ABSTRACT

Conclusion:

Both caged and cageless LIF methodologies are associated with elevated fusion rates. The autograft group demonstrates certain limited advantages in terms of fusion, whereas the caged group exhibits some benefits primarily related to alignment. The implementation of cageless autograft LIF, marked by its straightforwardness, simplicity, and cost-effectiveness, appears to be an underappreciated surgical technique within the current context.

Results:

Within our study, the late-stage fusion rates were determined to be 96.3% in the caged group and 96.7% in the cageless autograft group. No alterations were identified in segmental and lumbar lordosis angles across both groups. Notably, the caged group exhibited a propensity for late-stage cage embedding, while graft migration was the most prevalent complication within the autograft group.

Method:

This retrospective comparative study was conducted on patients who underwent surgical procedures at our institution’s neurosurgery clinic between 2011 and 2018, with the sanction of the ethics committee from the same institution. The study group (Group 1, n=27) comprised patients who underwent L4-5 single-level lumbar instrumentation and transforaminal LIF using a banana cage alongside autograft for the purpose of fusion. The control group (Group 2, n=31) consisted solely of cases that underwent posterior LIF operations with the utilization of autografts. Corticocancellous bone fragments sourced from posterior structures during decompression were utilized as autografts. The study parameters encompassed fusion rates, segmental and lumbar lordosis angles, disc height, ipsilateral and contralateral foramen heights, as well as slip distance.

Objective:

Lumbar interbody fusion (LIF) entails the placement of a bone graft within the intervertebral space, with or without the use of a cage, subsequent to discectomy. While numerous studies have explored caged LIF methods within the literature, limited attention has been given to direct comparisons between caged and cageless LIF techniques. This study aims to scrutinize the delayed outcomes of interbody fusion involving peek-caged and cageless laminar autografts. The investigation was specifically carried out at the L4-5 level.

Introduction

Lumbar interbody fusion (LIF) is a surgical procedure involving the placement of a bone graft within the intervertebral space, often accompanied by the use of a cage, subsequent to a discectomy. In contemporary practice, LIF is approached through five primary surgical techniques: Posterior lumbar interbody fusion (PLIF), transforaminal lumbar interbody fusion (TLIF or MI-TLIF), oblique lumbar interbody fusion/anterior psoas (OLIF/ATP), anterior lumbar interbody fusion (ALIF), and lateral lumbar interbody fusion (LLIF). While a discernible array of advantages and disadvantages have been documented for each of these modalities, it remains noteworthy that extant literature does not yet furnish definitive evidence substantiating the unequivocal superiority of any single approach over the others (1). The related literature is replete with numerous investigations that meticulously juxtapose the divergent dimensions of caged LIF techniques (2-8). It is, however, discernible that a paucity of comprehensive studies exists in relation to the comparative evaluation between caged and cageless LIF methodologies (9-12). This emerging trend underscores the progressive propensity towards the integration of intervertebral cages as a de facto standard within the LIF procedure.

The utilization of autografts for LIF has conventionally entailed their extraction from either the iliac crest or the posterior osseous structures during decompression procedures. Iliac crest grafts manifest optimal graft attributes owing to their corticocancellous architecture; nevertheless, they are concomitantly linked to pronounced wound morbidity (13). Conversely, grafts acquired from the lamina and spinous processes during decompression, while susceptible to challenges in sustaining structural integrity, obviate the necessity for supplementary surgical interventions and proffer inherent cost-related benefits. The deployment of autograft material confers an advantageous edge in fusion dynamics when juxtaposed with allograft and synthetic alternatives, primarily attributed to its heightened tissue compatibility.

Interbody cages exhibit a diverse array of structural profiles and configurations, commonly crafted from titanium or polyetheretherketone (PEEK) materials. The osteoconductive attributes of titanium cages are prominently evident, accentuating their capacity to foster optimal bone integration. However, it is noteworthy that the inherent rigidity of titanium constructs may engender an elevated susceptibility to implant embedding, which is a recognized concern (5). In contrast, PEEK lattice architectures offer a distinctive advantage characterized by a closer approximation to the mechanical elasticity of osseous tissue. Nevertheless, the advantageous mechanical harmony offered by PEEK structures is counterbalanced by certain challenges, such as their inherently smooth and hydrophobic surfaces, which may, in turn, impede the process of fusion (5).

In this study, it was aimed a comprehensive comparative analysis of long-term outcomes of interbody fusion employing PEEK cages versus cageless laminar autografts. Potential advantages and disadvantages of two methods were evaluated by meticulous assessment of demographic attributes, radiological metrics, and clinical presentations. The investigation was meticulously circumscribed to a patient cohort exclusively encompassing those who underwent only single-level L4-L5 interbody fusion. This methodological constraint was carefully instituted to confer precision and specificity to the study outcomes, enabling a more incisive examination of the parameters under consideration.

Materials and Methods

This retrospective comparative study was undertaken within the purview of the neurosurgery clinic at our esteemed institution, spanning the duration encompassing 2011 to 2018. The study was executed subsequent to obtaining the requisite endorsement from the ethics committee affiliated with the same institution. Consent was obtained from all participants. The study cohort, herein referred to as Group I, was composed of patients who underwent TLIF procedures. This involved the application of single-level lumbar instrumentation at the L4-5 level, accompanied by the utilization of a banana-shaped cage housing autograft material. The control group, denoted as Group II, exclusively encompassed cases subjected to PLIF interventions. The autograft was derived from corticocancellous bone fragments sourced from posterior osseous structures during the process of decompression in both groups. The patients were randomly divided into two groups.

The investigation encompassed an extensive review of the hospital registry system and patient records. During this process, meticulous attention was directed towards capturing pivotal demographic attributes, encompassing elements such as operation duration, intraoperative hemorrhage volume, body mass index (BMI), comorbid conditions, as well as postoperative complications. The clinical parameters of interest were supported by the inclusion of preoperative metrics, including visual analog scale (VAS) scores and Oswestry disability index (ODI) assessments.

At distinct time intervals, a comprehensive evaluation of patient outcomes was conducted. This entailed administering questionnaires at the 3-month juncture postoperatively, followed by a subsequent assessment extending beyond the span of 2 years post-surgery. For the sake of methodological rigor, participants who were deceased or rendered uncontactable were judiciously excluded from the analytical framework, thereby fortifying the integrity of the study cohort.

Inclusion Criteria

Individuals who had undergone surgical intervention involving L4-L5 pedicle screw instrumentation in conjunction with interbody fusion were enrolled into the study. This procedural selection was contingent upon a diagnostic framework characterized by degenerative grade 1 listhesis, accompanied by demonstrable clinical and radiological indicators of instability. Furthermore, a key criterion necessitated that these patients had exhausted conservative therapeutic modalities, thus warranting surgical intervention as a subsequent step in their clinical management.

Exclusion Criteria

Exclusion criteria encompassed cases involving procedures beyond the confines of the L4-L5 spinal segment, as well as instances involving multi-level operations. Moreover, patients with a documented history of prior instrumentation, those subjected to either solitary or supplementary posterolateral fusion procedures, and individuals afflicted by malignancy, traumatic injuries, or severe osteoporosis, were methodologically precluded from participation. Additionally, participants meeting the unfortunate outcome of deceased status or deemed non-compliant with the requisite follow-up protocol were excluded from the study. Furthermore, a stipulation was imposed mandating the availability of postoperative lumbar imaging records for a minimum duration of two years subsequent to the surgical intervention. Patients for whom lumbar computerized tomography (CT) and radiographic data were unavailable within the hospital’s radiological record system were systematically excluded from the cohort under consideration.

Radiological Investigations

The radiological investigations constituted an integral facet of this study, engaging a comprehensive array of parameters to discern and quantify pertinent anatomical variables. The timeline of measurement encompassed preoperative, early postoperative, and late postoperative stages, extending over a minimum of two years. These assessments were meticulously executed employing both CT imaging and standing X-ray examinations.

Noteworthy metrics subject to meticulous quantification included segmental and lumbar lordosis angles, disc height dimensions, ipsilateral and contralateral foramen heights, as well as slip distances. The determination of segmental lordosis (SL) was methodologically anchored in the calculation of angular deviation from the lower L4 and upper L5 endplates. In parallel, the computation of lumbar lordosis was derived from lines tangentially projected from the upper L1 to S1 endplates.

Concurrently, disc heights were ascertained through meticulous measurement along the anterior, middle, and posterior planes tangential to the respective endplates. The computation of the mean disc height necessitated the division of the cumulative measurements by a factor of three. Correspondingly, foramen heights were discerned through the measurement of distances between lines spanning from the inferior aspect of the L4 pedicle to the superior region of the L5 pedicle (Figure 1).

The framework established by Lee et al., as adapted for this context, constituted the basis for evaluating the fusion status (14). This framework discerns the presence or absence of bridging bony trabeculae, graft-to-bone space, and dynamic motion (≥3°) as evident in radiographs acquired during dynamic movements. The meticulous examination of dynamic radiographs for motion, as indicated in Table 1, underpinned the fusion evaluation.

Ensuring methodological rigor, measurements were conducted by two independent neurosurgeons, with their findings subjected to subsequent averaging. Instances of interpretational discrepancies concerning fusion assessments were judiciously arbitrated through a consensus-driven resolution process (Figure 2).

Surgical Technique

All procedural interventions were undertaken by the co-authors of this study. Employing a posterior midline approach, the surgical field encompassed exposure of the distal facet of L3-4, the entire facet of L4-5, and the initial region of the transverse process. Predominantly, pedicle screws were deployed, save for specific cases involving advanced stenosis where alternative measures were considered. Following this, decompression procedures were administered either unilaterally or bilaterally contingent upon the direction of pressure. In scenarios necessitating unilateral decompression, an extended approach was favored to ensure optimal bone graft capacity.

Post-discectomy and thorough endplate preparation, the implantation of a banana cage was executed at a height commensurate with physiological norms, thereby avoiding encroachment upon the posterior one-third of the intervertebral disc space. It is germane to note that facet osteotomy was executed with judicious precision, calibrated to facilitate the placement of the banana cage. This strategic choice was underscored by the imperative to mitigate the risk of potential instability in prospective revision scenarios. This surgical modality, in essence, mirrors an adapted version of the TLIF technique. Conspicuously, the strategy deployed forgoes complete L4 total laminectomy, a decision rooted in the intention to preclude the exacerbation of superior adjacent segment disease. Instead, an approach characterized by laminectomy of a scope adequate to accommodate microdiscectomy and the subsequent introduction of the cage is employed. Prior to the final placement of rods, meticulous maneuvering of the operating table is undertaken to optimize lumbar lordosis. Importantly, the procedural execution refrains from aggressive compression, as the potential for foraminal stenosis dictates caution.

In instances where the only autograft technique is embraced, an allocation of 20 to 30 bone fragments, varying in size, is thoughtfully inserted. Conversely, the cage group encompasses the utilization of approximately 10 bone graft pieces, with half of the allocation seamlessly integrated within the confines of the cage apparatus.

Statistical Analysis

Descriptive statistical measures encompassing mean, standard deviation, median, minimum, maximum, frequency, and ratio were employed to elucidate the inherent characteristics of the dataset. The distributional properties of the variables were assessed via the Kolmogorov-Smirnov test. To expound upon the analysis of quantitative independent data, both the Independent Sample t-test and the Mann-Whitney U test were judiciously administered. In parallel, the scrutiny of dependent quantitative data entailed the application of the Paired-Sample t-test and the Wilcoxon test, aptly tailored to accommodate the investigative context. Qualitative independent data underwent rigorous evaluation through the application of the chi-square test, thereby affording insights into the interrelationships within this stratum of variables. All statistical analyses were conducted utilizing the SPSS version 28.0 software.

Results

The distribution of patients’ age, gender, blood loss amount, operation time, and follow-up period demonstrated no significant disparities between Groups I and II (p<0.05), as evidenced through the BMI distribution. Similarly, no substantial variation in fusion values was observed between two groups (p<0.05) (Table 2).

Disc Height

The preoperative mean disc height (DH) value revealed no substantial distinction (p>0.05) between Group I and Group II. Notably, early and late postoperative mean DH values in Group II were significantly diminished relative to those in Group I (p<0.05). Conversely, Group I exhibited a noteworthy increase in early and late postoperative mean DH values compared to the preoperative values (p<0.05). Meanwhile, Group II experienced a significant increase in early postoperative mean DH value compared to the preoperative baseline (p<0.05), while the late postoperative mean DH value in Group II exhibited no significant difference from the preoperative metric (p>0.05). Within Group II, the preoperative/early postoperative and preoperative/late postoperative mean DH increments were significantly lower in comparison to Group I (p<0.05) (Table 3).

Lumbar Lordosis

No substantial disparity emerged in preoperative, early postoperative, and late postoperative lumbar lordosis (LL) values between Group I and Group II (p>0.05). Furthermore, early postoperative and late postoperative LL values exhibited no significant differentiation in either Group I or Group II relative to their preoperative values (p>0.05) (Table 4).

SL

Preoperative, early postoperative, and late postoperative SL values demonstrated no notable discrepancies between Group I and Group II (p>0.05). However, in both groups, the early postoperative SL value exhibited no significant deviation from the preoperative benchmark (p>0.05), while the late postoperative SL value underwent a significant reduction (p<0.05) (Table 4).

Foraminal Height (FH)

Preoperative and late postoperative FH values yielded no considerable distinction between Group I and Group II (p>0.05). However, the early postoperative FH value in Group II was significantly lower compared to Group I (p<0.05). In both groups, the early postoperative FH values significantly increased in comparison to the preoperative measures (p<0.05), while the late postoperative FH values returned to the preoperative values (p>0.05). The preoperative/early postoperative and preoperative/late postoperative FH changes exhibited no substantial differences between Group I and Group II (p>0.05) (Table 4). Contralateral and ipsilateral foramen height yielded analogous statistical results, as indicated in Table 4.

Slip Measurements

The preoperative, early postoperative, and late postoperative slip values displayed no discernible differentiation between Group I and Group II (p>0.05). In Group I, both early postoperative and late postoperative deviation values exhibited a significant decrease in contrast to the preoperative values (p<0.05). Parallelly, Group II showcased a similar pattern (p<0.05). The preoperative/early postoperative and preoperative/late postoperative slip reduction, however, was markedly lower in Group II (p<0.05) compared to Group I (Table 5).

Functional Outcomes

No noteworthy differences emerged in preoperative, early postoperative, and late postoperative ODI values between Group I and Group II (p>0.05). In both groups, both early postoperative and late postoperative values displayed a significant decline compared to the preoperative measures (p<0.05). The preoperative/early postoperative and preoperative/late postoperative ODI changes presented a similar trend across both groups (p>0.05) (Table 5).

Analogous patterns were observed in the assessment of leg VAS scores. Specifically, preoperative, early postoperative, and late postoperative leg VAS scores exhibited no substantial differentiation between Group I and Group II (p>0.05). In both groups, both early postoperative and late postoperative leg VAS scores registered a significant reduction compared to the preoperative values (p<0.05). The preoperative/early postoperative and preoperative/late postoperative leg VAS score reductions were analogous between Group I and Group II (p>0.05) (Table 6). Analogously, the back VAS score results echoed the trends observed in leg VAS scores across both groups (Table 6).

Complications

A total of 23 complications were seen within the caged group (Group I) with 11 cases of cage embedding, while 14 complications with 4 cases of graft displacement were noted in the cageless cohort. No substantial intergroup disparities emerged concerning dural injury, new-onset loss of strength, reoperation, or the incidence of adjacent segment disease.

Discussion

The central impetus behind employing interbody cages or grafts in lumbar degenerative disease surgeries is to augment fusion rates. Notably, the literature has consistently demonstrated that PLIF and TLIF yield higher fusion rates and superior clinical outcomes when juxtaposed with posterolateral fusion (PLF) approaches (15,16). This trend has culminated in the integration of PLIF and TLIF methodologies as near-standard practices, with interbody cage utilization becoming an inherent component of any posterolateral interbody fusion (PIF) intervention. Contemporary studies, in lieu of pitting against PLF, are now inclined to contrast PLIF and TLIF with innovative minimally invasive modalities such as ALIF, LLIF, and OLIF (17).

Despite the pervasive prevalence of PIF procedures and the amplification of the data corpus in scientific literature, studies on only interbody fusion employing autografts remain underrepresented. Autografts sourced from patients possess an array of merits, devoid of immunological complications and characterized by elevated fusion rates. The application of iliac crest grafts as a graft source in PIF has also been explored (18). Regrettably, despite the benefits, this ideal graft comes with the associated burdens of pain, bleeding, and infection due to secondary incisions, which has led surgeons to seek more minimally invasive alternatives, such as avoiding the utilization of iliac crest grafts.

Bone decompression stands as a necessity in the majority of surgeries related to degenerative spondylolisthesis. Even in cases where minimally invasive methods like ipsilateral contralateral decompression are pursued, ample bone fragments can be sourced to facilitate fusion over a single disc distance. This prompts the inquiry: Can these naturally obtained bone fragments be judiciously placed within the discectomy space, effectively fostering fusion? Does the prevailing attention granted to interbody cages align with their true significance?

Within our study, late fusion rates were observed to reach 96.3% in the caged group and 96.7% in the cage-free autograft group, concordant with existing literature. This echoes findings in independent studies that employed laminar bone fragments, where fusion outcomes correlated with our results. However, divergences in evaluation timeframe and criteria necessitate caution when drawing direct comparisons (9,10).

Although notable fusion rates exhibited by both groups in this study, more definitive evidence of fusion was observed in the autograft group (Group II). In the caged group (Group I), the device appeared to constrict the fusion area, engendering a hypodense space between the device and bone. Fusion was realized through autografts positioned anteriorly and posteriorly to the cage, rather than within the cage itself. These hypodense regions between the cage and end plates might potentially show a disadvantage for PEEK material in terms of fusion. Noteworthy, certain studies have reported the superiority of titanium cages over PEEK cages in terms of fusion efficacy (19).

Furthermore, even though posterolateral fusion was not a primary objective in both groups, significant fusion was observed within the facets, hinting at how PIF methodologies indirectly foster posterior fusion via the rigid construct formed anteriorly. While results pertaining to fusion alignment in both groups were comparable, the autograft group (Group II) exhibited a relative advantage due to its simplicity and cost-effectiveness. Additionally, the diminished facetectomy requirements associated with autograft interbody fusion mitigate the risk of potential instability upon instrument removal. Following the preliminary outcomes of this study, cageless autograft interbody fusion has evolved into the standard modality in our clinical setting for cases focused solely on achieving fusion.

These findings collectively underscore the significance of exploring alternative avenues for achieving successful interbody fusion, with a particular emphasis on harnessing autografts and simplifying procedural approaches, while maintaining an eye on long-term stability and cost-effectiveness.

An integral motivation underlying the adoption of interbody cages is the mitigation of root compression stemming from potential foramen stenosis, accomplished through the preservation of disc and foramen height. Related literature underscores the utility of employing solid interbody cages to sustain disc height (20). Conversely, another study reported a reduction in disc height when autografts were utilized (21). In the current study, significant early postoperative augmentation in mean disc height within the caged group failed to persist in the long-term. In other words, although late-phase mean disc height exhibited superiority within the caged group, noteworthy collapse ensued in both caged and cage-free groups, deviating from anticipated findings. The incapacity to maintain early postoperative mean disc height increment within the caged group could potentially be attributed to the predominantly elderly female patient cohort, where probable osteoporosis was a contributing factor. The majority of cases avoided complete excision of both ipsilateral and contralateral facets, opting instead for a physiological-sized device. The lack of supraphysiological devices might explain the failure to sustain late-phase disc height, despite prior literature suggesting that such devices are more intrinsically integrated (22,23).

In the realm of lumbar PIF procedures, the paramount objective of employing interbody devices pertains to the rectification of lordosis angles and the achievement of optimal alignment. A systematic review investigating lumbar angle enhancements following PIF revealed mean corrections of 4.67, 4.47, and 3.89 degrees for ALIF, LLIF, and TLIF, respectively (24). In our investigation, no substantive enhancements in lumbar or SL emerged in preoperative, early postoperative, and late postoperative phases within or between both study groups. Notably, even the autograft group (Group II) exhibited a significant decline in late postoperative periods. This divergence from existing literature could stem from a multitude of factors such as a homogeneous patient population without kyphotic deformity, midline device placement rather than anterior placement, usage of interbody cages without angles, not to be performed ipsilateral and contralateral facetectomy, as well as not to be performed compression during rod insertion to avoid foraminal stenosis. Furthermore, the studies reporting some values that should be performed for lordosis correction were usually the studies evaluating operations of diverse levels and multi-level lumbar interbody fusions (LIF), whereas our study exclusively enrolled single-level fusion cases. These findings underscore the pivotal role of procedural application in effecting lumbar lordosis correction, transcending the choice of interbody fusion method. Consequently, our study has prompted an inclination toward the preferential use of interbody cages for lumbar lordosis correction, predominantly employing angled and supraphysiological dimensions. This entails ipsilateral and contralateral facet osteotomies, coupled with enforced compression to facilitate lordotic correction.

Evaluation of slip distances in our statistical analyses revealed a substantial decline in both groups during the early and late postoperative phases. Notably, intergroup scrutiny underscored superior deviation correction within the device group compared to its counterpart. This effect can be attributed to a more aggressive discectomy approach in the device group, coupled with the corrective influence exerted by distraction during cage insertion on the slippage.

Past studies have accentuated the connection between successful fusion and clinical contentment. Moreover, it is postulated that the preservation and enhancement of disc height could theoretically correlate with the amelioration of leg pain (25). In line with this notion, both groups within our study demonstrated notable improvement in VAS and ODI values during early and late postoperative stages, displaying remarkable congruence. While foramen heights were relatively better preserved in the caged group, no significant divergence in leg VAS values was observed. This observation aligns with existing literature (9-11). However, we posit that a larger prospective study would be indispensable to reveal any discernible disparity.

In summary, our findings underscore the multifaceted nature of lumbar PIF interventions, particularly in relation to the deployment of interbody cages. The interplay between patient characteristics, procedural nuances, and device attributes collectively shapes clinical outcomes, thereby advocating for a comprehensive approach that integrates existing evidence with nuanced clinical judgment.

This elevated frequency of cage embedding potentially correlates with the high prevalence of female patients possibly afflicted with osteoporosis, mirroring the diminished maintenance of disc height. However, the challenge of comparing this complication across existing literature persists due to disparities in mean age and gender distribution. A second salient complication manifested in the cageless autograft group, characterized by four instances of graft displacement. This outcome underscores the possibility that solitary reliance on autografts may not invariably yield structurally robust outcomes. A potential remedy lies in modulating the surgical approach, necessitating the creation of a more conservative window within the posterior longitudinal ligament for cageless PIF scenarios. Additionally, judicious placement of grafts posterior to the disc space involves the utilization of substantial bone fragments, preferentially deposited in the anterior and middle regions of the disc space.

Noteworthy, no substantial intergroup disparities emerged concerning dural injury, new-onset loss of strength, reoperation, or the incidence of adjacent segment disease. Anticipating the contours of prospective inquiries, the investigation of larger cohorts emerges as a beneficial avenue, particularly concerning the intricacies of adjacent segment disease. In summary, the intricacies of complications within both interbody cage and cageless autograft scenarios reinforce the necessity of context-sensitive approaches, adeptly reconciling patient-specific attributes, procedural intricacies, and prior scholarly findings.

Study Limitations

The study is constrained by its relatively modest sample size and its retrospective design, which collectively curtail its robustness. The retrospective acquisition of clinical data, encompassing metrics like ODI and VAS at distinct time points, introduces certain vulnerabilities in terms of data comprehensiveness and reliability.

Study Strengths

In contrast to akin investigations in the literature, the distinctive attribute of our study rests in its meticulous focus on cases pertaining exclusively to a single fusion level (L4-5). Moreover, the anatomical dimensions of segmental and lordosis angles were meticulously gauged from standing radiographs, while the evaluation of fusion was grounded in comprehensive CT scans. We posit that the fusion assessment conducted through CT scans yields results of greater significance. The multifaceted nature of parameters inherent to PIF procedures, including factors like cage height, angle, osteotomy configuration, and rod positioning involving compression or distraction, imparts complexity to the interpretative realm. The insight garnered from our study holds potential to serve as a guiding compass for fledgling surgeons, elucidating the nuances behind the dearth of lumbar lordosis enhancement within our study cohort.

Conclusion

Both the caged and cageless autograft LIF cohorts evinced comparably elevated fusion rates, paired with commendable clinical outcomes. Upon juxtaposition, the cageless contingent exhibited certain modest advantages in terms of fusion outcomes, while the caged group showcased superior alignment attributes. The method of autograft only LIF surfaces as an unheralded surgical modality, marked by its streamlined, uncomplicated, and cost-effective attributes.