Extreme lateral lumbar interbody fusion: Do the cons outweigh the pros?

BACKGROUND: Major factors prompted the development of minimally invasive (MIS) extreme lateral interbody fusion (XLIF; NuVasive Inc., San Diego, CA, USE) for the thoracic/lumbar spine. These include providing interbody stabilization and indirect neural decompression while avoiding major visceral/vessel injury as seen with anterior lumbar interbody fusion (ALIF), and to avert trauma to paraspinal muscles/facet joints found with transforaminal lumbar interbody fusion (TLIF), posterior lumbar interbody fusion (PLIF), and posterior-lateral fusion techniques (PLF). Although anticipated pros of MIS XLIF included reduced blood loss, operative time, and length of stay (LOS), they also included, higher fusion, and lower infection rates. Unanticipated cons, however, included increased morbidity/mortality rates.
METHODS: We assessed the pros and cons (e.g., risks, complications, comparable value/superiority/inferiority, morbidity/mortality) of MIS XLIF vs. ALIF, TLIF, PLIF, and PLF.
RESULTS: Pros of XLIF included various biomechanical and technical surgical advantages, along with multiple cons vs. ALIF, TLIF, PLIF, and PLF. For example, XLIF correlated with a considerably higher frequency of major neurological deficits vs. other constructs; plexus injuries 13.28%, sensory deficits 0-75% (permanent in 62.5%), motor deficits 0.7-33.6%, and anterior thigh pain 12.5-25%. XLIF also disproportionately contributed to other major morbidity/mortality; sympathectomy, major vascular injuries (some life-ending others life-threatening), bowel perforations, and seromas. Furthermore, multiple studies documented no superiority, and the potential inferiority of XLIF vs. ALIF, TLIF, PLIF, and PLF.
CONCLUSION: Reviewing the pros of XLIF (e.g. radiographic, technical, biomechanical) vs. the cons (inferiority, increased morbidity/mortality) vs. ALIF, TLIF, PLIF, and PLF, we question whether XLIF should remain part of the lumbar spinal surgical armamentarium.

INTRODUCTION

Minimally invasive surgery (MIS) consisting of extreme lateral interbody fusion procedures (XLIF) were devised to afford maximal disc excision and end plate availability for interbody fusion, while providing indirect decompression of the neural elements. Aims of MIS XLIF included avoiding the major visceral/vessel injuries seen with anterior lumbar interbody fusion (ALIF), and trauma to the posterior elements (e.g. paraspinal muscles/facet joints) seen with transforaminal lumbar interbody fusion (TLIF), posterior lumbar interbody fusion (PLIF), and posterolateral fusion (PLF). Although further pros included the reduction of operative time, blood loss, length of stay (LOS), and duration of surgery, with potentially higher fusion and lower infection rates, there were also unanticipated cons of MIS XLIF included a disproportionate increase in the neurological/complications of spinal surgery vs. other constructs; i.e. plexus injuries 13.28%, sensory deficits 0–75% (permanent in 62.5%), motor deficits 0.7–33.6%, and anterior thigh pain 12.5–25%.[9101112] Other general complications of XLIF included; major vascular injuries (e.g., some life-threatening, others life-ending), bowel perforations, sterile seromas, and instrumentation failures.[9101112] Here, we reviewed the pros of XLIF (e.g., radiographic, technical, biomechanical, and potential comparability/superiority) and cons (potential inferiority with increased morbidity/mortality) vs. other procedures (e.g. ALIF, TLIF, PLIF, and PLF) to determine whether XLIF should remain part of the spinal surgical armamentarium.

PROS AND CONS OF XLIF: X-RAY/COMPUTED TOMOGRAPHY (CT) AND BIOMECHANICS

Pros of X-ray/CT documentation of indirect decompression with extreme lateral interbody fusion

Postoperative X-rays and computed tomography (CT) studies documented that MIS XLIF with or without posterior instrumentation provided increased maximal disc removal/end plate availability for interbody fusion while affording indirect decompression of the spinal canal (degenerative stenosis or scoliosis) [Table 1].[81820] When 30 MIS XLIF were performed with posterior instrumentation in adults with degenerative lumbar scoliosis in the series by Caputo et al., X-rays showed an increase in neuroforaminal height (80.3%), neuroforaminal width (7.4%), disc height (116.7%), segmental lordosis at L4-L5 (14.1%), and global lordosis (11.5%) [Table 1].[8] For MIS XLIF interbody fusions performed at 43 levels (stand alone for degenerative lumbar stenosis) performed in 21 patients averaging 67.6 years of age in a study by Oliveira et al., radiographs documented an increase of 41.9% disc height, 13.5% in foraminal height, 24.7% in foraminal area, and 33.1% in central canal diameter.[20] Utilizing 2-day postoperative CT scans, Malham et al. further documented increased postoperative disc height (89%), foraminal height (38%)/area (45.1%) for 52 patients (average age 66.4) undergoing 79-level MIS XLIF.[18] Of interest was the disparity in the percentage of decompression provided by MIS XLIF for degenerative stenosis/scoliosis provided by different X-ray and CT evaluations; disc height on X-rays was 116.7% vs. 89% on CT, foraminal height was 80.3% on X-ray vs. 38% on CT.

Table 1

Radiographic (X-ray, MR, CT), cadaveric, and biomechanical considerations for MIS lumbar XLIF

Biomechanical pros, cons, and comparability of minimally invasive surgeries (MIS) extreme lateral interbody fusion (XLIF)/lateral lumbar interbody fusion (LLIF) vs. other techniques

Several studies explored the biomechanical pros, cons, and comparability of MIS LLIF/XLIF vs. other procedures (e.g. ALIF, TLIF, PLIF, and PLF) (e.g., greater end plate/disc removal, restoration of sagittal balance and/or lordosis, but early cage settling) [Table 1].[12222627] Tatsumi et al. found that for four fusions (ALIF, PLIF, TLIF, and XLIF) performed in 8 cadavers (24 disc spaces and 48 end plates from L2-L5), MIS XLIF provided the most extensive end plate preparation (58.3%) and disc removal (90%), whereas less disc was removed for the other constructs (e.g. 65% for TLIF, 43% for PLIF, and 40% ALIF groups).[26] Comparing MIS lateral lumbar interbody fusions (LLIF) to ALIF, TLIF, and PLF, Sembrano et al. noted on standing pre and 6-week postoperative X-rays (147 patients; 212 levels fused), that all constructs demonstrated comparable improvement in sagittal balance, but that ALIF provided better segmental/general correction.[22] For Tohmeh et al., 140 patients undergoing MIS XLIF/pedicle screw fixation at 223 levels (followed for 15.5 postoperative months), despite increases in overall lordosis (4.0 to 8.1) and segmental lordosis (10.7 to 13.7), cage settling (e.g. >1 mm or more) occurred in 20% of patients immediately postoperatively and in 62% of the patients within 1 postoperative year.[27] Although this led to the recommendation to use wider and longer cages, this maneuver would potentially increase the risk of “interbody spacer overhang” and contralateral foraminal nerve root compromise or ligamentous rupture, particularly if the device were placed too anteriorly. Note, Epstein's previous review cited a 45% risk of cage-overhang if MIS XLIF were applied in the anterior 1/3 of the vertebral body.[12]

Pros of bilateral vs. unilateral pedicle screw fixation with minimally invasive surgery (MIS) extreme lateral interbody fusion (XLIF)

Several authors found that supplementing MIS XLIF with unilateral or bilateral pedicle screw fixation both increased lordosis, but bilateral instrumentation provided greater stabilization [Tables 1 and 2].[11321] When Alimi et al. performed MIS XLIF plus unilateral instrumented pedicle/screw fusions (91%) to treat unilateral radiculopathy in 23 patients (91%) (e.g., 61% with degenerative scoliosis and 43% with prior surgery), they effectively successfully resolved radicular complaints and maintained increased unilateral foraminal height for up to 11 ± 3.7 postoperative months.[1] However, when Fogel et al. compared the efficacy of MIS XLIF stand alone cages vs. MIS XLIF with varying combinations of lateral plates, unilateral/bilateral pedicle screws, and spinous process plates in 7 cadavers at the L4-L5 level (with/without degenerative spondylolisthesis (DS)), they concluded bilateral pedicle screws provided the greatest stability, while spinous process plates afforded the least.[13] Similarly, Phillips et al., in 2013, found that, for 107 patients (average age 68) undergoing average 3 (1–6 levels) level MIS XLIF with/without pedicle screw/rod fusions for degenerative scoliosis (2-year period), the best radiographic results were achieved utilizing bilateral pedicle screws (e.g. best correctin of the Cobb angle (average 15.2 degrees at 2 years)).[21] Certainly, the majority of surgeons would utilize bilateral pedicle screw fixation if they were utilizing instrumentation to supplement MIS XLIF.

Table 2

Comparison of safety/efficacy of MIS XLIF/LLIF/DLIF Vs. open/MIS PLIF, TLIF, ALIF, and PLF

Summary of computed tomography (CT)/X-rays and biomechanics minimally invasive surgery (MIS) extreme lateral interbody fusion (XLIF) pros/cons

Pros for MIS XLIF vs. ALIF, TLIF, PLIF, and PLF included more disc removal/end plate availability for interbody fusion, and greater indirect neural decompression by increasing disc height/foraminal height/area/canal diameter.[81820] Cons, however, included a high risk for neurological injury and general complications, along with graft/cage settling (e.g., 20% immediately, 62% at one year).[12222627]

CONS OF MINIMALLY INVASIVE SURGERY EXTREME LATERAL INTERBODY FUSION

High complication rate for minimally invasive surgery surgery (MIS) extreme lateral interbody fusion (XLIF) vs. Other minimally invasive surgery constructs (ALIF, TLIF, PLIF, PLF)

Neurological complications of extreme lateral interbody fusion vs. other procedures

Neurological complications frequently followed thoracic and lumbar MIS XLIF vs. other constructs that some preferred to label as “anticipated” risks rather than “complications” [Tables 1 and 2].[479101120] Historically, open spinal procedures (discectomy/laminectomy/with or without fusion) incur a 0–2% incidence of root injuries; their frequency was equal to MIS TLIF (2%), but less than MIS PLIF (7.8%), or MIS ALIF (15.8%), and substantially lower than XLIF (23.8%: sustained root/plexus deficits).[910] A focused review of neurological complications for MIS XLIF procedures included; plexus injuries (13.28%), sensory deficits (0–75%: permanent in 62%), 5 motor deficits (0.7–33.6%), and anterior thigh pain (12.5–25%).[11] Oliveira et al. found that, in their series of 21 patients undergoing 43-level MIS XLIF alone (degenerative lumbar stenosis), that 3 (14.3%) patients developed new iliopsoas weakness/deficits.[20] Berjano et al. took it even a step further, recommending prophylactic preoperative steroids to address their too frequent postoperative plexus injuries that continued to occur despite technical improvements for MIS XLIF approaches.[4] When Caputo et al. evaluated the efficacy of 30 MIS XLIF (127 levels; T10-L5 (average 4.2 levels) with ALIF (L5S1; 11 patients)) and pedicle screw/rod fixation, postoperative anterior thigh/pain/numbness was so common that they recommended it no longer be considered a “complication” of MIS XLIF, but rather an “anticipated” risk (e.g. postoperative factors).[7]

Cadaver and magnetic resonance/dynamically-evoked electromyography offer technical improvements for minimally invasive surgery extreme lateral interbody fusion procedures, but neurological deficits persist

Two studies, one performed in cadavers and the other performed utilizing magnetic resonsnce images (MR), sought to limit the common MIS XLIF postoperative lumbar plexus deficits [Table 1].[469101120] Utilizing 12 cadavers and 24 lumbar plexuses/psoas muscle exposures, Spivak et al. found the “safe” area to avoid MIS XLIF-related lumbar nerve root/plexus injuries between the L2-L4 levels (most susceptible); it was best to place the retractor in the anterior half of the disc.[24] However, if an interbody MIS XLIF spacer is placed within the anterior one-third of the disc, the risk of contralateral root compromise is reportedly high, and this may, therefore, not be a viable solution.[911] When Buric et al. carefully studied preoperative MR examinations (e.g., lumbar plexus shape/position) and additionally utilized intraoperative dynamically-evoked electromyography to perform 29 MIS XLIF (average age: 59 years; 1.6 level MIS XLIF at 47 levels; 83% used pedicle screws) between the L2-3 and L4-L5 levels, on postoperative day 1, 10 (34%) patients still had anterior thigh/groin pain.[6]

Fluoroscopy and computed tomography (CT) studies offer technical improvements for minimally invasive surgery (MIS) extreme lateral interbody fusion (XLIF) procedures, but neurological deficits still persist

Utilizing intraoperative X-ray/fluoroscopy in combination with postoperative CT examinations helped guide the performance of thoracic and lumbar MIS XLIF procedures [Table 1].[230] Arnold et al. utilized intraoperative fluoroscopic guidance to place sequential tubes/dilators for perform MIS XLIF; they recommended utilizing a true lateral position, with incision of the mid or somewhat anterior portion of the disc, but still observed a persistent high rate of “ neural injuries, psoas weakness, and thigh numbness.”[2] In a series by Youssef et al. involving 84 MIS XLIF fusions, 68 (81%) were fused on both postoperative CT and dynamic X-rays; notably, the fusion rate was comparable to fusion rates for MIS ALIF, MIS TLIF, and MIS PLIF.[30] Nevertheless, despite MIS XLIF correlating with shorter operative times, reduced blood loss and shorter length of stay, they resulted in an increased incidence of lumbar plexus deficits. Clearly, the MIS XLIF approach inherently places major neurological structures at risk, and the multiple studies developed to limit these risks have not succeeded.

Neurological complications of minimally invasive surgery surgery (MIS) extreme lateral interbody fusion (XLIF) with bone morphogenetic protein: Reported vs. “obfuscated” results

Additional unique complications occurred when bone morphogenetic protein (rhBMP-2) was utilized to supplement MIS XLIF/LLIF constructs [Tables 1 and 2].[51729] In a series by Lykissas et al., over a period of 6 years, MIS LLIF were performed with (rhBMP-2; 72 patients) vs. without rhBMP-2 (72 patients, autograft/allograft).[17] BMP clearly contributed to both short and long-term direct damage to the lumbosacral plexus; long-term sensory deficits were noted in 29 patients who received rh-BMP-2 vs. 20 without; persistent motor deficits were observed in 35 patients with vs. 17 without rh-BMP-2; and anterior thigh/groin pain was observed in 8 patients with vs. 0 without BMP. On the contrary, when Wang et al. evaluated the treatment of adjacent segment stenosis in 21 patients (average age 61 following prior anterior or posterior fusions) undergoing 1-2 level MIS XLIF (17, single level, 4, two level; interbody spacers) with BMP without pedicle/screw fixation, CT studies showed a 100% fusion rate and no complications.[29] The total absence of neurological complications and 100% fusion rate were signals in this manuscript that, at best, the documentation was inadequate, and at worst, the data were “obfuscated.” In direct contrast to the perfect MIS XLIF (without instrumentation) fusion rate, I would offer the study by Berjano et al., in which the authors assessed the fusion rates utilizing CT studies (more accurate than X-rays) 1 year following MIS XLIF utilizing different bone graft supplements to fill cages [Table 2].[5] Fusion was documented in a much lower number of patients (e.g. just 68 of 78 patients (87.1%)).

Non-neurological complications of minimally invasive surgery surgery (MIS) extreme lateral interbody fusion (XLIF) vs. other procedures

Multiple additional medical/surgical complications, excluding neurological deficits, were attributed to MIS XLIF [Tables 1 and 2].[7812192030] Epstein observed the following major complications of MIS XLIF (e.g. likely vastly underreported due to our medicolegal system); sympathectomy, major vascular injuries (some life ending, others life-threatening), bowel perforations, seromas, malpositioning of MIS XLIF cages with extrusion or contralateral foraminal nerve root compression (e.g. cage-overhang).[12] In a study by Meredith et al., 18 patients had thoracic MIS XLIF procedures at 32 levels; 22% (4 of 18) of the patients exhibited major surgical complications, and there were 5 medical complications.[19] The complication rate in a 2012 study by Caputo et al. involving 30 MIS XLIF/pedicle screw fixation (127 levels (average 4.2 levels) from T10-L5 with MIS ALIF (L5S1; 11 patients)) was 26.6%. This included a 11.8% pseudoarthrosis rate, with 6 (20%) other major (in part overlapping) complications; 2 (6.7%) of whom required further surgery; 1 lateral incisional hernia, 2 ruptures of the anterior longitudinal ligament (ALL), 2 wound breakdowns, 1 pedicle fracture, 1 nonunion, 1 cardiac instability.[78] For the 43 MIS XLIF performed in 21 patients in Oliveira et al. series, 2 (9.5%) patients required secondary surgery for stenosis.[20] In a study by Youssef et al., complication rates for MIS XLIF peri and postoperatively were 2.4% and 6.1%, respectively.[30] Isaacs et al. examined radiographic outcomes of 107 MIS XLIF (average 4.4 levels/patient; average age of 68 years) for adult scoliosis (18.7% no instrumentation, 75.7% pedicle screws, 5.6% lateral fixation). Complications (1 or more) occurred in 9% of the patients undergoing MIS XLIF without instrumentation, whereas 20.7% had complications with open posterior instrumented procedures (e.g. including 3 deep wound infections). Certainly, the increased surgical/medical risks of so many multilevel MIS XLIF should prompt spine surgeons to ask why so many older patients are being subjected to such extensive multilevel MIS XLIF (e.g. 4.2 and 4.4 levels/patient in two studies) leading to such high complication rates with/without additional instrumentation.

Lack of safety, efficacy, and superiority (some say inferiority) of minimally invasive surgery (MIS) extreme lateral interbody fusion (XLIF) over other constructs

Multiple studies demonstrated a lack of safety or efficacy of MIS XLIF over other available fusion constructs (e.g. MIS, ALIF, TLIF, PLIF, and PLF) (e.g. particularly regarding perioperative neurological/other morbidity) [Table 2].[141623] When Lee et al. analyzed complications (including 3 transient lumbosacral plexus palsies) for 74 mini-open lateral approaches for 1–4 level MIS ALIF, they advised that, prior to trialing MIS XLIF or MIS DLIF, the enhanced risk of the latter approaches which placed the lumbosacral plexus at risk, should be further investigated.[16] In 2016, Hartl et al. compared the safety and efficacy of adding intraoperative neural monitoring (IONM) to perform MIS lumbar XLIF (24 case series; 18 used IONM) vs. MIS ALIF (8 randomized controlled trials and 1 case study), and one combined MIS XLIF/ALIF study [Table 2].[14] MIS XLIF had a two-fold greater neurologic complication rate (8.92%) vs. MIS ALIF (4.96 %). Sembrano et al., in 2016, compared outcomes for treating low-grade degenerative spondylolisthesis (DS) with stenosis (SS) over a two-year period utilizing MIS XLIF (29 patients) vs. MIS TLIF (26 patients) [Table 2].[23] Results for the two procedures were similar; average opeative time for MIS XLIF vs. MIS TLIF (171 vs. 186 minute), and identical 2-day length of stay (LOS). However, there was significant less blood loss for MIS XLIF vs. MIS TLIF. Critically, however, new iliopsoas weakness occurred in 31% of MIS XLIF vs. 0% of MIS TLIF procedures. The data in this latter study further highlight the significant neurological risks posed by MIS XLIF.

Lack of superiority and potential inferiority of minimally invasive surgery surgery (MIS) extreme lateral interbody fusion (XLIF) vs. other constructs

Multiple studies emphasized either the lack of superiority of MIS XLIF over other constructs or in some cases, MIS XLIF's lesser performance [Tables 1 and 2].[32528] In a review by Barbagallo et al. (e.g. only 6 quality articles of 258) regarding the relative safety, efficacy, and outcomes of 1 or more level MIS LLIF with/without instrumentation vs. MIS PLIF/TLIF for degenerative lumbar disease, they concluded there was “insufficient evidence of the comparative effectiveness of MIS LLIF versus MIS PLIF/TLIF surgery.”[3] When Talia et al. compared the strengths and weaknesses of different MIS surgical techniques, comparing XLIF with TLIF, and ALIF, they concluded there were no adequate long-term data confirming the benefit/efficacy/safety of any these approaches over another.[25] Furthermore, in a review of 23 articles (19 study cohorts, 720 patients) utilizing different MIS interbody fusion techniques (MIS ALIF, MIS XLIF, MIS P/TLIF), Uribe et al. discovered “significant gains in both weighted average lumbar lordosis and segmental lordosis. following MIS interbody fusion,” but did not single out XLIF.[28] Again, XLIF did not uniquely offer benefits over spinal constructs.

CONCLUSION

MIS XLIF were originally devised to provide increased end plate availability for interbody spinal fusion to better facilitate arthrodesis rates while providing indirect neural decompression. Anticipated major advantages included avoiding major vessel/visceral injuries seen with MIS ALIF, trauma to the posterior elements, and reduced operative time, blood loss, LOS vs. MIS TLIF/PLIF and PLF. Nevertheless, these multiple studies failed to document the safety, efficacy, or superiority of the MIS XLIF vs. the multiple other surgical alternatives. In fact, they documented the increased neurological and surgical/medical complication rates for XLIF that were in some instances life-threatening, or even, life-ending.[9101112] Shouldn’t we, therefore, conclude that the cons of MIS XLIF outweigh its pros, and move to strike it from our surgical armamentarium?

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.