NATURAL HISTORY OF LUMBAR CANAL STENOSIS: CLINICAL ASPECTS AND SAGITTAL BALANCE

ABSTRACT Objective To compare the sagittal alignment (SA) parameters in individuals with LCS and surgical indication with a control group and to study the correlations between SA parameters and ODI, VAS and EQ-5D in individuals with LCS and surgical indication. Methods In this multicenter cross-sectional case-control study, the individuals were allocated as follows. A stenosis group (SG) composed by patients with LCS confirmed by magnetic resonance imaging with surgical indication, treated between July 2010 and August 2016 and a control group (CG), without LCS. All subjects underwent anamnesis, completed the Health-related Quality of Life (HRQoL) and total spine radiographs were taken. Clinical data, HRQoL and radiographic parameters were correlated. Results Sixty-four individuals formed the SG and 14 the CG. The SG had higher values of mean age, coronal imbalance, sagittal vertical axis (SVA), pelvic tilt (PT), sacrofemoral distance (SFD), overhang (OH), PI-LL mismatch, Oswestry Disability Index (ODI) and Visual Analog Scale (VAS) for pain and smaller thoracic kyphosis (TK), total (TLPL) and regional lumbopelvic lordosis (RLPL) in all vertebrae, sagittal offset (SO) in all evaluated vertebrae and EuroQol-5D (EQ-5D) with p <0.05. In the SG, the only significant correlations (p <0.05) were between TK and ODI and EQ-5D; all the other sagittal parameters did not correlated with VAS, ODI or EQ-5D. Conclusion SG had SA parameters altered in relation to CG. There was a direct correlation between decrease in TK and worsening of ODI and EQ-5D in SG. Level of evidence: III; Case Control Study.


INTRODUCTION
Degenerative lumbar stenosis (DLS) is a narrowing of the lumbar spinal canal and/or the vertebral foramens, characterized by low back pain, radicular pain, and/or claudication. It is a common clinical condition, affecting more than 200,000 people in the USA, 1 being the main reason for indications of spinal surgery in patients above 65 years of age and affecting 1 in every 1000 people per year in this age group. 2 The increasing prevalence of disease, considered an exacerbation of the degenerative physiological process of aging, is expected as the life expectancy of the population continues to rise. 3,4 Disc degeneration is often associated with DLS, 3 usually occurring at the lowest lumbar levels 5 and causing loss of disc height from dehydration. Consequently, there is an overload on the facet joints, with arthritis and facet joint hypertrophy, thus causing a loss of lumbar lordosis (LL). 6 There is also a dynamic component, since the canal space decreases with extension and increase with the flexion-distraction of the trunk. 7 Flexion of the trunk increases the area of the lumbar foramens by 12% and in extension there is a reduction of 15% of their sectional area. 8 There is an association between structural changes of the spine and antalgic position, 9 which culminates in a posture of anteriorization of the trunk. Compensatory mechanisms, such as an increase in pelvic tilt (PT), are activated in an attempt to reduce the loss of sagittal alignment (SA). 10 Extensive literature supports the importance of SA and the recognition of its compensatory mechanisms in the treatment of degenerative lumbar spine diseases in cases where surgery is indicated. This becomes even more important when intraoperative sagittal correction with instrumentation and vertebral fusion is necessary. [11][12][13][14] The objectives of this study were to evaluate whether there were any changes in the SA parameters in DLS patients indicated for surgical treatment as compared to a control population and to study the correlation of the SA parameters with the Visual Analog Scale (VAS), 15 Oswestry Disability Index (ODI), 16 and EuroQol -5 Dimensions (EQ--5D) 17 in DLS patients with an indication of surgical treatment.

METHODS
Study and sample design: This was an original cross-sectional multicentric case-control study.
The Stenosis Group (SG) was comprised of patients with DLS confirmed by Magnetic Resonance Imaging with indication of surgical treatment (refractory symptoms after at least 4 months of the best conservative treatment). The Control Group (CG) included healthy individuals without any diagnosed spinal diseases. Data collection took place during the period from 07/03/2010 to 08/12/2016.
Individuals who refused to participate in the study, those with a history of spinal surgery, previously known disease of the hips and/ or pelvis, and individuals with a prior diagnosis of neuropathy were excluded from the study.
Clinical data such as data of evaluation, age, weight, height, tobacco use, and comorbidities were collected, in addition to VAS, ODI, and EQ-5D scores.
This study was approved by the Institutional Review Board (protocol number 50833515.3.0000.5404) and the data were collected after the study participants had signed the Informed Consent Form.
Digitalized frontal and lateral full spinal radiographs were taken of the study subjects in a standing position with hips and knees in extension. In the frontal radiographs, the subjects let their arms hang de Discapacidad de Oswestry (ODI), Escala Visual Analógica (VAS) para el dolor y menos cifosis torácica (CT), lordosis lumbopélvica total (LLPT) y regional (LLPR) en todas las vértebras, offset sagital (OS) en todas las vértebras evaluadas y EuroQol-5D (EQ-5D), con p < 0,05. En el GE, las únicas correlaciones significativas (p < 0,05) fueron entre TC y ODI y EQ-5D; todos los demás parámetros sagitales no se correlacionaron con EVA, ODI o EQ-5D. Conclusión: El GE tuvo parámetros de AS alterados en relación con el GC. Hubo correlación directa entre la disminución de TC y el agravamiento del ODI y EQ-5D en el GE. down close to the body and in the lateral radiographs the arms were flexed at 45 degrees in relation to the plane of the body. We used 30 x 90 cm film and the radiographs were centered on T12 during inhalation, with a distance between the film and the focus of 230 cm. The images were digitalized and the parameters were measured and reviewed by an orthopedist familiar with the indices described, using the Arya PACS Aurora v.1.9.19 program (São Bernardo do Campo, Brazil).
Vertebral and spinopelvic parameters measured are shown in Figure 1 and described in Chart 1.

Chart 1. Radiological parameters used in this study.
A TK (thoracic kyphosis): angle between T4 ptp and T12 dtp LL (lumbar lordosis): angle between L1 and ptp S1 SS (sacral slope): angle between S1 ptp and hl PT (pelvic tilt): angle between line that joins chr to the midpoint of the S1 ptp and vl PI (pelvic incidence): angle between line that touches the chr and midpoint of the S1 ptp and the line orthogonal to the S1 ptp

Statistical analysis
The radiographic measurements and personal characteristics of the CG and SG were described as mean and standard deviation (SD) or median, minimum, and maximum and compared using the Student's T-test or Mann-Whitney test. For the comparison of qualitative measurements, the Fisher's exact test was used. To check the correlation between the radiographic measurements and the questionnaires evaluated for each group, the Spearman correlation was calculated. 18 The analyses were conducted with the use of IBM-SPSS v. 20.0 software (Chicago, USA) and tabulated using Microsoft-Excel 2003 v. 11.0 software (Redmond, USA). All tests were performed with a level of significance of 5% (p < 0.05).

RESULTS
During the study period, 64 individuals were included in the SG and 14 in the CG. In the SG there were 33 men and 31 women, ranging in age from 34 to 82 years (mean of 60 years). In the CG there were 4 men and 10 women, ranging in age from 25 to 62 years (mean of 49.1 years). Table 1 summarizes the differences in personal characteristics between the two groups and we observed that the only statistically relevant variable was age, lower in the CG. Table 2 shows that patients in the SG had statistically higher values for coronal imbalance, SVA, PT, SFD, OH, PI-LL mismatch, ODI, and VAS and lower values for TK, TLPL, RLPL in all vertebrae, T1SO, T9SO, and EQ-5D (p < 0.05). Table 3 Table 4 correlates the radiographic parameters measured to the VAS, ODI, and EQ-5D questionnaires in the SG. It shows that TK presented an inverse correlation with the ODI (r = -0.273, p=0.038) and a direct correlation with the EQ-5D (r = 0.428, p=0.001), while there was no relationship observed between parameters such as SVA, SSA, TPA, and PI-LL and the questionnaires in the SG.
It was not possible to visualize the proximal terminal plate of T4, and therefore not possible to measure TK, in 6 out of the 64 individuals participating in the study. Of the 58 individuals had TK T4-T12 measured, 4 were hypokyphotic, 47 had normal kyphosis, and 7 were hyperkyphotic, considering normal to be between 20 and 40. 19 Table 5 shows the mean TK, ODI, VAS, and EQ-5D by TK classification group.

DISCUSSION
Today there is extensive discussion about the real importance of the SA parameters and whether they are able to predict quality of life in the patient with DLS. [20][21][22] The data found in our study are in agreement with those reported by Cavali et al. in terms of the loss of SA in the SG as compared to the CG. The same authors also found that the population of DLS patients was older than the randomly selected control group. 19 Several studies argue that an increase in SVA is associated  with a poorer quality of life and an increase in axial pain. [23][24][25] In 2013, Schwab concluded that serious clinical disability (ODI > 40) is associated with an increase in SVA, PI-LL mismatch, and PT. 26 Lafage also concluded that an increase in PT is correlated to a poorer quality of life and that T1SPi is more highly correlated to the quality of life questionnaires than the SVA. 27 In the literature it is also noted that abnormal TPA values are associated with a worse quality of life in adults with degenerative scoliosis 28,29 and that an increase in C7PL/SFD is related to a worse quality of life in degenerative lumbar disease. 30 Nevertheless, most studies that report a correlation between the SA parameters and preoperative quality of life use univariate analysis or multivariate analysis adjusted only for age, [23][24][25]31 which creates a strong confusion bias.
Takemoto compared SA parameters with the ODI in 204 cases of adult spinal deformity pre-and postoperatively to deformity correction surgery. In the multivariate analysis, the SA parameters were not significantly correlated with the preoperative ODI. However, he observed that the postoperative improvement in SA resulted in a better ODI. He concluded, therefore, that in surgical cases, reestablishing the SA should be sought with the goal of achieving an improved postoperative ODI, although changes in the SA should not be decisive in a surgical indication. 32 The lack of a correlation observed between the SA parameters and quality of life in the above-mentioned study is compatible with the data obtained in our study. This raises questions about the reproducibility of the results reported in studies that claim that the SA parameters are predictors of quality of life in individuals with DSL.
Besides the debatable value of using SA parameters to predict quality of life in DSL patients, said parameters are not correlated to non-specific 33   low back and are not useful as a tool for screening this condition. 34 We did not find a study that correlated TK and quality of life in DSL. A recent descriptive analytical study evaluated 34 women with osteoporosis and observed that a decrease in TK had an inverse correlation to gait performance and to the SF-36, 35 results similar to those found in our study.
One limitation of our study was comparing a population of DSL patients with surgical indication to a control group with a lower mean age.

CONCLUSION
This study showed a significant loss in SA parameter values in patients with DSL indicated for surgery as compared to a control population. It also revealed a direct correlation between a decrease in TK and a worsening of the ODI and EQ-5D quality of life indicators in these patients.