Degenerative lumbar scoliosis is a common spinal disorder with variable presentation and treatment options. Degenerative scoliosis has important differences from adult idiopathic scoliosis, including a new onset of deformity, a lumbar predominance in the curve pattern, degenerative changes within the deformity, and a steady progression of deformity with age. Degenerative scoliosis is an important spinal disorder affecting the aging spine, and an evidence-based approach to the nonoperative and operative care of the disorder has not been well established. This chapter discusses the evidence related to the relative merits of a combined anterior and posterior approach to treatment of degenerative scoliosis compared with a posterior approach to deformity correction.
A 57-year-old woman had progressive spinal deformity with imbalance in the sagittal and coronal plane. She reported that the onset of her spinal deformity was in adulthood and that she had had back pain for over 30 years. Over the past 5 years she had noted progression of leg pain, with the left leg more affected than the right. She reported that her left leg frequently gave way with walking and that she had a walking tolerance of less than two blocks. She had undergone several treatments with epidural injections and selective nerve blocks with only temporary and partial improvement of her pain. In the past year she was bothered most by her ribs’ abutting the pelvis on her right side. She also noted mild subjective weakness as well as numbness and tingling distally in her lower extremities, with pain and paresthesia symptoms worse in the left lower extremity. Other nonoperative care included physical therapy, yoga, and chiropractic care, which led to no significant improvement in her pain or disability. Her medications included extended-release oxycodone (OxyContin) and hydrocodone/acetaminophen (Vicodin) for pain.
PMH: Medication-controlled hypertension
Exam: The patient had an obvious deformity with a right rib hump. While standing she had a significant sagittal plane deformity with significant kyphosis as well as left lumbar level prominence. She had abutment of her right ilium into the anterior aspect of her right costal margin. On supine examination she was noted to have significantly less kyphosis. Her thoracolumbar spine area was nontender. Findings of motor, sensory, and reflex examinations were normal. No gait abnormalities were noted.
Imaging: Radiographs obtained included full-length images of the spine showing marked scoliosis, with lumbar scoliosis measuring 68 degrees with the left-sided apex at L2 ( Figure 17-1 ). The images demonstrated the sagittal deformity with plumb line 8 cm to the right and a positive sagittal imbalance of 17 cm. For preoperative planning traction radiographs and lateral radiographs with the patient over a bolster were also obtained, which revealed some reduction of her kyphosis ( Figure 17-2 ).
The patient with symptomatic degenerative scoliosis characteristically has symptoms of back and leg pain, and progression of deformity, including sagittal and coronal plane imbalance. Consistent findings in a review of data for 200 patients by Pritchett and Bortel included degenerated facet joints, a loss of lumbar lordosis, and end-plate sclerosis with vertebral rotation. Rotational subluxation within the deformity and sagittal plane decompensation are radiographic findings that are associated with significant compromise of health-related quality of life in affected patients.
The etiology of degenerative lumbar scoliosis is multifactorial. The relationship between bone density and degenerative scoliosis has been found to be variable, and there is not a clear causal link to bone mineral density alone. Because Pritchett and Bortel found certain consistent areas of anatomic degeneration, Tribus suggests that an asymmetric degeneration of the anterior disk space or of the posterior facet joints may lead to the rotational and translational deformities associated with degenerative scoliosis. This degenerative cascade, when combined with ligamentum flavum buckling, may explain the more prominent neurogenic symptoms in addition to back or deformity complaints from patients with the disorder.
In adult spinal deformity, there are important correlations between curve and deformity characteristics and patient clinical appearance. Global sagittal plane alignment is the most important radiographic characteristic associated with health status compromise, including pain and functional limitations. Schwab and colleagues proposed a clinical impact classification for adult scoliosis. They demonstrated that radiographic factors associated with disability and surgery include lumbar hypolordosis and intervertebral subluxation.
Radiographic characteristics alone have a poor correlation with the need for operative or nonoperative approaches to deformity. The decision to pursue operative versus nonoperative care is importantly based upon patient self-assessment of health status, and patient and physician combined decision making. Patient self-assessment of disability is an important determinant of the decision to pursue operative versus nonoperative care for spinal deformity. Glassman and associates demonstrated that patients who were treated with surgery for adult scoliosis had more back and leg pain than patients treated nonoperatively. Patients treated nonoperatively had significantly more comorbid conditions than patients treated operatively. Pekmezci and colleagues demonstrated that functional domain scores, including walking in the Oswestry Disability Index (ODI) and vitality in the Scoliosis Research Society (SRS) instrument, were more important than back or leg pain in distinguishing between operative and nonoperative care in adults with scoliosis.
Bess and co-workers studied factors that differentiated between operative and nonoperative approaches to treatment of deformity in a cohort of 290 adults. They demonstrated that among patients older than age 40, pain and disability were significantly higher in patients who chose operative care rather than nonoperative care. In contrast, among younger patients, curve size and coronal deformity were higher in patients who were treated with surgery rather than nonoperative care, and pain and disability were similar in the operative and nonoperative groups. These findings demonstrate that radiographic considerations may be a more important factor in determining treatment in younger patients, and that health-related quality of life is a more important determinant of treatment course in adults older than age 50.
Spinal deformity and global coronal or sagittal plane imbalance may be related to regional malalignment within the spine or may have extraspinal causes. In the peripheral skeleton, hip and knee flexion contracture may cause global sagittal imbalance and should lead to evaluation of those joints, because osteoarthritis of the hip and knee is common in the older age group. A complaint of altered gait should also lead to evaluation for abnormalities of the cervical spine as a cause of the patient’s symptoms. For those who complain primarily of back pain, it is more challenging to determine if the deformity is the source of the pain. Briard and associates found that the convex region of the curve was the area of greatest pain in 75%, with the second most common location being the concave region of the curve. Primary complaints associated with deformity are often related to functional loss from a flat back deformity or a “rib in pelvis” deformity, which can also be accompanied by muscular fatigue and pain due to attempts to compensate for the sagittal and coronal imbalance.
Evaluation begins with history taking and physical examination, with emphasis on neural evaluation as well as spinal alignment and peripheral joint abnormality. Radiographic evaluation should include full-spine standing radiographs to assess segmental, regional, and global alignment of the spine. The plain radiographs should also be reviewed for other pathologic conditions, because degenerative spondylolisthesis and rotational subluxation may be important causes of stenosis and neural compromise. Bending or traction radiographs will aid in determining the flexibility of the curve. Anteroposterior (AP) and lateral tractions radiographs with or without the patient over a bolster at the apex of the deformity may be useful in preoperative planning for larger curves or rigid deformities that may partially correct during operative positioning.
Advanced imaging, including magnetic resonance imaging (MRI) or computed tomography (CT) myelography, is useful for the evaluation of intraspinal lesions and neural compression. CT myelography may also be valuable in measuring segmental ankylosis and facet joint defects. Evaluation of bone mineral density with a dual energy x-ray absorptiometry (DEXA) scan is particularly important in patients with degenerative scoliosis, because osteoporosis is an important comorbid condition that may affect surgical approaches and fixation strategies. The role of diskography remains controversial in scoliosis. Kostuik has found diskography to be useful in choosing distal fusion levels. However, Grubb and Lipscomb found high rates of concordant pain reproduction and concluded that decision making for surgery was not influenced reliably by diskography results.
An understanding of the natural history of disease progression is important in order for the patient and the physician to participate in informed decision making in adult scoliosis management. Informed choice in the decision to pursue operative or nonoperative care for scoliosis requires information on the outcomes of each option. There is significant variability in clinical and radiographic outcomes depending on treatment, preoperative patient factors, and comorbid conditions. Schwab and associates identified the following as factors that predicted a good outcome from surgery in scoliosis: older age, lower apex of deformity, and greater self-assessment of disability. Patients with high degrees of sagittal imbalance and patients treated with long fusion to the sacrum are most likely to experience perioperative complications. Rates of surgical complications in adult deformity surgery have been reported to be over 70%, and reoperation rates may be greater than 25% at 2 years. An understanding of the risks and benefits of operative and nonoperative treatment is important to guide an evidence-based approach to care. Smith and colleagues reported that complication rates were significantly higher in older patients than in younger patients, with complication rates of 17% in patients younger than age 65 and 71% in patients older than age 65. However, older patients experienced significantly greater improvements in pain and disability, which suggests that the risk/benefit ratio may be similar across all ages.
Little evidence exists to demonstrate the long-term effectiveness of nonoperative care of adult scoliosis in improving patient self-assessment of quality of life. Glassman and associates reported that the average yearly cost of nonoperative care of adult scoliosis was more than $10,800 over 2 years, and there was no significant decrease in pain, improvement in function, or reduction in disability in patients treated nonoperatively.
Clinical studies comparing operative and nonoperative care are limited. There is no published prospective randomized comparison of operative and nonoperative care of adult scoliosis. In a study comparing change in quality of life in patients treated operatively or nonoperatively for adult scoliosis, Bridwell and co-workers demonstrated significant improvement in self-assessed quality of life in adults treated operatively but no significant change in quality of life in adults treated nonoperatively. Similarly, Smith and colleagues studied operative versus nonoperative care in adults with back pain, leg pain, and scoliosis. These authors demonstrated a significant improvement in ODI and numeric rating scale scores for leg pain and back pain in the operative care group but no significant change in the nonoperative care group. These studies are limited by dissimilarities at baseline between the operative and nonoperative treatment cohorts.
The decision to pursue operative versus nonoperative care in the management of adult scoliosis is as important as the surgical strategy or technique selected in pursuing surgery. An informed choice requires information on the expected outcomes of operative and nonoperative care. Operative care results in greater improvement of self-assessed health-related quality of life in adults with scoliosis compared with nonoperative care. However, a direct comparison of outcomes for patients with similar health status and comorbid conditions at baseline remains to be completed.
Indications for surgery in adult scoliosis include progression of deformity, neural impairment, and pain or functional limitations that persist despite nonoperative care. There is significant variability in the operative plan for treatment of adult deformity, and the spectrum of surgical approaches may include decompression alone, limited fusion, and fusion of the structural deformity in its entirety. For fusion procedures, approaches may include anterior only, posterior only, and combined anterior and posterior approaches. An evidence-based approach to surgical technique is based on factors such as curve location and stiffness; medical comorbidities, including poor bone quality; and the preferences of the patient and surgeon.
Surgical treatment for the patient described in the Case Presentation must address the following questions: Where is the stenosis that causes the lower extremity pain? Which levels are potentially responsible for the back pain? Is the deformity responsible for some of the patient’s disability, and to what extent does it need to be corrected to achieve a satisfactory result for the patient?
Isolated decompression is best performed in a patient with radicular symptoms or symptoms related to spinal stenosis but with minimal back pain. Advanced deformity may be a relative contraindication for isolated decompression, because the destabilization inherent in decompression may worsen the course of the scoliosis in curves already known to be predisposed to progression. Patients should be counseled that back pain symptoms may not improve and may even worsen after isolated decompression. Bridwell presents some guidelines for stratifying surgical treatment based on the amount of deformity in three dimensions as well as the severity of stenosis. Patients with primarily neural symptoms and limited rotational subluxation or spondylolisthesis, as well as patients with ankylosis of motion segments, may be appropriate candidates for decompression alone. The presence of larger bony structures, such as a wide pars interarticularis, and stabilizing osteophytes can help prevent iatrogenic instability. These patients should be followed closely postoperatively for recurrence of neurologic symptoms, iatrogenic instability, and progression of deformity.
A limited fusion may be a useful option to permit decompression of the neural elements, with the decompressed levels stabilized but the extent of the fusion restricted. Posterior decompression and fusion of a portion of the curve or sometimes even an area below the apex of the curve may be a viable option. The reasons for taking this approach vary, including the need to alleviate iatrogenic instability caused by necessary or aggressive decompression, treatment of arthritic levels responsible for back pain, and avoidance of a larger procedure in a patient who may not tolerate longer time under anesthesia or the metabolic demands of a longer fusion. Once again, as in the case of isolated decompression, the patient must be followed to detect worsening neurologic symptoms and postoperative progression of deformity. In addition, with spinal fusion, adjacent-segment disease may become a concern.
Posterior spinal fusion for deformity correction with decompression of the stenotic levels is often adequate surgical treatment even in the face of significant deformity. It has the advantage of addressing affected levels from a single approach and allowing direct decompression of compressed neural structures. Posteriorly based osteotomies, including Smith-Peterson osteotomies, pedicle subtraction osteotomies, and vertebral column resection, present the opportunity to address coronal and sagittal plane imbalance. For less severe deformity, multiple Smith-Peterson osteotomies may allow enough flexibility in the curve for adequate correction to be accomplished with instrumentation and fusion. The degree of correction of sagittal plane deformity per Smith-Peterson osteotomy is 5 to 8 degrees per level. Pedicle subtraction osteotomies allow greater sagittal correction through a single level and, via asymmetric osteotomy, may allow significant coronal plane correction. Vertebral column resection is the most powerful osteotomy, but carries greater risk to the neural elements. Vertebral column resection may also be less relevant for degenerative scoliosis, in which deformity occurs over several levels, than for a focal deformity, which vertebral column resection is ideally designed to treat.
Anterior approaches are often combined with posterior approaches, and this allows for greater correction of deformity with anterior release. In addition, anterior interbody grafting increases fusion potential. Anterior interbody grafting permits indirect decompression when foraminal height is restored. Performing the anterior stage first and delaying the posterior stage can allow evaluation for resolution of radicular symptoms, which may obviate the need for decompression at the time of posterior instrumentation and fusion.
Since the apex of lumbar degenerative scoliosis is usually in the upper lumbar spine, the lumbosacral junction is not always surgically addressed. In cases in which stenosis is present at the L5-S1 level or long constructs end at L5 or S1, there is the question of whether the L5-S1 disk space can tolerate the increased stress. There is biomechanical and clinical evidence to suggest that when the choice has been made to fuse the L5-S1 joint, augmentation with iliac fixation is beneficial. Anterior interbody grafting at the L5-S1 interspace may be necessary simply to improve the fusion rate at this vulnerable level. In this chapter the two approaches to be compared are posterior fusion with lumbosacral junction grafting only (without addressing other levels anteriorly) and a combined anterior and posterior approach.
On the cephalad end there is the question of how high to fuse. Since many curves will have their apex in the upper lumbar spine, fusion constructs that completely address the curve may end at the thoracolumbar junction. Multiple studies have found the potential risk of proximal junctional kyphosis to be 24% to 60%. In general, the fusion should extend proximally to include the entire curve in the coronal plane and cephalad, above the thoracolumbar junction when segmental kyphosis of the thoracolumbar junction is more than 5 degrees.