Radiol Oncol 1993; 27: 286-92.
Antitumor effect of interferon-a administered by different routes
of treatment
Srdjan Novakovic1 and W. Robert Fleischmann Jr.2
'Institute of Oncology, Ljubljana, Slovenia, 2University of Texas Medical Branch, Department of Microbiology, Galveston, Texas, USA
Besides the fact that interferons were identified as factors capable of inhibiting viral infections, they have proved to be antiproliferative, immunomodulatory and differentiation-inducing factors. On the basis of these activities, they have been employed clinically far treatment of various tumors. The study was performed to determine whether there was different antitumor effect of recombinant human interferon-a A/D (rHulFN-a A/D) when it was given as a local or systemic therapeutical agent. Two different tumor models, i.e. subcutaneous (s.c.) and intraperitoneal (i.p.) B-16 melanoma on C57Bl/6 mice, were employed in these experiments. Experimental mice were treated locally or systemically with different doses of rHulFN-a AID; the treatment was begun 24 hours afier tumor cell inoculation and continued through five consecutive days. lntraperitoneal treatment of animals with i.p. tumors resulted in significantly longer survival time in comparison with control group or with subcutaneously treated animals (p<0.001). Similarly, the delay of tumor detection and tumor growth in mice with s.c. tumors treated subcutaneously with rHulFN-a AID was significantly greater than in intraperitoneally treated animals (p<0.01). According to these results we can conclude that rHulFN-a AID is much more patent antitumor agent when it is used locally. However, systemic treatment with higher doses was effective in both tumor models and it is still more convenient for treatment of some tumor lesions which are not accessible for local treatment.
Key words: melanoma, experimental-drug therapy; interferon-alpha; drug administration routes
Introduction
Interferons are glycoproteins which were identified as factors capable of inhibiting viral infections. 1 ' 2 Besides, interferons have proved to be
Correspondence to: Novakovic Srdjan, MSc., Institute of Oncology, Zaloska 2, 61105 Ljubljana, Slovenia, Tel. + 386 61 323 063 ext. 29 33, Fax + 386 61 131 41 80.
UDC: 616-006.81-085
antiproliferative, immunomodulatory and differentiation inducing factors.3' 4' 5 Other putative functions include antioncogene activity and mobilisation of energy stores during sickness. 6'7
Three subtypes of interferons (IFN a, p and y) have been identified, differing in terms of their cell surface receptors, their acid stability, their primary sequence and their chromosomal location and organisation.3' 8 Interferon-a and interferon-p produced by leukocytes and fibro-blasts, respectively, are acid stable and share the same receptor, while interferon-y is produ-
Antitumor effect of interferon-a administered by different routes of treatment
287
ced by T lymphocytes, is acid labile and has a different receptor.4' 8
The precise mechanisms of action for the antitumor effects of interferons are not fully explained. They involve both direct (antiproliferative effects, cytotoxic effects and enhancement of celi surface antigen expression on tumor cells) and indirect antitumor action (activation of macrophages/monocytes, activation of T cells, activation of NK cells and modulation of antibody production).9 10
More then 20 subtypes of interferon-a are known, but only few of them are used systemi-cally or locally in the treatment of neoplasms as hairy celi leukemia, AIDS - related Kaposi sarcoma, Hodgkin's disease, 11011 - Hodgkin's lymphomas, oral cancer, malignant melanoma, renal celi carcinoma and bladder cancer.11-16
In our experiments we investigated the relative capability of local versus systemic treatment with rHuIFN-a A/D as an antitumor agent against B-16 melanoma. To address this question we used two different tumor models: i.p. and s.c. B-16 melanoma tumors.
Materials and methods
Animals
Six to eight weeks old female C57Bl/6 mice were used in the experiments. Mice were purchased from Jackson Laboratories (Bar Harbor, USA) and held in a pathogen free animal colony. The adaptation period before use was two to three weeks. At least nine healthy animals with normal body weight were included in each experimental group.
Tumor models
Subcutaneous (s.c.) and intraperitoneal (i.p.) tumors were employed as tumor models. Subcutaneous tumors were induced subcutaneously in the left lower abdomen with 106 B-16 melanoma cells in 0.1 ml EMEM (Eagle's minimal essential medium) supplemented with 2 % fetal calf serum (FCS), while mice for i.p. tumors were
inoculated with the same number of viable cells intraperitoneal^. In the experiments with s.c. tumors the day of tumor detection was monitored and tumor growth was followed by measuring two tumor diameters with a vernier caliper. The tumor burden was calculated by the standard formula for a prolate sphere V = n/ 6xd1xd} (d2<d.j). Mice with i.p. tumors were monitored for the day of death and the increased life span (ILS) was calculated. Also, mice with s.c. tumors were monitored for the day of tumor development and the increased tumor detection span (ITDS) was determined as shown below.
ILS
ITDS
av. day of death for IFN treated micc -av. day of death for control average day of death for control
av. day of tu. det. for IFN treated mice - av. day of tu. det. for control average day of tumor detection for control
x 100
x 100
Tumor cells
Murine B-16 melanoma cells (done Fl)17 were grown in EMEM supplemented with 10 % FCS, penicillin (100 units/ml), streptomycin (100 ftg/ ml) and gentamycin (11 [i,g/ml). The final celi suspension for inoculation (106 viable tumor cells per 0.1 ml) was prepared with EMEM supplemented with 2 % FCS and antibiotics as indicated above.
Interferon
Recombinant human interferon-a A/D (rHulFN-a A/D) used in this study was generously provided by Dr. Michael Brunda of Hoffman-La Roche (Nutley, New Jersey) and had a specific activity of 6.4x107 U/mg of protein. Recombinant human IFN-a A/D is a recombinant molecular hybrid of two subtypes of HuIFN-a which exerts antiviral, antitumor and myelotoxic activities in mice.18'19 The interferon was diluted in phosphate buffered saline (PBS) supplemented with 0.3 % bovine serum albumin (BSA, Sigma Chemical Company) and frozen at -70°C until used for treatment.
288
Novakovic S. and Fleischmann R. Jr.
Treatment
Interferon treatment was started 24 hours after tumor cell inoculation and continued daily through five consecutive days. Animals with i.p. tumors were treated with different IFN doses (3000 U/0.2ml, L0000 U/0.2ml or 30000 U/0.2ml per animal per day) intraperitoneally (locally) or subcutaneously (systemically). Subcutaneous tumor bearing animals were treated with the same doses as mentioned above, but in this case subcutaneous treatment was performed as Jocal and intraperitoneal as systemic treatment. The animals in the control group were injected subcutaneously or intraperitoneally with 0.2 ml of PBS supplemented with 0.3% BSA.
Statistical analysis
The data were evaluated for significance using Student's T-test.
Results
Experiments were performed to determine whether the antitumor effect of rHuIFN-a A/D was different when the agent was administered locally or systemically. Two different tumor models were employed: s.c. B-16 melanoma and i.p. B-16 melanoma.
Antitumor effect on s.c. tumors
Mice were inoculated s.c. with B-16 melanoma tumor cells and randomly divided in eight groups:
control group treated subcutaneously with PBS/BSA;
-	control group treated intraperitoneally with PBS/BSA;
-	group treated subcutaneously with 3000 U (3 KU) of rHuIFN-a A/D;
group treated intraperitoneally with 3000 U (3 KU) of rHuIFN-a A/D;
-	group treated subcutaneously with 10000 U (10 KU) of rHuIFN-a A/D;
-	group treated intraperitoneally with 10000 U (10 KU) of rHulFN-a A/D;
-	group treated subcutaneously with 30000 U (30 KU) of rHuIFN-a A/D and
-	group treated intraperitoneally with 30000 U (30 KU) of rHuIFN-a A/D.
Table l. Average day of tumor dctcction for s.c. B-16 melanoma bearing mice treated subcutaneously or intrapcritoncaly with rHuIFN-a A/D.
	Average	SD*	p-value	p-value
	day of		(comparing (comparing	
	tumor		to the	the same
	detection		control)	closes)
i.p. control	9.9	2.1		
i.p. 3 KU	12.7	2.9	0.0039	
i.p. 10KU	12.2	2.5	0.0087	
i.p. 30 KU	14.7	4.8	0.001	
s.c. control	10.4	2.2		
s.c. 3 KU	16.9	5.2	0.0001	0.005
s.c. 10 KU	19.9	3.9	0.0001	0.0001
s.c. 30 KU	20.2	3.8	0.0001	0.001
*SD - Standard deviation
IFN-t, Concentrations (U/injection)
Figure l. Increase in tumor detection span (ITDS) for subcutaneous B-16 melanoma tumors. Tumors were implanted using 106 viable tumor cells, and 24 hours later treated subcutaneously or intraperitoneally with different coneentrations of rHuIFN-a A/D during five consecutive days.
Antitumor effect of interferon-a administered by different routes of treatment
289
Starting 24 hours after tumor cell inoculation and continuing for five days, mice were injected s.c. or i.p. with either rHuIFN-a A/D (different concentrations) or PBS/BSA.
The day of tumor detection and tumor growth were monitored. Table 1 presents the average results of two identical experiments that gave similar results. Both subcutaneous (local) and intraperitoneal (systemic) treatments caused a significant antitumor effect at all interferon concentrations employed. However, it can be seen that local treatment was more effective than systemic treatment.
Locally treated mice with 3 KU of rHuIFN-a A/D developed tumors in 16.9 days on average, with 10 KU in 19.9 days, and with 30 KU in 20.2 days; these periods being 62.5 % , 91.3 % and 94.2 % longer than those in the control group (Table l, Figure 1).
Systemically treated mice with 3 KU developed tumors in 12.7 days in average, with 10 KU in 12.2 days, and with 30 KU in 14.7 days; those periods being 28.3 % , 23.2 % and 48.5 % longer than those in the control group (Table 1, Figure 1). Tumor growth kinetics was the same as in the control group, while local treatment slowed down the tumor growth in all treated groups (Figure 2).
Statistically significant differences in the day of tumor detection were observed between the two routes of interferon administration for all treatment doses (p =0.005 with 3 KU, p = 0.0001 for 10 KU, and p = 0.001 for 30 KU of rHuIFN-a A/D).
Antitumor effects on i.p. tumors
It was important to consider previous data from s.c. tumor model in order to asses whether the differential responsiveness of the tumors would be observed on i.p. tumor model after different routes of treatment with rHuIFN-a A/D. To address this point, mice were inoculated i.p. with B-16 melanoma tumor cells and randomly distributed (as mice with s.c. tumors), into eight groups. Mice were also treated locally (intraperitoneally) and systemically (subcuta-
Days after 1 ymor Cell inoculation
Figure 2. Growth kinetics of subcutaneous B-16 melanoma implanted in the left lower abdomen using 106 viable tumor cells, and treated subcutaneously or intraperitoneal^ with rHuIFN-a A/D.
neously) with different doses of rHulFN-a A/D or PBS/BSA. Treatment schedule was the same as the one described above for s.c. tumors. Mice were monitored for the day of death.
Table 2 presents the average results of two identical experiments which gave similar results. It can be seen that also in i.p. tumors local treatment was more effective than systemic
Table 2. Average day of death for i.p. B-16 melanoma bearing mice treated subcutaneously or intraperito-nealy with rHuIFN-a A/D.
		Average	SD*	p-valuc	p-valuc
		clay of		(comparing (comparing	
		death		to the	the same
				control)	closes)
s.c.	control	19.9	1.8	-----	
S.C.	3 KU	19.6	1.7	0.5688	
S.C.	10 KU	20.9	l.7	0.0714	
S.C.	30 KU	21.5	1.8	0.0098	
i.p.	control	19.3	1.3	—	
¡.p.	3KU	22.6	2.5	0.0001	0.0001
i.p.	10 KU	23.3	1.9	0.000]	0.0004
i.p.	30KU	24.8	3.1	0.0001	0.0002
*SD - Standard deviation
290
Novakovic S. and Fleischmann R. Jr.
treatment (Table 2, Figure 3). Percent of increase in life span ( %ILS±SE) for locally treated .mice with rHuIFN-a A/D was 17.3±2.99 (3 KU), 20.5±2.27 (10 KU) and 28.7±3.68 (30 KU); for systemically treated mice the %ILS was -1.7±1.96 (3 KU), 5.2+1.98 (10 KU) and
100 -
75 -
IP 3KU IFN- a
IP 10KU IFN- <
50
25
20	25	30	35
Days after tumor cell inoculation
100
75 -
50
25 -
	Control
........O-.....	SC 3KU IFN-a
----o----	SC 1 0KU IFN-a
----6----	SC .30KU IFN- a
20
25
—r-
30
—I
35
Days after tumor cell inoculation
Figure 3. Survival curves for intraperitoneal B-16 melanoma bearing mice (C57B1/6); tumors were induced intraperitoneally with 106 viable tumor cells and 24 hours later treated intraperitoneally (upper figure) or subcutaneously (lower figure) with different concentrations of rHuIFN-a A/D during five consecutive days.
10000
IFN-u concentration (U/Day)
Figure 4. Increasc in life span (ILS) for intraperitoneal B-16 melanoma bearing micc treated systemically or locally with rHuIFN-a A/D.
7.9±2.08 (30 KU) (Figure 4). Owing to a higher agressivness of i.p. tumors, systemic (subcutaneous) treatment did not statistically significantly affect the average day of death in comparison with control mice, except when the mice were treated with the highest dose (30 KU) (Table 2, Figure 3).
However, statistically significant increase in life span was observed when we compared locally (intraperitoneally) treated animals to the ones treated systemically (subcutaneously) with the same dose of rHuIFN-a A/D; the p-value for mice treated with 3 KU was 0.0001, for mice treated with 10 KU 0.0004 and for mice treated with 30 KU 0.0002. Moreover, the animals that received a threefold lower dose (10 KU) of rHuIFN-a A/D locally (intraperitoneally) survived significantly longer than those treated systemically (subcutaneously) with 30 KU (p = 0.005).
Based on the results obtained in both tumor models, it is clear that maximal antitumor activity occurred when rHuIFN-a A/D was given locally. Systemic treatment was moderately effective: more effective in s.c. tumors when rHuIFN-a A/D was administered intraperito-neally than on i.p. tumors when it was administered subcutaneously.
.10
10
000
00000
o
Antitumor effect of interferon-a administered by different routes of treatment
291
Discussion
Previous experimental findings demonstrated that IFN-a has reproducible antiproliferative effects in vitro20-22 and in vivo.23 On the basis of these findings, IFN-a has been employed clinically for treatment of various tumors.
Today, IFN-a is approved as an antiproliferative agent for the treatment of hairy cell leukemia and AIDS related Kaposi sarcoma.510 Ne-vertheles, IFN-a as a single agent has been reported to induce clinical remission in many hematological malignancies and solid tu-mors.16'24'25 Moreover, IFN-a has reproducible activity against malignant melanoma, a tumor for which conventional chemotherapy has poor efficacy. 10'2-5
The present study was undertaken to assess which route of administration is more suitable ' for IFN-a treatment. Therefore, we chose s.c. and i.p. B-16 melanoma tumor models. Mice were treated locally and systemically for five consecutive days with different doses of rHuIFN-a ^D. Systemic treatment of s.c. tumors was performed in the form of intraperito-neal injection, while locally treated animals were injected subcutaneously. In contrast, in the i.p. tumor model intraperitoneal administration was performed as a local and subcutaneous as a systemic treatment. In both cases local treatment proved to be significantly superior to systemic. An interesting observation was that systemic (intraperitoneal) treatment of s.c. tumors resulted in a statistically significant delay in tumor detection at all interferon concentrations examined, while systemic treatment (subcutaneous) of i.p. tumors did not significantly increase the life span of treated animals (except 30 KU). The fact that developed tumors in systemically treated animals continued growing at the same rate as tumors in control mice, suggests that systemically administered rHuIFN-a A/D exerts antitumor effect only on a very small tumor burden.
In accordance with our observations, rHulFN-a ^D is more effective when given as a local therapeutical agent. Nevertheless, when we have to use rHuIFN-a ^D systemically, it
is much more advisable to administer it intrape-ritoneally than subcutaneously. This is also in agreement with previous pharmacokinetic findings that intraperitoneal administration of IFN-a has good bioavailability (30 times higher) compared to the intravenous route.26
The role of IFN-a in the treatment of malignancies has not yet been fully established. Many questions remain to be answered concerning the optimal strategy for incorporating IFN-a into anticancer therapy, and one of them is the optimal route of its administration. However, the future of IFN-a usage in oncology seems to be in its local (and also systemical) use as adjuvant therapy after the tumor burden has been reduced by other therapeutic modalities.
Acknowledgement
This work was supported by U.S. Public Health Service Grant (National Cancer Institute).
References
1.	Isaacs A, Lindenmann J. Virus interference. I. The interferon. Proc R Soc Lond Ser B 19.57; 147: 258-67.
2.	Lindenmann J. Induction of chick interferon: procedures of the original experiments. Methods En-zymo/ 1981; 78: 181-8.
3.	Friedman MR, Vogel NS. Interferons with special emphasis on immune system. Adv Immuno/ 1983; 34: 97-[40.
4.	Baron S et al. The Interferons. Mechanisms of action and clinical applications. JAMA 1991; 266: 1375-83.
5.	Elsdssr-Beile U, von Kleist S. Cytokines as therapeutic and diagnostic agents. Tumor Biol 1993; 14: 69-94.
6.	Kurzrock R, Talpaz M, Gutterman JU. Interferons a, |3, y: basic principles and preclinical studies. In: DeVita VT Jr., Hellman S, Rosenberg A eds. Biological therapy of cancer. New York: J.B. Lippincott Co. 1991: 247-75.
7.	Patton JS et al. Interferons and tumor necrosis factors have similar catabolic effects on 3T3L1 cells. Proc Nat/ Acad Sci USA 1986; 83: 8313-8.
8.	Itri ML. The interferons. Cancer 1992; 70 (Suppl 4): 940-5.
292
Novakovic S. and Fleischmann R. Jr.
9. Flcischmann WR Jr., Flcischmann CM. Mechanisms of interferons antitumor action. In: Baron S et al cds. Interferon: Principles and medical applications. Galveston: UTMB 1992: 299-309.
10.	Dorr RT. Interferon-a in malignant and viral diseases. Drugs 1993 ; 45(2): 177-21 l.
11.	Rao SV, Wadler S. Curent use of interferons. Contemp Oncol 1992; 3: 44-9.
12.	Kuo JY et al. Impaired interferon-a production in whole blood cultures from bladder cancer patients. Urol Res 1991; 19: .51-6.
13.	Rassiga-Pidot AL, Mclntyre OR. In vitro leucocyte interferon production in patients with Hod-gkin's disease. Cancer Res 1974; 34: 2995-3002.
14.	Ho AD, Moritz T, Rensch K, Hunstein W, Kirchner H. Deficiency in interferon production of peripheral blood leucocytes from patients with non-Hodgkin lymphoma. J Interferon Res 1988; 8: 405-13.
15.	Jamkar AV, Banerjae AC, Gore MM, Sathe PS, Ghosh SN. Interferon producing capacity of peripheral mononuclear cells in oral cancer patients. Indian .! Cancer 1989; 26: 76-84.
16.	Wadler S. The role of interferons in the treatment of solid tumors. Cancer 1992; 70 (Suppl 4): 949-58.
17.	Fidler IJ. Selection of successive tumor lines for metastasis. Nature New Biol 1973; 242: 148-9.
18.	Kramer MJ et al. Cell and virus sensitivity studies with recombinant human alpha interferons. J Interferon Res 1983; 3: 425-35.
19.	Rosenthal GJ et al. Organ-specific hematopoetic changes induced by a recombinant human interferon alpha in mice. Fundam Appl Toxicol 1990; 14: 666-7.5.
20.	Paucker K, Cantell K, Henle W. Quantitative studies of viral interference in suspended L cells, III. Effect of interfering viruses and interferon on the growth rate of cells. Virology 1962; 17: 324-8.
21.	Pfeffer LM, Murphy JS, Tamm l. Interferon effects on the growth and division of human fibro-blasts. Exp Cell Res 1979; 121: 111-5.
22.	Jezersek B, Novakovic S, Sersa G, Auersperg M, Fleischmann WR Jr. Interactions of interferon and vinblastine on experimental tumor model melanoma B-16 in vitro. In print (Anti-Cancer Drugs).
23.	Jezersek B, Novakovic S, Sersa G, Cemazar M, Auersperg M, Fleischmann WR Jr. Interactions of interferon and vinblastine on experimental tumor model melanoma B-16 in vivo. Radiol Oncol 1993; 27: 275-9.
24.	Wandl UB, Niederle N, Kranzhoff M, Seeber S. Clonogenic assay is not predictive but reflects therapeutic efficacy of interferons in the treatment of chronic myelogenous leukemia. Int J Cell Cloning 1992; 10: 292-8.
25.	Kirkwood MJ. Studies of interferons in the therapy of melanoma. Semin Oncol 1991; 18 (Suppl 7): 83-90.
26.	Schuller J et al. Pharmacokinetic aspects of interferon alfa-2b after intrahepatic or intraperitoneal administration. Semin Oncol 1992; 19 (Suppl 3): 98-104.