Author + information
- Received September 23, 2005
- Revision received December 20, 2005
- Accepted December 30, 2005
- Published online June 6, 2006.
- David C. Isbell, MD⁎,
- Stuart S. Berr, PhD†,‡,
- Alicia Y. Toledano, ScD‖,
- Frederick H. Epstein, PhD†,‡,
- Craig H. Meyer, PhD‡,
- Walter J. Rogers, PhD⁎,†,
- Nancy L. Harthun, MD§,
- Klaus D. Hagspiel, MD†,
- Arthur Weltman, PhD⁎ and
- Christopher M. Kramer, MD⁎,†,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Christopher M. Kramer, University of Virginia Health System, Departments of Medicine and Radiology, Lee Street, Box 800170, Charlottesville, Virginia 22908
Objectives In this study we intend to characterize phosphocreatine (PCr) recovery kinetics with phosphorus-31 (31P) magnetic resonance spectroscopy in symptomatic peripheral arterial disease (PAD) patients compared with control subjects and determine the diagnostic value and reproducibility of this parameter.
Background Due to the inconsistent relationship between flow and function in PAD, novel techniques focused on the end-organ are needed to assess disease severity and measure therapeutic response.
Methods Fourteen normal subjects (5 men, age 45 ± 14 years) and 20 patients with mild-to-moderate symptomatic PAD (12 men, age 67 ± 10 years, mean ankle brachial index 0.62 ± 0.13) were studied. Subjects exercised one leg to exhaustion while supine in a 1.5-T magnetic resonance scanner using a custom-built plantar flexion device. Surface coil-localized, free induction decay acquisition localized to the mid-calf was used. Each 31P spectrum consisted of 25 signal averages at a repetition time of 550 ms. The PCr recovery time constant was calculated by monoexponential fit of PCr versus time, beginning at exercise completion.
Results Median exercise time was 195.0 s in normal subjects and 162.5 s in PAD patients (p = 0.06). Despite shorter exercise times in patients, the median recovery time constant of PCr was 34.7 s in normal subjects and 91.0 s in PAD patients. Area under the receiver-operating characteristic curve was 0.925 ± 0.045. Test-retest reliability was excellent.
Conclusions The PCr recovery time constant is prolonged in patients with symptomatic PAD compared with normal subjects. The method is reproducible and may be useful in the identification of disease. Further study of this parameter’s ability to track response to therapy as well as its prognostic capability is warranted.
Peripheral arterial disease (PAD) is a condition characterized by atherosclerotic obstruction of the arteries supplying the lower limbs. While peripheral blood flow at rest may be adequate to preserve tissue viability, there is insufficient flow to meet increased metabolic demands with exercise. Currently, between 8 and 12 million Americans are affected by PAD, and the incidence is expected to rise as the population ages (1). Aside from the increased risk of both cerebrovascular and coronary events, PAD patients may experience pain, diminished exercise capacity, and significant tissue loss, with some ultimately requiring amputation (2). Despite a growing appreciation of its contribution to vascular morbidity and mortality, PAD remains a clinical entity that is underdiagnosed, undertreated, and understudied (3,4).
Modalities currently utilized for the diagnosis and study of PAD have a number of well-recognized limitations (5,6). Although the presence of macrovascular atherosclerosis can generally be established, these tests provide little information on the downstream consequences of proximal arterial obstruction. Furthermore, increasing evidence suggests that factors intrinsicto skeletal muscle influence the physiologic impact of arterial insufficiency and may contribute to the poor correlation observed between hemodynamic parameters and peak functional capacity in many patients with claudication (7–9). Because of the sometimes inconsistent relationship between flow and function, diagnostic techniques that focus on the skeletal muscle may be better suited than conventional methods for characterizing disease severity, monitoring its progression, and objectively measuring therapeutic response.
For over three decades, phosphorus-31 magnetic resonance spectroscopy (31P MRS) has been utilized for the non-invasive study of muscle metabolism. Through the measure of high-energy phosphorylated compounds, insight into cellular energetics can be achieved with precision (10,11). Although the kinetics for a number of compounds can be characterized, phosphocreatine (PCr) recovery after exercise may be the most important parameter in patients with PAD. By relying on both the arterial oxygen supply andthe tissue’s capacity to utilize substrate, measurements of PCr regeneration incorporate many of the adaptive and/or maladaptive cellular changes developed in response to a chronic limitation in blood flow (12). The purpose of this study was to examine the utility of magnetic resonance spectroscopy for determining PCr recovery time constant after exercise in the magnetic resonance scanner in patients with established PAD as compared with normal subjects.
Patients between the ages of 30 to 85 years with symptoms of intermittent claudication without critical limb ischemia and an ankle brachial index (ABI) between 0.4 and 0.9 were eligible for this study. Normal human subjects without risk factors for atherosclerosis were recruited from the community to serve as control subjects. The protocol was approved by the Human Investigation Committee at the University of Virginia, and all participants signed informed consent.
31P spectra were acquired with a Siemens Sonata 1.5-T magnetic resonance scanner (Siemens Medical Solutions, Erlangen, Germany) using a single-pulse, surface coil localized, 512-ms free induction decay acquisition with the coil centered on the mid-calf. A standard 31P surface coil in the patient table was employed. A coil signal intensity profile with a phosphorus phantom demonstrated excellent signal within 7 to 8 cm. This distance fully encompasses the gastrocnemius and soleus of the average patient’s calf, the muscles worked during plantar flexion.
Siemens spectroscopy software on a Leonardo workstation (Siemens Medical Solutions) was used to estimate relative concentrations of adenosine triphosphate (ATP), PCr, and inorganic phosphate (Pi). Free induction decays were multiplied by an exponential with a time constant of 110 ms and zero-filled from 1,024 to 2,048 points before Fourier transformation. Phase and baseline corrections were made, followed by Lorentzian fitting and integration of the spectral peaks. The PCr recovery time constant (R), which characterizes the PCr recovery time, was calculated by monoexponential fit of the PCr levels versus time (beginning at the end of exercise), using the equation: y= a(b− e−t/R), where yis the fitted PCr level, tis the time after cessation of exercise, and aand bare the fitted relative levels of the PCr recovery curve. Tissue pH was calculated from the chemical shift between Pi and PCr (δ) using the following formula:
All subjects were placed supine in the magnetic resonance scanner with the calf at the magnet isocenter. Monitoring of the electrocardiogram and blood pressure was performed with an InVivo 3155MVS (Intermagnetics Companies, Orlando, Florida). After shimming to minimize non-uniformity of the magnetic field, five baseline spectra were acquired. A magnetic resonance image-compatible plantar flexion exercise device was then affixed to the magnetic resonance table. The subject was instructed to push the pedal at a steady rate until exhaustion or limiting symptoms developed, exercising the calf muscles. After four preparation pulses, 25 signal averages at a repetition time of 550 ms were acquired for a total acquisition time of 16 s per spectrum. Two spectra were acquired before the end of exercise, followed by the acquisition of 18 spectra commencing at the end of exercise. Three PAD patients and eight normal subjects returned at 315 ± 188 days after initial study for studies of test-retest reliability.
Subject characteristics were summarized as mean and standard deviation for continuous variables, and by frequencies for other variables. Characteristics were compared between control subjects and patients using ttests for continuous variables and chi-square tests for other variables. Exercise times, PCr recovery time constant, pH, and PCr/ATP were summarized as median and 25th percentile and 75th percentile. The Wilcoxon rank sum test was used to compare the centers of the PCR recovery times, pH, and PCr/ATP for patients versus control subjects. A smooth binormal receiver-operating characteristic (ROC) curve was estimated. Reproducibility was analyzed using the method of Bland and Altman (13). Statistical analyses were performed in Splus 2000 Professional Release 2 (MathSoft, Inc., Cambridge, Massachusetts).
Twenty patients (12 men) and 14 control subjects (5 men) were studied. Demographic characteristics are summarized in Table 1.The patients were older than the control subjects (67 ± 10 years vs. 45 ± 14 years, respectively, p < 0.0001). Baseline clinical characteristics of PAD patients are provided in Table 2.Among PAD patients, the average ABI was 0.62 ± 0.13. Mean systolic blood pressure of patients was 141 ± 19 mm Hg. Results from non-invasive vascular studies are shown in Table 3.Eleven of the patients underwent X-ray angiography, and nine of these had complete occlusions of at least one vessel (six superficial femoral, three posterior tibial, two anterior tibial, one iliac, and one common femoral) with extensive collateralization.
Median exercise time was 195.0 s in normal subjects (25th and 75th percentiles = 161.2 and 333.8 s) and 162.5 s in PAD patients (25th and 75th percentiles = 138.8 and 183.8 s; p = 0.06). Representative PCr spectra (Fig. 1)and recovery plots (Fig. 2)are shown. The median recovery time constant of PCr was 34.7 s in the control subjects (25th and 75th percentiles = 25.9 and 51.2 s) and 91.0 s in the PAD patients (25th and 75th percentiles = 65.0 and 134.5 s; p < 0.0001) (Fig. 3).A ROC curve was constructed to determine the discriminatory power of the PCr recovery time constant for detecting PAD (Fig. 4).The area under the fitted curve was 0.925 ± 0.045.
The PCr/ATP ratios were similar between the two groups both at rest (p = 0.50) and at end exercise (p = 0.54). The median pH measured at the completion of exercise was 7.02 in PAD and 6.95 in control subjects (p = 0.19). No significant correlation between PCr recovery time and age was detected within either subgroup (p = 0.31 for control subjects and p = 0.56 for PAD patients) or within the entire cohort (p = 0.12); PCr recovery time was also not significantly correlated with exercise time, within either subgroup (p = 0.27 for control subjects, p = 0.71 for patients) or within the entire cohort (p = 0.17). No relationship was observed between PCr recovery time and ABI based on linear regression analysis (p = 0.68). Furthermore, exercise time did not correlate with ABI in patients (p = 0.27).
Clinical follow-up was completed in 100% of PAD subjects at a mean of 435 ± 93 days. One patient died of cardiovascular complications after non-vascular surgery. Six of the limbs studied required revascularization or amputation. Percutaneous revascularization was performed in two subjects and surgical revascularization in four. In all but two cases, progressive worsening of claudication was the indication for intervention, with one experiencing an acute occlusion and another recurrent infection and tissue loss. Median PCr recovery was 121.0 s (25th and 75th percentiles = 91.8 and 220.0 s) in patients that died during follow-up or had limbs that required surgery, angioplasty, or amputation compared with 71.9 s in survivors that did not require limb intervention (25th and 75th percentiles = 51.0 and 112.8 s; p = 0.057). For PAD patients in the highest 75th percentile of PCr recovery kinetics (recovery time constant >118 s), the relative risk of subsequent surgery, angioplasty, amputation, or death was 3.11 (95% confidence interval 0.98 to 9.84). Median ABI was similar among PAD patients who died or had limbs that required revascularization or amputation (0.58; 25th and 75th percentiles = 0.53 and 0.65) compared with survivors free from limb intervention during follow-up (0.60; 25th and 75th percentiles = 0.52 and 0.69; p = NS).
Studies of test-retest reliability were performed in eight control patients and three PAD patients. The PCr recovery time constant was highly reproducible with an intraclass correlation coefficient r = 0.9505. Bland-Altman analysis is shown in Figure 5.One additional PAD subject (data not included in analysis or Fig. 5) had undergone surgical revascularization before a repeat study performed 226 days after the initial study. In addition to a near doubling of exercise time (135 vs. 230 s), PCr recovery time improved from 273 s at baseline to 60 s after revascularization.
There are four important findings in this study: 1) the kinetics of PCr recovery can be characterized at the completion of exercise in the magnetic resonance environment among patients with PAD; 2) the PCr recovery time constant is prolonged in patients with symptomatic PAD when compared with normal individuals despite shorter exercise times and similar post-exercise pH with an area under the ROC curve of 0.925; 3) PCr recovery time is reproducible and may represent a parameter with which PAD patients can be both identified and followed; and 4) a measure of post-exercise PCr recovery may identify PAD patients at higher risk of death and/or clinical deterioration requiring intervention at one year.
Non-invasive modalities employed for the diagnosis and study of PAD have a number of recognized limitations, including inaccuracy among patients with aortoiliac disease, diabetes mellitus, and collateralized territories (5,6). Heavily calcified arteries alone account for normalABI values in as many as 1 in 20 PAD patients (14). Both X-ray angiography and magnetic resonance angiography are limited to visualizing the vascular lumen. With these techniques, macrovascular abnormalities serve as a surrogate marker of tissue ischemia, largely ignoring adaptations in cellular metabolism and within the microvasculature that develop during the evolution of vascular insufficiency and influence end-organ response (12,15).
Experimental evidence from histopathologic and clinical studies suggests that skeletal muscle is not a passive bystander during development of peripheral vascular obstruction. Cellular alterations in PAD are comparable to those in subjects with mitochondrial myopathies and include altered mitochondrial enzyme expression, accumulation of the intermediates of oxidative metabolism, and mitochondrial deoxyribonucleic acid damage (15,16). Both biochemical and ultrastructural changes may, in part, explain why revascularization does not normalize measures of clinical performance in many PAD patients (17). While some metabolic changes may be maladaptive, others, such as an increase in mitochondrial density and more efficient glycolytic metabolism (18), can lead to improvement in symptoms and exercise capacity in the absence of revascularization and may underlie the benefit of physical training in PAD (19).
The major role of PCr in muscle metabolism is as a transport molecule and reservoir of high-energy phosphate bonds, the sole currency of cellular energetics (20). Phosphocreatine regeneration occurs exclusively within the mitochondria, and, because this depends entirely on the cell’s capacity for oxidative phosphorylation and oxygen supply (21), it represents an ideal parameter for detecting metabolic evidence of tissue ischemia. Relationships between peak oxygen uptake, maximal oxygen consumption (Vo2max), and PCr recovery kinetics have been well documented (22,23).
In the present study, PCr recovery time clearly distinguished symptomatic PAD patients from normal control subjects. These findings are in accordance with other studies that have demonstrated delayed regeneration of PCr as a reliable marker of arterial obstruction (12,17,24,25). When compared with normal subjects in our study, there was an approximate three-fold increase in the recovery time constant among patients. The area under the fitted curve of the ROC analysis is 0.925, demonstrating that PCr recovery kinetics could be employed to discriminate mild-to-moderate symptomatic PAD patients from normal subjects.
The median pH at end exercise was 7.02 in PAD patients and 6.95 in control subjects (p = 0.19). Hence, prolongation in PCr recovery was not the result of greater acidosis at end exercise among PAD patients. This is important as tissue pH can influence PCr recovery kinetics. Not only is creatine kinase a pH-dependent enzyme (21), but as much as 40% of cellular ATP may be consumed in ion pumping reactions during significant acidosis (23). With this ATP no longer available to facilitate PCr regeneration, recovery time is prolonged. Furthermore, PCr/ATP ratios were similar at both rest and peak exercise suggesting that differences in PCr recovery times were not the consequence of greater energy debt among PAD patients. Moreover, significant left ventricular systolic dysfunction was not responsible for prolonged PCr recovery times.
During clinical follow-up, revascularization was performed on six patients, and one individual died. Although only a trend, our data suggest longer PCr recovery time constants among mild-to-moderate PAD at greater risk for death, subsequent limb revascularization, or amputation (p = 0.057). This was most apparent among patients in the highest quartile of PCr recovery kinetics (relative risk = 3.11). No such trend was noted between these relevant clinical events and the resting ABI. This lack of correlation between the ABI and clinical events may be due to inherent inadequacies in the test itself as it evaluates primarily macrovascular hemodynamics rather than true tissue perfusion. Although an excellent tool for identifying PAD with a sensitivity and specificity of 90% and 98%, respectively (26), the ABI frequently fails to predict functional capacity, quality of life, symptoms, and clinical deterioration among the subgroup of patients with mild-to-moderate disease (14,27–29). A diagnostic tool capable of predicting clinical deterioration or death among patients with mild-to-moderate PAD would be invaluable, and, thus, the relationship between PCr recovery kinetics and clinical outcomes warrants further study.
In PAD patients undergoing angiography, there was a high incidence of complete vascular occlusion. With respect to other clinical characteristics, subjects in our protocol had a high incidence of coronary artery disease, diabetes mellitus, and tobacco use, making this group quite representative of the PAD population. The high-risk nature of mild-to-moderate PAD patients is reflected in the number of vascular events (peripheral angioplasty, vascular bypass, amputation, myocardial infarction, stroke, and vascular death) experienced during the clinical follow-up period (11 events in nine patients).
In order to optimize temporal resolution and retain adequate signal to measure both pH and PCr/ATP ratios, a surface-coil acquisition localized to the mid-calf was employed with the understanding that both exercising and non-exercising muscle groups would be incorporated into the signal, which might result in the underestimation of PCr depletion in the most stressed muscle groups (30). With the knowledge that PCr recovery time is the most relevant parameter and resolution of Pi and ATP peaks less important, echo-planar spectroscopic imaging could be used to interrogate individual muscle groups with adequate temporal resolution to assess PCr recovery kinetics (31).
Baseline age and gender differences between control subjects and patients might account for differences in PCr recovery time. However, others have demonstrated a lack of correlation between age and PCr recovery time (32,33), and no age-related differences were noted within the present cohort. While age-related changes in muscle mass, fiber type, and even oxidative capacity have been reported (34), a number of studies have demonstrated that active muscle perfusion remains intact among healthy older individuals (35,36). Hence, baseline differences in age may only be relevant under conditions in which blood flow is minimally perturbed (submaximal exercise) and mitochondrial function becomes the rate-limiting step in PCr regeneration (17). Also, age-related differences may have gone undetected because of our small sample size and non-selective spectroscopic technique.
Another potential limitation is that exercise ABIs, which may correlate better with disease severity than resting measures, were not performed. Total work and the rate and degree of plantar flexion were not controlled in our protocol. Prolongation of PCr regeneration does not pinpoint the level of vascular obstruction in PAD patients and has been described in other disease states where substrate supply may be limited or a primary metabolic defect exists (37,38). For example, in heart failure patients, delayed PCr recovery has been attributed to decreased mitochondrial volume and impaired oxidative metabolism rather than central hemodynamic factors (37). Many of the PAD patients had coronary artery disease. Myocardial ischemia and/or left ventricular dysfunction that may have developed during the exercise protocol could have confounded our results. However, no patients experienced chest pain or other clinical events during the exercise protocol. Furthermore, mean resting ejection fraction was preserved among patients in whom it was measured clinically before study (ejection fraction 56 ± 12%). With respect to clinical events, the sample size is too small to draw any definitive conclusions regarding the prognostic power of PCr recovery kinetics.
The 31P MRS has a number of potential applications in the PAD population. The relationship between PCr recovery kinetics and clinical events involving the lower limbs deserves further exploration in a larger cohort of patients. Investigation of asymptomatic patients and those who experience atypical symptoms may provide insight into the biochemical characteristics that distinguish these groups. Whether differences in PCr recovery time result entirely from changes in tissue blood flow, alterations in skeletal muscle at a cellular level, or a combination of both deserves further investigation. We are developing magnetic resonance measures of perfusion to compare with or add to spectroscopy in terms of their correlation, interdependence, and prognostic value. Another promising application of this technique is as a platform to study the surgical, percutaneous, and medical interventions currently employed as well as the investigational therapies of the future.
Supported by NIH, NHLBI, RO1 HL075792, and AHA 0425674U.
- Abbreviations and Acronyms
- ankle brachial index
- adenosine triphosphate
- peripheral arterial disease
- inorganic phosphate
- 31P MRS
- phosphorus-31 magnetic resonance spectroscopy
- receiver-operating characteristic
- Received September 23, 2005.
- Revision received December 20, 2005.
- Accepted December 30, 2005.
- American College of Cardiology Foundation
- American Heart Association
- Weitz J.I.,
- Byrne J.,
- Clagett G.P.,
- et al.
- Bhat H.K.,
- Hiatt W.R.,
- Hoppel C.L.,
- Brass E.P.
- Hiatt W.R.,
- Regensteiner J.G.,
- Hargarten M.E.,
- Wolfel E.E.,
- Brass E.P.
- Schunk K.,
- Romaneehsen B.,
- Mildenberger P.,
- Kersjes W.,
- Schadmand-Fischer S.,
- Thelen M.
- Hiatt W.R.,
- Regensteiner J.G.,
- Wolfel E.E.,
- Carry M.R.,
- Brass E.P.
- Bessman S.P.,
- Geiger P.J.
- Thompson C.H.,
- Kemp G.J.,
- Rajagopalan B.,
- Radda G.K.
- Wiener D.H.,
- Maris J.,
- Chance B.,
- Wilson J.R.
- Richardson R.S.,
- Haseler L.J.,
- Nygren A.T.,
- Bluml S.,
- Frank L.R.
- Horska A.,
- Fishbein K.W.,
- Fleg J.L.,
- Spencer R.G.
- Proctor D.N.,
- Newcomer S.C.,
- Koch D.W.,
- Le K.U.,
- MacLean D.A.,
- Leuenberger U.A.
- Martin W.H. III.,
- Ogawa T.,
- Kohrt W.M.,
- et al.