient-reported outcome questionnaires for single-level and 2-level PLATE and CAGE at preoperative and postoperative time points. NDI indicates neck disability index; SWAL-QOL, swallowing quality of life questionnaire; VAS, visual analog scale; VR12, Veteran’s RAND 12-Item Health Survey.

Radiographic Outcomes

At 1 year, 20 PLATE (83.3%) and 20 CAGE (90.9%) patients presented for radiographic evaluation, corresponding to a total of 60 operative levels and a total loss to follow-up of 13% (Table 2). At 1 year, arthrodesis was observed in 90% (18/20) of PLATE and 90% (18/20) of CAGE patients (P=0.999). There were no differences between groups with regard to magnitude of correction of segmental disc heights, segmental lordosis, or sagittal cervical alignment, nor was there a significant decay in these parameters seen in either the CAGE or PLATE group over the study period (Appendix A, Supplemental Digital Content 1, https://links.lww.com/CLINSPINE/A222). Between 6 weeks and 1 year postoperatively, we observed minor amounts of disc height correction loss without differences between groups with regard to the anterior (P=0.348), middle (P=0.763), or posterior (P=0.664) disc spaces of the fusion segment. In addition, there were no differences in the magnitude of disc height degeneration at adjacent levels. Minor interbody device subsidence was observed across the study with no differences between PLATE and CAGE cohorts at the superior (0.69±0.64 vs. 0.86±0.44 mm, P=0.637) and inferior endplates (0.55±0.59 vs. 0.95±0.65 mm, P=0.065). Categorical subsidence >2 mm was observed in no patients in the PLATE group and 2 patients (10%) in the CAGE cohort.

TABLE 2 - Postoperative Radiographic Changes in Indexed Disc Heights, Subsidence, and Sagittal Cervical Alignment Measured From 6 Weeks to 1 Year Postoperatively for Combined Single-Level and 2-Level PLATE and CAGE

Variable	All Patients	PLATE	CAGE	P
Disc heights (mm)
Operative segment (n=60)
Anterior disc space	−0.80±1.68	−0.48±2.35	0.12±1.10	0.348
Middle disc space	−1.33±2.00	−0.51±1.72	−0.35±1.41	0.763
Posterior disc space	−0.77±1.72	−0.38±1.41	−0.21±0.94	0.664
Cephalad adjacent segment
Anterior disc space	0.34±0.78	0.39±0.85	0.27±0.74	0.824
Middle disc space	0.24±0.52	0.27±0.62	0.20±0.41	0.929
Posterior disc space	0.31±0.87	0.29±1.05	0.35±0.62	0.999
Caudal adjacent segment
Anterior disc space	−0.75±2.48	−1.28±2.99	−0.03±1.42	0.934
Middle disc space	−1.08±1.95	−1.22±2.10	−0.90±1.84	0.620
Posterior disc space	−0.93±1.22	−1.05±1.43	−0.76±0.91	0.869
Subsidence (mm)
Superior endplate (n=60)	0.76±0.56	0.69±0.64	0.86±0.44	0.637
Inferior endplate (n=60)	0.72±0.64	0.55±0.59	0.95±0.65	0.065
Sagittal cervical alignment
FSL	0.43±3.04	−0.24±3.23	1.25±2.63	0.058
Cervical lordosis	0.01±5.91	−1.01±5.19	0.95±6.55	0.169
C2-C7 SVA (mm)	1.82±2.85	1.88±3.38	1.77±2.33	0.898
T1 slope	0.47±4.94	0.08±3.59	0.88±6.25	0.580

Negative values denote loss of correction.

FSL indicates fusion segment lordosis; SVA, sagittal vertical axis.

Single-Level ACDF

Patient-Reported Outcomes

Baseline PROs were similar between single-level PLATE and CAGE across all metrics with the exception of worse NDI (56.00±13.62 vs. 39.25±13.44, P=0.038) among PLATE patients (Fig. 3). After surgery, both cohorts of patients reported similar disability and pain scores at 6 weeks, 6 months, and 1 year postoperatively, although PLATE patients reported worse scores on the VR12 physical health component at 6 months (26.27±5.35 vs. 38.38±9.69, P=0.033). Despite the difference in baseline NDI, there was no significant difference detected between PLATE and CAGE in the magnitude of improvement at 1 year (−30.29±23.02 vs. −18.20±12.45, P=0.181). When evaluating for swallow function through the SWAL-QOL survey, PLATE patients reported greater dysfunction than CAGE patients at 6 weeks (71.31±14.09 vs. 87.86±11.13, P=0.050) and 6 months postoperatively (80.49±9.23 vs. 92.05±7.89, P=0.042), although there was no significant difference at 1 year follow-up (84.66±11.27 vs. 85.78±14.21, P=0.783).

FIGURE 3: Results from patient-reported outcome questionnaires for single-level PLATE and CAGE at preoperative and postoperative time points. NDI indicates neck disability index; SWAL-QOL, swallowing quality of life questionnaire; VAS, visual analog scale; VR12, Veteran’s RAND 12-Item Health Survey.

Radiographic Outcomes

At 1 year, 10 single-level PLATE (83.3%) and 10 CAGE (83.3%) patients presented for radiographic evaluation, for a total loss to follow-up of 17% (Table 3). At 1 year, arthrodesis was observed in 90% (9/10) of PLATE patients and 100% (10/10) of CAGE patients (P=0.305). There were no differences between groups regarding magnitude of correction of segmental disc heights, segmental lordosis, or sagittal cervical alignment, nor was there a significant decay in these parameters seen in either the CAGE or PLATE group over the study period (Appendix B, Supplemental Digital Content 1, https://links.lww.com/CLINSPINE/A222). Between 6 weeks and 1 year postoperatively, we observed minor amounts of disc height correction loss without difference between groups with regard to the anterior (P=0.859), middle (P=0.450), or posterior disc spaces (P=0.859). In addition, measurements of interbody subsidence into the superior (0.87±0.66 vs. 0.85±0.63 mm, P=0.754) or inferior endplates (0.50±0.57 vs. 0.96±0.67 mm, P=0.189) were similar between cohorts. Categorical subsidence >2 mm was only recorded in 1 patient (10%) in the CAGE group. Similarly, there was no difference between groups in radiographic disc degeneration at the cephalad or caudal adjacent segments.

TABLE 3 - Postoperative Radiographic Changes in Indexed Disc Heights, Subsidence, and Sagittal Cervical Alignment Measured From 6 Weeks to 1 Year Postoperatively for Single-Level PLATE and CAGE

Variable	All Patients	PLATE	CAGE	P
Disc space heights (mm)
Operative segment
Anterior	−0.66±1.32	−0.80±1.09	−0.47±1.62	0.859
Middle	−0.62±1.86	−0.96±1.92	−0.20±1.82	0.450
Posterior	−1.19±1.41	−1.34±1.30	−1.01±1.60	0.859
Cephalad adjacent segment
Anterior	−0.07±0.87	−0.18±0.89	0.06±0.89	0.268
Middle	0.17±0.68	0.20±0.85	0.14±0.49	0.630
Posterior	0.50±1.10	0.74±0.75	0.19±1.42	0.214
Caudal adjacent segment
Anterior	−0.58±2.00	−1.02±2.49	−0.02±1.08	0.594
Middle	−0.91±1.55	−1.16±1.86	−0.59±1.09	0.594
Posterior	−0.67±1.21	−0.95±1.53	−0.32±0.56	0.534
Subsidence (mm)
Superior endplate	0.86±0.63	0.87±0.66	0.85±0.63	0.754
Inferior endplate	0.70±0.64	0.50±0.57	0.96±0.67	0.189
Sagittal cervical alignment
FSL	0.46±3.24	−0.32±2.86	1.12±3.51	0.262
Cervical lordosis	0.60±5.82	−0.70±5.58	1.51±6.10	0.282
C2-C7 SVA (mm)	1.63±2.69	1.08±2.62	2.18±2.82	0.401
T1 slope	−1.58±4.60	−2.17±3.70	−0.98±5.66	0.522

Negative values denote loss of correction.

FSL indicates fusion segment lordosis; SVA, sagittal vertical axis.

Complications and Reoperation

No significant adverse events occurred during the perioperative period. In the single-level PLATE cohort, 1 patient-reported recurrent radiculopathy at the fused segment 8 months postoperatively. Another experienced new onset radiculopathy at an adjacent level at 5 months, and X-ray at the 6-month follow-up demonstrated screw fracture, although the fusion segment appeared well healed. Both patients were successfully treated conservatively and did not require revision surgery. A third patient presented with recurrent radiculopathy at the operated level, and imaging revealed screw fracture with pseudarthrosis for which posterior fusion and decompression was performed 14 months after the index procedure. In the CAGE cohort, 1 patient experienced recurrent radicular symptoms 3 months postoperatively after injury sustained in a motor vehicle collision. Two more patients had new onset radiculopathy at adjacent levels 4 and 6 months postoperatively. All patients were managed conservatively. No patients developed clinical or radiographic evidence of pseudarthrosis, adjacent segment degeneration leading to ASD, and there were no reoperations.

Two-Level ACDF

Patient-Reported Outcomes

There were no differences in baseline PROs between 2-level PLATE and CAGE cohorts (Fig. 4). At 6 weeks postoperative, patients reported similar improvement in all PRO metrics, with the exception of NDI, for which PLATE patients reported greater improvement than that seen in the CAGE group, which transiently worsened compared with preoperative state (22.00±13.57 vs. 52.00±25.58 mm, P=0.037). Similarly, patients in the PLATE cohort reported better outcomes at 6 months on NDI (12.00±8.37 vs. 39.33±21.34, P=0.017) and VAS neck (0.96±0.60 vs. 5.47±3.03, P=0.010). However, both groups reported similar scores by 1 year follow-up. Swallow function assessment demonstrated worse SWAL-QOL scores in PLATE patients than CAGE patients at 6 weeks postoperatively (76.54±7.34 vs. 91.34±8.22, P=0.038), although patients reported similar swallow function outcomes at all other time points evaluated.

FIGURE 4: Results from patient-reported outcome questionnaires for 2-level PLATE and CAGE at preoperative and postoperative time points. NDI indicates neck disability index; SWAL-QOL, swallowing quality of life questionnaire; VAS, visual analog scale; VR12, Veteran’s RAND 12-Item Health Survey.

Radiographic Outcomes

Ten 2-level PLATE (83.3%) and 10 CAGE (100%) patients presented for radiographic evaluation at 1 year follow-up, for a total loss to follow-up of 9% (Table 4). At 1 year, arthrodesis was observed in 90% (9/10) of PLATE patients and 80% (8/10) CAGE patients (P=0.531). There were no differences between groups regarding magnitude of correction of segmental disc heights, segmental lordosis, or sagittal cervical alignment, nor was there a significant decay in these parameters seen in either the CAGE or PLATE group over the study period (Appendix C, Supplemental Digital Content 1, https://links.lww.com/CLINSPINE/A222). Between 6 weeks and 1 year postoperatively, we observed minor amounts of disc height correction loss without difference between groups with regard to the anterior (P=0.715), middle (P=0.999), and posterior (P=0.647) disc spaces of the cephalad operative fusion segment, as well as at the anterior (P=0.286), middle (P=0.831), and posterior (P=0.201) spaces of the caudal operative fusion segment. In addition, there were no differences between the CAGE and PLATE groups with regard to radiographic disc degeneration at adjacent levels. Minor amounts of interbody device subsidence were seen across the study without differences between the CAGE and PLATE groups at the cephalad level superior (0.75±0.62 vs. 0.95±0.09 mm, P=0.999) and inferior endplates (0.85±0.74 vs. 1.38±0.54 mm, P=0.285), and at the caudal level superior (0.32±0.54 vs. 0.81±0.22 mm, P=0.093) and inferior endplates (0.35±0.43 vs. 0.59±0.57 mm, P=0.444). Categorical subsidence >2 mm was seen in 1 patient (10%) in the 2-level CAGE group.

TABLE 4 - Postoperative Radiographic Changes in Indexed Disc Heights and Sagittal Cervical Alignment Measured From 6 Weeks to 1 Year Postoperatively for 2-Level PLATE and CAGE

Variable	All Patients	PLATE	CAGE	P
Disc space heights (mm)
Cephalad operative segment
Anterior	−0.87±2.23	−0.94±2.35	−0.79±2.34	0.715
Middle	−1.36±2.09	−1.24±2.24	−1.50±2.13	0.999
Posterior	−0.49±1.85	−0.40±2.46	−0.61±0.97	0.647
Caudal operative segment
Anterior	0.30±1.45	0.66±1.76	−0.24±0.75	0.286
Middle	−1.55±1.45	−1.35±1.16	−1.85±1.96	0.831
Posterior	−0.31±1.54	0.22±1.39	−1.10±1.59	0.201
Cephalad adjacent segment
Anterior	0.07±1.20	−0.01±1.48	0.18±0.91	0.999
Middle	0.14±0.56	0.27±0.76	−0.01±0.12	0.999
Posterior	0.42±0.99	0.23±1.32	0.65±0.42	0.855
Caudal adjacent segment
Anterior	−1.15±2.43	−1.75±3.04	−0.45±1.44	0.715
Middle	−1.26±1.77	−1.07±1.74	−1.48±1.99	0.855
Posterior	−0.88±1.10	−0.78±1.20	−1.01±1.11	0.999
Subsidence (mm)
Cephalad superior endplate	0.83±0.48	0.75±0.62	0.95±0.09	0.999
Cephalad inferior endplate	1.06±0.69	0.85±0.74	1.38±0.54	0.285
Caudal superior endplate	0.55±0.48	0.32±0.54	0.81±0.22	0.093
Caudal inferior endplate	0.46±0.49	0.35±0.43	0.59±0.57	0.444
Sagittal cervical alignment
Cephalad FSL	0.74±2.74	0.40±3.30	1.17±1.98	0.534
Caudal FSL	0.63±2.37	0.49±2.62	0.82±2.22	0.950
Cervical lordosis	−0.83±6.20	−1.31±5.18	−0.16±8.02	0.465
C2-C7 SVA (mm)	2.13±3.20	3.16±4.34	1.10±1.24	0.465
T1 slope	2.69±4.45	2.33±1.60	3.12±6.78	0.584

Negative values denote loss of correction.

FSL indicates fusion segment lordosis; SVA, sagittal vertical axis.

Complications and Reoperation

There were no significant perioperative adverse events in either group of 2-level patients. One patient in the PLATE cohort developed new onset disease at adjacent segments cephalad to the operated levels which presented as new onset radiculopathy 2 years postoperatively. However, there was evidence of subclinical disease on preoperative radiographs which progressed after surgery. The patient subsequently experienced symptomatic improvement with conservative therapy. A second patient presented with recurrent radicular symptoms at 1 year and was lost to follow-up after a brief period of conservative management. In the CAGE group, 1 patient experienced slowly improving persistent radiculopathy but has been successfully treated conservatively up until the most recent follow-up. A second patient had an incidental finding of aseptic screw loosening and interbody subsidence leading to C5 vertebral body fracture at 18 months (Fig. 5). Postoperatively, the patient had continued to experience minor symptoms, though significantly improved compared with her preoperative state. At the time of this publication, the patient had not undergone reoperation, although revision surgery with posterior fusion was being considered. No patients developed clinical or radiographic evidence of pseudarthrosis, adjacent segment degeneration leading to ASD, or required reoperation.

FIGURE 5: A, Demonstrates a 2-level CAGE at C4-C6 at 6 weeks follow-up. B, The same patient experienced screw pullout and fracture of the C5 vertebral body 18 months postoperatively. NDI indicates neck disability index; VAS, visual analog scale; VR12, Veteran’s RAND 12-Item Health Survey.

DISCUSSION

Benefits of traditional anterior cervical plating include enhanced immediate structural support and higher arthrodesis rates. However, literature also suggests that the traditional plates used in ACDF increase the risk for complications, especially when multiple levels are being fused.^7–9 For this reason, there has been increased interest in low or zero-profile stand-alone interbody cages which theoretically provide the biomechanical stability conferred by plating while mitigating the complications associated with anterior instrumentation. In order to investigate differences in postoperative clinical and radiographic outcomes between these constructs, this study prospectively enrolled patients undergoing single-level and 2-level ACDF, randomizing them to either an interbody device with the addition of a static anterior plate, or with an integrated stand-alone interbody spacer. For single-level ACDF, our results demonstrate comparable symptomatic improvement and structural integrity between constructs, although anterior plating led to greater early (but transient) swallow dysfunction. For 2-level ACDF, patients experienced similarly greater early transient swallowing dysfunction, but also had greater early improvement in neck pain and neck disability compared with those treated with integrated interbody devices. These PRO differences were not seen at 1 year. In addition, we were not able to detect differences between the CAGE and PLATE groups with regard to arthrodesis rate, subsidence rate, or segmental and global alignment parameters.

This study has several limitations which must be taken into consideration when interpreting the results. First is the study’s small sample size, which limits the ability to confirm statistical equivalence between groups (ie, increases the likelihood of type II error). The reason for this small sample size is that the study was stopped early because of several reasons, including the costs of continuing the trial and the finding of several significant differences between groups. The senior author subsequently changed practice and no longer performs 2-level stand-alone procedures because of concern over worse early neck pain and disability, potentially attributed to worse early biomechanical stability. Given the small sample size, we could not confirm that the treatments are truly equivalent with regard to longer term outcomes such as ASD, arthrodesis, etc. Limited sample size may also have contributed to our finding of minimal change from preoperative to 1 year postoperative VR12 physical component scores between groups, although there were no differences between groups. Additional limitations include loss to follow-up of several patients with regard to radiographic data, and that the full data set was limited to 1 year after surgery. Furthermore, our evaluation of postoperative dysphagia was limited to patient-reported swallow dysfunction. While the literature supports the use of the SWAL-QOL and has validated it as a tool for the quantification of dysphagia severity, it has its limitations. A past study has shown that the SWAL-QOL assessment is most sensitive for the detection of deficits in the oral phase of swallowing, although detection of deficits in pharyngeal transit is less reliable. Thus, a valuable direction for future comparative analysis could include correlating dysphagia on PROs to clinical swallowing assessments following ACDF with PLATE and CAGE constructs. Despite these limitations, this prospective randomized trial adds significant findings to the existing literature on a topic for which there is no consensus.

Given that successful arthrodesis following ACDF depends largely on rigid fixation, the biomechanical stability of the final construct is crucial. A biomechanical study by Scholz et al²⁹ evaluating the stability of single-level ACDF with a locking plate to integrated spacers demonstrated similar stability in all tested ranges of motion. The same authors later performed a similar study investigating the constructs for 2-level and 3-level ACDF. While they observed decreased segmental stability as the number of instrumented levels increased regardless of the configuration, the locking plate and cage construct demonstrated superior rigidity in all testing modes than anchored, integrated devices.³⁰ However, the data on how this may translate to the clinical setting with regards to fusion, patient outcomes, and structural integrity is highly variable. A meta-analysis by Fraser and colleagues reporting the pooled outcomes of 21 studies including a total of 2682 patients found that the use of an anterior plate increases the rate of successful arthrodesis, regardless of the motion segments fused.³¹ In contrast, several retrospective studies have reported similar fusion rates with stand-alone cages for single-level and multilevel ACDF, leaving a lack of unanimity regarding the optimal construct.^9,32

Our findings corroborate the current literature supporting that patients may experience similar clinical outcomes with an integrated interbody device following single-level ACDF. Radiographic evaluation from the 6-week postoperative follow-up to the final 1 year time point demonstrated similar maintenance of fusion segment heights, adjacent segment heights, and sagittal cervical alignment. PRO data at all time points demonstrated similar positive outcomes between groups. These observations in conjunction with the comparable arthrodesis rates, magnitude of implant subsidence, segmental and global alignment suggest that stand-alone devices may provide enough stability in the single-level setting to achieve successful clinical and radiographic outcomes. This may not be true in the 2-level setting where we found that CAGE patients experienced a lower magnitude of symptomatic reduction on several PRO metrics at 6 weeks and 6 months follow-up. Seeing as CAGE patients experienced similar fusion rates and comparable postoperative degenerative changes, it is possible that any differences in symptomatic improvement disappeared upon successful arthrodesis and that the early differences may be attributed to improved early rigidity with PLATE. However, given the low power of the study, it is also possible that we simply did not detect persistent differences at one year. Although we evaluated symptomatic and radiographic changes, dynamic range of motion differences were not assessed between the CAGE and PLATE cohorts. Therefore, future studies could further investigate the relationship between postoperative range of motion, symptoms, and successful arthrodesis. Moreover, it is worth noting that there was a trend toward greater subclinical interbody subsidence into the superior endplate of the caudal fusion segment, making it difficult to draw definitive conclusions until longer-term follow-up has been obtained.

The manipulation of the soft-tissue structures in the anterior neck during ACDF can lead to intraoperative local tissue damage followed by postoperative inflammation, edema, and hyperemia, all of which may contribute to the development of postoperative dysphagia.^15,16 Stretch neuropraxia leading to neuromuscular discoordination is likely another contributing factor. However, swallowing dysfunction is usually transient, and most patients typically experience symptomatic resolution after 3 months, although long-term dysphagia is possible.^33–35 Past studies have demonstrated that the use of zero-profile integrated interbody cages can decrease the incidence of postoperative dysphagia when compared with plate and cage constructs,^36,37 while others have demonstrated that plate thickness is also a significant contributing factor.³⁸ Similarly, we found that single-level PLATE resulted in significantly worse outcomes through the SWAL-QOL assessment tool, although the effect was temporary and resolved at 1 year follow-up. In contrast, 2-level PLATE patients reported worse swallow function at several postoperative time points, although the difference was only significant at 6-week follow-up. Thus, our findings corroborate past studies which have reported that multilevel ACDF increases the risk for transient postoperative dysphagia,^39,40 but show that the additional instrumentation associated with anterior plating may further increase the risk.

The findings of this study are challenging to translate to clinical practice, since there were both advantages and disadvantages seen with each treatment group. The PLATE construct leads to worse early swallow dysfunction likely as a product of increased anterior soft-tissue manipulation and mechanical effect. However, the supplemental rigid fixation conferred by the addition of an anterior static plate may improve early stability at the fused segments, as suggested by past biomechanical studies, which may translate to greater neck pain and disability in the early postoperative period. Given that swallow function, neck pain, and disability were equivalent between single-level and 2-level PLATE and CAGE groups 1 year postoperatively, surgeons should weigh the advantages and disadvantages of both constructs at their own discretion, taking patient preferences during preoperative counseling into consideration.

CONCLUSION

At 1 year follow-up, we did not detect differences in construct integrity, arthrodesis rates, or cervical alignment for 1-level and 2-level ACDF procedures comparing PLATE and CAGE configurations. However, patients treated with a traditional anterior plate construct for 1-level or 2-level ACDF procedures experienced significantly worse early postoperative dysphagia which was transient. This was counterbalanced in the 2-level cohort by better early neck pain and neck disability PROs in the PLATE group, which possibly indicates better immediate clinical biomechanical stability. Since PLATE and CAGE produced similar outcomes at 1 year, surgeons should weight the early postoperative risks and benefits when choosing between constructs. Further large-scale studies are warranted to further investigate these differences.