Chun-Ming Chai

Cardiac metabolism measured noninvasively by hyperpolarized (13)C MRI.

Pyruvate is included in the energy production of the heart muscle and is metabolized into lactate, alanine, and CO2 in equilibrium with HCO  3− . The aim of this study was to evaluate the feasibility of using 13C hyperpolarization enhanced MRI to monitor pyruvate metabolism in the heart during an ischemic episode. The left circumflex artery of pigs (4 months, male, 29–34 kg) was occluded for 15 or 45 min followed by 2 hr of reperfusion. Pigs were examined by 13C chemical shift imaging following intravenous injection of 1‐13C pyruvate. 13C chemical shift MR imaging was used in order to visualize the local concentrations of the metabolites. After a 15‐min occlusion (no infarct) the bicarbonate signal level in the affected area was reduced (25–44%) compared with the normal myocardium. Alanine signal level was normal. After a 45‐min occlusion (infarction) the bicarbonate signal was almost absent (0.2–11%) and the alanine signal was reduced (27–51%). Due to image‐folding artifacts the data obtained for lactate were inconclusive. These studies demonstrate that cardiac metabolic imaging with hyperpolarized 1‐13C‐pyruvate is feasible. The changes in concentrations of the metabolites within a minute after injection can be detected and metabolic maps constructed. Magn Reson Med 59:1005–1013, 2008.

MR coronary angiography in pigs with intraarterial injections of a hyperpolarized (13)C substance.

A new diagnostic application of a water‐soluble contrast medium (CM) based on the hyperpolarization of a 13C substance is introduced. The degree of polarization achieved is >30%, which is about a factor of 105 higher than the thermal equilibrium polarization level at 1.5 T. Imaging of hyperpolarized (HP) CM during a cardiac interventional MRI procedure was studied. Catheters were positioned in the left and right coronary arteries of pigs. A coil tuned to 13C was used for nonproton imaging. The HP‐13C CM (∼5 ml, 0.5 M, ∼30% polarization) was injected during projection imaging using a fully balanced steady‐state free precession (SSFP) pulse sequence with and without cardiac gating. The contrast agent‐filled catheter was clearly visible during the procedure. The coronary arteries were well depicted and the signal‐to‐noise ratios (SNRs) were in the range of 10–40. The use of HP‐13C CM may provide a new diagnostic procedure for interventional MRI. Magn Reson Med, 2006.

Passive catheter tracking during interventional MRI using hyperpolarized 13C.

Interventional procedures in MRI can be performed preclinically using active or passive catheter‐tracking methods. A novel passive nonproton technique is suggested that uses a catheter filled with a hyperpolarized 13C contrast agent. A prototype three‐lumen catheter was built with two closed lumens containing a flowing hyperpolarized 13C contrast agent. Entire‐length 13C catheter projection visualization could be performed in vivo with a catheter SNR of ∼80, one dual projection frame per ∼700 ms, and an in‐plane resolution of 2 × 2 mm2 while traveling through the aorta of a pig. The traveling path of the 13C catheter was visualized after back‐projection catheter reconstruction and after image fusion with an anatomical offline proton road map. Catheter length visualization was aided by an oblique planar visualization mode. The high catheter signal demonstrated, together with the entire catheter length visualization and high surrounding soft‐tissue contrast, warrants further development into a real‐time technique. Magn Reson Med 57:1140–1147, 2007.

Nephrotoxicity after renal angiography using iodine and gadolinium contrast media in pigs with renal damage

Iodine contrast medium (I-CM) induced nephrotoxicity is recognized for many years and is especially frequent in patients with pre-existing renal insufficiency alone or combined with diabetes mellitus. The use of non-ionic low osmolar contrast media (LOCM) has reduced the risk of renal injury after X-ray arteriography in patients with renal impairment (1). The importance of nephrotoxicity may increase in the future, due to an anticipated higher number of examinations of elderly subjects, who often have impaired renal function. Furthermore, interventional radiology uses large doses of contrast media (CM). These risk factors make it necessary to adopt some strategies to overcome the problems with nephrotoxicity associated with CM administration in patients with renal damage, for instance the use of non-ionic LOCM (1) and adequate

Iodixanol 320 results in better renal tolerance and radiodensity than do gadolinium-based contrast media: arteriography in ischemic porcine kidneys.

PURPOSE To prospectively compare nephrotoxicity and radiodensity of plasma hyperosmotic gadolinium chelates (attenuation-osmotic ratio of 1:1) with those of plasma iso-osmotic iodine-based contrast media (attenuation-osmotic ratio of 3:1 or 6:1) after renal arteriography in ischemic porcine kidneys. MATERIALS AND METHODS The local animal care committee approved this study. The following contrast media were used: (a) iodixanol (150 mg of iodine per milliliter and 320 mg I/mL, 0.29 osm/kg H(2)O), (b) iopromide (150 mg I/mL, 0.34 osm/kg), (c) 0.5 mol/L gadodiamide (0.78 osm/kg), and (d) 1.0 mol/L gadobutrol (1.6 osm/kg). After left-sided nephrectomy, contrast media (3 mL per kilogram of body weight) were injected (20 mL/min) in a noncrossover design into the right renal artery of pigs during a 10-minute ischemic period. There were eight pigs in each group and one group for each contrast medium. We compared histomorphology, radiographic contrast medium excretion, subjective radiodensity of nephrograms (70 kVp) at the end of injection, and contrast medium plasma half-life elimination times 1-3 hours after injection. Longer elimination times resulted in lower glomerular filtration rates. RESULTS Gadobutrol caused extensive tubular necrosis and moderate glomerular necrosis; gadodiamide and iopromide, minimal to mild tubular necrosis; and iodixanol, no necrosis. Gadobutrol was the only contrast medium to show no sign of excretion, and its plasma half-life elimination time (median, 1103 minutes; P < .001) was significantly longer than that of other contrast agents. Gadodiamide had a significantly longer plasma half-life elimination time (median, 209 minutes; P = .01) than did iodine-based contrast media (median, 136-142 minutes). The 320 mg I/mL dose of iodixanol had the highest radiodensity, whereas gadodiamide had the lowest radiodensity. The radiodensity of the 320 mg I/mL dose of iodixanol was greater than that of the 150 mg I/mL dose of iodixanol, which was equal to the radiodensities of the 150 mg I/mL dose of iopromide and 1.0 mol/L gadobutrol, which in turn were greater than that of 0.5 mol/L gadodiamide. CONCLUSION Plasma iso-osmotic iodine-based contrast media used at commercially available concentrations have superior attenuation and nephrotoxic profiles compared with equal volumes of hyperosmotic nonionic 0.5-1.0 mol/L gadolinium-based contrast media when performing renal arteriographic procedures.

Adding Sodium and Calcium Ions to the Contrast Medium Iodixanol Reduced the Risk of Ventricular Fibrillation During Perfusion of the Left Coronary Artery in Pigs: Effects of Electrolytes, Viscosity, and Chemotoxicity of an Isotonic Perfusate 1

RATIONAL AND OBJECTIVES The effects of electrolytes, viscosity, and chemotoxicity of a plasma-isotonic iodine contrast medium iodixanol were compared with regard to its propensity to cause ventricular fibrillation (VF). MATERIALS AND METHODS The left coronary artery of pigs was perfused with five isotonic solutions: iodixanol 320 mg I/mL with 19 mmol/L NaCl + 0.3 mmol/L CaCl2, Iod 320+Mann (iodixanol 320 mg I/mL + 50 mmol/L mannitol), Mann+Na/Ca (240 mmol/L mannitol with 19 mmol/L NaCl + 0.3 mmol/L CaCl2), Mann (275 mmol/L mannitol) and Ringer. The first two solutions have at 37 degrees C a viscosity of approximately 13 mPa x s while the others have a viscosity < 1 mPa x s. In eight pigs, each test solution was injected twice into the left coronary artery in random order for 10 seconds (injection volume, 20 mL). In 15 pigs, each of the solutions was injected in random order for 11-40 seconds through the end-hole of a wedged 5F balloon catheter in left coronary artery. Injection rate was 0.5 mL/sec until VF occurred. If VF occurred, injection was stopped and the heart was defibrillated. If VF did not occur, the perfusion period was 40 seconds. RESULTS The 10-second perfusions caused no VF. The 40-second perfusions with iodixanol 320 mg I/mL with 19 mmol/L NaCl + 0.3 mmol/L CaCl2 or Ringer caused no VF (0%). Iod 320+Mann caused nine VF (60%) after 35 +/- 4 seconds (SEM). Mann+Na/Ca caused 14 VF (93%) after 30 +/- 2 seconds. Mann caused 15 VF (100%) after 24 +/- 2 seconds. Iodixanol 320 mg I/mL with 19 mmol/L NaCl + 0.3 mmol/L CaCl2 and Ringer caused fewer VF than all other solutions (P < .05-.001). Iod 320+Mann caused fewer VF than Mann (P < .05). Iod 320+Mann caused VF later than Mann+Na/Ca or Mann (P < .02 and P < .01). Mann+Na/Ca caused VF later than Mann (P < .05). CONCLUSION The results fit with a concept that VF starts when the electrolyte composition of the interstitial fluid in the myocardium is sufficiently nonphysiologic. The more physiologic the electrolyte composition of the perfusion fluid, and the higher its viscosity, the slower the composition of the interstitial fluid will be changed, and VF will occur later (or not at all).

Histomorphological Changes after Renal X-Ray Arteriography Using Iodine and Gadolinium Contrast Media in an Ischemic Porcine Model.

Background: Gadolinium contrast media (Gd-CM) are regarded as non-nephrotoxic or considerably less nephrotoxic than iodine contrast media (I-CM), and have therefore come to be used as a substitute for I-CM in patients with renal insufficiency in a variety of radiographic examinations. Purpose: To investigate renal histomorphological changes caused by Gd-CM in comparison with I-CM after renal X-ray arteriography in an ischemic porcine model, and to evaluate these changes in relation to the nephrotoxicity of the CM used. Material and Methods: Test solutions: gadopentetate, gadodiamide, iohexol, gadobutrol, iopromide, iodixanol, mannitol, and saline. The experiments were performed on 152 animals. Each pig was randomized to receive one test solution injected into the balloon-occluded (10 min) right renal artery. The kidneys were evaluated histomorphologically. The severity of histomorphological changes was graded subjectively: 1 = minimal, 2 = mild, 3 = moderate, and 4 = marked. Results: The main histological changes were 1) proximal tubular and glomerular necrosis, 2) hemorrhage/congestion of the cortex, medulla, and glomeruli, 3) proximal tubular vacuolation, and 4) protein-filled tubules in the cortex and medulla. Necrosis and hemorrhage/congestion were more frequent after injections with gadopentetate, mannitol solution iso-osmotic to gadopentetate, and gadobutrol compared to all other groups (P<0.001). The degree of necrosis and hemorrhage/congestion was related to the degree of impairment of renal function, but inversely related to vacuolation and tubular protein filling. Conclusion: In ischemic porcine kidneys, the histomorphological changes caused by Gd-CM are similar to those caused by I-CM. Vacuolation appears to be independent of the osmolality and viscosity of the CM, and does not seem to be an indicator of renal impairment. “High-osmolal” Gd-CM are more nephrotoxic than “low- and iso-osmolal” I-CM when compared in equal volumes of concentrations, resulting in equal X-ray attenuation.

Effects on erythrocyte aggregation and blood coagulation from iohexol solutions with and without sodium chloride. An in vitro study on the role of ion concentration and osmolality.

Solutions of the nonionic monomeric contrast medium iohexol (300 mg I/ml) with and without added NaCl were investigated for effects on red blood cell aggregation and blood coagulation. Three volumes of a test solution were mixed in test tubes with one volume of human blood. During 30 min samples of the mixture were taken for investigation. Six test solutions were used: 1) iohexol, 2) iohexol+glucose 280 mM, 3) iohexol+NaCl 150 mM, 4) glucose 280 mM, 5) glucose 140 mM+NaCl 75 mM, 6) NaCl 150 mM. Test solutions with NaCl caused no aggregation. Test solutions without NaCl always caused macroscopic red cell aggregates. These aggregates always disappeared when saline was added to the sample. The macroscopic red cell aggregates could be dispersed to microscopic aggregates by shaking the test tubes. During the next 30 min macroscopic aggregates returned in the glucose solution but not in the iohexol solutions. In 30 min, blood mixed with iohexol solutions never coagulated while blood layered on top of the same iohexol solutions always coagulated. Blood mixed with solutions 5 and 6, both without iohexol, always coagulated. It is concluded that adding 150 mM NaCl to iohexol did not eliminate its ability to anticoagulate whole blood, but inhibited its ability to aggregate red cells. This inhibition was not caused by the osmotic effects of the added NaCl.

Incidence of Ventricular Fibrillation during Left Coronary Arteriography in Pigs : Comparison of a Solution of the Nonionic Dimer Iodixanol with Solutions of Five Different Nonionic Monomers

Background: Solutions of iodine contrast media (CM) used for selective coronary arteriography (CA) should have minimal propensity to cause ventricular fibrillation (VF). Commonly used CM for CA are nonionic monomers or dimers. Purpose: To compare VF propensity of ready-to-use solutions of one nonionic dimer, iodixanol, and five nonionic monomers, iobitridol, iopamidol, iomeprol, iopromide, and ioversol. Material and Methods: Twenty milliliters of each CM was injected into the left coronary artery (LCA) through an inflated balloon catheter (0.5 ml/s) in 14 pigs; the longest period of injection was 40 s. If VF occurred before 40 s, the injection was stopped and the heart was defibrillated. After VF, there was a delay of 40 min before the next injection. Hemodynamic parameters and vector electrocardiography (VECG) were monitored. A CM with a lower frequency of VF and a longer period between start of injection and start of VF was considered to have a lower VF propensity. Results: Following 14 injections, each of the five nonionic monomers caused 14 VF, whereas iodixanol caused three VF (P<0.01). When VF occurred after iodixanol, it occurred later than after the other CM (P<0.001). Iodixanol caused less prolongation in QRS time (P<0.01) and QTc time (P<0.05) than the other CM. Prolongations in QRS and QTc times caused by CM parallel the VF propensities of the CM. Conclusion: Ready-to-use solutions of the dimer iodixanol have lower VF propensity than solutions of the five monomeric CM. This is related to the fact that the solutions of the dimer iodixanol have lower osmolality, higher viscosity, and higher concentrations of NaCl and CaCl2 than solutions of the five monomers.

Biliary and total extrarenal clearance of inulin and iohexol in pigs. A source of error when determining gfr as body clearance.

Biliary clearance, total extrarenal clearance, body and renal clearance of inulin and iohexol were determined in 11 normal and 11 nephrectomized pigs. The biliary clearance of inulin, calculated as biliary excretion divided by the plasma concentration, was 0.04 and 0.01 ml min–1 10 kg–1 and of iohexol 0.21 and 0.1 ml min–1 10 kg–1 , in normal, respectively, nephrectomized pigs (p < 0.05). The extrarenal clearance of inulin, calculated as body minus renal clearance, was 2.7 and 0.7 ml min–1 10 kg–1 and of iohexol 3.7 and 0.7 ml min–1 10 kg–1 in normal, respectively, nephrectomized pigs (p < 0.05). Some hours after injection of the markers their plasma concentrations were much higher in the nephrectomized pigs. This higher plasma concentration was not matched by an equally higher biliary excretion and therefore biliary clearance decreased. The smaller total extrarenal clearance in nephrectomized pigs, i.e. the overestimation of GFR when calculated as body clearance, indicates that this source of error decreases with decreasing renal function.