Background
The term lumbar spinal stenosis (LSS) is commonly used to describe patients with symptoms related to anatomical reduction in lumbar spinal canal. Among older individuals, LSS is a highly disabling condition [
1] and is the most common reason for spinal surgery [
2,
3]. The most common procedure involves a decompressive laminectomy of the structures thought to be causing nerve root irritation.
The challenge to the anatomically based determination is that while necessary for the diagnosis of LSS, it is not sufficient to determine the severity of symptoms that leads a patient to seek treatment [
4]. The extent of narrowing of the spinal canal correlates poorly with symptom severity, and radiologically significant lumbar spinal stenosis can be found in asymptomatic individuals [
4‐
7]. As a consequence, correlating symptoms and physical examination findings with decompression levels based on common imaging results is not reliable. In patients who have no concordance between radiological and clinical symptoms, the surgical levels determined by conventional magnetic resonance imaging (MRI) and neurogenic examination (NE) may lead to a more extensive surgery and significant complications. It is important to avoid inadequacies of MRI (MRI cannot precisely determine the lesion levels of lumbar spinal stenosis) in clinical practice.
Diffusion tensor imaging (DTI) is more sensitive than conventional MRI for precise determining the extent of spinal disorders via non-invasive, longitudinal examinations, in both humans and animal models. Moreover, the analysis of the fractional anisotropy (FA) proves more useful than other diffusional indices because of its simplicity, accuracy, and ability to reveal diverse spinal cord disorders, especially in clinical situations [
8]. The quantification of the nerve root using the proposed methodology of DTI can identify the specific site of any degenerative and inflammatory changes along the nerve roots of patients with lower back pain [
9].
Paraspinal mapping (PM) is a technique for needle electromyography (EMG) of the paraspinal muscles that has been the subject of several studies [
10‐
12]. Although conventional imaging studies have a high false positive rate (a level with anatomical stenosis that is clinically irrelevant) for disc herniations, PM rarely produces evidence of radiculopathy in individuals without pain [
12]. Theoretically, a single insertion into the location of each root level would assess for lesion in each root [
13]. The PM is a sensitive method in the diagnosis of lumbar spinal stenosis and reflects physiology of the nerve roots better than the limb EMG [
14]. Therefore, in the lumbar spinal stenosis, changing of DTI parameters (FA) or PM scores may possibly reflect the lesions of the cauda equina and/or spinal nerve roots more accurately than conventional MRI.
We hypothesized that the use of DTI and PM techniques can further prevent the occurrence of false positives with conventional MRI, distinguish which are clinically relevant from levels of cauda equina and/or nerve root lesions based on MRI, and determine and reduce the decompression levels of lumbar spinal stenosis than MRI + NE, while ensuring or improving surgical outcomes.
Discussion
Yamashita et al. [
30] have demonstrated the feasibility of whole-body MR neurography with the use of DWI that can depict tissues with an impeded diffusion, such as tumors, brain, spinal cord, and peripheral nerves. MR neurography by using DWI can clearly show lumbar spinal nerves, and the mean ADC in the nerve root entrapment with foraminal stenosis is higher than in the intact nerve roots by using MR imaging at 1.5 T [
31]. The ADC map is limited because the tissue contrast between the nerves and surrounding tissues is poor [
32]. FA had a much higher sensitivity and specificity (73.3 and 100 %) in the detection of the spinal cord abnormalities compared with T2-weighted FSE imaging (46.7 and 100 %) and ADC (13.4 and 80 %) [
33].
A few recent DTI studies of lumbar spinal nerve were demonstrated by Balbi et al. [
34] at 1.5 T and van der Jagt et al. [
35] and Budzik et al. [
36] at 3 T. Also, DTI studies of the cauda equina were demonstrated by Tsuchiya et al. and Filippi et al. [
37,
38]. all these showed that DTI can determine the FA of the spinal nerves and/or cauda equina in patients and healthy volunteers.
In this study, averages of reference FA values were cauda equina, 0.437 ± 0.028; left nerve root, 0.457 ± 0.026; and right nerve root, 0.467 ± 0.026. Averages of the FA values of negative levels and levels were cauda equina, 0.408 ± 0.045; left nerve root, 0.484 ± 0.072; and right nerve root, 0.487 ± 0.055. Our FA values of nerve roots were not comparable to those obtained in the study of lumbar spinal nerves by Balbi et al. [
34] (0.218), van der Jagt et al. [
35] (0.31), and Budzik et al. [
36], which might be due to the different software calculation methods. Our reference FA value and negative FA value of cauda equina were larger than the gray matter (0.32 ± 0.11), less than the white matter (0.63 ± 0.08) [
39,
40], as well as lower than the average of cauda equina (0.492) [
38]; because at the L1 level, the FA value we measured was actually a FA value of mixture of gray matter, white matter, and cerebrospinal fluid; at L2—S1 levels, the FA value we measured was actually a FA value of mixture of cauda equina nerve and cerebrospinal fluid. the cerebrospinal fluid would reduce the FA value.
By contrast, the FA values of positive levels compared with the reference FA value that were decreased with statistically significant differences which showed that reduction of the FA value ≥0.1 than the reference FA value was of statistical significance. Eguchi et al. [
27] showed that the mean FA of the proximal nerve roots on the side of entrapment was 0.128 ± 0.036, which is significantly lower than the 0.213 ± 0.042 on the intact side, and the mean FA of the distal lumbar spinal nerve roots on the side of entrapment was 0.131 ± 0.014, significantly lower than the 0.242 ± 0.032 seen on the intact side (
p ≥ 0.001). The difference between normal side and entrapment side values was about 0.1; our pre-experiment also showed that the FA value of cauda equina and/or nerve roots of the level was ≤0.1 than that of the normal level which was clinically meaningful. According to the schema above, we set the standard as follows: If the FA value of lumbar cauda equina and/or nerve roots of the narrow level decreased ≥0.1 than that of the non-stenotic and normal level (commonly taken T12–L1 cauda equina and nerve root value as reference), it was positive and the level should be treated surgically.
Although the mechanisms of decreasing FA in nerve roots have been controversial, these findings suggest that diffusion in the tissue had become more isotropic because of edema, in which fluid is trapped in the tissue, creating an isotropic environment and a reduction in FA. These hypotheses have been supported by previous experimental studies. Beaulieu et al. [
41,
42] reported that Wallerian degeneration after peripheral nerve injury reduces the anisotropy of water diffusion. Several studies indicated that the FA of peripheral nerves was strongly correlated with the axonal degeneration and regeneration in rat and mouse sciatic nerves [
43,
44]. The decrease in the FA values may reflect the degree of microstructural disorganization of the spinal cord, suggesting either local extra-cellular edema or a smaller number of fibers matching a larger extracellular space, or both. On the other hand, minor lesions and edema with roughly preserved fibrillary microstructure of the spinal cord are not associated with major FA changes, which opposes to the demyelination, cavitations, and necrotic changes [
45]. Thus, the high FA values suggest that the microstructure of the spinal cord is preserved, even in cases with high signal intensity of the spinal cord on T2-weighted images, maybe so does the cauda equina.
Our 3.13 × 2.54 × 3.0 mm
3 voxel size was larger than that in the previous study (1.1 × 1.6 × 3.0 mm
3), and therefore spatial resolution was unlikely to account for the difference [
38]; it might be due to attempts to increase resolution by decreasing voxel size would lead to a bad result in lumbar nerve root imaging. The FA values of the cauda equina were typically lower than the actual values which might be due in part to volume averaging with cerebrospinal fluid (CSF) in each voxel. All the above affected FA values of the reference and narrow levels but not their difference.
The PM scores of positive compared with the negative PM scores obviously increased with statistically significant differences (left nerve root, p = 0.000; right nerve root, p = 0.000) which showed that the standard of PM was statistically significant.
Levels of decompression determined by MRI + (PM or DTI) in the experimental group were less, statistically significant than that determined by MRI + NE in the control group which demonstrated that the use of PM and DTI can further prevent the occurrence of false positives with conventional MRI, distinguish which are clinically relevant from the cauda equina and nerve root lesions based on MRI, and determine and reduce the decompression levels of lumbar spinal stenosis than MRI + NE.
A positive EMG, based on spontaneous activity findings, can reassure clinicians that a lesion seen on an imaging study is indeed a pain generator [
46]. Haig et al. [
47] argued that imaging does not differentiate between symptomatic from asymptomatic individuals, whereas electrodiagnosis does. They believe that the radiographic findings alone are insufficient to justify the treatment for spinal stenosis. In chronic degenerative myelopathy caused by disc herniation or degenerative spinal canal stenosis, significant decrease of FA has been found, including cases with no visible changes in the spinal cord on plain MRI [
45,
48,
49]. In recently published reports on the contribution of DTI in cervical myelopathy, the authors have claimed that DTI proved to be more sensitive than conventional T2-weighted images in the assessment of cervical degenerative myelopathy [
45,
49‐
51]. Our results further indicated that DTI or PM can accurately identify the cauda equina and/or nerve root lesions than MRI in lumbar spinal stenosis, aviode the occurences of false positive with MRI.
The positive and the negative predictive values of (PM or DTI) in distinguishing which are clinically relevant from the decompression levels determined by MRI were all 100 % which demonstrated good diagnostic effect.
Because decompression levels in the experimental group were statistically significantly reduced compared with the control group, the corresponding surgical blood loss, surgical time, and surgical transfusion in the experimental group were also statistically significantly reduced than that in the control group. Apparently on reducing the decompression levels, the surgical dissection and complexity of the surgical procedure were reduced, which in turn reduced the amount of bleeding, surgical time, and surgical transfusion. The experimental group reported three cases of leg dysesthesia because of surgical complications, and no such events were reported in the control group; however, there was no statistically significant difference in terms of complications between the two groups.
In the follow-up, the averages of postoperative VAS-BP, VAS-LP, and ODI scores were comparable between the two groups; in the other words, the experimental group not only decreased decompression levels, surgical time, blood loss, and surgical transfusion but also achieved results of operations equal with that of the control group, thus obviously at an advantage. Although the postoperative VAS-BP, VAS-LP, and ODI scores in some cases had some fluctuation, the average was toward improvement and none of the patients’ symptoms recurred or exacerbated and required a repeat surgery, thus the effect of surgical treatment will stand for the test of time. All these also suggested that the use of PM and DTI to determine surgical levels will not miss the level which should be operated.
To our knowledge, this is the first study of the use of (DTI or PM) + MRI to look for decompression levels of patients with lumbar spinal stenosis. If patients who have no concordance between MRI and NE or decompression levels are longer (≥2) according to MRI + NE, in addition to the use of (DTI or PM), it can further determine and reduce decompression levels and avoid an extensive surgery, therefore reducing surgical trauma and hospitalization expenses etc. Ways to look for the responsibility level of surgery in patients with lumbar spinal stenosis already have provocative discography, discography, temporary external fixation, and facet joint blocks or zygapophyseal joint blocks. However, the disadvantages of these procedures are with invasion, low accuracy and complications [
52], and no using (DTI or PM).
Limitations
We acknowledge that our study has some limitations. One is that a small number of subjects were investigated and had limited follow-up. Further studies are needed to investigate whether our findings remain valid in a larger population and longer follow-up. Another, we could not repeat the DTI and PM after surgery because of spinal instrumentation artifacts, such as those from pedicle screw systems (affecting DTI) and surgical scar (affecting PM).
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
H-BC carried out the patient collection, participated in the design of the study, performed the statistical analysis, and drafted the manuscript. QW carried out the radiological examination. Q-FX carried out the paraspinal mapping examination. YC participated in the design of the study. BB conceived of the study and participated in its design. All authors read and approved the final manuscript.