Abstract
Iodinated contrast media (ICM) are widely used for diagnostic and interventional procedures in radiology and cardiology. Ideally, they should not interact with blood cells or vascular wall cells to avoid deteriorations of the blood circulation. However, it is well known that ICM can affect erythrocytes as well as endothelial cells which consequently might perturb especially the microcirculation. In former studies the influence of two ICM (iodixanol versus iopromide) on the vascular system, the development of blood stasis, on changes in renal resistive index (RRI) and vascular diameters, and on the post-mortem distribution of iodine as marker for ICM in the explanted kidneys was examined. The modus of ICM application into the supra-renal aorta followed the regime in interventional cardiology, so that 10 bolus injections were administered at steady intervals (iopromide 4,32 ml / iodixanol 5 ml) accompanied by infusion of 500 ml isotonic NaCl-solution.
In the present study, the post-mortem X-ray analysis revealed that there were no differences in iodine content in the regions of the mid-cortex and the medullo-pelvic transition zone of the kidneys after application of both ICM. Remarkable differences, however, were found in the region of the capsule-near cortex, where the application of iopromide led to a significantly lower iodine content in the microcirculation. This is in good agreement with former studies, in which a maldistribution in this area, presumably due to a decrease in arteriolar inflow as a result of stasis/occlusion was shown.
Introduction
ICM are widely used to visualize vascular structures and parenchymal changes e.g. in radiology and cardiology. Their effects on blood vessels should be minimal [1, 2] so that the object of the measurement is not influenced by the ICM molecules. However, all ICMs exhibit some effect on endothelial cells [2–6] as well as on blood cells [7–13], and iopromide led to significantly stronger impairment of the morphology and function of blood cells, cells of the vascular wall and of the microcirculation than iodixanol. Furthermore, after 10 bolus injections of iopromide, blood stasis was detected in vessels of all porcine kidney regions and of all vascular diameters. After iodixanol, there were significantly fewer observations of stasis, and none at all in the mid-cortex region [14]. Scanning Electron Microscopy (SEM) revealed, in only a few cases, mural platelet aggregates within minimal fibrin meshes, but only after the application of iopromide.
Contrast enhanced ultrasound (CEUS) with perfusion analysis showed that time-to-peak (TTP) values in the cortex in the iopromide-group were significantly prolonged after the tenth CM bolus compared to the iodixanol-group [15, 16]. A post-hoc analysis showed that the mean times to peak-intensity were significantly longer and thus the regional blood flow lower in the iopromide-group compared to the iodixanol-group after the tenth CM bolus (p < 0.001).
This was in line with results from in vivo measurements of the oxygen partial pressure in the cortico-medullary region. While the intra-renal pO2-values showed a tendency to increase after iodixanol injection, the pO2-values showed a tendency to decrease after the tenth iopromide injection and then to remain at a lower level without repeat increase of pO2-values [17], although the temporarily increased Renal Resistive Index values (RRI-values) had returned to starting values.
The different measurement methods used consistently show that the administration of several iopromide bolus injections resulted in a maldistribution of the microcirculation in the cortex of the pig kidney, which became evident in vivo by the marbling of the kidney surface.
The reduced blood flow in the cortex - both blood velocity and vessel diameters - were reduced after application of iopromide compared to iodixanol [15], could be the reason for a lower iodine content in this region as less iodine was transported into the microvasculature The present study examined, therefore, whether a difference in the postmortem iodine content in the kidney cortex could be proven.
Materials and methods
Study design
The study was performed as a prospective, randomized examination to compare the effects of two CM (iodixanol vs iopromide) on n = 16 German Landrace pigs with a body weight between 30–35 kg, approximately three months of age. The weight of the explanted kidneys ranged between 73 and 96 g. There were no significant age, body or kidney weight differences between both groups.
A detailed description of the study design was published earlier [15]. Group I received iodixanol (n = 8 animals), group II received iopromide (n = 8 animals). Simulating the clinical procedure, each animal received a total of 10 CM injections into the aorta, either with 5 ml iodixanol per injection or with 4.32 ml iopromide per injection, at a volume rate of 10 ml/sec, so that both groups received equal amounts of iodine [18]. The 10 injections were applied to the suprarenal part of the distal abdominal aorta. Five minutes passed between two injections. Each animal received a total of 500 ml NaCl throughout the entire examination, which also follows the clinical procedure. All examinations were carried out under general anesthesia, which was administered by a board-certified anesthetist with great experience in pig anesthesia. The Bavarian Institutional Animal Care and Use Committee approved the study protocol for the experiments performed in this study (number: 54-2532.1-31/13). All procedures were carried out in accordance with the EU Directive 2010/63/EU for animal experiments.
Iodinated contrast media
Two standard CM were applied, iodixanol 320 mg iodine/ml, GE Healthcare, Boston, MA, USA, and iopromide 370 mg iodine/ml, Bayer/Schering, Berlin, Germany.
Radiography
Because of the observed remanence of X-ray contrast medium (ICM) in kidney tissues, one focus of the investigation was whether ICM-associated iodine is found in kidney tissue in the period of 15 to 30 minutes after application of 10 ICM bolus injections. This allowed a direct comparison of the iodine distribution of the two ICMs in the kidney tissues using the “dual energy CT” detection methods (sectional image and total image with anterior-posterior positioning of the organs in the CT) (see Fig. 1).
Transverse section through the middle region of the right kidney by dual energy CT after 10 ICM bolus injections and 20 minutes after explantation.
Using dual energy CT (Somatom, Definition Flash CT Scanner, Siemens, Erlangen. Slice thickness 1 mm, 128 lines), the ICM transport through the kidney was followed. This was done in identically sized “regions of interest” (ROI) in the capsule-near cortex, mid cortical region and medulla-pelvis transition, in which the iodine density was estimated according to the available grey values. The iodine distribution in kidney tissues was recorded, in each case in five equidistant sections of the kidney.
On the one hand, this made it possible to follow ICM transport through large renal vessels, glomeruli, peri-tubular capillaries, tubules and collecting tubes/renal pelvis. On the other hand, the post-explantation iodine density in different kidney regions between the distal and proximal kidney pole could be measured. The Hounsfield scale is used in CT to describe the attenuation of X-ray radiation in tissues and is represented in grey scales (HU: Hounsfield Unit). The images of the explanted kidneys were taken at low-energy setting (30kVp).
Depending on the type and content, X-ray contrast media have values between 100 and 500 [HU] in humans. HU values were recorded and averaged in 5 sectional planes (A–E) in 3 different regions (capsule-near cortex, mid-cortical region, medulla-pelvis transition) of the kidney - according to Fig. 2.
X-ray of the kidney with the 5 sectional planes.
Data are described with arithmetic mean and standard deviation. For the comparison of the effects of the two iodinated contrast media the t-test for independent samples was used. P-values lower than 0.05 were considered significant.
Results
Iodine contents in [Hounsfield units, HU] were quantified in five different sectional planes equidistantly spaced from the proximal to the distal pole (Table 1).
Iodine contents in Hounsfield units [HU] in five different sectional planes A-E from proximal to distal pole and in the capsule-near cortex, the mid-cortical region and the medulla-pelvic transition, respectively (mv: mean value)
Iodine contents in Hounsfield units [HU] in five different sectional planes A-E from proximal to distal pole and in the capsule-near cortex, the mid-cortical region and the medulla-pelvic transition, respectively (mv: mean value)
HU in the capsule-near cortex were significantly higher after 10 intra-aortic injections of iodixanol with 151±15.2 HU than after iopromide with 123.6±8.4 HU (p = 0.004), whereas there were neither differences in the mid-cortical region (p = 0.139) nor in the medulla-pelvis transition (p = 0.931).
The study revealed that the X-ray-assessed iodine content in [HU] in the capsule-near cortex of animals exposed to 10 bolus injections of iopromide into the suprarenal aorta was significantly lower (18.2%) than after iodixanol application where no significant changes occurred (p = 0.004). A reduced arterial inflow, as shown in a study with contrast-enhanced ultrasound, is the most likely cause [15]. Furthermore, a SEM study showed that serious damages of vascular walls with the development of thrombi in meshes of fibrin fibers occurred only after iopromide application (not after iodixanol application) [15]. Whether occlusions in arteriolar vessels happened alone or also in venular vessels could not be ascertained in this study. Occlusions in arteriolar vessels would certainly affect a retardation/decrease in blood inflow into the kidney cortex. The supply of the cortical microcirculation pathway in the kidney is provided exclusively via end arteries [19]. Occlusions in the supplying arteries would therefore lead to a failure in the downstream microcirculation. This has already been demonstrated intravitally visually by transient marbling of the renal surface and postmortem by mammography [15].
Ultrasound measurements by CCDS of the renal resistance index (RRI) in the renal artery of the pig showed that the flow resistance increased after iopromide application, but did not remain elevated permanently and returned to normal after a few minutes [20]. Treitl had shown comparable RRI developments over time after ICM application in patients [18]. Chronologically, the decrease of the RRI increase after iopromide application was in good agreement with the disappearance of the marbling on the kidney surface [15] and the normalization of blood flow assessed by CCDS [15]. Comparable time courses of capillary perfusion at the ipsilateral nail fold of coronary patients after ICM application in vivo were assessed microscopically, and the ICM-induced decrease of the oxygen partial pressure in the myocardium of pigs in vivo was assessed oxymetrically [21, 22]. It was initially assumed that the changes in blood circulation that occurred after ICM application practically completely returned to baseline values.
However, post-mortem SEM-examinations in the porcine kidney showed that not all vascular changes were reversible [14]. Especially in the areas of the cortico-medullary transition and the mid-medulla, stasis (also occlusions, but without fibrin formation) was still detectable post-mortem - however, significantly more often after iopromide than after iodixanol [14]. In contrast, the frequency of stasis was relatively low in the mid-cortical region and even less in the capsule-near cortex. There were practically no differences between the local iodinated iodine contents [HU] found in the middle cortex after iodixanol and after iopromide application. In the capsule-near cortex, however, clear differences in the local iodine contents occurred after application of the two ICM (see Table 1). It can be speculated that the performance of the genetically differing glomerular endothelial cells could play a decisive role here. Hennigs et al. reported that the glomerular endothelial cells are capable of a significant increase in NO release, e.g. through expression of the Gata5 gene. [23]. This effect could lead to a reduced influence on thrombus formation with simultaneous vasodilation in the downstream vascular regions - from the middle cortex to the medulla. [14]. In addition, iopromide is known to induce platelet activation and aggregation [24, 25], which could contribute to an increased thrombus formation after iopromide administration [14].
Also, the increasing iodine levels from the proximal to the distal kidney pole for both groups are very striking (Table 1). This could be related to a better blood supply due to the arterial vascularization of the proximal pole. While the proximal pole is regularly supplied by two arterial branches, the distal pole is usually supplied by the ramus principalis anterior, i.e. only by one artery [26, 27]. The extent to which this leads to better irrigation of this renal area must remain a hypothesis here.
Conclusion
In this examination in pigs, it could be clearly shown via post-mortem X-ray analysis that after ten bolus applications of a ICM into the suprarenal aorta, iopromide (4,32 ml bolus) led to a significantly lower iodine content only in the microcirculation of the capsule-near cortex compared to iodixanol (5 ml bolus). This is in good agreement with a maldistribution in this area, presumably due to a decrease in arteriolar inflow as a result of stasis/occlusion.
Funding
The study was supported by a grant from GE Healthcare (AZ.: 54-2532.1-31/13 University of Regensburg).