Clinical Outcomes Following Three- and Four-Level Anterior Cervical Discectomy and Fusion: A Systematic Review and Meta-Analysis

Abstract

Study Design/Setting

Systematic review and meta-analysis.

Objectives

Anterior cervical discectomy and fusion (ACDF) is a well-established treatment for degenerative cervical pathology. Although outcomes following 1- and 2-level ACDF are favorable, increasing construct length has been associated with worse outcomes. Despite this, 3- and 4-level ACDF are often grouped as “multilevel” constructs, though emerging evidence suggests 4-level ACDF may carry greater risk. This study systematically reviewed the literature and performed a meta-analysis comparing outcomes between 3- and 4-level ACDF.

Methods

A systematic search of MEDLINE, Embase, and Cochrane databases was performed per PRISMA guidelines. Studies reporting outcomes following three- or four-level ACDF were included. Complex etiology and registry-based studies were excluded. Analysis included patient-reported outcomes, perioperative characteristics, and postoperative fusion and complication rates. Data were synthesized, and random-effects meta-analysis was performed using Cohen’s d for continuous outcomes and log odds ratios for dichotomous outcomes.

Results

Both 3- and 4-level ACDF improved pain, disability, and neurologic function. Meta-analysis showed 4-level ACDF had less improvement in axial neck pain, with similar arm pain improvement. Fusion rates were lower with 4-level ACDF. Perioperative outcomes favored 3-level constructs, with 4-level ACDF associated with longer operative time, greater blood loss, and longer hospital stay. No significant differences were observed in complication or revision rates, though confidence intervals were wide.

Conclusions

Both procedures improved pain and function. However, 4-level ACDF showed less neck pain improvement, lower fusion rates, and greater perioperative morbidity. Meta-analysis reveals clinically meaningful differences that should inform surgical planning and patient counseling.

Keywords

anterior cervical discectomy and fusion clinical outcomes systematic review meta-analysis

Introduction

Anterior cervical discectomy and fusion (ACDF) is a well-established surgical treatment for degenerative cervical spine pathology resulting in radiculopathy and myelopathy.^1-4 By achieving neural decompression and interbody fusion, ACDF improves radicular symptoms, prevents neurologic deterioration associated with cervical myelopathy, and restores spinal stability.^5-7 While outcomes following one- and two-level ACDF are generally favorable, increasing construct length has been associated with higher complication rates, diminished improvement in patient-reported outcomes (PROs), and lower rates of fusion.^8-12

As construct length increases, ACDF is associated with greater operative morbidity, including longer operative time, increased estimated blood loss, higher rates of dysphagia, and increased risk of pseudoarthrosis and revision surgery.^8,9,13-15 Despite this, 3- and 4-level ACDF are often grouped together as “multilevel” constructs in both clinical practice and literature, which implicitly suggests equivalence in outcomes.^16,17 However, emerging evidence suggests that this assumption may be invalid. Recent large administrative database studies have demonstrated higher rates of adverse events, reoperation, and dysphagia following 4-level ACDF compared to shorter constructs, whereas 3-level ACDF does not show such differences.¹⁸

Although a prior meta-analysis has compared 3- and 4-level ACDF in patients with cervical spondylotic myelopathy, a comprehensive systematic review of patient-reported outcomes, fusion success, and perioperative morbidity across the broader spectrum of degenerative cervical pathology has not been performed.¹⁹ As such, the primary objective of this study was to systematically evaluate outcomes following 3- and 4-level ACDF, with secondary aims including meta-analysis of directly comparative cohorts to clarify clinically meaningful difference between these constructs and better define the risks associated with increasing construct length. We hypothesized that 3- and 4-level ACDF would not yield equivalent outcomes and that 4-level ACDF would be associated with inferior patient-reported outcomes, lower fusion rates, and greater perioperative morbidity compared with 3-level ACDF.

Methods

Protocol and Registration

This systematic review and meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The protocol was not prospectively registered in PROSPERO; however, the research question, inclusion and exclusion criteria, search strategy, and outcome measures were defined a priori prior to data extraction.

Search Strategy

Three online databases (PubMed, Ovid Medline, and EMBASE) were queried for articles discussing clinical outcomes following 3- or 4-level anterior cervical discectomy and fusion (ACDF) from database inception to August 2024. The PubMed/MEDLINE Boolean string was: ((“anterior cervical discectomy and fusion”[Title/Abstract] OR “ACDF”[Title/Abstract] OR “cervical fusion”[Title/Abstract] OR “cervical discectomy”[Title/Abstract]) AND (“multilevel”[Title/Abstract] OR “multi-level”[Title/Abstract] OR “three-level”[Title/Abstract] OR “3-level”[Title/Abstract] OR “four-level”[Title/Abstract] OR “4-level”[Title/Abstract]) AND (“outcomes”[Title/Abstract] OR “complications”[Title/Abstract] OR “fusion rate”[Title/Abstract] OR “patient-reported outcomes”[Title/Abstract])). Equivalent strings were translated to Ovid MEDLINE and Embase syntax using Emtree and MeSH mappings. Reference lists of included studies were screened to identify additional eligible articles.

Study Screening

Study screening for the titles, abstracts, and full-text articles was performed by three independent reviewers (R.P., R.S., A.B.) and in accordance with PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines.

Assessment of Study Eligibility

The research question and criteria for eligible studies were established a priori. The inclusion criteria were: English language studies, human studies, level of evidence I to IV, and studies investigating clinical outcomes in patients following 3- or 4- level ACDF. Studies that combined data for 3- or 4-level ACDF with 1- or 2-level surgeries were excluded. Studies that did not report clinical outcomes were excluded. Studies with cohorts that contained revision surgeries, surgery for traumatic injury, and surgery for infectious or oncologic etiologies were excluded. Registry-based studies were excluded due to frequently overlap with one another which creates a high risk of double-counting when pooled with smaller traditional cohort studies. They often are restricted to ICD- and CPT-derived endpoints which limits granularity in patient outcomes. Complication and reoperation rates are also subject to coding-driven misclassification that is heterogeneous with the clinical adjudication used in single- and multi-center cohort studies. For these reasons, including registry-based analyses alongside traditional cohort studies would have introduced methodological heterogeneity not amenable to pooled comparison. Their findings are nevertheless explicitly contextualized in the discussion section.

Data Abstraction & Statistical Analysis

Data was collected by two reviewers and recorded using Microsoft Excel (Version 2019, Microsoft, Redmond, WA, USA). Author information, year of publication, study design, and sample sizes were abstracted. Demographics, clinical outcome scores, complication data, and fusion rates were collected. Surgical construct type, cage material, and graft material were also recorded. Specific clinical outcome measures that were extracted included visual analog scale (VAS) scores, Neck Disability Index (NDI), and Japanese Orthopaedic Association (JOA) Scores. Fusion rate was also reported for each study. Other variables, including hospital length-of-stay (LOS), estimated blood loss (EBL), and operative time were also reported.

Data were extracted and reported as means, ranges, and percentiles, directly from the results of each individual study. Meta-analysis was performed to compare several outcome measures between the 3- and 4-level ACDF cohorts: a) Overall VAS, b) NDI, c) JOA, d), fusion rate, e) LOS, f) EBL, and g) operative time. Studies were included in the meta-analysis if they directly compared 3- and 4-level ACDF, and if data were provided in an appropriate format for statistical analysis. Specifically, continuous variables were required to be presented with a mean, a measure of variability (i.e. standard deviation or standard error), and sample size. Cohorts without proper statistical measures for pooled analysis were excluded from the meta-analysis and included descriptively in the systematic review tables.

For continuous variables, a random-effects, inverse-variance weighted model was used to generate a pooled effect size across studies. A restricted maximum-likelihood (REML) estimator was used to assess heterogeneity variance. Pooled effect size was reported as Cohen’s d value. For categorical variables, a random-effects, inverse-variance model was used to generate pooled effect size across studies. An REML estimator was again used to assess heterogeneity across studies. Pooled effect size was reported as log risk ratio. Egger’s regression test and Funnel plots were used to assess for presence of publication or other systemic bias across studies. Statistical significance was defined as p<0.05. All statistical analyses were completed using IBM SPSS Version 24.0 (IBM Corp).

Results

Study Characteristics

An initial 1047 publications were identified across the electronic database search. A total of 65 cohorts across 49 studies were included in the final quantitative analysis (Figure 1), comprising 27 cohort studies, 21 case series, and 1 randomized control trial (Supplemental Table 1). Of the included studies, 44 were retrospective and 5 were prospective. Publication years ranged from 1997 to 2022. The systematic review included 41 3-level ACDF cohorts and 24 4-level ACDF cohorts. Among these, 6 studies directly comparing 3- and 4-level constructs were eligible for inclusion in meta-analysis.

Figure 1.

Prisma flowchart of study eligibility

Patient Demographics and Pathology

Baseline demographics were broadly comparable between 3- and 4-level cohorts (Table 1). Mean patient age, sex distribution, smoking status, body mass index, and diabetes mellitus prevalence, and mean follow-up were similar between groups.

Table 1.

Baseline Demographic Characteristics of Included Cohorts

Variable	3-Level ACDF	4-Level ACDF
Number of cohorts	41	24
Sample size per cohort	42.0 (10–129)	30.5 (4–130)
Total sample size	1720	733
Mean patient age (years)	56.3 (46.1–65.5) (n=26)	56.2 (48.0–62.0) (n=17)
Male patients (%)	58.2 (36.2–97.4) (n=27)	56.0 (25.0–100.0) (n=17)
Mean follow-up (months)	39.7 (7.6–99.7) (n=28)	30.6 (7.6–66.0) (n=15)
Current smokers (%)	20.7 (4.1–43.3) (n=12)	24.0 (12.0–48.0) (n=8)
Prior smokers (%)	22.7 (6.4–40.0) (n=5)	24.6 (11.5–42.0) (n=5)
Mean BMI (kg/m²)	25.8 (21.1–31.0) (n=6)	27.3 (25.8–30.2) (n=3)
Diabetes mellitus (%)	11.6 (6.8–18.8) (n=7)	11.2 (10.0–12.2) (n=3)

Footnote: Values reported as mean (range). Mean values represent the average of reported cohort means and are descriptive rather than pooled estimates. n indicates number of cohorts reporting the variable.

The distribution of surgical pathology is summarized in Table 2. Across both construct lengths, cohorts included patients with myelopathy, radiculopathy, and mixed degenerative pathology, with mixed presentations making up the majority of cohorts.

Table 2.

Surgical Pathology Distribution

Pathology	3-Level ACDF (n)	4-Level ACDF (n)
Myelopathy only	368 (n=8)	122 (n=3)
Mixed degenerative pathology	654 (n=19)	479 (n=13)
Radiculopathy only	48 (n=1)	15 (n=1)

Surgical Technique and Construct

All studies evaluated anterior cervical discectomy with a fusion construct length of 3 or 4 levels. Hybrid surgery procedures and posterior approaches were excluded. Surgical constructs varied and included cage-half plate constructs (5 cohorts), cage–full plate constructs (20 cohorts), plate-and-screw constructs (2 cohorts), and stand-alone cages (15 cohorts).

Cage materials included PEEK cages (32 cohorts) and titanium cages (4 cohorts). Graft materials consisted of allograft (20 cohorts) and autograft (18 cohorts). Cortico-cancellous (3 cohorts) and cadaveric (2 cohorts) were specified in allograft. Iliac crest (15 cohorts) and local bone graft (4 cohorts) were specified in autograft. Bone morphogenic protein (BMP) was reportedly used in 2 cohorts.

Patient Reported Outcomes

PROs across included cohorts are summarized in Table 3. Both 3- and 4-level ACDF were associated with significant postoperative improvement in pain, disability, and neurologic function. Across cohorts reporting VAS, NAS-neck pain, VAS-arm pain, NDI, and JOA/mJOA scores, mean postoperative values were consistently improved relative to preoperative baselines in both 3-level and 4-level groups. Because the systematic-review summaries in Table 3 reflect averages of cohort-level means rather than pooled patient-level data, inferential comparisons between 3- and 4-level groups were not performed within the table itself; statistical significance for between-group differences in directly comparative studies is reported in the corresponding meta-analyses (Figures 2 and 3).

Table 3.

Patient-Reported Outcomes Across Included Cohorts

Outcome	3-Level ACDF	4-Level ACDF
VAS preoperative	7.5 (6.76–8.2) (n=3)	7.1 (n=1) ^*
VAS postoperative	2.0 (0.95–3.24) (n=3)	3.4 (n=1) ^*
VAS-AP preoperative	6.8 (4.4–8.5) (n=5)	5.7 (1.7–8.6) (n=4)
VAS-AP postoperative	1.5 (0.2–2.5) (n=7)	2.2 (1.5–3.1) (n=4)
VAS-NP preoperative	6.0 (1.6–8.7) (n=8)	6.8 (5.2–8.4) (n=4)
VAS-NP postoperative	2.0 (0.1–3.1) (n=10)	3.1 (2.3–3.9) (n=4)
NDI preoperative	28.4 (19.7–39.1) (n=14)	31.5 (26.5–39.4) (n=5)
NDI postoperative	10.5 (1.9–23.3) (n=16)	14.0 (8.3–22.5) (n=5)
JOA/mJOA preoperative	9.6 (7.3–13.4) (n= 12 )	10.2 (7.5–13.3) (n=5)
JOA/mJOA postoperative	14.2 (13.3–15.7) (n= 12 )	14.1 (12.9–15.9) (n=5)

*Reported JOA/mJOA values represent the average of cohort means and reflect studies that variably used the modified JOA (mJOA, 0–18) and the original JOA scale (JOA, 0-17). Cohort-level values within each individual study did not exceed the maximum of the scale used in that study.

*1 study reported VAS scores out of 20, which was divided by 2 for comparison.

Figure 2.

Forest plot of VAS-NP for three-level ACDF compared with four-level ACDF (derived from meta-analysis of three studies). VAS-NP, visual analog score-neck pain; ACDF, anterior cervical discectomy and fusion

Figure 3.

Forest plot of VAS-AP for three-level ACDF compared with four-level ACDF (derived from meta-analysis of two studies). VAS-AP, visual analog score-arm pain; ACDF, anterior cervical discectomy and fusion

While descriptive data demonstrated higher mean preoperative and postoperative pain and disability scores among 4-level cohorts, these findings were primarily descriptive and reflect cohort-level averages rather than pooled comparative estimates. However, among studies directly comparing 3- and 4-level ACDF, meta-analysis demonstrated significantly less improvement in VAS-NP following 4-level ACDF (Cohen’s d −0.42; 95% CI −0.83 to −0.02; p=0.041) (Figure 2). In contrast, no significant difference in improvement of VAS-AP was observed between 3- and 4-level ACDF on meta-analysis (Cohen’s d −0.03; 95% CI −0.53 to 0.47; p=0.913) (Figure 3).

Fusion Outcomes

Fusion outcomes are summarized in Table 4. Reported fusion rates were slightly lower in the 4-level ACDF cohorts, with similar variability seen in both groups. Meta-analysis of six studies directly comparing 3- and 4-level ACDF demonstrated a significantly lower fusion rate in 4-level constructs (log odds ratio −0.76, 95% CI −1.42 to −0.10; p=0.024) (Figure 4).

Table 4.

Fusion Outcomes

Variable	3-Level ACDF	4-Level ACDF
Fusion rate (%)	88.8 (57.6–100) (n=24)	84.4 (50.0–100) (n=8)
Time to fusion (months)	8.1 (3.8–14.9) (n=5)	7.5 (4.0–13.3) (n=3)

Figure 4.

Forest plot of fusion rate for three-level ACDF compared with four-level ACDF (derived from meta-analysis of six studies). ACDF, anterior cervical discectomy and fusion

Fusion rates stratified by construct type are summarized in Table 5. In 3-level ACDF cohorts, cage–plate constructs demonstrated higher reported fusion rates compared with stand-alone cages. A similar pattern was observed in 4-level ACDF, with higher reported fusion rates among cage–plate constructs. Within construct type, fusion rates were comparable between 3- and 4-level ACDF, although the number of 4-level stand-alone cage cohorts was limited. These findings reflect descriptive trends from the systematic review and were not subjected to pooled comparative analysis.

Table 5.

Fusion Outcomes by Construct Type

Construct	3-Level ACDF	4-Level ACDF
Cage-Plate Fusion Rate (%)	97.1 (91.7–100) (n=8)	98 (94–100) (n=3)
Stand-Alone Cage Fusion Rate (%)	91.1 (72–100) (n=5)	87.5 (n=1)

Complications

Complication outcomes are summarized in Table 6. Overall complication rates were similar between 3- and 4-level constructs and overlapped significantly. The most common complications were adjacent segment disease, dysphagia, and pseudoarthrosis, followed by graft or hardware failure, hematoma, C5 palsy, and infection. Readmissions were reported in two cohorts (119 total patients; 3-level: 97, 4-level: 22), with a total of three readmissions (3-level: 2; 4-level: 1). Reoperations were reported in 21 cohorts with 1075 patients (3-level: 699; 4-level: 376), with 77 reoperations recorded (3-level: 47; 4-level: 30).

Table 6.

Complication Outcomes

Complication	3-Level ACDF	4-Level ACDF
Overall complications (%)	23.3 (2.3–42.6) (n=14)	31.8 (3.1–76.5) (n=10)
Adjacent segment disease (%)	30.2 (4.8–64.3) (n=7)	5.3 (3.8–6.3) (n=3)
Dysphagia (%)	25.0 (6.1–69.2) (n=19)	27.9 (4.6–80.0) (n=13)
Pseudoarthrosis (%)	22.4 (2.3–42.9) (n=9)	21.7 (0–56.3) (n=11)
Graft/hardware failure (%)	6.0 (0–36.4) (n=7)	5.8 (0–16.7) (n=4)
Hematoma (%)	1.9 (0–4.8) (n=7)	2.6 (0–7.7) (n=6)
C5 palsy (%)	1.5 (0–4.8) (n=9)	2.3 (0–6.5) (n=10)
Infection (%)	0.0 (0–4.8) (n=7)	2.0 (0–4.0) (n=4)

Meta-analysis of overall complication rates did not demonstrate a statistically significant difference between construct lengths, although confidence intervals were wide and the number of comparative studies was limited (log odds ratio 1.37, 95% CI −1.10 to 3.85; p=0.28) (Figure 5).

Figure 5.

Forest plot of total complication rate for three-level ACDF compared with four-level ACDF (derived from meta-analysis of two studies). ACDF, anterior cervical discectomy and fusion

Among studies directly comparing 3- and 4-level ACDF, meta-analysis demonstrated no statistically significant difference in revision rates between groups (log odds ratio 0.08, 95% CI −0.70 to 0.86; p=0.84) (Figure 6). Similarly, no significant difference in postoperative dysphagia rates was observed between 3- and 4-level constructs (log odds ratio 0.35, 95% CI −0.64 to 1.34; p=0.49) (Figure 7).

Figure 6.

Forest plot of revision rate for three-level ACDF compared with four-level ACDF (derived from meta-analysis of five studies). ACDF, anterior cervical discectomy and fusion

Figure 7.

Forest plot of dysphagia rate for three-level ACDF compared with four-level ACDF (derived from meta-analysis of three studies). ACDF, anterior cervical discectomy and fusion

Perioperative Outcomes

Across studies directly comparing 3- and 4-level ACDF, meta-analysis demonstrated that 4-level ACDF was associated with significantly longer operative time compared with 3-level constructs (Cohen’s d 1.19, 95% CI 0.81 to 1.58; p<0.001) (Figure 8) and significantly greater estimated blood loss (Cohen’s d 0.91, 95% CI 0.62 to 1.19; p<0.001) (Figure 9). Similarly, 4-level ACDF was associated with a modest but statistically significant increase in hospital length of stay compared with 3-level constructs (Cohen’s d 0.29, 95% CI 0.02 to 0.56; p=0.04) (Figure 10).

Figure 8.

Forest plot of operative time for three-level ACDF compared with four-level ACDF (derived from meta-analysis of four studies). ACDF, anterior cervical discectomy and fusion

Figure 9.

Forest plot of EBL for three-level ACDF compared with four-level ACDF (derived from meta-analysis of three studies). EBL, estimated blood loss; ACDF, anterior cervical discectomy and fusion

Figure 10.

Forest plot of hospital length of stay for three-level ACDF compared with four-level ACDF (derived from meta-analysis of three studies). ACDF, anterior cervical discectomy and fusion

Discussion

This systematic review and meta-analysis provides a focused synthesis evaluating outcomes following 3- and 4-level ACDF, with particular emphasis placed on studies directly comparing these constructs. Both 3- and 4-level ACDF resulted in meaningful improvements in pain, disability, and neurologic function, reinforcing the role of multilevel ACDF as an effective treatment option for appropriately selected patients with degenerative cervical pathology. However, when outcomes were directly compared through meta-analysis, increasing construct length was associated with significantly less improvement in neck pain, lower fusion rates, and increased perioperative morbidity.

Findings from the systematic review component of our study demonstrate that patients undergoing both 3- and 4-level ACDF experience substantial reductions in pain and disability, consistent with prior studies evaluating multilevel anterior constructs.^13,14,20 Across cohorts, mean postoperative VAS, VAS-neck pain, VAS-arm pain, NDI, and JOA/mJOA scores significantly improved from preoperative values in both groups. However, 4-level ACDF cohorts tended to have higher preoperative disability and pain scores and, despite improvement, higher postoperative scores compared with 3-level cohorts. This pattern suggests that while both constructs provide clinically meaningful benefit, patients requiring 4-level fusion represent a population with greater baseline disease burden.

These descriptive systematic review findings are supported by the meta-analysis of directly comparative studies, which demonstrated that 4-level constructs were associated with significantly less improvement in axial neck pain compared with 3-level constructs, while no difference was observed for radicular arm pain. This distinction aligns with the systematic-review data, where postoperative VAS-neck pain values remained higher in 4-level cohorts, whereas improvements in arm pain were comparable. Taken together, these findings suggest that while neural decompression achieved through ACDF is effective regardless of construct length, axial neck pain may be more sensitive to increasing fusion length and greater associated biomechanical stresses.

Conversely, several individual cohort studies, including Kim et al. and Canseco et al., reported postoperative improvement in PROs without statistically significant differences between construct length.^21,22 However, these studies were limited by modest sample sizes and variability in follow-up length, potentially limiting their ability to detect smaller effect sizes.

One of the most clinically relevant findings of our meta-analysis was the significantly lower fusion rate observed in 4-level ACDF compared with 3-level constructs, supporting findings from previous studies that fusion rates decrease with increasing construct length. Wewel et al. reported decreased segmental fusion rates with longer constructs, while De la Garza-Ramos et al. found higher pseudoarthrosis rates in 4-level ACDF, although these findings did not always reach statistical significance in isolation.^20,23 In contrast, the meta-analysis published by Wang et al. did not identify a statistically significant difference in fusion rates between 3- and 4-level ACDF.¹⁹ This discrepancy likely reflects the lower number of studies included in the meta-analysis of fusion rate in the Wang et al. study (two cohorts), versus six in the present analysis, suggesting that increasing construct length from three to four levels may impose biomechanical demands that meaningfully affect fusion success. Within the present data, the systematic-review subgroup analysis (Table 5) suggests that supplemental anterior plating may be associated with higher fusion rates than stand-alone cage constructs at both 3 and 4 levels, although these descriptive trends were not subjected to formal subgroup meta-analysis. Beyond construct selection, conclusions regarding patient-level biologic optimization (e.g., smoking cessation, bone-density management, bone morphogenetic protein use) cannot be drawn from the present dataset and represent a hypothesis for future prospective study.

Regarding fusion outcomes by construct type, the systematic review demonstrated higher reported fusion rates for cage–plate constructs compared with stand-alone cages in both 3- and 4-level ACDF. Fusion rates within each construct type were similar between 3- and 4-level procedures, though data for 4-level stand-alone cage constructs were limited. These findings are consistent with prior literature suggesting slightly higher fusion rates with supplemental anterior plating.²⁴ When considered alongside the lower fusion rates observed with increasing construct length in the present meta-analysis, these descriptive results suggest that construct selection may become increasingly relevant as the number of fused levels increases.

Meta-analysis of perioperative outcomes demonstrated a consistent and intuitive pattern. 4-level ACDF was associated with significantly longer operative time, greater estimated blood loss, and increased length of hospital stay. These findings align with previous individual comparative cohort studies and meta-analyses, all of which reported increased operative burden with 4-level ACDF when comparted to 3-level constructs.^19,21,23,25

Meta-analysis of perioperative complication rates did not demonstrate a statistically significant difference between 3- and 4-level constructs. However, confidence intervals were wide, and the number of comparative studies was limited, suggesting that this analysis may be underpowered to detect differences in relatively infrequent adverse events. Prior large database studies have reported higher short-term complication and reoperation rates with increasing fusion length, particularly in 4-level ACDF. While such studies offer improved statistical power, they lack the clinical and radiographic detail available in cohort-level investigations and were therefore excluded from the present analysis.

Overall, the findings of this meta-analysis complement those reported in large national database analyses. Joo et al. demonstrated that 4-level ACDF was associated with higher rates of 90-day adverse events, reoperation, and dysphagia compared with lower-level constructs, while 3-level ACDF did not consistently demonstrate increased risk when compared to shorter constructs.¹⁸ Although administrative database studies may overestimate complication rates due to coding limitations and lack of clinical granularity, the directionality of their findings aligns with the present pooled analysis; namely, that 4-level ACDF represents a distinct risk category relative to 3-level constructs.

This study has several limitations. The majority of included studies were retrospective, and there was heterogeneity in fusion definitions, outcome measure, complication reporting, and follow-up duration. To mitigate this heterogeneity, several methodological steps were taken. For the meta-analysis, random-effects models with restricted maximum-likelihood estimation were used to accommodate between-study variance, and pooled comparisons were restricted to studies that directly compared 3- and 4-level cohorts, thereby reducing the impact of inter-manuscript variation in surgical technique, indication, and outcome definitions. Funnel plots and Egger’s regression were used to evaluate small-study effects, with the recognized limitation that publication-bias assessment is unreliable when fewer than ten studies contribute to a given pooled estimate. Only a small number of studies directly compared 3- and 4-level constructs, limiting the statistical power of some meta-analyses. Despite these limitations, the consistency of findings across both descriptive systematic review data and comparative meta-analysis data supports the clinical relevance of presented findings.

Conclusion

In summary, both 3- and 4-level ACDF were associated with significant postoperative improvement in pain and function. However, meta-analysis revealed that 4-level ACDF demonstrated significantly less improvement in neck pain, lower rates of fusion, and greater perioperative morbidity when compared with 3-level constructs. While individual cohort studies may not consistently detect such differences, meta-analysis reveals clinically significant distinctions that should inform surgical planning and patient counseling regarding outcomes following 4-level ACDF. Future prospective, adequately powered studies are needed to better define the long-term clinical implications of multilevel anterior cervical discectomy and fusion.

Supplemental Material

Supplemental material - Clinical Outcomes Following Three- and Four-Level Anterior Cervical Discectomy and Fusion: A Systematic Review and Meta-Analysis

Supplemental material for Clinical Outcomes Following Three- and Four-Level Anterior Cervical Discectomy and Fusion: A Systematic Review and Meta-Analysis by Tyler M. Compton, Ravi A. Patel, Mark A. Plantz, Rohan M. Shah, Theodore A. Joaquin, Jacob R. Staub, Daniel E. Herrera, Cade F. Bennett, Anitesh Bajaj, Abhinav R. Balu, Matthew W. Colman, Wellington K. Hsu, Alpesh A. Patel, Srikanth N. Divi in Global Spine Journal

Supplemental Material

Supplemental material - Clinical Outcomes Following Three- and Four-Level Anterior Cervical Discectomy and Fusion: A Systematic Review and Meta-Analysis

Footnotes

ORCID iDs

Theodore A. Joaquin

Jacob R. Staub

Daniel E. Herrera

Ethical Considerations

This study is IRB exempt due to its design. No patient data was accessed.

Author Contributions

TC: conceptualization, data curation, formal analysis, investigation, methodology, writing – original draft, writing – review and editing.

RP: conceptualization, data curation, formal analysis, investigation, methodology, writing – original draft, writing – review and editing.

MP: data curation, formal analysis, methodology, writing – original draft.

RS: data curation, methodology, writing – original draft.

TJ: data curation, methodology, writing – original draft, writing – review and editing.

JS: conceptualization, writing – original draft.

DH: conceptualization, writing – original draft.

CB: data curation, writing – original draft.

AB: data curation, writing – original draft.

ARB: data curation, writing – original draft.

MC: supervision, writing – review and editing.

WH: conceptualization, supervision, writing – review and editing.

AP: conceptualization, supervision, writing – review and editing.

SD: conceptualization, supervision, writing – review and editing.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

All data was accessed from current medical literature. The collection of studies investigated can be accessed per reasonable request to the corresponding author.*

Supplemental Material

Supplemental material for this article is available online.

References

Bible

Goss

. To Use or Not Use Intraoperative Neuromonitoring: Utilization of Neuromonitoring During Spine Surgeries and Associated Conflicts of Interest, a Cross-Sectional Survey Study. J Am Acad Orthop Surg Glob Res Rev. 2022;6(3):e21.00273. doi:10.5435/JAAOSGlobal-D-21-00273.

Grochulla

. Anterior Cervical Discectomy and Fusion. In: Vieweg

Grochulla

, eds. Manual of Spine Surgery. Berlin Heidelberg: Springer; 2012:127-133.

Toci

Lambrechts

Karamian

, et al. Patients with radiculopathy have worse baseline disability and greater improvements following anterior cervical discectomy and fusion compared to patients with myelopathy. Spine J. 2023;23(2):238-246. doi:10.1016/j.spinee.2022.10.005.

Fraser

Härtl

. Anterior approaches to fusion of the cervical spine: a metaanalysis of fusion rates. J Neurosurg Spine. 2007;6(4):298-303. doi:10.3171/spi.2007.6.4.2.

Schroeder

Kurd

Millhouse

Vaccaro

Hilibrand

. Performing an Anterior Cervical Discectomy and Fusion. Clin Spine Surg. 2016;29(5):186-190. doi:10.1097/BSD.0000000000000383.

Emami

Coban

Changoor

, et al. Comparing Mid-Term Outcomes Between ACDF and Minimally Invasive Posterior Cervical Foraminotomy in the Treatment of Cervical Radiculopathy. Spine (Phila Pa 1976). 2022;47(4):324-330. doi:10.1097/BRS.0000000000004140.

Stull

Goyal

DKC

Mangan

, et al. The Outcomes of Patients With Neck Pain Following ACDF: A Comparison of Patients With Radiculopathy, Myelopathy, or Mixed Symptomatology. Spine (Phila Pa 1976). 2020;45(21):1485-1490. doi:10.1097/BRS.0000000000003613.

Veeravagu

Cole

Jiang

Ratliff

. Revision rates and complication incidence in single- and multilevel anterior cervical discectomy and fusion procedures: an administrative database study. Spine J. 2014;14(7):1125-1131. doi:10.1016/j.spinee.2013.07.474.

Epstein

. A Review of Complication Rates for Anterior Cervical Diskectomy and Fusion (ACDF). Surg Neurol Int. 2019;10:100. doi:10.25259/SNI-191-2019.

10.

Ong

Ghozy

, et al. Dysphagia rates in single- and multiple-level anterior cervical discectomy and fusion surgery: a meta-analysis. J Spine Surg. 2020;6(3):581-590. doi:10.21037/jss-20-506.

11.

Danto

DiCapua

Nardi

, et al. Multiple cervical levels: increased risk of dysphagia and dysphonia during anterior cervical discectomy. J Neurosurg Anesthesiol. 2012;24(4):350-355. doi:10.1097/ANA.0b013e3182622843.

12.

Wang

Yang

, et al. Factors predicting dysphagia after anterior cervical surgery: A multicenter retrospective study for 2 years of follow-up. Medicine (Baltimore). 2017;96(34):e7916. doi:10.1097/MD.0000000000007916.

13.

Laratta

Reddy

Bratcher

McGraw

Carreon

Owens

. Outcomes and revision rates following multilevel anterior cervical discectomy and fusion. J Spine Surg. 2018;4(3):496-500. doi:10.21037/jss.2018.06.16.

14.

Koller

Hempfing

Ferraris

Maier

Hitzl

Metz-Stavenhagen

. 4- and 5-level anterior fusions of the cervical spine: review of literature and clinical results. Eur Spine J. 2007;16(12):2055-2071. doi:10.1007/s00586-007-0398-7.

15.

Lee

Chen

Guo

Chen

. Comparison of radiological outcomes and complications between single-level and multilevel anterior cervical discectomy and fusion (ACDF) by using a polyetheretherketone (PEEK) cage-plate fusion system. Medicine (Baltimore). 2019;98(5):e14277. doi:10.1097/MD.0000000000014277.

16.

Albanese

Certo

Visocchi

Barbagallo

GMV

. Multilevel Anterior Cervical Diskectomy and Fusion with Zero-Profile Devices: Analysis of Safety and Feasibility, with Focus on Sagittal Alignment and Impact on Clinical Outcome: Single-Institution Experience and Review of Literature. World Neurosurg. 2017;106:724-735. doi:10.1016/j.wneu.2017.06.051.

17.

Luo

Lim

Chiu

Lai

Niu

. The surgical outcome of multilevel anterior cervical discectomy and fusion in myelopathic elderly and younger patients. Sci Rep. 2022;12(1):4495. doi:10.1038/s41598-022-08243-8.

18.

Joo

Zhu

Kammien

Gouzoulis

Arnold

Grauer

. Clinical outcomes following one-two-three-and four-level anterior cervical discectomy and fusion: a national database study. Spine J. 2022;22(4):542-548. doi:10.1016/j.spinee.2021.11.002.

19.

Wang

Long

Guo

Hou

. Comparison between 3-level and 4level anterior cervical discectomy and fusion in the treatment of cervical spondylotic myelopathy: A meta-analysis. Heliyon. 2023;9(11):e21595. doi:10.1016/j.heliyon.2023.e21595.

20.

Wewel

Kasliwal

Adogwa

Deutsch

O'Toole

Traynelis

. Fusion rate following three- and four-level ACDF using allograft and segmental instrumentation: A radiographic study. J Clin Neurosci. 2019;62:142-146. doi:10.1016/j.jocn.2018.11.040.

21.

Kim

Alan

Sansosti

Agarwal

Wecht

. Complications After 3- and 4-Level Anterior Cervical Diskectomy and Fusion. World Neurosurg. 2019;130:e1105-e1110. doi:10.1016/j.wneu.2019.07.099.

22.

Canseco

Minetos

Karamian

, et al. Comparison Between Three- and Four-Level Anterior Cervical Discectomy and Fusion: Patient-Reported and Radiographic Outcomes. World Neurosurg. 2021;151:e507-e516. doi:10.1016/j.wneu.2021.04.073.

23.

De la Garza-Ramos

Ramhmdani

, et al. Long-term clinical outcomes following 3- and 4-level anterior cervical discectomy and fusion. J Neurosurg Spine. 2016;24(6):885-891. doi:10.3171/2015.10.SPINE15795.

24.

Cohn

Carbone

Smith

, et al. Radiographic fusion and subsidence rates for stand-alone cage versus anterior cage-plate construct in ACDF. Seminars in Spine Surgery. 2024;36(3):101105. doi:10.1016/j.semss.2024.101105.

25.

McClure

Desai

Shabo

, et al. A single-center retrospective analysis of 3- or 4-level anterior cervical discectomy and fusion: surgical outcomes in 66 patients. J Neurosurg Spine. 2021;34(1):45-51. doi:10.3171/2020.6.SPINE20171.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.22 MB

0.21 MB