Interleukin-3 gene therapy for cancer

Abstract

A method for treating of tumors, in particular malignant solid tumors, using adenovirus-derived material and IL-3, so that the adenovirus preferably encodes IL-3 activity, which is given systematically to a mammal, optionally in an isolated perfusion setting. Preferably, IL-3 activity is combined with other cytotoxic activity.

Description

CROSS-REFERENCE TO RELATED APPLICATION

[0001] This application is a continuation of co-pending application Ser. No. 09/481,201 filed Jan. 11, 2000, now U.S. Pat. No. ______, which is a continuation of International Application No. PCT/NL98/00406, filed on Jul. 13, 1998 designating the United States of America, the contents of which is incorporated by this reference.

FIELD OF THE INVENTION

[0002] The present invention lies in the field of anticancer (gene) therapy. In particular, the invention relates to selective killing of (solid) tumor cells in a mammal by gene delivery via the blood circulation.

BACKGROUND

[0003] Many different kinds of solid tumors occur in the body of mammals, including humans. In many cases, these tumors are extremely difficult to treat, especially in advanced cancer with metastases. Currently available therapies include surgery, radiation therapy, chemotherapy, radio-immunotherapy, cytokine treatment and hyperthermia. All these therapies have important limitations and disadvantages. For example, surgery can only be performed on localized, accessible tumors; radiation and chemotherapy are associated with both acute and latent toxicity, and responses are often limited; radio-immunotherapy and hyperthermia have limited application and effectivity; and cytokine administration is often associated with toxicity and evokes many pleiotropic side effects. Often those therapies are combined to improve efficacy and to decrease toxic side effects. However, in general, the effectivity of those therapies and their combinations is still unsatisfactory.

[0004] More recently, gene therapy has been proposed as a novel approach to treat malignancies. The concept of gene therapy comprises the introduction of a molecule carrying genetic information into cells of a host, whereby the genetic information has a functional format. Genetic information may comprise a nucleic acid molecule that encodes a protein. In this case, "functional format" means that the protein can be expressed by the machinery of the host cell. The genetic information may also comprise or encode nucleic acid molecules with a sequence that is complementary to that of a nucleic acid molecule present in the host cell. The functional format in this case is that the introduced nucleic acid molecule or copies made thereof in situ are capable of base pairing with the complementary nucleic acid molecule present in the host cell. Genetic information may furthermore comprise a nucleic acid molecule that encodes or is itself a so-called ribozyme or deoxyribozyme. In this case, "functional format" means that the nucleic acid molecule or copies made thereof in situ are capable of specifically cleaving a nucleic acid molecule present in the host cell. The genetic information may furthermore comprise a nucleic acid molecule that encodes or is itself a so-called decoy molecule. In this case, "functional format" means that the nucleic acid molecule or copies made thereof in situ (nucleic acid molecules or proteins) are capable of specifically binding a peptide molecule present in the host cell.

[0005] Introduction of a molecule carrying genetic information into cells of a host is achieved by various methods known in the art. Those methods include, but are not limited to, direct injection of naked DNA constructs, bombardment with gold particles loaded with constructs, and macromolecule-mediated gene transfer using, e.g., liposomes, biopolymers, and the like. Preferred methods use gene delivery vehicles derived from viruses, including but not limited to adenoviruses, retroviruses, vaccinia viruses and adeno-associated viruses. Because of the much higher efficiency as compared to, e.g., vectors derived from retroviruses, vectors derived from adenoviruses (so-called adenoviral vectors) are the preferred gene delivery vehicles for transferring nucleic acid molecules into host cells in vivo.

[0006] The adenovirus genome is a linear double-stranded DNA molecule of approximately 36,000 base pairs ("bp"). The adenovirus DNA contains identical Inverted Terminal Repeats (ITR) of approximately 100 base pairs with the exact length depending on the serotype. The viral origins of replication are within the ITRs exactly at the genome ends. Adenoviruses can be rendered replication defective by deletion of the early-region 1 (E1) of their genome. Vectors derived from human adenoviruses (so-called adenoviral vectors), in which at least the E1 region has been deleted and replaced by a gene of interest, have been used extensively for gene therapy experiments in both preclinical and clinical phases. Apart from replication defective adenoviral vectors, helper independent or replication competent vectors, either or not containing a gene of interest, can also be used for gene therapy purposes. Adenoviral vectors have a number of features that make them particularly useful for gene therapy for malignancies. These features include (1) the biology of adenoviruses is characterized in detail, (2) adenoviruses are not associated with severe human pathology, (3) adenoviruses are extremely efficient in introducing their DNA into host cells, (4) adenoviruses can infect a wide variety of cells and have a broad host-range, (5) adenoviral vectors allow insertion of relatively large fragments of foreign DNA, (6) adenoviruses can be produced in large quantities with relative ease, and (7) adenoviral vectors are capable of transferring nucleic acid molecules very efficiently into host cells in vivo (Brody and Crystal, Ann. N.Y. Acad. Sci. 716:90-101, 1994).

[0007] The present inventors and their coworkers as well as others have demonstrated that recombinant adenoviral vectors efficiently transfer nucleic acid molecules to the liver of rats (Herz and Gerard, Proc. Natl. Acad. Sci. U.S.A., 96:2812-2816,1993) and to airway epithelium of rhesus monkeys (Bout et al., Gene Ther., 1:385-394,1994; Bout et al., Hum. Gene Ther., 5:3-10,1994). In addition, the present inventors, their coworkers and others have observed a very efficient in vivo adenoviral vector-mediated gene transfer into a variety of established solid tumors in animal models (lung tumors, glioma) and into human solid tumor xenografts in immune-deficient mice (lung) (Haddada et al., Biochem. Biophys. Res. Comm. 195:1174-1183, 1993; Vincent et al., Hum. Gene Ther., 7:197-205, 1996; reviewed by Blaese et al., Cancer Gene Ther., 2:291-297, 1995). Thus, preferred methods for in vivo gene transfer into tumor cells of nucleic acid molecules that encode molecules that can be used to kill tumor cells make use of adenoviral vectors as gene delivery vehicles.

[0008] The molecules that can be used to kill tumor cells include but are not restricted to suicide enzymes that convert a nontoxic prodrug into a toxic compound (e.g., the HSV-tk/ganciclovir system), cytokines, antisense nucleic acid molecules, ribozymes, and tumor suppressor proteins. In addition, treatment of cancer by gene therapy methods also includes the delivery of replicating vectors that are toxic to the tumor cells by themselves.

[0009] Gene therapy by introduction of nucleic acid molecules encoding suicide enzymes has been widely tested on a variety of tumor models. Especially the transfer of the Herpes simplex virus thymidine kinase (HSV-tk) gene into tumor cells in conjunction with systemic administration of the nontoxic substrate ganciclovir has proven to be an effective way of killing tumor cells in vivo (Esandi et al., Gene Ther., 4:280-287, 1997; Vincent et al., J. Neurosurg., 85:648-654,1996; Vincent et al., Hum. Gene Ther., 7:197-205, 1996). An important advantage of the HSV-tk/ganciclovir system is that, upon ganciclovir treatment, HSV-tk transduced tumor cells mediate a significant killing effect on neighboring untransduced tumor cells, the so-called bystander effect (Culver et al., Science 256:550-1552, 1992). Thus, using this approach, there is no absolute need for gene transfer into every individual cell in a solid tumor to achieve successful gene therapy. A limitation of this approach, however, is that the effect remains local. Consequently, the HSV-tk gene needs to be delivered into every individual solid tumor or metastasis throughout the body.

[0010] Gene therapy for cancer by the introduction of nucleic acid molecules encoding cytokines is based on the concept of enhancing the immune response against the tumor cells. The ultimate goal of this approach is to obtain regression of the treated tumor and simultaneously induce such a high degree of immunity that coexisting metastases are also destroyed. The mechanism by which the cytokine enhances the immune response against the tumor cells most likely in many cases involves eliciting an inflammatory type cell infiltration that results in improved antigen presentation. During the local inflammation, invading cells may lyse the tumor cells, releasing tumor antigens in a form that can be presented by other subpopulations of the invaders to T lymphocytes. These, in their turn, could act against coexisting metastases. Compared to administration of a cytokine protein, the gene transfer approach has the important advantage of high-level production of the cytokine at the site of the tumor, while systemic concentrations of the cytokine remain low. This avoids any pleiotropic and toxic side effects associated with cytokine. Signs of (partially) successful cancer treatment have been obtained with tumor cells expressing interleukin 2 (IL-2) (Fearon et al., Cell 60:397-401, 1990); Gansbacher et al., J. Exp. Med. 172:1217,1990), IL-4 (Golumbek et al., Science 254:713-716, 1991; Platzer et al., Eur. J. Immunol. 22:1729-1733, 1992), interferon-gamma (Gansbacher et al., Cancer Res. 50:7820-7824,1990), interferon-alpha (Ferrantini et al., Cancer Res. 53:1107-1112, 1993), TNF alpha (Blankenstein et al., J. Exp. Med. 173:1047-1052, 1991), IL-7 (Hock et al., J. Exp. Med. 174:1291-1298,1991; McBride et al., Cancer Res. 52:3931-3937,1992), G-CSF (Colombo et al., J. Exp. Med. 173:889-897, 1991), GM-CSF (Dranoff et al., Proc. Natl. Acad. Sci. U.S.A. 90:3539-3543, 1993), IL-12 (Tahara et al., Cancer Res. 54:182-189,1994), IL-1 (Apte et al., In: Cytokine-induced tumor immunogenicity, Acad. Press London, pp. 97-112, 1994; Apte et al., Folia Biol. Praha 40:1-18, 1994; Douvdevani et al, Int. J. Cancer 51:822-830, 1992; Nakata et al., Cancer Res. 48:584-588,1988; Zoller et al., Int. J. Cancer 50:443-449,1992) and IL-3 (McBride et al., Folia Biol. Praha 40:62-73, 1993; Pulaski et al., Cancer Res. 53:2112-2117,1993). The present inventors and their coworkers have previously observed partial regression of a nonimmunogenic solid tumor (L42 nonsmall cell lung cancer; Kal et al., NCI Monographs 6:111-114,1988; Kal et al., Radiother. Oncol. 6:231-238, 1986; Kal et al., J. Natl. Cancer Inst. 76:943-946, 1986) growing subcutaneously in WAG/Rij rats after intratumor injection of adenoviral vectors expressing IL-1a or IL-3. This regression occurred both in the injected tumor and in an untreated distant (contralateral) L42 tumor (patent application EP 96.202725, incorporated herein by reference).

[0011] Interleukin-3 (IL-3) is a cytokine well described as a hematopoietic growth factor that has a wide range of target cells including progenitor cells of every lineage, excluding cells committed to the T and B lymphoid lineage (Schrader et al., In: Lymphokines, Acad. Press, San Diego, 1988). The main production of IL-3 by activated T cells has led to the hypothesis that IL-3 is not involved in steady-state hematopoiesis but functions as a link between, on one hand, the T lymphocytes of the immune system, which sense invasion of the body by foreign materials, and, on the other hand, the hematopoietic system which generates the cellular elements that mediate defense and repair responses (Ihle, In: Immunoregulatory cytokines and cell growth, Karger, Basel, 1989). IL-3 exerts a broad spectrum of biological properties (Ihle et al., J. Immunol. 131:282, 1983), including stimulatory activity on several myeloid leukemia cell lines, formation of granulocyte-macrophage colonies, mast cell growth factor activity, P cell-stimulating activity and histamine-producing cell-stimulating factor activity. In addition, IL-3 is capable of promoting the proliferation of megakaryocyte colony-forming cells, of supporting the differentiation of eosinophils and pre-B-cell precursors, of supporting proliferation of Natural Cytotoxic (NC) cells but not Natural Killer (NK) cells and of promoting the formation of osteoblasts. IL-3 also stimulates the effector functions of monocytes, eosinophils and basophils, thereby having the potential to regulate inflammation and allergy (Elliott et al., J. Immunol. 145:167, 1990; Haak-Frendesco et al., J. Clin. Invest. 82:17, 1988; Lopez et al., J. Cell. Physiol. 145:69, 1990). Human endothelial cells express the IL-3 receptor, which expression is enhanced by tumor necrosis factor alpha (TNF-.alpha.). IL-3 stimulation of TNF-.alpha.-activated endothelial cells enhances IL-8 production, E-selectin expression and neutrophil transmigration (Korpelainen et al., Proc. Natl. Acad. Sci. U.S.A. 90:11137, 1993). This suggests that IL-3 plays a role in inflammation not only by stimulating effector functions of mature leukocytes but also by regulating their localization to sites of inflammation through its action on the endothelium.

[0012] There are several ways to administer recombinant adenoviral vectors with therapeutic genes into solid tumors that grow in a mammalian animal body. Currently, cancer gene therapy protocols predominantly use direct injection of the recombinant vector into the tumor (e.g., Haddada et al., Biochem. Biophys. Res. Comm. 195:1174-1183, 1993; Vincent et al., Hum. Gene Ther. 7:197-205, 1996). The major disadvantage of this application route is that metastases, and in particular micrometastases, in advanced cancer are practically impossible to reach with this approach. Therefore, such a gene therapy relies solely on a distant (immune mediated) effect of the introduced genetic information. Using current technology, the distant effect may not be expected to be complete and, consequently, may not be expected to cure the disease.

[0013] An alternative, and possibly better way of delivering genetic material into solid tumors and/or their metastases, could be by administering the recombinant adenoviral vector via the blood or lymphatic circulation. All established tumors, both primary and metastatized, that are larger than a few millimeters in diameter are vascularized (Folkman et al., J. Nat. Cancer Inst. 82:4, 1990; Folkman and Shing, J. Biol. Chem. 267:10931-10934,1992). In addition, distant metastases usually emerge after migration of tumor cells from the primary tumor through the blood or lymphatic circulation. Thus, all solid tumors are in close contact with the circulation and, in principle, could be reached via the circulation. Moreover, killing of a solid tumor does not necessarily depend on gene transfer into the tumor cells themselves. Gene therapy strategies have been proposed where genetic material (e.g., the HSV-tk gene) is introduced into endothelial cells of the tumor vasculature (e.g., W096/21416). This should result in destruction of the tumor vasculature, ultimately leading to tumor necrosis.

[0014] The total capillary surface area in an adult human is approximately 100 m.sup.2 comprising approximately 1.times.10.sup.12 endothelial cells, whereas the endothelial cell content of the vasculature of a solid tumor is about 4-log less (Chan and Harris, In: The Internet Book of Gene Therapy; Cancer Therapeutics, eds. R. E. Sobol and K. J. Scanlon, 1995, Appleton & Lange, CT, pp. 211-227). Based on these estimations, intravascularly administered adenoviral vectors only have a 0.01% chance of interacting with endothelial cells in the vasculature of a distant tumor. The proliferation index of endothelial cells in the vasculature of a tumor is about 100-fold higher than that of normal endothelial cells (Hobson and Denekamp, Br. J. Cancer, 1984,49:405-413). Thus, if gene delivery would preferentially occur into actively proliferating cells, the gene transfer efficiency into the chosen target cells could be raised to approximately 1%. However, because adenoviral vectors, in contrast to retroviral vectors, transduce both replicating and nonreplicating cells, the estimate of 0.01% gene transfer into cells of the tumor vasculature is more realistic. In any event, administering the adenoviral vector via the circulation is expected to result in at least 99% of the adenoviral vectors interacting and possibly being taken up by cells in normal tissues. This is highly undesirable with respect to toxic side effects of the procedure. For example, the introduction and expression of a suicide gene or an inflammation-eliciting cytokine gene should obviously not take place in the endothelium of the normal vasculature. Therefore, until the present invention, the common belief in the field has been that administering the adenoviral vector to the tumor via the circulation requires some sort of specific targeting of the adenoviral vector to the tumor or its vasculature (e.g., PCT International Publication No. WO 96/25947). Specific targeting may include specific interaction with and uptake by the intended target cells, as well as specific expression of the introduced genetic information in the intended target cells. Specific targeting was felt to be necessary to ensure efficient gene transfer and to avoid toxic side effects in other tissues. Many different molecules that are specifically expressed or upregulated on the cell surface of tumor cells or their vascular endothelial cells have been proposed as targets for specific uptake of gene transfer vectors. Examples of such molecules are carcinoembryonic antigen (CEA; Walther et al., Head-Neck 15:230-235, 1993), surface-bound vascular endothelial growth factor (VEGF; Plate et al., Int. J. Cancer 59:520-529, 1994; Brown et al., Hum. Pathol. 26:86-91, 1995), the avb.sub.3 integrin (Brooks et al, Science 264:569-571, 1994), endosialin (Rettig et al., Proc. Natl. Acad. Sci. USA 89:10832-10826, 1992) and radiation-induced E-selectin (WO 96/25947). However, specific interaction with and uptake by the intended target cells is extremely difficult to achieve, for two reasons; i.e. (1) most of the proposed target molecules are also expressed on normal tissue, albeit at lower levels, and (2) it is difficult to construct targeted gene delivery vehicles. Many years of research have been invested by many different investigators in devising targeted gene delivery vehicles for this purpose, without significant success. Perhaps eventually this goal will be reached, but not without a major research effort and significant investment.

[0015] As a further alternative way to accomplish functional expression of genetic material in the vasculature of tumors, it has been proposed to transfer genetic material into cultured endothelial cells ex vivo, followed by administration of cultured endothelial cells via the circulation (PCT International Publication No. WO 93/13807). This should result in selective incorporation of cultured endothelial cells at sites of active angiogenesis, including the vasculature of solid tumors. However, such a selective incorporation into the vasculature of solid tumors has not been shown to occur. Furthermore, the disadvantage of this approach is that it involves the isolation, ex vivo manipulation, and readministration of endothelial cells.

BRIEF SUMMARY OF THE INVENTION

[0016] The present invention provides an effective and relatively safe treatment of (solid) tumors in the bodies of mammals. This is accomplished by administration via the circulation of recombinant adenoviral vectors with a wild-type infection spectrum that carry an interleukin-3 gene in a functional format.

[0017] Thus, the invention provides the use of a recombinant adenoviral vector encoding IL-3 activity for manufacturing a pharmaceutical composition for the systemic treatment of tumors. For the present invention, IL-3 activity is defined as the protein itself, derivatives and/or fragments thereof having at least but preferably most or all of the biological functions of IL-3, although the amounts of activity displayed by these derivatives and/or fragments may vary. It is preferred that the systemic treatment is restricted to certain tissues, organs, or extremities, or certain combinations thereof, because adenovirus is, in principle, capable of infecting almost any cells in the host, so that the restriction enables one to avoid unnecessary infection, as well as a higher probability of infection of the proper targets. Thus, in a preferred embodiment, the invention provides a systemic treatment which includes isolated tissue perfusion. Tissue perfusion is intended to read on isolated tissues as well as organs and/or extremities or any combination thereof. Two approaches of isolated perfusion are provided, one whereby the isolated perfunded tissue includes the tumor and one whereby the isolated perfunded tissue excludes the tumor. In the second case, organs or body parts which are liable to be damaged by the treatment or which are likely to influence the uptake of virus by the target cells can be excluded from the system to which the adenoviral vector encoding IL-3 activity is provided. A preferred organ to be excluded according to the invention is the liver.

[0018] In the other isolated perfusion route, the vector is delivered to the isolated part only. It is preferred to deliver the vector in the form of a virus-like particle. This means that the vector is packed in an adenovirus shell. The most preferred virus-like particle is the human homolog of recombinant adenovirus IG.Ad.CMV.rIL-3 deposited at the ECACC under accession number V96071634 or a functional derivative thereof.

[0019] IL-3 is not only capable of inducing regression of tumors, but it is also capable of retarding or halting the growth of tumors over prolonged periods of time. Many cytostatic agents are also capable of accomplishing regression of tumors, but are not capable of holding the regressed tumor in check over a prolonged period of time. It is, therefore, advantageous to make combinations of IL-3 activity and other cytostatic activity to have the best of both worlds. Regression of the tumor by administration of one or a number of doses of a cytotoxic agent and obtaining further regression as well as retarding or halting the growth of the regressed tumor is accomplished by providing IL-3 activity.

[0020] Thus, the invention further provides a means for treating tumors comprising a pharmaceutical composition comprising IL-3 activity and a pharmaceutical composition comprising cytostatic activity. Preferably, the IL-3 activity is provided by a recombinant adenoviral vector (preferably in a virus-like particle) encoding activity to be given systemically, either in an isolated perfusion format or not. Preferably, the pharmaceutical composition comprising cytostatic activity is in a single dosage unit for injection into a solid tumor, to be given once or several times until the required dosage is reached.

[0021] Typically, the virus-like particle is present in an amount of from about 1.times.10.sup.6 to 5.times.10.sup.9 i.u. in a perfusion fluid. It is, of course, also possible that both activities are present in one composition.

[0022] Preferably, the cytostatic or cytotoxic activity is TNF-activity, Melphalan, or adriamycin. The invention further provides a pharmaceutical composition for systemic treatment of tumors comprising IL-3 activity provided by a recombinant adenoviral vector encoding such activity, whereby the pharmaceutical composition is a perfusion fluid. Preferably, the recombinant adenoviral vector is provided in the form of virus-like particles. Preferably, the virus-like particles are present in an amount of about 1.times.10.sup.6 to 5.times.10.sup.9 i.u.

[0023] The most preferred virus is the human homolog of recombinant adenovirus IG.Ad.CMV.rIL-3 deposited at the ECACC under accession number V96071634 or a functional derivative thereof.

[0024] The invention further provides a kit of parts for the treatment of tumors comprising a pharmaceutical composition comprising IL-3 activity, means for isolating certain tissues, and means for perfunding the isolated tissues. Hereby, the essential elements for performing a method of treatment according to the invention are given. The means for perfunding are preferably heart-lung machines or other equipment capable of perfunding and preferably oxygenating. Means for excluding certain organs, limbs and/or tissues are known in the art and references thereto can be found herein. If it is possible to exclude, then it is, of course, also possible to limit perfusion to organs, tissues or limbs which can be excluded. Of course, the IL-3 activity in the kit of parts is again preferably provided by a recombinant adenoviral vector-encoding activity, preferably in the form of a virus-like particle, preferably present in an amount of about 1.times.10.sup.6 to 5.times.10.sup.9 i.u. Preferably the kit of parts further comprises a pharmaceutical composition comprising cytostatic activity for the reasons already disclosed herein.

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0025] In the drawings, which illustrate what is currently considered to be the best mode for carrying out the invention:

[0026] FIG. 1 is a schematic presentation of the Ad5.pCLIP vector.

[0027] FIG. 2 depicts isolated limb perfusion or direct intratumoral injection with an adenoviral vector carrying the thymidine kinase gene, followed by treatment of ganciclovir. Rats bearing a BN175 tumor in their hind limb underwent ILP (.circle-solid., n=5) or were directly injected intratumoral (.largecircle., n=9) with an adenoviral vector carrying the thymidine kinase gene (TK), followed by treatment with ganciclovir (GCV). Sham ILP rats underwent a control ILP without adenovirus (.box-solid., n=2). Data represent the average.+-.SEM. When the error bars are not visible, errors are within the symbol size.

[0028] FIG. 3 illustrates isolated limb perfusion or direct intratumoral injection of BN175 sarcoma with an adenoviral vector carrying the interleukin-3 gene. Rats bearing a BN175 tumor in their hind limb underwent ILP (.circle-solid., n=9) or were directly injected intratumoral (.largecircle., n=11) with an adenoviral vector carrying the interleukin-3 gene (IG.Ad.CMV.rIL-3.beta.). Sham ILP rats underwent a control ILP without adenovirus (.box-solid., n=5). Nontreated BN175 bearing rats were not treated (.quadrature., n=4). Data represent the average.+-.SEM. When the error bars are not visible, errors are within the symbol size. Curves are fitted using an exponential model.

[0029] FIG. 4 illustrates isolated limb perfusion or direct intratumoral injection of ROS-1 osteosarcoma with an adenoviral vector without a therapeutic gene. Rats bearing a ROS-1 tumor in their hind limb underwent ILP (.circle-solid., n=4) or were directly injected intratumoral (.largecircle., n=4) with 1.times.10.sup.9 i.u. of a control adenoviral vector without a therapeutic gene (Ad5.PL+C). Sham ILP rats underwent a control ILP without adenovirus (.box-solid., n=6). Nontreated ROS-1 bearing rats were not treated (.quadrature., n=8). Data represent the average.+-.SEM. When the error bars are not visible, errors are within the symbol size. Curves are fitted using an exponential model.

[0030] FIG. 5 depicts isolated limb perfusion or direct intratumoral injection of ROS-1 osteosarcoma with an adenoviral vector carrying the interleukin-3 gene. Rats bearing a ROS-1 tumor in their hind limb underwent ILP (.circle-solid., n=9) or were directly injected intratumoral (.largecircle., n=6) with an adenoviral vector carrying the interleukin-3 gene (IG.Ad.CMV.rIL-3.beta.). Sham ILP rats underwent a control ILP without adenovirus (.box-solid., n=6 (this curve represents the same group as described in FIG. 4)). Data represent the average.+-.SEM. When the error bars are not visible, errors are within the symbol size. Curves are fitted using an exponential model.

[0031] FIG. 6 illustrates isolated limb perfusion or direct intratumoral injection of ROS-1 osteosarcoma with an adenoviral vector carrying the interleukin-3 gene driven by a weaker promoter, namely the MLP promoter. Rats bearing a ROS-1 tumor in their hind limb underwent ILP (.box-solid., n=4) or were directly injected intratumoral (.quadrature., n=4) with 1.times.10.sup.9 i.u. of an adenoviral vector carrying the interleukin-3 gene driven by the MLP promoter (IG.Ad.MLP.rIL-3.beta.). Sham ILP rats underwent a control ILP without adenovirus (.circle-solid., n=6) and nontreated rats were implanted with ROS-1 tumors but not treated (.largecircle., n=8) (the latter two curves represent the same group as described in FIG. 4). Data represent the average.+-.SEM. When the error bars are not visible, errors are within the symbol size. Curves are fitted using an exponential model.

[0032] FIG. 7 illustrates the antitumor effect of varying doses of an adenoviral vector carrying the interleukin-3 gene administered via the circulation (isolated limb perfusion). Rats bearing a ROS-1 tumor in their hind limb underwent ILP with 1.times.10.sup.5 i.u. (.largecircle., n=7), 1.times.10.sup.7 i.u. (.circle-solid., n=6), 1.times.10.sup.8 i.u. (.quadrature., n=6) or 1.times.10.sup.10 i.u. (.DELTA., n=2) of IG.Ad.CMV.rIL-3. The sham ILP curve (.diamond-solid., n=6) and the 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 ILP curve (.box-solid., n=9) are obtained from example 7 and are shown as reference curves. Data represent the average.+-.SEM. When the error bars are not visible, errors are within the symbol size. Curves are fitted using an exponential model.

[0033] FIG. 8 illustrates antitumor effect of doses below 1.times.10.sup.9 i.u. administered via the circulation (isolated limb perfusion) of an adenoviral vector carrying the interleukin-3 gene. Rats bearing a BN175 tumor in their hind limb underwent ILP with 1.times.10.sup.5 (.circle-solid., n=5), 1.times.10.sup.7 i.u. (.largecircle., n=5) of an adenoviral vector carrying the interleukin-3 gene. The sham ILP curve (.quadrature., n=6) and the 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 ILP curve (.box-solid., n=9) are obtained from example 7 and are shown as reference curves. Data represent the average.+-.SEM. When the error bars are not visible, errors are within the symbol size. Curves are fitted using an exponential model.

[0034] FIG. 9 depicts growth curves of ROS-1 tumors after isolated limb perfusions with 2.times.10.sup.8 i.u. and 5.times.10.sup.8 i.u. with an adenoviral vector carrying the interleukin-3 gene. Rats bearing a ROS-1 tumor in their hind limb underwent ILP for 15 minutes with 2.times.10.sup.8 i.u. (.tangle-solidup., n=5) or 5.times.10.sup.8 i.u. (.DELTA., n=4) IG.Ad.CMV.rIL-3. The 15 minutes sham ILP (.circle-solid., n=6) and 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 (.largecircle., n=9) ILP curves are obtained from example 7 and shown as reference curves. Data represent the average.+-.SEM. When the error bars are not visible, errors are within the symbol size. Curves are fitted using an exponential model.

[0035] FIG. 10 depicts growth curves of ROS-1 tumors after a 30 minute isolated limb perfusion with 1.times.10.sup.5 i.u. or 1.times.10.sup.7 i.u. of an adenoviral vector carrying the interleukin-3 gene. Rats bearing a ROS-1 tumor in their hind limb underwent ILP for 30 minutes with 1.times.10.sup.5 (.largecircle., n=2) or 1.times.10.sup.7 i.u. (.quadrature., n=2) of IG.Ad.CMV.rIL-3. The sham ILP curve (.circle-solid., n=6) is obtained from example 7 and shown as a reference curve. Data represent the average.+-.STD. When the error bars are not visible, errors are within the symbol size. Curves are fitted using an exponential model.

[0036] FIG. 11 illustrates growth curves of ROS-1 tumors after an isolated limb perfusion with an adenoviral vector carrying the interleukin-3 gene and the cytostatic melphalan. Rats bearing a ROS-1 tumor in their hind limb underwent ILP for 15 minutes with 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 in combination with 40 .mu.g melphalan (.box-solid., n--5) or with 40 .mu.g melphalan alone (.quadrature., n=5). The sham ILP (.circle-solid., n=6) and 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 curves (.largecircle., n=9) are obtained from example 7 and shown as reference curves. Data represent the average.+-.STD. When the error bars are not visible, errors are within the symbol size. Curves are fitted using an exponential model.

[0037] FIG. 12 depicts growth curves of ROS-1 tumors after isolated limb perfusion with an adenoviral vector carrying the interleukin-3 gene and TNF.alpha.. Rats bearing a ROS-1 tumor in their hind limb underwent ILP for 15 minutes with 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 in combination with 50 .mu.g TNF (.box-solid., n=5), with 50 .mu.g TNF.alpha. alone (.quadrature., n=4) or with 50 .mu.g TNF.alpha. and 40 .mu.g melphalan (.tangle-solidup., n=5). The sham ILP (.circle-solid., n=6) and 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 curves (.largecircle., n=9) are obtained from example 7 and shown as reference curves. Data represent the average.+-.STD. When the error bars are not visible, errors are within the symbol size. Curves are fitted using an exponential model.

[0038] FIG. 13 illustrates growth curves of ROS-1 tumors after an isolated limb perfusion with an adenoviral vector carrying the interleukin-3 gene and the cytostatic doxorubicin. Rats bearing a ROS-1 tumor in their hind limb underwent ILP for 15 minutes with 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 and 200 .mu.g doxorubicin (.circle-solid., n=4) or 200 .mu.g doxorubicin alone (.quadrature., n=5). The sham ILP (.box-solid., n=6) and 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 (.largecircle., n=9) curves are obtained from example 7 and shown as reference curves. Data represent the average.+-.STD. When the error bars are not visible, errors are within the symbol size. Curves are fitted using an exponential model.

[0039] FIG. 14 illustrates growth curves of ROS-1 tumors after isolated limb perfusions for 2 or 5 minutes with an adenoviral vector carrying the interleukin-3 gene. Rats bearing a ROS-1 tumor in their hind limb underwent ILP for 2 minutes (.largecircle., n=1) or 5 minutes (.quadrature., n=6) with 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 or underwent a sham ILP for 5 minutes (.box-solid., n=6). The 15 minute sham ILP (.diamond-solid., n=6) and 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 (.circle-solid., n=9) ILP curve is obtained from example 7 and shown as a reference curve. Data represent the average.+-.SEM. When the error bars are not visible, errors are within the symbol size. Curves are fitted using an exponential model.

[0040] FIG. 15 depicts generation of pMLPI.TK.

[0041] FIG. 16 is a map of the adapter plasmid pAd5/L420-HSA.

[0042] FIG. 17 shows the results of the expression assay and biological activity assay of the hIL-3 transgene of the IG.Ad5.CLIP.hIL-3 adenoviral vector.

[0043] FIG. 17A depicts an hIL-3 ELISA. Since the A.sub.450 nm is 0.541, it was concluded that the hIL-3 production was >2000 pg/ml (high production). hIL-3 transgene is produced.

[0044] FIG. 17B depicts a TF-1 bioactivity assay. TF-1 cells are, for their growth, dependent on the presence of hIL-3 in the culture medium. The less hIL-3 is present, the less well the cells grow. By a colorimetric assay with MTS/PMS (Promega), the proliferation can be monitored at A.sub.490 nm. A high proliferation (enough hIL-3) correlates with a high A.sub.490 nm. A dilution range of the samples is made to determine the hIL-3 present. The sample dilution table shows that the cells are proliferating (dilution 1: A 0.989) and that by dilution of the supernatant containing the hIL-3 protein, the proliferation of the cells decreases. Functional hIL-3 is formed.

DETAILED DESCRIPTION OF THE INVENTION

[0045] Despite the high potential of cancer gene therapy, the results of experimental treatment of solid tumors have until now been very disappointing. Direct injection of gene delivery vectors, mostly adenoviral vectors, carrying therapeutic genetic information into solid tumors has resulted in efficient gene transfer into tumor cells and has shown some, although still incomplete, tumor regression (e.g., see patent applications WO 95/05835 and EP 0 707071). The major limitation of this approach, however, has been that every solid tumor has to be individually injected. This makes clinical application of such a treatment far from realistic for most cancers, in particular for advanced cancers with metastases. The alternative approach, i.e., therapeutic gene delivery via the circulation after systemic intravascular administration of the gene delivery vector, has been associated with extremely low gene transfer efficiency into the tumor. The common belief in the field has been, therefore, that the gene transfer efficiency should be increased to obtain a significant therapeutic effect. It is also generally accepted that this should not be done by administering more gene delivery vectors, but by promoting the specific uptake of the gene delivery vector into the tumor cells or into the endothelial cells aligning the tumor vasculature. The reason for this is that high concentrations of untargeted gene delivery vectors cause (1) a stronger immune response, and (2) more toxicity due to delivery of the antitumor gene to other tissues. Furthermore, it is difficult and expensive to produce extremely high concentrations of gene delivery vectors.

[0046] The present inventors have made the surprising observation that adenoviral vector-mediated delivery of an interleukin-3 gene through administration via the circulation into the vasculature of solid tumors results in a very effective cancer treatment. Circulation is meant to include both the blood circulation and the lymphatic circulation. An adenoviral vector is not treated in any way to promote its specific uptake by the solid tumor cells or the endothelial cells aligning the tumor vasculature. The therapeutic effect of delivery is much more dramatic than could be expected from the low transduction efficiency (less than 1% transduced cells) that is obtained with administration via the circulation. Established solid tumors growing in relevant animal models regressed completely. The therapeutic effect is shown to be dependent on both administration via the circulation and the biological activity of the interleukin-3 encoded by the introduced gene.

[0047] The present invention, among other things, provides a recombinant adenoviral vector that carries a nucleic acid molecule that encodes interleukin-3 or a functional derivative or a fragment thereof. The nucleic acid molecule is provided with a format that allows functional expression of interleukin-3 in solid tumor cells and/or in endothelial cells of the vasculature of a solid tumor in the body of a mammal after administration of a recombinant adenoviral vector to the circulation of the mammal. The term "functional expression" is understood to mean production of interleukin-3 with biological activity that leads to killing of solid tumor cells. The format is conferred upon a nucleic acid molecule by including upstream of a nucleic acid molecule an activator (promoter and/or enhancer) nucleic acid molecule that preferably interacts with one or more transactivating transcription factors that are present in tumor cells or in cells of the vasculature of a tumor and downstream of the nucleic acid molecule a eukaryotic polyadenylation signal. An activator molecule may be derived from the adenovirus used to construct an adenoviral vector or from a different adenovirus. Alternatively, an activator molecule is of exogenous origin. Useful activator molecules in this aspect of the invention are derived from, e.g., the Cytomegalovirus Immediate Early promoter/enhancer or the Rous Sarcoma Virus LTR promoter/enhancer, but may also be derived from other activator molecules known in the art. In this aspect of the invention, it is preferred that the nucleic acid molecule encoding interleukin-3 is a functional derivative from or includes at least a functional fragment of a nucleic acid molecule isolated from the same species as a mammal. Because in most applications of the invention the mammal is a human, it is in most applications of the invention preferred that the nucleic acid molecule is a functional derivative from or includes at least a functional fragment of a nucleic acid molecule isolated from a human. The terms "functional derivative" and "functional fragment" are used here to indicate that the nucleic acid molecule encodes a peptide molecule with the same biological activity in kind, but not necessarily in amount, as interleukin-3. Many different examples of nucleic acid molecules encoding mutants of human interleukin-3 with functional interleukin-3 activity are given in European patent EP 0 413 383. It is furthermore preferred that the biological activity of interleukin-3 includes the elicitation of an intense local inflammation associated with an inflammatory-type cell infiltration. The recombinant adenoviral vectors according to the invention may be derived from any wild-type adenovirus serotype that allows the functional expression of interleukin-3 in solid tumor cells and/or in endothelial cells of the vasculature of a solid tumor in the body of a mammal after administration of the recombinant adenoviral vector to the circulation of the mammal. In the examples given infra to illustrate the present invention, recombinant adenoviral vectors are derived from human adenovirus type 5. It is to be understood, however, that those skilled in the art will be able to apply other recombinant adenoviral vectors without departing from the invention. Methods for the construction of recombinant adenoviral vectors according to the invention and for their propagation on useful packaging cells have been described in patent applications EP 0 707 071 and WO 97/00326, incorporated herein by reference. Other examples of vectors and packaging systems useful in the invention include, but are not limited to, those given in patent applications WO 93/19191, WO 94/28152, WO 96/10642, and WO 97/04119.

[0048] The present invention furthermore provides a pharmaceutical composition that comprises the recombinant adenoviral vector defined supra in combination with a diluent that is not toxic to the recipient mammal at the dosage used and that retains sufficient stability of the infectivity of the recombinant adenoviral vector for a time long enough to allow uptake of the recombinant adenoviral vector into the solid tumor cells and/or endothelial cells of the vasculature of a solid tumor after administration of the composition to the circulation of the recipient mammal. A typical nonlimiting example of a diluent according to this aspect of the invention is an isotonic saline solution that is sterile and that is buffered at a physiological pH. Preferably, the diluent furthermore contains serum-substituting ingredients. In the examples given infra to illustrate the present invention, Haemaccel (Behring Pharma) is used as a suitable diluent. It is to be understood, however, that those skilled in the art will be able to apply other diluents without departing from the invention. For some applications of the invention, it is furthermore preferred that the pharmaceutical composition is oxygenated prior to administration. Optionally, the recombinant adenoviral vector (or virus) is prepared in lyophilized form. In the latter case, the recombinant adenoviral vector is suspended in solution to obtain the pharmaceutical composition before administering the pharmaceutical composition to the circulation of the recipient mammal. Typically, a pharmaceutical composition comprising one dose contains at least about 1.times.10.sup.6, preferably about 1.times.10.sup.8 infectious units (i.u.) of the adenoviral vector of the invention, but in certain conditions it is preferred that it contains at least about 1.times.10.sup.9, more preferred 1.times.10.sup.10, or even more preferred 1.times.10.sup.11 i.u. The amount of virus to be provided depends on many parameters. As disclosed herein, only a very limited portion of the administered virus actually infects the target cells. This may be one reason to increase the amount of virus to be administered. Also the size of the tumor and/or the degree of its vascularization will influence the amount of virus required to get an effect. Another important aspect is, of course, the amount of IL-3 activity expressed by a cell infected with one or more viruses. This, of course, depends on the cell, but also on the promoter that drives the expression and its interaction with cell components of the expression machinery, etc.

[0049] Based on the rat studies, where a CMV promoter is driving the rat IL-3 gene, antitumor activity was measured after perfusion with 1.times.10.sup.9 i.u. of IG.Ad.CMV.rIL-3. Perfusion time was 15 minutes. The size of this tumor was approximately 1 cm.sup.3. Dose-finding studies are performed, where the range of the administered [perfused] adenoviral vector will increase from 1.times.10.sup.6 Up to 1.times.10.sup.10 i.u. Antitumor activity is measured according to the methods described. The lowest dose resulting in a maximal antitumor effect will be used to calculate the dose to be delivered to human tumors, assuming that the same promoter is used in the adenoviral vector harboring the human IL-3 gene or a derivative thereof. It is assumed that the infection of human cells by recombinant adenoviral vectors is 10.times. more efficient than rat cells. Furthermore, we assume the vascular bed or the tumor to be proportional to the tumor volume. Therefore, the optimal dose assessed in the rat model is extrapolated to the human situation by the following calculation: 1 dose delivered to humans = [ effective dose in rat ] .times. [ tumor volume ( human ) ] [ 10 .times. tumor volume ( rat ) ]

[0050] In another aspect, the invention provides a method to deliver the nucleic acid molecule that encodes interleukin-3 to solid tumor cells and/or endothelial cells of the vasculature of a solid tumor in the body of a mammal, whereby the adenoviral vector or pharmaceutical composition defined supra is administered to a site in the circulation of the mammal. "Circulation" is meant to include both the blood circulation and the lymphatic circulation. Thus, the administration is performed to any site in the body of the recipient mammal where the blood or lymph fluids of the mammal pass. Preferred sites of administration are intravenous or intraarterial, where it is further preferred that administration is into an artery located upstream of the tumor vasculature. There are several means to perform administration to the circulation. One of the means is by injection using, for example, a syringe, a catheter or another infusion system known in the art. Preferably, injection is performed at a controlled infusion rate. A much preferred means to perform administration to the circulation is by perfusion. Perfusion is a technique whereby the administered pharmaceutical composition is caused to pass through the circulation or through a part of the circulation. When the administration is performed by perfusion, it is furthermore preferred that perfusion is done multiple times by creating a closed circuit and repassaging the pharmaceutical composition through the circulation or part of the circulation. Typically, the causing to pass is done by using a pump device and perfusion is performed at a rate depending on the species of the mammal to which the pharmaceutical composition is being administered. For humans, the rate is often in the range of approximately 40-80 ml/min and perfusion is continued for a period of 60-90 minutes, but depending on the patient, type of tumor, and location thereof, these parameters may vary. For short treatment times (approximately 5-30 minutes) with the adenoviral construct, an anoxic perfusion can be performed by those skilled in the art by using balloon catheters to make a closed circuit. No heart-lung machine is necessary.

[0051] In this aspect of the invention, part of the circulation comprises the vasculature of the tumor or tumors to which gene delivery is performed. For optimal delivery of the nucleic acid molecule that encodes interleukin-3 to solid tumor cells and/or endothelial cells of the vasculature of a solid tumor, it is preferred that the adenoviral vector or composition of the invention does not pass through the liver or a part of the liver of the recipient mammal. Thus, part of the circulation does preferably not include the circulation of the liver or of a part of the liver, except when the tumor is located in or very close to the liver. For optimal delivery of the nucleic acid molecule that encodes interleukin-3 to solid tumor cells and/or endothelial cells of the vasculature of a solid tumor, it is furthermore preferred that the blood of the mammal is first washed away from the closed circuit (e.g., by precirculation with the diluent of the pharmaceutical composition only) before the pharmaceutical composition is administered. Optionally, the blood that is washed away is collected and readministered at the end of the procedure. Surgical techniques for perfusion of parts of the circulation according to the present invention are under development and are already available for various specific parts of the circulation, such as, e.g., the liver (Fraker et al., Circulatory shock, 44, p. 45-50, 1994), the lung (Progrebniak et al., Ann. Thorac. Surg., 57, p. 1477-83, 1994), and the kidney (Veen van de et al, Eur. J. Surg. Oncol. 20, p. 404-405, 1994). A typical nonlimiting example of a routine perfusion technique useful in the invention is isolated limb perfusion (ILP), where a closed circuit is created between the femoral artery and the femoral vein. Alternatively, essentially the same perfusion techniques can be employed in the invention to exclude the delivery of the nucleic acid molecule to a part or parts of the circulation. In this aspect of the invention, the part or parts of the circulation to which the delivery is unwanted are perfused with a diluent according to the invention while the pharmaceutical composition is administered to the circulation systemically (hence, outside the perfusion circulation). An important example of this embodiment of the invention is exclusion of the liver circulation from delivery of the nucleic acid molecule.

[0052] The invention furthermore provides genetically modified solid tumor cells and cells of the vasculature of a solid tumor expressing the interleukin-3 in the body of a mammal. These cells expressing the interleukin-3 are obtained by administering the composition containing the adenoviral vector according to the invention using the method according to the invention via the circulation of a mammal. The expression of interleukin-3 in a solid tumor cell or cells of the vasculature of a solid tumor results in an effective killing of the cells. Thus, the present invention also provides a gene therapy treatment for solid tumors. All tumors that are in close contact with the circulation can be treated according to the invention. Although leukemias and lymphomas are not excluded, vascularized solid tumors are especially suited for treatment according to the invention. Examples of types of solid tumors include, but are not limited to, carcinomas (e.g., of the lung, bladder, kidney, breast, stomach, pancreas, urogenital tract, and intestine), sarcomas (e.g., soft tissue sarcomas, osteogenic sarcomas, or Kaposi's sarcoma), gliomas and melanomas. Also benign types oftumors, such as, e.g., angiomas and fibrocytomas, can be treated according to the invention. It is to be understood, however, that the scope of the present invention is not to be limited to the treatment of any particular type of tumor.

[0053] It is furthermore to be understood that the cancer treatment according to the invention may be combined with other methods of cancer treatment known in the art. Such treatment combinations are also part of the present invention.

[0054] The invention is illustrated by means of the following examples. It is to be understood that the examples are not meant to limit the scope of the invention in any way.

[0055] Example 1 teaches the production of adenoviral vectors and pharmaceutical compositions according to the invention.

[0056] Example 2 teaches the cloning and production of an adenoviral vector with the human IL-3 gene and the pharmaceutical composition according to the invention.

[0057] Examples 3 and 4 show the gene transfer efficiency that is obtained when adenoviral vectors are administered to a solid tumor via the circulation or by direct intratumor injection, as well as the unwanted gene transfer into nontumor cells in both cases, and the type of cells in the tumor that are transduced using these administration methods. It is shown that the direct injection results in approximately 87 times more expression of the introduced gene in the tumor than administration via the circulation. The direct injection efficiently transduces many tumor cells along the needle tract, whereas administration via the circulation mainly transduces endothelial cells of the tumor vasculature, a few solid tumor cells adjacent to the vascular endothelial cells and some cells in or near the capsule of the tumor. Gene transfer into tissues other than the tumor hardly occurs using either method.

[0058] Examples 5, 6, 7 and 8 clearly demonstrate the effective antitumor effect that is accomplished by administering an adenoviral vector carrying the interleukin-3 gene into two types of solid tumors via the circulation. Complete regression of the tumors occurs. Experiments show that this effective antitumor effect is not obtained by direct intratumor injection or by using an adenoviral vector that expresses the IL-3 gene at low levels.

[0059] Control isolated limb perfusion experiments show that this effective antitumor effect is not obtained by an isolated limb perfusion without the addition of the adenoviral vector with the interleukin-3 gene or by treatment with an adenoviral vector without effector gene. After the latter treatment, some delay in tumor growth is observed when ROS-1 osteosarcomas were used, but the tumors do not regress. The latter growth delay is not observed when BN175 tumors were treated.

[0060] Example 5 shows that the antitumor effect is specific for the activity of the interleukin-3 gene. Anticancer treatment by administering adenoviral vectors expressing the HSV-tk gene via the circulation followed by ganciclovir injections shows only incomplete effects.

[0061] Examples 9, 10 and 12 demonstrate that in the rats with the two tumor models studied, the optimal dose for administering of the adenoviral vector carrying the interleukin-3 gene via the circulation is 1.times.10.sup.9 i.u. (infectious units), and that 15 minutes perfusion results in good antitumor effects.

[0062] Example 11 clearly demonstrates that the administration via the circulation of 1.times.10.sup.9 i.u. of the adenoviral vector carrying the interleukin-3 gene is at least as efficient as the established combination therapy with TNF (and Melphalan).

EXAMPLES

Example 1

Generation of Recombinant Adenoviral Vectors and Production of Pharmaceutical Compositions Containing the Recombinant Adenoviral Vectors for Administration via the Circulation of a Mammal

[0063] The cloning, sequence analysis and generation of E1-deleted adenoviral vectors has been described in detail in patent application EP 95 20 2213 incorporated herein by reference.

[0064] The adenovirus vector is deleted for the E1, but the E3 region was retained in this vector. The gene is driven by the Cytomegalovirus promoter (CMV) or the adenovirus-2 derived Major late promotor (MLP). The names of the viruses are IG.Ad.CMV.rIL-3 (this vector contains the rat IL-3 cDNA), IG.Ad.MLP.Luc (this vector contains the luciferase marker gene), IG.Ad.CMV.LacZ (this vector contains the LacZ marker gene) and Ad.CMV.TK (this vector contains the TK (thymidine kinase) gene).

[0065] Recombinant adenovirus IG.Ad.CMV.rIL-3 has been deposited at the ECACC under accession number V96071634.

[0066] The generation and propagation of these vectors on E1-complementing cell lines has been described in European patent application EP 95 20 2213 and in references Esandi et al. (1997) and Vincent et al. (1996a, 1996b). Propagation of the vectors on E1-complementing PER.C6 is described in patent application WO 97100326. The PER.C6 cell line has been deposited at the ECACC under deposition number 96022940. After propagation, the recombinant viruses were purified by CsC 1 density centrifugation and dialyzed according to standard procedures. Titration of the viruses was performed by end-point dilution on 911 cells. The vectors are stored in phosphate-buffered saline (PBS) supplemented with 10% (v/v) glycerol or 5% (w/v/) sucrose and stored at -80.degree. C.

Example 2

Generation of a Recombinant Adenoviral Vector with the Human IL-3 Gene (pAd5.CLIP.hIL-3) and Production of Pharmaceutical Compositions Containing the Recombinant Adenoviral Vector (IG.Ad5.CLIP.hIL-3) for Administration via the Circulation

[0067] 2.1.1. Generation of the Adenoviral Vector pAd5.CLIP:

[0068] The pAd5.CLIP adenoviral vector contains a deleted E1 gene, but the E3 region was retained in this vector. The adenoviral vector consists of a Cytomegalovirus promoter (CMV), polylinker, intron and polyA sequence. For those skilled in the art, it is possible to insert in the polylinker site of the pAd5.CLIP vector any piece of DNA of interest. In this case, the human interleukin-3 coding sequence was inserted (as described in example 2.2).

[0069] The cloning strategy of the pAd5.CLIP adenoviral vector is shown in FIG. 1 and described below.

[0070] PcDNA1 (Life Technologies) was digested with the restriction enzymes HhaI and AvrII. The sticky ends of the 1567 bp fragment were filled in with the enzyme T4-polymerase. The plasmid pAd/L420-HSA (the generation of this adapter plasmid is described in the next subparagraph of this example (2.1.2)) was digested with the restriction enzymes AvrII and BglII, followed by treatment with Klenow polymerase (Life Technologies), resulting in a 5.5 kb DNA fragment. The purified 5.5 kb pAd/L420-HSA fragment was dephosphorylated with Tsap (Thermo Sensitive Alkaline Phosphatase, Life Technologies) and ligated with the purified 1.5 kb pcDNA1/Amp/HhaI/AVrII fragment. The ligation product was added to transformation-competent DH5 (E. coli cells) and plated on ampicillin-containing plates. The resulting pAd5.CLIP plasmid DNA was isolated and checked by restriction digestion analysis.

[0071] 2.1.2. Construction of the Adapter Plasmid pAd/L420-HSA:

[0072] The absence of sequence overlap between the recombinant adenovirus and E1 sequences in the packaging cell line is essential for safe, RCA-free generation and propagation of new recombinant viruses. The adapter plasmid pMLPI.TK was designed for use according to the invention in combination with the improved packaging cell lines of the invention. The plasmid pAd/L420-HSA is derived out of the adapter plasmid pMLPI.TK. The construction strategy is described below (subparagraphs 2.1.2.1 and 2.1.2.2).

[0073] 2.1.2.1. Construction of pMLPI.TK:

[0074] The recombinant adenovirus vectors used (pE1A.E1B, pMLP.TK, see patent application EP 95/202213) are deleted for E1 sequences from nt. 459 to 3328. As construct pE1A.E1B contains Ad5 sequences nt. 459 to 3510, there is a sequence overlap of 183 nt between E1B sequences in the packaging construct pIG.E1 A.E.sub.1B and recombinant adenoviruses, such as, for example, IG.Ad.MLP.TK. The overlapping sequences were deleted from the new adenoviral vectors. In addition, noncoding sequences derived from LacZ, which are present in the original constructs, were deleted as well. This was achieved (see FIG. 15) by PCR amplification of the SV40 poly(A) sequences from pMLP.TK using primers SV40-1 (5'-GGGGGATCCGAACTTGTTTATTGCAGC-3' (SEQ ID NO: 1): introduces a BamHI site) and SV40-2 (5'-GGGAGATCTAGACATGATAAGATAC-3' (SEQ ID NO:2): introduces a BglII site). In addition, Ad5 sequences present in this construct were amplified from nt. 2496 using primer Ad5-1 (GGGAGATCTGTACTGAAATGTGTGGGC-3' (SEQ ID NO:3): introduces a BglII site) to nt. 2779 using primer Ad5-2 (5'-GGAGGCTGCAGTCTCCAACGGCGT-3' (SEQ ID NO:4)). Both PCR fragments were digested with BglII and were ligated. The ligation product was PCR amplified using primers SV40-1 and Ad5-2 (sequence described above). The PCR product obtained was cut with BamHI and AflII and was ligated into pMLP.TK predigested with the same enzymes. The resulting construct, named pMLPI.TK, contains a deletion in adenovirus El sequences from nt. 459-3510.

[0075] 2.1.2.2. Construction of the pAd/L420-HSA Adapter Plasmid:

[0076] The pMLPI.TK plasmid was used as the starting material to make a new vector in which nucleic acid molecules comprising specific promoter and gene sequences can be easily exchanged. First, a PCR fragment was generated from pZip(Mo+PyF101(N-) template DNA (described in PCT/NL96/00195) with the following primers: LTR-1:5'-CTGTACGTACCAGT GCACTGGCCTAGGCATGGAAAAATACATAACTG 3' (SEQ ID NO: 5) and LTR-2:5'-GCGGATCCTTCGAACCATGGTAAGCTTGGTACCGCTAGCGTTAACCGGGCGACTC AGTCAATCG-3' (SEQ ID NO:6). Pwo DNA polymerase (Boeringer Mannheim) was used according to the manufacturer's protocol with the following temperature cycles: once 5 minutes at 95.degree. C.; 3 minutes at 55.degree. C.; and 1 minute at 72.degree. C., and 30 cycles of 1 minute at 95.degree. C., 1 minute at 60.degree. C., 1 minute at 72.degree. C., followed by once 10 minutes at 72.degree. C. The PCR product was then digested with BamHI and ligated into a pMLP10 (Levrero et al., Gene 101: 195-202, 1991) vector digested with PvuII and BamHI, thereby generating vector pLTR10. This vector contains adenoviral sequences from bp 1 up to bp 454 followed by a promoter which includes part of the Mo-MuLVLTR in which the wild-type enhancer sequences are replaced by the enhancer from a mutant polyoma virus (PyF101). The promoter fragment was designated L420.

[0077] Next, the coding region of the murine HSA-gene was inserted. pLTR10 was digested with BstBI followed by Klenow treatment and digestion with NcoI. The HSA-gene was obtained by PCR amplification of pUC18-HSA (Kay et al., J. Immunol. 145:1952-1959, 1990) using the following primers: HSA 1:5'-GCGCCACCATGGGCAGAGCGATGGTGG C-3' (SEQ ID NO:7) and HSA2:5'-GTTAGATCTAAGCTTGTCGACATCGATCTACTAACAGTAGAGA TGTAGAA-3' (SEQ ID NO:8). The 269 bp amplified fragment was subcloned in a shuttle vector using the NcoI and BglII sites. Sequencing confirmed incorporation of the correct coding sequence of the HSA-gene, but with an extra TAG insertion directly following the TAG stop codon. The coding region of the HSA-gene, including the TAG duplication, was then excised as an NcoI (sticky)-SalI(blunt) fragment and cloned into the 3.5 kb NcoI (sticky)/BstI (blunt) fragment from pLTR10, resulting in pLTR-HSA10.

[0078] Finally, pLTR-HSA10 was digested with EcoRI and BamHI after which the fragment containing the left ITR, packaging signal, L420 promoter and HSA-gene was inserted into vector pMLPI.TK and digested with the same enzymes, thereby replacing the promoter and the gene sequences. This resulted in the new adapter plasmid pAd/L420-HSA (FIG. 16) that contains convenient recognition sites for various restriction enzymes around the promoter and gene sequences. SnaBI and AvrII can be combined with HpaI, NheI, KpnI, or HindIII to exchange promoter sequences, while the latter sites can be combined with the ClaI or BamHI sites 3' from the USA-coding region to replace genes in this construct.

[0079] 2.2. Generation of pAd5.CLIP.hIL-3:

[0080] PcDNA3.hIL-3 (PCT International Publication No. W088/04691, Gist Brocades (GB)) was digested with the restriction enzymes HindIII and BamHI to obtain a 1 kb fragment that contains the functional hIL-3 sequence (hIL-3/BamHI/HindIII (GB)). The plasmid pAd5.CLIP was digested with BamHI and HindIII (in the multiple cloning site) and purified. Next, the pAd5.CLIP/BamHI/HindIII DNA was dephosphorylated by Tsap. The pcDNA3.hIL-3/BamHI/HindIII and pAd5.CLIP/BamHI/HindIII DNA fragments were ligated and the ligation product was added to transformation-competent DH5 E. coli cells. The pAd5.CLIP.hIL-3(GB) plasmid DNA was isolated and checked by digestion with EcoRI. Because of possible unknown sequences at the HindIII site of the hIL-3 gene (pcDNA3.hIL-3 plasmid generated previously by Gist Brocades) and the presence of a polyA tail on the BamHI site, the hIL-3 cDNA coding sequence is amplified out of the pAd5.CLIP.huIL-3 (GB) by means of PCR. The primers used were: huIL3-forward: CCCCAAGCTTGCCACCATGAGCCGCCTGCCCGTC (SEQ ID NO:9) and huIL3-reverse: GCGGGATCCTCAAAAGATCGCGAGGC (SEQ ID NO:10) (Life Technologies). The PCR product (484 bp) was digested with the restriction enzymes BamHI and HindIII and the huIL-3/BamHI/HindIII (PCR) fragment (475 bp) was cloned in pAd5.CLIP.HindIII/BamHI. Plasmid DNA was obtained via transformation of competent DH5 E. coli cells. The 7131 bp construct is termed: pAd5.CLIP.hIL-3. The DNA was checked by restriction enzyme and sequence analysis. The expression and biological activity of the hIL-3 transgene was shown by transfection of the human PERC6 cell line (patent number WO 97/1000326, ECACC no. 96022940) with the pAd5.CLIP.hIL-3 plasmid followed by an hIL-3 protein ELISA (Quantikine kit, R & D) and hIL-3 bioactivity assay using TF-1 cells (human IL-3 protein dependent for their growth) of the secreted hIL-3 protein in the PER.C6 culture medium. The data of these two experiments (huIL-3 ELISA, TF-1 bioactivity assay) are shown in FIG. 17.

[0081] 2.3.Generation of Recombinant Adenovirus (IG.pAd5.CLIP.hIL-3):

[0082] A general protocol as outlined below and meant as a nonlimiting example of the present invention has been performed to produce the IG.pAd5.CLIP.hIL-3 recombinant adenovirus. Adenoviral packaging cells (PER.C6) were seeded in 25 cm.sup.2 flasks, and the next day when they were at approx. 80% confluency, the cells were transfected with a mixture of DNA and lipofectamine agent (Life Technologies) as described by the manufacturer. Routinely, 40 .mu.l lipofectamine, 4 .mu.g adapter plasmid and 4 .mu.g of the complementing adenovirus genome fragment AFLII-rITR were used. Two days later, cells were passaged to 80 cm.sup.2 flasks and further cultured. Approximately five days later, a cytopathic effect (CPE) was observed, indicating that functional adenovirus has been formed. Cells and medium were harvested upon full CPE and recombinant IG.pAd5.CLIP.hIL-3 adenovirus was released by freeze-thawing. An extra amplification step in an 80 cm.sup.2 flask was performed to increase the yield. After amplification, viruses were harvested and plaque purified on PER.C6 cells. Individual plaques were tested for production of viruses with the active IL-3 transgene by the hIL-3 ELISA and TF-1 bioactivity assay as described above. Functional IG.pAd5.CLIP.hIL-3 adenovirus was formed with an active IL-3 transgene.

[0083] Propagation of the vector on E1-complementing PER.C6 cells is described in PCT International Publication No. WO 97/100326. The PER.C6 cell line has been deposited at the ECACC under deposition number 96022940.

[0084] After propagation, the recombinant IG.pAd5.CLIP.hIL-3 adenovirus was purified by CsCl density centrifugation and dialyzed according to standard procedures. Titration of the viruses was performed by end-point dilution on 911 cells. The vector is stored in phosphate buffered saline (PBS) supplemented with 10% (v/v) glycerol or 5% (w/v) sucrose and stored at -80.degree. C.

Example 3

Comparison of Gene Transfer Efficiencies That are Achieved Using Recombinant Adenoviral Vectors That are Administered Either by Perfusion of the Circulation of a Tumor or by Direct Intratumor Injection

[0085] 3.1. Tumor models

[0086] The BN175 sarcoma (Marquet et al. (1983), Kort et al. (1984)) originated as a spontaneous tumor in the pancreatic, retroperitoneal region of a BN rat. The BN175 was implanted subcutaneously in the flank of donor BN rats and passaged serially. BN175 is a nonimmunogenic (Manusama (1996)), rapidly growing and metastazing tumor with a tumor doubling time of 2 days.

[0087] The rapidly growing ROS-l osteosarcoma originated spontaneously in the tibia of a Wag/Rij rat (Barendsen et al. (1987)). The ROS-1 was implanted subcutaneously in the flank of donor Wag/Rij rats and passaged serially. The ROS-1 has a tumor doubling time of 5 days.

[0088] Fragments of the BN 175 or ROS-1 (osteo)sarcoma were subcutaneously implanted into the right hind limb of the experimental animals just above the ankle. The tumor size was measured regularly by caliper measurement in two dimensions. When the tumor reached a volume of 180-524 mm.sup.3 (diameter between 7 and 10 mm) an isolated limb perfusion (ILP) with the recombinant adenoviral vector was performed. As a control, tumor-bearing rats underwent ILP without the addition of the recombinant adenoviral vectors (termed "sham ILP").

[0089] Another group of tumor-bearing rats underwent a direct intratumor injection with the recombinant adenoviral vector. The tumor volume was between 180-524 mm.sup.3. After ILP or direct intratumor injection, the tumor size was measured every (Mon-Fri)day.

[0090] 3.2. Surgical and Perfusion Techniques.

[0091] Surgical procedures were performed under Hypnorm anaesthesia (Janssen Pharmaceutica, Tilburg, The Netherlands). For isolated limb perfusion (ILP), a modification of the perfusion technique originally described by Brenckhuijsen et al. (1982) was used (Manusama et al. (1996)). After an incision parallel to the inguinal ligament, the femoral and vein were approached and cannulated with silastic tubing (0.30 mm ID, 0.64 mm OD; 0.64 mm, 1.19 OD, respectively, Degania Silicone, Degania Bet, Israel). Collaterals were temporarily occluded by the application of a tourniquet around the groin, which was fixed to the inguinal ligament. An oxygenation reservoir and a roller pump (Masterflex) were included in the vascularly isolated circuit, which was, initially, perfused with Haemaccel (Behring Pharma, Amsterdam, The Netherlands) for 3 minutes at a flow speed of 2 ml/min to wash out the blood. After the first wash-out step, recirculation was performed with recombinant adenoviruses (50 .mu.l-1 ml, containing approximately 1.times.10.sup.9 infectious units (i.u.)) dissolved in 2.5-3.5 ml Haemaccel at the same flow rate for a time period ranging from 5 to 30 minutes, followed by a second perfusion step of 5 minutes to wash out the nonbound virus with Haemaccel. During the perfusion and recirculation steps, the rat hind leg was kept at a constant temperature of 37-38.degree., and a warm water mattress was applied around the leg. After the second wash-out step, the vascularly isolated circuit was discontinued and, after cannule removal, the femoral vessels were ligated. Previous experiments have shown that the collateral circulation via the internal iliac artery to the leg is so extensive that ligation of the femoral vessels can be performed without detrimental effects.

[0092] 3.3. Results: The ILP or a direct intratumor injection was performed as described above with 50 .mu.l (containing approx. 5.times.10.sup.8 i.u.) of IG.Ad.MLP.Luc. Two days after the ILP or direct intratumoral injection, the rats were sacrificed and the tumor was removed. The luciferase activity was determined as described before (Fortunati et al., 1996). The luciferase activity was expressed in an amount of luciferase units per total volume of tumor lysate.

1TABLE I Effectivity of the transfer of adenoviral vector containing the Luciferase marker gene to a BN175 tumor via ILP or direct intratumor injection. perfusion time: (min) luciferase activity: No. rats 5 2039 .+-. 1197 6 15 4835 .+-. 2448 6 30 5647 .+-. 3308 6 intratumoral injection 422296 .+-. 271179 6

[0093] The luciferase activity is given as the mean value of 6 experiments.+-.S.D.

[0094] Conclusion: Direct intratumor injection results in an 87-fold higher expression of the luciferase activity in the tumor than ILP for 15 min.

[0095] ILP for 15 minutes results in an acceptable level of luciferase activity in the tumor compared to 30 minutes of ILP. Therefore, 15 min. of ILP is used throughout the experiments.

2TABLE II Transfer of adenoviral vector containing the Luciferase marker gene to other organs after ILP or direct intratumor injection. perfusion time: Luciferase activity: No. of rats: Skeletal muscle of (min) Liver: the isolated limb: 5 65 138 6 15 51 178 6 30 69 211 6 intratumor injection 38 196 6

[0096] The luciferase activity is given as the mean value of 6 experiments.

[0097] Conclusion: No high uptake of IG.Ad.MLP.Luc by the liver or skeletal muscle of the isolated limb after ILP or intratumor injection.

Example 4

Histocytological Examination of the Transduced Cells in a Tumor After Administration of a Recombinant Adenoviral Vector by Either Perfusion of the Circulation of the Tumor or Direct Injection into the Tumor

[0098] The ILP or direct intratumor injection of BN175 tumor-bearing BN rats with 5.times.10.sup.8 or 1.times.10.sup.10 i.u. IG.Ad.CMV.LacZ was performed as described above. Two days after treatment, the animals were sacrificed and tumors were removed. In these tissues the lacZ positive cells were localized by staining with X-gal as described in detail before (Bout et al., 1994).

[0099] Briefly, tumors were cut into slices of approximately 5 mm thickness, fixed for 2-3 hours in PBS containing 2% paraformaldehyde and 0.25% glutaraldehyde. After staining with X-gal, the tissue was post-fixed in 4% phosphate-buffered formalin and embedded in paraffin. 5 .mu.m sections were prepared according to routine histochemical methods. The sections were examined microscopically for the presence of blue (=LacZ positive) cells.

[0100] Results: Direct injection of the tumors with IG.Ad.CMV.LacZ resulted in staining along the track of the needle.

[0101] Staining of the tissues after ILP showed no blue color in the tumor. The color was restricted to the areas directly adjacent to the blood vessels of the tumor including the endothelial cells and in or near the capsule of the tumor.

[0102] The amount of blue-stained cells was much larger in the direct intratumoral-injected tumors than in the ILP-treated tumors.

Example 5

Effect of Direct Intratumor Injection or Administration Via the Circulation (Isolated Limb Perfusion) of an Adenoviral Vector Carrying the Thymine Kinase Gene

[0103] Rats bearing a BN175 tumor in their hind limb (as described in example 3.1) underwent ILP with 100 .mu.l of IG.Ad.CMV.TK (approx. 1.times.10.sup.9 i.u.) followed by intraperitoneal injection twice a day of ganciclovir (GCV). The tumor sizes were followed in time. The results are depicted in FIG. 2. Another group of BN rats was injected intratumorally in the BN175 tumor with 100 .mu.l of IG.Ad.CMV.TK followed by intraperitoneal injection twice a day of ganciclovir (GCV). The tumor sizes were followed in time. The results are depicted in FIG. 2.

[0104] Other rats underwent ILP without the addition of virus (control perfusion termed "sham ILP"). The tumor sizes were followed in time. The results are depicted in FIG. 2.

[0105] Results and conclusions: Isolated limb perfusion (ILP) with TK/GCV (suicide gene therapy) is not effective, neither is a control ILP (sham ILP) without the addition of virus.

Example 6

Effect of Direct Intratumoral Injection or Administration Via the Circulation (Isolated Limb Perfusion) of BN175 Sarcomas with an Adenoviral Vector Carrying the Interleukin-3 Gene

[0106] Rats bearing a BN175 tumor in their hind limb (as described in example 3.1) were injected intratumorally (IT) with 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3.

[0107] An other group of rats bearing a BN175 tumor in their hind limb underwent ILP for 15 minutes with 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 or perfusion buffer alone (the latter control perfusion is termed: "sham ILP"). The tumor sizes were followed in time. The results are depicted in FIG. 3.

[0108] Results and conclusions: Delivery of rIL-3.beta. (via an adenoviral vector) results in a relative delay of the tumor growth of 15 days (determined for an arbitrarily chosen tumor volume of 1000 mm.sup.3 and compared to the control sham ILP) in 8/9 of the treated BN 175 tumors when the vector is delivered via ILP. One out of nine treated tumors shows no antitumor response (normal growth). Direct intratumoral injection of the IG.Ad.CMV.rIL-3.beta. does not influence the tumor growth.

Example 7

Effect of Direct Intratumoral Injection or Administration Via the Circulation (Isolated Limb Perfusion) of ROS-1 Osteosarcomas with an Adenoviral Vector Carrying the Interleukin-3 Gene

[0109] ROS-1 tumor bearing rats (as described in example 3.1) were injected intratumorally (IT) with 1.times.10.sup.9 i.u. Ad5.PL+C.

[0110] Ad5.PL+C is an adenovirus that is E1 deleted and has no tg containing adenoviral sequences and carries no gene (=an "empty" control vector).

[0111] Another group of rats underwent ILP for 15 minutes with 1.times.10.sup.9 i.u. Ad5.PL+C.

[0112] The tumor sizes were followed in time. The results are depicted in FIG. 4.

[0113] Another group of rats bearing a ROS-1 tumor in their hind limb was injected intratumorally (IT) with 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3.

[0114] Other ROS-1-bearing rats underwent ILP for 15 minutes with 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 or perfusion buffer alone (the latter control perfusion is termed "sham ILP").

[0115] The tumor sizes were followed in time. The results are depicted in FIG. 5.

[0116] Results and conclusions: Delivery of IG.Ad.CMV.rIL-3 results in an antitumor effect (regression of the tumor growth) of all treated ROS-1 tumors when the adenovirus is delivered by ILP. Direct intratumoral injection with the same amount of IG.Ad.CMV.rIL-3 (10.sup.9 i.u.) is not effective.

[0117] The observed antitumor response is caused by the IL-3 gene and not by the adenovirus itself since the control virus Ad5.PL+C does not influence the tumor growth.

[0118] A mock ILP without the addition of adenovirus or drug (termed "sham ILP") delays the tumor growth compared to the nontreated tumors by 4 days (at the arbitrary chosen tumor volume of 2000 mm.sup.3). Compared to the sham ILP, the ILP treatment with 10.sup.9 i.u. IG.Ad.CMV.rIL-3 results in a 9-day delay in tumor growth. The observed delay in tumor growth after a sham ILP is explained by the wash out of growth factor(s) or the need for oxygen for the growth of the ROS-1 tumor and is characteristic for ROS-1.

[0119] In conclusion: Similar antitumor responses are observed after isolated limb perfusion with 1.times.10.sup.9 i.u. of IG.Ad.CMV.rIL-3 in two types of tumors.

Example 8

Antitumor Effect of an Adenoviral Vector Harboring the IL-3 Gene Driven by a Weaker Promoter, Namely the MLP Promoter

[0120] Rats bearing the ROS-1 tumor in their hind limb (as described in example 3.1) underwent ILP for 15 minutes with 1.times.10.sup.9 i.u. IG.Ad.MLP.rIL-3 (IG.Ad.MLP.rIL-3 is an adenovirus similar to IG.Ad.CMV.rIL-3; the only difference is the approximately 10-fold weaker MLP (major late promoter) promoter).

[0121] Another group of rats bearing the ROS-1 tumor in their hind limb was injected intratumorally with 1.times.10.sup.9 i.u. IG.Ad.MLP.rIL-3. The tumor sizes were followed in time. The results are depicted in FIG. 6.

[0122] Results and conclusions: The results show that isolated limb perfusions or direct intratumoral injections of ROS-1 osteosarcoma with an adenoviral vector with the IL-3 gene driven by a (10-fold) weaker promoter (IG.Ad.MLP.rIL-3.beta.) is not effective.

Example 9

Antitumor Effect of Varying Doses of an Adenoviral Vector Carrying the Interleukin-3 Gene Administered Via the Circulation (Isolated Limb Perfusion)

[0123] Rats bearing a ROS-1 tumor in their hind limb (as described in example 3.1) underwent ILP with 1.times.10.sup.5 i.u., 1.times.10.sup.7 i.u., 1.times.10.sup.8 i.u. and 1.times.10.sup.10 i.u. of IG.Ad .CMV.rIL-3 for 15 minutes. The tumor sizes were followed in time. The results are depicted in FIG. 7.

[0124] Another group of rats with BN175 sarcomas in their hind limb (as described in chapter 3.1) underwent ILP for 15 minutes with 1.times.10.sup.5 i.u. and 1.times.10.sup.7 i.u. of IG.Ad.CMV.rIL-3. The tumor sizes were followed in time. The results are depicted in FIG. 8.

[0125] Other rats with ROS-1 tumors underwent ILP for 15 minutes with 2.times.10.sup.8 i.u. or 5.times.10.sup.8 i.u. of IG.Ad.CMV.rIL-3. The tumor sizes were followed in time. The results are depicted in FIG. 9.

[0126] Results and conclusions: The results show that doses of 1.times.10.sup.5 i.u. or 1.times.10.sup.7 i.u. IG.Ad.CMV.rIL-3 result in no regression of the tumor growth after ILP both for ROS-1 and BN175 tumors.

[0127] Isolated limb perfusions of ROS-1 tumors with 1.times.10.sup.8 i.u. of IG.Ad.CMV.rIL-3 show a small regression of the tumor growth. ILP of ROS-1 with 2.times.10.sup.8 i.u. and 5.times.10.sup.8 i.u. (performed with another breed of rats, but with the same virus batch) did not show a regression of the tumor growth.

[0128] Isolated limb perfusions of ROS-1 osteosarcomas with 1.times.10.sup.10 i.u. of IG.Ad.CMV.rIL-3 is lethal for the rats. At day 8 after treatment, both rats were dead, probably due to severe leucocytosis (leucocytes increased approx. 17-fold to 200-300.times.10.sup.3 leucocytes/mm.sup.3). At this dose an arrest in tumor growth is observed.

[0129] In conclusion: the most optimal dose of IG.Ad.CMV.rIL-3 tested for isolated limb perfusion so far is 1.times.10.sup.9 i.u.

Example 10

Antitumor Effect of 30-Minute Isolated Limb Perfusions with a Low Dose of an Adenoviral Vector Carrying the Interleukin-3 Gene

[0130] ROS-1 tumor-bearing rats (as described in example 3.1) underwent ILP for 30 minutes with 1.times.10.sup.7 i.u. or 1.times.10.sup.5 i.u. of IG.Ad.CMV.rIL-3. The tumor sizes were followed in time. The results are depicted in FIG. 10.

[0131] Results and conclusions: An increase of the perfusion time from 15 to 30 minutes does not result in a better antitumor effect when a dose of 1.times.10.sup.5 i.u. or 1.times.10.sup.7 i.u. IG.Ad.CMV.rIL-3 is used.

Example 11

Antitumor Effect of Administration Via the Circulation (Isolated Limb Perfusion) of an Adenoviral Vector Carrying the Interleukin-3 gene and TNF.alpha. or Melphalan

[0132] Rats bearing a ROS-1 osteosarcoma in their hind limb (as described in example 3.1) underwent ILP with a mixture of a dose of 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 and a dose of 40 .mu.g (=effective dose) of the cytostatic melphalan (Alkeran.RTM., Glaxo Wellcome, UK) per rat. The drug and recombinant adenoviral vector were added to the oxygenation chamber of the perfusion system immediately after each other. The tumor sizes were followed in time.

[0133] Another group of rats was ILP treated with a total dose of 40 .mu.g melphalan per rat. The results are depicted in FIG. 11.

[0134] Other ROS-1 bearing rats underwent ILP with a mixture of a dose of 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 and a total dose of 50 .mu.g TNF.alpha. (Boehringer Ingelheim, Germany) per rat or a dose of 50 .mu.g TNF.alpha. alone (=the effective concentration). The tumor sizes were followed in time.

[0135] Another group of rats was ILP treated with a mixture of a dose of 50 .mu.g TNF.alpha. and a dose of 40 .mu.g melphalan per animal. The tumor sizes were followed in time. The results are depicted in FIG. 12.

[0136] Another group of rats bearing a ROS-1 tumor in their hind leg under went ILP with 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 and a dose of 200 .mu.g doxorubicin (Adriblastina RTU, Farmitalia Carlo Erba) per rat. Rats underwent ILP with a dose of 200 .mu.g doxorubicin per animal. The tumor sizes were followed in time. The results are depicted in FIG. 13.

[0137] Results and conclusions: ILP with a monotherapy of 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3 in the ROS-1 osteosarcoma model is at least as efficient as the established TNF.alpha./melphalan combination therapy (similar antitumor effects).

[0138] Addition of TNF.alpha., melphalan or doxorubicin to the IG.Ad.CMV.rIL-3 perfusion does not influence the antitumor response. This indicates that there is no additional benefit of these established antitumor agents when used at their indicated effective concentrations.

Example 12

Effect of Shorter Perfusion Times on the Antitumor Effect of an Adenoviral Vector Carrying the Interleukin-3 Gene

[0139] Rats bearing a ROS-1 osteosarcoma in their hind limb (as described in example 3.1) underwent ILP (as described in example 3.2) with 1.times.10.sup.9 i.u. of IG.Ad.CMV.rIL-3 for 5 or 2 minutes. As a control, a group of rats underwent ILP for 5 minutes with perfusion buffer alone (=sham ILP). The tumor sizes were followed in time. The results are depicted in FIG. 14.

[0140] Results and conclusions: The results show that regression of the tumor growth comparable to a 15-minute ILP is not observed when the perfusion time is reduced to 5 or 2 minutes.

Example 13

In Vitro Experiments with Human Cells with an Adenoviral Vector Carrying the Human Interleukin-3 Gene

[0141] Human melanoma, sarcoma, Karposi sarcoma cells and human umbilical vein endothelial cells (HUVEC) are infected with an adenovirus with the human interleukin-3 gene (IG.Ad5.CLIP.hIL-3 (described in example 2)). The effect of adenoviral infection and hIL-3 protein expression is studied by the determination of the growth curve of the cells in the presence and absence of IG.Ad5.CLIP.hIL-3 for a period of 3-4 weeks after the infection.

[0142] Other cultures of the above-described cells are infected in a similar way. The protein production of the huIL-3 transgene is determined for 3 weeks by means of an huIL-3 ELISA (Quantikine, R & D systems) and a TF-1 cell activity assay.

Example 14

In Vitro Experiments with Rat Tumor Cells with an Adenoviral Vector Carrying the Rat Interleukin-3 Gene

[0143] Rat sarcoma (BN175) and osteosarcoma (ROS-1) cell lines are infected with an adenovirus with the rat interleukin-3 gene (IG.Ad.CMV.rIL-3 (described in example 1)). The effect of adenoviral infection and rIL-3 protein expression is studied by the determination of the growth curve of the cells in the presence and absence of IG.Ad.CMV.rIL-3 adenovirus for a period of 3-4 weeks.

[0144] Other cultures of the same tumor cells are infected in a similar way. The protein production of the rat IL-3 transgene is determined for 3 weeks by means of an FDCP-1 cell activity assay.

Example 15

[0145] In vivo experiments with tumor-bearing rats with an adenoviral vector carrying the human interleukin-3 gene. ROS-1 osteosarcoma-bearing rats are treated via ILP (as described in section 3.2) for 15-30 minutes with 1.times.10.sup.9 i.u. of IG.Ad5.CLIP.hIL-3. Blood is sampled and tumor sizes are measured in time. The amount of huIL-3 in the rat blood is determined by means of an huIL-3 ELISA (Quantikine, R&D systems).

Example 16

Effect of Administration Via the Circulation (Isolated Limb Perfusion or Intraarterial Infusion) of an Adenoviral Vector Carrying the Rat Interleukin-3 Gene with the Liver Excluded From the Circulation

[0146] 16.1. Isolated Liver Perfusion Technique.

[0147] Surgical procedures are performed under ether anaesthesia. For isolated liver perfusion, the protocol as described in detail by Marinelli et al. (1990) was used.

[0148] Isolated liver perfusion (ILP) involves complete vascular isolation of the liver during perfusion. For this, a mid-line abdominal incision is made and two limbs of inflow were established by inserting cannulas into the pyloric branch of the portal vein and into the gastroduodenal branch of the common hepatic artery with their tips into the portal vein and the hepatic artery, respectively. The outflow limb is a cannula inserted into the caval vein. For a complete vascular isolation of the liver, all normal in- and outflow routes are clamped, the caval vein between the liver and diaphragm and between the cannula and the renal veins, the aorta proximal of the coeliac axis, the common hepatic artery and portal vein just proximal of the cannulas. The liver is perfused with the isotone perfusion fluid Haemmaccel (Behring Pharma, The Netherlands) for a time period of 10-45 minutes depending on the study protocol.

[0149] 16.2. Experimental Setup:

[0150] Rats receive implants in their hind limb with ROS-1 or BN175 tumors (as described in example 3.1). The liver is isolated from the blood circulation by means of an isolated liver perfusion as described in example 16.1. At the time the liver is excluded from the circulation, the rats are intravenously or intra-arterially injected with 1.times.10.sup.5-10.sup.10 i.u. of IG.Ad.CMV.rIL-3.

[0151] The isolated liver perfusion is performed for 10-45 minutes, after which the blood circulation in the liver is restored. Another group of tumor-bearing rats is intravenously or intraarterially injected with 1.times.10.sup.5-10.sup.10 i.u. of IG.Ad.CMV.rIL-3 without an isolated liver perfusion.

[0152] The tumor size development of the treated rats is followed in time. At the end, organ pathology is performed.

Example 17

Effect of a Longer Perfusion Time with 2.times.10.sup.8 i.u. on the Antitumor Effect of an Adenoviral Vector Carrying the Interleukin-3 Gene

[0153] ROS-1 osteosarcoma-bearing rats are treated via ILP (as described in section 3.2) for 30 Minutes with 2.times.10.sup.8 i.u. of IG.Ad5.CLIP.hIL-3. The antitumor effect is determined by daily tumor size measurement.

Example 18

Effect of an Intravenous or Intraarterial Injection of 1.times.10.sup.9 i.u. of an Adenoviral Vector Carrying the Interleukin-3 Gene on the Organ Pathology

[0154] BN rats are injected intravenously or intraarterially with 1.times.10.sup.9 i.u. IG.Ad.CMV.rIL-3. Health and behavior of the animals are monitored daily. The organ pathology is studied at day 0, 3, 7, 14, 28 after injection.

Example 19

Phase I Study Synopsis in Patients

[0155] 2.1. Compound: IG.Ad5.CLIP.hIL-3

[0156] 2.2. Study title: Isolated limb perfusion (ILP) with an adenoviral vector containing the IL-3 gene in patients with extremity sarcoma or melanoma, a dose escalation study.

[0157] 2.3. Development phase: Phase I-II

[0158] 2.4. Centers and countries: NL, CU, Other TED

[0159] 2.5. Study Objectives:

[0160] Primary: determine the regional and the systemic tolerability of escalating doses up to 1.times.10.sup.11 particles of IG.Ad5.CLIP.hIL-3 administered in conjunction with ILP to patients with extremity melanoma or sarcoma, assessed by clinical and laboratory parameters.

[0161] Secondary: determine the biological activity of IG.Ad5.CLIP.hIL-3 after ILP-assessing clinical, radiological and laboratory parameters, determine pathological and clinical tumor response, assessed by histology and clinical parameters.

[0162] 2.6. Design: Prospective, open label, dose escalation, multicenter study.

[0163] 2.7. Patients: Key inclusion criteria: age 18-80 years, failure of standard treatment of sarcoma or melanoma, measurable disease, ability to give informed consent, any other anticancer therapy completed at least 4 weeks prior to study entry, fertile patients willing to practice contraception during 3 months following the gene therapy.

[0164] Key exclusion criteria: any active or recent (within 7 days) infection, previous gene therapy of any kind, hematological disorder, autoimmune disease, plans for any additional anticancer therapy within 4 weeks after IG.Ad5.CLIP.hIL-3.

[0165] 2.8. Sample size: Sequential cohorts of 3 patients will be entered at each of the planned dose levels. At the dose where .gtoreq.2/3 patients show a complete response (CR) or a partial response (PR), additional patients will be treated up to a total sample size of 24 of that level. If the MTD or the highest planned dose level is reached and <2/3 patients at a given dose level show a CR or PR, then additional patients, to a sample size of 5, will be treated at the MTD or the highest level. If .gtoreq.2/5 patients show CR or PR, the sample size is expanded to a total of 24 at that dose; if <2/5 show PR or CR, the study is closed.

[0166] 2.9. Dose level of IG.Ad5.CLIP.hIL-3: 1.times.10.sup.8, 5.times.10.sup.8, 1.times.10.sup.9, 5.times.10.sup.9, 1.times.10.sup.10, 5.times.10.sup.11 and 1.times.10.sup.11 vector particles added to the perfusate used for ILP and perfused for 90 minutes.

[0167] 2.10. Dose escalation plan: Dose escalation will proceed until 1.times.10.sup.11 particles; if, however, at any dose level grade 3 or 4 (severe) systemic and/or .gtoreq.grade 4 local adverse events judged to be probably or definitely related to IG.Ad5.CLIP.hIL-3 occur in .gtoreq.2/3 patients in the 7 days following the ILP, then the MTD of IG.Ad5.CLIP.hIL-3 shall be defined as the dose below the one where these adverse events occurred.

[0168] 2.11. Safety criteria: Physical examination, vital signs, laboratory evaluation, adverse events and concomitant medication usage will assess safety and tolerability. Regional toxicity in the affected limbs will be graded according to Wieberdink.

[0169] 2.12. Efficacy criteria: Tumor response will be established at month 3 after the ILP. Complete response (CR) is defined as the disappearance of all evidence of disease with no new areas of diseases appearing within the perfusion field. Partial response (PR) is defined as a greater than 50% decrease in the sum of the perpendicular diameters of all measurable lesions with no single lesion increasing in the size and no new lesions appearing in the perfusion field. No change (NC) is defined as regression of less than 50% of the sum of diameters or progression of less than 25% and progressive disease (PD) is defined as greater than 25% increase in the sum of the diameters. Disease outside the perfused limb will also be measured and assessed according to standard WHO criteria.

[0170] 2.13. Follow-up: Patients will have regular clinical and laboratory examinations during 3 months following the gene therapy and will then be followed for survival only lifelong.

REFERENCES

[0171] 1. Esandi, M. d. C., van Someren, G. G., Vincent, A. J. P. E., van Bekkum, D. W., Valerio, D., Bout, A., Noteboom, J. L. (1997). Treatment of malignant mesothelioma in an immunocompetent rat model using a recombinant adenovirus expressing the HSV-tk gene. Gene Therapy 4: 280-287.

[0172] 2. Vincent, A. J. P. E., Esandi, M. d. C., van Someren, G. D., Noteboom, J. L., Vecht, C. J. J. C., Smitt, P. A. E. S., van Bekkum, D. W., Valerio, D., Hoogerbrugge, P. M., Bout, A. (1996a). Treatment of Leptomeningeal metastasis in a rat model using a recombinant adenovirus containing the HSV-tk gene. J. Neurosurgery 85: 648-654.

[0173] 3. Vincent, A. J. P. E., Vogels, R., van Someren, G., Esandi, M. d. C., Noteboom, J. L., Avezaat, C. J. J., Vecht, V. C., van Bekkum, D. W., Valerio, D., Bout, A., Hoogerbrugge, P. M. (1996b). Herpes Simplex Virus Thymidine kinase gene therapy for rat malignant brain tumors. Human Gene Therapy 7: 197-205.

[0174] 4. Manusama, E. R., Nooijen, P. T. G. A., Stavast, J., Durante, N. M. C., Marquet, R. L., Eggermont, A. M. M. (1996). Synergistic antitumor effect of recombinant human tumor necrosis factor a with melphalan in isolated limb perfusion in the rat. British Journal of Surgery 83: 511-555.

[0175] 5. Benckhuijsen, C., van Dijk, W. J., van't Hoff, S. C. (1982). High-flow isolation perfusion of the rat hind limb in vivo. Journal of Surgical Oncology 21: 249-257.

[0176] 6. Fortunati, E., Bout, A., Zanta, M. A., Valerio, D., Scarpa, N. (1996). In vitro and in vivo gene transfer to pulmonary cells mediated by cationic liposomes. Biochim. Biophys. Acta. in press.

[0177] 7. Bout, A., Perricaudet, N., Baskin, G., Imler, J. L., Scholte, B. J., Pavirani, A., Valerio, D. (1994). Lung gene therapy: in vivo adenovirus mediated gene transfer to rhesus monkey airway epithelum. Human Gene Therapy 5: 3-10.

[0178] 8. Barendsen, G. W., Janse, H. C. Differences in effectiveness of combined treatments with ionizing radiation and vinblastine, evaluation for experimental sarcomas and squamous cell carcinomas in rats. Int. J. Radiat. Oncol. Biol. Phys (1987) 4: 95-102.

[0179] 9. Marquet, R. L., Schellekens, H., Westbroek, D. L., Jeekel, J. Effect of treatment with interferon and cyclophosphamide on the growth of a spontaneous liposarcoma in rats. Int. J. Cancer (1983) 31: 323-226.

[0180] 10. Kort, W. J., Zondervan, P. E., Hulsman L. O., Weijma, I. M., Westbroek, D. L. Incidence of spontaneous tumors in a group of retired breeder female brown norway rats. J. Natl. Cancer Inst (1984) 72: 709-713.

[0181] 11. Marinelli, A. W. K. S., Van de Velde, C. J. H., Kuppen P. J. K., Franken, H. C. M., Souverij, J. H. M., Eggermont, A. M. M. A comparative study of isolated liver perfusion versus hepatic artery infusion with Mitomycin C in rats. Br. J. Cancer (1990) 62, 891-896.

Sequence CWU 1

1

10 1 27 DNA Unknown Organism Description of Unknown OrganismPRIMER SV40-1 1 gggggatccg aacttgttta ttgcagc 27 2 25 DNA Unknown Organism Description of Unknown OrganismPRIMER SV40-2 2 gggagatcta gacatgataa gatac 25 3 27 DNA Unknown Organism Description of Unknown OrganismPRIMER AD5-1 3 gggagatctg tactgaaatg tgtgggc 27 4 24 DNA Unknown Organism Description of Unknown OrganismPRIMER AD5-2 4 ggaggctgca gtctccaacg gcgt 24 5 47 DNA Unknown Organism Description of Unknown OrganismPRIMER LTR-1 5 ctgtacgtac cagtgcactg gcctaggcat ggaaaaatac ataactg 47 6 64 DNA Unknown Organism Description of Unknown OrganismPRIMER LTR-2 6 gcggatcctt cgaaccatgg taagcttggt accgctagcg ttaaccgggc gactcagtca 60 atcg 64 7 28 DNA Unknown Organism Description of Unknown OrganismPRIMER HSA1 7 gcgccaccat gggcagagcg atggtggc 28 8 50 DNA Unknown Organism Description of Unknown OrganismPRIMER HSA2 8 gttagatcta agcttgtcga catcgatcta ctaacagtag agatgtagaa 50 9 34 DNA Unknown Organism Description of Unknown OrganismPRIMER huIL3-forward 9 ccccaagctt gccaccatga gccgcctgcc cgtc 34 10 26 DNA Unknown Organism Description of Unknown OrganismPRIMER hu-IL3-reverse 10 gcgggatcct caaaagatcg cgaggc 26

Claims

What is claimed is:

1. A localized perfusion system for treating sarcoma and osteosarcoma tumors comprising a pharmaceutical composition having IL-3 activity and a pharmaceutical composition having cytostatic activity.

2. The localized perfusion system of claim 1, wherein the IL-3 activity is provided by a recombinant adenoviral vector encoding said activity.

3. The localized perfusion system of claim 1, wherein the pharmaceutical composition having cytostatic activity is in a single dosage unit for injection into a solid tumor.

4. The localized perfusion system of claim 1, wherein the pharmaceutical composition comprising IL-3 activity is a perfusion fluid.

5. The localized perfusion system of claim 1, wherein the perfusion fluid comprises a recombinant adenoviral vector in the form of a virus-like particle.

6. The localized perfusion system of claim 5, wherein the virus-like particle is present in an amount of from about 1.times.10.sup.6 to 5.times.10.sup.9 iu.

7. The localized perfusion system of claim 5, wherein both activities are present in a single combination.

8. The localized perfusion system of claim 1, wherein the cytostatic activity is at least one of TNF-activity, Melphalan and adriamycin.

9. A pharmaceutical composition for localized treatment of sarcoma and osteosarcoma tumors, said pharmaceutical composition consisting of: a recombinant adenoviral vector comprising genetic information encoding IL-3 activity; and a diluent.

10. The pharmaceutical composition of claim 9, wherein the recombinant adenoviral vector is in the form of virus-like particles

11. The pharmaceutical composition of claim 10, wherein the virus-like particles are present in said pharmaceutical composition an amount of about 1.times.10.sup.6 to 5.times.10.sup.9 in.

12. The pharmaceutical composition of claim 9, wherein a virus-like particle is a human homolog of a recombinant adenovirus IG.Ad.CMV.rIL-3 deposited at the ECACC under accession number V96071634.

13. A kit of parts for the reduction of growth of sarcoma and osteosarcoma tumors, said kit of parts comprising a pharmaceutical composition having IL-3 activity and cytostatic activity; means for isolating certain tissues; and means for perfunding said isolated tissues.

14. The kit of parts of claim 13, wherein said cytostatic activity is TNF-activity.

15. The kit of parts of claim 13, wherein the pharmaceutical composition having cytostatic activity is a single dosage unit for injection into a solid tumor.

16. The kit of parts of claim 13, wherein the pharmaceutical composition comprises at least one of Melphalan and adriamycin.