Degenerative lumbar scoliosis (DLS) is believed to develop as the result of asymmetric degeneration of disc, osteoporosis, and vertebral body compression fractures[16] , and the etiology of DLS remains difficult to be determined. DLS may start with instability at any lumbar intervertebral space. During initial wedging, the remaining vertebrae consequentially wedged toward the opposite apex to maintain balance, and therefore the initial wedging retrogressed[7]. DLS rarely presented as progression of adolescent idiopathic scoliosis, however, it appeared the deformity of the spinal column, by altering the mechanical loading conditions, can accelerate the degenerative cascade.[3] Schwab et al[17]suggested that the initiating cause of DLS was disc degeneration at lumbar level and illustrated the pathological degenerating process of DLS, and this progression may be further accelerated by aging or osteoporosis, particularly in post-menopausal female patients. DLS typically presented with symptoms of LBP, and approximately 90% of patients reported pain as their primary complaint[18]. Pain that localizes over the convexity of the curve is often axial and diffuse in nature which may attribute to muscle fatigue and/or spasm of the paraspinal muscle. However, pain on the concavity of the curve may be the result of the back and nerve roots[8, 19]. In conclusion, LBP was the most frequent complaint in DLS patients.
Surgical regime should be offered to patients with pain-related deformity who have failed conservative treatments. Arthrodesis appears beneficial for stabilization, correction, or prevention of curve progression. However, correction of the sagittal balance should be assessed because the loss of lumbar curve has been shown to be associated with poor outcomes[17]. Patients with lumbar curves more than 40°[19], kyphosis greater than 5°and a persistent neurological deficit (radiculopathy or neurogenic claudication) should be considered surgical treatment. Consequently, we expected that radiographic parameters for DLS should be standardized to serve as a basis of analysis and perhaps eventually as treatment guidelines; and surgeons should be aware of the epidemiology, prevalence and related radiological changes [20]
The question that arose, therefore, was whether radiographic parameters can be identified in an adult with DLS that will serve a predictive purpose in clinical practice. In this investigation, we illustrated radiographic parameters such as age, Cobb angle, among others.
In agreement with widely accepted assumption, this study demonstrated that the deformity become more prevalent with age advance, especially in the populations of 50 years or older. This increase seemed to be the corollary of lumbar spine degenerative changes during adulthood.
Few studies assessed the prevalence of DLS in LBP patients. According to a study of 671 patients with low back pain carried out by Perennou and his colleagues[21],7.5% of patients met the criteria for DLS. In addition, their data showed that the incidence of DLS was 15% among patients of over 60 years, 6% in 45 to 60 age group and 2% in patients younger than 40 years of age. With a much large sample size, our study suggested a DLS prevalence of 9.7% in patients with LBP, which was similar to the Perennou’s report . We also noticed a linear relationship between the prevalence and age. However, the deformity incidence in patients over 60 years old was 20.9% in our study, dramatically higher than the Perennou’s result. The ratio of men's prevalence to women's in our study was 1 to 1.9, which was similar to Murata’s observation[22] (1 to 2.36) and suggested that women were more affected by DLS.
Although the estimated increase in Cobb angle with age advanced was very slow, the significant relationship between the Cobb angle and age was a linear regression, which suggested that the increase was also constant[20]. But in our study, the Cobb angle was similar between age and gender groups. Data suggested the mean of Cobb angle was 15.3°, smaller than that of adolescent idiopathic scoliosis. The Cobb angles of patients in our study covered from 10° to 54.5°, which was not a significant criterion in this group of DLS patients, but it correlated with pain as Schwab and his colleagues considered[17]. Cobb angle may reveal lumbar instability, which may cause pain, but researches[23, 24] revealed a often smaller degrees of coronal Cobb angle in DLS than that of idiopathic scoliosis, and there was no significant relationship between the Cobb angle and the clinical symptoms.
As our analyses indicated, coronal Cobb angle had not been affected by imaging postures as well as apical rotation and lateral olisthesis. The curve was rigid and without significant movement, which was different from idiopathic scoliosis, so radiological parameters were similar between standing and supine groups except lumbar curve related to patient’s gravity when standing. Rigid curve and loss of flexibility in DLS patients because osteophytes at the facet joints and at the vertebral endplates, and calcification of the ligamentum flavum and joint capsules. But because of distinct loadings, lumbar curvature of supine patients normally showed obviously lower degrees than that of standing patients. Although clinical symptoms and sagittal curvature could not be veritably identified with supine posture and most patients were recommended to be examined erectly. It had been reported that DLS often accompanied loss of lumbar lordosis[25], which may become the underlying factor in LBP. In our study, lumbar lordosis decreased both in standing and supine groups.
Differing from previous reports[21], the patients enrolled in our study performed more left scoliosis (138 cases), which was different from the adolescent idiopathic scoliosis, but showed no statistic difference of curve direction between age and gender groups. Previous studies[1, 2] indicated that most degenerative curves involved a short segment in lumbar region (T11-S1), and the apex frequently located in L3/4 or L2/3 intervertebral space. In our study, DLS involved 3 to 5 vertebrae, and the apical vertebra was mostly identified as L3.
The most significant radiographic parameters were rotary subluxation or lateral translation between adjacent lumbar vertebrae on the frontal plane, curve magnitude and its apical vertebra. DLS is a complex three-dimensional change with axial rotational disorder and coronal and sagittal tilting of the vertebra, and lateral olisthesis of the apical vertrbra is related to the progression. When osteoporotic facet joints and vertebral endplates can not compensate the progression of asymmetrical deformity, vertebrae start to rotate and olisthesis in the coronal plane or to shift in the sagittal plane. Common radiographic findings in DLS patients include degenerative changes, most commonly at L5/S1, as well as obliquity at L4/5 and rotary subluxation or lateral translation at L3/4[26], which may indicate severe pain and bad prognosis. Lateral olisthesis appears a late complication of DLS which is more common in those with severe and moderate degrees of scoliosis, and we found maximal Lateral olisthesis with a mean of 6.6mm most located at L3/4or L4/5, without difference between age and gender groups. In Perennou’s[21] study, the correlation was demonstrated between anterior redicular pain and dislocated lumbar scoliosis, especially at the L3/L4 and L4/L5 levels. Degenerative instability and unfavorable lumbar vertebral alignment were related to symptoms rather than lumbar curvature or the cause of the original scoliosis[17] .
Shift in the sagittla plane often occurred in spondylolisthesis, an unisegmental form of unstable spinal stenosis, which was found in 95 patients with scoliosis, of which 46 patients forward slipped mostly at L4/5 or L5/S1 and 49 patients slapped backward in (mostly at L2/3 or L3/4). In the 200 adult patients with scoliosis examined by Pritchett and Bortel[27], degenerative spondylolisthesis coexisted in 111. However, the incidence of spondylolisthesis was lower in our study. Murata et al[22] found that DLS was associated with a decrease in segmental lumbar lordosis. Sagittal imbalance occurred gradually with the progression of DLS, which was related to symptoms.
Unlike idiopathic scoliosis which usually has no correlation between DLS and relevant radiographic parameters, data from our study revealed that there were positive correlations among apex vertebra rotation, the maximal lateral listhesis and coronal Cobb angle. Because of significant postural effects, when studying the lumbar curvature, we analyzed two posture groups respectively and found a negative correlation between lumbar curvature and the Cobb angle. Therefore, we suggested that three dimensional deformities consisted of vertebral rotation, lateral olisthesis, sagittal imbalance, and mostly the deformities were concurrence and interacting with each other.
We believed that our study may be applied to evaluation and characteristics for epidemiology of DLS, and further study focused on DLS in the normal population will help to more accurately understand the epidemiology and find out more closer association between DLS and LBP.