Discussion
The strength of this study was its prospective study design for both radiological and clinical methods. The main finding of our study was that there is no linear correlation in the radiological degree of severity of LSS and clinical findings. In contrast, according to the visual evaluations of the central canal LSS, leg pain measured by VAS was higher in the moderate stenosis group than in severe stenosis group. Additionally, the walking distance achieved was shorter in the patients with moderate stenosis on visual evaluation compared with the patients with severe stenosis. This finding was consistent for both the analysis performed using the maximal degree of stenosis and among patients with multilevel stenosis. We did not find any correlation between objective quantitative radiological measures and patient symptoms, which also supports the paradoxical finding based on visual evaluation.
Our results indicate that LSS is not solely an anatomical disorder, but that this disease may have other underlying pathobiological mechanisms to be discovered. Indeed, we found that the correlation between the severity of LSS and clinical findings is complex, with milder symptoms in patients with more severe stenosis. In a cross-sectional study setting, we cannot provide a definite explanation for our findings. Our results raise the possibility that the pain of patients could resolve spontaneously across time and that this adaptation could possibly explain the longer walking distance achieved in this group of patients regardless of the progression in the severity of the central LSS. One possible explanation to our apparently paradoxical findings could be the decreased lumbar spine instability in patients with advanced facet joint hypertrophy and large end-plate osteophytes, which in turn would provide pain relief and allow higher walking capacity. Accordingly, degenerative hypertrophy could be a protective mechanism against the disc degeneration typically found in patients with advanced age. Porter and Ward hypothesized that central stenosis at two levels or central stenosis at one level with lover root canal stenosis may cause venous congestions and may explain neurogenic claudication.[16] We are not aware of the methods to assess venous congestion on the MRI, which could be a potential target of future research.
Sirvanci et al. found no correlation between the severity of spinal stenosis and ODI. The aforementioned study, however, was retrospective and patient symptoms were evaluated only by the ODI scale. Moreover, no data of the experience of the subjects performing the radiological analysis were provided in that study.[5] Accordingly, it is difficult to evaluate the reliability of the radiological analysis. In the study by Geisser et al., no correlation was found between the quantitative measurement of central spinal canal AP diameter and clinical symptoms.[3] Assessment of spinal canal AP diameter may be problematic in the context of LSS because, according to our experience, the most common reason for LSS is facet joint hypertrophy that causes bilateral stenosis of the dural sac and does not influence the mid-sagittal level. Jonsson et al. found a weak positive correlation between the central spinal canal AP diameter and reduction of the patient's estimated walking ability; however, that correlation was not statistically significant.[4] Haig et al. evaluated the LSS by measuring the area of the minimal dural sac cross-sectional area, as also performed in our study, and they found no difference in the degree of stenosis between symptomatic and asymptomatic subjects.[2] Interestingly, in the one other study using validated methods to record patient symptoms, patients with multilevel spinal stenosis had significantly better scores in the general health items of the Short Form-36, and similar to our findings, moderate lower leg pain measured with the VAS-scale.[17] Further, Park et al. found that there was less pain radiation and pseudoclaudication in patients with three- and two-level spinal stenosis compared with patients with one-level stenosis only.[18] In contrast, Ogikubo et al. found lowered preoperative walking capacity, higher leg and back pain and reduced quality of life in LSS patients with smaller dural sac cross-sectional area.[19] Furthermore, Yukawa et al. found a positive correlation between the preoperative dural cross-sectional area in magnetic resonance imaging and with a better postoperative ODI score.[20] The aforementioned and many other studies[2–5,17] did not analyze spinal canal stenosis visually, which we considered an elemental part of the image analysis, especially in patients with stenosis at the upper part of the lumbar spine. The amount of neural tissue at L1–2 and L2–3 levels is considerably greater than at the L4–5 or at the presacral measurements, and thus, by performing dural sac cross-sectional area measurements only, subjects with reduced space for neural tissue may not be correctly recognized.
We did not execute intra-rater and inter-rater repeatability of MRI evaluation, and this is the main weakness of this study. However, we consider that such a measurement was not related to the present study aim. Notably, reliability of the qualitative grading of LSS has been described and evaluated previously, and it was shown to have substantial intraobserver and moderate interobserver agreement in a multicenter study setting. In the present study methods, a 7-grade classification has been used.[21] However, Lurie et al. used a 4-grade classification in their study and showed moderate to substantial reliability.[22] Moreover, we have recently extended the method of the assessment of lateral stenosis using a 3-grade classification, which has been demonstrated to have acceptable repeatability for research purposes.[23] Future objective would be standardized studies of visual assessment of the LSS and to analyze how these findings correlate with patient symptoms and surgical outcomes.
The results of the current study relate to routine clinical MRI with patients lying in the supine position. Imaging studies of patients in this position is a limitation because patient symptoms may worsen in an upright position. Further, the anatomy of the neural canal may appear altered when patients are in an upright position. Accordingly, the upright position would be the most appropriate imaging acquisition posture to link imaging findings to patient symptoms.[24–27] Hiwatashi et al. found that axial loading while performing imaging studies could even influence to treatment decisions.[27]
The incidence of LSS is increasing probably because of the better quality and availability of radiological imaging equipment, and facilities, added to increasing aging population,[28,29] which reflect in a higher number of LSS surgery. However, selection of patients for surgical treatment still remains challenging. Our results strengthen the classical conception that the diagnosis of this syndrome is constituted by the clinical history, clinical symptoms and radiographic evidence of a demonstrable stenosis.[30–32]
MRI evaluations are thus needed to establish the level (s) and severity of stenosis. However, MRI images cannot be the only decision-making factor of surgical treatment selection for LSS patients. The degree of the severity of the disease cannot be judged based solely on MRI either. Ohtori et al. found that proinflammatory cytokine levels in the cerebrospinal fluid of patients with LSS correlated with the severity of the stenosis.[33] Sairyo et al. found that hypertrophy of the lumbar ligamentum flavum is associated with inflammation-related genes.[34] Moon et al. pointed out that fibrosis and scarring during inflammatory reaction is the major pathomechanism of ligamentum flavum hypertrophy.[35] The present study adds to the current knowledge by showing that there is no straightforward association between stenosis of dural sac and patient symptoms or functional capacity, which indicates that dural sac stenosis is not the only key in the pathophysiology of LSS. It is not justified to select patients for surgery based solely on the degree of central stenosis either.
BMC Musculoskelet Disord. 2014;15(348) © 2014 BioMed Central, Ltd.
© 1999-2006 BioMed Central Ltd
