- Email: [email protected]
Malignant mesothelioma: the antiproliferative effect of cytokine combinations on three human mesothelioma cell lines
205
Cancer Letters, 58 (1991) 205-210 Elsevier Scientific
Publishers
Ireland
Ltd.
Malignant mesothelioma: the antiproliferative combinations on three human mesothelioma A.M. Hand”, K. Husgafvel-Pursiainen”, K. Linnainmaaa
L. Tammilehtoa7b,
effect of cytokine cell lines
K. Mattsonb and
‘Institute of Occupational Health, Topeliuksenkatu, 41 aA, 00250, Helsinki and bDepartment Helsinki University Central Hospital, Haartmaninkatu 4. 00290 Helsinki (Finland) (Received (Accepted
rTNF; Keywords: thelioma; in vitro
(TNF) combined (IFNy) is known to haoe antiproliferatioe effects on many tumour cells, both in oitro and in uiuo. We investigated whether human mesothelioma cells would respond in a similar way. Mesothelioma cell lines established from primary tumours did not respond significantly in uitro to either TNF or lFNy alone but were inhibited by combinations of TNF and IFNy at concentrations as low as 5 ng/ml. In contrast, a mesothelioma cell line established from a metastatic tumour was sensitive to IFIVy both alone and in combination with TNF but not to TNF alone. We also looked at the responses of one primary tumour cell line and the metastatic tumour cell line to alpha-interferon (IFNor) both alone and in combination with TNF. Both cell lines were Tumour
necrosis
more
sensitiue
to
We conclude that these low dose combinations of cytokines are worth further investigation in the deuelopment of mesothelioma therapy.
Correspondence to: Dr. K. Linnainmaa, Ph.D., Institute of Occupational Health, Dept Industrial Hygiene and Toxicology, Laboratory,
0304-3835/91/$03.50 Published and Printed
rIFNa,;
rlFNy;
meso-
factor
gamma-interferon
sensitiue to IFNol but were the THF/lFNor combination,
Mutagen Finland.
Medicine,
26 March 1991) 25 April 1991)
Summary
with
of Pulmonary
Topeliuksenkatu
41 a A, 00250
0 1991 Elsevier Scientific in Ireland
Helsinki,
Publishers
Introduction Mesothelioma is a tumour of the lining of the pleural and peritoneal cavities, thought to originate from mesothelial cells. It has been established that asbestos exposure is almost always associated with the development of diffuse malignant mesothelioma [ 11. With a latent period of 30-40 years, the recent increase in the incidence of mesothelioma may be expected to continue, following the increasing use of asbestos between the Second World War and the 1970s. There is no standard treatment, but multi-modality therapy, including surgical resection radiotherapy and chemotherapy, is currently practised in many centres. Survival in this disease is generally poor; typically less than 1 year from the initiation of treatment [Z]. The incidence of this disease in Finland has risen lo-fold in the last decade to 50 new cases per year in 1987 [3]. Tumour necrosis factor (TNF) is a cytokine derived from activated monocytes known to have cytostatic and cytotoxic activity against a variety of tumour cells as well as a range of other biological activities [4,5]. This activity can be enhanced by gamma-interferon (IFNy) Ireland Ltd.
206
and synergistic antiproliferative effects of combinations of TNF and IFNy have been described in a number of tumour cell lines in vitro [6,7] and tumour xenografts in vivo [8]. However, the use of these cytokines individually in cancer therapy has produced variable results and dose-limiting side effects [9- 1 l] and they are even less well tolerated clinically in combination [ 12,131. that human We sought to establish mesothelioma cells in vitro are sensitive to very low doses of human recombinant tumor necrosis factor (rTNF) combined with human recombinant gamma interferon (rIFNy) . We also looked at the responses of some of the cell lines to recombinant human alpha-interferon (rIFNa) both alone and in combination with rTNF as some tumour types which are not sensitive to r1FN-y can be sensitive to rIFNar [14]. Materfals
and Methods
The three human mesothelioma cell lines used in these experiments were originally established and characterised in our laboratory from fresh tumour tissue samples [15]. Cell lines M14K and M38K were established from primary tumour tissue from untreated patients. M9K was established from a metastasis of a patient who had undergone chemotherapy. From continuously growing monolayer cultures. l.O- 1.5 x lo5 cells were plated into 6-well plates (Nunc) in 3 ml medium (RPMI- 1640 supplemented with 10 ok; foetal calf serum, 0.03% L-glutamine, 100 units/ml penicillin and 100 pg/ml streptomycin, all from GIBCO). The plates were incubated at 37°C in a humidified atmosphere with 5% CO2 for 40-50 h to establish exponentially growing cultures [15]. The cultures were then treated with the cytokines separately and in combination and incubated for a further 72 h before harvesting. The number of living cells was then assessed by the trypan blue exclusion method. To test the in vitro effects of rTNF and rIFNy, the cultures were treated with rTNF rIFNy or rTNF + rIFNy over a range of
6-7000 ng/ml for rTNF, l-1000 ng/ml for rIFNy or 6-7000 ng/ml for rTNF + rIFNy In the second series of experiments cell lines M38K and M9K were treated with rTNF, rIFNy, rIFNar, rTNF + rIFNr or rTNF + rIFNa all in the range 10-100 ng/ml for M38K, 0.1-100 ng/mI for M9K. The cytokines, human rTNFcr (spec. act. 6 x lo7 units/mg), human rIFNr (spec. act. 2 x lo7 units/mg) and human rIFNa-2C (spec. act. 3;2 x 10’) were all obtained from Boehringer Ingelheim. The results of the treatments are presented as the percentage of the number of control cells that survived (% survival) against the amounts of rTNF, rIFNr and rIFNar used. Each point on the figures represents the mean of 2 experiments of duplicate cultures (3 experiments in Fig. 2) f standard deviation. The multiplicative model was used to evaluate the synergistic/additive interaction of rTNF with rIFNrv and rIFNr [16,17]. Results The cell lines generally showed similar responses to treatment with the cytokines. The number of viable cells in the confluent control cultures after 5 days ranged from 5.8 x lo5 to 9.0 x 105. All three cell lines were insensitive to rTNF (< 30% inhibition of growth). Two cell lines (M14K and M9K) were mildly sensitive to r1FN-y alone (30-50% inhibition of growth) but the third cell line (M38K) was not. All the cell lines were sensitive to rTNF in combination with r1FN-y (> 50% inhibition of growth) (Fig. 1). The two cell lines established from primary tumours showed a synergistic response to the cytokine combinations, as assessed by the multiplicative model [16] (Table I). The metastatic cell line (M9K) showed only an additive response, however. The cytokine combination response was dose dependent with 10 ng/ml of each cytokine being the effective minimum dose. At lower doses the cytokines, both alone and in combination, had no inhibitory effect and at very low doses (0.1 proliferative moderate effects, rig/ml)
207
a primary tumour and M9K, from a metastasis) to alpha-interferon and gamma-interferon, both alone and in combination with tumour necrosis factor. Both cell lines were mildly sensitive to rIFNa, alone (30-50’S inhibition) but the combination of rIFNa and rTNF was more effective (60% inhibition) (Fig. 2). The response of the metastatic cell line M9K to the
120
M36K
loo-
80.
m _f gl
60-
E 40.
20.
10'
Concentration
102
(nglml)
40.
Fis. 1. The effect of tumour necrosis factor (TNF) and gamma-interferon (IFNy), both alone and in combination, on 3 mesothelioma cell lines. ( l - l) TNF alone, ( x - x ) 1FN-y alone, ( o - o ) TNF + IFNy in the ratio of 111 (w/w).
ZO-30% increase in growth over the controls, were observed (data not shown). The second series of experiments was designed to compare and contrast the responses of two of the cell lines (M38K. from
ZO-
I.
Id
IO0 Concentration
10'
102
(nglml)
Fig. 2. The effect of alpha-interferon (IFNa) and gamma-interferon (IFNy), both alone and in combination with tumour necrosis factor (TNF), on 2 mesothelioma cell lines. ( n - m) IFNar alone, (X - x ) IFNy alone, ( l - l ) TNF alone, ( o - q) IFNa + TNF (in the ratio of l:l), (o - O) IFNy + TNF (in the ratio of 1:l).
208
Table
1.
recombinant Cell line
The interaction of recombinant tumour necrosis factor (rTNF), recombinant gamma-interferon (r1FN-r) and alpha-interferon (rIFNcr) on 3 mesothelioma cell lines presented as the survival fraction (SF) of the cells. Dose (ng/mf)
Observed SF,,
survival SF,,,,
Predicted SF,iW +
IFNyl
Additive SFTNF
8.3 83 830 8.3 83 830 5 10 50
M14K
M38K
M9K
5 10 50 5 10 50
M38K
M9K
1.06 0.96 0.81 0.73 0.69 0.70 0.88 0.85 0.82
0.71 0.53 0.47 0.82 0.89 0.88 0.54 0.50 0.48
0.29 0.18 0.14 0.55 0.34 0.29 0.45 0.39 0.31
SFTNF
SFIFN,
SF (TNF +
0.91 0.90 0.84 0.88 0.85 0.87
0.73 0.60 0.58 0.53 0.50 0.45
0.50 0.37 0.31 0.42 0.41 0.30
x
survival model Sh,
Synergistic Synergistic Synergistic Additive Synergistic Synergistic Additive Additive Additive
0.75 0.51 0.38 0.66 0.61 0.62 0.48 0.43 0.39 IFNa)
SF,,,
0.66 0.54 0.49 0.47 0.43 0.37
Type of interactiona
x
SF,,,,
Synergistic Synergistic Synergistic Additive Additive Additive
survival fraction with TNF SF,,, survival fraction with TNF alone; SFrFN., survival fraction with fFNy alone; SF(r,, + rFNYI and IFNy; SF,,,,, survival fraction with INFcy alone; SFtrNF + rFNb,survival fraction with TNF + IFNa. ‘Synergism exists if SFcrNF + rrNi is fess than SFrNr x SFfrN. The interaction is additive if SFcrxr + tFNris equal to SFrNr x
=,FN
WI
rTNF/rIFNcz combination was only an additive interaction similar to this cell line’s response to the rTNF/rIFNy combination. There was slight synergism in the response of M38K to the rTNF/rIFNcu combination but not such a marked response as with rTNF/rIFNy combination (Table I). To summarise our results, mesothelioma cells were inhibited by combinations of tumour necrosis factor with either alpha- or gammainterferon, although they were not sensitive to tumour necrosis factor alone and showed variable sensitivity to alpha- and gammainterferon alone. Recombinant human tumour necrosis factor plus recombinant human gamma-interferon had a synergistic inhibitory effect on mesothelioma cells. The combinations were potent inhibitors in our system at doses as low as 10 ng/ml of each cytokine.
Discussion Tumour necrosis factor regulates the growth of many normal cells as well as having a cytotoxic effect on many malignant cells [4,5]. It is thought to exert its cytostatic effect on tumour cells by stopping the cell cycle at the G2 phase [US]. Interferons exert their antitumour effects in a variety of different ways, both directly on the tumour cells and indirectly, by influencing the host’s response to the tumour [ 191. The synergism between tumour necrosis factor and interferon has been reported for various tumours both in in vitro and experimental in vivo studies [7,20,21]. Our results show that human malignant mesothelioma cell lines are also sensitive to the effects of rTNF combined with either rIFNa! or rIFNy, despite variable sensitivity to the dif-
209
ferent interferons alone. In parallel to the work of Lewis et al. [21] we have also observed that low-dose combinations can be inhibitory when the same dose of the agents alone produces a stimulatory effect. The mechanism of action of this synergy is unclear. IFNy may increase the cells’ sensitivity to TNF although this cannot be solely a function of increased number of TNF receptors (221 because other workers have shown that TNF/IFNy synergism can be independent of receptor availability [23,24]. Kavoussi et al. [25] have demonstrated that IFNy pretreatment of renal cell carcinoma cell lines improves the cells’ response to TNF but that the timing of the pretreatment is crucial and varies between the different cell lines. We have added the interferons at the same time as tumour necrosis factor, but by culturing with the cytokines for 72 h, approximately 2 cell doublings, we have taken into account possible variations between the cell lines in the time course of the response [16]. Tumour necrosis factor, alpha-interferon and gamma-interferon have been used separately and together in various clinical regimes [9,10,26]. The success of these regimes has however been variable and the use of these cytokines often associated with moderate to severe toxicity. In particular TNF combined with IFNy can be especially difficult to administer clinically, producing more severe side effects than the cytokines separately. However, there is evidence that the side effects can be controlled and that they are more severe in patients with haematological malignancies than in patients with solid tumours [12,13]. Los et al. [27] reported successful administration of chemotherapeutic agents directly into the peritoneal cavity in an ovarian cancer model. This method increases the drug concentration at the site of the tumour and reduces the systemic side effects of intravenous administration. There is evidence that tumour cell lines which are, or have become, resistant to chemotherapeutic drugs can be susceptible to treatment with combinations of TNF and IFNy
to overcome their resistance [17,28]. This points to a role for combined TNFAFN treatment in cases of chemotherapy failure such be experienced in malignant as may mesothelioma. Furthermore there is some clinical evidence that interferons should be used in combination with cytostatic drugs [26,29]. None of the chemotherapeutic agents currently in use for mesothelioma treatment is particularly effective so there is a need to test other agents for mesothelioma activity [30]. We aim to extend our in vitro studies to investigate the effective cytokine combinations reported here with the cytostatic drugs currently in use for mesothelioma treatment, to see if an improved response were possible. In conclusion, these experiments indicate that combinations of low doses of recombinant tumour necrosis factor and recombinant interferons effectively inhibit the growth of mesothelioma cells in vitro. Such combinations have been shown to be effective against other cancers but we believe this to be the first time that they have been shown to be effective against mesothelioma cells. The low doses required may offset the predicted clinical toxicities. The that this cytokine proposal combination could be used effectively to treat mesothelioma is attractive, given the current poor prognosis for mesothelioma patients and the resistance of mesothelioma to conventional chemotherapy, as well as multimodality schemes. Acknowledgements We would like to thank Mr. Hans Sarelin of Boehringer Ingelheim for providing the recombinant cytokines, and Ms. Madeleine Mattson, Ms. Satu Suhonen and Ms. Marjatta Vallas for their technical assistance. References Wagner. J.C.. Sleggs, C.A. and Marchand. P. (1960) Diffuse pleural mesotheliomas and asbestos exposure in the Northwestern Cape Province. Br. J. Ind. Med., 17, 260-271. Antman. K. and Aisner, J. (1987) Asbestos-Related Malignancy. Grune and Stratton, Orlando.
210
3
Huuskonen, Asbesti
M.S.,
Koskinen,
merkittava
Suomessa?
H.,
Roto, P. et al. (1987)
terveyshaitta
Suomen
Lalkarilehti.
vie12 17,
1669-
1676
B.J., Aggarwal,
Recombinant
5
with
of normal
230,
Semenzato,
Hass, P.E. et al. (1985)
human tumour necrosis factor-a:
proliferation Science,
B.B.,
and
transformed
multiple
Tumour
necrosis factor: a cytokine
activities.
Br.
L.,
Van
der Heyden,
J.
Cancer,
Recombinant
J., Ruysschaert,
18
R. and
on a variety of nor-
19 20
22, 419-426. Synergistic antitumor effects of tumor necrosis fac-
21
Cell
Gresser, I. (1989)
In vitro
human lung cancer
tumour
necrosis factor by virtue of
B.,
Carswell,
cycle-specific
1280.
E.A.
and
effects of tumour
Balkwill,
Antitumour
Balkwill,
F.R.,
Lee, A.,
F.R.,
Sherman,
M.L.,
Lewis, G.D., tumour
Aldan,
L. et al. (1986)
Ward,
B.G.,
tumor
22
Arthur, K.A.
et al. (1988)
23
6,
patients. J. Clin. Oncol., Kettlehut,
I.C.,
A.,
in cancer 25
A.L.
(1987)
Ruggiero,
84, 4273-4277.
Fiedler,
GBtz,
G.,
and Hossfeld.
J. Clin. Oncol.,
Jakubowski,
Larchian,
plus interferon-gamma
27
W.,
Starnes,
(IV) Tumour
Necrosis Factor
in patients with advanced
(Meeting Abstract) Proc. Annu.
C.
136, 2445.
M. and Pfizenmaier.
and characterization
K.
of high-affinity
receptors for tumor necrosis factor on human 38,
Kavoussi.
Hudson,
L.R.,
Effect
Ruesing, of tumor
on human
R.A.,
127-133.
necrosis factor
renal carcinoma
M.A. and
et al.
interferon
cell line growth.
J.
142, 875--878.
Strander,
H.A.
(1989)
Clinical effects of interferon therapy
Los, G.,
Ruevekamp,
28
M.,
Bosnie,
growth
N. et al. (1989)
and chemotherapy
Eur. J. Cancer Clin. Oncol..
Price,
G.,
Cytotoxic
cancer.
tumor Brenner,
K.,
Prentice,
effects of tumour
26,
1857-1866.
H.G.
et al.
(1987)
necrosis factor and gamma-
interferon on acute myeloid leukaemia
Meet. Am. Sot. Clin. On-
In-
in a rat
blasts. Br. J. Can-
cer, 55, 287-290.
Williamson, B.D., and Old, L. (1983)
Carswell, E., Rubin, B., Prendergast,
J.
29
human interferon. 5397-5401.
Proc.
Natl.
Acad.
K.. Husgafvel-Pursiainen,
Sci.
U.S.A.,
Ann. 30
L. et al. (1990) Asbestos-related malignant mesothelioma: Growth, tumorigenicity and CYtologY, consistent chromosome findings in cell lines from five patients. Car11, 673-681.
non-small
80.
K., Tammilehto,
Bowman,
A., Fergusson.
Potentiation
Human tumor necrosis factor produced
by human B-cell lines: Synergistic cytotoxic interaction with
cinogenesis,
J. Immunol.,
U., Kronke.
leukemic cell lines. Int. J. Cancer,
model.
F. et al. (1988)
col., 7. A634.
Pelin-Enlund,
P.. &er,
traperitoneal A.,
2441-2450.
J., Fiers, W. and Baglioni.
28, 355-362. 6,
1787-8. Phase-I trial of intravenous
136,
for tumour
with special emphasis on antitumour efficacy. Acta Oncol.,
D.K.
Increased toxicity of tumor necrosis factor/Gamma
interferon in patients with leukocytosis.
Res., 47. Interferon-
of cellular receptors
Tavernies.
Quantification
Ural.. 26
H.J.
V.,
Scheurich,
gamma
inhibitors. Proc. Natl. Acad.
Weh,
cells by human
Cancer
Induction of the synthesis of tumor necrosis factor
(1989)
The
necrosis factor in uiuo and their
Sci. U.S.A.,
Eessalu, T.E. et al. (1987)
and Vilcek, J. (1986)
expression
membrane
R. et al. (1986) gamma
4, 137-146.
prevention by cyclooxygenase W.,
Bhalla,
interferon
Fiers, W. and Goldberg,
toxic effects of tumor
B.B.,
Tsujimoto. M., Yip, Y.K.
(1986) Al-Katib,
Aggarwas.
receptors by interferon-y. 24
344-350. S.,
human ovarian cancer. Cancer
necrosis factor and interferon.
(1986)
as a 5-day continuous infusion in cancer patients: Phase 1
Phase 1 trial of recombinant
E. and Fiers. W.
of the growth of transformed
y enhances
with interferons.
human tumour necrosis factor administered
Vadhan-Raj,
Moodie,
potential of tumour necrosis factor and
necrosis factor. J. Immunol.,
Spriggs, D.R.,
Acta
5382-5385.
Human
human
effects of interferon.
28, 347-353.
Modulation
cell lines.
toxicity and effects on lipid metabolism. J. Clin. Oncol.,
15
8,
Res., 47, 4755-4758.
Bittner, G., Storer, B. and Willson, J.K.V.
Recombinant
14
human
Williamson,
(1984)
Oncologica.
Cancer Res.. 46, 3990-3993.
13
L.J.
Z..
interferon in experimental
(1988)
Y. et al. (1987)
human interferon-gamma
Oncol.,
Schiller, J.H.,
necrosis factor alone and in combination
12
recombinant
(1987) Therapeutic
tumor xerografts treated with recombinant
11
Saijo, N., Sasaki,
Darzynkiewicz, Old,
61,
Cancer Res.. 47, 2809-2813.
10
W.S.,
mal and transformed human cell lines. Eur. J. Cancer Clin.
tor and -y-interferon on human colon carcinoma
9
Hong,
and/or
tumor necrosis factor: its ef-
fect and its synergism with interferon-y
8
Ther.,
necrosis factor. Cancer Res., 44, 83-90.
Fransen,
(1987)
Pharmacol.
collateral sensitivity. Jpn J. Cancer Res., 78, 1274-
biological
Fiers, W. (1986)
7
In vitro systems for evaluation of
cell lines by recombinant
effects on
cells in vitro.
354-361. 6
(1980)
chemotherapy.
growth inhibition of cisplatin-resistant
943-945.
G. (1990)
R.L.
21-35.
(in 17
Sugarman,
Momparler, combination
Finnish). 4
16
2000-luvun
Allan, S.G.
et al. (1990)
of cisplatin by alpha-interferon
R.J.,
in advanced
cell lung cancer
Oncol..
(NSCLC):
1. 351-353.
Aisner, J. and Sigman, L.M. therapy
in Treatment
Asbestos-Related Antman
A phase II study.
(1987)
of Malignant
Malignancy.
The Role of ChemoMesothelioma.
pp. 385-401.
and J. Aisner. Grune and Stratton,
In:
Editors: K. Orlando.
Cancer Letters, 58 (1991) 205-210 Elsevier Scientific
Publishers
Ireland
Ltd.
Malignant mesothelioma: the antiproliferative combinations on three human mesothelioma A.M. Hand”, K. Husgafvel-Pursiainen”, K. Linnainmaaa
L. Tammilehtoa7b,
effect of cytokine cell lines
K. Mattsonb and
‘Institute of Occupational Health, Topeliuksenkatu, 41 aA, 00250, Helsinki and bDepartment Helsinki University Central Hospital, Haartmaninkatu 4. 00290 Helsinki (Finland) (Received (Accepted
rTNF; Keywords: thelioma; in vitro
(TNF) combined (IFNy) is known to haoe antiproliferatioe effects on many tumour cells, both in oitro and in uiuo. We investigated whether human mesothelioma cells would respond in a similar way. Mesothelioma cell lines established from primary tumours did not respond significantly in uitro to either TNF or lFNy alone but were inhibited by combinations of TNF and IFNy at concentrations as low as 5 ng/ml. In contrast, a mesothelioma cell line established from a metastatic tumour was sensitive to IFIVy both alone and in combination with TNF but not to TNF alone. We also looked at the responses of one primary tumour cell line and the metastatic tumour cell line to alpha-interferon (IFNor) both alone and in combination with TNF. Both cell lines were Tumour
necrosis
more
sensitiue
to
We conclude that these low dose combinations of cytokines are worth further investigation in the deuelopment of mesothelioma therapy.
Correspondence to: Dr. K. Linnainmaa, Ph.D., Institute of Occupational Health, Dept Industrial Hygiene and Toxicology, Laboratory,
0304-3835/91/$03.50 Published and Printed
rIFNa,;
rlFNy;
meso-
factor
gamma-interferon
sensitiue to IFNol but were the THF/lFNor combination,
Mutagen Finland.
Medicine,
26 March 1991) 25 April 1991)
Summary
with
of Pulmonary
Topeliuksenkatu
41 a A, 00250
0 1991 Elsevier Scientific in Ireland
Helsinki,
Publishers
Introduction Mesothelioma is a tumour of the lining of the pleural and peritoneal cavities, thought to originate from mesothelial cells. It has been established that asbestos exposure is almost always associated with the development of diffuse malignant mesothelioma [ 11. With a latent period of 30-40 years, the recent increase in the incidence of mesothelioma may be expected to continue, following the increasing use of asbestos between the Second World War and the 1970s. There is no standard treatment, but multi-modality therapy, including surgical resection radiotherapy and chemotherapy, is currently practised in many centres. Survival in this disease is generally poor; typically less than 1 year from the initiation of treatment [Z]. The incidence of this disease in Finland has risen lo-fold in the last decade to 50 new cases per year in 1987 [3]. Tumour necrosis factor (TNF) is a cytokine derived from activated monocytes known to have cytostatic and cytotoxic activity against a variety of tumour cells as well as a range of other biological activities [4,5]. This activity can be enhanced by gamma-interferon (IFNy) Ireland Ltd.
206
and synergistic antiproliferative effects of combinations of TNF and IFNy have been described in a number of tumour cell lines in vitro [6,7] and tumour xenografts in vivo [8]. However, the use of these cytokines individually in cancer therapy has produced variable results and dose-limiting side effects [9- 1 l] and they are even less well tolerated clinically in combination [ 12,131. that human We sought to establish mesothelioma cells in vitro are sensitive to very low doses of human recombinant tumor necrosis factor (rTNF) combined with human recombinant gamma interferon (rIFNy) . We also looked at the responses of some of the cell lines to recombinant human alpha-interferon (rIFNa) both alone and in combination with rTNF as some tumour types which are not sensitive to r1FN-y can be sensitive to rIFNar [14]. Materfals
and Methods
The three human mesothelioma cell lines used in these experiments were originally established and characterised in our laboratory from fresh tumour tissue samples [15]. Cell lines M14K and M38K were established from primary tumour tissue from untreated patients. M9K was established from a metastasis of a patient who had undergone chemotherapy. From continuously growing monolayer cultures. l.O- 1.5 x lo5 cells were plated into 6-well plates (Nunc) in 3 ml medium (RPMI- 1640 supplemented with 10 ok; foetal calf serum, 0.03% L-glutamine, 100 units/ml penicillin and 100 pg/ml streptomycin, all from GIBCO). The plates were incubated at 37°C in a humidified atmosphere with 5% CO2 for 40-50 h to establish exponentially growing cultures [15]. The cultures were then treated with the cytokines separately and in combination and incubated for a further 72 h before harvesting. The number of living cells was then assessed by the trypan blue exclusion method. To test the in vitro effects of rTNF and rIFNy, the cultures were treated with rTNF rIFNy or rTNF + rIFNy over a range of
6-7000 ng/ml for rTNF, l-1000 ng/ml for rIFNy or 6-7000 ng/ml for rTNF + rIFNy In the second series of experiments cell lines M38K and M9K were treated with rTNF, rIFNy, rIFNar, rTNF + rIFNr or rTNF + rIFNa all in the range 10-100 ng/ml for M38K, 0.1-100 ng/mI for M9K. The cytokines, human rTNFcr (spec. act. 6 x lo7 units/mg), human rIFNr (spec. act. 2 x lo7 units/mg) and human rIFNa-2C (spec. act. 3;2 x 10’) were all obtained from Boehringer Ingelheim. The results of the treatments are presented as the percentage of the number of control cells that survived (% survival) against the amounts of rTNF, rIFNr and rIFNar used. Each point on the figures represents the mean of 2 experiments of duplicate cultures (3 experiments in Fig. 2) f standard deviation. The multiplicative model was used to evaluate the synergistic/additive interaction of rTNF with rIFNrv and rIFNr [16,17]. Results The cell lines generally showed similar responses to treatment with the cytokines. The number of viable cells in the confluent control cultures after 5 days ranged from 5.8 x lo5 to 9.0 x 105. All three cell lines were insensitive to rTNF (< 30% inhibition of growth). Two cell lines (M14K and M9K) were mildly sensitive to r1FN-y alone (30-50% inhibition of growth) but the third cell line (M38K) was not. All the cell lines were sensitive to rTNF in combination with r1FN-y (> 50% inhibition of growth) (Fig. 1). The two cell lines established from primary tumours showed a synergistic response to the cytokine combinations, as assessed by the multiplicative model [16] (Table I). The metastatic cell line (M9K) showed only an additive response, however. The cytokine combination response was dose dependent with 10 ng/ml of each cytokine being the effective minimum dose. At lower doses the cytokines, both alone and in combination, had no inhibitory effect and at very low doses (0.1 proliferative moderate effects, rig/ml)
207
a primary tumour and M9K, from a metastasis) to alpha-interferon and gamma-interferon, both alone and in combination with tumour necrosis factor. Both cell lines were mildly sensitive to rIFNa, alone (30-50’S inhibition) but the combination of rIFNa and rTNF was more effective (60% inhibition) (Fig. 2). The response of the metastatic cell line M9K to the
120
M36K
loo-
80.
m _f gl
60-
E 40.
20.
10'
Concentration
102
(nglml)
40.
Fis. 1. The effect of tumour necrosis factor (TNF) and gamma-interferon (IFNy), both alone and in combination, on 3 mesothelioma cell lines. ( l - l) TNF alone, ( x - x ) 1FN-y alone, ( o - o ) TNF + IFNy in the ratio of 111 (w/w).
ZO-30% increase in growth over the controls, were observed (data not shown). The second series of experiments was designed to compare and contrast the responses of two of the cell lines (M38K. from
ZO-
I.
Id
IO0 Concentration
10'
102
(nglml)
Fig. 2. The effect of alpha-interferon (IFNa) and gamma-interferon (IFNy), both alone and in combination with tumour necrosis factor (TNF), on 2 mesothelioma cell lines. ( n - m) IFNar alone, (X - x ) IFNy alone, ( l - l ) TNF alone, ( o - q) IFNa + TNF (in the ratio of l:l), (o - O) IFNy + TNF (in the ratio of 1:l).
208
Table
1.
recombinant Cell line
The interaction of recombinant tumour necrosis factor (rTNF), recombinant gamma-interferon (r1FN-r) and alpha-interferon (rIFNcr) on 3 mesothelioma cell lines presented as the survival fraction (SF) of the cells. Dose (ng/mf)
Observed SF,,
survival SF,,,,
Predicted SF,iW +
IFNyl
Additive SFTNF
8.3 83 830 8.3 83 830 5 10 50
M14K
M38K
M9K
5 10 50 5 10 50
M38K
M9K
1.06 0.96 0.81 0.73 0.69 0.70 0.88 0.85 0.82
0.71 0.53 0.47 0.82 0.89 0.88 0.54 0.50 0.48
0.29 0.18 0.14 0.55 0.34 0.29 0.45 0.39 0.31
SFTNF
SFIFN,
SF (TNF +
0.91 0.90 0.84 0.88 0.85 0.87
0.73 0.60 0.58 0.53 0.50 0.45
0.50 0.37 0.31 0.42 0.41 0.30
x
survival model Sh,
Synergistic Synergistic Synergistic Additive Synergistic Synergistic Additive Additive Additive
0.75 0.51 0.38 0.66 0.61 0.62 0.48 0.43 0.39 IFNa)
SF,,,
0.66 0.54 0.49 0.47 0.43 0.37
Type of interactiona
x
SF,,,,
Synergistic Synergistic Synergistic Additive Additive Additive
survival fraction with TNF SF,,, survival fraction with TNF alone; SFrFN., survival fraction with fFNy alone; SF(r,, + rFNYI and IFNy; SF,,,,, survival fraction with INFcy alone; SFtrNF + rFNb,survival fraction with TNF + IFNa. ‘Synergism exists if SFcrNF + rrNi is fess than SFrNr x SFfrN. The interaction is additive if SFcrxr + tFNris equal to SFrNr x
=,FN
WI
rTNF/rIFNcz combination was only an additive interaction similar to this cell line’s response to the rTNF/rIFNy combination. There was slight synergism in the response of M38K to the rTNF/rIFNcu combination but not such a marked response as with rTNF/rIFNy combination (Table I). To summarise our results, mesothelioma cells were inhibited by combinations of tumour necrosis factor with either alpha- or gammainterferon, although they were not sensitive to tumour necrosis factor alone and showed variable sensitivity to alpha- and gammainterferon alone. Recombinant human tumour necrosis factor plus recombinant human gamma-interferon had a synergistic inhibitory effect on mesothelioma cells. The combinations were potent inhibitors in our system at doses as low as 10 ng/ml of each cytokine.
Discussion Tumour necrosis factor regulates the growth of many normal cells as well as having a cytotoxic effect on many malignant cells [4,5]. It is thought to exert its cytostatic effect on tumour cells by stopping the cell cycle at the G2 phase [US]. Interferons exert their antitumour effects in a variety of different ways, both directly on the tumour cells and indirectly, by influencing the host’s response to the tumour [ 191. The synergism between tumour necrosis factor and interferon has been reported for various tumours both in in vitro and experimental in vivo studies [7,20,21]. Our results show that human malignant mesothelioma cell lines are also sensitive to the effects of rTNF combined with either rIFNa! or rIFNy, despite variable sensitivity to the dif-
209
ferent interferons alone. In parallel to the work of Lewis et al. [21] we have also observed that low-dose combinations can be inhibitory when the same dose of the agents alone produces a stimulatory effect. The mechanism of action of this synergy is unclear. IFNy may increase the cells’ sensitivity to TNF although this cannot be solely a function of increased number of TNF receptors (221 because other workers have shown that TNF/IFNy synergism can be independent of receptor availability [23,24]. Kavoussi et al. [25] have demonstrated that IFNy pretreatment of renal cell carcinoma cell lines improves the cells’ response to TNF but that the timing of the pretreatment is crucial and varies between the different cell lines. We have added the interferons at the same time as tumour necrosis factor, but by culturing with the cytokines for 72 h, approximately 2 cell doublings, we have taken into account possible variations between the cell lines in the time course of the response [16]. Tumour necrosis factor, alpha-interferon and gamma-interferon have been used separately and together in various clinical regimes [9,10,26]. The success of these regimes has however been variable and the use of these cytokines often associated with moderate to severe toxicity. In particular TNF combined with IFNy can be especially difficult to administer clinically, producing more severe side effects than the cytokines separately. However, there is evidence that the side effects can be controlled and that they are more severe in patients with haematological malignancies than in patients with solid tumours [12,13]. Los et al. [27] reported successful administration of chemotherapeutic agents directly into the peritoneal cavity in an ovarian cancer model. This method increases the drug concentration at the site of the tumour and reduces the systemic side effects of intravenous administration. There is evidence that tumour cell lines which are, or have become, resistant to chemotherapeutic drugs can be susceptible to treatment with combinations of TNF and IFNy
to overcome their resistance [17,28]. This points to a role for combined TNFAFN treatment in cases of chemotherapy failure such be experienced in malignant as may mesothelioma. Furthermore there is some clinical evidence that interferons should be used in combination with cytostatic drugs [26,29]. None of the chemotherapeutic agents currently in use for mesothelioma treatment is particularly effective so there is a need to test other agents for mesothelioma activity [30]. We aim to extend our in vitro studies to investigate the effective cytokine combinations reported here with the cytostatic drugs currently in use for mesothelioma treatment, to see if an improved response were possible. In conclusion, these experiments indicate that combinations of low doses of recombinant tumour necrosis factor and recombinant interferons effectively inhibit the growth of mesothelioma cells in vitro. Such combinations have been shown to be effective against other cancers but we believe this to be the first time that they have been shown to be effective against mesothelioma cells. The low doses required may offset the predicted clinical toxicities. The that this cytokine proposal combination could be used effectively to treat mesothelioma is attractive, given the current poor prognosis for mesothelioma patients and the resistance of mesothelioma to conventional chemotherapy, as well as multimodality schemes. Acknowledgements We would like to thank Mr. Hans Sarelin of Boehringer Ingelheim for providing the recombinant cytokines, and Ms. Madeleine Mattson, Ms. Satu Suhonen and Ms. Marjatta Vallas for their technical assistance. References Wagner. J.C.. Sleggs, C.A. and Marchand. P. (1960) Diffuse pleural mesotheliomas and asbestos exposure in the Northwestern Cape Province. Br. J. Ind. Med., 17, 260-271. Antman. K. and Aisner, J. (1987) Asbestos-Related Malignancy. Grune and Stratton, Orlando.
210
3
Huuskonen, Asbesti
M.S.,
Koskinen,
merkittava
Suomessa?
H.,
Roto, P. et al. (1987)
terveyshaitta
Suomen
Lalkarilehti.
vie12 17,
1669-
1676
B.J., Aggarwal,
Recombinant
5
with
of normal
230,
Semenzato,
Hass, P.E. et al. (1985)
human tumour necrosis factor-a:
proliferation Science,
B.B.,
and
transformed
multiple
Tumour
necrosis factor: a cytokine
activities.
Br.
L.,
Van
der Heyden,
J.
Cancer,
Recombinant
J., Ruysschaert,
18
R. and
on a variety of nor-
19 20
22, 419-426. Synergistic antitumor effects of tumor necrosis fac-
21
Cell
Gresser, I. (1989)
In vitro
human lung cancer
tumour
necrosis factor by virtue of
B.,
Carswell,
cycle-specific
1280.
E.A.
and
effects of tumour
Balkwill,
Antitumour
Balkwill,
F.R.,
Lee, A.,
F.R.,
Sherman,
M.L.,
Lewis, G.D., tumour
Aldan,
L. et al. (1986)
Ward,
B.G.,
tumor
22
Arthur, K.A.
et al. (1988)
23
6,
patients. J. Clin. Oncol., Kettlehut,
I.C.,
A.,
in cancer 25
A.L.
(1987)
Ruggiero,
84, 4273-4277.
Fiedler,
GBtz,
G.,
and Hossfeld.
J. Clin. Oncol.,
Jakubowski,
Larchian,
plus interferon-gamma
27
W.,
Starnes,
(IV) Tumour
Necrosis Factor
in patients with advanced
(Meeting Abstract) Proc. Annu.
C.
136, 2445.
M. and Pfizenmaier.
and characterization
K.
of high-affinity
receptors for tumor necrosis factor on human 38,
Kavoussi.
Hudson,
L.R.,
Effect
Ruesing, of tumor
on human
R.A.,
127-133.
necrosis factor
renal carcinoma
M.A. and
et al.
interferon
cell line growth.
J.
142, 875--878.
Strander,
H.A.
(1989)
Clinical effects of interferon therapy
Los, G.,
Ruevekamp,
28
M.,
Bosnie,
growth
N. et al. (1989)
and chemotherapy
Eur. J. Cancer Clin. Oncol..
Price,
G.,
Cytotoxic
cancer.
tumor Brenner,
K.,
Prentice,
effects of tumour
26,
1857-1866.
H.G.
et al.
(1987)
necrosis factor and gamma-
interferon on acute myeloid leukaemia
Meet. Am. Sot. Clin. On-
In-
in a rat
blasts. Br. J. Can-
cer, 55, 287-290.
Williamson, B.D., and Old, L. (1983)
Carswell, E., Rubin, B., Prendergast,
J.
29
human interferon. 5397-5401.
Proc.
Natl.
Acad.
K.. Husgafvel-Pursiainen,
Sci.
U.S.A.,
Ann. 30
L. et al. (1990) Asbestos-related malignant mesothelioma: Growth, tumorigenicity and CYtologY, consistent chromosome findings in cell lines from five patients. Car11, 673-681.
non-small
80.
K., Tammilehto,
Bowman,
A., Fergusson.
Potentiation
Human tumor necrosis factor produced
by human B-cell lines: Synergistic cytotoxic interaction with
cinogenesis,
J. Immunol.,
U., Kronke.
leukemic cell lines. Int. J. Cancer,
model.
F. et al. (1988)
col., 7. A634.
Pelin-Enlund,
P.. &er,
traperitoneal A.,
2441-2450.
J., Fiers, W. and Baglioni.
28, 355-362. 6,
1787-8. Phase-I trial of intravenous
136,
for tumour
with special emphasis on antitumour efficacy. Acta Oncol.,
D.K.
Increased toxicity of tumor necrosis factor/Gamma
interferon in patients with leukocytosis.
Res., 47. Interferon-
of cellular receptors
Tavernies.
Quantification
Ural.. 26
H.J.
V.,
Scheurich,
gamma
inhibitors. Proc. Natl. Acad.
Weh,
cells by human
Cancer
Induction of the synthesis of tumor necrosis factor
(1989)
The
necrosis factor in uiuo and their
Sci. U.S.A.,
Eessalu, T.E. et al. (1987)
and Vilcek, J. (1986)
expression
membrane
R. et al. (1986) gamma
4, 137-146.
prevention by cyclooxygenase W.,
Bhalla,
interferon
Fiers, W. and Goldberg,
toxic effects of tumor
B.B.,
Tsujimoto. M., Yip, Y.K.
(1986) Al-Katib,
Aggarwas.
receptors by interferon-y. 24
344-350. S.,
human ovarian cancer. Cancer
necrosis factor and interferon.
(1986)
as a 5-day continuous infusion in cancer patients: Phase 1
Phase 1 trial of recombinant
E. and Fiers. W.
of the growth of transformed
y enhances
with interferons.
human tumour necrosis factor administered
Vadhan-Raj,
Moodie,
potential of tumour necrosis factor and
necrosis factor. J. Immunol.,
Spriggs, D.R.,
Acta
5382-5385.
Human
human
effects of interferon.
28, 347-353.
Modulation
cell lines.
toxicity and effects on lipid metabolism. J. Clin. Oncol.,
15
8,
Res., 47, 4755-4758.
Bittner, G., Storer, B. and Willson, J.K.V.
Recombinant
14
human
Williamson,
(1984)
Oncologica.
Cancer Res.. 46, 3990-3993.
13
L.J.
Z..
interferon in experimental
(1988)
Y. et al. (1987)
human interferon-gamma
Oncol.,
Schiller, J.H.,
necrosis factor alone and in combination
12
recombinant
(1987) Therapeutic
tumor xerografts treated with recombinant
11
Saijo, N., Sasaki,
Darzynkiewicz, Old,
61,
Cancer Res.. 47, 2809-2813.
10
W.S.,
mal and transformed human cell lines. Eur. J. Cancer Clin.
tor and -y-interferon on human colon carcinoma
9
Hong,
and/or
tumor necrosis factor: its ef-
fect and its synergism with interferon-y
8
Ther.,
necrosis factor. Cancer Res., 44, 83-90.
Fransen,
(1987)
Pharmacol.
collateral sensitivity. Jpn J. Cancer Res., 78, 1274-
biological
Fiers, W. (1986)
7
In vitro systems for evaluation of
cell lines by recombinant
effects on
cells in vitro.
354-361. 6
(1980)
chemotherapy.
growth inhibition of cisplatin-resistant
943-945.
G. (1990)
R.L.
21-35.
(in 17
Sugarman,
Momparler, combination
Finnish). 4
16
2000-luvun
Allan, S.G.
et al. (1990)
of cisplatin by alpha-interferon
R.J.,
in advanced
cell lung cancer
Oncol..
(NSCLC):
1. 351-353.
Aisner, J. and Sigman, L.M. therapy
in Treatment
Asbestos-Related Antman
A phase II study.
(1987)
of Malignant
Malignancy.
The Role of ChemoMesothelioma.
pp. 385-401.
and J. Aisner. Grune and Stratton,
In:
Editors: K. Orlando.