FROM:
European Spine Journal 2019 (Feb); 28 (2): 259–269 ~ FULL TEXT
Arnold Y. L. Wong, Eric C. Parent, Sukhvinder S. Dhillon, Narasimha Prasad, Dino Samartzis, Gregory N. Kawchuk
Department of Physical Therapy,
University of Alberta,
Edmonton, AB, Canada.
arnold.wong@polyu.edu.hk
PURPOSE: Our prior study revealed that people with non-specific low back pain (LBP) who self-reported a > 30% improvement in disability after SMT demonstrated significant post-treatment improvements in spinal stiffness, dynamic muscle thickness and disc diffusion, while those not having self-reported improvement did not have these objective changes. The mechanism underlying this differential post-SMT response remains unknown. This exploratory secondary analysis aimed to determine whether persons with non-specific LBP who respond to spinal SMT have unique lumbar magnetic resonance imaging (MRI) findings compared to SMT non-responders.
METHODS: Thirty-two participants with non-specific LBP received lumbar MRI before and after SMT on Day 1. Resulting images were assessed for facet degeneration, disc degeneration, Modic changes and apparent diffusion coefficient (ADC). SMT was provided again on Day 4 without imaging. SMT responders were classified as having a ≥ 30% reduction in their modified Oswestry disability index at Day 7. Baseline MRI findings between responders and non-responders were compared. The associations between SMT responder status and the presence/absence of post-SMT increases in ADC values of discs associated with painful/non-painful segments as determined by palpation were calculated. In this secondary analysis, a statistical trend was considered as a P value between 0.05 and 0.10.
RESULTS: Although there was no significant between-group difference in all spinal degenerative features (e.g. Modic changes), SMT responders tended to have a lower prevalence of severely degenerated facets (P = 0.05) and higher baseline ADC values at the L4–5 disc when compared to SMT non–responders (P = 0.09). Post hoc analyses revealed that 180 patients per group should have been recruited to find significant between–group differences in the two features. SMT responders were also characterized by significant increases in post–SMT ADC values at discs associated with painful segments identified by palpation (P < 0.01).
CONCLUSIONS: The current secondary analysis suggests that the spines of SMT responders appear to differ from non–responders with respect to degeneration changes in posterior joints and disc diffusion. Although this analysis was preliminary, it provides a new direction to investigate the mechanisms underlying SMT and the existence of discrete forms of treatment–specific LBP. These slides can be retrieved under Electronic Supplementary Material.
KEYWORDS: Apparent diffusion coefficient; Degeneration; Facet joint; Low back pain; Spinal manipulative therapy
TRIAL REGISTRATION: ClinicalTrials.gov identifier: NCT01761838
From the FULL TEXT Article:
Introduction
Low back pain (LBP) is the most common disabling health
condition in the world. [1] Despite its high prevalence,
approximately 90% of these cases are diagnosed as nonspecific
LBP due to unknown aetiology. [2] Accordingly,
patients with non–specific LBP are thought to be a heterogeneous
collection of various aetiologies [2] as demonstrated
by variable responses to any given intervention. [3–7]
Notably, spinal manipulative therapy (SMT) is a common
treatment for non–specific LBP that has shown a range
of therapeutic effects. [3–5] Our recent study confirmed
this variation in outcomes. Some participants demonstrated
clinically significant improvements in LBP–related disability
after 1 week of SMT treatments displayed post–SMT
decreases in spinal stiffness, increases in lumbar multifidus
contraction and increases in water diffusion at the L4–5 and
L5–S1 discs. [3] Conversely, no such changes were observed
in SMT non–responders, untreated non–specific LBP controls
or untreated asymptomatic controls. [3] Divergent
physical and clinical responses in patients with non–specific
LBP following an identical SMT application highlight the
importance of matching patient subgroups to the right LBP
intervention. [8] Importantly, these findings also offer a
unique opportunity to identify whether the differential treatment
responses arise from patients with distinct underlying
lumbar problems—a long–held premise with little empirical
evidence.
Since our prior results have shown that only SMT
responders demonstrated post–treatment improvement in
water diffusion in multiple lumbar discs [3], it may imply
that SMT non–responders may display other spinal degenerative
features (e.g. vertebral subchondral bone marrow
lesions) that differ from SMT responders. [9–13] Additionally,
because post–SMT improvement in disc diffusion
may benefit disc health and repair [13, 14], an exploration
of the association between post–SMT improvement in disc
diffusion at certain segments (especially discs associated
with painful segments as determined by palpation) and
corresponding clinical improvements would help determine
whether post–SMT enhancement of diffusion in
certain discs on magnetic resonance imaging (MRI) was
related to improved clinical outcomes.
Thus, our primary objective was to perform an exploratory
secondary analysis of our previous data to determine
whether SMT responders differed in their lumbar degeneration
status from SMT non–responders, which might
explain differential therapeutic responses of patients with
LBP after SMT. An additional objective was to determine
whether the presence/absence of increase in disc diffusion
of painful segments following the first application of SMT
was related to SMT responder/non–responder status after
1 week. It was hypothesized that SMT non–responders
displayed significantly more degenerative spinal features
than SMT responders. It was also hypothesized that an
improvement in disc diffusion of pain segments after the
first application of SMT was associated with favourable
clinical improvements after 1 week.
Materials and methods
Study design and participants
We performed an exploratory secondary analysis of a prospective
non–randomized clinical trial investigating post–
SMT differential responses of individuals with LBP [3]
(ClinicalTrials.gov identifier: NCT01761838). Participants
aged 18–60 years with non–specific LBP were recruited from
multidisciplinary settings. All enrolled participants had a
minimum LBP intensity of 2 on the 11–point numeric pain
rating scale and at least 20% on the modified Oswestry Disability
Index (mODI). [3] Exclusion criteria included but
were not limited to: prior lumbosacral surgery, prior SMT
within the past 4 weeks, scoliosis, pregnancy, spinal fractures/
tumours or any contraindications for MRI. [3]
To minimize the risk of unknowingly recruiting an
excessive number of SMT responders or non–responders,
we attempted to balance these groups by employing a clinical
prediction rule for identifying potential SMT responders
during enrolment. [15] Individuals with ≤ 2 items in the
clinical prediction rule were recruited as potential SMT
non–responders, while those with ≥ 4 characteristics were
recruited as potential responders (“Appendix 1”). [3, 4]
Individuals with exactly three characteristics were excluded
as per prior research. [3, 4] All participants gave written
informed consent for the study approved by the institutional
research ethics board.
Procedure
Figure 1
|
Thirty–two participants with LBP attended three morning
sessions in 1 week (Figure 1). On session 1, participants provided
demographics and completed the mODI. The lumbar
spine was palpated at the spinous processes to identify painful
and non–painful segment(s). [16] The cephalic and caudal
intervertebral discs (IVDs) associated with the painful
segment were noted. Additionally, participants underwent
lumbar MRI (L1–S1), spinal stiffness assessments and multifidus
ultrasonography before, and after, the application of
SMT in the first session. [3] The mean time between scans
was 1 h. On session 2 (3–4 days later), participants repeated
these procedures but without MRI. On session 3 (Day 7),
only the mODI score, spinal stiffness and multifidus function
were completed; no SMT was provided. Participants were
then dichotomized as SMT responders and non–responders
based on ≥ 30% and < 30% reduction in baseline mODI
scores. This cut–off is recognized as the minimal clinically
important improvement in LBP disability [17] although different
minimal clinically important improvement values have
been suggested by different studies. [18–20]
Spinal manipulative therapy
Participants received standardized SMT described elsewhere. [3] Briefly, it involved a long–lever application of a posteroinferior
thrust to the pelvis (“Appendix 2”). During sessions
1 and 2, a maximum of two SMT applications were allowed
for each side of the participant.
Imaging
Table 1
|
Participants' lumbar spines were scanned by a 1.5 Tesla MRI
unit (MAGNETOM Symphony, Siemens Medical Solutions)
using multi–element spine coils. The pre–SMT imaging protocol
included sagittal TI– and T2–weighted turbo spin echo,
axial T2–weighted turbo spin echo, sagittal STIR sequence
and sagittal diffusion–weighted imaging (DWI) sequences
(Table 1), while post–SMT imaging only involved DWI. [3]
Sagittal DWI were acquired using a single shot, dual spin
echo, echo–planar imaging acquisition with multi–element
spine coils and abdominal coils. The image processing system
of the scanner constructed mean apparent diffusion coefficient
(ADC) maps.
Grading of MRI findings
A certified radiologist, blinded to the ADC findings,
reviewed each participants' imaging and provided scoring
for: facet joint degeneration, IVD degeneration and Modic
changes (MCs) between the L3 and S1 segments. These segments
were chosen as LBP was commonly noted in the lower
lumbar spine. [21, 22]
The facet joint degeneration at L3–L5 levels was graded
by a validated 4–point scale having a high intra–observer reliability
(kappa 0.70–0.77) and clinical relevance. [23] Facet
degeneration was graded based on joint space, the presence
of osteophyte, articular process hypertrophy, subarticular
erosion and subchondral cyst. [23] A zero score denoted no
degeneration, while 3 represented severe degeneration. [24]
Disc degeneration at all lumbar levels was graded by
the 5–point Pfirrmann grading system that is based on disc
height, homogeneity of the disc structure, differentiation
between nucleus and annulus, and T2-weighted signal intensity. [25] This grading system has demonstrated excellent
intra-rater reliability (kappa from 0.84 to 0.90). [25] Higher
grades indicate more severe degeneration.
The MCs of vertebrae were graded by a method with substantial
to almost perfect intra-observer reliability (kappa
from 0.77 to 1.00) [26]. Specifically, type I MC (fibrovascular
replacement of subchondral bone) was characterized by
decreased T1 and increased T2 signal intensity. [27] Type 2
MC (endplate fissures and fatty replacement of subchondral
bone) showed hyperintense T1 and isointense/hyperintense
T2 signals. [27] Type 3 MC (sclerosis of subchondral bone)
appeared as hypointense T1 and T2 signals. [27] If multiple
types of MCs coexisted at the endplate/subchondral bone, it
was labelled as mixed type 1/2 or 2/3 MC. The locations and
sizes (endplate area, maximum height, volume or anteroposterior
diameter) of MCs were recorded. [28, 29]
ADC measurements
An examiner blinded to the radiologist's grading used a customized
programme to measure mean ADC values (ADC
mean) of five lumbar discs on the midsagittal ADC map.
Notably, a 40 mm2 circular region of interest (ROI) was
placed in the central, nuclear portion of a disc on the midsagittal
ADC map to minimize the noise signals from adjacent
vertebral structures. The ADC mean of a disc was estimated
from signal intensity within the ROI. If the diameter
of the ROI exceeded the disc height, the disc was excluded
from analysis. [3] A high ADC value represents higher IVD
diffusion. [13]
Intra-observer reliability of assessments
Three weeks after initial imaging, the radiologist and the
examiner independently repeated the grading of MRI findings
and ADC measurements on the scans of randomly
selected participants. Both readers were blinded to all measurement
results.
Statistical analysis
The exploratory, secondary analyses of underpowered data
were conducted using SPSS 22.0 software (IBM, Armonk,
NY). As some categories of the MRI findings showed low
prevalence of specific grades (< 10%) (e.g. type 3 MC), each
variable was dichotomized to calculate the intra-observer
reliability using kappa analysis. [30] Specifically, Pfirrmann
grading was dichotomized into “no/mild disc degeneration”
(grade 1–3) and “severe disc degeneration” (grade 4 or 5). [31] MC was dichotomized as “no” and “yes” (regardless
of types or locations). Facet joint degeneration was dichotomized
as “no/mild degeneration” (grade 0 and 1) and “severe
degeneration” (grade 2 and 3). [32] Kappa coefficients were
interpreted as follows: poor (≤ 0), slight (0.01–0.20), fair
(0.21–0.40), moderate (0.41–0.60), substantial (0.61–0.80)
and excellent (0.81–1.00). [33] The intra-observer reliability
of ADC measurements was reported elsewhere [3] and
demonstrated excellent intra-class correlation coefficients
(ICC3,1) estimates (0.97–0.98) (“Appendix 3”). [3, 34]
Chi-square/Fisher's exact tests assessed differences in the
frequency of dichotomized MRI findings between responders
and non-responders. Demographic data and baseline
ADC mean of all lumbar IVDs in responders and nonresponders
were checked for normality using Shapiro–Wilk
tests. If continuous data were not normally distributed,
Mann–Whitney U tests would replace independent t tests for
between-group comparisons. The effect sizes of all betweengroup
comparisons were calculated. [35] Additionally, to
determine the baseline ADC mean cut-off of each IVD that
could differentiate responders and non-responders [36], area
under the curve (AUC) of the receiver operating characteristic
curve and Youden index were used. [34, 36] An AUC
value of 0.5 indicates that the baseline ADC mean of a disc
is no better than chance to distinguish between responders
and non-responders, while an AUC of 1.00 represents the
baseline ADC mean that could always identify the responders. [34] The cut-off was chosen if Youden index value
(J = maxc [Se(c) + Sp (c) – 1]) was maximum, where Se and
Sp were the probability of truly identifying responders and
non-responders, respectively, at a given cut-off (c). [36]
The differences in post-SMT changes in ADC mean of
IVDs associated with painful/non-painful segments between
responders and non-responders were examined by separate
Chi-square tests. To dichotomize the post-SMT changes in
ADC mean into increase or no change/decrease in ADC
mean, values of the minimal detectable change at the 95%
confidence interval (“Appendix 4”) [37] derived from intraobserver
reliability of ADC measurements at the respective
disc levels were used as the cut-offs. Cramer's V tests were
used to estimate the strength of the association. Cramer's
V values were interpreted as weak (< 0.10), moderate
(0.11–0.33) and strong (> 0.30) [38]. The significance level
was set at 0.05 for all tests. For this exploratory secondary
analysis, a P value between 0.05 and 0.10 was considered as
a trend towards significance.
Results
Table 2
Figure 2
Figure 3
Figure 4
Table 3
Table 4
|
Based on mODI scores, 15 participants were classified
as SMT responders and 17 as non-responders. The baseline
demographics of these two groups were comparable
(Table 2). The intra-observer reliability of the various
lumbar degeneration grading was excellent (κ = 0.80–0.84)
(“Appendix 3”). [33]
Responders' and non-responders' MRI findings
Compared to SMT responders, SMT non-responders displayed
a trend of higher prevalence of severely degenerated
facet joints (χ2 = 3.86, P = 0.05, Cramer's V = 0.14) (Figure 2).
From the post hoc analysis using G * Power program [39],
180 participants per group should have been recruited to
detect a significant between-group difference in facet joint
degeneration (Power = 0.80, α = 0.05). There was no statistically
significant between-group difference in the prevalence
of IVD or MC grading. While most IVDs in both
subgroups were not degenerated, more degenerated discs
were found at the L4–5 and L5–S1 levels (Figure 3). Modic
changes were observed in 46.7% and 58.8% of responders'
and non-responders' endplates, respectively. For responders,
the prevalence estimates of types 1, 2 and 3 MCs in
endplates were 9.5%, 4.8% and 1.0%, respectively (Figure 4).
Similarly, type 1, type 2, type 3 and mixed type 1/2 MCs
were noted in 6.7%, 5.0%, 0.0% and 5.9% of non-responders'
endplates, respectively (Fig. 4). The anterior superior endplate
showed the highest relative frequency of MCs (33.3%
to 34.5%) in both subgroups although it did not significantly
differ from other locations. The majority of the observed
MCs involved only endplates (88.9% in responders, 68.9%
in non-responders).
Baseline ADC mean of responders and non-responders
Five out of 160 IVDs were excluded from the ADC measurements
because the disc space was smaller than the ROI diameter.
All the excluded discs were from five non-responders at
L1–2 (n = 2) and L4–5 (n = 1) and L5–S1 discs (n = 2). There
was no significant difference in ADC mean of all intervertebral
discs between responders and non-responders. However,
compared to responders, non-responders had a trend of lower
estimated baseline ADC mean of the L4–5 disc (U = 78.00,
P = 0.09, r = 0.30) (Table 3). The post hoc analysis revealed
that a sample size of 43 per group should have been recruited
to detect a significant between-group difference of this magnitude
(Power = 0.80, α = 0.05). Similarly, while AUCs for
baseline ADC mean of discs at each level were not statistically
significant, the AUC of the L4–5 disc was 0.70 (95%
CI: 0.50–0.89; P = 0.07). Specifically, patients with baseline
ADC mean higher than 2.07 × 10–3 mm2/s at the L4–5
disc were more likely to be responders (sensitivity = 0.71;
specificity = 0.68).
Comparing post-SMT ADC mean changes in discs at painful/non-painful segments
between responders and non-responders
Compared to non-responders, a significantly higher proportion
of discs identified as painful in responders demonstrated
immediate post-SMT increases in ADC mean (Cramer's
V = 0.50, P < 0.01) (Table 4). Conversely, both responders
and non-responders did not show significant differences in
the patterns of post-SMT changes in ADC mean of discs
associated with non-painful segments (P > 0.01) (Table 4).
Discussion
While the current secondary analysis has revealed no statistically
significant differences in spinal structural features
between SMT responders and non-responders, lumbar
degeneration status appears to be related to post-SMT treatment
outcomes. Compared to SMT responders, the nonresponders
demonstrated a trend of a higher prevalence of
severe facet degeneration (P = 0.05) and relatively lower
baseline ADC mean of the L4–5 disc (P = 0.09). Given the
exploratory nature of our analyses, it is important to consider
all potential findings that may explain mechanisms underlying
the differential responses between SMT responders
and non-responders. Our results suggest that spinal degeneration
(especially the discs) may be a potential treatment
effect modifier for SMT in patients with non-specific LBP
although further studies are warranted to test this hypothesis.
Distinct baseline ADC mean and post-SMT disc diffusion
responses were noted between responders and nonresponders.
Our current data suggest that ADC mean of
IVDs may be more sensitive to the microstructural intradiscal
changes that may not be quantified by qualitative
grading on MRI. [40] Since a lower ADC mean indicates
greater disc degeneration [13, 41], our findings suggest that
non-responders may have more degenerated L4–5 discs,
which may be sources of LBP. [42] A prior ovine study has
found that a spinal segment with a degenerated disc displays
less movement during SMT as compared to a nondegenerated
spine. [43] Such distinct mechanical responses
may explain the differential SMT responses. Additionally,
Vieira-Pellenz found that male LBP patients with degenerative
discs who demonstrated greater increases in body height
(a proxy of improved disc height) after a single session of
SMT were associated with less LBP prevalence. [44] Given
these results, further studies with larger sample sizes are
warranted to determine the causal relations among spine
degeneration, post-SMT mechanical responses and LBP.
Interestingly, our findings highlighted that responders
were characterized by immediate post-SMT increases in
ADC mean of discs associated with painful segments on
session 1. This finding implies that favourable clinical outcomes
arising from SMT may be partly due to improved
diffusion of discs associated with painful segments. Since
the significant relation between level-specific post-SMT
increases in ADC mean of discs and improved clinical outcomes
was only found in SMT responders, it corroborates
the importance of conducting further studies to clarify the
association between spinal degeneration and SMT outcomes,
especially when SMT non-responders displayed a the trend
of more lumbar degenerative features. Notwithstanding the
strong association between post-SMT increase in ADC mean
of discs on session 1 and clinical improvements on session
3, improved disc diffusion might not be the sole reason for
the differential clinical responses of our participants. Since
26.8% of discs at the painful segments of the responders
showed no change/decrease in the post-SMT ADC mean
(Table 4), their clinical improvements may be related to the
mechanical and/or neurophysiological effects of SMT on
spinal tissues other than improved disc diffusion. Similarly,
if IVDs were not the sole source of LBP or the dosage of
SMT was insufficient to elicit significant clinical improvements
in the non-responders, enhanced post-SMT disc diffusion
would not improve clinical outcomes. Furthermore,
SMT is unlikely to benefit non-biomechanical pathologies,
e.g. inflammation of vertebral subchondral bone [45] or
increased nociceptive innervation in IVDs/vertebral endplates. [46]
Additionally, our findings underscore the presence of
treatment-specific types of LBP. As non-specific LBP is a
constellation of heterogeneous conditions that have varied
mechanisms of pain generation [47, 48], the application of
a single LBP treatment to patients with non-specific LBP
would be expected to result in diverse clinical outcomes as
is the case in our work and that of others. [48, 49] Although
there have been many suggestions as to why this diverse
range of responses exist to a given treatment [50–53], this
secondary analysis suggests that a selective response to a
particular intervention does exist and has a potential physiological
explanation consisting of distinct biological and
biomechanical characteristics. Our findings indicate a novel
research direction for investigating the influence of structural
features in affecting treatment responses, which may guide
certain clinical decision in future. However, it is not recommended
to order routine MRI scans for the sole purpose of
determining the application of SMT.
The current study has inherent limitations. First, as the
current exploratory, secondary analysis was not powered for
the comparison of MRI findings between responders and
non-responders. As such, no statistically significant difference
can be drawn, and our results should be interpreted
with caution. Future research with a larger sample size is
warranted to validate the current findings. Second, although
SMT responders and non-responders were dichotomized
based on a 30% decrease in baseline mODI scores at the
third visit, the absolute change in mODI score of a SMT
responder could be as small as 6 points if the participant's
baseline mODI score was only 20 points. While 6 points
were still larger than some suggested minimal detectable
changes in patients with LBP (ranging from 3.2 to 5.2
points) [18], this value was much smaller than other suggested
minimal detectable changes (15–19 points). [19]
Therefore, some SMT responders' mODI improvements
might not demonstrate minimal clinical important differences
had other cut-off methods been used. Third, since the
present study investigated the association between post-SMT
changes in ADC mean of IVDs on session 1 and the clinical
outcomes on session 3, the relation between serial changes
in ADC mean of IVDs and the respective clinical outcomes
remains unknown.
As prior research suggests that SMT may
have a positive dose–response effect on clinical outcomes [54], it is conceivable that SMT may have similar cumulative
effects on the disc diffusion changes. Future research
should measure both the post-SMT ADC mean and clinical
outcomes at successive time points to clarify this relation.
Fourth, while our results appeared to suggest distinct MRI
spinal findings in SMT responders and non-responders, MRI
scans are rarely ordered in clinical practice for patients with
non-specific LBP. Future studies should investigate whether
differential spinal degenerative features exist in plain radiographs
(i.e. X-rays) between SMT responders and nonresponders,
which may inform clinical decision-making and
further assist in more refined patient selection for further
imaging assessments.
Conclusions
There was no significant difference in structural features between SMT responders and non-responders. However, SMT responders demonstrate a trend of a lower prevalence of severely degenerated facets and relatively high baseline ADC values of the L4–5 discs. Interestingly, SMT responders demonstrate post-SMT increases in apparent diffusion coefficient (ADC) values of discs associated with painful segments. These results suggest that SMT responses may be related to some underlying structural responses. Our findings provide a new hypothesis/direction for further investigating the underlying nature of the differential response to SMT, possible mechanisms of SMT and the existence of treatment-specific forms of LBP.
Appendix 1: Clinical prediction rule
Clinical characteristics of the clinical prediction rule for identifying people
who benefit from spinal manipulative therapy
Clinical characteristics Definition of a positive finding
1. Duration of the current Less than 16 days
episode of low back pain
2. Distal symptoms No symptoms distal to the knee
3. Fear avoidance beliefs Less than 19 points
questionnaire work subscale
4. Lumbar stiffness At least one lumbar segment is
determined to be hypomobile
by the examiner using a manual
posteroanterior spinal mobility
test
5. Hip internal rotation At least one hip with 35° or
range of motion greater as measured by an
inclinometer in prone
Appendix 2: Spinal manipulative therapy procedure
The supine participant crossed and put his/her fingers behind
the neck. The clinician stood opposite to the side to be
manipulated and side bent the participant's trunk towards the
side of the pelvis to be manipulated, and rotated the trunk
in the opposite direction. Then the clinician applied a highvelocity,
low-amplitude thrust to the pelvis in a posteroinferior
direction. The clinician delivered spinal manipulative
therapy to each side at each given session. If the first attempt
did not result in cavitation, a second spinal manipulation was
allowed for each side. A maximum of two spinal manipulations
would be given to each side within a session. In the
current study, only two out of 64 sessions required a second
spinal manipulation on the one side (one for responder and
one for non-responder).
Appendix 3: Intra-observer reliability of the dichotomized degeneration variables
Appendix 4: Intra-observer reliability of apparent diffusion coefficient measurement
References:
Global Burden of Disease 2013 Collaborators (2013)
Global, Regional, and National Incidence, Prevalence, and Years Lived with Disability for 301 Acute
and Chronic Diseases and Injuries in 188 Countries, 1990-2013: A Systematic Analysis
for the Global Burden of Disease Study 2013
Lancet. 2015 (Aug 22); 386 (9995): 743–800
Haldeman S, Kopansky-Giles D, Hurwitz EL, et al.
Advancements in the Management of Spine Disorders
Best Pract Res Clin Rheumatol. 2012 (Apr); 26 (2): 263–280
Wong AYL, Parent EC, Dhillon SS, Prasad N, Kawchuk GN.
Do Participants with Low Back Pain who Respond to Spinal
Manipulative Therapy Differ Biomechanically From
Nonresponders, Untreated Controls
or Asymptomatic Controls?
Spine (Phila Pa 1976). 2015 (Sep 1); 40 (17): 1329–1337
Fritz JM, Koppenhaver SL, Kawchuk GN, Teyhen DS, Hebert JJ, Childs JD.
Preliminary Investigation of the Mechanisms Underlying the Effects of Manipulation:
Exploration of a Multivariate Model Including Spinal Stiffness,
Multifidus Recruitment, and Clinical Findings
Spine (Phila Pa 1976). 2011 (Oct 1); 36 (21): 1772-1781
Koppenhaver SL, Fritz JM, Hebert JJ et al (2011)
Association between changes in abdominal and lumbar multifidus muscle thickness and clinical improvement
after spinal manipulation.
J Orthop Sports Phys Ther 41:389–399.
https://doi.org/10.2519/jospt .2011.3632
Long A, Donelson R, Fung T (2004)
Does it matter which exercise? A randomized control trial of exercise for low back pain.
Spine 29:2593–2602
Beneciuk JM, Robinson ME, George SZ (2015)
Subgrouping for patients with low back pain: a multidimensional approach incorporating cluster analysis
and the STarT Back Screening Tool.
J Pain 16:19–30.
https://doi.org/10.1016/j.jpain .2014.10.004
Hebert JJ, Koppenhaver SL, Walker BF (2011)
Subgrouping patients with low back pain: a treatment-based approach to classification.
Sports Health 3:534–542.
https://doi.org/10.1177/1941738111 41504 4
Jensen TS, Karppinen J, Sorensen JS et al (2008)
Vertebral endplate signal changes (Modic change): a systematic literature review of prevalence
and association with non-specific low back pain.
Eur Spine J 17:1407–1422.
https://doi.org/10.1007/s00586-008-0770-2
Kjaer P, Korsholm L, Bendix T et al (2006)
Modic changes and their associations with clinical findings.
Eur Spine J 15:1312–1319.
https://doi.org/10.1007/s0058 6-006-0185-x
Jensen RK, Leboeuf-Yde C, Wedderkopp N et al (2012)
Is the development of Modic changes associated with clinical symptoms?
A 14-month cohort study with MRI.
Eur Spine J 21:2271–2279.
https://doi.org/10.1007/s0058 6-012-2309-9
Jensen OK, Nielsen CV, Sørensen JS, Stengaard-Pedersen K (2014)
Type 1 Modic changes was a significant risk factor for 1-year outcome in sick-listed low back pain patients:
a nested cohort study using magnetic resonance imaging of the lumbar spine.
Spine J 14:2568–2581.
https://doi.org/10.1016/j.spinee.2014.02.018
Antoniou J, Demers CN, Beaudoin G et al (2004)
Apparent diffusion coefficient of intervertebral discs related to matrix composition and integrity.
Magn Reson Imaging 22:963–972.
https://doi.org/10.1016/j.mri.2004.02.011
Huang Y-C, Urban JPG, Luk KDK (2014)
Intervertebral disc regeneration: do nutrients lead the way?
Nat Rev Rheumatol 10:561–566.
https://doi.org/10.1038/nrrhe um.2014.91
Flynn T, Fritz J, Whitman J et al (2002)
A clinical prediction rule for classifying patients with low back pain who demonstrate shortterm improvement
with spinal manipulation.
Spine 27:2835–2843.
https://doi.org/10.1097/01.BRS.00000 35681 .33747 .8D
Haneline MT, Young M (2009)
A review of intraexaminer and interexaminer reliability of static spinal palpation: a literature synthesis.
J Manip Physiol Ther 32:379–386.
https://doi.org/10.1016/j.jmpt.2009.04.010
Ostelo RWJG, Deyo RA, Stratford P et al (2008)
Interpreting change scores for pain and functional status in low back pain: towards international
consensus regarding minimal important change.
Spine 33:90–94.
https://doi.org/10.1097/BRS.0b013e3181 5e3a1 0
Suarez-Almazor ME, Kendall C, Johnson JA, Skeith K, Vincent D (2000)
Use of health status measures in patients with low back pain in clinical settings.
Comparison of specific, generic and preference-based instruments.
Rheumatology 39:783–790.
https://doi.org/10.1093/rheum atolo gy/39.7.783
Davidson M, Keating JL (2002)
A Comparison of Five Low Back Disability Questionnaires: Reliability and Responsiveness
Physical Therapy 2002 (Jan); 82 (1): 8–24
Hagg O, Fritzell P, Nordwall A (2003)
The clinical importance of changes in outcome scores after treatment
for chronic low back pain.
Eur Spine J 12:12–20
DePalma MJ, Ketchum JM, Saullo T (2011)
What is the source of chronic low back pain and does age play a role?
Pain Med 12:224–233
Kjaer P, Leboeuf-Yde C, Korsholm L et al (2005)
Magnetic resonance imaging and low back pain in adults: a diagnostic imaging
study of 40-year-old men and women.
Spine 30:1173–1180
Weishaupt D, Zanetti M, Boos N, Hodler J (1999)
MR imaging and CT in osteoarthritis of the lumbar facet joints.
Skelet Radiol 28: 215–219
Kettler A, Wilke HJ (2006)
Review of existing grading systems for cervical or lumbar disc and facet joint degeneration.
Eur Spine J 15:705–718
Pfirrmann C, Metzdorf A, Zanetti M et al (2001)
Magnetic resonance classification of lumbar intervertebral disc degeneration.
Spine 26:1873–1878
Jensen TS, Sørensen JS, Kjaer P (2007)
Intra- and interobserver reproducibility of vertebral endplate signal (modic) changes
in the lumbar spine: the Nordic Modic Consensus Group classification.
Acta Radiol 48:748–754
Modic MT, Steinberg PM, Ross JS et al (1988)
Degenerative disk disease: assessment of changes in vertebral body marrow with MR imaging.
J Bone Joint Surg Am 166:193–199
Maataoui A, Vogl TJ, Khan MF (2015)
Magnetic resonance imaging-based interpretation of degenerative changes in the
lower lumbar segments and therapeutic consequences.
World J Radiol 7:194–197
Määttä JH, Karppinen JI, Luk KDK et al (2015) Phenotype
profiling of Modic changes of the lumbar spine and its association
with other MRI phenotypes: a large-scale population-based
study. Spine J 15:1933–1942. https://doi.org/10.1016/j.spine
e.2015.06.056
Gjorup T (1988) The kappa coefficient and the prevalence of a
diagnosis. Methods Inf Med 27:184–186
Kovacs FM, Arana E, Royuela A et al (2014) Disc degeneration
and chronic low back pain: an association which becomes nonsignificant
when endplate changes and disc contour are taken into
account. Neuroradiology 56:25–33. https://doi.org/10.1007/s0023
4-013-1294-y
Kalichman L, Hodges P, Li L et al (2010) Changes in paraspinal
muscles and their association with low back pain and spinal
degeneration: CT study. Eur Spine J 19:1136–1144. https://doi.
org/10.1007/s0058 6-009-1257-5
Landis JR, Koch GG (1977) The measurement of observer agreement
for categorical data. Biometrics 33:159–174. https://doi.
org/10.2307/25293 10
Portney LG, Watkins MP (2009) Foundations of clinical research
applications to practice, 3rd edn. Pearson Education Inc, New
Jersey
Fritz CO, Morris PE, Richler JJ (2012) Effect size estimates:
current use, calculations, and interpretation. J Exp Psychol Gen
141:2–18. https://doi.org/10.1037/a0024 338
Fluss R, Faraggi D, Reiser B (2005) Estimation of the Youden
index and its associated cutoff point. Biom J 47:458–472. https://
doi.org/10.1002/bimj.20041 0135
Eliasziw M, Young SL, Woodbury MG, Fryday-Field K (1994)
Statistical methodology for the concurrent assessment of interrater
and intrarater reliability: using goniometric measurements as an
example. Phys Ther 74:777–788
Healey JF (2007) The essentials of statistics: a tool for
social research. Nelson Education, Toronto. https://doi.
org/10.1179/10773 52018 00339 434
Erdfelder E, Faul F, Buchner A (1996) GPower: a general
power analysis program. Behav Res Methods Instrum Comput
1996(28):1–11. https://doi.org/10.3758/BF032 03630
Teraguchi M, Yoshimura N, Hashizume H et al (2014) Prevalence
and distribution of intervertebral disc degeneration over the
entire spine in a population-based cohort: the Wakayama Spine
Study. Osteoarthr Cartil 22:104–110. https://doi.org/10.1016/j.
joca.2013.10.019
Yu HJ, Bahri S, Gardner V, Muftuler LT (2015) In vivo quantification
of lumbar disc degeneration: assessment of ADC value using
a degenerative scoring system based on Pfirrmann framework.
Eur Spine J 24(11):2442–2448. https://doi.org/10.1007/s0058
6-014-3721-0
Weiler C, Lopez-Ramos M, Mayer HM et al (2011) Histological
analysis of surgical lumbar intervertebral disc tissue provides
evidence for an association between disc degeneration and
increased body mass index. BMC Res Notes 4(1):497. https://doi.
org/10.1186/1756-0500-4-497
Colloca CJ, Gunzburg R, Freeman BJ et al (2012) Biomechancial
quantification of pathologic manipulable spinal lesions: an in vivo
ovine model of spondylolysis and intervertebral disc degeneration.
J Manip Physiol Ther 35:354–366. https://doi.org/10.1016/j.
jmpt.2012.04.018
Vietra-Peltenz F, Olivia-Pascual-Vaca A, Rodriguez-Blanco C
et al (2014) Short-term effect of spinal manipulation on pain perception,
spinal mobility, and full height recovery in male subjects
with degenerative disk disease: a randomized controlled trial.
Arch Phys Med Rehabil 95:1613–1619. https://doi.org/10.1016/j.
apmr.2014.05.002
Albert HB, Lambert P, Rollason J et al (2013) Does nuclear tissue
infected with bacteria following disc herniations lead to Modic
changes in the adjacent vertebrae? Eur Spine J 22:690–696. https
://doi.org/10.1007/s0058 6-013-2674-z
Samartzis D, Liebenberg EC, Fong DY (2014) Innervation of
pathologies in the lumbar vertebral end plate and intervertebral
disc. Spine J 14:513–521. https://doi.org/10.1016/j.spine
e.2013.06.075
Dankaerts W, O'Sullivan PB, Straker LM et al (2006) The interexaminer
reliability of a classification method for non-specific
chronic low back pain patients with motor control impairment.
Man Ther 11:28–39. https://doi.org/10.1016/j.math.2005.02.001
Delitto A (2005) Research in low back pain: time to stop seeking
the elusive “magic bullet”. Phys Ther 85:206–208
McCarthy CJ, Arnall FA, Strimpakos N et al (2013) The biopsychosocial
classification of non-specific low back pain: a systematic
review. Phys Ther Rev 9:17–30. https://doi.org/10.1179/10833
19042 25003 955
O'Sullivan P (2005) Diagnosis and classification of chronic low
back pain disorders: maladaptive movement and motor control
impairments as underlying mechanism. Man Ther 10:242–255.
https://doi.org/10.1016/j.math.2005.07.001
McKenzie RA, May S (2003) Mechanical diagnosis and therapy:
the lumbar spine, 2nd edn. Spinal Publications, Waikanae. https
://doi.org/10.2519/jospt .2004.34.3.105
Waddell G, Newton M, Henderson I et al (1993) A Fear-Avoidance
Beliefs Questionnaire (FABQ) and the role of fear-avoidance
beliefs in chronic low back pain and disability. Pain 52:157–168
Mondloch M (2001) Does how you do depend on how you think
you'll do? A systematic review of the evidence for a relation
between patients” recovery expectations and health outcomes.
CMAJ 165:174–179
Bialosky JE, Bishop MD, Price DD et al (2009) The mechanisms
of manual therapy in the treatment of musculoskeletal
pain: a comprehensive model. Man Ther 14:531–538. https://doi.
org/10.1016/j.math.2008.09.001
Return to LOW BACK PAIN
Return to DISC HERNIATION
Since 2-12-2019
|