Collagen Chemoembolization: Pharmacokinetics and Tissue -Diamminedichloroplatinum(II) in Porcine Liver

Collagen Chemoembolization: Pharmacokinetics and Tissue
Tolerance of cis-Diamminedichloroplatinum(II) in Porcine Liver
and Rabbit Kidney
John R. Daniels, Mark Sternlicht and AnnaMarie Daniels
Cancer Res 1988;48:2446-2450.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/48/9/2446
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected]
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected]
Downloaded from cancerres.aacrjournals.org on June 9, 2014. © 1988 American Association for Cancer Research.
[CANCER RESEARCH 48, 2446-2450, May 1, 1988]
Collagen Chemoembolization: Pharmacokinetics and Tissue Tolerance of
cif-Diamminedichloroplatinum(II) in Porcine Liver and Rabbit Kidney1
John R. Daniels,2Mark Sternlicht, and AnnaMarie Daniels
Kenneth Norris, Jr., Research Institute and Cancer Hospital, University of Southern California, Los Angeles 90033 [J. R. D.J, and Target Therapeutics, Inc.,
Los Angeles, California 90025 [J. R. />., M. S., A. D.J
ABSTRACT
Pharmacokinetics of Chemoembolization with a fibrous collagen carrier
was studied in rabbit kidney and porcine liver models. Cisplatin (1 mg/
ml) Chemoembolization of liver and kidney was compared with i.v. and
intraarterial cisplatin infusion. Tissue platinum concentration [Pt] was
measured at 2.5 h by atomic absorption spectrometry. Renal platinum
retention and Angiostat (collagen for embolization) concentration were
linearly related (r = 0.87, p < 0.001). At 10 mg/ml collagen for emboli
zation, chemoembolized kidney [Pt| was 220 ±SO (SE; u = 4) times
contralateral kidney [Pt], and 62 and 23 times renal [Pt] by i.v. and
intraarterial infusion, respectively. At 10 mg/ml collagen for emboliza
tion, chemoembolized
liver [Pt] was 2 times hepatic |Pt] by i.v. and
intraarterial infusion. Since hepatic tumor vasculature is end arterial,
Chemoembolization should yield high [Pt] in tumor (as in kidney) but
lower levels in normal liver.
INTRODUCTION
Chemoembolization refers to the intraarterial coadministration of chemotherapeuiie and vascular occlusive agents to treat
malignant disease. Vascular occlusion prolongs dwell time of
the antineoplastic agent within the tumor, increasing first pass
fractional extraction of drug. Additional therapeutic effect may
be obtained by tumor ischemia. In liver, Chemoembolization
offers a special opportunity for selective drug delivery. The
blood supply of tumor in liver is principally arterial, while
hepatocytes are also bathed by portal blood flow (1). The portal
blood flow supports hepatocytes during ischemia and clears
drug from hepatocytes but not from areas of malignant disease
in liver.
Chemoembolization has been studied in several organ sites,
using a variety of embolie materials and drugs (2-7). The
majority of clinical hepatic Chemoembolization experiences
have used delivery of drug within biodegradable microspheres
and microcapsules (8-13). Other embolie materials which have
been explored include coaxial balloon catheters (14), Gel foam.
IvaIon, and stainless steel coils (15). To avoid tissue infarction,
embolie materials are generally either large enough to occlude
proximal to preexisting collateral vessels or are labile and
produce only transient occlusion.
We have developed a collagen particle [Angiostat (CFE3)]
specifically to address the requirements for efficient Chemoem
bolization of malignant disease. Design considerations included
degree of dwell time prolongation, distal penetration within the
vascular bed, interaction between drug and particle, degree of
ischemie injury, and reversibility of vascular occlusion. The
Received 6/24/87; revised 12/1/87; accepted 1/27/88.
The costs of publication of this article were defrayed in pan by the payment
of page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1Supported in part by Bristol Myers Company, Collagen Corporation, and Eli
Lilly Company.
2To whom requests for reprints should be addressed, at 2100 S. Sepulveda
Boulevard, Los Angeles, ÇA9002S.
3The abbreviations used are: CFE, collagen for embolization (Angiostat;
Target Therapeutics, Los Angeles, CA); HPLC, high-performance liquid chromatography; AAS, atomic absorption spectrometry; CDDP, cis-diamminedichloroplatinum(II); DDTC, diethyldithiocarbamate; i.a., intraarterial.
collagen particles are 5- x 75-/tm fibers precipitated from en
zyme solubilized bovine dermal collagen and cross-linked with
glutaraldehyde.
We have previously characterized the response of canine liver
to arterial embolization with this collagen (16). The particles
penetrate distal to intrahepatic arterial collateral channels to
the pa-capillary sphincter. Hepatic ischemie injury is transient
and is reversed at 48 to 72 h by recanalization. Over the ensuing
2 to 3 months the collagen is removed, and normal vascular
anatomy is restored.
In this communication we examine the consequences of using
collagen as a drug delivery system. For cisplatin, we establish
that drug binding to collagen is limited. A renal Chemoemboli
zation model is used to demonstrate the dose effect relationship
between collagen concentration and drug retention within an
end-artery system, while similar studies in porcine liver show
only modest increase in normal hepatocyte drug deposition.
The porcine liver is used to establish safe concentration ranges
for Chemoembolization in humans.
MATERIALS AND METHODS
Reagents. HPLC grade organic solvents were obtained from J. T.
Baker Chemical Co. All other chemical reagents and standards were
purchased from Sigma Chemical Co. unless otherwise specified.
Analytic Methods. Platinum levels in tissue homogenates, fluids, and
in ultrafiltrates of fluids were determined by flamelcss AAS, utilizing a
IVrkin Elmer Model 2380 atomic absorption spectrophotometer and a
heated graphite atomizer with an HGA-400 programmer. Sample aliquots of 20 fi\ were introduced and carried through a four-stage furnace
heating program. Liquid phase evaporation was achieved by tempera
ture ramp from ambient to 90°Cover 45 s and held for another 60 s.
Complex sample matrices were then volatilized in a charring step by
ramping the furnace temperature to 1450°Cover 40 s and holding for
30 s. During this 70-s interval, automatic base-line correction occurred
at 62 s, the argon gas flow (300 ml/min) was interrupted at 65 s, and a
spectrophotometer-read cycle was initiated at 69 s, i.e., 8, 5, and 1 s
prior to atomization, respectively. The atomization step used "max
power" (zero ramp time), heating to 2650°Cwhich was maintained for
5 s under gas flow arrest. Integrated absorbance (peak area) was
obtained during atomization, and sample platinum concentration was
determined from a machine calibration curve. A final 6-s 2650°Cburnoff under resumed gas flow conditions was added to reduce residual
analyte between samples.
Aqueous platinum standards were prepared from a standard solution
(1000 iig/ml in 0.01 N HC1) by serial dilution with 0.2% HNO3.
Preweighed tissue samples were homogenized in a glass homogenizer
with a known volume of 0.25% Triton X (New England Nuclear) and
aliquots were taken for triplicate ASS determinations. Tissue concen
trations are expressed as ^g platinum per g wet weight of tissue.
Reactive cisplatin (Bristol Laboratories) in fluids was estimated by
measuring the DDTC derivative, using HPLC. The sample extraction
procedure was adapted from Andrews et al. (17). To 0.5-ml sample
aliquots were added 5 n\ of 0.55 mg/ml NiCl2, internal standard, and
50 fil freshly prepared 10% DDTC in 0.1 N NaOH (w/v). The mixture
was incubated for 30 min at 37°C,then chilled on ice. The platinum
and nickel DDTC derivatives were extracted with 0.2 ml CHC13 by
thorough vortex mixing for 1 min. Separation into aqueous and organic
layers was obtained by centrifugation at 1000 x g for 10 min. The
2446
Downloaded from cancerres.aacrjournals.org on June 9, 2014. © 1988 American Association for Cancer Research.
PHARMACOKINETICS
OF COLLAGEN CHEMOEMBOLIZATION
chloroform (bottom) layer was filtered (0.22 Aim-filler) and a 20 ^1
aliquot was injected into the HPLC.
Aqueous samples were chromatographed on a Waters HPLC system
consisting of 2 Model 501 solvent delivery pumps, a Model U6K
injector, Model 680 automated gradient controller, Model 441 absorb
ante detector, Model 740 data module, and a 3.9-mm x 30-cm C,KfiBondapak column. A 4:1 methunol:water (v/v) mobile phase was
delivered at 1 ml/min, and column effluent was monitored at 254 nm
with a sensitivity of 0.01 absorbance unit full scale. Retention times for
platinum and nickel DDTC complexes were 7.0 and 8.0 min, respec
tively. Sample concentrations were determined from a standard curve
of the ratio of peak areas for the Pt(DDTC)2 and Ni(DDTC)2 peaks.
The standard curve was linear (r = 0.999) over the range of 1 to 10 /ig/
ml CDDP.
Drug Dialysis. To measure potential binding between cisplatin and
Angiostat, CFE (Target Therapeutics), the rate of drug efflux from a
dialysis bag containing either drug alone or drug and collagen was
determined. Drug and drug-CFE mixtures were prepared in diatrizoate
meglumine 66% and diatrizoate sodium 10% (Hypaque-76, WinthropBreon Laboratories). The iodinated contrast media were included to
model conditions of in vivo administration. Dialysis tubing (Union
Carbide; putative pore size, 12,000 to 14,000 daltons) was prepared by
boiling in multiple changes of water with EDTA. Dialysis bags con
tained 1 ml of the mixture to be studied. Final concentrations were 10
mg/ml CFE, l mg/ml CDDP, and 54% diatrizoate. Control bags were
prepared without collagen. Dialysis with stirring was with 200 ml
phosphate-buffered saline, pH 7.4, at 37°C.Dialysate was sampled at
0, 10, 30, 60, 120,180, and 240 min. CDDP levels were determined by
HPLC.
This design approximates the transfer of a solute within a closed
two-compartment system such that equilibrium is approached at a
single exponential rate. Half-times were determined from the slope of
the natural log of the difference between the concentration-time data
and the expected equilibrium concentrations, namely, the total amount
of drug in the system divided by the system volume. Because some
measurements exceeded the expected equilibrium value, slightly higher
asymptotes were chosen for calculations.
Renal Chemoembolization. Rabbit kidney was used to model chemoembolization in an end-artery system and to measure the effect of
collagen concentration upon tissue chemotherapy retention. Tissue
platinum deposition was compared following administration of 0.4 mg
CDDP by i.v. infusion (3 rabbits), left renal i.a. infusion (4 rabbits),
and left renal Chemoembolization with variable concentrations of CFE
(24 rabbits). Male New Zealand White rabbits weighing 4 to 5 kg were
preanesthetized with intramuscular injections of 30 mg/kg sodium
pentobarbital (Nembutal; Abbott Laboratories) and 30 mg/kg ketamine
HC1 (Bristol Laboratories). Anesthesia was maintained with i.v. sodium
pentobarbital in a running normal saline line. The animals in the i.v.
treatment group received 0.4 ml of l mg/ml CDDP in 76% iodinated
contrast infused into the ear vein. In the remaining animals, the right
femoral artery was dissected free, an arteriotomy was performed, and a
graded stiffness infusion catheter (Tracker infusion catheter; Target
Therapeutics) and 0.018-inch guide wire (Hi-Torque floppy guide wire;
Advanced Cardiovascular Systems) were introduced. The catheter was
placed in the left renal artery under fluoroscopic control, and 0.4 ml of
1 mg/ml CDDP with 0, 1, 2, 3, 5, 7.5, or 10 mg/ml CFE in 76%
iodinated contrast were delivered into four animals each. The effect of
non-ionic contrast (Hexabrix; Mallinckrodt) was evaluated under sim
ilar conditions in four animals. Animals were sacrificed 2.5 h following
administration and the right and left kidneys and liver were removed
and weighed for total platinum determinations by AAS.
Hepatic Chemoembolization. Duroc and Hampshire pigs weighing
20 to 40 kg were preanesthetized with an i.m. injection of 0.05 mg/kg
atropine sulfate (Invenex Laboratories) and 0.6 mg/kg acepromazine
(Ceva Laboratories), followed after 10 min with 20 mg/kg ketamine
HC1 and 0.4 mg/kg xylazine (Rompun; Miles Laboratories) adminis
tered i.m. Anesthesia was maintained with i.v. sodium pentobarbital in
a running normal saline line. The right femoral artery was dissected
free and a distal ligature was placed. An arteriotomy was performed
and a graded stiffness infusion catheter and 0.016-inch Taper steerable
guide wire (Target Therapeutics) were introduced. At this point, 4000
units heparin (Invenex Laboratories) and 250 mg ampicillin sodium
(Wyeth Laboratories) were administered by i.a. injection. The catheter
was placed within a segmental hepatic artery under fluoroscopic guid
ance. Chemoembolic mixtures were prepared by solubili/in*; CDDP in
76% iodinated contrast at 2 x final concentration and mixing 1:1 (v/v)
with CFE at 20 mg/ml by exchange between syringes, using a 3-way
stopcock. The chemoembolic mixture was administered until retrograde
flow was observed. The catheter was then removed, a proximal ligature
was placed for hemostasis, and the wound site was closed with 3-0 silk.
After Chemoembolization, antibiotic coverage consisted of 1300 units/
kg/day penicillin G procaine (E. R. Squibb & Sons) by i.m. injection
for 3 days. Blood samples were drawn for hematological and serum
chemistry evaluations at base line and at 3, 7, 14, and 21 days following
Chemoembolization. Animals were sacrificed on day 21 for gross ex
amination and for histológica! evaluation of liver, gallbladder, duo
denum, spleen, pancreas, and kidney. Percentage of liver damage was
estimated by visualization of gross specimens and, wherever possible,
by weight. Two additional animals underwent hepatic Chemoemboli
zation with 1 mg/ml CDDP and 10 mg/ml CFE with repeat Chemoem
bolization at 2 weeks. These animals were sacrificed 5 weeks following
initial Chemoembolization, and cholangiography was performed to
evaluate patency of the biliary tree.
Plasma clearance of total platinum and acute (2.5 h) tissue deposition
were determined in nine Duroc pigs weighing 23 to 27 kg. Platinum
levels were determined following administration of 0.4 mg/kg (1 mg/
ml) CDDP in three animals each by i.v. infusion, hepatic artery infusion,
and hepatic Chemoembolization with 4 mg/kg (10 mg/ml) CFE. All
drug mixtures were prepared in 66% diatrizoate meglumine and 10%
diatrizoate sodium and were delivered at approximately 2 ml/min.
Blood samples were obtained from a venous line immediately following
drug delivery and at 10, 20, 30, 40, 50, 60, 90, 120, and 150 min
thereafter. Serum was separated immediately at 1000 x g and removed
for AAS measurement of total platinum. Animals were sacrificed at 2.5
h, and liver, spleen, pancreas, and both kidneys were removed, weighed,
and prepared for [Pt] determination.
RESULTS
In Vitro Drug Binding. Apparent drug binding by CFE was
limited. The concentration-time curve for appearance of drug
into dialysate was slightly retarded in the presence of CFE in
the dialysis bag. The half-times for CDDP alone and for drug
plus CFE were 35 and 64 min, respectively. At all time points
examined, the presence of CFE affected dialysate drug concen
tration by no more than 21%, with an average difference of
11%.
Renal ChemoembolizatÃ-on. Chemoembolization with CFE
and cisplatin led to substantial increases in drug retention by
the target kidney. Renal and hepatic mean platinum levels
following i.v. CDDP infusion, left renal i.a. infusion, and Chem
oembolization are shown in Table 1. Systemic CDDP admin
istration produced an equivalent distribution between left and
right kidneys with 0.5% of the administered dose recovered in
each kidney at 2 h. Following i.a. infusion, the target left kidney
retained 1.4% of the administered dose with a mean platinum
level 5.4 times that of contralateral kidney and 3 times the
concentration after i.v. administration. In Chemoembolization
groups, increased target tissue drug retention and decreased
delivery to nontarget areas was a function of CFE concentration
(Fig. 1). Regression analysis and calculation of the F statistic
indicate a direct linear relationship (r = 0.87; P < 0.001)
between target tissue platinum and collagen concentration. At
10 mg/ml CFE, the chemoembolized kidney retained 38% of
the administered dose, 220 times the concentration in the
contralateral kidney, and 62 and 23 times renal levels obtained
following i.v. and i.a. infusion, respectively. Similar results were
2447
Downloaded from cancerres.aacrjournals.org on June 9, 2014. © 1988 American Association for Cancer Research.
PHARMACOKINETICSOF COLLAGEN CHEMOEMBOLIZATION
Table 1 Left renal chemoembolization (0.4 ml of 1 mg/ml CDDP and variable CFE)
SE)Right
Angiostatmg/ml0
deposition 0»gPt/g wet wt) (mean ±
(R)0.157
kidney
kidney ratio
(mean
SE)1.03±
(L)0.163
kidney
0.0200.431
±
±0.1541.49
±0.2181.93
±0.3905.96
1.246.67
±
1.167.79
±
1.3110.1
±
±0.540Liver0.126
±0.0150.102
(i.V.)0(i.a.)12357.510(mg)000.40.81.2234N34444444Tissue
0.0260.162
±
0.0440.1
±
±0.0110.076
16
0.0150.079
+
0.0080.073
±
0.0050.050
±
±0.006Left
0.0070.063
+
0.0190.057
+
0.0180.064
+
0.0080.054
±
0.0080.049
±
0.0050.039
±
0.0060.022
±
±0.001L:R
±0.0395.44
2.9710.1
±
1.2816.6
±
3.2384.4
±
22.483.6
±
8.66110
±
±23.8221
±51.3
10.0
IV (n=3)
IA (n=3)
CE (n=3)
8.0
6.0
*
4.0
1
:
E
O)
2.0
•
L. Kidney
o R. Kidney
o Liver
0.2
I
E
S
a.
O.I
o.o
Liver
0.0
2468
IO
Collagen (mg/ml)
Fig. I. Tissue platinum deposition 2.5 h following left renal chemoemboliza
tion with 0.4 ml of 1 mg/ml cisplatin and variable collagen (CFE) in the rabbit.
The efficiency with which cisplatin is retained within the left kidney and reduced
within right kidney and liver, is a function of CFE concentration.
Kidney
Pancreas
Spleen
Fig. 2. Porcine tissue platinum deposition 2.5 h following administration of
0.4 mg/kg (1 mg/ml) CDDP by i.v. infusion, hepatic i.a. infusion, and hepatic
chemoembolization (CE) with 10 mg/ml CFE. Relative to i.v. infusion, hepatic
deposition was unchanged by i.a. infusion, and doubled following chemoemboli
zation. Modest changes in nontarget tissue levels were also observed.
Table 2 Cisplatin pharmacokinetics: effect of i.v. infusion, hepatic i.a. infusion,
and hepatic chemoembolization (mean ±SE)
clearanceRouteInfusion Plasma Pt
Pt deposition
wt)Liver0.457
Gig Pt/g wet
Pt Cone.
(ng/ml)
Pt/ml)3.53
Gig
+ 0.092
+ 0.106
i.V.
±2.38
±37.6
0.457 ±0.081 0.859 ±0.027
Infusion i.a.
3 1.09 ±0.24 46.4 ±4.5
0.907 ±0.296Kidney0.660
0.437 ±0.034
Chemoembo
0.40 ±0.03AUC"(MK/ml/min)88.6
37.6 ±1.3Tissue
3Peak
lizationN3
°AUC, area under the plasma clearance curve.
ä-ä-B-n-
obtained by using ionic and non-ionic contrast media.
Hepatic Chemoembolization. Drug distribution following por
cine hepatic chemoembolization with cisplatin is presented in
Table 2 and in Figs. 2 and 3. No difference in normal hepatic
tissue platinum levels was observed between i.v. and hepatic i.a.
infusion groups, whereas chemoembolization resulted in a 2fold increase in hepatic platinum deposition. At 10 mg/ml
collagen, hepatic retention of platinum at 2.5 h was 4% of the
administered dose for i.v. and i.a. infusions, and 7% following
chemoembolization. The mean coefficient of variance between
specimens obtained from the same liver was about 10% for i.v.
and i.a. infused groups and 30% for chemoembolized liver.
Compared with i.v. administration, chemoembolization
re
sulted in a 10 to 30% reduction of tissue platinum in noninfused
organs. Splenic and pancreatic tissue levels following hepatic
chemoembolization were slightly higher than those observed
following i.v. infusion, probably reflecting inadvertent chem
oembolization of those organs. A rapid early serum half-time
of 17 min was observed for both i.v. and i.a. groups and not
seen following chemoembolization. Consistent with the reduc-
60
90
Time
(min)
120
150
Fig. 3. Porcine plasma platinum clearance curves following cisplatin by i.v.
infusion (D), hepatic i.a. infusion (O), and hepatic chemoembolization (A). Peak
levels were substantially reduced by hepatic regional delivery. Late phase clearance
was similar for all groups.
tion of platinum deposition in nontarget organs was a reduction
by 60% of the area under the plasma clearance curve for total
platinum following chemoembolization as compared with i.v.
delivery. Late phase clearance was similar for all groups.
Dose-effect relationships for liver toxicity were examined in
the porcine liver model for various concentrations of cisplatin
at 10 mg/ml CFE. When present, liver damage at day 21 was
characterized by necrosis which ranged from focal to massive,
intrahepatic fibrous tissue, and adhesions to adjacent structures.
Lesions were often encapsulated with a friable core. Liver
damage was scored on the basis of percentage of liver with
macroscopic damage. Increasing concentrations of CDDP re
sulted in a higher frequency and larger area of tissue damage
(Fig. 4).
Transient liver dysfunction measured by serum enzyme levels
was generally resolved by 1 to 3 weeks. Minimal elevations in
WBC and decreases in platelets were resolved by day 21. The
2448
Downloaded from cancerres.aacrjournals.org on June 9, 2014. © 1988 American Association for Cancer Research.
PHARMACOKINETICS OF COLLAGEN CHEMOEMBOLIZATION
20
LU
CA
6 n-3
10-
10
15
Cisplatin (mg/ml)
Fig. 4. Hepatic tolerance to cisplatin chemoembolization with Angiostat (10
mg/ml). The mean percentage of liver with macroscopic damage is shown as a
function of cisplatin concentration.
severity of serum enzyme abnormalities correlated well with the
degree of liver damage present at necropsy.
Hepatic chemoembolization was generally tolerated without
significant liver damage at up to 7 mg/ml CDDP.
Repeat chemoembolization 2 weeks following initial hepatic
administration of CFE:CDDP 10:1 mg/ml resulted in minimal
liver necrosis in two animals. In one case, transient 5- and 10fold elevations in total bilirubin were observed 1 week following
initial and repeat procedures, respectively. A postmortem cholangiogram showed stenotic areas to be present throughout the
biliary tree.
Three animals were not included in the results due to inad
vertent chemoembolization of the gastroduodenal artery which
became apparent at necropsy. These animals survived or were
sacrificed less than 1 week following chemoembolization. Se
vere gastroduodenal infarction was present in all cases.
DISCUSSION
Chemoembolization provides both enhanced confinement of
drug within the tumor bed and direct ischemia due to vascular
occlusion. The relative advantage of regional drug delivery is a
function of first pass extraction and of systemic half-life of the
drug (18). For several chemotherapeutics, only limited thera
peutic advantage is gained by intraarterial as compared to
systemic administration (18-20). Chemoembolization increases
first pass extraction by increasing drug dwell time. As regional
delivery becomes more efficient, however, local toxicity may
become the dose limiting factor. Thus, in hepatic artery chemo
embolization the relative distribution of drug between normal
hepatocytes and cancer in liver may determine the therapeutic
index and ultimate effectiveness. After hepatic artery chemoem
bolization, we hypothesize that portal blood flow will selectively
reduce retention of drug by normal liver as compared with
malignant disease in liver where the portal system has been
displaced.
In vitro studies indicate that CDDP is not significantly bound
by CFE, making it suitable as a passive carrier for this agent.
CFE (10 mg/ml) may be safely delivered with up to 7 mg/ml
CDDP as indicated by hepatic tolerance studies. It is suggested
that tolerance to hepatic chemoembolization is improved with
antibiotic protection.4 In general, transient elevation of liver
function tests was resolved by 3 weeks. Higher drug concentra
tions resulted in liver necrosis.
4 Unpublished observations.
In the present study we have measured platinum deposition
in normal liver and have used normal kidney to model the endartery kinetics of drug delivery to cancer in liver. Administration
of cisplatin by i.v. and hepatic artery i.a. resulted in similar
hepatic platinum levels of 0.46 ¿tg/gtissue at 2.5 h. After
chemoembolization, hepatic levels were increased 2-fold to 0.91
¿tg/g.Renal artery chemoembolization, however, resulted in
much higher 2.5-h renal platinum levels of 10.1 j/g/g, resulting
in an apparent 11:1 ratio for drug delivery by chemoemboliza
tion in end artery as compared with circulations influenced by
portal flow. In a concurrent clinical trial, CDDP concentrations
of 10 mg/ml have been achieved. Extrapolating from the model
data presented here, vascularized tumor tissue should achieve
platinum levels of 100 Mg/g wet weight compared with 0.2 to
0.6 Mg/g>the range noted in several tissues following i.v. ad
ministration, a 200-fold increase in cisplatin delivered to tumor.
The general validity of these estimates is supported by deter
mination of Pt distribution in patient livers containing primary
and metastatic cancer.4 Livers from five patients were obtained
at necropsy or upon surgical resection. Chemoembolization
with 10 mg/ml collagen and 1 mg/ml cisplatin had been per
formed prior to death or surgery. Liver computer-assisted to
mography scans at completion of embolization demonstrated
contrast arrest within the periphery of the tumor. Platinum
deposition generally was highest at the tumor rim, successively
lower in tumor core and normal liver, and lowest in extrahepatic
tissues.
Limited hepatic vascular capacitance coupled with increased
first pass extraction of drug results in relatively low systemic
exposure at drug concentrations within the range of hepatic
tolerance. We documented decreased levels of platinum in
serum and nontarget organs following hepatic chemoemboli
zation. In initial clinical trials patients have received an average
volume of 8 ±7 ml at concentrations as high as CFE:CDDP
10:10 mg/ml.4 We calculate that the systemic exposure required
to achieve these high tumor concentrations
mately 50% of a typical systemic dose.5
of drug is approxi
Chemoembolization with other agents, namely mitomycin C,
doxorubicin, fluorodeoxyuridine,
l,3-bis(2-chloro)-l-nitrosourea, and etoposide, and with drug combinations is being stud
ied in a similar manner.4 The effect of antibiotic protection on
hepatic tolerance to chemoembolization are currently being
studied in both animal models and concurrent clinical trials. It
is unclear at present what effects vascular occlusive changes in
physiological microenvironment such as hypoxia, reduced pH,
and energy status have on drug uptake and cytotoxicity. These
considerations warrant additional future study.
The present study documents a differential relative advantage
in drug deposition in end-artery systems over that observed for
normal liver. The establishment of safe concentration ranges
for hepatic chemoembolization has made possible the initiation
of Phase I clinical trials. Data from these trials are being
assembled and will be reported separately.
ACKNOWLEDGMENTS
The authors would like to thank Tim and Sue Moore of Irish Farms
(Norco, CA) who supplied and cared for the animals in this study.
5Assuming a typical cisplatin i.v. dose of 2 mg/kg in a 70-kg recipient results
in a 140-mg total body exposure. The maximum chemoembolic dose will be (8
ml) x (10 mg/ml) or 80 mg. Seven % regional entrapment further reduces this to
a 74-mg dose equivalent or 53%.
2449
Downloaded from cancerres.aacrjournals.org on June 9, 2014. © 1988 American Association for Cancer Research.
PHARMACOKINETICS OF COLLAGEN CHEMOEMBOLIZATION
REFERENCES
1. Breedis, C., and Young, G. Blood supply of neoplasms in the liver. Am. J.
Pathol., 30: 969-986, 1954.
2. I.incieli, B., Aronsen, K. F., Nosslin, B., and Rothman, U. Studies in pharmacokinetics and tolerance of substances temporarily retained in the liver by
microsphere embolization. Ann. Surg., 187:95-99, 1978.
3. Karakousis, C. P., Kanter, P. M., Lopez, R., Moire, R., and Holyoke, E. D.
Modes of regional chemotherapy. J. Surg. Res., 26: 134-141, 1979.
4. Anderson, J. H., Gianturco, C., and Wallace, S. Experimental transcatheter
infusion-occlusion chemotherapy. Invest. Radio!., 76:496-500, 1981.
5. Yoshioka, T., Hashida, M., Muranashi, S., and Sezaki, H. Specific delivery
of mitomycin C to the liver, spleen and lung: nano- and microspherical
carriers of gelatin. Int. J. Pharm. (Amst.), 81: 131-141, 1981.
6. Kudo, S., Wright, K. C., Chuang, V. P., Wallace, S., Mir, S., and Bechtel,
W. Experimental evaluation of intraarterial occlusion infusion chemother
apy. AJR, 143: 1069-1073, 1984.
7. I.orelius, L. E., Benedetto, A. R., Blumhardt, R., Gaskill, H. V., Lancaster,
J. L., and Stridbeck, H. Enhanced drug retention in VX2 tumors by use of
degradable starch microspheres. Invest. Radici!., /'/- 212-215, 1984.
8. Kato, T., Nemoto, R., Mori, H., Takahashi, M., Tamakawa, Y., and Harada,
M. Arterial chemoembolization with microencapsulated anticancer drug.
JAMA, 245: 1123-1127,1981.
9. Gyves, J. W., Ensminger, W. D., VanHarken, D., Niederhuber, J., Stetson,
P., and Walker, S. Improved regional selectivity of hepatic arterial mitomycin
in starch microspheres. Clin. Pharmacol. Ther., 34: 259-265, 1983.
10. Ohnishi, K., Tsuchiya, S., Nakayama, Y., Hiyama, Y., Iwama, S., Goto, N.,
Takashi, M., Ohtsuki, T., Kono, K., Nakajima, Y., and Okuda, K. Arterial
chemoembolization of hepatocellular carcinoma with mitomycin C microcapsules. Radiology, 152: 51-55, 1984.
11. Fujimoto, S., Miyazaki, M., Endoh, F., Takahashi, ().. Okuii, K., and
Morimmo, Y. Biodegradable mitomycin C microspheres given intra-arterially for inoperable hepatic cancer. Cancer (Phila.), 56: 2404-2410,1985.
12. Ensminger, W. D., Gyves, J. W., Stetson, P., and Walker-Andrews, S. Phase
I study of hepatic arterial degradable starch microspheres and mitomycin.
Cancer Res., 45:4464-4467, 1985.
13. Pfeifle, C. E., Howell, S. B., Ashbum, W. L., Barone, R. M., and Bookstein,
J. J. Pharmacologie studies of intra-hepatic artery chemotherapy with de
gradable starch microspheres. Cancer Drug Delivery, 3: 1-13, 1986.
14. El-Domeiri, A. A., and Mojab, K. Intermittent occlusion of the hepatic artery
and infusion chemotherapy for carcinoma of the liver. Am. J. Surg., 135:
771-775, 1978.
15. Wallace, S., Charnsangavej, C., Carrasco, C. H., Bechtel, W., Wright, K. C.,
and Gianturco, C. Percutaneous transcatheter infusion and infarction in the
treatment of human cancer. Curr. Prob. Cancer, 8(17): 1-62; 8(18): 1-76,
1984.
16. Daniels, J. R., Kerlan, R. K., Jr., Dodds, L., McLaughlin, P., LaBerge, J.
M., Harrington, D., Daniels, A. M., and Ring, E. J. Peripheral hepatic
embolization with crosslinked collagen fibers. Invest. Radiol., 22: 126-131,
1987.
17. Andrews, P. A., Wung, W. E., and Howell, S. B. A high performance liquid
Chromatographie assay with improved selectivity for cisplatin and active
platin(II) in plasma ultrafiltrate. Anal. Biochem., 143: 46-56, 1984.
18. Chen, H-S. G., and Gross, J. F. Intra-arterial infusion of anticancer drugs:
theoretic aspects of drug delivery and review of responses. Cancer Treat.
Rep., 64: 31-40, 1980.
19. Hu, E., and Howell, S. B. Pharmacokinetics of intraarterial mitomycin C in
humans. Cancer Res., 43:4474-4477, 1983.
20. Campbell, T. N., Howell, S. B., Pfeifie, C. E., Wung, W. E., and Bookstein,
J. Clinical pharmacokinetics of intraarterial cisplatin in humans. J. Clin.
Oncol., 1: 755-762, 1983.
2450
Downloaded from cancerres.aacrjournals.org on June 9, 2014. © 1988 American Association for Cancer Research.