| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1999 ARRS Executive Council Award |
1
The Russell H. Morgan Department of Radiology and Radiological Science, The
Johns Hopkins Medical Institutions, 600 N. Wolfe St., Baltimore, MD
21287.
2
Department of Surgery, The Johns Hopkins Medical Institutions, Baltimore, MD
21287.
Received March 3, 1999;
accepted after revision July 1, 1999.
Address correspondence to E. K. Fishman.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Seven VX2 hepatic tumors in five rabbits were imaged. Dynamic contrast-enhanced CT was performed at a single level centered over the lesions at 5-sec intervals for 119 sec after injection of 2 ml/kg IV contrast material at 2 ml/sec. Attenuation was measured over time within regions of interest in the tumor and normal liver, aorta, inferior vena cava, and portal vein. Lesion conspicuity, defined as the difference between the attenuation of the uninvolved liver and neoplasm, was calculated.
RESULTS. The mean diameter of the tumors on CT was 10 mm (range, 6-15 mm). The tumors appeared as low-attenuation lesions with progressive enhancement during the arterial phase and early portal phase. Peak mean lesion attenuation was 60 ± 27 H (enhancement, 23 H) at 64 sec. Peak mean lesion conspicuity was 80 ± 18 H at 39 sec, occurring 10 sec before the peak mean hepatic attenuation of 135 ± 15 H (enhancement, 67 H) at 49 sec. Relative lesion conspicuity paralleled relative enhancement of the liver throughout the imaging period.
CONCLUSION. Although low-level tumor enhancement during the arterial phase and early portal phase reduced the conspicuity of small hypovascular tumors in this animal model, our results support the use of maximum liver enhancement as a marker for peak lesion conspicuity.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A large body of research has been published aimed at understanding the contrast enhancement kinetics of the normal liver [5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]. Many investigators have plotted the enhancement of the liver and aorta as a function of time with various injection protocols to characterize the time and degree of peak hepatic enhancement and the equilibrium phase, at which point these two curves become parallel. Foley [6] has suggested that during the equilibrium phase lesion conspicuity decreases. The first assumption underlying this concept is that the enhancement of hypovascular tumors is negligible before the equilibrium phase and therefore lesion conspicuity is primarily a function of the enhancement of surrounding normal liver. The second assumption is that lesion conspicuity decreases during the equilibrium phase because contrast material accumulates within the tumor [5, 6]. These assumptions are widely accepted and serve as the basis for current contrast enhancement protocols. However, the studies in which the enhancement of hepatic tumors was actually measured [20, 21, 22] were not performed with a degree of temporal resolution that was sufficient to optimize scanning technique using modern subsecond helical scanners and multidetector scanners [23], which can image the entire liver in a matter of seconds. In patient studies, the entire liver is typically imaged, thus limiting temporal resolution for evaluating the enhancement of any individual lesion.
An animal model provides a means of addressing this issue by allowing dynamic contrast enhancement studies at a single level centered over the tumor of interest. The purpose of this study was to evaluate the effect of lesion enhancement on the conspicuity of small hypovascular hepatic tumors in an animal model.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Five New Zealand white rabbits weighing 3.38-4.24 kg were used in this study. Before tumor implantation and imaging, rabbits received an intramuscular injection of acepromazine maleate and ketamine hydrochloride (Ketaject; Phoenix Scientific, St. Joseph, MO). Intravenous access was then acquired via a marginal ear vein. Animals were intubated with a noncuffed endotracheal tube and placed on a small-animal ventilator (model 55-0798; Harvard Apparatus, South Natick, MA). Anesthesia was maintained using IV sodium pentothol.
A tumor was implanted into the liver of each of the five rabbits using a midline laparotomy. Hepatic tumors were created by injecting fragments of VX2 tumor measuring approximately 1 mm3 into the right (n = 2), left (n = 1), or both (n = 2) lobes of the liver using a 16-gauge angiocath, with the needle removed, and a guidewire as a pusher.
CT
CT was performed 13-24 days after implantation using a Somatom Plus 4
scanner (Siemens, Iselin, NJ). Before imaging, the animals were anesthetized,
intubated, and placed on the small-animal ventilator. They were paralyzed
using IV pancuronium immediately before imaging; paralysis was reversed after
imaging was completed using IV neostigmine methylsulfate and atropine sulfate.
Animals were weighed before imaging to determine the appropriate dose of
contrast material based on 2 ml/kg.
An initial unenhanced helical study of the entire liver was performed using the following parameters: 3-mm collimation, 3-mm/sec table speed in the craniocaudal direction, 0.75-sec gantry rotation, 292 mA, and 120 kVp. Ventilation was suspended for the entire helical study. Axial reconstructions were performed every 4 mm using 180° linear interpolation. The level of the lesions was determined from the unenhanced helical study; dynamic contrast-enhanced examination was performed at that level after IV injection of 8-9 ml of nonionic iodinated contrast material (iohexol, 300 mg I/ml; dose, 2 ml/kg of body weight; iodine, 2.4-2.7 g) at 2 ml/sec via an ear vein using a power injector. Once the contrast material had been injected, dynamic scanning was begun using 2-mm collimation, 143 mA, 140 kVp, and a 100- to 164-mm field of view. Scans were obtained at 5-sec intervals from 9 to 119 sec after the initiation of the contrast injection, with delayed images of up to 199 sec. During imaging, ventilation was suspended for approximately 20-sec intervals alternated with 8-sec periods of hyperventilation. Periods of ventilation were varied between animals to ensure that sufficient data were available for all time points.
Image Analysis
CT data sets were transferred to a freestanding Sparc20 workstation (Sun
Microsystems, Mountain View, CA) running MagicView software (Siemens). A
radiologist carefully measured the attenuation within the regions of interest
containing 21 or 37 pixels in the aorta, inferior vena cava, and portal vein;
in three locations in the right, middle, and left liver distant from the
tumor; and in the center of the tumor. Focal areas of low attenuation within
the tumors that were suggestive of necrosis were avoided. Areas containing
beam-hardening artifacts or volume averaging with adjacent structures were
avoided in all measurements. Attenuation values that could not be measured
because of respiration were estimated using linear interpolation of adjacent
values. The resulting attenuation data from all animals were averaged for each
time point and plotted as a function of time. Tumor conspicuity at each time
point was calculated by subtracting the mean lesion attenuation from the mean
liver attenuation.
Definitions
The arterial phase of hepatic enhancement is the period between the rapid
upslope in aortic enhancement that is due to contrast material bolus and the
later rapid increase in enhancement of the portal vein. During the arterial
phase, hepatic enhancement primarily results from the contrast material that
is supplied by the hepatic artery. The portal phase of hepatic enhancement is
the period between the rapid upslope in portal vein enhancement that is due to
the contrast material bolus and the onset of the equilibrium phase, which we
discuss later in this article. During the portal phase, hepatic enhancement
primarily results from the contrast material that is supplied by the portal
vein; peak hepatic enhancement occurs during this phase. The equilibrium phase
of hepatic enhancement is the point at which the attenuation curves for the
aorta and liver become parallel and decline at an equal rate
[6,
7]. Conspicuity is the
attenuation of normal liver parenchyma minus the attenuation of the tumor of
interest.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
The peak mean hepatic attenuation of 135 ± 15 H (enhancement, 67 H) occurred at 49 sec (Fig. 2). At baseline before infusion of contrast material, the tumors measured 37 ± 13 H. Peak mean lesion attenuation was 60 ± 27 H (enhancement, 23 H) at 64 sec. Tumor enhancement began during the arterial phase and continued through the early portal phase; tumor enhancement plateaued during the mid portal phase and appeared to change minimally during the late portal phase and equilibrium phase, which began at approximately 100 sec (Fig. 3). The peak mean aortic attenuation of 910 ± 274 H (859 H enhancement) occurred at 19 sec after the initiation of the injection of contrast material. Peak mean portal vein attenuation was 231 ± 58 H (enhancement, 189 H) at 29 sec. Peak mean inferior vena cava attenuation was 271 ± 37 H (enhancement, 223 H) at 29 sec. A difference of less than 10 H between the attenuation of the aorta and inferior vena cava, (the definition of the equilibrium phase of enhancement proposed by Burgener and Hamlin [9]) occurred at 29 sec—well before peak hepatic enhancement and peak lesion conspicuity (Fig. 2).
|
|
Maximum lesion conspicuity occurred before maximum hepatic enhancement in four (57%) of seven lesions and at the time of peak hepatic enhancement in the remaining three lesions (43%). Peak mean lesion conspicuity was 80 ± 18 H at 39 sec, thus occurring 10 sec before the peak mean hepatic enhancement. The mean lesion conspicuity at peak hepatic enhancement was 76 ± 17 H—that is, 4 H (5%) less than peak conspicuity. Peak conspicuity occurred at approximately the same time that attenuation within the aorta and portal vein became equal (39 sec) (Fig. 2). The attenuation curves for the aorta and portal vein were nearly identical after this point. Relative conspicuity paralleled relative hepatic enhancement throughout the imaging period (Fig. 4).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The hypovascular tumors in this animal model enhanced well into the portal phase of hepatic enhancement. This tumor enhancement, combined with decreasing liver attenuation from peak hepatic enhancement to the onset of the equilibrium phase, resulted in a substantial decline in lesion conspicuity before the equilibrium phase of hepatic enhancement. These results confirm that helical CT protocols for imaging hypovascular liver tumors should cover the liver as rapidly as possible during peak hepatic enhancement. Acquisition that is completed just before the equilibrium phase but well after peak hepatic enhancement results in an average lesion conspicuity that is significantly diminished compared with a very rapid acquisition completed at peak lesion conspicuity. Potential strategies for imaging the liver as rapidly as possible include performing subsecond scanning, imaging with multidetector-array CT, and increasing the pitch. However, it is known that increasing the pitch increases the effective slice thickness—therefore reducing the conspicuity of small tumors because of volume-averaging effects.
Peak lesion conspicuity occurred when the attenuation in the aorta, inferior vena cava, and portal vein became equal and began to decrease together (Fig. 2). This finding corresponds to the onset of equilibrium in the concentration of the contrast material within the intravascular space. It is reasonable to postulate that lesion conspicuity increases up to that point because of differences in the relative amounts of blood being supplied by the hepatic artery and portal vein to the tumor and uninvolved liver parenchyma combined with differences in the concentration of contrast material within those two vessels. Once equilibrium between the concentrations of contrast material in the arterial system and the portal vein has been established, differences in blood supply to the tumor and liver can no longer be exploited to enhance lesion conspicuity.
The concept of an equilibrium phase of contrast enhancement has played a key role in the formulation of contrast enhancement protocols for CT detection of hypovascular liver tumors to date. Although the point at which the equilibrium phase begins has been defined differently, the underlying concept is that at later times after injection of a contrast bolus, hypovascular liver tumors become substantially less conspicuous. This phenomenon has been suggested to be a result of diffusion of contrast material into the extravascular space of the tumor [6, 7]. The concept of the equilibrium phase was defined by Burgener and Hamlin [9] as "a difference in blood iodine concentration between the aorta and inferior vena cava of less than 10 Hounsfield units." Foley [6] and Cox et al. [7] subsequently proposed an alternative definition of equilibrium as "when the two enhancement curves [aortic and hepatic] become parallel and decline at an equal rate," which they suggest occurs "when intravascular and extravascular contrast material equilibrate." In our study, equilibrium as defined by Burgener and Hamlin occurred before peak lesion conspicuity and well before peak hepatic enhancement. Our data suggest that enhancement of the tumor has plateaued and conspicuity is substantially lower than its maximum value by the onset of the equilibrium phase as defined by Foley and Cox et al.
The main potential limitation of applying the results of this study to patient care is that liver tumor enhancement characteristics in humans vary with tumor type and may differ from those seen in this animal model. Additionally, we studied a small number of hepatic tumors in a small number of animals. Larger studies are needed in humans to accurately characterize the enhancement characteristics of hepatic tumors as a function of both size and tumor type. The 100- to 164-mm field of view used in this animal study is smaller than that used in routine clinical practice. This smaller field of view provided better spatial resolution than standard scanning techniques used in human patients. Finally, the effects of the rate of contrast material administration and the total dose of contrast material administered on lesion enhancement and conspicuity were not evaluated.
In conclusion, the small hypovascular tumors in this animal model exhibited low-grade enhancement during the arterial phase and early portal phase of hepatic enhancement. Although peak tumor conspicuity occurred slightly before peak hepatic enhancement, conspicuity paralleled hepatic enhancement throughout the imaging period. Our results support the use of peak hepatic enhancement as a marker for peak lesion conspicuity. High-temporal resolution studies of tumor enhancement kinetics are needed in humans to optimize imaging protocols for detecting small hepatic tumors with modern subsecond and multidetector helical scanners.
Acknowledgments
We thank Nancy L. Spangler for her technical assistance.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1.5 cm) malignant hepatic neoplasms. AJR
1995;164:879-884
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. J. Karlik Visualizing Radiologic Data Am. J. Roentgenol., March 1, 2003; 180(3): 607 - 619. [Full Text] [PDF]
|
|
|
|
|
|
B. S. Kuszyk, F. M. Corl, F. N. Franano, D. A. Bluemke, L. V. Hofmann, B. J. Fortman, and E. K. Fishman Tumor Transport Physiology: Implications for Imaging and Imaging-Guided Therapy Am. J. Roentgenol., October 1, 2001; 177(4): 747 - 753. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), aorta (
), inferior vena cava (x), and portal vein
(
) plotted against time after initiation of contrast material injection.
Peak aortic attenuation of 910 H at 19 sec is not included on graph. Peak
conspicuity occurs when attenuation of aorta and portal vein are equal.
