Key-words: Laminectomy - low back pain - modic changes - outcome - posterior spinal fusion
Introduction
Modic changes (MC) are visible on magnetic resonance imaging (MRI) as lesions of the
vertebral endplate and alterations in signal intensity of the bone marrow adjacent
to a degenerated disc. There are three types of MCs, namely signs of bone marrow edema
(Modic Type I), fat (Modic Type II), and osseous sclerosis (Modic Type III).[[1 ]],[[2 ]]
Numerous studies have shown an association between MCs and chronic low back pain (LBP)
and MC Type I had a stronger association than other MC types.[[3 ]],[[4 ]],[[5 ]],[[6 ]],[[7 ]],[[8 ]] Chronic LBP is one of the leading causes of disability in adults.[[9 ]] The most frequent problem in LBP is degenerative disc disorder (DDD) which is also
associated with MCs, especially Type I and II.[[5 ]],[[6 ]],[[10 ]]
Patients with chronic LBP with DDD and severe symptoms undergo different surgeries
mainly spinal fusion. Studies have indicated that MC existence affects the outcome
of surgery and levels of pain reduction.[[11 ]],[[12 ]],[[13 ]],[[14 ]] Some studies have also reported satisfactory and high successful results following
lumbar disc replacement or fusion in patients with MC especially type I MC.[[15 ]],[[16 ]]
The studies in this regard are few, and still, we cannot make certain conclusions
on the effects of the MC on surgery of chronic LBP outcome; also, there are few data
available regarding the differences between different types of MC and surgery outcome.
In this study, we aim to evaluate the effect of preoperative MC on the outcome of
patients with LBP following posterior spinal fusion (PSF) or laminectomy.
Materials and Methods
Between January 2016 and February 2017, 162 patients with chronic back pain and disc
extrusion or lumbar canal stenosis or spondylolisthesis candidate of surgery with
MC Type 1 and II in MRI visiting neurosurgery clinics of Shohada-e-Tajrish Hospital,
Tehran, Iran, were recruited for this study. All patients had MC Type I and II in
MRI studies.
Inclusion criteria were patients between 20 and 70-year-old with axial pain, disc
extrusion, lumbar canal stenosis, or spondylolisthesis not responding to medical therapy
and MC Type 1 and II in MRI undergoing laminectomy or laminectomy with PSF. Patients
with a history of trauma to spine, previous lumbar spine surgery, diabetes mellitus
or other diseases causing neuropathy, rheumatologic, infectious, inflammatory or collagen
vascular diseases, cardiovascular, pulmonary, and hepatic comorbidities that could
affect the decision on the surgery type, global or regional deformity in spine including
scoliosis, kyphosis, sagittal, or coronal imbalance were excluded. The study was approved
by Ethics Committee of Shahid Beheshti University of Medical Sciences, and all patients
gave written informed consent.
Patients' demographic findings, medical history, and duration of the disease were
recorded. Pain intensity, before and 3 months after surgery, was evaluated using visual
analog scale (VAS).
In our center, laminectomy and spinal cord decompression without PSF is performed
for (1) disc extrusion with compression on spinal cord causing back pain or radicular
pain, resistant to medical therapy or with neurologic deficits; (2) central or lateral
lumbar canal stenosis, but no need for resection of more than half of the medial facet
to decompress; (3) with no regional deformity or instability in spinal cord. Laminectomy
and spinal cord decompression with PSF is the choice for cases (1) in need of discectomy
in more than two adjacent levels because of the possible subsequent instability and
increasing the back pain; (2) symptomatic spondylolisthesis in lumbar vertebrae with
resistant medical therapy; (3) multilevel lateral recess stenosis in need of medial
facetectomy; and (4) instability in need of surgical correction diagnosed in dynamic
images before surgery. The standard operating procedure technique was followed in
each surgery, and all surgeries were performed by a single neurosurgeon.
All patients underwent MRI of spinae before surgery. MCs are bone marrow and endplate
lesions visible on MRI. We classified MC using the original classification by Modic
et al.:[[17 ]]
MC Type I: hypointense on T1WI and hyperintense on T2WIMC Type II: hyperintense on
T1WI and isointense or hyperintense on T2WIMC MC Type III: hypointense on T1WI and
hypointense on T2WI.
Statistical analysis
All data were analyzed using SPSS22 software (IBM SPSS Statistics Version 22, International
Business Machines Corp., Armonk, NY, USA). Results are expressed as mean ± standard
deviation or percentage. All normally distributed continuous data were analyzed using
unpaired t-tests and expressed as the means and standard deviations. P < 0.05 was
considered statistically significant.
Results
We studied 162 patients with chronic back pain and MC Type I and II before and after
surgery. Patients underwent laminectomy (n = 72) or PSF (n = 90) surgeries. Patients
had MC Type I in 75 (46.3%) (29 in laminectomy and 46 cases in PSF groups) and Type
II in 87 (53.7%) (43 in laminectomy and 44 in PSF group). Pain score using VAS was
significantly decreased following surgery (7.93 ± 1.27–5.98 ± 1.57, P < 0.001).
Patients' baseline findings with changes in pain severity are shown in [[Table 1 ]]. The only significant difference we observed was lower VAS 1 month after surgery
in PSF group.
Table 1: Patient’s baseline findings with changes in pain severity after surgery
There was no difference in pain VAS score between MC Type I and II before (7.86 ±
1.32 vs. 8.00 ± 1.23, P = 0.51) and after treatment (5.89 ± 1.72 vs. 6.05 ± 1.44,
P = 0.51). The mean VAS percentage of change also had no significant difference between
MC Type I and II (−23.01 ± 26.63% vs.−22.54 ± 22.42%, P = 0.90).
We intended to compare the preoperative MC role in surgery outcome, so patients were
evaluated separately in groups of Type I and Type II changes [[Table 2 ]]. Among MC Type I, there was a significantly more changes in pain score in PSF compared
to laminectomy, but the changes in modic Type II were similar between groups.
Table 2: Pain score using visual analog scales before and after surgery between different
types of modic changes
Discussion
Previous studies have reported the associations between MC with disc degeneration,
DD severity, and disc herniations.[[5 ]],[[18 ]],[[19 ]] MC is related to discogenic LBP[[20 ]] and they would be representative of an underlying pathology that should be a target
for therapy.[[21 ]] These patients also have higher pain intensity than LBP patient without MC, which
could also affect the surgical treatment outcome.
In this study, we evaluated the outcome of PSF and laminectomy in 162 patients with
MC I and II and observed significant improvement in pain intensity following surgery.
Previous studies have also reported significant improvement in pain score following
surgical treatment.[[22 ]],[[23 ]]
It is possible the surgical treatment improves patients' status regardless of MC.
Chin et al.[[24 ]] reported no significant difference between patients with and without MC in microdiscectomy
outcome, but found a higher tendency for better improvement in those without MC. Sørlie
et al.[[14 ]] observed that patients with MC type I would have lower improvement compared to
other types of MC or those without MC, although the improvement is significant. Thus,
Lurie et al.[[25 ]] indicated that MC Type 1 may be predictors of surgical treatment.
We compared the pain VAS before and after treatment between MC I and II and found
no significant difference between groups. Similar to our findings, Ghodsi et al.[[22 ]] in their study on the efficacy of posterolateral fusion in patients with unstable
lumbar spine with MC also observed no significant difference in surgical outcome between
different types of MC. In another study, Yu et al.[[23 ]] found no difference in surgical outcome among patients with different modic types
in patients with disc degeneration and MC.
Although Kwon et al.[[15 ]] reported significant improvement in pain VAS after treatment in both MC Type I
and Type II, in their study, two types were not compared. Esposito et al.[[26 ]] also evaluated the effect of lumbar fusion in patients with chronic discogenic
LBP. They observed significant improvement in pain in MC Type I, but there was no
significant improvement in patients with MC Type II. However, they did not compare
the difference in improvement between groups.
It is reported that MC Type I changes make patients more prone to spinal segmental
instability than MC Type II.[[27 ]] Eser et al.[[28 ]] have also concluded that MC Type I is indicative of an ongoing active degenerative
and inflammatory process, while MC Type II is representative of more stable and chronic
process. They suggested that posterior dynamic stabilization could be an effective
treatment in MC Type I. Similarly Vital et al.[[29 ]] reported conversion in MC Type I to Type II or normal following posterolateral
fusion.
We also observed that among patients with Type I changes, PSF had significantly better
improvement in pain score compared to laminectomy, while among Type I patients, the
improvement was comparable. This is indicative of the better efficacy of PSF in MC
Type I.
Conclusion
Surgical treatment in patients with LBP with MC accompanies with significant improvement
in pain. PSF seems better treatment in patients with MC Type I.
