Contrast Enhancement in Cardiovascular MDCT: Effect of Body Weight, Height, Body Surface Area, Body Mass Index, and Obesity

Vascular Imaging Original Research Bae et al. Contrast Enhancement in Cardiovascular MDCT Vascular Imaging Original Research Kyongtae T. Bae 1 Brian A. Seeck 2 Charles F. Hildebolt 3 Cheng Tao 1 Fang Zhu 1 Masayuki Kanematsu 4 Pamela K. Woodard 3 Bae KT, Seeck BA, Hildebolt CF, et al. Keywords: aorta, cardiac CT, CT angiography, vascular imaging DOI:10.2214/AJR.07.2765 Received June 21, 7; accepted after revision September 26, 7. 1 Department of Radiology, University of Pittsburgh School of Medicine, Lothrop St., Ste. 4895, Pittsburgh, PA 15213. Address correspondence to K. T. Bae (baek@upmc.edu). 2 Division of Cardiology, Washington University School of Medicine, St. Louis, MO. 3 Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, MO. 4 Department of Radiology, Gifu University School of Medicine, Gifu, Japan. AJR 8; 190:777 784 0361 803X/08/1903 777 American Roentgen Ray Society Contrast Enhancement in Cardiovascular MDCT: Effect of Body Weight, Height, Body Surface Area, Body Mass Index, and Obesity OBJECTIVE. The purpose of our study was to evaluate the effect of body weight, height, body surface area (BSA), body mass index (BMI), and obesity on aortic contrast enhancement in cardiac MDCT. MATERIALS AND METHODS. Seventy-three consecutive patients underwent cardiac CT angiography on a 64-MDCT scanner. Seventy-five ml of contrast medium ( mg I/ ml) was injected at 4.5 ml/s, followed by a 40-mL saline flush at 4.5 ml/s. The scanning delay of CT was determined with a bolus tracking technique. Aortic attenuation was measured over the aortic-root lumen. BMI and BSA were calculated from the patient s body weight and height. The patients were divided into low- (BMI < 30) and high- ( 30) BMI groups. Associations of aortic attenuation with body weight, height, BMI, and BSA were evaluated with regression analysis and the Student s t test. RESULTS. Strong inverse correlations were seen between aortic attenuation and body weight (r = 0.73), height (r = 0.47), BMI (r = 0.63), and BSA (r = 0.74) (p < 0.001 for all). The regression formula of aortic attenuation versus body weight suggests that 1.0 ml/kg of contrast medium would yield a mean aortic attenuation of 355 H. The mean aortic attenuation was significantly higher in the low-bmi (352.6 ± 59.1 H) than in the high-bmi (286.2 ± 55.5 H) group. The regression formula for aortic attenuation on body weight was aortic attenuation = 586 3.1 body weight (p < 0.001) for the low-bmi group and aortic attenuation = 485 1.9 body weight (p < 0.001) for the high-bmi group, suggesting that the amount of contrast medium required with increased body weight is less in the high-bmi group. This group difference was less pronounced for the regression of aortic attenuation on BSA. CONCLUSION. To achieve a consistent contrast enhancement in cardiac CT angiography (CTA), contrast-medium dose should be adjusted with the body weight or the BSA (which accounts for both the body weight and height factors) to provide adjustment of iodine dose over a wide range of body sizes. M DCT angiography is increasingly used for the diagnostic evaluation of patients with clinically suspected obstructive coronary artery disease or abnormal cardiac or coronary anatomy. Advances in CT and ECGgating technology allow us to acquire highresolution, motion-free images of the heart and coronary arteries during a single short breath-hold [1, 2]. Markedly improved temporal resolution of CT has contributed to reducing motion artifacts and improving contrast enhancement and image quality [3 6]. Fast CT such as 64-MDCT allows us to obtain an entire scan of the coronary artery and heart in 10 seconds or less. As a result, scan timing becomes far more critical and challenging than with older and slower CT scanners [2, 7 9]. The other consideration with this challenge is that the short scanning time of MDCT may provide us an opportunity to improve contrast enhancement and to use the contrast medium more efficiently [9, 10]. With MDCT, the amount of contrast medium injected during some clinical applications may be reduced without decreasing contrast enhancement. Diverse protocols have been used for contrast administration and scan timing in cardiac and coronary CT angiography [8, 11 28]. Contrast-medium volumes ranging from 45 to 160 ml and injection rates ranging 2.5 to 5 ml/s have been used. This diversity in contrast-medium administration protocols reflects the rapid technical evolution of cardiac CT technology. As scanning duration shortens AJR:190, March 8 777

Bae et al. from 30 to 40 seconds with 4-MDCT to 6 12 seconds with 64-MDCT, there is a trend for using less contrast medium typically, 75 100 ml with fast MDCT for an averagesize adult [13, 15, 18]. As CT technology evolves, contrast medium injection protocols must be adjusted and optimized. A patient s body weight and the amount of contrast medium are closely related to the degree of contrast enhancement [10, 29 36]. When consistent contrast enhancement is desired, the amount of administered iodine should be adjusted according to the patient s body weight. A large patient needs more iodine than a small patient to achieve the same magnitude of enhancement. One simple scheme for adjusting the amount of iodine mass with the body weight is to use a simple linear scale for example, to double the iodine mass when the patient s body weight doubles. However, this simple body-weight-based linearity may not provide an accurate estimate of the required contrast medium dose, particularly in obese patients who have a lot of body fat that is not metabolically active and thus contributes little to dispersing or diluting the contrast medium in the blood. Linear-weightbased dosing may overestimate the amount of contrast medium needed in these patients. Other body size parameters such as height, body mass index (BMI), and body surface area (BSA) should be considered for optimal dosing of contrast medium. To our knowledge, however, the association of various body size parameters with the degree of contrast enhancement for a given iodine dose has not been previously addressed. Therefore, the purpose of our study was to evaluate the effect of body weight, height, BMI, BSA, and obesity on aortic contrast enhancement in 64-MDCT cardiac angiography. Materials and Methods Patients This study was approved by our institutional internal review board, and informed consent was obtained from each patient. Patients who had prior coronary artery bypass surgery, valve replacement surgery, atrial fibrillation, or contraindications to contrast-enhanced CT were excluded. The study group consisted of 73 consecutive patients (41 men and 32 women; mean age, 59.4 ± 11 [SD] years; age range, 30.9 80.9 years) who were scheduled to undergo catheter-based conventional coronary angiography and who agreed to participate in coronary CT angiography as a part of the research protocol. The patients body weight and height were recorded. CT All CT was performed on a 64-MDCT scanner (Sensation 64, Siemens Medical Solutions). The heart rate was measured in each patient before the CT examination. If the heart rate was higher than 65 beats per minute, IV β-blockers were administered to lower the heart rate. The scanning protocol consisted of 120 kvp, 750 850 effective mas, tube current modulation, 64 0.4 mm section collimation (z-flying focal spot), 330-millisecond gantry rotation time, and retrospective ECG gating. The images were reconstructed with a standard soft-tissue-kernel algorithm at 0.6-mm section thickness (without overlap) at every 10% of the R-R intervals. Contrast enhancement was achieved by administering mg I/mL contrast medium (Optiray [ioversol], Tyco Health/Mallinckrodt) into an antecubital vein through a 20-gauge angiocatheter. A fixed volume of 75 ml of con trast medium followed by a 40-mL saline flush was continuously injected at an injection rate of 4.5 ml/s for all patients. The scanning delay for diagnostic CT was determined using a bolus tracking method: A circular region of interest of 5 10 mm in diameter was placed over the ascend ing aorta, and diagnostic CT was started auto matically 5 seconds after the contrast enhancement exceeded a predefined threshold of 100 H. Aortic Attenuation Measurement The diagnostic cardiac CT images were retrieved from the institutional PACS and sent to a clinical workstation (Leonardo, Siemens Medical Solutions). Multiplanar reformatted images with 10-mm transverse section thickness were generated from the CT images reconstructed at 70% of the R-R interval of each cardiac cycle. Mean CT values (H) of the ascending aorta on these thick, transverse CT images were measured by a radiologist who had 12 years of experience in interpreting chest CT images. A 1.0-cm 2 circular region of interest was used to make measurements. Body Mass Index The BMI is defined as the weight in kilograms divided by the square of the height in meters. Although the BMI calculation does not take into account factors such as frame size and body tissue compositions, BMI categories are generally used as a means of estimating adiposity and assessing how much an individual s body weight departs from what is normal or desirable for a person of the same height. A BMI greater than or equal to 30 is generally considered to indicate obesity [37]; thus, in our study, for the evaluation of the effect of BMI, patients were divided into two groups for analysis: low BMI (BMI < 30) and high BMI (BMI 30). Body Surface Area BSA is a commonly used index in clinical practice to correct for patient size differences in various physiologic measurements and in calculating drug dosage. BSA is a better indicator of metabolic mass than body weight because it is less associated with excessive body fat. Various formulas have been proposed to estimate the BSA from a patient s weight and height, and these formulas result in slightly different values [38 43]. The most commonly used formula in day-to-day clinical practice is the Mosteller formula [42]: BSA (m 2 ) = (square root of product of weight [kg] height [cm]) / 60. This formula is simplified from a formula produced by Gehan and George [40] and has become a common standard because it is easy to memorize, and its use requires only a handheld calculator. More recently, however, a new formula was developed by Livingston and Lee [43] to relate BSA to body weight alone without using height data: BSA (m 2 ) = 0.1173 weight (kg) to the power of 0.6466. This scaling formula was reported to provide a more accurate estimate of the BSA in obese patients. Both the Mosteller and the scaling formulas were used in our study to calculate the BSA for each patient. Statistical Analysis The patient data for the following variables were fitted with normal distributions and tested for normality using the Shapiro-Wilk W tests: patient age, body weight, height, BMI, BSA, and aortic attenuation. Aortic attenuation values and ages for low- and high-bmi groups were also tested for normality and for equality of variances using the O Brien, Brown-Forsythe, Levene, and Bartlett tests. Differences in sex between the lowand high-bmi groups were tested using the Fisher s exact test. Means and 95% CIs were calculated for variables. To evaluate the effect of the patient s body size parameters on contrast enhancement, we performed linear regression analyses between aortic attenuation versus each of the following: body weight, height, BMI, and BSA. Ninety-five percent CIs were plotted for the regression lines. We used the Pearson product moment correlation coefficient (r) to assess the strengths of associations involving normal data distributions and the Spearman s rank (ρ) correlation coefficient to assess strengths of associations involving nonnormal data distributions. In addition, linear regression analyses were performed separ ately for the low- and high-bmi groups and 95% CIs were calculated for the resulting regression slopes. Differences in aortic attenuation between the lowand high-bmi groups were tested for statistical significance using the Student s t test. Alpha was set at 0.05. Statistical analyses were performed with 778 AJR:190, March 8

Contrast Enhancement in Cardiovascular MDCT patients were in the high-bmi group (BMI 30). The mean aortic attenuation was 325.3 ± 66.1 H (range, 167 488 H). Cardiac CT images from three patients (small, medium, and large body size) are presented in Figure 1 to illustrate the different degrees of aortic attenuation. JMP Statistical Software (version 6, SAS Institute) and StatXact 7 Statistical Software for Exact Nonparametric Inference (Cytel). Results Data Distributions: Age, Weight, Height, BMI, BSA, and Aortic Attenuation Data were normally distributed for age, height, BSA, and aortic attenuation (Shapiro Wilk W test, p > 0.05). Body weight and BMI were not normally distributed (Shapiro Wilk W test, p < 0.05). For the low- and high-bmi groups, aortic attenuation and body weight were normally distributed (Shapiro Wilk W test, p > 0.05). For the lowand high-bmi groups, aortic attenuation and body weight had equal variances (O Brien, Brown-Forsythe, Levene, and Bartlett tests, p > 0.05). No significant difference was seen in the patient age and sex between the lowand high-bmi groups (p 0.34). A wide range of patient ages (30.9 80.9 years), weights (50.0 154.0 kg; mean, 88.5 ± 22.0 kg), heights (151 192 cm; mean, 170.7 ± 9.5 cm), BMI (19.6 48.4; mean, 30.2 ± 6.1), and BSA (1.46 2.85 m 2 ; mean, 2.04 ± 0.29m 2 ) were observed. Forty-three patients were in the low-bmi group (BMI < 30) and 30 A C Fig. 1 Transverse CT images are shown for patients having three body sizes in whom different degrees of aortic attenuation were seen (image display window width, 800 H; center, H). A, 59-year-old woman weighing 55.8 kg and having body mass index (BMI) of 21.1 and body surface area (BSA) of 1.59 m 2 (aortic attenuation, 469 H). B, 61-year-old man weighing 91.4 kg and having BMI of 30.6 and BSA of 2.09 m 2 (aortic attenuation, 379 H). C, 59-year-old man weighing 154.0 kg and having BMI of 42.4 and BSA of 2.85 m 2 (aortic attenuation, 172 H). 40 50 60 70 80 B Effect of Body Weight, Height, BMI, and BSA on Aortic Attenuation Strong inverse correlation was noted between the aortic attenuation and body weight (r = 0.73, ρ = 0.74, p < 0.001), indicating reduced aortic attenuation in heavier patients (Fig. 2). The regression formula (aortic attenuation [H] = 520 2.2 weight [kg]) suggests that 1.0 ml/kg (e.g., 75 ml for a 75-kg patient) of mg I/mL contrast medium would yield the aortic attenuation of 355 H. A moderately strong inverse correlation was noted between aortic attenuation and height (r = 0.47, p < 0.001), which indicates reduced aortic attenuation in taller patients (Fig. 3). The regression formula (aortic attenuation [H] = 882 3.3 height [cm]) (p < 0.001) suggests that, for each 10-cm increase in height, there is a decrease in enhancement of 33 H in the aortic attenuation. A relatively strong inverse correlation existed between aortic attenuation and the BMI (r = 0.63, ρ = 0.64, p < 0.001). This indicates reduced aortic attenuation in patients with higher BMI (Fig. 4). The regression formula was aortic attenuation (H) = 529 6.8 BMI (p < 0.001). The strongest inverse correlation (r = 0.74, p < 0.001) existed between the aortic attenuation 90 100 110 120 130 140 160 Weight (kg) Fig. 2 Plot of aortic attenuation (H) versus body weight (kg). Strong inverse correlation existed between aortic attenuation and body weight (r = 0.73, ρ = 0.74, p < 0.001). This indicates aortic attenuation is reduced in heavier patients. Regression formula (aortic attenuation [H] = 520 2.2 weight [kg]) indicates that 1.0 ml/kg (e.g., 75 ml for a 75-kg patient) of mg I/mL contrast medium is required to achieve aortic attenuation of 355 H. + = low-body mass index (BMI) group (BMI < 30), o = high-bmi group (BMI 30). Ninety-five percent CIs (dotted fitting lines) are fit to regression line (solid fitting line). AJR:190, March 8 779

Bae et al. Fig. 3 Plot of aortic attenuation (H) versus height (cm). Moderately strong inverse correlation existed between aortic attenuation and body weight (r = 0.47, p < 0.001). This indicates aortic attenuation is reduced in taller patients. Regression formula (aortic attenuation [H] = 882 3.3 height [cm]) indicates that, for each 10-cm increase in height, there is decrease in enhancement of 33 H in aortic attenuation. + = low-body mass index (BMI) group (BMI < 30), o = high-bmi group (BMI 30). Ninety-five percent CIs (dotted fitting lines) are fit to regression line (solid fitting line). and the BSA (estimated with the Mosteller formula). This indicated reduced aortic attenuation with increased BSA (Fig. 5). The regression analysis was repeated with the BSA values, which were estimated using the scaling formula [43] from the body weight data only. The computed r value was nearly identical. The regression formula was aortic attenuation (H) = 674 171.4 BSA (m 2 ) (p < 0.001). This formula implies that, with our study injection protocol, the aortic attenuation of 382 H would be achieved for a patient with an average adult BSA of 1.7 m 2. The BSA that corresponds to 155 160 165 170 20 25 Height (cm) 175 180 185 190 195 the aortic attenuation of 355 H (i.e., the aortic attenuation achieved with 1.0 ml/kg of the weight-based dose) is 1.86 m 2. Comparison Between the Low- and High-BMI Groups The mean aortic attenuation for the low- BMI group (352.6 ± 59.1 H) was significantly higher (p < 0.001) than that for the high-bmi group (286.2 ± 55.5 H) (Fig. 6). This trend was anticipated from the strong inverse correlation observed in the aortic attenuation versus BMI, as shown in Figure 3. 30 35 40 45 50 Body Mass Index Fig. 4 Plot of aortic attenuation (H) versus body mass index (BMI). Relatively strong inverse correlation existed between aortic attenuation and BMI (r = 0.63, ρ = 0.64, p < 0.001). This indicates aortic attenuation is reduced in patients with higher BMI. Regression formula was aortic attenuation (H) = 529 6.8 BMI (p < 0.001). + = low-bmi group (BMI < 30), o = high-bmi group (BMI 30). Ninety-five percent CIs (dotted fitting lines) are fit to regression line (solid fitting line). For the low-bmi group, the regression formula for aortic attenuation on body weight was aortic attenuation (H) = 586 3.1 weight (kg) (r = 0.60, p < 0.001); for the high-bmi group, the regression formula was aortic attenuation (H) = 485 1.9 weight (kg) (r = 0.69, p < 0.001) (Fig. 7). Note that the regression slope (H/kg) of the high-bmi group (1.9 [2.6 1.1, 95% CI]) was less steep than that for the low-bmi group (3.1 [4.4 1.8]), but the 95% CIs overlap. The lower slope in the high-bmi group suggests a trend for a slower decline of contrast enhancement with increased body weight and for less contrast medium to be required per kilogram of body weight to maintain contrast enhancement. The regression formula for aortic attenuation on BSA was aortic attenuation (H) = 706 187 BSA (m 2 ) (r = 0.58, p < 0.001) for the low- BMI and aortic attenuation (H) = 627 151.0 BSA (m 2 ) (r = 0.71, p < 0.001) for the high- BMI group (Fig. 8). Note that the regression slope (H/m 2 ) of the high-bmi group (151) was less steep than that for the low-bmi group (187), just as was the trend observed in the regression of the aortic attenuation on body weight. However, the relative ratio of the slopes for the high- to low-bmi groups for the BSA regression is smaller (151/187 [81%]) than that for the body weight regression (1.9/3.1 [61%]). This suggests that BSA may serve better than body weight for the adjustment of iodine dose over a wide range of body sizes. Discussion The magnitude of contrast enhancement for CTA depends on a number of patientrelated and injection-related factors, including body weight and cardiac output, contrast medium volume and concentration, injection rate, type of contrast medium, and saline flush [10]. The most important patient-related factor affecting the magnitude of vascular and parenchymal contrast enhancement is body weight [30, 34, 35, 44]. The significance of body weight on contrast enhancement was confirmed by our study, which showed a strong trend of declining aortic attenuation proportional to patient weight for a given contrast medium administration protocol. To achieve a consistent degree of contrast enhancement, the amount of administered contrast medium should be adjusted according to the patient s body weight. To devise a contrast medium dosing protocol, we first must determine the desirable level of contrast enhancement for aortic and coronary CT angiography. We consider attenuations of H 780 AJR:190, March 8

Contrast Enhancement in Cardiovascular MDCT 1.4 1.6 Fig. 5 Plot of aortic attenuation (H) versus body surface area (BSA) (m 2 ). Strongest inverse correlation (r = 0.74, p < 0.001) existed between aortic attenuation and BSA (estimated with Mosteller formula [42]). This indicates aortic attenuation is reduced as BSA increases. Regression formula was aortic attenuation (H) = 674 171.4 BSA (m 2 ) (p < 0.001). + = low body mass index (BMI) group (BMI < 30), o = high-bmi group (BMI 30), Ninety-five percent CIs (dotted fitting lines) are fit to regression line (solid fitting line). (i.e., H net contrast enhancement) to be diagnostically adequate for aortic and coronary CT angiography [17]. Although slightly lower attenuations ( H) were advocated for coronary CT angiography to differentiate the enhanced lumen from calcification (> H) [24], higher vascular attenuations (> H) were reported as necessary to improve visualization of small coronary arteries and stenoses [17]. On the basis of our results, we think that an aortic attenuation of 355 H (i.e., approximately H net contrast enhancement) would be achieved with 1.0 ml/kg 1.8 2 2.2 2.4 2.6 2.8 3 Body Surface Area (m 2 ) of mg I/mL contrast medium injected at 4.5 ml/s (i.e., 0.35 g I/kg of contrast medium injected at 1.6 g I/s), followed by a saline flush. Using this weight-based protocol, we are be able to estimate the amount of required iodine loads for large patients (i.e., use of higher volumes or longer injections) or for small patients to generate equivalent degrees of contrast enhancement. Alternatively, we can modify iodine delivery rates (i.e., use different injections or different iodine concentrations in contrast mediums) to achieve different degrees of contrast enhancement. < 30 30 Body Mass Index Fig. 6 Ninety-five percent mean diamond plots of aortic attenuation (H) for low- and high-body mass index (BMI) groups. Mean aortic attenuation of low-bmi group (352.6 ± 59.1 H) was significantly higher (p < 0.001) than that of high-bmi group (286.2 ± 55.5 H). + = low-bmi group (BMI < 30), o = high-bmi group (BMI 30). Horizontal line is grand mean. Heights of diamonds represent 95% CIs and widths of diamonds are proportional to sample sizes. Although numerous studies have been conducted to investigate the effect of body weight on contrast enhancement [9, 30, 31, 34, 35, 45 48], the effect of the patient s height on contrast enhancement has rarely been studied. Our study revealed a moderately strong inverse correlation between aortic attenuation and height. This indicates reduced aortic attenuation in taller patients. Although the correlation is weaker than that for the body weight, this inverse relationship is expected because height and weight are proportional (i.e., taller people are generally heavier than shorter people). This is supported by a recent study that showed that for weights less than 80 kg, height and weight are highly correlated and both increase proportionally: height (cm) = 33.34 weight (kg) to the power of 0.3922 (r = 0.99) [43]. This study also reported that the height weight relationship is complex because beyond 80 kg, weight increases without any significant increase in height, as commonly seen in patients with obesity. Considering this nonlinear relationship and individual height variation in adult patients being less than body weight variation, it is not surprising that the correlation of aortic attenuation with height was weaker than that with body weight. Nevertheless, to precisely determine the required volume of contrast medium for a consistent enhancement, not only body weight but also height and obesity should be taken into consideration. Although BMI is commonly used because it is easy to calculate and is strongly correlated with total body fat content in adults, its accuracy in relation to actual levels of body fat is easily distorted by such factors as fitness level, muscle mass, bone structure, sex, and ethnicity. Muscular people have high BMIs because muscle is denser than fat. Despite its limitations, the BMI has been widely applied for the identification, evaluation, and treatment of obesity in adults [37]. Our study result show a relatively strong inverse correlation between aortic attenuation and BMI. The mean aortic attenuation was significantly lower in high- rather than low-bmi patients. These findings intuitively make sense because contrast enhancement is expected to be lower in obese patients than in nonobese patients. However, the findings are less obvious from the definition of BMI (i.e., weight divided by the square of height) because both weight and height showed an inverse correlation with aortic attenuation. Our explanation for this is that although the square of the height constitutes the denominator for the calculation of the BMI, the contribution of AJR:190, March 8 781

Bae et al. 40 50 Fig. 7 Plot of aortic attenuation (H) versus body weight (kg) with bivariate fits for low- and high-body mass index (BMI) groups. Regression formula of aortic attenuation versus body weight was aortic attenuation (H) = 586 3.1 weight (kg) (p < 0.001) for low-bmi group (+, solid fitting line) and aortic attenuation (H) = 485 1.9 weight (kg) (p < 0.001) for high-bmi group (o, dotted fitting line). Regression slope (H/kg) of high-bmi group (1.9 [2.6 1.1, 95% CI]) was less steep than that for low-bmi group (3.1 [4.4 1.8]), but 95% CIs overlap. So that regression lines can be better seen, 95% CIs are not included in figure. weight is greater than that of height. This is because height is proportional to the 0.3922 power of weight for weights < 80 kg [43]. The BMI still has a net positive correlation with body weight and thus has an inverse correlation with the aortic attenuation. The correlation of aortic attenuation with BSA was higher than those with body weight, height, and BMI. Although the relationship between contrast enhancement and BSA has not been addressed in previous studies, this strong relationship is not surprising because 1.4 1.6 1.8 60 70 80 90 100 110 120 130 140 160 Weight (kg) for many physiologic and clinical purposes, BSA has been used as a better indicator of metabolic mass than body weight or other indicators. In our study, we found no difference between the BSA values estimated with the Mosteller [42] and the scaling [43] formulas for the correlation with aortic attenuation. When we consider contrast medium as an IV drug that distributes and disperses pharmacokinetically throughout the body, it is conceivable that contrast medium could be dosed according to the patient s BSA rather 2 2.2 2.4 2.6 2.8 3 Body Surface Area (m 2 ) Fig. 8 Plot of aortic attenuation (H) versus body surface area (BSA) (m 2 ) with bivariate fit of low- and high- body mass index (BMI) groups. Regression formulas of aortic attenuation versus BSA were aortic attenuation (H) = 706 187 BSA (m 2 ) (p < 0.001) for low-bmi group (+, solid fitting line) and aortic attenuation (H) = 627 151.0 BSA (m 2 ) (p < 0.001) for high-bmi group (o, dotted fitting line). Note that these two regression lines are more closely approximated than two regression lines in Figure 7. This suggests that discrepancy between high- and low-bmi groups in decline rate of aortic attenuation was less pronounced with BSA than with body weight. than other body size parameters. With a direct 1:1 weight-based linear proportionality, contrast medium may be overdosed for obese patients and underdosed in pediatric patients. Support for a BSA-based contrast medium dosing scheme is as follows. In our study, the aortic attenuation of 355 H was obtained with 75 ml of contrast medium for a patient with a BSA of 1.86 m 2. This 75 ml of contrast medium normalized by the BSA corresponds to 40.3 ml/m 2. In our study population, BSA ranged from 1.45 to 2.85 m 2. When these are multiplied by 40.3 ml/m 2, the volumes of contrast medium corresponding to the aortic attenuation of 355 H equal 58.7 ml for the minimum BSA and 115.0 ml for the maximum BSA. This estimation based on BSA can be measured against that based on body weight as follows: It is assumed that through linear proportionality, a weight-based dosing scheme can be used to estimate the amount of contrast volume required to achieve a certain degree of enhancement (i.e., 1.0 ml/kg of mg I/mL contrast medium results in an aortic attenuation of 355 H). In our study population, the body weight ranged from 50 to 154 kg; thus, the estimated volumes of contrast medium would be 50 and 154 ml. When these are compared with the volumes estimated with BSA, the lightest patient would be underdosed by 8.7 ml, and the heaviest patient overdosed by 39 ml. Instead of linear proportionality, because the BSA is directly related to weight (kg) to the power of 0.6466 according to the scaling formula [43], weight scaling can be used as a proportionality to estimate the contrast medium dose. If 75 kg is used as the reference weight and 1.0 ml/(kg) to the power of 0.6466 as the proportionality, the amount of contrast medium required for a 50-kg patient is 57.5 ml (i.e., 75 ml [50 / 75] to the power of 0.6466) and that required for a 154-kg patient is 119.4 ml (i.e., 75 ml [154 / 75] to the power of 0.6466). Not surprisingly, these values are similar to the 58.7 and 115.0 ml that were estimated from the BSA. For obese patients, the relationship between iodine dose and the degree of contrast enhancement is not well established. An obese patient may have a high proportion of body fat and thus a relatively small blood volume and a small, well-perfused extracellular compartment. As a result, in obese patients, if the amount of iodine dose is estimated and increased linearly in proportion to body weight, the resulting contrast enhancement may be 782 AJR:190, March 8

Contrast Enhancement in Cardiovascular MDCT higher than that of nonobese patients, who receive the iodine dose with the same linear body weight proportionality. We propose a contrast medium dosing scheme for which a reference dose is used for a standard body size to achieve a clinically desirable degree of contrast enhancement, and subsequent contrast-medium doses are adjusted using the two-thirds power of the weight, as suggested in the BSA scaling formula [43]. This scaling rule of adjusting contrast medium dose with the two-thirds power of the weight is applicable not only across a wide range of human body sizes (from infants to a markedly obese adults) but also across species ranging in size from rats to cows [49]. However, further clinical studies are necessary to validate this. Because the blood volume per body weight is relatively smaller in obese patients, the rate of declining aortic attenuation with body weight in obese patients (BMI 30) is expected to be slower than that in nonobese patients (BMI < 30). This was confirmed in our results that the regression slope of the high-bmi patients was less than that for the low-bmi patients. In other words, for patients with high BMIs compared with those with low BMIs, a smaller decrease in contrast enhancement occurred with increased body weight (i.e., less contrast medium was required per body weight to maintain contrast enhancement). Furthermore, we observed that, with BSA instead of weight as the regression variable, the regression slope discrepancy between the high- and low-bmi patients was less pronounced. This implies that BSA is a better parameter than body weight for estimating contrast medium dose for both the high- and low-bmi patients. Our study had some limitations. First, it was a retrospective study and was not specifically designed to test the clinical impact of different contrast enhancements on the accuracy of diagnosing coronary artery disease or stenosis. No follow-up or further clinical evaluation was performed to investigate the extent to which the image (diagnostic quality) was associated with the degree of contrast enhancement. Second, we used BMI to divide the patients into obese and nonobese groups. Although the BMI is easy to use and readily calculated from the weight and height data, it is limited in characterizing obesity because it does not distinguish between muscle and fat. Percentage of body fat is a better measure to quantify adipose tissue; however, it requires techniques such as skin-fold measurements and specialized body fat scales, which were not available in our study. Third, although aortic attenuation is affected by cardiovascular function and cardiac output [10, 50], this was not specifically investigated in our study. We recognize that because of variations in cardiac output among patients, it is critical to individualize the scanning delay for CT angiography. We used a bolus tracking method to determine the scanning delay. We think that the effect of cardiac output variation on aortic attenuation is minimized by means of an individualized scanning delay and use of a fixed injection duration. Also, the attenuation value we measured does not necessarily represent the peak attenuation value but rather attenuation values obtained with diagnostic scans using a specific injection protocol. In our clinical diagnostic scan of 5- to 7-seconds duration, we could neither explore the temporal course of the enhancement nor determine peak enhancement. Fourth, our CT attenuation measurements were limited to the ascending aorta. Although the attenuation or contrast enhancement of the coronary arteries is clinically valuable, we think that the ascending-aortic attenuation measurement is highly reliable and representative of the coronary artery attenuation measurements. The measurement of CT attenuation in small arteries is technically difficult and highly variable and is affected by the severity of vascular disease. In conclusion, on the basis of our specific contrast injection protocol and study results, we found that the degree of aortic contrast enhancement in MDCT angiography declines with increases in body weight, height, body mass index, and body surface area. This finding may be interpreted and used in clinical practice as follows: To achieve a consistent aortic and coronary artery enhancement, the amount of contrast medium should be adjusted according to the patient s body weight or body surface area. Body surface area accounts for both the body weight and height factors and perhaps is more appropriate for obese patients because a 1:1 direct increase in contrast volume with body weight may overestimate the contrast medium dose needed in obese patients. We propose that one approach to adjust the amount of administered contrast medium with respect to the patient s body surface area is to estimate the required amount of contrast medium proportional to the two-thirds power of the body weight. This proposed method would require further prospective investigative work for its validation. References 1. Ohnesorge BM, Hofmann LK, Flohr TG, Schoepf UJ. CT for imaging coronary artery disease: defining the paradigm for its application. Int J Cardiovasc Imaging 5; 21:85 104 2. Schoepf UJ, Becker CR, Ohnesorge BM, Yucel EK. CT of coronary artery disease. Radiology 4; 232:18 37 3. Achenbach S, Ropers D, Kuettner A, et al. Contrast-enhanced coronary artery visualization by dual-source computed tomography: initial experience. Eur J Radiol 6; 57:331 335 4. Hoffmann MH, Lessick J. Multidetector-row computed tomography for noninvasive coronary imaging. Expert Rev Cardiovasc Ther 6; 4:583 594 5. Cury RC, Nieman K, Shapiro MD, Nasir K, Cury RC, Brady TJ. Comprehensive cardiac CT study: evaluation of coronary arteries, left ventricular function, and myocardial perfusion is it possible? J Nucl Cardiol 7; 14:229 243 6. Herzog C, Zangos S, Zwerner P, Costello P, Vogl TJ, Schoepf UJ. CT of coronary artery disease. J Thorac Imaging 7; 22:40 48 7. Bae KT. Test-bolus versus bolus-tracking techniques for CT angiographic timing. Radiology 5; 236:369 370; author reply 370 8. Cademartiri F, Nieman K, van der Lugt A, et al. Intravenous contrast material administration at 16-detector row helical CT coronary angiography: test bolus versus bolus-tracking technique. Radiology 4; 233:817 823 9. Bae KT, Tao C, Gurel S, et al. Effect of patient weight and scanning duration on contrast enhancement during pulmonary multidetector CT angiography. Radiology 7; 242:582 589 10. Bae KT, Heiken JP. Scan and contrast administration principles for MDCT. Eur Radiol 5; 15[suppl 5]:E46 E59 11. Funabashi N, Suzuki K, Terao M, et al. New acquisition method to exclusively enhance the left side of the heart by a small amount of contrast material achieved by multislice computed tomography with 64 data acquisition system. Int J Cardiol 7; 114:265 269 12. Rist C, Nikolaou K, Kirchin MA, et al. Contrast bolus optimization for cardiac 16-slice computed tomography: comparison of contrast medium formulations containing and milligrams of iodine per milliliter. Invest Radiol 6; 41:460 467 13. Hamoir XL, Flohr T, Hamoir V, et al. Coronary arteries: assessment of image quality and optimal reconstruction window in retrospective ECG-gated multislice CT at 375-ms gantry rotation time. Eur Radiol 5; 15:296 304 14. Haberl R, Tittus J, Bohme E, et al. Multislice spiral computed tomographic angiography of coronary arteries in patients with suspected coronary artery disease: an effective filter before catheter AJR:190, March 8 783

Bae et al. angiography? Am Heart J 5; 149:1112 1119 15. Leber AW, Knez A, von Ziegler F, et al. Quantification of obstructive and nonobstructive coronary lesions by 64-slice computed tomography: a comparative study with quantitative coronary angiography and intravascular ultrasound. J Am Coll Cardiol 5; 46:147 154 16. Datta J, White CS, Gilkeson RC, et al. Anomalous coronary arteries in adults: depiction at multi-detector row CT angiography. Radiology 5; 235:812 818 17. Cademartiri F, Mollet NR, van der Lugt A, et al. Intravenous contrast material administration at helical 16-detector row CT coronary angiography: effect of iodine concentration on vascular attenuation. Radiology 5; 236:661 665 18. Achenbach S, Ropers D, Pohle FK, et al. Detection of coronary artery stenoses using multi-detector CT with 16 0.75 collimation and 375 ms rotation. Eur Heart J 5; 26:1978 1986 19. Nikolaou K, Knez A, Sagmeister S, et al. Assessment of myocardial infarctions using multidetector-row computed tomography. J Comput Assist Tomogr 4; 28:286 292 20. van Ooijen PM, Dorgelo J, Zijlstra F, Oudkerk M. Detection, visualization and evaluation of anomalous coronary anatomy on 16-slice multidetectorrow CT. Eur Radiol 4; 14:2163 2171 21. Kuettner A, Trabold T, Schroeder S, et al. Noninvasive detection of coronary lesions using 16- detector multislice spiral computed tomography technology: initial clinical results. J Am Coll Cardiol 4; 44:1230 1237 22. Kopp AF, Kuttner A, Trabold T, Heuschmid M, Schroder S, Claussen CD. Multislice CT in cardiac and coronary angiography. Br J Radiol 4; 77[spec no 1]:S87 S97 23. Hong C, Chrysant GS, Woodard PK, Bae KT. Coronary artery stent patency assessed with instent contrast enhancement measured at multidetector row CT angiography: initial experience. Radiology 4; 233:286 291 24. Becker CR, Hong C, Knez A, et al. Optimal contrast application for cardiac 4-detector-row computed tomography. Invest Radiol 3; 38:690 694 25. Pannu HK, Flohr TG, Corl FM, Fishman EK. Current concepts in multi-detector row CT evaluation of the coronary arteries: principles, techniques, and anatomy. RadioGraphics 3; 23[spec no]:s111 S125 26. Vogl TJ, Abolmaali ND, Diebold T, et al. Techniques for the detection of coronary atherosclerosis: multi-detector row CT coronary angiography. Radiology 2; 223:212 220 27. Nieman K, Cademartiri F, Lemos PA, Raaijmakers R, Pattynama PM, de Feyter PJ. Reliable noninvasive coronary angiography with fast submillimeter multislice spiral computed tomography. Circulation 2; 106:2051 2054 28. Kopp AF, Schroeder S, Kuettner A, et al. Non-invasive coronary angiography with high resolution multidetector-row computed tomography: results in 102 patients. Eur Heart J 2; 23:1714 1725 29. Dean PB, Violante MR, Mahoney JA. Hepatic CT contrast enhancement: effect of dose, duration of infusion, and time elapsed following infusion. Invest Radiol 1980; 15:158 161 30. Kormano M, Partanen K, Soimakallio S, Kivimaki T. Dynamic contrast enhancement of the upper abdomen: effect of contrast medium and body weight. Invest Radiol 1983; 18:364 367 31. Berland LL, Lee JY. Comparison of contrast media injection rates and volumes for hepatic dynamic incremented computed tomography. Invest Radiol 1988; 23:918 922 32. Chambers TP, Baron RL, Lush RM. Hepatic CT enhancement. Part I. Alterations in the volume of contrast material within the same patients. Radiology 1994; 193:513 517 33. 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 34. Heiken JP, Brink JA, McClennan BL, Sagel SS, Crowe TM, Gaines MV. Dynamic incremental CT: effect of volume and concentration of contrast material and patient weight on hepatic enhancement. Radiology 1995; 195:353 357 35. Platt JF, Reige KA, Ellis JH. Aortic enhancement during abdominal CT angiography: correlation with test injections, flow rates, and patient demographics. AJR 1999; 172:53 56 36. Bae KT. Peak contrast enhancement in CT and MR angiography: when does it occur and why? Pharmacokinetic study in a porcine model. Radiology 3; 227:809 816 37. NHLBI Obesity Education Initiative Expert Panel. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: the evidence report. Bethesda, MD: National Institutes of Health, National Heart, Lung, and Blood Institute, 1998 38. DuBois D, DuBois EF. A formula to estimate the approximate surface area if height and weight are known. Arch Intern Med 1916; 17:863 871 39. Boyd E. The growth of the surface area of the human body. Minneapolis, MN: University of Minnesota, 1935 40. Gehan EA, George SL. Estimation of human body surface area from height and weight. Cancer Chemother Rep 1970; 54:225 235 41. Haycock GB, Schwartz GJ, Wisotsky DH. Geometric method for measuring body surface area: a height weight formula validated in infants, children, and adults. J Pediatr 1978; 93:62 66 42. Mosteller RD. Simplified calculation of bodysurface area. N Engl J Med 1987; 317:1098 43. Livingston EH, Lee S. Body surface area prediction in normal-weight and obese patients. Am J Physiol Endocrinol Metab 1; 281: E586 E591 44. Bae KT. Technical aspects of contrast delivery in advanced CT. Appl Radiol 3; 32[suppl]:12 19 45. Bae KT, Heiken JP, Brink JA. Aortic and hepatic contrast medium enhancement at CT. Part I. Prediction with a computer model. Radiology 1998; 207:647 655 46. Yamashita Y, Komohara Y, Takahashi M, et al. Abdominal helical CT: evaluation of optimal doses of intravenous contrast material prospective randomized study. Radiology 0; 216:718 723 47. Takeshita K. Prediction of maximum hepatic enhancement on computed tomography from dose of contrast material and patient weight: proposal of a new formula and evaluation of its accuracy. Radiat Med 1; 19:75 79 48. Awai K, Hori S. Effect of contrast injection protocol with dose tailored to patient weight and fixed injection duration on aortic and hepatic enhancement at multidetector-row helical CT. Eur Radiol 3; 13:2155 2160 49. Dawson TH. Engineering design of the cardiovascular systems of mammals. Englewood Cliffs, NJ: Prentice Hall, 1991 50. Bae KT, Heiken JP, Brink JA. Aortic and hepatic contrast medium enhancement at CT. Part II. Effect of reduced cardiac output in a porcine model. Radiology 1998; 207:657 662 784 AJR:190, March 8