(CANCER RESEARCH 51. 6643-6649. December 15. 199I|
Antitumor Activity of -y-Interferon in Ascitic and Solid Tumor Models of Human
Ovarian Cancer
Saleem T. A. Malik, Richard G. Knowles, Nick East, Danielle Lando, Gordon Stamp, and Frances R. Balkwill
Imperial Cancer Research Fund, Lincoln's Inn Fields, London WC2A3PX, Uniled Kingdom fS. T. A. M., N. E., G. S., F. R. B.J; Wellcome Research Laboratories,
Langley Court, Beckenham, Kent BR3 3BS, United Kingdom [R. (i. K.}; and Centre de Recherche, Roussel UCLAF.lll Route de Noisy, Romainville, 93230, France
[D. L.I
ABSTRACT
We have investigated the action of recombinant human 7-interferon
(rHuIFN-7) against human ovarian cancer xenografts growing as ascites
or as bulky solid i.p. tumors in nude mice. Both forms of the disease
responded to i.p. rHuIFN-7 with significant increases in mouse survival
time, and in 2 of 3 ascitic models the mice were cured of peritoneal
disease. The activity of rHuIFN-7 was dose and schedule dependent, and
xenografts derived from 3 different patients showed a heterogeneity of
response. Peak i.p. levels of rHuIFN-7 in nude mice bearing multiple i.p.
solid tumors were similar to those found in ovarian cancer patients
receiving i.p. rHuIFN-7, but clearance was more rapid in the mice. Rat
7-interferon had no antitumor activity at the same doses and schedules
although it had some biological activity in the nude mice. Histológica!
examination of treated tumors revealed increased necrosis and loss of
cellular organization with large areas of hypocellular epithelial mucin.
These changes were preceded by a fall in tumor tryptophan and a rise in
tumor kynurenine. We conclude that ri lui 1 V> has a direct dose related
antitumor effect on ovarian cancer xenografts that is preceded by in
creased metabolism of tryptophan.
INTRODUCTION
Ovarian cancer is the commonest fatal gynecological cancer
in the developed world (1). The major contributory factor to
poor prognosis in patients with ovarian cancer is the late stage
of presentation (2). Over 60% of patients present with disease
that has spread beyond the pelvis and into the peritoneal cavity
(3). Despite advances in the surgical and chemotherapeutic
management of this disease, the 5-year survival for patients
with disease spread into the peritoneal cavity is in the region of
5% (4).
The propensity of ovarian cancer to spread within the con
fines of the peritoneal cavity makes it a suitable cancer for
locoregional therapy. The theoretical advantages of i.p. treat
ment with relatively large molecular weight molecules such as
cytokines include the possibility of achieving high local concen
trations of cytokine without high serum levels, therefore dimin
ishing systemic toxicity (5).
Several trials of i.p. cytokine therapy of ovarian cancer have
been reported. TNF1 and IFN-0 have both been reported to
eradicate ascites in patients with advanced ovarian cancer, but
there was no evidence of an antitumor response (6, 7). More
recently, adoptive i.p. immunotherapy with IL-2/LAK. cells led
to one partial response in ten patients with advanced ovarian
cancer, but repeated therapy was associated with peritoneal
fibrosis (8). In a similar study, 6 responses were noted in 22
patients but peritoneal fibrosis prevented repeated administra
tion (9).
Received 7/18/91; accepted 10/1/91.
The costs of publication of this article were defrayed in part 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.
' The abbreviations used are: TNF, tumor necrosis factor; IFN, interferon;
rHuTNF, recombinant human tumor necrosis factor; rHulFN-7, recombinant
human 7-interferon; rRalFN-7, recombinant rat IFN->; IDO, indoleamine
dioxygenase.
IFN-c,' is not active in ovarian cancer patients when given
systemically (e.g., Ref. 10). However, 2 trials of i.p. IFN-a
therapy noted objective complete and partial responses, partic
ularly in patients with residual tumor <5 mm (11, 12).
There are also some reports of activity of IFN-7 in ovarian
cancer. Although no responses were noted in 2 trials of i.p.
IFN-7 therapy in patients with advanced ovarian cancer (13,
14), Welander et al. (15) noted 4 responses in 14 patients
receiving i.v. IFN-7. More recently, 15 complete and 2 partial
responses were noted in 58 évaluablepatients after i.p. therapy
with 20 x IO6 units/m2 IFN-7 twice weekly for 2-4 months
(16). Sixteen of these responses were seen in patients with
residual disease <2 cm. The available clinical data do not
provide information on ways of optimizing this action of IFN7 or ways of defining markers predictive of response.
We have previously described the establishment of i.p. ovar
ian cancer xenograft models and the effects of i.p. TNF therapy
(17). In 2 of the 3 models, i.p. TNF therapy led to the formation
of bulky solid tumors in the mice with eradication of ascites,
whilst control or diluent treated mice had ascitic tumor only,
with no intraabdominal solid tumors. This observation may be
relevant to the biology of human ovarian cancer, since we
subsequently found that tumor cells from human ovarian cancer
patient biopsies were producing TNF mRNA and protein in
situ (18) and that cells transfected with the TNF gene and
constitutively producing its product showed increased invasiveness when implanted i.p. (19).
In this paper, we describe investigations of the antitumor
effects of i.p. IFN-7 therapy in these xenograft models of ascitic
ovarian cancer and the 2 models in which solid intraabdominal
i.p. carcinomatosis can be generated by a short course of TNF
treatment.
MATERIALS
AND METHODS
Mice and Xenografts. Specific pathogen free female nu/nu (nude)
mice of mixed genetic background, 6-12 weeks old, were maintained
as described previously (17). Ovarian cancer xenografts OS, LA, and
HU were established from primary human tumors as described previ
ously (17). OS originated from a 51-year-old woman with a moderately
differentiated serous cystadenocarcinoma, LA from a 72-year-old
woman with a poorly differentiated mucinous cystadenocarcinoma, and
HU from a 23-year-old woman with a moderately differentiated serous
cystadenocarcinoma. OS was used at passage 63, LA between passages
63 and 77, and HU between passages 46 and 58. Survival time of tumor
bearing mice and the biological behavior of the tumor was reasonably
constant during this time. The tumors were passaged in the nude mice.
Ascitic Tumors. At the start of each experiment mice were given i.p.
injections of 0.1 ml of ascitic tumor (approximately 1 x 10' cells)
diluted 1+ 1 in RPMI 1640. Therapy started 7 days later, six-8 mice
were used in each group in experiments where survival was evaluated.
Solid i.p. Tumors. At the start of each experiment mice were given
i.p. injections of tumor cells as described above. Seven days later mice
were treated with l ¿tgrHuTNF daily for 7 days i.p. Four mice were
then selected at random and examined postmortem to check that the
tumor had implanted. IFN-7 therapy was started 1 day later. Eight
6643
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1991 American Association for Cancer Research.
f-INTERFERON
AND OVARIAN CANCER XENOGRAFTS
mice were used in each group where survival was evaluated, and 3-5
mice were used at each time point for evaluation of tumor weights or
measurement of tryptophan/kynurenine levels.
Cytokines. rHuTNF was provided by BioGent, Ghent, Belgium, and
was more than 99% pure. Endotoxin levels were less than 0.2 ng/mg
and the specific activity was 2 x IO7 units/mg. rHuTNF was diluted in
analysis of tryptophan and kynurenine as follows. Perchloric acid
extracts of plasma or tumors, 40-100 n\, were injected onto a high
performance liquid chromatography system consisting of a Zorbax (',«
column (4 x 80 mm) with a 12.5-mm guard column of the same
material. The mobile phase was 100 mM ammonium acetate adjusted
to pH 4.5 with acetic acid and 1% (v/v) methanol at 1.5 ml/min.
Kynurenine was detected by absorption at 350 nm at a retention time
of 1.4 min, and tryptophan was detected by fluorescence, excitation at
285 nm, emission 340 nm, at a retention time of 3.5 min. The results
are expressed as MMtryptophan or kynurenine, assuming a volume of
0.7 ml/g tumor tissue.
Statistics. The Mann-Whitney U test was used to evaluate survival
results. Student's i test was used to evaluate tumor weight data experi
calcium and magnesium free phosphate buffered saline plus 3 mg/ml
bovine serum albumin (Sigma, Poole, England) and stored in single
dose aliquots at —¿
20°Cuntil required. Mice were given O.l-ml injections
of rHuTNF or diluent. Recombinant human IFN-7 (RU 42369) Lot
L:400 (endotoxin, <0.025 ng/mg; specific activity, 2 x IO7 units/mg),
rat IFN-7 (RU 53561) Lot:22 649-70 (endotoxin, <0.01 ng/mg; spe
cific activity, IO7 units/mg), and control diluent (RU 55648) Lot CB
22 658-73 were provided by Roussel UCLAF as lyophilized prepara
ments, the pharmacokinetic experiments, and the tryptophan metabo
tions. The biological activity of rat IFN-7 was measured in Ratee cells
lism experiments.
infected with vesicular stomatitis virus and in mouse L929 cells. There
was a good correlation between the antiviral activity of rat IFN-7 ¡n
RESULTS AND DISCUSSION
the mouse and rat cells. The IFNs were diluted in sterile distilled water
to give a concentration of 1.5 x 10" units/ml and stored in aliquots at
Pharmacokinetics of i.p. rHuIFN-7. The pharmacokinetics of
-20°C.The control diluent was diluted in the same volume of distilled
water.
Pharmacokinetic Study. Mice with no tumor, ascitic tumor, or solid
tumor were given i.p. injections of 1.5 x IO5units rHuIFN-7 as 0.1 ml
of the 1.5 x IO6 units/ml solution. At appropriate time intervals (0-72
h after injection) blood was obtained by cardiac puncture, and peritoneal
washes were obtained by injecting the mice with 2 ml calcium and
magnesium free phosphate buffered saline. There were 5 mice/time
point group. IFN-7 was assayed in the plasma and peritoneal fluid
using the Centocor 7-interferon radioimmunoassay (Centocor, Malvern, PA) as recommended by the supplier with minor modifications.
Standard curve and mouse control samples were prepared with RU
42369 and the results are expressed as RU 42369-like immunoreactivity. Taking into account the volume of the samples, the limit of
quantification was 5.0 units/ml.
Within-assay and between-assay coefficients of variation were be
tween 4.4 and 11.6%.
Histology. Biopsies were fixed in formol-saline and processed to
paraffin wax. Periodic acid-Schiff diastase and Alcian blue stains were
used to study mucin production.
Measurement of Tryptophan and Kynurenine. At various time inter
vals after IFN-7 therapy, mice were killed by cervical dislocation and
tumors immediately snap frozen in liquid N2. Plasma was obtained
from EDTA blood taken by cardiac puncture. After cold spinning, the
plasma was snap frozen on solid COz. After thawing, tumors were
minced with scissors and homogenized (0-4°C)in 3 ml/g tissue homogenization buffer (0.32 M sucrose-50 mvi Tris-1 m\i EDTA-1 mM
dithiothreitol-10 jig/ml soybean trypsin inhibitor-2 Mg/ml aprotinin,
pH 7) using a Ystral homogenizer (Dottingen, Germany). Homogenate
(or plasma), 250 ¿il,
was acidified with 25 n\ 20% (v/v) perchloric acid,
and after 5 min on ice the perchloric acid extracts were centrifuged
(10,000 x g for 5 min at 4°C)to remove precipitated protein. The
supernatant
was used for high performance liquid chromatography
i.p. rHuIFN-7 were assessed in 3 groups of mice: mice with no
tumor; mice bearing ascitic HU tumor; and mice bearing solid
i.p. tumors of the HU xenograft. We chose a dose of 1.5 x 10s
units/mouse because this would be equivalent to 20 x 106/m2
in humans when differences in surface area:volume ratios be
tween mouse and humans are taken into account (20). As shown
in Fig. 1, the presence of i.p. tumor influenced the pharmaco
kinetics of the rHuIFN-7, as measured by radioimmune assay.
After a single injection of 1.5 x 10s units i.p., peak mean levels
of 33,617 units/ml were found at 0.5 h in normal mice. These
levels fell rapidly to 1071 units/ml at 6 h and <5 units/ml at
24 h. In mice bearing solid tumors, peak i.p. levels of IFN-7
were significantly lower (16,287 units/ml at 0.5 h; P < 0.02)
and then declined rapidly as in normal mice. In mice bearing
ascites peak levels of 41,944 units/ml were achieved at 1 h but
the clearance from the peritoneum was much slower, with
appreciable levels of IFN-7 (229 and 55 units/ml) still meas
urable at 48 and 72 h. The area under the curve in the peritoneal
fluid pharmacokinetics was 8 times greater in ascites bearing
mice compared with normal mice and 16 times greater com
pared with solid tumor bearing mice.
rHuIFN-7 was also detected in the plasma of mice. Plasma
pharmacokinetics in solid tumor bearing and control mice were
similar, but like the peritoneal levels, the plasma levels declined
more slowly in mice bearing ascites. The plasma levels were
always at least 1 log lower than those in the peritoneum.
The lower peritoneal levels achieved in mice bearing solid
tumors could have been due to binding of the human IFN to
the tumor tissue. Because this human IFN would not be ex
pected to bind in a specific manner to any mouse tissues, it
B
100000
Fig. 1. Pharmacokinetics of rHuIFN-y in
nude mice. At time 0, 1.5 x 10*units rHuIFN•¿y
were injected i.p. Five mice were used for
each time point in each group. I. mice without
tumor; B, mice with ascitic xenografts; C, mice
with solid i.p. tumors. D. plasma levels; •¿
peritoneal levels. There was a significant de
crease in peak levels in mice in C compared to
A and A(/><0.02).
b..
1
H
Vv,
024
U»n
6 -#^-/f£^ì.
24 "48
72
0246
24
time after injection (h)
6644
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1991 American Association for Cancer Research.
48
72
Y-INTERFERON AND OVARIAN CANCER XENOGRAFTS
B
therapy
therapy
Fig. 2. Survival of mice bearing ascitic xenografts of human ovarian cancer. Mice were
given i.p. injections of tumor cells. There were
6-8 mice in each experimental group.
rHuIFN-7 therapy started 7 days later for 8
<0
>
weeks. A, OS xenograft; B, HU xenograft; C,
LA xenograft. -.-.-,
diluent treated mice;
O, rHuIFN-i, 1.5 x 10' units twice weekly; •¿, 3
in
rHuIFN-i, 1.5 x IO5 units daily. There was a
significant increase in survival in all treated
mice in B and C (P < 0.001). Twice weekly
therapy was less effective than daily therapy in
C (P < 0.002).
therapy
lOOo-O—(
10000
b
»-1
60
9
40
150
200
250
¿---o
100
150
200
250
100
150
200
250
days after tumor injection
would essentially be targeted to the xenograft. The xenografts
in the solid tumor groups probably have the highest tumor
burden because they have been growing in the mice for 14 days,
as opposed to 7 days for the ascitic tumors. In mice bearing
ascitic tumors, factors such as the binding of IFN-7 to mucinous
ascites may have contributed to the prolonged retention of IFN7-
The i.p. pharmacokinetics of rHuIFN-7 in the nude mice
showed some similarities with a comparable dose (20 x IO6
units/m2) in human patients. In a study of 6 patients2 peak i.p.
levels of approximately 13,000 units/ml were achieved at the
end of a 10-min infusion, but the clearance in humans was
much slower and levels of 1150 units/ml could be still be
detected at 48 h. Unlike the mice, only very low levels rHuIFN7 (1-2 units/ml) which were close to the detection limit of the
assay, were detected in the serum of the patients treated i.p.
Thus, although peak levels were similar between patients with
minimal residual ovarian cancer and nude mice bearing solid
tumors, the subsequent clearance of the cytokine was markedly
different. One factor that may have influenced this is the volume
of instillation of IFN-7 which was proportionately much greater
in the clinical trials. Our data suggested that to achieve the
same i.p. exposure to IFN-7 as that achieved in the clinical
trial, we would have to dose more frequently, e.g., daily com
pared with twice weekly.
Effects of rHuIFN-7 on Ascitic Ovarian Tumor Xenografts.
The survival of mice bearing three different ovarian cancer
xenografts, OS, HU, and LA, growing as ascites, was assessed
in an experiment to comparing daily with twice weekly rHuIFN7 therapy (Fig. 2). In the OS xenograft there was no statistically
significant improvement in survival but there appeared to be
some antitumor activity when rHuIFN-7 was administered
daily. In contrast, the survival of mice bearing the HU xenograft
was significantly increased with both daily and twice weekly
therapy (P< 0.001). There was no difference in survival between
the two schedules of rHuIFN-7. The control mice died of ascitic
disease in the peritoneal cavity with no signs of solid i.p. tumor
or of disease at any other site. The rHuIFN-7 treated mice died
of s.c. tumors that grew at the site of tumor cell injection. There
was no intraabdominal tumor in IFN-7 treated mice at post
mortem examination. In view of the length of survival of mice
at this time point, it is not unreasonable to assume that mice
were in fact cured of their i.p. disease by IFN-7 treatment.
rHuIFN-7 also increased survival in mice bearing the LA
xenograft (P < 0.001 daily or twice weekly treatment versus
2 Roussel UCLAF, unpublished data.
diluent treated) but daily therapy was significantly better than
twice weekly (P < 0.002). Most of the IFN treated LA bearing
mice died of s.c. tumors at the site of tumor injection and there
was no intraabdominal tumor in the mice that received daily
therapy.
Effects of rHuIFN-7 on Solid i.p. Tumors. As previously
described (17) i.p. TNF therapy of the HU and LA xenografts
eradicated ascitic disease and promoted the formation of mul
tiple solid i.p. tumor deposits that closely resembled the histo
lógica!appearance of stage 3 ovarian cancer in women. After 7
days of therapy with 1 pg rHuTNF i.p., a solid tumor burden
of 300 mg-1 g could be excised postmortem. All therapy
stopped for 24 h, and at this point no TNF was detectable in
peritoneal washouts. rHuIFN-7 treatment was started. Typical
experiments with the HU and LA xenografts are shown in Fig.
3. Daily, but not twice weekly, therapy prolonged survival of
mice bearing the HU xenograft (P < 0.005), but neither sched
ule was active against the LA xenograft.
We then assessed the effect of rHuIFN-7 therapy on solid
tumor burden in the HU xenograft. In two of three experiments
there was a significant difference in tumor weight in the
rHuIFN-7 treated mice compared with diluent treated mice.
This was first seen 4 days after therapy began. Fig. 4A shows
an experiment in which the effect of 1.5 x IO5 units rHuIFN7, and 5 times that amount, was assessed on tumor burden over
21 days of therapy. With the higher dose there was a significant
difference compared to diluent treated mice at 4 days (P <
0.005) and thereafter at all time points and both doses, rHuIFN7 treated tumor burden was less than that of control mice (P<
0.09-P < 0.003). Although the tumor burden in mice treated
with 7.5 x 10s units was lower than those treated with 1.5 X
10' units, the differences between the groups were not
significant.
Thus treatment with rHuIFN-7 caused primarily a tumor
stasis as measured by i.p. tumor burden. It is important to note
that tumor burden did not actually decrease in the group treated
with 1.5 x IO5 units IFN-7. In the high dose group the tumor
weight at day 0 was 443 ±50 (SE) mg, and at day 21 it was
317 ±85 mg, but this difference was not statistically significant.
Dose and Scheduling of rHuIFN-7. The experiment in Fig. 5
demonstrated that significant effects on survival could be
achieved in mice bearing the HU xenograft with even 1 week
of therapy (P < 0.05) although 4 and 8 weeks of therapy gave
greater survival (P < 0.01). We next investigated whether the
LA xenograft was responsive to higher doses of rHuIFN-7.
Although approximately 50% of mice died fairly rapidly, the
6645
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1991 American Association for Cancer Research.
T-INTERFERON
AND OVARIAN CANCER XENOGRAFTS
other 50% were long-term survivors when mice were treated at
twice or 5 times the doses used in the other experiments (P <
0.05) (Fig. 6). In this instance, a significant difference in tumor
weight was seen at 4 weeks in the mice treated with the highest
dose of rHuIFN-7 (7.5 x IO5units/mouse; P< 0.05) compared
with mice treated with control diluent. Once again tumor bur
den was not decreased from starting levels (Fig. 4B).
Effect of rRaIFN-7 on the Human Xenografts. We also as
sessed the effect of a preparation of that has biological activity
on murine cells in vitro and in vivo. At equivalent doses and
schedules, rRaIFN-7 did not increase the life span of mice
bearing solid tumors of the H U and LA xenografts (data not
shown). The survival of mice bearing the HU xenograft was, in
fact, slightly but significantly decreased in the groups treated
with daily and twice weekly rat IFN-7 (P < 0.05), but we could
find no explanation for this after detailed postmortem
examination.
Histology of rHuIFN-7 Treated Tumors. Solid tumors were
excised at regular intervals throughout the experimental period
and examined histologically. Sections were stained with hematoxylin and eosin and with periodic acid-Schiff diastase and
Alcian blue stains for mucins. All sections were coded and
assessed "blind" for general histopathological features, partic
ularly architecture and cytological features including mitosis,
inflammation, individual cell death (apoptosis), and areas of
extensive necrosis. Low power optical fields were quartered
over the whole tumor section for the assessment. The LA
xenograft was composed of nodular aggregates of mucin se
creting adenocarcinoma cells arranged in compact trabecular
and cribriform islands within pools of extracellular epithelial
acidic and neutral mucin. There were foci of apoptotic cell
death in control tumors, with abundant nuclear debris, occu
pying between 10 and 30% of the cross-sectional area of the
tumors, but frank necrosis was not prominent. After 1 week of
IFN-7 therapy, particularly 3 and 7.5 x IO5 units daily, there
was an increase in necrosis, the tumor cells appeared more
degenerate and less cohesive and did not form the even trabec
ular and cribriform arrangement seen in the control tumors.
These changes became more prominent with time and there
was always more necrosis, less organization, and more evidence
of apoptosis in the tumors treated with the higher doses of IFNy. Large foci of hypocellular mucin were seen in treated but
not in control tumors (Fig. 7). There was no detectable altera
tion in inflammatory cell infiltrate in treated tumors. The tumor
cells in the HU xenograft were larger, formed irregular clusters
and coarser clumps, and also produced acidic and neutral
epithelial mucin. In the treated groups, particularly after 2
weeks of therapy, the cells were more dispersed and showed
generalized loss of cohesion. There was an overall reduction in
the number of epithelial cells after IFN-7 therapy, again with
large areas of hypocellular mucin. The histológica! changes
resulting in large areas of acellular mucin offer an explanation
for the fact that a survival benefit can sometimes be achieved
without a reduction in tumor weight.
B
therapy
therapy
1000—Q-
90
Fig. 3. Survival of mice bearing solid peri
toneal tumor xenografts of human ovarian can
cer. Mice were given i.p. injections of tumor
cells. There were 8 mice in each experimental
group. After 7 days, mice received rHuTNF, 1
¿ig/mousei.p. for 7 days. After 1 day of rest, 8
weeks of rHuIFN-i therapy began. A, HU
xenograft; B, LA xenograft. -.-.-,
diluent
treated mice; O, rHuIFN-f, 1.5 x 10' units
twice weekly; •¿,
rHuIFN--y 1.5 x 10' units
daily. In A daily IFN therapy gave a significant
increase in survival (P < 0.005).
10l
'
90
80
(0
0)
70
70
60
60
SO
SO
40
40
30
30
20
20
10
10
o
20
40
60
80
Jj
100 120 140 160
)
20
ç
i
40
60
80
100 120 140 160
days after tumor injection
B
Fig. 4. Effect of rHulFN-7 therapy on solid
i.p.tumor weights. At weekly intervals after the
start of therapy mice were selected at random
and all solid tumor was excised postmortem
and weighed. There were 3-5 mice in each
experimental group. A, HU xenograft. O, di
luent treated; •¿,
rHu!FN--y, 1.5 x 10' daily;
B,rHuIFN--x,7.5 x 10'daily. B, LA xenograft.
O, diluent treated; •¿,
rHuIFN-i, 1.5 x 10*
daily; D, rHuIFN--y, 3.0 x 10' daily; •¿.
rHuIFN--y, 7.5 x 10' daily. »,P < 0.05 com
pared to diluent treated.
10
15
10
15
20
25
days after start of therapy
6646
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1991 American Association for Cancer Research.
30
7-INTERFERON
AND OVARIAN CANCER XENOGRAFTS
B
Fig. 5. Scheduling of rHuIFN-7 treatment
of solid i.p. tumors. Mice were given i.p. injec
tions of HU tumor cells. There were 8 mice in
each experimental group. After 7 days, mice
received rHuTNF, 1 iig/mouse i.p. for 7 days.
After 1 day of rest, rHuIFN-7 therapy (1.5 x
IO5 units) began. A, diluent treated versus 1
and 2 weeks rHuIFN-7 therapy. -.-.-,
diluent treated; •¿,
1 week daily therapy; O, 2
weeks daily therapy. B, diluent treated versus
4 and 8 weeks rHuIFN-7 therapy. -.-.-,
diluent treated; •¿,
4 weeks daily therapy; O, 8
weeks daily therapy. There was a significant
increase in survival in all treated groups (P <
0.05 for 1 week therapy, P< 0.01 for the other
groups).
T!
,•
¿
I,
iI•!.........
m>"g3CO2r*90807060SO403020100;
ii
oo11
.
20
¡'•
, 1 ,
•¿â€¢y*ìiio111•-»
40
60
80
100
100
days after tumor injection
Tryptophan Levels in the Tumor. The induction of the enzyme
IDO which metabolizes tryptophan to kynurenine, thus deplet
ing the cellular pool of this amino acid, is one mechanism by
which IFN-7 inhibits the growth of some tumor cell lines in
vitro (21-24). We therefore investigated levels of tryptophan
and kynurenine in rHuIFN-7 treated tumors. Fig. 8 shows the
mean results from two separate experiments that gave similar
results. These experiments were with the HU xenograft treated
as a solid tumor with daily injections of rHuIFN-7. There was
no change in tryptophan or kynurenine levels 2 h after the first
IFN-7 treatment, but there was a significant fall in tryptophan
(P < 0.001) at 24 h and a corresponding rise in kynurenine (P
< 0.005) (Fig. 8). Significant but lesser changes were seen at 2
and 4 days. Although the overall decrease in tryptophan con
centration in the tumor cells as a whole was only 20-30%, the
decrease of tryptophan in the tumors may be much greater than
this because of the presence of a significant extracellular volume
in the tumors (see above). By 7 days there was no significant
difference between diluent and rHuIFN-7 treated tumors. This
may reflect the disappearance of viable cells from the treated
therapy
>
Ico
40
60
80
100
120
140
160
days after tumor injection
Fig. 6. The effect of dose of rHuIFN-i on survival of mice bearing the LA
xenograft. Mice were given i.p. injections of LA tumor cells. There were 8 mice
in each experimental group. After 7 days, mice received rHuTNF, 1 ¿ig/mouse
i.p. for 7 days. After 1 day of rest, rHuIFN-i therapy began. -.-.-,
diluent
treated; •¿,
1.5 x 10* units rHuIFN-7; A, 3.0. x 10' units rHuIFN--y; »,7.5 x
10* units rHuIFN-7. There was a significant increase in survival in mice treated
with 3 and 7.5 x 10' units rHuIFN-7 (P < 0.05).
tumors, as is suggested by the histológica! observations, or may
suggest a tachyphylaxis to the effects of IFN-7.
The changes in tryptophan/kynurenine after rHuIFN-7 treat
ment of the tumor were compared with levels in plasma. When
there were significant changes in tumor levels, there were no
changes in the mouse plasma (data not shown). In contrast,
rRaIFN-7 given i.p. at the same dose decreased plasma tryp
tophan and raised plasma kynurenine. For instance, 24 h after
i.p. injection, the plasma tryptophan levels in control or diluent
treated mice were 92 ±4.3 and 100 ±3.4 M, respectively. In
rHuIFN-7 treated mice levels were 95 ±3.8 M, but in rRalFN7 treated mice they were 63 ±6.5 M (a significant difference; P
< 0.01). Similarly the plasma kynurenine levels were 3.3 ±
0.43, 2.6 ±0.44, and 3.9 ±0.59 for control, diluent treated,
and rHuIFN-7 treated mice but 7.6 ±1.15 in the rRaIFN-7
treated mice (P < 0.005). These data also confirm that the rat
IFN-7 has biological activity in the nude mice and are analogous
to those showing a decrease in serum tryptophan and an in
crease in urinary kynurenine in human patients treated with
human IFN-7 (25).
These results show for the first time that tryptophan can be
depleted by conversion to kynurenine in tumors in vivo on
treatment with IFN-7, as it is in vitro in tumor cell culture
following induction of IDO by IFN-7 (21, 22). Induction of
IDO has been observed in allografted tumor cells undergoing
rejection, probably as a consequence of production of endoge
nous IFN-7 (26, 27). These observations support the hypothesis
that at least part of the antitumor effects of IFN-7 may be
mediated by tryptophan depletion, although it is important to
note that the changes in tryptophan levels occurred only in the
first few days of therapy, whereas prolonged exposure to IFN7 was necessary for maximal therapeutic benefit. Depletion of
tryptophan may represent one necessary step in the pathway of
IFN-7 mediated cytostasis and cytotoxicity in vivo. Our results
have two main implications. First induction of IDO may be a
marker predictive of response. We are currently establishing a
panel of ovarian cancer xenografts which will be used to inves
tigate this possibility. Unpublished data from earlier studies are
of interest in this respect. IFN-7 had had only weak antitumor
activity in a breast cancer xenograft that was growing s.c. in
nude mice and had no effect on tryptophan and kynurenine
levels. The second implication is that modulating this pathway
of tryptophan metabolism may have therapeutic potential. It
will also be of interest to look at the effects of a combination
of rat and human IFN-7 in these models inasmuch as systemic
6647
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1991 American Association for Cancer Research.
•¿/-INTERFERON
AND OVARIAN CANCER XENOGRAFTS
1-3mm
1-3mm
A-
;.
A
B
Fig. 7. Histology of rHuIFN-7 treated LA tumors. Mice were given i.p. injections of LA tumor cells. After 7 days, mice received rHuTNF, 1 «ig/mousei.p. for 7
days. After 1 day of rest, rHuIFN-f therapy (7.5 x 10* units daily) began. Tumors were excised after 2 weeks IFN-i treatment. A, diluent treated; B, rHuIFN--x
treated, x 6.4.
A
Fig. 8. Measurement of tumor tryptophan
and kynurenine. Mice were given i.p. injections
of HU tumor cells. There were 5 mice in each
experimental group. After 7 days, mice re
ceived rHuTNF, 1 pg/mouse i.p. for 7 days.
After 1 day of rest, rHuIFN-f therapy (1.5 x
10' units) began. Tumors were excised and
immediately snap frozen. I. tumor tryptophan.
O, diluent treated; •¿,
rHuIFN-y, 1.5 x 10s
units daily. li. tumor kynurenine. O, diluent
treated; »,rHuIFN-7 1.5 x 10' units daily. **,
different from diluent treated (P < 0.001); *,
different from diluent treated (P < 0.005).
B
C
n
CL
O
O
E
D
so
days IFN treatment
depletion of tryptophan may exacerbate the effects occurring in
the tumors. This would also be a better model for the action of
IFN-7 on tryptophan metabolism in human ovarian cancer
patients where effects on both blood and tumor tryptophan
levels would be expected.
This study of rHuIFN--y therapy in human ovarian cancer
xenograft models has demonstrated several points that may be
clinically relevant. There appeared to be a differential sensitivity
between ascitic free floating tumors and solid tumor implants
to the antitumor effects of rHuIFN-7. Thus in the HU xenograft
model, the median survival of mice bearing ascitic tumor only
and treated with daily rHuIFN-7 was greater than 250 days,
compared to a median survival of 100 days in mice bearing
solid HU tumor implants. Similarly, daily therapy of ascitic LA
xenografts with rHuIFN-7 at a dose of 1.5 x IO5 units/day for
8 weeks prolonged the survival of ascites bearing mice, but the
same dose schedule was ineffective in mice bearing solid LA
tumor implants. The explanation for this may be related to the
greater access of rHuIFN-7 to the tumor cells in the ascitic
state and the prolonged retention of rHuIFN-7 in the peritoneal
cavity of mice with ascites. The importance of continued re
peated exposure to rHuIFN-7 to its antitumor effects was
suggested by the greater antitumor effects of daily therapy
compared to twice weekly therapy in the ascitic LA and OS
xenograft models and in mice bearing solid HU tumor implants.
However, in the solid tumor model, there was evidence that the
survival benefit stops with cessation of treatment and that
maximal benefit was seen with 4 weeks of therapy. Mice treated
for 8 weeks survived a shorter time after cessation of therapy
than did mice treated for 4 weeks. In fact, there was no differ
ence in survival between mice treated for 4 and 8 weeks in the
experiments described here. One possible explanation is that
some cells within the tumor become resistant to the cytostatic/
cytotoxic action of IFN-7 with continued treatment. Fig. 4A
suggests that toward the end of a 3-week course of therapy, the
growth rate of the tumors is beginning to increase again. Ex-
6648
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1991 American Association for Cancer Research.
7-INTERFERON
AND OVARIAN CANCER XENOGRAFTS
periments with IFN-7 maintenance therapy for indefinite pe
riods and an assessment of the response of tumors transplanted
from IFN-7 treated mice would be useful in further exploring
this issue. There was a clear difference in the sensitivity of the
different xenografts to rHuIFN-7 therapy but there was no
apparent correlation with the histological grade of tumor and
sensitivity to rHuIFN-7.
In modeling the action of anticancer agents in animals it is
important to attain similar circulating levels in the mice and
humans. We used the conversion factor of Friereich et al. (20)
which was developed for chemotherapeutic drugs. Our results
show that this is a valid way to calculate cytokine doses to be
administered to mice.
Dose-response experiments in the solid tumor models
showed that the antitumor effect not apparent at standard doses
of rHuIFN-7 could be seen at higher doses but that there was
a limit to this effect in the LA xenograft model. The duration
of treatment was also shown to be important in the HU xeno
graft model.
In conclusion the study of i.p. rHuIFN-7 therapy in three
xenograft models suggested that the antitumor effects of
rHuIFN-7 in i.p. therapy in clinical practice may be best seen
with continuous, prolonged, and high dose therapy with this
cytokine. Further experiments are under way to assess param
eters, such as dose intensity, and modulations of tryptophan
levels, that may improve the efficacy of rHuIFN-7 therapy in
these models. If the results are extrapolated to clinical practice,
antitumor effects should be best seen in patients with ascitic
tumors. However, although current trials of i.p. rHuIFN-7
therapy show that tumor responses are generally noted only in
patients with small tumor volumes, the xenograft data suggest
that with appropriate dose scheduling, tumor stasis and even
regression may also be seen in patients with bulky intraabdominal disease.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
ACKNOWLEDGMENTS
We thank Hazel Holdsworth for excellent technical assistance and
M. M. Mouren for valuable help with the pharmacokinetic studies. We
also thank Professor W. Fiers for the recombinant human TNF.
21.
22.
REFERENCES
1. Silverberg, E.. Boring, C. C, and Squires, T. S. Cancer statistics 1990. CaCancer J. Clin., 40: 9-26, 1990.
2. Bargmann, F. Carcinoma of the ovary, a clinicopathological study of 86
autopsied cases with special reference to the mode of spread. Acta Gynecol.
Scand., 45:211-221, 1966.
3. Tobias, J. S., and Griffiths, C. T. Management of ovarian carcinoma. N.
Engl. J. Med., 294: 877-882. 1976.
4. Oozols, R., GarvÃ-n,A. J., and Costa, J. Advanced ovarian cancers: correlation
of histologie grade with response to therapy and survival. Cancer (Phila.),
45:572-581, 1980.
5. Berek. J. S. Intraperitoneal adoptive immunotherapy for peritoneal cancer.
J. Clin. Oncol.. A: 1610-1612. 1991.
6. Rath, U., Kaufmann, M., Schmid, H„Hofmann, J., Wiedenmann. B.. Kist.
A., Kempeni. J., Schlick, E., Basiert. G.. Kommerell, B., and Mannel, D.
Effect of intraperitoneal recombinant human tumor necrosis factor a on
malignant ascites. Eur. J. Cancer, 27: 121-125, 1991.
7. Rambaldi, A., Introna, M., Coletta, F.,Landolfo, S., Colombo, N., Mangioni,
C., and Mantovani, A. Intraperitoneal administration of interferon beta in
ovarian cancer patients. Cancer (Phila.), 56: 294-301, 1985.
8. Stewart, J. A., Belinson, J. L., Moore, A. L., Dorighi, J. A., Grant. B. VV.,
23.
24.
25.
26.
27.
Haugh, L. D., Roberts, J. D., Albertini. R. J.. and Brandan, R. F. Phase I
trial of intraperitoneal recombinant interleukin-2/lymphokine-activated
killer cells in patients with ovarian cancer. Cancer Res., 50: 6302-6310,
1990.
Steis, R. G., Urba, W. J., VanderMolen. L. A.. Bookman. M. A.. Smith II,
J. W., Clark, J. W.. Miller, R. L., Crum. E. D., Beckner, S. K.. McKnight,
J. E., Ozols, R. F., Stevenson, H. C., Young, R. C., and Longo, D. L.
Intraperitoneal lymphokine-activated killer-cell and interleukin-2 therapy for
malignancies limited to the peritoneal cavity. J. Clin. Oncol.. 8: 1618-1629.
1990.
Welander. C. E. Interferon in the treatment of ovarian cancer. Semin. Oncol.,
/5(Suppl. 5): 26-29, 1988.
Berek, J. S., Hacker, N. F., Lichtenstein, A., Jung, T., Spina, C., Knox, R.
M., Brady, J., Greene, T., Ettinger, L. M., Lagasse, L. D„Bonnern, E. M.,
Spiegel, R. J.. and Zighelboim. J. Intraperitoneal recombinant a-interferon
for "salvage" immunotherapy in stage III epithelial ovarian cancer: a gyne
cologic oncology group study. Cancer Res., 45:4447-4453, 1985.
Willemse, P. H. B.. De Vries, E. G. E., Mulder, N. H., Aalders, J. G.,
Bouma, J.. and Sleijfer. D. T. Intraperitoncal human recombinant interferon
alpha-2b in minimal residual ovarian cancer. Eur. J. Cancer, 26: 353-358,
1990.
Allavena, P., Peccatori, F., Maggioni. D.. Erroi, A.. Sironi. M.. Colombo,
N., Lissoni, A., Galazka, A., Meiers. W., Mangioni. C., and Mantovani, A.
Intraperitoneal recombinant 7-interferon in patients with recurrent ascitic
ovarian carcinoma: modulation of cytotoxicity and cytokine production in
tumor-associated effectors and of major histocompatibility antigen expres
sion on tumor cells. Cancer Res.. 50: 7318-7323. 1990.
D'Acquisto, R., Markman, M., Hakes, T., Rubin. S., Hoskins, W., and
Lewis, J. L. A phase I trial of intraperitoneal recombinant 7-interferon in
advanced ovarian carcinoma. J. Clin. Oncol.. 6: 689-695, 1988.
Welander, C. E., Homesley, H. D.. Reich, S. D., and Levin, E. A. A phase
II study of the efficacy of recombinant interferon 7 in relapsing ovarian
adenocarcinoma. Am. J. Clin. Oncol., //: 465-469. 1988.
Pujade-Lauraine, E., Guastella, J. P., Colombo. N.. Namer, N.. Fumoleau,
P., Monnier, A.. Nooy, M. A., Falkson. G., Mignot. L., Bugat. R., Oliveira,
C. M. D., Mousseau, M., Netter, G., Oberling. F., Coiffier, B., and Brandely,
M. Intraperitoneal human r-IFN 7 as treatment of residual carcinoma (OC)
and laparotomy (SLL). Proc. Am. Soc. Clin. Oncol.. 10: 195, 1991.
Malik, S. T. A., Griffin. D. B.. Fiers. W.. and Balkwill. F. R. Paradoxical
effects of tumor necrosis factor in experimental ovarian cancer. Int. J.
Cancer., ¥4:918-925, 1989.
Naylor, M. S.. Malik, S. T. A., Stamp. G. W. H.. Jobling, T.. and Balkwill,
F. R. In situ detection of tumor necrosis factor in human ovarian cancer
specimens. Eur. J. Cancer, 26: 1027-1030, 1990.
Malik, S. T. A., Naylor, M. S., East, N.. Oliff. A., and Balkwill, F. R. Cells
secreting tumor necrosis factor show enhanced metastasis in nude mice. Eur.
J. Cancer, 26: 1031-1034, 1990.
Friereich, E. J. Gehan, R. A.. Rail. D. A.. Schmidt, L. H., and Skipper, H.
E. Quantitative comparison of toxicity of anticancer agents in mouse, rat,
hamster, dog. monkey and man. Cancer Chemother. Rep.. 50: 219-244,
1966.
Takikawa. O.. Kuroiwa. T.. Yamazaki. F.. and Kido, R. Mechanism of
interferon action—characterization of indolcamine 2.3-dioxygenase induc
tion in cultured human cells by ¡nterferon-7 and evaluation of the enzyme
catalyzed tryptophan degradation in the anticellular activity. J. Biol. Chem.,
263: 2041-2048. 1988.
Ozaki, Y., Edelstein, M. P., and Duch. D. S. Induction of indolcamine 2,3dioxygenase: a mechanism of the antitumor activity of interferon 7. Proc.
Nati. Acad. Sci. USA, 85: 1242-1246, 1988.
de la Maza, L. M., and Peterson, E. M. Dependence of the in vitro antiproliferative activity of recombinant human 7-interferon on the concentration
of tryptophan in culture media. Cancer Res., 48: 346-350, 1988.
Knowles, R. G., Clarkson. N. A., Pogson, C. L, Salter, M., Duch, D. S., and
Edelstein, M. P. The role of tryptophan and kynurenine transport in the
catabolism of tryptophan through indoleamine 2.3-dioxygenase. In: R.
Schwarcz, S. Young, and R. R. Brown (eds.). Kynurenine and Serotonin
Pathways: Progress in Tryptophan Research, pp. 161-166. New York:
Plenum Press, 1991.
Schiller. J. H.. Storer. B.. Pualnock, D. M.. Brown. R. R.. Datta. S. P.. Witt.
P. L., and Borden, E. C. A direct comparison of biological response modu
lation and clinical side effects of interferon-/^,. interferon-7. or the combi
nation of interferons ß„,
and y in humans. J. Clin. Invest., 86: 1211-1221,
1990.
Yoshida, R.. Park. S. W., Yasui, H., and Takikawa, O. Tryptophan degra
dation in transplanted tumor cells undergoing rejection. J. Immun»]..141:
2819-2823, 1988.
Takikawa, O., Habara-Phkubo. A., and Yoshida, R. IFN-7 is the inducer of
indoleamine 2,3-dioxygenase in allografted tumor cells undergoing rejection.
J. Immunol.. 145: 1246-1250, 1990.
6649
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1991 American Association for Cancer Research.
Antitumor Activity of γ-Interferon in Ascitic and Solid Tumor
Models of Human Ovarian Cancer
Saleem T. A. Malik, Richard G. Knowles, Nick East, et al.
Cancer Res 1991;51:6643-6649.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/51/24/6643
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 14, 2017. © 1991 American Association for Cancer Research.