Use of magnetic resonance imaging and p-31 magnetic resonance spectroscopy to detect and quantify muscle dysfunction in the amyopathic and myopathic variants of dermatomyositis.

ARTHRITIS & RHEUMATISM Volume 38
Number 1, January 1995, pp 68-77
0 1995, American College of Rheumatology
68
USE OF MAGNETIC RESONANCE IMAGING AND
P-3 1 MAGNETIC RESONANCE SPECTROSCOPY TO
DETECT AND QUANTIFY MUSCLE DYSFUNCTION
IN THE AMYOPATHIC AND MYOPATHIC
VARIANTS OF DERMATOMYOSITIS
JANE H. PARK, NANCY J. OLSEN, LLOYD KING, JR, TERRI VITAL, ROBERT BUSE,
SURESH KARL MARTA HERNANZ-SCHULMAN, and RONALD R. PRICE
Objective. To investigate the use of magnetic
resonance imaging (MRI) and P-31 magnetic resonance
spectroscopy (MRS) in characterizing the metabolic and
functional status of muscles in patients with amyopathic
dermatomyositis (DM) and to compare the findings with
those in patients with classic myopathic DM.
Methods. Nine patients with amyopathic DM, 11
patients with myopathic DM, and 11normal individuals
were studied. MRI images of thigh muscles were obtained, and T1 and T2 relaxation times were calculated.
Biochemical status was quantitated with P-31 MRS, by
determining concentrations of phosphate metabolites
during rest and exercise.
Results. Patients with amyopathic DM showed no
muscle inflammation, and MRS data obtained during
rest were normal. During exercise at 25% and 50%
maximum voluntary contractile force, the MRS data
revealed significant differences between amyopathic DM
patients and control subjects indicating inefficient metabolism. In contrast, muscles of patients with myopathic DM showed inflammation and metabolic abnormalities even during rest.
Conclusion. Metabolic deficiencies in patients
with amyopathic DM were unmasked by exercise, sugDr. Olsen’s and Dr. Park’s work was supported in part by
a Clinical Research Grant from the Arthritis Foundation and by NIH
grant AR-43156. Dr. King’s work was supported by funds from the
Bureau of Veterans Affairs.
Jane H. Park, PhD, Nancy J. Olsen, MD, Lloyd King, Jr.,
MD, PhD, T e m Vital, BS, Robert Buse, MD, Suresh Kari, BS, BE,
Marta Hernanz-Schulman, MD, Ronald R. Price, PhD: Vanderbilt
University School of Medicine, Nashville, Tennessee.
Address reprint requests to Nancy J. Olsen, MD, Department of Medicine, T-3219 MCN, Vanderbilt University School of
Medicine, Nashville, TN 37232.
Submitted for publication February 22, 1994; accepted in
revised form July 13, 1994.
gesting that the 2 DM syndromes may share muscle
abnormalities. MRYMRS may be useful in diagnosis
and optimization of treatment.
Dermatomyositis (DM) is an autoimmune disease characterized by a typical erythematous, photosensitive rash and severe proximal muscle weakness
associated with an inflammatory myopathy (1,2). A
subset of DM patients has been described in which
typical cutaneous features are present but no muscle
weakness or only minor muscle involvement is clinically apparent and serum levels of muscle enzymes are
usually normal (3-7). In 1979, Pearson suggested the
term amyopathic DM to refer to this group of DM
patients, in whom no apparent myopathy developed
over at least 2 years of observation (8). Euwer and
Sontheimer (6) conducted a systematic review of amyopathic DM starting with the first report by Krain (4),
who characterized 6 patients without initial muscle
involvement. Recently, Stonecipher et a1 reported on
the evaluation and followup of 13 additional patients
with typical cutaneous findings but absent or minimal
muscle disease (9).
In the patient series described by Krain (4), the
most commonly reported symptoms were lethargy and
fatigue, followed by photosensitivity and then arthralgias. Among the 6 patients described by Euwer and
Sontheimer, the most common symptom was photosensitivity, while lethargy and fatigue were present in
50% of the group (6). This finding of lethargy and
fatigue in the absence of objective weakness or other
supporting diagnostic data suggested the possibility of
subtle defects in muscle function which might not be
detected by the usual clinical evaluation. In order to
examine this possibility, we evaluated the muscles of 9
69
MRUMRS IN AMYOPATHIC DM
Table 1. Clinical features of 9 patients with amyopathic dermatomyositis (DM) studied by magnetic resonance imaging (MRI) and P-31
magnetic resonance spectroscopy (MRS), and of comparison groups*
Agelsedrace
Amyopathic DM
Patient 1
Patient 2
Patient 3
Patient 4
Patient 5
Patient 6
Patient 7
Patient 8
Patient 9
Myopathic DM
Normal subjects
27/F/W
58IFlW
49/F/W
35/F/W
47/F/W
34/F/W
35/F/W
42/F/W
39/F/W
34 f 5/7F, 4M
44 f 5/6F, 5M
Duration
(years)t
Gottron’s
papules
ADL
score
CPK
(IU/ml)$
Prednisone
(mg/day)§
ANA
titer
4
12
10.5
0.87
2.3
0.37
4.5
2.3
1.37
2.2 f 0.5
Yes
Yes
Yes
No
Yes
No
Yes
No
Yes
Rash (all)
No
NA
1
1
1
1
1.13
1
1
NA
1.54 f 0.3
-
27
51/80
53/83
NA
4011 19
78/97
NA
27/39
281197#
916 f 490
3012 10
10.0
2.5
1.25
None
None
4.0
None
5.0
3.0
13 2 5
None
1 :40
<1:40
1:160
1 :40
1:320
1:40
1:160
1:80
1:40
?1:160
<1:40
-
* Values for the myopathic DM patients and the normal subjects are the mean f SEM (n = 1 1 in each group). ADL
Living; CPK = creatine phosphokinase; ANA = antinuclear antibody; NA = data not available.
t From appearance of rash to MRUMRS examination.
$ At or before/after time of MRUMRS examination.
§ Dosage at time of MRIiMRS examination.
7 Subsequently followed up for 2 years without demonstrable weakness.
# Values for this patient obtained at a laboratory with a normal range of 50-160 IU/ml.
patients with amyopathic DM, using noninvasive magnetic resonance imaging (MRI) and P-31 magnetic
resonance spectroscopy (MRS) (10-12). These techniques were utilized because of their high sensitivity
and previous success in characterizing patients with
classic DM (13-15) and in evaluating responses of
these patients to treatment regimens (16-19).
Our results indicate the presence of underlying
metabolic abnormalities in muscles of patients whose
DM has been designated as amyopathic, a finding
which could have an impact on diagnosis and optimization of treatment in individual patients. Furthermore, the data suggest that the 2 variants of DM are
closely related and may share elements of pathophysiology which are variably expressed.
PATIENTS AND METHODS
Patient groups and normal subjects. The amyopathic
DM group included 9 patients and the myopathic group
included 11 patients, all of whom were referred by either
rheumatologists or dermatologists affiliated with Vanderbilt
University Medical School. The patients with amyopathic
DM (Table 1) were all white women ranging in age from 27 to
58 years. Clinical features included disease duration of up to
12 years; 8 of the 9 patients had positive antinuclear antibody, generally in low titers. All amyopathic DM patients
had a characteristic photosensitive, erythematous rash, and
6 had typical Gottron’s papules. Only 1 of these patients was
receiving prednisone at 210 mg daily (mean f SEM for the
group 3 f 1 mg), although 4 patients had received higher
doses of prednisone in the past. Hydroxychloroquine was
being used by 1 patient at the time of study, and 3 patients
had received courses of this drug in the past.
=
Activities of Daily
Data on creatine phosphokinase (CPK) values were
available for 7 of the 9 amyopathic DM patients, and all
values were repeatedly in the normal range with the exception of patient 9, who had normal values on 3 tests and 1
isolated value of 197 IU/ml in a test performed at a laboratory which had a normal CPK range of 50-160 IU/ml. Muscle
biopsies were performed in 2 patients (patients 3 and 5 ) , and
both showed negative results. Five subjects had electromyographic (EMG) testing, and in 2 (patients 3 and 9), the EMG
results were consistent with myositis. One of these EMG
tests had been obtained 10 years previously, and during the
intervening period, no weakness developed (patient 3).
Thus, patient 9 had some findings suggestive of myopathic
DM, but she never had significant muscle weakness. Eight of
the 9 amyopathic DM patients had significant fatigue, which
was quantitated by 3 patients on a 10-cm visual analog scale,
with values of 4,7, and 7 cm, respectively. Nevertheless, all
9 individuals were employed or were active homemakers.
The 7 patients tested had a mean f SEM activities of daily
living (ADL) score of 1.02 2 0.02, indicating good functional
status (20).
The myopathic DM group consisted of 1 black and 10
white patients with ages ranging from 22 to 75 years. There
were 7 women and 4 men. Typical features of a group of
myopathic DM patients in our institution have been detailed
previously (19). For comparative purposes, features of the
myopathic DM group in the present study are summarized in
Table 1. All myopathic DM patients experienced both fatigue and weakness, and the 6 patients tested had a mean
ADL score of 1.54 f 0.30. Nine myopathic DM patients
underwent muscle biopsies, all of which had results that
were consistent with the diagnosis of DM. Seven patients
had EMGs; in 6, the findings were positive for myopathy,
and the results were equivocal in 1. Six subjects were
receiving prednisone at 2 1 0 mg per day (mean f SEM for
the group 13 f 5 mg/day). Four patients were being treated
70
PARK ET AL
Figure 1. T2-weighted images (repetition time/echo time = 2,0001
88) of the thigh muscles of a healthy subject and of patients with
amyopathic or myopathic dermatomyositis (DM). A , Healthy subject. VL = vastus lateralis; VI = vastus intermedius; VM = vastus
medialis; RF = rectus femoris; BF = biceps femoris. B, Patient with
amyopathic DM. There is a homogeneous, low signal with an
intensity equivalent to that of normal muscles. C, Patient with
severe myopathic DM. Signal intensity is significantly increased
(brightness), with irregular distribution in the RF, VL, and VI
(arrows), consistent with inflammation.
with methotrexate or azathioprine. Clinical status was assessed by the referring physicians.
The control group consisted of 11 healthy, nonobese
individuals (6 women and 5 men). Their ages ranged from 27
to 75 years, with a mean k SEM of 44 ? 5 years. Among
these control subjects, no significant differences in the MRS
data between women and men were observed. Furthermore,
comparison of the patient groups with either the normal
female group or the combined female and male group gave
the same statistical results, but, as expected, the P values
were more significant when all 11 control subjects were
included as the comparison group. MRI/MRS examinations
were performed over the same time periods for the control
subjects and study patients. These studies were approved by
the Committee for the Protection of Human Subjects of the
Vanderbilt Institutional Review Board.
Procedure for obtaining T1- and TZweighted images.
The thigh of the subject was positioned in an extremity coil
with a 20-cm diameter. Images were acquired with T1- and
TZweighted spin-echo (SE) sequences in the transverse
orientation with a 1.5T Magnetom (Siemens Medical Systems, Iselin, NJ) (Figure 1). The T1-weighted images were
obtained using a Carr, Purcell, Meiboom, Gill (CPMG) SE
pulse sequence500/22, 5 5 , 88 (repetition time [TR] rnsec/
echo time [TE] m s e c k w i t h a 90" flip angle, 256 x 256
matrix, and a 10-mm section thickness separated by 10-mm
gaps. The T1-weighted (500/22) images with high resolution
were used for identification of specific muscles, including the
vastus lateralis (VL), vastus intermedius (VI), vastus medialis, rectus femoris (RF), and biceps femoris (BF), and for
evaluation of fat distribution. TZweighted images were
acquired with an S E pulse sequence (2,000/22/55/88). Inflammation was identified in diseased muscles as nonhomogeneous regions of high signal intensity (brightness) on
T2-weighted images (2,000/88) (Figure 1C). Regions of inflammation are characterized by high signal intensity on only
the TZweighted images, while areas of fat infiltration are
most easily observed by brightness on T1-weighted images.
Calculation of T1 and T2 relaxation times. For quantitative verification of the visual analyses, T1 and T2 relaxation times were calculated as previously described, using
the data acquired with the 2 CPMG pulse sequences
(13,19,21). The regions of interest in the VL and BF muscles
were selected for evaluation of T1 and T2 values, as reported
MRI/MRS IN AMYOPATHIC DM
previously (13,19). For an individual subject, the mean T1 or
T2 relaxation time was the average of 8 determinations for
the VL and 4 determinations for the BF, which generally had
a more uniform signal intensity. In regions of inflammation,
both T1 and T2 values are elevated above normal due to
increased water content. In contrast, where there is superficial fat or fat infiltration, TI values are very low, but T2
values are elevated.
Exercise protocol. The initial maximum voluntary
contraction (MVC) for each subject’s quadriceps muscles
was determined outside of the magnet, as previously described (13,19). Increasingly heavy weights were strapped to
the ankle of the seated subject, and the foot was raised by
full extension of the lower leg, a motion involving the
quadriceps muscles. Thereafter, the subject was positioned
in the magnet. Spectra were acquired every minute during
the initial 6-minute rest period. A nonmagnetic weight equivalent to 25% of the MVC was then secured on the ankle by
Velcro straps, and the subject lifted the weight by extension
of the lower leg once every 5 seconds over a 6-minute
interval. Next, without an intervening rest period, a 50%
MVC weight was placed on the ankle and exercise was
continued for another 6-minute period. By intermittent lifting of the weight, partial muscle recovery is possible between lifts. This relatively easy exercise protocol was completed by all normal subjects and all but 1 of the 20 study
patients, regardless of the level of fitness or training.
P-31 MRS. Subjects were placed in a supine position
in the magnet, and the thigh was centered and firmly
stabilized with straps. The knee was flexed over an arch, and
the surface coil was secured over the quadriceps muscles.
The magnetic field was adjusted for maximum proton resolution (-0.1 ppm). Spectra of the resting quadriceps muscles
were acquired with a Siemens surface coil (8 cm diameter) at
25.7 MHz with a rectangular pulse of 500 psec duration, a
TR of 3 seconds, and a resolution of 1.95 Hz. The surface
coil monitors a relatively large volume of muscle, and the
spectral data represented a weighted average of normal
muscle or diseased muscle with focal, nonhomogeneous
inflammation. Spectra were acquired during each minute of
the 6-minute periods for rest and for exercise at 25% and
50% MVC (Figure 2). Only the last 4 minutes of each
6-minute exercise period were summed for calculation of
phosphocreatine/inorganic phosphate (PCrPi) ratios and
P-31 metabolite concentrations, in order to assure that
measurements were obtained during periods of steady-state
kinetics. Phasing of spectra and baseline corrections were
performed as previously described (13,19).
Resonance areas of Pi, PCr, and ATP were corrected
for saturation effects, and relative concentrations of P-31
metabolites were determined from the corrected resonance
areas as detailed in previous reports (13,19,22,23). This
quantitative approach has been verified in other studies of
control subjects, patients, and athletes (13,19,22,24-26).
Quality assurance and validation of results were obtained by
evaluation of a Siemens P-31 standard before each MRS
examination. The variability in the resonance area of this
standard over 1 year was +5%. Two control subjects tested
frequently over 10 months had stable metabolite levels at
rest, with a variation of +6%.
The PCr/Pi ratio was determined since it is a sensi-
71
tive indicator of the muscle’s efficiency in generating and
maintaining high-energy phosphate compounds (PCr and
ATP). A lower-than-normal PCr/Pi ratio indicates a defect in
the production and/or utilization of PCr and ATP. To correlate biochemical and functional abnormalities during exercise, the total oxidative capacity (Vmax)of the muscles was
determined according to the equations reported by Chance
et a1 (24), i.e., V,
= V[(0.53 x PCr/Pi) + 11 where V =
work in joules as calculated from the exercise data. V/V,,,
represents the fraction of the oxidative capacity utilized
during exercise.
The energy reserve of the muscles was determined as
the phosphorylation potential (PP), where PP = ATP/(Pi x
ADP). The ATP and Pi concentrations were determined as
described above. Concentrations of cytosolic free ADP are
in pmoles and below the level of sensitivity of in vivo P-31
MRS measurements. Therefore, the ADP concentration was
calculated from the equilibrium equation for the creatine
kinase reaction, as formulated and verified by Chance et a1 (24).
Statistical analysis. Statistical comparisons of data for
the normal subjects and the 2 patient groups were performed
by Student’s 2-tailed I-test. P values less than 0.05 were
considered significant.
RESULTS
T1- and T2-weighted MR images. The T1weighted images of the thigh muscles of all patients
with amyopathic DM showed no abnormalities, as
noted in previous studies of myopathic DM patients
(13,14,19). When the signal intensities of the T2weighted images from a normal subject were compared with those from a patient with amyopathic DM,
no abnormalities were detected in the patient’s muscles (Figures 1A and B). In contrast, the T2-weighted
image from the patient with myopathic DM showed
focal regions of high signal intensity in the quadriceps
muscles (RF, VL, and VI), indicative of inflammation
(Figure 1C). The posterior muscles in the regions of
the hamstrings (BF) and the adductors were generally
less affected than the quadriceps muscles, as noted
previously (13,14,19).
T1 and T2 relaxation time. The qualitative visual analyses of the images were quantified with
calculated TI and T2 relaxation times (Table 2). The
mean TI and T2 values for the VL and BF muscles in
the healthy subjects and the amyopathic DM patients
showed no statistically significant differences. SEM
values were also relatively small in both of these
groups. In the myopathic DM group, the VL muscles
showed the highest mean TI value and the largest
SEM (1,621 +- 84), significantly different from corresponding values in the control group (P < 0.03) and in
the amyopathic DM group (P < 0.01). Inflammation
72
PARK ET AL
0
CONTROL
AMYOPATHIC
MYOPATHIC
REST
REST
PCr
>
E
‘600 r(
Q
c
en
400
ATP
200
0
-
w
I
/I \
P;
5’
I
0
I
-5
I
-10
I
I
-15
d
5’
I
-5
I
-10
I
-15
I
PPm
PPm
Figure 2. Comparison of P-31 spectra of the quadriceps muscles of a control subject at rest (A) and of
patients with amyopathic and myopathic dermatomyositis (DM) (B). The spectrum for the patient with
amyopathic DM shows slightly reduced levels of phosphocreatine (PCr) and ATP. The spectrum for the
patient with myopathic DM and severe proximal muscle weakness demonstrates significantly reduced levels
of PCr and ATP. Pi = inorganic phosphate.
was less severe in the BF and adductor muscles. Only
4 severely ill patients with myopathic DM showed high
signal intensity and T1 values >1,500 msec in the
biceps. For the T2 values, only the data for the VL
muscle were significantly higher in myopathic DM
patients compared with either the control group (P<
0.01) or the amyopathic DM group (P < 0.02). For
subcutaneous fat, the T1 value was low (330 msec) and
the T2 value high (55 msec). Since the T1 and T2
values were both found to be high in the diseased VL
muscle (Table 2), the increased signal (brightness) in
Figure 1C cannot be attributed to fatty infiltration.
PCr/Pi ratios during rest and exercise. The mean
PCr/Pi ratios in the amyopathic DM group were not
significantly different from the control values either
during rest or during exercise at 25% MVC (Figure 3).
However, during exercise at 50% MVC, the PCr/Pi
ratio was significantly lower for the amyopathic DM
group compared with the controls (P< 0.05), indicating metabolic deficiencies which became apparent at
the higher work load. In muscles of the myopathic DM
patients, the mean PCr/Pi ratios were significantly
lower than those in the normal or the arnyopathic DM
group during rest (P < 0.01 and P < 0.005, respectively) and during exercise at 25% MVC (P < 0.001
and P < 0.01, respectively). The low PCr/Pi ratios can
be explained on the basis of the relative concentrations
of P-3 1 metabolites, discussed below.
Table 2. T1 and T2 values in the thigh muscles of patients with
amyopathic and myopathic dermatomyositis (DM) and in normal
controls *
T2 (msec)
T1 (msec)
Group
VL
BF
BF
VL
~
Controls
1,384 f 29
(n = 10)
Amyopathic DM 1,302 f 42
(n = 9)
Myopathic DM 1,621 2 84t
(n = 9)
* Values
1,316 f 50 38 t 0.4
39
2
1,397 f 50 40
38
f 0.6
42
f 2.0
1,454 2 47
f
0.8
51 f 4.0$
0.6
are the mean f SEM. P values were calculated by
student’s t-test. VL = vastus lateralis muscle; BF = biceps femoris
muscle.
t P < 0.03 versus control group; P < 0.01 versus amyopathic DM
group.
$ P < 0.01 versus control group; P < 0.02 versus amyopathic DM
PUP.
73
MRUMRS IN AMYOPATHIC DM
Normal
0 Amyopathic
ae
m-
B Myopathic
6-
'=
0
4 -
**
**
2-
0REST
25% MVC
50% MVC
Figure 3. PCdPi ratios in the quadriceps muscles of control subjects and of patients with amyopathic or
myopathic DM during rest and exercise (25% and 50% maximum voluntary contraction [MVC]). The PCrlPi
ratios were calculated from spectroscopic determinations of the concentrations of PCr and Pi (Table 3) and
are shown as the mean and SEM. + = P < 0.05 versus normal subjects; * = P < 0.01 versus normal subjects
and P < 0.005 versus patients with amyopathic DM; ** = P < 0.001 versus normal subjects and, for 25%
MVC only, versus patients with amyopathic DM. See Figure 2 for other definitions.
Concentrations of P-31 metabolites during rest
and exercise. The mean PCr/Pi ratios for the amyopathic DM group during rest or during exercise at 25%
MVC were not significantly different from normal
values, because the mean PCr and Pi concentrations
were both somewhat lower in the muscles of patients
with amyopathic DM than in controls (Table 3). At
higher work loads (50% MVC), the proportionately
greater loss of PCr from the amyopathic muscles was
reflected in a statistically significant decrease in the
PCr/Pi ratio when compared with normal muscles (P<
0.05). In contrast, very low PCr/Pi ratios, largely due
to low levels of PCr, were seen in the myopathic
muscles at all stages of the protocol (Figure 3 and
Table 3). For example, during rest, the mean concentration of PCr in myopathic muscles was approximately 40% lower than the normal value (P < 0.001)
and 30% below that of the amyopathic DM group (P<
0.005), whereas the Pi concentration remained equivalent to control values. During exercise at 25% MVC,
the PCr decrease was proportionately greater in the
myopathic than in the amyopathic muscles, and the P
value for the comparison of PCr/Pi ratios between
myopathic DM patients and control subjects was decreased accordingly (P < 0.001).
In general, the concentrations of ATP (pphosphate peak) followed the same pattern as those
for PCr (Table 3). Thus, the average ATP levels in the
muscles of patients with amyopathic DM were somewhat lower than normal, but the differences were not
statistically significant. In contrast, the relative ATP
levels in the muscles of patients with myopathic DM
during rest were well below the normal or amyopathic
DM values (by 35% and 25%, respectively) (P< 0.001
and P < 0.005, respectively). During exercise (25%
MVC), ATP was lost from the myopathic muscles but
not from the amyopathic muscles, and the difference
between the 2 patient groups reached a greater level of
statistical significance (P < 0.001).
V,,,,
V/V,,,,
and phosphorylation potential.
The metabolic data were correlated with functional
capacity by calculation of the V,,
and VIV,,, values
(Table 4). The work (V) performed by the amyopathic
DM and myopathic DM patients at 25% MVC was not
significantly different. However, the calculated V values for both patient groups were substantially lower
than that for the controls (P < 0.005 and P < 0.001,
respectively). The total oxidative capacity (V,,,) was
the parameter that most clearly differentiated the 3
PARK ET AL
74
Table 3. Levels of P-31 metabolites in muscles of patients with amyopathic and myopathic dermatomyositis (DM) and normal controls, during
rest and exercise*
25% MVC
Rest
50% MVC
Group
Pi
Controls
(n = 11)
Amyopathic DM
3.120.2
24.521.1
5.420.3
6.020.4
21.1t1.1
5.420.3
7.9r0.5
19.2r1.2
5.3t0.3
2.5 f 0.2$
22.1 t 1.4
4.7 t 0.2
5.9 2 0.6
18.2 f 1.0
4.7 2 0.3
8.5 f 0.6
15.6
4.6
f
0.2
2.6
15.4 t 1.41
3.5 t 0.35
5.4 2 0.5
11.6 f 1.17
3.1 I 0 . 3 7
7.0
10.6 f 1.11
3.4
f
0.31
(n
PCr
ATPt
Pi
PCr
ATPt
Pi
PCr
f
1.0
ATPt
= 9)
Myopathic DM
(n = 11)
f
0.2
* Values are the mean f SEM mmoles/kg (wet weight). MVC = maximum voluntary contraction; Pi
phosphocreatine.
t Calculated as the pphosphate peak, which contained no adenine nucleotide impurities.
$ P < 0.05 versus control group.
1 P < 0.001 versus control group; P < 0.005 versus amyopathic DM group.
7 P < 0.001 versus control group and versus amyopathic DM group.
groups of subjects. The oxidative capacity of the
amyopathic DM and myopathic DM patients was,
respectively, 30% and 55% below that of the normal
subjects (P < 0.002 and P < 0.001, respectively).
These decreased V,, values in the case of the amyopathic DM group were due to the low V values, while
for myopathic patients a combination of low V values
and low PCr/Pi ratios contributed to the reduced V,,.
data points for all subjects are
Individual V and V,
values at the higher
shown in Figure 4. The V,
exercise level (50% MVC) also statistically differentiated the 3 groups of subjects. Thus, the MRS procedures were able to detect subtle abnormalities in the
muscles of patients with amyopathic DM, particularly
during exertion.
The calculation of V/V,,,,
which represents
the fraction of the oxidative capacity utilized during
exercise, gave values of 35%, 38%, and 47% for the
controls, amyopathic DM patients, and myopathic DM
f 0.5
=
inorganic phosphate; PCr
patients, respectively. There was no significant difference between the control and the amyopathic DM
groups, but a significant difference was noted between
each of these groups and the myopathic DM group
(P < 0.01 for each). The V/V,,, ratio in the amyopathic DM group was normal because both the V and
the V,
values were significantly lower in these
patients than in controls. Another useful measure is
the phosphorylation potential, which is defined as the
free energy available for phosphorylation. The mean -+
SEM phosphorylation potential of muscles of patients
in the amyopathic DM group (28.7 iz 5.0) was almost
twice as high as that of muscles of patients in the
myopathic DM group (16.8 ? 2.4; P < 0.05). The V,
and PP values indicate that the oxidative capacity of
the amyopathic DM group was significantly greater
than that of the myopathic DM group, providing
further quantitative substantiation of differences in
muscle function.
_ _ _
Table 4. Total oxidative capacity (Vmax) and Dhosohorvlation Dotential (PP) in patients with
amyopathic and myopathic dermatomyositis (DM) and normal controls, during exercise at 25%
maximum voluntary contraction*
~
~~
Group
V
Controls
(n = 11)
Amyopathic DM
(n = 9)
Myopathic DM
81.4 f 3.5
235.2 f 12.9
0.35
60.4
4.8$
163.5 t 14.01
0.38 2 0.03
51.1 f 7.011
107.1 2 13.0#
0.47
(n
&
Vmax
VNmax
?
?
0.01
0.02#
= 11)
* Values are the mean 2 SEM.
t Expressed as mmoles~'.
$ P < 0.005 versus control group.
5 P < 0.002 versus control group.
=
ll P < 0.001 versus control group.
# P < 0.001 versus control group; P < 0.01 versus amyopathic DM group.
** P < 0.002 versus control group; P < 0.05 versus amyopathic DM group.
PPt
53.9 2 11.1
28.7
f
5.0
16.8 f 2.4**
75
MRUMRS IN AMYOPATHIC DM
B
A
1
300 0
lZO
100
-2
80
.-c
a
1
Q
c
i
-
.-2 200
-
E
E
1
-ga 60 -
-a
\
0
:
-J0
7
0
2
:
100 -
0
E
>
2o
1
0
0
Control
Amyopathic
Myopathic
Control
Amyopathic
1
Myopathic
Figure 4. Individual values for work load (V) (A) and total oxidative capacity (VmaX)(B) in control subjects,
amyopathic dermatomyositis (DM) patients, and myopathic DM patients studied during exercise at 25%
maximum voluntary contraction. Although there is some overlap among the groups, there is a clear
downward trend in these values as the severity of muscle involvement increases. Statistical analyses of these
data are shown in Table 4.
DISCUSSION
The criteria for classification of patients with
amyopathic DM have been the subject of debate for 2
decades (1,6,9). One reason for this may be that
patients with minimal or no clinical evidence of muscle
disease present problems with regard to diagnostic
procedures such as biopsy or EMG. If the serum levels
of muscle enzymes such as CPK are not elevated, the
risk and discomfort of invasive procedures may not
seem warranted in clinical practice (6). The question
then arises as to whether the diagnosis of amyopathic
DM can be accurately made in the absence of muscle
biopsy and EMG (1,2). Precise evaluation in these
cases assumes importance as it may affect the decision
as to whether to initiate aggressive immunosuppressive therapy. Another difficulty is that a patient’s
condition may change from an amyopathic to a myopathic state very slowly over time, or there may be a
flare of the myopathic state that was not detected by
clinical or laboratory investigations. As a partial solution to these ongoing dilemmas, we have used noninvasive MRI/MRS procedures to acquire unique data
on the morphologic/metabolic status of muscles of
patients with presumed amyopathic DM (10-12).
In earlier investigations, we and others have
shown that MRI is a very sensitive method for localization and characterization of the nonhomogeneous
inflammation of myopathic DM (13-15). In the present
study, none of our patients with amyopathic DM
showed visible inflammation on T2-weighted images of
the thigh muscles. In a recent longitudinal study of 11
patients with myopathic DM, inflammation was initially observed by MRI in most patients, but during
treatment, the extent and severity of inflammation
decreased at varying rates, which were not necessarily
concordant with rates of change in CPK levels or the
more prolonged metabolic abnormalities detected by
P-31 MRS (19). In cases of discordant longitudinal
MRI/MRS data, our experience suggests that MRS
provides the more useful information for evaluation of
disease status (1619). These findings may explain why
some DM patients become slowly but progressively
weaker based on physical testing but show little or no
evidence of inflammation or muscle damage.
Although the PCr/Pi ratio is not diagnostic for
specific muscle diseases, it serves as an important
measure of defective bioenergetics in given muscle
disorders (26-28). This ratio is particularly useful for
analysis of data acquired during exercise and recovery, when muscles are more dependent on oxidative
rather than glycolytic pathways for energy production.
In the present group of patients with amyopathic DM,
the PCr/Pi ratio was normal during rest and during
exercise at low levels (25% MVC) (Figure 3). However, at 50% MVC, the PCr/Pi ratio in the amyopathic
DM patients was significantly below normal, indicating deficient bioenergetic status. The inability to generate and/or utilize PCr and ATP at the higher work
76
loads is consistent with subjective fatigue and lethargy
noted in earlier reports (4,6) and reported by 8 of our
9 amyopathic subjects. In contrast, the patients with
myopathic DM showed serious abnormalities even
during rest, as demonstrated by the low PCr/Pi ratios
resulting from low PCr levels. For the patients with
amyopathic DM, whose energy metabolism was statistically superior to that of the myopathic DM group,
the relatively high PCr levels were the major factor in
determining the high values for the ratio. The Pi level
remained more constant, with essentially equivalent
values in all 3 groups of subjects.
Underlying mechanisms responsible for the observed biochemical abnormalities are not defined by
the present studies. However, it is of interest to note
that muscle biopsy samples from DM patients have
been shown to have significantly reduced capillary
density, even in the absence of inflammation (29).
Reduced capillary density may lead to decreased tissue oxygenation, which might in turn contribute to
metabolic abnormalities such as decreased ATP synthesis in the Krebs cycle during exercise and recovery.
It should be noted that in previous studies of patients
with myopathic DM, resolution of inflammation on MRI
was not always concordant with normalization of MRS
values, suggesting that metabolic changes are, at least
in part, independent of inflammatory processes (19).
We have previously demonstrated that patients
with myopathic DM have significantly lower levels of
PCr than normal subjects, and that during the course
of treatment, increases in PCr levels correlate with
improved clinical status (16-19). The relative changes
in PCr levels were greater than those of ATP; therefore, PCr was the better quantitative indicator of
changes in disease status. In the amyopathic DM
patients in the present study, PCr levels were lower
than, but not statistically different from, normal values
(Table 3). When the 2 DM groups were compared, the
amyopathic DM patients were found to have significantly greater levels of PCr than the myopathic DM
patients. Moreover, PCr levels were maintained more
effectively during exercise in the amyopathic DM
patients than in the myopathic DM patients, consistent
with the former group’s greater capacity for daily
activities and employment. ATP concentrations in the
amyopathic DM patients were slightly lower than, but
not statistically different from, control values, and
were significantly higher than values in the weaker
patients with myopathic DM.
Subtle defects in amyopathic muscles were best
detected by correlations between functional capacity
PARK ET AL
during exercise and biochemical parameters, calculated as the total oxidative capacity (V,,,) shown in
Table 4. The data clearly demonstrated the superior
oxidative capacity (endurance) of the controls as compared with the amyopathic DM patients. Thus, the
symptoms of fatigue and lack of endurance noted by
patients with amyopathic DM were substantiated by
calculations. In turn, the superior capacity
the V,
for sustained energy generation by amyopathic DM
patients as compared with the myopathic DM group
was also documented with the calculations of V,,
values and phosphorylation potential. The severe
weakness and fatigue of the myopathic DM patients
correlated with the very low V,,, and PP values
(Table 4).
In practical terms, these data indicate that
patients with amyopathic DM may perform efficiently
at low work loads, but due to lack of oxidative
capacity and endurance, they would not be able to
sustain the activity as long as normal individuals.
Thus, during the MRS protocol with exercise of short
duration, muscle metabolism (PCr/Pi ratios) of these
patients may appear normal at low work loads. In
contrast, myopathic DM patients are not able to
efficiently generate energy for muscle contraction at
any time during the exercise protocol, or in some
cases, even during rest. Thus, on a quantitative basis,
the patients with myopathic DM have more severe
metabolic abnormalities and dysfunction than do those
with the amyopathic variant.
During the course of this study, patient 5
(Table 1) showed a sudden flare in disease activity.
The precipitous onset of myopathy in this patient
required hospitalization and substantial changes in her
treatment regimen, as previously reported (12,19).
After the development of myopathy, the MRI examination revealed inflammation of the thigh muscles, and
MRS data showed the expected serious decline in her
metabolic status. Her response to immunosuppressive
therapy was monitored by MR procedures to determine when the medication should be tapered (12,19). It
was important to accurately evaluate these metabolic
changes in order to determine as soon as possible what
therapeutic interventions might be indicated to avoid
further functional deterioration.
In conclusion, this investigation shows the
unique capability of MRVMRS examinations to detect
subtle defects in patients with amyopathic DM. These
metabolic defects may be responsible for clinical findings such as fatigue and may provide objective data on
this otherwise purely subjective symptom. The nonin-
MRUMRS IN AMYOPATHIC DM
vasive MR tests can be readily repeated for longitudinal evaluations and are very useful in making the
decision to start, modify, or stop immunosuppressive
therapy, thereby minimizing complications and maximizing therapeutic efficacy (12,19).
ACKNOWLEDGMENTS
These studies would not have been possible without
the support of our refemng physicians, Drs. John S. Johnson, Joseph Huston, Porter Meadors, and s. Bob0 Tanner.
We would also like to thank the many people who assisted
with the performance of these studies, including Drs. Bany
Allen, Peter Baumgartel, Toby Cole, Lewis Gilpin, Daniel
Golwyn, John McGue, and Philip Moyers, Ms Kara Rader,
and Mr. Kevin King. Dr. Lauren Adams assisted in acquisition of clinical data on the patients. We appreciate the
helpful comments of Dr. Charles R. Park during the preparation of the manuscript. We are also grateful for the
administrative assistance of Ms Pat Runsvold and the photographic contributions of Mr. John Bobbitt.
REFERENCES
1. Bohan A, Peter JB: Polymyositis and dermatomyositis. N Engl
J Med 292:344347, 1975
2. Dalakas MC: Polymyositis, dermatomyositis and inclusion body
myositis. N Engl J Med 325:1487-1498, 1991
3. Bohan A, Peter JB, Bowman RL, Pearson CM: A computerassisted analysis of 153 patients with polymyositis and dermatornyositis. Medicine (Baltimore) 56:255-286, 1977
4. Krain LS: Dermatomyositis in six patients without initial muscle involvement. Arch Dermatol 3:241-245, 1975
5 . Euwer RL, Sontheimer RD: Amyopathic dermatomyositis (dermatomyositis sine myositis). J Am Acad Dermatol 24959-966,
1991
6. Euwer RL, Sontheimer RD: Amyopathic dermatomyositis.
J Am Acad Dermatol 26506-508, 1992
7. Callen JP, Jorizzo JL: Amyopathic dermatomyositis (dermatomyositis sine myositis). J Am Acad Dermatol 26505-506, 1992
8. Pearson CM: Polymyositis and dermatomyositis, Arthritis and
Allied Conditions. Ninth edition. Edited by DJ McCarty. Philadelphia, Lea and Febiger, 1979
9. Stonecipher MR, Jorizzo JL, White WL, Walker FO, Prichard
E: Cutaneous changes of dermatomyositis in patients with
normal muscle enzymes: dermatomyositis sine myositis? J Am
Acad Dermatol 28:951-956, 1993
10. Vital TV, Olsen NJ, King L, Baumgartel P, Stein S, Partain CL,
Park JH: New classification of myositis confirmed with MRI and
P-3 I magnetic resonance spectroscopy: dermatomyositis and
amyopathic dermatomyositis (abstract). SOCMagn Reson Med
11:2727, 1992
11. Buse R, Park JH, Olsen NJ, Vital TL, Kari S, Ryder NM, King
LE: Quantitative comparisons of oxidative capacity in myopathic and amyopathic patients. Radiology 183:1182, 1993
12. King LE, Park JH, Adams L, Vital TL, Olsen NJ: Evaluation of
muscles in a patient with suspected amyopathic dermatomyositis by magnetic resonance imaging and P-31 spectroscopy. J Am
Acad Dermatol 30: 137-138, 1994
13. Park JH, Vansant JP, Kumar NG, Gibbs SJ, Curvin MS, Price
RR, Partain CL, James AE: Dermatomyositis: correlative mag-
77
netic resonance imaging and P-3 1 spectroscopy for quantitative
characterization of inflammatory disease. Radiology 177:473479, 1990
14. Fraser DD, Frank JA, Dalakas M, Miller FW, Hicks JE, Plotz
P: Magnetic resonance imaging in the idiopathic inflammatory
myopathies. J Rheumatol 18: 1693-1699, 1991
15. Newman ED, Kurland RJ: P-31 magnetic resonance spectroscopy in polymyositis and dermatomyositis: altered energy utilization during exercise. Arthritis Rheum 35: 199-203, 1992
16. Hernanz-Schulman M, Park JH, Olsen NJ, Stein SM, Kumar N,
Park CR, Price RR, Partain CL, James AE: MRI and P-31 MR
spectroscopy for evaluation of therapeutic responsiveness of
dermatomyositis patients (abstract). SOCMagn Reson Imaging
1:162, 1991
17. Park JH, Stein SM,Hernanz-Schulman M, Olsen NJ, Price RR,
Partain CL, James AE: MR imaging and P-31 spectroscopy for
long-term evaluation of immunosuppressant therapy for inflammatory myopathies (abstract). Radiology 181:107, 1991
18. Ryder N, Park JH, Olsen NJ, Finney A, Buse RO, Lawton A,
Hernanz-Schulman M: Advantages of MRI and P-31 MR spectroscopy for longitudinal evaluation of prednisone treatment of
juvenile dermatomyositis (abstract). SOC Magn Reson Med
12:1213, 1993
19. Park JH, Vital TL, Ryder NM, Hernanz-Schulman M, Partain
CL, Price RR, Olsen NJ: Magnetic resonance imaging and P-31
magnetic resonance spectroscopy provide unique quantitative
data useful in the longitudinal management of patients with
dermatomyositis. Arthritis Rheum 37:736-746, 1994
20. Pincus T, Callahan LF, Brooks RH, Fuchs HA, Olsen NJ, Kaye
JJ: Self-report questionnaire scores in rheumatoid arthritis
compared with traditional physical, radiographic, and laboratory measures. Ann Intern Med 110:259-266, 1989
21. Kjos BO, Ehman RL, Brant-Zawadski M, Kelley WM, Norman
D, Newton TH: Reproducibility of relaxation times and spin
density from routine MR imaging sequence. Am J Neuroradiol
6:271-283, 1985
22. Taylor DJ, Fore PJ, Styles P, Gadian DG, Radda GK: Bioenergetics of intact human muscles: a P-31 nuclear magnetic
resonance study. Mol Biol Med 1:77-94, 1983
23. Hams RC, Hultman E, Nordesjo LO: Glycogen, glycolytic
intermediates and high energy phosphates determined in biopsy
samples of musculus quadriceps femoris of man at rest. Scand J
Clin Lab Invest 33:385-394, 1974
24. Chance B, Leigh JS, Clark BJ, Maris J, Kent J, Nioka S, Smith
D: Control of oxidative metabolism and oxygen delivery in
human skeletal muscle: a steady-state analysis of the work/
energy cost transfer function. Proc Natl Acad Sci U S A
82:8384-8388, 1985
25. Park JH, Brown RL, Park CR, Cohn M, Chance B: Energy
metabolism of the untrained muscle of elite runners as observed
by P-31 magnetic resonance spectroscopy: evidence suggesting
a genetic endowment for endurance exercise. Proc Natl Acad
Sci U S A 85:878&8784, 1988
26. Chance B, Younkin DP, Kelly R, Bank WJ, Berkowitz HD,
Argov Z, Donlon E, Boden B, McCully K, Buist NMR, Kennaway N: Magnetic resonance spectroscopy of normal and
diseased muscles. Am J Med Genet 25:659-679, 1986
27. Bottomly PA: Human in vivo nuclear magnetic resonance
spectroscopy in diagnostic medicine: clinical tool or research
probe? Radiology 170:1-15, 1989
28. Radda GK, Taylor DJ: Applications of nuclear magnetic resonance in pathology. Int Rev Exp Pathol 27: 1-58, 1985
29. Emslie-Smith A, Engel AG: Microvascular changes in early and
advanced dermatomyositis: a quantitative study. Ann Neurol
27:343-356, 1990