Extreme lateral interbody fusion (ELIF) represents

Transcription

1 J Neurosurg Spine 20: , 2014 AANS, 2014 Radiological and clinical outcomes following extreme lateral interbody fusion Clinical article Marjan Alimi, M.D., 1 Christoph P. Hofstetter, M.D., Ph.D., 1 Guang-Ting Cong, B.S., 3 Apostolos John Tsiouris, M.D., 2 Andrew R. James, M.B.B.S., F.R.C.S., 1 Danika Paulo, B.S., 1 Eric Elowitz, M.D., 1 and Roger Härtl, M.D. 1 1 Weill Cornell Brain and Spine Institute, Department of Neurological Surgery; 2 Department of Neuroradiology; and 3 Weill Cornell Medical College, NewYork-Presbyterian Hospital, New York, New York Object. Extreme lateral interbody fusion (ELIF) is a popular technique for anterior fixation of the thoracolumbar spine. Clinical and radiological outcome studies are required to assess safety and efficacy. The aim of this study was to describe the functional and radiological impact of ELIF in a degenerative disc disease population with a longer follow-up and to assess the durability of this procedure. Methods. Demographic and perioperative data for all patients who had undergone ELIF for degenerative lumbar disorders between 2007 and 2011 were collected. Trauma and tumor cases were excluded. For radiological outcome, the preoperative, immediate postoperative, and latest follow-up coronal Cobb angle, lumbar sagittal lordosis, bilateral foraminal heights, and disc heights were measured. Pelvic incidence (PI) and PI lumbar lordosis (PI-LL) mismatch were assessed in scoliotic patients. Clinical outcome was evaluated using the Oswestry Disability Index (ODI) and visual analog scale (VAS), as well as the Macnab criteria. Results. One hundred forty-five vertebral levels were surgically treated in 90 patients. Pedicle screw and rod constructs and lateral plates were used to stabilize fixation in 77% and 13% of cases, respectively. Ten percent of cases involved stand-alone cages. At an average radiological follow-up of 12.6 months, the coronal Cobb angle was 10.6 compared with 23.8 preoperatively (p < ). Lumbar sagittal lordosis increased by 5.3 postoperatively (p < ) and by 2.9 at the latest follow-up (p = 0.014). Foraminal height and disc height increased by 4 mm (p < ) and 3.3 mm (p < ), respectively, immediately after surgery and remained significantly improved at the last follow-up. Separate evaluation of scoliotic patients showed no statistically significant improvement in PI and PI-LL mismatch either immediately postoperatively or at the latest follow-up. Clinical evaluation at an average follow-up of 17.6 months revealed an improvement in the ODI and the VAS scores for back, buttock, and leg pain by 21.1% and 3.7, 3.6, and 3.7 points, respectively (p < ). According to the Macnab criteria, 84.8% of patients had an excellent, good, or fair functional outcome. New postoperative thigh numbness and weakness was detected in 4.4% and 2.2% of the patients, respectively, which resolved within the first 3 months after surgery in all but 1 case. Conclusions. This study provides what is to the authors knowledge the most comprehensive set of radiological and clinical outcomes of ELIF in a fairly large population at a midterm follow-up. Extreme lateral interbody fusion showed good clinical outcomes with a low complication rate. The procedure allows for at least midterm clinically effective restoration of disc and foraminal heights. Improvement in coronal deformity and a small but significant increase in sagittal lordosis were observed. Nonetheless, no significant improvement in the PI-LL mismatch was achieved in scoliotic patients. ( Key Words extreme lateral interbody fusion foraminal height disc height Cobb angle lumbar lordosis Oswestry Disability Index visual analog scale Abbreviations used in this paper: ALIF = anterior lumbar interbody fusion; BMI = body mass index; DDD = degenerative disc disease; ELIF = extreme lateral interbody fusion; LL = lumbar lordosis; MCID = minimum clinically important difference; ODI = Oswestry Disability Index; PI = pelvic incidence; VAS = visual analog scale. Extreme lateral interbody fusion (ELIF) represents an alternative technique for managing many different pathologies of the thoracic and lumbar spine such as degenerative disc disease (DDD), deformity, tumor, trauma, and infection. 1,3,4,8,13,24,34,36,40,50,55 This approach developed as a combination of traditional anterior 623

2 M. Alimi et al. lumbar interbody fusion (ALIF) procedures, minimally invasive laparoscopic techniques, and tubular lumbar decompression surgery. In 2006, Ozgur et al. 40 reported on a mini-open technique from a direct lateral transpsoas approach for the treatment of pathologies affecting the midlumbar spine, in which they used electrophysiological monitoring to avoid nerve damage in the placement of structural interbody fusion cages. This approach was based on Pimenta s work (Pimenta L, paper presented at the Brazilian Spine Society Meeting, 2001), who in 2001 presented his preliminary results with this technique. Ozgur and colleagues called the approach extreme lateral interbody fusion. Their technique, which involves triggered electromyographic nerve monitoring and a table-mounted split-blade retractor system, has become the procedure commonly used for lateral access to the midlumbar spine, providing access to the levels between T-4 and L-5. Although initial results were very promising, longer-term results and the overall safety profile of this technique must still be evaluated. The aim of this study was to describe the functional and radiological impact of ELIF in a DDD population with a longer follow-up and to assess the durability of this procedure. Patient Population Methods Using the prospectively maintained clinical database of the hospital, including the Epic, Eclipsys, and PACS systems, we performed an institutional review board approved retrospective cohort study. From this database, we obtained the demographic and perioperative data for all patients with degenerative lumbar disorders who had undergone ELIF performed by two different surgeons at a single center between 2007 and Trauma and tumor cases were excluded. Age, sex, body mass index (BMI), smoking status, history of diabetes, steroid medication use, history of previous lumbar spine surgeries, and indication for surgery were recorded. Operative reports for each case were reviewed to extract data on the number and levels of vertebrae treated surgically, presence and type of additional instrumentation, timing of instrumentation placement in relation to ELIF (same vs different day), diameter and height of interbody cage, and occurrence of CSF leakage. Length of hospital stay was determined by reviewing patient discharge documents. Operative Procedure The procedure has been described in great detail elsewhere. 15,40 In brief, patients under general anesthesia were positioned laterally, and fluoroscopy was brought in for localization. A small incision was made on the flank overlying the target level. Sharp and blunt dissections were used to cut through the abdominal muscles and into the retroperitoneal space. Intraoperative neurophysiological monitoring was performed throughout the procedure to keep the surgeon in the safe region and to avoid the possible risks of lumbar plexus or nerve root injury. The psoas muscle was gently penetrated with a blunt dilator. The lateral disc area was identified and exposed using the NuVasive MaXcess minimally invasive retractor. Discectomy was performed, and a NuVasive polyetheretherketone (PEEK) cage filled with silicated calcium phosphate (Actifuse, Apatech, Baxter) was placed into the discectomy defect. For anterior instrumentation cases, a DePuy Synthes anterior tension band system with 4 bone screws was placed. For posterior instrumentation cases, open or percutaneous, uni- or bilateral pedicle screw instrumentation was placed, and posterolateral fusion was performed. In selected cases an additional laminectomy was performed. Intraoperative spinal navigation (Brain- LAB) based on intraoperative 3D images was frequently used to implant the posterior instrumentation. Radiological Evaluation Radiological parameters were measured on preoperative, immediate postoperative, and most recent follow-up images. Computed tomography scans, radiographs, or MR images (first, second, and third priority, respectively) were used for radiological measurements, depending on each patient s available imaging studies. To measure the coronal Cobb angle, the superior and inferior end vertebrae of the surgical region were determined and the angle between the cephalad endplate of the top vertebra and the caudal endplate of the bottom vertebra was measured. Lumbar sagittal lordosis was measured between the cephalad endplate of L-1 and the cephalad endplate of the sacrum (Fig. 1A and B). Foraminal heights were measured on both the right and left side. When using CT scans and MR images, we attempted to select the best cut on each side, showing the highest foraminal height with a completely sharp bone border (Fig. 1C). Disc heights were measured on lateral views by determining the mean value of the anterior and posterior disc heights for each level (Fig. 1D). To specifically evaluate the impact of ELIF on alignment, pelvic incidence (PI) was measured in scoliotic patients, that is, those who by definition had a preoperative coronal Cobb angle 10. Lateral lumbar spine radiographs at each time point were used to measure the PI, the angle between a line drawn perpendicular from the middle point of the sacral plate and a line drawn from the same point to an average point between the centers of the two femoral heads. Pelvic incidence lumbar lordosis (PI-LL) mismatch was then calculated in this subgroup of patients. We specifically performed this analysis because previous multicenter prospective studies have shown a strong correlation between PI-LL mismatch and outcome in patients with spinal deformity. 48,51,52 A PI-LL mismatch 11 has been associated with a significantly higher level of disability. 48 An independent radiologist (J.T.) evaluated bony fusion on CT scans or flexion and extension lateral radiographs. Fusion on CT scans was determined through the detection of bridging bone on a minimum of 2 contiguous sections and in at least 2 of 3 planes (axial, sagittal, or coronal). 11 Fusion on flexion and extension radiographs was determined when less than 4 of angular motion was observed. 9,33 For positive bony fusion, the absence of a dark halo surrounding the implant or instrumentation fracture was required. 37,46 Fusion rate was assessed in 73% of the levels (109 levels) with a minimum follow-up of 1 year. 624

3 Extreme lateral interbody fusion outcomes Fig. 1. Methods for radiological measurements. A: Coronal Cobb angle. B: Lumbar sagittal lordosis. C: Foraminal height. D: Disc height. E: Pelvic incidence. Clinical Outcome Evaluation We compared preoperative and most recent followup values for the Oswestry Disability Index (ODI) 16 and visual analog scale (VAS) scores for back, buttock, and leg pain (on both sides for the latter two). The minimum clinically important difference (MCID) was defined as a 12-point or more change in the ODI 7,19,22 and a 3-point or more change in the VAS score. 17,18,22 The Macnab criteria were used to evaluate functional outcome. 35 For each patient, the postoperative inpatient notes, first follow-up office notes, as well as the most recent office visit notes were reviewed to determine the neurological symptoms and signs. In the event of neurological symptoms and signs in a patient, the last preoperative office note was also reviewed for comparison. Only new thigh numbness and motor weakness were recorded as complications on the grounds that these symptoms had been present in a number of patients prior to surgery as a result of their specific pathology and thus could not have been attributed to the surgical procedure. Statistical Analysis Statistical analysis was performed to evaluate the impact of ELIF on each radiological parameter immediately postoperatively and at the latest follow-up. Continuous variables are shown as the means ± standard deviation from the means. For continuous variables with repeated measurements, differences were assessed using the paired t-test. Correlations between continuous variables were examined using scatter diagrams and the Pearson correlation coefficient. For categorical variables, percentages were calculated. All analyses were performed using appropriate statistical software (SPSS version , SPSS Inc.). Patient Demographics Results Ninety patients, consisting of 55 females and 35 males, with a mean age of 64 years (range years) at the time of surgery were included in the study cohort (Table 1). The mean BMI was 27.6 kg/m 2. The main indications for surgery were DDD (35.6%), spondylolisthesis (26.7%), and degenerative scoliosis (24.4%). There was a positive history for smoking, diabetes, and steroid medication use in 10%, 14.4%, and 6.7% of the patients, respectively. A history of spine surgery was present at the index level(s) in 26.7% of the patients. Surgical Details Among 145 spine levels, L4 5 (40.7%) was the level most commonly treated, followed by L3 4 (27.6%) and L2 3 (22.1%; Table 2). Single-level surgery was performed in 57.8% of the patients, whereas 2-, 3-, and 4-level surgeries were performed in 18.9%, 15.6%, and 7.8%, respectively. Additional posterior instrumentation consisting of pedicle screws and rods was placed in the majority of cases (76.6%). Lateral plates were used in 12 patients (13.3%), and the remaining 9 patients (10%) re- 625

4 M. Alimi et al. TABLE 1: Characteristics of 90 patients who underwent ELIF for degenerative lumbar disorders ceived stand-alone cages. For patients with posterior instrumentation, unilateral screws were used to reinforce the constructs in 34 patients (49.3%), and 35 patients (50.7%) had bilateral screws. Additional laminectomy was performed in 29 patients (32.2%). Among the 145 levels, 18-mm-diameter cages were implanted in 82 levels (56.6%), and 22-mm-diameter cages were used in the other 63 levels (43.4%). Implanted cage heights varied from 8 to 14 mm. The most frequently used cage height was 10 mm (49.6% of the levels), followed by 8 mm (21.4%) and 12 mm (15.9%). There was no relationship between cage size and vertebral level treated. However, the cage height correlated with the preoperative disc height (p = 0.005, r = 0.297), as well as with the immediate postoperative and latest follow-up disc height (p < 0.001, r = 0.393; p < 0.001, r = 0.380, respectively). However, cage height did not correlate with subsidence at the latest follow-up. The mean hospital stay was 4.5 days (range 1 31 days), and 34.1% of the patients were discharged on their 1st postoperative day. Radiological Outcome Variable No. (%) mean age at surgery (yrs)* 64.4 ± sex male 35 (38.9) female 55 (61.1) mean BMI (kg/m 2 )* 27.6 ± 4.88 smoking yes 9 (10) no 81 (90) diabetes yes 13 (14.4) no 77 (85.6) steroid medication use yes 6 (6.7) no 84 (93.3) history of spine surgery yes 24 (26.7) no 66 (73.3) indication for surgery DDD 32 (35.6) spondylolisthesis 24 (26.7) degenerative scoliosis 22 (24.4) adjacent-level disease 7 (7.8) postlaminectomy syndrome 3 (3.3) instrumentation failure/nonfusion 2 (2.2) * Value is expressed as the mean ± standard deviation. The mean follow-up duration for radiological outcome was 12.6 months (range months). The images used for radiological measurements were CT scans TABLE 2: Surgical summary for 145 spine levels treated with ELIF* Variable No. (%) level of surgery (no. of cases) T (0.7) T12 L1 2 (1.4) L (7.6) L (22.1) L (27.6) L (40.7) no. of treated vertebral levels/patient 1 52 (57.8) 2 17 (18.9) 3 14 (15.6) 4 7 (7.8) additional instrumentation inserted (no. of patients) pst 69 (76.6) unilat screws 34 (49.3) bilat screws 35 (50.7) lat plate 12 (13.3) stand-alone cage 9 (10) timing of pst instrumentation placement in relation to ELIF (no. of patients) same day 52 (75.4) different day 17 (24.6) cage width in mm (no. of levels) (56.6) (43.4) cage height in mm (no. of levels) 8 31 (21.4) 9 7 (4.8) (49.6) 11 9 (6.2) (15.9) 13 1 (0.7) 14 2 (1.4) * pst = posterior. for 109 levels (75%), MR images for 3 levels (2%), and radiographs for 33 levels (23%). The mean coronal Cobb angle decreased by 13.5 ± 10.6 (p < ) postoperatively and by 13.1 ± 11 (p < ) at the latest follow-up, compared with the preoperative value (Table 3). The median increase in lumbar sagittal lordosis was 5.3 postoperatively (range -32 to 42, p < ) and 2.9 at the latest follow-up (range -29 to 41, p = 0.014; Fig. 2). Comparison between preoperative and immediate postoperative mean foraminal height showed a 4 ± 8.56 mm increase (p < ). At the latest follow-up foraminal height remained increased by 2 ± 2.5 mm (p < ). The preoperative mean disc height increased by 3.3 ± 2.0 mm immediately postoperatively (p < ). Given subsidence of 0.97 ± 1.20 mm at the 626

5 Extreme lateral interbody fusion outcomes TABLE 3: Radiological outcome for 90 patients who underwent ELIF* Variable Preop Postop p Value Last FU p Value mean coronal Cobb angle in degrees 23.8 ± ± 8.55 < ± 8.85 < mean lumbar sagittal lordosis in degrees 39.7 ± ± < ± mean foraminal height in mm 15.4 ± ± 4.1 < ± 3.7 < mean disc height in mm 4.1 ± ± 2.17 < ± 1.84 < * FU = follow-up. A p < 0.05 was considered statistically significant. latest follow-up, the average increase in disc height was 2.4 ± 2.2 mm at the latest follow-up, which showed that the disc height was still significantly higher than the preoperative value (p < ; Fig. 3). Radiological Subgroup Analyses Deformity Cases PI and PI-LL Mismatch. Twenty patients had had a preoperative coronal Cobb angle 10. The mean preoperative PI in these scoliotic patients was 50.5 ± 9, which changed to 51.8 ± 10.7 immediately postoperatively (p = 0.435) and to 50.8 ± 12.5 at the latest follow-up (p = 0.497). The average preoperative PI- LL mismatch was 14.1 ± 12.5, which decreased to 9.6 ± 15.3 immediately postoperatively (p = 0.256) and to 13.5 ± 13.7 at the latest follow-up (p = 0.860). Thus, there was no statistically significant improvement in PI and PI-LL mismatch either immediately postoperatively or at the latest follow-up. Deformity Versus Nondeformity Degenerative Cases. We performed a subgroup analysis to compare outcome in the deformity cases (20 cases with preoperative coronal Cobb angle 10 ) with that in the nonscoliotic degenerative cases (Table 4). Unsurprisingly, improvement in the coronal Cobb angle was significantly higher in the deformity group; however, there were no other statistically significant differences in any other outcome parameters, either radiographic or clinical, between the two groups. One-Level Versus Two-Level Degenerative Cases. Similarly, a subgroup analysis was performed to compare the outcome of 1-level versus 2-level degeneration cases (Table 5). No statistically significant differences were found between these two groups in any of the radiographic or clinical parameters. Stand-Alone Cage Versus Circumferential Fusion. Subgroup analysis was performed to compare the outcome of stand-alone versus circumferential fusion cases (Table 6). There were no statistically significant differences between the two groups in any radiological or clinical parameters, except for the latest follow-up coronal Cobb angle, which was significantly higher in circumferential fusion cases than in the stand-alone cases, suggesting that coronal alignment correction would be more effectively maintained over time with posterior instrumentation. Foraminal height and disc height subsidence (immediate postoperative to the latest follow-up decrease of the mentioned parameters) were not statistically different between the two groups (Figs. 4 and 5). To determine if the observed subsidence would have a clinical impact, additional statistical tests were performed; we assessed Fig. 2. Comparison among preoperative, postoperative, and last follow-up coronal Cobb angle (dark gray diamonds) and lumbar sagittal lordosis (light gray squares). Values on the y-axis represent degrees. *p < 0.05, compared with the preoperative value. Fig. 3. Comparison among preoperative, postoperative, and last follow-up foraminal height (dark gray diamonds) and disc height (light gray squares). Values on the y-axis represent millimeters. *p < 0.05, compared with the preoperative value. 627

6 M. Alimi et al. TABLE 4: Radiological and clinical variables in deformity compared with nondeformity degenerative cases Variable Deformity Nondeformity Degenerative p Value total no. of cases mean coronal Cobb angle in degrees preop to postop decrease 14.7 ± ± 0.6 <0.001* preop to latest FU decrease 14.3 ± ± 0.7 <0.001* mean lumbar sagittal lordosis in degrees preop to postop increase 5.7 ± ± preop to latest FU increase 2.4 ± ± mean foraminal height in mm preop to postop increase 3.0 ± ± preop to latest FU increase 2.2 ± ± mean disc height in mm preop to postop increase 3.5 ± ± preop to latest FU increase 2.7 ± ± preop to latest FU improvement mean ODI (%) 15.1 ± ± mean VAS back pain score 3.5 ± ± mean VAS buttock pain score 3.2 ± ± mean VAS leg pain score 3.0 ± ± * A p < 0.05 was considered statistically significant. the correlation between foraminal height subsidence and radicular pain on each side, as well as the average disc height subsidence and radicular pain. Interestingly, no correlations were found between subsidence and the development of recurrent radicular pain. Stand-Alone Versus Unilateral Instrumented Versus Bilateral Posterior Pedicle Screw Fixation Cases. To assess the impact of instrumentation type on outcome, we compared foraminal and disc height subsidence as well as clinical outcome among the 3 groups of stand-alone, unilaterally instrumented, and bilateral posterior pedicle screw fixation cases (Table 7). Statistical analysis showed no significant differences in any of the parameters among the 3 groups. Given the notable dissimilarity of the 3 groups in terms of sample size, additional specific statistical tests were performed to assess the presence of statis- TABLE 5: Radiological and clinical variables in 1- compared with 2-level degeneration cases Variable 1-Level Degeneration 2-Level Degeneration p Value no. of cases mean coronal Cobb angle in degrees preop to postop decrease 6.2 ± ± preop to latest FU decrease 5.3 ± ± mean lumbar sagittal lordosis in degrees preop to postop increase 14.3 ± ± preop to latest FU increase 3.1 ± ± mean foraminal height in mm preop to postop increase 3.0 ± ± preop to latest FU increase 1.4 ± ± mean disc height in mm preop to postop increase 3.1 ± ± preop to latest FU increase 1.9 ± ± preop to latest FU improvement mean ODI (%) 21.1 ± ± mean VAS back pain score 3.3 ± ± mean VAS buttock pain score 3.2 ± ± mean VAS leg pain score 3.8 ± ±

7 Extreme lateral interbody fusion outcomes TABLE 6: Radiological and clinical variables in stand-alone versus circumferential fusion groups Variable Stand-Alone Cage Circumferential Fusion p Value no. of cases 9 81 mean coronal Cobb angle in degrees preop to postop decrease 12.8 ± ± preop to the latest FU decrease 5.8 ± ± * mean lumbar sagittal lordosis in degrees preop to postop increase 5.2 ± ± preop to latest FU increase 6.3 ± ± mean foraminal height in mm preop to postop increase 3.8 ± ± preop to latest FU increase 1.9 ± ± mean disc height in mm preop to postop increase 3.2 ± ± preop to latest FU increase 2.5 ± ± preop to latest FU improvement mean ODI (%) 18.5 ± ± mean VAS back pain score 4.3 ± ± mean VAS buttock pain score 4.5 ± ± mean VAS leg pain score 3.0 ± ± immediate postop to latest FU decrease mean foraminal height in mm 1.8 ± ± mean disc height in mm 0.74 ± ± * A p < 0.05 was considered statistically significant. tically significant differences with homogeneous groups. Homogeneous subsets did not show any significant differences either. In 109 of 145 spine levels, the fusion rate was 89% at an average follow-up of 15.7 ± 10.1 months. The current study was based on ELIF cases treated by two different surgeons, one who evaluated fusion using CT scans and the other who did so only via follow-up radiographs; therefore, fusion could not be evaluated for the 25% of the levels with follow-up imaging other than CT. To partially make up for this lack of data, we performed a subgroup analysis in which we reported the clinical outcome for the nonevaluated levels (Table 8). In summary, all 4 radiographic parameters, including coronal Cobb angle, lumbar sagittal lordosis, foraminal Fig. 4. Foraminal height restoration outcome in stand-alone (dark gray diamonds) versus instrumented (light gray squares) cases: comparison among preoperative, postoperative, and last follow-up in each group. Values on the y-axis represent millimeters. NS = not significant. Fig. 5. Disc height restoration outcome in stand-alone (light gray triangles) versus instrumented (dark gray circles) cases: comparison among preoperative, postoperative, and last follow-up in each group. Values on the y-axis represent millimeters. NS = not significant. 629

8 M. Alimi et al. TABLE 7: Impact of instrumentation type on subsidence and clinical outcome Parameter Stand-Alone Unilat Instrumented Bilat Pst Pedicle Screw Fixation p Value no. of cases immediate postop to latest FU decrease mean foraminal height in mm 1.8 ± ± ± mean disc height in mm 0.7 ± ± ± preop to latest FU improvement mean ODI (%) 18.6 ± ± ± mean VAS back pain score 4.3 ± ± ± mean VAS buttock pain score 4.6 ± ± ± mean VAS leg pain score 3.0 ± ± ± height, and disc height, showed significant improvement in all patients immediately after surgery and at the latest follow-up compared with preoperative values (Table 3). However, in the scoliotic group of patients, comparisons of PI and PI-LL mismatch at the 3 time points showed no statistically significant differences. Clinical Outcome The mean follow-up for clinical outcome in 70 evaluable patients was 17.6 ± 12.4 months (range months). The average preoperative ODI was 50% ± 17.5%, which was improved by 21.12% ± 19.20% at the latest follow-up (p < ; Fig. 6 and Table 9). An MCID, defined as an improvement in the ODI by 12 points, 7,19,22 was achieved in 72.8% of the patients. The preoperative average VAS scores for back, buttock, and leg pain were 7 ± 2.89, 6.1 ± 3.54, and 6.4 ± 3.29 points, respectively; at the last follow-up they had improved by 3.76 ± 3.95, 3.61 ± 4.02, and 3.76 ± 4.10 points, respectively (p < for all 3 parameters; Fig. 6 and Table 9). An MCID, defined as improvement in the VAS score by 3 points, 17,18,22 was found in 67.1% of patients for back pain, in 57.1% for buttock pain, and in 61.4% for leg pain. TABLE 8: Clinical outcome for 36 vertebral levels in which fusion evaluation had not been possible Statistical tests showed no correlations between the amount of improvement in the VAS scores and the extent of increase in the foraminal and disc height at the latest follow-up. According to the Macnab criteria, 35 which provided subjective functional outcome assessment, 84.8% of 46 patients had an excellent (54.3%), good (19.6%), or fair (10.9%) outcome. Surgeries were generally well tolerated, and there were no intraoperative complications, no cases of femoral nerve paralysis, bowel injury, or abdominal flank bulge. Nonneurological inpatient complications included one case of myocardial infarction and one case of adynamic ileus. Two patients (2.2%) suffered from new postoperative lower-extremity weakness and decreased sensation on the 2nd day after surgery. In one patient, emergent imaging revealed a small bone fragment in the right L2 3 foramen. The patient was taken back to the operating room emergently, the bony fragment was removed, and nerve root decompression was performed. Thereafter, the patient clinically improved. In the second patient, the weakness did not improve in the short term, although it had completely resolved at the 3-year follow-up. Four patients (4.4%) experienced postoperative thigh numbness, which Parameter Preop Last FU p Value ODI (%) mean 42.7 ± ± 23.7 <0.0001* cases w/ MCID 77% VAS back score mean 6.5 ± ± 3.4 <0.0001* cases w/ MCID 69.2% VAS buttock score mean 5.9 ± ± 2.3 <0.0001* cases w/ MCID 65.4% VAS leg score mean 6.7 ± ± 3.2 <0.0001* cases w/ MCID 61.5% * A p < 0.05 was considered statistically significant. Fig. 6. Comparison between preoperative (dark gray bars) and last follow-up (light gray bars) ODI and VAS scores for back, buttock, and leg pain. Values on the y-axis represent scores. *p < 0.05, compared with the preoperative value. 630

9 Extreme lateral interbody fusion outcomes TABLE 9: Clinical outcome in 70 patients who underwent ELIF and had available preoperative and FU scores Parameter Preop Last FU p Value ODI (%) 50 ± ± <0.0001* patients w/ MCID 51 (72.8%) VAS back score 7 ± ± 3.16 <0.0001* patients w/ MCID 47 (67.1%) VAS buttock 6.1 ± ± 2.96 <0.0001* patients w/ MCID 40 (57.1%) VAS leg 6.4 ± ± 3.15 <0.0001* patients w/ MCID 43 (61.4%) * A p < 0.05 was considered statistically significant. resolved by 3 months (range months) after surgery in all cases. Reoperation was performed in 12 patients for nonunion (8 patients), adjacent-level disease (3 patients), and postlaminectomy syndrome and radiculopathy (1 patient). Statistical tests were performed to find out if any specific parameters in the nonunion cases were different from the rest of the study group. No significant differences were found between these two groups in terms of age, sex, history of smoking, diabetes, steroid medication use, history of previous surgery, number of surgical levels, presence or absence of instrumentation, preoperative coronal deformity, preoperative ODI, and preoperative VAS back, buttock, and leg pain scores. Discussion The extreme lateral approach for interbody fusion was introduced to treat various pathologies of the thoracolumbar spine. 1,3,4,8,13,14,24,26,34,36,40,42,44,50,55 Like many other minimally invasive techniques, ELIF has been associated with less blood loss, shorter hospitalizations, and earlier mobilization of patients than similar open surgery. 3,4,8,13,14,34,39,44,47 The particular advantages of ELIF can be attributed to its power in decompressing neural elements and restoring disc height without violating the stabilizing structures of the spine. The procedure reintroduced the concept of indirect decompression of the spinal canal, which had been described in ALIF surgery. 23 Similar results were reported for ELIF with a significant increase in dural sac dimensions, possibly due to stretching and unbuckling of the spinal ligaments and to a decrease in intervertebral disc bulging. 15 The extreme lateral approach to the lumbar spine allows the placement of larger interbody cages than those possible with transforaminal lumbar interbody fusion (TLIF) and, unlike ALIF, avoids the need to retract great vessels. Our data showed that ELIF can partially restore many of the radiological parameters and that corrections are durable at the midterm follow-up. Comparison of Our Results With Previously Published Findings Oswestry Disability Index results in our study corroborated and were comparable to findings in previous studies with relatively similar follow-up durations. Our data showed that on average there was 21% improvement in the ODI 18 months after surgery. 1,4,13,15,29,30,32 Visual analog scale scores were also comparable with previously published data on ELIF outcome. With our clinical follow-up of 18 months, the average improvements in back, buttock, and leg VAS scores were 3.8, 3.6, and 3.6, respectively. 1,4,13,15,29,32,54 It seems that the smaller decrease in ODI and VAS scores, as compared with those in other studies, 44 may be due to the patient population (Table 10); almost one-third of our patients had undergone previous spinal surgery, and many patients underwent multilevel front and/or back procedures for relatively severe pathology. We observed an 83% increase in the intervertebral disc space and a 26% increase in foraminal height on the immediate postoperative imaging, which compare favorably to data previously published by Oliveira et al. (41.9% and 13.5%, respectively; Table 11). 39 Kepler and colleagues measured preoperative and postoperative foraminal area in 29 ELIF patients at an average follow-up of 4 (9 patients) and 18 months (20 patients). 30 They observed a 35% increase in the foraminal area. It is important to note that they found no significant differences in foraminal area improvement between the 4- and 18-month follow-up groups, suggesting that the radiographic improvement in ELIF was durable. This finding corroborates our results, which showed significant improvement in disc and foraminal height on both the immediate postoperative and the latest follow-up imaging. Interestingly, we did not see a difference in radiological outcome between stand-alone cases and patients with additional lateral or posterior fixation. This result may be attributable to the small number of stand-alone cases and is currently being studied further. The amount of subsidence in stand-alone and instrumented cases was 11.6% and 11.4%, respectively, at the latest follow-up. Our findings were comparable to but slightly lower than the values found in Pimenta et al. s prospective study on stand-alone ELIF cases. 43 They reported 17% subsidence after 36 months. The higher subsidence rate in their study may be partially attributed to their longer follow-up. In general, the results on disc height in our study agreed with those in similar previous studies (Table 11). 1,4,30,32,39,54,57 In our study we found no correlations between foraminal height and disc height subsidence and the development of recurrent radicular pain. Despite a partial loss in foraminal and disc height, in general patients were doing well clinically, suggesting that subsidence does not necessarily lead to recurrent radicular pain. However, one must keep in mind that symptoms may start if the extent of subsidence exceeds a certain threshold. Results in the current study are consistent with those of Acosta et al., who found that ELIF was associated with significant improvement in segmental, regional, and global coronal plane alignment in patients with degenerative lumbar disease. 1 Their study showed a mean global coronal alignment of 19.1 preoperatively and 12.5 on immediate postoperative imaging (p < 0.05). The improvement in coronal plane alignment was less than but comparable to that in our study, with preoperative and postoperative coronal alignments of 23.8 and 10.3, respectively (p < 631

10 M. Alimi et al. TABLE 10: Literature review of clinical outcome for ELIF surgeries* Authors & Year No. of Cases FU Duration (mos) ODI Improvement VAS Improvement Kotwal et al., Kepler et al., NA Acosta et al., Elowitz et al., Kepler et al., Pimenta et al., Anand et al., Dakwar et al., Tormenti et al., NA 5.3 current study * NA = not applicable. Minimum ). In contrast to their results, which were very similar to Johnson et al. s findings, 25 ELIF in our study also led to a significant albeit small improvement in lumbar sagittal lordosis. This finding agreed with results in the Kepler et al. and Watkins et al. studies, both of which showed significant improvement in lumbar lordosis after surgery. 28,57 The observed discrepancy regarding sagittal improvement may in part be attributed to the smaller number of cases in the Acosta et al. and Johnson et al. studies. Prior studies on scoliotic patients with coronal deformity as their main indication for surgery revealed ELIF as an efficacious procedure for achieving coronal alignment. 3,4,54 In Anand et al. s study, the coronal Cobb angle was corrected from the preoperative value of 22 to 7 postoperatively. 4 Tormenti et al. showed preoperative and postoperative Cobb angles of 38.5 and 10, respectively. 54 The improvements in these reports are higher than that in the current study, probably because we did not limit our patients to only those with scoliosis. Our findings were more comparable with those of Kotwal et al., who showed 11.2 of coronal alignment correction at the latest follow-up. 32 Although Kotwal and colleagues had a slightly larger study group than ours, their radiological results were somewhat limited given that only radiographs were used for evaluation and that the patients did not have follow-up CT scans. Nonetheless, the subanalysis of scoliotic patients in our study showed that, although the PI-LL mismatch was slightly corrected immediately after surgery, the amount of correction was not statistically significant. The difference was also insignificant at the latest follow-up. In sum, it seems that although ELIF can partially improve lumbar lordosis both immediately after surgery and at the latest follow-up, no overall PI- LL mismatch improvement is achieved by this procedure specifically in a subset of scoliotic patients. In summary, ELIF appears to provide appropriate anterior column support and allows for correction of coronal and, to a lesser degree, sagittal alignment (Table 6). Data in the current study indicated that ELIF is associated with a significant increase in foraminal height and disc height, both postoperatively and at the latest followup, showing minimal subsidence. TABLE 11: Literature review of radiographic outcome for ELIF surgeries Improvements Authors & Year No. of Cases FU Duration in Mos (no. of patients) Global Lumbar Lordosis Global Coronal Cobb Angle Foraminal Height Disc Height Watkins et al., º NA NA 2.2 mm Kotwal et al., NA 11.2º NA 3 5 mm* Kepler et al., (9); 18 (20) 3.7º NA 35% 58% ant; 70% pst Oliveira et al., NA NA 13.5% 41.9% Acosta et al., º 6.6º NA NA Anand et al., NA 15º NA NA Tormenti et al., NA 28.5º NA NA current study º 13.5º 26% (4 mm) 3.3 mm (83%) current study º 13.1º 15% (2 mm) 2.4 mm (66%) * At different levels. ant = anterior. Foraminal area. Immediate postoperative. 632

11 Extreme lateral interbody fusion outcomes Complications of ELIF Several recent studies have focused on early and late complications associated with ELIF. 5,6,20,21,27,41 Iliopsoas muscle weakness and femoral nerve injury are wellknown complications of ELIF, and L4 5 carries a higher risk of nerve injury than upper thoracolumbar levels. The incidence of motor nerve deficit in our study was 2.2%, similar to results from Cahill et al. (1.7%). 10 Ahmadian et al. 2 performed a systematic review on the complications of ELIF. They found the rate of motor weakness ranged from 0.7% to 33.6%. Most studies report higher rates of motor nerve deficit than that in the current study. 3,4,12, 13,24,27,31,38,44,49,53,54,56 Nevertheless, the deficits tend to be transient, and the rate of permanent nerve injury is low. Other early complications of ELIF include sensory deficits, particularly thigh numbness, and studies have reported a wide range of rates (0% 75%) for these complications. 2 In the current series, the rate of new thigh numbness after surgery was 4.4%. The sensory symptoms resolved in all patients within 3 months. Our diligent attention to the relation between lumbar plexus and the iliopsoas muscles resulted in a lower rate of neurological complications. We attribute the relatively low complication rates in our study to the fact that both surgeons had already had significant experience with the procedure by the time data for this trial were collected, and that only the development of new thigh numbness and motor weakness were recorded as complications. If patients had similar signs and symptoms prior to surgery, those were not recorded as complications of the procedure. Complications in the current study were comparable to those in a large study performed by Pumberger et al. in 2012 in 235 ELIF patients. 45 The prevalence of sensory deficits, psoas muscle mechanical deficits, and lumbar plexus related deficits were 1.6%, 1.6%, and 2.9%, respectively. The most significant limitation of the current study lies in the fact that it was performed retrospectively. Computed tomography scans were not available for the entire study group at the latest follow-up. Conclusions To our knowledge, this study provides what is to date the most comprehensive set of radiological and clinical outcomes of ELIF in a fairly large study group at a midterm follow-up. We provided comparisons of radiological results from an entire database at 3 time points, assessing the durability of radiological improvement. Extreme lateral interbody fusion allows partial restoration of intervertebral disc and foraminal height and improves coronal and sagittal alignment of the spine. Our radiographic follow-up data demonstrate that ELIF constructs are durable and allow good functional outcome at a midterm follow-up. Acknowledgment We acknowledge Mr. Paul Cristos for helping with the statistical analysis in the current study. Disclosure The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. This study was supported by a Baxter clinical research grant (to M.A.). Mr. James was supported by an AOSpine fellowship grant. Dr. Elowitz has received support from NuVasive for non study-related clinical or research effort. Dr. Härtl is a consultant for DePuy-Synthes, BrainLab, Lanx, and SpineWave. Author contributions to the study and manuscript preparation include the following. Conception and design: Härtl, Alimi, Hofstetter, James, Elowitz. Acquisition of data: Alimi, Hofstetter, Cong, Tsiouris, Paulo. Analysis and interpretation of data: Alimi, Hofstetter, Tsiouris. Drafting the article: Alimi. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Statistical analysis: Alimi, Hofstetter. Administrative/technical/material support: Alimi. Study supervision: Härtl, Hofstetter, Tsiouris, James, Elowitz. References 1. 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. Clinical article. J Neurosurg Spine 15:92 96, Ahmadian A, Deukmedjian AR, Abel N, Dakwar E, Uribe JS: Analysis of lumbar plexopathies and nerve injury after lateral retroperitoneal transpsoas approach: diagnostic standardization. A review. J Neurosurg Spine 18: , Anand N, Baron EM, Thaiyananthan G, Khalsa K, Goldstein TB: Minimally invasive multilevel percutaneous correction and fusion for adult lumbar degenerative scoliosis: a technique and feasibility study. J Spinal Disord Tech 21: , Anand N, Rosemann R, Khalsa B, Baron EM: Mid-term to long-term clinical and functional outcomes of minimally invasive correction and fusion for adults with scoliosis. Neurosurg Focus 28(3):E6, Banagan K, Gelb D, Poelstra K, Ludwig S: Anatomic mapping of lumbar nerve roots during a direct lateral transpsoas approach to the spine: a cadaveric study. Spine (Phila Pa 1976) 36:E , Benglis DM, Vanni S, Levi AD: An anatomical study of the lumbosacral plexus as related to the minimally invasive transpsoas approach to the lumbar spine. Laboratory investigation. J Neurosurg Spine 10: , Beurskens AJ, de Vet HC, Köke AJ, van der Heijden GJ, Knipschild PG: Measuring the functional status of patients with low back pain. Assessment of the quality of four disease-specific questionnaires. Spine (Phila Pa 1976) 20: , Billinghurst J, Akbarnia BA: Extreme lateral interbody fusion XLIF. Curr Orthop Pract 20: , Bono CM, Khandha A, Vadapalli S, Holekamp S, Goel VK, Garfin SR: Residual sagittal motion after lumbar fusion: a finite element analysis with implications on radiographic flexion-extension criteria. Spine (Phila Pa 1976) 32: , Cahill KS, Martinez JL, Wang MY, Vanni S, Levi AD: Motor nerve injuries following the minimally invasive lateral transpsoas approach. Clinical article. J Neurosurg Spine 17: , Carreon LY, Djurasovic M, Glassman SD, Sailer P: Diagnostic accuracy and reliability of fine-cut CT scans with reconstructions to determine the status of an instrumented posterolateral fusion with surgical exploration as reference standard. Spine (Phila Pa 1976) 32: , Cummock MD, Vanni S, Levi AD, Yu Y, Wang MY: An analysis of postoperative thigh symptoms after minimally invasive transpsoas lumbar interbody fusion. Clinical article. J Neurosurg Spine 15:11 18,

12 M. Alimi et al. 13. Dakwar E, Cardona RF, Smith DA, Uribe JS: Early outcomes and safety of the minimally invasive, lateral retroperitoneal transpsoas approach for adult degenerative scoliosis. Neurosurg Focus 28(3):E8, Eck JC: Minimally invasive corpectomy and posterior stabilization for lumbar burst fracture. Spine J 11: , Elowitz EH, Yanni DS, Chwajol M, Starke RM, Perin NI: Evaluation of indirect decompression of the lumbar spinal canal following minimally invasive lateral transpsoas interbody fusion: radiographic and outcome analysis. Minim Invasive Neurosurg 54: , Fairbank JC, Couper J, Davies JB, O Brien JP: The Oswestry low back pain disability questionnaire. Physiotherapy 66: , Farrar JT, Portenoy RK, Berlin JA, Kinman JL, Strom BL: Defining the clinically important difference in pain outcome measures. Pain 88: , Farrar JT, Young JP Jr, LaMoreaux L, Werth JL, Poole RM: Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain 94: , Fritz JM, Irrgang JJ: A comparison of a modified Oswestry Low Back Pain Disability Questionnaire and the Quebec Back Pain Disability Scale. Phys Ther 81: , Galloway GM: Direct lateral transpsoas approach to interbody fusion may be risky after all. J Clin Neurophysiol 28: , Guérin P, Obeid I, Gille O, Bourghli A, Luc S, Pointillart V, et al: Safe working zones using the minimally invasive lateral retroperitoneal transpsoas approach: a morphometric study. Surg Radiol Anat 33: , Hägg O, Fritzell P, Nordwall A: The clinical importance of changes in outcome scores after treatment for chronic low back pain. Eur Spine J 12:12 20, Inoue S, Watanabe T, Hirose A, Tanaka T, Matsui N, Saegusa O, et al: Anterior discectomy and interbody fusion for lumbar disc herniation. A review of 350 cases. Clin Orthop Relat Res (183):22 31, Isaacs RE, Hyde J, Goodrich JA, Rodgers WB, Phillips FM: A prospective, nonrandomized, multicenter evaluation of extreme lateral interbody fusion for the treatment of adult degenerative scoliosis: perioperative outcomes and complications. Spine (Phila Pa 1976) 35 (26 Suppl):S322 S330, 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 20: , Karikari IO, Nimjee SM, Hardin CA, Hughes BD, Hodges TR, Mehta AI, et al: Extreme lateral interbody fusion approach for isolated thoracic and thoracolumbar spine diseases: initial clinical experience and early outcomes. J Spinal Disord Tech 24: , Kepler CK, Bogner EA, Herzog RJ, Huang RC: Anatomy of the psoas muscle and lumbar plexus with respect to the surgical approach for lateral transpsoas interbody fusion. Eur Spine J 20: , Kepler CK, Huang RC, Sharma AK, Meredith DS, Metitiri O, Sama AA, et al: Factors influencing segmental lumbar lordosis after lateral transpsoas interbody fusion. Orthop Surg 4:71 75, Kepler CK, Sharma AK, Huang RC: Lateral transpsoas interbody fusion (LTIF) with plate fixation and unilateral pedicle screws: a preliminary report. J Spinal Disord Tech 24: , Kepler CK, Sharma AK, Huang RC, Meredith DS, Girardi FP, Cammisa FP Jr, et al: Indirect foraminal decompression after lateral transpsoas interbody fusion. Clinical article. J Neurosurg Spine 16: , Knight RQ, Schwaegler P, Hanscom D, Roh J: Direct lateral lumbar interbody fusion for degenerative conditions: early complication profile. J Spinal Disord Tech 22:34 37, Kotwal S, Kawaguchi S, Lebl D, Hughes A, Huang R, Sama A, et al: Minimally invasive lateral lumbar interbody fusion: clinical and radiographic outcome at a minimum 2-year follow-up. J Spinal Disord Tech [epub ahead of print], Lehmann TR, Spratt KF, Tozzi JE, Weinstein JN, Reinarz SJ, el-khoury GY, et al: Long-term follow-up of lower lumbar fusion patients. Spine (Phila Pa 1976) 12:97 104, Lubansu A: [Minimally invasive spine arthrodesis in degenerative spinal disorders.] Neurochirurgie 56:14 22, 2010 (Fr) 35. Macnab I: Negative disc exploration. An analysis of the causes of nerve-root involvement in sixty-eight patients. J Bone Joint Surg Am 53: , Madhok R, Kanter AS: Extreme-lateral, minimally invasive, transpsoas approach for the treatment of far-lateral lumbar disc herniation. Report of 2 cases. J Neurosurg Spine 12: , McAfee PC: Interbody fusion cages in reconstructive operations on the spine. J Bone Joint Surg Am 81: , Moller DJ, Slimack NP, Acosta FL Jr, Koski TR, Fessler RG, Liu JC: Minimally invasive lateral lumbar interbody fusion and transpsoas approach-related morbidity. Neurosurg Focus 31(4):E4, Oliveira L, Marchi L, Coutinho E, Pimenta L: A radiographic assessment of the ability of the extreme lateral interbody fusion procedure to indirectly decompress the neural elements. Spine (Phila Pa 1976) 35 (26 Suppl):S331 S337, Ozgur BM, Aryan HE, Pimenta L, Taylor WR: Extreme Lateral Interbody Fusion (XLIF): a novel surgical technique for anterior lumbar interbody fusion. Spine J 6: , Park DK, Lee MJ, Lin EL, Singh K, An HS, Phillips FM: The relationship of intrapsoas nerves during a transpsoas approach to the lumbar spine: anatomic study. J Spinal Disord Tech 23: , Pimenta L, Díaz RC, Guerrero LG: Charité lumbar artificial disc retrieval: use of a lateral minimally invasive technique. Technical note. J Neurosurg Spine 5: , Pimenta L, Marchi L, Oliveira L, Coutinho E, Amaral R: A prospective, randomized, controlled trial comparing radiographic and clinical outcomes between stand-alone lateral interbody lumbar fusion with either silicate calcium phosphate or rh-bmp2. J Neurol Surg A Cent Eur Neurosurg 74: , Pimenta L, Oliveira L, Schaffa T, Coutinho E, Marchi L: Lumbar total disc replacement from an extreme lateral approach: clinical experience with a minimum of 2 years follow-up. Clinical article. J Neurosurg Spine 14:38 45, Pumberger M, Hughes AP, Huang RR, Sama AA, Cammisa FP, Girardi FP: Neurologic deficit following lateral lumbar interbody fusion. Eur Spine J 21: , Ray CD: Threaded fusion cages for lumbar interbody fusions. An economic comparison with 360 degrees fusions. Spine (Phila Pa 1976) 22: , Rodgers WB, Gerber EJ, Rodgers JA: Lumbar fusion in octogenarians: the promise of minimally invasive surgery. Spine (Phila Pa 1976) 35 (26 Suppl):S355 S360, Schwab FJ, Blondel B, Bess S, Hostin R, Shaffrey CI, Smith JS, et al: Radiographical spinopelvic parameters and disability in the setting of adult spinal deformity: a prospective multicenter analysis. Spine (Phila Pa 1976) 38:E803 E812, 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 24: , Shen FH, Samartzis D, Khanna AJ, Anderson DG: Minimally invasive techniques for lumbar interbody fusions. Orthop Clin North Am 38: , Smith JS, Klineberg E, Schwab F, Shaffrey CI, Moal B, Ames 634