AJR 2003; 180:681-686
© American Roentgen Ray Society
Using Slow-Infusion MR Arteriography and an Implantable Port System to Assess Drug Distribution at Hepatic Arterial Infusion Chemotherapy
Hiroshi Seki1,
Toshirou Ozaki2,
Satoshi Takaki2,
Hiroyuki Ooi2,
Junichi Oda1 and
Makoto Shiina1
1 Department of Radiology, Niigata Cancer Center Hospital, 2-15-3,
Kawagishi-cho, Niigata 951-8566, Japan.
2 Division of Radiation Oncology, Department of Molecular Genetics, Course for
Molecular and Cellular Medicine, Niigata University Graduate School of Medical
and Dental Sciences, 757, 1-bancho, Asahimachi-dori, Niigata 951-8510,
Japan.
Received June 24, 2002;
accepted after revision August 20, 2002.
Address correspondence to H. Seki.
Abstract
OBJECTIVE. The purpose of this study was to assess perfusion
patterns seen on slow-infusion MR arteriography using the hepatic arterial
infusion system compared with those seen on CT arteriography.
SUBJECTS AND METHODS. In 37 patients with liver metastases who had
implantable port systems for hepatic arterial infusion chemotherapy,
slow-infusion MR arteriography using an infusion rate of 10 mL/hr through an
implantable port and CT arteriography using an injection rate of 0.7 mL/sec
were performed. In 15 of 37 patients, we evaluated enhancement patterns of
tumors of the liver and visceral organs using slow-infusion MR arteriography.
In all 37 patients, we compared slow-infusion MR arteriography with CT
arteriography concerning intra- and extrahepatic perfusion patterns.
RESULTS. On slow-infusion MR arteriography performed 10-20 min after
initiation of infusion, tumors of the liver revealed significant enhancement
with only a slight effect of systemic enhancement. In seven (19%) of 37
patients, intrahepatic distributions on slow-infusion MR arteriography
differed from those on CT arteriography. In eight patients, the patterns of
extrahepatic perfusion into the duodenum and the pancreas head differed on
slow-infusion MR arteriography from those seen on CT arteriography. In
addition, strong artifact caused by platinum coils in the gastroduodenal
artery interfered with the evaluation of perfusion in the area around the
coils on CT arteriography, whereas no imaging artifact was seen on
slow-infusion MR arteriography.
CONCLUSION. We believe that slow-infusion MR arteriography reflects
the actual distribution of infused drugs more accurately than CT
arteriography. When clinical complications occur during treatment,
slow-infusion MR arteriography should be used to assess perfusion
abnormalities.
Introduction
Hepatic arterial infusion chemotherapy using an implantable port system is
now widely used as a regional therapy for unresectable metastases of the liver
[1,
2,
3]. However, a good response to
chemotherapy requires perfusion of the entire liver, and extrahepatic
perfusion of infused drugs is associated with a high incidence of drug
toxicity [4,
5,
6]. In general, hepatic
arterial perfusion scintigraphy and CT arteriography using an indwelling
catheter have proven valuable in evaluating perfusion patterns via the
implantable port system. In hepatic arterial perfusion scintigraphy, slow
infusion of radionuclide through the implantable port resembles the patient's
blood flow distribution more closely than does a large volume of contrast
medium injected during CT arteriography
[7,
8]. However, CT arteriography
has an advantage in confirming the site of the perfusion abnormality over a
radionuclide study because CT arteriography has a higher spatial resolution
[9,
10].
Recently, we reported a new method in assessing perfusion patterns during
hepatic arterial infusion chemotherapy with MR arteriography using an
indwelling catheter [11]. In
this study, MR arteriography was more helpful than CT arteriography in
depicting the perfusion abnormality after transcatheter arterial embolization
using platinum coils because the coils produced no imaging artifact on MR
arteriography, whereas a strong artifact occurred on CT arteriography. For
both MR arteriography and CT arteriography, we used the same high injection
rate of contrast medium. Slow-infusion MR arteriography in which contrast
medium is injected at a slow rate, one that is similar to the rate used during
hepatic arterial perfusion scintigraphy, has been reported recently
[12]. In addition, MR imaging
has higher resolution compared with scintigraphy. To our knowledge, however,
the perfusion patterns seen on slow-infusion MR arteriography have not been
compared with those visible on CT arteriography.
In this study, we evaluated the optimal scanning phase of slow-infusion MR
arteriography and compared perfusion patterns of slow-infusion MR
arteriography with those of CT arteriography. Then we attempted to determine
the value of slow-infusion MR arteriography in assessing perfusion
abnormalities during hepatic arterial infusion chemotherapy.
Subjects and Methods
Between February 2001 and June 2002, 37 patients (27 men and 10 women; age
range, 43-83 years; mean age, 63 years) with unresectable metastases of the
liver who had implantable port systems for hepatic arterial infusion
chemotherapy were included in this prospective study. The examination was
initiated after patients and families were fully informed and gave their
consent. Twenty-two patients had metastases of the liver from colorectal
cancer, 13 from gastric cancer, and two from breast cancer. For catheter
placement, aberrant hepatic arteries were embolized using platinum coils
(FPC35 Pt-Max, Vortex, or both; Boston Scientific Japan, Tokyo, Japan) to
redistribute the hepatic arterial blood flow: the replaced left hepatic artery
arising from the left gastric artery in one patient, the replaced right
hepatic artery arising from the superior mesenteric artery in one, and both in
four. To prevent drug perfusion into the stomach, duodenum, and pancreas, we
embolized the gastroduodenal artery and the right gastric artery using
platinum coils. Occlusion of the right gastric artery was not necessary in 13
patients who had undergone subtotal gastrectomy. In addition, the accessory
left gastric artery arising from the left hepatic artery was embolized in six
patients, the posterosuperior pancreaticoduodenal artery arising from the
proper hepatic artery in four, and the dorsal pancreatic artery arising from
the common hepatic artery in one. In all patients, an end-closed and side-hole
catheter (Anthron PU catheter; Toray Industries, Tokyo, Japan) was inserted
into the hepatic artery using the percutaneous procedure in which the side
hole was placed in the common hepatic artery and the endclosed catheter tip
was fixed in the gastroduodenal artery using platinum coils
[13]. The catheter was
inserted via the left subclavian artery in 27 patients and the inferior
epigastric artery in 10. The catheter was connected with the port
(Soph-A-Port; Sophysa, Orsay, France) that was implanted in the subcutaneous
space.
Slow-infusion MR arteriography was performed 4-12 days after catheter
placement in 31 patients, 3 months in one, 6 months in three, and 1 year in
two. MR imaging was performed using a 1.5-T super-conducting magnet (Magnetom
Vision and Symphony; Siemens, Erlangen, Germany) with a phased array coil. MR
images were obtained in the axial plane using a fast low-angle shot sequence.
When the Magnetom Vision unit was used, the MR imaging parameters were as
follows: TR/TE of 200/4.1, an 80° flip angle, a 94 x 256 matrix, an
8-mm section thickness, a 2-mm intersection gap, and a rectangular field of
view of 219 x 350 mm. When the Magnetom Symphony unit was used, the MR
imaging parameters were as follows: 140/4.75, a 90° flip angle, a 192
x 256 matrix, an 8-mm section thickness, a 2-mm intersection gap, and a
rectangular field of view of 240 x 320 mm. Multisections encompassing
the entire liver were obtained during one breath-hold. A Huber-type
ferromagnetic needle (Coreless needle; Nipro, Osaka, Japan) was used for
access to the implantable port; this port was placed at the left anterior
chest wall or at the lower abdominal wall apart from the upper abdomen. The
contrast solution was composed of 10 mL of 0.5 mol/L gadopentetate dimeglumine
(Magnevist; Nihon Schering, Osaka, Japan) diluted with 5 mL of normal saline.
The solution was continuously injected through the implantable port at a rate
of 10 mL/hr using an infusion pump (Terufusion TE-331; Terumo, Tokyo, Japan).
This pump was placed apart from the magnet unit and wrapped in aluminum foil
(Mitsubishi Foil; Mitsubishi Aluminum, Tokyo, Japan) to protect against the
magnetic force. In 15 patients who initially entered this study, MR imaging
was performed for 30 min to evaluate enhancement patterns of the tumor, liver,
and other organs. Each set of the scans was obtained every 1 min for 20 min
and every 2 min for the subsequent 10 min. The first 10 min, the subsequent 10
min, and the last 10 min were defined as phase 1, phase 2, and phase 3,
respectively. In the next 22 patients, each set of the scans was obtained 11,
13, 15, 17, and 19 min after the initiation of infusion.
CT arteriography was performed 1-7 days before slow-infusion MR
arteriography in 10 patients and 1-3 days after MR arteriography in 27. We
used two types of equipment scanners (LightSpeed QX/i and HiSpeed Advantage
SG; General Electric Medical Systems, Milwaukee, WI). With a detector row
configuration of 4 x 2.5 mm, the LightSpeed QX/i scanner allowed the
acquisition of four images per gantry rotation; 5-mm axial images were
reconstructed at 5-mm intervals. The CT parameters were as follows: a pitch of
6:1, a gantry rotation speed of 0.8 sec, a table speed of 15 mm per gantry
rotation (18.75 mm/sec), 120 kVp, 300 mA, and a field of view of 30 cm. When
the HiSpeed Advantage scanner was used, single-detector scanning was performed
and contiguous axial images were reconstructed with a 7-mm thickness with no
gap. The CT parameters were as follows: a collimation of 7 mm, a table speed
of 7 mm/sec, 140 kVp, 200 mA, and a field of view of 30 cm.
CT images were obtained through the entire liver. The contrast medium was
diluted with normal saline to contain 100 mg I/mL of iopamidol (Iopamiron;
Bracco, Milan, Italy), and the solution was injected via the implantable port
at a rate of 0.7 mL/sec. CT data acquisition began 20 sec after the initiation
of injection.
During hepatic arterial infusion chemotherapy, 5-fluorouracil was given by
continuous infusion at a flow rate of 10 mL/hr: 1000 mg/m2 for 5 hr
every week in 22 patients with metastases of the liver from colorectal cancer,
and 330 mg/m2 for 3 hr every week in 15 patients with gastric or
breast cancer. In addition, in patients with metastases of the liver from
gastric or breast cancer, 2.7 mg/m2 of mitomycin and 30
mg/m2 of epirubicin were given by bolus injection every 2 weeks and
every 4 weeks, respectively.
Image Analysis
In the initial 15 patients, a small (1-5 mm2) region of interest
was placed in an enhanced part of the liver tumor and parenchyma of the liver,
pancreas body, spleen, and kidney on MR images. In slow-infusion MR
arteriography, the enhancement ratio was obtained for each region of interest
as follows: (Sn - S0) x 100 /
S0, where Sn indicates the signal
intensity at each scan phase and S0, the signal intensity
on the image obtained before administration of the contrast medium. We
assessed enhancement patterns of each part and then determined the optimal
scanning phase of slow-infusion MR arteriography.
Image Interpretation
In all 37 patients, intra- and extrahepatic perfusion patterns were
evaluated using slow-infusion MR arteriography and CT arteriography.
Slow-infusion MR arteriography was interpreted with no knowledge of the
results of CT arteriography. Intrahepatic perfusion patterns were categorized
as homogeneous distribution, lobar hyperperfusion, segmental hyperperfusion,
peripheral hyperperfusion, or a peripheral perfusion defect. Homogeneous
distribution was defined as uniform perfusion in the entire liver. Lobar and
segmental hyperperfusion was located according to Couinaud segmental anatomy
and was defined as an area of hyperperfusion compared with the remainder of
the liver parenchyma. Peripheral hyperperfusion was characterized as a small
peripheral area of hyperperfusion without a more proximal tumor nodule. A
peripheral perfusion defect was defined as a peripheral area of the liver with
no perfusion. When perfusion abnormalities were found, preoperative
arteriography was reviewed, and postoperative angiography was performed using
the implantable port or transfemoral arterial approach to determine the cause
of these abnormalities.
Results
Enhancement Patterns of the Liver Tumor and Visceral Organs
The enhancement ratios of the liver tumor and liver parenchyma, pancreas
body, spleen, and kidney on slow-infusion MR arteriography in the initial 15
patients are shown in Figure 1.
Tumors of the liver were recognized as hypointense areas in the liver on
unenhanced MR images in all patients. The liver tumors showed rapid
enhancement on MR arteriography during the latter half of phase 1 and
significant enhancement during phases 2 and 3 (Fig.
2A,
2B,
2C). In most cases, the
enhancement ratio of the liver parenchyma was less than that of tumors of the
liver in phases 2 and 3. Then, enhancement of tumors of the liver was clearly
revealed in phases 2 and 3. On the other hand, although enhancement of the
pancreas body and the spleen on MR arteriography was slight, the kidney had
gradual enhancement by systemic circulation of the contrast medium. In
particular, the renal parenchyma was significantly enhanced during phase 3 MR
arteriography compared with phases 1 and 2 (Fig.
2A,
2B,
2C). Regarding these findings,
we believe that slow-infusion MR arteriography during phase 2 is suitable for
exhibiting the perfusion pattern via the implantable port system with only a
slight effect of systemic enhancement.
View larger version (164K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B. —Liver metastases from sigmoid colon cancer in 48-year-old
man. Slow-infusion MR arteriograms obtained during phase 2 (B) and
phase 3 (C) show that enhancement of liver tumor (arrow) is
greater than that of liver parenchyma. Kidney is significantly enhanced in
phase 3 image compared with phase 2 image.
|
|
View larger version (160K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C. —Liver metastases from sigmoid colon cancer in 48-year-old
man. Slow-infusion MR arteriograms obtained during phase 2 (B) and
phase 3 (C) show that enhancement of liver tumor (arrow) is
greater than that of liver parenchyma. Kidney is significantly enhanced in
phase 3 image compared with phase 2 image.
|
|
Intrahepatic Perfusion Patterns
Intrahepatic perfusion patterns on slow-infusion MR arteriography and CT
arteriography are presented in Table
1. In seven (19%) of 37 patients, intrahepatic distributions on
slow-infusion MR arteriography differed from those seen on CT arteriography.
Of 23 patients with homogeneous distribution on CT arteriography, lobar
hyperperfusion was revealed on slow-infusion MR arteriography in three
patients (Fig. 3A,
3B). Of seven patients with
lobar hyperperfusion on CT arteriography, the perfusion pattern on
slow-infusion MR arteriography changed to hyperperfusion of the opposite lobar
side in one patient and homogeneous distribution in another. Of three patients
with segmental hyperperfusion on CT arteriography, homogeneous distribution
was seen on slow-infusion MR arteriography in one patient (Fig.
4A,
4B). In one patient with
peripheral hyperperfusion, an area of hyperperfusion decreased in size on
slow-infusion MR arteriography compared with CT arteriography. On the other
hand, in three patients with a peripheral perfusion defect, the same pattern
was seen on slow-infusion MR arteriography and CT arteriography.
On angiography, occlusion or stenosis of the hepatic artery was not seen in
any of the 37 patients. In two patients with lobar hyperperfusion of the right
hepatic lobe on slow-infusion MR arteriography and CT arteriography, occlusion
of the right branch of the portal vein was found to result from compression by
the hepatic tumor. In three patients with a peripheral perfusion defect on
slow-infusion MR arteriography and CT arteriography, collateral blood supply
to the tumor of the liver was revealed. In the remaining patients, no finding
that would cause intrahepatic perfusion abnormalities was depicted on
angiography.
Extrahepatic Perfusion Patterns
Of 37 patients, enhancement of the duodenum and the pancreas head was seen
in eight patients on slow-infusion MR arteriography and in 10 on CT
arteriography. Of eight patients with perfusion in this area on slow-infusion
MR arteriography, the enhanced area increased in size on CT arteriography in
four patients and could not be evaluated using CT arteriography in two
patients because of a strong artifact caused by platinum coils in the
gastroduodenal artery (Fig. 5A,
5B). Of 10 patients with
enhancement in this area on CT arteriography, no extrahepatic perfusion was
seen on slow-infusion MR arteriography in four patients (Fig.
6A,
6B), whereas vessels of the
duodenum and the pancreas head were found on postoperative angiography using
the implantable port in two of these four patients. In addition, on CT
arteriography, an imaging artifact was caused around the gastroduodenal artery
by platinum coils. This artifact interfered to various degrees with the
assessment of the perfusion into the duodenum and the pancreas head. On the
other hand, on slow-infusion MR arteriography, the platinum coils produced no
magnetic susceptibility artifact and did not affect the quality of the MR
images (Figs. 5A,
5B and
6A,
6B).
View larger version (167K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A. —Liver metastases from rectal cancer in 65-year-old woman with
gastroduodenal toxicity. On CT arteriogram, imaging artifacts
(arrows) caused by platinum coils in gastroduodenal artery are too
strong to evaluate perfusion in area around coils.
|
|
View larger version (140K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B. —Liver metastases from rectal cancer in 65-year-old woman with
gastroduodenal toxicity. Slow-infusion MR arteriogram obtained during phase 2
reveals enhancement of duodenum and pancreas head (arrows). Platinum
coils (arrowhead) can be seen as signal loss.
|
|
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6A. —Liver metastases from ascending colon cancer in 66-year-old
woman with no clinical problem during treatment. CT arteriogram shows
enhancement of duodenal wall and pancreas head (arrows).
|
|
View larger version (128K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B. —Liver metastases from ascending colon cancer in 66-year-old
woman with no clinical problem during treatment. On slow-infusion MR
arteriogram obtained during phase 2, no enhancement can be seen in this
area.
|
|
During hepatic arterial infusion chemotherapy, gastroduodenal toxicity that
required dosage reduction of the drug was seen in four of eight patients with
enhancement of the duodenum and the pancreas head on slow-infusion MR
arteriography and two of 10 patients with perfusion in this area on CT
arteriography. In four patients with enhancement in this area seen on CT
arteriography but not on slow-infusion MR arteriography, no clinical problem
was seen. In addition, in three of four patients in whom perfusion in this
area had decreased in size on slow-infusion MR arteriography compared with CT
arteriography, gastroduodenal toxicity was mild or absent.
Discussion
When a patient is undergoing hepatic arterial infusion chemotherapy, the
evaluation of the perfusion patterns via the implantable port system is
important for effective treatment. Several authors have reported that CT
arteriography is useful in confirming the sites of perfusion abnormalities
through an implantable port system because CT arteriography has good spatial
resolution [9,
10]. However, perfusion
patterns on CT arteriography using a high injection rate may differ from those
of infused drugs. A report mentioned that intra- and extrahepatic perfusion
could not be evaluated on CT arteriography using a low infusion rate because
of poor contrast enhancement
[10].
In our study, various perfusion patterns could be clearly depicted on
slow-infusion MR arteriography using a low infusion rate that is similar to
that used for chemotherapy. Intrahepatic perfusion patterns on slow-infusion
MR arteriography differed from those on CT arteriography in 19% of our
patients. In these cases, no abnormal finding was discovered on angiography.
These differences may be caused by the laminar flow that results from poor
mixing of infused materials with hepatic arterial blood
[7,
14]. These findings suggest
that slow-infusion MR arteriography reflects the actual distribution of
infused drugs more accurately than CT arteriography.
Slow-infusion MR arteriography may lead to an enhancement effect by
systemic circulation of the contrast medium. In this study, on slow-infusion
MR arteriography performed during phase 2, tumors of the liver revealed
significant enhancement with only a slight enhancement effect of extrahepatic
organs. Therefore, we believe that slow-infusion MR arteriography should be
undertaken 10-20 min after injection.
In patients receiving hepatic arterial infusion chemotherapy,
gastroduodenal toxicity is one of the major complications
[5,
6]. Several investigators have
mentioned that gastroduodenal toxicity may result from misperfusion of infused
drugs into the stomach and the duodenum through minor and accessory vessels
[6]. Therefore, the assessment
of misperfusion into the stomach and the duodenum is important in predicting
gastroduodenal toxicity. In our study, perfusion into the duodenum and the
pancreas head was overestimated in several cases on CT arteriography, whereas
perfusion in these areas was not overestimated on slow-infusion MR
arteriography. In addition, extrahepatic perfusion in this area was associated
more closely with gastroduodenal toxicity seen on slow-infusion MR
arteriography than that seen on CT arteriography. On the other hand, a strong
artifact caused by platinum coils in the gastroduodenal artery interfered with
the evaluation of perfusion into the duodenum on CT arteriography, whereas no
imaging artifact related to magnetic susceptibility was seen on slow-infusion
MR arteriography because platinum is a nonferromagnetic material
[11]. We think that
slow-infusion MR arteriography is superior to CT arteriography in assessing
misperfusion into the duodenum.
In 81% of the patients in our study, intrahepatic perfusion patterns on CT
arteriography were identical to those on slow-infusion MR arteriography. In
addition, conventional CT is generally used to evaluate therapeutic response
of intra- and extrahepatic diseases during chemotherapy. Therefore, we believe
that CT arteriography with conventional CT is preferable for follow-up studies
in patients with no clinical problems instead of slow-infusion MR
arteriography or hepatic arterial perfusion scintigraphy.
In conclusion, slow-infusion MR arteriography is a reliable examination in
exhibiting perfusion patterns similar to those of infused drugs. When the drug
distribution is initially evaluated after catheter placement or when clinical
complications occur during treatment, slow-infusion MR arteriography should be
used to assess perfusion abnormalities. We believe that slow-infusion MR
arteriography in combination with CT arteriography is useful in assessing
perfusion abnormalities during hepatic arterial infusion chemotherapy.
References
- Kemeny N, Daly J, Reichman B, Geller N, Botet J, Oderman P.
Intrahepatic or systemic infusion of fluorodeoxyuridine in patients with liver
metastases from colorectal carcinoma. Ann Intern Med
1989;107:459
-465
- Rougier P, Laplanche A, Huguier M, et al. Hepatic arterial infusion
of floxuridine in patients with liver metastases from colorectal carcinoma:
long-term results of a prospective randomized trial. J Clin
Oncol 1992;10:1112
-1118[Abstract]
- Arai Y, Inaba Y, Takeuchi Y, Ariyoshi Y. Intermittent hepatic
arterial infusion of high-dose 5FU on a weekly schedule for liver metastases
from colorectal cancer. Cancer Chemother Pharmacol
1997;40:526
-530[Medline]
- Seki H, Kimura M, Yoshimura N, Yamamoto T, Ozaki K, Sakai K.
Development of extrahepatic arterial blood supply to the liver during hepatic
arterial infusion chemotherapy. Eur Radiol
1998;8:1613
-1618[Medline]
- Pozniak MA, Babel SG, Trump DL. Complications of hepatic arterial
infusion chemotherapy. RadioGraphics
1991;11:67
-79[Abstract]
- Hohn DC, Stagg RJ, Price DC, Lewis BJ. Avoidance of gastroduodenal
toxicity in patients receiving hepatic arterial 5-fluoro-2'-deoxyuridine.
J Clin Oncol
1985;3:1257
-1260[Abstract/Free Full Text]
- Bledin AG, Kim EE, Chuang VP, Wallace S, Haynie TP. Changes of
arterial blood flow patterns during infusion chemotherapy, as monitored by
intra-arterially injected technetium 99m macroaggregated albumin.
Br J Radiol
1984;57:197
-203[Abstract]
- Civelek AC, Sitzmann JV, Chin BB, Venbrux A, Wagner HN Jr, Grochow
LB. Misperfusion of the liver during hepatic artery infusion chemotherapy:
value of preoperative angiography and postoperative pump scintigraphy.
AJR
1993;160:865
-870[Abstract/Free Full Text]
- Seki H, Kimura M, Kamura T, Miura T, Yoshimura N, Sakai K. Hepatic
perfusion abnormalities during treatment with hepatic arterial infusion
chemotherapy: value of CT arteriography using an implantable port system.
J Comput Assist Tomogr
1996;20:343
-348[Medline]
- Morimoto M, Satake M, Sekiguchi R, Haruno M, Moriyama N. Optimal
infection protocol for CT evaluation during hepatic arterial infusion
chemotherapy. Invest Radiol
1999;34:744
-750[Medline]
- Seki H, Ozaki T, Takano T, et al. MR arteriography using an
implantable port system: a new method in assessing perfusion abnormalities
during hepatic arterial infusion chemotherapy. J Comput Assist
Tomogr 2000;24:890
-892[Medline]
- Takizawa K, Saeki M, Imamura K, et al. Ultraslow infusion dynamic
MRI using an infusion pump: a new method for the evaluation of drug
distribution in arterial infusion chemotherapy [in Japanese].
Nippon Acta Radiol
2001;61:246
-248
- Arai Y, Inaba Y, Takeuchi Y. Interventional techniques for hepatic
arterial infusion chemotherapy. In: Castañeda-Zúñiga WR,
ed. Interventional radiology, 3rd ed. Baltimore:
Williams & Wilkins, 1997:192
-205
- Lutz RJ, Miller DL. Mixing studies during hepatic artery infusion
in an in vitro model. Cancer
1988;62:1066
-1073[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
) exceeded that of
liver parenchyma (
), kidney (•), spleen (x), and pancreas
body (
) during phase 2 (10-20 min).
