- Email: [email protected]
Evaluation of Renal Arteries with Use of Gadoterate Meglumine-, CO2-, and Iodixanolenhanced DSA Measurements versus Histomorphometry in Renal Artery Restenosis in Rabbits
Evaluation of Renal Arteries with Use of Gadoterate Meglumine-, CO2-, and Iodixanolenhanced DSA Measurements versus Histomorphometry in Renal Artery Restenosis in Rabbits Alain-Ferdinand Le Blanche, MD, PhD, Marc J. Bazot, MD, Michel Bonneau, PhD, Maria-Theresa Farres, MD, Michel Wassef, MD, Bernard Levy, MD, PhD, Jean-Michel Bigot, MD, and Frank Boudghene, MD, PhD
PURPOSE: To experimentally evaluate gadolinium (Gd)-, carbon dioxide (CO2)-, and iodixanol-enhanced digital subtraction angiography (DSA) versus histomorphometry in the assessment of renal artery stenosis. MATERIALS AND METHODS: Fifteen male New Zealand White rabbits weighing 4.0 kg underwent percutaneous catheterization. Renal artery stenosis was induced by bilateral overdilation-deendothelialization (balloon diameter ⴝ 2 mm). The percentage of artery overdilation was 33%. After 4 weeks, the rabbits were randomized into two groups: group A underwent right-sided therapeutic percutaneous transluminal renal angioplasty (PTRA) (balloon diameter ⴝ 1.5 mm). After another 4 weeks, the renal arteries were evaluated by gadoterate-, iodixanol-, and CO2-enhanced selective quantitative DSA. The rabbits were then killed and renal arteries were perfusion-fixed for 60 minutes. Serial orcein-stained 4-um-thick slices were prepared for histomorphometry. RESULTS: Based on morphometric data of single-stenosis versus post-PTRA restenosis lesions, no significant difference was observed between Gd- and iodixanol-enhanced quantitative DSA (r2 > 0.95), although the iodine/Gd density ratio was equal to 3.5. Carbon dioxide less reliably allowed quantitative DSA (r2 < 0.75). CONCLUSION: Gd-based contrast agents represent a highly reliable alternative in experimental quantitative DSA evaluation of renal artery restenosis. Index terms: Animal model
•
Angiography, contrast media
•
Renal arteries, angioplasty
•
Renal arteries, stenosis or obstruction
J Vasc Interv Radiol 2001; 12:747–752 Abbreviations: DOTA ⫽ 1, 4, 7, 10-tetraazacyclodecane-N,N1,N11,N111-tetraacetic acid, DSA ⫽ digital subtraction angiography, EL ⫽ elastic lamina, Gd ⫽ gadolinium, PTRA ⫽ percutaneous transluminal renal angioplasty
RECURRENT stenosis, or restenosis, is a mid-term complication of percutaFrom the Department of Radiology and Medical Imaging (A.F.L.B.), Charles Foix–Jean Rostand University Hospital, Ivry-sur-Seine; Centre de Recherche en Imagerie d’Intervention (M.B.), AP-HP and Institut National de la Recherche Agronomique, Jouy-en-Josas; Department of Radiology and Medical Imaging (M.J.B., M.T.F., J.M.B., F.B.), Hoˆpital Universitaire Tenon; and Department of Pathology (M.W.) and Institut National de la Sante´ et de la Recherche Me´dicale (INSERM) (B.I.L.), Hoˆpital Universitaire Larboisie`re, Paris, France. Received September 28, 2000; revision requested November 10; revision received January 5, 2001; accepted January 12. Address correspondence to A.F.L.B., Departments of Radiology and Medical Imaging, Charles Foix–Jean Rostand University Hospital, AP-HP 7 Avenue de la Republique, F-94205 Ivry-sur-Seine, France; E-mail: [email protected] © SCVIR, 2001
neous transluminal renal angioplasty (PTRA) (1). The immediate post-PTRA status of renal arteries is evaluated by iodine-enhanced digital subtraction angiography (DSA), but iodinated contrast agents can impair renal function in patients with renal insufficiency (2) or may induce adverse reactions (3). Renovascular disease is the underlying cause of renal insufficiency in as many as 14.3% of patients who enter hemodialysis programs each year (4). PTRA is appropriate for renal revascularization, but its outcome would be optimized by the use of a renal-safe contrast agent, allowing reliable stenosis measurements by quantitative DSA. Alternative contrast agents exist, CO2 being currently used for arterial opacification, but air leaks
may, rarely, induce a risk of gas embolism for the patient (5). Because CO2 angiography was initially restricted to a few specific centers because of the expensive dedicated injector required, a new technique involving hand injections with a plastic bag delivery system has been developed (6). Overall, the most crucial issue is the reliability of CO2 in detecting significant renal artery stenosis. Recently, gadolinium (Gd)-enhanced angiography was successfully performed in lower limb or selective angiography (7,8). Although cases of acute renal failure after arteriography with a Gd-based contrast agent have been reported, it was not clear whether renal insufficiency resulted from extracellular dehydration by concomitant overload of furo-
747
748
•
Gadolinium, CO2, and Iodinated Contrast for Renal Artery Stenosis
semide (9), unexpected worsening of preexisting transplant rejection (10), or cholesterol emboli during endovascular procedures (11). Indeed, Gd complexes have been shown to have no significant influence on renal function (12–14). Despite their relatively high cost compared to nonionic iodinated contrast agents, it seems clinically relevant to experimentally evaluate Gd complexes in vascular interventional radiology to possibly protect frail patients from the adverse effects of iodinated contrast agents, the potential expense of dialysis complications, or the hazardous use of CO2. A new model of endovascularly induced renal artery stenosis was recently developed in rabbits (15). This in vivo experimental study was designed to evaluate induced renal artery restenosis with use of DSA with gadoterate, iodixanol, or CO2 enhancement.
MATERIALS AND METHODS Animals and DSA Survey Experiments were performed in accordance with the European Community rules of animal care (16). Generally, the higher the voltage used during performance of contrast-enhanced DSA, the sharper the image pixel delineation. We performed a pilot study involving five adult male New Zealand White rabbits to determine the voltage providing ideal contrast in gadoterate-enhanced DSA. Seventy-three kilovolts and a cathode current energy of 100 mA were found to be adequate values and constituted the standard parameters used thereafter. Fifteen adult male New Zealand White rabbits weighing 4.0 kg (range ⫽ 3.5– 4.2), fed a regular rabbit chow ad libitum, were then included in the study. Model Induction After fasting for 24 hours, the rabbits were sedated by intramuscular injection of 30 mg/kg of acepromazine (Calmivet; Vétoquinol, Lure, France), and were anesthetized by intravenous auricular injection of 15 mg/kg of sodium thiopental (Nesdonal; Rhône-Poulenc Rorer, Paris, France), diluted to 10% in saline solution. A previously reported model of endovascularly induced renal artery stenosis (15) was adapted to
avoid arteriotomy. In our variant, a femoral approach was performed percutaneously with a 18-gauge needle, then a 0.021-inch guide-wire and 4-F sheath (Terumo, Tokyo, Japan) were introduced into the right mid-femoral artery. The femoral artery puncture site was determined by palpation. Aortography was then performed above the renal arteries with a 4-F angiography catheter, including two perpendicular projections, with injection of 8 mL iodixanol (Visipaque; Nycomed Amersham, Oslo, Norway) at a concentration of 320 mgI/mL with a flow rate of 2 mL/sec. Multiple renal arteries may influence the vessel selection process and were systematically searched, but we have never observed any accessory renal arteries during our 4-year experience with renal DSA in rabbits (n ⫽ 110). PTRA procedures were performed in the center of the radiographic beam with use of a 14-cm field of view to achieve precise balloon placement. Overdilation and de-endothelialization were performed bilaterally in the renal arteries with a monorail 2-mm-diameter ⫻ 20-mm Manfield coronary angioplasty balloon catheter (Boston Scientific/Medi-tech, Watertown, MA) inflated to 10 atm with a pure commercially available solution of Gd and 1,4,7,10-tetraazacyclodecaneN,N1,N11,N111-tetraacetic acid (DOTA) concentrated at 0.5 mmol/mL (Dotarem; Laboratoire Guerbet, Aulnaysous-Bois, France). Three inflations were performed at 10 atm for 30 seconds, separated by 1-minute deflation periods to restore renal perfusion. The balloon was advanced into the renal artery over a 0.014-inch guide wire (Terumo) so that its middle (an annular thin opaque marker on the catheter shaft) extended 20 mm beyond the ostium. The distance separating the ostium from the proximal extremity of the balloon was 10 mm. The inflated balloon was then alternately pushed 5 mm forward and pulled 5 mm backward under fluoroscopic control, but never completely withdrawn when inflated within the first 5 mm of the renal artery, to preserve a 5-mm-long nonstenotic reference segment. After completion of the third inflation, the angioplasty balloon was deflated and withdrawn. The 4-F angiography catheter was then replaced and another aortogram was acquired to verify the immediate postprocedural patency of the renal arteries and intrarenal vasculature. Gadoterate-enhanced
June 2001
JVIR
DSA of the suprarenal aorta was performed with injection of 6 mL GdDOTA at a 2 mL/sec flow rate. The arterial access material was then withdrawn and manual soft compression of the groin was applied for 10 minutes. After recovery, the rabbits were kept in under unchanged dietary conditions. Therapeutic PTRA at 4 Weeks After 4 weeks, the rabbits were prepared as described earlier. Kidneys and renal arteries were visualized by gadoterate-enhanced DSA. The rabbits were then randomized to two groups for unilateral treatment of previously induced renal artery stenoses: group A (n ⫽ 8) underwent right PTRA and group B (n ⫽ 7) underwent left PTRA, and the results were compared to the untreated contralateral stenoses. The induced stenoses were visualized and measured by renal angiography. Unilateral therapeutic PTRA was then performed on the renal artery determined by randomization with use of a monorail 1.5-mm-diameter ⫻ 20-mm Cruiser II coronary angioplasty balloon catheter (Nycomed Amersham) inflated to 10 atm for 30 seconds. After three inflations, a second angiogram was then acquired to control the immediate postprocedural result of the dilation. Eight-week Evaluation of Stenoses by Angiography and PerfusionFixation Four weeks later, the rabbits were prepared as described earlier. Renal arteries and kidneys were visualized by selective DSA with injection of GdDOTA, iodixanol, and CO2, in the supine position. Gd-enhanced renal selective DSA was performed with automatic injection of 2 mL gadoterate meglumine at a 2 mL/sec flow rate. Iodine-enhanced DSA was performed with use of 2 mL iodixanol. A pneumatic chamber dedicated to CO2 angiography, including two three-way stopcocks, was used to obtain a sealed, air-free CO2 delivery system (6). Fifteen to twenty mL of CO2 were injected manually with a 20-mL syringe (Luerlock; Becton Dickinson, San Jose, CA). The rabbits were then deeply anesthetized. The renal arteries were exposed by blunt dissection through a midline abdominal incision, which was extended
Volume 12
Number 6
Le Blanche et al
•
749
was calculated as a percentage. Calibration software (Polytron T.O.P.; Siemens, Erlangen, Germany) takes into account the object-to-image-intensifier distance to perform accurate diameter measurement and automatic stenosis calculation with negligible distortion, provided the lumen is positioned in the central beam with a table-beam angulation less than 15°. The software takes into account the maximum opacification density in cases of positive contrast agents (iodine, Gd) versus the minimum density in cases of negative contrast agents (CO2). For each rabbit, software validation was obtained when the diameter value of the 4-F angiography catheter diameter corresponded to 1.4 mm. These measurements were repeated three times and averaged for each stenosis quantification. Figure 1. Correlation graph of stenosis measurement by various contrast agents. Twenty-seven stenoses were measured (% stenosis) by automated quantitative angiography. (a) When Gd-DOTA (ordinate) is compared with iodixanol (abscissa), no significant difference is observed between the contrast agents (r2 ⬎ 0.95). (b) Conversely, when CO2-opacification (ordinate) is compared with iodixanol (abscissa), the correlation is revealed to be lower (r2 ⬍ 0.75).
caudally and laterally at the level of the kidneys. The rabbits were killed by exsanguination at the suprarenal aorta. The supra- and infrarenal aorta and vena cava were isolated by ligation and cannulated with a 14-gauge angiocatheter (Introcan; B. Braun, Melsungen, Germany), then 10% buffered formalin (Sigma Chemical, St. Louis, MO) was infused into the aorta for 60 minutes at 100 mm Hg to fix the arteries in expansion and kidneys, both of which were left in place inside the abdomen. Renal veins were cut to evacuate formalin. The supra- and infrarenal aorta, renal arteries, and kidneys were then removed and placed in 10% buffered formalin for pathologic and morphometric examinations. The right kidney was marked by a sagittal cut. Analysis of the Angiographic Data Two senior interventional radiologists (A.F.L.B, M.J.B.) with daily involvement in peripheral vascular practice and who were blinded to the morphometric results, independently
Density Measurements Densitometric data were directly available from the stenosis measurement software by measurement of pixel values. For each reference and stenosis segment, the density of the opacified lumen area was simultaneously displayed when stenosis measurement was activated. Figure 2. Comparison of stenosis measurement by automated quantitative angiography versus histomorphometry by regression analysis (n ⫽ 27). When iodixanol (a) or Gd-DOTA (b) are compared with morphometric data, the estimation of the stenoses is highly correlated (0.94 ⬍ r2 ⱕ 0.99). (c) CO2-opacification is revealed to provide less reliable measurements (r2 ⬍ 0.73).
analyzed the angiograms. They noted which branches of the intrarenal arterial vasculature were opacified with the three contrast agents. Stenosis Measurements Artery diameters were measured by an automated edge detection algorithm. Because the balloon was deflated before being withdrawn from the renal artery during induction of stenosis, the reference segment of the stenotic artery was measured between the angioplasty site and the ostium. All diameter measurements were then obtained at the opacification peak. The preprocedural reference diameter was measured and the postprocedural stenosis of the renal artery lumen
Tissue Analysis The sections were examined by two experienced observers (M.W., B.I.L.) blinded to the side of restenosis and the results of the correlative study. Tissue Harvest and Histology An embedding technique was used. Each renal artery was cut into seven (right artery) or eight (left artery) 3– 4mm-long segments. An aorta-sided ink label indicated the proximal pole of the sample. Another similar baseline landmark represented the distal extremity of each segment. Each identified lesion of the renal artery was therefore localized (in mm from the ostium) and the balloon middle area was easily correlated. Four-m-thick serial sections of renal artery were sampled transversely (one series of eight slices per 250-m-long portion of each artery) and dehydrated in graded ethanol. Among all slices, 40 were part of the renal artery segment close to the balloon middle, and two were reference segments. Twenty slices (plus one reference segment) were stained with he-
750
•
Gadolinium, CO2, and Iodinated Contrast for Renal Artery Stenosis
matoxylin-eosin-safranin for histologic study, and 20 slices (plus one reference segment) were stained with orcein for histomorphometric analysis. Each specimen was evaluated for the presence of neointima formation, medial dissection, alteration of the internal elastic lamina (EL) and external elastic lamina, and morphologic appearance of the adventitial, medial, and neointimal layers. Sections were also evaluated for the presence of intraluminal thrombus, intraluminal hemorrhage, and inflammatory cells.
was calculated as an injury severity index (17). The extent of stenosis was determined as percent stenosis ⫼ reference segment ⫽ 100% ⫻ (1 ⫺ stenosis lumen area ⫼ reference segment lumen area). The medial area-to-luminal area ratio (hypertrophy index) was calculated in reference segments and injured renal artery segments to assess the degree of neointimal formation after 8 weeks.
Histomorphometry
DSA evaluation of the postprocedural response with use of various contrast agents was compared by a Wilcoxon signed rank test. Data were expressed as means ⫾ SD. Linear regression was used to (i) compare Gd-DOTA and CO2 with iodixanol angiographic data and (ii) compare the morphometric restenosis data directly with angiographic stenoses determined for the three agents. A value of probability P ⬍ .05 was required to reject the null hypothesis at the 95% confidence level. Statistical analysis was performed with StatView 4.5 version (Abacus Concepts, Berkeley, CA).
A stenotic lesion identified on microscopy was easily visualized on angiography. Prestenotic (ie, reference segment) and stenotic segments of the injured renal arteries were examined and compared to prestenotic and stenotic segments of the treated injured renal arteries. Neointima was defined as any tissue layer circumscribed by the internal EL, excluding intraluminal fibrin deposits and larger organized mural thrombus material. Other specific changes were defined as reshaping of the medial layer, tearing of one of the elastic laminae, and/or presence of adventitial fibrosis. All transverse sections were digitized via a Cohu-CCD video camera, then morphometrically analyzed with use of MicroVision software (Evry, France). In each artery, measurements were performed in the proximal healthy reference segment and in 20 pathologic slices. The areas of intima, media, and residual lumen as well as circumferences demarcated by internal and external EL were determined. The length of the internal EL fracture through the neointima, extending from one end of the dissected internal EL to the other, was used as a marker of the severity of the injury. Area measurements were performed. The vessel perimeter was delineated by the external EL, thereby defining the vessel area. The lumen perimeter defined the lumen area, and the internal EL perimeter defined an area including neointima and lumen (internal EL area). Media area was obtained as the difference between vessel area and internal EL area, and neointima area was obtained as the difference between internal EL area and lumen area. The neointima-to-internal fracture length ratio
Data Processing
RESULTS Stenosis Quantification No significant difference was observed between gadoterate- and iodixanol-enhanced DSA for automated quantification of stenoses (Fig 1a). Unfortunately, automated quantification was not highly correlated (r2 ⬍ 0.75) when the arteries were opacified by CO2 (Fig 1b). Based on morphometric data, the estimation of the stenoses was highly correlated (0.94 ⬍ r2 ⱕ 0.99) with iodixanol (Fig 2a) and GdDOTA (Fig 2b), whereas CO2 (Fig 2c) proved to be less reliable (r2 ⬍ 0.73). Comparison of Contrast Media in DSA Gd-DOTA allowed visualization of the main renal arteries and collateral branches (Fig 3a). Gd-DOTA visualized arteries of the superior renal pole in all rabbits just as well as the iodinated gold standard, iodixanol. Intrarenal vasculature was visible as far as third-generation branches, whereas the fourth generation was more easily
June 2001
JVIR
identified by iodixanol opacification (Fig 3b) on all examinations. Stenosis and poststenotic dilation were evidenced by CO2-enhanced angiography, whereas only second-generation arterioles were identified inside the renal parenchyma (Fig 3c). Density Quantification Under similar injection conditions, iodixanol was found to be 3.5 times as dense as the Gd complex (Table). However, this density gradient between the two contrast agents did not significantly interfere with the measurement of stenoses, and the regression coefficient (Fig 1a) was high (r ⬎ 0.97).
DISCUSSION To date, renal opacification by gadoterate, iodixanol, and CO2 has not been simultaneously quantitatively evaluated in vivo in a model of restenosis. Kidneysafe alternatives to iodine-enhanced DSA do exist, represented by CO2-enhanced DSA (10,11,18) or MR angiography (19). Unfortunately, CO2-enhanced DSA provides few advantages in terms of contrast and measurement reliability (20). The use of CO2 was not so appropriate in this animal study, particularly in the challenge of evaluating stenoses in small renal arteries. Our data may be related to the size of the study subjects and may not be completely exportable to humans. However, it appeared interesting to compare contrast agents on the basis of experimental conditions of challenge in terms of vessel size. In particular, the poor contrast of CO2 prevented accurate measurement of stenoses in the present study. Its widespread use is associated with restricted indications, such as in vascular interventional procedures in patients with renal impairment or who have contraindications to magnetic resonance (MR) imaging. Although the large majority of daily use of CO2 appears to be very safe, recent studies evaluated the risk of air leak leading to intraarterial gas embolism (5). This risk should lead to a recommendation of its careful use in organs with poor anastomotic arterial blood supply (eg, kidney, brain) (21), as it has been responsible, albeit rarely, for severe complications such as mesenteric ischemia (5,22,23). MR angiography has therefore become the preferred noninvasive modality to evaluate all arterial
Volume 12
Number 6
Le Blanche et al
•
751
Figure 3. Comparison of selective angiograms of a 72% renal artery stenosis (on the basis of histomorphometric data). (a) The Gd complex allowed opacification of the main and polar vessels. Intrarenal arterial vascularization is visible as far as the third-generation branches (thin black arrows). Evidence of poststenotic dilation is clearly demonstrated (arrowheads). Automated quantification allows reliable measurement of the stenosis (white line), compared to the reference segment (black line). (b) Iodine-enhanced angiography provides three-fold higher density and visualizes fourth generation branches (arrowheads). Automated quantification was nevertheless identical to the previous measurement. (c) With CO2 angiography, stenosis (black arrowhead), poststenotic dilation (white arrowhead), and polar artery (white arrow) remain visible, whereas only 2nd generation arterioles are identified inside the renal parenchyma (black arrow).
DSA Comparative Contrast Density Data of the Renal Artery Reference Lumen (N ⴝ 27) Contrast Agent Gd-DOTA Iodixanol Iodixanol/ Gd-DOTA*
Density 364 ⫾ 129 1167 ⫾ 355 3.49 ⫾ 1.25
Note.—Values expressed as means ⫾ SD. * Wilcoxon signed rank test, P ⬍ .0001.
territories. Nevertheless, MR evaluation of small artery stenoses remains difficult because this type of imaging is dependent on flow velocity. It remains less effective than DSA in the measurement of stenoses, particularly in case of distal arteries, and the diameter of a stenosis may be either over- or underestimated (19). Conversely, automatic stenosis quantification performed with DSA data has become very reliable, based on geometric and densitometric data. In the case of renovascular hypertension or ischemic nephropathy, when there is a clinical indication for renal revascularization, therapeutic decisions regarding PTRA and/or stent placement in renal arteries are based on stenosis grade, even though transstenotic pressure measurements, the most accurate method for determining the significance of renal artery stenoses, can be safely performed in almost all circumstances. In light of all
these arguments, it seems attractive to use the currently available techniques to develop DSA enhanced by a noniodinated, nongaseous radiographic dense contrast agent, and Gd complexes appear to be an interesting alternative (10,11,18,24,25). However, the efficacy of Gd-enhanced DSA depends predominantly on digital subtraction and would be dramatically impaired in conventional angiography. This study demonstrates that Gd-based contrast agents may reliably replace iodinated contrast agents for stenosis quantification in selective DSA. When the use of iodinated contrast agents is hazardous for renal function, Gd-enhanced DSA may represent an attractive alternative for the measurement of stenotic lesions before and after PTA, in contrast with CO2, which fails to provide accurate measurement of stenosis. Acknowledgments: The authors thank Jean-Pierre Djemat for animal care, Professor Raymond Ardaillou from the National Academy of Medicine for experimental support, Suzette Pereira-Freire and Sandrine Boucheteil for literature research assistance, and Anthony Saul, MD, for manuscript review. References 1. van de Ven PJ, Kaatee R, Beutler JJ, et al. Arterial stenting and balloon angioplasty in ostial atherosclerotic renovascular disease: a randomised trial. Lancet 1999; 353:282–286. 2. Solomon R. Contrast-medium-induced renal failure. Kidney Int 1998; 53:230–242.
3. Lasser EC, Lyon SG, Berry CC. Reports on contrast media reactions: analysis of data from reports of the US food and drug administration. Radiology 1997; 203:605– 610. 4. O’Neil EA, Hansen KJ, Canzanello VJ, Pennell TC, Dean RH. Prevalence of ischemic nephropathy in patients with renal insufficiency. Am Surg 1992; 58: 485– 490. 5. Yang X, Manninen H, Soimakallio S. Carbon dioxide in vascular imaging and intervention. Acta Radiol 1995; 36: 330 –337. 6. Hawkins IF, Caridi JG, Kerns SR. Plastic bag delivery system for hand injection of carbon dioxide. AJR Am J Roentgenol 1995; 165:1487–1489. 7. Kinno Y, Odagiri K, Andoh K, Itoh Y, Tarao K. Gadopentetate dimeglumine as an alternative contrast material for use in angiography. AJR Am J Roentgenol 1993; 160:1293–1294. 8. Fobbe F, Wacker F, Wagner S. Arterial angiography in high-kilovoltage technique with gadolinium as the contrast agent: first clinical experience. Eur Radiol 1996; 6:224 –229. 9. Gemery J, Idelson B, Reid S, et al. Acute renal failure after arteriography with a gadolinium-based contrast agent. AJR Am J Roentgenol 1998; 171:1277–1278. 10. Spinosa DJ, Matsumoto AH, Angle JF, et al. Gadolinium-based contrast and carbon dioxide angiography to evaluate renal transplants for vascular causes of renal insufficiency and accelerated hypertension. J Vasc Interv Radiol 1998; 9:909 –916. 11. Spinosa DJ, Matsumoto AH, Angle JF, Hagspiel KD, McGraw JK, Ayers C. Renal insufficiency: usefulness of ga-
752
12.
13.
14.
15.
16.
•
Gadolinium, CO2, and Iodinated Contrast for Renal Artery Stenosis
dodiamide-enhanced renal angiography to supplement CO2-enhanced renal angiography for diagnosis and percutaneous treatment. Radiology 1999; 210:663– 672. Katzberg RW, Sahler LG, Duda SW, et al. Renal handling and physiologic effects of the paramagnetic contrast medium GdDOTA. Invest Radiol 1990; 25:714–719. Schuhmann-Giampieri G, Krestin GP. Pharmacokinetics of Gd-DTPA in patients with chronic renal failure. Invest Radiol 1991; 26:975–979. Prince MR, Arnoldus C, Frisoli JK. Nephrotoxicity of high-dose gadolinium compared with iodinated contrast. J Magn Reson Imaging 1996; 6:162–166. Le Blanche AF, Sibony M, Kollar A, Callard P, Bigot JM, Boudghene F. A new model of endovascularly-induced renal artery stenosis in hypercholesterolemic versus normocholesterolemic rabbits. Invest Radiol 1998; 33:322–328. EEC Directive No.86/609. Official J Eur Communities 1986; 358:1–28.
17. Waksman R, Robinson KA, Crocker IR, et al. Intracoronary radiation decreases the second phase of intimal hyperplasia in a repeat balloon angioplasty model of restenosis. Int J Radiat Oncol Biol Phys 1997; 39:475– 480. 18. Spinosa DJ, Angle JF, Hagspiel KD, et al. Feasibility of gadodiamide compared with dilute iodinated contrast material for imaging of the abdominal aorta and renal arteries. J Vasc Interv Radiol 2000; 11:733–737. 19. Prince MR, Schoenberg SO, Ward JS, Londy FJ, Wakefield TW, Stanley JC. Hemodynamically significant atherosclerotic renal artery sclerosis: MR angiographic features. Radiology 1997; 205:128 –136. 20. Moresco KP, Patel N, Johnson MS, Trobridge D, Bergan KA, Lalka SG. Accuracy of CO2 angiography in vessel diameter assessment: a comparative study of CO2 versus iodine contrast material in an aortoiliac flow model. J Vasc Interv Radiol 2000; 11:437– 444.
June 2001
JVIR
21. Hawkins IF, Mladinich CRJ, Storm B, et al. Short-term effects of selective renal arterial carbon dioxide administration on the dog kidney. J Vasc Interv Radiol 1994; 5:149 –154. 22. Caridi JG, Hawkins IF. CO2 digital substraction angiography: potential complications and their prevention. J Vasc Interv Radiol 1997; 8:383–391. 23. Spinosa DJ, Matsumoto AH, Angle JF, Hagspiel KD, Hooper TN. Transient mesenteric ischemia: a complication of carbon dioxide angiography. J Vasc Interv Radiol 1998; 9:561–564. 24. Matchett WJ, McFarland DR, Russell DK, Sailors DM, Moursi MM. Azotemia: gadopentetate dimeglumine as contrast agent at digital substraction angiography. Radiology 1996; 201:569–571. 25. Spinosa DJ, Angle JF, Hagspiel KD, Schenk WG III, Matsumoto AH. CO2 and gadopentetate dimeglumine as alternative contrast agents for malfunctioning dialysis grafts and fistulas. Kidney Int 1998; 54:945–950.
PURPOSE: To experimentally evaluate gadolinium (Gd)-, carbon dioxide (CO2)-, and iodixanol-enhanced digital subtraction angiography (DSA) versus histomorphometry in the assessment of renal artery stenosis. MATERIALS AND METHODS: Fifteen male New Zealand White rabbits weighing 4.0 kg underwent percutaneous catheterization. Renal artery stenosis was induced by bilateral overdilation-deendothelialization (balloon diameter ⴝ 2 mm). The percentage of artery overdilation was 33%. After 4 weeks, the rabbits were randomized into two groups: group A underwent right-sided therapeutic percutaneous transluminal renal angioplasty (PTRA) (balloon diameter ⴝ 1.5 mm). After another 4 weeks, the renal arteries were evaluated by gadoterate-, iodixanol-, and CO2-enhanced selective quantitative DSA. The rabbits were then killed and renal arteries were perfusion-fixed for 60 minutes. Serial orcein-stained 4-um-thick slices were prepared for histomorphometry. RESULTS: Based on morphometric data of single-stenosis versus post-PTRA restenosis lesions, no significant difference was observed between Gd- and iodixanol-enhanced quantitative DSA (r2 > 0.95), although the iodine/Gd density ratio was equal to 3.5. Carbon dioxide less reliably allowed quantitative DSA (r2 < 0.75). CONCLUSION: Gd-based contrast agents represent a highly reliable alternative in experimental quantitative DSA evaluation of renal artery restenosis. Index terms: Animal model
•
Angiography, contrast media
•
Renal arteries, angioplasty
•
Renal arteries, stenosis or obstruction
J Vasc Interv Radiol 2001; 12:747–752 Abbreviations: DOTA ⫽ 1, 4, 7, 10-tetraazacyclodecane-N,N1,N11,N111-tetraacetic acid, DSA ⫽ digital subtraction angiography, EL ⫽ elastic lamina, Gd ⫽ gadolinium, PTRA ⫽ percutaneous transluminal renal angioplasty
RECURRENT stenosis, or restenosis, is a mid-term complication of percutaFrom the Department of Radiology and Medical Imaging (A.F.L.B.), Charles Foix–Jean Rostand University Hospital, Ivry-sur-Seine; Centre de Recherche en Imagerie d’Intervention (M.B.), AP-HP and Institut National de la Recherche Agronomique, Jouy-en-Josas; Department of Radiology and Medical Imaging (M.J.B., M.T.F., J.M.B., F.B.), Hoˆpital Universitaire Tenon; and Department of Pathology (M.W.) and Institut National de la Sante´ et de la Recherche Me´dicale (INSERM) (B.I.L.), Hoˆpital Universitaire Larboisie`re, Paris, France. Received September 28, 2000; revision requested November 10; revision received January 5, 2001; accepted January 12. Address correspondence to A.F.L.B., Departments of Radiology and Medical Imaging, Charles Foix–Jean Rostand University Hospital, AP-HP 7 Avenue de la Republique, F-94205 Ivry-sur-Seine, France; E-mail: [email protected] © SCVIR, 2001
neous transluminal renal angioplasty (PTRA) (1). The immediate post-PTRA status of renal arteries is evaluated by iodine-enhanced digital subtraction angiography (DSA), but iodinated contrast agents can impair renal function in patients with renal insufficiency (2) or may induce adverse reactions (3). Renovascular disease is the underlying cause of renal insufficiency in as many as 14.3% of patients who enter hemodialysis programs each year (4). PTRA is appropriate for renal revascularization, but its outcome would be optimized by the use of a renal-safe contrast agent, allowing reliable stenosis measurements by quantitative DSA. Alternative contrast agents exist, CO2 being currently used for arterial opacification, but air leaks
may, rarely, induce a risk of gas embolism for the patient (5). Because CO2 angiography was initially restricted to a few specific centers because of the expensive dedicated injector required, a new technique involving hand injections with a plastic bag delivery system has been developed (6). Overall, the most crucial issue is the reliability of CO2 in detecting significant renal artery stenosis. Recently, gadolinium (Gd)-enhanced angiography was successfully performed in lower limb or selective angiography (7,8). Although cases of acute renal failure after arteriography with a Gd-based contrast agent have been reported, it was not clear whether renal insufficiency resulted from extracellular dehydration by concomitant overload of furo-
747
748
•
Gadolinium, CO2, and Iodinated Contrast for Renal Artery Stenosis
semide (9), unexpected worsening of preexisting transplant rejection (10), or cholesterol emboli during endovascular procedures (11). Indeed, Gd complexes have been shown to have no significant influence on renal function (12–14). Despite their relatively high cost compared to nonionic iodinated contrast agents, it seems clinically relevant to experimentally evaluate Gd complexes in vascular interventional radiology to possibly protect frail patients from the adverse effects of iodinated contrast agents, the potential expense of dialysis complications, or the hazardous use of CO2. A new model of endovascularly induced renal artery stenosis was recently developed in rabbits (15). This in vivo experimental study was designed to evaluate induced renal artery restenosis with use of DSA with gadoterate, iodixanol, or CO2 enhancement.
MATERIALS AND METHODS Animals and DSA Survey Experiments were performed in accordance with the European Community rules of animal care (16). Generally, the higher the voltage used during performance of contrast-enhanced DSA, the sharper the image pixel delineation. We performed a pilot study involving five adult male New Zealand White rabbits to determine the voltage providing ideal contrast in gadoterate-enhanced DSA. Seventy-three kilovolts and a cathode current energy of 100 mA were found to be adequate values and constituted the standard parameters used thereafter. Fifteen adult male New Zealand White rabbits weighing 4.0 kg (range ⫽ 3.5– 4.2), fed a regular rabbit chow ad libitum, were then included in the study. Model Induction After fasting for 24 hours, the rabbits were sedated by intramuscular injection of 30 mg/kg of acepromazine (Calmivet; Vétoquinol, Lure, France), and were anesthetized by intravenous auricular injection of 15 mg/kg of sodium thiopental (Nesdonal; Rhône-Poulenc Rorer, Paris, France), diluted to 10% in saline solution. A previously reported model of endovascularly induced renal artery stenosis (15) was adapted to
avoid arteriotomy. In our variant, a femoral approach was performed percutaneously with a 18-gauge needle, then a 0.021-inch guide-wire and 4-F sheath (Terumo, Tokyo, Japan) were introduced into the right mid-femoral artery. The femoral artery puncture site was determined by palpation. Aortography was then performed above the renal arteries with a 4-F angiography catheter, including two perpendicular projections, with injection of 8 mL iodixanol (Visipaque; Nycomed Amersham, Oslo, Norway) at a concentration of 320 mgI/mL with a flow rate of 2 mL/sec. Multiple renal arteries may influence the vessel selection process and were systematically searched, but we have never observed any accessory renal arteries during our 4-year experience with renal DSA in rabbits (n ⫽ 110). PTRA procedures were performed in the center of the radiographic beam with use of a 14-cm field of view to achieve precise balloon placement. Overdilation and de-endothelialization were performed bilaterally in the renal arteries with a monorail 2-mm-diameter ⫻ 20-mm Manfield coronary angioplasty balloon catheter (Boston Scientific/Medi-tech, Watertown, MA) inflated to 10 atm with a pure commercially available solution of Gd and 1,4,7,10-tetraazacyclodecaneN,N1,N11,N111-tetraacetic acid (DOTA) concentrated at 0.5 mmol/mL (Dotarem; Laboratoire Guerbet, Aulnaysous-Bois, France). Three inflations were performed at 10 atm for 30 seconds, separated by 1-minute deflation periods to restore renal perfusion. The balloon was advanced into the renal artery over a 0.014-inch guide wire (Terumo) so that its middle (an annular thin opaque marker on the catheter shaft) extended 20 mm beyond the ostium. The distance separating the ostium from the proximal extremity of the balloon was 10 mm. The inflated balloon was then alternately pushed 5 mm forward and pulled 5 mm backward under fluoroscopic control, but never completely withdrawn when inflated within the first 5 mm of the renal artery, to preserve a 5-mm-long nonstenotic reference segment. After completion of the third inflation, the angioplasty balloon was deflated and withdrawn. The 4-F angiography catheter was then replaced and another aortogram was acquired to verify the immediate postprocedural patency of the renal arteries and intrarenal vasculature. Gadoterate-enhanced
June 2001
JVIR
DSA of the suprarenal aorta was performed with injection of 6 mL GdDOTA at a 2 mL/sec flow rate. The arterial access material was then withdrawn and manual soft compression of the groin was applied for 10 minutes. After recovery, the rabbits were kept in under unchanged dietary conditions. Therapeutic PTRA at 4 Weeks After 4 weeks, the rabbits were prepared as described earlier. Kidneys and renal arteries were visualized by gadoterate-enhanced DSA. The rabbits were then randomized to two groups for unilateral treatment of previously induced renal artery stenoses: group A (n ⫽ 8) underwent right PTRA and group B (n ⫽ 7) underwent left PTRA, and the results were compared to the untreated contralateral stenoses. The induced stenoses were visualized and measured by renal angiography. Unilateral therapeutic PTRA was then performed on the renal artery determined by randomization with use of a monorail 1.5-mm-diameter ⫻ 20-mm Cruiser II coronary angioplasty balloon catheter (Nycomed Amersham) inflated to 10 atm for 30 seconds. After three inflations, a second angiogram was then acquired to control the immediate postprocedural result of the dilation. Eight-week Evaluation of Stenoses by Angiography and PerfusionFixation Four weeks later, the rabbits were prepared as described earlier. Renal arteries and kidneys were visualized by selective DSA with injection of GdDOTA, iodixanol, and CO2, in the supine position. Gd-enhanced renal selective DSA was performed with automatic injection of 2 mL gadoterate meglumine at a 2 mL/sec flow rate. Iodine-enhanced DSA was performed with use of 2 mL iodixanol. A pneumatic chamber dedicated to CO2 angiography, including two three-way stopcocks, was used to obtain a sealed, air-free CO2 delivery system (6). Fifteen to twenty mL of CO2 were injected manually with a 20-mL syringe (Luerlock; Becton Dickinson, San Jose, CA). The rabbits were then deeply anesthetized. The renal arteries were exposed by blunt dissection through a midline abdominal incision, which was extended
Volume 12
Number 6
Le Blanche et al
•
749
was calculated as a percentage. Calibration software (Polytron T.O.P.; Siemens, Erlangen, Germany) takes into account the object-to-image-intensifier distance to perform accurate diameter measurement and automatic stenosis calculation with negligible distortion, provided the lumen is positioned in the central beam with a table-beam angulation less than 15°. The software takes into account the maximum opacification density in cases of positive contrast agents (iodine, Gd) versus the minimum density in cases of negative contrast agents (CO2). For each rabbit, software validation was obtained when the diameter value of the 4-F angiography catheter diameter corresponded to 1.4 mm. These measurements were repeated three times and averaged for each stenosis quantification. Figure 1. Correlation graph of stenosis measurement by various contrast agents. Twenty-seven stenoses were measured (% stenosis) by automated quantitative angiography. (a) When Gd-DOTA (ordinate) is compared with iodixanol (abscissa), no significant difference is observed between the contrast agents (r2 ⬎ 0.95). (b) Conversely, when CO2-opacification (ordinate) is compared with iodixanol (abscissa), the correlation is revealed to be lower (r2 ⬍ 0.75).
caudally and laterally at the level of the kidneys. The rabbits were killed by exsanguination at the suprarenal aorta. The supra- and infrarenal aorta and vena cava were isolated by ligation and cannulated with a 14-gauge angiocatheter (Introcan; B. Braun, Melsungen, Germany), then 10% buffered formalin (Sigma Chemical, St. Louis, MO) was infused into the aorta for 60 minutes at 100 mm Hg to fix the arteries in expansion and kidneys, both of which were left in place inside the abdomen. Renal veins were cut to evacuate formalin. The supra- and infrarenal aorta, renal arteries, and kidneys were then removed and placed in 10% buffered formalin for pathologic and morphometric examinations. The right kidney was marked by a sagittal cut. Analysis of the Angiographic Data Two senior interventional radiologists (A.F.L.B, M.J.B.) with daily involvement in peripheral vascular practice and who were blinded to the morphometric results, independently
Density Measurements Densitometric data were directly available from the stenosis measurement software by measurement of pixel values. For each reference and stenosis segment, the density of the opacified lumen area was simultaneously displayed when stenosis measurement was activated. Figure 2. Comparison of stenosis measurement by automated quantitative angiography versus histomorphometry by regression analysis (n ⫽ 27). When iodixanol (a) or Gd-DOTA (b) are compared with morphometric data, the estimation of the stenoses is highly correlated (0.94 ⬍ r2 ⱕ 0.99). (c) CO2-opacification is revealed to provide less reliable measurements (r2 ⬍ 0.73).
analyzed the angiograms. They noted which branches of the intrarenal arterial vasculature were opacified with the three contrast agents. Stenosis Measurements Artery diameters were measured by an automated edge detection algorithm. Because the balloon was deflated before being withdrawn from the renal artery during induction of stenosis, the reference segment of the stenotic artery was measured between the angioplasty site and the ostium. All diameter measurements were then obtained at the opacification peak. The preprocedural reference diameter was measured and the postprocedural stenosis of the renal artery lumen
Tissue Analysis The sections were examined by two experienced observers (M.W., B.I.L.) blinded to the side of restenosis and the results of the correlative study. Tissue Harvest and Histology An embedding technique was used. Each renal artery was cut into seven (right artery) or eight (left artery) 3– 4mm-long segments. An aorta-sided ink label indicated the proximal pole of the sample. Another similar baseline landmark represented the distal extremity of each segment. Each identified lesion of the renal artery was therefore localized (in mm from the ostium) and the balloon middle area was easily correlated. Four-m-thick serial sections of renal artery were sampled transversely (one series of eight slices per 250-m-long portion of each artery) and dehydrated in graded ethanol. Among all slices, 40 were part of the renal artery segment close to the balloon middle, and two were reference segments. Twenty slices (plus one reference segment) were stained with he-
750
•
Gadolinium, CO2, and Iodinated Contrast for Renal Artery Stenosis
matoxylin-eosin-safranin for histologic study, and 20 slices (plus one reference segment) were stained with orcein for histomorphometric analysis. Each specimen was evaluated for the presence of neointima formation, medial dissection, alteration of the internal elastic lamina (EL) and external elastic lamina, and morphologic appearance of the adventitial, medial, and neointimal layers. Sections were also evaluated for the presence of intraluminal thrombus, intraluminal hemorrhage, and inflammatory cells.
was calculated as an injury severity index (17). The extent of stenosis was determined as percent stenosis ⫼ reference segment ⫽ 100% ⫻ (1 ⫺ stenosis lumen area ⫼ reference segment lumen area). The medial area-to-luminal area ratio (hypertrophy index) was calculated in reference segments and injured renal artery segments to assess the degree of neointimal formation after 8 weeks.
Histomorphometry
DSA evaluation of the postprocedural response with use of various contrast agents was compared by a Wilcoxon signed rank test. Data were expressed as means ⫾ SD. Linear regression was used to (i) compare Gd-DOTA and CO2 with iodixanol angiographic data and (ii) compare the morphometric restenosis data directly with angiographic stenoses determined for the three agents. A value of probability P ⬍ .05 was required to reject the null hypothesis at the 95% confidence level. Statistical analysis was performed with StatView 4.5 version (Abacus Concepts, Berkeley, CA).
A stenotic lesion identified on microscopy was easily visualized on angiography. Prestenotic (ie, reference segment) and stenotic segments of the injured renal arteries were examined and compared to prestenotic and stenotic segments of the treated injured renal arteries. Neointima was defined as any tissue layer circumscribed by the internal EL, excluding intraluminal fibrin deposits and larger organized mural thrombus material. Other specific changes were defined as reshaping of the medial layer, tearing of one of the elastic laminae, and/or presence of adventitial fibrosis. All transverse sections were digitized via a Cohu-CCD video camera, then morphometrically analyzed with use of MicroVision software (Evry, France). In each artery, measurements were performed in the proximal healthy reference segment and in 20 pathologic slices. The areas of intima, media, and residual lumen as well as circumferences demarcated by internal and external EL were determined. The length of the internal EL fracture through the neointima, extending from one end of the dissected internal EL to the other, was used as a marker of the severity of the injury. Area measurements were performed. The vessel perimeter was delineated by the external EL, thereby defining the vessel area. The lumen perimeter defined the lumen area, and the internal EL perimeter defined an area including neointima and lumen (internal EL area). Media area was obtained as the difference between vessel area and internal EL area, and neointima area was obtained as the difference between internal EL area and lumen area. The neointima-to-internal fracture length ratio
Data Processing
RESULTS Stenosis Quantification No significant difference was observed between gadoterate- and iodixanol-enhanced DSA for automated quantification of stenoses (Fig 1a). Unfortunately, automated quantification was not highly correlated (r2 ⬍ 0.75) when the arteries were opacified by CO2 (Fig 1b). Based on morphometric data, the estimation of the stenoses was highly correlated (0.94 ⬍ r2 ⱕ 0.99) with iodixanol (Fig 2a) and GdDOTA (Fig 2b), whereas CO2 (Fig 2c) proved to be less reliable (r2 ⬍ 0.73). Comparison of Contrast Media in DSA Gd-DOTA allowed visualization of the main renal arteries and collateral branches (Fig 3a). Gd-DOTA visualized arteries of the superior renal pole in all rabbits just as well as the iodinated gold standard, iodixanol. Intrarenal vasculature was visible as far as third-generation branches, whereas the fourth generation was more easily
June 2001
JVIR
identified by iodixanol opacification (Fig 3b) on all examinations. Stenosis and poststenotic dilation were evidenced by CO2-enhanced angiography, whereas only second-generation arterioles were identified inside the renal parenchyma (Fig 3c). Density Quantification Under similar injection conditions, iodixanol was found to be 3.5 times as dense as the Gd complex (Table). However, this density gradient between the two contrast agents did not significantly interfere with the measurement of stenoses, and the regression coefficient (Fig 1a) was high (r ⬎ 0.97).
DISCUSSION To date, renal opacification by gadoterate, iodixanol, and CO2 has not been simultaneously quantitatively evaluated in vivo in a model of restenosis. Kidneysafe alternatives to iodine-enhanced DSA do exist, represented by CO2-enhanced DSA (10,11,18) or MR angiography (19). Unfortunately, CO2-enhanced DSA provides few advantages in terms of contrast and measurement reliability (20). The use of CO2 was not so appropriate in this animal study, particularly in the challenge of evaluating stenoses in small renal arteries. Our data may be related to the size of the study subjects and may not be completely exportable to humans. However, it appeared interesting to compare contrast agents on the basis of experimental conditions of challenge in terms of vessel size. In particular, the poor contrast of CO2 prevented accurate measurement of stenoses in the present study. Its widespread use is associated with restricted indications, such as in vascular interventional procedures in patients with renal impairment or who have contraindications to magnetic resonance (MR) imaging. Although the large majority of daily use of CO2 appears to be very safe, recent studies evaluated the risk of air leak leading to intraarterial gas embolism (5). This risk should lead to a recommendation of its careful use in organs with poor anastomotic arterial blood supply (eg, kidney, brain) (21), as it has been responsible, albeit rarely, for severe complications such as mesenteric ischemia (5,22,23). MR angiography has therefore become the preferred noninvasive modality to evaluate all arterial
Volume 12
Number 6
Le Blanche et al
•
751
Figure 3. Comparison of selective angiograms of a 72% renal artery stenosis (on the basis of histomorphometric data). (a) The Gd complex allowed opacification of the main and polar vessels. Intrarenal arterial vascularization is visible as far as the third-generation branches (thin black arrows). Evidence of poststenotic dilation is clearly demonstrated (arrowheads). Automated quantification allows reliable measurement of the stenosis (white line), compared to the reference segment (black line). (b) Iodine-enhanced angiography provides three-fold higher density and visualizes fourth generation branches (arrowheads). Automated quantification was nevertheless identical to the previous measurement. (c) With CO2 angiography, stenosis (black arrowhead), poststenotic dilation (white arrowhead), and polar artery (white arrow) remain visible, whereas only 2nd generation arterioles are identified inside the renal parenchyma (black arrow).
DSA Comparative Contrast Density Data of the Renal Artery Reference Lumen (N ⴝ 27) Contrast Agent Gd-DOTA Iodixanol Iodixanol/ Gd-DOTA*
Density 364 ⫾ 129 1167 ⫾ 355 3.49 ⫾ 1.25
Note.—Values expressed as means ⫾ SD. * Wilcoxon signed rank test, P ⬍ .0001.
territories. Nevertheless, MR evaluation of small artery stenoses remains difficult because this type of imaging is dependent on flow velocity. It remains less effective than DSA in the measurement of stenoses, particularly in case of distal arteries, and the diameter of a stenosis may be either over- or underestimated (19). Conversely, automatic stenosis quantification performed with DSA data has become very reliable, based on geometric and densitometric data. In the case of renovascular hypertension or ischemic nephropathy, when there is a clinical indication for renal revascularization, therapeutic decisions regarding PTRA and/or stent placement in renal arteries are based on stenosis grade, even though transstenotic pressure measurements, the most accurate method for determining the significance of renal artery stenoses, can be safely performed in almost all circumstances. In light of all
these arguments, it seems attractive to use the currently available techniques to develop DSA enhanced by a noniodinated, nongaseous radiographic dense contrast agent, and Gd complexes appear to be an interesting alternative (10,11,18,24,25). However, the efficacy of Gd-enhanced DSA depends predominantly on digital subtraction and would be dramatically impaired in conventional angiography. This study demonstrates that Gd-based contrast agents may reliably replace iodinated contrast agents for stenosis quantification in selective DSA. When the use of iodinated contrast agents is hazardous for renal function, Gd-enhanced DSA may represent an attractive alternative for the measurement of stenotic lesions before and after PTA, in contrast with CO2, which fails to provide accurate measurement of stenosis. Acknowledgments: The authors thank Jean-Pierre Djemat for animal care, Professor Raymond Ardaillou from the National Academy of Medicine for experimental support, Suzette Pereira-Freire and Sandrine Boucheteil for literature research assistance, and Anthony Saul, MD, for manuscript review. References 1. van de Ven PJ, Kaatee R, Beutler JJ, et al. Arterial stenting and balloon angioplasty in ostial atherosclerotic renovascular disease: a randomised trial. Lancet 1999; 353:282–286. 2. Solomon R. Contrast-medium-induced renal failure. Kidney Int 1998; 53:230–242.
3. Lasser EC, Lyon SG, Berry CC. Reports on contrast media reactions: analysis of data from reports of the US food and drug administration. Radiology 1997; 203:605– 610. 4. O’Neil EA, Hansen KJ, Canzanello VJ, Pennell TC, Dean RH. Prevalence of ischemic nephropathy in patients with renal insufficiency. Am Surg 1992; 58: 485– 490. 5. Yang X, Manninen H, Soimakallio S. Carbon dioxide in vascular imaging and intervention. Acta Radiol 1995; 36: 330 –337. 6. Hawkins IF, Caridi JG, Kerns SR. Plastic bag delivery system for hand injection of carbon dioxide. AJR Am J Roentgenol 1995; 165:1487–1489. 7. Kinno Y, Odagiri K, Andoh K, Itoh Y, Tarao K. Gadopentetate dimeglumine as an alternative contrast material for use in angiography. AJR Am J Roentgenol 1993; 160:1293–1294. 8. Fobbe F, Wacker F, Wagner S. Arterial angiography in high-kilovoltage technique with gadolinium as the contrast agent: first clinical experience. Eur Radiol 1996; 6:224 –229. 9. Gemery J, Idelson B, Reid S, et al. Acute renal failure after arteriography with a gadolinium-based contrast agent. AJR Am J Roentgenol 1998; 171:1277–1278. 10. Spinosa DJ, Matsumoto AH, Angle JF, et al. Gadolinium-based contrast and carbon dioxide angiography to evaluate renal transplants for vascular causes of renal insufficiency and accelerated hypertension. J Vasc Interv Radiol 1998; 9:909 –916. 11. Spinosa DJ, Matsumoto AH, Angle JF, Hagspiel KD, McGraw JK, Ayers C. Renal insufficiency: usefulness of ga-
752
12.
13.
14.
15.
16.
•
Gadolinium, CO2, and Iodinated Contrast for Renal Artery Stenosis
dodiamide-enhanced renal angiography to supplement CO2-enhanced renal angiography for diagnosis and percutaneous treatment. Radiology 1999; 210:663– 672. Katzberg RW, Sahler LG, Duda SW, et al. Renal handling and physiologic effects of the paramagnetic contrast medium GdDOTA. Invest Radiol 1990; 25:714–719. Schuhmann-Giampieri G, Krestin GP. Pharmacokinetics of Gd-DTPA in patients with chronic renal failure. Invest Radiol 1991; 26:975–979. Prince MR, Arnoldus C, Frisoli JK. Nephrotoxicity of high-dose gadolinium compared with iodinated contrast. J Magn Reson Imaging 1996; 6:162–166. Le Blanche AF, Sibony M, Kollar A, Callard P, Bigot JM, Boudghene F. A new model of endovascularly-induced renal artery stenosis in hypercholesterolemic versus normocholesterolemic rabbits. Invest Radiol 1998; 33:322–328. EEC Directive No.86/609. Official J Eur Communities 1986; 358:1–28.
17. Waksman R, Robinson KA, Crocker IR, et al. Intracoronary radiation decreases the second phase of intimal hyperplasia in a repeat balloon angioplasty model of restenosis. Int J Radiat Oncol Biol Phys 1997; 39:475– 480. 18. Spinosa DJ, Angle JF, Hagspiel KD, et al. Feasibility of gadodiamide compared with dilute iodinated contrast material for imaging of the abdominal aorta and renal arteries. J Vasc Interv Radiol 2000; 11:733–737. 19. Prince MR, Schoenberg SO, Ward JS, Londy FJ, Wakefield TW, Stanley JC. Hemodynamically significant atherosclerotic renal artery sclerosis: MR angiographic features. Radiology 1997; 205:128 –136. 20. Moresco KP, Patel N, Johnson MS, Trobridge D, Bergan KA, Lalka SG. Accuracy of CO2 angiography in vessel diameter assessment: a comparative study of CO2 versus iodine contrast material in an aortoiliac flow model. J Vasc Interv Radiol 2000; 11:437– 444.
June 2001
JVIR
21. Hawkins IF, Mladinich CRJ, Storm B, et al. Short-term effects of selective renal arterial carbon dioxide administration on the dog kidney. J Vasc Interv Radiol 1994; 5:149 –154. 22. Caridi JG, Hawkins IF. CO2 digital substraction angiography: potential complications and their prevention. J Vasc Interv Radiol 1997; 8:383–391. 23. Spinosa DJ, Matsumoto AH, Angle JF, Hagspiel KD, Hooper TN. Transient mesenteric ischemia: a complication of carbon dioxide angiography. J Vasc Interv Radiol 1998; 9:561–564. 24. Matchett WJ, McFarland DR, Russell DK, Sailors DM, Moursi MM. Azotemia: gadopentetate dimeglumine as contrast agent at digital substraction angiography. Radiology 1996; 201:569–571. 25. Spinosa DJ, Angle JF, Hagspiel KD, Schenk WG III, Matsumoto AH. CO2 and gadopentetate dimeglumine as alternative contrast agents for malfunctioning dialysis grafts and fistulas. Kidney Int 1998; 54:945–950.