Abstract
-
Objective
Adjacent segment disease (ASD) is a recognized long-term complication after anterior cervical discectomy and fusion (ACDF). Cervical disc arthroplasty (CDA) was introduced as a motion-preserving alternative, thought to reduce stress at adjacent levels and lower ASD risk. This study aimed to compare the risk of ASD requiring reoperation after ACDF and CDA and identify factors associated with ASD.
-
Methods
This retrospective cohort study used prospectively collected data from the Swedish Spine Registry (Swespine) and included adult patients undergoing anterior cervical surgery for degenerative cervical disease between 2006 and 2026. ASD was defined as a secondary operation at an adjacent level ≥12 months after the index procedure. Cumulative incidence was estimated using Kaplan-Meier analysis, and associations with ASD were evaluated using multivariable Cox proportional hazards regression. Radiological or conservatively treated ASD were not assessed.
-
Results
A total of 9,338 patients were included (ACDF, n=9,005; CDA, n=333). During 45,666 person-years follow-up, 509 ASD events occurred (incidence rate 11.1 per 1,000 person-years). Cumulative incidence was 4.9% at 5 years and 10.4% at 10 years. CDA was associated with lower ASD hazard than ACDF (hazard ratio [HR], 0.61) though this association was not significant after propensity score matching (HR, 0.88). Multilevel procedures (HR, 0.61), increasing age (HR, 0.96 per year), and radiculopathy were associated with lower risk, while myelopathy or stenosis was associated with higher risk (HR, 1.22).
-
Conclusion
The long-term risk of ASD requiring reoperation was lower than previously reported. After adjustment, CDA was associated with lower ASD risk compared to ACDF, although the absolute difference was small, and the clinical significance remains uncertain.
-
Keywords: Adjacent segment disease, Anterior cervical discectomy and fusion, Total disc replacement, Cervical disc arthroplasty, Intervertebral disc degeneration
INTRODUCTION
Anterior cervical discectomy and fusion (ACDF) is one of the most commonly used surgical procedures for cervical disc diseases (CDD) causing radiculopathy or myelopathy [
1]. ACDF is an anterior approach, involving removal of the damaged disc and fusion of the operated segment [
2]. ACDF is a well-established surgical procedure for CDD, however, long-term complications such as adjacent segment disease (ASD) have been described in studies [
1,
3]. Loss of range of motion at the index level after fusion is thought to shift the load to the adjacent levels, which may result in ASD [
4,
5]. This is however, complicated by the age-related natural degeneration [
6,
7].
The definition of ASD varies across studies. Some studies define ASD as radiological with degenerative changes at levels adjacent to the index operation, while others define it as clinical ASD, when radiological findings correspond with the development of new clinical symptoms [
3].
Current rates vary across studies due to variations in definition and follow-up time, which limit the understanding of true risk for developing ASD [
8,
9]. The reported ASD incidence over a 10-year period, following ACDF, ranges between 16% and 36% for symptomatic ASD and with up to 21% of patients requiring reoperation [
3,
10,
11]. This variation highlights the challenge of estimating the true clinical ASD burden.
Cervical disc arthroplasty (CDA) was introduced as a motion-preserving alternative to ACDF, replacing the degenerated disc with an artificial prosthesis and thereby maintaining segmental motion [
12]. Thus, reducing biomechanical stress at adjacent levels and potentially decreasing the risk of ASD. Randomized trials and meta-analyses have compared ACDF and CDA regarding risk of developing ASD and have reported mixed results. Some studies suggest that CDA leads to fewer reoperations due to ASD, while other studies suggest that CDA has no superiority in reducing the risk of ASD compared to ACDF [
13,
14]. However, studies with longer follow-up remain scarce.
There remains a need for population-based data with more than 10-year follow-up to provide more accurate estimates of ASD risk after ACDF and CDA. In this nationwide study using data from the Swedish Spine Registry (Swespine) [
15], we aim to estimate and compare the long-term risk of ASD after ACDF versus CDA. Secondary aims were to describe the incidence of ASD over time and to explore factors associated with ASD risk.
MATERIALS AND METHODS
1. Study Design and Data Source
This is a retrospective nationwide observational cohort study including adult patients who underwent cervical spine surgery between 2006 and 2026. Data were obtained from the Swespine registry [
16-
18]. Data extraction occurred in February 2026. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [
19].
Swespine is a nationwide quality registry that prospectively collects information on spinal procedures performed in Sweden [
16-
18]. During the study period, the registry included more than 95% of spine surgery units nationwide and captured approximately 75%–85% of eligible procedures [
20]. The registry contains patient and surgical characteristics, implant data, reoperations, and standardized patient-reported outcomes during routine follow-up [
20-
22].
The study population consisted of adult patients (≥18 years) registered in Swespine who underwent cervical spine surgery for degenerative cervical disease during the study period. The index procedure was defined as the first recorded cervical operation for each patient within the registry, and only one index procedure per patient was included.
Patients were excluded if surgery was performed for nondegenerative indications such as trauma, tumor, infection, or inflammatory disease. Patients with incomplete follow-up data to determine time at risk were also excluded.
3. Surgical Groups
Procedures were classified based on surgeon-reported procedure codes in Swespine. The analyses were restricted to anterior cervical procedures and focused on comparisons between ACDF and CDA. Single- and multilevel procedures were defined according to the reported number of operated cervical levels.
4. Outcome and Follow-up
ASD was operationally defined as a reoperation at a level adjacent to the index cervical procedure for a degenerative indication occurring ≥12 months after the index surgery. Reoperations performed for nondegenerative causes were not classified as ASD. To focus on degenerative ASD rather than early postoperative complications, only ASD events occurring at least 12 months after the index procedure were included in the analyses. For patients with multiple reoperations, only the first qualifying adjacent-level degenerative reoperation was counted as an ASD event.
Patients were followed from the index procedure until the occurrence of ASD. Patients without any reoperation were censored at the last available follow-up in the registry. For patients with reoperations that did not meet the ASD criteria, follow-up was censored at the latest available date in the registry, defined as the later of the last recorded reoperation date or the last follow-up date. Since Swespine does not record date of death, mortality could not be accounted for, and patients were censored at their last available registry contact.
5. Statistical Analysis
Missing values were not imputed. Analyses involving variables with missing data were conducted using complete cases, and the number of patients included in each analysis is reported. Descriptive statistics were presented as means with standard deviations for continuous variables and counts with percentages for categorical variables. Follow-up time was calculated from the date of the index procedure. However, only qualifying adjacentlevel reoperations occurring ≥12 months after the index procedure were classified as ASD events. Incidence rates of ASD were calculated from the index surgery as events per 1,000 person-years, with 95% confidence intervals (CIs) estimated using Poisson methods. Cumulative incidence at 5 and 10 years was estimated using Kaplan-Meier analysis and presented as one minus the survival function. Group comparisons were performed using the log-rank test on complete follow-up.
Associations with ASD were evaluated using Cox proportional hazards models. The primary exposure was surgical technique (CDA vs. ACDF), with ACDF as the reference category. Covariate selection for the multivariable model followed a purposeful variable selection approach as described by Hosmer et al. [
23] and detailed by Bursac et al. [
24] Age, sex, and procedure extent (single-vs. multilevel) were included as fixed variables regardless of statistical significance based on clinical relevance. First, candidate variables (body mass index [BMI], smoking status, and preoperative diagnosis) were screened using univariable Cox regression analyses with a p<0.25 threshold for inclusion. Second, variables meeting this threshold were retained as independent predictors if p<0.1, while those with p≥0.1 but <0.25 were assessed as potential confounders and retained if they changed any remaining coefficient by >20%. Third, previously excluded variables were individually reintroduced and retained as confounders if they altered any coefficient by >20%. Preoperative diagnosis met criteria for retention as an independent predictor. BMI and smoking status did not meet criteria for inclusion as predictors or confounders. The final multivariable model included surgical technique, age, sex, procedure extent, and preoperative diagnosis. The multivariable Cox regression analysis was performed using complete cases with available data for all included covariates. The proportional hazards assumption was assessed using Schoenfeld residuals. A granular classification by number of operated levels (1, 2, or ≥3) is presented in baseline characteristics. However, this classification was not pursued in multivariable analyses, as only 2 CDA patients underwent 3 or more levels, precluding meaningful subgroup comparisons.
As a sensitivity analysis, propensity score matching was performed to further address potential confounding by indication. Patients were matched 2:1 (ACDF:CDA) using nearest neighbor matching on age, procedure extent, and preoperative diagnosis. Kaplan-Meier analysis and a univariable Cox proportional hazards model were used to assess ASD risk in the matched cohort.
Additional analyses of age were performed. Kaplan-Meier curves were stratified by age group at the time of index surgery (18–45, 46–65, and ≥66 years). To further explore the association between age and ASD risk, a Cox proportional hazards model was fitted with age modeled using splines. The model was adjusted for surgical technique, sex, procedure extent, and preoperative diagnosis. Hazard ratios were expressed relative to the median age.
A 2-sided p-value <0.05 was considered statistically significant. All analyses were performed using R ver. 4.5.2 (R Foundation for Statistical Computing, Austria).
6. Ethical Considerations
Patient inclusion in the Swespine registry follows an opt-out model, therefore informed consent is not required. Participation in follow-up questionnaires is voluntary. This study was performed in accordance with the ethical standards of the Declaration of Helsinki. The study was approved by the Swedish Ethical Review Authority (Dnr: 2020-00193 and 2021-04773).
RESULTS
1. Study Population and Baseline Characteristics
A total of 20,707 patients were identified in Swespine during the study period. After exclusion of procedures other than ACDF or CDA (n=11,038) and patients with follow-up less than 12 months (n=331), 9,338 patients were included in the final analyses (
Supplementary Fig. 1). Among these, 9,005 patients underwent ACDF and 333 underwent CDA.
Baseline characteristics are presented in
Table 1. Among patients undergoing anterior cervical procedures, the mean age was 51.9 years in the ACDF group and 45.6 years in the CDA group, and approximately half of the patients in each group were male. Single-level surgery was the most common procedure extent in both groups. Among ACDF patients, 59% underwent single-level, 32% 2-level, and 4% 3 or more levels. Procedure extent was missing for 416 patients (4.6%). Among CDA patients, 74% underwent single-level, 23% 2-level, and less than 1% 3 or more levels. Procedure extent was missing for 6 patients (1.8%). Radiculopathy was the most common surgical indication in both groups. Compared with patients treated with ACDF, patients undergoing CDA were younger, and more frequently underwent single-level procedures.
Among patients at risk in the ACDF and CDA cohort, a total of 509 ASD reoperations occurred during 45,666 person-years of follow-up (
Table 2). The incidence rate of ASD was 11.1 events per 1,000 person-years (95% CI, 10.2–12.2). The cumulative incidence of ASD was 4.9% (95% CI, 4.4%–5.4%) at 5 years and 10.4% (95% CI, 9.4%–11.5%) at 10 years after surgery. Among the 509 patients who developed ASD, the most common secondary surgery was a new ACDF (n=301, 59%), followed by posterior foraminotomy (n=141, 28%) and laminectomy (n=27, 5.3%).
Kaplan-Meier analyses from the index procedure showed a gradual increase in cumulative ASD incidence over time (
Fig. 1A). When comparing surgical techniques, the cumulative incidence curves for ACDF and CDA overlapped throughout follow-up, with no statistically significant difference between groups (logrank p=0.17) (
Fig. 1A). When comparing single-versus multilevel procedures, a separation of curves between the 2 was observed (
Fig. 1B). Single-level surgery was associated with a higher cumulative incidence of ASD compared with multilevel surgery, and the difference between groups was statistically significant (log-rank p<0.001) (
Fig. 1B). When stratified by age group, younger patients showed a higher cumulative incidence of ASD over time compared with older patients (
Fig. 2A). Patients aged 18–45 years had the highest cumulative incidence, followed by those aged 46–65 years, whereas patients aged ≥66 years had the lowest incidence. Detailed baseline characteristics stratified by age group, including surgical technique, procedure extent, and preoperative diagnosis, are presented in
Supplementary Table 1.
The multivariable Cox model included 8,696 patients and 509
ASD events; 642 patients were excluded because of missing covariate data. In multivariable Cox proportional hazards regression analysis, CDA was associated with a significantly lower hazard of ASD compared with ACDF (HR, 0.61; 95% CI, 0.40–0.93; p=0.022) (
Table 3). Multilevel procedures were associated with a significantly lower hazard of ASD compared with single-level procedures (HR, 0.61; 95% CI, 0.50–0.76; p<0.001). Increasing age was associated with a lower hazard of ASD (HR, 0.96 per year; 95% CI, 0.96–0.97; p<0.001). In the exploratory spline-based Cox model, the association between age and ASD risk appeared approximately linear, with no clear deviation (
Fig. 2B). Preoperative diagnosis of myelopathy or stenosis was associated with a higher hazard of ASD compared with radiculopathy (HR, 1.22; 95% CI, 1.01–1.48; p=0.044). Sex was not associated with ASD risk.
In a sensitivity analysis using propensity score matching (2:1, ACDF:CDA) on age, procedure extent, and preoperative diagnosis, 325 CDA patients were matched to 650 ACDF patients (n=975 total). Baseline characteristics were well balanced after matching, with all standardized mean differences below 0.01. Full details are provided in
Supplementary Table 2, which presents baseline characteristics before and after matching. In the matched cohort, 83 ASD events occurred. Kaplan-Meier analysis showed no statistically significant difference between groups (log-rank p=0.62) (
Supplementary Fig. 2). Univariable Cox regression in the matched cohort showed no statistically significant difference in ASD risk between CDA and ACDF (HR, 0.88; 95% CI, 0.54–1.44; p=0.62) (
Supplementary Table 3).
DISCUSSION
ASD is a known long-term complication following anterior cervical surgery [
1,
3]. In this study, we used the Swespine registry to evaluate rates of ASD requiring reoperation, defined as a secondary operation at a level adjacent to the index procedure occurring ≥12 months after the index surgery. We aimed to assess the risk of developing ASD after ACDF versus CDA. Our findings suggest that the cumulative incidence of ASD increased with time, with a low absolute risk in both groups. After adjusting for covariates, CDA was associated with a significantly lower hazard of ASD compared with ACDF, although the absolute risk difference was small. This difference did not reach statistical significance in the propensity score matched sensitivity analysis. Additionally, single-level procedures were associated with higher ASD risk than multilevel surgery, increasing age was associated with lower risk of ASD, and preoperative diagnosis of myelopathy or stenosis was associated with higher risk compared to radiculopathy.
ASD is a complication that gradually increases over time, becoming clinically relevant many years after the index operation [
3]. Thus, studies are often limited by too short follow-up, rarely extending beyond 10 years [
25], which may not reflect the true burden of ASD. To the best of our knowledge, this study provides one of the longest follow-up periods, with up to 18.8 years. Similar to prior studies, the cumulative incidence of ASD was shown to increase steadily over time [
3,
26,
27]. This further strengthens the need for a longer follow-up period to accurately estimate the risk of developing ASD. Hilibrand et al. [
3] originally reported symptomatic ASD after ACDF at an incidence of 2.9% per year, with Kaplan-Meier analysis predicting 26% of patients to develop ASD within 10 years. Additionally, a large single-surgeon survivorship analysis reported a 10-year cumulative incidence of 22% [
27], and a prospective study with more than 10-year follow-up after ACDF surgery reported that up to 21% of patients required a secondary surgery due to ASD [
11]. However, our findings suggest that the incidence of ASD requiring surgery is lower, with a cumulative incidence of 4.9% at 5 years and 10.4% at 10 years. Direct comparison with prior studies is limited, as most existing reports either define ASD narrowly as a new ACDF at the adjacent level, report crude proportions rather than cumulative incidence, or are based on small single-center cohorts. Large registry-based studies using a consistent, broad definition of ASD remain scarce, and further research is needed to establish reliable reference estimates.
ASD requiring reoperation was defined as a secondary operation occurring ≥12 months after the index procedure. By excluding events occurring before the 12-month period, we aimed to only include degenerative ASD rather than early postoperative complications, such as infection, bleeding, and residual symptoms. The requirement of secondary surgery was made to only capture clinically severe cases of ASD.
When comparing ACDF and CDA, Kaplan-Meier curves overlapped, and the unadjusted log-rank test did not reach statistical significance. This is expected, as the unadjusted analysis does not account for the substantial baseline differences between groups, including younger age and higher proportion of radiculopathy in the CDA group. After adjusting for covariates in the multivariable proportional hazards regression analysis, CDA was associated with a significantly lower hazard of ASD compared with ACDF. These findings are similar to several randomized trials and recent meta-analyses reporting lower rates of reoperation and ASD after CDA compared to ACDF [
28,
29]. However, results across studies remain inconsistent [
30]. Despite the relative difference, the absolute risk difference between these techniques remained small, as both groups showed a low overall incidence of ASD. These results should be interpreted with caution since the CDA group was much smaller than the ACDF group, limiting the ability to reliably compare these techniques. Compared to ACDF, CDA is a newer method, still in development [
31], and as implant materials and methods continue to improve, future outcomes may differ. Additionally, CDA is more often used in single-level procedures and younger patients, which could influence rates of ASD and reflect potential selection bias [
32]. To further address this, a propensity score matched sensitivity analysis was performed. No statistically significant difference was detected in the matched cohort, likely reflecting the considerably reduced sample size and statistical power following matching, as only 83 ASD events were available. Although CDA was associated with a significantly lower risk of ASD reoperation in the primary adjusted analysis, the association was attenuated and no longer statistically significant after propensity score matching. Therefore, the apparent protective association of CDA should be interpreted cautiously. Notably, the direction of the association remained consistent with the primary analysis, with CDA showing a numerically lower hazard compared with ACDF. It is also important to consider that while our findings suggest that CDA may be associated with a lower risk of ASD compared with ACDF in a nationwide Swedish setting, the low absolute risk in both groups means that the clinical significance of this difference remains uncertain. With a cumulative incidence of 4.9% at 5 years and 10.4% at 10 years, the overall burden of ASD requiring reoperation was limited regardless of surgical technique. Although CDA was associated with a statistically significant lower hazard after adjustment, the absolute risk reduction is modest, and this difference may not be sufficient to guide surgical decision-making independently of other clinical factors. Other factors, such as patient age and disease characteristics remain important considerations when choosing between ACDF and CDA. These findings are further supported by a Swedish randomized controlled trial (RCT) comparing CDA and ACDF, which found no significant difference in clinical outcomes at 2-year follow-up [
33], consistent with the 10-year results of another Swedish RCT showing no advantage of CDA over fusion regarding adjacent segment pathology [
30]. Nevertheless, as a large population-based registry study, our findings reflect real world outcomes in routine clinical practice. Larger, possibly multinational cohorts may be required to further evaluate the performance of these methods.
There remains a challenge in differentiating between ASD because of the surgical intervention and ASD due to natural degeneration caused by aging. Anterior cervical surgery, particularly fusion procedures has been thought to alter biomechanics, reducing motion at the index level and thus increasing mechanical demands in adjacent segments [
4,
5,
12]. This shift in load to adjacent levels may accelerate or induce degenerative processes. However, degeneration is a naturally occurring process increasing with age and patients undergoing surgery may already show degenerative changes in adjacent levels [
6,
7]. Hence it is difficult to determine whether ASD is accelerated or induced by index surgery or a cause of natural degeneration that would have occurred regardless of surgical intervention.
In the Cox regression analysis, multilevel procedures were associated with a lower risk of ASD compared to single level. This aligns with prior studies reporting lower rates of ASD after multilevel ACDF, which could possibly be explained by fewer remaining segments at risk after multilevel operation [
3]. Additionally, when performing a single-level procedure, there is a risk of preexisting degenerative changes in levels adjacent to the index segment, which could later develop into ASD.
Furthermore, in our cohort, the most common multilevel procedure involved the C5–7 segments. These levels are among the most susceptible to degenerative changes in the cervical spine [
34], and operating on these segments during the index procedure may therefore reduce the number of adjacent segments remaining at risk for ASD. Additionally, other factors such as patient selection or disease distribution may also contribute to this difference. Likewise, older age was associated with lower hazard of ASD. This was also supported by the spline analysis, which showed a similar pattern with decreasing ASD risk at older age, without any clear nonlinear trend. Although it is known that the incidence of radiographic degeneration increases with age, the same cannot be said for reoperation due to symptomatic ASD [
10,
35]. This may be explained by a higher threshold for surgery in older, often more comorbid patients. Thus, younger patients might undergo surgery at higher rates, skewing the demographic burden of ASD. Furthermore, younger patients tend to have higher functional demands and longer cumulative exposure to the risk of developing ASD. In addition, preoperative diagnosis also emerged as a significant predictor for ASD. Patients operated for myelopathy or stenosis exhibited a 22% higher hazard of developing ASD compared to those operated for radiculopathy. To our knowledge, few studies have directly compared ASD risk based on preoperative diagnosis, and direct comparisons are therefore limited. One possible explanation is that myelopathy and cervical stenosis often represent a more generalized degenerative process affecting multiple levels, rather than the focal disc pathology more typical of radiculopathy [
3]. Patients with stenotic disease may therefore have preexisting degenerative changes at adjacent segments at the time of the index procedure, which could later progress to symptomatic ASD [
3,
36]. Moreover, the extensive decompression often required in myelopathic patients might contribute to increased biomechanical stress on neighboring levels, potentially accelerating the degenerative process [
4,
5].
This study has several limitations that must be acknowledged. First, this retrospective study with prospectively collected registry data from Swespine is limited to the data contained therein, and the retrospective design limits the ability to infer causal relationships. Second, lack of radiographic imaging might exclude patients with ASD that do not present with symptoms severe enough to be eligible for reoperation and therefore excluded from the ASD group. Third, individual medical records were not available for review, and the classification of surgical indications relied on surgeon-reported data in the Swespine registry. Fourth, the disparity in number of ACDF procedures compared to CDA limits the ability to reliably compare these methods, and multinational cohort studies may be needed to further observe the performance of these techniques. Fifth, since Swespine does not record the date of death, mortality could not be accounted for in the survival analyses. In studies with long followup, death may act as a competing risk and could potentially bias cumulative incidence estimates, particularly in older age groups. Sixth, because person-time was calculated from the index procedure while ASD events were restricted to those occurring ≥12 months postoperatively, the reported incidence rates should be interpreted within this operational definition. Finally, although Swespine covers more than 95% of spine surgery units nationwide, approximately 75%–85% of all eligible procedures are captured [
20], and a proportion of cases may be missing from the registry. If reporting patterns are influenced by specific patient or surgical characteristics, this could introduce selection bias. However, previous studies on Swedish (Swespine), Norwegian, and Danish spine registries have shown that the nonrespondents did not significantly influence study outcomes [
37-
40]. Furthermore, our findings reflect a nationwide Swedish setting and may not represent populations outside Sweden, which limits external validity [
20].
CONCLUSION
In conclusion, the long-term risk of ASD requiring reoperation after anterior cervical surgery was lower than previously suggested in the literature. However, radiological or conservatively treated ASD were not assessed in this study. In the principal adjusted analysis, CDA was associated with a reduced risk of ASD reoperation, but after propensity score matching, the association weakened and was no longer statistically significant. Therefore, the protective effect of CDA should be interpreted cautiously. Single-level procedures, younger age, and preoperative diagnosis of myelopathy or stenosis were associated with higher ASD risk. These findings suggest that while surgical technique may influence ASD risk, the overall burden of clinically relevant ASD requiring reoperation remains low in a nationwide Swedish setting.
Supplementary Materials
Supplementary Fig. 1.
Flowchart illustrating patient inclusion process in the study cohort. ACDF, anterior cervical discectomy and fusion; CDA, cervical disc arthroplasty; ASD, adjacent segment disease.
Supplementary Fig. 2.
Kaplan-Meier curves illustrating the cumulative incidence of ASD requiring reoperation in the propensity score matched cohort (2:1, ACDF:CDA), stratified by surgical technique. Numbers at risk are shown below the x-axis. Logrank p=0.62. Cumulative incidence of ASD at 5 years: ACDF 5.1%, CDA 2.4%; at 10 years: ACDF 8.7%, CDA 8.5%. ASD, adjacent segment disease; ACDF, anterior cervical discectomy and fusion; CDA, cervical disc arthroplasty.
NOTES
-
Conflict of Interest
The authors declare that the article and its contents were composed in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Adrian Elmi-Terander and Victor Egon Staartjes – although the authors are guest editors of the journal, there was no involvement with the peer review process for this article.
-
Funding/Support
This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
-
Author Contribution
Conceptualization: ZN, AET, VES; Data curation: ZN, VGEH; Formal analysis: ZN, VES; Methodology: ZN, VES; Project administration: ZN, VES; Visualization: ZN; Writing – original draft: ZN; Writing – review & editing: ZN, VGEH, EE, AET, VES.
Fig. 1.(A) Kaplan-Meier curves illustrating the cumulative incidence of ASD after the index procedure, stratified by surgical technique ACDF versus CDA. Numbers at risk are shown below the x-axis. Log-rank test performed on complete follow-up. Cumulative incidence of ASD at 5 years: ACDF 5.0%, CDA 2.4%; at 10 years: ACDF 10.5%, CDA 8.2%. (B) Kaplan-Meier curves illustrating the cumulative incidence of ASD requiring reoperation after the index procedure, stratified by procedure extent (single- vs. multilevel). The analysis included 8,916 patients with available information on the number of operated cervical levels at the index procedure (5,531 single-level, 3,385 multilevel); 422 patients (416 ACDF, 6 CDA) with missing or unknown procedure extent were excluded from this subgroup analysis. Numbers at risk are shown below the x-axis. Log-rank test performed on complete follow-up. Cumulative incidence of ASD at 5 years: single 6.3%, multi 3.1%; at 10 years: single 12.3%, multi 8.3%. ASD, adjacent segment disease; ACDF, anterior cervical discectomy and fusion; CDA, cervical disc arthroplasty.
Fig. 2.(A) Kaplan-Meier curves illustrating the cumulative incidence of ASD requiring reoperation after the index procedure, stratified by age group at the time of surgery. Numbers at risk are shown below the x-axis. Cumulative incidence of ASD at 10 years: 18–45 years 14.1%, 46–65 years 9.3%, ≥66 years 4.5%. (B) Cox proportional hazards models with age modeled using natural cubic splines, showing the association between age at index surgery and hazard of ASD. HR are presented relative to the median age (51 years, dotted vertical line). The shaded area represents 95% CI, and the dashed horizontal line indicates HR=1. The model is adjusted for surgical technique, sex, procedure extent, and preoperative diagnosis. ASD, adjacent segment disease; HR, hazard ratio; CI, confidence interval.
Table 1.Baseline characteristics by procedure group
Table 1.
|
Variable |
ACDF (N = 9,005) |
CDA (N = 333) |
|
Age (yr) |
51.9 ± 10.1 |
45.6 ± 7.9 |
|
BMI (kg/m2) |
27.3 ± 8.4 |
25.7 ± 3.6 |
|
Male sex |
4,257 (47) |
153 (46) |
|
Smoker |
996 (11) |
52 (16) |
|
Procedure extent |
|
|
|
1 Level |
5,284 (59) |
247 (74) |
|
2 Levels |
2,923 (32) |
78 (23) |
|
≥ 3 Levels |
382 (4) |
2 (0.6) |
|
Missing/unknown |
416 (4.6) |
6 (1.8) |
|
Preoperative diagnosis |
|
|
|
Radiculopathy |
5,148 (57) |
245 (74) |
|
Myelopathy or stenosis |
3,714 (41) |
87 (26) |
|
Other or unknown |
98 (1.0) |
1 (0.3) |
|
Missing |
45 (0.5) |
0 (0) |
Table 2.Incidence of ASD requiring reoperation after the first postoperative year in the ACDF and CDA cohort, including cumulative incidence and incidence rates
Table 2.
|
Variable |
ACDF and CDA cohort |
|
Patients at risk |
9,338 |
|
ASD events |
509 |
|
Person-years from index surgery |
45,666 |
|
5-Year cumulative incidence, % (95% CI) |
4.9 (4.4–5.4) |
|
10-Year cumulative incidence, % (95% CI) |
10.4 (9.4–11.5) |
|
Incidence rate per 1,000 person-years (95% CI) |
11.1 (10.2–12.2) |
Table 3.Multivariable Cox proportional hazards regression analysis evaluating associations with ASD in the CDA and ACDF cohort
Table 3.
|
Variable |
HR |
95% CI |
p-value |
|
Surgical technique (CDA vs. ACDF) |
0.61 |
0.40–0.93 |
0.022 |
|
Age (per year) |
0.96 |
0.96–0.97 |
< 0.001 |
|
Male sex (vs. female) |
0.93 |
0.78–1.11 |
0.44 |
|
Multilevel procedure (vs. single) |
0.61 |
0.50–0.76 |
< 0.001 |
|
Myelopathy or stenosis (vs. radiculopathy) |
1.22 |
1.01–1.48 |
0.044 |
REFERENCES
- 1. Matz PG, Holly LT, Groff MW, et al. Indications for anterior cervical decompression for the treatment of cervical degenerative radiculopathy. J Neurosurg Spine 2009;11:174-82.
- 2. Rhee JM, Ju KL. Anterior cervical discectomy and fusion. JBJS Essent Surg Tech 2016;6:e37.
- 3. Hilibrand AS, Carlson GD, Palumbo MA, et al. Radiculopathy and myelopathy at segments adjacent to the site of a previous anterior cervical arthrodesis. J Bone Joint Surg Am 1999;81:519-28.
- 4. Park DH, Ramakrishnan P, Cho TH, et al. Effect of lower two-level anterior cervical fusion on the superior adjacent level. J Neurosurg Spine 2007;7:336-40.
- 5. Eck JC, Humphreys SC, Lim TH, et al. Biomechanical study on the effect of cervical spine fusion on adjacent-level intradiscal pressure and segmental motion. Spine (Phila Pa 1976) 2002;27:2431-4.
- 6. Lundine KM, Davis G, Rogers M, et al. Prevalence of adjacent segment disc degeneration in patients undergoing anterior cervical discectomy and fusion based on pre-operative MRI findings. J Clin Neurosci 2014;21:82-5.
- 7. Tao Y, Galbusera F, Niemeyer F, et al. Radiographic cervical spine degenerative findings: a study on a large population from age 18 to 97 years. Eur Spine J 2021;30:431-43.
- 8. Kong L, Cao J, Wang L, et al. Prevalence of adjacent segment disease following cervical spine surgery: a PRISMA-compliant systematic review and meta-analysis. Medicine (Baltimore) 2016;95:e4171.
- 9. Verma K, Gandhi SD, Maltenfort M, et al. Rate of adjacent segment disease in cervical disc arthroplasty versus singlelevel fusion: meta-analysis of prospective studies. Spine (Phila Pa 1976) 2013;38:2253-7.
- 10. Goffin J, Geusens E, Vantomme N, et al. Long-term followup after interbody fusion of the cervical spine. J Spinal Disord Tech 2004;17:79-85.
- 11. Buttermann GR. Anterior cervical discectomy and fusion outcomes over 10 years: a prospective study. Spine (Phila Pa 1976) 2018;43:207-14.
- 12. Park DK, Lin EL, Phillips FM, et al. Index and adjacent level kinematics after cervical disc replacement and anterior fusion: in vivo quantitative radiographic analysis. Spine (Phila Pa 1976) 2011;36:721-30.
- 13. Goedmakers CMW, de Vries F, Bosscher L, et al. Long-term results of the NECK trial-implanting a disc prosthesis after cervical anterior discectomy cannot prevent adjacent segment disease: five-year clinical follow-up of a double-blinded randomised controlled trial. Spine J 2023;23:350-60.
- 14. Zhang Y, Lv N, He F, et al. Comparison of cervical disc arthroplasty and anterior cervical discectomy and fusion for the treatment of cervical disc degenerative diseases on the basis of more than 60 months of follow-up: a systematic review and meta-analysis. BMC Neurol 2020;20:143.
- 15. Strömqvist B, Fritzell P, Hägg O, et al. Swespine: the Swedish spine register: the 2012 report. Eur Spine J 2013;22:953-74.
- 16. El-Hajj VG, Singh A, Blixt S, et al. Evolution of patient-reported outcome measures, 1, 2, and 5 years after surgery for subaxial cervical spine fractures, a nation-wide registry study. Spine J 2023;23:1182-8.
- 17. Buwaider A, El-Hajj VG, MacDowall A, et al. Machine learning models for predicting dysphonia following anterior cervical discectomy and fusion: a Swedish Registry Study. Spine J 2025;25:419-28.
- 18. El-Hajj VG, Stenimahitis V, Singh A, et al. The effect of concomitant spinal cord injury on postoperative health-related quality of life after traumatic subaxial cervical spine injuries: a nationwide registry study. Arch Phys Med Rehabil 2024;105:1069-75.
- 19. Cuschieri S. The STROBE guidelines. Saudi J Anaesth 2019;13(Suppl 1):S31-4.
- 20. Fritzell P, Hägg O, Gerdhem P, et al. Uppföljning AV Ryggkirurgi Utförd I Sverige ÅR 2019. Swespine 2019.
- 21. El-Hajj VG, Ranganathan S, de Wilde D, et al. A nationwide registry study of surgical and patient-reported outcomes following anterior cervical discectomy and fusion: Part 1 - Synthetic cages versus autologous crista bone grafts. Eur Spine J 2026;35:510-9.
- 22. de Wilde D, El-Hajj VG, Vigren P, et al. A nationwide registry study of surgical and patient-reported outcomes following anterior cervical discectomy and fusion: Part 2 - cage with versus without local bone graft. Acta Neurochir (Wien) 2025;167:319.
- 23. Hosmer DW, Lemeshow S, Sturdivant RX, et al. Applied logistic regression. 3rd ed. Hoboken (NJ): Wiley; 2013.
- 24. Bursac Z, Gauss CH, Williams DK, et al. Purposeful selection of variables in logistic regression. Source Code Biol Med 2008;3:17.
- 25. Xia XP, Chen HL, Cheng HB, et al. Prevalence of adjacent segment degeneration after spine surgery: a systematic review and meta-analysis. Spine (Phila Pa 1976) 2013;38:597-608.
- 26. Wu JC, Liu L, Wen-Cheng H, et al. The incidence of adjacent segment disease requiring surgery after anterior cervical diskectomy and fusion: estimation using an 11-year comprehensive nationwide database in Taiwan. Neurosurgery 2012;70:594-601.
- 27. Lee JC, Lee SH, Peters C, et al. Risk-factor analysis of adjacent-segment pathology requiring surgery following anterior, posterior, fusion, and nonfusion cervical spine operations: survivorship analysis of 1,358 patients. J Bone Joint Surg Am 2014;96:1761-7.
- 28. Zhang Y, Ju J, Wu J, et al. Comparison of cervical disc arthroplasty versus anterior cervical discectomy and fusion for the treatment of single-segment cervical degenerative disc disease with a minimum of 4-year follow-up: a systematic review and meta-analysis of randomized controlled trials. J Orthop Surg Res 2025;20:758.
- 29. Byvaltsev VA, Stepanov IA, Riew DK, et al. Mid-term to longterm outcomes after total cervical disk arthroplasty compared with anterior diskectomy and fusion: a systematic review and meta-analysis of randomized controlled trials. Clin Spine Surg 2020;33:192-200.
- 30. Kontakis M, Marques C, Löfgren H, et al. Artificial disc replacement and adjacent-segment pathology: 10-year outcomes of a randomized trial. J Neurosurg Spine 2021;36:945-53.
- 31. Derman PB, Zigler JE. Cervical disc arthroplasty: rationale and history. Int J Spine Surg 2020;14(s2):S5-13.
- 32. de Leeuw CN, Ryu WHA, Yoo J, et al. Cervical disc arthroplasty versus anterior cervical discectomy and fusion in the treatment of degenerative cervical myelopathy: patient characteristics and surgical outcomes in a national administrative database. Global Spine J 2025;15:3637-47.
- 33. Skeppholm M, Lindgren L, Henriques T, et al. The Discover artificial disc replacement versus fusion in cervical radiculopathy-- a randomized controlled outcome trial with 2-year follow-up. Spine J 2015;15:1284-94.
- 34. Suzuki A, Daubs MD, Hayashi T, et al. Patterns of cervical disc degeneration: analysis of magnetic resonance imaging of over 1000 symptomatic subjects. Global Spine J 2018;8:254-9.
- 35. Shahzad H, Alvarez PM, Pallumeera M, et al. Exploring the incidence and risk factors of reoperation for symptomatic adjacent segment disease following cervical decompression and fusion. N Am Spine Soc J 2023;17:100305.
- 36. Ishihara H, Kanamori M, Kawaguchi Y, et al. Adjacent segment disease after anterior cervical interbody fusion. Spine J 2004;4:624-8.
- 37. Ingebrigtsen T, Aune G, Karlsen ME, et al. Non-respondents do not bias outcome assessment after cervical spine surgery: a multicenter observational study from the Norwegian registry for spine surgery (NORspine). Acta Neurochir (Wien) 2023;165:125-33.
- 38. Højmark K, Støttrup C, Carreon L, et al. Patient-reported outcome measures unbiased by loss of follow-up. Single-center study based on DaneSpine, the Danish spine surgery registry. Eur Spine J 2016;25:282-6.
- 39. Endler P, Ekman P, Hellström F, et al. Minor effect of loss to follow-up on outcome interpretation in the Swedish spine register. Eur Spine J 2020;29:213-20.
- 40. Solberg TK, Sørlie A, Sjaavik K, et al. Would loss to follow-up bias the outcome evaluation of patients operated for degenerative disorders of the lumbar spine? Acta Orthop 2011;82:56-63.
, Victor Gabriel El-Hajj1
