Serum creatine kinase (CK) levels and the grade of signal intensity of findings in antimyosin scintigraphy, T2-weighted magnetic resonance imaging (MRI), and T1-weighted MRI at diagnosis and at follow-up in 17 patients with myositis. Pre indicates before treatment; post, after treatment.
Increased uptake of antimyosin antibody in scintigraphy (A), increased signal intensity in T2-weighted magnetic resonance imaging (MRI) (B), and pathological signal intensity in T1-weighted MRI (C) in the thighs of patient 2 at diagnosis. At follow-up the findings in scintigraphy (D) and T2-weighted MRI (E) decreased, but the lesions in T1-weighted MRI (F) are slightly more pronounced.
Accumulation of antimyosin antibody (A) and increased signal intensity in T2-weighted sequences in magnetic resonance imaging (MRI) (B) in the calves of patient 17 at diagnosis when the patient was bedridden, and 8 months later at follow-up (C and D) when she was mostly independent in activities of daily living and able to walk small distances without support. T1-weighted MRI findings remained normal (E and F).
Löfberg M, Liewendahl K, Lamminen A, Korhola O, Somer H. Antimyosin Scintigraphy Compared With Magnetic Resonance Imaging in Inflammatory Myopathies. Arch Neurol. 1998;55(7):987-993. doi:10.1001/archneur.55.7.987
To compare indium In 111 altumomab pentetate–labeled antimyosin scintigraphy with magnetic resonance imaging (MRI) in the diagnosis and follow-up of patients with myositis.
Design and Methods
Sixteen patients with polymyositis and 1 patient with dermatomyositis, all verified with biopsy samples, were examined during diagnostic evaluation with antimyosin antibody scintigraphy and low-field MRI of the thighs and calves using T1- and T2-weighted sequences. Both examinations were repeated 6 to 22 months after therapeutic intervention with anti-inflammatory drugs. The performance of the 2 methods for the assessment of the severity of muscle inflammation was evaluated using comparison with clinical examination and the serum creatine kinase level.
At diagnosis all patients had increased uptake of antimyosin antibody in the thighs and/or calves. In T2-weighted MRI images, increased signal intensity changes reflecting intramuscular edema and inflammation were seen in all patients in at least 1 muscle group in the thighs or calves. After anti-inflammatory drug therapy, the mean uptake of antibody and the mean signal intensity changes in T2-weighted MRI had decreased. However, in T1-weighted MRI the signal intensity changes reflecting intramuscular fatty degeneration were more pronounced in the follow-up study. The level of serum creatine kinase had decreased markedly by the second examination except in 1 patient who also had more accumulation of antibody in the calves after than before treatment. The clinical condition improved in 8 patients and remained unchanged in 9 patients.
Antimyosin scintigraphy and T2-weighted MRI are feasible tools for the detection and follow-up of lesions in patients with myositis. Scintigraphy findings correlate with serum creatine kinase activity and seem to reflect disease activity better than T2-weighted MRI changes, whereas secondary degenerative intramuscular lesions are only detectable using T1-weighted MRI.
POLYMYOSITIS AND dermatomyositis are acquired muscle diseases with proximal muscle weakness, atrophy, and tenderness as the main clinical findings. Increased sarcoplasmic enzyme activities in the serum and typical findings from needle electromyography are the other characteristic features. The diagnosis is confirmed histologically when inflammatory infiltrates are found in skeletal muscle. In polymyositis the inflammatory cells are mainly T lymphocytes attacking muscle cells, whereas in dermatomyositis the pathomechanism is humoral immunity and the cells found in perivascular infiltrates in muscle are typically B lymphocytes.1 Inclusion body myositis differs from other inflammatory myopathies clinically by simultaneous affection of distal muscles and histologically by the presence of vacuolated muscle fibers and intracellular amyloid deposits or 15- to 18-nm tubulofilaments shown using electron microscopy.2
Imaging methods have been used to characterize the distribution of mesenchymal changes in myositis.3 Muscle wasting in advanced polymyositis can be detected using computed tomography and ultrasonography. In childhood dermatomyositis, computed tomography can identify subcutaneous and intramuscular calcifications, which may also result in foci of increased muscle echogenicity shown in ultrasonography. Computed tomography and ultrasonography are, however, less sensitive than magnetic resonance imaging (MRI) in detecting muscle edema, a typical sign of active inflammation in skeletal muscle.3
Scintigraphy using technetium Tc 99m chelates or gallium citrate Ga 67 has been used in the diagnosis of inflammatory myopathies. In studies including 6 or more patients with active myositis, the sensitivity of these scintigraphies was about 90%.3 Increased uptake of indium In 111 altumomab pentetate–labeled antimyosin antibody was reported by De Geeter et al4 in 2 patients with dermatomyositis and by us5 in 2 patients with polymyositis and 1 patient with dermatomyositis. In this study we compared 111In-labeled antimyosin scintigraphy with MRI in the diagnosis and follow-up of 17 patients with inflammatory mypathies.
All patients except patient 17 had slowly progressive muscle weakness for several months before the diagnostic examinations. Most patients could walk without assistance but needed the aid of a handrail while climbing stairs. Patient 14 felt weakness only in her thighs while walking longer distances. Patients 8 and 11 needed support while walking. Patient 17, the most affected, was diagnosed as having an aggressive and rapidly progressing disease, and the patient was bedridden and needed assistance in all activities of daily living. The age of the patients ranged from 36 to 81 years (mean [±SD], 62±13 years; SEM, 3 years).
In all patients the serum creatine kinase (CK) level was increased, and needle electromyography showed changes typical of myositis or myopathy. The diagnosis was confirmed using muscle biopsy samples; all patients except patient 2 had inflammatory myopathy with T-lymphocyte predominance typical of polymyositis. Some histological features of inclusion body myositis (rimmed vacuoles or inclusion bodies) were seen in 5 of these patients (patients 5, 7, 10, 12, and 16) but the diagnostic criteria for inclusion body myositis2 were not fulfilled. Patient 2 had skin lesions and B-lymphocyte predominance shown through analysis of muscle biopsy samples and was therefore considered to have dermatomyositis. Magnetic resonance imaging and antimyosin antibody imaging were performed during the diagnostic evaluation before onset of drug therapy and repeated after 6 to 22 months (mean [±SD], 10±4 months; SEM, 1 month).
The first choice of an anti-inflammatory drug was prednisone, with a starting dose of 1 mg/kg. The initial response to corticosteroid therapy was good in all patients as judged by a decrease in the serum CK level, but in some patients this therapy caused adverse effects. Therefore, to maintain good clinical response, sufficient levels of prednisone could not be continued for more than 4 to 6 weeks. At this point azathioprine, 100 mg/d (in patient 9, only 50 mg/d), was added to the therapy in 9 patients, and the corticosteroid dosage was lowered to a minimum. All patients were examined and treated at the Helsinki University Central Hospital, Helsinki, Finland, in accordance with the Declaration of Helsinki II.
Magnetic resonance imaging of patients' thighs and lower legs was performed with a 0.1-T MRI system (Merit, Picker Nordstar Inc, Helsinki). Two sets of axial slices were obtained, one set positioned on thigh musculature and the other set on calf muscles, with a slice thickness of 10 mm, an interslice gap of 10 mm, and an imaging matrix of 256×256. The field-of-view was 460×460 mm. The number of slices was adjusted to cover the whole length of the muscle bundles in the craniocaudal direction, usually with a set of 16 to 18 slices. Both T1- and T2-weighted images were acquired with identical slice positioning. For T1-weighted images a partial saturation sequence with a repetition time of 150 milliseconds, an echo time of 15 milliseconds, and 6 signal acquisitions was used, and for T2-weighted images, a repetition time of 1500 milliseconds, an echo time of 60 milliseconds, and 2 signal acquisitions were used. In patient 1, imaging only of the calves was performed at diagnosis.
Two radiologists (A.L. and O.K.) independently analyzed the MRI scans in both phases of the study without knowledge of the clinical status, serum CK values, or scintigraphy findings of the patient. Synergistic muscle groups, ie, extensors and flexors, were analyzed as a single entity for both thighs and calves. Interobserver agreement was first assessed using κ statistics; to facilitate further statistical evaluation, consensus grades were obtained using joint reanalysis and discussion in the cases in which there was initial disagreement between the 2 readers.
The MRI signal intensity changes were semiquantitatively graded as follows. The T1-weighted image sets: grade 0, normal muscle signal intensity and homogeneous hypointense signal contrasting sharply with subcutaneous and intermuscular fat; grade 1, hyperintense, patchy but not confluent intramuscular signal in a single muscle of a synergistic muscle group; grade 2, hyperintense, patchy, and confluent intramuscular signal, with widespread but less than total involvement of the synergistic muscle group; and grade 3, homogeneously hyperintense, confluent intramuscular signal, with total involvement of the synergistic muscle group.
Because fat suppression sequences were unavailable in the current imaging software, the T2-weighted image sets were analyzed taking into account the signal changes caused by fatty degeneration in the exactly corresponding T1-weighted slices: grade 0, normal and homogeneous hypointense signal; grade 1, limited changes and slightly increased morphologically localized intramuscular signal intensity in T2-weighted images compared with T1-weighted images; grade 2, moderate changes morphologically more widespread, but patchy increases in intramuscular signal intensity in T2-weighted images; and grade 3, widespread changes and extensive, homogeneous, and widespread increase in intramuscular signal intensity in T2-weighted images.
For scintigraphy, a 0.5 mg-dose of murine antimyosin Fab fragments conjugated with diethylenetriaminepentaacetic acid (Myoscint, Centocor BV, Leiden, the Netherlands) and labeled with 111In (range, 65-140 MBq; mean, 102 MBq) was administered intravenously. This antibody, which was raised against cardiac myosin, has been found to cross-react with skeletal myosin.4- 6 No adverse effects were observed in the patients administered millipore-filtered labeled antibody. Labeling was performed according to the instructions of the manufacturer (Centocor BV). The labeling efficiency is 96% to 99% in thin-layer chromatography studies in our laboratory.
Scintigraphy was performed with a single detector gamma camera (Picker SX 300), connected to a computer (PDP 11/74), or a single-head gamma camera (Picker Prism XP 1500) connected to a computer (Odyssey VP computer, Picker International, Philadelphia, Pa). The cameras were equipped with medium-energy parallel-hole collimators. Patients' lower extremities were imaged 3 to 4 hours and 20 to 24 hours after the injection of radiolabeled antimyosin by collecting 400000 to 600000 counts per image (imaging time, 15 minutes) using 2 energy windows (171 keV and 245 keV; 20% windows). The lesion-to-nonlesion activity uptake ratio was higher at 24 hours than at 4 hours. The scintigrams acquired at 24 hours were selected for comparison with MRI.
For grading of immunoscintigraphic findings in the lower extremities, the intensity of radioactivity in the muscular areas above and below the knee were compared with that in the main blood vessels in thighs and calves, respectively. Blinded interpretation of images taken at 20 to 24 hours was performed by 2 investigators (M.L. and K.L.) with a 4-point scale (0 to +++) Interobserver variation was assessed using κ statistics. Discrepancies of interpretation were resolved using joint reanalysis and discussion, and consensus values were used for statistics. For grade 0, the finding was considered normal if muscular radioactivity was homogeneously distributed and the level of radioactivity was lower than in the blood vessels; grade+, patchy or focal distribution of muscular radioactivity, the level of which was similar to that in the blood vessels; grade++, patchy or focal distribution of muscular radioactivity, the level of which was higher than in the blood vessels; and grade+++, more widespread intense focal distribution of muscular radioactivity, the level of which was much higher than in the blood vessels. A standard source of radioactivity (0.74 MBq/100 mL of 111In) was inserted between the thighs and calves to facilitate the comparison of intensities of pathological findings in different gamma camera fields.
The nonparametric Spearman rank correlation test was used for correlation of MRI signal intensity changes, serum CK level, and uptake of antimyosin antibody. Kappa statistics were used to assess interobserver grading differences between the 2 observers.
By the time of the follow-up MRI and scintigraphy studies, the serum CK levels had decreased from 1664±564 U/L to 238±59 U/L (mean±SEM) (Figure 1). The response to drug therapy was generally moderate as judged by the improvement of muscle strength. None of the patients' conditions worsened during treatment and the need for assistance decreased or remained unchanged. The degree of disability evaluated by the Vignos grade7 initially created for muscular dystrophies remained unchanged in 9 patients and improved in 8 patients.
Before treatment all patients had increased signal intensity on T2-weighted MRI scans in at least 1 muscle group. On T1-weighted MRI scans 5 patients had no signal abnormalities at diagnosis, and most of the other patients had more widely distributed or more severe lesions in the T2-weighted than in the T1-weighted images. The results are presented in detail in Table 1.
On the follow-up T2-weighted MRI scans, the signal intensity changes had decreased, but in only 1 patient had all lesions disappeared. On the contrary, on T1-weighted MRI scans, the average grade of signal intensity changes had increased by the second examination. In patients 3, 13, and 17, T1-weighted MRI findings had remained normal. The results of the follow-up examination are presented in Table 2.
At diagnosis all patients had varying degrees of increased uptake of antimyosin antibody in the lower limbs. In some patients, the uptake was more pronounced in the calves than in the thighs, and 7 patients also had increased uptake in the buttocks.
In the follow-up scans the muscular uptake of antimyosin antibody had decreased, but all patients still had increased antimyosin uptake in at least 1 limb. However, patient 8, whose CK level had not changed during therapy, actually had more accumulation of antibody in her calves in the second study. Magnetic resonance imaging and scintigraphy findings of patient 2 and patient 17 at diagnosis and at follow-up are presented (in Figure 2 and Figure 3, respectively).
There was a tendency for positive correlation between T2-weighted MRI imaging and uptake of antimyosin antibody before treatment (Rs=0.36; P=.15) when using a mean value for all the muscle groups studied in each patient. At follow-up this correlation could not be found. Uptake of antimyosin correlated with serum CK levels both at diagnosis (Rs=0.48; P=.05) and at follow-up (Rs=0.71; P=.004). There was no correlation between findings of T2-weighted MRI and CK activity. The agreement between the mean pretherapy and posttherapy uptake of antibody and mean signal intensity on T2-weighted MRI is shown in Figure 1.
The κ values for interobserver variation were 0.71 (good) for MRI grading and 0.55 (moderate) for scintigraphy grading.
Both scintigraphy with antimyosin antibody and T2-weighted MRI were capable of detecting inflammatory lesions in muscles of patients with polymyositis and dermatomyositis. Scintigraphy and T2-weighted MRI also provided information on the efficacy of drug therapy to suppress inflammation in these patients. However, the good correlation between the uptake of antimyosin antibody and serum CK activity was not found between T2-weighted MRI and CK levels, favoring scintigraphy as the method of choice for the assessment of the severity of inflammation. The plausible explanation for increased uptake of antimyosin antibody is that the inflammatory process disturbs the integrity of muscle cell membranes and reveals the insoluble myosin to the antibody. Notably, the progression of intramuscular fatty degeneration causing increased signal intensity on T1-weighted MRI scans was not reflected in antimyosin scintigraphy. A further advantage of MRI compared with scintigraphy was the possibility of distinguishing between lesions in extensor and flexor muscles.
Kaufman et al8 identified abnormalities in T1-weighted MRI signal intensity and fat replacement that were claimed to correlate with clinical disease activity in 13 patients with polymyositis and dermatomyositis. Fujino et al9 concluded from studies on 8 patients with active myositis that the high-intensity lesions in the T2-weighted images represented both inflammation and edema of muscles and that, therefore, T2-weighted imaging appeared to be more important for the diagnosis of inflammatory myopathies than T1-weighted images. Reimers et al,10 assessing the value of gadolinium–gadopentetate dimeglumine in inflammatory myopathies of adults, found T2-weighted images to be more sensitive than contrast-enhanced T1-weighted images. Of 58 patients, the sensitivity of MRI for detection of abnormal muscle structure was about 80%, with this method being especially sensitive for the detection of edemalike abnormalities.
Only a few patients with inflammatory myopathies have been serially followed up with MRI. Fraser et al11 reported that the increased signal intensity in fat suppressive images paralleled disease activity using needle biopsy in the 3 patients studied. When 10 patients with polymyositis were examined twice with MRI, 3 patients showed definitive changes at follow-up in imaging compared with the first study, an observation that correlated with clinical and biochemical findings in 2 of them.12 In juvenile dermatomyositis, changes in T2-weighted imaging that rapidly parallel disease activity or with a delay of 2 months have been reported.13- 15
Antimyosin scintigraphy, which has an established role in the evaluation of cardiac muscle damage, has also been used to detect traumatic skeletal muscle lesions.5,16 The accumulation of antimyosin antibody is focal and often quite intensive in rhabdomyolysis, compared with the more diffuse or patchy antibody accumulation in myositis. In muscular dystrophies the better preserved muscle bulk in the calves accumulated more antibodies than the atrophic thigh muscles, which had more pronounced increases in signal intensity in T1-weighted MRI, reflecting replacement of muscle by fat.5 Similar observations on the effect of fatty degeneration on MRI findings were made in our study on patients with myositis.
The follow-up of patients with inflammatory myopathies is based on clinical examination and serum CK levels.1 Repeated muscle biopsy is not used often since it is an invasive and expensive method for evaluating the efficacy of therapy. Muscle imaging has the advantage of noninvasiveness and the capability to reveal changes in all muscles of interest. Antimyosin scintigraphy correlated with serum CK levels, both at diagnosis and at follow-up, in our patients with myositis. T1-weighted MRI images were required for revealing secondary changes, particularly fatty degeneration, in patients with chronic myositis. In conclusion, our study has demonstrated that relevant information on the extent and severity of lesions in patients with myositis can be obtained with antimyosin scintigraphy and low-field MRI, reducing the need for follow-up muscle biopsy.
Accepted for publication December 1, 1997.
This work was supported by a grant from the Paulo Foundation, Helsinki, Finland (Dr Löfberg).
Reprints: Mervi Löfberg, MD, Institute of Neurosciences, Department of Neurology, Helsinki University Central Hospital, 00290 Helsinki, Finland.