Introduction
Lumbar interbody fusion is a widely performed surgical procedure used to treat diverse spinal pathologies, including degenerative disc disease, spinal stenosis, recurrent disc herniation, instability, spondylolisthesis, and spinal deformities [1]. A key component to this procedure is the use of interbody cages, which are implanted in the disc space following the removal of the intervertebral disc [2]. The insertion of interbody cages facilitates indirect neural decompression through ligamentotaxis [3], enhances anterior column stabilization, promotes successful fusion [4], and enables surgeons to restore lumbar lordosis and sagittal alignment [5].
Various types of interbody cages have been developed to support lumbar interbody fusion [6–8]. The placement of two cages in the intervertebral space has been considered optimal, except in cases involving severe asymmetric disc space narrowing, unilateral adhesions from previous surgeries, or anatomical nerve root variations [9]. Several clinical and biomechanical studies have raised questions regarding the necessity of bilateral cage placement [10,11]. The existing literature reports favorable outcomes for unilateral cage placement compared to bilateral cage placement, particularly regarding operative time and intraoperative blood loss; however, findings regarding the rate of successful arthrodesis remain inconsistent [12,13].
This study aimed to systematically review and meta-analyze the most current high-quality evidence comparing unilateral and bilateral intervertebral cages for lumbar interbody fusion in terms of successful spinal fusion, radiological outcomes, surgery-related outcomes, patient-reported outcomes, complication rates, and cost-effectiveness.
Materials and Methods
Registration and protocol
This study was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement [14]. A completed PRISMA checklist is presented in Supplement 1. Furthermore, this study was registered in the Prospective Register of Systematic Reviews (PROSPERO) System (CRD42024583337).
Search strategy
The following Population, Intervention, Comparison, Outcomes model was used for formulating clinical questions in this systematic review: (1) Population: Patients undergoing lumbar interbody spinal fusion; (2) Intervention: Use of unilateral interbody cage; (3) Comparison group: Use of bilateral interbody cages; (4) Outcomes: The primary outcome was the spinal fusion rate. Secondary outcomes included postoperative complications, perioperative blood loss, length of hospital stay, operative time, Visual Analog Scale (VAS) score, Oswestry Disability Index (ODI) score, cost-related outcomes, and postoperative radiographic parameters.
To identify eligible studies, a systematic search of the PubMed (MEDLINE), Scopus, and Cochrane (CENTRAL) electronic databases was conducted, covering the period from database inception to August 20, 2024 (Supplement 2). A manual search of the reference lists of relevant articles was also performed. The primary database search was independently performed by two investigators (S.S. and D.S.) who screened and assessed all available publications. Any discrepancies were resolved through consultation with a third investigator (C.K.).
Inclusion and exclusion criteria
Specific inclusion and exclusion criteria were established prior to initiating the literature search. Studies were included based on the following inclusion criteria: (1) randomized control trials (RCTs) comparing lumbar interbody fusion using unilateral cage placement versus bilateral cage placement in human participants; (2) a postoperative follow-up duration >1 year; (3) inclusion of any surgical approach (e.g., posterior lumbar interbody fusion [PLIF], transforaminal lumbar interbody fusion [TLIF], anterior lumbar interbody fusion [ALIF], extreme lateral interbody fusion [XLIF]), provided the study directly compared unilateral versus bilateral cage placement.
The exclusion criteria were as follows: (1) studies that did not report outcome data for either the intervention or the comparison group; (2) studies investigating patients with a previous history of spinal fracture, infection, or malignancy; (3) studies with a follow-up period of less than 1 year; (4) Non-human, cadaveric, or animal studies; (5) computational finite element studies; (6) case series, cohort studies, and case reports; and (7) studies not published in English.
Data extraction
Two researchers (S.S. and D.S.) reviewed the included studies. To ensure accuracy and minimize human error, a third researcher (C.K.) independently reviewed the extracted data against the original information reported in the studies. The following data were extracted and recorded: (1) first author; (2) year of publication; (3) country where the study was conducted; (4) sample size, including the total number of participants and their allocation to the unilateral or bilateral cage groups; (5) intervention details, including the type of surgical approach used; (6) mean age and sex distribution of patients in each group; (7) fusion rate; (8) patient-reported outcomes, including VAS scores for back and leg pain, ODI scores, and patient satisfaction scores; (9) surgery-related parameters, such as operation time, length of hospital stay, and estimated blood loss; (10) complication rates, including data on cage migration and other procedure-related complications; (11) cost-related outcomes, including implant and total surgery costs; and (12) postoperative radiographic parameters, including disc height and segmental lordosis. Definitions and measurement methods for fusion success, disc height, segmental lordosis, and cage migration are provided in Supplement 3. For studies reporting multiple follow-up time points, data from the final available follow-up (typically at 12 or 24 months) were extracted for each outcome. Patients in the included studies were divided into two groups based on the number of intervertebral cages implanted.
Risk of bias and study quality assessment
The included studies were independently evaluated for quality by the same two authors. Any controversies or disagreements between the two authors were reviewed and resolved by a third author (C.K.) to achieve a final consensus. The risk of bias in the included studies was evaluated using the revised Cochrane Risk of Bias tool for RCTs (RoB ver. 2.0; Cochrane, London, UK), a standardized tool designed to assess the methodological quality of RCTs [15]. This approach utilizes a fixed set of bias domains that address specific aspects of trial design, conduct, and reporting. Within each domain, “signaling questions” are used to identify study characteristics that may introduce bias. Based on the responses, each domain is assigned a rating (“low risk,” “high risk” or “some concerns”). The outcomes of the risk of bias assessment are presented in Supplement 4.
Statistical analysis
All analyses were performed using the Review Manager ver. 5.4 software (Cochrane). The presence of heterogeneity was assessed using the Cochrane chi-square test. The extent of heterogeneity was quantified using the I2 statistic; heterogeneity was considered low (I2 <30%), moderate (I2=30%–60%), or high (I2 >60%). A random-effects model was applied when I2 ≥30%, while a fixed-effects model was used when I2 <30%. (Mantel-Haenszel model). Associations are reported as odds ratios (OR), risk ratios (RR), and risk differences (RD) with 95% confidence intervals (CI) for quality measurements and mean differences (MD) or standardized mean differences (SMD) for quantity measurements (±standard errors) with 95% CI. Specifically, for the analysis of clinical outcome ratings, the 4-point grading systems employed across studies—such as the Odom’s and Kirkaldy-Willis criteria—were consolidated into a single unified measure referred to as the “patient rate of successful clinical outcomes.” Although these two systems are distinct, they are conceptually aligned in both structure (each stratifies outcomes as excellent, good, fair, and poor) and clinical interpretation, thereby supporting their synthesis for pooled analysis. The four categories were dichotomized into two groups: outcomes rated “excellent” or “good” were classified as successful, while those rated “fair” or “poor” were considered unsuccessful. A dichotomous model was then employed for statistical analysis. For rare events, such as cage migration, the RD was calculated to yield a more reliable estimate. Regarding cost analysis, due to significant variations in cost scales across studies from different countries, the SMD was employed to account for differences in currency and cost reporting, thereby ensuring comparability of effect sizes. A p-value of <0.05 was considered to indicate statistical significance. The assessment of publication bias was not performed, as each outcome included fewer than 10 studies, limiting the reliability of such analyses [16]. Sensitivity analysis was performed to evaluate the strength and validity of the pooled results by sequentially excluding each individual study.
To further investigate the impact of the surgical approach on fusion success, a subgroup analysis was conducted based on the approach used: Only PLIF and TLIF were included, as these were the only approaches with extractable data.
Results
Study characteristics
Our search strategy yielded 895 potentially relevant studies. After excluding the duplicates (n=216), 679 records were screened based on title and abstract. A full-text assessment was performed for 19 studies, of which 11 were excluded for the following reasons: (1) 10 were comparative but non-RCT studies and (2) one lacked sufficient data to complete 2×2 contingency tables. A flowchart diagram outlining the search strategy is presented in Fig. 1.
Eight studies met the predefined inclusion criteria and were ultimately included in both the qualitative and quantitative analyses [17–24]. These studies were published between 2002 and 2024. The countries in which they were conducted were China (n=3), Japan (n=1), South Korea (n=1), Germany (n=1), Egypt (n=1), and the Netherlands (n=1). With respect to study design, all seven studies were RCTs. A total of 509 patients (215 males and 294 females) were included in the analysis. The unilateral cage group (U group) comprised of 255 patients, while the bilateral cage group (B group) comprised of 254 patients. The descriptive characteristics of the included studies are summarized in Table 1. All included studies involved lumbar interbody fusion performed at the L4–L5 and L5–S1 spinal levels. Seven studies evaluated single-level lumbar interbody fusion procedures [17–23], while one study included patients who underwent two-level fusions [24]. Baseline variables, including age, sex, and follow-up duration, were compared between the unilateral and bilateral cage groups for outcomes included in the meta-analysis. Regarding these outcomes, no statistically significant differences in age or sex were observed between the groups, indicating comparability at baseline (Table 2).
Fusion rate
Seven studies involving 374 patients (188 in the U group and 186 in the B group) reported the outcomes related to successful fusion [17–19,21–25] (Table 3). No significant difference was found between the two groups in terms of successful fusion (OR, 0.74; 95% CI, 0.35 to 1.53; p=0.41; I2=0%) (Fig. 2).
Subgroup analysis based on the surgical approach did not change the results. Specifically, the PLIF subgroup included two studies [17,19], with a total of 116 patients (60 in the U group and 56 in the B group), while the TLIF subgroup comprised of two studies [18,24], with a total of 115 patients (57 in the U group and 58 in the B group). In both subgroups, the analyses revealed no significant difference in fusion success rates between the unilateral and bilateral cage groups (OR, 0.85; 95% CI, 0.22 to 3.32; p=0.81 and OR, 0.46; 95% CI, 0.11 to 1.96; p=0.30, respectively) (Fig. 2).
Other radiological parameters
Only one study revealed postoperative radiological outcomes other than fusion [23] (Table 3), making pooled data analysis impractical. This study found no significant difference between the unilateral and bilateral cage groups in terms of postoperative disc height (MD, 1.20; 95% CI, −1.22 to 3.62; p=0.33) or segmental lordosis (MD, 1.20; 95% CI, −1.33 to 3.73; p=0.35).
Surgery-related parameters
Length of stay
Four studies involving 170 patients (85 in the U group and 85 in the B group) reported on the length of hospital stay [17,21,22,24–26] (Table 4). No significant difference in the duration of hospitalization was reported between the two groups (MD, −0.51; 95% CI, −1.02 to 0.00; p=0.05; I2=0%) (Fig. 3A).
Estimated blood loss
Seven studies including 462 patients (231 in the U group and 231 in the B group) evaluated the intraoperative blood loss [17–21,23,24] (Table 4). The U group patients experienced significantly lower blood loss compared to those in the B group (MD, −125.00; 95% CI, −187.68 to −62.32; p<0.0001; I2=86%) (Fig. 3C).
The patient-reported outcomes
Complications
Other complications
All eight studies provided data on complications excluding cage migration [17–24] (Table 3). Compared to the B group, the U group demonstrated a statistically significant reduction in the risk of complications (RR, 0.66; 95% CI, 0.45 to 0.96; p=0.03; I2=0%) (Fig. 5B). The specific complications classified as “other complications” in each study are outlined in Table 3.
Overall complications
The pooled analysis of overall complications indicated a statistically significant reduction in risk for the U group (RR, 0.67; 95% CI, 0.47 to 0.97; p=0.03; I2=0%) (Fig. 5C).
Cost-related outcomes
Two studies, encompassing 95 patients (48 in the U group and 47 in the B group), compared implant costs [17,24], while two other studies [17,21], including 55 patients (28 in the U group and 27 in the B group), evaluated differences in overall operating costs (Table 6). However, one study provided insufficient data on implant costs, limiting our ability to draw definitive conclusions from this source [24]. The remaining study reported a significant difference in implant costs favoring unilateral cage placement. In terms of overall cost, the analysis revealed a significant advantage for the unilateral cage group (SMD, −2.78; 95% CI, −3.55 to −2.01; p<0.00001; I2=0%) (Fig. 6).
Sensitivity analysis
A comprehensive outlier sensitivity analysis was undertaken to evaluate the robustness of the results. This process involved sequentially omitting each individual study in turn to determine whether any single study disproportionately influenced the overall results. The analysis validated that excluding any single study did not cause alteration in the statistical significance of the primary and secondary outcomes.
Discussion
Bilateral cage placement theoretically offers greater biomechanical stability, an increased probability of achieving spinal fusion and improved correction of spinal alignment [9]. Consequently, it has been routinely adopted as a standard approach in lumbar interbody fusion procedures. More recently, the use of a single cage for lumbar interbody fusion has gained attention due to its advantages, including reduced blood loss and a shorter operative time. The existing evidence on optimal instrumentation of the intervertebral space in spinal fusion is limited and primarily consists of low-quality studies with conflicting results—particularly concerning fusion rate [28,29], segmental spinal correction [30,31], clinical outcomes [9,32], and complication rates [9,17].
Two prior meta-analyses compared the outcomes of unilateral and bilateral cage implantation in lumbar interbody fusion. Liu et al. [12] reported comparable fusion rates between the two groups; however, the single cage group exhibited shorter operative times, reduced blood loss, and lower complication rates. Daher et al. [13] found that the bilateral cage group demonstrated a higher fusion rate and lower subsidence in studies utilizing the TLIF approach and in those combining TLIF and PLIF approaches. However, this advantage was not observed in studies that exclusively used the PLIF approach [13]. They also reported shorter operative times and lower blood loss in the unilateral group when analyzing studies employing the PLIF approach as well as studies that concurrently used the TLIF and PLIF approaches. Finally, they reported no difference in complication rates (excluding cage subsidence), VAS for back pain, and ODI scores. They attributed the higher fusion rates and lower subsidence rates to the increased contact area and reduced stress concentration provided by the two cages. Additionally, they observed that the higher subsidence rate in the unilateral group might be linked to the use of banana-shaped cages—commonly employed in unilateral TLIF procedures—which have been associated with an increased risk of subsidence.
Fusion represents the primary outcome of spinal instrumentation. In our study, the pooled data from available RCTs revealed no significant differences in fusion rates between the two groups. This finding is consistent with the meta-analysis by Liu et al. [12] but contrasts with the more recent meta-analysis by Daher et al. [13]. Furthermore, the subgroup analysis based on approach (PLIF and TLIF) demonstrated that the choice of surgical approach does not significantly influence spinal fusion success between unilateral and bilateral cage placement. This finding aligns with those of previous studies [33]. This result likely reflects the multifactorial determinants of successful lumbar interbody fusion, which extend beyond the number of cages or the cage–endplate contact area. Key determinants of fusion success include cage material and design, bone graft material, thorough filling of the disc space with graft material, supplemental stabilization with bilateral posterior pedicle screws, and quality of endplate preparation [10,34].
Regarding surgery-related parameters, our results indicate that the unilateral cage group experienced significantly shorter operative times and reduced intraoperative blood loss, while the length of hospital stay remained unaffected by the number of cages used. This finding may be attributed to the additional time and surgical manipulation required for preparing the disc space and inserting a second cage, which inherently increases both duration of the operation and blood loss. In addition, PLIF necessitates a bilateral surgical approach to accommodate the placement of two interbody cages, which prolongs operative time and leads to increased blood loss.
Patient-reported outcomes—including back or leg pain (VAS score), disability (ODI), and overall satisfaction—do not appear to be affected by the number of cages used for anterior instrumentation in elective lumbar spinal fusion. The results for postoperative back pain and disability were consistent with those reported in previous meta-analyses [13].
Concerning complication rates, the risk of cage migration was comparable between the two groups. Previous studies have shown that a smaller cage-to-bone contact area (use of a single cage or smaller cage size) serves as a potential risk factor for cage migration [35,36]. Based on this rationale, it is anticipated that the bilateral cage group would exhibit a lower risk of cage migration. However, our analysis did not confirm this hypothesis. Furthermore, the risk of both non-migration-related complications and overall complications was lower in the unilateral cage group. This finding contrasts with the results reported in previous meta-analyses [12,13]. The increased incidence of these complications in the bilateral group may be attributed to multiple factors. Bilateral cage placement typically entails an extended operative time and is linked to an increased blood loss, both of which are known risk factors for postoperative complications. Additionally, the placement of two cages necessitates greater manipulation of the neural elements, potentially elevating the risk of dural tears and nerve injuries [23,37]. Finally, the use of additional hardware in bilateral procedures may increase the likelihood of implant-related complications [24].
The difference in the overall procedural cost could be attributed to the shorter surgical duration, which in turn reduces the need for anesthetic agents and associated perioperative resources. Moreover, reduced blood loss reduces the likelihood of requiring transfusions. Finally, using a single cage instead of two lowers implant-related expenses.
The present study exhibits several strengths, including its comprehensive scope and focus on RCTs to provide high-quality evidence. To our knowledge, this is the first meta-analysis exclusively based on RCTs that directly compares unilateral versus bilateral cage placement in lumbar interbody fusion. Adherence to the PRISMA guidelines ensures a transparent and standardized approach, while the inclusion of a broad range of outcomes, including fusion rates, operative time, blood loss, and patient-reported outcomes, offers a holistic assessment. The inclusion of subgroup analyses based on different surgical approaches (PLIF and TLIF) enhances the depth of the findings, and the risk of bias was systematically evaluated using the Cochrane Risk of Bias Tool, ensuring a transparent and standardized evaluation of study quality. In addition, the inclusion of cost analysis provides valuable practical insights into healthcare decision-making.
However, the study also has certain limitations, notably the small number of included studies reporting on specific outcomes (e.g., fusion rate subgroup analysis, length of stay, VAS leg, patient rate of successful clinical outcomes, cage migration, and cost-related outcomes). This limited data may reduce statistical power, hinder generalizability of the results, and introduce potential bias. The limited radiological results reported in the included studies restricted our ability to further evaluate potential differences between unilateral and bilateral cage placement in restoring segmental lordosis and overall sagittal balance. Additionally, most studies had a follow-up duration limited to 2 years. Although this time frame is sufficient to determine the fusion status, which was our primary outcome, it limits the ability to draw conclusions regarding long-term functional outcomes. Furthermore, radiological parameters beyond the fusion status were inadequately reported in the included studies, limiting a more comprehensive assessment of interbody fusion efficacy. The exclusion of non-English studies may also have introduced selection bias. In addition, publication bias could not be formally assessed because of the limited number of studies per outcome (<10), which restricts the reliability of such evaluations. Finally, the scarcity of data on cost differences between unilateral and bilateral cage placement in the included studies limited the comprehensiveness of the economic cost analysis.
Conclusions
This meta-analysis indicates that, compared to bilateral cage placement in lumbar interbody fusion, unilateral cage placement provides notable benefits—including shorter operative time, reduced estimated blood loss, and lower complication rates—while maintaining comparable fusion success rates and patient-reported outcomes. Although bilateral cages are traditionally favored for their potential to enhance biomechanical stability and improve fusion outcomes, our findings suggest that unilateral cage placement is equally effective in clinical practice. Moreover, the cost savings associated with unilateral cage placement—stemming from shorter operative time and reduced resource utilization—underscore its potential as a more cost-effective alternative. These findings may inform surgical decision-making and encourage further research aimed at refining and optimizing lumbar interbody fusion techniques.
Key Points
Unilateral cage placement provides outcomes comparable to bilateral placement, with equivalent fusion rates, pain relief, functional improvement, and satisfaction levels.
Unilateral cage placement is associated with shorter operative time and decreased intraoperative blood loss.
Unilateral cage placement was linked to a lower overall complication rate, without an increased incidence of cage migration.
From a financial standpoint, using a single cage is more cost-effective without compromising clinical effectiveness.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement
