➢ Degenerative lumbar scoliosis is a common condition, with a higher prevalence in older populations, that can lead to substantial pain and disability.
➢ After a course of nonoperative treatment, surgery is indicated to address neurological deficits and/or disability resulting from deformity with or without stenosis.
➢ Surgical intervention may include decompression with or without fusion and can be performed from an anterior, lateral, posterior, or combined approach.
➢ Despite a relatively high prevalence of perioperative complications, the overall long-term outcomes of surgical interventions for degenerative lumbar scoliosis are favorable in well-selected, medically optimized patients.
Degenerative lumbar scoliosis is a progressive coronal plane deformity that includes sagittal malalignment and usually occurs in patients older than forty years of age1. This is in contrast to adult idiopathic scoliosis, which is a result of untreated adolescent idiopathic scoliosis and is more commonly seen in the thoracic spine. Degenerative lumbar scoliosis involves fewer segments and less severe curves than adult idiopathic scoliosis; however, both diagnoses can present with similar symptoms of back pain, lower-limb pain, and functional impairment. The prevalence is difficult to determine, possibly because of the diverse patient populations studied; it has ranged from 1% to 68% in the literature1-8, but most studies have demonstrated prevalences of about 7.5% to 15%6. There are radiographic findings that may predict which patients will develop scoliosis (defined as a coronal Cobb angle of >10°), as one community cohort demonstrated that asymmetric disc degeneration and unilateral osteophyte formation led to an increased prevalence of scoliosis5,9.
The pathogenesis of degenerative lumbar scoliosis is unknown. However, recently researchers have investigated markers that could help establish an etiology or risk for progression of this condition10. Biomarkers that signify type-II-collagen synthesis and degradation have been studied in patients with osteoarthritis, and similar findings may be seen in patients with degenerative lumbar scoliosis. Compared with controls, patients with degenerative lumbar scoliosis were found to have higher serum levels of keratan sulfate, procollagen type-II C-propeptide (CPII), and collagen-type-II cleavage, with positive correlations between the Cobb angle and the CPII level11, suggesting an increased turnover of the components of the nucleus pulposus and facet joint cartilage in patients with degenerative lumbar scoliosis. Certain N-methyl-D-aspartate (NMDA) receptor genes have also been associated with degenerative lumbar scoliosis12. Although no definitive conclusions can be drawn from these findings at this time, this promising research may establish the etiology or risk for progression of degenerative lumbar scoliosis as well as possible areas for therapeutics.
There are conflicting data in the literature regarding the association between osteoporosis and scoliosis. Patients with scoliosis have lower hip bone mineral density than nonscoliotic patients13; however, bone mineral density may not necessarily be an independent predictor of curve magnitude8. Age and ethnicity may also predict the prevalence of degenerative lumbar scoliosis. In a retrospective review of 2973 dual x-ray absorptiometry (DXA) scans, the prevalence of scoliosis was highest in white subjects (11.11% versus 6.49% for black subjects and 4.04% for other races grouped together)4. In the same study, the prevalence increased with age with no difference based on sex4. In a Chinese Han population, the prevalence of scoliosis was even higher than that reported for white subjects (13.3% for those older than forty years), with female sex, age, and osteopenia correlating with the prevalence of degenerative lumbar scoliosis14. A lower body mass index has also been found to be an independent risk factor for scoliosis8.
The progression of degenerative lumbar scoliosis may be accompanied by structural changes in the local anatomy. The psoas and paravertebral muscles may hypertrophy on the convex side of the curve and degenerate on the concave side of the curve in patients with degenerative lumbar scoliosis15-17. Pedicle morphology is also likely to be altered in patients with degenerative lumbar scoliosis, with narrower diameters on the concave side18, and it is important to be aware of this during surgical planning. There are anatomic changes in the foramina of patients with degenerative lumbar scoliosis as well. The foraminal height, interval between the pedicle and superior articular process, and foraminal area are smaller in patients with degenerative lumbar scoliosis compared with controls. These measurements are also smaller on the concave side of the curves19. Certain levels of lumbar nerve roots may be more affected depending on the pattern of curvature. L3 and L4 are more likely to be compressed in the foramen on the concave side of the curve, and the lower roots (L5 and S1) are more likely to be affected on the convex side by lateral recess stenosis18,20.
A radiographic classification system for adult scoliosis based on lumbar lordosis and obliquity of the end plate of L3 in the frontal plane was developed by Schwab et al.21. It was modified to include pelvic and sagittal parameters as well as the thoracic curves22 (Table I). This classification has good to excellent intraobserver and interobserver reliability22,23 and has been validated with parameters based on health-related quality-of-life outcome scores. Curves are classified according to the location of the major coronal curve (>30°)—i.e., thoracic only (apical level at T9 or cephalad), lumbar or thoracolumbar (apical level at T10 or caudal), double curve, or type N (no curve of >30°). Sagittal modifiers are based on three parameters: pelvic incidence minus lumbar lordosis, sagittal vertical axis, and pelvic tilt (Table I). These spinopelvic parameters have been found to be important predictors of patient outcomes24.
Patients with degenerative lumbar scoliosis typically present with low back pain and difficulty maintaining an upright posture, with variable degrees of lower-limb pain. The pain typically becomes worse after standing for a long period of time and with fatigue and is relieved by lying down. Symptoms due to neurogenic claudication are variable depending on the concavity or convexity of the curve and the curve level. A thorough neurological examination as well as observation of posture and gait are important in the initial evaluation of patients with degenerative lumbar scoliosis. It is important to view patients from the side to obtain an overall sense of sagittal alignment as well as any compensatory mechanisms.
Standing posteroanterior and lateral 36-in (0.9-m) radiographs are necessary to evaluate alignment on initial and subsequent visits to follow curve progression. Progression is variable but averages about 3° per year25. Flexion and extension lateral radiographs are also important for surgical decision-making, as any instability manifested as spondylolisthesis should be treated with fusion to prevent further progression. The degree of osteophyte formation may help predict instability in the lumbar spine26. Magnetic resonance imaging (MRI) is useful for evaluating nerve root compression, disc degeneration, and modic changes (changes in signal intensity of the vertebral end plates and subchondral bone)27,28. Computed tomography (CT) myelography is particularly useful for patients with more severe deformity, to define the neural elements and areas of compression as well as anatomic changes associated with lateral listhesis29.
A trial of nonoperative management is considered appropriate for most patients presenting with symptomatic degenerative scoliosis30,31. Patients usually present with symptoms related to either radiculopathy or claudication due to stenosis or to pain and functional impairment assumed to be a result of degeneration and deformity, either sagittal or coronal or both. Despite the usual practice of trying nonoperative treatment for degenerative scoliosis, we are not aware of any studies addressing its efficacy32,33.
Medication can be used for relief of pain related to deformity and/or degeneration. Traditional recommendations include nonsteroidal anti-inflammatory drugs, muscle relaxers, and narcotics. A trial of physical therapy or its equivalent is appropriate for most patients32,34. Goals generally include optimization of core fitness, flexibility, and body mechanics as well as sensible modification of activity. Chiropractic evaluation and treatment can be considered for acute episodic pain, but evidence of its efficacy is mainly limited to case reports and small series35-38. More comprehensive treatment incorporating cognitive behavioral therapy, when available, may be optimal for chronic pain, as is the case for other types of degenerative conditions39,40. Interventional treatment, including epidural steroid injections, facet injections, or sacroiliac injections, are controversial because of questions about their long-term benefits, but they can be helpful for diagnostic as well as therapeutic purposes. Transforaminal epidural steroid injections may provide short-term relief for most patients, but their effectiveness decreases over time41.
For neurogenic symptoms due to stenosis, nonoperative treatment similar to that used for spinal stenosis alone is reasonable42,43. Rehabilitation options and medications are similar to those described above, and membrane stabilizers, such as gabapentin, can be included as well. Epidural steroid injections in particular are more efficacious for neurogenic symptoms arising from nerve root compression41.
Operative treatment can be considered for patients with stenosis and neurogenic symptoms that are not relieved by a course of nonoperative treatment. Most neurogenic symptoms related to stenosis are thought to be stable. While there is no firm time point at which to consider nonoperative treatment a failure, a minimum of six weeks is usually considered adequate. Accelerated surgical decision-making is indicated for progressive neurological deficits, especially cauda equina syndrome.
Most operative treatment involves fusion to address the deformity along with decompression of the stenosis44. In isolated circumstances, limited decompression alone may be appropriate. For patients with risk factors for progression, such as a Cobb angle of >30°, lateral listhesis of ≥6 mm, or substantial apical vertebral body rotation, decompression alone may lead to progression25. There are no rigid guidelines for determining levels of fusion. At a minimum, the levels of decompression should be included. The remaining levels depend on the severity of the curve, the resulting sagittal and/or coronal imbalance, and any concurrent spondylolisthesis or lateral listhesis.
Other indications for operative treatment are more ambiguous and may include intractable pain, patient dissatisfaction with the cosmetic appearance, a neurological deficit, and instability45-53. Patients with deformity-related pain, which is back pain due to postural abnormalities and muscular fatigue, should undergo a more extensive trial of nonoperative treatment; an operation before six months of dedicated treatment is rarely appropriate for such patients. When operative treatment is considered in the absence of radiculopathy, fusion is always indicated. Medical comorbidities as well as bone mineral density should be optimized and controlled, and modifiable risk factors such as smoking and morbid obesity should be corrected prior to operative treatment54,55. Bone mineral density can be evaluated with a DXA scan, and osteoporosis treatment regimens may be indicated. Symptoms should be correlated to the deformity and radiographic evidence of neurological compression, although this is more difficult than it is for neurogenic conditions. An operation to address progression of scoliosis is reasonable. There is no strict degree of progression or curvature that warrants operative treatment, but a lumbar curve of greater than 30° to 40° or progression of the curve of more than 10° have been cited30,31. In addition to progression over time, features such as spondylolisthesis, lateral listhesis, and positive sagittal balance beyond 5 cm merit consideration for surgery. The overall deformity is often addressed more completely in these patients compared with those with primarily radicular symptoms due to stenosis. Again there are no strict guidelines, and selection of the upper and lower levels of the operative reconstruction remains debatable56-59. Each patient must be evaluated individually, and levels must be selected on an individual basis. Patients should be well informed about the risks of operative treatment, which are far from infrequent or trivial. However, outcomes can be expected to be favorable in patients with optimized risk factors and substantial functional impairment despite nonoperative treatment and whose symptoms can be attributable to their deformity30,60.
Operative management can be considered for patients with degenerative scoliosis after conservative therapy has failed45-51. Many options are available for patients with degenerative scoliosis, and therefore treatment planning and intervention must be individualized. The goals of surgery include decompression of neural tissue and curve correction for restoration of spinal alignment and stability45-47,52,61-69.
Operative techniques for thoracolumbar scoliosis are varied and can be performed through an anterior or posterior approach or both45-49,51,61,62,66,70-78. Posterior approaches are often sufficient for younger adult patients with flexible curves or in patients with balanced curves. Anterior-only surgery runs the risk of leading to kyphosis or loss of sagittal balance. A combined anterior-posterior approach is often useful in patients with rigid deformity. This combination offers the benefits of anterior procedures such as anterior release of the spine with deformity correction and fusion, including placement of interbody devices. This combination is particularly common at the lumbosacral junction, where the risk of nonunion is the highest and the benefit in terms of sagittal alignment is the greatest. In recent years, minimally invasive techniques as well as the direct lateral or transpsoas approach have increased the surgeon’s armamentarium for the management of degenerative scoliosis45-49,51,53,61,62,66-74,76-79.
Decompression without fusion is an option for neurogenic claudication or radiculopathy in patients who have minimal deformity without instability or those who cannot tolerate more extensive surgery45-47. This is ideally done unilaterally and usually involves a laminectomy and/or foraminotomy without instrumentation. Complication rates have been reported to be lower in patients who undergo decompression alone; however, this strategy should be reserved for patients without instability, with minimal scoliosis, and with involvement of a limited number of spinal levels45-48,53,80.
Anterior surgery, either alone or in combination with a posterior procedure, is a well-described approach for treatment of deformity45,46,48,49,51,53,61,62,66,70,71,76,79. The most common anterior surgery is anterior lumbar interbody fusion, combined with anterior and/or posterior instrumentation. Anterior lumbar interbody fusion is most useful in the lower thoracic and lumbar regions. In selected cases, the anterior approach provides deformity correction and satisfactory patient outcomes. The anterior approach allows direct visualization of the disc space, along with the ability to place large disc spacers, which enhances fusion, maintains distraction, and aids in the restoration of lordosis. Potential disadvantages of this procedure include the risk of vascular injury, abdominal hernia, ureteral damage, retrograde ejaculation, and graft dislodgement.
The transpsoas interbody fusion, a variant of the anterior lumbar interbody fusion, offers many of the same advantages as an anterior approach, such as the placement of a large interspace graft and the ability to achieve lordotic correction (Fig. 1, A and B). Transpsoas interbody fusion can be performed with a minimally invasive technique, and pedicle screws can be placed percutaneously. However, access to the L4-L5 and L5-S1 disc spaces is limited. The main risks associated with transpsoas interbody fusion are nerve root injury and postoperative thigh numbness. In a multicenter study of 107 patients who underwent transpsoas interbody fusion, thirty-six (33.6%) had some weakness after the operation; twenty-nine of the thirty-six had isolated hip weakness81. This weakness was transient in twenty-five (86.2%) of the twenty-nine patients. Several recent studies have shown good outcomes with high fusion rates, even over several levels, in patients treated with transpsoas interbody fusion and pedicle fixation for degenerative scoliosis45-47,51,53,61,62,66-74,76-79. Modest improvements in lumbar lordosis can be expected after transpsoas interbody fusion82,83. In one series of 107 patients, lumbar lordosis was corrected from a mean of 27.7° preoperatively to a mean of 33.6° at the time of a two-year follow-up84. Another study demonstrated 5.3° of improvement in lumbar lordosis immediately postoperatively, but this amount of improvement decreased to 2.9° at one year, with no significant improvements in pelvic incidence or in pelvic incidence-lumbar lordosis mismatch85.
The mainstay of spinal deformity surgery is the posterior approach. Posterior surgery involves the use of instrumentation, usually pedicle screws with or without hooks, in combination with transforaminal or posterior lumbar interbody fusion51,62,67,70-73,75. Posterior surgery also offers the option of performing an osteotomy with posterior release as well as vertebrectomy with cage placement72-74. A major benefit of the posterior approach is the ability to attain curve correction over a large spinal area; this can be enhanced with judicious rod bending, placement of interbody spacers for correction in the coronal and sagittal planes, and distribution of force over the entire construct. In addition, pedicle screws and rods allow control of all three columns of the spine in a single construct and may allow fewer levels of fusion51,53,62,67,70-75.
Several studies evaluating posterior surgery for degenerative scoliosis have shown improved patient outcomes, high fusion rates, good deformity correction, and acceptable complication rates48,62,67,71,73-75,77,78. The major risks associated with posterior surgery for curve correction include screw malplacement, pseudarthrosis, spinal cord injury, and infection. Patients undergoing an operation for deformity correction should have spinal cord monitoring during the procedure48,50,53,63,71-75.
Proximal junctional kyphosis is another known complication after fusion for degenerative scoliosis, and it may be related to the level of the upper instrumented vertebrae86-88. In a study of ninety consecutive patients followed for two years after instrumented fusion to the sacrum, a change in the lumbar lordosis of >30° and a preexisting thoracic kyphosis of >30° were found to be independent risk factors for developing proximal junctional kyphosis89.
Adjuncts to standard interbody fusions include osteotomies67-69,74,75. Smith-Petersen osteotomy is technically the easiest osteotomy to perform and may allow up to 10° of lordotic correction per level90. The amount of correction depends on the mobility of the disc space and is often less in these degenerative spines. Pedicle subtraction osteotomy is a more challenging surgical procedure but allows a greater degree of correction, up to 35°91. Use of vertebral column resection is generally limited to the thoracic spine. It is the most technically demanding osteotomy and is normally used in cases that cannot be corrected with pedicle subtraction osteotomy or Smith-Petersen osteotomy, or when the primary deformity is in the thoracic spine49,51,53,63,67-69,72-75.
Combined anterior-posterior surgery for large curve correction provides deformity reduction and acceptable fusion rates and patient outcomes49,51,52,62,63,69,72,77,78 (Fig. 2, A and B). However, these complex procedures are often associated with higher complication rates, as the patient is exposed to longer operative times and potentially greater blood loss50,53,61,63-65,67,70,71,76. Rates of postoperative complications after thoracolumbar surgery for degenerative scoliosis are generally reported to be higher than rates after other elective spine operations, primarily because of the extent of the operation as well as a greater number of comorbidities in patients with thoracolumbar deformity. There is a wide range of reported complication rates (7% to 68%)50,61,63,64,68-70,92, likely because of variation in patient populations and criteria for defining complications. One large Scoliosis Research Society database of 10,329 patients who underwent operative treatment for degenerative lumbar scoliosis demonstrated complications in 719 patients (7%), but other smaller studies showed higher rates92. Complications may also be related to the approach and surgical technique. Direct neurological complications include cord manipulation, traction injury, and compression injury due to positioning. Indirect causes primarily result from ischemia. Some early postoperative complications include pulmonary embolism, urinary tract infection, delirium, epidural hematoma, wound infection, respiratory distress syndrome, neurological deficit, and ileus. Late complications can include pseudarthrosis, hardware failure, adjacent segment disease, infection, and systemic complications61,63-65,69,76,93,94. Older patients may have a slightly higher risk of complications, but the rates appear to be acceptable. In a study of forty-seven patients treated with decompression and instrumented pedicle screw fixation (mean age, 66.1 ± 5.0 years), there were fourteen early complications and eighteen late complications93, leading to revision surgery in seven patients. An age of more than sixty-five years and two or more comorbidities were associated with higher early complication rates, but the differences were not significant (p = 0.053 and 0.052, respectively). Increased blood loss was a significant predictor of early complications (p = 0.02, odds ratio = 1.01). In another study of patients older than sixty-five years (n = 98), the most common major complication was wound infection (10%), and the most common minor complication was urinary tract infection (34%). These rates increased with increasing age as well as with the extent of the procedure (increased estimated blood loss, longer operating room time, and more levels fused)94.
In recent years, advances in operative techniques, including minimally invasive approaches, intraoperative navigation, implant design, and an enhanced ability to achieve fusion, have contributed to the overall safety and improved outcomes in adults undergoing surgery for deformity correction48,49,51,61,62,67,70-75,78. Another consideration for operative treatment is the number of levels to fuse, including the proximal and distal levels of the construct49,53,77,78,95. Overall recent trends in large published series have shown improvements in standard outcome metrics and the ability to achieve and maintain correction, even in the face of an increasingly elderly population with several comorbidities48,50,61,66,67,70,75,76.
We are not aware of any prospective studies that compared operative and nonoperative treatment of degenerative lumbar scoliosis. Retrospective studies comparing operative and nonoperative treatment of adult scoliosis have shown that, compared with the patients who were treated nonoperatively, the patients who underwent surgery had less back pain, lower-limb pain, and disability and greater improvement in general health status after two years of follow-up80,96.
Numerous studies have assessed the clinical outcomes of various forms of surgical management of degenerative lumbar scoliosis. However, it is difficult to compare the results of these studies as the operative techniques have varied widely. The primary categories of operative treatment for degenerative lumbar scoliosis include decompression alone (i.e., to treat associated stenosis and/or radiculopathy), selective fusion, and fusion of the entire degenerative curve. Transfeldt et al.97 compared these three treatment options in patients with degenerative lumbar scoliosis and radiculopathy with a minimum of two years of follow-up. The authors found that the degenerative lumbar curve (average, 25°) did not progress after decompression alone or selective fusion. Although full curve fusion was associated with an improvement in the degree of curvature, it was also associated with a higher complication rate (reported to be 56%, with n = 20) and less improvement from the baseline Oswestry Disability Index (ODI) compared with decompression alone and selective fusion97.
Clinical and radiographic results of conventional, open posterior fusion for lumbar degenerative scoliosis are mainly limited to observational studies with relatively small patient numbers. Simmons and Simmons98 studied the results at an average of forty-four months following decompression and posterior instrumented fusion for degenerative lumbar scoliosis and stenosis in forty patients with an average age of 61.5 years. The authors reported a substantial decrease in back pain, with 93% of the patients reporting mild to no back pain after the surgery. The mean curve correction was 19°. In a prospective observational study of thirty-one patients who underwent posterior lumbar decompression and fusion for degenerative lumbar scoliosis, Keorochana et al.99 reported significant improvement in the scores on visual analog scales for pain, the ODI, and the Roland Morris score after an average duration of follow-up of thirty-two months.
There have been a number of recent studies on the safety and benefit of minimally invasive anterior lumbar interbody fusion through a lateral approach for the treatment of degenerative lumbar scoliosis. In a multicenter study of 107 patients who underwent interbody fusion through a lateral approach for degenerative lumbar scoliosis, Isaacs et al.81 concluded that this approach minimizes the morbidity of the surgical management of this condition. Khajavi and Shen100 prospectively assessed the clinical and radiographic outcomes of the minimally invasive lateral approach to anterior interbody fusion for the treatment of degenerative lumbar scoliosis in twenty-four patients. There was no comparison group. These authors reported a significant improvement, compared with baseline, in visual analog scale scores for back and lower-limb pain, the ODI, and Short Form-36 (SF-36) scores at an average of two years postoperatively. The mean improvement in the Cobb angle was from 27.7° preoperatively to 16.6° at the time of follow-up, and there was a 39% increase in lumbar lordosis. A meta-analysis by Dangelmajer et al.101 compared minimally invasive lateral interbody fusion (eight studies) with decompression without fusion (four studies) for the treatment of degenerative lumbar scoliosis. Patients undergoing decompression without fusion had a mean preoperative Cobb angle of 16.7°, which improved to 18° postoperatively. Patients undergoing lateral interbody fusion had a mean preoperative Cobb angle of 22.3°, which improved to 9.2° postoperatively. The change between the preoperative and postoperative Cobb angles differed significantly between the two approaches (p = 0.014); however, no clinical outcomes were measured in this study. Decompression surgery alone may be adequate for patients with mild curves and symptoms due to stenosis102, but fusion should be performed if the patient has risk factors for progression.
Sagittal parameters, such as the sagittal vertical axis, as well as spinopelvic parameters have proved to be predictive of patient disability. Schwab et al.24 found that severe disability, as defined by an ODI of >40, was likely in patients with pelvic tilt of ≥22°, a sagittal vertical axis of ≥47 mm, and a pelvic incidence-lumbar lordosis mismatch of ≥11°.
Our current understanding of the treatment of degenerative lumbar scoliosis and the incremental benefit that surgery may provide over nonoperative treatment lags behind that of other degenerative conditions of the lumbar spine. The study of this condition is complicated by the wide range of presenting symptoms, degrees of deformity, and available surgical techniques for treatment.
Degenerative lumbar scoliosis is a common condition that can lead to substantial functional disability. With an aging population, it is important to further investigate the pathogenesis, nonoperative and operative treatment, and outcomes in these patients. Advances in operative techniques continue to improve outcomes and limit morbidity when nonoperative treatment is unsuccessful. Prospective, comparative studies are needed to better define the most effective treatment options.
Source of Funding: This study was designed and prepared through the Association for Collaborative Spine Research (ACSR).
Investigation performed at the Association for Collaborative Spine Research (ACSR)
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. Also, one or more of the authors has had another relationship, or has engaged in another activity, 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