Introduction
Cervical radiculopathy is a clinical syndrome resulting from cervical nerve root compression, characterized by pain and sensory-motor symptoms in the neck and upper extremity [1,2]. Although most cases can be managed conservatively, surgical intervention is indicated for patients with persistent, debilitating pain and progressive motor weakness [1]. The two most commonly used surgical approaches are anterior cervical discectomy and fusion (ACDF) and posterior cervical foraminotomy, with the choice of procedure depending on several factors [3–5]. However, selecting the surgical level for cervical radiculopathy can be challenging when multilevel foraminal stenosis (FS) is present, due to highly variable nerve root anatomy and patient symptoms [6]. The course of cervical nerve roots often varies, inside the dura and at the brachial plexus, making it difficult to identify the symptomatic level [6,7]. Therefore, surgeons often need to operate on multiple levels that may contribute to the patient’s symptoms, as failing to treat a potentially symptomatic level can lead to inadequate symptom improvement [8].
However, increasing the surgical extent for treating cervical radiculopathy has significant drawbacks. With more levels of ACDF, the risk of acute complications, such as hematoma, airway compromise, and dysphagia, increases after surgery [9,10]. Moreover, fusion rates decrease significantly with multilevel fusions, potentially leading to poor outcomes [11,12]. The risk of interbody spacer failure and adjacent segment disease (ASD) also increases with multilevel ACDF compared to single-level surgery [12,13]. Moreover, cervical range of motion (ROM) decreases, and the financial burden increases with more extensive ACDF procedures [10]. Therefore, limiting the operation to only the levels most likely contributing to the patient’s symptoms is crucial.
Nevertheless, it remains unclear whether residual foraminal stenosis after ACDF adversely affects surgical outcomes, given the highly variable nerve root anatomy. Furthermore, the risk of ASD may increase if the underlying pathology is untreated, necessitating further investigation. Therefore, this study aimed to determine (1) whether residual FS at other levels not addressed by ACDF adversely affects postoperative outcomes and (2) whether the incidence of ASD increases in the presence of residual FS.
Materials and Methods
Study design and population
This retrospective cohort study was approved by the Institutional Review Board (IRB) of Asan Medical Center (IRB approval no., 2024–0338). Informed consent was waived due to the retrospective nature of the study. A total of 301 patients who underwent ACDF for treatment of degenerative cervical radiculopathy between March 2014 and December 2020, with a minimum follow-up period of 2 years, were reviewed. Exclusion criteria included surgery for trauma, tumor, or infection, previous cervical spine surgery, and a follow-up period under 2 years (Fig. 1).
Patients in whom all levels with foraminal stenosis were operated, and there was no residual foraminal stenosis on the same side as the initial symptoms, were classified as the no-foraminal stenosis group (No-FS group). Those with residual foraminal stenosis at levels not covered by ACDF on the same side as the initial radiculopathy symptoms were classified as the FS group. FS from C4 to T2 levels were considered. In patients with bilateral radiculopathy, those with residual foraminal stenosis on either side were included in the FS group.
Surgical level selection and procedures
Preoperative evaluation included detailed history taking, physical examination, and electromyography. Surgical level selection was determined considering potential variations in cervical nerve roots [6]. Surgical level selection did not rely on typical “Netter diagram” or characteristic symptoms of each nerve root compression, as research has shown that these typical clinical pictures are often flawed [8]. Considering previous studies based on myotomes and electrophysiological studies, levels that could potentially contribute to patient’s symptoms were included in the operation level [14,15]. However, levels with FS that were unlikely to be the cause of patient symptoms were not operated. For example, in patients with deltoid weakness, C4–5 and C5–6 levels were operated on if they had FS, since both levels can contribute to patient’s symptom (Fig. 2) [14]. However, FS at the C6–7 level was not treated since it is less likely to cause the patient’s symptoms. When unsure, the level was included in the fusion to ensure adequate treatment.
ACDF was performed via the standard Smith-Robinson approach. After disc and cartilage removal, endplates were prepared using a high-speed burr until bleeding bone was observed. Posterior compressive pathologies, including bone spurs, disc, and ossified posterior longitudinal ligament, were removed. Uncinate process hypertrophy contributing to FS was resected using a high-speed burr [3]. The interbody space was packed with allograft and autogenous local bone grafts [16]. An anterior cervical plate was applied for stabilization.
Variables
Patient data were collected through electronic medical chart reviews. Patient-reported outcome measures, including neck pain Visual Analog Scale (VAS) scores, arm pain VAS scores, and Neck Disability Index (NDI), were assessed preoperatively and at 3 months and 2 years postoperatively. Postoperative arm pain VAS scores were recorded based on symptoms on the same side as the initial presentation.
Radiographic assessments were performed twice by a spine fellowship-trained surgeon with 3 years of experience. FS was defined as a foraminal dimension less than 50% of the nerve root dimension, based on the criteria by Kim et al. [17]. C2–C7 lordosis was measured using the Cobb angle between the lower endplates of C2 and C7. Cervical ROM was calculated as the difference in C2–C7 lordosis between lateral radiographs taken in extension and flexion. The C2–C7 sagittal vertical axis (SVA) was measured as the distance from the posterosuperior corner of C7 to a vertical line drawn from the center of the C2 body. Radiographic assessments were performed preoperatively and at 3 months and 2 years postoperatively. ASD was diagnosed based on the following criteria: (1) radiographic evidence of FS at the adjacent level (foramen dimension <50% of the nerve root dimension); and (2) new-onset radiculopathy symptoms corresponding to FS at an adjacent level [17,18]. Fusion status was assessed at 2 years postoperatively using the following criteria: (1) interspinous motion of <1 mm on lateral dynamic radiographs (Fig. 2C); and (2) bone bridging formation on computed tomography (CT) [19,20].
Statistical analysis
A power analysis indicated that 116 patients would be required in the FS group to detect a clinically significant difference in postoperative NDI scores with 80% power and an alpha level of 0.05. However, the FS group in this study included only 26 patients, potentially underpowering it to detect small differences in postoperative NDI.
Normality of continuous variables was assessed using the Shapiro-Wilk test. Group comparisons (No-FS vs. FS group) were made using the independent t-test for continuous variables and the chi-square test for categorical variables. Intra-observer reliability for radiographic assessments was evaluated using interclass correlation coefficients, which were 0.812 for C2–C7 lordosis, 0.735 for cervical ROM, and 0.769 for C2–C7 SVA. Statistical analyses were performed using IBM SPSS Statistics for MAC, ver. 29.0 (IBM Corp., Armonk, NY, USA). All p-values of <0.05 were considered indicative of statistical significance.
Results
Study population
A total of 188 patients met the inclusion criteria. Among them, 162 patients (84.8%) without residual foraminal stenosis after ACDF were included in the No-FS group (mean age: 57.0±9.7 years; 78 men, 48.1%). The remaining 26 patients (13.6%) with at least one residual foraminal stenosis at the side of initial symptoms were included in the FS group (mean age: 61.3±9.1 years; 16 men, 61.5%). Patients in the FS group were significantly older than those in the No-FS group (p=0.018). Furthermore, the FS group more frequently presented with bilateral radiculopathy compared to the No-FS group (p=0.030) (Table 1).
Complications in the No-FS group included one case of hematoma (0.6%) requiring evacuation, one dural tear (0.6%), two mild screw migration (1.2%), and one temporary C5 palsy (0.6%). In the FS group, one patient had screw breakage (3.8%) and another had C5 palsy (3.8%) that recovered within 3 months.
In the FS group, four patients (15.4%) had two levels of residual foraminal stenosis, and 22 patients (84.6%) had one level. A total of 30 root compressions were left untreated, with C7 root compression being the most common (11/30, 36.7%). All 11 patients with untreated C7 nerve root compression had primary symptoms involving the deltoid muscle. Seven patients with untreated C5 nerve root compression had hand intrinsic muscle weakness. Furthermore, seven patients with untreated C4 nerve root compression had minimal neck pain and C4 radiculopathy was deemed less likely to contribute to their symptoms (Table 2).
Radiographic measurements
Preoperative radiographic parameters were not significantly different between the No-FS and FS groups. Both groups showed a significant improvement in C2–C7 lordosis after the operation (p<0.001 and p=0.033, respectively). C2–C7 lordosis at postoperative 3 months (p=0.583) and 2 years (p=0.384) was not significantly different between the two groups. Both groups showed a significant decrease in C2–C7 SVA after the operation (p<0.001 and 0.023, respectively), with no significant intergroup differences at postoperative 3 months (p=0.289) and postoperative 2 years (p=0.617). Similarly, both groups showed a significant decrease in cervical ROM after ACDF (p<0.001 and 0.030, respectively), with no significant intergroup differences at postoperative 3 months (p=0.135) and postoperative 2 years (p=0.214) (Table 3). Finally, there were no significant intergroup differences with respect to fusion rates at 2 years after the operation when assessed by CT bone bridging (p=0.853) and interspinous motion (p=0.987).
Patient-reported outcome measures
Preoperative neck pain VAS scores (p=0.728), arm pain VAS scores (p=0.074), and NDI scores (p=0.682) were not significantly different between the No-FS group and FS group. Both groups showed a significant improvement in neck pain VAS score after (p<0.001 and 0.035, respectively), with no significant intergroup difference at postoperative 3 months (p=0.227) and postoperative 2 years (p=0.538). Arm pain VAS scores also significantly improved after ACDF in both groups (p<0.001 and <0.001, respectively), with no significant intergroup difference at postoperative 3 months (p=0.869) and 2 years (p=0.273). Furthermore, NDI significantly improved after surgery in both groups (p<0.001 and 0.043, respectively). While NDI significantly improved between 3 months and 2 years postoperatively in the No-FS group (p=0.035), the improvement was not significant in the FS group (p=0.337), possibly due to the small sample size in the FS group. NDI scores assessed at postoperative 3 months (p=0.781) and 2 years (p=0.319) were not significantly different between the two groups (Table 4, Fig. 3). On logistic regression analysis, none of the preoperative demographic or radiographic factors showed a significant association with the NDI improvement rate of >50% (Table 5).
ASD
Nine patients in the No-FS group (5.6%), and two patients in the FS group (7.7%) were diagnosed with ASD during 2 years of follow-up (p=0.652). Furthermore, the incidence of ASD did not differ significantly between the two groups when analyzed based on the number of levels fused (Table 6). Four patients (2.7%) in the No-FS group and one patient (3.8%) in the FS group underwent revision surgery due to ASD. The revision rate was not significantly different between the two groups (p=0.529).
Discussion
In this study, 13.6% of patients had residual FS after ACDF for cervical radiculopathy. Despite this, patient symptoms, including neck pain, arm pain, and NDI, significantly improved after surgery in the FS group. Furthermore, postoperative outcomes were not significantly different between the No-FS and FS groups. Additionally, the rates of ASD and revision surgery due to ASD were not significantly different between the two groups despite residual FS.
Selecting the surgical level for cervical radiculopathy is a complex process [6] due to the variable course of the condition and nerve root interconnections within the dura and brachial plexus [21,22]. Unlike the lumbar spine, where symptoms often follow a predictable dermatomal pattern, cervical radiculopathy can present differently [23–26]. McAnany et al. [8] found that only 54% of patients exhibit a standard dermatomal pain pattern, making it challenging to pinpoint the affected level. Therefore, a highly selective approach for choosing surgical levels based on traditional symptom descriptions may risk leaving symptomatic levels untreated, leading to suboptimal outcomes [6–8,25]. Thus, surgical levels should include not only typically symptomatic levels, but also levels that potentially contribute to patient symptoms [6]. Only levels with FS that are unlikely to contribute to symptoms can be omitted from surgery.
Previous studies have identified methods to determine root levels unlikely to contribute to patient symptoms. Hong and Nam [25] found that while dermatomal symptoms (pain) often deviate from standard radiculopathy patterns, myotomal symptoms (motor weakness) tend to follow expected patterns. Notably, severe motor weakness (grade <3) is more likely caused by the expected level based on the typical cervical radiculopathy symptom description [25]. Electromyography-based studies have established connections between nerve roots and upper extremity muscles [15,27–29]. The electromyography-based myotome mapping by Furukawa et al. [15] summarizes previous research, enabling surgeons to identify levels unlikely to contribute to patient symptoms by referencing these studies and mappings [14,15]. In the present study, we included all levels with FS likely to contribute to patient symptoms, considering possible variations in cervical nerve root course, and excluded levels very unlikely to cause symptoms. However, it is unclear whether treating all pathologic levels with FS would lead to better outcomes or reduce ASD rates compared to selectively performing ACDF on potentially symptomatic levels. Fusion performed at adjacent levels may increase stress at remaining levels with FS, potentially hindering symptom improvement or increasing revision surgery rates for ASD, which warrants clarification.
The study’s findings suggest that leaving some levels of FS untreated during ACDF surgery does not negatively impact clinical outcomes, as long as potentially symptomatic levels are properly addressed. FS and no-FS groups showed significant improvement in neck pain, arm pain, and NDI scores after surgery, with no significant intergroup differences during follow-up. Therefore, surgeons need not perform operations for all levels with FS when a certain level seems unlikely to contribute to patient symptoms. Expanding surgical levels increases the risk of acute complications, including hematoma, dysphagia, and airway problems, as well as delayed complications like pseudarthrosis, graft-related complications, and instrument failure [9–12]. Furthermore, increasing ACDF levels inevitably incur greater costs and loss of ROM [10].
The incidence of clinical ASD and revision surgery due to ASD did not significantly differ between the no-FS and FS groups. This suggests that prophylactic fusion of nonsymptomatic levels to prevent future ASD may not be necessary. Previously reported risk factors for ASD after ACDF include younger age, decreased cervical lordosis, and developmental canal narrowing [13,18,30]. However, Song et al. [31] found that greater degenerative change at the adjacent level of the fused segment increases the risk of ASD, which contradicts the results of the present study. Further research is needed to clarify the relationship between adjacent segment degeneration and ASD rates, given the discrepancy between studies.
Considering the cost of increasing the level of fusion with ACDF and the favorable postoperative clinical course of those with residual FS after surgery, it would be reasonable to limit the operation level to those that potentially contribute to preoperative patient symptoms. Detailed history taking, physical examination, and electrophysiological studies should be performed to identify levels that may potentially contribute to patient symptoms [7,8,23,28]. Moreover, the risk and benefit of increasing the surgical extent should be balanced, considering the patient’s age, desire, and underlying medical conditions [6].
In this study, FS was defined as a foramen dimension of less than 50%, corresponding to the most severe grade in most grading systems [17,32,33]. Thus, further subdivision of FS was not feasible. This definition was adopted because mild or moderate stenosis has less clinical significance, and surgeons may face challenges in deciding whether to exclude levels with severe but asymptomatic FS.
Some limitations of this study should be acknowledged. First, being a retrospective cohort study, it cannot fully capture various cervical radiculopathy symptoms, making it challenging to detail the reasons for level selection. However, level selection was conducted uniformly by a single surgeon, considering patient symptoms, physical examination findings, radiographic findings, and electrophysiological studies. Second, the study did not clearly divide the levels with FS into symptomatic and non-symptomatic. However, considering the highly variable anatomy and symptoms of cervical radiculopathy [6,8], such classification is often not possible in clinical practice. Third, the study had a relatively short follow-up period. The possibility that the incidence of ASD may differ between the FS and no-FS groups beyond the 2-year postoperative period cannot be ruled out [30]. More long-term follow-up studies are required to clarify the relationship between residual FS and ASD [30]. Finally, this study had a small sample size relative to the estimated requirement from the power analysis, especially for the FS group. Therefore, the results should be interpreted with caution. Since our findings suggest that retaining asymptomatic levels of FS may not complicate the postoperative course of ACDF, further verification with a larger sample size is warranted.
Conclusions
The study found that residual FS at levels not addressed by ACDF surgery did not negatively impact outcomes. Patients with residual FS showed significant clinical improvement 2 years after surgery, similar to those without residual FS. Furthermore, residual FS did not increase the ASD rate or revision rate due to ASD. This suggests that surgically addressing all pathologic levels, including those unlikely to contribute to patient symptoms, may not provide additional benefits compared to limiting surgery to levels potentially contributing to patient symptoms.
Key Points
Residual foraminal stenosis at levels not included in anterior cervical discectomy and fusion (ACDF) does not negatively impact outcomes when symptomatic levels are appropriately addressed.
Untreated foraminal stenosis at non-symptomatic levels did not increase adjacent segment disease or revision surgery rates.
Limiting ACDF to levels contributing to patient symptoms minimizes surgical risks, reduces costs, and preserves cervical range of motion without compromising clinical outcomes.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
