J Neurol DOI 10.1007/s00415-015-7726-3
ORIGINAL COMMUNICATION
MRI of lumbar trunk muscles in patients with Parkinson’s disease and camptocormia N. G. Margraf1 • A. Rohr2 • O. Granert1 • J. Hampel1 A. Drews2 • G. Deuschl1
•
Received: 2 March 2015 / Revised: 25 March 2015 / Accepted: 26 March 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract Camptocormia in Parkinson’s disease (PD) is an axial postural disorder usually accompanied by histopathological changes in the paravertebral muscles of unknown etiology. The diagnostic potential of magnetic resonance imaging (MRI) of back muscles in camptocormia has not been systematically assessed. Our objective was to characterize pathological muscle changes with MRI and to develop radiological criteria for camptocormia. The criteria edema, swelling and fatty degeneration in 20 idiopathic PD patients with camptocormia were assessed using MRI (T1w and short tau inversion recovery (STIR) sequences) of the lumbar trunk muscles and compared with 20 group-matched PD patients without camptocormia. Edema and fatty degeneration of the paravertebral muscles were significantly more frequent in camptocormia. Edema correlated negatively and fatty degeneration positively with the duration of camptocormia and not PD. Swelling of the paravertebral muscles, edema and swelling of the quadratus lumborum muscle and rare edema of the psoas muscle were only found in camptocormia patients. In this case–control study the defined MRI criteria distinguish the group of PD patients with camptocormia versus those without. Our findings suggest dynamic changes in the MRI signals over N. G. Margraf and A. Rohr contributed equally to the manuscript. & N. G. Margraf [email protected] & A. Rohr 1
Department of Neurology, University Hospital SchleswigHolstein, Campus Kiel, Schittenhelmstr. 10, 24105 Kiel, Germany
2
Institute of Neuroradiology, University Hospital SchleswigHolstein, Campus Kiel, Schittenhelmstr. 10, 24105 Kiel, Germany
time in the paravertebral muscles: edema and swelling are found initially, followed by fatty atrophic degeneration 2–3 years after the beginning of camptocormia. Muscle MRI qualifies as a tool for categorizing phases of camptocormia as acute or chronic, with potential consequences for therapeutic approaches. The involvement of muscles beyond an isolated impairment of the paravertebral muscles implies a more systemic view with a deregulation of lumbar trunk muscles. Keywords Idiopathic Parkinson’s disease Camptocormia Paravertebral muscles Bent spine syndrome Muscle MRI
Introduction Camptocormia is an involuntary, marked flexion of the thoracolumbar spine, which is apparent during standing or walking and disappears in the supine position. This axial postural disorder has been identified in a broad spectrum of neurological diseases [1–3] but seems to be most common in movement disorders and particularly Parkinson’s disease (PD). The prevalence of camptocormia in PD ranges widely between 3 and 17.6 % [4–7]. On investigating myopathological changes in the paravertebral muscles of camptocormia patients suffering from PD, remarkable similarities to the lesion pattern of muscles with tendon tears were observed [8]. Although CT and MR imaging criteria have been developed in the orthopaedic diagnostics of muscles following tension tears [9–11], there is limited experience with the use of MRI in detecting corresponding changes in the back muscles in camptocormia [4, 12, 13]. Our objective was to examine the diagnostic capability of MRI of lumbar trunk muscles
123
J Neurol
in PD patients with camptocormia, and to develop radiological criteria suitable to characterize pathological changes in these muscles.
(inversion time) 2500/70/170 ms, TSE factor 15. Axial T1w and STIR images were acquired in two stacks, covering the whole thoracic and lumbar spine as the area of the clinically obvious changes.
Methods
Evaluation of the MRI of the lumbar trunk muscles
Patients and controls
We analyzed MRI changes in three lumbar trunk muscles: the paravertebral muscles, the quadratus lumborum and the psoas muscles. We defined the criteria ‘‘edema’’, ‘‘swelling’’ and ‘‘fatty degeneration’’ as described below. Relevant pathological changes of the spine were recorded.
Patients with idiopathic PD (according to the British Parkinson’s Disease Society Brain Bank criteria [14]) and camptocormia were recruited at the Department of Neurology, Kiel University between 2010 and 2013. Camptocormia (with or without laterodeviation) was defined as an at least 30° thoracolumbar flexion apparent when standing or walking and resolving in the recumbent position [15]. Laterodeviation was defined as visually obvious deflection from the normally horizontal line between the two shoulders. Control patients with idiopathic PD without camptocormia were recruited and group matched for sex, age, subtype, duration and severity of PD (Hoehn & Yahr stage, UPDRS III motor score, L-dopa equivalent dose). Exclusion criteria for both groups were: contra-indications for MRI examination, clinical or radiological signs suggesting an atypical Parkinsonian syndrome or dystonia. Informed consent was obtained from all participants. The ethics committee of the Medical Faculty of the Christian-Albrechts University of Kiel approved the study and it was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. Clinical and MRI examination of the participants A medical history was obtained from all patients and their caregivers. To measure the burden of camptocormia in daily life, patients used a numerical rating scale (NRS) from 1 to 10 with higher numbers indicating a greater impairment. Patients were asked to rate the severity of back pain on a NRS from 1 to 10 with higher numbers representing more severe pain. The neurological examination included the Unified Parkinson’s Disease Rating Scale, part III [16]. The patients in our study were scanned in a 1.5 T MRI Scanner (Philips Achieva, The Netherlands) with a spine phased array coil, using identical MR protocols with standard sequences used for evaluation of muscles [17]. This protocol consisted of: 1. T2 weighted (T2w) turbo spin echo (TSE) sagittal, slice thickness (SL) 3 mm, repetition time/echo time (TR/TE) 3500/120 ms, TSE factor 21, covering the thoracic and lumbar spine, 2. T1 weighted (T1w) spin echo (SE) axial (no angulation), SL 5 mm, TR/TE 514/15 ms, and 3. Short tau inversion recovery (STIR) axial (no angulation), SL 5 mm, TR/TE/TI
123
Edema Muscle edema was defined as a distinct hyperintense signal located within muscle tissue in at least two contiguous axial slices in the STIR sequence. The STIR sequence is sensitive to edema with a fat-suppressing method and is widely used to detect edemas in different organs. The edema severity (Fig. 1a–d) was semiquantitatively and independently judged by two raters (A.R., neuroradiologist; N.G.M., neurologist trained in the evaluation of muscle MRI) who were blinded to clinical data and categorized severity as: not present (0), mild (1, slight hyperintensity) or severe (2, pronounced hyperintensity). In cases of disagreement a consensus decision was made. For the paravertebral muscles, the axial STIR sequences covering the upper sacral region up to T1 were evaluated with respect to the occurrence of edema and the edema severity was rated. An edema score was calculated by multiplying the edema severity (score 0–2) by the cranio-caudal extensiveness of the edema, quantified by the number of vertebrae defining the length of the edema. This score was calculated for each side separately (unilateral score) and for both sides combined as a sum score (bilateral score). This semiquantitative assessment of edema severity was also used for the psoas and the quadratus lumborum muscles (score 0–2). Due to the smaller volume of these two muscles, the spatial distribution of the edema was not described and an edema score not calculated. Swelling This criterion was defined by two essential factors: (1) presence of muscle edema in the STIR sequence as previously defined, and (2) obviously increased size of the muscle compared to its contralateral counterpart on an axial T1w scan (Fig. 1e). Swelling indicated a pronounced muscle edema and was evaluated as being absent or present for the three muscles on level L3, and for the paravertebral muscles additionally on T12.
J Neurol
Fig. 1 Examples for the assessment of the criteria edema, swelling and fatty degeneration. a–d shows an MRI of a PD patient with a camptocormia duration of 11 months. Axial MR imaging at the level of L3 (see dotted line in a) shows an increase in size (swelling) of the left quadratus lumborum muscle compared to the right side (crosshair in T1w image in b), linked to an edematous signal increase in the STIR image (c). Edemas are present in both erector spinae (asterisk) and there is also significant pathological signal increase after i.v. gadolinium application (asterisk), which was not given regularly in patients or controls (d). e Gives an example of edema assessment: In the coronal STIR image of a PD patient with a camptocormia duration of 16 months, a left-sided strong edema (severity score 2) of a large
spatial distribution (see crosshair) of the paravertebral muscles is shown. The crosshair indicates also the position in the muscle as demonstrated in the smaller axial T1w image. The edema (severity score 2) extends from T12–L5 (6 segments), which is calculated as a unilateral edema score of 12. f–g shows an example of the segmentation technique: For the quantitative assessment of the amount of fat in the paravertebral muscles (see method section), we used axial T1w images (f). Applying an individual threshold derived from the signal of pure retroperitoneal fat, muscle and fat tissue in the muscle ROI were separated and colored red (muscle) and green (fat) as indicated in (g). In this PD patient with a camptocormia duration of 44 months, about 33–35 % of the muscle fascia contained fat
Fatty degeneration
intensity profile for the cut-off value. This value was defined as the mean intensity minus twice the standard deviation of the intensity distribution within this fat region. These cut-off values were determined for each patient image separately, and for each side (right/left) and slice (L3/T12), to account for the variations in the MR signal. The thoracolumbar fascia surrounding the paravertebral muscles was graphically lined out by manual processing. This was conducted on each side and both levels (L3 and T12) using the axial T1w sequences. Every voxel in this region of interest (ROI) with intensity below the defined cut-off fat value was considered to represent muscle tissue, and every voxel above was considered to represent fat tissue (Fig. 1f–g). As a measure for fatty replacement of
Fatty replacement was identified by hyperintense signals in T1w images within the muscle. Unlike in T2w images, edema produces a signal decrease in T1w images and can thus not be confused with fatty degeneration [18]. A quantitative measurement of fatty degeneration was applied for the paravertebral muscles as follows: The T1w intensity profiles of pure fat and lean muscle tissue show non-overlapping curves (with a low signal for muscle tissue and a high signal for fat tissue). We used individual intensity profiles from fat tissue as a reference to define a cutoff value for separating muscle and fat tissue. A region within the patient’s retroperitoneal fat provided the
123
J Neurol
muscle, we calculated, in percent, the area of fat within the ROI relative to the area of the whole muscle fascia (ROI). The described method is comparable with that used by Kim et al. [19]. We used a semiquantitative assessment for the psoas and the quadratum lumborum muscles. Fatty replacement of these muscles on level L3 was characterized as absent or minimal (score 0), compromising more than 25 % of the cross-section of the muscle (score 1) or more than 50 % (score 2). Statistical analysis Comparisons between the two examined groups were performed using a two-sample Wilcoxon rank sum test, or a Fisher’s exact test when comparing proportions. Correlations were found using non-parametric Spearman’s rank correlation tests. An alpha level p B 0.05 was regarded as significant difference. Inter-rater reliability was tested with the weighted Fleiss-Cohen’s kappa method [20]. All analyses were conducted using R statistics (http:// www.r-project.org/; version 2.15.2) and the additional Visualizing Categorical Data package (vcd; version 1.2-13).
indicated by a higher bilateral and unilateral edema score. Swelling of the paravertebral muscles was found exclusively in the group of camptocormia patients and was unilateral in 7–8 patients in whom swelling was seen. All of the 8 camptocormia patients with swelling of the paravertebral muscles had a clinical laterodeviation. The swelling was always contralateral to the side of the laterodeviation, with the exception of two cases (one bilateral and one unilateral swelling of the same side of the laterodeviation). For the quadratus lumborum muscle, edema was detected in 8/19 (42 %) of PD patients with camptocormia, which was unilateral in all but one patient. In 7/9 patients, edema was combined with a swelling of this muscle. As in the paravertebral muscles, all patients with swelling of the quadratus lumborum muscle also had a laterodeviation. In 7 out of 7 cases, this swelling was contralateral to the side of the laterodeviation. None of the patients in the control group showed any edema or swelling of the quadratus lumborum muscle. Regarding the psoas muscle, only one camptocormia patient showed a unilateral edema without swelling, while no abnormalities of this muscle were detected in the control group. Fatty degeneration
Results Clinical data Twenty patients with idiopathic PD and camptocormia were enrolled. The baseline characteristics are presented in Table 1. The mean duration of camptocormia was 28.2 months (±20.9) and the mean forward bending angle 55.0° (±13.7). Five patients showed forward bending only and fifteen patients had additional laterodeviation. Back pain was reported by 85 % of our patients with a moderate to severe strength. The control group of 20 PD patients without camptocormia showed no statistically relevant difference in the matching criteria. The only significant difference was a lower score on the pain numerical rating scale. Muscle MRI
There was no difference between the two groups with regard to the mean total size of the ROI restricted by the thoracolumbar fascia compromising the paravertebral muscles at levels L3 and T12, nor with regard to the mean of the signal intensities of the retroperitoneal fat reference region that served to determine the signal cut-off values (Table 2). The amount of fat calculated in % of the crosssection of the muscle was significantly larger in the group of PD patients with camptocormia on L3 and T12 than in the control group. For the quadratus lumborum and the psoas muscles, we did not find any fatty degeneration in the semiquantitative rating in any PD patient with or without camptocormia. Radiological secondary findings These findings are summarized in Table 1. There were no statistically relevant differences between the two groups.
Swelling and edema The inter-rater reliability of the two raters in regard to the radiological criteria edema and swelling was good to excellent (Table 2). As shown in Table 2, edema of the paravertebral muscles in the PD patients with camptocormia was more frequent, of higher signal intensity and of a more extended spatial distribution, which was also
123
Correlations between clinical and radiological data for the PD patients with camptocormia The bilateral edema score of the paravertebral muscles was negatively correlated with the duration of camptocormia (q = -0.46; p = 0.050). While both components of the edema score contributed to the correlation with the
J Neurol Table 1 Baseline characteristics and radiological secondary findings of the PD patients with and without camptocormia PD patients without camptocormia (n = 20) mean (±SD)
p valuea
Characteristic
PD patients with camptocormia (n = 20) mean (±SD)
Sex (F/M)
8/12
8/12
1
Age (years)
66.4 (±6.2)
69.3 (±5.8)
0.176
Duration of PD (years) L-dopa equivalent (mg/day)
11.5 (±3.9) 977.5 (±403.8)
12.9 (±4.5) 935.2 (±292.5)
0.406 0.645
UPDRS-III (ON)
23.4 (±10.3)
20.5 (±8.8)
0.229
Hoehn and Yahr (ON)
3.0 (±0.8)
2.6 (±0.7)
0.207
Body mass index (BMI)
24.4 (±4.8)
25.5 (±3.0)
0.09
NRS back pain (points)
6.4 (±2.5)
3.1 (±3.0)
0.003*
Dyskinesia (number of affected patients)
8 (40 %)
13 (65 %)
0.205
Motor fluctuations
13 (65 %)
17 (85 %)
0.273
Freezing
9 (45 %)
12 (60 %)
0.527
On-dystonia
12 (60 %)
6 (30 %)
0.110
Autonomic symptoms
11 (55 %)
11 (55 %)
1
Time between the diagnosis of PD and camptocormia (years) Age at the beginning of camptocormia (years)
9.2 (±4.5)
Not applicable for patients without camptocormia
Angle of camptocormia (deg)
55.0 (±13.7)
Duration of camptocormia (months)
28.2 (±20.9)
NRS impact of camptocormia on daily life (points)
7.4 (±2.0)
64.0 (±6.9)
Radiological secondary findings Pathological overall (number of patients)
14 (70 %)
12 (60 %)
0.741
Slight degree
3 (15 %)
5 (25 %)
0.695
Severe degree
11 (55 %)
7 (35 %)
0.341
Osteochrondrosis and activated osteochondrosis
9 (45 %)
7 (35 %)
0.748
Activated spondylathrosis
4 (20 %)
0 (0 %)
0.106
Spinal canal stenosis (absolute, relative)
4 (20 %)
4 (20 %)
1
Former fracture or surgical treatment
4 (20 %)
4 (20 %)
1
Pseudospondylolisthesis
3 (15 %)
5 (25 %)
0.695
NRS numerical rating scale (1–10 points), PD idiopathic Parkinson’s disease * Significant difference (alpha level p B 0.05) a Group differences of nominal variables were tested by the Fisher test, all other group comparisons were done by the Mann–Whitney U Test (asymptotic, two-sided)
duration of camptocormia, the role of edema severity was slightly more prominent (q = -0.52; p = 0.022) than the cranio-caudal extensiveness of the edema (q = -0.40; p = 0.091). Additionally, we found a strong positive correlation between fatty degeneration on both examined levels and the duration of camptocormia (q = 0.74; p = 0.000 at level L3 and q = 0.75; p = 0.001 at level T12; Fig. 2). Contrastingly, we did not find a significant correlation between the duration of PD and any of the radiological variables. As the bilateral edema score was negatively and fatty degeneration positively correlated to the duration of camptocormia, we aimed to use these parameters for a classification into acute or chronic stage of camptocormia.
Unilateral edema scores and the percentage values of fatty degeneration at level L3 and level T12 of the paravertebral muscles were considered pathologic, if they were outside two standard deviations of the mean of the control group. We considered those patients in an acute stage of camptocormia who exhibited pathologic muscle edema but no fatty degeneration. Patients with fatty degeneration, regardless of whether they still had partial edema, were defined as chronic stage of camptocormia. Eighteen out of 20 patients were classified according to this method. Two patients were excluded because of lack of data for this analysis. Ten patients (56 %) were identified as being in an acute stage and suffering from camptocormia for 4 up to 31 months (Fig. 3). Six patients (33 %) were classified as
123
J Neurol Table 2 Radiological data for edema, swelling and fatty degeneration of the paravertebral, quadratus lumborum and psoas muscles Edema and swelling of the paravertebral muscles
PD patients with camptocormia Mean (SD) or percentage (n = 19)a
PD patients without camptocormia Mean (SD) or percentage (n = 19)a
p value
Agreement between two ratersd
Bilateral edema score (points)e
21.4 (±18.8)
4.4 (±9.8)
0.000*b
q = 0.93
c
j = 0.84 j = 0.72
Presence of edema Presence of swelling
84 % 42 %
26 % 0%
Ratio of uni- to bilateral swelling
7:1
0:0
0.001* 0.001*c
Edema and swelling of the quadratus lumborum muscle
Percentage (n = 19)a
Percentage (n = 19)a
p valuec
Agreement between two raters (%)
Presence of edema Presence of swelling
42 % 37 %
0% 0%
0.001* 0.008*
90 95
Ratio of uni- to bilateral swelling
7:0
0:0
Edema and swelling of the psoas muscle
Percentage (n = 19)a
Percentage (n = 19)a
p valuec
Presence of edema
5%
0%
0.49
Presence of swelling
0%
0%
1
Agreement between two raters (%) 98 100
Fatty degeneration of the paravertebral muscles (level L3)
Mean (SD) (n = 19)a
Mean (SD) (n = 20)
p valueb
Total area muscle compartment (mm2) Cut-off intensity value to separate fat from muscle tissue
5536.3 (±1205.6) 359.7 (±77.9)
5556.0 (±895.3) 397.2 (±63.3)
0.99 0.184
19.5 (±11.9)
13.7 (±6.7)
0.050*
Fat / compartment (%)
p valueb
Fatty degeneration of the paravertebral muscles (level T12)
Mean (SD) (n = 17)a
Mean (SD) (n = 20)
Total area muscle compartment (mm2)
3363.1 (±667.7)
3191.1 (±831.3)
0.270
378.8 (±68.3)
393.1 (±41.4)
0.641
14.7 (±12.5)
5.8 (±3.3)
0.005*
Cut-off intensity value to separate fat from muscle tissue Fat / compartment (%) PD idiopathic Parkinson’s disease * Significant difference (alpha level p B 0.05) a
Partly sequences could not be analyzed for technical reasons
b
Based on a Wilcoxon rank sum test (asymptotic, two-sided)
c
Fisher’s exact test
d
Kappa for dichotomous data (values [0.5 indicates strong agreement) or Spearman’s rank correlation rho for continues scores
e
The bilateral edema score is a sum score of both sides, constituted of the unilateral edema score for the left and right side, respectively. For each side, a unilateral edema score is calculated by multiplying the edema severity (score 0–2) with the cranio-caudal extensiveness of the edema given by the number of vertebrae defining the length of the edema. Data is based on the consensus of two raters; last row indicates agreement before consensus
chronic stage (five patients only with fatty degeneration, one patient with additional partial edema). Their duration of camptocormia ranged between 37 and 84 months. Two camptocormia patients (11 %) were classified by this scheme as normal, because they did not show any pathological values according to the defined cut-off criteria. In the control group, two patients showed edematous muscle changes according to the named cut-off value: one of these also exhibited an increased amount of fat at level L3.
123
Discussion In this case–control study, we defined MRI criteria to characterize pathological changes in the lumbar trunk muscles. These criteria distinguished PD patients with camptocormia versus those without in the so far largest systematically MRI-analyzed camptocormia patient cohort. Our MRI findings on muscle changes suggest a dynamic process in camptocormia characterized by an initial acute
J Neurol
Fig. 2 Non-parametric correlation between the duration of camptocormia and the main radiological variables (bilateral edema score and percent fat at level L3 and T12). Significant difference at an alpha level p B 0.05
phase of edema and swelling of the paravertebral and quadratus lumborum muscles in patients with a shorter duration of camptocormia, and fatty degeneration of the paravertebral muscles as the hallmark of advanced camptocormia in the chronic stage. Acute changes only—without fatty degeneration—were detected in our cohort up to 31 months after the beginning of camptocormia. While pathological changes in paravertebral muscles in terms of camptocormia have been described previously, our present data show the involvement of other lumbar trunk muscles, possibly indicating the need for a more systemic view on camptocormia. Muscle MRI criteria for camptocormia So far, the potential for using muscle MRI as a non-invasive diagnostic tool in PD patients with camptocormia has not been analyzed systematically. The existing literature on muscle MRI in these patients has focused on degenerative changes [4, 7, 12]. Acute changes are missed because of the absence of the STIR sequence. Here, distinguishing between PD patients with and without camptocormia, we can define MRI criteria on the basis of T1w and STIR sequences and, to our best knowledge, offer for the first time an interpretation of these findings with regard to clinical data. Interestingly, the MR changes in paravertebral muscles in camptocormia patients are similar to those seen in muscles after tendon tears [11]. Edema and swelling While we found edema of the paravertebral muscles significantly more pronounced in camptocormia patients compared to controls, swelling as the hallmark of severe edema in the paravertebral muscles was exclusively present only in camptocormia patients. Moreover, we observed
edema and/or swelling of the quadratus lumborum muscle solely in camptocormia patients, as well as the rare edema of the psoas muscle. Hyperintensities in the STIR sequence are non-specific as they represent stationary body fluid such as edema and therefore indicating an acute process. Generally, they can be found in myositis, trauma, denervations as well as acute and chronic muscle diseases [21–27]. Previously, two cases of PD patients with camptocormia were reported showing similar radiological findings and were diagnosed bioptically with focal myositis of the paravertebral muscles [28, 29]. However, as shown pathologically [8], myositis is only a rare cause of camptocormia in PD. Nevertheless, in cases of exclusive edema and swelling in MRI, a biopsy to rule out myositis should be considered. Muscle wasting and fatty degeneration Our quantitative approach to estimate the proportion of fatty replacement in muscle tissue confirms that PD patients with camptocormia show a significantly higher proportion of fat in the paravertebral muscles than PD patients without camptocormia. The existence of fatty degeneration in camptocormia has been reported previously [7, 12, 13] but was regarded as an unspecific finding. We interpret these findings to be the result of a degenerative process of paravertebral muscles that appears similar to the well known degenerative process in muscles following tendon tears. These observations do not necessarily indicate a primary muscle disorder but rather imply a deregulation in the proprioceptive feedback and movement regulation [30]. Acute and chronic camptocormia Patients with a camptocormia duration of up to 31 months displayed edema and swelling only, which we call ‘‘acute
123
J Neurol b Fig. 3 Radiological images of the dynamic process from acute to
chronic stages of camptocormia in PD. In a PD patient without camptocormia (A1/2), erector spinae (asterisk), quadratus lumborum (curved arrows) and psoas muscles (triangles) appear normal on axial images at the level of L3, with muscle tissue being hypointense on T1w images compared to fat (A1) and being isointense compared to fat on STIR images (A2). In a PD patient with a 12-month history of camptocormia, lumbar trunk muscles at the level of T12 appear normal on T1w images (B1) but STIR images reveal slightly asymmetrical muscle edemas (asterisk in B2): presumed acute stage. In a PD patient with a 42-month history of camptocormia, unilateral muscle atrophy with fatty degeneration is seen, especially in the medial compartment (arrow in C1), and there is muscle edema on the contralateral side (asterisk in C2): presumed chronic stage. In a PD patient with a 55-month history of camptocormia, asymmetric muscle atrophy is seen not only involving the erector spinae (arrows in D1), but also the quadratus lumborum (curved arrow) and psoas muscles (triangle), however, there is no muscle edema (D2): presumed chronic stage without active pathological muscle changes
This interpretation of time-related pathological muscle changes in MRI is supported by the results of our current observational cohort study of a group of 25 idiopathic PD patients with camptocormia treated with deep brain stimulation (DBS) in the subthalamic nucleus [33]. Based on multifactorial analysis we found that all PD patients with a camptocormia duration of up to 1.5 years showed a beneficial effect of DBS on the axial disorder, patients between 1.5 and *3 years showed mixed results, but none with a camptocormia duration of more than 40 months improved. We hypothesize that PD patients with camptocormia entering a stage of chronic muscle changes as shown in MRI are unlikely to benefit from DBS. The paravertebral muscle changes could be regarded as indicating a certain time point in a process of at present unknown etology and not necessarily a primarily myopathy. Deregulation of lumbar trunk muscles in camptocormia
changes’’. In comparison to the control group, after 3 years patients in our camptocormia cohort started to present fatty degeneration: we describe this stage as ‘‘chronic’’. If confirmed by longitudinal data, these observations suggest that MRI is a useful tool in determining stages of camptocormia. Following this hypothesis, the paravertebral muscles seem to enter a stage of chronic and possibly degenerative change with inferior therapeutic prospects of success after that time. This might explain why different treatment approaches, for example with deep brain stimulation, have shown inconsistent results [31, 32]. Consequently, prospective studies should include the duration of camptocormia in their design.
123
Previous research in camptocormia has focused on the paravertebral muscles. The question of other affected muscles has hardly been addressed [28, 34]. Our data indicate an involvement of the quadratus lumborum muscle in nearly half of the cases for the variables ‘‘edema’’ and ‘‘swelling’’. Rarely, the psoas muscle also seems to be affected. We do not have a conclusive explanation for this at the present time but these findings argue against camptocormia as an isolated process in the paravertebral muscles. Instead our findings suggest a need for a more systemic view in the search for the etiology of camptocormia, considering a deregulation of lumbar trunk muscles by a central mechanism as indicated by the possibly positive effects of DBS.
J Neurol Acknowledgments The authors express their deep thanks to Sari Munser for her valuable help in gathering the data of this study, and Klarissa Hanja Stu¨rner and Walter J. Schulz-Schaeffer for very helpful discussion of the manuscript. The study was supported by a grant from the Medical Faculty of the Christian-Albrechts University of Kiel. Conflicts of interest On behalf of all authors, the corresponding author states that there is no conflict of interest.
References 1. Doherty KM, van de Warrenburg BP, Peralta MC, SilveiraMoriyama L, Azulay JP, Gershanik OS, Bloem BR (2011) Postural deformities in Parkinson’s disease. Lancet Neurol 10:538–549 2. Finsterer J, Strobl W (2010) Presentation, etiology, diagnosis, and management of camptocormia. Eur Neurol 64:1–8 3. Upadhyaya CD, Starr PA, Mummaneni PV (2010) Spinal deformity and Parkinson disease: a treatment algorithm. Neurosurg Focus 28:E5 4. Lepoutre AC, Devos D, Blanchard-Dauphin A, Pardessus V, Maurage CA, Ferriby D, Hurtevent JF, Cotten A, Destee A, Defebvre L (2006) A specific clinical pattern of camptocormia in Parkinson’s disease. J Neurol Neurosurg Psychiatry 77:1229–1234 5. Tiple D, Fabbrini G, Colosimo C, Ottaviani D, Camerota F, Defazio G, Berardelli A (2009) Camptocormia in Parkinson disease: an epidemiological and clinical study. J Neurol Neurosurg Psychiatry 80:145–148 6. Ashour R, Jankovic J (2006) Joint and skeletal deformities in Parkinson’s disease, multiple system atrophy, and progressive supranuclear palsy. Mov Disord 21:1856–1863 7. Abe K, Uchida Y, Notani M (2010) Camptocormia in Parkinson’s disease. Parkinsons Dis 8. Wrede A, Margraf NG, Goebel HH, Deuschl G, Schulz-Schaeffer WJ (2012) Myofibrillar disorganization characterizes myopathy of camptocormia in Parkinson’s disease. Acta Neuropathol 123:419–432 9. Goutallier D, Postel JM, Bernageau J, Lavau L, Voisin MC (1994) Fatty muscle degeneration in cuff ruptures. Pre- and postoperative evaluation by CT scan. Clin Orthop Relat Res 78–83 10. Fuchs B, Weishaupt D, Zanetti M, Hodler J, Gerber C (1999) Fatty degeneration of the muscles of the rotator cuff: assessment by computed tomography versus magnetic resonance imaging. J Shoulder Elbow Surg 8:599–605 11. Hoffmann A, Mamisch N, Buck FM, Espinosa N, Pfirrmann CW, Zanetti M (2011) Oedema and fatty degeneration of the soleus and gastrocnemius muscles on MR images in patients with Achilles tendon abnormalities. Eur Radiol 21:1996–2003 12. Bonneville F, Bloch F, Kurys E, du Montcel ST, Welter ML, Bonnet AM, Agid Y, Dormont D, Houeto JL (2008) Camptocormia and Parkinson’s disease: MR imaging. Eur Radiol 18:1710–1719 13. Schabitz WR, Glatz K, Schuhan C, Sommer C, Berger C, Schwaninger M, Hartmann M, Hilmar Goebel H, Meinck HM (2003) Severe forward flexion of the trunk in Parkinson’s disease: focal myopathy of the paraspinal muscles mimicking camptocormia. Mov Disord 18:408–414 14. Gibb WR, Lees AJ (1988) The relevance of the Lewy body to the pathogenesis of idiopathic Parkinson’s disease. J Neurol Neurosurg Psychiatry 51:745–752
15. Margraf NG, Wrede A, Rohr A, Schulz-Schaeffer WJ, Raethjen J, Eymess A, Volkmann J, Mehdorn MH, Jansen O, Deuschl G (2010) Camptocormia in idiopathic Parkinson’s disease: a focal myopathy of the paravertebral muscles. Mov Disord 25:542–551 16. Elton S. F. a. R (1987) Unified Parkinson’s disease rating scale. In: Fahn S, M. C, Caine DB, Goldstein ME (eds) Recent developments in Parkinson’s disease. Macmillan Health Care Information, Florham Park, NJ, USA pp 153–163, 293–304 17. Beese MS, Winkler G, Nicolas V, Maas R, Kress D, Kunze K, Bucheler E (1993) The diagnosis of inflammatory muscular and vascular diseases using MRT with STIR sequences. Rofo 158:542–549 18. Ranson CA, Burnett AF, Kerslake R, Batt ME, O’Sullivan PB (2006) An investigation into the use of MR imaging to determine the functional cross sectional area of lumbar paraspinal muscles. Eur Spine J 15:764–773 19. Kim H, Lee CK, Yeom JS, Lee JH, Cho JH, Shin SI, Lee HJ, Chang BS (2013) Asymmetry of the cross-sectional area of paravertebral and psoas muscle in patients with degenerative scoliosis. Eur Spine J 22:1332–1338 20. Cohen J (1968) Weighted kappa: nominal scale agreement with provision for scaled disagreement or partial credit. Psychol Bull 70:213–220 21. West GA, Haynor DR, Goodkin R, Tsuruda JS, Bronstein AD, Kraft G, Winter T, Kliot M (1994) Magnetic resonance imaging signal changes in denervated muscles after peripheral nerve injury. Neurosurgery 35:1077–1085 (discussion 1085–1076) 22. McDonald CM, Carter GT, Fritz RC, Anderson MW, Abresch RT, Kilmer DD (2000) Magnetic resonance imaging of denervated muscle: comparison to electromyography. Muscle Nerve 23:1431–1434 23. Tasca G, Monforte M, Iannaccone E, Laschena F, Ottaviani P, Leoncini E, Boccia S, Galluzzi G, Pelliccioni M, Masciullo M, Frusciante R, Mercuri E, Ricci E (2014) Upper girdle imaging in facioscapulohumeral muscular dystrophy. PLoS One 9:e100292 24. Diekman EF, van der Pol WL, Nievelstein RA, Houten SM, Wijburg FA, Visser G (2014) Muscle MRI in patients with longchain fatty acid oxidation disorders. J Inherit Metab Dis 37:405–413 25. Subhawong TK, Wang X, Machado AJ, Mammen AL, Christopher-Stine L, Barker PB, Carrino JA, Fayad LM (2013) 1H Magnetic resonance spectroscopy findings in idiopathic inflammatory myopathies at 3 T: feasibility and first results. Invest Radiol 48:509–516 26. Hayashi D, Hamilton B, Guermazi A, de Villiers R, Crema MD, Roemer FW (2012) Traumatic injuries of thigh and calf muscles in athletes: role and clinical relevance of MR imaging and ultrasound. Insights Imaging 3:591–601 27. Tomasova Studynkova J, Charvat F, Jarosova K, Vencovsky J (2007) The role of MRI in the assessment of polymyositis and dermatomyositis. Rheumatology (Oxford) 46:1174–1179 28. Wunderlich S, Csoti I, Reiners K, Gunthner-Lengsfeld T, Schneider C, Becker G, Naumann M (2002) Camptocormia in Parkinson’s disease mimicked by focal myositis of the paraspinal muscles. Mov Disord 17:598–600 29. Charpentier P, Dauphin A, Stojkovic T, Cotten A, Hurtevent JF, Maurage CA, Thevenon A, Destee A, Defebvre L (2005) Parkinson’s disease, progressive lumbar kyphosis and focal paraspinal myositis. Rev Neurol (Paris) 161:459–463 30. Dietz V (2002) Proprioception and locomotor disorders. Nat Rev Neurosci 3:781–790 31. Asahi T, Taguchi Y, Hayashi N, Hamada H, Dougu N, Takashima S, Tanaka K, Endo S (2011) Bilateral subthalamic deep brain stimulation for camptocormia associated with Parkinson’s disease. Stereotact Funct Neurosurg 89:173–177
123
J Neurol 32. Capelle HH, Schrader C, Blahak C, Fogel W, Kinfe TM, Baezner H, Krauss JK (2011) Deep brain stimulation for camptocormia in dystonia and Parkinson’s disease. J Neurol 258:96–103 33. Schulz-Schaeffer WJ, Margraf NG, Munser S, Wrede A, Buhmann C, Deuschl G, Oehlwein C (2015) Effect of neurostimulation on camptocormia in Parkinson’s disease depends on symptom duration. Mov Disord (epub ahead of print)
123
34. Spuler S, Krug H, Klein C, Medialdea IC, Jakob W, Ebersbach G, Gruber D, Hoffmann KT, Trottenberg T, Kupsch A (2010) Myopathy causing camptocormia in idiopathic Parkinson’s disease: a multidisciplinary approach. Mov Disord 25:552–559
ORIGINAL COMMUNICATION
MRI of lumbar trunk muscles in patients with Parkinson’s disease and camptocormia N. G. Margraf1 • A. Rohr2 • O. Granert1 • J. Hampel1 A. Drews2 • G. Deuschl1
•
Received: 2 March 2015 / Revised: 25 March 2015 / Accepted: 26 March 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract Camptocormia in Parkinson’s disease (PD) is an axial postural disorder usually accompanied by histopathological changes in the paravertebral muscles of unknown etiology. The diagnostic potential of magnetic resonance imaging (MRI) of back muscles in camptocormia has not been systematically assessed. Our objective was to characterize pathological muscle changes with MRI and to develop radiological criteria for camptocormia. The criteria edema, swelling and fatty degeneration in 20 idiopathic PD patients with camptocormia were assessed using MRI (T1w and short tau inversion recovery (STIR) sequences) of the lumbar trunk muscles and compared with 20 group-matched PD patients without camptocormia. Edema and fatty degeneration of the paravertebral muscles were significantly more frequent in camptocormia. Edema correlated negatively and fatty degeneration positively with the duration of camptocormia and not PD. Swelling of the paravertebral muscles, edema and swelling of the quadratus lumborum muscle and rare edema of the psoas muscle were only found in camptocormia patients. In this case–control study the defined MRI criteria distinguish the group of PD patients with camptocormia versus those without. Our findings suggest dynamic changes in the MRI signals over N. G. Margraf and A. Rohr contributed equally to the manuscript. & N. G. Margraf [email protected] & A. Rohr 1
Department of Neurology, University Hospital SchleswigHolstein, Campus Kiel, Schittenhelmstr. 10, 24105 Kiel, Germany
2
Institute of Neuroradiology, University Hospital SchleswigHolstein, Campus Kiel, Schittenhelmstr. 10, 24105 Kiel, Germany
time in the paravertebral muscles: edema and swelling are found initially, followed by fatty atrophic degeneration 2–3 years after the beginning of camptocormia. Muscle MRI qualifies as a tool for categorizing phases of camptocormia as acute or chronic, with potential consequences for therapeutic approaches. The involvement of muscles beyond an isolated impairment of the paravertebral muscles implies a more systemic view with a deregulation of lumbar trunk muscles. Keywords Idiopathic Parkinson’s disease Camptocormia Paravertebral muscles Bent spine syndrome Muscle MRI
Introduction Camptocormia is an involuntary, marked flexion of the thoracolumbar spine, which is apparent during standing or walking and disappears in the supine position. This axial postural disorder has been identified in a broad spectrum of neurological diseases [1–3] but seems to be most common in movement disorders and particularly Parkinson’s disease (PD). The prevalence of camptocormia in PD ranges widely between 3 and 17.6 % [4–7]. On investigating myopathological changes in the paravertebral muscles of camptocormia patients suffering from PD, remarkable similarities to the lesion pattern of muscles with tendon tears were observed [8]. Although CT and MR imaging criteria have been developed in the orthopaedic diagnostics of muscles following tension tears [9–11], there is limited experience with the use of MRI in detecting corresponding changes in the back muscles in camptocormia [4, 12, 13]. Our objective was to examine the diagnostic capability of MRI of lumbar trunk muscles
123
J Neurol
in PD patients with camptocormia, and to develop radiological criteria suitable to characterize pathological changes in these muscles.
(inversion time) 2500/70/170 ms, TSE factor 15. Axial T1w and STIR images were acquired in two stacks, covering the whole thoracic and lumbar spine as the area of the clinically obvious changes.
Methods
Evaluation of the MRI of the lumbar trunk muscles
Patients and controls
We analyzed MRI changes in three lumbar trunk muscles: the paravertebral muscles, the quadratus lumborum and the psoas muscles. We defined the criteria ‘‘edema’’, ‘‘swelling’’ and ‘‘fatty degeneration’’ as described below. Relevant pathological changes of the spine were recorded.
Patients with idiopathic PD (according to the British Parkinson’s Disease Society Brain Bank criteria [14]) and camptocormia were recruited at the Department of Neurology, Kiel University between 2010 and 2013. Camptocormia (with or without laterodeviation) was defined as an at least 30° thoracolumbar flexion apparent when standing or walking and resolving in the recumbent position [15]. Laterodeviation was defined as visually obvious deflection from the normally horizontal line between the two shoulders. Control patients with idiopathic PD without camptocormia were recruited and group matched for sex, age, subtype, duration and severity of PD (Hoehn & Yahr stage, UPDRS III motor score, L-dopa equivalent dose). Exclusion criteria for both groups were: contra-indications for MRI examination, clinical or radiological signs suggesting an atypical Parkinsonian syndrome or dystonia. Informed consent was obtained from all participants. The ethics committee of the Medical Faculty of the Christian-Albrechts University of Kiel approved the study and it was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. Clinical and MRI examination of the participants A medical history was obtained from all patients and their caregivers. To measure the burden of camptocormia in daily life, patients used a numerical rating scale (NRS) from 1 to 10 with higher numbers indicating a greater impairment. Patients were asked to rate the severity of back pain on a NRS from 1 to 10 with higher numbers representing more severe pain. The neurological examination included the Unified Parkinson’s Disease Rating Scale, part III [16]. The patients in our study were scanned in a 1.5 T MRI Scanner (Philips Achieva, The Netherlands) with a spine phased array coil, using identical MR protocols with standard sequences used for evaluation of muscles [17]. This protocol consisted of: 1. T2 weighted (T2w) turbo spin echo (TSE) sagittal, slice thickness (SL) 3 mm, repetition time/echo time (TR/TE) 3500/120 ms, TSE factor 21, covering the thoracic and lumbar spine, 2. T1 weighted (T1w) spin echo (SE) axial (no angulation), SL 5 mm, TR/TE 514/15 ms, and 3. Short tau inversion recovery (STIR) axial (no angulation), SL 5 mm, TR/TE/TI
123
Edema Muscle edema was defined as a distinct hyperintense signal located within muscle tissue in at least two contiguous axial slices in the STIR sequence. The STIR sequence is sensitive to edema with a fat-suppressing method and is widely used to detect edemas in different organs. The edema severity (Fig. 1a–d) was semiquantitatively and independently judged by two raters (A.R., neuroradiologist; N.G.M., neurologist trained in the evaluation of muscle MRI) who were blinded to clinical data and categorized severity as: not present (0), mild (1, slight hyperintensity) or severe (2, pronounced hyperintensity). In cases of disagreement a consensus decision was made. For the paravertebral muscles, the axial STIR sequences covering the upper sacral region up to T1 were evaluated with respect to the occurrence of edema and the edema severity was rated. An edema score was calculated by multiplying the edema severity (score 0–2) by the cranio-caudal extensiveness of the edema, quantified by the number of vertebrae defining the length of the edema. This score was calculated for each side separately (unilateral score) and for both sides combined as a sum score (bilateral score). This semiquantitative assessment of edema severity was also used for the psoas and the quadratus lumborum muscles (score 0–2). Due to the smaller volume of these two muscles, the spatial distribution of the edema was not described and an edema score not calculated. Swelling This criterion was defined by two essential factors: (1) presence of muscle edema in the STIR sequence as previously defined, and (2) obviously increased size of the muscle compared to its contralateral counterpart on an axial T1w scan (Fig. 1e). Swelling indicated a pronounced muscle edema and was evaluated as being absent or present for the three muscles on level L3, and for the paravertebral muscles additionally on T12.
J Neurol
Fig. 1 Examples for the assessment of the criteria edema, swelling and fatty degeneration. a–d shows an MRI of a PD patient with a camptocormia duration of 11 months. Axial MR imaging at the level of L3 (see dotted line in a) shows an increase in size (swelling) of the left quadratus lumborum muscle compared to the right side (crosshair in T1w image in b), linked to an edematous signal increase in the STIR image (c). Edemas are present in both erector spinae (asterisk) and there is also significant pathological signal increase after i.v. gadolinium application (asterisk), which was not given regularly in patients or controls (d). e Gives an example of edema assessment: In the coronal STIR image of a PD patient with a camptocormia duration of 16 months, a left-sided strong edema (severity score 2) of a large
spatial distribution (see crosshair) of the paravertebral muscles is shown. The crosshair indicates also the position in the muscle as demonstrated in the smaller axial T1w image. The edema (severity score 2) extends from T12–L5 (6 segments), which is calculated as a unilateral edema score of 12. f–g shows an example of the segmentation technique: For the quantitative assessment of the amount of fat in the paravertebral muscles (see method section), we used axial T1w images (f). Applying an individual threshold derived from the signal of pure retroperitoneal fat, muscle and fat tissue in the muscle ROI were separated and colored red (muscle) and green (fat) as indicated in (g). In this PD patient with a camptocormia duration of 44 months, about 33–35 % of the muscle fascia contained fat
Fatty degeneration
intensity profile for the cut-off value. This value was defined as the mean intensity minus twice the standard deviation of the intensity distribution within this fat region. These cut-off values were determined for each patient image separately, and for each side (right/left) and slice (L3/T12), to account for the variations in the MR signal. The thoracolumbar fascia surrounding the paravertebral muscles was graphically lined out by manual processing. This was conducted on each side and both levels (L3 and T12) using the axial T1w sequences. Every voxel in this region of interest (ROI) with intensity below the defined cut-off fat value was considered to represent muscle tissue, and every voxel above was considered to represent fat tissue (Fig. 1f–g). As a measure for fatty replacement of
Fatty replacement was identified by hyperintense signals in T1w images within the muscle. Unlike in T2w images, edema produces a signal decrease in T1w images and can thus not be confused with fatty degeneration [18]. A quantitative measurement of fatty degeneration was applied for the paravertebral muscles as follows: The T1w intensity profiles of pure fat and lean muscle tissue show non-overlapping curves (with a low signal for muscle tissue and a high signal for fat tissue). We used individual intensity profiles from fat tissue as a reference to define a cutoff value for separating muscle and fat tissue. A region within the patient’s retroperitoneal fat provided the
123
J Neurol
muscle, we calculated, in percent, the area of fat within the ROI relative to the area of the whole muscle fascia (ROI). The described method is comparable with that used by Kim et al. [19]. We used a semiquantitative assessment for the psoas and the quadratum lumborum muscles. Fatty replacement of these muscles on level L3 was characterized as absent or minimal (score 0), compromising more than 25 % of the cross-section of the muscle (score 1) or more than 50 % (score 2). Statistical analysis Comparisons between the two examined groups were performed using a two-sample Wilcoxon rank sum test, or a Fisher’s exact test when comparing proportions. Correlations were found using non-parametric Spearman’s rank correlation tests. An alpha level p B 0.05 was regarded as significant difference. Inter-rater reliability was tested with the weighted Fleiss-Cohen’s kappa method [20]. All analyses were conducted using R statistics (http:// www.r-project.org/; version 2.15.2) and the additional Visualizing Categorical Data package (vcd; version 1.2-13).
indicated by a higher bilateral and unilateral edema score. Swelling of the paravertebral muscles was found exclusively in the group of camptocormia patients and was unilateral in 7–8 patients in whom swelling was seen. All of the 8 camptocormia patients with swelling of the paravertebral muscles had a clinical laterodeviation. The swelling was always contralateral to the side of the laterodeviation, with the exception of two cases (one bilateral and one unilateral swelling of the same side of the laterodeviation). For the quadratus lumborum muscle, edema was detected in 8/19 (42 %) of PD patients with camptocormia, which was unilateral in all but one patient. In 7/9 patients, edema was combined with a swelling of this muscle. As in the paravertebral muscles, all patients with swelling of the quadratus lumborum muscle also had a laterodeviation. In 7 out of 7 cases, this swelling was contralateral to the side of the laterodeviation. None of the patients in the control group showed any edema or swelling of the quadratus lumborum muscle. Regarding the psoas muscle, only one camptocormia patient showed a unilateral edema without swelling, while no abnormalities of this muscle were detected in the control group. Fatty degeneration
Results Clinical data Twenty patients with idiopathic PD and camptocormia were enrolled. The baseline characteristics are presented in Table 1. The mean duration of camptocormia was 28.2 months (±20.9) and the mean forward bending angle 55.0° (±13.7). Five patients showed forward bending only and fifteen patients had additional laterodeviation. Back pain was reported by 85 % of our patients with a moderate to severe strength. The control group of 20 PD patients without camptocormia showed no statistically relevant difference in the matching criteria. The only significant difference was a lower score on the pain numerical rating scale. Muscle MRI
There was no difference between the two groups with regard to the mean total size of the ROI restricted by the thoracolumbar fascia compromising the paravertebral muscles at levels L3 and T12, nor with regard to the mean of the signal intensities of the retroperitoneal fat reference region that served to determine the signal cut-off values (Table 2). The amount of fat calculated in % of the crosssection of the muscle was significantly larger in the group of PD patients with camptocormia on L3 and T12 than in the control group. For the quadratus lumborum and the psoas muscles, we did not find any fatty degeneration in the semiquantitative rating in any PD patient with or without camptocormia. Radiological secondary findings These findings are summarized in Table 1. There were no statistically relevant differences between the two groups.
Swelling and edema The inter-rater reliability of the two raters in regard to the radiological criteria edema and swelling was good to excellent (Table 2). As shown in Table 2, edema of the paravertebral muscles in the PD patients with camptocormia was more frequent, of higher signal intensity and of a more extended spatial distribution, which was also
123
Correlations between clinical and radiological data for the PD patients with camptocormia The bilateral edema score of the paravertebral muscles was negatively correlated with the duration of camptocormia (q = -0.46; p = 0.050). While both components of the edema score contributed to the correlation with the
J Neurol Table 1 Baseline characteristics and radiological secondary findings of the PD patients with and without camptocormia PD patients without camptocormia (n = 20) mean (±SD)
p valuea
Characteristic
PD patients with camptocormia (n = 20) mean (±SD)
Sex (F/M)
8/12
8/12
1
Age (years)
66.4 (±6.2)
69.3 (±5.8)
0.176
Duration of PD (years) L-dopa equivalent (mg/day)
11.5 (±3.9) 977.5 (±403.8)
12.9 (±4.5) 935.2 (±292.5)
0.406 0.645
UPDRS-III (ON)
23.4 (±10.3)
20.5 (±8.8)
0.229
Hoehn and Yahr (ON)
3.0 (±0.8)
2.6 (±0.7)
0.207
Body mass index (BMI)
24.4 (±4.8)
25.5 (±3.0)
0.09
NRS back pain (points)
6.4 (±2.5)
3.1 (±3.0)
0.003*
Dyskinesia (number of affected patients)
8 (40 %)
13 (65 %)
0.205
Motor fluctuations
13 (65 %)
17 (85 %)
0.273
Freezing
9 (45 %)
12 (60 %)
0.527
On-dystonia
12 (60 %)
6 (30 %)
0.110
Autonomic symptoms
11 (55 %)
11 (55 %)
1
Time between the diagnosis of PD and camptocormia (years) Age at the beginning of camptocormia (years)
9.2 (±4.5)
Not applicable for patients without camptocormia
Angle of camptocormia (deg)
55.0 (±13.7)
Duration of camptocormia (months)
28.2 (±20.9)
NRS impact of camptocormia on daily life (points)
7.4 (±2.0)
64.0 (±6.9)
Radiological secondary findings Pathological overall (number of patients)
14 (70 %)
12 (60 %)
0.741
Slight degree
3 (15 %)
5 (25 %)
0.695
Severe degree
11 (55 %)
7 (35 %)
0.341
Osteochrondrosis and activated osteochondrosis
9 (45 %)
7 (35 %)
0.748
Activated spondylathrosis
4 (20 %)
0 (0 %)
0.106
Spinal canal stenosis (absolute, relative)
4 (20 %)
4 (20 %)
1
Former fracture or surgical treatment
4 (20 %)
4 (20 %)
1
Pseudospondylolisthesis
3 (15 %)
5 (25 %)
0.695
NRS numerical rating scale (1–10 points), PD idiopathic Parkinson’s disease * Significant difference (alpha level p B 0.05) a Group differences of nominal variables were tested by the Fisher test, all other group comparisons were done by the Mann–Whitney U Test (asymptotic, two-sided)
duration of camptocormia, the role of edema severity was slightly more prominent (q = -0.52; p = 0.022) than the cranio-caudal extensiveness of the edema (q = -0.40; p = 0.091). Additionally, we found a strong positive correlation between fatty degeneration on both examined levels and the duration of camptocormia (q = 0.74; p = 0.000 at level L3 and q = 0.75; p = 0.001 at level T12; Fig. 2). Contrastingly, we did not find a significant correlation between the duration of PD and any of the radiological variables. As the bilateral edema score was negatively and fatty degeneration positively correlated to the duration of camptocormia, we aimed to use these parameters for a classification into acute or chronic stage of camptocormia.
Unilateral edema scores and the percentage values of fatty degeneration at level L3 and level T12 of the paravertebral muscles were considered pathologic, if they were outside two standard deviations of the mean of the control group. We considered those patients in an acute stage of camptocormia who exhibited pathologic muscle edema but no fatty degeneration. Patients with fatty degeneration, regardless of whether they still had partial edema, were defined as chronic stage of camptocormia. Eighteen out of 20 patients were classified according to this method. Two patients were excluded because of lack of data for this analysis. Ten patients (56 %) were identified as being in an acute stage and suffering from camptocormia for 4 up to 31 months (Fig. 3). Six patients (33 %) were classified as
123
J Neurol Table 2 Radiological data for edema, swelling and fatty degeneration of the paravertebral, quadratus lumborum and psoas muscles Edema and swelling of the paravertebral muscles
PD patients with camptocormia Mean (SD) or percentage (n = 19)a
PD patients without camptocormia Mean (SD) or percentage (n = 19)a
p value
Agreement between two ratersd
Bilateral edema score (points)e
21.4 (±18.8)
4.4 (±9.8)
0.000*b
q = 0.93
c
j = 0.84 j = 0.72
Presence of edema Presence of swelling
84 % 42 %
26 % 0%
Ratio of uni- to bilateral swelling
7:1
0:0
0.001* 0.001*c
Edema and swelling of the quadratus lumborum muscle
Percentage (n = 19)a
Percentage (n = 19)a
p valuec
Agreement between two raters (%)
Presence of edema Presence of swelling
42 % 37 %
0% 0%
0.001* 0.008*
90 95
Ratio of uni- to bilateral swelling
7:0
0:0
Edema and swelling of the psoas muscle
Percentage (n = 19)a
Percentage (n = 19)a
p valuec
Presence of edema
5%
0%
0.49
Presence of swelling
0%
0%
1
Agreement between two raters (%) 98 100
Fatty degeneration of the paravertebral muscles (level L3)
Mean (SD) (n = 19)a
Mean (SD) (n = 20)
p valueb
Total area muscle compartment (mm2) Cut-off intensity value to separate fat from muscle tissue
5536.3 (±1205.6) 359.7 (±77.9)
5556.0 (±895.3) 397.2 (±63.3)
0.99 0.184
19.5 (±11.9)
13.7 (±6.7)
0.050*
Fat / compartment (%)
p valueb
Fatty degeneration of the paravertebral muscles (level T12)
Mean (SD) (n = 17)a
Mean (SD) (n = 20)
Total area muscle compartment (mm2)
3363.1 (±667.7)
3191.1 (±831.3)
0.270
378.8 (±68.3)
393.1 (±41.4)
0.641
14.7 (±12.5)
5.8 (±3.3)
0.005*
Cut-off intensity value to separate fat from muscle tissue Fat / compartment (%) PD idiopathic Parkinson’s disease * Significant difference (alpha level p B 0.05) a
Partly sequences could not be analyzed for technical reasons
b
Based on a Wilcoxon rank sum test (asymptotic, two-sided)
c
Fisher’s exact test
d
Kappa for dichotomous data (values [0.5 indicates strong agreement) or Spearman’s rank correlation rho for continues scores
e
The bilateral edema score is a sum score of both sides, constituted of the unilateral edema score for the left and right side, respectively. For each side, a unilateral edema score is calculated by multiplying the edema severity (score 0–2) with the cranio-caudal extensiveness of the edema given by the number of vertebrae defining the length of the edema. Data is based on the consensus of two raters; last row indicates agreement before consensus
chronic stage (five patients only with fatty degeneration, one patient with additional partial edema). Their duration of camptocormia ranged between 37 and 84 months. Two camptocormia patients (11 %) were classified by this scheme as normal, because they did not show any pathological values according to the defined cut-off criteria. In the control group, two patients showed edematous muscle changes according to the named cut-off value: one of these also exhibited an increased amount of fat at level L3.
123
Discussion In this case–control study, we defined MRI criteria to characterize pathological changes in the lumbar trunk muscles. These criteria distinguished PD patients with camptocormia versus those without in the so far largest systematically MRI-analyzed camptocormia patient cohort. Our MRI findings on muscle changes suggest a dynamic process in camptocormia characterized by an initial acute
J Neurol
Fig. 2 Non-parametric correlation between the duration of camptocormia and the main radiological variables (bilateral edema score and percent fat at level L3 and T12). Significant difference at an alpha level p B 0.05
phase of edema and swelling of the paravertebral and quadratus lumborum muscles in patients with a shorter duration of camptocormia, and fatty degeneration of the paravertebral muscles as the hallmark of advanced camptocormia in the chronic stage. Acute changes only—without fatty degeneration—were detected in our cohort up to 31 months after the beginning of camptocormia. While pathological changes in paravertebral muscles in terms of camptocormia have been described previously, our present data show the involvement of other lumbar trunk muscles, possibly indicating the need for a more systemic view on camptocormia. Muscle MRI criteria for camptocormia So far, the potential for using muscle MRI as a non-invasive diagnostic tool in PD patients with camptocormia has not been analyzed systematically. The existing literature on muscle MRI in these patients has focused on degenerative changes [4, 7, 12]. Acute changes are missed because of the absence of the STIR sequence. Here, distinguishing between PD patients with and without camptocormia, we can define MRI criteria on the basis of T1w and STIR sequences and, to our best knowledge, offer for the first time an interpretation of these findings with regard to clinical data. Interestingly, the MR changes in paravertebral muscles in camptocormia patients are similar to those seen in muscles after tendon tears [11]. Edema and swelling While we found edema of the paravertebral muscles significantly more pronounced in camptocormia patients compared to controls, swelling as the hallmark of severe edema in the paravertebral muscles was exclusively present only in camptocormia patients. Moreover, we observed
edema and/or swelling of the quadratus lumborum muscle solely in camptocormia patients, as well as the rare edema of the psoas muscle. Hyperintensities in the STIR sequence are non-specific as they represent stationary body fluid such as edema and therefore indicating an acute process. Generally, they can be found in myositis, trauma, denervations as well as acute and chronic muscle diseases [21–27]. Previously, two cases of PD patients with camptocormia were reported showing similar radiological findings and were diagnosed bioptically with focal myositis of the paravertebral muscles [28, 29]. However, as shown pathologically [8], myositis is only a rare cause of camptocormia in PD. Nevertheless, in cases of exclusive edema and swelling in MRI, a biopsy to rule out myositis should be considered. Muscle wasting and fatty degeneration Our quantitative approach to estimate the proportion of fatty replacement in muscle tissue confirms that PD patients with camptocormia show a significantly higher proportion of fat in the paravertebral muscles than PD patients without camptocormia. The existence of fatty degeneration in camptocormia has been reported previously [7, 12, 13] but was regarded as an unspecific finding. We interpret these findings to be the result of a degenerative process of paravertebral muscles that appears similar to the well known degenerative process in muscles following tendon tears. These observations do not necessarily indicate a primary muscle disorder but rather imply a deregulation in the proprioceptive feedback and movement regulation [30]. Acute and chronic camptocormia Patients with a camptocormia duration of up to 31 months displayed edema and swelling only, which we call ‘‘acute
123
J Neurol b Fig. 3 Radiological images of the dynamic process from acute to
chronic stages of camptocormia in PD. In a PD patient without camptocormia (A1/2), erector spinae (asterisk), quadratus lumborum (curved arrows) and psoas muscles (triangles) appear normal on axial images at the level of L3, with muscle tissue being hypointense on T1w images compared to fat (A1) and being isointense compared to fat on STIR images (A2). In a PD patient with a 12-month history of camptocormia, lumbar trunk muscles at the level of T12 appear normal on T1w images (B1) but STIR images reveal slightly asymmetrical muscle edemas (asterisk in B2): presumed acute stage. In a PD patient with a 42-month history of camptocormia, unilateral muscle atrophy with fatty degeneration is seen, especially in the medial compartment (arrow in C1), and there is muscle edema on the contralateral side (asterisk in C2): presumed chronic stage. In a PD patient with a 55-month history of camptocormia, asymmetric muscle atrophy is seen not only involving the erector spinae (arrows in D1), but also the quadratus lumborum (curved arrow) and psoas muscles (triangle), however, there is no muscle edema (D2): presumed chronic stage without active pathological muscle changes
This interpretation of time-related pathological muscle changes in MRI is supported by the results of our current observational cohort study of a group of 25 idiopathic PD patients with camptocormia treated with deep brain stimulation (DBS) in the subthalamic nucleus [33]. Based on multifactorial analysis we found that all PD patients with a camptocormia duration of up to 1.5 years showed a beneficial effect of DBS on the axial disorder, patients between 1.5 and *3 years showed mixed results, but none with a camptocormia duration of more than 40 months improved. We hypothesize that PD patients with camptocormia entering a stage of chronic muscle changes as shown in MRI are unlikely to benefit from DBS. The paravertebral muscle changes could be regarded as indicating a certain time point in a process of at present unknown etology and not necessarily a primarily myopathy. Deregulation of lumbar trunk muscles in camptocormia
changes’’. In comparison to the control group, after 3 years patients in our camptocormia cohort started to present fatty degeneration: we describe this stage as ‘‘chronic’’. If confirmed by longitudinal data, these observations suggest that MRI is a useful tool in determining stages of camptocormia. Following this hypothesis, the paravertebral muscles seem to enter a stage of chronic and possibly degenerative change with inferior therapeutic prospects of success after that time. This might explain why different treatment approaches, for example with deep brain stimulation, have shown inconsistent results [31, 32]. Consequently, prospective studies should include the duration of camptocormia in their design.
123
Previous research in camptocormia has focused on the paravertebral muscles. The question of other affected muscles has hardly been addressed [28, 34]. Our data indicate an involvement of the quadratus lumborum muscle in nearly half of the cases for the variables ‘‘edema’’ and ‘‘swelling’’. Rarely, the psoas muscle also seems to be affected. We do not have a conclusive explanation for this at the present time but these findings argue against camptocormia as an isolated process in the paravertebral muscles. Instead our findings suggest a need for a more systemic view in the search for the etiology of camptocormia, considering a deregulation of lumbar trunk muscles by a central mechanism as indicated by the possibly positive effects of DBS.
J Neurol Acknowledgments The authors express their deep thanks to Sari Munser for her valuable help in gathering the data of this study, and Klarissa Hanja Stu¨rner and Walter J. Schulz-Schaeffer for very helpful discussion of the manuscript. The study was supported by a grant from the Medical Faculty of the Christian-Albrechts University of Kiel. Conflicts of interest On behalf of all authors, the corresponding author states that there is no conflict of interest.
References 1. Doherty KM, van de Warrenburg BP, Peralta MC, SilveiraMoriyama L, Azulay JP, Gershanik OS, Bloem BR (2011) Postural deformities in Parkinson’s disease. Lancet Neurol 10:538–549 2. Finsterer J, Strobl W (2010) Presentation, etiology, diagnosis, and management of camptocormia. Eur Neurol 64:1–8 3. Upadhyaya CD, Starr PA, Mummaneni PV (2010) Spinal deformity and Parkinson disease: a treatment algorithm. Neurosurg Focus 28:E5 4. Lepoutre AC, Devos D, Blanchard-Dauphin A, Pardessus V, Maurage CA, Ferriby D, Hurtevent JF, Cotten A, Destee A, Defebvre L (2006) A specific clinical pattern of camptocormia in Parkinson’s disease. J Neurol Neurosurg Psychiatry 77:1229–1234 5. Tiple D, Fabbrini G, Colosimo C, Ottaviani D, Camerota F, Defazio G, Berardelli A (2009) Camptocormia in Parkinson disease: an epidemiological and clinical study. J Neurol Neurosurg Psychiatry 80:145–148 6. Ashour R, Jankovic J (2006) Joint and skeletal deformities in Parkinson’s disease, multiple system atrophy, and progressive supranuclear palsy. Mov Disord 21:1856–1863 7. Abe K, Uchida Y, Notani M (2010) Camptocormia in Parkinson’s disease. Parkinsons Dis 8. Wrede A, Margraf NG, Goebel HH, Deuschl G, Schulz-Schaeffer WJ (2012) Myofibrillar disorganization characterizes myopathy of camptocormia in Parkinson’s disease. Acta Neuropathol 123:419–432 9. Goutallier D, Postel JM, Bernageau J, Lavau L, Voisin MC (1994) Fatty muscle degeneration in cuff ruptures. Pre- and postoperative evaluation by CT scan. Clin Orthop Relat Res 78–83 10. Fuchs B, Weishaupt D, Zanetti M, Hodler J, Gerber C (1999) Fatty degeneration of the muscles of the rotator cuff: assessment by computed tomography versus magnetic resonance imaging. J Shoulder Elbow Surg 8:599–605 11. Hoffmann A, Mamisch N, Buck FM, Espinosa N, Pfirrmann CW, Zanetti M (2011) Oedema and fatty degeneration of the soleus and gastrocnemius muscles on MR images in patients with Achilles tendon abnormalities. Eur Radiol 21:1996–2003 12. Bonneville F, Bloch F, Kurys E, du Montcel ST, Welter ML, Bonnet AM, Agid Y, Dormont D, Houeto JL (2008) Camptocormia and Parkinson’s disease: MR imaging. Eur Radiol 18:1710–1719 13. Schabitz WR, Glatz K, Schuhan C, Sommer C, Berger C, Schwaninger M, Hartmann M, Hilmar Goebel H, Meinck HM (2003) Severe forward flexion of the trunk in Parkinson’s disease: focal myopathy of the paraspinal muscles mimicking camptocormia. Mov Disord 18:408–414 14. Gibb WR, Lees AJ (1988) The relevance of the Lewy body to the pathogenesis of idiopathic Parkinson’s disease. J Neurol Neurosurg Psychiatry 51:745–752
15. Margraf NG, Wrede A, Rohr A, Schulz-Schaeffer WJ, Raethjen J, Eymess A, Volkmann J, Mehdorn MH, Jansen O, Deuschl G (2010) Camptocormia in idiopathic Parkinson’s disease: a focal myopathy of the paravertebral muscles. Mov Disord 25:542–551 16. Elton S. F. a. R (1987) Unified Parkinson’s disease rating scale. In: Fahn S, M. C, Caine DB, Goldstein ME (eds) Recent developments in Parkinson’s disease. Macmillan Health Care Information, Florham Park, NJ, USA pp 153–163, 293–304 17. Beese MS, Winkler G, Nicolas V, Maas R, Kress D, Kunze K, Bucheler E (1993) The diagnosis of inflammatory muscular and vascular diseases using MRT with STIR sequences. Rofo 158:542–549 18. Ranson CA, Burnett AF, Kerslake R, Batt ME, O’Sullivan PB (2006) An investigation into the use of MR imaging to determine the functional cross sectional area of lumbar paraspinal muscles. Eur Spine J 15:764–773 19. Kim H, Lee CK, Yeom JS, Lee JH, Cho JH, Shin SI, Lee HJ, Chang BS (2013) Asymmetry of the cross-sectional area of paravertebral and psoas muscle in patients with degenerative scoliosis. Eur Spine J 22:1332–1338 20. Cohen J (1968) Weighted kappa: nominal scale agreement with provision for scaled disagreement or partial credit. Psychol Bull 70:213–220 21. West GA, Haynor DR, Goodkin R, Tsuruda JS, Bronstein AD, Kraft G, Winter T, Kliot M (1994) Magnetic resonance imaging signal changes in denervated muscles after peripheral nerve injury. Neurosurgery 35:1077–1085 (discussion 1085–1076) 22. McDonald CM, Carter GT, Fritz RC, Anderson MW, Abresch RT, Kilmer DD (2000) Magnetic resonance imaging of denervated muscle: comparison to electromyography. Muscle Nerve 23:1431–1434 23. Tasca G, Monforte M, Iannaccone E, Laschena F, Ottaviani P, Leoncini E, Boccia S, Galluzzi G, Pelliccioni M, Masciullo M, Frusciante R, Mercuri E, Ricci E (2014) Upper girdle imaging in facioscapulohumeral muscular dystrophy. PLoS One 9:e100292 24. Diekman EF, van der Pol WL, Nievelstein RA, Houten SM, Wijburg FA, Visser G (2014) Muscle MRI in patients with longchain fatty acid oxidation disorders. J Inherit Metab Dis 37:405–413 25. Subhawong TK, Wang X, Machado AJ, Mammen AL, Christopher-Stine L, Barker PB, Carrino JA, Fayad LM (2013) 1H Magnetic resonance spectroscopy findings in idiopathic inflammatory myopathies at 3 T: feasibility and first results. Invest Radiol 48:509–516 26. Hayashi D, Hamilton B, Guermazi A, de Villiers R, Crema MD, Roemer FW (2012) Traumatic injuries of thigh and calf muscles in athletes: role and clinical relevance of MR imaging and ultrasound. Insights Imaging 3:591–601 27. Tomasova Studynkova J, Charvat F, Jarosova K, Vencovsky J (2007) The role of MRI in the assessment of polymyositis and dermatomyositis. Rheumatology (Oxford) 46:1174–1179 28. Wunderlich S, Csoti I, Reiners K, Gunthner-Lengsfeld T, Schneider C, Becker G, Naumann M (2002) Camptocormia in Parkinson’s disease mimicked by focal myositis of the paraspinal muscles. Mov Disord 17:598–600 29. Charpentier P, Dauphin A, Stojkovic T, Cotten A, Hurtevent JF, Maurage CA, Thevenon A, Destee A, Defebvre L (2005) Parkinson’s disease, progressive lumbar kyphosis and focal paraspinal myositis. Rev Neurol (Paris) 161:459–463 30. Dietz V (2002) Proprioception and locomotor disorders. Nat Rev Neurosci 3:781–790 31. Asahi T, Taguchi Y, Hayashi N, Hamada H, Dougu N, Takashima S, Tanaka K, Endo S (2011) Bilateral subthalamic deep brain stimulation for camptocormia associated with Parkinson’s disease. Stereotact Funct Neurosurg 89:173–177
123
J Neurol 32. Capelle HH, Schrader C, Blahak C, Fogel W, Kinfe TM, Baezner H, Krauss JK (2011) Deep brain stimulation for camptocormia in dystonia and Parkinson’s disease. J Neurol 258:96–103 33. Schulz-Schaeffer WJ, Margraf NG, Munser S, Wrede A, Buhmann C, Deuschl G, Oehlwein C (2015) Effect of neurostimulation on camptocormia in Parkinson’s disease depends on symptom duration. Mov Disord (epub ahead of print)
123
34. Spuler S, Krug H, Klein C, Medialdea IC, Jakob W, Ebersbach G, Gruber D, Hoffmann KT, Trottenberg T, Kupsch A (2010) Myopathy causing camptocormia in idiopathic Parkinson’s disease: a multidisciplinary approach. Mov Disord 25:552–559
