WO1987006830A1

WO1987006830A1 - Lipopolysaccharide and natural factor compositions for anti-tumor therapy and method of treatment

Info

Publication number: WO1987006830A1
Application number: PCT/US1987/001050
Authority: WO
Inventors: Michael K. Hoffmann; Myung Chun; Ulrich Hammerling
Original assignee: Sloan-Kettering Institute For Cancer Research
Priority date: 1986-05-09
Filing date: 1987-05-07
Publication date: 1987-11-19
Also published as: JPS63503308A; AU7397887A; EP0271521A4; EP0271521A1

Abstract

Bacterial lipopolysaccharide (LPS) can be used to induce release of factors into the serum enabling rejection of tumors and increasing the efficacy of added anti-tumor agents. Thus exogenous TNF, when used together with LPS or other factors induced by LPS can be used at lower non-toxic doses to cause tumor rejection. These factors can be combined in anti-tumor therapy with other chemotherapeutic agents such as indomethacin or cyclophosphamide or natural agents such as platelet factor 4 which block negative feedback responses suppressing the immune system, thereby mediating the body resistance to the therapy.

Description

LIPOPOLYSACCHARIDE AND NATURAL FACTOR COMPOSITIONS FOR ANTI-TUMOR THERAPY AND METHOD OF TREATMENT

This application concerns an invention developed with U.S. government funds under grant No. CA 11673. Therefore the United States Government has certain rights in this invention.

SUMMARY Lipopolysaccharide (LPS) or variants thereof can be used together with such anti-tumor agents as exogenous TNF as effective anti-tumor agents at non-toxic levels.

Inhibitors such as indomethacin which block agents, such as prostaglandin which act against immune system activation or agents which act as against suppressor cell activation such as cyclophosphamide can be used as well.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 shows synergism in causing necrosis and regression of tumors with LPS and TNF, also with TNF and LPS with INDO (Indomethacin).

Fig. 2 shows correlation between tumor necrosis and subsequent regression of tumors.

DESCRIPTION DESCRIPTION OF THE DRAWINGS Fig. 1: Synergism between TNF and LPS is causing necrosis and regression of tumor transplants in mice. Animals were treated in addition to TNF and LPS with INDO. Fig. 2: Correlation between tumor necrosis and subsequent regression of tumors, the data obtained with LPS (0) relates the degrees of necrosis in individual mice to the regression of the same tumors. In the experiments with TNF we had not assessed individual mice for necrosis and regression but only groups of identically treated animals. The percent of tumor regression in these animals was related to the group's average degree of necrosis.

Lipopolysaccharide (LPS) prepared from the outer membranes of gram negative bacteria has been shown to have strong anti-tumor effects in mice (I. Parr et al,

Brit.J.Cancer, 27:370-389, 1973; M.J. Shear, J.Nat. Cancer

Inst., 4: 461-476, 1944). It causes acute necrosis of some

(but not all) experimental tumor transplants in mice and, in many cases, complete regression of the tumor. The latter phase, tumor rejection, is in contrast to the former, tumor necrosis, an immunological phenomenon (M.J. Berendt et al,

J.Exp.Med., 148:1550-1559, 1978). LPS has long been known to elicit anti-tumor reaction in experimental animals and man. LPS causes acute partial necrosis of certain murine and human tumor transplants in mice and allows one third of the treated animals to completely reject the tumor. The response can be divided into two phases, an early non-immunologic necrosis phase and a subsequent immunologic rejection phase.

The application of LPS as a therapeuticum has been hampered by its prohibitive toxicity. It was learned that LPS causes the release of several factors into the serum of treated animals and such serum, called tumor necrosis serum (TNS), caused tumor necrosis and frequent rejection of tumors without apparent toxicity to the tumor host (L.J. Old et al. Current enigmas in cancer research, Harvey Lecture Series 67:273-315, Academic Press, New York and London, 1973; E.A. Carswell et al, Proc.Natl.Acad.Sci.USA, 72:3666-3670, 1975). The molecule in TNS responsible for tumor necrosis, tumor necrosis factor (TNF), is now produced in recombinant bacteria (D. Pennica et al. Nature 312:724-729, 1984). TNF is a very efficient inducer of tumor necrosis but is also very toxic to animals. LPS-induced shock seems to be mediated by TNF (B. Beutler et al, Nature 316:552-554, 1985). For the cure of a tumor a mouse seems to have to pay a smaller penalty on toxicity when TNF is administered within TNS than when administered in purified form. TNF kills tumor cells in tissue culture whereas LPS has no effect on cultured tumor cells directly. Our findings suggest LPS acts on tumors through different pathways only one of which is mediated by TNF. In combination the different defense systems are more effective than each by itself. TNS contains other factors which affect the fate of tumors. The increased presence of IL-1 or IL-2 like factors and; of interferon has been identified and indirect evidence suggests the presence of other, as yet unidentified factors.

TNS has been found to stimulate the activity of NK cells, a cell that kills tumor cells on contact. It induces also the production of substances such as prostaglandins which paralyze the defense capacity of the tumor host. Interleukin 1 (IL-1), is a lymphokine released by LPS-reactive macrophages, facilitates the activation of lymphocytes which are the carriers of the immunological response against tumors. IL-1 facilitates the recruitment of lymphocytes, causes them to produce lymphokines which participate in the process of the immune response (such as IL-2 and TRF) and is therefore essential for the generation of effector cells which produce either antibodies to tumor associated surface membrane structures or which seek out tumor cells via these membrane antigens and kill them. Besides their upregulatory function lymphokines also initiate downregulatory mechanisms which block the immune response. This downregulatory pathway is mediated by cells called suppressor cells, of which suppressor T-cells are the most prominent.

Taken together, LPS can be viewed as inducing reactions that inhibit malignant growth (TNF production, enhancement of immune function) and reactions which favor malignant growth (immune suppression). It would appear, that effective immunotherapy must be designed to support the former reaction and to curb the latter. Our laboratory has identified negative feedback responses which favor tumor growth and we have demonstrated that these feedback mechanisms can be controlled leading to substantial improvement of tumor rejection. Another relevant objective of immunotherapy is the least possible damage caused to the treated tumor host, that is to say low toxicity to the host. Tumor necrosis factor is a highly toxic molecule and its therapeutic dose in mice and humans is limited by its lethal effects. This laboratory has established a treatment regimen in which TNF can be applied in low (not toxic) doses in conjunction with other reagents in non toxic doses. LPS itself is a good candidate to complement TNF function since LPS induces in the host the release of several lymphokines such as IL-1, IFN, IL-2 and those not identified as yet which may replace LPS. Non toxic doses of LPS combined with non toxic doses of TNF, both ineffective when given alone cause tumor elimination in approximately one third of the animals tested.

LPS presumably through mediators such as IFN causes increased production of prostaglandins (PG) in the tumor host. PG inhibits the function of the immune system (including NK cell activity). Inhibition of PG synthesis with Indomethacin increases the tumor rejection induced by TNF and LPS from 30 to 60 percent.

In parallel to stimulating immune effector cells LPS causes the generation of suppressor cells. Suppressor cell activity can be inhibited by administration of cyclophosphamide and this drug given together with TNF, LPS and INDO increases tumor rejection in mice up to 100%.

Bacterial lipopolysaccharide (LPS) induces the release of factors into the serum which enable mice to reject experimental tumors. One such factor is tumor necrosis factor (TNF) which causes acute necrosis of syngeneic sarcoma transplants in mice. Effective therapeutic use of TNF is limited, however, by its toxicity. We show here, that the efficacy of TNF can be substantially increased by combining its application with low doses of LPS. Our data suggests that LPS exerts its anti-tumor effects by engaging more than one defense mechanism.

Characteristic for the activation of a biological system is a concommitant induction of negative feedback mechanisms which antagonize the initial stimulus. Interference with the negative feedback response may substantially increase biological reactions. We show here that the blocking of two negative feedback responses occurring as a consequence of treatment with LPS namely the production of prostaglandin E (PGE) and the generation of suppressor T lymphocytes, increases dramatically the ability of mice to reject tumor transplants.

Thus, through appropriate combination of different factors one may reduce the dose of each below toxic levels and through interference with negative feedback responses increase the efficacy of anti-tumor reagents. We consider our findings in the context of an effective immunotherapy of malignancies.

It is also possible to use analogues of LPS in low doses to stimulate the systems to produce natural anti-tumor factors. One can vary chain length, lipid or sugar moieties to vary the effect and enhance it. The examples below serve to illustrate the invention without limiting it.

TUMORS. Meth A tumors were maintained in BALB/c mice by weekly passages. For transplants, 10 cells/0.2 ml, were injected intradermally into two sites of the shaved abdominal skin of (BALB/c x C57BL/6) F1 mice with a 30 gauge needle. The grade of tumor necrosis was scored 48 hours after LPS treatment in a grading system consisting of grades 0,1,2 and 3 according to Carswell et al. Proc.Natl. Acad. Sci., Supra. Tumor rejections were scored four weeks later. Grade 3 reflects total or near total necrosis of the tumor, grade 2 covers necrosis areas down to approximately one-half of the tumor, and grade 1, covers all necrosis areas smaller than one-half of the tumor.

MICE. BALB/c mice and (BALB/c x C57BL/6) F1 mice were purchased from Jackson Laboratories, Bar Harbor, Maine.

MATERIALS. We obtained TNF (D. Pennica et al, Nature, Supra) (recombinant human TNF, 0.49 mg/ml, 5.0 x 10⁷ u/mg where = micro) from Genentech, Cal; LPS (E:coli 0111:B4) from Ribi, Hamilton, MA; Indomethacin from Sigma Corporation in Missouri and Cyclophosphamide from Johnson and Co., Evansville, IL. TREATMENT. TNF or LPS was injected intraveneously. Indomethacin (100 ug/animal) was injected intraveneously together with LPS and/or TNF and subsequently added to the drinking water (20 ug/ml). Cyclophosphamide (30 mg/kg) was injected interperitoneally. Freshly prepared BALB/c mouse serum was injected intraveneously twice, two and 5 days after LPS injection.

EXAMPLE I

SYNERGISM BETWEEN LPS AND TUMOR-NECROSIS FACTOR

We wished to know whether TNF was the exclusive mediator of the anti-tumor effects of LPS or whether other mediators identified in LPS-induced tumor necrosis serum played also a role. We investigated this question by combining TNF and LPS in the treatment of tumor-bearing mice. Syngeneic Meth A sarcomas were transplanted into the skin of (BALB/c x C57 BL/6) F1 mice and the animals were treated with TNF and/or LPS 7 days later when tumors had grown to the size of approximately 1 cm in diameter.

The results are shown in Table 1. TNF is an effective inducer of tumor necrosis but its therapeutic dose range is small. Substantial degrees of necrosis can be achieved with the injection of 10 ug TNF. There is a rapid decline of necrosis at lower doses. Higher doses were toxic and not administered here. LPS, in a non-toxic dose range, is a less potent inducer of tumor necrosis than purified TNF. Added together, however, the two drugs synergize. They cause a more effective tumor necrotization together than either one of the two substances alone.

A synergism between LPS and TNF becomes also apparent if one measures the proportion of tumor-bearing mice which reject the tumor after treatment (Table 1). By necrotizing the tumors so effectively, TNF produces initially the appearance of tumor rejection within a week or two of treatment. Some tumor cells seem nevertheless to survive TNF treatment and resume growth subsequently leading to sizable tumors 4 weeks after treatment. In order to obtain reliable data in terms of tumor rejection as a consequence of TNF treatment it is important to wait at least 4 weeks before treatment results are assessed. Our results shown in Table 1 are obtained in this way. Both, LPS and TNF may cause tumor rejection alone, but both substances are substantially more active when added together. One microgram of LPS, for example, has no effect alone but, injected together with TNF, it increases the rejection rate substantially above that induced by TNF alone. EXAMPLE II

NEGATIVE FEEDBACK EFFECTS IN THE LPS-INDUCED ANTI-TUMOR RESPONSE

LPS, as mentioned above, induces the release of TNF and other mediators which mobilize the body's defense mechanisms. One such factor is interferon (IFN) (W.A. Stinebring, et al, Nature, 204:712-715, 1964), which is a potent inducer of NK cell activity (G. Trinchieri, et al, J. Exp. Med., 147: 1314-1332, 1978; M. Chun et al, J. Exp. Med., 150: 426-431, 1979; M. Chun et al, J. Immunol., 12:331-334, 1981). Another factor is interleukin-1, a mediator that stimulates the antigen dependent response of lymphocytes (M.K. Hoffman et al, Nature, 263:416-417, 1976; S.K. Durmm et al. Am. Rev. Immunol., 3:263-368, 1985; M.K. Hoffman et al, J. Immunol., 122:1371-1375, 1979).

IFN activates NK cells. We have considered the possibility that NK cell activation contributes to the response of tumor-bearing mice to LPS, and studied the activation of NK cells in vitro. We reported that IFN has a dual effect on NK cells; it increases their cytotoxic activity and induces at the same time the production of prostaglandin E (PGE) which blocks NK cell activation (M. Chun et al, Lymphokine Res., 1:91-98, 1982; M. Chun et al, J. Scan. Immunol., 22: 375-381, 1985). The production of PGE may be blocked with indomethacin (INDO) (H.R. Strausser et al, Int. J. Cancer, 15:724-730, 1975) and suppression of PGE production by INDO increases NK cell activation by IFN 5-50 times (M. Chun et al, J. Scan. Immunol., Supra).

EXAMPLE III

We were interested to know whether LPS induces

PGE-mediated negative feedback also in vivo and treated tumor bearing mice with both, LPS and INDO. Table 2 shows that INDO injected in conjunction with LPS doubles the rate of tumor rejection as compared to the result obtained with

LPS alone. The degree of necrosis induced with LPS is also increased by treatment of tumor bearing mice with INDO.

EXAMPLE IV

We reexamined the synergism between LPS and TNF when administered in conjunction with INDO (Fig. 1). Besides the fact that INDO increases the efficacy of both reagents our findings were consistent with those obtained without INDO; TNF and LPS synergize in causing acute necrosis of tumor transplants in mice (Fig. 1a) and synergize in allowing animals to reject tumors (Fig. 1b). It should be noted that no data is given for the combination of 10 ug TNF and 1 ug LPS per mouse. In this combination we observed toxicity and we lost some experimental animals.

LPS, may also stimulate antigen-specific lymphocyte reactions through the release of IL-1 (M.K. Hoffman et al. Nature, 263:416-417, 1976; S.K. Durmm et al. Am. Rev. Immunol., 3 : 262-368 , 1985; M.K. Hoffman et al, J. Immunol., 122:1371-1375, 1979). An immune response of T lymphocytes against tumor-associated antigens has been described by Berendt et al. as a consequence of LPS treatment (M.J. Berendt et al, J. Exp. Med., 148:1550-1559, 1978; M.J. Berendt et al, J. Exp. Med., 151:69-80, 1980). Not only immunologically specific effector T cells are generated but also tumor antigen related suppressor T cells whose activity was shown to antagonize the function of effector T cells and the ability of mice to reject tumors. The function of the suppressor cells was diminished by treatment of mice with low doses of cyclophosphamide (M.J. Berendt et al, J. Exp. Med., 151:69-80, 1980; P.W. Arkenase, J. Exp. Med., 141:697-702, 1975).

We wished to know whether a reduction of suppressor cell function may further increase LPS-mediated tumor rejection above the level already achieved by combination treatment with INDO. Table 2 shows that this is the case. Cyclophosphamide (Cy) injected two days after LPS and INDO allows almost all experimental animals to reject their tumors. We chose to delay injection of Cy as compared to the injection of LPS and INDO because preliminary experiments had shown Cy to interfere with the formation of tumor necrosis induced by LPS.

EXAMPLE V

A factor which reverses the immune suppression in tumor bearing mice has recently been discovered by Katz et al. in normal serum (I.R. Katz et al, J. Immunol., 134:3199-3203, 1985). The factor does not occur in plasma or in the serum of platelet deficient individuals and seems to be identical with platelet factor 4 as per all available evidence. Fifty microliters of normal mouse serum are reported to contain enough of this factor to reverse immune suppression in tumor bearing mice. We have examined this serum factor in our tumor system and confirmed its activity (Table 3). The serum factor increased the ability of tumor bearing mice to reject the malignancy in conjunction with LPS, TNF and INDO.

EXAMPLE VI

CORRELATION BETWEEN TUMOR NECROSIS AND TUMOR REGRESSION

Inspecting tumor bearing animals after treatment with either LPS or TNF we noted that TNF induces relatively consistent degrees of necrosis at a given dose. Plotting averaged degrees of necrosis against the percentage of tumor rejection associated with that necrosis, we found a linear relationship (Fig. 2). Such a clear correlation can not be seen with animals treated with LPS. Injected with 20 ug of LPS, mice may exhibit all degrees of necrosis from 0 to 3. It is interesting that, while TNF-treated animals regressed only necrotized tumors, LPS-treated mice rejected tumors regardless of whether they were necrotized or not. Only a high degree of necrosis favored rejection.

Bacterial products and in particular LPS have long been known to cause damage to tumors in animals and men (I. Parr et al, Brit. J. Cancer, 27:370-389, 1973; M.J. Shear, J. Nat. Cancer Inst., 4:461-476, 1944). Their therapeutical use has been hampered, however, by the high degree of toxicity to the tumor host. The discovery that tumor necrosis serum (TNS), the serum of LPS-treated animals, mediates the anti-tumor effects of LPS without apparent toxicity to the treated animals seemed to offer a solution to the toxicity problem (L.J. Old et al, Current enigmas in cancer research, Lectures, Series 67:273-315, Academic Press, New York and London, 1973; E.A. Carswell et al, Proc. Natl. Acad. Sci., 72:3666-3670, 1975). However, when the factor in TNS responsible for tumor necrosis, tumor necrosis factor (TNF), became available in purified form as product of recombinant bacteria, it was noted that it exhibits considerable toxicity as in mice. A single injection of 20 ug TNF may kill a mouse. Substantial anti-tumor activity on the other hand can not be achieved with less than 10 ug per mouse. This shows, that lethal and therapeutic doses of TNF are dangerously close together.

TNF is remarkably effective in causing necrosis of

Meth-A tumor transplants in mice but we noted that in order to cause lasting tumor elimination TNF must completely necrotize tumors. This is in striking contrast to LPS which causes tumor elimination often with minimal or no necrotization. Therefore the action mechanism of LPS/TNS may be different from the action mechanism of TNF and TNS may not exclusively act through TNF. In fact, in view of the long established notion that LPS-induced tumor necrosis and

LPS-induced tumor regression are two distinct phases of LPS mediated anti-tumor activities (M.J. Berendt et al, J. Exp. Med., 148:1550-1559, 1978; L.J. Old et al, Current enigmas in cancer research. Lectures, Series 67:273-315, Academic Press, New York and London, 1973), it would appear that TNF mediates only the early phase of tumor necrosis induction.

The experiments described here are based on the hypothesis that LPS-induced tumor necrosis serum contains a number of factors which engage different defense mechanisms with one of them involving TNF. Besides causing necrosis, TNS can also engage the immune system through IL-1 (M.K.

Hoffman et al. Nature, 263:416-417, 1976; S.K. Durmm et al.

Am. Rev. Immunol. 3:263-368, 1985; M.K. Hoffman et al,

J. Immunol., 122:1371-1375, 1979) and NK cells through IFN (G. Trinchieri et al, J. Exp. Med., 147:1314-1332, 1978; M. Chun et al, J. Exp. Med., 150:426-431, 1979). Each factor activates a defense mechanism which may have some anti-tumor activity alone and may cause tumor elimination but with the penalty of toxicity as we see in the case of TNF. Together they may be effective when applied at doses below toxicity levels.

The data presented here support this hypothesis in many ways. Both, TNF and LPS cause tumor necrosis but at a given degree of necrosis LPS is by far more effective in causing the tumor's complete rejection. In other words LPS not only causes necrosis of tumors but also enhances the ability of the tumor host to eliminate those tumor cells which escaped necrotization. If one considers the LPS-mediated anti-tumor reaction as consisting of two phases, the acute necrosis phase and the subsequent immunological rejection phase it would appear that TNF is active only in the first phase and LPS in both phases.

TNF may not be the only factor in LPS-induced TNS to account for effective necrotization of tumors because TNF synergizes with LPS also in causing tumor necrosis. It is unlikely that this can be attributed to the release of additional TNF. LPS causes necrosis only in rather high concentrations but synergizes with TNF in low concentrations. This would suggest that in order to produce TNF, LPS is required in high doses, while in low doses it causes the release of factors that synergize with TNF (LPS itself does not act directly on tumor cells) (E.A. Carswell et al, Proc. Nat. Acad. Sci. USA., 72:3666-3670, 1975). One likely candidate for such a synergistic factor is interferon which occurs in the serum of LPS-treated mice (W.A. Stinebring, et al, Nature, 204:712-715, 1964; M. Chun et al, J. Exp. Med., 150:426-431, 1979; M. Chun et al, J. Immunol., 12:331-334, 1981) and whose ability to synergize with TNF has been demonstrated (B.D. Williamson et al, Proc. Nat. Acad. Sci., 80:5397-5401/1983). The LPS-mediated elimination of tumor cells which survive the attack of TNF requires an intact immune system and therefore seems to be an immunological process. Berendt et al. described the generation of T cells specifically reactive with tumor associated antigens and the concommitant increase in suppressor T lymphocytes (M.J. Berendt et al, J. Exp. Med., 148:1550-1559, 1978; M.J. Berendt et al, J. Exp. Med., 151:69-80, 1980). We consider it likely that IL-1, of which LPS is such a strong inducer (Hoffman et al. Nature, 263:416-417, 1976; S.K. Durmm et al, Am. Rev. Immunol., 3:263-368, 1985; M.K. Hoffman et al, J. Immunol.,

122:1371-1375, 1979) is instrumental in the stimulation of tumor related T cell responses.

In considering immunotherapy as a treatment of malignancies one must not overlook the fact that an immune response automatically induces its own negative feedback. The feedback response is important to keep the response from overshooting but, when initiated prematurely or when stimulated in a chronic fashion - as in patients with advanced cancer (M.K. Hoffman et al, J. Immunol. Methods, 55:327-336, 1982) - it may completely block immune functions. Beneficial as feedback reactions may be for maintaining a balanced defense system, it may be desirable to block them at least temporarily when the task of eliminating malignancies demands it. There are at least two negative feedback reactions whose prevention in animals facilitates the rejection of experimental tumors. One inhibitory pathway is mediated presumably by prostaglandin E which we conclude from the fact that the pathway can be blocked by indomethacin. LPS, presumably via interferon, is a strong inducer of PGE synthesis (L.M. Pelus et al, J. Immunol., 123:2118-2125, 1979). We have previously observed a strong synergism between LPS or IFN and INDO in the generation of NK cells in vitro (M. Chun et al, J. Scan. Immunol. 22:375-381, 1985). In. that response IFN induces the activation of NK cells and simultaneously the production of PGE which inhibits NK cell activation. Blocking of PGE synthesis with INDO increases NK cell activation by IFN 5-50 times. In the tumor-bearing mouse INDO seems to block LPS-induced PGE production and it improves the animal's ability to reject the tumor.

Interference with the suppressor T cell-mediated negative feedback response has been shown previously to improve the animal's ability to eliminate tumor transplants (M.J. Berendt et al, J. Exp. Med., 151:69-80, 1980) and we show here that in concert with other measures to block negative feedback responses it increases the rejection rate to almost 100 percent. Cyclophosphamide has been shown to preferentially inhibit suppressor T cells at the dose applied in the experiments here (M.J. Berendt et al, J. Exp. Med., 151:69-80, 1980). In higher doses it affects also other lymphoid cell functions. Factors activated in serum during the clotting process and released by platelets have been shown to reverse cancer related suppression of the humoral immune response (I.R. Katz et al, J. Immunol., 134:3199-3203, 1985) and our data shows that such factors do also reduce suppressive effects on the host response against its tumor.

Applied in the right fashion LPS is thus an extremely effective anti-tumor therapeutic substance in mammals as shown against Meth-A tumor transplants in mice. Its main disadvantage, its toxicity, seems to be related to the production of TNF (B. Beutler et al, Nature, 316:552-554, 1985) whose sufficient production (judged by the degree of necrosis achieved) requires high doses of LPS. When TNF is used from an external source much lower doses of LPS are needed to complement TNF action. By apparently utilizing different defense systems of the body, LPS provides an example as to how these defense systems may interact to achieve an optimized attack against malignancies and the study of LPS action may help in designing new strategies in the treatment of malignancies by activating the body's means of defense in a concerted fashion. We also expect this method may prove effective when used together with other chemotherapeutic agents or chemical equivalents or variants of the natural agents induced by LPS. Thus purified natural TNF, IL-I, IL-2 or IFN may be used or recombinant material.

Therefore combination therapy between LPS and factors whose: release it induces such as TNF, IL-1, IL-2 or IFN is indicated as is combination therapy between factors whose release LPS causes: TNF, IL-1, IL-2, IFN.

We note that the combination of moderate doses-of TNF with low doses of LPS seem particularly important, since the high (and toxic) doses of LPS required for effective tumor treatment seem to be needed for endogenous TNF production. Lower (non toxic) doses of LPS are sufficient for full effectivity with exogenous TNF application.

Combination therapy involves direct (TNF) effects on tumor cells and indirect effects (mediated by the immune system). It does not relate to a combined effect of different factors directly on tumor cells such as seen in the synergism between TNF and IFN on L cells in vitro.

TNF effects may also be shown by other factors with the ability to cause necrosis of tumor cells and which may be released by different cell types. LPS effects may also be shown by other microbial or non microbial products which cause the release of immunoregulatory factors in mammals. There is contemplated use of other chemotherapeutic agents or natural factors which block the negative feedback inhibition against immune system activation. Indomethacin is only one example from a group of prostaglandin inhibitors such as aspirin or Voltaren (Ciba-Geigy Co., Basel, Switzerland).

Table 1. Tumor necrosis factor and LPS synergize in their anti-tumor activities.

Treatment Necrosis* Rejection^xx (ug/mouse)

TNF (10) 2.0 1/10

( 3) 1.4 1/10

( 1) 0.6 0/20 LPS (20) 1.1 6/20

(10) 0.4 0/20

( 1) 0 0/20

( 1) + TNF (3) 2.3 14/20

( 1) + TNF (1) 1.1 2/20

* averaged degrees at 48 hours after treatment. x^xnumber of rejections per tumors transplanted. Data combined from two experiments. Table 2. Indonethacin and Cyclopbosphamide enhance the anti-tumor effects of LPS.

Treatment Hecrosis* Hegression^xx (ug/mouse)

LPS (20) 1.1 21/64 (33%)

LPS (20) + INDO⁺ 1.9 43/64 (67%)

LPS (20) + INDO + Cy⁺⁺ 1.9 20/24 ( 83%)

INDO 0 0 Cy 0 0

* average degree of necrosis x^xnumber of rejections per tumors transplanted, + see materials and methods, +⁺30 mg/kg mouse, 2 days after LPS and INDO.

Data combined from all of 8 experiments. Table 3. Antisuppressor oell factor in normal mouse serum enhances the rejection rate of tumor transplants in mice.

Rejections

Treatment Saline NMS (ug/mouse) (50 ul/mouse) (50 ul/mouse)

LPS (1) TNF ( 3) 11/16* (69%) 15/16 (93%)

LPS (1) TNF ( 1) 3/16 (19%) 10/16 (63%) LPS (1) TNF (0.3) 0/8 ( 0%) 0/8 ( 0%)

*number of rejections per tumors transplanted.

Data combined from two experiments performed in this fashion.

Claims

1. Composition comprising pharmacologically effective, synergistically active, non-toxic combinations of LPS and natural factors induced by LPS for anti-tumor therapy in mammals.

2. Composition of claim 1 comprising a pharmacologically effective, synergistically active, non-toxic combination of LPS and one or more natural factors selected from the group consisting of TNF, IL-1, IL-2 and IFN for anti-tumor therapy in mammals.

3. Composition comprising a pharmacologically effective, synergistically active, non-toxic combination of TNF plus LPS plus one or more of the natural factors induced by LPS for anti-tumor therapy in mammals.

4. Composition of claim 3 wherein the natural factors induced by LPS are selected from one or more of the group consisting of IFN, IL-1 and IL-2.

5. Composition comprising a pharmacologically effective, synergistically active, non-toxic combination of TNF and one or more of the natural factors induced by LPS for anti-tumor therapy in mammals.

6. Composition of claim 5 wherein the natural factors induced by LPS are selected from one or more of the group consisting of IFN, IL-1 and IL-2 for anti-tumor therapy in mammals.

7. Composition comprising a pharmacologically effective, synergistically active, non-toxic combination of LPS, natural agents induced by LPS and chemotherapeutic agents.

8. Composition of claim 7 wherein the natural agent is selected from one or more of the group consisting of IL-1, IL-2, IFN and TNF.

9. Composition of claim 7 wherein the chemotherapeutic agent is selected from one or more of the group consisting of indomethacin, cyclophosphamide and platelet factor 4.

10. Composition comprising a pharmacologically effective, synergistically active, non-toxic combination of natural agents induced by LPS and chemotherapeutic agents.

11. Composition of claim 10 wherein the natural agent is selected from one or more of the group consisting of IL-1, IL-2, IFN and TNF.

12. Composition of claim 10 wherein the chemotherapeutic agent is selected from one ore more of the group consisting of indomethacin, cyclophosphamide and platelet factor 4.

13. Composition comprising a pharmacologically effective, synergistically active, non-toxic combination of LPS plus a first agent which stimulates the activity of the immune system and further comprising a second agent which blocks the negative feedback response against stimulation of the immune system.

14. Composition of claim 13 wherein the first agent is selected from one or more of a group of natural factors induced by LPS.

15. Composition of claim 14 wherein the natural factors are selected from one or more of the group consisting of TNF, IL-1, IL-2 and IFN.

16. Composition of claim 13 wherein the second agent is selected from one or more of the group consisting of indomethacin, cyclophosphamide and platelet factor 4.

17. Method for treatment of tumors to lower toxicity of anti-tumor agents which comprises treating the tumor with a pharmacologically effective, non-toxic amount of lipopolysaccharide when used serially or together with a. pharmacologically effective, non-toxic amount of an anti-tumor agent.

18. Method of claim 17, wherein the anti-tumor agent is selected from one or more of the group of natural factors induced by LPS.

19. Method of claim 18 wherein the natural factors are selected from one or more of the group consisting of TNF, IL-1, IL-2, and IFN.

20.. Method of claim 17 wherein the anti-tumor agent is a chemotherapeutic agent or other natural material not induced by LPS.

21. Method of claim 20 wherein the chemotherapeutic agent is indomethacin or cyclophosphamide.

22. Method of claim 20 wherein the natural anti-tumor agent not induced by LPS is platelet factor 4.

23. Method of claim 17 wherein the anti-tumor effect is synergistic.

24. Method for treatment of tumors to lower toxicity of anti-tumor agents which comprises treating the tumor with a pharmacologically effective, non-toxic amount of LPS when used serially or together with a first agent which stimulates, the activity of the immune system and a second agent which blocks the negative feedback response against stimulation of the immune system.

25. Method of claim 24 wherein the first agent is selected from one or more of a group of natural factors induced by LPS.

26. Method of claim 25 wherein the natural factors are selected from one or more of the group consisting of TNF, IL-1, IL-2 and IFN.

27. Method of claim 24 wherein the second agent is selected from one or more of the group consisting of indomethacin, cyclophosphamide and platelet growth factor 4.

28. Method of claim 24 wherein the second agent is selected from the group of prostaglandin inhibitors.

29. Method of claim 28 wherein the prostaglandin inhibitors are indomethacin, aspirin or Voltaren.