Methemoglobin Modulation as an Intravascular Contrast Agent for Magnetic Resonance Imaging: Proof of Concept

Objective: The aim of this feasibility study was to investigate methemoglobin modulation in vivo as a potential magnetic resonance imaging (MRI) gadolinium based contrast agent (GBCA) alternative. Recently, gadolinium tissue deposition was identified and safety concerns were raised after adverse effects were discovered in canines and humans. Because of this, alternative contrast agents are warranted. One potential alternative is methemoglobinemia induction, which can create T1-weighted signal in vitro. Canines with hereditary methemoglobinemia represent a unique opportunity to investigate methemoglobin modulation. Our objective was to determine if methemoglobinemia could create high intravascular T1-signal in vivo with reversal using methylene blue. Methods: To accomplish this study, a 1.5-year-old male-castrated mixed breed canine with hereditary methemoglobinemia underwent 3T-MRI/MRA with T1-weighted sequences including 3D-T1-weighted Magnetization Prepared Rapid Acquisition Gradient Echo (MPRAGE) and 3D-Time-Of-Flight (TOF). Images were acquired during baseline methemoglobinemia and rescued using intravenous methylene blue (1 mg/kg). Intravascular T1-signal was compared between baseline methemoglobinemia and post-methylene blue. N = 10 separate T1-signal measurements were acquired for each vascular structure, normalized to muscle. Significance was determined using paired two-tailed t-tests and threshold alpha = 0.05. Fold-change was also calculated using the ratio of T1-signal between methemoglobinemia and post-methylene blue states. Results: At baseline, methemoglobin levels measured 19.5% and decreased to 4.9% after methylene blue. On 3D-T1-weighted MPRAGE, visible signal change was present in internal vertebral venous plexus (IVVP, 1.34 ± 0.09 vs. 0.83 ± 0.05, p < 0.001, 1.62 ± 0.06-fold) and external jugular veins (1.54 ± 0.07 vs. 0.87 ± 0.06, p < 0.001, 1.78 ± 0.10-fold). There was also significant change in ventral spinal arterial signal (1.21 ± 0.11 vs. 0.79 ± 0.07, p < 0.001, 1.54 ± 0.16-fold) but not in carotid arteries (2.12 ± 0.10 vs. 2.16 ± 0.11, p = 0.07, 0.98 ± 0.03-fold). On 3D-TOF, visible signal change was in IVVP (1.64 ± 0.14 vs. 1.09 ± 0.11, p < 0.001, 1.50 ± 0.11-fold) and there was moderate change in external jugular vein signal (1.51 ± 0.13 vs. 1.19 ± 0.08, p < 0.001, 1.27 ± 0.07-fold). There were also small but significant differences in ventral spinal arterial signal (2.00 ± 0.12 vs. 1.78 ± 0.10, p = 0.002, 1.13 ± 0.10-fold) but not carotid arteries (2.03 ± 0.17 vs. 1.99 ± 0.17, p = 0.15, 1.02 ± 0.04-fold). Conclusion: Methemoglobin modulation produces intravascular contrast on T1-weighted MRI in vivo. Additional studies are warranted to optimize methemoglobinemia induction, sequence parameters for maximal tissue contrast, and safety parameters prior to clinical implementation.