WO2001003745A2 - Therapy and diagnosis of colorectal cancer with radiopharmaceuticals - Google Patents

Abstract

Methods and composition for therapy, prevention of recurrence, and diagnosis of colorectal cancer with radiopharmaceuticals. A prodrug is administered of a desired active cytotoxic pharmaceutical, specifically a radiopharmaceutical. The prodrug is converted to the active pharmaceutical at the desired site of the malignancy or potential malignancy. For example, glycoside derivative prodrugs of 5-[125]iodo-2'-deoxyuridine are converted to the desired active form in the lumen of the intestine by the glycosidases produced by intestinal bacterial flora.

Description

TITLE: THERAPY, PREVENTION OF RECURRENCE, AND

DIAGNOSIS OF COLORECTAL CANCER WITH RADIOPHARMACEUTICALS

CROSS-REFERENCE TO RELATED APPLICATION

This application claims priority from U.S. Patent Application Serial No. 60/143,232 filed July 9, 1999.

BACKGROUND OF THE INVENTION Colon cancer, the second most common adult malignancy in the

United States, remains essentially incurable in 50 to 60% of cases. The effective treatment of patients not responding to currently available therapies continues to be a major clinical problem.

The majority of colorectal tumors originate from the mucosa. Shamsuddin, A.K.M., Weiss, L., Phelps, P.C., Trump, B.F. Colon epithelium. IV. Human colon carcino gene sis. Changes in human colon mucosa adjacent to and remote from carcinomas of the colon. J. Natl. Cancer ist. 1981; 66:413-419; Bleiberg, H. Natural history of the colorectal cancer. Bull. Cancer. 1993; 80:1051-1057; Robins, S.L. In: Robins, S.L. and Cotran, R.S., Eds. Pathologic basis of disease. 2nd Edition, Philadelphia, W.B. Saunders, 1979; pp. 946-952. The cells of the normal mucosal epithelium of the human gastrointestional (GI) tract are continuously replaced by cells that migrate from the deep portion of the crypts to the surface. As these cells differentiate and mature, DNA synthesis and cell division cease. In contrast, the DNA synthesis pattern in colon cancer is characterized by displacement of proliferating cells onto the lumenal surface of the crypts. This shift of the proliferating cell zone represents the earliest evidence of abnormality in adenomatous polyps, the putative precursors of adenocarcinoma. Shpitz, B., Bomstein, Y., Mekori, Y., Cohen, R., Kaufman, Z., Neufeld, D., Galkin, M., Bernheim, J. Aberrant crypt foci in human colons: distribution and histomorphologic characteristics. Hum. Pathol. 1998; 29:469-75; Polyak, K., Hamilton, S.R., Vogelstein, B., Kinzler, K.W. Early alteration of cell-cycle-regulated gene expression in colorectal neoplasia. Am. J. Pathol. 1996; 149: 381-7; Shpitz, B., Bomstein, Y., Mekori, Y., Cohen, R., Kaufman, Z., Grankin, M., Bernheim, J. Proliferating cell nuclear antigen as a marker of cell kinetics in aberrant crypt foci, hyperplastic polyps, adenomas, and adenocarcinomas of the human colon. Am. J. Surg. 1997; 174:425-30; Melhem, M.F., Meisler, A.I., Finley, G.G., Bryce, W.H., Jones, M.O., Tribby, I.I., Pipas, J.M., Koski, R.A. Distribution of cells expressing myc proteins in human colorectal epithelium, polyps, and malignant tumors. Cancer Res. 1992; 52: 5853-64. Deschner, E.E., Raicht, R.F. Kinetic and morphologic alterations in the colon of a patient with multiple polyposis. Cancer. 1981; 47:2440-5. Malignant colorectal neoplasms manifest a further transition of the major site of DNA synthesis from the lower third of the crypts to the upper regions and a development of crypts with hyperproliferative activities. High labeling indices (LI), a faster migration of S-phase cells to the lumenal surface, and elevated proliferative activities are all found in colorectal cancer biopsies. Balzi, M., Becciolini, A., Mauri, P., Larosa, V., Bechi, P. Proliferative activity in normal colon mucosa and tumor tissue: clinical implications. In Vivo. 1993; 7:635-637; Bleiberg, H., Galand, P. In vitro autoradiographic determination of cell kinetic parameters in adenocarcinomas and adjacent healthy mucosa of the human colon and rectum. Cancer Res. 1976; 36:325-8; Maskens, A.P., Deschner, E.E. Tritiated thymidine incorporation into epithelial cells of normal- appearing colorectal mucosa of cancer patients. J. Natl. Cancer Inst. 1977; 58:1222-4. Rew, D.A., Wilson, G.D., Taylor, I., Weaver, P.C. Proliferation characteristics of human colorectal carcinomas measured in vivo. Br. J. Surg. 1991; 78:60-6. The S-phase duration in adenocarcinomas is significantly prolonged when compared to the S-phase duration in the normal, unaffected mucosa of colon cancer patients. Shimomatsuya, T., Tanigawa, N., Muraoka, R. Proliferative activity of human tumors: assessment using bromodeoxyuridine and flow cytometry. Jpn. J. Cancer Res. 1991; 82:357-62; Bleiberg, H., Buyse, M., Galand, P. Cell kinetic indicators of premalignant stages of colorectal cancer. Cancer. 1985; 56:124- 9. The presence of hyperactive crypts is implicated in the increase of the LI in colon cancer patients and in populations at high risk of developing colorectal carcinoma. De Franceschi, L., Conti, L., Moro, L., Francavilla, V., Valente, F., Casole, P.M., Grassi, A., Casale, V., Gandolfo, G.M. Study of cell kinetics in the normal colorectal mucosa of subjects at risk of developing colorectal carcinoma. Oncology. 1997; 54:129-33; Bourry, J., Gioanni, J., Ettore, F., Giacomini, M.A., Simon, J.M., Courdi, A. Labeling index and labeling distribution in the colonic crypts: a contribution to definition of patients at high risk for colorectal cancer. Biomed. Pharmacother. 1987; 41:151-5.

Such altered cell kinetics and the increased salvage and de noυo synthetic activities for pyrimidines in colorectal cancer are compatible with therapeutic strategies involving cell-cycle-dependent agents such as 5-iodo- 2'-deoxyuridine (IUdR). Target-specific prodrugs of the cell-cycle-dependent agents, such as 5-[¹²⁵I]iodo-2'-deoxyuridine, may be beneficial adjuvant therapeutics.

The characteristics of colorectal cancer are compatible with therapeutic strategies involving cell-cycle-dependent agents such as IUdR. IUdR is incorporated by cells only during DNA synthesis. It follows metabolic pathways of thymidine and can replace thymidine as a substrate. When DNA-incorporated IUdR contains iodine- 125, the radioisotope decays and releases Auger- and Coster-Kroning electrons directly in the most radiosensitive target within the cell, the DNA. This causes a highly localized deposition of energy which results in the production of double strand breaks in the DNA molecule and kills the cell. Ludwikow, G., Hofer, KG., Bao, S.P., Ludwikow, F. The effect of ¹²⁵I decay at different stages of S- phase on survival, expression of micronuclei and chromosome aberrations in CHO cells. Int. J. Radiat. Biol. 1996; 70:177-87; Chadwick, K.H., Leenhouts,

H.P. On the linearity of the dose-effect relationship of DNA double strand breaks. Int. J. Radiat. Biol. 1994; 66:649-52; Krisch-RE; Krasin-F; Sauri-CJ. DNA breakage, repair, and lethality accompanying ¹²⁵l decay in microorganisms. Curr. Top. Radiat. Res. Q. 1978; 12:355-68. ¹²⁵IUdR appears ideally suited to the detection and treatment of cancer in the colonic epithelium. However, two events in the anabolic and catabolic processes of ¹²⁵IUdR greatly reduce its clinical efficacy: (1) degradation of ¹²⁵IUdR to 5- [¹²⁵I]iodouracil by thymidine phosphorylase in the liver and in the peripheral circulation; and (2) deiodination of ¹²⁵IUdR triphosphate to deoxyuridine triphosphate by thymidylate synthetase. Both of these processes rapidly reduce the circulating concentration of ¹²⁵IUdR (tι/2<10 min). Tjuvajev, J.G., Macapinlac, H.A., Daghighian, F., Scott, A.M., Ginos, J.Z., Finn, R.D., Kothari, P., Desai, R., Zhang, J., Beattie, B., et al. Imaging of brain proliferative activity with iodine-131-iododeoxyuridine. J. Nucl. Med. 1994; 35:1407-17; Kinsella, T.J., Collins, J., Rowland, J., Klecker, R., Jr, Wright, D., Katz, D., Steinberg, S.M., Glastein, E. Pharmacology and phase I/II study of continuous intravenous infusions of iododeoxyuridine and hyperfractionated radiotherapy in patients glioblastoma multiforme. J. Clin. Oncol. 1988; 6:871-9. Furthermore, since ¹²⁵IUdR does not possess any selectivity or specificity for tumor cells, any cell in the process of making DNA, normal or malignant, can use ¹²⁵IUdR in place of thymidine. The kinetics of normal and neoplastic colonic epithelium are completely compatible with a prodrug approach. The lumenal surface of colorectal tumors expresses the highest level of DNA synthesis. Initially, all tumor cells in microscopic adenomas and carcinomas can replicate with the growth fraction of one. Deschner, E.E., Raicht, R.F. Kinetic and morphologic alterations in the colon of a patient with multiple polyposis. Cancer. 1981; 47:2440-5; Maskens, A.P., Deschner, E.E. Tritiated thymidine incorporation into epithelial cells of normal-appearing colorectal mucosa of cancer patients. J. Natl. Cancer Inst. 1977; 58:1221-4; Bleiberg, H., Buyse, M., Galand, P. Cell kinetic indicators of premalignant stages of colorectal cancer. Cancer 1985; 56:124-9; Lieb, L.M., Lisco, H. In vitro uptake of tritiated thymidine by carcinoma of the human colon. Cancer Res. 1996; 26:733-40. Macroscopic carcinomas are also characterized by a high LI (13%-33%). Bleiberg, H., Buyse, M., Galand, P. Cell kinetic indicators of premalignant stages of colorectal cancer. Cancer 1985; 56:124-9; Deschner, E.E. In: Moyer, M.P., Poste, G.H., Eds. Colon cancer cells. New York, Academic Press, 1990; Chapter 3, Kinetics of normal, preneoplastic, and neoplastic epithelium, pp. 41-61. The distribution of S-phase cells favors the tumor periphery over the tumor center. The LI can exceed 30% at the lumenal surface, the site of ¹²⁵IUdR release, while in the tumor center the LI is about 5%. Lieb, L.M., Lisco, H. In vitro uptake of tritiated thymidine by carcinoma of the human colon. Cancer Res. 1966; 26:733-40. It is this difference in cell proliferation patterns that is exploited in the current prodrug approach to therapy of colorectal tumors. The delivery of prodrugs to tumors from the mucosal surface rather than the blood stream targets ¹²⁵IUdR to the population of malignant cells actively in DNA synthesis. This route of administration also averts the problem of systemic degradation and lowers the radiation burden on the rest of the body. The present invention addresses the deficiencies of ¹²⁵IUdR by converting it into prodrugs whose activation sites are restricted to the lumenal surfaces of colorectal tumors. This site-targeted delivery is based on -β-D-glycoside derivatives which release the active drug in the presence of glycosidases produced by the bacterial flora of the large intestine. Haeberlin, B., Friend, D.R. In: Friend, D.R., Ed. Oral-colon specific drug delivery. CRC Press, Boca Raton, Florida. 1992; pp. 1-35; Hawksworth, G., Drasar, D.S., Hill, M.J. J. Med. Microbiol. 1971; 4:451-7; Friend, D.R.,

Chang, G.W. A colon-specific drug delivery system based on drug glycosides and the glycosidases of colonic bacteria. J. Med. Chem. 1984; 27:261-266; Lieb, L.M., Lisco, H. In vitro uptake of tritiated thymidine by carcinoma of the human colon. Cancer Res. 1966; 26: 733-40.. ¹²⁵IUdR is not released until enzymatic hydrolysis of the glycosidic bond has been completed.

Consequently, ¹²⁵IUdR reaches high concentrations in the lumen of the large intestine that is the site of cancer. Such a mechanism of the ¹²⁵IUdR liberation allows oral or rectal routes of administration.

SUMMARY OF THE INVENTION An object of the invention is to establish a new and effective treatment of colorectal cancer.

Another object of the invention is to provide a prodrug approach to treatment, prevention, and/or diagnosis of malignancy by administering ¹²⁵IUdR in the form of a glycoside derivative which allows the radioactivity to be targeted to malignant cells.

A further object of the invention to provide a method for presenting radiopharmaceuticals to targeted sites that lowers the radiation burden on the body as compared to systemic delivery.

These and other objects, features, and advantages will become apparent after review of the following description and claims of the invention which follow.

The present invention provides controlled delivery of 5-iodo-2'- deoxyuridine (IUdR) prodrugs to specific sites of the gastrointestinal (GI) tract for radiotherapy of colon cancer and metastatic liver disease. By virtue of their chemical structure and microbial-controlled release of the parent IUdR, the IUdR prodrugs and their unconventional routes of administration offer advantages such as (1) superior targeting of the diseased site; (2) sustained concentrations of the drug at the targeted site; and (3) reduction of the administered dose. IUdR is incorporated into the DNA during its synthesis in place of thymidine. IUdR is an effective radiosensitizing and radiotherapeutic agent, but its clinical usefulness is limited because of the rapid degradation and nonspecific uptake by any cell in the S-phase or during the unscheduled DNA synthesis. The present invention's prodrugs and methods address problems related to nonspecificity and rapid systemic degradation. The site-targeted delivery system is based on glycoside, specifically -β-D-glycoside, derivatives of IUdR which release the active drug in the presence of bacterial glycosidases in the large intestine. Friend, D.R., Chang, G.W. A colon-specific drug delivery system based on drug glycosides and the glycosidases of colonic bacteria. J. Med. Chem. 1984; 27:261-266; Lieb, L.M., Lisco, H. In vitro uptake of tritiated thymidine by carcinoma of the human colon. Cancer Res. 1966; 26:733-40. The active form of the drug, i.e., IUdR, is not released until enzymatic hydrolysis of the glycosidic bond is completed. Consequently, ¹²⁵IUdR reaches high concentrations in the lumen of the large intestine, i.e., the site of the tumor. This mechanism of IUdR liberation allows oral or rectal administration of prodrugs. Additionally, the problem of hepatic degradation is averted because IUdR is delivered to the tumor from the lumenal surface rather than the blood stream, where dehalogenation is so rapid. As a result, targeting of the colon is achieved.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 shows chemical structures of ¹²⁵IUdR glycoside prodrugs that were tested in vitro in normal mice and in mice bearing carcinogen- induced tumors.

Figure 2 is a graph of in vitro degradation of ¹²⁵IUdR-5'-β-D- galactoside (closed circles) and ¹²⁵IUdR-5'-β-D-glucoside in the presence of β- D-glucosidase (open circles) and α-D-glucosidase (open triangles). Each prodrug was incubated (n=3) with indicated amounts of enzyme at 37°C and reaction mixtures analyzed on silica gel TLC plates and Cie reversed phase or carbohydrate (85°C) HPLC columns. Figure 3 is a graph of ex vivo release of ¹²⁵IUdR from prodrugs in homogenates of the content of proximal small intestine (PSI), distal small intestine (DSI), and caecum. Shaded bars represent data obtained for ¹²⁵IUdR-5'-β-D-galactoside; open bars are for ¹²⁵IUdR-5'-β-D-cellobioside (similar results were observed for ¹²⁵IUdR-5'-β-D-glucoside). Each prodrug was incubated at 37°C for 60 min. in contents collected from three mice; bars represent means of two experiments (standard deviations indicated as a line above the bar).

Figure 4 is a graph of blood clearance curves after oral (left panel) and i.v. (right panel) administration of ¹²⁵IUdR (closed circles) and ¹²⁵IUdR- 5'-β-D-glycoside (open circles). Each point is a mean of two experiments (n=8). Bars represent standard deviations.

Figure 5 is a graph of uptake of radioiodine in the thyroid after oral and i.v. administration of ¹²⁵IUdR (closed circles) or ¹²⁵IUdR-5'-β-D-glycoside (open circles). Each point is a mean of two experiments (n=8). Bars represent standard deviations.

Figure 6 is planar images of mice after oral administration of ¹²⁵IUdR-5'-β-D-galactoside (80 μCi) and ¹²⁵IUdR (70 μCi). Images acquried 90 min. after administration. Note the prominent uptake of radioiodine in the stomach, thyroid, and bladder of a mouse treated with ¹²⁵IUdR. In an animal receiving a prodrug, the only prominent feature are stomach and large bowel.

Figure 7 is a graph of stomach uptake of iodine- 125 after oral administration of either the parent drug ¹²⁵IUdR or its prodrugs ¹²⁵IUdR- Glu. The accumulation of radioactivity in the stomach of ¹²⁵IUdR-treated mice is indicative of rapid degradation of ¹²⁵IUdR and reabsorption of free iodide- 125 to the stomach. On the other hand, the ¹²⁵IUdR-Glu (prodrug) leaves the stomach with the residence time typical of the transit time in mice and does not undergo deiodination, hence there is no uptake in the stomach. One of the main problems with using ¹²⁵IUdR in therapy is the rapid dehalogenation and a half- life in circulation of less than 10 min. The intact prodrugs do not seem to escape the lumen of the GI tract.

Figure 8 is a graph of uptake of ¹²⁵IUdR after a single oral dose of prodrugs (5 μCi each, glucoside and galactoside; shaded bars) or "free" ¹²⁵IUdR (10 μCi; open bars) in mice treated with ten weekly doses of 1,2- dimethylhydrazine and in control mice. Necropsy preformed 24 h after treatment (n=5); distal colon includes rectum. Figure 9 is a graph of changes in weights of mice in different treatment groups. The weight is a good indicator of animal well-being. Note a very slow growth and a dramatic decline in weights of mice treated with 1,2-DMH after about 14 weeks since the first 1,2-DMH dose. The weight of mice treated with ¹²⁵IUdR-Glu and ¹²⁵IUdR-Gal remains at the levels of non-treated controls. Also, mice receiving only radiopharmaceuticals without the carcinogen treatment are maintaining the normal weight changes indicating the lack of any radiation-related problems. Figure 10 is a graph of the proliferative index in selected tissues in mice treated with 1,2-DMH only once a week for 10 weeks and 1,2-DMH once a week for 10 weeks with the therapeutic doses of ¹²⁵IUdR-Glu and ¹²⁵IUdR-Gal (PO; 4 weekly doses 0.005 mCi each per dose). Proliferative index was measured using a thymidine analog, ¹³¹IUdR, and calculated as a ratio of the tissue uptake of ¹³¹IUdR (i.v. administration) in mice receiving indicated above treatments to the ¹³¹IUdR tissue uptake in normal, non- treated controls. Proliferative index of 1 indicates that there is no difference between normal mice, not treated with either the carcinogen or the prodrug, and the treated animals. Figure 11 shows two sections of large intestine of mice treated with

10 weekly doses of 1,2-DMH alone (left); four oral doses of ¹²⁵IUdR plus 10 weekly doses of 1,2-DMH (center), and four oral doses of a mixture of ¹²⁵IUdR-Glu and ¹²⁵IUdR-Glu, plus 10 weekly doses of 1,2-DMH (right). Note the normal appearance of the large intestine of the prodrug-treated mice and multiple 1,2-DMH-induced tumors in control mice.

Figure 12 shows a graph of uptake of ¹²⁵IUdR from orally administered prodrugs or the parent drug during the carcinogen treatment at 4 hours (gray bars) and 24 hours (black bars) after oral administration of the drugs. The uptake is calculated as a ratio of the ¹²⁵I uptake in livers of 1,2-DMH-treated animals to the uptake in normal mice. The relative uptake of 1 indicates no preferential accumulation of the radioactivity from the tested drug. It appears that ¹²⁵IUdR-Glu and ¹²⁵IUdR-Gal, unlike the parent ¹²⁵IUdR, are able to deliver the radioactivity to abnormally proliferating liver.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

In order to address problems of the prior art, a prodrug approach to site-targeted delivery of an active radiopharmaceutical was developed. A goal of the invention was development of novel adjuvant therapy protocols for the treatment and detection of microscopic disease and prevention of recurrence of colorectal cancer. Cytotoxic agents specifically taken up by colon cancer cells in the S-phase, i.e., during the DNA synthesis, should be beneficial as an adjuvant in a therapy regimen for colon cancer, and 5'-β-D- glycosides of ¹²⁵IUdR will constitute the basis of such adjuvant therapies. The kinetics of normal and neoplastic colonic epithelium are completely compatible with a prodrug approach. The lumenal surface of colorectal tumors expresses the highest level of DNA synthesis. Initially, all tumor cells in microscopic adenomas and carcinomas can replicate with the growth fraction of one. Deschner, E.E., Raicht, R.F. Kinetic and morphologic alterations in the colon of a patient with multiple polyposis. Cancer. 1981; 47:2440-5; Maskens, A.P., Deschner, E.E. Tritiated thymidine incorporation into epithelial cells of normal-appearing colorectal mucosa of cancer patients. J. Natl. Cancer Inst. 1977; 58:1221-4; Bleiberg, H., Buyse, M., Galand, P. Cell kinetic indicators of premalignant stages of colorectal cancer. Cancer 1985; 56:124-9; Lieb, L.M., Lisco, H. In vitro uptake of tritiated thymidine by carcinoma of the human colon. Cancer Res. 1996;

26:733-40. Macroscopic carcinomas are also characterized by a high LI (13%-33%). Bleiberg, H., Buyse, M., Galand, P. Cell kinetic indicators of premalignant stages of colorectal cancer. Cancer 1985; 56:124-9; Deschner, E.E. In: Moyer, M.P., Poste, G.H., Eds. Colon cancer cells. New York. Academic Press, 1990; Chapter 3, Kinetics of normal, preneoplastic and neoplastic epithelium, pp. 41-61. The distribution of S-phase cells favors the tumor periphery over the tumor center. The LI can exceed 30% at the lumenal surface, the site of ¹²⁵IUdR release, while in the tumor center the LI is about 5%. Lieb, L.M., Lisco, H. In vitro uptake of tritiated thymidine by carcinoma of the human colon. Cancer Res. 1966; 26:733-40. It is this difference in cell proliferation patterns that we exploit in the prodrug approach to therapy of colorectal tumors. The delivery of prodrugs to tumors from the mucosal surface rather than the blood stream targets ¹²⁵IUdR to the population of malignant cells actively in DNA synthesis, averts the problem of systemic degradation, and lowers the radiation burden on the rest of the body.

Two -β-D-glycoside-based prodrugs of ¹²⁵IUdR were tested in carcinogen-induced colorectal cancer in mice. Criteria for evaluation of the treatment efficacy included the number, size, and location of gastrointestinal (GI) lesions. Orally administered 5'-β-D-glucopyranoside and 5'-β-D-galactopyranoside of ¹²⁵IUdR completely halted the development of tumors. Mice treated with these prodrugs remained healthy for the duration of the experiment. All control mice developed tumors of the GI tract.

The targeting of ¹²⁵IUdR to the colon and its controlled release into the lumen of the large intestine is determined by the interaction of 5'-β-D- glycosides with bacterial glycosidases. -β-D-Glucosidase and -β-D- galactosidase are primary glycosidases produced by bacteria found in the gut of humans and other mammals. Friend, D.R., Chang, G.W. A colon- specific drug delivery system based on drug glycosides and the glycosidases of colonic bacteria. J. Med. Chem. 1984; 27:261-266; Hawksworth, G.,

Drasar, D.S., Hill, M.J. Intestinal bacteria and the hydrolysis of glycosidic bonds. J. Med. Microbiol. 1971; 4:451-7. Two substrates for these enzymes, ¹²⁵IUdR-glucopyranoside (¹²⁵IUdR-Glu) and ¹²⁵IUdR-galactopyranoside (¹²⁵IUdR-Gal) were synthesized and evaluated (Figure 1). The ability of the colonic microflora to release intact ¹²⁵IUdR was tested in vitro and in Swiss

Webster mice. In vitro, both prodrugs released ¹²⁵IUdR in the presence of either pure -D-glycosidases or in homogenates of the murine GI content. Unlike ¹²⁵IUdR, glycosidic prodrugs of ¹²⁵IUdR were stable in the gastric contents. Baranowska-Kortylewicz, J., Kortylewicz, Z.P., Hoffman, D., Dalrymple, G.V. Prodrugs in a site-selective delivery of radiopharmaceuticals. Q. J. Nucl. Med. 1997; 41:127-139; Baranowska- Kortylewicz, J., Kortylewicz, Z.P., Hoffman, D., Winoto, A., Lai, J., Dalrymple, G.V. Colon-specific prodrugs of 5-radioiodo-2'-deoxyuridine. Acta Oncologica. 1996; 35:959-964. The efflux of prodrugs from the GI tract after oral administration in mice was minimal because of the hydrophilic nature of glycosides. Baranowska-Kortylewicz, J., Kortylewicz, Z.P., Hoffman, D., Dalrymple, G.V. Prodrugs in a site-selective delivery of radiopharmaceuticals. Q. J. Nucl. Med. 1997; 41:127-139; Baranowska- Kortylewicz, J., Kortylewicz, Z.P., Hoffman, D., Winoto, A., Lai, J., Dalrymple, G.V. Colon-specific prodrugs of 5-radioiodo-2'-deoxyuridine. Acta Oncologica. 1996; 35:959-964; Sayler, A.A., Leedle, J.A.Z. In: Henteges, D.J., Ed. Human Intestinal microβora in health and disease. Academic Press, New York, 1983; pp 129-140; Haeberlin, B., Friend, D.R. In: Friend, D.R., Ed. Oral Colon-specific drug delivery. CRC Press, Boca Raton, Florida, 1992; pp. 1-35. In the case of prodrugs, the total systemic distribution of radioactivity was less than 2% of the administered dose at the time point corresponding to the steady state levels in blood. Baranowska-Kortylewicz, J., Kortylewicz, Z.P., Hoffman, D., Winoto, A., Lai, J., Dalrymple, G.V. Colon-specific prodrugs of 5-radioiodo-2'-deoxyuridine. Acta Oncologica. 1996; 35:959-964. At these concentrations, even assuming that all released radioisotope is in the form of ¹²⁵IUdR, its uptake in tissues that may have been engaged in DNA synthesis is expected to be negligible. The DNA incorporation of ¹²⁵I is directly proportional to the extracellular concentration of ¹²⁵IUdR. Narra, V.R., Howell, R.W., Thanki, K.H., Rao-DV. Radiotoxicity of ¹²⁵I-iododeoxyuridine in pre-implantation mouse embryos Int. J. Radiat. Biol. 1991; 60:525-32; Makrigiorgos, G., Kassis, A.I.,

Baranowska-Kortylewicz, J., McElvany, K.D., Welch, M.J., Sastry, K.S.R., Adelstein, S.J. Radiotoxicity of 5-[¹²³I]iodo-2'-deoxyuridine in V79 cells: A comparison with 5-[¹²⁵I]iodo-2'-deoxyuridine. Radiat. Res. 1989; 118:532- 539. The prodrugs and metabolites containing radioiodine (the intact prodrug, iodouracil, and free iodide) are either rapidly excreted or are accumulated by the thyroid. Biodistribution studies did not point to any specific tissue other than the GI tract, as the targeted site. The uptake in all tested tissues was minimal and paralleled radioactivity levels in blood. Baranowska-Kortylewicz, J., Kortylewicz, Z.P., Hoffman, D., Winoto, A., Lai, J., Dalrymple, G.V. Colon-specific prodrugs of 5-radioiodo-2'- deoxyuridine. Acta Oncologica. 1996; 35:959-964. Stated another way, the anticipated targeting of the GI tract was accomplished with high effectiveness using glycosidic prodrugs of ¹²⁵IUdR.

These results indicate that the use of prodrugs in the site-directed delivery of radiopharmaceuticals is valid and provides a distinct pharmacologic advantage over unmodified ¹²⁵IUdR. The oral drugs of

¹²⁵IUdR may be useful in the treatment of adenomatous polyps, the putative precursor of adenocarcinoma, and may prevent progression to colorectal cancer. The total arrest of tumor development also bears on the mechanisms underlying carcinogenesis and prevention of colorectal cancer. A comparison of the present therapeutic modality with results reported for conventional therapies is difficult even when the same cancer model is employed. The total administered ¹²⁵IUdR in this study was approximately 0.1 μg/kg body weight whereas most chemotherapeutic agents are tested in mice at mg/kg body weight or more. Similar difficulties arise when the radiation doses used in conventional radiotherapy and doses from ¹²⁵I are compared. The decay of ¹²⁵I is associated with 35.5 keV gamma-emission. One mCi of ¹²⁵I deposits an estimated radiation dose of 2 x 10"² Gy in the large intestine, assuming a mean residence time of 9.5 h in the lumen of the large intestine. The tolerance doses for therapeutic external beam radiotherapy are 55 Gy for the colon and 60 Gy for the rectum. Rubin, P., Constine, L.S., Nelson, D.F. In: Perez, C.A., Brady, L.W., Eds. Principle and practice of radiation oncology. 2nd ed. J.B. Lippincott Co., Philadelphia, 1992, pp. 24-161. The total radiation dose delivered by orally administered prodrugs of 1 mCi ¹²⁵IUdR is over 2,700 times lower than the tolerance dose. The prodrugs that may be used in the present invention are any which will be active at the desired location, i.e., the tumor. The preferred prodrugs are glycoside derivatives of the active drug. The preferred active drug of the present invention is IUdR. Specifically, the prodrugs used were 5'-β-D-glucopyranoside and 5'-β-D-galactopyranoside of ¹²⁵IUdR. Other examples of glycosides are dextrans and cellobioses. A single prodrug or a mixture of prodrugs may be administered. One of ordinary skill would be able to determine prodrugs appropriate for a desired application.

It is believed the prodrugs can be administered alone or with other treatments. One of ordinary skill in the art would be able to determine which, if any, treatments would be effective in combination with the current invention.

The prodrugs of the current invention can be made in a variety of ways. For examples, a method similar to the one described by Watanabe et al. for preparation of glucuronides can be used substituting the glucuronide derivatives with appropriate O-acetyl protected bromo-glycosides (K.A.

Watanabe, A. Matsuda, M.J. Halat, D.H. Hollenberg, J.S. Nesselbaum, and J.J. Fox. Nucleosides. 114. 5'-0-Glucoronides of 5-fluorouridine and 5- flurorocytidine. Masked precursors of anticancer nucleosides. J. Med. Chem. 1981; 24:893-897). One of ordinary skill in the art will be able to make the desired prodrugs.

The present invention administers the prodrugs in any form that is effective. One of ordinary skill would be able to determine forms which will be effective. For example, in mice, the prodrug is administered as a solution in physiologic saline or phosphate buffered saline. In a clinic, we can substitute these with any juice, water, or a soft drink (much the same way the iodide-131 is given for thyroid ablation and/or imaging). The prodrugs may be mixed with standard pharmaceutical carriers.

The present invention administers the prodrugs by any route that is effective. Preferably, the prodrugs are administered to a mucosal surface, such as orally or rectally. One of ordinary skill would be able to determine other routes which will effectively administer the prodrugs.

The present invention administers an effective amount of the prodrugs. One of ordinary skill in the art would be able to determine what an effective amount of the prodrug is for a given case. The effective therapeutic amounts are determined after a diagnostic dose is administered and the uptake evaluated. Example approximate dosages of the prodrugs are:

The present invention may be used several ways. Perisurgical administration to prevent tumor recurrence is possibly the most important of clinical applications for ¹²⁵IUdR prodrugs. Local/regional tumor growth represents the most common form of relapse following curative resection of colorectal cancer. Re-implantation of viable exfoliated cells present within the colonic lumen is considered the major cause of this recurrence. Hohenberger, P. Recent Results Cancer Res. 1998; 146:127-40. Prodrugs administered prior to surgery is expected to allow incorporation of ¹²⁵IUdR into the DNA of proliferating tumor cells. This process should kill proliferating cells and subsequently reduce the number of potentially implantable viable cells.

¹²⁵IUdR prodrugs are also expected to facilitate detection and treatment of occult/microscopic colorectal cancer. High concentrations of ¹²⁵IUdR released from prodrugs at the lumenal surface will provide a sustained source of ¹²⁵IUdR to rapidly dividing cells within a microscopic disease. Detection of the radioactivity incorporated into sites with increased proliferative activities may be more advantageous than radioimmuno uided surgery where only sites with a high antigen expression are detected.

In addition to treatment of adenocarcinomas of the GI tract, the methods of the current invention are expected to work for inoperable adenomateous familial polyposis and similar conditions.

For treatment of liver conditions and the like, such as metastatic liver disease and neovasculature associated with angiogenesis process in the developing cancers, the prodrugs, radiosensitizers, can be selectively delivered to abnormally proliferating cells. Similar to the other methods of the invention the prodrugs can be dosed orally, such as in a liquid. An example oral dosage for radiosensitization is 1 g to 100 g of halogenated prodrug. Figure 12 shows an example of uptake of ¹²⁵IUdR from orally administered prodrugs or the parent drug during the carcinogen treatment at 4 hours (gray bars) and 24 hours (black bars) after oral administration of the drugs. The uptake is calculated as a ratio of the ¹²⁵I uptake in livers of 1,2-DMH-treated animals to the uptake in normal mice. It appears that ¹²⁵IUdR-Glu and ¹²⁵IUdR-Gal, unlike the parent ¹²⁵IUdR, are able to deliver the radioactivity to an abnormally proliferating liver. Our autoradiography data indicates that not only are the tumor cells radiolabeled but many of the new blood vessel cells also contain radioactivity which is probably killing the blood supply to the tumor. Based on our results, we believe that:

(1) ¹²⁵IUdR prodrugs will eradicate occult/microscopic colorectal cancer and will establish a foundation for novel adjuvant therapies of colorectal cancer. Analysis of the DNA synthesis pattern in colon cancer and related abnormalities reveals displacement of proliferating cells into the upper third, i.e., lumenal surface, of the crypt. Microscopic adenomas and adenocarcinomas have high labeling indices. Initially, all tumor cells in a small neoplastic lesions can replicate. It is this difference in cell proliferation patterns that we exploit in our prodrug approach to therapy and detection of microscopic tumors. High concentrations of ¹²⁵IUdR released from its prodrugs at the lumenal surface will provide a sustained source of this agent to dividing cells in a microscopic disease.

(2) Targeting of radiosensitizing ¹²⁷IUdR to metastatic liver disease via its oral prodrugs will lower the rate of adverse effects and reduce the administered dose needed to elicit the desired pharmacologic response. ¹²⁷IUdR is an effective radiosensitizer of liver metastases. However, to achieve IUdR levels relevant in the radiation therapy, e.g., high thymidine replacement ratios, days of continuous systemic administration are required. Our preliminary studies indicate that IUdR arrives intact into the diseased sites in the liver with a minimal systemic distribution.

We evaluated the following:

(1) Therapeutic and diagnostic potential of oral ¹²⁵IUdR prodrugs in preneoplastic and neoplastic lesions during 1,2-dimethyιhydrazine- induced colorectal carcinogenesis. (2) Measurement of levels of thymidine replacement by IUdR in liver of carcinogen-treated mice.

(3) Design, synthesis, and testing of additional prodrugs of ¹²⁵IUdR that mitigate any potential shortcomings of already available agents.

Having established the stability and enzymatic hydrolysis of ¹²⁵IUdR prodrugs in vitro and in vivo in Swiss Webster mice, we desired to study two aspects of ¹²⁵IUdR prodrugs in the management of colon cancer: (1) the role of ¹²⁵IUdR prodrugs in eradication of a microscopic disease and an established cancer and (2) the function of IUdR prodrugs in the treatment of colorectal cancer metastatic to the liver. We also planned to study the distribution, uptake and cellular localization of ¹²⁵IUdR in tumor-bearing mice to estimate radiation doses delivered to normal tissues. Finally, using the experience gained upon completion of initial experiments, we will design and synthesize, as warranted, additional derivatives of ¹²⁵IUdR with improved in vivo characteristics.

EXAMPLES

METHODS AND MATERIALS

Vertebrate Animals

Six week old female Swiss albino mice were injected subcutaneously with 1,2-DMH (10 weekly doses) to generate carcinogenesis for biodistribution and therapy studies. Additionally, nontreated controls were used in each study. All animals received oral doses of prodrugs. Animals used in biodistribution studies also were injected i.v. with ¹³¹IUdR to determine proliferative capacity of various tissues. Mice were sacrificed at different times by cervical dislocation (or CO2 asphyxiation) and blood, tumors and all the major organs (up to 16 tissues per mouse) were collected, wet-weighed using an analytical balance, and counted in a gamma scintillation counter. The percentage of the injected dose per gram (%ID/g) for each organ were determined, and relative proliferative ratios as well as the subcellular localization were measured.

The choice between a mouse or a rat model was made on purely economical grounds. The maintenance and handling of mice is less expensive and challenging. The total number of mice in this study to date was 395. The investigators were guided and assisted by the Veterinarian Director of the Animal Resource Facility and his associates regarding handling, analgesics and euthanasia for these studies. All procedures incorporated in these studies were reviewed and approved by the University of Nebraska Medical Center Animal Review Committee. The Animal Resource Facility is fully accredited by the American Association for the Accreditation of Laboratory Animal Care for the use of warm-blooded animals in research, training or other activities sponsored by grants, awards and contracts. Only momentary pain was anticipated during the subcutaneous injection of 1,2-DMH and during the intravenous injection of ¹³¹IUdR. Therefore, no pain control was planned. Animals were be sacrificed when progression of the disease is such that the rectal/anal tumors are visible and there is blood in feces to reduce potential for any pain. It is our experience that mice at this stage of cancer development do not exhibit any symptoms of being in pain.

Mice under study were euthanized by cervical dislocation and exsanguination to obtain enough blood to get accurate weights and counts and to minimize blood-borne activity in the major organs or through a C0₂ asphyxiation when the blood samples are not needed. Ketamine-xylazine at

0.09 and 0.018 mg/gm, respectively, were administered i.p. immediately prior to exsanguination. This method is one being used at the National Institutes of Health and is consistent with the recommendations of the Panel on Euthanasia of the American Veterinary Medical Association.

Carcinogen-induced colon cancer in mice

The DMH-induced colorectal cancer model in mice or rats is the only animal model available for the evaluation of orally administered radiotherapeutic agents for treatment of colorectal cancer. 1,2-DMH-induced colon cancer in mice is an ideal tumor model for therapy studies of a cell-cycle dependent agent such as IUdR. The distribution of S-phase cells within crypts in mice treated with 1,2-DMH reveals the extension of the proliferative compartment to the lumenal surface (stage I defect), a shift of the predominant zone of DNA synthesis to the upper third of the glands (stage II defect), and crypts with the labeling indices >15% (stage III defect). Bleiberg, H., Buyse, M., Galand, P. Cell kinetic indicators of premalignant stages of colorectal cancer. Cancer. 1985; 56:124-9; Deschner, E.E., Maskens, A.P. Significance of the labeling index and labeling distribution as kinetic parameters in colorectal mucosa of cancer patients and DMH treated animals. Cancer. 1982; 50:1136-41. Cell kinetics in biopsies from various groups of patents at high risk for colon cancer indicate similar three stages leading to development of neoplasia. Melhem, M.F., Meisler, A.I., Finley, G.G., Bryce, W.H., Jones, M.O., Tribby, 1.1., Pipas, J.M., Koski, R.A. Distribution of cells expressing myc proteins in human colorectal epithelium, polyps, and malignant tumors. Cancer Res. 1992; 52:5853-64. Deschner, E.E., Raicht, R.F. Kinetic and morphologic alterations in the colon of a patient with multiple polyposis. Cancer. 1981; 47:2440-5; Balzi, M., Becciolini, A., Mauri, P., Larosa, V., Bechi, P. Proliferative activity in normal colon mucosa and tumor tissue: clinical implications. In Vivo. 1993; 7:635-637; Bourry, J., Gioanni, J., Ettore, F., Giacomini, M.A., Simon, J.M., Courdi, A. Labeling index and labeling distribution in the colonic crypts: a contribution to definition of patients at high risk for colorectal cancer. Biomed. Pharmacother. 1987; 41:151-5. Mice treated with 1,2-DMH develop carcinomas of the colon that are similar histologically to human tumors. Deschner, E.E., Maskens, A.P. Significance of the labeling index and labeling distribution as kinetic parameters in colorectal mucosa of cancer patients and DMH treated animals. Cancer. 1982; 50:1136-41. There is a progressively increased crypt hyperplasia and displacement of a cell division zone with continued exposure to the carcinogen. Early changes in the dynamics of crypt cell population and labeling parameters throughout the carcinogen treatment appear to fit the classic multistep model of carcinogenesis. Two kinetic parameters are characteristic of the tumor-bearing colorectal mucosa of man and mouse: upward shift in the distribution of DNA synthesizing cells and an extremely elevated labeling index easily detectable in ³H-thymidine-labeled biopsies. Bleiberg, H., Galand, P. In vitro autoradiographic determination of cell kinetic parameters in adenocarcinomas and adjacent healthy mucosa of the human colon and rectum. Cancer Res. 1976; 36:325-8; Maskens, A.P., Deschner, E.E. Tritiated thymidine incorporation into epithelial cells of normal-appearing colorectal mucosa of cancer patients. J. Natl. Cancer Inst. 1977; 58:1222-4 (¹²⁵IUdR is a thymidine analog.) Crypts in 1,2-DMH treated mice have proliferative defects similar to those observed in the colonic mucosa of colon cancer patents. Several reports have shown a positive correlation between cell proliferation in the colonic crypts and an increased incidence or risk of colon cancer even in cases of compensatory proliferative responses to surgical procedures. Allendorf, J.D., Bessler, M., Kayton, M.L., Oesterling, S.D., Treat, M.R., Nowygrod, R., Whelan, R.L.

Increased tumor establishment and growth after laparotomy vs. laparoscopy in a murine model. Arch. Surg. 1995; 130:649-53; Eggermont, A.M., Steller, E.P., Sugarbaker, P.H. Laparotomy enhances intraperitoneal tumor growth and abrogates the antitumor effects of interleukin-2 and lymphokine- activated killer cells. Surgery. 1987; 102:71-8.

Treatment

All ¹²⁵I-labeled agents were administered orally in physiologic saline. The age-matched, female Swiss albino mice (six weeks old) were divided into six groups: five controls and one experimental group. Control mice were subjected to the following treatments: None; ¹²⁵IUdR alone (IUdR group); a mixture of ⁱ^IUdR-Glu and ¹²⁵IUdR-Gal (IUdR-Glu/IUdR-Gal group); 1,2- DMH alone (DMH group); 1,2-DMH followed by ⁱ^IUdR (IUdR/DMH group). Adhering to the identical schedule, the experimental group was injected with 1,2-DMH and received a mixture of ¹²⁵IUdR-Glu and ¹²⁵IUdR-

Gal (IUdR-Glu/IUdR-Gal/DMH group). To avoid any interference from agents other than these being tested, the traditional thyroid block with potassium iodide was not included in the protocol. Forty-eight hours after every other oral dose of radiopharmaceuticals, three mice were randomly selected from each of the control groups and five mice from the experimental group. These animals were asphyxiated with CO2, necropsy performed, and the radioactive content of selected tissues measured. After the fourth dose, the thyroid uptake of ¹²⁵I in the IUdR and IUdR/DMH groups exceeded 0.3 μCi at 48 h post-administration. At this time the treatment with orally administered radiopharmaceuticals was terminated to avoid any complications that may have resulted from impaired thyroid function. Overall, the mice received approximately 40 μCi of either ¹²⁵IUdR or a mixture of ¹²⁵IUdR-Glu and ¹²⁵IUdR-Gal. During the course of the experiments, mice were randomly selected from each group and injected i.v. via a tail vein with 10 μCi ¹³¹IUdR, a marker of proliferation, to assess the DNA synthesis in the GI tract and other selected tissues. ¹³¹IUdR was used to allow simultaneous determination of the uptake of both radioiodine isotopes. Two days later, these mice were also asphyxiated and necropsy performed. This animal protocol was approved by the University of Nebraska Medical Center Institutional Animal Care and Use Committee.

Example 1

Two substrates for the bacterial enzymes, ¹²⁵IUdR-glucopyranoside (¹²⁵IUdR-Glu) and ¹²⁵IUdR-galactopyranoside (¹²⁵IUdR-Gal) were synthesized and evaluated (Figure 1). The ability of the colonic microflora to release intact ¹²⁵IUdR was tested in vitro and in Swiss Webster mice. In vitro both prodrugs released ¹²⁵IUdR in the presence of either pure -D- glycosidases or in homogenates of the murine GI content. Unlike ¹²⁵IUdR, glysoidic prodrugs of ¹²⁵IUdR were stable in the gastric contents. Baranowska-Kortylewicz, J., Kortylewicz, Z.P., Hoffman, D., Dalrymple, G.V. G. J. Nucl. Med. 1997; 41:127-139; Baranowska-Kortylewicz, J., Kortylewicz, Z.P., Hoffman, D., Winoto, A., Lai, J., Dalrymple, G.V. Acta Oncologica. 1996; 35:959-964. The efflux of prodrugs from the GI tract after oral administration in mice was minimal because of the hydrophilic nature of glycosides. Baranowska-Kortylewicz, J., Kortylewicz, Z.P., Hoffman, D., Winoto, A., Lai, J., Dalrymple, G.V. Acta Oncologica. 1996; 35:959-964; Haeberlin, B., Friend, D.R. In: Friend, D.R., Ed. Oral Colon-specific drug delivery. CRC Press, Boca Raton, Florida, 1992; pp. 1-35. The total systemic distribution of radioactivity was less than 2% of the prodrug's dose at the time point corresponding to the steady state level in blood. Baranowska-Kortylewicz, J., Kortylewicz, Z.P., Hoffman, D., Winoto, A., Lai, J., Dalrymple, G.V. Acta Oncologica. 1996; 35:959-964. At these concentrations, even assuming that all released radioisotope is in the form of ¹²⁵IUdR, its uptake in tissues that may have been engaged in DNA synthesis is expected to be negligible because DNA incorporation of ¹²⁵I is directly proportional to the extracellular concentration of ¹²⁵IUdR.

Makrigiorgos, G., Kassis, A.I., Baranowska-Kortylewicz, J., McElvany, K.D., Welch, M.J., Sastry, K.S.R., Adelstein, S.J. Radiat. Res. 1989; 118:532-539. The prodrugs and metabolites containing radioiodine are either rapidly excreted or are accumulated by the thyroid. Biodistribution studies did not point to any specific tissue other than the GI tract, as the targeted site. The uptake in all tested tissues was minimal and paralleled radioactivity levels in blood. Baranowska-Kortylewicz, J., Kortylewicz, Z.P., Hoffman, D., Winoto, A., Lai, J., Dalrymple, G.V. Acta Oncologica. 1996; 35:959-964. The desired targeting of the GI tract was accomplished with high effectiveness by means of glycosidic prodrugs of ¹²⁵IUdR.

Example 2 In Vitro Studies

Having established the site-specific release of ¹²⁵IUdR, the glycosides were tested in a mouse model of colorectal cancer. The 1,2- dimethylhydrazine (l,2-DMH)-induced colon cancer was selected to approximate human colon carcinogenesis. The tumor-bearing colorectal mucosa of humans and 1,2-DMH-treated mice have two kinetic parameters in common: (1) the upward shift in the distribution of DNA synthesizing cells and (2) the elevated LI. Richards, T.C. Cancer Res. 1977; 37:1680-5; Deschner, E.E., Maskens, A.P. Cancer. 1982; 50:1136-41; Chang, W.W.L. J. Natl. Cancer Inst. 1978; 60:1405-18. Mice treated with 1,2-DMH develop carcinomas of the colon that are similar histologically to human tumors. Chang, W.W.L. J. Natl. Cancer Inst. 1978; 60:1405-18; Shamsuddin, A.M. In: Moyer, M.P., Poste, G.H., Eds. Colon cancer cells. Academic Press, New York, 1990; pp. 15-40. There is progressively increased crypt hyperplasia and displacement of the cell division zone with continued exposure to the carcinogen. The 1,2-DMH-induced tumors in mice appear in the greatest numbers in the distal colon in an area between the descending colon and anus. This precisely matches the sites of occurrence of colon cancer in humans.

Glycoside-based prodrugs of ¹²⁵IUdR address problems related to nonspecificity and rapid systemic degradation of ¹²⁵IUdR. This delivery system is based on glycoside derivatives of IUdR which release the active drug in the presence of bacterial glycosidases in the large intestine.

Consequently, IUdR reaches a high concentration in the lumen of the large intestine, i.e., the site of the tumor, and its incorporation into the dividing cancer cells is facilitated. The problem of hepatic degradation is averted because IUdR is delivered to the tumor from the mucosal surface rather than the blood stream, where dehalogenatiion is so rapid. As a result, targeting of the colon is achieved. This mechanism allows the prodrug to be administered orally. The oral delivery of IUdR in chemotherapy/radiosensitization trials has already been explored for some hepatotropic IUdR derivatives. Kinsella, T.J., Kunugi, K.A., Vielhuber, K.A., McCulloch, W., Liu, S.H., Cheng, Y.C. An in vivo comparison of oral 5- iodo-2'-deoxyuridine and 5-iodo-2-pyrimidinone-2'-deoxyribose toxicity, pharmacokinetics, and DNA incorporation in athymic mouse tissues and the human colon cancer xenograft, HCT-116. Cancer Res. 1994; 54:2695-700.

To date, we have synthesized and conducted limited testing of three prodrugs: ¹²⁵IUdR-5'-β-D-glucopyranoside (IUdR-Glu), ¹²⁵IUdR-5'-β-D- galactopyranoside (IUdR-Gal), and ⁱ^IUdR-δ'-β-D- cellobioside (IUdR-Cel). Chemical structures of prodrugs tested to date are shown in Figure 1.

Glycosidase activity along the GI tract originates from two separate sources: bacteria and the host organism. Site-specific release of ¹²⁵IUdR from its prodrug glycosides into the lumen of the large bowel is dependent on the recognition of ¹²⁵IUdR prodrugs by specific enzymes. Two prevailing glycosidases, -D-glucosidase and -D-galactosidase, are produced by bacteria found in the gut of humans and mice. Therefore, our preliminary studies concentrated on two substrates for these enzymes: ¹²⁵IUdR-glucopyranoside and ¹²⁵IUdR- galactopyranoside. The ability of colonic microflora to release ¹²⁵IUdR was tested in vitro and in the content of the GI tract of Swiss

Webster mice. The activity of β-D-glucosidase and β-D-galactosidase toward the p-nitrophenyl-based substrates was first measured ex vivo to verify the presence of these enzymes in the isolated contents of the GI tract. The results are summarized in Table 1 below.

Table 1. Ex Vivo p-N0₂-phenyl-β-D- p-N02-phenyl-β-D- activities of -β-D- glucopyranoside (nmole galactopyranoside glucosidase and -β-D- min ¹ gram content ¹) (nmole min -¹ gram galactosidase in the GI content ¹) contents of Swiss Webster mice mean stds mean stds stomach 49.4 6.5 5.3 0.6

proximal small 217.0 31.3 18.0 0.8 intestine

distal small intestine 129.7 8.9 15.2 1.1

colon/caecum 1965.9 158.5 2563.2 217.8

Spectroscopic, chromatographic, and enzymatic analyses confirmed that no-carrier-added glycosides of ¹²⁵IUdR have the desired -D- stereochemistry. In vitro, prodrugs released ¹²⁵IUdR in the presence of either isolated enzymes (Figure 2) or contents of the GI tract (Figure 3). The release was nearly quantitative and ¹²⁵IUdR was the only product detected. Incubation of IUdR-Glu with -D-glucosidase did not produce any appreciable amount of ¹²⁵IUdR (open circles in Figure 2). On average less than 1% of the radioactivity was in the form of ¹²⁵IUdR, approximately 1% in the form of free iodide and about 98% remained as the starting glycoside. This experiment confirmed the stereochemistry of prodrugs and the recognition of β-D-glycoside residues by specific enzymes.

A similar residue recognition was observed in ex vivo experiments conducted in homogenates of the murine GI content. Unlike ¹²⁵IUdR, glycosidic prodrugs of ¹²⁵IUdR were stable in the stomach content (pH 1-3). Luminal enzymes of proximal and distal small intestines in mice degraded less than 5% of IUdR-Glu, 20% of IUdR-Gal, and about 9% of IUdR-Cel. Contents of the human small intestine have only a few bacteria. Kinsella, T.J., Collins, J., Rowland, J., Klecker, R., Jr., Wright, D., Katz, D., Steinberg, S.M., Glastein, E. Pharmacology and phase I/II study of continuous intravenous infusions of iododeoxyuridine and hyperfractionated radiotherapy in patients glioblastoma multiforme. J. Clin. Oncol. 1988; 6:871-9; Sayler, A.A., Leedle, J.A.Z. In: Henteges, D.J., Ed. Human Intestinal microflora in health and disease. Academic Press, New York, 1983; pp 129-140. Therefore, we can anticipate selective release of ¹²⁵IUdR in the luminal content of the large intestine in patients. Bacterial enzymes from the colonic/caecal lumen of mice released nearly 100% of ¹²⁵IUdR during the time period (60 min.) approximately equivalent to the murine large intestine residence time. ¹²⁵IUdR freed by bacterial glycosidases was stable in the GI content. No significant amount of other metabolites was observed (total radioactivity recovered as by-products was less than 10%).

Example 3 Biodistribution studies in normal mice

In vivo after the i.v. injection, neither of the two tested prodrugs was dehalogenated by the liver. This was indicated by rapid blood clearance, lack of the uptake by the thyroid, and the analyses of circulating metabolites. As expected the efflux of prodrugs from the GI tract after oral administration was minimal, probably because of the hydrophilic nature of glycosides. The detection of any intact prodrug that may have entered the systemic circulation was difficult because its blood half-life is on the order of seconds. Blood clearance curves shown in Figure 4 clearly illustrate the advantage of an oral prodrug over ¹²⁵IUdR. Systemic distribution of radioactivity in the case of a prodrug is minimal (less than 2% at the time point corresponding to a steady state concentration in blood). At this concentration, even assuming that all released radioisotope is in the form of ¹²⁵IUdR, the uptake in tissues that may be engaged in the DNA synthesis will be negligible. Any other possible form of radioiodine (intact prodrug, iodouracil, free iodide) is either rapidly excreted or deposited in thyroid. Delayed presentation of ¹²⁵IUdR and free iodide (lag time about 30 min.) in systemic circulation indicates that at least part of the freed ¹²⁵IUdR may have been taken up by lymphatic capillary beds. Hydrophilic molecules are taken up primarily into superior and inferior mesenteric veins, which drain into the hepatic-portal circulation, more hydrophobic drugs (e.g., ¹²⁵IUdR) are transported from the large intestine via lymphatics. Deschner, E.E. In: Moyer, M.P., Poste, G.H., Eds. Colon cancer cells. New York. Academic Press, 1990; Chapter 3:Kinetics of normal, preneoplastic, and neoplastic epithelium, pp. 41-61. Planar gamma-camera images of mice after i.v. administration of

¹²⁵IUdR, IUdR-Glu, and IUdR-Gal confirm results of the biodistribution studies (data not shown). At 15 min. after an i.v. injection, both prodrugs are rapidly cleared via the renal excretion pathway and are present only in kidneys (glucopyranoside) or bladder (galactopyranoside), whereas radioactivity in mice receiving ¹²⁵IUdR (not in the prodrug form) is homogeneously distributed into the extracellular space and prominent in the stomach and bladder in the form of free iodide.

Images acquired 90 min. after oral administration (Figure 6) show radioactivity in the GI tract of a mouse treated with a prodrug and only thyroid- and stomach-associated radioactivity in the form of free iodide, in an animal receiving oral ¹²⁵IUdR. The distribution of orally and i.v. administered ¹²⁵IUdR is virtually identical and corresponds to the sites of deposition of free radioiodide. In contrast, orally administered prodrugs do not leak into systemic circulation and are predominantly localized in the GI tract.

Of particular significance is the stomach clearance curve for i.v. administered ¹²⁵IUdR and prodrugs. Five minutes after oral dose of ¹²⁵IUdR treatment there is already a significant redistribution of radioactivity from stomach to systemic circulation and extracellular spaces. At 30 minutes post-treatment the radioactivity in stomach increases indicating that free iodide is being reabsorbed in its usual metabolic pathways (Figure 7). In contrast, the decline of radioactivity in stomach of mice treated with oral doses of IUdR-Glu is accompanied by increased radioactivity in the intestine. The uptake in other tissues is minimal and parallels radioactivity levels in the blood. Unquestionably, the desired targeting of the GI tract is accomplished with glycosidic prodrugs of ¹²⁵IUdR. The mean residence time of prodrugs in the intestine is about 170 minutes (not corrected for decay). The pharmacologically relevant concentrations persist for at least 60 minutes in the large intestine and can be modified by either altering the sugar residue, for example, by using other oligosaccharides, dextrans or various cyclodextrins, or by adjusting dietary intake, an event known to modify the intestinal microflora.

Example 4 Biodistribution in mice treated with carcinogen

Experiments conducted in female Swiss albino mice treated with 1,2- dimethylhydrazine (1,2-DMH), a carcinogen known to produce neoplastic changes in the GI tract confirmed colon-directed delivery of ¹²⁵IUdR glycosides. Six-week old mice were divided into four groups, 10 mice per group (groups A and B: carcinogen and its control for ¹²⁵IUdR prodrug treatment, and groups C and D: carcinogen and its control for ¹²⁵IUdR treatment) and were given subcutaneous injections of a total of 0.2 mg 1,2- DMH/kg body weight in ten equal doses once a week for 10 weeks. Twenty and 24 hours after the eighth, ninth, and tenth 1,2-DMH injection, mice in groups A and B were given orally a mixture of 5 μCi IUdR-Glu and 5 μCi IUdR-Gal. Mice in groups C and D were treated identically with 10 μCi of orally administered ¹²⁵IUdR. Twenty-four hours after the last dose of radiopharmaceutical mice were euthanized, necropsy performed, radioactive content of selected tissues determined and stomach, small intestine, distal and proximal colon, caecum and rectum preserved in 10% formalin, and processed for autoradiography. Abnormal proliferative activities in colon, rectum and caecum are positively identified by enhanced uptake of ¹²⁵IUdR freed from its prodrugs (Figure 8). Similar effects are not observed in tissues collected from animals treated with ¹²⁵IUdR not in the form of a prodrug. The "unprotected" ¹²⁵IUdR rapidly escapes the GI tract or is degraded in the stomach content and is not able to reach the site of cancer. The relative uptake of radioactivity from prodrugs in distal colon is about six times greater than the uptake in control mice and three times greater than in mice treated with ¹²⁵IUdR (the relative uptake of one indicates identical behavior of the drug in carcinogen-treated and control animals). The results of these pilot experiments substantiate our hypothesis that glycosides of ¹²⁵IUdR are colon- directed and during their residence at the site of cancer, these prodrugs release intact ¹²⁵IUdR which is preferentially taken up by abnormally dividing cells. The design of this experiment (dictated by economics; long term animal studies are expensive) did not allow evaluation of individual prodrugs. Nevertheless, the cumulative uptake of ¹²⁵IUdR released from two prodrugs, glucoside and galactoside, administered at the total dose equivalent to the dose of ¹²⁵IUdR, decisively confirms the potential value of oral radiotherapeutics in the treatment of residual or occult colorectal cancer. Additionally, there appears to be an enhanced uptake of radioactivity in the liver. It indicates that intact ¹²⁵IUdR reaches the liver and is taken up by DNA-synthesizing cells. Since this phenomenon is observed in neither glycoside-treated controls nor in ¹²⁵IUdR-treated groups C and D, this abnormal uptake may be a result of 1,2-DMH-induced preneoplastic changes in the liver. This must be confirmed in full autoradiography and histopathology studies of liver and other tissues known to be susceptible to this carcinogen's action, harvested from mice with established cancer. This result is also relevant to the potential use of halogenated pyrimidines as radiosensitizers. An oral prodrug of IUdR or

BUdR able to deliver pharmacologically relevant doses to liver metastases will eliminate current protocols of continuous infusion (days at doses as high as 1 g/m²) and reduce the risk of side effects associated with such high doses.

Example 5 Therapy studies in mice treated with carcinogen

Having established the site-specific release of the intact ¹²⁵IUdR from its glycoside carriers, two prodrugs underwent testing in the mouse model of colorectal cancer. The 1,2-DMH-induced colon cancer was selected because it approximates human colon carcinogenesis. The tumor-bearing colorectal mucosa of humans and mice have two kinetic parameters in common: (1) the upward shift in the distribution of DNA synthesizing cells and (2) the elevated LI. Richards, T.C. Early changes in the dynamics of crypt cell populations in mouse colon following administration of 1,2- dimethylhydrazine. Cancer Res. 1977; 37:1680-5; Deschner, E.E., Maskens, A.P. Significance of the labeling index and labeling distribution as kinetic parameters in colorectal mucosa of cancer patients and DMH treated animals. Cancer. 1982; 50:1136-41; Chang, W.W.L. Histogenesis of symmetrical 1,2-dimethylhydrazine-induced neoplasms of the colon in the mouse. J. Natl. Cancer Inst. 1978; 60:1405-18. The distribution of S-phase cells within crypts in mice treated with 1,2-DMH reveals the extension of the proliferative compartment onto the lumenal surface, a shift of the predominant zone of DNA synthesis to the upper third of the glands and hyperproliferative crypts with LI >15%. Chang, W.W.L. Histogenesis of symmetrical 1,2-dimethylhydrazine-induced neoplasms of the colon in the mouse. J. Natl. Cancer Inst. 1978; 60:1405-18. This is analogous to cell kinetics demonstrated in biopsies from various groups of patients at high risk for colon cancer. Shpitz, B., Bomstein, Y., Cohen, R., Kaufman, Z., Neufeld, D., Galkin, M., Bernheim, J. Aberrant crypt foci in human colons: distribution and histomorphologic characteristics. Human Pathol. 1998;

29:469-75; Deschner, E.E., Raicht, R.F. Kinetic and morphologic alterations in the colon of a patient with multiple polyposis. Cancer. 1981; 47:2440-5; Balzi, M. Becciolini, A., Mauri, P., Larosa, V., Bechi, P. Proliferative activity in normal colon mucosa and tumor tissue: clinical implications. In Vivo. 1993; 7:635-637; Bleiberg, H., Galand, P. In vitro autoradiographic determination of cell kinetic parameters in adenocarcinomas and adjacent healthy mucosa of the human colon and rectum. Cancer Res. 1976; 36:325-8; Maskens, A.P., Deschner, E.E. Tritiated thymidine incorporation into epithelial cells of normal-appearing colorectal mucosa of cancer patients. J. Natl. Cancer Inst. 1977; 58:1221-4; Shimomatsuya, T., Tanigawa, N., Muraoka, R. Proliferative activity of human tumors: assessment using bromodeoxyuridine and flow cytometry. Jpn. J. Cancer Res. 1991; 82:357-62. Mice treated with 1,2-DMH develop carcinomas of the colon that are similar histologically to human tumors. Charalambous, D., Farmer, C, O'Brien, P.E. Sulindac and indomethacin inhibit formation of aberrant crypt foci in the colons of dimethyl hydrazine treated rats. J. Gastroenterol. Hepatol.

1996; 11:88-92; Shamsuddin, A.M. In: Moyer, M.P., Poste, G.H., Eds. Colon cancer cells. Academic Press, New York, 1990; Normal and pathological anatomy of the large intestine, pp 15-40. There is a progressively increased crypt hyperplasia and displacement of the cell division zone with continued exposure to the carcinogen. The 1,2-DMH-induced tumors in mice appear in the greatest numbers in the distal colon in an area between the descending colon and anus. This precisely matches the sites of occurrence of human colon cancer.

One hundred sixty female Swiss albino mice (six-weeks old), 80 of whom were given ten weekly subcutaneous injections of 1,2-DMH were randomly assigned to six groups: five controls and one experimental group. The treatment schedule is outlined in Table 2. Table 2. Summary of the treatment schedule and group designation. Mice received ten subcutaneous weekly doses of 1,2-dimethylhydrazine dihydrochloride (left shoulder; 0.33 μmole base [20μg] per gram body weight in 0.01 mL phosphate buffered saline, pH adjusted to 7.2 with sodium hydroxide). Twenty-four to 30 hours after 1,2-DMH, a 10-μCi oral dose of ¹²⁵IUdR (4.54 pmole/dose) or the ¹²⁵IUdR-Glu and ¹²⁵IUdR-Gal mixture (5 μCi each; 2.27 pmole each/dose) was administered in 5 μL PBS. The administration of radiopharmaceuticals was discontinued after the 4th dose when the thyroid uptake in ¹²⁵IUdR and ¹²⁵IUdR/l,2-DMH mice exceeded 0.3 μCi.

O O

The control, age-matched mice received either no treatment (NT group), orally administered ¹²⁵IUdR (IUdR group), or a mixture of IUdR-Glu and IUdR-Gal (IUdR-Glu/IUdR-Gal group). The controls in the 1,2-DMH- treated category received either 1,2-DMH alone (DMH group) or 1,2-DMH followed by oral ¹²⁵IUdR (IUdR/DMH group). Adhering to the identical schedule, the experimental group was injected with 1,2-DMH and received orally a mixture of IUdR-Glu and IUdR-Gal (IUdR-Glu/IUdR-Gal/DMH group). To avoid any interference from agents other than those being tested, the traditional thyroid block with potassium iodide was not included in the protocol. Forty-eight hours after every other oral dose of radiopharmaceuticals three mice were randomly selected from each of the control groups and five mice from the experimental group. These animals were asphyxiated with C0₂₎ necropsy performed, and the radioactive content of selected tissues measured. After the fourth dose, the thyroid uptake of ¹²⁵I in the IUdR and IUdR/DMH groups exceeded 0.3 μCi at 48 h post-administration. At this time the treatment with orally administered radiopharmaceuticals was terminated to avoid any complications that may have resulted from impaired thyroid function. Overall, the mice received total of 40 μCi of either ¹²⁵IUdR or a mixture of ¹²⁵IUdR-Glu and ¹²⁵IUdR- Gal. During the course of the experiments, additional groups of mice were injected i.v. via a tail vein with 10 μCi ¹³¹IUdR, a marker of proliferation, to assess the DNA synthesis in the GI tract and other selected tissues. ¹³¹IUdR was used to allow simultaneous determination of the uptake of both radiodine isotopes. Two days later, these mice were also asphyxiated and necropsy performed.

The food consumption and body weight gain were similar in all groups during the initial four weeks of the experiment (Figure 9). After the sixth dose of 1,2-DMH the growth of mice in the DMH group appeared to be arrested and eventually their weight started to decline at eight weeks of the carcinogen treatment. A similar drop in weight at eight weeks was also observed for all other mice treated with 1,2-DMH, controls as well as the experimental. The IUdR-Glu/IUdR-Gal DMH mice recovered their weight by week 10 and continued to gain at the rate comparable to the control mice (NT, IUdR and IUdR-Glu/IUdR-Gal groups). Within the same time frame, mice in the DMH group lost nearly all of the weight gained during the initial four weeks of the experiment and did not recover their weight. The weight of the IUdR/DMH mice plateaued at about 130% of the initial weight at eleven weeks and remained virtually constant for the duration of the experiment. The slight weight increase, about 1%, in the DMH and IUdR/DMH groups around week 20 is the effect of the increased tumor mass. The necropsies indicated that a total tumor burden in several animals in these two groups had already exceeded 0.5 g. The observed changes in the weight correspond to the health status of the mice. The IUdR-Glu IUdR-Gal/DMH mice gained weight at the rate virtually identical to animals in all control groups including normal mice of the same sex and age housed in the same facility but which were not exposed to any of the drugs.

Abnormal proliferative activities in colon, rectum and caecum were positively identified by the enhanced uptake of ¹³¹IUdR in mice from DMH and IUdR/DMH groups. The tissue uptake of ¹³¹I in mice receiving i.v. 10 μCI ¹³¹IUdR eight weeks after the last dose of 1,2-DMH indicates that the proliferation is significantly increased in all GI segments in mice treated with 1,2-DMH plus oral ¹²⁵IUdR relative to controls treated with oral ¹²⁵IUdR only (Figure 10). The relative proliferative activity of several tissues in IUdR/DMH mice is depressed when compared to animals from the DMH group (Figure 10). As an example, the uptake of i.v. administered

¹³¹IUdR is greater in the liver, kidney, and caecum of mice treated with 1,2- DMH alone. This may be an indication that in these tissues, which are known to be the additional sites of 1,2-DMH-induced tumor development, the intact ¹²⁵IUdR escaping the GI tract halted or delayed the formation of tumors. There were no significant differences in the levels of ¹³¹IUdR in tissues harvested from any of the NT, IUdR, and IUdR-Glu/IUdR-Gal animals as compared to IUdR-Glu/IUdR-Gal/DMH mice (Figure 10) except in the large intestine and caecum. The relative uptake of ¹³¹I in these tissues was 82% ± 11% and 74% ± 13% (means ± stds) of the control levels, respectively. The suppression of the crypt LI in the 1,2-DMH-treated rodents is typically recorded at about seven to ten hours after 1,2-DMH. One day later a slight elevation of the LI, approximately 2-4%, is observed and soon after, within the 36-48 hours, the LI returns to the normal values. Sunter, J.P. In: Appleton, D.R., Sunter, J.P., Watson, A.J., Eds. Cell

Proliferation in the gastrointestinal tract. Pitman, Turnbridge Wells, Kent. 1980; Experimental carcinogenesis and cancer in the rodent gut. pp. 255-277. The ³ⁱIUdR uptake in IUdR-Glu/IUdR-Gal/DMH mice was evaluated several weeks after the last dose of the carcinogen; therefore any 1,2-DMH- related inhibition of the LI is not anticipated. On the contrary, the expected LI should either reach levels similar to these of the 1,2-DMH mice in the event of the failure of the therapy, or remain at the levels observed for the control group if therapy succeeds. It appears that the inhibitory effect of ¹²⁵IUdR-Glu and ¹²⁵IUdR-Gal after 1,2-DMH treatment persists for at least eight weeks after the last carcinogen dose and 12 weeks after the last oral administration of the ¹²⁵IUdR-Glu and ¹²⁵IUdR-Gal mixture. The duration of this effect may be longer lasting but in our experimental design the final group of mice was tested for changes in the proliferative activities eight weeks following the last dose of 1,2-DMH. At the present time, the mode of the observed effect and the role of

¹²⁵IUdR prodrugs are highly speculative. It appears that the uptake of ¹²⁵IUdR into the DNA, either during DNA repair or during scheduled DNA synthesis, suppresses the 1,2-DMH-induced colon carcinogenesis, prevents tumor development and growth. It has been established that stimulation of mitotic activity during the initiation stage of carcinogenesis enhances the carcinogenic process, it follows than that the elimination of cells affected during the initiation stage may reduce or stop the development of colon tumors. ¹²⁵IUdR released from its prodrugs in the lumen of the GI tract can be taken up only by cells in the process of making DNA. The cells most likely to use ¹²⁵IUdR are these with the elevated LI index, prolonged S- phase and engaged in the repair processes, that is cells from which the 1,2- DMH-induced colorectal cancer originates. The comparison of the ¹³¹I radioactivity levels in tissues harvested from the DMH mice with those obtained from either the IUdR-Glu/IUdR-Gal/DMH or NT groups verifies highly elevated proliferative activities in all segments of the large intestine in the DMH group (Figure 10). This parallels the ¹³¹I levels in mice treated with ¹²⁵IUdR plus 1,2-DMH and confirms the capability of ¹²⁵IUdR-Glu and ¹²⁵IUdR-Gal to intervene in the progression of the typical multistep carcinogenesis proposed for 1,2-DMH-treated rodents.

This capability of ¹²⁵IUdR-Glu and ¹²⁵IUdR-Gal to interfere with the process of cancer formation is remarkably underscored by the results of autopsies. The experiment was terminated on compassionate grounds when three mice in the DMH control groups developed large tumors protruding from the anus and nearly all of the 1,2-DMH-treated mice had blood in the feces. This occurred 21 weeks after the first 1,2-DMH treatment. Table 3 summarizes the tumor incidence and tumor burden for IUdR-Glu/IUdR- Gal/DMH mice and collectively for DMH and IUdR/DMH control groups.

Table 3. Tumor sites and frequency in mice treated with ¹²⁵IUdR-5'-β-D-glucopyranoside and ¹²⁵IUdR-5'-β-D- galactopyranoside. Mice were sacrificed 21 weeks after the first dose of 1,2-DMH at the age of 27-weeks.

IUdR-Glu/ IUdR-Gal/DMH DMH Controls

Number of mice 12 14

Mice with tumors 12

Dead None 9**

Total number of GI and other tumors 129

Tumors in distal large intestine and rectum 97

Tumors at sites other than distal large intestine and rectum none caecum (1); liver (1); stomach

(2); small intestine (24); proximal large intestine (4)

Tumors < 1 mm 16

Tumors 1-3 mm none 77

Tumors 3-5 mm none 30

Tumors > 5 mm none 6

Average tumor burden Small Intestine (mg/mouse; standard deviation) 0 111(32) Large Intestine (mg/mouse; standard deviation) 3* 260(90)

* total weight (mg) of tumors in one mouse ** necropsy not performed

All of the mice in both DMH control groups developed tumors of the GI tract. Seventy-five percent of tumors were located in the distal large intestine. Nearly 19% of tumors were found in the small intestine which appears to be the secondary site. There were some differences between the DMH and IUdR/DMH groups in the total number of tumors and their location. The DMH mice had a total of 79 tumors, of these 24 were located in the small intestine. The IUdR/DMH animals did not develop any tumors in this area. The primary tumor sites in the IUdR/DMH group were the distal colon (44 tumors; 88%) and proximal colon (4 tumors; 8%) indicating that a small fraction on intact ¹²⁵IUdR escapes degradation in the stomach, passes into the small intestine and may prevent tumor development in this area of the GI tract. The tumor incidence in the large intestine for these two control groups was comparable, 53 in the DMH group and 44 in the IUdR/DMH group. The tumor burden was lower in mice which received PO ¹²⁵IUdR during the carcinogen treatment. Twenty-four percent of mice in the IUdR/DMH group had tumors 1 mm as compared to 5% in the mice treated with 1,2-DMH alone. The incidence of tumors >3 mm was nearly identical in both groups (18 vs. 19). Almost 71% (56) of tumors found in the distal large intestine of mice treated with 1,2-DMH were within the 3 to 5 mm diameter compared to 42% (21 tumors) in the IUdR/DMH mice. Figure

11 illustrates a typical appearance of the large intestine in mice from two control and one experimental group.

Mice treated with 1,2-DMH followed by oral doses of ¹²⁵IUdR-Glu and ¹²⁵IUdR-Gal showed virtual absence of tumors. Only a single mouse presented with three small tumors (< 1 mm) in the rectal region. Glycosides of ¹²⁵IUdR are colon-directed. During their residence at the site of cancer, ¹²⁵IUdR released from carrier glycosides is preferentially taken up by dividing cells or cells engaged in unscheduled DNA synthesis such as repair of lesions produced by 1,2-DMH. The cumulative uptake of ¹²⁵IUdR released from two prodrugs, glucoside and galactoside, administered at the total dose of 40 μCi, equivalent to the dose of ¹²⁵IUdR, prevented formation of 1,2- DMH-induced tumors and decisively confirmed the potential value of these oral radiotherapeutics in the treatment of residual or occult colorectal cancer.

Although the 1,2-DMH-induced tumors have a very low incidence of metastasis to the liver, up to 36% incidence of mesenteric lymph node metastases has been reported. House, A.K., Maley, M.A. Colorectal carcinoma in a rat model: suppression of tumor development and altered host immune status following treatment with anti-B-lymphocyte serum. J. Surg. Oncol. 1986; 32:256-62. In colon cancer patients lymph node metastases are uniformly associated with a poor outcome. For example, approximately 40% node-positive patients develop recurrence versus only 6% in patients with metastasis-negative nodes. Malassagne, B., Valleur, P., Serra, J., Sarnacki, S., Galian, A., Hoang, C, Hautefeuille, P. Relationship of apical lymph node involvement to survival in resected colon carcinoma. Dis. Colon Rectum. 1993; 36:645-53; Cohen, A., Tremiterra, S., Candela, F., Thaler, H.T., Signurdson, E.R. Prognosis of node-positive colon cancer. Cancer. 1991; 67:1859-61. IUdR prodrugs may be useful in preventing lymph-node metastases. There is no direct systemic drainage from any region of the large intestine, the potential absorption of freed IUdR through the lymphatic capillary beds would drain IUdR in a retrograde manner, ultimately emptying it into the systemic circulation at the left internal jugular vein. Uptake into the lymphatics would allow IUdR to participate in the DNA synthesis of dividing tumor cells in lymph nodes bypassing immediate transport to the liver where a first-pass-effect can occur.

Example 6

(Prophetic)

The aim of this experiment is to evaluate the therapeutic and diagnostic potential of oral ¹²⁵IUdR prodrugs in preneoplastic and neoplastic lesions during 1,2-dimethylhydrazine-induced colorectal carcinogenesis. We will evaluate the therapeutic and diagnostic potential and determine the metabolic fate of oral ¹²⁵IUdR prodrugs in colorectal cancer in mice. The results of the preliminary studies discussed earlier in this application established foundation for further experiments and clearly confirmed the validity of this approach. The main experimental end-points which will be considered are (1) therapy/survival studies, (2) the subcellular distribution of ¹²⁵IUdR in various tissues after carcinogen treatment; and (3) histopathology of tissues with significant uptake of radioactivity. To accomplish these end-points we will expand the experimental group of animals to allow for therapy studies with parallel biodistribution studies at time points significant to tumor development such as 10-16 weeks: hyperplastic changes; 16-22 weeks: macroscopic polyps and tumors; and 30 weeks: up to 10% of mice already in disease-related distress (termination of the experiment if the event of failed therapeutic regimen). We will use randomly bred female Swiss albino mice from a colony maintained at University of Nebraska Medical Center. It is our experience that the rate and frequency of carcinogen-induced tumors is identical in male and female mice. However, female mice are less aggressive and easier to handle in long-term experiments planned in this study. Mice will be housed in plastic cages on granular cellulose bedding separated into groups of 5 and fed Tekland sterilizable rodent diet (Harlan Tekland, Madison, WI). The carcinogen, 1,2-dimethylhydrazine dihydrochloride will be given weekly for 10 weeks via subcutaneous injections of 20 μg 1,2-DMH base/g body weight in 0.01 mL physiologic saline in the interscapular region with a tuberculin syringe using a 24-gauge needle. Animals will be checked daily and their weight recorded at weekly intervals. Any gross pathological changes will be recorded.

To determine the therapeutic potential of ¹²⁵IUdR prodrugs in established colon cancer mice will receive 10 weekly doses of 1,2-DMH. Upon completion of the carcinogen treatment all animals will receive either

¹²⁵IUdR-prodrug or no treatment as outlined below. Additional mice will be treated with prodrugs alone. After eight 1,2-DMH injections the proliferative zone height of colonic mucosa does not return to control levels even after 22-week recovery period, therefore, we expect that any reduction in the number of tumors in treated animals will correlate with the administration of prodrugs. Group 1: 10 weekly SC doses of 1,2-DMH followed one week later by 10 weekly oral does of 10 μCi ¹²⁵IUdR prodrug in normal saline. Group 2: 10 weekly SC doses of 1,2-DMH followed one week later by 10 weekly oral doses of normal saline. Group 3: 10 weekly SC doses of saline followed one week later by 10 weekly oral doses of 10 μCi ¹²⁵IUdR prodrug in normal saline. Group 4: 10 weekly SC doses of saline; normal untreated mice.

Each group of mice will be checked daily to determine the number of surviving animals and to note any changes in normal habits. The experiment will be terminated on or before week 30, depending of the health status of animals and all surviving mice will be sacrificed. The number of tumors in the GI tract will be determined. The data will be analyzed for three end-points: (1) median and absolute survival using Wilcoxon signed rank test; (2) total number of tumor sites in surviving mice; and (3) the size of tumors. From each group of mice, a randomly selected sample of six mice will receive i.v. ¹³¹IUdR 10 weeks into the 1,2-DMH treatment to determine the proliferative activities of selected tissues and to conduct histophatology studies. Additionally, six mice from each group will be autopsied at 6, 24, and 48 hours after the administration of ¹²⁵IUdR prodrugs. The total radioactive content and the DNA-bound radioactivity in each tissue will be measured and the radiation doses determined using a MIRDOSE2 computer program. Selected tissues will be processed for autoradiography. Tissues dissected from the therapy experiment (liver, bone marrow, tongue, spleen, kidney, lungs, brain, heart, blood, thyroid) will be examined for obvious signs of cancer and their radioactive content determined. All tumors and abnormal tissues will be examined macroscopically, fixed in 10% buffered formalin, and stained with hematoxylin/eosin. Selected tumors and tissues will undergo histopathologic examination.

To determine the usefulness of this adjuvant therapy at various stages of tumor development, groups of mice treated with 1,2-DMH and controls as outlined above will be treated with multiple doses of prodrugs during the carcinogen treatment, immediately after the course of the carcinogen, and 2, 4, 8, 16, and 20 weeks after the last dose of 1,2-DMH. Mice will be killed and autopsied 12, 24, 48, and 72 hours after the last dose of the prodrug (10 mice per time point; the number of animals per group is increased because of the variability in tumor development particularly at earlier time points after 1,2-DMH). The total radioactive content and DNA- bound radioactivity in each tissue will be measured, radiation doses determined using a MIRDOSE2 computer program, and tissues with high radioactivity uptake processed for autoradiography and histopathologic examination.

One area of concern is the radiation dose accumulated during the transit of the prodrug through the GI tract. A dose calculation for ¹²³IUdR administered intravesically into the bladder and allowed to accumulate decays for 60 min. indicates that only about 6 x 10 ² cGy/mCi is expected (Dr. Peter Leichner, private communication). Therefore, the γ-ray dose received from ¹²⁵IUdR prodrugs (this would be the 35 keV γ-ray emission from ¹²⁵I) by the entire gastrointestinal tract should also be negligible. We suppose that since the tolerance dose (TD5/5) for therapeutic fractioned external beam doses are 5500 cGy for colon and 6000 cGy for rectum (Cohen, A., Tremiterra, S., Candela, F., Thaler, H.T., Signurdson, E.R. Prognosis of node-positive colon cancer. Cancer. 1991; 67:1859-61), the radiation dose delivered by orally administered prodrugs of ¹²⁵IUdR should not produce any of the damage observed with the external radiation treatment (the dose from 10 mCi extracellular ¹²³I is about 100,000 times lower than TD₅/5). Nevertheless, we plan to evaluate radiation burden on all tissues with the abnormally high uptake of radioactivity. Some of the proposed treatments are outlined below.

For biodistribution studies, one set of mice will receive orally 10 μCi of iodine-125-labeled agent and 10 μCi of iodine-131-labeled IUdR 24-30 hours after the last 1,2-DMH dose; one set of animals will receive orally one or doses of 10 μCi of iodine- 125-labeled agent and 10 μCi of iodine-131- labeled IUdR 24-30 hours after the last saline injection; the untreated mice will be injected only with 10 μCi of iodine-131-labeled IUdR. The therapy groups will undergo the same treatment with the ¹³¹IUdR.

For biodistribution studies, one set of mice will receive orally 10 μCi of iodine-131-labeled IUdR 20 weeks after the last 1,2-DMH dose; one set of animals will receive orally 10 μCi of iodine- 125-labeled agent and 10 μCi of iodine-131-labeled IUdR 20 weeks after the last saline injection; the untreated mice will be injected only with 10 μCi of iodine- 131-labeled IUdR. The therapy groups will undergo the same treatment without the ¹³¹IUdR. In both approaches, the therapy of established tumors and prevention of new tumor formation, for biodistribution studies tissues will be harvested, radioactive content determined in the gamma-counter (15-75 keV ¹²⁵I minus 14.4% ¹³ I spillover (eff. 75%); 260-470 keV ¹³¹I (eff 47.5%), 15-2000 keV total. All sections of GI tract, liver and any tissue with elevated radioactive content will be analyzed for ¹²⁵I and ¹³¹I contents in DNA as well examined histologically.

Adhering to the identical schedule, mice in the therapy studies will be treated with a single or multiple doses of prodrugs. They will be observed for 30 weeks or until the experiment is terminated on compassionate grounds. Any dead animal will undergo necropsy and the sites of a macroscopic disease in the GI tract determined. Twenty-four hours prior to euthanasia during the week 20 all surviving mice will receive i.v. injection of ¹³¹IUdR; selected tissues will be harvested, radioactive content determined in a gamma-counter; macroscopic disease in the GI tract will be determined; sections of the GI tract, liver, and other tissues with elevated labeling index will be processed for histopathology and autoradiography; remaining sections will be processed for ¹²⁵I-DNA content.

Example 7 (Prophetic)

The aim of this experiment will be to measure levels of thymidine replacement by IUdR in liver of carcinogen-treated mice. IUdR and its brominated analog, 5-bromo-2'-deoxyuridine, are used as radiosensitizers in therapies of colorectal cancer metastatic to the liver. The metabolic processing and lack of specificity for tumor cells discussed earlier is responsible for systemic toxicity and other side effects when the continuous infusion is employed to achieve high replacement of thymidine with these radiosensitizers. The alternative route of delivery offered by our prodrugs has a potential to reduce the systemic toxicity. The targeting of ¹²⁵IUdR to the liver observed in 1,2-DMH-treated mice indicate that IUdR freed from its prodrugs in the large intestine escapes the GI tract and is able to reach the liver in its intact form. This observation prompted us to undertake further studies into the oral delivery of IUdR to liver for radiotherapy with the external beam.

The molecular mechanism of tumor radiosensitization by IUdR is uncertain but it seems to rely on the generation of highly reactive free radicals within DNA and analogously to radioiodinated IUdR, requires incorporation of into the DNA to produce the desired biologic effect.

Radiosensitization appears to be linearly related to IUdR incorporation throughout the range of conditions assessed. Lawrence, T.S., Davis, M.A., Maybaum, J., Steston, P.L., Ensminger, W.D. The dependence of halogenated pyrimidine incorporation and radiosensitization on the duration of drug exposure. Int. J. Radiat. Oncol. Biol. Phys. 1990; 18:1393- 8. Most clinical trials with IUdR involve continuous, prolonged infusions of 1 g IUdR/m²/day for a week or longer. Epstein, A.H., Lebovics, R.S., Goffman, T., Teague, D., Fuetsch, E.S., Glatstein, E., Okunieff, P., Cook, J.A. Treatment of locally advanced cancer of the head and neck with 5'- iododeoxyuridine and hyperfractionated radiation therapy: measurement of cell labeling and thymidine replacement. J. Natl. Cancer Inst. 1994; 86: 1775-80; Rodriguez, R., Ritter, M.A., Fowler, J.F., Kinsella, T.J. Kinetics of cell labeling and thymidine replacement after continuous infusion of halogenated pyrimidines in vivo. Int. J. Radiat. Oncol. Biol. Phys. 1994; 29:105-13; Cook, J.A., Glass, J., Lebovics, R., Bobo, H., Pass, H., DeLaney,

T.F., Oldfield, E.H., Mitchell, J.B., Glatstein, E., Goffman, T.E. Measurement of thymidine replacement in patients with high grade gliomas, head and neck tumors, and high grade sarcomas after continuous intravenous infusions of 5-iododeoxyuridine. Cancer Res. 1992; 52:719-25. The optimal clinically relevant thymidine replacement has not been clearly defined, but values of 3-10% of thymidine replacement by IUdR in all tumor cells are expected to be within the required range. Early clinical experience with IUdR was disappointing because normal tissue toxicity abolished any potential for therapeutic gain. Current clinical trials employ local/regional routes of administration and various dosing regimens but still the outcome is often inconclusive or negative. The problem lies, in part, in the patient selection, i.e., poorly radioresponsive high grade gliomas or sarcomas. But, the primary reason for failure is prolonged dosing regimen dosing regimen needed to improve IUdR uptake. This allows for the tumor cell repopulation during protracted treatment. Bypassing the usual catabolic pathways should produce similar thymidine replacement ratio at a fraction of the time, i.e., increase pharmacologically relevant concentrations of IUdR at the site of the tumor.

Because this set of experiments will establish the thymidine replacement ratio after oral administration of prodrugs, the pharmacologic advantage of the oral route of administration must be determined with the conventional multiple bolus i.v. injections of IUdR (the continuous infusion is impossible in this animal model. When the relative biological advantage of prodrugs over IUdR in this context is established, we plan to expand these studies to the model of regenerating liver in swine. Liver after partial hepatectomy displays increased proliferative capacity similar to the metastatic liver cancer). Mice will be treated as described above with the following changes: the total amount of administered drugs will be on the order of μmoles (not pmoles as in the case of no-carrier-added radiotherapeutic prodrugs); each dose of the prodrug will be radioiodinated with the trace amount of iodine- 125; each dose of IUdR will be radioiodinated with the trace amount of iodine-131. Both drugs, i.e., oral ^127/125IUdR-prodrug and ^{127 131}IUdR will be administered simultaneously to the same animal to determine the relative thymidine replacement of each under the identical set of conditions. The determination of the metabolites and ¹²⁵IUdR in blood will be done using TLC and HPLC. The tissue incorporation of ¹²⁵I and ¹³¹I will be measured in a gamma-counter. Some samples will be digested and the radioactivity in DNA will be measured. Additionally, the autoradiography will be used to verify the nuclear localization of IUdR. When the targeting of the liver and pharmacologic advantage of prodrugs over IUdR is established, the escalating doses of oral drugs will be administered to determine toxicity.

Example 8 (Prophetic)

The aim of this experiment will be the designing, synthesizing, and testing additional prodrugs of ¹²⁵IUdR that mitigate any potential shortcomings of already available agents. This will depend upon the results established in therapy studies described in the two previous prophetic examples. We will synthesize additional radioactive derivatives as needed and proceed with the in vitro testing and biodistribution in normal mice. Some of the contemplated prodrugs will incorporate ¹²⁵IUdR into the structure of dextrans and cyclodextrans or low molecular weight azo compounds.

The bioactivation of azo drugs is a result of the reductive environment of the ceacum. Singh, S., Das, M., Khanna, S.K. Azo reductase activity of microbial population from gastrointestinal tract segments of various animals species. Indian J. Exp. Biol. 1998; 36:99-103. In this instance, as in the case of glycosides, microflora of the GI tract is responsible for the release of IUdR. Since it has already been established that the bacteria-mediated release of the radioactive drugs is highly effective for targeting of colon, it appears that the azo derivatives of IUdR may provide an alternative mode of delivery. The advantage of the azo derivatives over the glycosides-based prodrugs may be in the reliance on the different bacteria population as well as the already available results of the extensive testing in humans. Watkinson, G. Sulphasalazine: a review of 40 years' experience. Drugs. 1986; 32 Suppl 1:1-11; Brown, J.P., McGarraugh, G.V., Parkinson, T.M., Wingard, R.E., Jr., Onderdonk, A.B. A polymeric drug for treatment of inflammatory bowel disease. J. Med. Chem. 1983; 26:1300-7; Zinberg, J., Molinas, S., Das, KM. Double-blind placebo- controlled study of olsalazine in the treatment of ulcerative colitis. Am. J. Gastroenterol. 1990; 85:562-6.

Dextrans, used in clinical practice since 1943, have a wealth of published information on its clinical uses (e.g., Thoren, L. Dextran as a plasma volume substitute. In: Jamieson, G.A., Greenwalt, T.J., Eds. Blood substitutes and plasma expanders. Alan R. Liss, New York. 1978; pp 19-27. The fate of dextrans in the GI tract is also well established and suggest that when longer residence time may be required such a prodrug will serve this purpose. Nielsen, L.S., Weibel, H., Johansen, M., Larsen, C. Macromolecular prodrugs. XX. Factors influencing model dextranase- mediated depolymerization of dextran derivatives in vitro. Acta Pharm. Nord. 1992; 4:23-30. Dextranase activity has been demonstrated in the intestinal mucosa of rat and pig, in human fecal samples and in anaerobic bacteria cultured from human feces. Nielsen, L.S., Weibel, H., Johansen, M., Larsen, C. Macromolecular prodrugs. XX. Factors influencing model dextranase-mediated depolymerization of dextran derivatives in vitro. Acta Pharm. Nord. 1992; 4:23-30; Aberg, B. Breakdown of dextran in human feces. Scan. J. Clin Lab Invest. 1953; 5:37-41; Dahlqvist, A. The location of carbohydrases in the digestive tract of the pig. Biochem. J. 1961; 78:282-6. By varying the size of carrier, dextran it is possible to tailor the GI residence time. Dextran conjugates have been synthesized from a variety of drugs including nonsteroidal anti-inflammatory agents (Larsen, C, Jensen, B.H., Olesen, H.P. Bioavailability of ketoprofen from orally administered ketoprofen-dextran ester prodrugs in the pig. Acta Pharm. Nord. 1991; 3:71- 6; Harboe, E., Larsen, C, Johansen, M., Olesen, H.P. Macromolecular prodrugs. XV. Colon-targeted delivery-bioavailability of naproxen from orally administered dextran-naproxen ester prodrugs varying in molecular size in the pig. Pharm. Res. 1989; 6:919-23), daunomycin (Mann, J.S., Huang, J.C., Keana, J.F. Molecular amplifiers: synthesis and functionalization of a poly(aminopropyl)dextran bearing a uniquely reactive terminus for univalent attachment to biomolecules. Bioconjug. Chem. 1992; 3:154-9), adriamycin (Munechika, K; Sogame, Y.; Kishi, N.; Kawabata, Y.; Ueda, Y.; Yuamanouchi, K.; Yokoyama, K. Tissue distribution of macromolecular conjugate, adriamycin linked to oxidized dextran, in rat and mouse bearing tumor cells. Biol. Pharm. Bull. 1994; 17:1193-8) and others.

Drugs may be attached directly to dextran through a linker containing carboxylic acid residues. Various active esters of ¹²⁵IUdR and other derivatives reactive toward aldehydes are already available in our laboratories and were successfully used in the preparation of protein-IUdR conjugates. The periodate oxidation of dextran produced an aldehyde that readily reacts with drugs containing amino groups to form a Schiff base. We have several IUdR derivatives containing amino acids and small peptides in their structure. This intermediate can be used directly in the preparation of the dextran IUdR conjugates via a Schiff base. The flexibility of radioiodination based on our method of radiohalodestannylation (Baranowska-Kortylewicz, J. U.S. Patent 5,468,853; Methods for making 5- radiohalo-2'-deoxyuridines. 1995; Baranowska-Kortylewicz, J., Helseth,

L.D., Lai, J., Schneiderman, M.H., Schneiderman, G.S., Dalrymple, G.V. Radiolabeling kit/generator for 5-radiohalogenated uridines. J. Labelled Compds. Radiopharm. 1994; 34:513-521) allows for the introduction of the radioiodine during the last step of the synthetic scheme. This approach facilitates preparation of no-carrier-added derivatives.

Having described the invention with reference to particular compositions, theories of effectiveness, and the like, it will be apparent to those of skill in the art that it is not intended that the invention be limited by such illustrative embodiments or mechanisms, and that modifications can be made without departing from the scope or spirit of the invention, as defined by the appended claims. It is intended that all such obvious modifications and variations be included within the scope of the present invention as defined in the appended claims. The claims are meant to cover the claimed components and steps in any sequence which is effective to meet the objectives there intended, unless the context specifically indicates to the contrary.

Claims

What is claimed is:

1. A method of detecting the presence of tumor cells comprising: administering an effective amount of a prodrug of a desired radiopharmaceutical wherein the prodrug will be converted into the desired active radiopharmaceutical in or adjacent to desired targeted cells and measuring the uptake of the radiopharmaceutical into the cells.

2. The method of claim 1 wherein the desired active radiopharmaceutical is 5-[¹²⁵I]iodo-2'-deoxyuridine.

3. The method of claim 1 wherein the prodrug is a glycoside derivative of 5-[¹²⁵I]iodo-2'-deoxyuridine.

4. The method of claim 1 wherein the desired active radiopharmaceutical contains a radioisotope selected from the group consisting of iodine-131, iodine-123, iodine-124 and all radioisotopes of bromine.

5. The method of claim 4 wherein the prodrug is [¹²³I]iodo-2'- deoxyuridine-β-D-glycoside and the amount is about 0.1 mCi to about 20 mCi.

6. The method of claim 4 wherein the prodrug is [⁷⁷Br]bromo-2'- deoxyuridine-β-D-glycoside and the amount is about 0.1 mCi to about 20 mCi.

7. The method of claim 4 wherein the prodrug is [^80mBr]bromo-2'- deoxyuridine-β-D-glycoside and the amount is about 0.1 mCi to about 20 mCi.

8. The method of claim 4 wherein the prodrug is [¹³¹I]iodo-2'- deoxyuridine-β-D-glycoside and the amount is about 3 mCi.

9. The method of claim 4 wherein the prodrug is [¹²⁴I]iodo-2'- deoxyuridine-β-D-glycoside and the amount is about 3 mCi.

10. The method of claim 1 wherein the prodrug is administered orally or rectally.

11. The method of claim 1 wherein the desired targeted cells are cells suspected of being tumor cells.

12. The method of claim 1 wherein the desired targeted cells are in the gastrointestinal tract.

13. A method of selectively targeting and killing tumor cells comprising: administering an effective amount of a prodrug of a desired cytotoxic pharmaceutical wherein the prodrug will be converted into the desired active cytotoxic pharmaceutical in or adjacent to desired targeted cells.

14. The method of claim 13 wherein the cytotoxic pharmaceutical is a r adiop harm ace utical .

15. The method of claim 14 wherein the radiopharmaceutical is 5- [¹²⁵I]iodo-2'-deoxyuridine.

16. The method of claim 15 wherein the prodrug is a glycoside derivative of 5-[¹²⁵I]iodo-2'-deoxyuridine.

17. The method of claim 14 wherein the desired radiopharmaceutical contains a radioisotope selected from the group consisting of Auger-electron- emitting halide isotopes.

18. The method of claim 17 wherein the of Auger-electron-emitting halide isotopes are selected from the group consisting of bromine-77, bromine-80m, iodine-123, iodine-131, and astatine-211.

19. The method of claim 16 wherein the prodrug is [¹²⁵I]iodo-2'- deoxyuridine-β-D-glycoside and the amount of prodrug is about 1 mCi to about 50 mCi.

20. The method of claim 18 wherein the prodrug is [⁷⁷Br]bromo-2'- deoxyuridine-β-D-glycoside and the amount of prodrug is about 1 mCi to about 50 mCi.

21. The method of claim 18 wherein the prodrug is [^80mBr]bromo-2'- deoxyuridine-β-D-glycoside and the amount of prodrug is about 1 mCi to about 50 mCi.

22. The method of claim 18 wherein the prodrug is [¹²³I]iodo-2'- deoxyuridine-β-D-glycoside and the amount of prodrug is about 1 mCi to about 50 mCi.

23. The method of claim 18 wherein the prodrug is [²¹¹At]astato-2'- deoxyuridine-β-D-glycoside and the amount of prodrug is about 10 mCi.

24. The method of claim 13 wherein the desired targeted cells are cells suspected of being tumor cells.

25. The method of claim 13 wherein the desired targeted cells are in the gastrointestinal tract.

26. The method of claim 25 wherein the cells are in the large intestine.

27. The method of claim 13 wherein the prodrug is administered to a mucosal surface.

28. The method of claim 27 wherein the prodrug is administered orally or rectally.

29. A method for preventing recurrence of a tumor comprising administering an effective amount of a prodrug of a desired cytotoxic pharmaceutical wherein the prodrug will be converted into the desired active cytotoxic pharmaceutical in or adjacent to desired targeted cells.

30. The method of claim 29 wherein the cytotoxic pharmaceutical is a radiopharmaceutical.

31. The method of claim 29 wherein administration is peri- and/or post- surgical.

32. A composition for treating, preventing the recurrence of, and/or diagnosing cancer comprising an effective amount of a glycoside derivative of 5-iodo-2'-deoxyuridine.

33. The composition of claim 32 wherein the 5-iodo-2'-deoxyuridine is a radioisotope of 5-iodo-2'-deoxyuridine.

34. The composition of claim 33 wherein the radioisotope of 5-iodo-2'- deoxyuridine is 5-[¹²⁵I]iodo-2'-deoxyuridine.

35. The composition of claim 32 further comprising a pharmaceutically acceptable carrier.

36. The composition of claim 35 wherein the pharmaceutically acceptable carrier is a carrier suitable for oral or rectal administration.

37. The composition of claim 35 wherein the composition is in a dosage form selected from the group consisting of capsules, tablets, suppositories, troches, suspensions, and liquids.

38. A method for delivery of radiosensitizers to abnormally proliferating cells in the liver comprising administering an effective amount of a prodrug of a desired radiopharmaceutical wherein the prodrug will be converted into the desired active radiopharmaceutical in the liver.

39. The method of claim 38 wherein the amount of prodrug 1 g to 100 g and wherein the administration is oral.

40. The method of claim 39 wherein the prodrug is added to a liquid for administration.