Peripheral arterial disease (PAD) is mainly caused by atherosclerosis, and is associated with an impaired blood flow and decreased oxygen supply to distal muscle tissue. Both blood oxygenation level-dependent (BOLD) MR imaging¹ and exercise phosphorus-31 MR spectroscopy (³¹P-MRS)² have been used to study the effect of PAD on skeletal musculature. BOLD MRI provides a window into blood oxygenation at tissue level, whereas ³¹P-MRS measures phosphocreatine (PCr) recovery kinetics that reflect mitochondrial (dys)function and oxygen supply. Typically, dynamic MR studies are limited to measurements of only one parameter, whereas obtaining multiple parameters simultaneously during a single experiment would provide more insight into pathophysiology.^3,4 The purpose of this work is to implement and test interleaved acquisitions of quantitative T₂^* maps for assessments of the BOLD response and ³¹P-MR spectra for measuring PCr recovery kinetics during a calf muscle exercise-recovery protocol.

The study protocol was approved by the institutional review board, and participants (PAD patients, n = 10; age-matched healthy volunteers, n = 9) gave their written informed consent.

Exercise MR protocol: All experiments were conducted on a Philips Ingenia 3.0 Tesla MR system (Philips Medical Systems, Best, The Netherlands) equipped with a linearly-polarized ³¹P-MR surface transmit/receive coil (∅ 14 cm, 51.8 MHz; Philips). For shimming and proton MRI, the quadrature body coil was used (∅ 70 cm, 127.8 MHz). Subjects were positioned supine in the MR scanner, with one foot strapped into a custom-build plantar flexion exercise device affixed to the patient table. The ³¹P-MR surface coil was centered underneath the calf muscle. Following MR planning and shimming, the dynamic MR protocol was started and consisted of 200 scans with a multi-echo gradient-echo sequence for proton T₂^* mapping (transversal; field of view: 192×192 mm; matrix: 64×64; slice thickness: 10 mm; flip angle: 15°; 15 echoes; TE: 1.11 ms + 1.8 ms/step; TR: 28 ms), and 200 pulse-acquire ³¹P-MRS scans (adiabatic excitation; 2048 data points; bandwidth: 3000 Hz) in an interleaved^3,5 fashion (Figure 1). The effective repetition time, and hence the temporal resolution of both datasets, was 3 seconds. First, 2 × 20 scans were acquired under resting conditions, after which the subject was instructed to push the pedal at a steady rate of 1 repetition/second under audiovisual guidance until exhaustion, or until the increasing inorganic phosphate (P_i) peak amplitude leveled with the decreasing PCr peak amplitude, typically after 1-3 minutes of exercise. Measurements continued during recovery, completing the dynamic series of 10 minutes.

Data analyses: ³¹P-MRS data were analyzed as described previously.⁶ The PCr recovery time constant (τ_PCr) was determined by fitting a mono-exponential curve to [PCr] during recovery. From the multi-echo MR images, proton T₂^* maps were calculated using Philips scanner software. Mean proton T₂^* for the gastrocnemius muscle was determined in a region of interest drawn in an image acquired at rest after rigid-body registration of consecutive images using Elastix.⁷ The post-exercise BOLD response was quantified by determining the amplitude of T₂^* increase relative to end-recovery values.³

Figure 2 (healthy volunteer) and Figure 3 (PAD patient) show time curves for muscle PCr and P_i­ concentrations measured with ³¹P-MRS at rest, during exercise and subsequent recovery, and simultaneously acquired proton T₂^* values for the gastrocnemius muscle at rest and during recovery. The temporal resolution of 3 seconds was sufficient to capture both the dynamics of high-energy phosphate concentrations as well as the post-exercise BOLD response. PCr recovery time constants were higher in PAD patients compared to healthy controls (54 ± 14 s vs. 29 ± 8 s, p < 0.001). Likewise, the amplitude of the BOLD response was higher in PAD patients (4.4 ± 2.1% vs. 2.4 ± 1.9%, p < 0.05). Results for all subjects are collected in Figure 4.In this preliminary work, we demonstrate that with interleaved acquisitions of proton T₂^* maps and ³¹P-MR spectra, it is feasible to dynamically assess both tissue oxygenation as well as muscle energy metabolism in the human calf muscle during a single exercise session. Impaired blood flow in PAD patients was reflected by an altered BOLD response to exercise in the gastrocnemius muscle, combined with a slower PCr recovery. In a cohort of patients, quantitative MR measures should be compared with current indices of PAD severity used in the clinic, such as the ankle-brachial index (ABI) or 6-min walk distance.⁸ Such studies will reveal whether this multiparametric approach will be practical to monitor disease progression or effects of therapeutic interventions.