- Email: [email protected]
Does approach matter? A comparative radiographic analysis of spinopelvic parameters in single-level lumbar fusion
Accepted Manuscript Title: Does approach matter? a comparative radiographic analysis of spinopelvic parameters in single level lumbar fusion. Author: Seth Ahlquist, Howard Y. Park, Jonathan Gatto, Ayra N. Shamie, Don Y. Park PII: DOI: Reference:
S1529-9430(18)30125-6 https://doi.org/10.1016/j.spinee.2018.03.014 SPINEE 57633
To appear in:
The Spine Journal
Received date: Revised date: Accepted date:
8-12-2017 19-3-2018 26-3-2018
Please cite this article as: Seth Ahlquist, Howard Y. Park, Jonathan Gatto, Ayra N. Shamie, Don Y. Park, Does approach matter? a comparative radiographic analysis of spinopelvic parameters in single level lumbar fusion., The Spine Journal (2018), https://doi.org/10.1016/j.spinee.2018.03.014. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
Does Approach Matter? A Comparative Radiographic Analysis of Spinopelvic Parameters in Single Level Lumbar Fusion.
Seth Ahlquist BS1, Howard Y. Park, MD1, Jonathan Gatto, BS1, Ayra N. Shamie, MD1, Don Y. Park MD1 1. Department of Orthopaedic Surgery, David Geffen School of Medicine at UCLA, 1250 16th Street, Santa Monica, CA, 90404, USA
Corresponding Author Don Y. Park, MD Department of Orthopaedic Surgery David Geffen School of Medicine at UCLA, 1250 16th Street Santa Monica, CA, 90404, USA Email: [email protected] Phone: 424-259-9829 Fax: 424-259-6599
Abstract BACKGROUND CONTEXT: Lumbar fusion is a popular and effective surgical option to provide stability and restore anatomy. Particular attention has recently been focused on sagittal alignment and radiographic spinopelvic parameters that apply to lumbar fusion as well as spinal deformity cases. Current literature has demonstrated the effectiveness of various techniques of lumbar fusion, however comparative data of these techniques is limited. PURPOSE: To directly compare the impact of various lumbar fusion techniques (ALIF, LLIF, TLIF, PLF) based on radiographic parameters. STUDY DESIGN/SETTING: A single-center retrospective study examining pre-operative and post-operative radiographs. PATIENT SAMPLE: A consecutive list of lumbar fusion surgeries performed by multiple spine surgeons at a single institution from 2013-2016 were identified. 1 Page 1 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
OUTCOME MEASURES: Radiographic measurements utilized included segmental lordosis (SL), lumbar lordosis (LL), pelvic incidence (PI), pelvic incidence-lumbar lordosis mismatch (PI-LL), anterior and posterior disk height (DH-A, DH-P respectively), and foraminal height (FH). METHODS: Radiographic measurements were performed on pre-operative and post-operative lateral lumbar radiographs on all single-level lumbar fusion cases. Demographic data was collected including age, gender, approach, diagnosis, surgical level, and implant lordosis. Paired sample t-test, one-way ANOVA, McNemar Test, and independent sample t-test were used to establish significant differences in the outcome measures. Multiple linear regression was performed to determine a predictive model for lordosis from implant lordosis, fusion technique, and surgical level. RESULTS: There were 164 patients (78 males, 86 females) with a mean age of 60.1 years and average radiographic follow up time of 9.3 months. These included 34 ALIF, 23 LLIF, 63 TLIF, and 44 PLF surgeries. ALIF and LLIF significantly improved SL (7.9° & 4.4°), LL (5.5° & 7.7°), DH-A (8.8 mm & 5.8 mm), DH-P (3.4 mm & 2.3 mm), and FH (2.8 mm & 2.5 mm), respectively (p ≤ .003). TLIF significantly improved these parameters, albeit to a lesser extent: SL (1.7°), LL (2.7°), DH-A (1.1 mm), DH-P (0.8 mm), and FH (1.1 mm), p ≤ .02. PLF did not significantly alter any of these parameters while significantly reducing FH (-1.3 mm, p = .01). One-way ANOVA showed no significant differences between ALIF and LLIF other than ALIF with greater ΔDH-A (3.0 mm, p = .02). Both ALIF and LLIF significantly outperformed PLF in pre-operative to post-operative change in all parameters p ≤ .001. Additionally, ALIF significantly outperformed TLIF in the change in SL (6.2°, p < .001) and LLIF significantly outperformed TLIF in the change in LL (5.0°, p = .02). Both outperformed TLIF in ΔDH-A (7.7 mm & 4.7 mm) and ΔDH-P (2.6 mm & 1.5 mm), respectively (p ≤ .02). ALIF was the only fusion technique that significantly improved the proportion of patients with a PI-LL < 10° (0.41 to 0.66, p = .02). Lordotic cages had superior improvement of all parameters as compared to non-lordotic cages (p <.001). Implant lordosis (m = 1.1), fusion technique (m = 6.8), and surgical level (m = 6.9) significantly predicted post-operative SL (p < .001, R2 = .56). CONCLUSIONS: This study demonstrated that these four lumbar fusion techniques yield divergent radiographic results. ALIF and LLIF produced greater improvements in radiographic measurements post-operatively as compared to TLIF and PLF. ALIF was the most successful in improving PI-LL mismatch, an important parameter relating to sagittal alignment. Lordotic implants provided better sagittal correction and surgeons should be cognizant of the impact that these differing implants and techniques produce after surgery. Surgical technique is an important determinant of post-operative alignment and has ramifications upon sagittal alignment in lumbar fusion surgery. Keywords: Lumbar interbody fusion, ALIF, LLIF, TLIF, PLF, sagittal alignment, segmental lordosis, lumbar lordosis, PI-LL
2 Page 2 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
Introduction Degenerative conditions of the lumbar spine are becoming increasingly prevalent with an aging population and result in significant reductions to quality of life due to immobility, radicular pain, and muscle spasm. Lumbar fusion through various techniques is an effective intervention for restoring the stability of these spinal segments, alleviating compression of neural elements, and reestablishing spinal anatomy. The critical importance of sagittal alignment and pelvic parameters is well established in spinal deformity and lumbar degenerative surgery alike. High pelvic incidence-lumbar lordosis (PI-LL) mismatch (greater than 10°) is associated with adjacent segment degeneration (ASD) and a tenfold risk of revision surgery [1-3], worsened post-operative residual symptoms [4], reduced quality of life [5], severe disability [6, 7], and decreased recovery rate [8]. Similarly, lower postoperative lumbar lordosis (LL) is significantly associated with increased ASD [3], disability [9], and pain [10]. There are multiple studies in the literature that examine the effects of the various lumbar fusion techniques have on spinopelvic parameters, albeit with conflicting results. Transforaminal lumbar interbody fusion (TLIF) has been shown to improve segmental lordosis (SL) [11, 12] and LL [13, 14] in previous studies. However, a recent retrospective study and randomized controlled trial demonstrated conflicting results as TLIF did not result in superior SL improvement over posterolateral fusion (PLF) [15, 16]. Lateral lumbar interbody fusion (LLIF) was initially shown to only improve SL [17-21], however later studies revealed the technique can also improve LL [22-26]. Anterior lumbar interbody fusion (ALIF) outperformed TLIF in the improvement of SL and LL in multiple studies [27-31]. However, few studies directly compared the radiographic outcomes of all the available lumbar fusion techniques together, and none included PI-LL mismatch as an outcome measure. Watkins et al compared ALIF, LLIF, and TLIF and determined that only ALIF and LLIF significantly improved SL [32]. Sembrano et al compared ALIF, LLIF, TLIF, and PLF and found that the former three approaches increased SL and LL, with ALIF to the greatest extent, emphasizing the importance of interbody cages [33]. A recent systematic review found that ALIF had significantly higher postoperative SL than LLIF, TLIF, and PLIF [34]. The geometry of the interbody cage itself may influence sagittal alignment, with wedge-shaped lordotic cages becoming more commonly used over non-lordotic cages since the lordotic cages can significantly increase SL and LL [35]. Lordotic cages in LLIF have been shown to provide a significant increase in SL (p =.01) while non-lordotic cages do not (p = .71) [36-37]. Hyperlordotic cages and ALL resection may further impact sagittal alignment as placement of increasingly lordotic cages without the restraint of the ALL led to greater gains in SL and LL [38-41]. In addition to cage geometry, cage position and insertion level are factors that may also influence the change in SL [42]. The literature regarding the various lumbar fusion techniques and interbody cages has demonstrated satisfactory fusion rates and clinical success; however, their comparative effects on sagittal alignment and pelvic parameters, including SL, LL, and PI-LL mismatch remain unclear. 3 Page 3 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
The goal of this study was to compare the capacities of ALIF, LLIF, TLIF, and PLF to change SL, LL, PI-LL mismatch, disk height (DH), and foraminal height (FH), as well as to develop a predictive model for lordosis based on the surgical approach, insertion level, and cage lordosis. Materials and Methods A consecutive list of lumbar fusion surgeries performed by multiple spine surgeons at a single institution from 2013-2016 were identified and retrospectively analyzed with Institutional Review Board approval (IRB#16-000175). Inclusion criteria included all single level lumbar fusion surgeries (n = 164). Exclusion criteria included revision surgeries, multilevel fusions, inclusion of spinal osteotomies, fusion performed for spinal trauma, tumor, and infections (n = 36). Osteotomies were specifically excluded to reduce the variables examined in this study and focus the investigation to the lumbar fusion techniques themselves. 164 patients fit the inclusion criteria while 36 patients were excluded from the study. Radiographic measurements were performed on pre-operative and post-operative neutral, lateral lumbar radiographs on all single-level lumbar fusion cases using the PACS system. These measurements were performed independently by two trained observers. Radiographic measurements obtained included segmental lordosis (SL), lumbar lordosis (LL), pelvic incidence (PI), pelvic incidence-lumbar lordosis mismatch (PI-LL), disk height (DH), and foraminal height (FH) (Fig 1). Post-op radiographs were obtained at the standardized time points (i.e. 6 weeks, 3 months, 6 months, 1 year post-surgery), although there was some variability in these time points based on the patient’s followup. Demographic data was collected including age, gender, surgical approach, diagnosis, surgical level, and implant lordosis. Paired sample t-test, one-way ANOVA with Tukey/Games Howell post hoc analysis, exact McNemar Test, and independent sample t-test were used to establish significant differences in the outcome measures. Multiple linear regression was performed to determine a predictive model for post-operative segmental lordosis from implant lordosis, fusion technique, and surgical level. Comparisons between groups and within groups were deemed statistically significant at the p < .05 threshold. Statistical analyses were performed using SPSS 24 software (IBM Corp., Armonk, NY, USA).
Results Demographic data There were 164 patients [78 males (47.6%), 86 females (52.4%)] with a mean age of 60.1 years (range 25-88) and average radiographic follow up time of 9.3 months. These patients underwent 34 ALIF (20.7%), 23 LLIF (14.0%), 63 TLIF (38.4%), and 44 PLF surgeries (26.8%). Follow up time was not significantly different for any of the approaches (p = .10). The most common primary indication for surgery was spondylolisthesis (61.0%), followed by degenerative disk disease (15.2%) and isthmic spondylolisthesis (9.1%) (Table 1). All techniques had spondylolisthesis as the most frequent indication for surgery, except for LLIF which had adjacent
4 Page 4 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
level disease. Also of note is that ALIF had the highest proportion of degenerative disk disease cases (41.1%) and TLIF the highest proportion of isthmic spondylolisthesis cases (19.0%). Change in Radiographic Parameters by Surgical Technique Paired sample t-test was performed to examine whether each individual surgical approach significantly altered the aforementioned radiographic parameters from the pre-operative to postoperative radiographs. ALIF and LLIF significantly improved SL (7.9° & 4.4°, respectively; p < .001), LL (5.5° & 7.7°, respectively; p < .001), DH-A (8.8 mm & 5.8 mm, respectively; p < .001), DH-P (3.4 mm & 2.3 mm, respectively; p < .001), and FH (2.8 mm & 2.5 mm, respectively; p <.001 and p =.003) (Fig. 2A, 2B). TLIF significantly improved these parameters as well, however to a lesser extent: SL (1.7°, p = .02), LL (2.7°, p = .01), DH-A (1.1 mm, p = .002), DH-P (0.8 mm, p = .002), and FH (1.1 mm, p = .002) (Fig. 2C). PLF did not significantly alter SL (-0.3°, p = .65), LL (-0.5°, p = .66), DH-A (0.1 mm, p = .75), or DH-P (0.1 mm, p = .70), and significantly decreased FH (-1.3 mm, p = .01) (Fig. 2D). Contribution of Disk Height to Lordosis Overall disk height was analyzed to determine the contribution of pre-operative disk height to the change in lordosis post-operatively using an average of the anterior and posterior disk height values across all surgical techniques. A halfway cutoff of 6.0mm was determined to differentiate between a more collapsed disk height group (n=78) and a less collapsed disk height group (n=86) (Table 6). SL improvement was significantly greater in the more collapsed disk height group than the less collapsed disk height group (p < .001). No significant difference was detected in LL improvement between the two groups (p=0.34). Comparison of Radiographic Parameter Changes Between Surgical Techniques A one-way ANOVA was utilized to determine whether a particular surgical approach significantly outperformed the others in its ability to improve the radiographic parameters of SL, LL, DH-A/P and FH. No significant differences were found between ALIF and LLIF in these parameters except for ALIF having a significantly greater ΔDH-A (3.0 mm, p = .02) (Fig. 3C). Both ALIF and LLIF significantly outperformed PLF in pre-operative to post-operative change in SL (8.2° & 4.7°), LL (6.0° & 8.2°), DH-A (8.9 mm & 5.9 mm), DH-P (3.5 mm & 2.4 mm), and FH (4.1 mm & 3.8 mm), respectively, with all p ≤ .001 (Fig. 3A-D). Additionally, ALIF significantly outperformed TLIF in the change in SL (6.2°, p < .001) and LLIF significantly outperformed TLIF in the change in LL (5.0, p = .02). Both outperformed TLIF in ΔDH-A (7.7 mm & 4.7 mm, respectively; p ≤ .001) and ΔDH-P (2.6 mm, p <.001 & 1.5 mm, p =.02, respectively) (Fig. 3A-D). TLIF was significantly more successful than PLF in terms of FH improvement (2.4 mm, p = .001) (Fig. 3D) but not compared to the ALIF and LLIF. Capacity of Surgical Techniques to Increase Proportion of Patients with a PI-LL < 10° McNemar Test was performed to see whether any of the surgical techniques significantly altered the proportion of patients with a PI-LL < 10° from pre-op to post-op. ALIF was the only technique to significantly improve the proportion of patients with a PI-LL < 10° (0.41 to 0.66, p = .02), while the other techniques did not significantly alter the pre-operative proportions: LLIF 5 Page 5 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
0.55 to 0.59 (p = 1.0), TLIF 0.46 to 0.56 (p = .21), and PLF 0.43 to 0.50 (p = .38) (Fig. 4). Only one patient who had a pre-operative PI-LL of < 10 who underwent ALIF ended up having a PILL > 10° after surgery. Comparison of Radiographic Parameter Changes between Lordotic and Non-Lordotic Cages 58 of the cages implanted during surgery were lordotic and 106 were non-lordotic (Table 2). All of the ALIF procedures utilized a lordotic implant. The majority of LLIF procedures utilized a lordotic implant while the majority of TLIF procedures did not. Independent sample t-test was performed in order to determine whether there were significant differences in the ability of lordotic cages to improve radiographic parameters over that of non-lordotic cages. Analysis determined that lordotic cages improved SL (5.0°), LL (4.7°), DH-A (6.3 mm), DH-P (1.2 mm), and FH (2.3 mm) to a significantly greater degree as compared to non-lordotic cages, with all p < .001 (Fig. 5). Predictive Equation for Post-Operative Segmental Lordosis The most common surgical level overall for lumbar fusion during the study period was L4-5 (50.6%), however, LLIF was most frequently performed at the L3-4 level (15.9% overall), and ALIF was most frequently performed at the L5-S1 level (26.8% overall) (Table 3). ALIF, LLIF, and TLIF had different profiles of lordotic cages used with the majority of ALIF being > 8° (11.3 ± 3.3°), LLIF being < 8° (5.6 ± 3.2°), and TLIF being = 0° (0.6 ± 2.1°). Multiple linear regression was performed on all cases with lordotic implants (n = 58) to derive a predictive equation for post-operative SL from implant lordosis, surgical level, and surgical approach. The multiple regression model statistically significantly predicted SLpost-op, F(3, 54) = 22.753, p < .001, R2 = .56 (Eq.1). ALIF was the only technique that statistically significantly increased SLpost-op, and thus surgical approach is accounted for as either ALIF (approach = 1) or not ALIF (approach = 0). Similarly, fusions performed at the L4-5 and L5-S1 levels (surgical level = 1) afforded significantly higher SLpostop than those performed at L2-3 and L3-4 (surgical level = 0). Equation 1. SLpost-op = 1.1*(Implant Lordosis) + 6.8*(Approach) + 6.9*(Surgical Level) + 11.4
All three variables added significantly to the prediction, p < .05. Regression coefficients and standard errors can be found in Table 5.
Discussion Lumbar fusion is the 5th most common surgical procedure performed in United States hospitals, with over 450,000 cases performed annually [43]. Controversy exists over which lumbar fusion technique is optimal in providing the greatest improvement of radiographic parameters. Although these techniques can achieve fusion, there are differences in each technique’s ability to affect sagittal alignment, an important predictor of clinical success. There are few studies that directly compare all the available surgical techniques for lumbar fusion and no prior studies investigating PI-LL mismatch as an outcome measure, which has been repeatedly shown to 6 Page 6 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
correlate with clinical outcomes. This study aimed to compare the radiographic outcomes of various techniques performed to achieve single level lumbar fusion. This study demonstrated that ALIF, LLIF, and TLIF techniques can each alter and influence sagittal alignment, although in different ways. We found that ALIF improved SL by 7.9°, p < .001 (Fig. 2A), while prior studies found lower magnitude changes in SL, ranging from 3.7° to 4.5° [30-33] (p < .05). We were able to achieve greater improvements in SL with the ALIF technique, which may be due to greater attention to sagittal alignment by the surgeons and greater availability of lordotic cages. ALIF significantly improved all radiographic measurements in this study, which is in line with the results from prior studies [30-34, 44]. LLIF improved SL by 4.4° in our results, p < .001 (Fig. 2B), as compared to 1.2° to 12.0° in prior studies [17- 20, 22-23, 26, 32-33] (p < .05). Manwaring et al. found a substantially higher change in SL (12.0°), which may be explained by the inclusion of anterior column release cases (ACR) that can substantially increase lordosis. In addition, the study was limited by a small sample size of 9 cases, and included multilevel fusions [26]. Our study of single level fusions found LLIF improved LL by 7.7°, p < .001 (Fig. 2B), which was in line with prior studies (2.5° to 12.2°) [22-24, 33]. A systematic review of LLIF demonstrated a ΔSL of 2.4° and ΔLL of 7.5° (p < .05), which is in line with our results that the LLIF technique is effective in changing sagittal alignment [25]. LLIF significantly improved DH-A, DH-P and FH, which is in agreement with prior studies [33, 45, 46]. Our study demonstrated that TLIF did not change the radiographic parameters to the same extent as ALIF or LLIF, which may be attributed to the vast majority of TLIF cages being non-lordotic. TLIF improved SL by only 1.7° in our study, p = .02 (Fig. 2C), as compared to prior studies ranging from 1.9 to 9.3° [11, 12, 30-31, 33]. Watkins et al. found a non-significant change in SL with TLIF measuring 0.8 (p > .05) [32]. In respect to the change in LL, we found that TLIF improved LL by 2.7, p = .01 (Fig. 2C), while other studies ranged from 2.1 to 10.5° [12, 13, 33] (p < .05). Kim et al. found non-significant changes with TLIF of -0.1 and 1.7 at the L4-5 (p = .98) and L5-S1 (p = .38) levels, respectively, which is critically important since the majority of lumbar lordosis occurs from L4 to S1 levels [31]. Overall, our results determined that TLIF underperformed radiographically as compared to the results of previous studies [30, 33]. In regards to PLF, we demonstrated that PLF was radiographically inferior in changing the sagittal alignment as compared to the other lumbar fusion techniques. Our results demonstrated that PLF did not significantly alter SL (-0.3, p = .65), LL (-0.5, p = .66), DH-A (0.1mm, p = .75), or DH-P (0.1mm, p = .70), and significantly decreased FH (-1.3, p = .01) (Fig. 2D). This is closely in line with work by Sembrano et al. who found a non-significant ΔSL of (0.7, p = .13), ΔLL (-0.5, p = .66), and ΔDH-A (0.3, p = .25), as well as a small but significant decrease in ΔDH-P (-0.4, p = .04) [33]. Posterior osteotomies with cantilevering compression techniques have been shown to improve lordosis in prior studies, which may influence the ability of TLIF to improve sagittal alignment [48]. No posterior osteotomies were included in our study as the aim of the study was to investigate the radiographic impact of the various lumbar fusion techniques themselves, which could have improved these radiographic results. However, our results demonstrate that TLIF and 7 Page 7 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
PLF as a standalone technique does not produce significant changes in sagittal alignment as compared to ALIF and LLIF. This may not be clinically relevant in patients who are already in balance pre-operatively, but surgeons should be cognizant of the limitations of these techniques in changing sagittal alignment. When comparing ALIF and LLIF, few studies in the literature have evaluated the improvement of radiographic parameters between these techniques. Watkins et al. found that ALIF significantly outperformed LLIF in SL improvement (2.3°, p < .05) [32]. Our study failed to find any difference in the changes of SL (p = .07) and LL (p = .64) when comparing ALIF and LLIF (Fig. 3A-B). Sembrano et al. also found no difference in changes in SL (p = .57) or LL (p = .17) [33]. Our results support the findings that ALIF and LLIF are statistically equivalent in changing the SL and LL from the pre-operative to post-operative value. However, the change of these parameters may not be as clinically important as the actual values of the parameters themselves. The actual value of SL and LL may provide satisfactory sagittal alignment and balance. There may be a potential confounding effect regarding the surgeon’s choice in selecting patients to undergo fusion with one technique versus another, including availability of vascular access surgeon, comfort and training with certain techniques, and the awareness of the pelvic parameters in preoperative planning. Severity of preoperative disk space narrowing and collapse may also influence decision making as one technique may be favored over another based on the severity of disk space collapse, especially posteriorly which may make TLIF difficult to perform. Our analysis of the pre-operative disk height and its contribution to lordosis demonstrated a significantly greater improvement in SL in the more collapsed disk height group than the less collapsed disk height group across all fusion techniques (p < .001). This reflects a greater change in the disk space after surgery in disk spaces that are more collapsed pre-operatively, thereby impacting lordosis to a greater degree. Incorporating the state of the disk space pre-operatively may be an important element in surgical decision-making, in addition to surgical technique and implant selection, to effectively improve alignment. This underscores the importance in preoperative planning in lumbar fusion surgeries beyond achieving fusion. Much attention has been placed on the importance of improving the PI-LL mismatch as it relates to improved clinical outcomes. We demonstrated that ALIF was the only technique that significantly increased the proportion of patients with a PI-LL < 10°, p = .02 (Fig. 4). Given the studies in the literature showing the importance of the PI-LL mismatch parameter [1-8], this is especially pertinent. Approximately 60% of lumbar lordosis is derived from the L4-S1 segments making these levels particularly important [47]. In this study, LLIF was more commonly used at the L3-4 level, which may explain why LLIF was less successful in restoring PI-LL balance. With greater attention to PI-LL mismatch by surgeons, the inferior results with LLIF and TLIF techniques may change, especially with lordotic cage designs, posterior osteotomy and compression techniques, and expandable cage technology. While the differences in SL and LL improvement between surgical techniques in our study were less than 10 degrees, these differences were statistically significant. These small differences may be clinically significant since pre-operative PI-LL mismatch values ranged from 9-13, and an increase in lordosis in several degrees may improve the post-operative PI-LL mismatch < 10°.
8 Page 8 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
While this result may seem intuitive, lordotic cages provided a significantly larger improvement in the radiographic parameters than non-lordotic cages (Fig. 5). Multiple prior studies have provided evidence for qualitative increases in SL with increasingly lordotic cages [38-42]. Interbody cages are now designed with lordosis with some providing hyperlordotic dimensions to provide even more sagittal correction. The vast majority of TLIF cages were non-lordotic in this study (6 lordotic TLIF cages/63 total TLIF cages) while the majority of ALIF and LLIF cages were lordotic. This may explain the significant difference in radiographic results between the various techniques. Based on our results, we derived a predictive model for post-operative SL as a function of implant lordosis, surgical level and surgical approach (Eq. 1). Notably, there was about one degree of SL gained for each degree increase in implant lordosis. As would be expected from our within-group analysis (Fig. 2A), ALIF was much more effective at increasing SL than either LLIF or TLIF (Table 5). There are greater increases in SL when the cage is implanted at the L4-5 or L5-S1 levels (+6.9°) than at the L2-3, L3-4 levels. It should be recognized that this equation will only be useful for predicting lordosis for surgeries in which osteotomies are not performed since the study did not include osteotomies in the analysis. This model has the potential to be a useful tool for spine surgeons to utilize in surgical planning. It will be most useful to surgeons who have a target post-operative SL in mind for the given spinal level with pathology were the model can help them select the appropriate lordosis for the interbody implant as well as understand whether a specific approach will be required to meet their goal. In this study, we demonstrated that ALIF and LLIF produced superior radiographic outcomes as compared to TLIF and PLF. There were several limitations in this study given that it was a retrospective analysis with a small sample size at a single institution. Many of our measurements involved measuring cobb angles on standardized digital radiographs on a PACS system, which has inherent standard error. Small changes in measurements may reflect these standard errors. In addition, clinical outcome measures were not simultaneously obtained to directly correlate with radiographic measurements. Complications were not determined in this study as the purpose of the study was to investigate the radiographic impact of these various lumbar fusion techniques and the complication profiles have been well described in the scientific literature. Ideally, a prospective study with longer follow up and multiple clinical outcome measures including complications rates would address the limitations of this study. Given the effect that lordotic implants have on radiographic parameters, a future study with all surgical techniques having similar lordotic implant profiles would be useful to determine the intrinsic potential for sagittal correction with each technique.
Conclusions The present study represents the largest comparison of lumbar fusion techniques and their effect upon spinopelvic radiographic outcomes to date. To our knowledge, it is the only study that directly compares the lumbar fusion techniques based on their ability to affect PI-LL mismatch. Finally, this study derived a predictive model for post-operative SL from surgical approach, surgical level, and implant lordosis. Our analysis demonstrates that ALIF and LLIF provide superior sagittal correction compared to TLIF and PLF, which may influence clinical outcomes. 9 Page 9 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
References [1] Schwab F, Patel A, Ungar B, Farcy JP, Lafage V. Adult spinal deformity-postoperative standing imbalance: how much can you tolerate? An overview of key parameters in assessing alignment and planning corrective surgery. Spine. 2010;35(25):2224-31. [2] Rothenfluh DA, Mueller DA, Rothenfluh E, Min K. Pelvic incidence-lumbar lordosis mismatch predisposes to adjacent segment disease after lumbar spinal fusion. Eur Spine J. 2015;24(6):1251-8. [3] Matsumoto T, Okuda S, Maeno T, et al. Spinopelvic sagittal imbalance as a risk factor for adjacent-segment disease after single-segment posterior lumbar interbody fusion. J Neurosurg Spine. 2017;26(4):435-440. [4] Aoki Y, Nakajima A, Takahashi H, et al. Influence of pelvic incidence-lumbar lordosis mismatch on surgical outcomes of short-segment transforaminal lumbar interbody fusion. BMC Musculoskelet Disord. 2015;16:213. [5] Makino T, Kaito T, Fujiwara H, et al. Risk factors for poor patient-reported quality of life outcomes after posterior lumbar interbody fusion: An analysis of two-year follow-up. Spine. 2017. [6] Than KD, Park P, Fu KM, et al. Clinical and radiographic parameters associated with best versus worst clinical outcomes in minimally invasive spinal deformity surgery. J Neurosurg Spine. 2016;25(1):21-5. [7] Schwab FJ, Blondel B, Bess S, et al. Radiographical spinopelvic parameters and disability in the setting of adult spinal deformity: a prospective multicenter analysis. Spine. 2013;38(13):E803-12. [8] Minamide A, Yoshida M, Iwahashi H, et al. Minimally invasive decompression surgery for lumbar spinal stenosis with degenerative scoliosis: Predictive factors of radiographic and clinical outcomes. J Orthop Sci. 2017;22(3):377-383. [9] Schwab F, Farcy JP, Bridwell K, et al. A clinical impact classification of scoliosis in the adult. Spine. 2006;31(18):2109-14. [10] Schwab FJ, Smith VA, Biserni M, Gamez L, Farcy JP, Pagala M. Adult scoliosis: a quantitative radiographic and clinical analysis. Spine. 2002;27(4):387-92. [11] Yang EZ, Xu JG, Liu XK, et al. An RCT study comparing the clinical and radiological outcomes with the use of PLIF or TLIF after instrumented reduction in adult isthmic spondylolisthesis. Eur Spine J. 2016;25(5):1587-94. [12] Ould-slimane M, Lenoir T, Dauzac C, et al. Influence of transforaminal lumbar interbody fusion procedures on spinal and pelvic parameters of sagittal balance. Eur Spine J. 2012;21(6):1200-6. [13] Kepler CK, Rihn JA, Radcliff KE, et al. Restoration of lordosis and disk height after singlelevel transforaminal lumbar interbody fusion. Orthop Surg. 2012;4(1):15-20. [14] Zhu Y, Liu HY, Wang B, et al. Long-term clinical outcomes of selective segmental transforaminal lumbar interbody fusion and posterior spinal fusion for degenerative lumbar scoliosis. Zhonghua Yi Xue Za Zhi 2013;93:3577-81. [15] Challier V, Boissiere L, Obeid I, et al. One-Level Lumbar Degenerative Spondylolisthesis and Posterior Approach: Is Transforaminal Lateral Interbody Fusion Mandatory?: A Randomized Controlled Trial With 2-Year Follow-Up. Spine. 2017;42(8):531-539. [16] Fujimori T, Le H, Schairer WW, et al. Does Transforaminal Lumbar Interbody Fusion Have Advantages over Posterolateral Lumbar Fusion for Degenerative Spondylolisthesis? Global Spine J 2015;5:102-9. 10 Page 10 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
[17] Sharma AK, Kepler CK, Girardi FP, Cammisa FP, Huang RC, Sama AA. Lateral lumbar interbody fusion: clinical and radiographic outcomes at 1 year: a preliminary report. J Spinal Disord Tech. 2011;24(4):242-50. [18] Acosta FL, Liu J, Slimack N, Moller D, Fessler R, Koski T. Changes in coronal and sagittal plane alignment following minimally invasive direct lateral interbody fusion for the treatment of degenerative lumbar disease in adults: a radiographic study. J Neurosurg Spine. 2011;15(1):92-6. [19] Kepler CK, Huang RC, Sharma AK, et al. Factors influencing segmental lumbar lordosis after lateral transpsoas interbody fusion. Orthop Surg. 2012;4(2):71-5. [20] Lee YS, Park SW, Kim YB. Direct lateral lumbar interbody fusion: clinical and radiological outcomes. J Korean Neurosurg Soc. 2014;55(5):248-54. [21] Johnson RD, Valore A, Villaminar A, Comisso M, Balsano M. Pelvic parameters of sagittal balance in extreme lateral interbody fusion for degenerative lumbar disc disease. J Clin Neurosci. 2013;20(4):576-81. [22] Khajavi K, Shen AY. Two-year radiographic and clinical outcomes of a minimally invasive, lateral, transpsoas approach for anterior lumbar interbody fusion in the treatment of adult degenerative scoliosis. Eur Spine J. 2014;23(6):1215-23. [23] Malham GM, Ellis NJ, Parker RM, et al. Maintenance of Segmental Lordosis and Disk Height in Stand-alone and Instrumented Extreme Lateral Interbody Fusion (XLIF). Clin Spine Surg. 2017;30(2):E90-E98. [24] Blizzard DJ, Gallizzi MA, Sheets C, et al. Sagittal Balance Correction in Lateral Interbody Fusion for Degenerative Scoliosis. Int J Spine Surg. 2016;10:29. [25] Phan K, Rao PJ, Scherman DB, Dandie G, Mobbs RJ. Lateral lumbar interbody fusion for sagittal balance correction and spinal deformity. J Clin Neurosci. 2015;22(11):1714-21. [26] Manwaring JC, Bach K, Ahmadian AA, et al. Management of sagittal balance in adult spinal deformity with minimally invasive anterolateral lumbar interbody fusion: a preliminary radiographic study. J Neurosurg Spine 2014;20:515-22. [27] Jiang SD, Chen JW, Jiang LS. Which procedure is better for lumbar interbody fusion: anterior lumbar interbody fusion or transforaminal lumbar interbody fusion?. Arch Orthop Trauma Surg. 2012;132(9):1259-66. [28] Phan K, Thayaparan GK, Mobbs RJ. Anterior lumbar interbody fusion versus transforaminal lumbar interbody fusion--systematic review and meta-analysis. Br J Neurosurg. 2015;29(5):705-11. [29] Ajiboye RM, Alas H, Mosich GM, Sharma A, Pourtaheri S. Radiographic and Clinical Outcomes of Anterior and Transforaminal Lumbar Interbody Fusions: A Systematic Review and Meta-analysis of Comparative Studies. Clin Spine Surg. 2017; [30] Hsieh PC, Koski TR, O'shaughnessy BA, et al. Anterior lumbar interbody fusion in comparison with transforaminal lumbar interbody fusion: implications for the restoration of foraminal height, local disc angle, lumbar lordosis, and sagittal balance. J Neurosurg Spine. 2007;7(4):379-86. [31] Kim JS, Lee KY, Lee SH, Lee HY. Which lumbar interbody fusion technique is better in terms of level for the treatment of unstable isthmic spondylolisthesis?. J Neurosurg Spine. 2010;12(2):171-7. [32] Watkins RG 4th, Hanna R, Chang D, et al. Sagittal alignment after lumbar interbody fusion: comparing anterior, lateral, and transforaminal approaches. J Spinal Disord Tech 2014;27:253-6. 11 Page 11 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
[33] Sembrano JN, Yson SC, Horazdovsky RD, et al. Radiographic Comparison of Lateral Lumbar Interbody Fusion Versus Traditional Fusion Approaches: Analysis of Sagittal Contour Change. Int J Spine Surg 2015;9:16. [34] Teng I, Han J, Phan K, Mobbs R. A meta-analysis comparing ALIF, PLIF, TLIF and LLIF. J Clin Neurosci. 2017. [35] Gödde S, Fritsch E, Dienst M, Kohn D. Influence of cage geometry on sagittal alignment in instrumented posterior lumbar interbody fusion. Spine. 2003;28(15):1693-9. [36] Barone G, Scaramuzzo L, Zagra A, Giudici F, Perna A, Proietti L. Adult spinal deformity: effectiveness of interbody lordotic cages to restore disc angle and spino-pelvic parameters through completely mini-invasive trans-psoas and hybrid approach. Eur Spine J. 2017. [37] Sembrano JN, Horazdovsky RD, Sharma AK, Yson SC, Santos ER, Polly DW. Do lordotic Cages Provide Better Segmental Lordosis Versus Non-lordotic Cages in Lateral Lumbar Interbody Fusion (LLIF)?. Clin Spine Surg. 2016. [38] Uribe JS, Harris JE, Beckman JM, Turner AW, Mundis GM, Akbarnia BA. Finite element analysis of lordosis restoration with anterior longitudinal ligament release and lateral hyperlordotic cage placement. Eur Spine J. 2015;24 Suppl 3:420-6. [39] Melikian R, Yoon ST, Kim JY, Park KY, Yoon C, Hutton W. Sagittal Plane Correction Using the Lateral Transpsoas Approach: A Biomechanical Study on the Effect of Cage Angle and Surgical Technique on Segmental Lordosis. Spine. 2016. [40] Uribe JS, Smith DA, Dakwar E, et al. Lordosis restoration after anterior longitudinal ligament release and placement of lateral hyperlordotic interbody cages during the minimally invasive lateral transpsoas approach: a radiographic study in cadavers. J Neurosurg Spine. 2012;17(5):476-85. [41] Hong TH, Cho KJ, Kim YT, Park JW, Seo BH, Kim NC. Does Lordotic Angle of Cage Determine Lumbar Lordosis in Lumbar Interbody Fusion?. Spine. 2017;42(13):E775-E780. [42] Anand N, Cohen RB, Cohen J, Kahndehroo B, Kahwaty S, Baron E. The Influence of Lordotic cages on creating Sagittal Balance in the CMIS treatment of Adult Spinal Deformity. Int J Spine Surg. 2017;11:23. [43] Fingar KR, Stocks C, Weiss AJ, & Steiner CA. Most frequent operating room procedures performed in US hospitals, 2003–2012. HCUP Statistical Brief 2014; 186. [44] Rao PJ, Maharaj MM, Phan K, Lakshan abeygunasekara M, Mobbs RJ. Indirect foraminal decompression after anterior lumbar interbody fusion: a prospective radiographic study using a new pedicle-to-pedicle technique. Spine J. 2015;15(5):817-24. [45] Kepler CK, Sharma AK, Huang RC, et al. Indirect foraminal decompression after lateral transpsoas interbody fusion. J Neurosurg Spine. 2012;16(4):329-33. [46] Pawar AY, Hughes AP, Sama AA, Girardi FP, Lebl DR, Cammisa FP. A Comparative Study of Lateral Lumbar Interbody Fusion and Posterior Lumbar Interbody Fusion in Degenerative Lumbar Spondylolisthesis. Asian Spine J. 2015;9(5):668-74. [47] Jackson RP, Mcmanus AC. Radiographic analysis of sagittal plane alignment and balance in standing volunteers and patients with low back pain matched for age, sex, and size. A prospective controlled clinical study. Spine. 1994;19(14):1611-8. [48] Rice JW, Sedney CL, Daffner SD, Arner JW, Emery SE France JC. Improvement of Segmental Lordosis in Transforaminal Lumbar Interbody Fusion: A comparison of Two Techniques. Global Spine J. 2016 May;6(3):229-33. 12 Page 12 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Fig 1. Radiographic Measurements. SL is the angle between the superior and inferior endplates of adjacent superior and inferior vertebral bodies, respectively (1). LL is the angle between the superior S1 and L1 endplates (2). Pelvic incidence is the angle between a line from the center of the S1 endplate to the center of the femoral heads and a line perpendicular to the center of the S1 endplate (3). Disk height was measured as the minimum distance (mm) between the superior and inferior end plates of adjacent vertebral bodies at their anterior and posterior surfaces (4). Foraminal height was measured as the maximum distance between the inferior margin of the pedicle of the superior vertebra and the superior margin of the pedicle of the inferior vertebra (5). Fig. 2. Paired sample t-test analysis of pre-operative vs. post-op radiographic parameters SL, LL, DHA/P, and FH by surgical approaches ALIF (A), LLIF (B), TLIF (C), PLF (D). Parameter ± SD, error bars reflect 95% CI. Fig. 3. ANOVA for mean degree change in radiographic parameters SL (A), LL (B), DH-A/P (C), and FH (D) by surgical approach. Fig. 4. Change in proportion of patients with PI-LL < 10° by surgical approach. Fig. 5. Relative degree change of radiographic parameters for lordotic (n = 58) vs. non-lordotic cages (n = 106) in ALIF, LLIF, and TLIF surgeries.
13 Page 13 of 27
1 2
Table 1. Frequency of primary diagnoses for each surgical approach including degenerative disk disease, spondylolisthesis, isthmic spondylolisthesis, adjacent level disease, pseudoarthrosis, and scoliosis.
ALIF
LLIF
TLIF
PLF
Total
Degenerative disk disease
14
4
5
2
25
Isthmic spondylolisthesis
2
0
12
1
15
Spondylolisthesis
17
7
41
35
100
Adjacent level disease
0
10
0
2
12
Pseudoarthrosis
1
0
1
1
3
Scoliosis
0
2
1
1
4
Other
0
0
3
2
5
Total
34
23
63
44
164
Primary Diagnosis
3 4 5
Table 2. Comparison of the number of lordotic and non-lordotic implants for each surgical approach.
ALIF
LLIF
TLIF
PLF
Total
Y
34
18
6
0
58
N
0
5
57
44
106
Total
34
23
63
44
164
Lordotic Implant
6 7 8
Table 3. Number of fusions at a given surgical level for each surgical approach.
ALIF
LLIF
TLIF
PLF
Total
L1-2
0
0
0
1
1
L2-3
0
6
2
2
10
L3-4
3
11
3
9
26
L4-5
8
6
43
26
83
L5-S1
23
0
15
6
44
Total
34
23
63
44
164
Surgical Level
9
14 Page 14 of 27
1
Table 4. Number of lordotic implants with a given degree lordosis for ALIF, LLIF, and TLIF.
ALIF
LLIF
TLIF
Total
0
0
5
57
62
5
3
0
4
7
6
1
5
1
7
7
0
7
0
7
8
4
5
0
9
9
2
0
0
2
10
4
1
0
5
12
8
0
1
9
14
2
0
0
2
15
10
0
0
10
Total
34
23
63
120
Implant Lordosis (°)
2 3 4 5
6 7 8
Table 5. Slope values for implant lordosis, surgical approach, and surgical level in predictive equation for post-operative segmental lordosis as derived from multiple linear regression.
β
Variable
B
SEB
Sig.
Intercept
11.406
4.789
Implant Lordosis (°)
1.089
.369
.342
.005
Approach*
6.835
2.739
.306
.016
Surgical Level**
6.874
2.575
.274
.010
.021
Note. * 0 = LLIF/TLIF, 1 = ALIF; ** 0 = L2-3/L3-4, 1 = L4-5/L5-S1 B = unstandardized regression coefficient; SEB = standard error of coefficient; β = standardized coefficient
9
15 Page 15 of 27
1
Table 6. Comparison of SL and LL based on pre-operative disk height
ΔSL
ΔLL
< 6.0 mm (n=78)
4.7
3.7
> 6.0 mm (n=86)
1.1
2.6
< .001
.34
Average Pre-op DH
P values 2 3 4 5 6 7 8 9 10 11 12
Eq 1. Multiple linear regression equation for predicting post-operative segmental lordosis from the variables of implant lordosis, surgical approach, and surgical level.
16 Page 16 of 27
1 2 3
TSJ LIF Figure 1.jpg
17 Page 17 of 27
1 2 3
TSJ LIF Figure 2A.jpg
18 Page 18 of 27
1 2 3
TSJ LIF Figure 2B.jpg
19 Page 19 of 27
1 2 3
TSJ LIF Figure 2C.jpg
20 Page 20 of 27
1 2 3
TSJ LIF Figure 2D.jpg
21 Page 21 of 27
1 2 3
TSJ LIF Figure 3A.jpg
22 Page 22 of 27
1 2 3
TSJ LIF Figure 3B.jpg
23 Page 23 of 27
1 2 3
TSJ LIF Figure 3C.jpg
24 Page 24 of 27
1 2 3
TSJ LIF Figure 4.jpg
25 Page 25 of 27
1 2 3
TSJ LIF Figure 5.jpg
26 Page 26 of 27
1 2
TSJ LIF Figure3D.jpg
27 Page 27 of 27
S1529-9430(18)30125-6 https://doi.org/10.1016/j.spinee.2018.03.014 SPINEE 57633
To appear in:
The Spine Journal
Received date: Revised date: Accepted date:
8-12-2017 19-3-2018 26-3-2018
Please cite this article as: Seth Ahlquist, Howard Y. Park, Jonathan Gatto, Ayra N. Shamie, Don Y. Park, Does approach matter? a comparative radiographic analysis of spinopelvic parameters in single level lumbar fusion., The Spine Journal (2018), https://doi.org/10.1016/j.spinee.2018.03.014. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
Does Approach Matter? A Comparative Radiographic Analysis of Spinopelvic Parameters in Single Level Lumbar Fusion.
Seth Ahlquist BS1, Howard Y. Park, MD1, Jonathan Gatto, BS1, Ayra N. Shamie, MD1, Don Y. Park MD1 1. Department of Orthopaedic Surgery, David Geffen School of Medicine at UCLA, 1250 16th Street, Santa Monica, CA, 90404, USA
Corresponding Author Don Y. Park, MD Department of Orthopaedic Surgery David Geffen School of Medicine at UCLA, 1250 16th Street Santa Monica, CA, 90404, USA Email: [email protected] Phone: 424-259-9829 Fax: 424-259-6599
Abstract BACKGROUND CONTEXT: Lumbar fusion is a popular and effective surgical option to provide stability and restore anatomy. Particular attention has recently been focused on sagittal alignment and radiographic spinopelvic parameters that apply to lumbar fusion as well as spinal deformity cases. Current literature has demonstrated the effectiveness of various techniques of lumbar fusion, however comparative data of these techniques is limited. PURPOSE: To directly compare the impact of various lumbar fusion techniques (ALIF, LLIF, TLIF, PLF) based on radiographic parameters. STUDY DESIGN/SETTING: A single-center retrospective study examining pre-operative and post-operative radiographs. PATIENT SAMPLE: A consecutive list of lumbar fusion surgeries performed by multiple spine surgeons at a single institution from 2013-2016 were identified. 1 Page 1 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
OUTCOME MEASURES: Radiographic measurements utilized included segmental lordosis (SL), lumbar lordosis (LL), pelvic incidence (PI), pelvic incidence-lumbar lordosis mismatch (PI-LL), anterior and posterior disk height (DH-A, DH-P respectively), and foraminal height (FH). METHODS: Radiographic measurements were performed on pre-operative and post-operative lateral lumbar radiographs on all single-level lumbar fusion cases. Demographic data was collected including age, gender, approach, diagnosis, surgical level, and implant lordosis. Paired sample t-test, one-way ANOVA, McNemar Test, and independent sample t-test were used to establish significant differences in the outcome measures. Multiple linear regression was performed to determine a predictive model for lordosis from implant lordosis, fusion technique, and surgical level. RESULTS: There were 164 patients (78 males, 86 females) with a mean age of 60.1 years and average radiographic follow up time of 9.3 months. These included 34 ALIF, 23 LLIF, 63 TLIF, and 44 PLF surgeries. ALIF and LLIF significantly improved SL (7.9° & 4.4°), LL (5.5° & 7.7°), DH-A (8.8 mm & 5.8 mm), DH-P (3.4 mm & 2.3 mm), and FH (2.8 mm & 2.5 mm), respectively (p ≤ .003). TLIF significantly improved these parameters, albeit to a lesser extent: SL (1.7°), LL (2.7°), DH-A (1.1 mm), DH-P (0.8 mm), and FH (1.1 mm), p ≤ .02. PLF did not significantly alter any of these parameters while significantly reducing FH (-1.3 mm, p = .01). One-way ANOVA showed no significant differences between ALIF and LLIF other than ALIF with greater ΔDH-A (3.0 mm, p = .02). Both ALIF and LLIF significantly outperformed PLF in pre-operative to post-operative change in all parameters p ≤ .001. Additionally, ALIF significantly outperformed TLIF in the change in SL (6.2°, p < .001) and LLIF significantly outperformed TLIF in the change in LL (5.0°, p = .02). Both outperformed TLIF in ΔDH-A (7.7 mm & 4.7 mm) and ΔDH-P (2.6 mm & 1.5 mm), respectively (p ≤ .02). ALIF was the only fusion technique that significantly improved the proportion of patients with a PI-LL < 10° (0.41 to 0.66, p = .02). Lordotic cages had superior improvement of all parameters as compared to non-lordotic cages (p <.001). Implant lordosis (m = 1.1), fusion technique (m = 6.8), and surgical level (m = 6.9) significantly predicted post-operative SL (p < .001, R2 = .56). CONCLUSIONS: This study demonstrated that these four lumbar fusion techniques yield divergent radiographic results. ALIF and LLIF produced greater improvements in radiographic measurements post-operatively as compared to TLIF and PLF. ALIF was the most successful in improving PI-LL mismatch, an important parameter relating to sagittal alignment. Lordotic implants provided better sagittal correction and surgeons should be cognizant of the impact that these differing implants and techniques produce after surgery. Surgical technique is an important determinant of post-operative alignment and has ramifications upon sagittal alignment in lumbar fusion surgery. Keywords: Lumbar interbody fusion, ALIF, LLIF, TLIF, PLF, sagittal alignment, segmental lordosis, lumbar lordosis, PI-LL
2 Page 2 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
Introduction Degenerative conditions of the lumbar spine are becoming increasingly prevalent with an aging population and result in significant reductions to quality of life due to immobility, radicular pain, and muscle spasm. Lumbar fusion through various techniques is an effective intervention for restoring the stability of these spinal segments, alleviating compression of neural elements, and reestablishing spinal anatomy. The critical importance of sagittal alignment and pelvic parameters is well established in spinal deformity and lumbar degenerative surgery alike. High pelvic incidence-lumbar lordosis (PI-LL) mismatch (greater than 10°) is associated with adjacent segment degeneration (ASD) and a tenfold risk of revision surgery [1-3], worsened post-operative residual symptoms [4], reduced quality of life [5], severe disability [6, 7], and decreased recovery rate [8]. Similarly, lower postoperative lumbar lordosis (LL) is significantly associated with increased ASD [3], disability [9], and pain [10]. There are multiple studies in the literature that examine the effects of the various lumbar fusion techniques have on spinopelvic parameters, albeit with conflicting results. Transforaminal lumbar interbody fusion (TLIF) has been shown to improve segmental lordosis (SL) [11, 12] and LL [13, 14] in previous studies. However, a recent retrospective study and randomized controlled trial demonstrated conflicting results as TLIF did not result in superior SL improvement over posterolateral fusion (PLF) [15, 16]. Lateral lumbar interbody fusion (LLIF) was initially shown to only improve SL [17-21], however later studies revealed the technique can also improve LL [22-26]. Anterior lumbar interbody fusion (ALIF) outperformed TLIF in the improvement of SL and LL in multiple studies [27-31]. However, few studies directly compared the radiographic outcomes of all the available lumbar fusion techniques together, and none included PI-LL mismatch as an outcome measure. Watkins et al compared ALIF, LLIF, and TLIF and determined that only ALIF and LLIF significantly improved SL [32]. Sembrano et al compared ALIF, LLIF, TLIF, and PLF and found that the former three approaches increased SL and LL, with ALIF to the greatest extent, emphasizing the importance of interbody cages [33]. A recent systematic review found that ALIF had significantly higher postoperative SL than LLIF, TLIF, and PLIF [34]. The geometry of the interbody cage itself may influence sagittal alignment, with wedge-shaped lordotic cages becoming more commonly used over non-lordotic cages since the lordotic cages can significantly increase SL and LL [35]. Lordotic cages in LLIF have been shown to provide a significant increase in SL (p =.01) while non-lordotic cages do not (p = .71) [36-37]. Hyperlordotic cages and ALL resection may further impact sagittal alignment as placement of increasingly lordotic cages without the restraint of the ALL led to greater gains in SL and LL [38-41]. In addition to cage geometry, cage position and insertion level are factors that may also influence the change in SL [42]. The literature regarding the various lumbar fusion techniques and interbody cages has demonstrated satisfactory fusion rates and clinical success; however, their comparative effects on sagittal alignment and pelvic parameters, including SL, LL, and PI-LL mismatch remain unclear. 3 Page 3 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
The goal of this study was to compare the capacities of ALIF, LLIF, TLIF, and PLF to change SL, LL, PI-LL mismatch, disk height (DH), and foraminal height (FH), as well as to develop a predictive model for lordosis based on the surgical approach, insertion level, and cage lordosis. Materials and Methods A consecutive list of lumbar fusion surgeries performed by multiple spine surgeons at a single institution from 2013-2016 were identified and retrospectively analyzed with Institutional Review Board approval (IRB#16-000175). Inclusion criteria included all single level lumbar fusion surgeries (n = 164). Exclusion criteria included revision surgeries, multilevel fusions, inclusion of spinal osteotomies, fusion performed for spinal trauma, tumor, and infections (n = 36). Osteotomies were specifically excluded to reduce the variables examined in this study and focus the investigation to the lumbar fusion techniques themselves. 164 patients fit the inclusion criteria while 36 patients were excluded from the study. Radiographic measurements were performed on pre-operative and post-operative neutral, lateral lumbar radiographs on all single-level lumbar fusion cases using the PACS system. These measurements were performed independently by two trained observers. Radiographic measurements obtained included segmental lordosis (SL), lumbar lordosis (LL), pelvic incidence (PI), pelvic incidence-lumbar lordosis mismatch (PI-LL), disk height (DH), and foraminal height (FH) (Fig 1). Post-op radiographs were obtained at the standardized time points (i.e. 6 weeks, 3 months, 6 months, 1 year post-surgery), although there was some variability in these time points based on the patient’s followup. Demographic data was collected including age, gender, surgical approach, diagnosis, surgical level, and implant lordosis. Paired sample t-test, one-way ANOVA with Tukey/Games Howell post hoc analysis, exact McNemar Test, and independent sample t-test were used to establish significant differences in the outcome measures. Multiple linear regression was performed to determine a predictive model for post-operative segmental lordosis from implant lordosis, fusion technique, and surgical level. Comparisons between groups and within groups were deemed statistically significant at the p < .05 threshold. Statistical analyses were performed using SPSS 24 software (IBM Corp., Armonk, NY, USA).
Results Demographic data There were 164 patients [78 males (47.6%), 86 females (52.4%)] with a mean age of 60.1 years (range 25-88) and average radiographic follow up time of 9.3 months. These patients underwent 34 ALIF (20.7%), 23 LLIF (14.0%), 63 TLIF (38.4%), and 44 PLF surgeries (26.8%). Follow up time was not significantly different for any of the approaches (p = .10). The most common primary indication for surgery was spondylolisthesis (61.0%), followed by degenerative disk disease (15.2%) and isthmic spondylolisthesis (9.1%) (Table 1). All techniques had spondylolisthesis as the most frequent indication for surgery, except for LLIF which had adjacent
4 Page 4 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
level disease. Also of note is that ALIF had the highest proportion of degenerative disk disease cases (41.1%) and TLIF the highest proportion of isthmic spondylolisthesis cases (19.0%). Change in Radiographic Parameters by Surgical Technique Paired sample t-test was performed to examine whether each individual surgical approach significantly altered the aforementioned radiographic parameters from the pre-operative to postoperative radiographs. ALIF and LLIF significantly improved SL (7.9° & 4.4°, respectively; p < .001), LL (5.5° & 7.7°, respectively; p < .001), DH-A (8.8 mm & 5.8 mm, respectively; p < .001), DH-P (3.4 mm & 2.3 mm, respectively; p < .001), and FH (2.8 mm & 2.5 mm, respectively; p <.001 and p =.003) (Fig. 2A, 2B). TLIF significantly improved these parameters as well, however to a lesser extent: SL (1.7°, p = .02), LL (2.7°, p = .01), DH-A (1.1 mm, p = .002), DH-P (0.8 mm, p = .002), and FH (1.1 mm, p = .002) (Fig. 2C). PLF did not significantly alter SL (-0.3°, p = .65), LL (-0.5°, p = .66), DH-A (0.1 mm, p = .75), or DH-P (0.1 mm, p = .70), and significantly decreased FH (-1.3 mm, p = .01) (Fig. 2D). Contribution of Disk Height to Lordosis Overall disk height was analyzed to determine the contribution of pre-operative disk height to the change in lordosis post-operatively using an average of the anterior and posterior disk height values across all surgical techniques. A halfway cutoff of 6.0mm was determined to differentiate between a more collapsed disk height group (n=78) and a less collapsed disk height group (n=86) (Table 6). SL improvement was significantly greater in the more collapsed disk height group than the less collapsed disk height group (p < .001). No significant difference was detected in LL improvement between the two groups (p=0.34). Comparison of Radiographic Parameter Changes Between Surgical Techniques A one-way ANOVA was utilized to determine whether a particular surgical approach significantly outperformed the others in its ability to improve the radiographic parameters of SL, LL, DH-A/P and FH. No significant differences were found between ALIF and LLIF in these parameters except for ALIF having a significantly greater ΔDH-A (3.0 mm, p = .02) (Fig. 3C). Both ALIF and LLIF significantly outperformed PLF in pre-operative to post-operative change in SL (8.2° & 4.7°), LL (6.0° & 8.2°), DH-A (8.9 mm & 5.9 mm), DH-P (3.5 mm & 2.4 mm), and FH (4.1 mm & 3.8 mm), respectively, with all p ≤ .001 (Fig. 3A-D). Additionally, ALIF significantly outperformed TLIF in the change in SL (6.2°, p < .001) and LLIF significantly outperformed TLIF in the change in LL (5.0, p = .02). Both outperformed TLIF in ΔDH-A (7.7 mm & 4.7 mm, respectively; p ≤ .001) and ΔDH-P (2.6 mm, p <.001 & 1.5 mm, p =.02, respectively) (Fig. 3A-D). TLIF was significantly more successful than PLF in terms of FH improvement (2.4 mm, p = .001) (Fig. 3D) but not compared to the ALIF and LLIF. Capacity of Surgical Techniques to Increase Proportion of Patients with a PI-LL < 10° McNemar Test was performed to see whether any of the surgical techniques significantly altered the proportion of patients with a PI-LL < 10° from pre-op to post-op. ALIF was the only technique to significantly improve the proportion of patients with a PI-LL < 10° (0.41 to 0.66, p = .02), while the other techniques did not significantly alter the pre-operative proportions: LLIF 5 Page 5 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
0.55 to 0.59 (p = 1.0), TLIF 0.46 to 0.56 (p = .21), and PLF 0.43 to 0.50 (p = .38) (Fig. 4). Only one patient who had a pre-operative PI-LL of < 10 who underwent ALIF ended up having a PILL > 10° after surgery. Comparison of Radiographic Parameter Changes between Lordotic and Non-Lordotic Cages 58 of the cages implanted during surgery were lordotic and 106 were non-lordotic (Table 2). All of the ALIF procedures utilized a lordotic implant. The majority of LLIF procedures utilized a lordotic implant while the majority of TLIF procedures did not. Independent sample t-test was performed in order to determine whether there were significant differences in the ability of lordotic cages to improve radiographic parameters over that of non-lordotic cages. Analysis determined that lordotic cages improved SL (5.0°), LL (4.7°), DH-A (6.3 mm), DH-P (1.2 mm), and FH (2.3 mm) to a significantly greater degree as compared to non-lordotic cages, with all p < .001 (Fig. 5). Predictive Equation for Post-Operative Segmental Lordosis The most common surgical level overall for lumbar fusion during the study period was L4-5 (50.6%), however, LLIF was most frequently performed at the L3-4 level (15.9% overall), and ALIF was most frequently performed at the L5-S1 level (26.8% overall) (Table 3). ALIF, LLIF, and TLIF had different profiles of lordotic cages used with the majority of ALIF being > 8° (11.3 ± 3.3°), LLIF being < 8° (5.6 ± 3.2°), and TLIF being = 0° (0.6 ± 2.1°). Multiple linear regression was performed on all cases with lordotic implants (n = 58) to derive a predictive equation for post-operative SL from implant lordosis, surgical level, and surgical approach. The multiple regression model statistically significantly predicted SLpost-op, F(3, 54) = 22.753, p < .001, R2 = .56 (Eq.1). ALIF was the only technique that statistically significantly increased SLpost-op, and thus surgical approach is accounted for as either ALIF (approach = 1) or not ALIF (approach = 0). Similarly, fusions performed at the L4-5 and L5-S1 levels (surgical level = 1) afforded significantly higher SLpostop than those performed at L2-3 and L3-4 (surgical level = 0). Equation 1. SLpost-op = 1.1*(Implant Lordosis) + 6.8*(Approach) + 6.9*(Surgical Level) + 11.4
All three variables added significantly to the prediction, p < .05. Regression coefficients and standard errors can be found in Table 5.
Discussion Lumbar fusion is the 5th most common surgical procedure performed in United States hospitals, with over 450,000 cases performed annually [43]. Controversy exists over which lumbar fusion technique is optimal in providing the greatest improvement of radiographic parameters. Although these techniques can achieve fusion, there are differences in each technique’s ability to affect sagittal alignment, an important predictor of clinical success. There are few studies that directly compare all the available surgical techniques for lumbar fusion and no prior studies investigating PI-LL mismatch as an outcome measure, which has been repeatedly shown to 6 Page 6 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
correlate with clinical outcomes. This study aimed to compare the radiographic outcomes of various techniques performed to achieve single level lumbar fusion. This study demonstrated that ALIF, LLIF, and TLIF techniques can each alter and influence sagittal alignment, although in different ways. We found that ALIF improved SL by 7.9°, p < .001 (Fig. 2A), while prior studies found lower magnitude changes in SL, ranging from 3.7° to 4.5° [30-33] (p < .05). We were able to achieve greater improvements in SL with the ALIF technique, which may be due to greater attention to sagittal alignment by the surgeons and greater availability of lordotic cages. ALIF significantly improved all radiographic measurements in this study, which is in line with the results from prior studies [30-34, 44]. LLIF improved SL by 4.4° in our results, p < .001 (Fig. 2B), as compared to 1.2° to 12.0° in prior studies [17- 20, 22-23, 26, 32-33] (p < .05). Manwaring et al. found a substantially higher change in SL (12.0°), which may be explained by the inclusion of anterior column release cases (ACR) that can substantially increase lordosis. In addition, the study was limited by a small sample size of 9 cases, and included multilevel fusions [26]. Our study of single level fusions found LLIF improved LL by 7.7°, p < .001 (Fig. 2B), which was in line with prior studies (2.5° to 12.2°) [22-24, 33]. A systematic review of LLIF demonstrated a ΔSL of 2.4° and ΔLL of 7.5° (p < .05), which is in line with our results that the LLIF technique is effective in changing sagittal alignment [25]. LLIF significantly improved DH-A, DH-P and FH, which is in agreement with prior studies [33, 45, 46]. Our study demonstrated that TLIF did not change the radiographic parameters to the same extent as ALIF or LLIF, which may be attributed to the vast majority of TLIF cages being non-lordotic. TLIF improved SL by only 1.7° in our study, p = .02 (Fig. 2C), as compared to prior studies ranging from 1.9 to 9.3° [11, 12, 30-31, 33]. Watkins et al. found a non-significant change in SL with TLIF measuring 0.8 (p > .05) [32]. In respect to the change in LL, we found that TLIF improved LL by 2.7, p = .01 (Fig. 2C), while other studies ranged from 2.1 to 10.5° [12, 13, 33] (p < .05). Kim et al. found non-significant changes with TLIF of -0.1 and 1.7 at the L4-5 (p = .98) and L5-S1 (p = .38) levels, respectively, which is critically important since the majority of lumbar lordosis occurs from L4 to S1 levels [31]. Overall, our results determined that TLIF underperformed radiographically as compared to the results of previous studies [30, 33]. In regards to PLF, we demonstrated that PLF was radiographically inferior in changing the sagittal alignment as compared to the other lumbar fusion techniques. Our results demonstrated that PLF did not significantly alter SL (-0.3, p = .65), LL (-0.5, p = .66), DH-A (0.1mm, p = .75), or DH-P (0.1mm, p = .70), and significantly decreased FH (-1.3, p = .01) (Fig. 2D). This is closely in line with work by Sembrano et al. who found a non-significant ΔSL of (0.7, p = .13), ΔLL (-0.5, p = .66), and ΔDH-A (0.3, p = .25), as well as a small but significant decrease in ΔDH-P (-0.4, p = .04) [33]. Posterior osteotomies with cantilevering compression techniques have been shown to improve lordosis in prior studies, which may influence the ability of TLIF to improve sagittal alignment [48]. No posterior osteotomies were included in our study as the aim of the study was to investigate the radiographic impact of the various lumbar fusion techniques themselves, which could have improved these radiographic results. However, our results demonstrate that TLIF and 7 Page 7 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
PLF as a standalone technique does not produce significant changes in sagittal alignment as compared to ALIF and LLIF. This may not be clinically relevant in patients who are already in balance pre-operatively, but surgeons should be cognizant of the limitations of these techniques in changing sagittal alignment. When comparing ALIF and LLIF, few studies in the literature have evaluated the improvement of radiographic parameters between these techniques. Watkins et al. found that ALIF significantly outperformed LLIF in SL improvement (2.3°, p < .05) [32]. Our study failed to find any difference in the changes of SL (p = .07) and LL (p = .64) when comparing ALIF and LLIF (Fig. 3A-B). Sembrano et al. also found no difference in changes in SL (p = .57) or LL (p = .17) [33]. Our results support the findings that ALIF and LLIF are statistically equivalent in changing the SL and LL from the pre-operative to post-operative value. However, the change of these parameters may not be as clinically important as the actual values of the parameters themselves. The actual value of SL and LL may provide satisfactory sagittal alignment and balance. There may be a potential confounding effect regarding the surgeon’s choice in selecting patients to undergo fusion with one technique versus another, including availability of vascular access surgeon, comfort and training with certain techniques, and the awareness of the pelvic parameters in preoperative planning. Severity of preoperative disk space narrowing and collapse may also influence decision making as one technique may be favored over another based on the severity of disk space collapse, especially posteriorly which may make TLIF difficult to perform. Our analysis of the pre-operative disk height and its contribution to lordosis demonstrated a significantly greater improvement in SL in the more collapsed disk height group than the less collapsed disk height group across all fusion techniques (p < .001). This reflects a greater change in the disk space after surgery in disk spaces that are more collapsed pre-operatively, thereby impacting lordosis to a greater degree. Incorporating the state of the disk space pre-operatively may be an important element in surgical decision-making, in addition to surgical technique and implant selection, to effectively improve alignment. This underscores the importance in preoperative planning in lumbar fusion surgeries beyond achieving fusion. Much attention has been placed on the importance of improving the PI-LL mismatch as it relates to improved clinical outcomes. We demonstrated that ALIF was the only technique that significantly increased the proportion of patients with a PI-LL < 10°, p = .02 (Fig. 4). Given the studies in the literature showing the importance of the PI-LL mismatch parameter [1-8], this is especially pertinent. Approximately 60% of lumbar lordosis is derived from the L4-S1 segments making these levels particularly important [47]. In this study, LLIF was more commonly used at the L3-4 level, which may explain why LLIF was less successful in restoring PI-LL balance. With greater attention to PI-LL mismatch by surgeons, the inferior results with LLIF and TLIF techniques may change, especially with lordotic cage designs, posterior osteotomy and compression techniques, and expandable cage technology. While the differences in SL and LL improvement between surgical techniques in our study were less than 10 degrees, these differences were statistically significant. These small differences may be clinically significant since pre-operative PI-LL mismatch values ranged from 9-13, and an increase in lordosis in several degrees may improve the post-operative PI-LL mismatch < 10°.
8 Page 8 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
While this result may seem intuitive, lordotic cages provided a significantly larger improvement in the radiographic parameters than non-lordotic cages (Fig. 5). Multiple prior studies have provided evidence for qualitative increases in SL with increasingly lordotic cages [38-42]. Interbody cages are now designed with lordosis with some providing hyperlordotic dimensions to provide even more sagittal correction. The vast majority of TLIF cages were non-lordotic in this study (6 lordotic TLIF cages/63 total TLIF cages) while the majority of ALIF and LLIF cages were lordotic. This may explain the significant difference in radiographic results between the various techniques. Based on our results, we derived a predictive model for post-operative SL as a function of implant lordosis, surgical level and surgical approach (Eq. 1). Notably, there was about one degree of SL gained for each degree increase in implant lordosis. As would be expected from our within-group analysis (Fig. 2A), ALIF was much more effective at increasing SL than either LLIF or TLIF (Table 5). There are greater increases in SL when the cage is implanted at the L4-5 or L5-S1 levels (+6.9°) than at the L2-3, L3-4 levels. It should be recognized that this equation will only be useful for predicting lordosis for surgeries in which osteotomies are not performed since the study did not include osteotomies in the analysis. This model has the potential to be a useful tool for spine surgeons to utilize in surgical planning. It will be most useful to surgeons who have a target post-operative SL in mind for the given spinal level with pathology were the model can help them select the appropriate lordosis for the interbody implant as well as understand whether a specific approach will be required to meet their goal. In this study, we demonstrated that ALIF and LLIF produced superior radiographic outcomes as compared to TLIF and PLF. There were several limitations in this study given that it was a retrospective analysis with a small sample size at a single institution. Many of our measurements involved measuring cobb angles on standardized digital radiographs on a PACS system, which has inherent standard error. Small changes in measurements may reflect these standard errors. In addition, clinical outcome measures were not simultaneously obtained to directly correlate with radiographic measurements. Complications were not determined in this study as the purpose of the study was to investigate the radiographic impact of these various lumbar fusion techniques and the complication profiles have been well described in the scientific literature. Ideally, a prospective study with longer follow up and multiple clinical outcome measures including complications rates would address the limitations of this study. Given the effect that lordotic implants have on radiographic parameters, a future study with all surgical techniques having similar lordotic implant profiles would be useful to determine the intrinsic potential for sagittal correction with each technique.
Conclusions The present study represents the largest comparison of lumbar fusion techniques and their effect upon spinopelvic radiographic outcomes to date. To our knowledge, it is the only study that directly compares the lumbar fusion techniques based on their ability to affect PI-LL mismatch. Finally, this study derived a predictive model for post-operative SL from surgical approach, surgical level, and implant lordosis. Our analysis demonstrates that ALIF and LLIF provide superior sagittal correction compared to TLIF and PLF, which may influence clinical outcomes. 9 Page 9 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
References [1] Schwab F, Patel A, Ungar B, Farcy JP, Lafage V. Adult spinal deformity-postoperative standing imbalance: how much can you tolerate? An overview of key parameters in assessing alignment and planning corrective surgery. Spine. 2010;35(25):2224-31. [2] Rothenfluh DA, Mueller DA, Rothenfluh E, Min K. Pelvic incidence-lumbar lordosis mismatch predisposes to adjacent segment disease after lumbar spinal fusion. Eur Spine J. 2015;24(6):1251-8. [3] Matsumoto T, Okuda S, Maeno T, et al. Spinopelvic sagittal imbalance as a risk factor for adjacent-segment disease after single-segment posterior lumbar interbody fusion. J Neurosurg Spine. 2017;26(4):435-440. [4] Aoki Y, Nakajima A, Takahashi H, et al. Influence of pelvic incidence-lumbar lordosis mismatch on surgical outcomes of short-segment transforaminal lumbar interbody fusion. BMC Musculoskelet Disord. 2015;16:213. [5] Makino T, Kaito T, Fujiwara H, et al. Risk factors for poor patient-reported quality of life outcomes after posterior lumbar interbody fusion: An analysis of two-year follow-up. Spine. 2017. [6] Than KD, Park P, Fu KM, et al. Clinical and radiographic parameters associated with best versus worst clinical outcomes in minimally invasive spinal deformity surgery. J Neurosurg Spine. 2016;25(1):21-5. [7] Schwab FJ, Blondel B, Bess S, et al. Radiographical spinopelvic parameters and disability in the setting of adult spinal deformity: a prospective multicenter analysis. Spine. 2013;38(13):E803-12. [8] Minamide A, Yoshida M, Iwahashi H, et al. Minimally invasive decompression surgery for lumbar spinal stenosis with degenerative scoliosis: Predictive factors of radiographic and clinical outcomes. J Orthop Sci. 2017;22(3):377-383. [9] Schwab F, Farcy JP, Bridwell K, et al. A clinical impact classification of scoliosis in the adult. Spine. 2006;31(18):2109-14. [10] Schwab FJ, Smith VA, Biserni M, Gamez L, Farcy JP, Pagala M. Adult scoliosis: a quantitative radiographic and clinical analysis. Spine. 2002;27(4):387-92. [11] Yang EZ, Xu JG, Liu XK, et al. An RCT study comparing the clinical and radiological outcomes with the use of PLIF or TLIF after instrumented reduction in adult isthmic spondylolisthesis. Eur Spine J. 2016;25(5):1587-94. [12] Ould-slimane M, Lenoir T, Dauzac C, et al. Influence of transforaminal lumbar interbody fusion procedures on spinal and pelvic parameters of sagittal balance. Eur Spine J. 2012;21(6):1200-6. [13] Kepler CK, Rihn JA, Radcliff KE, et al. Restoration of lordosis and disk height after singlelevel transforaminal lumbar interbody fusion. Orthop Surg. 2012;4(1):15-20. [14] Zhu Y, Liu HY, Wang B, et al. Long-term clinical outcomes of selective segmental transforaminal lumbar interbody fusion and posterior spinal fusion for degenerative lumbar scoliosis. Zhonghua Yi Xue Za Zhi 2013;93:3577-81. [15] Challier V, Boissiere L, Obeid I, et al. One-Level Lumbar Degenerative Spondylolisthesis and Posterior Approach: Is Transforaminal Lateral Interbody Fusion Mandatory?: A Randomized Controlled Trial With 2-Year Follow-Up. Spine. 2017;42(8):531-539. [16] Fujimori T, Le H, Schairer WW, et al. Does Transforaminal Lumbar Interbody Fusion Have Advantages over Posterolateral Lumbar Fusion for Degenerative Spondylolisthesis? Global Spine J 2015;5:102-9. 10 Page 10 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
[17] Sharma AK, Kepler CK, Girardi FP, Cammisa FP, Huang RC, Sama AA. Lateral lumbar interbody fusion: clinical and radiographic outcomes at 1 year: a preliminary report. J Spinal Disord Tech. 2011;24(4):242-50. [18] Acosta FL, Liu J, Slimack N, Moller D, Fessler R, Koski T. Changes in coronal and sagittal plane alignment following minimally invasive direct lateral interbody fusion for the treatment of degenerative lumbar disease in adults: a radiographic study. J Neurosurg Spine. 2011;15(1):92-6. [19] Kepler CK, Huang RC, Sharma AK, et al. Factors influencing segmental lumbar lordosis after lateral transpsoas interbody fusion. Orthop Surg. 2012;4(2):71-5. [20] Lee YS, Park SW, Kim YB. Direct lateral lumbar interbody fusion: clinical and radiological outcomes. J Korean Neurosurg Soc. 2014;55(5):248-54. [21] Johnson RD, Valore A, Villaminar A, Comisso M, Balsano M. Pelvic parameters of sagittal balance in extreme lateral interbody fusion for degenerative lumbar disc disease. J Clin Neurosci. 2013;20(4):576-81. [22] Khajavi K, Shen AY. Two-year radiographic and clinical outcomes of a minimally invasive, lateral, transpsoas approach for anterior lumbar interbody fusion in the treatment of adult degenerative scoliosis. Eur Spine J. 2014;23(6):1215-23. [23] Malham GM, Ellis NJ, Parker RM, et al. Maintenance of Segmental Lordosis and Disk Height in Stand-alone and Instrumented Extreme Lateral Interbody Fusion (XLIF). Clin Spine Surg. 2017;30(2):E90-E98. [24] Blizzard DJ, Gallizzi MA, Sheets C, et al. Sagittal Balance Correction in Lateral Interbody Fusion for Degenerative Scoliosis. Int J Spine Surg. 2016;10:29. [25] Phan K, Rao PJ, Scherman DB, Dandie G, Mobbs RJ. Lateral lumbar interbody fusion for sagittal balance correction and spinal deformity. J Clin Neurosci. 2015;22(11):1714-21. [26] Manwaring JC, Bach K, Ahmadian AA, et al. Management of sagittal balance in adult spinal deformity with minimally invasive anterolateral lumbar interbody fusion: a preliminary radiographic study. J Neurosurg Spine 2014;20:515-22. [27] Jiang SD, Chen JW, Jiang LS. Which procedure is better for lumbar interbody fusion: anterior lumbar interbody fusion or transforaminal lumbar interbody fusion?. Arch Orthop Trauma Surg. 2012;132(9):1259-66. [28] Phan K, Thayaparan GK, Mobbs RJ. Anterior lumbar interbody fusion versus transforaminal lumbar interbody fusion--systematic review and meta-analysis. Br J Neurosurg. 2015;29(5):705-11. [29] Ajiboye RM, Alas H, Mosich GM, Sharma A, Pourtaheri S. Radiographic and Clinical Outcomes of Anterior and Transforaminal Lumbar Interbody Fusions: A Systematic Review and Meta-analysis of Comparative Studies. Clin Spine Surg. 2017; [30] Hsieh PC, Koski TR, O'shaughnessy BA, et al. Anterior lumbar interbody fusion in comparison with transforaminal lumbar interbody fusion: implications for the restoration of foraminal height, local disc angle, lumbar lordosis, and sagittal balance. J Neurosurg Spine. 2007;7(4):379-86. [31] Kim JS, Lee KY, Lee SH, Lee HY. Which lumbar interbody fusion technique is better in terms of level for the treatment of unstable isthmic spondylolisthesis?. J Neurosurg Spine. 2010;12(2):171-7. [32] Watkins RG 4th, Hanna R, Chang D, et al. Sagittal alignment after lumbar interbody fusion: comparing anterior, lateral, and transforaminal approaches. J Spinal Disord Tech 2014;27:253-6. 11 Page 11 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
[33] Sembrano JN, Yson SC, Horazdovsky RD, et al. Radiographic Comparison of Lateral Lumbar Interbody Fusion Versus Traditional Fusion Approaches: Analysis of Sagittal Contour Change. Int J Spine Surg 2015;9:16. [34] Teng I, Han J, Phan K, Mobbs R. A meta-analysis comparing ALIF, PLIF, TLIF and LLIF. J Clin Neurosci. 2017. [35] Gödde S, Fritsch E, Dienst M, Kohn D. Influence of cage geometry on sagittal alignment in instrumented posterior lumbar interbody fusion. Spine. 2003;28(15):1693-9. [36] Barone G, Scaramuzzo L, Zagra A, Giudici F, Perna A, Proietti L. Adult spinal deformity: effectiveness of interbody lordotic cages to restore disc angle and spino-pelvic parameters through completely mini-invasive trans-psoas and hybrid approach. Eur Spine J. 2017. [37] Sembrano JN, Horazdovsky RD, Sharma AK, Yson SC, Santos ER, Polly DW. Do lordotic Cages Provide Better Segmental Lordosis Versus Non-lordotic Cages in Lateral Lumbar Interbody Fusion (LLIF)?. Clin Spine Surg. 2016. [38] Uribe JS, Harris JE, Beckman JM, Turner AW, Mundis GM, Akbarnia BA. Finite element analysis of lordosis restoration with anterior longitudinal ligament release and lateral hyperlordotic cage placement. Eur Spine J. 2015;24 Suppl 3:420-6. [39] Melikian R, Yoon ST, Kim JY, Park KY, Yoon C, Hutton W. Sagittal Plane Correction Using the Lateral Transpsoas Approach: A Biomechanical Study on the Effect of Cage Angle and Surgical Technique on Segmental Lordosis. Spine. 2016. [40] Uribe JS, Smith DA, Dakwar E, et al. Lordosis restoration after anterior longitudinal ligament release and placement of lateral hyperlordotic interbody cages during the minimally invasive lateral transpsoas approach: a radiographic study in cadavers. J Neurosurg Spine. 2012;17(5):476-85. [41] Hong TH, Cho KJ, Kim YT, Park JW, Seo BH, Kim NC. Does Lordotic Angle of Cage Determine Lumbar Lordosis in Lumbar Interbody Fusion?. Spine. 2017;42(13):E775-E780. [42] Anand N, Cohen RB, Cohen J, Kahndehroo B, Kahwaty S, Baron E. The Influence of Lordotic cages on creating Sagittal Balance in the CMIS treatment of Adult Spinal Deformity. Int J Spine Surg. 2017;11:23. [43] Fingar KR, Stocks C, Weiss AJ, & Steiner CA. Most frequent operating room procedures performed in US hospitals, 2003–2012. HCUP Statistical Brief 2014; 186. [44] Rao PJ, Maharaj MM, Phan K, Lakshan abeygunasekara M, Mobbs RJ. Indirect foraminal decompression after anterior lumbar interbody fusion: a prospective radiographic study using a new pedicle-to-pedicle technique. Spine J. 2015;15(5):817-24. [45] Kepler CK, Sharma AK, Huang RC, et al. Indirect foraminal decompression after lateral transpsoas interbody fusion. J Neurosurg Spine. 2012;16(4):329-33. [46] Pawar AY, Hughes AP, Sama AA, Girardi FP, Lebl DR, Cammisa FP. A Comparative Study of Lateral Lumbar Interbody Fusion and Posterior Lumbar Interbody Fusion in Degenerative Lumbar Spondylolisthesis. Asian Spine J. 2015;9(5):668-74. [47] Jackson RP, Mcmanus AC. Radiographic analysis of sagittal plane alignment and balance in standing volunteers and patients with low back pain matched for age, sex, and size. A prospective controlled clinical study. Spine. 1994;19(14):1611-8. [48] Rice JW, Sedney CL, Daffner SD, Arner JW, Emery SE France JC. Improvement of Segmental Lordosis in Transforaminal Lumbar Interbody Fusion: A comparison of Two Techniques. Global Spine J. 2016 May;6(3):229-33. 12 Page 12 of 27
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Fig 1. Radiographic Measurements. SL is the angle between the superior and inferior endplates of adjacent superior and inferior vertebral bodies, respectively (1). LL is the angle between the superior S1 and L1 endplates (2). Pelvic incidence is the angle between a line from the center of the S1 endplate to the center of the femoral heads and a line perpendicular to the center of the S1 endplate (3). Disk height was measured as the minimum distance (mm) between the superior and inferior end plates of adjacent vertebral bodies at their anterior and posterior surfaces (4). Foraminal height was measured as the maximum distance between the inferior margin of the pedicle of the superior vertebra and the superior margin of the pedicle of the inferior vertebra (5). Fig. 2. Paired sample t-test analysis of pre-operative vs. post-op radiographic parameters SL, LL, DHA/P, and FH by surgical approaches ALIF (A), LLIF (B), TLIF (C), PLF (D). Parameter ± SD, error bars reflect 95% CI. Fig. 3. ANOVA for mean degree change in radiographic parameters SL (A), LL (B), DH-A/P (C), and FH (D) by surgical approach. Fig. 4. Change in proportion of patients with PI-LL < 10° by surgical approach. Fig. 5. Relative degree change of radiographic parameters for lordotic (n = 58) vs. non-lordotic cages (n = 106) in ALIF, LLIF, and TLIF surgeries.
13 Page 13 of 27
1 2
Table 1. Frequency of primary diagnoses for each surgical approach including degenerative disk disease, spondylolisthesis, isthmic spondylolisthesis, adjacent level disease, pseudoarthrosis, and scoliosis.
ALIF
LLIF
TLIF
PLF
Total
Degenerative disk disease
14
4
5
2
25
Isthmic spondylolisthesis
2
0
12
1
15
Spondylolisthesis
17
7
41
35
100
Adjacent level disease
0
10
0
2
12
Pseudoarthrosis
1
0
1
1
3
Scoliosis
0
2
1
1
4
Other
0
0
3
2
5
Total
34
23
63
44
164
Primary Diagnosis
3 4 5
Table 2. Comparison of the number of lordotic and non-lordotic implants for each surgical approach.
ALIF
LLIF
TLIF
PLF
Total
Y
34
18
6
0
58
N
0
5
57
44
106
Total
34
23
63
44
164
Lordotic Implant
6 7 8
Table 3. Number of fusions at a given surgical level for each surgical approach.
ALIF
LLIF
TLIF
PLF
Total
L1-2
0
0
0
1
1
L2-3
0
6
2
2
10
L3-4
3
11
3
9
26
L4-5
8
6
43
26
83
L5-S1
23
0
15
6
44
Total
34
23
63
44
164
Surgical Level
9
14 Page 14 of 27
1
Table 4. Number of lordotic implants with a given degree lordosis for ALIF, LLIF, and TLIF.
ALIF
LLIF
TLIF
Total
0
0
5
57
62
5
3
0
4
7
6
1
5
1
7
7
0
7
0
7
8
4
5
0
9
9
2
0
0
2
10
4
1
0
5
12
8
0
1
9
14
2
0
0
2
15
10
0
0
10
Total
34
23
63
120
Implant Lordosis (°)
2 3 4 5
6 7 8
Table 5. Slope values for implant lordosis, surgical approach, and surgical level in predictive equation for post-operative segmental lordosis as derived from multiple linear regression.
β
Variable
B
SEB
Sig.
Intercept
11.406
4.789
Implant Lordosis (°)
1.089
.369
.342
.005
Approach*
6.835
2.739
.306
.016
Surgical Level**
6.874
2.575
.274
.010
.021
Note. * 0 = LLIF/TLIF, 1 = ALIF; ** 0 = L2-3/L3-4, 1 = L4-5/L5-S1 B = unstandardized regression coefficient; SEB = standard error of coefficient; β = standardized coefficient
9
15 Page 15 of 27
1
Table 6. Comparison of SL and LL based on pre-operative disk height
ΔSL
ΔLL
< 6.0 mm (n=78)
4.7
3.7
> 6.0 mm (n=86)
1.1
2.6
< .001
.34
Average Pre-op DH
P values 2 3 4 5 6 7 8 9 10 11 12
Eq 1. Multiple linear regression equation for predicting post-operative segmental lordosis from the variables of implant lordosis, surgical approach, and surgical level.
16 Page 16 of 27
1 2 3
TSJ LIF Figure 1.jpg
17 Page 17 of 27
1 2 3
TSJ LIF Figure 2A.jpg
18 Page 18 of 27
1 2 3
TSJ LIF Figure 2B.jpg
19 Page 19 of 27
1 2 3
TSJ LIF Figure 2C.jpg
20 Page 20 of 27
1 2 3
TSJ LIF Figure 2D.jpg
21 Page 21 of 27
1 2 3
TSJ LIF Figure 3A.jpg
22 Page 22 of 27
1 2 3
TSJ LIF Figure 3B.jpg
23 Page 23 of 27
1 2 3
TSJ LIF Figure 3C.jpg
24 Page 24 of 27
1 2 3
TSJ LIF Figure 4.jpg
25 Page 25 of 27
1 2 3
TSJ LIF Figure 5.jpg
26 Page 26 of 27
1 2
TSJ LIF Figure3D.jpg
27 Page 27 of 27