Assessment of the Use of a Saline Chaser to Reduce the Volume of Contrast Medium in Abdominal CT

Transcription

1 Saline Chaser in bdominal CT bdominal Imaging Original Research C D E M N E U T R Y L I M C I G O F I N G Florence Orlandini Stéphanie Boini Sandrine Iochum-Duchamps Toufik Batch Xavier Zhu lain Blum Orlandini F, Boini S, Iochum-Duchamps S, Batch T, Zhu X, Blum Keywords: abdominal CT, aorta, contrast media, liver, saline chaser DOI:0.4/JR Received June 5, 004; accepted after revision pril6,005. Service d Imagerie Guilloz, University Hospital of Nancy, 9, ave. Marechal de Lattre de Tassigny, Nancy 54 05, France. ddress correspondence to. Blum (a.blum@chu-nancy.fr). Service d Epidémiologie et évaluation cliniques, Centre d Epidémiologie Clinique, Hôpital Marin, University Hospital of Nancy, Nancy 54 05, France. JR 006; 87: X/06/87 5 merican Roentgen Ray Society ssessment of the Use of a Saline Chaser to Reduce the Volume of Contrast Medium in bdominal CT OBJECTIVE. The purpose of this study was to investigate, first, the effect on enhancement of the liver and aorta during abdominal CT of the use of saline solution as a partial substitute for or in addition to contrast medium when the dose of medium is determined by the patient s body weight and, second, whether use of a saline chaser allows a decrease in the dose of contrast medium to less than.5 ml/kg. SUBJECTS ND METHODS. We enrolled 407 patients undergoing abdominal exploration on MDCT, including an early arterial phase and a portal phase. received contrast medium at a dose of.5 ml/kg. received contrast medium at a dose of.5 ml/kg less 0 ml of iodine followed by 0 ml of saline solution. received the full.5 ml/kg dose of medium followed by a 0-mL saline chaser. ttenuation values were obtained from the aorta in the arterial phase and from the liver in the portal phase. RESULTS. The groups were comparable in mean body weight and heart rate. None of the differences between them in aortic enhancement in the early arterial phase were statistically significant (group, 06 ± H; group, 04 ± H; group, 09 ± 4 H). There was a negative correlation between weight and aortic enhancement (r = 0.4, p < 0.000) and a positive correlation between weight and hepatic enhancement (r = 0.9, p < 0.000). significant reduction (p = 0.000) in hepatic enhancement was observed in group (group, 5 H; group, 46 H; group, 54 H). Hepatic enhancement greater than 45 H was observed in all groups except for the group subgroup of patients weighing less than 55 kg. CONCLUSION. In MDCT, reducing the amount of contrast medium does not affect aortic enhancement in the early arterial phase but decreases hepatic enhancement at the portal phase. The saline chaser technique had no influence on the phases studied. Greater than.5 ml/kg reduction in the dose of contrast medium followed by saline flushing may not be advisable for liver CT in low- and medium-weight patients. bdominal CT is crucial in many clinical situations, such as detection of metastatic lesions of the liver and hypervascular liver neoplasms. This technique, however, has the disadvantages of exposing the patient to radiation and requiring administration of iodinated contrast medium. Many studies have addressed the question of how best to improve the image quality and diagnostic accuracy of CT. Imaging overkill may be tempting under certain circumstances but cannot be considered a serious option because of the risk of contrast-induced nephropathy, particularly among elderly patients. In addition to increasing safety, use of low doses of contrast medium may result in considerable cost savings. In a 00 study, Dorio et al. [] found that use of a saline chaser (also known as saline flushing) may allow a reduction in the amount of contrast medium needed in abdominal imaging examinations. Those investigators compared two fixed-dose injection protocols, one with 50 ml of contrast material and one with 00 ml of contrast material followed by a 50-mL saline chaser. Despite the reduced volume of contrast medium, there were no clinically significant differences in hepatic enhancement between the two protocols. dministration of a fixed amount of contrast medium can have limitations. volume insufficient for enhancement of the liver in a large patient may cause contrast medium induced nephropathy in someone thin and frail. It has therefore become usual practice to consider the patient s weight in decisions about the amount of medium to use [ 4]. The objective of this study was to explore the effects of using a saline JR:87, ugust 006 5

2 chaser in abdominal CT when the dose of contrast medium is based on the patient s weight. nother objective was to evaluate the use of saline solution when doses of contrast medium are reduced to less than.5 ml/kg. Subjects and Methods Patient Population and Protocol Four hundred seven patients (08 men and 99 women) consecutively referred for abdominal CT to include liver and vascular studies were enrolled to receive IV contrast medium according to one of the following three protocols (all injected agents were delivered at a rate of ml/s): group (n = 5), iobitridol (Xenetix 50, Guerbet).5 ml/kg; group (n = 5), iobitridol.5 ml/kg with 0 ml (7 g) of iodine replaced by 0 ml of saline chaser; group (n = ), iobitridol.5 ml/kg followed by 0 ml of saline chaser. The study was conducted in three - week periods during which each of the test protocols was used. Subjects in each group were subdivided into three categories by body weight: less than 55 kg, kg, and 80 kg or more. Twenty patients who for medical reasons underwent a repeat examination or weeks after the first were included in two of the groups. Exclusion criteria were age less than 8 years, known or suspected hypersensitivity to iodinated contrast medium, severe renal impairment (serum creatinine value > 0 µmol/l), liver cirrhosis, and compromised venous access. Image cquisition Contrast-enhanced MDCT (Sensation 6, Siemens Medical Solutions) was performed. Each examination included one acquisition of unenhanced images from the hepatic dome to the pelvis with the following parameters: detector row configuration, 6.5 mm; pitch, (table feed per gantry rotation, 4 mm/s); slice thickness, 8 mm at 0-mm intervals; 0 kv; and product of tube current and exposure time in milliampere-seconds (ms), equal to the patient s body weight. Early arterial phase acquisition with an automated bolus-tracking device (Carebolus, Siemens) was initialized when enhancement in the aorta exceeded 00 H. cquisition lasted approximately 6 seconds and was performed with a 6.5 mm detector row configuration and a pitch of (table feed per gantry rotation, 4 mm/s; gantry rotation time, 0.5 seconds). Images were reconstructed with a 5- mm slice thickness at 7-mm intervals. Forty seconds after the end of the first acquisition, data were acquired for the entire abdomen in the portal phase mm or 6.5 mm detector row configuration was chosen with a pitch of. Slice thickness was 5 mm at 7-mm intervals. Other slices were obtained for examination and interpretation elsewhere. In all cases, 0 kv was used; the product of tube current and exposure time in milliampere-seconds was equal to two to three times the body weight. Contrast medium was injected through an 8- gauge IV catheter into an antecubital vein. t the beginning of the study, the saline chaser technique required two interconnected power injectors, each holding a 00-mL syringe as described by Haage and colleagues [5]. The two long tubes on the injectors were connected in line with a nonreturn valve and then with a Y-adapter leading to the IV catheter. To save time, dual-head injectors (Stellant, Medrad) were used when they became available toward the end of the study. Reports in the literature indicate the volume of chaser saline solution used varies from 0 to 50 ml, and there is no consensus on the optimal amount. Bae et al. [6] estimated that the dead space caused by the brachial vein is approximately 40 ml in a 70-kg patient. Because we believe dead space is minimally influenced by the patient s weight, a fixed dose of 0 ml was adopted to simplify the procedure and make it reproducible. Data cquisition For each of the 407 scans, one radiologist measured arterial and hepatic enhancement after identification of consistent regions of interest with areas ranging from 5 to 0 mm. Measurements first were obtained from unenhanced images of the aorta and liver. In the arterial phase, circular regions of interest were identified at the level of the celiac trunk, the superior mesenteric artery, and the iliac bifurcation. In the portal phase, measurements free of artifacts and enhanced vessels were obtained in the hepatic parenchyma and the aorta. Statistical nalysis Comparability between groups was assessed with tests based on the chi-square test and Student s t test for categorical and continuous variables, respectively (level of type error, 5%). Quantitative data obtained with the various protocols were compared through one-way analysis of variance. Bonferroni adjustment was used for multiple comparisons. difference was considered statistically significant at p < Pearson s correlation coefficient was calculated to test the hypothesis that aortic enhancement and hepatic enhancement are related to patient weight. nalysis of variance was used to determine the effect of weight (< 55 kg, kg, and 80 kg) on aortic and hepatic enhancement. Statistical analysis was performed with SS system software (release 8., SS Institute [7]). Results Characteristics of the patient population are shown in Table. lthough they had different numbers of patients, the groups were TBLE : Patient Population Characteristic No. of patients 5 5 ge (y) Mean a Range % CI Body weight (kg) Mean Range % CI Heart rate (beats/min) Mean Range % CI Sex Male a Female a Note No significant differences in weight or heart rate were observed between groups. CI = confidence interval. a Compared with the other two groups, group patients were predominantly female (p = 0.0) and older (p = 0.0). comparable in terms of mean weight and heart rate. included a greater proportion of women than did groups and (p = 0.0), and the mean age was greater (p = 0.0). The findings for age and sex were adjusted to account for these differences. Contrast enhancement data for the three groups are summarized in Tables 5 and illustrated in Figures and. The magnitudes of observed enhancement were comparable in the aorta during the early arterial phase: group, 06 H; group, 04 H; and group, 09 H (Table ). Differences between groups did not reach statistical significance. negative correlation was observed between patient weight and aortic enhancement (r = 0.4, p < 0.000) (Table ). verage hepatic enhancement in the three groups was 5 H, 46 H, and 54 H. statistically significant reduction in hepatic enhancement was observed in group compared with groups and (6.8 H, p = 0.00; 7.8 H, p < 0.00). When doses of contrast medium were similar (groups and ), hepatic enhancement was unaffected by the saline chaser (Table 4). Patient weight and hepatic enhancement were positively correlated (r = 0.9, p < 0.000). Hepatic enhancement was greater than 45 H in all patient categories 5 JR:87, ugust 006

3 Saline Chaser in bdominal CT TBLE : ortic Enhancement (H) in the Early rterial Phase Value Mean aortic enhancement 95% CI Note None of the differences observed was statistically significant. CI = confidence interval. TBLE : ortic Enhancement (H) in the Early rterial Phase by Weight Class Patient Weight (kg) ll s < ± 7 ± 9 40 ± 8 8 ± ± 4 09 ± 4 ± 5 ± ± 6 7 ± 7 84 ± 7 78 ± 4 Note Values are mean ± SD. nalysis of variance showed all groups data had significant differences between weight categories (p < 0.000). TBLE 4: Hepatic Enhancement (H) in the Portal Phase Value Mean hepatic enhancement 95% CI Note No difference was observed between groups receiving.5 ml of iodine, regardless of use of saline flush. CI = confidence interval. TBLE 5: Hepatic Enhancement (H) in the Portal Phase by Weight Class Patient Weight (kg) ll s < ± 4 8 ± 4 47 ± 44 ± ± 5 45 ± 8 5 ± 50 ± ± 5 ± 60 ± 55 ± Note Values are mean ± SD. nalysis of variance showed hepatic enhancement superior to 45 H was observed in all groups other than the low-weight subgroup of group. other than group subjects weighing less than 55 kg. The only significant difference in hepatic enhancement was an increase among group patients weighing kg compared with the same weight category in group (p = 0.0) (Table 5). Discussion The advent of MDCT allows increased z- axis coverage with shorter scanning duration. Improved understanding of the phases of enhancement necessitates refinement of injection acquisition techniques [8]. shorter duration of injection implies a need for lower doses of contrast medium and consequent reductions in cost and risk of nephropathy [9, 0]. The total amount of contrast medium, however, must be sufficient to produce acceptable aortic and hepatic enhancement throughout acquisition. Consensus has not been reached about the ideal contrast agent or the optimal dose and rate of injection. Issues concerning the choice of monophasic or biphasic injection and scan delay after injection also remain to be fully addressed. Some authors have advocated use of a saline flush to improve vessel and parenchymal enhancement achieved with low doses of contrast medium []. Hopper et al. [] looked at flushing techniques as a way of reducing the dose of contrast medium in thoracic CT. Their hypothesis was that medium remaining in the brachiocephalic vein and the right-heart cavities provided few data and that saline pushing would reduce the transit time of the column of medium and decrease the amount of iodine in these so-called dead spaces. 75-mL dose of contrast medium pushed with saline solution was shown to enhance the thoracic vasculature to a degree equivalent to that obtained with 5 ml of contrast material alone and to cause significantly fewer beam-hardening artifacts. To the best of our knowledge, there have been few studies of the value of saline flushing in abdominal CT. Dorio and colleagues [] evaluated 86 hypoattenuating metastatic lesions of the liver in 6 patients who underwent CT in two sessions separated by a mean of 85 days. t the first session, 50 ml of contrast material (iohexol; Omnipaque 00, Nycomed) was administered at a rate varying from.0 to.0 ml/s depending on venous access. t the second session, 00 ml of contrast medium was followed by a 50-mL saline chaser. Liver and tumor attenuation was slightly better with the 50-mL dose, but the difference was of questionable clinical significance (95 H vs 89 H for the liver). Schoellnast and colleagues [] conducted a study in which 4 patients (average weight, 65 kg) with malignant tumors or inflammatory disease underwent two successive sessions of abdominal MDCT. The subjects received 00 ml of nonionic contrast medium (concentration, 00 mg/ml) either alone or pushed with 0 ml of saline solution through a double-syringe power injector at a rate of.5 ml/s. Mean enhancement of the liver improved 9 ± 9 H when saline solution was used. Itoh et al. [] studied the effects of a 5% dextrose flush on aortic, portal, and hepatic enhancement in 80 patients given. ml/kg of contrast medium (concentration, 00 mg/ml) in four different protocols. No statistically significant difference in hepatic enhancement in the portal phase was observed between protocols with or without 5% dextrose. Our results contradict those of Dorio et al. [] and Schoellnast et al. [] but are in accord with those of Itoh et al. []. Our data do not show that bolus flushing improves hepatic enhancement in the portal phase or allows significant reduction in the amount of iodine administered. Numerous factors, such as dose of contrast medium, injection rate, venous site of injection, and the liver disease involved, may account for the discrepancies. Dorio et al. and Schoellnast et al. used a fixed dose of contrast medium, whereas we determined the dose according to the patient s body weight. Our results suggest that hepatic enhancement depends on the total dose of contrast medium and not on whether a saline chaser is administered. In the portal phase, all the contrast material is available to contribute to hepatic enhancement because the dead spaces of the right-heart cavities and the brachiocephalic veins have already been washed out. One of the aims of this study was to determine whether the use of saline solution allows reduction in doses of contrast medium. Some investigators have devised strategies to lower iodine doses, but their findings have been contradictory. Most authors consider maximal hepatic enhancement greater than 50 H to be adequate. Bree and colleagues [4] found that 5 ml of iohexol 00 (4.7 g iodine) resulted in quantitatively optimal hepatic enhancement but that qualitatively optimal enhancement required 50 ml (45 g iodine). Baker et al. [5] found no significant difference in hepatic enhancement in a 68-patient study in which one group received 50 ml of iopamidol 00 (45 g iodine) and another received 5 ml of ioversol 0 (40 g iodine). Blumeke et al. [6] extrapolated from data on rabbits and concluded that 49 mg/kg is the lowest dose with which to produce a 0-H increase in hepatic enhancement. This figure is close to the 457 mg/kg used by Brink et al. []. In a large, clinical multicenter trial, Megibow and colleagues [] established that a JR:87, ugust 006 5

4 dose of 450 mg/kg (.5 ml/kg of iopromide at a concentration of 00 mg/ml) produces acceptable scans in most patients. The effect of patient weight has also been noted in previous studies. Brink s group [] found that in heavy patients an iodine dose of 8 g was required for adequate enhancement, whereas in thin patients doses as low as 6 g might be sufficient. In a prospective randomized study by Yamashita et al. [7], hepatic parenchymal enhancement at visual analysis was graded as Fig. rterial acquisition for 65-year-old woman in groups and. No significant differences were observed between enhancement of celiac trunk and hepatic artery.,, injection of.5 ml/kg of contrast medium. B,, injection of.5 ml/kg of contrast medium followed by 0 ml of saline flush at ml/s. Fig. Portal acquisition for 65-year-old woman in groups and.,, injection of.5 ml/kg of contrast medium. B,, injection of.5 ml/kg of contrast medium followed by 0 ml of saline solution at ml/s. Hepatic enhancement is similar to that in. good or excellent in 64% of patients receiving.5 ml/kg, 85% of those receiving ml/kg, 94% of those receiving.5 ml/kg, and 65% of those receiving a fixed dose. We chose a basic dose of.5 ml/kg because it is close to the minimum recommended by Megibow and colleagues []. fixed dose reduction of 0 ml for all patients in group was adopted because it was simple and reproducible. Moreover, this amount was considered a reasonable approximation of the dead space in the peripheral vein that does not contribute to aortic or hepatic enhancement. Because it corresponds to the vascular compartment, we considered it unlikely this amount would be significantly affected by body weight. These hypotheses were not confirmed because hepatic enhancement apparently was unaffected by flushing. The dose used for thin patients therefore is clearly too low. That reductions in contrast medium were not made on the basis of weight is a poten- B B 54 JR:87, ugust 006

5 Saline Chaser in bdominal CT tial limitation, because thin patients were disproportionately sensitive to dose lowering. The observation of less enhancement in thin patients in all groups is in accord with the finding by Megibow et al. [] that the larger percentage of unacceptable enhancement was found in patients in the lowweight group. We observed that the relation between a patient s weight and hepatic enhancement was not linear and that the highest levels of hepatic enhancement occurred in the highest weight category, even in groups and. The reason for this effect remains hypothetical. It may be that in large patients, the ratio between the diffusion space of contrast medium (i.e., interstitial and blood pools) and the intracellular pool is lower. It is interesting that hepatic enhancement greater than 45 H was seen in all subjects except for the thin and medium-sized patients in group (< 55 kg, hepatic enhancement, 8 H; kg, hepatic enhancement, 45 H). Even if it is agreed that maximal hepatic enhancement greater than 50 H produces 00% acceptable scans, important individual variability occurs in acceptability levels. Thus in group, scans were acceptable for large patients. Therefore, if hepatic enhancement alone is considered, one might use a lower dose of contrast medium for these patients. This practice, however, is not acceptable for vascular structures because aortic enhancement is lower in large patients. With regard to aortic enhancement, attenuation values remained high even when the dose of contrast medium was reduced. This finding is consistent with recent evidence that MDCT allows dramatic reduction in contrast medium in vascular studies [8]. In our study, there was no difference among the three groups. This finding is in agreement with the finding by Itoh et al. [] of no significant difference in aortic enhancement in the early arterial phase between four injection techniques. The lack of discrepancy between groups can probably be attributed to assessment of aortic enhancement before the arterial peak, which is known to occur (approximately) at the end of continuous injection after a continuous increase in arterial opacification due to recirculation effects []. With MDCT, acquisition time is very short. Because it lasted 6 seconds, the arterial phase in this study was completed before the end of injection in many patients. bsence of late arterial data was a limitation of the investigation. The effect on aortic enhancement of the volume injected and the chaser was probably greater in the late arterial phase, as suggested by the results of Itoh et al. []. In their study, a statistically significant increase in aortic enhancement was observed in the late arterial phase and corresponded to the aortic peak of enhancement, when saline flush was used. In this study, multiphasic acquisitions with early and late arterial phases were not clinically justified. In summary, reducing the volume of contrast medium in this context leads to a reduction in hepatic enhancement not compensated by saline flushing. In the portal phase, it is unlikely that contrast medium in the arm vein will be available for saline flushing. However, if most examinations are to be acceptable, greater than.5 ml/kg reduction in contrast bolus may not be advisable for scanning of the livers of thin and medium-sized patients. With regard to aortic enhancement in the early arterial phase, a low dose of contrast medium is sufficient to produce the same enhancement as.5 ml/kg with or without a bolus chaser. cknowledgments We thank Jean Bérard for his assistance and William Francis for reviewing the manuscript. References. Dorio P, Lee FT, Henseler K, et al. Using a saline chaser to decrease contrast media in abdominal CT. JR 00; 80: Megibow J, Jacob G, Heiken JP, et al. Quantitative and qualitative evaluation of volume of low osmolality contrast medium needed for routine helical abdominal CT. JR 00; 76: Brink J, Heiken J, Forman H, Sagel S, Molina P, Brown P. Hepatic spiral CT: reduction of dose of intravenous contrast material. Radiology 995; 97: Bonaldi VM, Garcia P, Bret PM. Parameters influencing hepatic parenchyma enhancement in X-ray computed tomography after intravenous injection of iodinated contrast medium [in French]. J Radiol 997; 78: Haage P, Schmitz-Rode T, Hubner D, Piroth W, Gunther RW. Reduction of contrast material dose and artifacts by a saline flush using a double power injector in helical CT of the thorax. JR 000; 74: Bae KT, Heiken JP, Brink J. ortic and hepatic contrast medium enhancement at CT. Part. Prediction with a computer model. Radiology 998; 07: SS Institute. SS/STT User s guide, release 8.. Cary, NC: SS Institute, Fleishman D. Use of high concentration contrast media: principles and rationale. Eur J Radiol 00; 45:S88 S9 9. Katzberg R. Urography into the st century: new contrast media, renal handling, imaging characteristics, and nephrotoxicity. Radiology 997; 04:97 0. Lautin E, Freeman NJ, Schoenfeld H, et al. Radiocontrast-associated renal dysfunction: a comparison of lower-osmolality and conventional high-osmolality contrast media. JR 99; 57: Hopper KD, Mosher TJ, Kasales CJ, TenHave TR, Tully D, Weaver JS. Thoracic spiral CT: delivery of contrast material pushed with injectable saline solution in power injector. Radiology 997; 05:69 7. Schoellnast H, Tillich M, Deutschmann H, Schaffler G, Schoellnast R, Uggowitzer M. Improvement of parenchymal and vascular enhancement using saline flush and power injection for multiple-detector-row abdominal CT. Eur Radiol 004; 4: Itoh S, Ikeda M, chiwa M, Ota T, Satake H, Ishigaki T. Multiphase contrast enhanced CT of the liver with a multislice CT scanner. Eur Radiol 00; : Bree RL, Parisky YR, Bernardino ME, Costello P, Leder R, Brown PC. Cost-effective use of low-osmolality contrast media for CT of the liver: evaluation of liver enhancement provided by various doses of iohexol. JR 994; 6: Baker ME, Beam C, Leder R, Gulliver D, Paine SS, Dunnick NR. Contrast material for combined abdominal and pelvic CT of the liver: can cost be reduced by increasing the concentration and decreasing the volume? JR 99; 60: Blumeke D, Fishman EK, nderson JH. Dose requirements for a nonionic contrast agent for spiral computed tomography of the liver in rabbits. Invest Radiol 994; 9: Yamashita Y, Komohara Y, Takahashi M, et al. bdominal helical CT: evaluation of optimal doses of intravenous contrast material a prospective randomized study. Radiology 000; 6: Rubin G. MDCT imaging of the aorta and peripheral vessels. Eur J Radiol 00; 45:S4 S49 JR:87, ugust