|Home | About | Journals | Submit | Contact Us | Français|
We aimed to investigate the pathophysiology of peripheral arterial disease (PAD) by examining magnetic resonance imaging (MRI) and spectroscopic (MRS) correlates of functional capacity.
Despite the high prevalence, morbidity, and cost of PAD, its pathophysiology is incompletely understood.
Eighty-five patients (age 68±10) with mild-to-moderate PAD (ankle brachial index (ABI) 0.69±0.14) had their most symptomatic leg studied by MRI/MRS. Percent wall volume in the superficial femoral artery was measured with black blood MRI. First-pass contrast-enhanced MRI calf muscle perfusion and 31P MRS phosphocreatine recovery time constant (PCr) were measured at peak exercise in calf muscle. All patients underwent MR angiography (MRA), treadmill testing with VO2 max measurement and a 6-minute walk test.
Mean MRA index of number and severity of stenoses was 0.84±0.68 (normal, 0), % wall volume 74±11% (normal, 46±7%), tissue perfusion 0.039±0.015 sec−1 (normal, 0.065±0.013 sec−1), and PCr 87±54s (normal, 34±16s). MRA index, % wall volume, and ABI correlated with most functional measures. PCr was the best correlate of treadmill exercise time, whereas calf muscle perfusion was the best correlate of 6-minute walk distance. No correlation was noted between PCr and tissue perfusion.
Functional limitations in PAD are multifactorial. As measured by magnetic resonance imaging and spectroscopy, atherosclerotic plaque burden, stenosis severity, and tissue perfusion and energetics all play a role. However, cellular metabolism is uncoupled from tissue perfusion. These findings suggest a potential role for therapies that regress plaque, increase tissue perfusion, and/or improve cellular metabolism.
Peripheral arterial disease (PAD) affects over 8 million Americans and its prevalence is expected to rise as the population ages(1). Apart from its strong association with cardiovascular mortality, PAD itself carries significant morbidity(2) and health care costs(3). Patients with PAD suffer from progressive functional decline and impaired quality of life(4;5). .
Despite the high prevalence, cost, and morbidity of lower extremity PAD, there are few effective medical therapies for the diseased limb, in part due to two factors. First, the pathophysiology of PAD is complex and incompletely understood. While the obstruction of blood flow due to large vessel atherosclerosis is critical, the degree of hemodynamic impairment does not consistently relate to functional limitation(6). Thus, factors independent of blood flow and downstream from the obstruction are believed to play an important role. Secondly, the current gold standard for testing the efficacy of therapeutic interventions, the exercise treadmill test(7), often lacks standardization and is plagued by intra-patient variability and a significant placebo response
Therefore, we aimed to further examine the pathophysiology of PAD and its primary clinical manifestation, claudication, by determining the relationship between large vessel obstruction, atherosclerotic plaque burden, tissue perfusion, cellular metabolism and functional capacity. To accomplish this, we used newly developed magnetic resonance imaging (MRI) and spectroscopy (MRS) methods to assess atherosclerotic plaque burden(8) and exercise measures of calf muscle perfusion(9) and metabolism(10) in patients with mild-to-moderate PAD. We also examined their relationship to functional capacity.
Subjects between the ages of 30 and 85 years with symptoms of intermittent claudication and an ankle-brachial index (ABI) between 0.4 and 0.9 were eligible for this study. The protocol was approved by the Human Investigation Committee at the University of Virginia Health System (UVA) and all participants signed informed consent. Subjects with rest pain, critical limb ischemia, or a contraindication to MRI were excluded as were those with comorbidities that severely limited their ability to perform a walking treadmill test. Normal human subjects without risk factors for atherosclerosis were recruited from the community to serve as controls for calf muscle perfusion measures.
The study was divided over two days to allow sufficient time between exercise portions of the protocol to avoid fatigue and/or ischemic preconditioning. Patients were admitted overnight to the General Clinical Research Center. Due to time constraints, only the most symptomatic leg was studied by MRI/MRS. Entry ABI’s were obtained at the UVA vascular laboratory using standard techniques.
MRI protocols were completed on an Avanto 1.5T scanner and MRS protocols on a Sonata 1.5T scanner (both Siemens Medical Solutions, Erlangen, Germany).
Wall volume in 15–20 cm of the superficial femoral artery (SFA) was assessed as previously described(8) using a custom-built surface coil array placed over the thigh of the subject’s most symptomatic leg and a fat-suppressed multi-slice turbo-spin-echo pulse sequence (Figure 1A). Wall volume, defined as total vessel area minus lumen area multiplied by the slice thickness, was measured with VesselMASS software (University of Leiden, Netherlands). Percentage wall volume was calculated by dividing the wall volume by the total vessel volume (wall + lumen)(11).
Calf muscle perfusion measured at peak exercise was performed with a MR-compatible foot pedal ergometer using first-pass contrast-enhanced imaging as previously described(9) (Figure 1C). Participants performed pedal ergometry at a steady rate (10–12 rpm) until exhaustion and/or limiting symptoms developed. Gadopentetate dimeglumine 0.1 mM Gd/kg (Magnevist, Berlex, Montville, NJ, USA) was immediately infused and imaging commenced. Time intensity curves were generated with Argus software (Siemens Medical Solutions) for regions of interest in the calf muscle and input artery (typically the popliteal). The slope of the time intensity curve of signal normalized to the proton density from the calf muscle group with the greatest signal intensity was defined as tissue perfusion. To assess microvascular blood flow, tissue perfusion was indexed to macrovascular blood flow into the calf by dividing it by the slope of the arterial input curve to obtain a perfusion index(9).
Phosphocreatine recovery time constant (PCr) was measured as previously described(10) by 31phosphorus MR spectroscopy during recovery from peak exercise using a single-pulse, surface coil localized, 512ms free induction decay acquisition with 20 averages centered on the mid-calf. PCr was then calculated using a monoexponential fit of phosphocreatine concentration versus time, beginning at cessation of exercise (Figure 1D). Calf muscle pH at end-exercise was measured as previously described(10).
Gadopentetate dimeglumine (0.2mM Gd/kg) enhanced MR angiography from the abdominal aorta to the foot was performed with a moving table/bolus chase technique in 3 stations (64–104 slices, field of view 500mm, repetition time 2.5–3.0, echo time 1.0–1.1, flip angle 20–25º, voxel size 1.6–2.0 × 1.0–1.3 × 1.0–1.5 mm) (Figure 1B). Image analysis was performed by 2 experienced observers by visual consensus using two different grading schemes(12;13). First, the arterial tree of the leg of interest was divided into 16 segments from the distal aorta to the runoff vessels. A five point ordinal scale (zero for 0–19% diameter stenosis, 1 for 20–49%, 2 for 50–74%, 3 for 75–99%, and 4 for total occlusion) was used to grade each segment. The total score was then indexed to the total number of segments (MRA index)(12). The runoff index (13) then assesses the resistance to blood flow by accounting for the length of occlusion and weighting each segment by its contribution to calf muscle perfusion.
Subjects walked up and down a 100 foot corridor for six minutes and walking distance and initial claudication distance were recorded.
Under the supervision of a physician and an exercise physiologist, subjects underwent a standardized graded Skinner-Gardner exercise treadmill test(14). Patients had VO2 and electrocardiogram monitoring during the exam. Metabolic measures were collected using standard open circuit spirometry (Viasys 229, Yorba Linda, CA). The test was terminated by the participant at volitional exhaustion, due to symptom limiting claudication or after 20 minutes. Total treadmill exercise time, time to claudication, and symptom limited maximal oxygen consumption (VO2) were measured. Post-exercise ABI was measured.
Physical activity was assessed using the Aerobics Center Longitudinal Study Physical Activity Questionnaire(15) modified for interview. Physical activity was calculated as metabolic equivalent hours per week using the Compendium of Physical Activities(16). Energy expenditure per week was estimated.
Baseline lipids were measured. High sensitivity C-reactive protein (hs-CRP) was assayed on a solid-phase, chemiluminescent immunometric analyzer (Immulite 2000, Diagnostics Products Corp, Los Angeles, CA). Interleukin-10 (IL-10) and tumor necrosis factor alpha (TNFα) were assayed by quantitative sandwich enzyme immunoassays.
All patients’ baseline characteristics, MRI and functional measures were presented as mean (SD) for continuous variables and N (%) for categorical variables. Standard statistical tests including the Shapiro-Wilk test and the normal probability plot were used to check the normality assumption for continuous variables. Bivariate relationships between each factor and outcomes were examined using either Pearson's correlation coefficient or Spearman’s rank correlation coefficient. Multivariable regression models were fit to examine the relationship between the functional capacity outcomes and the selected MRI factors including MRA index, PCr, tissue perfusion, and % wall volume. F-tests were used to identify the MRI factors which were strongly related with different functional outcomes. All statistical analyses were performed using SAS version 9.1.3 (SAS Inc., Cary, North Carolina).
Eighty-seven patients with mild-to-moderate PAD were enrolled; two withdrew prior to completion of the MRI due to claustrophobia, leaving 85 for data analysis. All 85 patients completed the 6 minute walk test, while 83 completed the treadmill test. Percent wall volume, MRA index, and runoff index data were obtained for all 85 patients, PCr from 83, tissue perfusion from 80 and perfusion index from 77 patients. Missing data was due to either technical problems with data acquisition or failure of the patient to complete the exercise. Seven normal subjects (age 47±5, 4 males) made up the cohort of controls for perfusion measures.
Table 1 describes the baseline characteristics of the patient population. The ranges for age, entry ABI, and body mass index were 41–82, 0.36–0.90, 17.1–53.3 respectively. Women and non-Caucasians comprised 44% and 32% of the population, respectively. Of the 19 patients who had prior revascularization, 12 revascularizations were of the leg studied by MRI. Of those 12, 9 had an iliac stent. The remaining 3 had an aorta-femoral, femoral-femoral, or femoral-popliteal artery bypass. Table 2 lists the results of MRI and functional measures, as well as the self-reported physical activity questionnaire and laboratory values. For reference, values from normal subjects for the MRI measures are as follows: % wall volume = 45.8±7.0%(17), tissue perfusion = 0.065±0.013 sec−1, perfusion index = 0.69±0.17(9), and PCr = 34±16sec(10). Normal values for the MRA are assumed (Runoff index = 4, MRA index = 0). The relationships between the noninvasive measures assessed by either Pearson's correlation coefficient or Spearman’s rank correlation coefficient are depicted in Table 3. The ABI correlated with each MRI/MRS measure as did the MRA index. Of note, PCr did not correlate with tissue perfusion or perfusion index (Figure 2).
Table 4 shows the relationship between functional performance and the noninvasive measures. All noninvasive measures correlated with treadmill exercise time, except for tissue perfusion and end-exercise pH. The individual correlations between tissue perfusion and 6-minute walk time and PCr with treadmill exercise time are shown in Figure 3 and Figure 4, respectively. All noninvasive measures other than PCr and perfusion demonstrated correlation to functional performance as measured by VO2 max (Table 4). Using a model that included tissue perfusion, % wall volume, PCr, and MRA index, the F-test identified PCr as being significantly related to the logarithm of treadmill exercise time (F=5.59, p=0.02), tissue perfusion with 6-minute walk distance (F=7.61, p<0.01), and % wall volume with VO2 (F=4.19, p<0.05).
When the relationship between baseline clinical characteristics and noninvasive and functional measures were examined, older age correlated with higher MRA index (r=0.34, p<.01) and lower tissue perfusion (r=−0.32, p<0.05) and perfusion index (r=−0.25, p<0.05), as well as shorter treadmill exercise time (r=−0.30, p<0.01) and 6-minute walk distance (r=−0.36, p<0.001). Non-Caucasian ethnicity was associated with a higher MRA index (r=0.28, p<0.01) and lower perfusion index (r=−0.32, p<0.01). Female gender correlated with higher wall volume (r=0.28, p<0.05), and lower perfusion index and VO2 (r=−0.25, p<0.05 for both).
A history of hypertension related to a higher % wall volume (r=0.28, p<0.05). Prior revascularization was associated with higher % wall volume (r=0.22, p<0.05), MRA index (r=0.26, p<0.05), and runoff index (r=0.32, p<0.01) and a worse 6-minute walk distance and VO2 (r=−0.25, p<0.05 for both). The presence of diabetes related to a worse VO2 (r=−0.22, p<0.05). CRP was the only measured serum or inflammatory marker to relate to any functional measure as it had a negative correlation with 6-minute walk distance (r =−0.25, p<0.05).
When the relationship between self-reported physical activity levels and noninvasive measures were examined, both physical activity and energy expenditure correlated with % wall volume (r=−0.25, p<0.05 for both). Estimated physical activity correlated with the 6-minute walk distance (r=0.27, p<0.05) and time to claudication (r=0.30, p<0.01).
This study used magnetic resonance imaging and spectroscopy to further elucidate the multifactorial pathophysiology of claudication. We found that the severity of macrovascular obstruction, atherosclerotic plaque burden, reduced tissue perfusion, and abnormal energy metabolism all relate to functional limitation in patients with claudication. Another important finding is that the phosphocreatine recovery time constant of calf muscle does not correlate with calf muscle perfusion, suggesting that cellular metabolism impacts exercise performance but is uncoupled from tissue perfusion. Taken together, these MRI/MRS measures allow an integrative assessment of large vessel atherosclerosis, blood flow, and cellular metabolism which each individually contribute to functional capacity in PAD.
The pathophysiology of claudication is complex and incompletely understood. In prior studies the degree of flow limitation has not correlated with functional impairment. ABI has not correlated with lower extremity functional strength(18). Maximal calf blood flow measured by strain gauge plethysmography does not predict treadmill walking distance(19). Furthermore, when patients with PAD undergo exercise training their functional capacity and calf blood flow improve but changes in flow have not correlated with changes in function(20).
In the present study calf muscle blood flow at peak exercise, as measured by first-pass contrast enhanced MRI, correlated with 6-minute walk distance. The discrepancy between previous studies and the present can likely be explained by the techniques used to measure calf muscle perfusion. The prior studies measured gross calf muscle blood flow. With the ability to visualize the individual muscle groups used in exercise, inactive muscles and intravascular flow can be excluded, allowing for a more direct analysis of maximal end organ tissue perfusion.
Factors independent of blood flow and intrinsic to the skeletal muscle likely play a role in the functional limitations seen in PAD. Patients with PAD have impaired mitochondrial function(21;22) as mitochondria are morphologically abnormal and are increased in number proportionally to the severity of occlusive disease(23). Phosphocreatine recovery rate post-exercise, an index of mitochondrial function, is significantly diminished in PAD(10). In the present study, PCr correlated inversely with treadmill exercise time, lending credence to the concept that mitochondrial dysfunction plays a role in the exercise limitation of PAD.
One of the aims of this study was to further explore the relationship between tissue perfusion and mitochondrial function (cellular metabolism). Previous studies that examined this relationship have inferred blood flow from angiographic findings and the ABI(24) or examined mitochondrial function under conditions of unchallenged blood flow(25). We demonstrated no correlation between contrast-enhanced measures of tissue perfusion with PCr, supporting the notion that mitochondrial impairment in PAD is uncoupled from reduced blood flow. However, one reason for the lack of correlation could be that PCr is measured from a slab of tissue across the entire calf whereas perfusion is measured in specific regions of interest in the calf muscle.
Several measures were identified in the present study that correlate with exercise performance and may be suitable as surrogate measures of functional capacity in PAD therapeutic trials. Currently, the accepted primary endpoint for new claudication therapies is an improvement in exercise performance measured by treadmill testing(7). Some have speculated that the failure of many investigational therapies to show benefit may be due to the known placebo response and significant intra-patient variability (26). A recent study has shown that simple office-based measures of functional capacity can predict mortality beyond that of the ABI(27). Longer follow-up is required to demonstrate whether an improvement in any of the identified imaging endpoints translate into clinical improvement. Additional techniques such as 1H MR spectroscopy to assess ischemia-induced reoxygenation may also offer potential in this regard(28).
Patient characteristics associated with worse imaging and functional performance in the present study included older age, non-Caucasian ethnicity, and female gender. Age is a known risk factor for the presence of PAD(29) as is African-American race(30) and both were shown to be associated with greater atherosclerosis burden and worse tissue perfusion at peak exercise in this study. We investigated the association between lipids and inflammatory markers with functional status. Apart from the modest relationship of CRP with the 6-minute walk distance, which has been previously demonstrated(31), no other measured laboratory marker correlated with any functional measure. CRP did not correlate with MRA index or % wall volume suggesting that the relationship with function is not merely a reflection of atherosclerosis burden.
Our patient population excluded those with PAD who were asymptomatic, had critical limb ischemia, or had significant co-morbid conditions. Thus, our results should not be generalized to these patient populations. PAD typically involves both lower extremities and functional capacity is dependent on both legs. However, in our study we only studied the most symptomatic leg due to the practicality and time constraints of acquiring several MRI measures.
Collateral blood flow plays an important role in the pathophysiology of PAD. However, the spatial resolution of MR angiography allows detection of only a fraction of collateral vessels. Endothelial function is abnormal in patients with peripheral vascular disease(32). Its contribution to the pathophysiology of claudication has not been thoroughly investigated. The present study did not include a measure of endothelial function.
Two of the methods in this study, MRI first-pass tissue perfusion and MR angiography, utilize gadolinium-based contrast agents. Recently the FDA has issued a warning regarding the use of these agents in patients with advanced kidney disease(33). To some extent this may limit the use of these agents in patients with renal failure, although early hemodialysis may prevent the development of this side effect(34).
Despite the significant prevalence, morbidity, and cost associated with PAD, there are few effective therapies, in part due to the incomplete understanding of pathophysiology. In the present study we have demonstrated that the functional limitation seen in PAD patients is multifactorial. Abnormal mitochondrial function in PAD is independent of tissue perfusion and it too correlates with functional impairment. These findings suggest a potential role for therapies that regress plaque, increase tissue perfusion, and/or improve cellular metabolism.
We gratefully acknowledge the important contributions of Jennifer R. Hunter RN and John Christopher RT to the completion of this work.
Supported by NHLBI R01 HL075792 (CMK), M01RR000847 from the National Center for Research Resources, and NIBIB T32 EB003841 (JDA, AMW)
Drs. Epstein, Meyer, Hagspiel and Kramer receive research support from Siemens Medical Solutions.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Clinical Trial Identifier: NCT00587678, Comprehensive Magnetic Resonance in Peripheral Arterial Disease