| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Experimental Therapeutics |
INSERM U370, Necker Faculty of Medicine, 75730 Paris [R. G., A. S. C., C. B., P-L. T.]; Laboratoire de Recherche en Chirurgie, INSERM 94–05, Hopital Cochin, 75014 Paris [J. C.]; Laboratoire de Recherche en Imagerie, INSERM U494, Necker Faculty of Medicine, 75730 Paris [M. L., O. C.]; and Service d’Anatomie Pathologique, Hopital Beaujon, 92110 Clichy [M-P. B.], France
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSToxicity of Ad.CMVtk and...DISCUSSIONREFERENCES |
|---|
-fetoprotein regulatory gene sequences, respectively. The
antitumor response was assessed by magnetic resonance imaging and by
autopsy and histological analysis following postmortem. Tumor growth
cessation was demonstrated by magnetic resonance imaging in large tumor
nodules of size 5–8 mm treated by intratumoral administration of
2 x 109 pfu Ad.CMVtk plus
i.p. treatment with GCV. We also show an antitumor efficacy in small
tumor nodules of size <3 mm treated with 2 x 109 pfu Ad.CMVtk plus GCV by the
intrahepatic artery route, albeit associated with an adverse toxicity.
In vivo targeting of the HSV-tk gene to
diethylnitrosamine-induced HCC cells with the recombinant
Ad.AFPtk suppresses the hepatic toxicity in the
nontumoral liver. The lower antitumor response would argue for the use
of multiple injections of such adenoviral constructs. These
observations may lead to potential approaches for designing gene
therapy destined for early treatment of dysplastic nodules or advanced
HCC in cirrhosis. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSToxicity of Ad.CMVtk and...DISCUSSIONREFERENCES |
|---|
, and p53 proteins into
HCC-derived cell lines can induce tumor growth inhibition
(1, 2, 3, 4, 5, 6, 7, 8)
. In vitro, data have been substantiated
by the ability of such vectors to inhibit in vivo the growth
of tumors in mice upon s.c. HCC cell injection (4
, 6
, 9)
. Although encouraging, such observations do not address a major issue in gene therapy designed for primary liver cancer, namely that it develops in most cases against a background of CAH and cirrhosis (10 , 11) . HCC, the most prevalent histological form of primary liver cancer, indeed occurs with a yearly rate of 3–5% in patients with cirrhosis (10) . Although present therapies of human HCC tumors, including liver transplantation, have been improved, the cure rate for this disease still remains low. In contrast with normal liver, the vascularization of cirrhotic livers is predominantly arterial, and this feature is reinforced in HCC-bearing patients, in whom up to 70% of the blood supply is dependent on the hepatic artery (12 , 13) . Cirrhosis and HCC tumors also show the so-called "sinusoid capillarization" whereby the normal fenestration of liver sinusoidal endothelial cells is abrogated (14 , 15) . The necrotic and inflammatory lesions that drive CAH and cirrhosis stimulate liver cell DNA replication. Hence, this latter process may favor targeting of toxic compounds to nontumor cells, such as those generated upon HSV-tk transduction and related toxicity mediated by phosphorylated GCV. Finally, given the generally poor liver function observed in patients with advanced liver cirrhosis, the potentially toxic effects of gene transfer may have marked deleterious consequences. These observations stress the need, when designing vectors and routes of administration, for the use of relevant animal models. Tumors that develop in immunodeficient mice are extremely useful for investigating the parameters of the therapeutic effect of HSV-tk gene expression, but they clearly do not enable us to address the specific problems encountered in human HCC due to the pattern of tumor vascularization.
To address this issue, we and others have been investigating rat HCC induced by DEN treatment. In this classical model of liver carcinogenesis, HCC develops against a background of increased liver cell proliferation in immunocompetent animals (16 , 17) . We have investigated in detail this model and have shown that, as with human tumors, HCC displays predominant arterial vascularization, thus pointing to the need for selective IHA or IT injections (18) . We have also shown the overall low transduction rate of tumor cells, although IHA injection of small nodules and dysplastic nontumor areas yielded higher efficacy (18) .
Despite this overall low transduction efficacy, intraportal injection of an adenoviral vector encoding HSV-tk under the control of a CMV promoter can promote, upon GCV injection, a potential antitumor efficacy leading to partial regression of HCC nodules, but at the expense of high toxicity (5) . The ubiquitous CMV promoter indeed induced tk expression and related toxicity mediated by phosphorylated GCV in nonneoplastic hepatic cells, which show increased DNA synthesis, as would occur in human CAH and cirrhotic tissues. In the present study, we investigated whether selective IHA or IT administration of recombinant Ad could restrict the extension of toxicity generated by the HSV-tk/GCV protocol. To circumvent tk expression in nonneoplastic tissue, specific gene expression can be achieved with the AFP regulatory sequences, the expression of which is predominantly observed in hepatoma cells (3 , 19 , 20) . Several reports have demonstrated that this hepatoma-specific expression of murine and human AFP regulatory sequences is retained in adeno- and retroviral vectors (19, 20, 21) . However, the efficacy and toxicity of this approach in an in vivo model such as DEN-induced rat HCC have not yet been evaluated. The present study includes an analysis of antitumoral efficacy and eventual related toxicity in HSV-tk/GCV gene therapy. Our results lead us to propose different modalities in gene therapy for the early treatment of dysplastic nodules or advanced HCC in cirrhotic patients.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSToxicity of Ad.CMVtk and...DISCUSSIONREFERENCES |
|---|
The E1-deleted recombinant Ad.CMVlacZ and Ad.CMVtk containing, respectively, Escherichia coli lacZ and HSV-tk genes, were kindly provided by Drs C. Qian and J. Prieto (5) .
All recombinant Ads were amplified on 293 cells and purified by double cesium chloride density gradient ultracentrifugation. Titers of the adenoviral stocks were determined by plaque assay on 293 cells.
Cell Culture
Two AFP-producing human hepatoma cell lines (HepG2 and HuH7) and
a human cervical cancer cell line (HeLa) were grown in DMEM (Life
Technologies, Inc., Gaithersburg, MD) containing 8% FCS, 2
mM glutamine, 100 units/ml penicillin, and 100 µg/ml
streptomycin.
To test the sensitivity to GCV of infected cells, HuH7, HepG2, and HeLa cells were plated in triplicate wells into a 96-well plate, and various MOIs of Ad.AFPtk were infected. Twenty h after infection, increasing concentrations (0, 1, 5, 10, and 20 µg/ml) of GCV (Synthex Laboratories, Inc., Palo Alto, CA) were added. Cell proliferation was then measured with a nonradioactive cell proliferation assay according to the manufacturer’s protocol (CellTiter 96 AQueous Non-Radioactive Cell Proliferation Assay; Promega, Madison, WI). The percentage of surviving cells is presented as a percentage of the absorbance observed in GCV-treated cells divided by that of cells without GCV treatment (mean ± SD).
Tumor Induction and Tumor Size Assessment
Male Wistar rats (6 weeks old, 150–180 g) were purchased from
Iffa Credo (l’Arbesles, France). They were acclimated for 15 days
before DEN treatment. Rats were treated at a dose of 10 mg/kg/day of
DEN (Sigma) for 60 days (Figs. 1
and 2)
. They received their dose of DEN in drinking water at 100 mg/liter
from a fresh DEN solution prepared weekly. Multifocal hepatic lesions,
including dysplastic nodules and tumors, were apparent in 60% of the
animals 10 weeks after beginning exposure to the carcinogen.
|
|
D. The percentage of tumor
growth (TSI) in surface between day 94 (S94) and day 80
(S80) was calculated as follows: TSI (%) = 100 x (S94 -
S80)/S80.
In Vivo Gene Therapy
All animal experiments were reviewed and approved by
institutional committees for animal care and use and for recombinant
DNA research (Ministère de l’Agriculture, Paris, France).
Evaluation of the efficacy of HSV-tk gene therapy was
performed on DEN-treated rats with tumors of size 5–8 mm or <3 mm.
Two recombinant Ads, Ad.CMVtk and Ad.AFPtk, were
assessed. Control animals received either Ad.CMVlacZ,
Ad.AFPlacZ, or saline buffer. Two experimental schedules
were designed for IT and IHA administrations as shown in Fig. 1, A and B
,
respectively.
Administration of Recombinant Ads Via the IT Route
Ad.CMVtk.
Two independent experiments on HSV-tk/GCV efficacy upon IT
administration of recombinant Ad were performed in tumor-bearing rats
using Ad.CMVtk and Ad.CMVlacZ, or saline buffer.
The experimental schedule is shown in Fig. 1A
. The rats were
monitored by MRI on day 80 to select homogeneous groups of animals with
advanced tumors of 5–8 mm in size. On the following day (day
81), they were randomized and laparotomized under ether
anesthesia. Then the selected tumor nodule received a volume of 100
µl containing either recombinant Ad or saline buffer as follows: in
the first experiment, seven rats received injections of 2 x 109 pfu Ad.CMVtk, and five
rats received injections of 2 x 109 pfu Ad.CMVlacZ; the second
experiment included eight rats receiving injections of 2 x 109 pfu Ad.CMVtk and seven
rats receiving injections of saline buffer. In both experiments, GCV
treatment by the i.p. route started on the following day and lasted
from days 82 to 91. Evaluation of tumor growth by MRI was assessed on
day 94, 3 days after the last GCV dose, and all animals were sacrificed
on day 95. During GCV treatment, the death of one rat in
Ad.CMVtk + GCV-treated rats was reported in each
experiment. Autopsy revealed i.p. hemorrhage, and histological
examination showed evidence of acute hepatitis. In control groups, one
case of decease in saline-injected rats was related to a hemorrhage of
the injected tumor. No deaths were reported in
Ad.CMVlacZ-treated rats.
Ad.AFPtk.
Two independent experiments were performed in the groups of
tumor-bearing rats treated with Ad.AFPtk. As a typical
experiment, 12 tumor-bearing rats were randomly distributed into two
groups and treated with either 5 x 109 pfu Ad.AFPtk
(n = 8) or saline buffer
(n = 4). The treatment schedule was the same
as previously described in Fig. 1A
. No clinical signs or
morbidity were observed during GCV treatment. Deaths of
Ad.AFPtk-treated (four of eight) and saline-injected (one of
four) rats were essentially due to their advanced pathology, with a
bleeding tumor provoking i.p. hemorrhage.
Selective Administration of Recombinant Ads Via the IHA Route
Ad.CMVtk.
Two independent experiments were performed with recombinant
Ad.CMVtk as scheduled in Fig. 1B
. Tumor-bearing
rats with multiple tumors of size <3 mm were selected by MRI on day 75
and randomly assigned to various groups. They were laparotomized under
ether anesthesia on day 76. After dissection of the hepatic pediculus,
the celiac artery was clamped. A 30 1/2-gauge needle was introduced
into the hepatic artery via the gastroduodenal artery.
Recombinant Ad.CMVtk or Ad.CMVlacZ or saline
buffer was then injected in a volume of 400 µl, followed by an
equivalent volume of saline buffer. After injections of both solutions,
the gastroduodenal artery was ligated, and blood flow into the hepatic
artery was re-established. The duration of the arterial blood flow
interruption varied between 5 and 15 min. On day 77, all rats received
a dose of 50 mg/kg/day GCV diluted in saline buffer by i.p.
administration for 10 days. The first experiment included 21 rats
injected with either a dose of 2 x 109 pfu Ad.CMVtk
(n = 8), 109 pfu
Ad.CMVtk (n = 7), 2 x 109 pfu Ad.CMVlacZ
(n = 3), or saline buffer
(n = 3); the second experiment was repeated
with 14 rats injected with either a dose of 2 x 109 pfu Ad.CMVtk
(n = 7), 2 x 109 pfu Ad.CMVlacZ
(n = 4), or saline buffer
(n = 3). Follow-up of the experiments
included controls of the morbidity, mortality, and a second MRI on day
105. In both experiments, treatment with HSV-tk/GCV induces
early toxicity manifested by mild to severe hepatitis. Mortality
recorded was due to HSV-tk-related toxicity and/or to the
advanced pathology. All rats were kept for survival recording up to day
125. The surviving rats were then killed. The evaluation of the
antitumor response was assessed by recording their body and liver
weights and the number of tumor nodules. Because tumor growth was
variable, the number of tumor nodules included tumor nodule sizes from
1–10 mm.
Ad.AFPtk.
The evaluation of the efficiency of Ad.AFPtk + GCV
on tumor growth and comparison with that of Ad.CMVtk + GCV was assessed by three experiments. A typical experiment is
reported herein: tumor-bearing rats with homogeneous tumor sizes <3
mm, as monitored by MRI, were randomly assigned to various groups on
day 75. After being ether-anesthetized and laparotomized, they were
injected by the IHA route with either a dose of 5 x 109 pfu Ad.AFPtk
(n = 6), 109 pfu
Ad.CMVtk (n = 4), or 5 x 109 pfu Ad.AFPlacZ
(n = 6), or saline buffer
(n = 6). No morbidity or mortality was
observed in Ad.AFPtk-injected rats. No sign of toxicity was
observed in Ad.AFPtk-, Ad.AFPlacZ-, or
saline-treated rats. Instead, high toxicity was observed in
Ad.CMVtk/GCV-treated rats during GCV treatment, leading to
death of all them. Autopsy of deceased animals in the control
groups showed that it was essentially due to their advanced pathology
because histological examination did not reveal evidence of hepatitis.
The others were kept for survival recording. They were then
sacrificed. Their liver weights were noted, and the number of tumor
nodules was recorded.
Toxicity Study in Tumor-free Rats
Twenty-five male Wistar rats (10–12 weeks old; 300 g) were
ether-anesthetized and laparotomized. They were injected through the
hepatic artery with a dose of either recombinant Ad 2 x 109 pfu Ad.CMVtk
(n = 15) or 2 x 109 pfu Ad.CMVlacZ
(n = 10), as described above. The following
day, 10 of 15 Ad.CMVtk- and 5 of 10
Ad.CMVlacZ-treated rats were treated with GCV at a dose of
50 mg/kg/day for 10 days. The remaining rats were left without GCV
administration. Three days after the end of GCV treatment, blood
samples were collected for assay of liver serum enzymes (AST and ALT),
and all animals were sacrificed for macroscopic examination. The liver
and various organs were removed and processed for histology.
Histology
Liver samples from tumoral and nontumoral areas of each
DEN-treated rat, and liver samples as well as various organs (spleen,
intestine, pancreas, kidney, and lung) from tumor-free rats were
collected on the day of sacrifice or after death of the animal. They
were fixed in 4% formaldehyde and then embedded in paraffin. Staining
of 3-µm-thick sections was performed with H&E.
Statistics
Comparisons of the survival rate, number of tumors,
liver weight, and tumor growth between rats from different groups were
performed using the nonparametric distribution Mann and Whitney test.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSToxicity of Ad.CMVtk and...DISCUSSIONREFERENCES |
|---|
8 mm. They were divided into
two groups with tumor sizes of 5–8 mm and <3 mm. Histological
examination of these neoplastic nodules (dysplastic and HCC nodules)
showed that they presented all features of cellular alterations
characteristic of a process of hepatocarcinogenesis (not shown), as
previously described (24)
.
Tumor Growth Delay following IT Administration of
Ad.CMVtk + GCV
The efficacy of HSV-tk/GCV treatments on DEN-induced
rat HCC was first evaluated by IT injection of large hepatic tumor
nodules of size 5–8 mm in diameter with either 2 x 109 pfu Ad.CMVtk, 2 x 109 pfu Ad.CMVlacZ, or saline buffer
into each selected tumor of each rat. The tumor sizes of injected and
noninjected tumors were evaluated by MRI 14 days after initiation of
the treatment. Fig. 2
depicts
representative magnetic resonance images of two individual rats
showing, for each animal, an antitumor response after treatment with
2 x 109 pfu
Ad.CMVtk + 50 mg/kg/day GCV, whereas no reduction
in tumor size was noted in surrounding untreated tumors. Table 1
pools together the results of two experiments (experiments I and II).
Each value represents separately the percentage of tumor growth as
calculated for each treated and surrounding untreated tumor in each
animal from each group of rats. In some rats, one or two noninjected
tumors were measured, whereas in others, the noninjected tumors were
not measurable due to their size below the detection threshold. In the
two experiments, the results showed that Ad.CMVtk-treated
tumors responded to HSV-tk + GCV treatments
as follows: 5 of 13 tumors had significant regression (-22
to -73% of tumor growth), 4 tumors had pronounced growth
cessation (+10–29% of tumor growth), and growth of 3 tumors was
delayed (+56–90% of tumor growth). In control groups, growth of
Ad.CMVlacZ treated-tumors fluctuated from +53 to +196% of
tumor growth, and that of saline-injected tumors fluctuated from +20 to
+880%. Likewise, growth of the surrounding untreated tumors from all
rats unevenly fluctuated. The mean values of tumor growth determined
for all treated and untreated tumors in each group of rats were,
respectively, +5% of tumor growth in Ad.CMVtk-treated
tumors, +125% in Ad.CMVlacZ-treated tumors, and +221% in
saline-injected tumors (P < 0.001; Table 1
).
Table 1
also showed no difference in the mean values of tumor growth
between all untreated tumors and saline-injected tumors.
|
. The inflammatory response was moderate
in Ad.CMVtk + GCV-treated tumors and was composed
predominantly of mononuclear cells (Fig. 3A)
. No evidence of
hepatic lesion was observed in the nontumoral liver parenchyma (not
shown). No inflammatory infiltrate was observed in any control groups.
In conclusion, treatment of DEN-induced HCC tumors of size 5–8 mm by
IT administration of Ad.CMVtk + GCV elicited tumor
growth cessation with low toxicity.
|
reports the effect of Ad.CMVtk/GCV on tumor burden as
measured by liver weight 50 days after initiation of the treatment.
Both groups of animals that were treated with either 2 x 109 or 109 pfu
Ad.CMVtk + GCV showed a decrease in tumor burden,
as measured by weight. In contrast, control rats injected with
Ad.CMVlacZ + GCV had an increase in liver weight
similar to that of saline-injected rats. The latter observation can be
correlated with the number of tumors listed for each liver (Table 2)
.
The mean number of tumors in rats treated with 2 x 109 pfu Ad.CMVtk was 8.8, and that of
rats treated with 109 pfu was 16.4, as compared
to a mean value of 29 in Ad.CMVlacZ-treated rats and 27 for
the saline-injected rats (P < 0.001). This
demonstrates that the high Ad.CMVtk dose had a more
efficient therapeutic effect.
|
. Deaths in the control
groups of rats injected with either 2 x 109 pfu Ad.CMVlacZ + GCV or
saline + GCV were essentially due to their advanced
pathology. The survival rate between the two groups of
Ad.CMVtk + GCV-treated rats was not significantly
different (P > 0.09), whereas the survival
rate between rats treated with 2 x 109 pfu Ad.CMVtk + GCV and
that of 2 x 109 pfu
Ad.CMVlacZ + GCV or saline-injected rats was
significantly different (P < 0.05). Thus,
high doses of Ad.CMVtk + GCV were found to be toxic for
the animals. Histological examination of liver sections from the surviving Ad.CMVtk-treated rats showed remnants of hepatic tumors associated with a few dysplastic areas, and the structure of large areas of liver parenchyma appeared almost normal, without evidence of inflammatory infiltration (not shown). No hepatic lesion due to HSV-tk/GCV-related toxicity was observed on the day of the killing. By contrast, control rats displayed multiple tumors in all liver lobes, associated with several dysplastic areas or HCC foci, and histologically conserved areas were limited (not shown). Thus, treatment of tumors <3 mm in DEN-induced rat HCC by a single IHA administration of Ad.CMVtk + GCV treatment could decrease tumor burden and delay tumor progression.
Targeted Expression of the HSV-tk Gene to HCC Cells
In Vitro Biological Activity of Recombinant
Ad.AFPtk Containing the HSV-tk Gene Under
the Control of the Rat AFP Gene Promoter.
Previous results from our laboratory showed the need for specific
targeting of the recombinant Ad.AFPlacZ to HCC cells
(20)
. We therefore constructed a recombinant Ad,
Ad.AFPtk, containing the HSV-tk gene controlled
by the rat AFP gene regulatory sequences. Two human HCC cell lines,
HuH7 and HepG2, and a non-HCC cell line, HeLa, were used as target
cells for the recombinant Ad infection. At a MOI of 500, all HCC and
HeLa cells transduced with Ad.AFPtk showed cytopathic
effects. After infection with adenoviral vector Ad.AFPtk,
cells were treated with varying doses of GCV for 5 days, and the number
of viable cells was determined by a cell proliferation assay. The
AFP-producing HuH7 and HepG2 cells were more sensitive to this effect
than HeLa cells. HuH7 cells infected with Ad.AFPtk displayed
GCV sensitivity at a concentration as low as 5 µg/ml, at a MOI of 30
(Fig. 4)
. HeLa cells infected with Ad.AFP.tk did not show any GCV
sensitivity (Fig. 4)
. These results indicate that
Ad.AFPtk-mediated transfer of the HSV-tk gene
resulted in GCV killing of only AFP-producing cells.
|
Antitumor therapy of DEN-induced HCC tumors was designed as scheduled
in Fig. 1A
. IT injection of Ad.AFPtk was
performed in 5-mm HCC tumor nodules. Each designated tumor received
either a dose of 1010 pfu Ad.AFPtk or
saline buffer, followed by i.p.-treatment with 50 mg/kg/day of
GCV. The results displayed as experiment III in Table 1
show a
pronounced growth cessation in two tumors (+2 and +23% of tumor
growth), whereas growth of the other two tumors was simply delayed. A
mean value of +66% of tumor growth in Ad.AFPtk-treated
tumors was compared with the mean values of +521% of tumor growth
in saline-injected tumors and +296% in untreated tumors
(P < 0.05). Histological examination of
Ad.AFPtk + GCV-treated tumors showed
moderate cell apoptosis and necrosis, with a limited inflammatory
response composed predominantly of mononuclear cells (Fig. 3C)
.
Rats bearing tumors with size <3 mm were also treated by the IHA route
with either a single dose of 5 x 109 pfu Ad.AFPtk or a single dose of
109 pfu Ad.CMVtk. Control rats
received, by the IHA route, either a dose of 5 x 109 pfu Ad.AFPlacZ or saline buffer. A
5-fold dose of the recombinant Ad.AFPtk was used, as
compared to that of Ad.CMVtk. All rats were then treated by
i.p. injection of GCV for 10 days. The results are summarized in Table 2
(experiment C). No toxic effect was observed in Ad.AFPtk-
and Ad.AFPlacZ-treated rats. No mortality was reported in
Ad.AFPtk-treated rats, whereas all four
Ad.CMVtk + GCV-treated rats died during GCV
treatment. Data show a weak antitumoral effect in
Ad.AFPtk-treated rats, as compared to that observed in
Ad.CMVtk/GCV-treated rats.
In conclusion, targeting HSV-tk to HCC under the control of AFP regulatory sequences abolished the hepatic toxicity induced in the nontumoral liver parenchyma by HSV-tk/GCV. The limited antitumor effect observed in rats, which were injected with a single treatment dose of recombinant Ad by IT or IHA routes, should advocate the use of multiple injections of recombinant Ad.
|
Toxicity of Ad.CMVtk and GCV in Tumor-free Rats |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSToxicity of Ad.CMVtk and...DISCUSSIONREFERENCES |
|---|
. This toxic effect was much less observed in animals
treated by direct IT injection of Ad.CMVtk. The enhanced and
dose-dependent toxicity of the higher dose was likely due to
transduction of normal hepatic cells by Ad.CMVtk and
resultant toxicity mediated by phosphorylated GCV. Our present data
show that absence of toxicity in the nontumoral liver parenchyma was
obtained when HSV-tk was placed under the control of AFP
regulatory sequences (not shown).
|
displays the mean values of serum levels of AST and
ALT, and the liver weight of the various groups of Ad-treated ± GCV rats. It showed, 14 days after initiation of the treatment,
an increase in serum liver enzymes in Ad.CMVtk + GCV-treated rats accompanied by a decrease in liver weight, indicating
severe liver damage.
|
. This severe hepatitis was diffused in all of the liver
lobes. In the remaining Ad.CMVtk + GCV-treated
rats, severe (n = 4) and moderate hepatitis
(n = 2) were recorded. The spleen of
Ad.CMVtk-treated rats were atrophic, whereas in the other
organs (intestine, pancreas, kidney, or lung), no abnormality was
reported. Finally, Table 3
summarizes the scoring of HSV-tk/GCV-related toxicity, as
observed in all experiments performed with either tumor-free or
DEN-induced HCC rats, following treatment with either
Ad.CMVtk + GCV or Ad.AFPtk + GCV
by the IHA and IT routes. We conclude that low toxicity of
HSV-tk + GCV could be obtained by using IT
administration of recombinant Ad and by targeting HSV-tk
gene expression to HCC.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSToxicity of Ad.CMVtk and...DISCUSSIONREFERENCES |
|---|
The results presented here may lead to an additional approach on the in vivo feasibility and impact of the HSV-tk/GCV suicide gene therapy for primary liver cancer. Our experiments provide strong support that the transfer of HSV-tk/GCV into DEN-induced rat HCC by IT and IHA injections of Ad.CMVtk elicited an antitumor response in tumor nodules, including small and large tumors, as assessed by MRI and histological analyses. We showed that antitumor efficacy associated with low toxicity was obtained by IT administration of Ad.CMVtk into large tumor nodules and GCV treatment. Our results also stressed the major problem of adverse toxicity of HSV-tk + GCV treatment despite the use of the selective IHA instead of the intraportal route of administration (5) .6 Finally, we showed that targeting the HSV-tk gene to HCC with Ad.AFPtk under the control of the AFP gene promoter suppressed adverse toxicity in the nontumoral liver parenchyma. Nonetheless, cell targeting with a tissue-specific promoter also induced a decrease in the antitumoral efficacy.
Because of the heterogeneity of tumor nodules in HCC induced by DEN treatment in rats, monitoring their sizes by MRI is an important step in the evaluation of the tumor growth. Thus, the combination of MRI with the liver weight, and histological examination at necropsy provide potential tools for its assessment. Moreover, the MRI approach allowed us to locate and select large tumor nodules for a direct IT injection into the tumor. Thus far, the growth of the selected tumor nodule can be analyzed and compared to that of untreated nodules, acting as internal controls. In addition, we were able to establish a reproducible selective IHA injection procedure that parallels therapeutic strategies presently used in humans for chemotherapy and chemoembolization.
Direct IT injection of large tumor nodules (5–8 mm) with 2 x 109 pfu Ad.CMVtk and GCV treatment revealed a significant bystander effect within the tumoral mass, leading to a tumor growth arrest, as compared to that of untreated nodules. It clearly elicited less toxicity in the nontumoral liver parenchyma than that observed by IHA administration of the recombinant Ad, although signs of mild hepatitis were present. The absence of antitumor effects on untreated tumor nodules suggests that, in this model of HCC, there are only a few IT arterial anastomoses to allow leakage of the vector and/or toxic metabolites generated into other tumor areas. This observation varies from those made in rats injected with the HCC-derived cell line (25) . In this model, tk gene expression was retrovirally transduced ex vivo in rat McA-RH8994 tumor cells before transplantation into the right hepatic lobe of syngeneic rats, whereas nonexpressing tk cells were transplanted into the left hepatic lobe of the same animals. These authors showed that following treatment with 150 mg/kg/day GCV, not only TKh+ tumors but also TKh- tumors regressed. They reported that the distant bystander effect observed was likely immune-mediated. Such observations were also reported in another model of rat using ex vivo-transduced HCC cells before transplantation into syngeneic rats. In both models, the introduction of transduced tumor cells in vivo could induce efficiently a tumor-specific immunity that was associated to CD8 T cells (27) . Instead, in the model of HCC tumors induced in situ in the rat liver, the activation of the immune response against transduced and nontransduced tumor cells would require more complex immune mechanisms.
Intrahepatic arterial injection of Ad.CMVtk plus i.p.
injection of GCV led to a significant antitumoral effect in small tumor
nodules <3 mm. The antitumor response observed in small nodules was
clearly due to tk gene expression and was not observed in
Ad.CMVlacZ- and saline-treated animals. It was also
dose-dependent because the dose of 2 x 109 pfu Ad.CMVtk showed significantly
higher efficacy, as compared to the dose of 109
pfu. We also attempted to treat large tumor nodules >5 mm by IHA
injection of 2 x 109 pfu
Ad.CMVtk plus i.p. injection of GCV. The low antitumor
response of large tumor nodules is in accordance with our previous
report that, despite the use of selective IHA injection, the
transduction efficiency of large HCC nodules remains extremely low
(18)
.7
However, doses of 2 x 109 and
109 pfu Ad.CMVtk both induced
hepatitis, although the toxicity of the former dose was high because 10
of 15 of the Ad.CMVtk-treated animals died during GCV
treatment. This toxic effect was likely due to the use of a strong and
ubiquitous CMV enhancer/promoter to drive expression of the
tk gene in both tumoral and nontumoral cells. GCV
metabolites generated from tk expression should theoretically show
toxic effects only in cells that replicate their DNA (28)
.
The high replication rate of nonmalignant hepatocytes induced by DEN
treatment could, in part, account for the observed toxicity in our
experiments and others (5)
. However, in agreement with a
previous report, we demonstrate that Ad.CMVtk + GCV also induces significant alterations in liver functions of
tumor-free normal rats leading to severe hepatitis and high mortality
rate (Fig. 6
; Ref. 29
). Similar results of
HSV-tk/GCV-related toxicity for resting cells have also been
reported by several laboratories (30, 31, 32)
. An histological
score of hepatotoxicity in both tumor-free and tumor-bearing rats
showed that it ranged from moderate to severe hepatitis and included
the usual pattern of HSV-tk/GCV-induced toxicity with
apoptotic and necrotic liver cells (Table 3
and Fig. 5, A and B
). Given the extremely low proportion of replicating
normal liver cells, the mechanisms implicated in this toxicity still
remain unclear, but this observation has important implications for the
design of clinical gene therapy.
Finally, the present study shows that the hepatic toxicity elicited in
the nontumoral liver parenchyma could be practically eliminated by the
use of the AFP regulatory sequences targeting tk gene
expression to hepatoma cells. We did observe an absence of toxicity
upon IHA injection of the Ad.AFPtk vector and GCV
administration, although up to 5 x 109 pfu was used. On the other hand, the
tissue-specific promoter produced a lower antitumoral response in both
IT and IHA administrations of the recombinant virus (Tables 1
and 2)
.
Two explanations could account for such results. It is obvious that the
AFP promoter sequence is markedly less efficient, both in
vitro and in vivo, as compared with the CMV promoter;
thus, the intracellular amount of thymidine kinase might not be
sufficient to generate complete toxic and bystander
effects.8
Furthermore, Ohguchi et al. (33)
have reported
the heterogeneity of the AFP expression as revealed by
immunohistochemistry and in situ hybridization within human
HCC. It is not excluded that the same process would also occur in the
DEN-induced rat HCC and contribute to the low antitumoral response.
Therefore, the absence of the hepatic toxicity in the nonmalignant
liver parenchyma will allow multiple injections of Ad.AFPtk
for in vivo treatment of HCC. Bramson et al.
(34)
have reported that preexisting immunity to Ad in the
Ad-immune mice did not prevent tumor regression following IT
administration of a recombinant Ad expressing the cytokine IL-12, but
inhibits virus dissemination. This should be examined in the
DEN-induced rat model under HSV-tk gene therapy. Our
histological results showed that within a short term following
Ad.CMVtk or Ad.AFPtk injection and GCV treatment,
the transduced tumors underwent rapidly an apoptotic and necrotic
process. In addition, the inflammatory response composed of
mononuclear cells may result in antitumor effects independent of the
transgene.
In conclusion, both IT and IHA injections of the recombinant Ad for in vivo treatment of HCC appear to be two potential and feasible routes of administration for human clinical trials. The combination of direct IT injections of large nodules with IHA injections allowing "sterilization" of small foci of dysplastic, "preneoplastic" liver cells would provide an attractive strategy for treatment of the multifocal tumor lesions in most primary liver cancer. Our data emphasizes the need, in the future, to further delineate hepatoma-specific regulatory sequences to improve targeting efficacy.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by the Institut National de la
Santé et de la Recherche Médicale, the Association de la
Recherche sur le Cancer, the Ligue Nationale de Recherche sur le
Cancer, the Association Française contre les Myopathies, and by a
fellowship of the Association de la Recherche sur le Cancer (to R. G.). 
2 To whom requests for reprints should be
addressed, at UMR 8532 CNRS, Gustave Roussy Institute, 39
rue Camille Desmoulins, 94805 Villejuif, France. Phone: 33-1-42114560;
Fax: 33-1-42115276; E-mail: pltran{at}igr.fr
3 Present address: UMR 8532 CNRS, Gustave Roussy
Institute, 39 rue Camille Desmoulins, 94805 Villejuif, France.
4 The abbreviations used are: HCC, hepatocellular
carcinoma; HSV-tk, herpes simplex virus-thymidine
kinase; CMV, cytomegalovirus; MOI, multiplicity of infection; CAH,
chronic active hepatitis; DEN, diethylnitrosamine; Ad, adenovirus; AFP,
-fetoprotein; pfu, plaque-forming unit(s); GCV, ganciclovir; MRI,
magnetic resonance imaging; IHA, intrahepatic artery; IT, intratumoral;
AST, aspartate amino-transferase; ALT, alanine amino-transferase.
5 P. L. Tran, A. Sa Cunha, and C. Brechot.
Centre National de la Recherche Scientifique (CNRS) under patent.
6 R. Gerolami and P. L. Tran, unpublished
results.
8 R. Gerolami and P. L. Tran, unpublished
results.
Received 11/10/99. Accepted 12/14/99.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSToxicity of Ad.CMVtk and...DISCUSSIONREFERENCES |
|---|
gene transfer. Gastroenterology, 112: 501-510, 1997.[Medline]
-fetoprotein gene promoter. Cancer Res., 55: 3105-3109, 1995.-fetoprotein and albumin genes in human hepatocellular carcinomas: limitations in the application of the genes for targeting human hepatocellular carcinoma in gene therapy. Hepatology, 28: 599-607, 1998.
This article has been cited by other articles:
|
|
 |
|
  C.-Q. Ling, B. Li, C. Zhang, D.-Z. Zhu, X.-Q. Huang, W. Gu, and S.-X. Li Inhibitory effect of recombinant adenovirus carrying melittin gene on hepatocellular carcinoma Ann. Onc., January 1, 2005; 16(1): 109 - 115. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Faivre, J. Clerc, R. Gerolami, J. Herve, M. Longuet, B. Liu, J. Roux, F. Moal, M. Perricaudet, and C. Brechot Long-Term Radioiodine Retention and Regression of Liver Cancer after Sodium Iodide Symporter Gene Transfer in Wistar Rats Cancer Res., November 1, 2004; 64(21): 8045 - 8051. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Haberkorn and A. Altmann Imaging Techniques for Gene Therapy: SPECT, PET, and MRI Journal of Pharmacy Practice, October 1, 2001; 14(5): 383 - 396. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
, HuH7 without treatment; 
,
Ad.AFPtk + GCV-treated HuH7; 
,
Ad.AFPtk alone-treated HuH7; 
, HeLa without
treatment; x, Ad.AFPtk + GCV-treated HeLa;
*, Ad.AFPtk alone-treated HeLa.
), necrosis, and
apoptosis (arrowhead) with a few mononuclear cells
(*). H&E staining; magnification, x200. B, in
tumor-free rat liver, IHA injection of Ad.CMVtk
(2 x 109 pfu) + GVC treatment
induces severe acute hepatitis with marked intralobular hepatocyte
necrosis and apoptosis (arrowhead), piece meal necrosis,
and infiltrates of mononuclear cells (*). H&E staining;
magnification, x200.