![Histological changes in liver tissue after RF ablation. Arrows indicated the location of ablated tissue and implant position. (A) After 24 h, the boundary between normal and ablated tissue is clearly defined (*). (B) At 48 h, the ring of inflammatory cells is clearly visible (#). (C) At 96 h, the inflammatory response gives way to fibroblasts and collagen formation (+). (D) At 144 h, the disorganized structure of the fibrous capsule is clearly visible (+). The original magnification was 100× for (A) and (B) and 200× for (C) and (D). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]](https://www.researchgate.net/profile/Agata-Exner/publication/9043880/figure/fig1/AS:267476931969117@1440782878863/Histological-changes-in-liver-tissue-after-RF-ablation-Arrows-indicated-the-location-of_Q320.jpg)
Content uploaded by Agata Exner
Author content
All content in this area was uploaded by Agata Exner on Jul 13, 2018
Content may be subject to copyright.
Local carboplatin delivery and tissue distribution in livers
after radiofrequency ablation
A. Szymanski-Exner,
1
A. Gallacher,
2
N. T. Stowe,
3
B. Weinberg,
1
J. R. Haaga,
4
J. Gao
1,4
1
Department of Biomedical Engineering, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, Ohio
2
Department of Analytical Chemistry, Ricerca, LLC, Concord, Ohio
3
Department of Surgery, Case Western Reserve University School of Medicine, Cleveland, Ohio
4
Department of Radiology, University Hospitals of Cleveland, Cleveland, Ohio
Received 9 December 2002; accepted 20 December 2002
Abstract: This study investigated the local drug pharma-
cokinetics of intralesional drug delivery after radiofre-
quency ablation of the liver. We hypothesized that the tissue
architecture damaged by the ablation process facilitates the
drug penetration in the liver and potentially enlarges the
therapeutic margin in the local treatment of cancer. The
delivery rate and tissue distribution of carboplatin, an anti-
cancer agent, released from poly(D,L-lactide-co-glycolide)
implants into rat livers after radiofrequency ablation were
quantified by atomic absorption spectroscopy. Results
showed that carboplatin clearance through blood perfusion
was significantly slower in the ablated livers, leading to a
more extensive tissue retention and distribution of the drug.
The concentration of Pt at the implant–tissue interface
ranged from 234 to 1440 g Pt/(g liver) in the ablated livers
over 144 h versus 56 to 177 g Pt/(g liver) in the normal
tissue. The maximum penetration distance at which Pt level
reached above 6 g/g (calculated based on a reported IC
90
value for carboplatin) was 8–10 mm and 4–6 mm in ablated
and normal liver, respectively. Histological analysis of the
necrotic lesions showed widespread destruction of tissue
structure and vasculature, supporting the initial hypothesis.
This study demonstrated that intralesional drug delivery
could provide a sustained, elevated concentration of anti-
cancer drug at the ablation boundary that has the potential
to eliminate residual cancer cells surviving radiofrequency
ablation. © 2003 Wiley Periodicals, Inc. J Biomed Mater Res
67A: 510-516, 2003
Key words: drug delivery; tissue distribution; atomic ab-
sorption spectroscopy; pharmacokinetics; radiofrequency
ablation
INTRODUCTION
Image-guided radiofrequency (RF) ablation has
been successfully used in recent years as a minimally
invasive therapy to treat unresectable liver tumors.
1–8
In this procedure, imaging methods such as X-ray
computed tomography, magnetic resonance imaging,
or ultrasound guides the percutaneous insertion of a
needle electrode to the site of a tumor. Electric current
at radio frequency is applied through the needle, lead-
ing to ionic agitation, increased temperature, and fi-
nally coagulative necrosis.
2–6
The procedure is simple,
quick, and can be performed on an outpatient basis.
3,4
Studies have shown that although RF ablation is able
to destroy majority of the tumor tissue, tumor recur-
rence has been reported in many cases because of the
incomplete elimination of all the cancer cells.
3,6,7
To
address this limitation, other minimally invasive in-
terventional procedures, such as intralesional drug
therapy, are being explored.
Currently, research efforts in our group focus on the
development of a combination therapy consisting of
RF ablation and local chemotherapy for the treatment
of solid tumors. The therapy includes the following
steps: first, a small drug-loaded cylindrical implant,
the polymer millirod, will be fabricated using estab-
lished methods.
9
This device is composed of an active
agent entrapped in a biodegradable poly(lactic-co-gly-
colic acid) (PLGA) matrix. Second, under image guid-
ance, a tumor will be treated with RF ablation fol-
lowed by implantation of millirod(s) directly into
tumor tissues to provide site-specific delivery of the
selected agent. The active agent will be released from
the polymer matrix directly into tumor tissue at a
controlled rate over time to eliminate the residual
cancer cells.
Local drug delivery is a powerful technique that has
the potential to become a mainstream alternative to
Correspondence to: J. Gao; e-mail: jmg23@po.cwru.edu
Contract grant sponsor: National Institutes of Health; con-
tract grant number: R21 CA93993
© 2003 Wiley Periodicals, Inc.
conventional systemic chemotherapy. This mode of
drug administration can deliver the drug to a site of
action at a significantly higher concentration than is
possible with intravenous injection or oral delivery. It
can also maintain the concentration within the thera-
peutic window to avoid unnecessary toxicity. Partic-
ularly, local delivery is beneficial for drugs (especially
anticancer drugs) with narrow therapeutic indices or
short in vivo half-lives to maximize their therapeutic
efficacy.
For a local drug therapy to attain maximum efficacy,
it must deliver the therapeutic dose of a drug to the
target area without substantially affecting the normal
tissues. Frequently, the drug released from the im-
plant must travel a significant distance and reach the
site of action in its active state, and its ability to do so
will depend highly on the properties of the drug as
well as on the property of the tissue. Other groups
have examined the effects of cell density on drug
distribution in tissue and have determined that a de-
creased cell density due to apoptosis enhances tissue
penetration.
10–12
Currently, data on local drug release
and tissue pharmacokinetics are limited but need to be
quantified for the successful development of a local
drug delivery system. This is true especially in com-
plex tissue environments, such as thermally ablated
tissue. Because the viable and ablated tissue structure
varies drastically, with the ablated region lacking vi-
able cells and vasculature, we predict that this differ-
ence will significantly affect the local drug pharmaco-
kinetics in tissues.
In this study, we examined the release kinetics and
tissue distribution of carboplatin as it was released
from the polymer millirod in ablated liver tissues. The
drug distribution profiles provide information on the
local drug pharmacokinetics in ablated tissues and
were correlated with tissue structure from histology
analysis. Results from this study provide the funda-
mental understanding of structure-property relation-
ships between in vivo drug transport properties and
liver structures, which is essential for the future de-
sign of sustained release formulations in conjunction
with RF ablation for the treatment of liver tumors.
MATERIALS AND METHODS
Materials
PLGA (lactide:glycolide ⫽1:1, 0.65 dL/g inherent viscos-
ity) was purchased from Birmingham Polymers, Inc. (Bir-
mingham, AL). Carboplatin (cis-diammine(1,1-cyclobutane-
dicarboxylato) platinum) and poly(vinyl alcohol) (13–23
kDa) was purchased from Sigma-Aldrich (Milwaukee, WI).
D(⫹)-Glucose was purchased from Fluka (Milwaukee, WI).
Phosphate-buffered saline (PBS), sodium hydroxide (NaOH,
0.2M), nitric acid (70%, trace metal grade), and methylene
chloride were obtained from Fisher Scientific (Pittsburgh,
PA). Single-element platinum standard (NIST traceable) was
purchased from CPI International (Santa Rosa, CA). Teflon
tubes were purchased from McMaster–Carr Supply Com-
pany (Cleveland, OH). Sprague–Dawley rats were obtained
from Charles River Laboratories (Wilmington, MA).
Polymer implant fabrication and in vitro
characterization
Millirod implants were fabricated according to a previ-
ously established compression-heat molding procedure.
9
Briefly, PLGA microspheres (approx. 4 m diameter) were
mixed with carboplatin to form a uniform mixture, and
D(⫹)-glucose was added to the mixtures in order to expedite
the rate of release. The homogeneously mixed powder was
placed in a mold and compressed at 4.6 ⫻10
6
Pa at 90°C for
2 h. The resulting cylindrical millirods have an average
diameter of 1.62 mm and a theoretical loading density of
10% carboplatin with 35% glucose (w/w).
The in vitro release of carboplatin from the millirods was
measured in PBS (pH 7.4). Typically, segments of millirods
(approx. 8 mm in length) were submerged in 10 mL of PBS
and placed in an orbital shaker (New Brunswick Scientific,
model C24) at 37°C and 100 rpm agitation. At each sampling
point, the millirod was removed from the vial and placed
into 10 mL of fresh PBS. The retained sample was analyzed
using flameless, graphite-furnace atomic absorption spec-
troscopy (GF-AAS, PerkinElmer model 4100ZL). This
method has been a widely accepted method of examining
levels of various metals in tissue and body fluids.
13–15
The
method is a spectrophotometric technique based on the ab-
sorption of radiant energy by atoms.
16
First, five single-
element platinum standards (NIST traceable) were prepared
in deionized water at concentrations of 10, 25, 50, 100, and
500 g Pt/L to establish the calibration curve. The linear
correlation coefficients for all the curves (area counts for Pt
at 265.9 nm vs. concentration) were greater than 0.999 before
analysis. The platinum concentration in the retained sample
(Pt, g/mL) was determined by quantitative serial dilution
(250⫻) in water followed by their interpolative assay against
the standard curve. The platinum content was then con-
verted to reflect the total carboplatin release based on the
weight percentage of Pt to the drug molecule. The average
amount of carboplatin remaining in the implants was calcu-
lated from this data and was standardized to the initial
measured length of each implant. The total concentration of
carboplatin was calculated based on measurements of the
drug in completely degraded implants.
RF ablation and millirod implantation in rat livers
Animal procedures followed an approved protocol by the
Institutional Animal Care and Use Committee. Male Spra-
gue–Dawley rats (250–350 g) were anesthetized using an
intraperitoneal injection of sodium pentobarbital (4 mg/100
g). The abdomen was swabbed with betadine solution, and
CARBOPLATIN DISTRIBUTION AFTER RF ABLATION 511
mercaine (bupivacaine, 0.1 mL) was injected subcutaneously
at the incision site prior to the start of surgery. The liver was
exposed through an incision in the midsection. In the ani-
mals undergoing ablation, the liver capsule of the medial
lobe was first perforated with an 18-gauge hypodermic nee-
dle, and the liver tissue was ablated with a 19-gauge needle
electrode (Radionics
威
, Burlington, MA 01803) at 90 ⫾3°C for
2 min. The ablation procedure followed a previously estab-
lished method.
8,17
After ablation, millirods (5–7 mm in
length) were implanted into the needle electrode track. Mil-
lirods of the same composition used in the ablated lobes
were implanted into nonablated livers for comparison. After
implantation, a small piece of cotton was sutured on top of
the implantation site to seal the wound and prevent the
implant from slipping out. The animals received Buprenex
(buprenorphine, 0.05–0.1 mg/kg) after the surgery and were
allowed to recover. The animals were then euthanized with
Fatal Plus (concentrated sodium pentobarbital), and livers
were removed and stored at ⫺80°C. Animals were divided
into six experimental groups with endpoints of 1, 6, 24, 48,
96, and 144 h. Each group consisted of six rats with implants
in normal livers (n⫽3) and ablated livers (n⫽3).
Sample preparation and analysis with atomic
absorption spectroscopy
The platinum content in livers and explanted millirods
was also analyzed with atomic absorption spectroscopy
(AAS). In preparation for AAS analysis, the livers were
divided into three segments around the center (implantation
site) and three 2-mm thick slices were sectioned from each
segment. These were cut into 2-mm wide strips and the six
strips closest to the implant (radial distance of 12 mm) were
retained for analysis. These samples were deposited into
previously weighed, 4-mL glass scintillation vials. The vials
were weighed again to determine the wet mass of each liver
segment. Sample digestion was conducted using a novel
dry-block procedure (compared to the standard microwave
digestion). The 4-mL vials were placed into dry-block incu-
bators and 100 L of 70% nitric acid was added to each vial.
The samples were heated to 70°C for 1 h and digestion was
aided by addition of 50 L hydrogen peroxide after 20, 30,
and 50 min. The clear sample solutions were transferred to
15-mL Falcon centrifuge tubes that were diluted to 5 mL
with distilled water and stored at 4°C until analysis. To
determine carboplatin release in vivo, the explanted milli-
rods were degraded in 2 mL of 0.2M NaOH at 37°C for 1
week. The solution was neutralized with addition of 2 mL of
0.2M nitric acid and filtered with 0.45-m syringe filters
before AAS analysis. The platinum concentrations in the
above solutions were determined by interpolative assay
against the standard curve (10–500 g Pt/L) as described in
the previous section. The g/L measurements were con-
verted to g/(g liver) by multiplication of the sample vol-
ume and division by the original wet tissue weight. The
explant measurements were converted to mg/cm as the
remaining quantity of carboplatin per unit length of the
millirod.
The t
1/2
was calculated based on exponential decay curve
fit to the average release data. The tissue penetration dis-
tance was calculated as the maximum distance at which the
Pt concentration reaches above 6 g/g, a value calculated
based on the IC
90
(10 g/mL)
18
carboplatin concentration to
the VX-2 cancer cells. The area under the concentration-time
curve (AUC) for carboplatin in ablated liver tissue over
144 h was calculated using the trapezoid rule for the im-
plant/tissue interface and ablation boundary. An unpaired,
two-tailed Student t test with a 95% confidence interval was
used to determined significant differences among data sets.
Significant outliers were discarded on basis of the Grubbs
test (␣⫽0.05).
Histological analysis
All liver samples underwent gross histological examina-
tion and microscopic analysis. The diameter of the visible
ablated area perpendicular to the ablation needle tract was
measured three times and an average measurement of the
ablation area was calculated. Representative sections of the
specimens were removed and fixed in 10% formalin solu-
tion. The preserved sections were embedded in paraffin,
sectioned into 5-m slices and stained with hematoxylin and
eosin. The hematoxylin and eosin sections were examined to
determine the normal tissue structure surrounding the im-
planted millirod and the damage inflicted by the implanta-
tion trauma and thermal ablation.
RESULTS
In vitro release of carboplatin in PBS buffer
Figure 1 shows the average release profile of carbo-
platin in PBS buffer (pH 7.4) at 37°C. The data are
represented as carboplatin remaining in the millirod
over time and is standardized by the length of each
implant (mg/cm). The cumulative release follows the
Higuchi model (C
drug
⫽kt
n
,n⫽0.5 ⫾0.03) with a
value of t
1/2
at 42 ⫾8 h. In addition, 2.0 ⫾0.2 and
3.8 ⫾0.3 mg carboplatin/cm were released after 24
and 96 h, respectively (n⫽4). The reproducibility of
the fabrication procedure is demonstrated by a rea-
sonably low standard deviation (⬍0.3 mg/cm, n⫽4)
at all time points.
In vivo release of carboplatin in rat livers
The in vivo release of carboplatin in ablated and
normal rat livers is shown in Figure 1. Compared with
the in vitro release, the standard deviation in in vivo
release data is significantly greater, particularly at ear-
lier time points (tⱕ24 h). Although carboplatin re-
lease appears to be faster in ablated livers than normal
livers in the first day after implantation, it should be
512 SZYMANSKI-EXNER ET AL.
noted the difference is not statistically significant (e.g.
the pvalue from the Student ttest is 0.13 and 0.28 at 6
and 24 h, respectively). Release half-life was deter-
mined as 51 ⫾10 and 44 ⫾15 h for normal and ablated
tissue, respectively. When examining the raw data, we
observed a comparable release amount of carboplatin
at longer time points in both systems, with 3.7 ⫾0.2
and 3.8 ⫾0.5 mg carboplatin/cm released after 96 h in
normal and ablated livers, respectively. This data
demonstrates similar overall release kinetics in the
two liver environments as well as in PBS buffer (3.8 ⫾
0.3 mg/cm at 96 h).
Distribution of carboplatin in normal and ablated
liver tissues
Figure 2(A,B) shows the tissue distribution of car-
boplatin in platinum concentrations as a function of
distance in ablated (n⫽9) and normal (n⫽9) liver
tissues. Data show significantly more carboplatin were
retained in the ablated tissue over the normal tissue. In
the livers treated with RF ablation [Fig. 2(A)], the
range of platinum in the tissue exceeds 1400 g Pt/(g
liver) at the millirod/tissue interface, and the concen-
tration rises steadily through 48 h and decreases there-
after. In contrast, the range of Pt in the normal liver
[Fig. 2(B)] consistently remains below 200 g Pt/(g
liver), and the concentration varies randomly through-
out the 6 days after implantation. The maximum tissue
penetration distance, at which the Pt concentration
reaches above 6 g/g liver, was found to be between
8–10 mm in the ablated liver at 1 and 96 h. During the
other time points, the maximum distance was between
4– 6 mm. In the normal liver, the maximum penetra-
tion distance was found to be between 0–2 mm at all
but one time point and between 4– 6 mm in the first
hour.
The change in carboplatin concentration over time
at the implant/tissue interface and the distal ablation
boundary is shown in Figure 2(C,D). At implant/
tissue interface, carboplatin in the ablated livers
reaches 1.4 ⫾0.6 ⫻10
3
g/g liver (C
max
,n⫽9) at 48 h
and is 24 times higher than that in the normal liver
[61 ⫾45 g/g, Fig. 2(C)]. At the approximate ablation
boundary (4– 6 mm), the carboplatin concentration
was at least two times higher in the ablated liver at all
time points, and 14 times higher at 144 h [Fig. 2(D)].
Although the differences in measured Pt are consis-
tent, it should be noted that some measurements (es-
pecially the Pt content in normal tissue) at this dis-
tance are below the detection limit of the atomic
absorption technique, and may not be accurate.
The AUC at the implant/tissue interface was calcu-
lated to be 234.6 and 28.5 (mg/g)*h for carboplatin in
ablated and normal liver, respectively. The AUC at the
ablation boundary (or 4– 6 mm from the implant) was
calculated to be 4.8 (mg/g)*h in ablated and 0.8 (mg/
g)*h in normal liver. All AUC values are reported in
mg carboplatin (not Pt) for clinical relevance.
Histological analysis
The gross morphology of the tissue surrounding the
millirod was significantly altered in the ablated liver
site. The ablation radius ranged from 3.8 to 5.8 mm
(average radius ⫽4.8 ⫾0.9 mm). Millirod implanta-
tion in the ablated lesion resulted in little or no addi-
tional injury to the tissue as observed by the easy
removal of the millirod from the implantation site.
Upon gross examination, the tissue at the millirod
boundary appeared to be identical to that further
away from the implantation site in the ablated area. In
comparison, the implantation site in normal liver
showed a slight morphological change in the sur-
rounding tissue. The tissue adjacent to the implant
appeared white and granular, and the implant was
difficult to remove, possibly indicating an early re-
sponse of the tissue to injury.
Microscopic examination of the implantation site
showed a vastly different tissue structure in the ab-
lated liver tissue with notable changes over 6 days
[Fig. 3(A,D)]. The tissue structure near the implanta-
tion site showed the same architecture as in the rest of
the ablated area [Fig. 3(A,B)]. The cellular structure
was different from normal liver cells and corresponds
Figure 1. Release of carboplatin from millirod implants in
vitro (n⫽4) and in vivo (n⫽3) in ablated (䊐) and normal (F)
liver. Dara are shown as weight of carboplatin remaining in
millirod and standardized based on the length of individual
implants. Portions of error bars (reflecting standard devia-
tion measurements) were omitted for clarity. In the in vitro
release (Œ), the standard deviation was below 0.3 mg/cm at
all time points.
CARBOPLATIN DISTRIBUTION AFTER RF ABLATION 513
to features of coagulative necrosis with picnotic nuclei
and disrupted and irregular cytoplasm. The sinusoi-
dal structure was destroyed and the interstitial space
was increased. At the distal edge of the ablated area,
and at the interface of the ablated and normal liver
tissue, a ring of inflammatory infiltrate, composed of
lymphocytes and monocytes, was observed. This ring is
more apparent at later time points starting at 24 h and
increasing through 144 h. The cell nuclei in the necrotic
area become increasingly sparse as the necrosis spreads,
and fibroblasts and granulation tissue are visible at the
distal ablation boundary at 96 and 144 h [Fig. 3(C,D)].
DISCUSSION
The therapeutic efficacy of the intralesional drug
delivery system addressed in this study is highly de-
pendent on the local drug pharmacokinetics in the
tissue environment surrounding the implantation site.
Characterization of the in vivo drug release kinetics
and drug penetration in ablated tissue is critical in
providing the necessary experimental data for the de-
velopment of these delivery systems.
The in vitro release kinetics of carboplatin in PBS
buffer (t
1/2
⫽42 ⫾8 h) agrees reasonably well with
Figure 2. Local drug pharmacokinetics of carboplatin in normal and ablated liver over 144 h. (A) Tissue distribution in
ablated liver. Carboplatin distribution increases initially and reaches a maximum at 48 h. The distribution data at 6 h
superimpose with the 96-h data and were omitted. (B) Tissue distribution in normal liver, where no obvious time-dependent
trend was observed. Error bars were omitted for clarity in (A) and (B). (C) Pt concentration at the implant/tissue interface in
normal (■) and ablated (E) livers. (D) Pt concentration at the distal ablation boundary in ablated livers. Pt concentration at
the same distance in normal liver is shown for comparison. Dashed line represents the IC
90
value of carboplatin to VX-2 cancer
cells from the literature. Error bars reflect standard error (n⫽9).
514 SZYMANSKI-EXNER ET AL.
those in normal liver (t
1/2
⫽51 ⫾10 h) and ablated
liver (t
1/2
⫽44 ⫾15 h) in vivo. The in vivo release
kinetics showed larger experimental variability, which
may reflect the heterogeneous liver environments in
different animals. Results from current study indicate
slightly faster release of carboplatin from polymer
millirods in ablated liver than normal liver; however,
this difference is not statistically significant. In com-
parison, the retention, distribution and penetration of
carboplatin are significantly different in the two tissue
environments. Carboplatin retention is approximately
one order of magnitude higher in the ablated liver
over normal liver [Fig. 2(A,B)]. In addition, drug pen-
etration is significantly greater in ablated liver with a
maximum penetration distance of 8–10 mm compared
to mostly 2 mm in the normal liver.
The different pharmacokinetics of carboplatin in ab-
lated and normal liver can be correlated to structural
changes in the liver following RF ablation. The histol-
ogy analysis documented these changes in the ablated
lesion over 6 days. During this time it is possible to
observe morphological changes in hepatocytes under-
going coagulative necrosis, where the cells shrink and
nuclei become increasingly sparse as they are cleared
out from the injured area. These results are in agree-
ment with those reported previously by our group
17
as well as others.
5,8
More importantly, the histological
analysis of ablated lesions confirms the severe destruc-
tion of organized sinusoidal capillary structure and
other liver vasculature visible in normal tissue (not
shown). The lack of carboplatin clearance through
blood perfusion in ablated tissue is a determinant
factor for the greater retention and penetration of car-
boplatin. Without drug loss from perfusion, drug dif-
Figure 3. Histological changes in liver tissue after RF ablation . Arrows indicated the location of ablated tissue and implant
position. (A) After 24 h, the boundary between normal and ablated tissue is clearly defined (*). (B) At 48 h, the ring of
inflammatory cells is clearly visible (#). (C) At 96 h, the inflammatory response gives way to fibroblasts and collagen formation
(⫹). (D) At 144 h, the disorganized structure of the fibrous capsule is clearly visible (⫹). The original magnification was 100⫻
for (A) and (B) and 200⫻for (C) and (D). [Color figure can be viewed in the online issue, which is available at www.inter-
science.wiley.com.]
CARBOPLATIN DISTRIBUTION AFTER RF ABLATION 515
fusion becomes the dominant process of transport in
ablated liver, and leads to significantly higher pene-
tration distance over normal liver. Wound healing
response to the heat-induced injury may also affect the
local drug pharmacokinetics. At the ablation boundary,
a ring of inflammatory cells is observed from 24 to 96 h
[Fig. 3(B,C)], and fibroblasts are observed forming a fi-
brous capsule around the ablated area after 96 h [Fig.
3(D)]. The formation of fibrous capsule may cause the
elevated platinum concentration at the ablation bound-
ary at 144 h [Fig. 2(D)] over the other time points.
In our proposed combination therapy, intralesional
drug delivery aims to eliminate the residual cancer
cells surviving RF ablation. Therefore, the ablation
boundary is the targeted site of action and the concen-
tration-time relationships at this location are impor-
tant for an effective carboplatin therapy. Our study
showed that carboplatin reached its therapeutic con-
centration (IC
90
of carboplatin to VX-2 cancer cells) at
the ablation boundary for all the time points over
144 h [Fig. 2(D)]. The AUC value at the ablation
boundary was determined to be 4.8 (mg/g)*h in ab-
lated liver, six times more than normal liver [0.8 (mg/
g)*h] at the same distance. Since clinical AUC values
are normally calculated from the plasma drug concen-
trations, the tissue measurements from this study can-
not be compared with the clinical data. Nonetheless, as
regional and local drug delivery becomes a more ac-
cepted therapy, these measurements may become essen-
tial in describing the local tissue pharmacokinetics and
evaluating the effectiveness of drug delivery systems.
CONCLUSIONS
The data obtained in this study successfully support
our hypothesis that radiofrequency ablation of tissue
increases the maximum penetration distance of a drug
and allows the drug to be consistently elevated above
therapeutic concentration at the site of action. The
tissue penetration was three times higher in the ab-
lated tissue and drug concentration was found to be
an order of magnitude higher at most distances from
the tissue/implant interface as compared with similar
implants in normal livers. The mode of action for this
striking difference is the destruction of vasculature and
loss of perfusion resulting from the ablation process,
which is supported by gross and microscopic examina-
tion of the histology sections. Although this study dem-
onstrates that the drug can reach the ablation boundary,
further studies on a tumor model are necessary to eval-
uate the increased efficacy of the combination therapy
(RF ablation ⫹local drug delivery) over RF ablation
alone. These studies are currently being conducted in
our group with the rabbit VX-2 tumor model.
The authors thank Dr. James Anderson for his help with
the histology analysis.
References
1. Becker GJ. 2000 RSNA annual oration in diagnostic radiology:
The future of interventional radiology. Radiology 2001;220:
281–292.
2. Lewin JS, Connell CF, Duerk JL, Chung YC, Clampitt ME,
Spisak J, Gazelle GS, Haaga JR. Interactive MRI-guided radio-
frequency interstitial thermal ablation of abdominal tumors:
clinical trial for evaluation of safety and feasibility. J Magn
Reson Imaging 1998;8:40–47.
3. Dodd GD III, Soulen MC, Kane RA, Livraghi T, Lees WR,
Yamashita Y, Gillams AR, Karahan OI, Rhim H. Minimally
invasive treatment of malignant hepatic tumors: at the thresh-
old of a major breakthrough. Radiographics 2000;20:9–27.
4. Rhim H, Dodd GD III. Radiofrequency thermal ablation of
liver tumors. J Clin Ultrasound 1999; 27:221–229.
5. Goldberg SN, Gazelle GS, Compton CC, Mueller PR, Tanabe
KK. Treatment of intrahepatic malignancy with radiofre-
quency ablation: radiologic-pathologic correlation. Cancer
2000;88:2452–2463.
6. Goldberg SN, Gazelle GS. Radiofrequency tissue ablation:
physical principles and techniques for increasing coagulation
necrosis. Hepatogastroenterology 2001;48:359–367.
7. Bartolozzi C, Crocetti L, Cioni D, Donati FM, Lencioni R.
Assessment of therapeutic effect of liver tumor ablation pro-
cedures. Hepatogastroenterology 2001;48:352–358.
8. Boaz TL, Lewin JS, Chung YC, Duerk JL, Clampitt ME, Haaga
JR. MR monitoring of MR-guided radiofrequency thermal ab-
lation of normal liver in an animal model. J Magn Reson
Imaging 1998;8:64–69.
9. Qian F, Szymanski A, Gao J. Fabrication and Characterization
of Controlled Release Poly(D,L-Lactide-co-Glycolide) Milli-
rods. J Biomed Mater Res 2001;55:512–522.
10. Au JLS, Jang SH, Wientjes MG. Clinical aspects of drug deliv-
ery to tumors. J Control Release 2002;78:81–95.
11. Kuh HJ, Jang SH, Wientjes MG, Weaver JR, Au JLS. Determi-
nants of Paclitaxel Penetration and Accumulation in Human
Solid Tumor. J Positron Emission Tomogr 1999;290:871–880.
12. Jang SH, Wientjes MG, Au JLS. Enhancement of paclitaxel
delivery to solid tumors by apoptosis-inducing pretreatment:
effect of treatment schedule. J Positron Emission Tomogr 2001;
296:1035–1042.
13. Duffull SB, Robinson BA. Clinical pharmacokinetics and dose
optimisation of carboplatin. Clin Pharmacokinet 1997;33:161–83.
14. Milacic R, Cemazar M, Sersa G. Determination of platinum in
tumour tissues after cisplatin therapy by electrothermal atomic
absorption spectrometry. J Pharm Biomed Anal 1997;16:343–348.
15. Meerum Terwogt JM, Tibben MM, Welbank H, Schellens JH,
Beijnen JH. Validated method for the determination of plati-
num from a liposomal source (SPI-77) in human plasma using
graphite furnace Zeeman atomic absorption spectrometry. Fre-
senius J Anal Chem 2000;366:298–302.
16. Shugar GJ. Atomic Absorption Spectroscopy. In: Shugar R, Bau-
man L, Shugar-Bauman R, editors. Chemical technicians’ ready
reference handbook, 3rd ed. New York: McGraw-Hill, Inc.; 1990.
p 751–763.
17. Szymanski-Exner A, Stowe NT, Lazebnik RS, Salem KA, Wilson
DL, Haaga JR, Gao J. Noninvasive monitoring of local drug
release in a rabbit radiofrequency (RF) ablation model using x-ray
computed tomography. J Control Release 2002;83:415–425.
18. Pauser S, Wagner S, Lippmann M, Pohlen U, Reszka R, Wolf
KJ. Evaluation of efficient chemoembolization mixtures by
magnetic resonance imaging therapy monitoring: An experi-
mental study on the VX2 tumor in the rabbit liver. Cancer Res
1996;56:1863–1867.
516 SZYMANSKI-EXNER ET AL.
