Clinical and radiological outcomes after three- and four-level anterior cervical discectomy and fusion: A single-center case series of 113 patients with a minimum 2-year follow-up Authors: Mehdi Darmoul, Amine Trifa, Ghassen Elkahla, Kais Maamri Published: 2026 ABSTRACT ABSTRACTBackground:Anterior cervical discectomy and fusion (ACDF) is a standard procedure for cervical spine disorders, although multilevel ACDF (≥3 levels) is less frequently performed. This study evaluates the long-term clinical outcomes and effects on cervical sagittal balance following three- and four-level ACDF.Materials and Methods:We retrospectively analyzed patients treated for cervical spondylotic myelopathy with multilevel ACDF over 11 years. Patients were divided into two groups: three-level (Group 1) and four-level (Group 2). Clinical and radiological outcomes were assessed postoperatively and at 3, 12, and 24 months. Outcome measures included the Visual Analog Scale, neck disability index, Japanese Orthopaedic Association (JOA) score, and Odom’s criteria. Radiological evaluation included C2–C7 Cobb angle, sagittal vertical axis (SVA), and C7 slope.Results:A total of 113 patients (mean age 54.35 years) were included in the study. Most presented with neck pain and brachial neuralgia, with moderate-to-severe myelopathy on JOA scoring. Preoperative imaging frequently showed loss of cervical lordosis or kyphosis, with magnetic resonance imaging confirming multilevel cord compression. Of the cohort, 102 patients underwent three-level ACDF and 11 underwent four-level ACDF. Complications occurred in 14.1% of cases, mainly transient dysphagia. At a mean follow-up of 35.6 months, both groups demonstrated significant improvements in clinical scores and sagittal alignment.Conclusion:Multilevel ACDF is a safe and effective treatment for cervical spondylotic myelopathy, offering sustained clinical and radiological improvement with acceptable complication rates. INTRODUCTION Anterior cervical discectomy and fusion (ACDF) is a well-established, safe, and effective procedure for treating cervical spine disorders.[12] It not only relieves neurological symptoms but also helps restore cervical lordosis. Although commonly performed for one- or two-level disease, multilevel degeneration involving three or more levels is frequent, leading to increased use of multilevel ACDF. However, evidence on three- and four-level procedures remains limited, with few published series.[23] This study aims to assess the long-term effects of three- and four-level ACDF on patients’ quality of life and cervical sagittal alignment, representing one of the largest cohorts reported to date. MATERIALS AND METHODS Study sample We conducted a retrospective study involving 113 patients who underwent multilevel (ACDF), including procedures with three and four levels, at our Neurosurgery Department. The study covered the period from January 2010 to December 2020. We identified a total of 125 cases, and 113 of them met the inclusion/exclusion criteria [Figure 1]. Inclusion/exclusion criteria of three- and four-level anterior cervical discectomy and fusion cases Data collection From clinical notes, we extracted patient data, including age, sex, body mass index, comorbidities, smoking history, medical and surgical history, occupation, and clinical information. The imaging of each patient was reviewed; to assess sagittal balance, we measured on standard radiographs the Cobb angle (C2-C7 angle), C2-7 sagittal vertical axis (SVA), and the C7 slope. Magnetic resonance imaging (MRI) parameters were also collected for each patient, such as the presence of cervical stenosis, disc herniation, T2 hyperintensity in the spinal cord, and its level. From operative notes, we extracted intraoperative data, encompassing discectomy levels, surgical duration, and the occurrence of intraoperative incidents. Outcomes were obtained on follow-up visits (3, 6, 12 and 24 months in most cases), including patient quality of life assessed by the visual analog scale (VAS), the neck disability index (NDI), the Japanese Orthopaedic Association (JOA) score, the Odom’s criteria, and we also noted the presence of late complications, the sagittal balance of the cervical spine assessed by measuring radiological parameters from follow-up standard radiographs, measured using the same technique described earlier. Surgical technique Patients were positioned supine on the operating table. The neck was appropriately extended using a shoulder bump. Shoulders were taped down to provide appropriate visualization on lateral fluoroscopy. A standard Smith-Robinson approach with a right-sided oblique incision was made at the appropriate level. Caspar pins were used to aid in distraction; the discectomy and decompression were completed under the microscope to aid in visualization. All fusions were completed using cervical polyetheretherketone interbody devices fusion. An appropriate 3–4-level cervical plate was placed with screws, and a drain was left postoperatively. Patients received a hard cervical collar postoperatively for all cervical fusion procedures. Collars were worn for a 2-week duration. Statistical analysis Patient variables are expressed as the mean ± standard deviation where applicable. Binary outcomes were compared using Chi-square tests. Statistical analyses were conducted using IBM software: SPSS version 25. P < 0.05 was considered statistically significant. RESULTS A total of 113 patients met the inclusion and exclusion criteria; 102 underwent three-level ACDF (Group 1), whereas 11 underwent four-level ACDF (Group 2). The mean age of the cohort was 54.35 years (range: 34–78 years). The most represented age group was 50–65 years, accounting for 57.5% of patients. The study population included 73 males and 40 females. No statistically significant differences were observed between the two groups regarding age, sex, smoking status, or comorbidities [Table 1]. Table 1Demographics of 113 patients who underwent anterior cervical discectomy and fusionVariableTotalThree-level ACDFFour-level ACDF P Total number of patients11310211Mean age (years)54.3553.9580.06Male sex64.658650.71Smoking history43.3641.244.70.47Comorbidities23.427.222.50.33Presenting symptoms Radiculopathy84.986.272.70.32 Myelopathy69.966.690.9 0.04 Weakness57.5256.863.30.18 Sensory deficit72.5675.445.40.62 Bladder dysfunction1515.690.6Values are presented as percentage unless indicated otherwise; Boldface type indicates statistical significance. ACDF - Anterior cervical discectomy and fusion Demographics of 113 patients who underwent anterior cervical discectomy and fusion Values are presented as percentage unless indicated otherwise; Boldface type indicates statistical significance. ACDF - Anterior cervical discectomy and fusion The mean duration between symptom onset and surgical intervention was 19 months, ranging from 15 days to 120 months. The most common presenting complaint was neck pain (86.7%), followed by radicular pain (76.1%), sensory disturbances (57.5%), and gait impairment (51.3%). Analysis of clinical scores – including the VAS, NDI, and JOA score - revealed no significant differences between the two groups. However, myelopathy was significantly more prevalent in the four-level ACDF group (90.9%) compared with the three-level group (66.6%; P = 0.04) [Table 1]. No significant differences were observed between the two groups regarding mean cervical sagittal balance parameters. The four-level ACDF group demonstrated a relatively straight cervical alignment (Cobb angle between −5° and 5°), whereas the three-level group exhibited a more pronounced lordotic curvature (>5°). Static cervical spine abnormalities (straightening or kyphosis) were identified in 50% of cases, with no significant difference in their distribution between groups. MRI showed that the most frequently affected disc level was C4–C5 (38%), followed by C5–C6 (29%). Intramedullary T2-weighted hyperintensity was present in 62.8% of patients, most commonly at the C4–C5 level (41%). In all patients undergoing four-level ACDF, discectomy was performed from C3 to C7. Among patients treated with three-level ACDF, discectomy involved C3–C6 in 43.13% of cases, C4–C7 in 43.13%, and C5–T1 in 13.72%. The mean operative time was 270 min and was significantly longer in the four-level group (360 min) compared to the three-level group (252 min; P < 0.01). Postoperative complications occurred in 16 patients (14.1%) during the perioperative period. Dysphagia was the most common complication, affecting 12 patients (10.6%) [Table 2]. Persistent dysphagia at final follow-up was observed in one patient in the four-level group and in two patients in the three-level group. At discharge, clinical outcomes showed improvement in 76.1% of patients, stability in 23%, and deterioration in 0.8%. No cases of vertebral artery injury were reported in this series. Table 2Postoperative complications of all patientsComplication3-Level ACDF4-Level ACDF P Deep wound infection2 (1.8)00.18Dysphagia9 (8.8)3 (27.2)0.08Sensory motor deficits1 (0.9)00.11Mean length of stay (days)760.19Values are presented as the number of patients (%) unless indicated otherwise. ACDF - Anterior cervical discectomy and fusion Postoperative complications of all patients Values are presented as the number of patients (%) unless indicated otherwise. ACDF - Anterior cervical discectomy and fusion The mean follow-up duration was 35.68 months, ranging from 24 to 84 months. The mean Neck Pain VAS score showed a significant reduction, decreasing from 7.24 preoperatively to 5.68 at the final follow-up (P < 0.01). Similarly, the mean NDI improved from 67.08% preoperatively to 55.70% at the last follow-up (P < 0.01). In contrast, the mean JOA score demonstrated a significant increase, rising from 12.31 preoperatively to 14.09 at the final follow-up (P < 0.01) [Table 3]. Table 3Quality of life parameters of 113 patients who underwent anterior cervical discectomy and fusionVariable3-level ACDF4-level ACDF P Mean VAS score Preoperative7.217.50.44 At last FU5.66.50.09 P <0.01 0.08Mean NDI score Preoperative66.4773.30.095 At last FU54.7865.10.07 P <0.01 0.04 Mean JOA score Preoperative12.3511.90.64 At last FU14.1713.30.4 P <0.01 <0.01 Odom’s criteria (%) Excellent39.2127.20.57 Good31.327.20.93 Fair21.536.30.6 Poor7.890.83VAS, NDI, JOA Score are presented as the mean; Boldface type indicates statistical significance. FU - Follow-up; VAS - Visual Analog Scale; NDI - Neck disability index; JOA - Japanese Orthopaedic Association Quality of life parameters of 113 patients who underwent anterior cervical discectomy and fusion VAS, NDI, JOA Score are presented as the mean; Boldface type indicates statistical significance. FU - Follow-up; VAS - Visual Analog Scale; NDI - Neck disability index; JOA - Japanese Orthopaedic Association At the final follow-up, analysis of cervical sagittal balance parameters demonstrated a significant increase in the cervical Cobb angle, rising from 5.38° preoperatively to 9.34° (P < 0.01) [Figures 2 and 3]. A significant reduction in the SVA was also observed, decreasing from 23.36 mm preoperatively to 21.78 mm (P < 0.01). In contrast, the change in the mean C7 slope did not reach statistical significance (P = 0.13). Cervical spine lateral radiograph before (a) and after (b) a 4-level anterior cervical discectomy and fusion operation: Improvement of Cobb angle Cervical spine lateral radiograph before (a) and after (b) a 3-level anterior cervical discectomy and fusion operation: Improvement of Cobb angle Subgroup analysis revealed that improvements in both the C2–C7 cervical Cobb angle and SVA were significant in the two ACDF groups (P < 0.01). However, a significant postoperative change in C7 slope was observed only in patients who underwent three-level ACDF [Table 4]. Furthermore, a statistically significant correlation was identified between the reduction in SVA and the increase in the JOA score at the final follow-up (P = 0.03). Table 4Cervical sagittal balance evaluation parameters of our seriesVariableThree-level ACDFFour-level ACDF P Cobb angle Preoperative6.41°−2.64°0.03 At last FU10.03°3.9°0.05 P<0.01<0.01SVA (mm) Preoperative21.7120.700.16 At last FU22.1019.250.21 P<0.01<0.01C7 slope Preoperative22.02°16.2°0.15 At last FU20.91°19.3°0.4 P0.020.16SVA - Sagittal vertical axis; FU - Follow-up; ACDF - Anterior cervical discectomy and fusion Cervical sagittal balance evaluation parameters of our series SVA - Sagittal vertical axis; FU - Follow-up; ACDF - Anterior cervical discectomy and fusion DISCUSSION Degenerative cervical spinal cord involvement represents a common cause of disability among adult and, especially in the elderly population.[1] ACDF procedures have been firmly established to enhance both radicular and myelopathic symptoms in patients appropriately chosen for treatment. However, three and four-level ACDF procedures are less common, and data regarding their clinical outcomes are limited, particularly for four-level ACDF,[23] This is believed to be due to a presumed higher rate of complications.[4] The findings of our series align with recent studies,[34567] demonstrating that the anterior approach for treating three or more cervical levels is an effective procedure. The anterior approach to the cervical spine provides multiple benefits, including direct decompression of the spinal cord and nerve roots, correction of hypermobility, and the restoration of cervical lordosis.[8] Multilevel ACDF procedures offer the advantage of increasing the height of the anterior column through intervertebral disc replacement.[4567] The use of an anterior cervical plate is crucial for enhancing fixation rates, particularly in multi-level constructs.[8910] In all patients in our series, a locking anterior cervical plate was used. The most commonly reported complication in the literature following ACDF is dysphagia, which can occur in up to 50% of cases in the early postoperative period.[11] The development of dysphagia mainly depends on the number of levels treated.[11] By 12 months postoperatively, the rate of moderate to severe dysphagia decreases to 3%.[12] In our series, postoperative dysphagia occurred in 12 cases; however, only 3 patients had persistent symptoms at the time of final follow-up. The duration of follow-up is essential for accurately assessing late complications and radiological outcomes, as emphasized by a recent meta-analysis examining postoperative sagittal balance parameters and their association with delayed complications. This analysis revealed that articles with longer follow-up periods tended to report higher complication rates.[10] Our study assessed surgical outcomes over an average follow-up of 36 months, with a minimum postoperative follow-up of 24 months. We observed a marked improvement in neck pain, demonstrated by a significant reduction in both VAS and NDI scores. These findings align with existing literature. Indeed, Basques et al.[13] observed no significant differences in patient quality of life scores based on the number of levels treated, while Shin[14] noted improvements in pain and disability scores following single-, double-, and triple-level ACDF without significant differences between the groups. A study by Lin et al.[15] on 27 patients undergoing four-level ACDF revealed a substantial improvement in NDI, dropping from 35.5% to 12.6% after an average follow-up of 36.9 months. Similarly, Wang et al.[16] observed significant NDI improvement from 26.6% to 14.2% in 26 patients undergoing the same procedure, after 2 years of follow-up. However, concerns arise regarding the long-term risks associated with four-level ACDF, including chronic neck pain, swallowing disorders,[17] and surgical revisions, with revision rates potentially reaching up to 35%.[418] In our cohort of 11 patients who underwent four-level ACDF, none required readmission or further surgery. However, chronic neck pain was noted in 36.3% of patients, and only one patient still experienced moderate dysphagia. Several studies have employed the JOA score to assess neurological recovery following ACDF procedures. In the study by Zhou et al.[19] which followed 113 patients undergoing ACDF for 1 year, a change of 1.25 points in the JOA score was defined as clinically significant. In addition, a recovery rate threshold of 31.67% was established to differentiate between good and poor recovery outcomes. In comparison, our cohort demonstrated a mean improvement of 1.78 points in the JOA score, with an average recovery rate of 37.95%, exceeding the proposed benchmarks. Regarding cervical spine sagittal balance, our study demonstrated an improvement in cervical spine statics among operated patients, as evidenced by the increase in the mean Cobb angle, which was significantly greater in patients undergoing three-level ACDF procedures (P = 0.05). Shamji et al.[20] found in their systematic review that multilevel ACDF offers superior correction of sagittal balance compared to corpectomy and hybrid discectomy-corpectomy procedures. In addition, Basques et al.[13] demonstrated that multi-level ACDF achieves more significant restoration of lordosis compared to single and two-level ACDF. Their study revealed an average difference in cervical lordosis (C2-C7) of 6.5° among patients undergoing three and four-level ACDF, aligning closely with our series’ findings (6.54°). Sakai et al.[21] and Hirai et al.[22] both reported that cervical sagittal balance, particularly the C2-C7 SVA, improved after ACDF, and this improvement correlated with enhanced postoperative quality of life enhancement. Our study revealed a significant decrease in the mean C2-C7 SVA at the final follow-up (P < 0.01). Interestingly, our analysis suggests that this postural improvement correlates with the improvement in JOA score (P = 0.03). This finding, supported by several studies, demonstrates that improvement in sagittal balance is positively associated with enhanced patient quality of life.[23] Our study did not find a significant difference in the mean C7 slope at the final follow-up. Indeed, Xu et al.[24] demonstrated a significant improvement in the mean C7 slope following multi-level ACDF, but this orthopedic effect was lost at the last follow-up, with means returning close to preoperative levels. This is explained by the fact that after fusion, the lower part of the cervical spine and intervertebral disc bear more weight due to changes in force distribution, leading to compensatory changes in the lower cervical spine and an increase in the C7 slope compared to preoperative levels. Our study faced several limitations inherent to retrospective studies, including variability in patient selection criteria and surgical indications. The lack of a clear international consensus on these matters introduces potential biases. In addition, our reliance on available data for patient quality of life evaluations and radiographic assessments at a minimum of 2 years postoperation may introduce selection bias. The small sample size, with a disparity in the number of patients undergoing different levels of ACDF, limits the strength of our analysis. Furthermore, the lack of late postoperative MRI data and variability in the quality of obtained radiographs contribute to limitations in our radiological analysis. This may have led to underestimation of certain complications and hindered assessment of the impact of sagittal balance correction, especially in kyphotic patients. CONCLUSION Three and four-level ACDF represent a reliable therapeutic option for the surgical treatment of cervical myelopathy. Our analysis demonstrates that this procedure enables correction of the sagittal balance of the cervical spine, which correlates with improved quality of life for patients. A more precise assessment of surgical outcomes may be achieved through extended clinical and radiographic monitoring, spanning up to 24 months. Conflicts of interest There are no conflicts of interest.