Quantitative assessment of cardiac phosphocreatine metabolism under physiological and pharmacological stress using CEST MRI.

Abstract

Background: Phosphocreatine (PCr) exhibits a distinct chemical exchange saturation transfer (CEST) resonance at ~2.5 ppm with slower exchange kinetics and lower pH sensitivity than creatine. This pilot study evaluated the feasibility of quantifying myocardial PCr-dominated concentration ([PCr]_d) using CEST cardiac MRI (CEST-CMR) and validated the measurements with phosphorus magnetic resonance spectroscopy (³¹P-MRS).

Methods: Phantoms with varied concentrations of PCr and Cr were scanned to characterize PCr CEST signals under physiological conditions. Experiments were conducted at both room temperature and 37 °C with full Z-spectra acquisition. For the study in vivo, CEST imaging was performed in normal canines (n = 13), at rest and after regadenoson vasodilation, and during dobutamine stress (n=5). In addition, healthy volunteers (n =9) underwent rest-exercise-recovery studies with in-magnet plantar-flexion exercise. Both CEST images and ³¹P-MRS were acquired in separate exercise sessions. WASSR-derived corrected The Z-spectra corrected by derived B₀ maps were fitted with a three-pool Bloch-McConnell model (water, PCr, magnetization transfer) to estimate [PCr]_d. For ³¹P-MRS, PCr/γATP ratios were converted to PCr concentrations using γATP as a reference. Rate-pressure product (RPP) was used as an indicator of myocardial oxygen consumption.

Results: In phantoms, PCr-dominated CEST contrast increased monotonically with PCr concentration across pure and mixed PCr/Cr solutions, while Cr-only phantoms did not produce artifactual PCr estimates. In canines, myocardial [PCr]_d was 12.0 ± 1.8mM at rest and decreased to 7.1 ± 1.6mM during dobutamine stress (p<0.001), while remaining unchanged with regadenoson vasodilation (11.8 ± 0.3 vs 11.8 ± 0.6mM). Changes in [PCr]_d correlate_d negatively with RPP (r = -0.75). In human subjects, CEST-derived [PCr]_d was 13.2 ± 1.4mM at rest, 7.1 ± 1.0mM during exercise (p<0.001), and 12.9 ± 1.5mM during hyperemia. Corresponding ³¹P-MRS estimates were 12.9 ± 1.4mM, 7.7 ± 2.3mM, and 12.2 ± 1.6mM, respectively. CEST-derived [PCr]_d showed a moderate negative correlation with RPP (r = -0.52).

Conclusion: CEST-CMR enables noninvasive estimation of myocardial [PCr]_d in vivo and detects physiologic energetic changes during stress, with measurements consistent with ³¹P-MRS.