Introduction: Clinical symptoms and degree of spinal stenosis
based on cross sectional dural sac area correlate only weakly in
lumbar spinal stenosis (LSS) patients. We conceived a four grade
classification system (A, B, C & D) based on the morphology of
the dural sac and its contents as seen on T2 axial MRI images.
The categories take into account the rootlet/CSF ratio. We
applied this grading to three patient groups: LSS scheduled for
surgery; LSS following conservative treatment and patients with
low back pain (LBP) without leg pain.
Materials/Methods: A total of 346 T2 axial MRI images taken
from LSS and LBP patients were included in this retroperspective
study. 37 patients had decompressive surgery (132
MRI images), 31 conservative treatment (116 MRI images) and
27 patients had unspecific LBP (98 MRI images). Dural sac
cross-sectional surface area and morphological grading of the
canal were measured digitally both at disc and pedicle level.
Intra- and inter-observer reliability were assessed (weighted
Cohen's kappa statistics) from 50 MRI images taken from the
surgery group.
Results: At the most severe disc level, grade A (mild stenosis)
was found in 3% of MRI images of the surgical group as opposed
to 51% in the conservatively treated group and 85% in the LBP
group. Grade B occurred in 8% of the surgical, 20% of the
conservative and was negligible in LBP group (below 1%).
Grade C and D (severe stenosis) was found in 89% of the
surgical group, as opposed to 30% in conservative group and
11% in LBP group. The grades of all groups were comparable at
the pedicle levels, exhibiting in 94% a grade A with a maximum
at the A1 grade. Pedicle and disc level cross-sectional area were
smallest in the surgery group and smaller in the conservative
group as compared to the LBP group at the levels L2, L3 and L4.
According to cross-sectional area measurements patients from the
surgery group seems to have smaller vertebral canal although this
was not related to smaller stature. Validation of grading: Average
intra-and inter observer kappas were 0.76 and 0.69 respectively,
for physicians working in the study originating institution.
Combining all observers the kappa values were 0.57 +/- 0.19. and
0.44 +/- 0.19 respectively. Dural sac cross-sectional area
measurements showed no statistically significant differences
between observers.
Conclusion: Since no specific measurement tools are needed the
grading suits everyday clinical practice and favours
communication of degree of stenosis between practising
physicians. In our institution Grade A stenosis was less likely to
require surgical treatment. This grading can therefore be an aid in
surgical patient selection in teaching units.