- Email: [email protected]
Effect of contrast concentration on abdominal enhancement in the rabbit: Spiral computed tomography evaluation
Effect of Contrast Concentration on Abdominal E n h a n c e m e n t in the Rabbit: Spiral Computed Tomography Evaluation David A. Bluemke, MD, PhD, Elliot K. Fishman, MD, J a m e s H. Anderson, PhD
Rationale and Objectives. We examined the effect of varying the concentration of a nonionic iodinated contrast agent on hepatic and abdominal vascular enhancement using spiral computed tomography (CT) scanning. M e t h o d s . Spiral CT scans of 10 New Zealand male rabbits were obtained after intravenous injection of 240 mg I/ml iohexol and 350 mg I/ml iohexol injected at a rate of 2 ml/sec. Each animal was studied at both concentrations using four different contrast volumes: 2.3, 1.7, 1.1, and 0.57 ml/kg. Enhancement values of the aorta, hepatic veins, inferior vena cava, and hepatic parenchyma were measured using a region-of-interest cursor. Results. At all contrast doses, equal volumes of contrast iohexol-350 resulted in statistically higher hepatic (p <_ .003, paired Student's t tes0 as well as vascular (p < .04) enhancement compared with iohexol-240. The slopes and intercepts of the enhancement curves for iohexol-350 and iohexol-240 were not statistically different (p >> .05) w h e n enhancement was plotted as a function of total grams of iodine administered. C o n c l u s i o n . With spiral CT scanning, appropriate contrast doses and relative costs per dose of iohexol should be considered on the basis of total iodine load administered rather than total volume administered.
Key W o r d s . Computed tomography (CT); helical (spiral) technology; contrast media; liver CT.
nfusion of iodinated contrast agents for abdominal computed tomography (CT) scanning generally results in an increased attenuation difference between pathologic and nonpathologic tissues. Higher levels of enhancement are desirable to increase the conspicuity of focal hepatic abnormalities. Previous researchers have evaluated contrast protocols for dynamic CT scanning [1-4], with an emphasis on hepatic enhancement. Using dynamic CT scanning, peak vascular and hepatic parenchymal enhancement increases with increased rates of injection [5-7] and increased contrast volumes [8]. Additionally, contrast distribution from the intravascular to the extravascular spaces affects lesion conspicuity in the liver [9, 10].
I
226
From the Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins Medical Institutions, Baltimore, MD. Support for this project was provided through a research grant from Sanofi-Winthrop Pharmaceuticals. Address reprint requests to D. A. Bluemke, MD, PhD, Department of Radiology, Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287. Received August 30, 1994, and accepted for publication after revision October 18, 1994. Acad Radio11995;2:226-231 © 1995, Association of University Radiologists
Vol. 2, No. 3, March 1995
EFFECT OF CONTRAST CONCENTRATION
The concentration (weight/volume, such as milligrams of iodine per milliliter) of a contrast agent m a y also be altered. Increasing the concentration of an iodinated contrast agent potentially m a y result in similar levels of e n h a n c e m e n t using lower volumes of contrast. This strategy m a y be useful to reduce the cost per dose of nonionic contrast agents [11] because the cost per dose is frequently based more in relation to volume than total iodine content. Thus, if the same n u m b e r of grams of iodine can be obtained in a smaller volume, the cost per dose will be reduced. Contrary results were obtained in two studies using human participants that examined the effect of concentration on e n h a n c e m e n t using dynamic CT scanning. Baker et al. [11] concluded that similar total iodine loads (total grams of iodine administered) result in similar hepatic enhancement, but Singer et al. [12] found no difference in e n h a n c e m e n t using different total iodine loads of iohexol. A study of ionic and nonionic CT contrast agents labeled with 1251 in rats s h o w e d that the iodine content of various tissues was independent of concentration up to 300 sec after administration [13]. With spiral CT scanning, the entire a b d o m e n can b e scanned within 20-30 sec. Contrast injection protocols therefore n e e d to optimize e n h a n c e m e n t levels over only a relatively brief interval c o m p a r e d with dynamic CT scanning. This is likely to require different strategies to optimize average e n h a n c e m e n t over the duration of the spiral CT examination. To date, evaluation of injection protocols for spiral CT scanning have focused on varying the volume or the rate of contrast injection [8, 14, 15]. In this study, w e c o m p a r e d abdominal enhancement using two different concentrations of a nonionic contrast using spiral CT scanning. Experiments were p e r f o r m e d using disease-free rabbits in a paired fashion (i.e., evaluating each animal at both concentrations). /
MATERIALS AND METHODS The protocol for this investigation was approved b y our institutional animal care and use committee. Eleven male New Zealand White rabbits (weight = 3-4 kg) were evaluated at each dose of contrast agent. Animal care and maintenance complied with National Institutes of Health guidelines. The animals were premedicated with acepromazine maleate (10 mg) and ketamine hydrochloride (150 mg) intramuscularly and were allowed to breathe r o o m air. Respiration was not controlled during the CT scanning procedure. Additional anesthesia was
provided using intravenous (IV) sodium thiamylal. W access was obtained using a 21-gauge butterfly needle in an ear vein. Animal weights and blood samples for blood urea nitrogen (BUN) and creatinine were obtained weekly immediately prior to contrast injection. T w o different concentrations of iohexol (Omnipaque, Nycomed, USA) were evaluated: 240 mg I/ml and 350 mg I/ml. Four different volumes doses of contrast agent at both concentrations were evaluated on the basis of the animal's weight: 2.3, 1.7, 1.1, and 0.57 ml iohexol/ kg. (For comparison purposes, these contrast volumes correspond to 160, 120, 80, and 40 ml of contrast, respectively, for a weight of 70 kg.) Because the relative iodine content per milliliter of the two iohexol concentrations differed, the total iodine load at each volume also varied. For iohexol-350, these corresponded to iodine loads of 800, 600, 400, and 200 mg I/kg, respectively. For iohexol-240, the iodine loads were 540, 410, 270, and 140 mg I/kg, respectively. Thus, eight different iodine loads (i.e., eight combinations of iohexol concentration and volume dose) w e r e evaluated. The CT scanning protocol consisted of evaluating one of the eight different iodine loads each week, for a total of 8 w e e k s of the investigation. Each week, 11 animals were evaluated and received the same iodine loads. One animal had a dynamic CT scan, and the other 10 had spiral CT scans, as discussed below. Each week, a dynamic CT scan was performed on one animal to determine the appropriate delay time after injection before beginning the spiral CT scan. The same animal was used for this experiment each w e e k at all contrast doses and concentrations. Dynamic CT scanning was performed at the same level in the midliver at a rate of 12 slices per minute for 200 sec after contrast injection. Attenuation profiles for the liver were then plotted as a function of time. The delay time was chosen as the interval b e t w e e n the beginning of the injection until the time for p e a k hepatic enhancement. Delay times were 45, 40, 30, and 25 sec for iohexol-350 at 2.3, 1.7, 1.1, and 0.57 ml/kg, respectively. For iohexol-240, delay times were 40, 35, 30, and 25 sec at 2.3, 1.7, 1.1, and 0.57 ml/kg, respectively. After the dynamic CT scan on one animal, the same contrast concentration and volume dose was then' administered to 10 animals for spiral CT scanning. For all experiments, the contrast agent was injected at a rate of 2 ml/sec. Before each spiral CT scan, an initial noncontrast scan was obtained in the midliver superior to the portal vein. The appropriate contrast dose was then injected. After a specified delay time (discussed earlier), 227
B L U E M K E ET AL.
spiral CT scanning began at the level of the diaphragm and continued caudally through the abdomen. On two occasions, the scheduled dose of contrast agent could not be administered to an animal because of lack of IV access. On a third occasion, a computer software error resulted in the CT scanner not starting at the required delay time. CT scans w e r e performed on a commercially available Siemens Somatom Plus scanner (Siemens Medical Systems, Iselin, NJ) with software version D and spiral CT option. The table s p e e d was 4 mm/sec, with 4-mm collimation. The scan duration was 24 sec. The total scanned length was 9.6 cm. Spiral CT scans were performed at 165 mA and 120 kVp. Attenuation values were measured for liver, aorta, inferior vena cava, and a middle hepatic vein in the central portion of the liver parenchyma. Attenuation values in Hounsfield units were obtained using region-of-interest (ROD analysis software on the Somatom Plus, sampiing away from streak or b e a m hardening artifacts. For the liver, two points on each CT scan slice were evaluated and averaged. Circular ROIs averaging 30-40 pixels in size were evaluated every 4 m m (corresponding to successive 1-sec intervals). The length of the liver was scanned within 8-12 sec (3.2-4.8 cm) in all cases, and these attenuation values were then used to determine the overall average liver attenuation during the spiral scan. For the aorta and inferior vena cava, overall average attenuation values were obtained b y averaging attenuation values over the duration of the spiral CT scan. For the hepatic veins, attenuation values were obtained over the initial 4--8 sec of the spiral scan and averaged as the mean. Mean values + 1 standard deviation were determined in all cases. Tissue enhancement was calculated by subtracting the noncontrast attenuation from the attenuation after contrast administration. Enhancement values for each animal were paired at each dose of iohexol to test the null hypothesis that equivalent volumes per kilogram of contrast agent would yield the same level of enhancement. A paired Student's t test was used to evaluate the null hypothesis. Differences were considered statistically significant if the probability level was less than or equal to .05. As discussed previously, there w e r e three cases in which a substitute animal was used because of lack of IV access or a technical CT scan failure. In these cases, there was no paired animal for statistical analysis. Enhancement as a function of total iodine load (grams of iodine administered) was also calculated for each animal for iohexol-240 and iohexol-350. Paired 228
Vol. 2, No. 3, March 1995
differences b e t w e e n the slopes and intercepts of each curve w e r e determined and evaluated with the Student's t test. Curve fitting was performed using CACricket Graph III vl.01 software (Computer Associates International, NY). RESULTS
All contrast injections were tolerated well without apparent side effects or evidence of contrast agent reaction. BUN and creatinine were normal for all animals throughout the study. Paired enhancement values for the liver at four different volume doses of contrast are shown in Figure 1. For three animals at the lowest contrast dose of 0.6 ml iohexol/kg, enhancement was slightly greater for iohexol-240 compared with iohexol350. In all other cases, hepatic enhancement was greater for iohexol-350 compared with iohexol-240 at the same volume dose. The differences between iohexol-350 and iohexol-240 were statistically significant (ps = .0008, .003, .005, and .003 for volume doses of 2.3, 1.7, 1.1, and 0.6 ml iohexol/kg, respectively). Average enhancements over the duration of the spiral scan for the liver, aorta, hepatic veins, and inferior vena cava for iohexol-350 and iohexol-240 are s h o w n in Figure 2. For the aorta, the difference in e n h a n c e m e n t b e t w e e n iohexol-350 and iohexol-240 was statistically significant (ps < .002 for all volume doses). For the inferior vena cava, differences were statistically significant (p < .04). For the hepatic veins, e n h a n c e m e n t could not be reliably measured for 9 of 10 animals at 0.6 ml iohexol-240/kg. For the other three larger volu m e doses, enhancement was significantly greater with iohexol-350 c o m p a r e d with iohexol-240 (p _< .002) Total iodine loads (grams of iodine administered) for each of the four doses of iohexol-350 and iohexol-240 were calculated and e n h a n c e m e n t values determined (Fig. 3). Confirming the visual a p p e a r a n c e of the curves, there was no statistical difference in the slopes and intercepts of the enhancement curves b e t w e e n iohexol-350 and iohexol-240 for the liver or for the vasculature (p >> .05). DISCUSSION
In this study, we evaluated the effect of changes in concentration of a nonionic iodinated contrast agent on the e n h a n c e m e n t of the liver and abdominal vasculature using spiral CT scanning. To our knowledge, this has b e e n evaluated in only two patient studies with dynamic CT scanning with opposing results [11, 12]. For
Vol. 2, No. 3, March 1995
Hepatic Enhancement:
EFFECT OF CONTRAST
Hepatic Enhancement:
2.3 m l i o h e x o l / k g
oillttllllll I
FIGURE 1. Hepatic enhancement for each animal for two concentrations of iohexol at four administered volumes of contrast agent. Variation between animals is approximately 10-15% of the mean: 2.3 ml iohexol/kg (A), 1.7 ml iohexol/kg (B), 1.1 ml iohexol/kg (C), and 0.6 ml iohexol/kg (D). HU = Hounefield units.
1.7 ml iohexol/kg
80
70
70
60
~ . 60
50
40
•
iohexo1350
30
[]
iohexo1240
20
~
30
[]
10 0
2
3
4
5 6 7 Animal
8
9
iohexo1240
~ 2o
1
1
CONCENTRATION
i
i 2
10
i 3
r 4
i 5
t 15
i 7
i 8
i 9
i 10
Animal
B
A Hepatic Enhancement:
Hepatic Enhancement:
1.1 m l i o h e x o l / k g
80
0.6 ml iohexol/kg
80
70
6oj
60
4O2o3O1ol 50
•
iohexol 350
•
iohexo1350
[]
iohexol 240
[]
iohexo1240
•
iohexo1350
[]
iohexo1240
"i Jitttttti
a=
~ 20 1
0
f
f
i 3
t 4
F i i 5 6 7 Animal
i 8
i 9
i l0
1
2
3
4
5 6 7 Animal
8
9
10
D
C Average Hepatic Enhancement
Average Aortic Enhancement
70
140
60 T
120
Y
100 •
iohexo1350
80
[]
iohexo1240
60
T
4O
0 2.3
1.7
, 1.1
o 2.3
0.6
i 1.7
i 1.1
0.6
Volume D o s e ( m ~ g )
Volume Dose(ml/kg)
A
B Average Hepatic Vein Enhancement
Average IVC
Enhancement
140 120
T
80 i
60
~
40
T
[]
iohexo1350
[]
iohexol 2A0
20 0
, 2.3
1.7
LI
2.3
0.6
1.1
0.6
Volume Dose (ml/kg)
Volume Dose (ml/kg)
C
IlL
1.7
•
iohexo1350
[]
iohexo1240
FIGURE 2. Average enhancement of the liver (A), aorta (B), hepatic veins (C), and inferior vena cava (D) for two concentrations of iohexol. At all volume doses, enhancement was statistically significantly superior with iohexol-350 compared with iohexol240. Vertical bars represent 1 standard deviation. IVC = inferior vena cava, HU = Hounsfield units.
D
the same volume of administered contrast agent, our data indicate that iohexol-350 resulted in statistically significant increased enhancement compared with iohexol-240 during spiral CT scanning. The enhance-
ment curves for the two concentrations of nonionic contrast agent had a similar slope and intercept w h e n vascular and hepatic enhancement was plotted as a function of the total administered iodine load (Fig. 3). 229
BLUEMKE
ET AL.
FIGURE 3. Enhancement curves for iohexol-350 and iohexol-240 plotted as a function of total administered dose of iodine per kilogram in the liver (A), aorta (B), hepatic veins (C), and inferior vena cava (D). No statistical difference in the slopes or intercept was present between the two curves at each concentration. Vertical bars represent I standard deviation. CT = computed tomography, HU = Hounsfield units.
vol. 2, No. 3, March 1995
Spiral CT: Liver Enhancement
70~ 60q ~ S04 404 i
T ~
T ±
T
~co 8o 60
80q 20q 10Q Ol
Spiral CT: Aorta Enhancement
140
T
~ ~
iobexo1350 iohexo124O
1
o
20O
do
do
IedlneLoad
(rag
80O
r~ 40 2
l&
o
~
~
~ iohexo1350 ----o---- iohexo1240
40o
20O
l/kg)
&
IodineLoad
A
(mg
8~
-S~
I/kg)
B Spiral CT: Hepatic Vein Enhancement
i4o
Spiral CT: IVC Enhancement
T
T
120t
± i= 8oq 604
S
±
----c--
404 20q off o
~ 2~o
do
&
IodineLoad
C
Our results for spiral CT scanning are similar to those of Baker et al. [11], w h o found the same degree of hepatic enhancement using 150 ml iopamidol-300 (45 g iodine, 168 patients) compared with 125 ml ioversol-320 (40 g iodine, 119 patients) in a heterogeneous but large patient population. Because two different contrast agents were used, it is not clear whether similar hepatic enhancement was attributable to variation in the contrast agent, similar iodine loads, or both. Also, small differences between the two groups might have been obscured by the heterogeneous patient population. In a smaller study, Singer et al. [12] evaluated 20 patients, w h o each received the same volume of iohexol-240 or iohexol-300. Singer et al. found no difference in hepatic or vascular enhancement for iohexol-240 and iohexol300. As suggested by Dodd and Baron [16], differences in a heterogeneous patient population may be obscured by interpatient variability. In addition, the results of Singer et al. and Baker et al. were based on dynamic CT scanning over an approximately 2-min interval. The applicability of those results to the short scan interval with spiral CT scanning is unclear. The consistent trends w e observed were likely aided by paired observations on the same experimental group of animals. This approach helps to control intrinsic variables that may influence hepatic and vascular enhancement [16, 17]. Factors such as renal function, the presence or absence of a disease state, including 230
(rag
iohexo1350 iobexo124O
do
T
----a---
iohexo1350
ioh~o1240
20O
~&
4~
do
80O
l~
IodineLoad(ragI/kg)
I/kg)
D that involving the liver or cardiovascular system, weight, and hydration are likely to affect the reproducibility of measurements of enhancement. For animal studies, the level of anesthesia also may affect cardiac output. Thus, we observed modest variation--standard deviations were typically 10-15% (Fig. 1 ) - - b e t w e e n animals in hepatic and vascular enhancement. Our data indicate that hepatic and vascular enhancement achieved during spiral CT scanning of the abdomen is proportional to the total iodine load administered. The enhancement curves shown in Figure 3, however, suggest a trend toward less enhancement per gram of iodine at the highest contrast loads. Because there were only four sample points for each concentration along the x-axis, there are insufficient data to determine the nature of the function (e.g., linear, logarithmic) relating the contrast dose to enhancement. A disease-free animal model was used in this study to provide a careful evaluation of hepatic enhancement over a range of contrast doses. Animal studies of nonionic contrast agents for CT scanning have accurately demonstrated the relations between contrast doses, injection rates, and timing protocols and CT enhancement [9, 18-20]. Direct extrapolation, however, of these results to human studies is limited; different volumes of distribution of iohexol and circulation times between species indicate that a direct correlation between administered iodine load and enhancement is unlikely.
Vol. 2, No. 3, March 1995
Our data are also limited to average enhancement values during the duration of the spiral CT scan; w e chose this as the plateau phase of enhancement following peak enhancement. Nevertheless, important principles for human trials have been established. In conclusion, our results indicate that hepatic and vascular enhancement measured by spiral CT scanning in the rabbit is significantly higher using iohexol-350 compared with iohexol-240 for the same volume dose. Furthermore, enhancement is roughly proportional to the total iodine load administered. At this point, our data indicate the appropriate patient dose should be evaluated on the basis of the total administered iodine load. The cost of contrast agents is frequently based more on the total volume rather than the total iodine load; if the same number of grams of iodine can be obtained in a smaller volume, the contrast agent cost per patient will be lower. ACKNOWLEDGMENTS
We would like to extend our thanks to Nancy Spangler, RT, and Carolyn Magee, BS, for their considerable efforts in helping to perform this study. We especially thank Paul Brown, MD, for his suggestions and encouragement. REFERENCES 1. Heiken JP, Brink JA, McClennan BL, Sagel SS, Forman HP, DiCroce J. Dynamic contrast-enhanced CT of the liver: comparison of contrast medium injection rates and uniphasic and biphasic injection protocols. Radiology 1993; 187:327-331. 2. Foley WD. Dynamic hepatic CT. Radiology 1989; 170:617-622. 3. Platt JF, Glazer GM. IV contrast material for abdominal CT: comparison of three methods of administration. AJR 1988;151:275-277. 4. Nelson RC, Moyers JH, Chezmar JL, et al. Hepatic dynamic sequential CT: section enhancement profiles with a bolus of ionic and nonionic con-
EFFECT OF CONTRAST
CONCENTRATION
trast agents. Radiology 1991 ;178:499-502. 5. Claussen CD, Banzer D, Pfretzschner C, Kalender WA, SchSrner W. Bolus geometry and dynamics after intravenous contrast medium injection. Radiology 1984; 153:365-368. 6. Berland LL, Lee JY. Comparison of contrast media injection rates and volumes for dynamic incremented computed tomography. Invest Radiol 1988;23:918-922. 7. Harmon BH, Berland LL, Lee JY. Effect of varying rates of Iow-osmolarity contrast media injection for hepatic CT: correlation with indocyanine green transit time. Radiology 1992; 184:379-382. 8. Small WC, Nelson RC, Bernardino ME, Brummer LT. Contrast-enhanced spiral CT of the liver: effect of different amounts and injection rates of contrast material on early contrast enhancement. AJR 1994;163:87-92. 9. Burgener FA, Hamlin DJ. Contrast enhancement in abdominal CT: bolus vs. infusion. AJR 1981 ;137:351-358. 10. Foley WD, Berland LL, Lawson TL, Smith DF, Thorsen MK. Contrast enhancement technique for dynamic hepatic computed tomographic scanning. Radiology 1983;147:797-803. 11. Baker ME, Beam C, Leder R, Gulliver D, Paine SS, Dunnick NR. Contrast material for combined abdominal and pelvic CT: can cost be reduced by increasing the concentration and decreasing the volume? AJR 1993;160: 637-641. 12. Singer AA, Tagliabue JR, Paushter DM, Borkowski GP, Einstein DM. Comparison of iohexol-240 versus iohexol-300 in abdominal CT. Gastrointest Radiol 1992; 17:122-124, 13. Dean PB, Kivisaari L, Kormano M. Does dilution of contrast media affect contrast enhancement? An experimental study in rats. Invest Radio11988; 23(suppl):S118-S121. 14. Foley WD, Hoffman RG, Quiroz FA, Kahn CEJ, Perret RS. Hepatic helical CT: contrast material injection protocol. Radiology 1994;192:367-371. 15. Bluemke DA, Fishman EK, Anderson JH. Dose requirements for a nonionic contrast agent for spiral computed tomography of the liver in rabbits. Invest Radio11994;29:195-200. 16. Dodd GD Ill, Baron RL. Investigation of contrast enhancement in CT of the liver: the need for improved methods. AJR 1993;160:643-645. 17. Chambers TP, Baron RL, Lush RM, Dodd GD, Miller WJ, Confer SR. Hepatic enhancement: a method to demonstrate reproducibility, Radiology 1993;188:627-631. 18. Dean PB, Violante MR, Mahoney JA. Hepatic CT contrast enhancement: effect of dose, duration of infusion, and time elapsed following infusion. Invest Radio11990;15:158-161. 19. Young SW, Turner RJ, Castellino RA. A strategy for the contrast enhancement of malignant tumors using dynamic computed tomography and intravascular pharmacokinetics. Radiology 1980; 137:137-147. 20. Dean PS, Kivisaafi L, Kormano M. Contrast enhancement pharmacokinetics of six ionic and nonionic contrast media. Invest Radio11983;18:368-374.
231
Rationale and Objectives. We examined the effect of varying the concentration of a nonionic iodinated contrast agent on hepatic and abdominal vascular enhancement using spiral computed tomography (CT) scanning. M e t h o d s . Spiral CT scans of 10 New Zealand male rabbits were obtained after intravenous injection of 240 mg I/ml iohexol and 350 mg I/ml iohexol injected at a rate of 2 ml/sec. Each animal was studied at both concentrations using four different contrast volumes: 2.3, 1.7, 1.1, and 0.57 ml/kg. Enhancement values of the aorta, hepatic veins, inferior vena cava, and hepatic parenchyma were measured using a region-of-interest cursor. Results. At all contrast doses, equal volumes of contrast iohexol-350 resulted in statistically higher hepatic (p <_ .003, paired Student's t tes0 as well as vascular (p < .04) enhancement compared with iohexol-240. The slopes and intercepts of the enhancement curves for iohexol-350 and iohexol-240 were not statistically different (p >> .05) w h e n enhancement was plotted as a function of total grams of iodine administered. C o n c l u s i o n . With spiral CT scanning, appropriate contrast doses and relative costs per dose of iohexol should be considered on the basis of total iodine load administered rather than total volume administered.
Key W o r d s . Computed tomography (CT); helical (spiral) technology; contrast media; liver CT.
nfusion of iodinated contrast agents for abdominal computed tomography (CT) scanning generally results in an increased attenuation difference between pathologic and nonpathologic tissues. Higher levels of enhancement are desirable to increase the conspicuity of focal hepatic abnormalities. Previous researchers have evaluated contrast protocols for dynamic CT scanning [1-4], with an emphasis on hepatic enhancement. Using dynamic CT scanning, peak vascular and hepatic parenchymal enhancement increases with increased rates of injection [5-7] and increased contrast volumes [8]. Additionally, contrast distribution from the intravascular to the extravascular spaces affects lesion conspicuity in the liver [9, 10].
I
226
From the Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins Medical Institutions, Baltimore, MD. Support for this project was provided through a research grant from Sanofi-Winthrop Pharmaceuticals. Address reprint requests to D. A. Bluemke, MD, PhD, Department of Radiology, Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287. Received August 30, 1994, and accepted for publication after revision October 18, 1994. Acad Radio11995;2:226-231 © 1995, Association of University Radiologists
Vol. 2, No. 3, March 1995
EFFECT OF CONTRAST CONCENTRATION
The concentration (weight/volume, such as milligrams of iodine per milliliter) of a contrast agent m a y also be altered. Increasing the concentration of an iodinated contrast agent potentially m a y result in similar levels of e n h a n c e m e n t using lower volumes of contrast. This strategy m a y be useful to reduce the cost per dose of nonionic contrast agents [11] because the cost per dose is frequently based more in relation to volume than total iodine content. Thus, if the same n u m b e r of grams of iodine can be obtained in a smaller volume, the cost per dose will be reduced. Contrary results were obtained in two studies using human participants that examined the effect of concentration on e n h a n c e m e n t using dynamic CT scanning. Baker et al. [11] concluded that similar total iodine loads (total grams of iodine administered) result in similar hepatic enhancement, but Singer et al. [12] found no difference in e n h a n c e m e n t using different total iodine loads of iohexol. A study of ionic and nonionic CT contrast agents labeled with 1251 in rats s h o w e d that the iodine content of various tissues was independent of concentration up to 300 sec after administration [13]. With spiral CT scanning, the entire a b d o m e n can b e scanned within 20-30 sec. Contrast injection protocols therefore n e e d to optimize e n h a n c e m e n t levels over only a relatively brief interval c o m p a r e d with dynamic CT scanning. This is likely to require different strategies to optimize average e n h a n c e m e n t over the duration of the spiral CT examination. To date, evaluation of injection protocols for spiral CT scanning have focused on varying the volume or the rate of contrast injection [8, 14, 15]. In this study, w e c o m p a r e d abdominal enhancement using two different concentrations of a nonionic contrast using spiral CT scanning. Experiments were p e r f o r m e d using disease-free rabbits in a paired fashion (i.e., evaluating each animal at both concentrations). /
MATERIALS AND METHODS The protocol for this investigation was approved b y our institutional animal care and use committee. Eleven male New Zealand White rabbits (weight = 3-4 kg) were evaluated at each dose of contrast agent. Animal care and maintenance complied with National Institutes of Health guidelines. The animals were premedicated with acepromazine maleate (10 mg) and ketamine hydrochloride (150 mg) intramuscularly and were allowed to breathe r o o m air. Respiration was not controlled during the CT scanning procedure. Additional anesthesia was
provided using intravenous (IV) sodium thiamylal. W access was obtained using a 21-gauge butterfly needle in an ear vein. Animal weights and blood samples for blood urea nitrogen (BUN) and creatinine were obtained weekly immediately prior to contrast injection. T w o different concentrations of iohexol (Omnipaque, Nycomed, USA) were evaluated: 240 mg I/ml and 350 mg I/ml. Four different volumes doses of contrast agent at both concentrations were evaluated on the basis of the animal's weight: 2.3, 1.7, 1.1, and 0.57 ml iohexol/ kg. (For comparison purposes, these contrast volumes correspond to 160, 120, 80, and 40 ml of contrast, respectively, for a weight of 70 kg.) Because the relative iodine content per milliliter of the two iohexol concentrations differed, the total iodine load at each volume also varied. For iohexol-350, these corresponded to iodine loads of 800, 600, 400, and 200 mg I/kg, respectively. For iohexol-240, the iodine loads were 540, 410, 270, and 140 mg I/kg, respectively. Thus, eight different iodine loads (i.e., eight combinations of iohexol concentration and volume dose) w e r e evaluated. The CT scanning protocol consisted of evaluating one of the eight different iodine loads each week, for a total of 8 w e e k s of the investigation. Each week, 11 animals were evaluated and received the same iodine loads. One animal had a dynamic CT scan, and the other 10 had spiral CT scans, as discussed below. Each week, a dynamic CT scan was performed on one animal to determine the appropriate delay time after injection before beginning the spiral CT scan. The same animal was used for this experiment each w e e k at all contrast doses and concentrations. Dynamic CT scanning was performed at the same level in the midliver at a rate of 12 slices per minute for 200 sec after contrast injection. Attenuation profiles for the liver were then plotted as a function of time. The delay time was chosen as the interval b e t w e e n the beginning of the injection until the time for p e a k hepatic enhancement. Delay times were 45, 40, 30, and 25 sec for iohexol-350 at 2.3, 1.7, 1.1, and 0.57 ml/kg, respectively. For iohexol-240, delay times were 40, 35, 30, and 25 sec at 2.3, 1.7, 1.1, and 0.57 ml/kg, respectively. After the dynamic CT scan on one animal, the same contrast concentration and volume dose was then' administered to 10 animals for spiral CT scanning. For all experiments, the contrast agent was injected at a rate of 2 ml/sec. Before each spiral CT scan, an initial noncontrast scan was obtained in the midliver superior to the portal vein. The appropriate contrast dose was then injected. After a specified delay time (discussed earlier), 227
B L U E M K E ET AL.
spiral CT scanning began at the level of the diaphragm and continued caudally through the abdomen. On two occasions, the scheduled dose of contrast agent could not be administered to an animal because of lack of IV access. On a third occasion, a computer software error resulted in the CT scanner not starting at the required delay time. CT scans w e r e performed on a commercially available Siemens Somatom Plus scanner (Siemens Medical Systems, Iselin, NJ) with software version D and spiral CT option. The table s p e e d was 4 mm/sec, with 4-mm collimation. The scan duration was 24 sec. The total scanned length was 9.6 cm. Spiral CT scans were performed at 165 mA and 120 kVp. Attenuation values were measured for liver, aorta, inferior vena cava, and a middle hepatic vein in the central portion of the liver parenchyma. Attenuation values in Hounsfield units were obtained using region-of-interest (ROD analysis software on the Somatom Plus, sampiing away from streak or b e a m hardening artifacts. For the liver, two points on each CT scan slice were evaluated and averaged. Circular ROIs averaging 30-40 pixels in size were evaluated every 4 m m (corresponding to successive 1-sec intervals). The length of the liver was scanned within 8-12 sec (3.2-4.8 cm) in all cases, and these attenuation values were then used to determine the overall average liver attenuation during the spiral scan. For the aorta and inferior vena cava, overall average attenuation values were obtained b y averaging attenuation values over the duration of the spiral CT scan. For the hepatic veins, attenuation values were obtained over the initial 4--8 sec of the spiral scan and averaged as the mean. Mean values + 1 standard deviation were determined in all cases. Tissue enhancement was calculated by subtracting the noncontrast attenuation from the attenuation after contrast administration. Enhancement values for each animal were paired at each dose of iohexol to test the null hypothesis that equivalent volumes per kilogram of contrast agent would yield the same level of enhancement. A paired Student's t test was used to evaluate the null hypothesis. Differences were considered statistically significant if the probability level was less than or equal to .05. As discussed previously, there w e r e three cases in which a substitute animal was used because of lack of IV access or a technical CT scan failure. In these cases, there was no paired animal for statistical analysis. Enhancement as a function of total iodine load (grams of iodine administered) was also calculated for each animal for iohexol-240 and iohexol-350. Paired 228
Vol. 2, No. 3, March 1995
differences b e t w e e n the slopes and intercepts of each curve w e r e determined and evaluated with the Student's t test. Curve fitting was performed using CACricket Graph III vl.01 software (Computer Associates International, NY). RESULTS
All contrast injections were tolerated well without apparent side effects or evidence of contrast agent reaction. BUN and creatinine were normal for all animals throughout the study. Paired enhancement values for the liver at four different volume doses of contrast are shown in Figure 1. For three animals at the lowest contrast dose of 0.6 ml iohexol/kg, enhancement was slightly greater for iohexol-240 compared with iohexol350. In all other cases, hepatic enhancement was greater for iohexol-350 compared with iohexol-240 at the same volume dose. The differences between iohexol-350 and iohexol-240 were statistically significant (ps = .0008, .003, .005, and .003 for volume doses of 2.3, 1.7, 1.1, and 0.6 ml iohexol/kg, respectively). Average enhancements over the duration of the spiral scan for the liver, aorta, hepatic veins, and inferior vena cava for iohexol-350 and iohexol-240 are s h o w n in Figure 2. For the aorta, the difference in e n h a n c e m e n t b e t w e e n iohexol-350 and iohexol-240 was statistically significant (ps < .002 for all volume doses). For the inferior vena cava, differences were statistically significant (p < .04). For the hepatic veins, e n h a n c e m e n t could not be reliably measured for 9 of 10 animals at 0.6 ml iohexol-240/kg. For the other three larger volu m e doses, enhancement was significantly greater with iohexol-350 c o m p a r e d with iohexol-240 (p _< .002) Total iodine loads (grams of iodine administered) for each of the four doses of iohexol-350 and iohexol-240 were calculated and e n h a n c e m e n t values determined (Fig. 3). Confirming the visual a p p e a r a n c e of the curves, there was no statistical difference in the slopes and intercepts of the enhancement curves b e t w e e n iohexol-350 and iohexol-240 for the liver or for the vasculature (p >> .05). DISCUSSION
In this study, we evaluated the effect of changes in concentration of a nonionic iodinated contrast agent on the e n h a n c e m e n t of the liver and abdominal vasculature using spiral CT scanning. To our knowledge, this has b e e n evaluated in only two patient studies with dynamic CT scanning with opposing results [11, 12]. For
Vol. 2, No. 3, March 1995
Hepatic Enhancement:
EFFECT OF CONTRAST
Hepatic Enhancement:
2.3 m l i o h e x o l / k g
oillttllllll I
FIGURE 1. Hepatic enhancement for each animal for two concentrations of iohexol at four administered volumes of contrast agent. Variation between animals is approximately 10-15% of the mean: 2.3 ml iohexol/kg (A), 1.7 ml iohexol/kg (B), 1.1 ml iohexol/kg (C), and 0.6 ml iohexol/kg (D). HU = Hounefield units.
1.7 ml iohexol/kg
80
70
70
60
~ . 60
50
40
•
iohexo1350
30
[]
iohexo1240
20
~
30
[]
10 0
2
3
4
5 6 7 Animal
8
9
iohexo1240
~ 2o
1
1
CONCENTRATION
i
i 2
10
i 3
r 4
i 5
t 15
i 7
i 8
i 9
i 10
Animal
B
A Hepatic Enhancement:
Hepatic Enhancement:
1.1 m l i o h e x o l / k g
80
0.6 ml iohexol/kg
80
70
6oj
60
4O2o3O1ol 50
•
iohexol 350
•
iohexo1350
[]
iohexol 240
[]
iohexo1240
•
iohexo1350
[]
iohexo1240
"i Jitttttti
a=
~ 20 1
0
f
f
i 3
t 4
F i i 5 6 7 Animal
i 8
i 9
i l0
1
2
3
4
5 6 7 Animal
8
9
10
D
C Average Hepatic Enhancement
Average Aortic Enhancement
70
140
60 T
120
Y
100 •
iohexo1350
80
[]
iohexo1240
60
T
4O
0 2.3
1.7
, 1.1
o 2.3
0.6
i 1.7
i 1.1
0.6
Volume D o s e ( m ~ g )
Volume Dose(ml/kg)
A
B Average Hepatic Vein Enhancement
Average IVC
Enhancement
140 120
T
80 i
60
~
40
T
[]
iohexo1350
[]
iohexol 2A0
20 0
, 2.3
1.7
LI
2.3
0.6
1.1
0.6
Volume Dose (ml/kg)
Volume Dose (ml/kg)
C
IlL
1.7
•
iohexo1350
[]
iohexo1240
FIGURE 2. Average enhancement of the liver (A), aorta (B), hepatic veins (C), and inferior vena cava (D) for two concentrations of iohexol. At all volume doses, enhancement was statistically significantly superior with iohexol-350 compared with iohexol240. Vertical bars represent 1 standard deviation. IVC = inferior vena cava, HU = Hounsfield units.
D
the same volume of administered contrast agent, our data indicate that iohexol-350 resulted in statistically significant increased enhancement compared with iohexol-240 during spiral CT scanning. The enhance-
ment curves for the two concentrations of nonionic contrast agent had a similar slope and intercept w h e n vascular and hepatic enhancement was plotted as a function of the total administered iodine load (Fig. 3). 229
BLUEMKE
ET AL.
FIGURE 3. Enhancement curves for iohexol-350 and iohexol-240 plotted as a function of total administered dose of iodine per kilogram in the liver (A), aorta (B), hepatic veins (C), and inferior vena cava (D). No statistical difference in the slopes or intercept was present between the two curves at each concentration. Vertical bars represent I standard deviation. CT = computed tomography, HU = Hounsfield units.
vol. 2, No. 3, March 1995
Spiral CT: Liver Enhancement
70~ 60q ~ S04 404 i
T ~
T ±
T
~co 8o 60
80q 20q 10Q Ol
Spiral CT: Aorta Enhancement
140
T
~ ~
iobexo1350 iohexo124O
1
o
20O
do
do
IedlneLoad
(rag
80O
r~ 40 2
l&
o
~
~
~ iohexo1350 ----o---- iohexo1240
40o
20O
l/kg)
&
IodineLoad
A
(mg
8~
-S~
I/kg)
B Spiral CT: Hepatic Vein Enhancement
i4o
Spiral CT: IVC Enhancement
T
T
120t
± i= 8oq 604
S
±
----c--
404 20q off o
~ 2~o
do
&
IodineLoad
C
Our results for spiral CT scanning are similar to those of Baker et al. [11], w h o found the same degree of hepatic enhancement using 150 ml iopamidol-300 (45 g iodine, 168 patients) compared with 125 ml ioversol-320 (40 g iodine, 119 patients) in a heterogeneous but large patient population. Because two different contrast agents were used, it is not clear whether similar hepatic enhancement was attributable to variation in the contrast agent, similar iodine loads, or both. Also, small differences between the two groups might have been obscured by the heterogeneous patient population. In a smaller study, Singer et al. [12] evaluated 20 patients, w h o each received the same volume of iohexol-240 or iohexol-300. Singer et al. found no difference in hepatic or vascular enhancement for iohexol-240 and iohexol300. As suggested by Dodd and Baron [16], differences in a heterogeneous patient population may be obscured by interpatient variability. In addition, the results of Singer et al. and Baker et al. were based on dynamic CT scanning over an approximately 2-min interval. The applicability of those results to the short scan interval with spiral CT scanning is unclear. The consistent trends w e observed were likely aided by paired observations on the same experimental group of animals. This approach helps to control intrinsic variables that may influence hepatic and vascular enhancement [16, 17]. Factors such as renal function, the presence or absence of a disease state, including 230
(rag
iohexo1350 iobexo124O
do
T
----a---
iohexo1350
ioh~o1240
20O
~&
4~
do
80O
l~
IodineLoad(ragI/kg)
I/kg)
D that involving the liver or cardiovascular system, weight, and hydration are likely to affect the reproducibility of measurements of enhancement. For animal studies, the level of anesthesia also may affect cardiac output. Thus, we observed modest variation--standard deviations were typically 10-15% (Fig. 1 ) - - b e t w e e n animals in hepatic and vascular enhancement. Our data indicate that hepatic and vascular enhancement achieved during spiral CT scanning of the abdomen is proportional to the total iodine load administered. The enhancement curves shown in Figure 3, however, suggest a trend toward less enhancement per gram of iodine at the highest contrast loads. Because there were only four sample points for each concentration along the x-axis, there are insufficient data to determine the nature of the function (e.g., linear, logarithmic) relating the contrast dose to enhancement. A disease-free animal model was used in this study to provide a careful evaluation of hepatic enhancement over a range of contrast doses. Animal studies of nonionic contrast agents for CT scanning have accurately demonstrated the relations between contrast doses, injection rates, and timing protocols and CT enhancement [9, 18-20]. Direct extrapolation, however, of these results to human studies is limited; different volumes of distribution of iohexol and circulation times between species indicate that a direct correlation between administered iodine load and enhancement is unlikely.
Vol. 2, No. 3, March 1995
Our data are also limited to average enhancement values during the duration of the spiral CT scan; w e chose this as the plateau phase of enhancement following peak enhancement. Nevertheless, important principles for human trials have been established. In conclusion, our results indicate that hepatic and vascular enhancement measured by spiral CT scanning in the rabbit is significantly higher using iohexol-350 compared with iohexol-240 for the same volume dose. Furthermore, enhancement is roughly proportional to the total iodine load administered. At this point, our data indicate the appropriate patient dose should be evaluated on the basis of the total administered iodine load. The cost of contrast agents is frequently based more on the total volume rather than the total iodine load; if the same number of grams of iodine can be obtained in a smaller volume, the contrast agent cost per patient will be lower. ACKNOWLEDGMENTS
We would like to extend our thanks to Nancy Spangler, RT, and Carolyn Magee, BS, for their considerable efforts in helping to perform this study. We especially thank Paul Brown, MD, for his suggestions and encouragement. REFERENCES 1. Heiken JP, Brink JA, McClennan BL, Sagel SS, Forman HP, DiCroce J. Dynamic contrast-enhanced CT of the liver: comparison of contrast medium injection rates and uniphasic and biphasic injection protocols. Radiology 1993; 187:327-331. 2. Foley WD. Dynamic hepatic CT. Radiology 1989; 170:617-622. 3. Platt JF, Glazer GM. IV contrast material for abdominal CT: comparison of three methods of administration. AJR 1988;151:275-277. 4. Nelson RC, Moyers JH, Chezmar JL, et al. Hepatic dynamic sequential CT: section enhancement profiles with a bolus of ionic and nonionic con-
EFFECT OF CONTRAST
CONCENTRATION
trast agents. Radiology 1991 ;178:499-502. 5. Claussen CD, Banzer D, Pfretzschner C, Kalender WA, SchSrner W. Bolus geometry and dynamics after intravenous contrast medium injection. Radiology 1984; 153:365-368. 6. Berland LL, Lee JY. Comparison of contrast media injection rates and volumes for dynamic incremented computed tomography. Invest Radiol 1988;23:918-922. 7. Harmon BH, Berland LL, Lee JY. Effect of varying rates of Iow-osmolarity contrast media injection for hepatic CT: correlation with indocyanine green transit time. Radiology 1992; 184:379-382. 8. Small WC, Nelson RC, Bernardino ME, Brummer LT. Contrast-enhanced spiral CT of the liver: effect of different amounts and injection rates of contrast material on early contrast enhancement. AJR 1994;163:87-92. 9. Burgener FA, Hamlin DJ. Contrast enhancement in abdominal CT: bolus vs. infusion. AJR 1981 ;137:351-358. 10. Foley WD, Berland LL, Lawson TL, Smith DF, Thorsen MK. Contrast enhancement technique for dynamic hepatic computed tomographic scanning. Radiology 1983;147:797-803. 11. Baker ME, Beam C, Leder R, Gulliver D, Paine SS, Dunnick NR. Contrast material for combined abdominal and pelvic CT: can cost be reduced by increasing the concentration and decreasing the volume? AJR 1993;160: 637-641. 12. Singer AA, Tagliabue JR, Paushter DM, Borkowski GP, Einstein DM. Comparison of iohexol-240 versus iohexol-300 in abdominal CT. Gastrointest Radiol 1992; 17:122-124, 13. Dean PB, Kivisaari L, Kormano M. Does dilution of contrast media affect contrast enhancement? An experimental study in rats. Invest Radio11988; 23(suppl):S118-S121. 14. Foley WD, Hoffman RG, Quiroz FA, Kahn CEJ, Perret RS. Hepatic helical CT: contrast material injection protocol. Radiology 1994;192:367-371. 15. Bluemke DA, Fishman EK, Anderson JH. Dose requirements for a nonionic contrast agent for spiral computed tomography of the liver in rabbits. Invest Radio11994;29:195-200. 16. Dodd GD Ill, Baron RL. Investigation of contrast enhancement in CT of the liver: the need for improved methods. AJR 1993;160:643-645. 17. Chambers TP, Baron RL, Lush RM, Dodd GD, Miller WJ, Confer SR. Hepatic enhancement: a method to demonstrate reproducibility, Radiology 1993;188:627-631. 18. Dean PB, Violante MR, Mahoney JA. Hepatic CT contrast enhancement: effect of dose, duration of infusion, and time elapsed following infusion. Invest Radio11990;15:158-161. 19. Young SW, Turner RJ, Castellino RA. A strategy for the contrast enhancement of malignant tumors using dynamic computed tomography and intravascular pharmacokinetics. Radiology 1980; 137:137-147. 20. Dean PS, Kivisaafi L, Kormano M. Contrast enhancement pharmacokinetics of six ionic and nonionic contrast media. Invest Radio11983;18:368-374.
231