Australas Radio1 1992; 36: 210-213
Liver CT: A Practical Approach to Dynamic Contrast Enhancement MICHAEL R. DITCHHELD, M.B., B.S. Radiology Registrar The Royal Melbourne Hospital ROBERTN. GIBSON, M.B., B.S., F.R.A.C.R., D.D.U. Associate Professor The University of Melbourne Radiology Department NEIL FAIRLIE, B .M., B.Ch. (Oxford), M.R.C.P., F.R.C.R. Radiology Registrar The Royal Melbourne Hospital
The protocols using 100 mls of contrast produced approximately twice the parenchymal and hepatic venous enhancement compared with those using 50 mls. Approximately 60-90% of examinations using 100 mls produced scans through the entire liver during the bolus or nonequilibrium phase, deemed the most sensitive for the detection of focal lesions, compared with 13-33% of those using 50 mls. Equally satisfactory results were obtained using the relatively inexpensive Biotel power injector preceded by a 40 ml hand injected bolus, compared with using an Angiomat angiography infusion Pump. We conclude that 100ml of contrast medium is necessary to achieve high quality liver CT scans and that this can be adequately delivered using a combination of a hand injection and an infusion from a relatively inexpensive pump.
ABSTRACT The aim of this study was to establish a practical, simple protocol that reliably produces high quality dynamic incremental computed tomography (CT) of the liver. We reviewed 90 patients randomly allocated into six different protocols. All had preliminary unenhanced scans followed by a dynamic incremental CT of the liver. An initial delay of 30 seconds was used from the commencement of the injection of Iopamiro 370. The groups were: 1. Pump infusion (a) 100 mls at 2 mls/sec scanning inferosuperiorl y. (b) 100 mls at 2 mIs/sec scanning superoinferiorly. ( c ) 100 mls at 1 ml/sec scanning inferosuperiorly. (d) 50 mls at 1 ml/sec scanning inferosuperiorly. 2. 40 mls hand injected bolus followed immediately by 60 ml pump infusion a t 1.3 mls/sec scanning inferosuperiorly. 3. 50 mls hand injected bolus scanning inferosuperiorly. The parameters recorded were the degree of hepatic parenchymal and hepatic venous enhancement and the aortic - IVC difference at the last slice through the liver, all measured in Hounsfield units.
INTRODUCTION Intravenous contrast medium enhancement of the liver can be performed in may ways with variations in contrast dose, technique of administration and time of scanning with reference to the contrast administration. When inappropriately used (eg. slow infusion technique) lesion detection may actually be impaired by the use of the contrast media compared with a non contrast scan (1).
Address for correspondence: Dr Michael R Ditchfield Department of Radiology C/o Post Office The Royal Melbourne Hospital Victoria 3050 Australia
MATERIALS AND METHODS We studied 90 patients referred for CT scans that would include the liver. They were scanned on a GE 9800 before and following IV Iopamidol 370. The contrast scans were obtained with an injection using either a 19g Butterfly or a 20g Jelco into a dorsal hand or antecubital vein. Scans of lcm
210
width and 1.5cm separation were obtained with a 30 second initial delay, an individual scan time of 2 seconds and a 6-7 second interscan delay. Patients were randomly placed into 6 groups of 15 patients. The groups were: 1. Using an Angiomat angiography infusion pump. (a) 100 mls at 2 mlsfsec scanning inferosuperiorly. (b)100 mls at 2 mls/sec scanning superoinferiorly. (c 100 mls at 1 ml/sec scanning inferosuperiorly. (d)50 mls at 1 ml/sec scanning inferosuperiorly. 2. 40 ml hand injected bolus followed immediately by 60 mls delivered by a Biotel power injector at 1.3 mls/sec scanning inferosuperiorly. 3. 50 mls injected by hand over 20-25 seconds scanning inferosuperiorly. The parameters documented were: 1. At a level immediately below the junction of the right hepatic vein and IVC (chosen as a relatively reproducible level passing through both left and right lobes of the liver) the degree of enhancement measured in Hounsfield units (HU) observed in: (a)The liver parenchyma of the right lobe. (b)The right hepatic vein. 2. At the last scan through the liver (ie. most inferior when scanning superoinferiorly and most superior when scanning inferosuperiorly) the difference in HU between the Submitted for publication on: 15th October, 1991 Accepted for publication on: 21st January, 1992
Australasian Radiology. Vol. 36, No. 3. August, 1992
LIVER C T A PRACTICAL APPROACH TO DYNAMIC CONTRAST ENHANCEMENT aorta and inferior vena cava. The study was assessed as satisfatory if the aorto-IVC difference was greater than 10 HU and hence the scans completed in the bolus or non-equilibrium phase (2). The first scans through the liver were assessed subjective 1y to determine if the degree of initial venous enhancement was satisfactory or not satisfactory. The time to complete all scans through the liver was also measured. RESULTS The results are summarised in Table 1. It is evident that 100 mls of contrast produces an average liver parenchymal enhancement in the four techniques used of 48-53 Hounsfield units. The enhancement of the right hepatic vein varied from 96-105 HU, with the lower figure representing that of the lowest infusion rate of 1 ml/sec. The number of studies completed during the bolus or non-equilibrium phase (ie. an aorto-IVC difference of greater that 10 HU) varied from 6093%, using 100 mls. Using 50 mls the liver parenchymal enhancement was approximately half that when 100 ml was used, varying from 21-28 HU. Similarly the right hepatic venous enhancement was approximately halved varying from 42-59 HU. The number of scans completed in the bolus or non-equilibrium phase was also considerably less varying from 13-33%. Overall, regardless of the technique used, 6-13% of studies showed unsatisfactory early hepatic venous enhancement in the initial scans through the liver. However, in no cases was this considered diagnostically important on subjective assessment. However, the initial scans when scanning inferosuperiorly only assess a small volume of the right lobe and this potential problem is therefore minimized using this technique. The time taken to complete the scans in all groups studied demonstrated that the vast majority (98%) could be completed within two minutes of cessation of the contrast bolus. DISCUSSION The aim of IV contrast is to improve the diagnostic accuracy of a range of pathology including mass lesions, biliary and vascular disease. The detection of a focal lesion is Australasian Radiology,Vol.36,No.3,August, 1992
Absolute Liver Parenchymal Enhancement (HU)
TABLE 1 Absolute Right Hepatic Vein Enhancement (HU)
Unsatisfactory Initial Venous Enhancement
Study Completed in Nonequilibrium Phase
IOOmls, 2ml/sec. InferosuDerior
50.2
105.1
13.3%
73%
lOOmls, lml/sec. Superoinferior
48.3
104.9
13.3%
73%
lOOmls, Iml/sec. Inferosuperior
52.9
96
6.7%
4Oml hand bolus, 6Oml at 1.3ml/sec. Inferosuperior
48.4
104.3
13.3%
60%
5Oml hand bolus, Inferosuperior
28.6
58.7
13.3%
33.3%
20.1
41.9
13.3%
13.3%
93.3%
5Om1, Iml/sec.
Inferosuperior
dependent on lesion size, section thickness and contrast differentiation (3). Contrast enhancement in CT scanning is of value in detecting hepatic mass lesions only if the attenuation difference between parenchyma and lesion is increased. This is optimised by scanning when there is a significant difference between the intravascular and extravascular concentration of contrast media. According to Burgener (2) there are three phases of contrast enhancement: Bolus phase - during which there is marked vascular and parenchymal enhancement which peaks at the end of a bolus injection. Non-equilibrium phase - when vascular enhancement declines rapidly. Equilibrium phase - when the concentration in the intra and extravascular space equalizes. The first two Dhases can be recognised when thk arteriovenous density difference is greater than 10 HU. The non-equilibrium phase lasts approximately 2 minutes following a bolus injection and is followed by the equilibrium phase (2). Lesion detection is increased by 15% and individual lesion size is maximised when scanning is performed during the bolus or nonequilibrium phase (1). The average difference in liver to lesion attenuation has been found to increase from 20 to 60 HU using this technique, making lesion detection much more certain (3) when at least 10 HU difference is required to detect a lesion with confidence (4,5). When scans are performed during the equilibrium phase lesion detection is actually lower than non contrast scans (1, 4, 6, 7, 8). However persis-
tent nonequilibrium achieved by a slow infusion does not guarantee lesion detectability equivalent to bolus contrast delivery (1). This relates to the rapid passage of contrast to the interstitial space occurring throughout an infusion. More than half of iodinated contrast injected via a systematic vein is transferred to the interstitial space within 1 minute and by 5 minutes 80% is transferred (9). A multitude of possible protocols have been used to optimise hepatic CT scanning. The possible variables for contrast enhancement are: 1. Contrast dose (volume & concentration); 2. Rate of administration; and 3. Time of scanning with reference to contrast administration. The choice of contrast dose requires a compromise between the greatest and most prolonged enhancement, and cost and patient risk. The volumes of contrast have varied from relatively low to very high doses (50 mls-300 mls) of a variable iodine concentration (4). Doubling the iodine dose resulted in an increase in liver parenchymal enhancement by a factor of 1.5-2.5 (lo), similar to our own results. Different rates of iodine infusion have also been used and the highest enhancement was always obtained from the most rapid infusion. Dean et a1 found shorter infusion times (30 sec) produced greater and earlier peak parenchymal enhancement (10). This suggests that the greatest improvement in visualisation of a lesion is obtained during and in the short time after bolus injection. A more rapid injection of a smaller contrast volume has also been found to produce equivalent hepatic peak 211
MICHAEL R. DITCHFIELD er a1 and overall contrast enhancement compared with a slower yet larger volume injection, actually allowing a reduction in contrast dose for similar enhancement (11). Foley et af suggest that the normal hepatic parenchyma is best enhanced by the rapid and sustained delivery of a large contrast bolus (45-50 gms of iodine administered over 2-2.5 mins), rather than by an infusion technique (42 gms of iodine delivered over 5-10 mins). Better liver to lesion contrast is achieved with bolus than infusion techniques because parenchyma is more enhanced and less contrast medium diffuses into the interstitial space of tumours. The timing of scans with respect to contrast administration is very important to ensure that contrast is in the appropriate region when scanned. Normal circulation time from cubital vein to central arteries is 13-22 seconds, but this is increased when cardiac output is impaired (4). Following a bolus injection the delay used in several studies before scanning commences has varied between 15-45 seconds (1, 6). Our results have shown that liver parenchymal enhancement, important in lesion detection, is greater with 100 mls of contrast and is greater with a rapid 50 ml bolus than with a slower infusion when only 50 mls of contrast is used. Parenchymal enhancement is similar regardless of the direction of the scan. Hepatic venous enhancement was also far greater when 100 mls of contrast was used and also greater with the more rapid bolus. The direction of scanning did not affect the absolute enhancement of the hepatic veins. However, some degree of poor early venous enhancement was found in 613% of studies regardless of the technique used. Only a small volume of the liver is scanned initially using the inferosuperior technique thereby minimizing the potential problem of an unenhanced hepatic vein mimicking a focal lesion. The hepatic studies were completed in 98% of patients within 2 minutes of the end of the bolus injection, meeting one of the criteria of scanning in the nonequilibrium phase. However only 13-30% of studies with 50 mls and 60-93% of studies using 100 rnls were completed within the nonequiiibnum phase when the arteriovenous density difference of >10 HU was used as a criterion. Clearly the length of the nonequilibrium phase varies with contrast dose, rate of 212
FIGURE 1 - The Biotel pump showing the 40 ml syringe connected by a 3 way tap enabling a rapid hand bolus followed immediately by the pump infusion.
infusion and patients’ cardiovascular status. Infusion pumps capable of delivering 100 rnls of contrast are in general expensive. Equally satisfactory scans can be obtained using a relatively inexpensive pump (Figure 1) holding 60 mls of contrast which i s immediately preceded by a 40ml contrast bolus injected by hand. CONCLUSION Good quality dynamic hepatic CT scans can be obtained using a 30 sec-
ond initial delay and at least 100 mls of contrast injected at a rate of 1-2 mlsfsec. Scanning inferosuperiorly is preferred so that the occasional case of poor early venous enhancement affects only a small volume of the liver. A relatively inexpensive Biotel power injector produces satisfactory results and the technique should therefore be achieveable by most radiological practices. ACKNOWLEDGEMENTS We would like to thank the CT staff and, in particular, Allison Lee-Tannock for their help and cooperation. Australasian Radiology, Vol.36. No. 3. August, 1992
LIVER CT: A PRACTICAL APPROACH TO DYNAMIC CONTRAST ENHANCEMENT REFERENCES 1. Paushter DM, Zeman RK, Scheibler ML, Cheyke PL, Jaffe MH, Clark LR. CT evaluation of suspected hepatic metastases: comparison of techniques for IV contrast enhancement. AJR 1989; 152: 267-27 I . 2. Burgener FA, Hamlin DJ. Contrast enhancement in abdominal CT: bolus versus infusion. AJR 1981; 137: 351-358. 3. Foley WD, Berland LL, Lawson TL, Smith DF, Thorsen MK. Contrast enhancement technique for dynamic hepatic computed tomographic scanning. Radiology 1983; 147: 797-803.
Austsulasiun Radiology, Val. 36, No. 3,August. 1992
4.
5.
6. 7.
8.
Albertyn LE. Rationales for the use of intravenous contrast medium in computed tomography. Australas Radiol 1989; 33: 1 29-33. Foley WD.Dynamic hepatic CT. Radiology 1989; 170 617-622. Foley WD. Dynamic hepatic CT scanning. kTR 1989; 152: 272-274. Burgener FA, Hamlin DJ. Contrast enhancement of hepatic tumours in CT: comparison between bolus and infusion techniques. AJR 1983; 140: 291-295. Moss AA, Schrumpf J, Schnyder P, Korobkin M, Shimshak RR. Computed tomography of focal hepatic lesions: a blind clinical evaluation of the effect of contrast enhancement. Radiology 1979; 131: 427-430.
9. Kormano M, Dean PM. Extravascular contrast material: the major components of contrast enhancement. Radiology 1976; 121: 379-382. 10. Dean PB, Violante MR, Mahoney JA. Hepatic CT contrast enhancement: effect of dose, duration of infusion, and time elapsed following infusion. Invest Radiol 1980; 2: 158-161. 11. Berland LL, Lee JY. Comparison of contrast media injection rates and volumes for hepatic dynamic incremented computed tomography. Invest Radiol 1988; 23: 9 18-922.
213
Liver CT: A Practical Approach to Dynamic Contrast Enhancement MICHAEL R. DITCHHELD, M.B., B.S. Radiology Registrar The Royal Melbourne Hospital ROBERTN. GIBSON, M.B., B.S., F.R.A.C.R., D.D.U. Associate Professor The University of Melbourne Radiology Department NEIL FAIRLIE, B .M., B.Ch. (Oxford), M.R.C.P., F.R.C.R. Radiology Registrar The Royal Melbourne Hospital
The protocols using 100 mls of contrast produced approximately twice the parenchymal and hepatic venous enhancement compared with those using 50 mls. Approximately 60-90% of examinations using 100 mls produced scans through the entire liver during the bolus or nonequilibrium phase, deemed the most sensitive for the detection of focal lesions, compared with 13-33% of those using 50 mls. Equally satisfactory results were obtained using the relatively inexpensive Biotel power injector preceded by a 40 ml hand injected bolus, compared with using an Angiomat angiography infusion Pump. We conclude that 100ml of contrast medium is necessary to achieve high quality liver CT scans and that this can be adequately delivered using a combination of a hand injection and an infusion from a relatively inexpensive pump.
ABSTRACT The aim of this study was to establish a practical, simple protocol that reliably produces high quality dynamic incremental computed tomography (CT) of the liver. We reviewed 90 patients randomly allocated into six different protocols. All had preliminary unenhanced scans followed by a dynamic incremental CT of the liver. An initial delay of 30 seconds was used from the commencement of the injection of Iopamiro 370. The groups were: 1. Pump infusion (a) 100 mls at 2 mls/sec scanning inferosuperiorl y. (b) 100 mls at 2 mIs/sec scanning superoinferiorly. ( c ) 100 mls at 1 ml/sec scanning inferosuperiorly. (d) 50 mls at 1 ml/sec scanning inferosuperiorly. 2. 40 mls hand injected bolus followed immediately by 60 ml pump infusion a t 1.3 mls/sec scanning inferosuperiorly. 3. 50 mls hand injected bolus scanning inferosuperiorly. The parameters recorded were the degree of hepatic parenchymal and hepatic venous enhancement and the aortic - IVC difference at the last slice through the liver, all measured in Hounsfield units.
INTRODUCTION Intravenous contrast medium enhancement of the liver can be performed in may ways with variations in contrast dose, technique of administration and time of scanning with reference to the contrast administration. When inappropriately used (eg. slow infusion technique) lesion detection may actually be impaired by the use of the contrast media compared with a non contrast scan (1).
Address for correspondence: Dr Michael R Ditchfield Department of Radiology C/o Post Office The Royal Melbourne Hospital Victoria 3050 Australia
MATERIALS AND METHODS We studied 90 patients referred for CT scans that would include the liver. They were scanned on a GE 9800 before and following IV Iopamidol 370. The contrast scans were obtained with an injection using either a 19g Butterfly or a 20g Jelco into a dorsal hand or antecubital vein. Scans of lcm
210
width and 1.5cm separation were obtained with a 30 second initial delay, an individual scan time of 2 seconds and a 6-7 second interscan delay. Patients were randomly placed into 6 groups of 15 patients. The groups were: 1. Using an Angiomat angiography infusion pump. (a) 100 mls at 2 mlsfsec scanning inferosuperiorly. (b)100 mls at 2 mls/sec scanning superoinferiorly. (c 100 mls at 1 ml/sec scanning inferosuperiorly. (d)50 mls at 1 ml/sec scanning inferosuperiorly. 2. 40 ml hand injected bolus followed immediately by 60 mls delivered by a Biotel power injector at 1.3 mls/sec scanning inferosuperiorly. 3. 50 mls injected by hand over 20-25 seconds scanning inferosuperiorly. The parameters documented were: 1. At a level immediately below the junction of the right hepatic vein and IVC (chosen as a relatively reproducible level passing through both left and right lobes of the liver) the degree of enhancement measured in Hounsfield units (HU) observed in: (a)The liver parenchyma of the right lobe. (b)The right hepatic vein. 2. At the last scan through the liver (ie. most inferior when scanning superoinferiorly and most superior when scanning inferosuperiorly) the difference in HU between the Submitted for publication on: 15th October, 1991 Accepted for publication on: 21st January, 1992
Australasian Radiology. Vol. 36, No. 3. August, 1992
LIVER C T A PRACTICAL APPROACH TO DYNAMIC CONTRAST ENHANCEMENT aorta and inferior vena cava. The study was assessed as satisfatory if the aorto-IVC difference was greater than 10 HU and hence the scans completed in the bolus or non-equilibrium phase (2). The first scans through the liver were assessed subjective 1y to determine if the degree of initial venous enhancement was satisfactory or not satisfactory. The time to complete all scans through the liver was also measured. RESULTS The results are summarised in Table 1. It is evident that 100 mls of contrast produces an average liver parenchymal enhancement in the four techniques used of 48-53 Hounsfield units. The enhancement of the right hepatic vein varied from 96-105 HU, with the lower figure representing that of the lowest infusion rate of 1 ml/sec. The number of studies completed during the bolus or non-equilibrium phase (ie. an aorto-IVC difference of greater that 10 HU) varied from 6093%, using 100 mls. Using 50 mls the liver parenchymal enhancement was approximately half that when 100 ml was used, varying from 21-28 HU. Similarly the right hepatic venous enhancement was approximately halved varying from 42-59 HU. The number of scans completed in the bolus or non-equilibrium phase was also considerably less varying from 13-33%. Overall, regardless of the technique used, 6-13% of studies showed unsatisfactory early hepatic venous enhancement in the initial scans through the liver. However, in no cases was this considered diagnostically important on subjective assessment. However, the initial scans when scanning inferosuperiorly only assess a small volume of the right lobe and this potential problem is therefore minimized using this technique. The time taken to complete the scans in all groups studied demonstrated that the vast majority (98%) could be completed within two minutes of cessation of the contrast bolus. DISCUSSION The aim of IV contrast is to improve the diagnostic accuracy of a range of pathology including mass lesions, biliary and vascular disease. The detection of a focal lesion is Australasian Radiology,Vol.36,No.3,August, 1992
Absolute Liver Parenchymal Enhancement (HU)
TABLE 1 Absolute Right Hepatic Vein Enhancement (HU)
Unsatisfactory Initial Venous Enhancement
Study Completed in Nonequilibrium Phase
IOOmls, 2ml/sec. InferosuDerior
50.2
105.1
13.3%
73%
lOOmls, lml/sec. Superoinferior
48.3
104.9
13.3%
73%
lOOmls, Iml/sec. Inferosuperior
52.9
96
6.7%
4Oml hand bolus, 6Oml at 1.3ml/sec. Inferosuperior
48.4
104.3
13.3%
60%
5Oml hand bolus, Inferosuperior
28.6
58.7
13.3%
33.3%
20.1
41.9
13.3%
13.3%
93.3%
5Om1, Iml/sec.
Inferosuperior
dependent on lesion size, section thickness and contrast differentiation (3). Contrast enhancement in CT scanning is of value in detecting hepatic mass lesions only if the attenuation difference between parenchyma and lesion is increased. This is optimised by scanning when there is a significant difference between the intravascular and extravascular concentration of contrast media. According to Burgener (2) there are three phases of contrast enhancement: Bolus phase - during which there is marked vascular and parenchymal enhancement which peaks at the end of a bolus injection. Non-equilibrium phase - when vascular enhancement declines rapidly. Equilibrium phase - when the concentration in the intra and extravascular space equalizes. The first two Dhases can be recognised when thk arteriovenous density difference is greater than 10 HU. The non-equilibrium phase lasts approximately 2 minutes following a bolus injection and is followed by the equilibrium phase (2). Lesion detection is increased by 15% and individual lesion size is maximised when scanning is performed during the bolus or nonequilibrium phase (1). The average difference in liver to lesion attenuation has been found to increase from 20 to 60 HU using this technique, making lesion detection much more certain (3) when at least 10 HU difference is required to detect a lesion with confidence (4,5). When scans are performed during the equilibrium phase lesion detection is actually lower than non contrast scans (1, 4, 6, 7, 8). However persis-
tent nonequilibrium achieved by a slow infusion does not guarantee lesion detectability equivalent to bolus contrast delivery (1). This relates to the rapid passage of contrast to the interstitial space occurring throughout an infusion. More than half of iodinated contrast injected via a systematic vein is transferred to the interstitial space within 1 minute and by 5 minutes 80% is transferred (9). A multitude of possible protocols have been used to optimise hepatic CT scanning. The possible variables for contrast enhancement are: 1. Contrast dose (volume & concentration); 2. Rate of administration; and 3. Time of scanning with reference to contrast administration. The choice of contrast dose requires a compromise between the greatest and most prolonged enhancement, and cost and patient risk. The volumes of contrast have varied from relatively low to very high doses (50 mls-300 mls) of a variable iodine concentration (4). Doubling the iodine dose resulted in an increase in liver parenchymal enhancement by a factor of 1.5-2.5 (lo), similar to our own results. Different rates of iodine infusion have also been used and the highest enhancement was always obtained from the most rapid infusion. Dean et a1 found shorter infusion times (30 sec) produced greater and earlier peak parenchymal enhancement (10). This suggests that the greatest improvement in visualisation of a lesion is obtained during and in the short time after bolus injection. A more rapid injection of a smaller contrast volume has also been found to produce equivalent hepatic peak 211
MICHAEL R. DITCHFIELD er a1 and overall contrast enhancement compared with a slower yet larger volume injection, actually allowing a reduction in contrast dose for similar enhancement (11). Foley et af suggest that the normal hepatic parenchyma is best enhanced by the rapid and sustained delivery of a large contrast bolus (45-50 gms of iodine administered over 2-2.5 mins), rather than by an infusion technique (42 gms of iodine delivered over 5-10 mins). Better liver to lesion contrast is achieved with bolus than infusion techniques because parenchyma is more enhanced and less contrast medium diffuses into the interstitial space of tumours. The timing of scans with respect to contrast administration is very important to ensure that contrast is in the appropriate region when scanned. Normal circulation time from cubital vein to central arteries is 13-22 seconds, but this is increased when cardiac output is impaired (4). Following a bolus injection the delay used in several studies before scanning commences has varied between 15-45 seconds (1, 6). Our results have shown that liver parenchymal enhancement, important in lesion detection, is greater with 100 mls of contrast and is greater with a rapid 50 ml bolus than with a slower infusion when only 50 mls of contrast is used. Parenchymal enhancement is similar regardless of the direction of the scan. Hepatic venous enhancement was also far greater when 100 mls of contrast was used and also greater with the more rapid bolus. The direction of scanning did not affect the absolute enhancement of the hepatic veins. However, some degree of poor early venous enhancement was found in 613% of studies regardless of the technique used. Only a small volume of the liver is scanned initially using the inferosuperior technique thereby minimizing the potential problem of an unenhanced hepatic vein mimicking a focal lesion. The hepatic studies were completed in 98% of patients within 2 minutes of the end of the bolus injection, meeting one of the criteria of scanning in the nonequilibrium phase. However only 13-30% of studies with 50 mls and 60-93% of studies using 100 rnls were completed within the nonequiiibnum phase when the arteriovenous density difference of >10 HU was used as a criterion. Clearly the length of the nonequilibrium phase varies with contrast dose, rate of 212
FIGURE 1 - The Biotel pump showing the 40 ml syringe connected by a 3 way tap enabling a rapid hand bolus followed immediately by the pump infusion.
infusion and patients’ cardiovascular status. Infusion pumps capable of delivering 100 rnls of contrast are in general expensive. Equally satisfactory scans can be obtained using a relatively inexpensive pump (Figure 1) holding 60 mls of contrast which i s immediately preceded by a 40ml contrast bolus injected by hand. CONCLUSION Good quality dynamic hepatic CT scans can be obtained using a 30 sec-
ond initial delay and at least 100 mls of contrast injected at a rate of 1-2 mlsfsec. Scanning inferosuperiorly is preferred so that the occasional case of poor early venous enhancement affects only a small volume of the liver. A relatively inexpensive Biotel power injector produces satisfactory results and the technique should therefore be achieveable by most radiological practices. ACKNOWLEDGEMENTS We would like to thank the CT staff and, in particular, Allison Lee-Tannock for their help and cooperation. Australasian Radiology, Vol.36. No. 3. August, 1992
LIVER CT: A PRACTICAL APPROACH TO DYNAMIC CONTRAST ENHANCEMENT REFERENCES 1. Paushter DM, Zeman RK, Scheibler ML, Cheyke PL, Jaffe MH, Clark LR. CT evaluation of suspected hepatic metastases: comparison of techniques for IV contrast enhancement. AJR 1989; 152: 267-27 I . 2. Burgener FA, Hamlin DJ. Contrast enhancement in abdominal CT: bolus versus infusion. AJR 1981; 137: 351-358. 3. Foley WD, Berland LL, Lawson TL, Smith DF, Thorsen MK. Contrast enhancement technique for dynamic hepatic computed tomographic scanning. Radiology 1983; 147: 797-803.
Austsulasiun Radiology, Val. 36, No. 3,August. 1992
4.
5.
6. 7.
8.
Albertyn LE. Rationales for the use of intravenous contrast medium in computed tomography. Australas Radiol 1989; 33: 1 29-33. Foley WD.Dynamic hepatic CT. Radiology 1989; 170 617-622. Foley WD. Dynamic hepatic CT scanning. kTR 1989; 152: 272-274. Burgener FA, Hamlin DJ. Contrast enhancement of hepatic tumours in CT: comparison between bolus and infusion techniques. AJR 1983; 140: 291-295. Moss AA, Schrumpf J, Schnyder P, Korobkin M, Shimshak RR. Computed tomography of focal hepatic lesions: a blind clinical evaluation of the effect of contrast enhancement. Radiology 1979; 131: 427-430.
9. Kormano M, Dean PM. Extravascular contrast material: the major components of contrast enhancement. Radiology 1976; 121: 379-382. 10. Dean PB, Violante MR, Mahoney JA. Hepatic CT contrast enhancement: effect of dose, duration of infusion, and time elapsed following infusion. Invest Radiol 1980; 2: 158-161. 11. Berland LL, Lee JY. Comparison of contrast media injection rates and volumes for hepatic dynamic incremented computed tomography. Invest Radiol 1988; 23: 9 18-922.
213
