| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Musculoskeletal Imaging |
1 From the Departments of Radiology (S.F.K., C.C.H., S.H.N., Chen-Chang Lee, Chih-Chia Lee), Thoracic and Cardiovascular Surgery (M.J.H.), Neurology (T.K.L.), and Public Health and Biostatistics (M.C.C.), Chang Gung University, College of Medicine, Chang Gung Memorial Hospital-Kaohsiung Medical Center, Kaohsiung, Taiwan; and Department of Chemistry, National Sun Yat-Sen University, Kaohsiung, Taiwan (L.L.). Received March 31, 2007; revision requested May 30; revision received June 11; accepted July 18; final version accepted September 7. S.F.K. supported by grant NSC 93-2314-B-182A-085 from the National Science Council, Taiwan. Address correspondence to S.F.K. (e-mail: sfatko{at}adm.cgmh.org.tw).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
Materials and Methods: With institutional review board approval and informed consent, resting and dynamic 31P MR spectroscopy were performed in 14 healthy volunteers (five men, nine women; mean age, 33 years; range, 23–48 years) and 16 MG patients (six men, 10 women; mean age, 37 years; range 18–50 years) before and after thymectomy. Patients were stratified into groups according to the modified Osserman classification: mild-MG group (classes I–IIA) and moderate-to-severe–MG group (classes IIB–IV). Variables compared among the three groups (Kruskal-Wallis test) included the inorganic phosphate (Pi)–adenosine triphosphate (ATP) (Pi/ATP) ratio, phosphocreatine (PCr)-ATP (PCr/ATP) ratio, Pi/PCr ratio, muscle pH at resting and at end-exercise 31P MR spectroscopy, rate constant for PCr recovery (kPCr), and maximum oxidative capacity (Vmax). These variables were also compared in MG patients before and after thymectomy (Wilcoxon signed rank test).
Results: There were no significant differences in resting Pi/ATP, PCr/ATP, and Pi/PCr ratios and resting muscle pH among the three groups (control group, 14; mild-MG group, nine; moderate-to-severe–MG group, seven). Comparison of the control group with the mild-MG group and comparison of the mild-MG group before thymectomy with the mild-MG group after thymectomy showed no significant differences in end-exercise Pi/ATP, PCr/ATP, and Pi/PCr ratios; end-exercise muscle pH; kPCr; and Vmax. Compared with the control and mild-MG groups, the moderate-to-severe–MG group had significantly higher end-exercise Pi/ATP and Pi/PCr ratios and significantly lower end-exercise muscle pH, kPCr, and Vmax before thymectomy (P 
.001), but these values showed significant restoration to normal after thymectomy (P = .018).
Conclusion: Mild-MG group patients have muscle oxidative metabolism similar to that of healthy control subjects, whereas moderate-to-severe–MG group patients have impaired Vmax during exercise and a noticeable shift to glycolytic metabolism, but these abnormalities are reversible after thymectomy.
© RSNA, 2008
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
The absolute amount and concentration of glycogen, glycolytic intermediates, and high-energy phosphates in the muscles can be quantified in the biopsy samples (10). On the other hand, phosphorus 31 (31P) magnetic resonance (MR) spectroscopy is capable of direct, continuous, in vivo noninvasive monitoring of bioenergetics during rest, exercise, and recovery from exercise (10–15). In vivo 31P MR spectroscopy has been reported to be useful in the evaluation of various neuromuscular disorders, such as mitochondrial or metabolic myopathies, myotonic dystrophy, and motor neuron denervating disorders (16–18). Although the presence of enzymatic adenosine triphosphatase alteration with disturbances in the muscular energy metabolism and mitochondrial dysfunction in MG have been demonstrated with histopathologic or immunochemical means (19–23), only a few studies with 31P MR spectroscopic evaluation of resting muscle have been reported (16–18). Thus, the purpose of our study was to prospectively assess muscle metabolism in MG patients before and after thymectomy by using 31P MR spectroscopy.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
Inclusion criteria for MG patients were as follows: (a) fulfillment of the criteria for the diagnosis of MG, including detailed history and characteristic clinical presentations and findings at physical examination validated by an experienced neurologist (T.K.L., with 15 years of experience) before 31P MR spectroscopy; (b) positive results of adjunctive tests, including the neostigmine test and a decremental electromyographic response at repetitive nerve stimulation and/or a test for acetylcholine receptor–specific autoantibodies in the patient's serum; and (c) the presence of thymic hyperplasia or thymoma at multidetector computed tomography of the thorax.
The severity of MG was determined according to the modified Osserman classification (1,8,9), as follows: class I, myasthenia limited to the ocular region only; class IIA, mild generalized MG with ocular involvement; class IIB, moderately generalized MG with ocular involvement and mild bulbar symptoms; class III, acute severe MG with severe bulbar symptoms; and class IV, delayed severe generalized MG.
31P MR Spectroscopy
Image-selected in vivo 31P MR spectroscopy was performed with a 1.5-T body imaging and/or spectroscopic instrument (Intera; Philips Medical Systems, Best, the Netherlands). The patients were placed in a supine feet-first position. A 60-mm single-loop spectroscopic surface coil (P-60; Philips Medical Systems) was attached to one of the straps of a hook-and-loop fastener (Velcro; Velcro, Manchester, NH) and was then fixed under the patient's calf in the location where the muscle was widest. The leg with the coil was subsequently placed in the central region of the MR bed to avoid field inhomogeneity and shimming problems. For resting 31P MR spectroscopy, survey images were obtained with the preset procedure with three stacks for transverse, sagittal, and coronal images. Then planscan (a tool of the MR system) volume was adjusted to be above the coil center and close to it, with a volume size of 40 x 40 x 80 mm3, and the long axis was along the leg axis. Thereafter, automatic shimming and manual tuning were performed.
Instead of exciting a voxel by using gradients and selective pulses, 31P MR spectroscopy relied on a specially tuned phosphorus 60-mm single-loop surface coil to excite a volume. An adiabatic single-forward half-passage excitation was used to apply a uniform 90° radiofrequency pulse to all the tissues in a cylinder with the same diameter (60 mm) as the coil and a depth approximately equal to the diameter. In addition, since the 31P T2 values are so short that even the shortest echo times are too long with the use of a technique with a 90°–180° radiofrequency pulse, free induction decay (FID) acquisition was used in 31P MR spectroscopy so that the signal could be acquired as soon as possible after the end of the excitation. Other imaging parameters for resting 31P MR spectroscopy included a repetition time of 6000 msec, offset frequency of –100 Hz, offset frequency effective at chemical shift only, FID volume selection, shim reference tissue as muscle, no presaturation, spectral bandwidth of 1500 Hz, data samples of 1024, and 64 measurements for one spectrum (number of rows, one).
For the dynamic 31P MR spectroscopic study, the recovery-kinetics approach was used to determine the rate of recovery of high-energy phosphates. The localization technique, FID volume selection, and 90° radiofrequency pulse single half-passage adiabatic excitation were similar to those of resting 31P MR spectroscopy. Other imaging parameters included repetition time of 5000 msec, offset frequency of –100 Hz, offset frequency effective at chemical shift only, FID volume selection, shim reference tissue as muscle, no presaturation, spectral bandwidth of 1500 Hz, data samples of 1024, automatic time series measurements, automatic time series start times, and two measurements for each spectrum (number of rows, 60). The temporal resolution was 10 seconds per spectrum.
Before the examination, the volunteers and the patients were taught to perform an isotonic exercise involving repeated plantar flexion of the ankle at a frequency of one flexion per 2 seconds. An MR technician accompanied each subject throughout the whole procedure to ensure that this exercise was optimally performed for 3 minutes. A total of six, 18, and 36 31P MR spectra were obtained during the 1 minute of rest, 3 minutes of exercise, and 6 minutes of recovery, respectively, with an overall acquisition time series of 10 minutes.
Postprocessing and Spectral Analysis of 31P MR Spectroscopic Data
The FID data were transferred to a personal computer (ASUS M5200A; ASUSTek Computer, Taipei, Taiwan), and further analyses were performed in consensus (S.F.K., C.C.H., and L.L., with 8, 5, and 20 years of experience in MR spectroscopy, respectively) with commercially available curve-fitting software (NutsPro-NMR Utility Transform Software; Acorn NMR, Livermore, Calif). Exponential multiplication corresponding to 10-Hz line broadening and conversion of zero-filling with 2048 samples was applied to FID data. After phasing and baseline correction, peaks corresponding to inorganic phosphate (Pi) and phosphocreatine (PCr), as well as the 
, 
, and β phosphate groups of adenosine triphosphate (ATP), were identified on the basis of their chemical shift locations. Areas of each spectral peak on resting and dynamic 31P MR spectroscopic images were quantified by using integration of the Lorentzian line with a least-squares method fitting routine. Recovery-versus-time curves for PCr were also created for determination of the rate constants.
Equations and Calculations
The calculations (10–13) were performed (S.F.K. and C.C.H.) in consensus, with the peak areas corrected for the appropriate saturation factors. The concentrations measured in millimoles per liter were calculated, with the assumption that ATP concentrations are a certain value for all tests and subjects, with the following equation: ATP = 8.2 mmol/L in resting skeletal muscles.
The sum of Pi and PCr concentrations was also assumed to remain constant during exercise (11–14), with the calculation of the following equation:
 |
The intracellular pH, or pHint, was calculated (11,15) from the 
chemical shift, or pH, where pH is the maximal frequency shift of the Pi peak during end exercise compared with resting status, by using the following equation:
 |
The postexercise PCr recovery and maximum oxidative capacity (Vmax) (11–14) were determined by using the following equations:
 |
 |
PCr is the remainder from PCr at the end of exercise subtracted from PCr at rest, with Vmax and the rate constant for PCr recovery (kPCr).
Clinical Follow-up
The MG patients were regularly followed up at 1, 3, 6, and 12 months after thymectomy by the surgeon (M.J.H., with 22 years of experience) and the neurologist (T.K.L., with 15 years of experience), and the severity of MG was reassessed (M.J.H. and T.K.L.) in consensus. Follow-up 31P MR spectroscopy was performed within 6–9 months after thymectomy.
Statistical Analysis
The age, sex, and body mass index between the healthy control subjects and MG patients were compared by using the Mann-Whitney U test for age and body mass index and the Fisher exact test for sex. For the statistical analysis in our study, we stratified our patients with classes I–IIA MG as the mild-MG group and those with classes IIB–IV MG as the moderate-to-severe–MG group. The metabolic variables were quantified, and values were expressed as the mean ± standard deviation unless otherwise stated. Our three groups were compared on the basis of the Pi/ATP, PCr/ATP, and Pi/PCr ratios; muscle pH; and äpH at resting and end-exercise 31P MR spectroscopy; kPCr; and Vmax. Statistical analyses were performed (S.F.K. and M.C.C.) in consensus with software (SYSTAT, version 11.0; SPSS, Chicago, Ill). A nonparametric approach was used in the analysis because of the limited sample sizes of the three groups. The Kruskal-Wallis test was used to determine the differences.
A difference with P .05 was considered statistically significant. The Mann-Whitney U test was employed for intergroup comparisons. For each of these intergroup comparisons (control group vs mild-MG group, control group vs moderate-to-severe–MG group, mild-MG group vs moderate-to-severe–MG group) for a total of three comparisons for each variable of interest, the significance level was set at approximately .02 (.05 ÷ 3 = .017) because multiple comparisons tend to increase type I errors. The preoperative and postoperative data were also compared with each other in the mild-MG group and in the moderate-to-severe–MG group by using the Wilcoxon signed rank test.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
Clinical Follow-up
Of our nine patients in the mild-MG group, three had class I MG, and the other six had class IIA MG. All three patients with class I MG and four of six patients with class IIA MG had a complete recovery after surgery, whereas the remaining two patients with class IIA MG had clinical improvement and MG was reclassified as class I. On the other hand, of our seven patients in the moderate-to-severe–MG group, four had class IIB MG, two had class III MG, and one had class IV MG. All seven patients had clinical improvement after thymectomy, with complete recovery in three patients with class IIB MG, improvement and reclassification from class III MG to class I MG in one patient, improvement and reclassification from classes IIB–IV MG to class IIA MG in two patients, and improvement and reclassification from class III MG to class IIB MG in the remaining one patient. Follow-up 31P MR spectroscopy was performed within 6–9 months (average, 7 months) after thymectomy.
Spectra
Analysis of our data about resting 31P MR spectroscopy revealed no significant differences with respect to resting Pi/ATP, PCr/ATP, and Pi/PCr ratios and resting muscle pH between control subjects and all 16 MG patients before and after thymectomy (Figure; Tables 1, 2). At dynamic 31P MR spectroscopy (Table 2), there were no significant differences between patients with mild MG and control subjects in end-exercise Pi/ATP, PCr/ATP, and Pi/PCr ratios; end-exercise muscle pH; kPCr; and Vmax. No significant changes were observed before and after thymectomy in the mild-MG group. On the other hand, in comparison with the control and mild-MG groups, the moderate-to-severe–MG group showed significantly higher end-exercise Pi/ATP and Pi/PCr ratios and significantly lower end-exercise muscle pH, kPCr, and Vmax (P .001). There was significant restoration of these dynamic 31P MR spectroscopic variables after thymectomy in the moderate-to-severe–MG group (P = .018).
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
Abnormally high Pi and/or, less often, abnormally low PCr levels (high Pi/PCr ratio) at resting 31P MR spectroscopy have been reported in 35%–85% of mitochondrial encephalomyopathies (16–18). Furthermore, high Pi/ATP, high Pi/PCr, and low PCr/ATP ratios at resting 31P MR spectroscopy have also been reported to be helpful in the assessment of muscle fiber distribution, muscle injury, or muscle inflammation (11–18). Type I fibers (predominantly oxidative metabolism, slow twitch and fatigue resistant) are characterized by relatively high Pi and relatively low PCr levels (ie, high Pi/PCr ratio) in comparison with type II fibers (predominantly glycolytic metabolism, fast high twitch and fatigue sensitive) (24). The type I and II fibers are generally supposed to be balanced in distribution in healthy individuals (15,24). In our study, despite that significant differences were encountered in the moderate-to-severe–MG group with exercise, there were no metabolite ratio differences between groups at rest. These results imply that the muscle fibers of MG patients with various levels of severity may not be different from those of healthy individuals in regard to distribution and may lack any inflammatory changes.
Dynamic 31P MR spectroscopy allows measurements of oxidative metabolism by using either a steady-state or recovery-kinetics approach (11–15,17,24). Steady-state measurement necessitates a steady state of exercise at difficult work levels during a relatively long data acquisition period, and such a requirement can be difficult for MG patients (11–13). In contrast, recovery-kinetics measurement enables determination of the rate of recovery of high-energy phosphates and the calculation of kPCr, Vmax, and the changes in muscle pH independent of the exercise intensity (12,13,17,24). The values of kPCr and Vmax may reflect the potential of oxidative phosphorylation, whereas the changes in muscle pH during exercise can provide an estimate of the rate of glycolysis (11–19). In our study, adequate data were successfully acquired in a 10-second time resolution of recovery test in all healthy control subjects and in 86% (19 of 22) of MG patients.
Normal PCr recovery depends on normal perfusion (or oxygen delivery) and normal mitochondrial function. Limited oxidative metabolism with impaired PCr recovery has been described in patients with heart failure and severe chronic obstructive pulmonary diseases, presumably ascribed to abnormal muscle blood flow and relatively poor oxygen delivery, respectively (25,26). Impaired ATP synthesis has been described in patients with Child-Pugh class B and C cirrhosis, presumably because of a decreased number of mitochondria in skeletal muscle (27). Our study found no significant difference in the Vmax between control subjects and patients with mild MG. However, the patients with moderate to severe MG had significantly lower kPCr and Vmax, reflecting relatively poor PCr recovery and impaired Vmax.
In our study, there were no significant differences between the healthy control subjects and MG patients with respect to age, sex, and body mass index. Most of the MG patients were 20–40 years of age, and none of them had pulmonary, cardiac, or peripheral vascular diseases or cirrhosis of the liver. Furthermore, since recovery measurement is independent of work intensity during exercise and only submaximal exercise was performed, regional perfusion was likely to be adequate. Therefore, the reduced muscle oxidative capacity demonstrated in our patients with moderate to severe MG may reflect the presence of mitochondrial dysfunction. In addition, significant improvement of oxidative metabolism in these patients was found after extended thymectomy, and this finding suggests that such mitochondrial dysfunction may be at least partially reversible. Nevertheless, elucidation of the precise relationship between thymomatous lesions and impaired Vmax in patients with moderate to severe MG requires further study.
Another major energy pathway through the glycolytic mechanism, with formation of lactate and production of protons, can be assessed by changes in muscle pH during exercise (12–15). The chemical shift between the PCr and the Pi peaks at 31P MR spectroscopy is determined by the pH environment of the Pi. Since most of the Pi is intracellular, a calculation of the pH from the Pi value can represent intracellular acidity (11,15). Sprint athletes compared with endurance athletes demonstrated a twofold faster rate of pH decline after a short duration of vigorous exercise (28). Some mitochondrial diseases may manifest lactic acidemia and abnormal exercise-induced intracellular acidosis at 31P MR spectroscopy (29). However, bioenergetic heterogeneities of mitochondrial encephalomyopathies with a relative resistance to intracellular acidosis have been documented in some patients (29,30). Our study findings revealed not only significantly higher end-exercise Pi/ATP and Pi/PCr ratios but also a significantly greater decline of end-exercise muscle pH in patients with moderate to severe MG compared with patients with mild MG and control subjects. These results suggest that reduced oxidative metabolism in such patients may result in a shift to glycolytic metabolism in energy production for exercise.
Our study had limitations. First, only 16 patients with various degrees of severity of MG with postoperative follow-up for 6–9 months could be recruited. Thus, for most end points when no statistically significant differences were found, this did not necessarily mean that no differences existed because the sample size was so small. Further studies in larger series of patients with a follow-up period of longer duration are necessary to clarify the clinical and prognostic implications of findings at 31P MR spectroscopy in patients with MG. Second, in our study, we did not perform histopathologic or electron microscopic assessments of the mitochondria, and, thus, correlation of mitochondrial abnormalities with 31P MR spectroscopic data was not possible. Third, although weakness is more prominent in the proximal muscles of patients with mild and moderate MG, the calf muscle was used for provision of a larger image volume and for expediency of data acquisition during exercise. Finally, because it was not ethical to alter potentially beneficial MG treatment regimens for the purpose of our study, the effect of different treatments on individual patients could not be assessed.
In summary, the results of this 31P MR spectroscopic study indicated that patients with moderate to severe MG exhibited muscular oxidative metabolic abnormalities, which occur during exercise with a shift to glycolytic metabolism, and these abnormalities and shift were reversible after thymectomy. These findings support the proposition that mitochondrial dysfunction may play a role in the muscle weakness in patients with moderate to severe MG.
|
ADVANCE IN KNOWLEDGE |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
|
FOOTNOTES |
|---|
Abbreviations: ATP = adenosine triphosphate • FID = free induction decay • kPCr = rate constant for PCr recovery • MG = myasthenia gravis • PCr = phosphocreatine • Pi = inorganic phosphate • Vmax = maximum oxidative capacity
See also Science to Practice in this issue.
Author contributions: Guarantors of integrity of entire study, S.F.K., C.C.H.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, S.F.K., C.C.H., Chih-Chia Lee; clinical studies, S.F.K., M.J.H., S.H.N., Chen-Chang Lee, T.K.L., L.L.; statistical analysis, S.F.K., C.C.H., M.C.C., L.L.; and manuscript editing, all authors
Authors stated no financial relationship to disclose.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  D. Lindquist What Can 31P MR Spectroscopy Tell Us about Muscle Disease? Radiology, April 1, 2008; 247(1): 1 - 2. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
