➢ High-grade spondylolisthesis may cause substantial disability, sagittal imbalance, pain, and neurologic dysfunction.
➢ Imaging of spondylolisthesis should include full-length posteroanterior and lateral radiographs of the spine with attention focused on pelvic incidence, sacral slope, pelvic tilt, and slip angle.
➢ Indications for operative treatment in high-grade spondylolisthesis include the presence of substantial deformity, pain, stenosis, neurologic deficit, and radiographic progression.
➢ The optimal operative technique involves decompression, reduction, and circumferential fusion with instrumentation as this allows for restoration of sagittal balance and an improved fusion rate and has an acceptable safety profile.
Spondylolisthesis is defined as a slippage of one vertebra relative to the adjacent lower vertebra1. Although the prevalence of spondylolysis approaches 6% in the general population2,3, the prevalence of high-grade spondylolisthesis remains difficult to define. As our understanding of the pathomechanics of spondylolisthesis and slip progression has evolved, so too has our approach to the management of these patients. This review will consider the clinical presentation, evaluation, pathomechanics, and evidence-based treatment options for these patients. The focus will be on the pediatric population presenting with dysplastic or acquired (i.e., isthmic) high-grade spondylolisthesis and spondyloptosis, as this has been a major area of investigation and controversy.
Clinical Presentation and Physical Examination
Adolescents and young adults with high-grade spondylolisthesis have highly variable clinical presentations ranging from back pain to radiculopathy to cauda equina syndrome. Neurologic symptoms may include radiating pain in the buttock or posterior thigh, bowel or bladder dysfunction, or weakness. This is especially true for patients with an intact pars interarticularis because of the increased risk of severe stenosis. In addition to neurologic dysfunction, patients may report difficulty walking and leaning forward and may express concern with their appearance. The presence of nighttime pain is rare and warrants further work-up for the possibility of infection or neoplasm.
The physical examination should include a detailed neurologic assessment, including an observational gait analysis. Radicular signs may be present, including motor weakness, numbness, and decreased reflexes, particularly along the L5/S1 distribution. Patients with cauda equina compression may have diminished sensation along the sacral nerve roots. With further progression, the pelvis may become unbalanced, allowing the sacrum to become clinically prominent as it develops a vertical orientation. As a result, there may be a midline step-off, heart-shaped buttocks, or an excessive lumbar lordosis proximally. Tight hamstrings are often present and may be due either to contracture as the pelvis rotates and the sacrum adopts a vertical position or to chronic nerve root irritation. This may contribute to the classic gait pattern with flexed knees and hips (Phalen-Dickson sign)4. As imbalance progresses further, there may be forward trunk leaning as global sagittal balance decompensates anteriorly. Scoliosis may be present as a result of spasm, asymmetric slippage, or idiopathic causes and was reported in 42% of patients in a study of eighty-four individuals with symptomatic spondylolisthesis5.
The initial imaging of the patient should consist of radiographs and should be focused on differentiating among spondylolysis, spondylolisthesis, and other causes of low back pain such as neoplasm. If initial posteroanterior and lateral radiographs are equivocal or negative6 but there is a high index of suspicion for a defect of the pars interarticularis, single photon emission computed tomography (SPECT), magnetic resonance imaging (MRI), or reverse-gantry computed tomography (CT) may be used to evaluate for spondylolysis7-13. However, high-grade spondylolisthesis is readily identified on initial radiographs, and, once confirmed, a systematic analysis should be followed14.
If not previously obtained at the initial evaluation, standing full-length posteroanterior and lateral radiographs of the spine and pelvis should be obtained to allow for the assessment of coronal plane deformity, global spinal balance, pelvic orientation, and pelvic morphology. Interpretation of the imaging is then done in a systematic fashion. First, attention is drawn to the local anatomy and the degree of slippage is graded according to the Meyerding classification15. Next, the local morphology is assessed, specifically evaluating for pars defects and/or elongation, dysplastic or dysraphic posterior elements, and dysplastic sacral morphology. The lumbosacral slip angle (lumbosacral kyphosis) is then measured to quantify the local deformity (Fig. 1). In normal patients with high pelvic incidence, the lumbosacral angle is approximately 10° lordotic16.
Finally, sacral slope, pelvic tilt, and pelvic incidence (Fig. 2) are measured to evaluate local and global balance because spinopelvic parameters are important in the treatment algorithm. Pelvic orientation is first assessed by measuring sacral slope and pelvic tilt. Both sacral slope and pelvic tilt are position-dependent and are directly interrelated such that as the sacral slope increases, pelvic tilt decreases, and vice versa. However, pelvic incidence is a fixed anatomic parameter that describes the pelvic morphology and is independent of the patient positioning. The final parameter to assess is the global sagittal balance, assessed by measuring the C7 plumb line. Global balance is confirmed when a vertical line drawn from the middle of the C7 body passes between the femoral heads and the sacrum14,17.
Additional imaging studies, such as MRI, can be obtained on a case-by-case basis if there is atypical osseous anatomy or there are signs and symptoms of stenosis. In patients with substantial symptoms or nighttime pain, MRI should be obtained to evaluate for the possibility of neoplasm.
Particularly in young patients, consideration should be given to reducing the cumulative dose of radiation, which may, in turn, decrease the risk of carcinogenesis18-21. For this reason, our preference is to use EOS biplanar slot-scanning digital radiography (EOS Imaging, Paris, France) for preoperative and postoperative imaging, as it yields excellent image quality with substantial reduction from the radiation of traditional radiographs22. CT scans, in particular, should not be obtained unless there is a clear indication. Berrington de González et al. reported that up to 1500 future malignancies may develop as a result of spine CT scans performed in the United States in 200719. This risk is increased in younger patients19,20. In general, we restrict the use of CT scans in high-grade slips to patients scheduled to undergo a surgical procedure. Prior to the surgical procedure, all patients undergo MRI and CT, as these allow for a detailed understanding of the stenosis, the morphology of the pars interarticularis and L5 pedicles, and the amount of sacral doming, all of which are essential for appropriate preoperative planning.
Historically, the two most frequently used classification systems were the Wiltse23 and Meyerding15 classifications. The Wiltse classification has five major types and describes morphology and etiology. Types I (dysplastic or developmental) and II (isthmic or acquired, subdivided as IIA [stress fracture], IIB [pars elongation], and IIC [acute pars fracture]) are the most common types in children and adolescents. Type III (degenerative) presents in adulthood and is a frequent cause of neurogenic claudication and back pain in older adults24. Types IV (posttraumatic) and V (pathologic) are uncommon. Unlike the Wiltse classification, the Meyerding classification is a straightforward system that considers the amount of slippage15. In this system, there are five grades based on the number of quadrants that one vertebra is slipped relative to another. Grades 1 and 2 are low-grade; Grade 1 is a slip of 1% to 25% and Grade 2 is a slip of 26% to 50%. Grades 3 and 4 are high-grade; Grade 3 is a slip of 51% to 75% and Grade 4 is a slip of 76% to 100%. Grade 5, termed spondyloptosis, is a complete dislocation of L5 in front of the sacrum (slippage of >100%).
More recently, Mac-Thiong et al. developed an improved classification system to reflect the Meyerding slip grade, the pelvic incidence, and the overall sagittal spinopelvic balance (Table I)25,26. The resulting Spinal Deformity Study Group (SDSG) classification17,25,27 has six types and stratifies low-grade slips (types 1 to 3) based on pelvic incidence and high-grade slips (types 4 to 6) based on spinopelvic balance. This system has demonstrated good intraobserver and interobserver reliability25 and is useful because it is organized in such a way that higher-risk slips are assigned a higher type. Therefore, this classification system may guide important management decisions, such as whether to perform a fusion in situ or with reduction16.
Pathomechanics, Progression, and Spinopelvic Balance
The forces acting at the lumbosacral joint are complex and contribute to the development and progression of spondylolisthesis (Fig. 3). The primary forces acting across this joint include axial loading, forward flexion, and truncal rotation. This combination creates a large amount of shear at the L5/S1 junction. In the normal spine, the shear is resisted by a posterior osseous and ligamentous tension band and an intact disc. In the patient with high-grade spondylolisthesis, the posterior tension band fails. With the loss of posterior restraint, the shearing forces predominate and there is anterior and inferior translation as the anterior column can no longer provide support. As the slip progresses, the primary posterior element deficiency gives way to secondary changes including lumbosacral kyphosis, L5 wedging, sacral doming, and lumbosacral disc degeneration. Ultimately, pelvic retroversion and global spinal imbalance develop as a result of these dysplastic changes.
As the pathomechanics of spondylolisthesis are linked with spinopelvic balance, there has been increased focus on the importance of assessing spinopelvic parameters (Fig. 1 and Fig. 2)28. This is due both to the recognition that imbalance adversely affects quality of life29-31 and to the understanding that these parameters may correlate with grade, progression, and symptoms and may also guide treatment.
Because it describes pelvic morphology, pelvic incidence has been studied as a predictor of slip progression. Although high pelvic incidence is a biomechanically important predictor of slip progression due to increased compression and shear at the L5/S1 junction32, this remains to be proven in clinical studies, which suggest instead that slip percentage, slip angle, and the presence of a high-grade slip are the important factors33-35. Nevertheless, high pelvic incidence (>60°) is especially important to recognize in patients with low-grade slips as shear-type morphology may increase the risk of progression (SDSG type 3) and require close monitoring17,32,36. This is especially true if additional risk factors are present, such as growth, female sex, dysplastic or dysraphic changes, or a small sacrum.
On the contrary, because patients with high-grade slips almost always have high pelvic incidence, the focus shifts to assessing pelvic and spinal balance16,37-39, as this is critical to developing a treatment algorithm17,27. The measurement of the sacral slope and pelvic tilt becomes important in the assessment of spinopelvic balance (Fig. 2). In a study of 133 patients with high-grade spondylolisthesis, Hresko et al. defined a nomogram for thresholds beyond which sacral slope and pelvic tilt values are abnormal16. Similar to what is observed in controls with high pelvic incidence, high sacral slope and low pelvic tilt were seen in balanced high-grade spondylolisthesis (mean pelvic incidence of 81°, sacral slope of 60°, pelvic tilt of 21°, and lumbosacral angle of 9° of kyphosis)16. Conversely, those patients with an unbalanced (i.e., retroverted) pelvis (SDSG type 5) will have low sacral slope and high pelvic tilt with more lumbosacral kyphosis (mean pelvic incidence of 77°, sacral slope of 40°, pelvic tilt of 37°, and lumbosacral angle of 29° of kyphosis)16. After assessment of pelvic balance, the global spinal balance must be evaluated by measurement of the C7 plumb line. The spine is in positive sagittal imbalance if this line passes in front of the femoral heads (SDSG type 6). In particular, patients with an unbalanced pelvis and/or unbalanced spine are those in whom reduction is indicated to negate the shear forces at the lumbosacral junction16,40,41.
Consideration must also be given to the slip angle, as local kyphosis is emerging as a determinant of quality of life and may be another criterion on which to consider the need to perform reduction33,35. The L5/S1 angle should be slightly lordotic and although absolute lumbosacral angle thresholds do not currently exist, Lundine et al. recently reported that a slip angle of >20° may be an indicator of worse prognosis in both operatively and nonoperatively treated patients33. Therefore, slip angle improvement should be considered the primary goal of reduction.
Although the prognostic role of the spinopelvic parameters remains uncertain, the current evidence suggests that patients are more likely to progress as the pelvic incidence increases, the pelvis retroverts, and the spine displaces anteriorly into imbalance. Treatment of these patients should aim to correct these underlying abnormalities.
Most patients with high-grade spondylolisthesis have symptoms, so it is relatively uncommon for them to undergo nonoperative management for an extended period of time. Reported long-term outcomes of those who are treated nonoperatively are variable, with substantial pain relief in 10% to 90% of patients42,43. In a recent study, Bourassa-Moreau et al. reported that scores on the Scoliosis Research Society-22 questionnaire44, a validated spine-specific outcome measure assessing pain, function, appearance, mental health, and satisfaction, remained constant over a two-year period in five patients with high-grade slips treated nonoperatively, leading them to conclude that nonoperative treatment is reasonable in select patients with good baseline quality-of-life scores, a normal neurologic examination, and imaging that remains unchanged over time45. The long-term outcomes of those managed without surgical procedures remains unknown.
Indications for Operative Treatment
Because of the high risk of progression, particularly in growing children, high-grade spondylolisthesis usually requires operative intervention. One absolute indication for operative treatment is cauda equina syndrome. However, motor weakness, back pain, postural decompensation, and lower-extremity radicular pain should be considered strong relative indications. Slip progression alone is a relative indication in the adolescent. In contrast, adults have a low risk of progression, so surgical indications are typically related to pain, stenosis, and sagittal imbalance.
General Principles and Controversies
The primary goals of operative management for spondylolisthesis are multiple: to achieve a solid fusion, to correct deformity to achieve global balance, and to perform adequate nerve root decompression. The methods used to achieve these goals will be different in pediatric and adult patients, depending on the comorbidities, stiffness of deformity, and presence of degenerative changes. Although controversial, in most cases, successful management of high-grade spondylolisthesis requires direct decompression, reduction, and instrumented posterior fusion with structural interbody support.
Is There Still a Role for Uninstrumented Fusion?
Prior to the widespread adoption of pedicle screws, uninstrumented in situ fusion was the mainstay of treatment. However, failure rates of up to 12% (six of forty-nine patients) for low-grade slips46 and 14% (three of twenty-one patients) for high-grade slips47 have been reported. Additionally, uninstrumented arthrodesis necessitates fusion in situ and frequently requires postoperative immobilization. Consequently, uninstrumented techniques have largely been abandoned.
Should One Attempt Reduction?
There has long been debate about whether or not to reduce high-grade spondylolisthesis due to the procedure’s difficulty, complications, and questions regarding clinical benefit48-60. Some authors have reported good results with in situ fusion, particularly in patients with maintained pelvic balance52,57. Others have found in situ fusion preferable to a technique of gradual reduction, distraction, and staged fusion59. At fifteen years, Poussa et al. reported improved Scoliosis Research Society scores and a lower Oswestry Disability Index in patients undergoing fusion in situ60.
Much of the original concern with reduction stemmed from the risk of neurologic injury, and although the rationale for reduction is multifactorial, there is emerging evidence that it is neurologically safe and biomechanically preferable. This may be especially true when aiming for complete reduction. Petraco et al. showed that the amount of nerve path lengthening rapidly increased as a slip of 100% was reduced the final 25%61. In contrast, the first half of the reduction produced only 29% of the total nerve root strain61. Although nerve root injury has been reported following reduction50,51,58,59, larger, more recent studies have demonstrated a prevalence of neurologic deficit in the range of 5% to 10%50,58,62. A recent query of the Scoliosis Research Society Morbidity and Mortality database demonstrated a new nerve root deficit in nine (10.2%) of eighty-eight pediatric patients58. Of the nine patients with nerve root deficits, one had a permanent deficit and four had full recovery58. Although neither direct nerve decompression nor reduction was a risk factor for a new deficit, osteotomy was a risk factor58. Longo et al. compared reduction to fusion in situ in a recent meta-analysis, showing similar neurologic deficit rates (p = 0.8) in patients treated with reduction and fusion (7.9% [thirteen of 165 patients]) and those treated with in situ fusion (8.9% [nine of 101 patients])50. Additionally, the ability to achieve direct canal decompression may actually prevent the risk of acute, postoperative cauda equina syndrome63. Although most surgeons perform some direct neural decompression58, the option to achieve reduction without decompression exists and has been used safely by some55.
The second major advantage of reduction is that it allows correction of the local anatomy. With reduction, the surgeon can directly influence the slip angle, which can be improved to <30°48,49,54,55, and the percentage, which can be improved to <40%48,54,56. The downstream effect is an improvement in the global sagittal alignment. Indeed, this has been used as the major rationale for reduction16,17. To test this hypothesis, Martiniani et al. performed a retrospective study in which patients with balanced high-grade spondylolisthesis (SDSG type 4) were treated with in situ posterolateral fusion and those with unbalanced deformities (SDSG type 5 or 6) were treated with reduction and posterior fusion with interbody cage augmentation52. The investigators demonstrated that the patients with balanced high-grade spondylolisthesis, although they received in situ fusion, healed without decompensating. Furthermore, those who underwent reduction and fusion healed with improvement in the sacral slope and pelvic tilt. This provided support to the rationale that patients with unbalanced high-grade spondylolisthesis benefit most from reduction16,17.
Finally, alteration of the local forces allows a decreased risk of pseudarthrosis. Several studies have now demonstrated an increased risk of pseudarthrosis in patients who do not undergo reduction50,51,53,54. In a recent meta-analysis, Longo et al. demonstrated a pseudarthrosis rate of 5.5% (nine of 165 patients) in those who underwent reduction compared with 17.8% (eighteen of 101 patients) in those who underwent fusion in situ (p = 0.004)50. This is likely due to the continued presence of shear forces across the lumbosacral disc space, which places excessive strain on the implants, leading to nonunion, loosening, and failure51,56.
At the present time, the best evidence indicates that reduction can be performed with an acceptable risk profile, allows for improved healing, and alters the overall spinal alignment in a beneficial way (Fig. 4, Fig. 5, Fig. 6, and Fig. 7).
What Levels (and Where) Should One Fuse?
The final preoperative decisions that must be made involve which levels to fuse, whether to incorporate the pelvis, and whether to provide anterior column support.
Monosegmental fusion has been utilized by some authors for high-grade spondylolisthesis56,64. However, even when combined with anterior column support, posterior fusion of only L5/S1 was associated with a 17% nonunion rate in a study of thirty-four patients64 and with an 11% rate of structural complications in another study of eighteen adolescents56. The authors of both studies therefore recommended fusion from L4 to S156,64. Although proximal fixation usually ends at L3 or L4, distal fixation may either terminate at S1 alone56,64-67 or incorporate the ilium68,69. Distal fixation to S1 or S2 frequently works well for children and adolescents, but older patients with high-grade slips, an open S1/S2 disc space, a connective tissue disorder, or poor sacral bone quality may benefit from iliac screw augmentation. In two-year68 and five-year69 data on patients treated for high-grade slips with iliac screw augmentation of S1 pedicle screws, the pelvic fixation was beneficial, with primary fusion rates of 96% (forty-seven of forty-nine patients) when combined with anterior column support69. Although 53% (eighteen) of thirty-four patients with spondylolisthesis required iliac screw removal, there was good function with few long-term problems of the sacroiliac joint69.
Multiple studies have demonstrated the advantages of anterior interbody support for high-grade slips68-74. The method of providing anterior column support is frequently combined in studies68,69 and can be achieved either with an interbody cage (Fig. 5 and Fig. 6) or with an S1-L5 strut (Fig. 7). The primary advantage to anterior support is the decreased pseudarthrosis rate compared with that for instrumented posterolateral fusion alone. Multiple small series (fewer than thirty patients each) have reported pseudarthrosis rates of 29%73 to 32%71 with instrumented posterolateral fusion, but with additional anterior support this decreases to between 0%73,74 and 7%71. Furthermore, circumferential fusion facilitates correction of local kyphosis73 and sagittal balance74. Two recent meta-analyses found that circumferential fusion took longer than instrumented posterolateral fusion but allowed for better restoration of alignment, fusion rate, and clinical satisfaction75,76.
In cases in which an interbody cage cannot be utilized, an alternative technique is to use a transsacral interbody fibular or titanium mesh strut (the modified Bohlman technique)64,66,67,77-80. Without requiring a forceful reduction, this is an effective method for both achieving union66,78,79 and correcting local kyphosis66,67. In a series of sixteen patients, Hart et al. reported one case of fibular allograft strut failure, but none of the eleven patients with a titanium mesh strut had an implant failure78. To our knowledge, there have been no current studies evaluating the results of fusion with an interbody cage compared with a transsacral strut.
The above techniques can be utilized for grade-3 and 4 spondylolisthesis, as well as for spondyloptosis. Because of the increased local deformity in spondyloptosis, reduction may be extremely challenging or impossible. In these cases, consideration is given to the Gaines procedure81-83. As originally described by Gaines and Nichols, an anterior approach was performed first to allow resection of the L5 body and adjacent intervertebral discs82. Two weeks later, a separate posterior approach was performed, the remaining L5 posterior elements were removed, distraction was applied, the L4 body was reduced onto the sacrum, and an instrumented fusion of L4 to S1 was accomplished81,82. This procedure may be modified to include partial L5 vertebrectomy or longer instrumentation segments posteriorly83. In general, this very challenging procedure can result in improved sagittal balance, function, and gait, but patients should be cautioned to expect a transient L5 neurologic deficit postoperatively81.
Our Preferred Technique
Our preferred technique is shown in Figure 5, Figure 6, and Figure 7. The patient is positioned prone on the Jackson table with the hips maximally extended and the knees flexed. A Gill procedure84 with exposure from L4 to S1/S2 is performed. The L5 nerve roots are identified and are frequently located either draped posteriorly to the L5 pedicle near the location of the pedicle screw starting point, or beneath the large osteophyte that may form on the residual L5 pedicle. Great care must be taken to identify the nerve root during the exposure so that full decompression and fixation may be safely performed. Bilateral iliac and pedicle screws are placed from L4 to S1. The S2 alar iliac approach may be utilized for screw placement to facilitate in-line rod connection and to avoid the need for removal of the posterior superior iliac spine85. Alternatively, a traditional iliac starting point may be used after creation of a 1 × 1-cm notch in the posterior superior iliac spine. For optimal fixation in a standard adolescent, 8 × 80-mm screws directed 1 cm above the sciatic notch aiming toward the anterior inferior iliac spine are often required. Next, temporary distraction is applied to allow for L5/S1 discectomy followed by sacral dome osteotomy. During this process, the posterior longitudinal ligament is excised, the posterior aspect of the anulus is removed, and a full discectomy is performed. The anterior aspect of the anulus is thinned but not removed and an extensive anterior release is not typically required. The sacral dome osteotomy is then performed. Distally, the starting point for the osteotome is limited by the location of the S1 foramen and screws. If the L5 body has substantial remodeling with anterior inferior beaking, the osteotome can be used to remove this prominence to facilitate reduction and provide a level surface for fusion. Once the L5/S1 disc space has been well decorticated, additional distraction and hip extension are applied to maximize the reduction, which can be further facilitated by using a disc shaver or an interbody fusion trial to lever L5 onto S1. An interbody titanium mesh cage with morselized local bone graft is placed. In cases with too much deformity or in which reduction does not allow an interbody cage, we perform a modified Bohlman procedure using a titanium mesh strut. Distraction is then released and rods are placed. Compression is applied to achieve lordosis and to restore the posterior tension band. Finally, a posterolateral fusion with a mixture of local autograft and corticocancellous allograft is performed. Neuromonitoring (including somatosensory and motor-evoked potentials) is used throughout the case, with direct electromyographic stimulation of the L5 nerve roots being utilized prior to decompression, immediately after decompression, and after reduction. The stimulation threshold should decrease after each of these maneuvers, typically from 3 mA or more before decompression to around 1 mA after full decompression and to near-normal (0.8 to 1 mA) after reduction.
Postoperatively, patients are specifically cautioned to avoid knee extension with hip extension so that the lumbar plexus is not stretched. Careful postoperative neurologic examinations for the first forty-eight hours focus on motor strength in the tibialis anterior, extensor hallucis longus, peroneals, and quadriceps. Narcotics are needed for pain relief and Valium (diazepam) is needed for spasm relief. Occasionally, gabapentin is useful to manage temporary radiculitis, which we have observed in approximately 30% of our patients. Most patients are discharged home on day 3 or 4. To date, we have received no reports of patients with permanent neurologic deficits.
Source of Funding: No funds were received for this study.
Investigation performed at the Department of Orthopaedic Surgery, Nemours/A.I. duPont Hospital for Children, Wilmington, Delaware
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. One or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. In addition, one or more of the authors has a patent or patents, planned, pending, or issued, that is broadly relevant to the work. No author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.
- Copyright © 2015 by The Journal of Bone and Joint Surgery, Incorporated