Radiology
Thoracic Imaging Lawrence R. Goodman, MD Meltem Gulsun, MD2 Paul Nagy, PhD Lacey Washington, MD Published online before print 10.1148/radiol.2343031871 Radiology 2005; 234:923–928 1
From the Department of Radiology, Medical College of Wisconsin, 9200 W Wisconsin Ave, Milwaukee, WI 532263596. Received November 21, 2003; revision requested February 6, 2004; revision received May 6; accepted June 2. Supported by a grant from the Amersham Corporation, Princeton, NJ. Address correspondence to L.R.G. (e-mail:
[email protected]). Current address: Department of Radiology, Hacettepe University, Faculty of Medicine, Ankara, Turkey.
2
CT of Deep Venous Thrombosis and Pulmonary Embolus: Does Iso-osmolar Contrast Agent Improve Vascular Opacification?1 PURPOSE: To prospectively compare the vascular attenuation achieved with the iso-osmolar dimeric contrast agent iodixanol with that achieved with the nonionic monomeric contrast agent iohexol for computed tomographic (CT) venography after CT pulmonary angiography. MATERIALS AND METHODS: Institutional review board approval and informed consent were obtained, and 51 consecutive patients undergoing CT pulmonary angiography and CT venography were recruited. A 130-mL dose of iodixanol 320 was injected intravenously at a rate of 4 mL/sec and followed by injection of 50 mL of saline. CT venography was performed after 3.5 minutes. From prior studies, 51 patients of similar weight were picked as control subjects. They received a similar iodine load with iohexol 300 and were studied with a similar technique. Section thickness was 1.25 mm for pulmonary emboli and 5 mm for deep venous thrombosis. Test and control group characteristics (ie, sex, age, and weight) were not significantly different (P ⬎ .05). Additionally, in test patients who had undergone CT pulmonary angiography and CT venography during the two preceding years, current and previously obtained CT scans were compared (ie, paired studies). Regions of interest were measured in four pulmonary artery and four lower extremity vein locations by two independent observers. RESULTS: Iodixanol increased average attenuation by 7 HU (P ⬍ .05) in the lower extremities and decreased average attenuation by 42 HU (P ⬍ .05) in the pulmonary arteries. In the 11 paired studies, similar results were obtained. CONCLUSION: Iodixanol caused a modest but statistically significant improvement in venous attenuation and a decrease in arterial attenuation. The diagnostic importance of this small increase in venous attenuation is not clear. ©
Author contributions: Guarantors of integrity of entire study, L.R.G., M.G.; study concepts, L.R.G.; study design, L.R.G., P.N.; literature research, M.G.; clinical studies, M.G., L.R.G., L.W.; data acquisition, M.G., L.R.G., L.W.; data analysis/interpretation, P.N., M.G.; statistical analysis, P.N.; manuscript preparation, L.R.G.; manuscript definition of intellectual content, L.R.G., P.N.; manuscript editing, revision/review, and final version approval, all authors ©
RSNA, 2005
RSNA, 2005
The evaluation of patients suspected of having venous thromboembolic disease requires the evaluation of the pulmonary arteries for pulmonary emboli and the pelvic and leg veins for deep venous thrombosis. At many institutions, helical computed tomography (CT) has become the primary imaging test after chest radiography in the diagnosis of pulmonary emboli (1,2,3). As an adjunct, many centers are now scanning the deep veins of the pelvis and legs to look for the source of clots (4 –9). The pulmonary arteries are studied during the first-pass phase (ie, the first pass of intravenous contrast agent), whereas the veins are studied in the equilibrium phase (ie, 3–31⁄2 minutes after injection of contrast agent). Average pulmonary artery attenuation is in the range of 200 –350 HU, whereas the average attenuation of the iliac, femoral, and popliteal veins is in the range of 85–110 HU (4,5,10). There is a subset of patients in whom pulmonary artery and deep venous attenuation is suboptimal. In a prospective study of 541 patients examined with CT, Cham et al (5) rated 95% of pulmonary emboli scans and 77% of deep venous thrombosis scans 923
Radiology
as “good to excellent.” There is no universal agreement in the literature as to what level of venous enhancement is adequate for diagnosis or exclusion of deep venous thrombosis. The decrease in contrast attenuation during the equilibrium phase is caused by a combination of dilution of the contrast agent by the entire blood volume, renal excretion of the contrast agent, osmotic hemodilution, and extravascular diffusion of the contrast agent. Iodixanol (Visipaque 320; Amersham Health, Princeton, NJ) is a nonionic dimer radiographic contrast agent that is iso-osmolar to plasma at all concentrations. Because the osmolarity of iodixanol is approximately one-half that of nonionic monomers, osmotic diuresis is reduced, and hemodilution should be reduced. This should lead to a higher retained vascular attenuation beyond the first pass. CT venous studies, which frequently are affected negatively by dilute or unpredictable contrast enhancement, should benefit most. Thus, the purpose of our study was to prospectively compare the vascular attenuation achieved with the iso-osmolar dimeric contrast agent iodixanol with that achieved with the nonionic monomeric contrast agent iohexol (Omnipaque 300; Amersham Health) for CT venography, which was performed after CT pulmonary angiography.
MATERIALS AND METHODS Amersham Health provided iodixanol for use in this study; however, the execution and reporting of the study, as well as the data, were controlled by the authors.
Subjects and Imaging This study was approved by the institutional review board. The procedure and possible advantages and disadvantages were discussed with patients prior to the study, and informed consent was obtained. Test patients were recruited from a series of consecutive adult patients undergoing CT for evaluation of pulmonary emboli, deep venous thrombosis, or both at a large university hospital during working hours over a 4-month period. Allergic patients, pregnant patients, and patients with renal impairment (serum creatinine level, ⬎1.5 mg/dL [⬎132.6 mol/L]) were excluded. Intubated patients and patients incapable of giving consent were also excluded. CT scanning was performed with 130 mL of iodixanol 924
䡠
Radiology
䡠
March 2005
(iodine, 320 mg/mL) injected at a rate of 4 mL/sec. The contrast agent was warmed to body temperature, and a 50-mL saline chaser (11) was administered after the injection, as per manufacturer instructions, to minimize the effects of the increased iodixanol viscosity. All CT scanning was performed with eight- and 16-section multi– detector row CT scanners (Light Speed; GE Medical Systems, Milwaukee, Wis). Pulmonary arteries were scanned after a 20 –28-second delay, and the pelvis was scanned after a 31⁄2-minute delay; 1.25-mm helical images were obtained continuously through the thorax, and 5-mm discontinuous transverse images were obtained every 2 cm, from the iliac crest to the knees. For the thorax, a rotation time of 0.5– 0.8 sec, a tube voltage of 140 kVp, and a tube current of 380 mA were used. For the deep pelvic and thigh veins, a rotation time of 3.8 sec, a tube voltage of 120 kVp, and a tube current of 280 mA were used. For patient care purposes, scans were interpreted in the usual manner by residents or fellows, and the examinations were staffed by faculty. In one patient, a small amount (approximately 20 mL) of iodixanol extravasated into the subcutaneous tissues, without apparent ill effects. In the 6 months prior to the start of this study, we were able to select 51 control patients from the 352 patients evaluated with CT for pulmonary emboli and deep venous thrombosis. These patients had been studied by using 140 mL of iohexol (iodine, 300 mg/mL), which provided an equivalent iodine load (42 g of iodine). The contrast agent was warmed to body temperature, but no saline chaser was administered, as per usual department protocol; otherwise, the technique was identical to the technique used for iodixanol injection. From past clinical observations, venous attenuation often appeared suboptimal in obese patients. Thus, one control patient was picked per test patient, and the weight of the control patient was within 10% of the weight of the test patient. Test patient weights were recorded at the time of scanning; control patient weights were determined from clinical charts. Other possible patient parameters, such as cardiac disease or renal status, were not examined. In a separate analysis in test patients who had undergone studies for venous thromboembolic disease with an identical technique in the 2 preceding years, test scans obtained with iodixanol were compared with the old scans obtained with iohexol. Of the 52 consecutive eligible patients who were approached, 51 consented to
TABLE 1 Patient Characteristics and Type of Contrast Agent Characteristic
Iodixanol
Iohexol
Sex* Male Female Age (y) Weight (kg)
23 28 56 ⫾ 17 84.0 ⫾ 23
27 24 61 ⫾ 16 85.4 ⫾ 25
Note.—Unless otherwise indicated, data are mean ⫾ standard deviation. * Data are number of patients.
participate in the study. Twenty-three patients in the test group and 27 in the control group were men; the average patient age was 56 (age range, 20 – 84 years) and 61 years (age range, 25–91 years), respectively. The average test patient weighed 3.2 lbs (1.5 kg) (1.7%) less than the average control patient (184.7 vs 187.9 lbs [84.0 vs 85.4 kg]) (Table 1). There was no significant difference (P ⬎ .05) between the groups. In the group of patients who underwent repeat scanning, there were four men and seven women, with an average age of 53 years (age range, 28 – 82 years).
Image Analysis Two independent observers (M.G. and a research technologist, with 5 and 9 years of experience, respectively) performed all measurements, were blinded to all patient identification, and did not know which subjects were in the control group. In the 11 patients who underwent prior studies, the results of test and control studies were evaluated 2 or more weeks apart. The observers measured regions of interest at four representative points in the pulmonary arteries (ie, main, interlobar, anterior segment of left upper lobe, and posterior basal segment of the right lower lobe) and at four representative points in the deep veins (ie, right common iliac, left common femoral, right femoral, and left popliteal) of the lower extremities. The regions of interest were obtained at a workstation by using an oval approximately two-thirds the area of the vessel in cross-section (Fig 1). If a vessel of interest was surgically missing, occluded with clot, or obscured by artifact, an equivalent branch was measured elsewhere. This was performed seven times with iohexol (two left upper lobe lobectomies, three pulmonary emboli, one deep venous thrombosis, and one metallic artifact) and nine times with iodixanol (three pulmonary emboli, two Goodman et al
Radiology Figure 1. Comparison of attenuation of the interlobar pulmonary artery and the left common femoral vein on transverse CT scans obtained in (a) the index case (iodixanol), (b) the control case (iohexol), and (c) the index patient 18 months before a (iohexol). (a) Interlobar pulmonary artery attenuation was 230 HU. Left common femoral vein attenuation was 88 HU. (b) Interlobar pulmonary artery attenuation was 282 HU. Left common femoral vein attenuation was 80 HU. (c) Interlobar pulmonary artery attenuation was 173 HU. Left common femoral vein attenuation was 83 HU.
deep venous thrombosis, and four metallic artifacts). No studies were excluded for technical reasons.
Statistical Analysis Mean attenuation ⫾ standard deviation at each station were calculated for iodixanol, which served as the contrast agent in test patients, and iohexol, which served as the contrast agent in control patients. Differences in attenuation between test patients and control patients were assessed with a two-sample t test for independent samples with equal variances. Differences in attenuation between iodixanol and iohexol in the same patient were assessed with a one-sample paired t test. In both, the null hypothesis assumed that attenuation with iodixanol was equal to attenuation with iohexol. A P value of less than .05 was considered to indicate a statistically significant difference. Arterial and venous attenuation was also correlated with patient weight (Excel 2000; Microsoft, Redmond, Wash).
RESULTS Measurements were obtained in 51 test patients and 51 control patients at four Volume 234
䡠
Number 3
venous and four arterial sites. Attenuation measurements, which were obtained by two observers at each vascular station, were similar. The coefficient of variation for all venous measurements was 4.7% for iodixanol and 6.6% for iohexol. For the arterial measurements, the coefficient of variation was 4.2% for iodixanol and 5.2% for iohexol. Thus, combined readings at each station were used for all computations. The average venous attenuation was 100 HU for iodixanol and 93 HU for iohexol. As shown in Table 2, at three of the four venous stations, iodixanol provided significantly higher attenuation than did iohexol by an average of 7 HU (7%) (mean difference range, 6 – 8 HU). The average arterial attenuation was 259 HU for iodixanol and 305 HU for iohexol. At all four arterial stations, iohexol provided significantly higher attenuation than iodixanol by an average of 42 HU (16%) (mean difference range, 35– 49 HU). For both venous and arterial enhancement, however, iodixanol resulted in less interobserver variability. The standard deviations for all venous and arterial measurements were less with iodixanol in veins (iodixanol, 18.1; iohexol, 25.2;
P ⬍ .05) and arteries (iodixanol, 75.4; iohexol, 113.5; P ⬍ .05). We also tested the hypothesis that weight correlates inversely with vascular opacification. As shown in Figures 2 and 3, correlation between weight and vascular attenuation was fairly low (r2 ⫽ 0.08 – 0.25). Eleven test patients were found to have undergone prior CT scanning for thromboembolic disease during the previous 2 years. In the legs of patients scanned with iodixanol, attenuation was an average of 9.9 HU higher than in patients scanned with iohexol. In the pulmonary arteries of patients scanned with iohexol, attenuation was an average of 25 HU higher than in patients scanned with iodixanol. In this small group, none of the differences were statistically significant (Table 3).
DISCUSSION Theoretically, an iso-osmolar contrast agent such as iodixanol should improve venous enhancement and provide more reliable venous opacification in the equilibrium phase of the study. This study was performed to determine whether iodixanol would improve venous enhance-
Deep Venous Thrombosis and Pulmonary Embolus
䡠
925
Radiology Figure 2.
Graphs show venous attenuation correlation with weight for examinations with (a) iodixanol and (b) iohexol.
Figure 3.
Graphs show arterial attenuation correlation with weight for examinations with (a) iodixanol and (b) iohexol.
ment during CT venography. The average increase in venous attenuation was 7 HU (7.4%) (P ⬍ .05). Statistically significant increased attenuation was demonstrated at three of four measured venous stations. The variability in venous enhancement was also less with iodixanol. In the small group of patients who had undergone previous examinations, the average increase in density was 9.9 HU (11%) (P ⬎ .05). Overall, there does appear to be a modest but statistically significant increase in enhancement and more consistent venous enhancement with iodixanol. Is the 7%–11% increase in attenuation clinically significant? There is, in fact, little discussion in the literature as to what constitutes adequate venous opacification to diagnose or rule out deep venous thrombosis on the basis of CT findings. Venous clots have been reported to have an average attenuation of 31 HU ⫾ 10 by Loud et al (4) and an average attenuation of 51 HU (95% con926
䡠
Radiology
䡠
March 2005
fidence interval: 48, 57; median, 55 HU) by Cham et al (5). Brink et al (12) report that experimental thrombi in pigs averages 50 HU. If a clot has an attenuation of 31 HU, venous attenuation in the range of 60 –70 HU is probably adequate for detection. If, as Cham et al (5) report, the median attenuation of a clot is 55 HU, half the clots will be have an attenuation that is greater than 55 HU. One might be able to diagnose a clot when the venous attenuation is in the 60 –70 HU range, but one could not exclude a clot with confidence. It is, therefore, reasonable but unproven that an attenuation of 80 HU or more would provide adequate contrast differentiation between the clot and opacified vessel. Brink et al (12) evaluated this problem for clots in the pulmonary arteries. They found that the task of discerning a clot totally obstructing a vessel and one partially obstructing a vessel are different and that detection is partially dependent on window and level settings.
Thus, absolute numeric requirements could not be established. This requires further study for deep venous thrombosis detection. There are several limitations to our study. In studying the pulmonary arteries, iohexol was injected in the conventional manner, whereas iodixanol, because of its higher viscosity, was injected with a 50-mL bolus chaser to clear the arm vein of contrast agent. Thus, two slightly different injection techniques in the pulmonary arteries are being compared. It is possible that slightly more contrast agent entered the pulmonary arteries in the iodixanol studies because of the chaser. Nonetheless, iohexol provided significantly better attenuation. Conversely, it is possible that the saline actually diluted the iodixanol and caused diminished pulmonary arterial attenuation. Katz et al (13) studied 19 patients with 125 mL of iohexol 300 and iohexol 320 and 19 patients with 125 mL of ioGoodman et al
TABLE 2 Arterial and Venous Attenuation in 51 Test and 51 Control Patients
Radiology
Pulmonary Arteries
Deep Veins
Attenuation and Statistical Value
Main
Anterior Segment of Left Upper Lobe
Interlobar
Posterior Basal Segment of Right Lower Lobe
Right Common Iliac
Left Common Femoral
Right Femoral
Left Popliteal
Attenuation (HU)* Iodixanol Iohexol Difference t Value P value
268 ⫾ 69 318 ⫾ 114 50 3.716 ⬍.001
256 ⫾ 73 305 ⫾ 119 49 3.546 ⬍.001
259 ⫾ 76 307 ⫾ 117 48 3.468 ⬍.001
254 ⫾ 82 289 ⫾ 103 35 2.677 .004
105 ⫾ 17 98 ⫾ 27 ⫺7 2.32 .01
103 ⫾ 18 96 ⫾ 30 ⫺7 1.557 .06
94 ⫾ 18 88 ⫾ 20 ⫺6 2.07 .02
101 ⫾ 18 93 ⫾ 21 ⫺8 2.777 .003
* Unless otherwise indicated, data are mean ⫾ standard deviation.
TABLE 3 Arterial and Venous Attenuation in 11 Repeat Observations Pulmonary Arteries Attenuation and Statistical Value
Main
Average attenuation (HU) Iodixanol 271 Iohexol 280 Difference* 10.2 ⫾ 101.6 t Value 0.333 t 10, .975† 2.201
Deep Veins
Anterior Segment of Left Upper Lobe
Interlobar
Posterior Basal Segment of Right Lower Lobe
265 258 9.2 ⫾ 104.7 0.249 2.306
251 287 40.7 ⫾ 93.3 1.379 2.228
255 274 33.0 ⫾ 105.8 0.986 2.228
Right Common Iliac
Left Common Femoral
Right Femoral
Left Popliteal
106 102 95 102 98 93 88 90 ⫺8.1 ⫾ 21.9 ⫺9.4 ⫾ 24.7 ⫺9.4 ⫾ 24.7 ⫺12.7 ⫾ 24.4 ⫺1.230 ⫺1.449 ⫺1.264 1.727 2.201 2.201 2.201 2.201
* Data are mean ⫾ standard deviation. † t Test value for 10 degrees of freedom, ␣ ⫽ .975.
dixanol 270 without a saline chaser and found no difference in pulmonary artery attenuation between the two contrast agents. Lee et al (14) found no diagnostic advantage with iodixanol 270 for abdominal CT, and Rienmuller et al (15) found no advantage for coronary angiography. During the venous phase of our study, it is possible but unlikely that the 50-mL saline chaser diluted in the patient’s 5-L blood volume and caused a noticeable decrease in attenuation of the veins. Another limitation was the choice of patient controls. Other factors, such as renal function and cardiac function, may have influenced the results in an unpredictable fashion. Since age, sex, and weight were similar in both groups, it is likely that these physiologic factors were self-canceling. On the basis of previous clinical perceptions, we used weight as a clinical control for picking patients. Our results show that there was in fact only a modest correlation between increasing weight and decreasing vascular attenuation. Finally, the sample size, especially for repeat studies, was very small. If more patients had undergone prior scanning, Volume 234
䡠
Number 3
it is possible that the venous attenuation differences might have reached statistical significance, with an average increased attenuation of approximately 9 HU (11%). This increase is slightly greater than the 7-HU increase (7%) in the main study, which was statistically significant in 51 patients. Iodixanol is less nephrotoxic than iohexol (16 –21). A recent study by Aspelin et al (16) confirmed the findings of earlier studies that iodixanol causes less nephrotoxicity than does iohexol in high-risk patients. In their study of patients with diabetic nephropathy, the creatinine concentration increased by 0.5 mg/dL (44.2 mol/L) or more 1–3 days after injection in two (3%) of 64 patients who received iodixanol and in 17 (26%) of 65 patients who received iohexol (16). Creatinine concentration increases of 1 mg/dL (88.4 mol/L) were found in 0% and 15% of patients, respectively. Mean creatinine elevation was 0.07 mg/dL (6.18 mol/L) versus 0.24 mg/dL (21.21 mol/L). Iodixanol has also been reported to cause less patient discomfort during systemic arterial studies (19). In conclusion, iodixanol increased ve-
nous attenuation by approximately 7 HU (7%), while it diminished pulmonary artery opacification by 42 HU (16%). Iodixanol is considerably more expensive than iohexol, and it is not clear, at this point, if the modest increase in venous attenuation justifies its routine use. The drop in pulmonary artery attenuation when iodixanol is used with a saline flush is detrimental, but pulmonary artery attenuation was generally well within the diagnostic range. Because of the cost differential, iodixanol should be considered for patients with marginal renal function who require CT for venous thromboembolic disease. Aspelin et al (16) found that patients with impaired renal function (creatinine level, 1.5–3.5 mg/dL [132.6 –309.4 mol/L]) who were well hydrated before and after contrast agent administration appeared to have significantly less renal test function impairment with iodixanol than with iohexol. Acknowledgments: Thanks to Sylvia Bartz for her help in preparing the manuscript and Maureen Levenhagen for her technical support.
Deep Venous Thrombosis and Pulmonary Embolus
䡠
927
Radiology
References 1. Remy-Jardin M, Remy J, Wattinne L, Giraud F. Central pulmonary thromboembolism: diagnosis with spiral volumetric CT with the single-breath-hold technique— comparison with pulmonary angiography. Radiology 1992; 185:381–387. 2. Goodman LR, Lipchik RJ. Diagnosis of acute pulmonary embolism: time for a new approach (editorial). Radiology 1996; 199:25–27. 3. Stein PD, Kayali F, Olson, RE. Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism. Am J Cardiol 2004; 93:1316 – 1317. 4. Loud PA, Katz DS, Bruce DA, Klippenstein DL, Grossman CD. Deep venous thrombosis with suspected pulmonary embolism: detection with combined CT venography and pulmonary angiography. Radiology 2001; 219:498 –502. 5. Cham MD, Yankelevitz DF, Shaham D, et al. Deep venous thrombosis: detection by using indirect CT venography. Radiology 2000; 216:744 –751. 6. Coche EE, Hamoir XL, Hammer FD, Hainaut P, Goffette PP. Using dual-detector helical CT angiography to detect deep venous thrombosis in patients with suspicion of pulmonary embolism: diagnostic value and additional findings. AJR Am J Roentgenol 2001; 176:1035–1039. 7. Garg K, Kemp JL, Wojcik D, et al. Thromboembolic disease: comparison of combined CT pulmonary angiography and venography with bilateral leg sonography
928
䡠
Radiology
䡠
March 2005
8.
9.
10.
11.
12.
13.
14.
in 70 patients. AJR Am J Roentgenol 2000; 175:997–1001. Ciccotosto C, Goodman LR, Washington L, Quiroz FA. Indirect CT venography following CT pulmonary angiography: spectrum of CT findings. J Thorac Imaging 2002; 17:18 –27. Yankelevitz DF, Gamsu G, Shah A, et al. Optimization of combined CT pulmonary angiography with lower extremity CT venography. AJR Am J Roentgenol 2000; 174:67– 69. Bruce D, Loud PA, Klippenstein D, Grossman ZD, Katz DS. Combined CT venography and pulmonary angiography: how much venous enhancement is routinely obtained? AJR Am J Roentgenol 2001; 176:1281–1285. Hopper KD, Mosher TJ, Kasales CJ, TenHave TR, Tully DA, Weaver JS. Thoracic spiral CT: delivery of contrast material pushed with injectable saline solution in a power injector. Radiology 1997; 205: 269 –271. Brink JA, Woodard PK, Horesh L, et al. Depiction of pulmonary emboli with spiral CT: optimization of display window settings in a porcine model. Radiology 1997; 204:703–708. Katz DS, Young W, Mueller R, Pollack S. Iohexol vs. iodixanol in multislice CT pulmonary angiography. Presented at the 103rd Meeting of the American Roentgen Ray Society, San Diego, Calif, May 4 –9, 2003. Lee FT Jr, Caroline DF, Thornbury JR, et al. A randomized comparison of iodixa-
15.
16.
17. 18.
19.
20.
21.
nol and iohexol in adult body computed tomography scanning. Acad Radiol 1996; 3(suppl 3):S500 –S506. Rienmuller R, Brekke O, Kampenes VB, Reiter U. Dimeric versus monomeric nonionic contrast agents in visualization of coronary arteries. Eur J Radiol 2001; 38: 173–178. Aspelin P, Aubry P, Fransson SG, Strasser R, Willenbrock R, Berg KJ. Nephrotoxic effects in high-risk patients undergoing angiography. N Engl J Med 2003; 348: 491– 499. Chalmers N, Jackson RW. Comparison of iodixanol and iohexol in renal impairment. Br J Radiol 1999; 72:701–703. Grynne BH, Nossen JØ, Bolstad B, Borch KW. Main results of the first comparative clinical studies on Visipaque. Acta Radiol Suppl 1995; 399:265–270. Pugh ND, Sissons GR, Ruttley MS, Berg KJ, Nossen JØ, Eide H. Iodixanol in femoral arteriography (phase III): a comparative double-blind parallel trial between iodixanol and iopromide. Clin Radiol 1993; 47:96 –99. ˘ , Berg KJ, Kjœrsgaard P, et al. Jakobsen JA Angiography with nonionic x-ray contrast media in severe chronic renal failure: renal function and contrast retention. Nephron 1996; 73:549 –556. Katzberg RW. Urography into the 21st century: new contrast media, renal handling, imaging characteristics, and nephrotoxicity. Radiology 1997; 204:297–312.
Goodman et al
