Degenerative lumbar spondylolisthesis is a common condition afflicting an estimated 13.6% of the United States population.1 For a subset of symptomatic patients in whom nonoperative treatment fails, surgery is efficacious.2 Over the past several decades, there has been a marked evolution in the surgical techniques employed to treat spondylolisthesis, with techniques evolving from an era of open decompression alone to the advent of decompression with spinal arthrodesis via a posterolateral approach and/or an interbody device to, more recently, an era including surgical decompression (with or without spinal arthrodesis) via minimally invasive (MI) approaches. One strategy, the MI transforaminal lumbar interbody fusion (MI-TLIF), was first described by Foley and Lefkowitz in 2002.3 Compared to the traditional open TLIF,4 which involves a midline posterior approach with paraspinal musculature dissection, the MI-TLIF involves a paramedian Wiltse plane approach that avoids extensive musculature dissection and limits incision size (Fig. 1).
Depiction of an MI-TLIF. © Andrew K. Chan, published with permission.
Proponents of MI-TLIF suggest that this approach, via decreased musculoligamentous injury,5,6 decreases pain, minimizes blood loss, expedites perioperative recovery, and has been linked to improved clinical outcomes in several studies.7–9 There have been multiple studies comparing MI-TLIF to traditional open TLIF in cohorts of heterogeneous degenerative diagnoses.10–20 However, studies focused solely on spondylolisthesis cohorts are comparatively limited and are typically small, single-center investigations with 12- to 24-month follow-up.21–28 To assess comparative effectiveness in a larger, multicenter, and contemporaneous spondylolisthesis cohort, we previously studied 24-month outcomes in a prospective registry cohort of 297 patients with Meyerding grade I degenerative lumbar spondylolisthesis.7 At 24 months, Chan et al.7 observed an association between MI-TLIF and superior outcomes for disability, quality of life, and patient satisfaction. However, studies investigating longer-term outcomes (5–10 years) are required to assess the comparative durability of either procedure in treating lumbar spondylolisthesis, an especially important consideration given the relative nascency of MI-TLIF.
To this end, we compared the outcomes of MI and open TLIF with a minimum 60-month follow-up in a prospective multicenter cohort of 297 patients with Meyerding grade I degenerative lumbar spondylolisthesis, the largest single-study cohort by sample size, from the Quality Outcomes Database (QOD).
Methods
This was a retrospective analysis of prospectively collected data using an augmented data set from the QOD lumbar module. Institutional review board approval was obtained from the University of California, San Francisco.
This augmented data set (QOD lumbar spondylolisthesis module) represents the coordinated effort of 12 high-enrolling sites that combined their QOD data and collected additional data points. The data were audited by a central team and by individual sites. The inclusion criteria for the QOD lumbar spondylolisthesis module have been described previously.7,29–34 Briefly, we queried the module for patients who had undergone single-segment TLIF surgery for grade I lumbar spondylolisthesis via either an MI or an open approach from July 2014 through June 2016. Preoperative radiographs (standing or dynamic) had been obtained and evaluated by surgeons at the participating sites to confirm the diagnosis of grade I spondylolisthesis as defined by the Meyerding classification.35 Our definition of MI surgery (MIS) has been published previously.7 Specifically, surgeries were classified as MI if there was utilization of percutaneous screw fixation and placement of an intervertebral body graft (MI-TLIF) via the Wiltse plane. Procedures that involved only a component of MIS (e.g., "partially" MIS involving an "open" component) were not classified as MIS. In addition to the QOD exclusion criteria,36 we excluded patients who had grade II or higher spondylolisthesis or who had "partially" MI-TLIF procedures (e.g., mini-open37,38).
Demographic, Clinical, and Surgical Variables
Demographic variables included age, sex, body mass index (BMI), ethnicity, insurance, education level, and employment; patient comorbidities included smoking, diabetes, anxiety, osteoporosis, depression, and American Society of Anesthesiologists (ASA) classification; clinical characteristics included dominant presenting symptom, ambulation status, symptom duration, and presence of a motor deficit; baseline and follow-up patient-reported outcomes (PROs) included the Oswestry Disability Index (ODI), EQ-5D, numeric rating scale (NRS) for leg pain (NRS-LP), NRS for back pain (NRS-BP), and North American Spine Society (NASS) satisfaction; and surgical variables included the use of MI-TLIF versus open TLIF, estimated blood loss, operative time, length of hospitalization, discharge disposition, return-to-work rate, and reoperation rate. Reoperations were recorded if they were deemed related to the index surgery.
Primary and Secondary Outcomes
We assessed outcomes at 60 months using validated questionnaires. The primary outcome of interest was the ODI. The threshold for reaching the ODI minimum clinically important difference (MCID) was defined as 14.3.29 Secondary outcomes of interest included the NRS-BP, NRS-LP, EQ-5D, and NASS satisfaction. The NASS satisfaction questionnaire assesses patient satisfaction after surgery via a survey with 4 answer choices scored 1 through 4, respectively: "surgery met my expectations," "I did not improve as much as I had hoped but I would undergo the same operation for the same results," "surgery helped but I would not undergo the same operation for the same results," and "I am the same or worse as compared to before surgery." Radiographic outcomes (fusion, spondylolisthesis reduction) were determined by a neuroradiologist not affiliated with the study team.
Statistical Analysis
Descriptive statistics were reported as means ± standard deviations or frequencies and percentages, as appropriate. For univariable analyses, t-tests and chi-square analyses were used. Yates’ correction for continuity was employed when appropriate. Fisher’s exact test was used in the 2 × 2 case when an expected frequency was below 1. For multivariable analyses, multivariable linear regression models were fitted for ODI, EQ-5D, NRS-BP, and NRS-LP 60-month mean and change scores (i.e., 60-month value minus baseline value). A logistic regression model was fit for the analysis of reaching the ODI MCID. An ordinal logistic regression model was fit for the NASS satisfaction questionnaire. For each model, covariates included factors reaching p < 0.20 on baseline univariable comparisons and baseline PRO values for each respective model. This analysis was conducted using R 2.15.2 (R Foundation for Statistical Computing). Missing values in the data were imputed using the missForest R package. The p values were two-tailed, and an alpha of 0.05 was considered statistically significant.
Results
Overall, 297 patients were included, 225 (75.8%) of whom had undergone open TLIF and 72 (24.2%) of whom had undergone MI-TLIF. Follow-up rates of at least 60 months were not significantly different between the cohorts (75.6% vs 86.1%, respectively; p = 0.06).
Baseline Characteristics
Table 1 includes baseline demographic and clinical variables for the two cohorts. The cohort that had undergone MI-TLIF had a lower mean BMI (29.5 ± 5.1 vs 31.3 ± 7.0, p = 0.0497). Clinically, patients who had undergone open TLIF had a higher rate of motor deficit at presentation (23.6% vs 5.6%, p < 0.001). Those who had undergone MI-TLIF had a higher proportion utilizing workers’ compensation (11.1% vs 1.3%, p < 0.001). Otherwise, the cohorts were well matched for baseline PROs including ODI, NRS-BP, NRS-LP, and EQ-5D (p > 0.05 for all). Moreover, the cohorts shared similar magnitudes of baseline mean listhesis, mean age, comorbidities, back pain– versus leg pain–predominant presentations, symptom duration, level of education, and use of private insurance (p > 0.05 for all).
Characteristics of patients undergoing surgery for grade I lumbar spondylolisthesis
| Characteristic | Open TLIF | MI-TLIF | p Value |
|---|---|---|---|
| No. of patients | 225 (75.8) | 72 (24.2) | |
| Age in yrs | 59.5 ± 11.7 | 62.1 ± 10.6 | 0.10 |
| ≥60-mo FU rate | 170 (75.6) | 62 (86.1) | 0.06 |
| Female | 143 (63.6) | 40 (55.6) | 0.22 |
| BMI | 31.3 ± 7.0 | 29.5 ± 5.1 | 0.0497 |
| Smoker | 32 (14.2) | 9 (12.5) | 0.71 |
| Comorbidity | |||
| Diabetes mellitus | 34 (15.1) | 7 (9.7) | 0.25 |
| Coronary artery disease | 24 (10.7) | 9 (12.5) | 0.67 |
| Anxiety | 46 (20.4) | 11 (15.3) | 0.33 |
| Depression | 56 (24.9) | 16 (22.2) | 0.65 |
| Osteoporosis | 14 (6.2) | 3 (4.2) | 0.72 |
| Baseline listhesis in mm | 6.5 ± 3.2 | 5.9 ± 3.4 | 0.42 |
| Dynamic instability at baseline | 24/78 (30.8) | 11/32 (34.4) | 0.71 |
| Dominant presenting symptom | 0.54 | ||
| Back pain dominant | 88 (39.1) | 33 (45.8) | |
| Leg pain dominant | 30 (13.3) | 10 (13.9) | |
| Back pain = leg pain | 107 (47.6) | 29 (40.3) | |
| Motor deficit at presentation | 53 (23.6) | 4 (5.6) | <0.001 |
| Independently ambulatory | 208 (92.4) | 64 (88.9) | 0.34 |
| Symptom duration* | 0.34 | ||
| <3 mos | 5 (2.2) | 0 (0) | |
| ≥3 mos | 211 (93.8) | 69 (95.8) | |
| ASA class* | 0.59 | ||
| 1 or 2 | 126 (56.0) | 42 (58.3) | |
| 3 or 4 | 91 (40.4) | 26 (36.1) | |
| Hispanic or Latino ethnicity | 9 (4.0) | 5 (6.9) | 0.48 |
| ≥4 yrs of college education | 75 (33.3) | 29 (40.3) | 0.28 |
| Employed or employed & on leave | 108 (48.0) | 39 (54.2) | 0.36 |
| Private insurance | 118 (52.4) | 47 (65.3) | 0.06 |
| Use of workers’ compensation | 3 (1.3) | 8 (11.1) | <0.001 |
| Baseline ODI | 48.0 ± 16.6 | 46.2 ± 16.3 | 0.41 |
| Baseline NRS-BP | 7.0 ± 2.3 | 6.9 ± 2.6 | 0.70 |
| Baseline NRS-LP | 6.6 ± 2.8 | 6.3 ± 2.8 | 0.41 |
| Baseline EQ-5D | 0.53 ± 0.22 | 0.58 ± 0.20 | 0.08 |
FU = follow-up.
Values are expressed as number (%) or mean ± standard deviation, unless indicated otherwise. Boldface type indicates statistical significance (p < 0.05).
Missing data.
Perioperative Parameters
MI-TLIF was associated with less estimated blood loss (108.8 ± 85.6 vs 299.6 ± 242.2 ml, p < 0.001) and a modest increase in operative time (228.2 ± 111.5 vs 189.6 ± 66.5 minutes, p < 0.001). However, the mean length of hospitalization (MI-TLIF 2.9 ± 1.8 vs open TLIF 3.3 ± 1.6, p = 0.08), rate of discharge to home or home healthcare (93.1% vs 91.1%, respectively; p = 0.60), and rate of 30-day complications (6.9% vs 7.6%, respectively; p = 0.86) were similar. The 30-day complications in the MI-TLIF cohort numbered 5, including 2 durotomies, 1 hematoma, 1 neurological deficit, and 1 myocardial infarction. The 18 complications in 17 patients in the open TLIF cohort included 7 durotomies, 5 surgical site infections, 3 neurological deficits, 2 urinary tract infections, and 1 myocardial infarction.
Univariable Analysis of Outcomes
For the MI-TLIF cohort, comparing 60-month outcomes to baseline, there were significant improvements for ODI (18.9 ± 18.4 vs 46.2 ± 16.3, p < 0.001), NRS-BP (2.8 ± 2.9 vs 6.9 ± 2.6, p < 0.001), NRS-LP (1.8 ± 2.8 vs 6.3 ± 2.8, p < 0.001), and EQ-5D (0.77 ± 0.25 vs 0.58 ± 0.20, p < 0.001).
For the open TLIF cohort, comparing 60-month outcomes to baseline, there were significant improvements for ODI (24.8 ± 19.0 vs 48.0 ± 16.6, p < 0.001), NRS-BP (3.7 ± 3.1 vs 7.0 ± 2.3, p < 0.001), NRS-LP (2.8 ± 3.1 vs 6.6 ± 2.8, p < 0.001), and EQ-5D (0.75 ± 0.20 vs 0.53 ± 0.22, p < 0.001).
Table 2 provides the results of univariable comparisons of the two cohorts. Not accounting for confounding differences between the cohorts, MI-TLIF was associated with a lower mean ODI at 60 months (18.9 ± 18.4 vs 24.8 ± 19.0, p = 0.04). However, there was no significant difference in the change in ODI (MI-TLIF −26.1 ± 21.5 vs open TLIF −21.8 ± 19.6, p = 0.16) or the proportion reaching the ODI MCID (MI-TLIF 74.2% vs open TLIF 65.7%, p = 0.22). Moreover, MI-TLIF was associated with a significantly lower mean NRS-BP score (2.8 ± 2.9 vs 3.7 ± 3.1, p = 0.048) and NRS-LP score (1.8 ± 2.8 vs 2.8 ± 3.1, p = 0.03) but similar changes in the NRS-BP score (MI-TLIF −4.0 ± 3.5 vs open TLIF −3.3 ± 3.3, p = 0.21) and NRS-LP score (MI-TLIF −4.3 ± 3.3 vs open TLIF −3.8 ± 3.8, p = 0.40) at 60 months. For NASS satisfaction, there was a statistically significant difference in the distribution of scores between the two cohorts (p = 0.04). This was largely driven by a twofold greater number of respondents (10.7% vs 4.8%) in the open TLIF cohort with a NASS score of 4, which indicates that surgery resulted in no improvement in the preoperative state. Otherwise, a similar proportion of respondents (MI-TLIF 69.8% vs open TLIF 72.3%) noted the highest satisfaction score (NASS score 1), the only response indicating that surgery met a patient’s preoperative expectations. Otherwise, there were no differences in EQ-5D (MI-TLIF 0.77 ± 0.25 vs open TLIF 0.75 ± 0.20, p = 0.45), change in EQ-5D ( +0.18 ± 0.27 vs +0.20 ± 0.26, respectively; p = 0.69), or cumulative related reoperation rate at 60 months (5.6% vs 11.6%, respectively; p = 0.14). The reasons for related reoperations are provided in Table 3.
Univariable comparison of 60-month outcomes for patients undergoing MI-TLIF versus open TLIF for grade I lumbar spondylolisthesis
| Characteristic | Open TLIF | MI-TLIF | Unadjusted p Value |
|---|---|---|---|
| No. of patients | 225 | 72 | |
| Mean ODI | 24.8 ± 19.0 | 18.9 ± 18.4 | 0.04 |
| Mean change in ODI | −21.8 ± 19.6 | −26.1 ± 21.5 | 0.16 |
| ODI MCID | 0.22 | ||
| Reached | 111/169 (65.7) | 46/62 (74.2) | |
| Not reached | 58/169 (34.3) | 16/62 (25.8) | |
| Mean NRS-BP score | 3.7 ± 3.1 | 2.8 ± 2.9 | 0.048 |
| Mean change in NRS-BP score | −3.3 ± 3.3 | −4.0 ± 3.5 | 0.21 |
| Mean NRS-LP score | 2.8 ± 3.1 | 1.8 ± 2.8 | 0.03 |
| Mean change in NRS-LP score | −3.8 ± 3.8 | −4.3 ± 3.3 | 0.40 |
| Mean EQ-5D score | 0.75 ± 0.20 | 0.77 ± 0.25 | 0.45 |
| Mean change in EQ-5D score | +0.20 ± 0.26 | +0.18 ± 0.27 | 0.69 |
| NASS satisfaction score | 0.04 | ||
| 1 | 115/159 (72.3) | 44/63 (69.8) | |
| 2 | 26/159 (16.4) | 12/63 (19.0) | |
| 3 | 1/159 (0.6) | 4/63 (6.3) | |
| 4 | 17/159 (10.7) | 3/63 (4.8) | |
| Reops related to index surgery at 60-mo FU | 26 (11.6) | 4 (5.6) | 0.14 |
Values are expressed as mean ± standard deviation or number (%), unless indicated otherwise. Boldface type indicates statistical significance.
Reasons for reoperations related to index surgery at the 60-month follow-up
| Characteristic | Open TLIF | MI-TLIF | p Value |
|---|---|---|---|
| No. of patients | 225 | 72 | |
| Patients requiring reop | 26 (11.6) w/ 29 reops | 4 (5.6) w/ 6 reops | 0.14 |
| Adjacent-segment disease | 16 (7.1) | 3 (4.2) | |
| Wound infection | 6 (2.7)* | 1 (1.4) | |
| Removal of painful instrumentation | 2 (0.9) | 0 (0) | |
| Instrumentation revision | 2 (0.9) | 0 (0) | |
| Pseudarthrosis | 2 (0.9) | 1 (1.4) | |
| Subsidence | 1 (0.4) | 0 (0) | |
| Medical | 0 (0) | 1 (1.4)† |
Values are expressed as number (%), unless indicated otherwise.
Two of 6 reoperations were associated with an intraoperative durotomy at the index procedure.
The index surgery, a planned MI-TLIF, was aborted after the decompression was completed because of an air embolism. The patient returned in 1 month for instrumentation.
As previously reported in the comparison of 24-month outcomes,7 there were no differences in radiographic fusion at 24 months (MI-TLIF 100% vs open TLIF 96.4%, p = 0.21) and no difference in the reduction of listhesis (change in millimeters) at 24 months (−2.8 ± 3.8 vs −2.3 ± 3.5 mm, respectively; p = 0.51). At 24 months, 100% of eligible MI-TLIF patients returned to work compared to 80% of eligible patients in the open TLIF cohort (p = 0.02).
Multivariable Comparison of Outcomes
To account for baseline differences between the cohorts that had undergone open and MI-TLIF, we conducted multivariable adjusted analyses (Table 4). In these analyses, MI-TLIF compared to open TLIF was associated with similar outcomes for 60-month ODI, ODI change, odds of reaching ODI MCID, NRS-BP, NRS-BP change, NRS-LP, NRS-LP change, EQ-5D, EQ-5D change, and NASS satisfaction (adjusted p > 0.05 for all).
Multivariable analysis of 60-month outcomes following surgery for grade I lumbar spondylolisthesis
| Outcome | Adjusted* β Coefficient (95% CI) | p Value |
|---|---|---|
| Primary | ||
| ODI, 60 mos | −2.4 (−7.0 to 2.2) | 0.31 |
| ODI, 60-mo change | −1.9 (−6.5 to 2.7) | 0.43 |
| ODI MCID, 60 mos | 1.3† (0.7 to 2.7) | 0.45 |
| Secondary | ||
| NRS-BP score, 60 mos | −0.5 (−1.3 to 0.3) | 0.23 |
| NRS-BP score, 60-mo change | −0.4 (−1.1 to 0.4) | 0.35 |
| NRS-LP score, 60 mos | −0.5 (−1.3 to 0.2) | 0.17 |
| NRS-LP score, 60-mo change | −0.4 (−1.2 to 0.4) | 0.36 |
| EQ-5D score,60 mos | −0.01 (−0.07 to 0.04) | 0.61 |
| EQ-5D score,60-mo change | −0.01 (−0.07 to 0.04) | 0.64 |
| NASS satisfaction, 60 mos | 1.0† (0.5 to 1.8) | 0.99 |
Reference = open TLIF. β coefficients are reported such that a negative value for ODI, NRS-BP, and NRS-LP and a positive value for EQ-5D represent more favorable outcomes at 60 months for MI-TLIF compared to open TLIF. Odds ratios are reported such that an OR < 1.0 for NASS satisfaction represents increased odds for greater satisfaction at 60 months for MI-TLIF compared to open TLIF.
Multivariable models adjusted for factors with p < 0.20 on univariable comparisons and respective baseline PRO values.
Odds ratio.
Discussion
We present a comparison of 60-month outcomes in the largest single study cohort of patients who underwent MI or open TLIF for grade I degenerative lumbar spondylolisthesis. In our present comparison, after adjusting for baseline differences, we did not observe an association between the use of MI techniques and disability, back pain, leg pain, quality of life, and patient satisfaction compared to those with open TLIF. Similarly, we did not observe a difference between the cohorts for 60-month reoperation rates, 24-month fusion rates, and 24-month reductions in the magnitude of listhesis. Regardless of the strategy used, both cohorts demonstrated mean improvements in disability, back pain, leg pain, and quality of life compared to the mean baseline condition.
There are many reports on the perioperative advantages of MI-TLIF compared to traditional open TLIF. Our observation that MI-TLIF was associated with less intraoperative blood loss has been extensively reported.7,21–23,25–28,39,40 The length of hospitalization has varied with some studies demonstrating no significant differences7,21,23,25 and others, including a pooled meta-analysis,8 demonstrating shorter hospital stays following MI-TLIF for spondylolisthesis.22,24,26–28,39,40 This inconsistency likely stems from differences in local practice patterns and patient expectations that drive durations of hospital stay more than interprocedural differences. For example, our cohort, which represents high-volume surgeons’ experience in the United States, had similar mean stays of 2.9 and 3.3 days for MI and open TLIF, respectively. On the other hand, in a single-surgeon experience in Korea,40 MI-TLIF was associated with shorter stays (7.5 vs 12.6 days), although both procedures had mean lengths of stay over 2 to 3 times greater than our experience. Somewhat more variably, the operative time for MI-TLIF has been reported to be shorter,18,28 similar,40 and even longer8 than for open TLIF. This likely reflects a variation in surgical skill due to the learning curve involved in MI-TLIF.
Less well established are the consequences of an MI approach to TLIF on long-term (i.e., > 5 years) patient-centered outcomes specifically for patients with grade I degenerative lumbar spondylolisthesis. We observed that both procedures were associated with similar significant improvements in ODI, NRS-BP, NRS-LP, and EQ-5D with no differences in the odds of reaching the ODI MCID and the highest odds of NASS satisfaction in the long term. Our findings provide prospective, large sample size support, with a high rate of follow-up, to a prior retrospective investigation of 2- and 10-year outcomes for MI and open TLIF in a population undergoing single-level lumbar spondylolisthesis surgery by a single surgeon. Comparing 108 MI-TLIF with 53 open TLIF—with 52 (48%) and 31 (58%) reaching the 10-year follow-up, respectively—Kwon et al.40 observed an initial superiority of MI-TLIF 2 years postoperatively but similar ODI and visual analog scale (VAS) back and leg pain scores at 10 years. They also observed no difference in radiographic fusion or reoperation. Similarly, our registry cohort demonstrated superior ODI outcomes for MI-TLIF at 24 months but not in the extended follow-up. Given our additional studied outcome measures, we were able to extend this 24-month versus 5- to 10-year outcome discrepancy to EQ-5D and NASS satisfaction as well (since, in our 24-month analysis, there was an initial 24-month superiority of EQ-5D and NASS satisfaction for MI-TLIF). Overall, the equivalency in outcomes despite our high 60-month follow-up rates (75.6%–86.1%) suggests that the long-term outcome parity observed by Kwon et al. may not be entirely due to a loss to follow-up. In the treatment of spondylolisthesis, there is no evidence to suggest that either procedure is differentially susceptible to clinical outcome deterioration over an extended follow-up.
Ultimately, these long-term studies extend the findings of multiple shorter-term studies of operative spondylolisthesis, which demonstrate equivalency in clinical outcomes for MI and open TLIF21–27,39,41 (except for Wu et al., who found superior 24-month pain outcomes for MI-TLIF28). Interestingly, in a meta-analysis of 6 studies with 394 patients undergoing single-level surgery for grades I and II spondylolisthesis (with 1- to 2-year follow-up), Qin et al.8 found a pooled association between MI-TLIF and a lower ODI (but similar VAS back pain scores). This finding suggests that the individual studies may have been underpowered. Outside of the short-term postoperative period, similar PROs for MI and open TLIF are also observed in most other studies in which nonspondylolisthesis-specific populations are studied.10–15,17,19,20,42,43
An important consideration in the comparison of surgical efficacy is durability. Here we report no significant difference in the cumulative reoperation rate between MI-TLIF (5.6%) and open TLIF (11.6%), consistent with the pooled analysis of 6 spondylolisthesis cohorts.8 Our rates are similar to those reported by other studies, including the one by Kwon et al., which had a minimum 10-year follow-up for their low-grade spondylolisthesis cohort (MI-TLIF 4.6% and open TLIF 5.7%).40 Additionally, we observed no difference in the comparative ability to achieve successful fusion (similar to multiple prior investigations8,23,26). In contrast to the other degenerative conditions often studied when comparing MI and open TLIF, a unique consideration during the treatment of spondylolisthesis is the achievement of listhesis reduction. We observed no difference in the ability of either procedure to achieve listhesis reduction. Still, there is evidence that listhesis reduction may not be associated with clinical outcomes following surgery for grade I degenerative lumbar spondylolisthesis.34 Nevertheless, concerns regarding decreased bony surface area for arthrodesis and heightened difficulty with listhesis reduction maneuvers during MI-TLIF are not supported by our data.
It is also important to consider current advances in MI-TLIF that may further alter the risk-benefit profile in future studies. First, as surgeons progress through the learning curve of MI-TLIF, operative times and outcomes are expected to further improve. For example, some reports associate percutaneous pedicle screw placement, as compared to open screw placement, with cranial facet joint violation.44 Leveraging techniques to minimize this violation will further decrease rates of adjacent-segment degeneration, thereby minimizing the need for reoperation. Second, one potential drawback of MI-TLIF for surgeons reliant on fluoroscopy includes the increased intraoperative fluoroscopic time compared to that with open TLIF.15,18 However, with the increased adoption of 3D-based navigation, augmented reality, and/or robotic platforms, this surgeon-associated risk may be minimized. Other surgeons have hypothesized that open TLIF may permit greater segmental and global lordosis correction than MI-TLIF.41 Interestingly, in a retrospective, single-center matched radiographic study (1:1) of 75 MI-TLIFs versus 75 open TLIFs in a population with grade I degenerative lumbar spondylolisthesis (average follow-up 55.3–57.8 weeks), Dibble et al.41 observed greater improvements in segmental lordosis following MI-TLIF. These authors suggest that technological innovations—among other operative techniques and sparing of the posterior tension band in MI-TLIF—may contribute to this finding. Still, there were no differences in global lordosis. Therefore, either approach, at the least, may be associated with similar radiographic outcomes with a meticulous surgical technique.
Taken together, our data, in the context of the literature, suggest that MI-TLIF is equivalent to open TLIF for long-term clinical outcomes. Given this equivalence in clinical outcomes, there is no strong evidence favoring one technique over the other. Rather, surgeons may employ either procedure based chiefly on other considerations, such as patient preference, surgeon skill set, specific radiographic and clinical characteristics, frailty, the need for expedited perioperative recovery or minimal blood loss, and the need to accommodate procedures without general anesthesia. It is worth noting that not every spine surgeon performs both procedures. For these surgeons in particular, our data suggest that either approach appears acceptable.
Study Limitations
This study is a retrospective analysis of a prospective cohort and thus is vulnerable to the associated biases. Specifically, given our use of a registry, there is no ability to control for surgeon and patient selection between open TLIF and MI-TLIF. Furthermore, though we provide a definition of MI and open TLIF, there may be variations in surgical technique (e.g., single vs bilateral facetectomies). Second, there were some key baseline differences between the cohorts including for BMI, motor deficits at presentation, and the use of workers’ compensation, and the findings should be interpreted accordingly. However, we adjusted for these factors in our multivariable analyses. Third, as with any longitudinal study with extended follow-up, there were patients lost to follow-up.31 Importantly, our 60-month follow-up rate (75.6% and 86.1% for the open TLIF and MI-TLIF cohorts, respectively) is notable in the context of other high-quality, prospective studies on surgical outcomes for degenerative lumbar spondylolisthesis. For example, in an 8-year analysis of patients with degenerative lumbar spondylolisthesis from the Spine Patient Outcomes Research Trial (SPORT), data were obtained for only 69% of the randomized cohort and 57% of the observational cohort.45 Continued investigation of our cohort, at an even longer-term follow-up than presently described, will shed further light on the very long-term outcomes following MI and open TLIF for degenerative lumbar spondylolisthesis. Lastly, there are certain covariates (e.g., baseline radiographic parameters) and outcomes (e.g., opioid medication use postoperatively, cost analysis) that are important in a comparison of MI and open TLIF but outside the scope of the current investigation. Specifically, though we included several radiographic outcomes (millimeters of listhesis reduction, radiographic fusion, reoperations for adjacent-segment disease, instrumentation revision, pseudarthrosis, and subsidence), we did not include other variables such as segmental or regional radiographic parameters and radiographic subsidence. Still, we hope that by capturing reoperations, we have included clinically relevant radiographic outcomes.
Conclusions
In a 60-month comparison of a prospective registry cohort of 72 and 225 patients who had undergone MI and open TLIF for grade I degenerative lumbar spondylolisthesis, respectively, we did not observe any differences for disability, back pain, leg pain, quality of life, patient satisfaction, and cumulative reoperation. Both procedures were associated with mean improvements in each measured PRO, compared to baseline. Therefore, there is no compelling evidence supporting the use of one technique over the other.
Acknowledgments
This research was supported by the NeuroPoint Alliance (NPA), the Neurosurgery Research & Education Foundation (NREF), and the Spine Section. The NPA is a 501(c)(6) affiliate nonprofit organization of the American Association of Neurological Surgeons (AANS) dedicated to the improvement of the quality of care in neurosurgical practice via the institution of national quality registries, such as the one utilized for this study. The NREF is the philanthropic arm of the AANS and has financially supported the creation and maintenance of the QOD. The Spine Section is a neurosurgical community formed in collaboration between the AANS and the Congress of Neurological Surgeons to advance spine and peripheral nerve patient care through education, research, and advocacy. Dr. Glassman reported institutional grants from the Scoliosis Research Society. Dr. Knightly reported nonfinancial support from NPA during the conduct of this study. Dr. P. Park reported grants from the International Spine Study Group (ISSG). Dr. Virk reported grants from the NIH outside the submitted work. Dr. Mummaneni reported grants from the NREF during the conduct of the study and grants from the NREF, ISSG, and NIH.
Disclosures
Dr. Bisson reported personal fees from Stryker, Medtronic, MiRus, and Proprio outside the submitted work. Dr. Glassman reported personal fees from Medtronic and K2M/Stryker outside the submitted work; a patent for Medtronic with royalties paid; institutional grants from Pfizer, Texas Scottish Rite Hospital, the Alan L. and Jacqueline B. Stuart Spine Research Foundation, Cerapedics, Medtronic, and Empirical Spine; and a chair position with the American Spine Registry. Dr. Foley reported personal fees from Medtronic outside the submitted work and a patent for instruments for stabilization of bony structures with royalties paid from Medtronic. Dr. C. Shaffrey reported personal fees from NuVasive, Medtronic, SI Bone, and Proprio outside the submitted work. Dr. Potts reported personal fees from Medtronic outside the submitted work. Dr. Coric reported personal fees from Medtronic, Globus Medical, Spine Wave, and Premia Spine outside the submitted work. Dr. P. Park reported personal fees and royalties from Globus; personal fees from NuVasive, Accelus, DePuy Synthes, and LifeNet; grants from Cerapedics, SI Bone, and Depuy Synthes; and travel and lodging fees from Medtronic outside the submitted work. Dr. Wang reported personal fees from DePuy Synthes, Stryker, Spineology, and Surgalign; and stock in Kinesiometrics, Medical Device Partners, and Innovative Surgical Designs outside the submitted work. Dr. Fu reported personal fees from Medtronic and DePuy Synthes outside the submitted work. Dr. Virk reported education from and consultancy for DePuy Synthes and stock in and consultancy for OnPoint Surgical. Dr. Agarwal reported royalties from Thieme Medical Publishers and Springer International Publishing. Dr. Chou reported personal fees from Orthofix and Globus outside the submitted work. Dr. Mummaneni reported personal fees from DePuy Synthes, Globus, NuVasive, and Stryker outside the submitted work; stock in Spinicity/ISD; grants from AO Spine; and royalties from Thieme Publishers and Springer Publishers.
Author Contributions
Conception and design: Chan, Glassman, CI Shaffrey, Potts, Knightly, P Park, Slotkin, Asher, Virk, Haid, Mummaneni. Acquisition of data: Chan, Bisson, Glassman, Foley, CI Shaffrey, Potts, ME Shaffrey, Coric, Knightly, P Park, Wang, Fu, Asher, Virk, Guan, Agarwal, Mummaneni. Analysis and interpretation of data: Chan, Foley, ME Shaffrey, Wang, Asher, Agarwal, Mummaneni. Drafting the article: Chan, CI Shaffrey, Virk, Haid, Agarwal. Critically revising the article: Chan, Bydon, Bisson, Glassman, Foley, CI Shaffrey, Potts, ME Shaffrey, P Park, Fu, Haid, Agarwal, Mummaneni. Reviewed submitted version of manuscript: Chan, Bydon, Bisson, Glassman, Foley, CI Shaffrey, Potts, Coric, Wang, Fu, Slotkin, Asher, Michalopoulos, Guan, Agarwal, Chou. Approved the final version of the manuscript on behalf of all authors: Chan. Statistical analysis: Chan. Administrative/technical/material support: Glassman, Potts, Wang, Michalopoulos, C Park. Study supervision: Bydon, Bisson, Potts, Knightly, Mummaneni.
Supplemental Information
Previous Presentations
This work was presented as an oral presentation at the Society for Minimally Invasive Spine Surgery Annual Forum held in Las Vegas, Nevada, on September 29–October 1, 2022.
References
- 1↑
Kalichman L, Kim DH, Li L, Guermazi A, Berkin V, Hunter DJ. Spondylolysis and spondylolisthesis: prevalence and association with low back pain in the adult community-based population. Spine (Phila Pa 1976). 2009;34(2):199–205.
- 2↑
Weinstein JN, Lurie JD, Tosteson TD, et al. Surgical versus nonsurgical treatment for lumbar degenerative spondylolisthesis. N Engl J Med. 2007;356(22):2257–2270.
- 3↑
Foley KT, Lefkowitz MA. Advances in minimally invasive spine surgery. Clin Neurosurg. 2002;49:499–517.
- 4↑
Harms J, Rolinger H. A one-stager procedure in operative treatment of spondylolistheses: dorsal traction-reposition and anterior fusion. Article in German. Z Orthop Ihre Grenzgeb. 1982;120(3):343–347.
- 5↑
Kawaguchi Y, Matsui H, Tsuji H. Back muscle injury after posterior lumbar spine surgery. A histologic and enzymatic analysis. Spine (Phila Pa 1976). 1996;21(8):941–944.
- 6↑
Kawaguchi Y, Yabuki S, Styf J, et al. Back muscle injury after posterior lumbar spine surgery. Topographic evaluation of intramuscular pressure and blood flow in the porcine back muscle during surgery. Spine (Phila Pa 1976). 1996;21(22):2683–2688.
- 7↑
Chan AK, Bisson EF, Bydon M, et al. A comparison of minimally invasive and open transforaminal lumbar interbody fusion for grade 1 degenerative lumbar spondylolisthesis: an analysis of the prospective Quality Outcomes Database. Neurosurgery. 2020;87(3):555–562.
- 8↑
Qin R, Liu B, Zhou P, et al. Minimally invasive versus traditional open transforaminal lumbar interbody fusion for the treatment of single-level spondylolisthesis grades 1 and 2: a systematic review and meta-analysis. World Neurosurg. 2019;122:180–189.
- 9↑
Shunwu F, Xing Z, Fengdong Z, Xiangqian F. Minimally invasive transforaminal lumbar interbody fusion for the treatment of degenerative lumbar diseases. Spine (Phila Pa 1976). 2010;35(17):1615–1620.
- 10↑
Adogwa O, Carr K, Thompson P, et al. A prospective, multi-institutional comparative effectiveness study of lumbar spine surgery in morbidly obese patients: does minimally invasive transforaminal lumbar interbody fusion result in superior outcomes? World Neurosurg. 2015;83(5):860–866.
- 11
Cheng JS, Park P, Le H, Reisner L, Chou D, Mummaneni PV. Short-term and long-term outcomes of minimally invasive and open transforaminal lumbar interbody fusions: is there a difference? Neurosurg Focus. 2013;35(2):E6.
- 12
Peng CW, Yue WM, Poh SY, Yeo W, Tan SB. Clinical and radiological outcomes of minimally invasive versus open transforaminal lumbar interbody fusion. Spine (Phila Pa 1976). 2009;34(13):1385–1389.
- 13
Price JP, Dawson JM, Schwender JD, Schellhas KP. Clinical and radiologic comparison of minimally invasive surgery with traditional open transforaminal lumbar interbody fusion: a review of 452 patients from a single center. Clin Spine Surg. 2018;31(2):E121–E126.
- 14
Rodríguez-Vela J, Lobo-Escolar A, Joven E, Muñoz-Marín J, Herrera A, Velilla J. Clinical outcomes of minimally invasive versus open approach for one-level transforaminal lumbar interbody fusion at the 3- to 4-year follow-up. Eur Spine J. 2013;22(12):2857–2863.
- 15↑
Seng C, Siddiqui MA, Wong KP, et al. Five-year outcomes of minimally invasive versus open transforaminal lumbar interbody fusion: a matched-pair comparison study. Spine (Phila Pa 1976). 2013;38(23):2049–2055.
- 16
Singh K, Nandyala SV, Marquez-Lara A, et al. A perioperative cost analysis comparing single-level minimally invasive and open transforaminal lumbar interbody fusion. Spine J. 2014;14(8):1694–1701.
- 17↑
Tian W, Xu YF, Liu B, et al. Computer-assisted minimally invasive transforaminal lumbar interbody fusion may be better than open surgery for treating degenerative lumbar disease. Clin Spine Surg. 2017;30(6):237–242.
- 18↑
Wong AP, Smith ZA, Stadler JA III, et al. Minimally invasive transforaminal lumbar interbody fusion (MI-TLIF): surgical technique, long-term 4-year prospective outcomes, and complications compared with an open TLIF cohort. Neurosurg Clin N Am. 2014;25(2):279–304.
- 19↑
Zairi F, Arikat A, Allaoui M, Assaker R. Transforaminal lumbar interbody fusion: comparison between open and mini-open approaches with two years follow-up. J Neurol Surg A Cent Eur Neurosurg. 2013;74(3):131–135.
- 20↑
Zhang W, Li X, Shang X, et al. Modified minimally invasive transforaminal lumbar interbody fusion using a trans-multifidus approach: a safe and effective alternative to open-TLIF. J Orthop Surg Res. 2015;10:93.
- 21↑
Archavlis E, Carvi y Nievas M. Comparison of minimally invasive fusion and instrumentation versus open surgery for severe stenotic spondylolisthesis with high-grade facet joint osteoarthritis. Eur Spine J. 2013;22(8):1731–1740.
- 22↑
Parker SL, Mendenhall SK, Shau DN, et al. Minimally invasive versus open transforaminal lumbar interbody fusion for degenerative spondylolisthesis: comparative effectiveness and cost-utility analysis. World Neurosurg. 2014;82(1-2):230–238.
- 23↑
Saetia K, Phankhongsab A, Kuansongtham V, Paiboonsirijit S. Comparison between minimally invasive and open transforaminal lumbar interbody fusion. J Med Assoc Thai. 2013;96(1):41–46.
- 24↑
Serban D, Calina N, Tender G. Standard versus minimally invasive transforaminal lumbar interbody fusion: a prospective randomized study. BioMed Res Int. 2017;2017:7236970.
- 25↑
Sulaiman WA, Singh M. Minimally invasive versus open transforaminal lumbar interbody fusion for degenerative spondylolisthesis grades 1-2: patient-reported clinical outcomes and cost-utility analysis. Ochsner J. 2014;14(1):32–37.
- 26↑
Wang J, Zhou Y, Zhang ZF, Li CQ, Zheng WJ, Liu J. Comparison of one-level minimally invasive and open transforaminal lumbar interbody fusion in degenerative and isthmic spondylolisthesis grades 1 and 2. Eur Spine J. 2010;19(10):1780–1784.
- 27↑
Wu AM, Hu ZC, Li XB, et al. Comparison of minimally invasive and open transforaminal lumbar interbody fusion in the treatment of single segmental lumbar spondylolisthesis: minimum two-year follow up. Ann Transl Med. 2018;6(6):105.
- 28↑
Wu MH, Dubey NK, Li YY, et al. Comparison of minimally invasive spine surgery using intraoperative computed tomography integrated navigation, fluoroscopy, and conventional open surgery for lumbar spondylolisthesis: a prospective registry-based cohort study. Spine J. 2017;17(8):1082–1090.
- 29↑
Asher AL, Kerezoudis P, Mummaneni PV, et al. Defining the minimum clinically important difference for grade I degenerative lumbar spondylolisthesis: insights from the Quality Outcomes Database. Neurosurg Focus. 2018;44(1):E2.
- 30
Bisson EF, Guan J, Bydon M, et al. Patient-reported outcome improvements at 24-month follow-up after fusion added to decompression for grade I degenerative lumbar spondylolisthesis: a multicenter study using the Quality Outcomes Database. J Neurosurg Spine. 2021;35(1):42–51.
- 31↑
Bisson EF, Mummaneni PV, Knightly J, et al. Assessing the differences in characteristics of patients lost to follow-up at 2 years: results from the Quality Outcomes Database study on outcomes of surgery for grade I spondylolisthesis. J Neurosurg Spine. 2020;33(5):643–651.
- 32
Chan AK, Bisson EF, Bydon M, et al. A comparison of minimally invasive transforaminal lumbar interbody fusion and decompression alone for degenerative lumbar spondylolisthesis. Neurosurg Focus. 2019;46(5):E13.
- 33
Chan AK, Bisson EF, Bydon M, et al. Laminectomy alone versus fusion for grade 1 lumbar spondylolisthesis in 426 patients from the prospective Quality Outcomes Database. J Neurosurg Spine. 2018;30(2):234–241.
- 34↑
Chan AK, Mummaneni PV, Burke JF, et al. Does reduction of the Meyerding grade correlate with outcomes in patients undergoing decompression and fusion for grade I degenerative lumbar spondylolisthesis? J Neurosurg Spine. 2022;36(2):177–184.
- 35↑
Meyerding HW. Diagnosis and roentgenologic evidence in spondylolisthesis. Radiology. 1933;20(2):108–120.
- 36↑
McGirt MJ, Speroff T, Dittus RS, Harrell FE Jr, Asher AL. The National Neurosurgery Quality and Outcomes Database (N2QOD): general overview and pilot-year project description. Neurosurg Focus. 2013;34(1):E6.
- 37↑
Amin BY, Tu TH, Mummaneni PV. Mini-open transforaminal lumbar interbody fusion. Neurosurg Focus. 2013;35(2 Suppl): Video 2.
- 38↑
Dhall SS, Wang MY, Mummaneni PV. Clinical and radiographic comparison of mini-open transforaminal lumbar interbody fusion with open transforaminal lumbar interbody fusion in 42 patients with long-term follow-up. J Neurosurg Spine. 2008;9(6):560–565.
- 39↑
Adogwa O, Parker SL, Bydon A, Cheng J, McGirt MJ. Comparative effectiveness of minimally invasive versus open transforaminal lumbar interbody fusion: 2-year assessment of narcotic use, return to work, disability, and quality of life. J Spinal Disord Tech. 2011;24(8):479–484.
- 40↑
Kwon JW, Park Y, Lee BH, et al. Ten-year outcomes of minimally invasive versus open transforaminal lumbar interbody fusion in patients with single-level lumbar spondylolisthesis. Spine (Phila Pa 1976). 2022;47(11):773–780.
- 41↑
Dibble CF, Zhang JK, Greenberg JK, et al. Comparison of local and regional radiographic outcomes in minimally invasive and open TLIF: a propensity score-matched cohort. J Neurosurg Spine. 2022;37(3):384–394.
- 42↑
Garg B, Mehta N. Minimally invasive transforaminal lumbar interbody fusion (MI-TLIF): A review of indications, technique, results and complications. J Clin Orthop Trauma. 2019;10(suppl 1):S156–S162.
- 43↑
Heemskerk JL, Oluwadara Akinduro O, Clifton W, Quiñones-Hinojosa A, Abode-Iyamah KO. Long-term clinical outcome of minimally invasive versus open single-level transforaminal lumbar interbody fusion for degenerative lumbar diseases: a meta-analysis. Spine J. 2021;21(12):2049–2065.
- 44↑
Babu R, Park JG, Mehta AI, et al. Comparison of superior-level facet joint violations during open and percutaneous pedicle screw placement. Neurosurgery. 2012;71(5):962–970.
- 45↑
Abdu WA, Sacks OA, Tosteson ANA, et al. Long-term results of surgery compared with nonoperative treatment for lumbar degenerative spondylolisthesis in the Spine Patient Outcomes Research Trial (SPORT). Spine (Phila Pa 1976). 2018;43(23):1619–1630.
