Outcomes of lumbar fusion surgery in isthmic versus degenerative A systematic review and meta-analysis

Authors: Pavlos Texakalidis, Stavros Matsoukas, Constantine L. Karras, Christopher S. Ahuja, Najib El Tecle, Tyler R. Koski, Nader S. Dahdaleh

Abstract

Lumbar spondylolisthesis can cause a significant disability due to back and lower extremity pain. Degenerative spondylolisthesis (DS) and isthmic spondylolisthesis (IS) are the two most common types. DS results from arthritic changes and degenerative disc disease, whereas IS occurs secondary to a pars interarticularis defect. This study compared surgical outcomes between IS and DS. A systematic literature review and meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Twenty studies comprising 3136 patients were included (DS: 1944; IS: 1192). No differences were identified in terms of estimated blood loss (mean difference [MD]: 5.79, 95% confidence interval [CI]: 42.88, 54.46; I^2^ = 81.7%), length of stay (MD: 0.24, 95% CI: 1.38–0.90; I^2^ = 87.7%), duration of operation (MD: 2.73, 95% CI: 20.61–15.16; I^2^ = 84.7%), reduction of Visual Analog Scale (VAS) for back pain before and after surgery (MD: 0.01, 95% CI: 0.68–0.65; I^2^ = 87.5%), reduction in Oswestry Disability Index before and after surgery (MD: 1.37, 95% CI: 2.39–5.12; I^2^ = 75.5%), return to operating room for any reason (odds ratio [OR]: 1.01, 95% CI: 0.48–2.15; I^2^ = 0%), successful fusion at the last follow-up (OR: 0.68, 95% CI: 0.26–1.77; I^2^ = 60.5%), adjacent segment disease during follow-up (OR: 0.45, 95% CI: 0.18–1.13; I^2^ = 0%), and subjective unsatisfactory outcome (OR: 1.53; 95% CI: 0.99–2.35; I^2^ = 0%). IS was associated with significantly higher odds for reporting subjective unsatisfactory outcome compared to DS (OR: 1.56; 95% CI: 1.00–2.43, I^2^: 0%) in the subgroup analysis of posterior lumbar interbody fusion/transforaminal lumbar interbody fusion. DS showed greater VAS back pain reduction than IS with posterolateral fusion alone (MD: -1.06, 95% CI: -1.90–0.22, I^2^:0%). Surgical outcomes are largely comparable between IS and DS, though etiology-specific differences in patient-reported outcomes may depend on fusion technique.

INTRODUCTION

Spondylolisthesis is a pathological condition of the spine, where there is a sagittal displacement of one vertebral body over the other. There are different types of spondylolisthesis, with the two most common types being degenerative spondylolisthesis (DS) and isthmic spondylolisthesis (IS). The former is of a degenerative nature, where spondylotic changes and degeneration of the intervertebral disc and facets are frequently seen in multiple levels of the spine. DS is multifactorial, with genetics having a role in the process.[1] The most commonly affected level is L4/L5. In contrast, IS is related to a defect in the pars interarticularis.[23] While it can be asymptomatic, a subset of IS patients can develop symptoms usually at an earlier age compared to DS. The most commonly affected level is L5/S1.[4] Nevertheless, patients with either condition present with similar symptoms, including low back pain and/or lumbar radiculopathy/neurogenic lower extremity pain.

Several studies have investigated the optimal treatment strategies for DS and IS. When conservative measures are exhausted, surgical intervention can be considered, which typically consists of spinal fusion. There are multiple different approaches that have been described, including posterolateral fusion (PLF), posterior lumbar interbody fusion (PLIF), transforaminal lumbar interbody fusion (TLIF), and anterior lumbar interbody fusion (ALIF), most commonly with or sometimes without posterior instrumentation.[5] Results across all these different approaches vary significantly within the literature.[6]

Even though DS and IS have different etiologies, the presenting symptoms are very similar, and studies overall report good to excellent improvement in back and leg pain as well as improvement in the quality of life measures.[78] Several comparative studies have investigated postoperative outcomes, given the different nature of the diseases and variability in the baseline characteristics of patients (IS patients usually younger, lower ASA score, and a lower body mass index compared to DS); however, the results are inconsistent across the literature. We performed a systematic literature review and identified studies comparing outcomes after the treatment of symptomatic IS versus DS. We subsequently performed a meta-analysis and synthesized outcomes after spinal fusion surgeries in these patients with a focus on intraoperative outcomes, postoperative symptom improvement, successful fusion at the last follow-up, as well as patient-reported unsatisfactory outcomes.

METHODS

This systematic review and meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.[9]

Search strategy and selection criteria

Systematic literature searches were conducted in PubMed/Medline. The search algorithm used for PubMed was the (isthmic AND degenerative). The search was conducted by two independent investigators. Any disagreements or discrepancies were resolved by consensus.

A study was included in this meta-analysis if it fulfilled the following predefined (i) prospective or retrospective observational analyses reporting outcomes after spinal fusion surgery separately for IS and DS and (ii) studies published up to October 2024. Studies that did not specifically provide outcomes of interest for both study groups were excluded as they could not be utilized in the analysis.

Data extraction

Data extracted included the first author, title, date of publication, study design, country of origin, patient number, demographics, age, enrollment years, sex, levels fused, Meyerding classification, preoperative comorbidities, duration of operation, estimated blood loss (EBL), length of stay (LOS), mean Visual Analog Scale (VAS) difference between pre- and postoperative for back and leg pain, mean Oswestry Disability Index (ODI) difference pre- and postoperative, subjective satisfaction as reported by the patients including definitions, postoperative complications, adjacent segment disease (ASD), and duration of follow-up. Risk of bias was assessed by two investigators (SM and PT) utilizing the ROBINS-1 tool for observational studies by Cochrane.[10]

Statistical synthesis and analysis

Odds ratios (ORs) with the corresponding 95% confidence intervals (CIs) were used for categorical outcomes using random-effects model meta-analysis. Mean difference (MD) was utilized for continuous outcomes with the 95% CI. Heterogeneity was assessed with the Higgins I^2^ statistic.[11] I^2^ > 50% indicated significant heterogeneity. Forest plots were used to graphically display the effect size in each study and the pooled estimates. P < 0.05 was considered statistically significant. STATA 14.1 (StataCorp, College Station, Texas, USA) was used as statistical software.

RESULTS

Literature search and characteristics of the included studies

The search algorithm yielded a total of 288 records after duplicates were removed. A total of 21 papers underwent full-text evaluation, and 20 studies in total were included as shown in the PRISMA flow diagram [Figure 1]. Risk of bias assessment according to the ROBINS-1 tool is presented in Table 1.

Six of the included studies were prospective,[121314151617] while the remaining were retrospective cohort analyses.[1819202122232425262728293031] In total, 3136 patients were included (DS: 1944; IS: 1192) in this systematic review and meta-analysis. The majority of patients had a Grade 1 or 2 spondylolisthesis at L4/L5 or L5/S1 and a weighted mean age of 57.9 years old. The mean follow-up duration was 36.3 months. Important baseline study and patient characteristics are presented in Table 2.

Outcomes of isthmic spondylolisthesis versus degenerative spondylolisthesis

No differences were identified in terms of EBL (MD: 5.79, 95% CI: 42.88–54.46; I^2^ = 81.7%) [Figure 2a], LOS (MD: 0.24, 95% CI: 1.38–0.90; I^2^ = 87.7%) [Figure 2b], duration of operation (MD: 2.73, 95% CI: 20.61–15.16; I^2^ = 84.7%) [Figure 2c], reduction of VAS for back pain before and after surgery (MD: 0.01, 95% CI: 0.68–0.65; I^2^ = 87.5%) [Figure 2d], reduction in ODI before and after surgery (MD: 1.37, 95% CI: 2.39–5.12; I^2^ = 75.5%) [Figure 2e], return to operating room for any reason (OR: 1.01, 95% CI: 0.48–2.15; I^2^ = 0%) [Figure 2f], ASD during follow-up (IS: 8.2%, DS: 14.5%; OR: 0.45, 95% CI: 0.18–1.13; I^2^ = 0%) [Figure 3a], subjective unsatisfactory outcome (IS: 13.2%, DS: 9.2%, OR: 1.53; 95% CI: 0.99–2.35; I^2^ = 0%) [Figure 3b], and successful fusion at the last follow-up (IS: 82%, DS: 91.3%, OR: 0.68, 95% CI: 0.26–1.77; I^2^ = 60.5%) [Figure 3c] between IS and DS.

Subgroup analyses

Posterolateral fusion only (isthmic spondylolisthesis versus degenerative spondylolisthesis)

Subgroup analyses were conducted by utilizing studies that only reported the use of PLF without an interbody and compared outcomes in IS versus DS. We found similar rates of fusion (IS: 90.3%, DS: 80%, OR: 2.27; 95% CI: 0.93–5.56, I^2^: 0%), ODI reduction between pre- and postoperative (MD: 2.36, 95% CI: 8.72–4.00, I^2^: 49.7%), and reporting subjective unsatisfactory outcome (IS: 21.6%, DS: 24.3%; OR: 0.89; 95% CI: 0.25–3.19, I^2^: 56.7%) between the two groups. The DS group was associated with significantly higher VAS back reduction between pre- and postoperative compared to IS (MD: 1.06, 95% CI: 1.90–0.22, I^2^: 0%) [Figure 4a].

Posterior lumbar interbody fusion/transforaminal lumbar interbody fusion only (isthmic spondylolisthesis versus degenerative spondylolisthesis)

Subgroup analyses were conducted by utilizing studies that only reported the use of interbody fusion either by PLIF or TLIF and compared outcomes in IS versus DS. No differences in terms of successful fusion (IS: 85.3%, DS: 95.8%, OR: 0.39; 95% CI: 0.14–1.09, I^2^: 18%), VAS back reduction between pre- and postoperative (MD: 0.09, 95% CI: 0.49–0.30, I^2^: 10.2%), and ODI reduction between pre- and postoperative (MD: 0.65, 95% CI: 3.69–4.99, I^2^: 18.7%) were identified between the two study groups. IS was associated with significantly higher odds for reporting subjective unsatisfactory outcome compared to DS (IS: 15.3%, DS: 10.2%; OR: 1.56; 95% CI: 1.00–2.43, I^2^: 0%) [Figure 4b]. Data were insufficient to perform subgroup analyses only for studies utilizing the TLIF approach.

DISCUSSION

To our knowledge, this is the first systematic review and meta-analysis synthesizing and comparing outcomes of spinal fusion in IS versus DS. In the cumulative analysis with all surgical approaches combined (TLIF/PLIF/ALIF/PLF), we found similar rates of EBL, LOS, duration of operation, reduction of VAS for back pain before and after surgery, reduction in ODI before and after surgery, return to OR for any reason, successful fusion at the last follow-up, ASD, and reporting subjective unsatisfactory outcome between IS and DS. Our subgroup analyses demonstrated that IS was associated with significantly higher odds for reporting subjective unsatisfactory outcomes compared to DS when interbody fusion was utilized either with PLIF or TLIF. In addition, the DS group was associated with significantly higher VAS back reduction between pre- and postoperative compared to IS when PLF without an interbody was used.

IS and DS are the two most common types of spondylolistheses. Although their etiologies and pathophysiology are different, they can both cause back and/or lower extremity pain. Importantly, DS patients are usually older than their IS counterparts, and DS has a predilection for female patients.[1618] Surgical approaches of spondylolisthesis, especially in the setting of abnormal motion on dynamic films, most commonly involve some type of fusion surgery. There are various approaches to this that can involve PLF with or without interbody fusion. When interbody fusion is utilized, most frequently it involves TLIF, PLIF, or ALIF at the L4-L5 and L5-S1 levels, which are most frequently affected. There are several systematic reviews and meta-analyses that have attempted to compare and identify whether one of these approaches is associated with superior outcomes in the setting of spondylolisthesis (either any type, or IS or DS only).[3233] To date, there is no known single surgical approach that has been universally associated with superior outcomes either for IS or DS, and therefore, the surgical approach is decided based on patient anatomy, patient and surgeon preference after discussion of risks, benefits, and alternatives.

The aim of our study was not to identify whether one of the surgical approaches was associated with superior outcomes, but rather we sought to compare intra- as well as postoperative outcomes in IS versus DS. In our cumulative analysis which included all approaches used (PLF, PLF, TLIF, and ALIF), we did not identify any differences in terms of EBL, LOS, duration of operation, reduction of VAS for back pain before and after surgery, reduction in ODI before and after surgery, return to OR for any reason, successful fusion at the last follow-up, ASD, and reported subjective unsatisfactory outcome between IS and DS. It is important to highlight that the surgical approaches utilized were the same across the groups described by each study (i.e. there were no studies that reported outcomes of TLIFs in IS versus PLFs in DS). Four studies reported the use of PLF only, ten studies reported the use of either TLIF or PLIF only, one used stand-alone ALIF, while one used either PLIF or ALIF with pedicle screws (circumferential fusion). The remaining four studies reported used of either PLF or some type of interbody technique; however, the outcomes were pooled and not reported based on the use of interbody or not. In this cumulative analysis, the development of ASD during follow-up was 8.2% and 14.5% for the IS and DS groups, respectively; however, statistical significance was not reached (OR: 0.45, 95% CI: 0.18–1.13; I^2^ = 0%). Unfortunately, only four of the included studies reported data on ASD, and therefore, it is likely that this analysis was underpowered to detect such a difference. An ASD rate of 8.2% versus 14.5% in IS versus DS, respectively, would be clinically meaningful to discuss with patients preoperatively. This potentially higher ASD rate in DS could be explained by the nature of the disease, with more diffuse spondylotic changes in adjacent segments that might not yet be symptomatic at the time of index surgery. With time, however, stress at these adjacent segments might worsen degeneration and potentially lead to ASD that may or may not require extension of fusion. In addition, the cumulative rates of subjective unsatisfactory outcome were 13.2% and 9.2% for IS and DS, respectively, and the analysis showed that this difference almost reached the level of significance with a lower 95 CI at 0.99 (OR: 1.53; 95% CI: 0.99–2.35; I^2^ = 0%). This could also be explained by an underpowered analysis, given only six studies reported relevant data for this outcome.

We additionally performed subgroup analyses to compare outcomes in the two study groups by including studies that performed PLF only and studies that either utilized a PLIF or TLIF. Interestingly, we found that for the PLF subgroup, DS was associated with significantly higher VAS back reduction between pre- and postoperative compared to IS (MD: 1.06, 95% CI: 1.90–0.22, I^2^: 0%). Moreover, in the PLIF/TLIF subgroup analysis, we found that IS was associated with significantly higher odds for reporting subjective unsatisfactory outcome compared to DS (OR: 1.56; 95% CI: 1.00–2.43, I^2^: 0%). The former means that the DS group had higher odds of reporting a significant VAS reduction for back pain by 1 additional unit compared to IS, when a PLF was performed. The latter finding means that when a TLIF/PLIF was performed, patients with IS had higher odds of reporting unsatisfactory outcome compared to DS. The common factor between reporting back pain scores and satisfaction after lumbar fusion is that both outcomes are subjective in nature. Back pain can fluctuate and is also multifactorial. Satisfaction after an operation is a product of set expectations, which are also multifactorial. Only speculations can be made as to why these differences exist; however, potential explanations include the subjective nature of these outcomes, unmet higher expectations for surgical outcomes, which could be a result of preoperative discussion with the surgeon. In addition, it is likely that IS constitutes a more challenging population to treat, given these patients are frequently young with fewer comorbidities and higher expectations with regard to their surgical outcome compared to DS patients. Accordingly, in the interbody subgroup analysis, we did not find a difference in the rates of fusion, VAS back, and ODI reduction between IS and DS, which points further to the subjective nature of “satisfaction” after a lumbar fusion surgery.

Another potential explanation for the lower satisfaction rate in IS patients in the TLIF/PLIF subgroup might be the fact that IS can be a more biomechanically demanding entity for reduction and fusion. A biomechanical analysis of spondylolisthesis reduction demonstrated anterior lever reduction to be a more effort-saving method compared to posterior lever reduction. Anterior required less torque value (116 N-cm versus 155 N-cm), while the total axial force received by the model was greater in the posterior versus anterior lever method (40.8 N versus 16.38 N).[34] Min et al. in a comparative study of ALIF versus PLIF showed similar outcomes; however, ALIF was associated with a lower rate of ASD (44% versus 82.6%).[35] Interestingly, Tye et al. demonstrated that ALIF with posterior percutaneous pedicle screws is associated with significantly greater improvement in EQ-5D scores at 1 year compared to TLIF in IS patients.[36] The ALIF group in that study also achieved greater segmental lordosis and disc height restoration compared to the TLIF group. Therefore, it would be very interesting to perform a subgroup analysis of the ALIF plus pedicle screw fixation population, comparing outcomes between IS and DS to investigate whether the difference in satisfaction rates is eliminated with this surgical approach. Unfortunately, data are currently insufficient in the literature to perform such an analysis.

Our study is limited by the retrospective design and nonrandomized nature of the included studies. Selection bias, surgeon preference, and experience cannot be adjusted or accounted for. In addition, even though we included 20 studies and 3136 patients overall, not all studies reported all outcomes of interest; therefore, each study was utilized for the outcomes for which it provided relevant data. In addition, definitions for the outcomes of fusion were not consistent between the studies with some studies reporting using radiographs (XRs) and others using computed tomographies (CTs) to demonstrate successful fusion; however, this was consistent between the same groups within each study. Similarly, reporting of unsatisfactory outcomes did not have a homogeneous definition across the included studies. In some instances, the definition just included the question whether the surgery helped or did not help, whereas in other studies, satisfaction was tied to the continuous use of opioids, persistence of pain, and level of physical activity achieved postoperatively; nevertheless, definitions were always the same for both groups within the same study.

CONCLUSIONS

The present systematic review and meta-analysis aimed to compare outcomes after lumbar fusion for IS versus DS. In the cumulative analysis with all surgical approaches combined (TLIF/PLIF/ALIF/PLF), we found similar rates of EBL, LOS, duration of operation, reduction of VAS for back pain before and after surgery, reduction in ODI before and after surgery, return to OR for any reason, successful fusion at the last follow-up, ASD, and reporting subjective unsatisfactory outcome between IS and DS. IS was associated with significantly higher odds for reporting subjective unsatisfactory outcome compared to DS when interbody fusion was utilized either with PLIF or TLIF. The DS group was associated with significantly higher VAS back reduction between pre- and postoperative compared to IS when PLF was used. Future large prospective studies would be required to confirm these results.

Conflicts of interest

There are no conflicts of interest.