Hematopoietic gene therapy restores thymidine phosphorylase activity in a cell culture and a murine model of MNGIE.

Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE) is an autosomal recessive disorder caused by mutations in the TYMP gene, which encodes thymidine phosphorylase (TP). TP dysfunction results in systemic thymidine (dThd) and deoxyuridine (dUrd) overload, which selectively impair mitochondrial DNA replication. Allogeneic hematopoietic transplantation has been used to treat MNGIE patients; however, this approach has serious adverse effects, including the toxicity of myeloablative conditioning, graft rejection and graft-versus-host disease. With the aim of testing the feasibility of gene therapy for MNGIE, we transduced TP-deficient B-lymphoblastoid cells from two MNGIE patients, with lentiviral vectors carrying a functional copy of the human TYMP DNA coding sequence. This restored TP activity in the cells, which reduced the excretion of dThd and dUrd and their concentrations when added in excess. Additionally, lentiviral-mediated hematopoietic gene therapy was used in partially myeloablated double Tymp/Upp1 knockout mice. In spite of the relatively low levels of molecular chimerism achieved, high levels of TP activity were observed in the peripheral blood of the transplanted mice, with a concomitant reduction of nucleoside concentrations. Our results suggest that hematopoietic gene therapy could be an alternative treatment for this devastating disorder in the future.

p53-dependent suppression of uridine phosphorylase gene expression through direct promoter interaction.

Uridine phosphorylase (UPase) is a key enzyme in the pyrimidine salvage pathway. It reversibly catalyzes the catabolism of uridine to uracil; controls the homeostatic regulation of uridine concentration in plasma and tissues; and plays a role in the intracellular activation of 5-fluorouracil. We cloned the murine UPase gene promoter, a 1703-bp fragment, and determined the transcription initiation sites located at +1 and +92 bp of the cDNA sequence. Through transient expression analysis of the 5'-flanking region of UPase gene, we have evaluated the promoter activity for a series of fragments with 5'- to 3'-deletion in murine breast cancer EMT-6 cells and immortalized murine fibroblast NIH 3T3 cells. Cotransfection of the UPase promoter constructs (from -1619 to -445) containing p53 binding motif with the wild-type p53 construct resulted in a significant reduction of luciferase activity; however, this effect disappeared with the additional deletion of the -445 to -274 sequence to suggest the existence in this promoter region of a putative p53 recognition element. Similar cotransfection in murine embryo fibroblasts p53-/- confirmed the inhibitory role of p53 on the UPase promoter activity. The specificity of the interaction is demonstrated by nuclear protein-specific binding to the putative p53 recognition sequence using gel mobility shift assay and DNase I footprinting analysis. These data indicate the UPase gene is a novel target of p53, and its expression is down-regulated by p53 at the promoter level.

(6R)-5,10-Dideaza-5,6,7,8-tetrahydrofolic acid effects on nucleotide metabolism in CCRF-CEM human T-lymphoblast leukemia cells.

(6R)-5,10-Dideaza-5,6,7,8-tetrahydrofolic acid [(6R)DDATHF] is a folate antimetabolite with activity specifically directed against de novo purine synthesis, primarily through inhibition of glycinamide ribonucleotide transformylase. This inhibition resulted in major changes in the size of the nucleotide pools in CCRF-CEM cells. After a 4-h incubation with 1 microM (6R)DDATHF, dramatic reductions in the ATP and GTP pools were observed, with almost no effect on CTP, UTP, and deoxyribonucleotide pools. When the incubation was continued in drug-free medium, recovery of ATP and GTP pools was protracted. ATP did not return to normal until 24-36 h, and GTP pools were only partially repleted by 48 h. The ATP and GTP pools were not affected when the initial 4-h incubation with (6R)DDATHF was conducted in the presence of 100 microM hypoxanthine. Addition of hypoxanthine to the medium after a 4-h incubation with (6R)DDATHF caused rapid recovery of the ATP and GTP pools. Similar effects were seen when the purine precursor aminoimidazole carboxamide was used in place of hypoxanthine. The effect of (6R)DDATHF on nucleotide pools and the capability of hypoxanthine or aminoimidazole carboxamide to prevent or reverse this phenomenon correlated directly with the inhibition of cell growth. Presumably as a consequence of the decrease in purine nucleotide triphosphate levels, the conversion of exogenously added uridine, thymidine, and deoxyuridine to nucleotides was markedly decreased. These effects were protracted for almost 48 h and were also reversed by hypoxanthine. Differential repletion of ATP and GTP pools after (6R)DDATHF pre-treatment demonstrated that diminished precursor phosphorylation is primarily a consequence of GTP rather than ATP starvation.

Expression, characterization, and detection of human uridine phosphorylase and identification of variant uridine phosphorolytic activity in selected human tumors.

Uridine phosphorylase (UPase) catalyzes the reversible phosphorolysis of uridine to uracil. We purified the enzyme from the murine colon 26 tumor using a two-step procedure through 5-amino-benzylacyclouridine affinity chromatography. Antibodies raised in rabbits against the purified protein revealed single bands in Western blots of normal human tissue and tumor extracts. The polyclonal antibody used to screen a human liver expression library allowed the isolation of a 1.2-kb clone that contained the entire open reading frame of the human UPase. The UPase cDNA has been expressed as a fusion protein in Escherichia coli using the pMal-C2 vector. The kinetic analysis demonstrated that the recombinant UPase preferentially uses uridine, 5-fluorouracil, and uracil as substrates, although lower levels of activity were observed with 2-deoxyuridine and thymidine. Clinical samples of human tumors and adjacent normal tissues were assayed for phosphorolytic activity and sensitivity to 5-benzylacyclouridine (BAU), a potent inhibitor of the enzyme presently in Phase I-II clinical trial. Activity in normal tissues appeared to be low but very sensitive to BAU (approximately 90% inhibition at 10 microM). Tumors had generally 2-3-fold greater activity compared with adjacent normal tissues. In breast cancer specimens and head-neck squamous carcinomas, however, uridine cleavage was only partially inhibited (40-60%) by 10 or 100 microM BAU. The BAU-insensitive activity requires phosphate and pH conditions similar to the normal enzyme, and the new phosphorolytic activity was independent from thymidine phosphorylase. The BAU-insensitive phosphorolytic activity in selected tumors, coupled with the potent inhibitory activity of BAU against the "classical" uridine phosphorylase in normal human tissues, provides the rationale for combining BAU with 5-fluorouracil in the treatment of breast and head-neck tumors.

Inherent resistance of human squamous carcinoma cell lines to methotrexate as a result of decreased polyglutamylation of this drug.

Three human squamous carcinoma cell lines (FaDu, A253, and SQCC/Y1) were tested for sensitivity to methotrexate (MTX) and trimetrexate, a second generation folate antagonist in clinical trials. Two of the three cell lines (A253 and SQCC/Y1) showed inherent resistance to methotrexate, when cytotoxicity was evaluated after short term exposure (4 and 24 h). In contrast, all three cell lines were markedly sensitive to trimetrexate, an antifolate which is taken up by cells by a different transport system than is methotrexate and which is not polyglutamylated. The basis for the natural resistance to methotrexate shown by two of the three cell lines was examined. Levels of dihydrofolate reductase activity, inhibition of this enzyme by methotrexate, and influx of MTX did not differ significantly between the three cell lines; however, resistance was correlated to the amounts of polyglutamates of methotrexate synthesized by the three cell lines. After a 24-h incubation with 10 microM MTX, the A253 cell line was able to form only 35.0 pmol/10(7) cells of polyglutamates, compared to 250 pmol/10(7) cells synthesized by the FaDu cell line, while the SQCC/Y1 cell line, intermediate in sensitivity to methotrexate, was able to form 145 pmol/10(7) cells of MTX polyglutamates. The A253 cell line contained less folylpolyglutamate synthetase activity compared to the FaDu and SQCC/Y1 cell lines. However, it is not clear if this difference is sufficient to explain the marked differences in polyglutamates of methotrexate found between the cell lines. We conclude that decreased polyglutamylation of methotrexate in some human squamous cell carcinomas may be the major contributing factor in inherent resistance to high dose pulse administration of this drug.

Brequinar potentiates 5-fluorouracil antitumor activity in a murine model colon 38 tumor by tissue-specific modulation of uridine nucleotide pools.

Modulation of pyrimidine metabolism or the metabolic fate of 5-fluorouracil by a number of different agents has permitted a significant increase in the response rate to this agent, particularly for colorectal cancers. Brequinar, a noncompetitive inhibitor of mitochondrial dihydroorotate dehydrogenase has been shown to achieve a tumor-specific modulation of the therapeutic effect of 5-fluorouracil. A selective decrease of uridine nucleotide pools in Colon tumor 38 compared to normal tissues of C57/BL6 mice was observed after Brequinar administration. This effect was achieved with very low nontherapeutic doses of Brequinar (8 to 27% of the maximum tolerated dose in this model). Pretreatment with Brequinar 4 and 24 h prior to administration of [3H]fluorouracil significantly increased incorporation of the fluoropyrimidine into Colon 38 tumor RNA, while minimal effects were seen in normal tissues of C57/BL6 mice. Brequinar (15, 30, and 50 mg/kg) was administered 4 h prior to fluorouracil (85 mg/kg) on a weekly basis in Colon 38-bearing mice. All combinations potentiated 5-fluorouracil antitumor activity and the lowest dose of Brequinar (15 mg/kg) showed a reduced toxicity (weight loss) compared to the same dose of 5-fluorouracil as a single agent. When Brequinar preceded fluorouracil by 24 h, greater toxicity and less antitumor activity were observed. A comparison of the optimal Brequinar-fluorouracil regimen with a previously optimized N-(phosphonoacetyl)-L-aspartic acid-fluorouracil combination in Colon 38 tumor indicated that Brequinar-fluorouracil was more effective and less toxic.

Genomic structure, chromosomal mapping, and promoter region analysis of murine uridine phosphorylase gene.

Uridine phosphorylase (UPase) plays an important role in the activation of 5-fluorouracil and in the regulation of tissue and plasma concentration of uridine, a potential biochemical modulator of 5-fluorouracil therapy. UPase expression is affected by the c-H-ras oncogene and various cytokines through unknown mechanisms. To understand its expression and regulation, we cloned the murine UPase gene, defined its genomic organization, determined its 5'- and 3'-end flanking sequences, and evaluated the promoter activity. The UPase gene contains nine exons and eight introns, spanning a total of approximately 18.0 kb. Its promoter lacks canonical TATA and CCAAT boxes, although a CAATAAAAA TATA-like box is seen from -41 to -49. Furthermore, IFN regulatory factor 1, c/v-Myb, and p53 binding sites are present in the promoter region, indicating that UPase expression may be directly regulated by cytokines and oncogene products. The 1.2-kb flanking fragment showed promoter activity driving the expression of the luciferase gene in various mammalian cells. A TGGGG repeat sequence is seen in the 3'-end flanking region. This element is considered to be a potential recombination consensus hot spot that may contribute to the encoding of different UPase isoforms present in different tissues, both normal and neoplastic.

In vivo effect of 5-ethynyluracil on 5-fluorouracil metabolism determined by 19F nuclear magnetic resonance spectroscopy.

Biochemical modulation of 5-fluorouracil (5-FU) has been used over the past 20 years to improve the therapeutic efficacy of this antineoplastic agent. Recently, modulation of the catabolic pathway of this fluoropyrimidine has been the focus of extensive preclinical and clinical investigation. Dihydropyrimidine dehydrogenase catalyzes the rate-limiting step in the catabolism of 5-FU and rapidly degrades 60-90% of the drug. An irreversible inactivating inhibitor of this enzyme, 5-ethynyluracil (EU), markedly improves the antitumor effect of 5-FU in animal models. Early clinical studies have shown a substantial alteration of the systemic disposition of 5-FU with an increase in 5-FU terminal half-life and have also indicated that EU allows safe oral administration of 5-FU by improving the oral bioavailability of the fluoropyrimidine, which is otherwise too erratic and unpredictable for a drug with such a limited therapeutic window. We evaluated the effect of EU on the metabolism of 5-FU in mice bearing colon 38 tumors using 19F nuclear magnetic resonance spectroscopy. Ex vivo measurements of tissue extracts from liver, kidney, and tumor indicated a >95% elimination of alpha-fluoro-beta-ureidopropionic acid and a-fluoro-beta-alanine signals in the tissues of mice that received 2 mg/kg of EU before administration of 5-FU. The spectra also showed an increased formation of fluoronucleotides in both normal and tumor tissues, a prolonged presence of 5-FU, and the accumulation of 5-fluorouridine that otherwise is undetectable, particularly in normal tissues. The in vivo NMR experiments on colon 38 tumors confirmed these findings, showing a complete elimination of the a-fluoro-beta-ureidopropionic acid and a-fluoro-beta-alanine signals in tumors treated with EU and a dramatic formation and accumulation of 5-fluorouridine mono-, di-, and triphosphates and 5-fluorouridine. Thus, by inactivating dihydropyrimidine dehydrogenase, EU prolonged the half-life for 5-FU, almost completely eliminated its catabolism for 4-6 h, which led to an increased accumulation of 5-fluorouridine mono-, di-, and triphosphates in both normal and tumor tissues.

Impaired polyglutamylation of methotrexate as a cause of resistance in CCRF-CEM cells after short-term, high-dose treatment with this drug.

Two methotrexate-resistant sublines, CCRF-CEM R3/7 and CCRF-CEM R30/6, were selected from the human leukemia T-lymphoblast cell line, CCRF-CEM, after repeated exposures (7 and 6 times, respectively) for 24 h to constant concentrations (3 and 30 microM) of the drug. Analysis of the mechanism of resistance revealed no differences in levels of dihydrofolate reductase activity, its binding affinity for methotrexate, or in methotrexate transport between the CCRF-CEM parent and methotrexate-resistant cell lines. The development of resistance to methotrexate was associated with a marked decrease in the intracellular level of methotrexate polyglutamates. Although the resistant sublines were able to form substantial amounts of folate polyglutamates when measured with [3H]folic acid, the level of polyglutamates formed was decreased to about 50% of that formed by the parent cell line. No qualitative differences in folate polyglutamates formed were noted between the parental and resistant sublines. This is the first example of a cell line which displays resistance which is solely attributable to defective methotrexate polyglutamate synthesis.

Multifactorial resistance to 5,10-dideazatetrahydrofolic acid in cell lines derived from human lymphoblastic leukemia CCRF-CEM.

5,10-dideaza-5,6,7,8-terrahydrofolic acid (DDATHF) is a potent antiproliferative agent in cell culture systems and in vivo in a number of murine and human xenograft tumors. In contrast to classical antifolates, which are dihydrofolate reductase inhibitors, DDATHF primarily inhibits GAR transformylase, the first folate-dependent enzyme along the pathway of de novo purine biosynthesis. The (6R) diastereomer of DDATHF (Lometrexol), currently undergoing clinical investigation, was used to develop CCRF-CEM human leukemia sublines resistant to increasing concentrations of the drug. Three cell lines were selected for ability to grow in medium containing 0.1 microM, 1.0 microM, and 10 microM of (6R)DDATHF, respectively. Impaired polyglutamylation was identified as a common mechanism of resistance in all three cell lines. A progressive decrease in the level of polyglutamylation was associated with diminished folylpolyglutamate synthetase activity and paralleled increasing levels of resistance to the drug. However, the expression of folylpolyglutamate synthetase RNA was not altered in the resistant cell lines compared to the parent cells. The most resistant cell subline also displayed an increased activity of gamma-glutamyl hydrolase. The sublines were scrutinized for other possible mechanisms of resistance. No alterations in drug transport or in purine economy were found. Modest increases were found in the activity of methylene tetrahydrofolate dehydrogenase but no alterations of other folate-dependent enzymes were observed. Increases in accumulation and conversion of folic acid to reduced forms, particularly 10-formyltetrahydrofolate, was also seen. The resistant cell lines were sensitive to dihydrofolate reductase inhibitors, methotrexate and trimetrexate, for a 72-h exposure period but showed cross-resistance to methotrexate for 4 and 24 h exposures. Cross-resistance was also shown toward other deazafolate analogues for both short- and long-term exposures.

The use of the L-plastin promoter for adenoviral-mediated, tumor-specific gene expression in ovarian and bladder cancer cell lines.

A 2.4-kb truncated L-plastin promoter was inserted either 5' to the LacZ gene (Ad-Lp-LacZ) or 5' to the cytosine deaminase (CD) gene (Ad-Lp-CD) in a replication-incompetent adenoviral vector backbone. Infectivity and cytotoxicity experiments with the LacZ and CD vectors suggested that the L-plastin promoter-driven transcriptional units were expressed at much higher levels in explants of ovarian cancer cells from patients and in established ovarian or bladder cancer cell lines than they were in normal peritoneal mesothelial cells from surgical specimens, in organ cultures of normal ovarian cells, or in the established CCD minimal deviation fibroblast cell line. Control experiments showed that this difference was not attributable to the lack of infectivity of the normal peritoneal cells, the normal ovarian cells, or the minimal deviation CCD fibroblast cell line, because these cells showed expression of the LacZ reporter gene when exposed to the replication-incompetent adenoviral vector carrying the cytomegalovirus (CMV)-driven LacZ gene (Ad-CMV-LacZ). The Ovcar-5 and Skov-3 ovarian cancer cell lines exposed to the Ad-Lp-CD adenoviral vector were much more sensitive to the prodrug 5-fluorocytosine (5FC), which is converted from the 5FC prodrug into the toxic chemical 5-fluorouracil, than was the CCD minimal deviation fibroblast cell line after exposure to the same vector. A mouse xenograft model was used to show that the Ad-Lp-CD vector/5FC system could prevent engraftment of ovarian cancer cells in nude mice. Finally, injection of the Ad-Lp-CD vector into s.c. tumor nodules generated a greater reduction of the size of the tumor nodules than did injection of the Ad-CMV-LacZ vectors into tumor nodules. The Ad-Lp-CD vectors were as suppressive to tumor growth as the Ad-CMV-CD vectors. These results suggest that an adenoviral vector carrying the CD gene controlled by the L-plastin promoter (Ad-Lp-CD) may be of potential value for the i.p. therapy of ovarian cancer.

Dose-escalation and pharmacodynamic study of topotecan in combination with cyclophosphamide in patients with refractory cancer.

PURPOSE: Based on preclinical data that demonstrated synergy between alkylating agents and topoisomerase (topo) I poisons, we determined the maximum-tolerated dose (MTD) of topotecan, using a 5 day bolus schedule, that could be given in combination with a single, fixed dose of cyclophosphamide. Pharmacodynamics of this combination were explored by analyzing biochemical effects of treatment in peripheral-blood mononuclear cells (PBMCs). PATIENTS AND METHODS: Patients with refractory cancer were treated with cyclophosphamide 600 mg/m2 on day 1, followed by topotecan given as a 30-minute infusion for 5 consecutive days. Cycles were repeated every 3 weeks. Once the MTD was defined, granulocyte colony-stimulating factor (G-CSF) was added to the regimen in an attempt to escalate further the dose of topotecan. Plasma concentrations of topotecan were determined during the first treatment cycle by high-performance liquid chromatography. PBMCs were sampled at baseline and throughout the 5-day treatment period for analysis of topo I protein concentrations and to determine drug-induced DNA fragmentation. RESULTS: Twenty-six patients were treated with topotecan at doses that ranged from 0.5 mg/m2/d to 1.2 mg/ m2/d for a total of 74 cycles. Reversible neutropenia was dose-limiting, with mild to moderate suppression of the other blood-cell elements commonly occurring. Transfusions of RBCs and platelets were required in 24% and 7% of treatment cycles, respectively. The most prominent nonhematologic toxicities were fatigue and weight loss. Compared with previously published data in which topotecan was administered alone, cyclophosphamide did not appear to alter the pharmacokinetics of topotecan. Significant increases in topo I concentration were identified in PBMCs following the administration of cyclophosphamide on day 1 and there was a significant decrease in topo 1 during the 5-day course of treatment (P < .01, sign test). DNA fragmentation as a result of drug treatment was identified in 11 of 15 (73%) cycles analyzed. CONCLUSION: For previously treated patients, the recommended dose of topotecan in this schedule is 0.75 mg/m2/d without growth factor support and 1.0 mg/ m2/d if it is administered with G-CSF. Biochemical changes in cells induced by exposure to camptothecins can be measured in vivo and these effects may have important implication in the design of combination therapies and the optimal scheduling of this class of agents.

Pharmacokinetic and phase I trial of intraperitoneal carboplatin and cyclosporine in refractory ovarian cancer patients.

PURPOSE: The feasibility and pharmacokinetics of cyclosporine (CsA) delivered intraperitoneally (IP) have not been previously explored. We performed a pharmacokinetic study of IP CsA followed by a phase I dose-escalation trial of the combination of IP CsA and carboplatin in refractory ovarian cancer patients. PATIENTS AND METHODS: A pilot study was performed of three patients who received 1, 10, and 20 mg/kg IP CsA alone. Subsequently, a phase I trial of 35 patients was performed between April 1990 and April 1993. Whole-blood and IP fluid CsA concentrations were measured at serial time points. The highest dose delivered IP was 34.6 mg CsA/kg in combination with carboplatin (250 mg/m2 or 300 mg/m2, depending on creatinine clearance), which was not dose-escalated. The area under the concentration-time curve (AUC) for CsA and half-life (T1/2) were calculated. Objective and serologic responses were noted, and toxicity was graded using the National Cancer Institute common toxicity criteria. RESULTS: The feasibility of delivering IP CsA alone was established. We observed a 1,000:1 ratio between IP fluid and blood concentrations at 20 mg CsA/kg. Pharmacokinetic analysis confirmed that at 20 mg CsA/kg, there was an IP fluid-to-blood AUC ratio of 600:1 in favor of peritoneal exposure. At the highest dose delivered, 34.6 mg CsA/kg, the mean IP CsA levels of 1,110 micrograms/ mL were tolerated moderately well and the IP fluid-to-blood ratio of 1,000:1 was maintained. Blood and IP CsA concentrations were analyzed in the presence and absence of IP carboplatin. At 20 mg CsA/kg, there was no difference in either mean blood CsA levels (0.9 microgram/ mL) or mean IP CsA concentrations (1,000 micrograms/mL) obtained in the absence or presence of carboplatin. The most common toxicity in the phase I study was anemia, seen in 66% of patients. Common toxicities at the maximum CsA dose delivered (34.6 mg/kg) were anemia, leukopenia, thrombocytopenia, and hypertension. In this trial, three objective responses (two complete and one partial) were observed for a duration of 3 to 11 months. Control of platinum-resistant ascites was an important feature, noted in five of eight patients. CONCLUSION: We have established the feasibility of delivering IP CsA up to doses of 34.6 mg/kg in conjunction with carboplatin, and the sustaining of IP fluid to blood ratios of 1,000:1. The IP administration of CsA resulted in a favorable ratio of exposure for the peritoneal cavity compared with systemic exposure, indicating a therapeutic advantage of this approach with a significant decrease in systemic toxicity. We recommend that 34.6 mg/ kg of IP CsA be tested as a phase II dose in combination with carboplatin in refractory ovarian cancer patients. This report provides the groundwork for future studies using IP CsA, both as a chemomodulator of platinum and of multidrug resistance.

Percutaneous hepatic vein isolation and high-dose hepatic arterial infusion chemotherapy for unresectable liver tumors.

PURPOSE: This prospective, nonrandomized trial evaluated a percutaneous isolated chemotherapy perfusion approach for treating advanced primary and metastatic liver tumors. Chemotherapy was administered via hepatic artery catheter and hepatic venous blood isolated by a novel percutaneous double-balloon inferior vena cava (IVC) catheter was passed through a detoxification/filtration cartridge in a venovenous bypass circuit. PATIENTS AND METHODS: Among 23 patients enrolled onto the study, 58 procedures were performed on 21 patients. Twelve patients received dose escalations of fluorouracil (5-FU) (1,000 mg/m2 to 5,000 mg/m2), and nine received dose escalations of doxorubicin (50 mg/m2 to 120 mg/m2). Pharmacokinetic studies included drug accumulation in the liver, extraction by detoxification filters, systemic exposure, and alterations of half-life. Each patient received two treatments at 3-week intervals. Those showing stabilization or response received additional treatments. RESULTS: There was a direct relationship between dose and peak concentration of drug entering the hepatic veins. The system functioned efficiently throughout the dose range, with extraction efficiencies ranging from 64% to 91% (P < .001). The hepatic vein drug levels showed a sixfold increase in 5-FU with dose escalation from 1,000 to 5,000 mg/m2, and a twofold increase in dox with dose escalation from 50 to 120 mg/m2 (P < .001, filter-mediated drug extraction). The treatments were accomplished with only an overnight hospital stay and no mortality. The common procedure-related toxicity was transient hypotension (grade I to II), due to catecholamine depletion by the filter. Dose-limiting toxicity (leukopenia) was observed in patients receiving 5-FU at a dose of 5,000 mg/m2 and doxorubicin at a dose of 120 mg/m2. Significant tumor response (> 95% reduction) was obtained in two patients receiving doxorubicin at 90 mg/m2 and 120 mg/m2. CONCLUSION: The use of a double-balloon catheter to isolate and detoxify hepatic venous blood during intraarterial therapy is technically feasible, safe, and allows administration of large doses of intrahepatic chemotherapy at short intervals. This approach should allow new dose-intensification strategies to increase tumor responses in primary and metastatic liver tumors.

Phase I clinical and pharmacological studies of benzylacyclouridine, a uridine phosphorylase inhibitor.

Benzylacyclouridine (BAU, IND 039655) is a potent and specific inhibitor of uridine phosphorylase (UrdPase; EC 2.4.2.3). This enzyme plays a major role in regulating uridine homeostasis and also catalyzes the conversion of fluoropyrimidine nucleosides to their respective bases. Inhibition of UrdPase enzyme activity 18-24 h after 5-fluorouracil (5-FU) administration increased plasma levels of uridine and enhanced the therapeutic index of 5-FU by rescuing normal tissues. Moreover, in vitro preclinical studies have also shown that inhibiting UrdPase enzyme activity by BAU prior to administration of 5-FU increased cytotoxicity in a number of human cancer cell lines. A series of preclinical studies was performed in dogs and pigs to evaluate the pharmacological and pharmacodynamic properties of BAU. These data showed a sustained elevation in plasma uridine concentration in both animal models. The rapid degradation of a tracer dose of uridine into uracil was virtually arrested by BAU administered both p.o. or i.v. The t1/2 of BAU was 1.8-3.6 h in dogs, with bioavailability levels of 85% (30 mg/kg) and 42.5% (120 mg/kg). In pigs, the half-life varied from 1.6 to 2.3 h, with a bioavailability of 40% at 120 mg/kg. The drug was distributed into most tissues with a tissue: plasma ratio of approximately 0.7. On the basis of these preclinical studies, we performed a Phase I clinical trial of BAU in patients with advanced cancer. Patients received 200, 400, 800, and 1600 mg/m2 BAU as a single oral dose. Toxicities included grade 2 anemia, grade 1 fever, grade 1 fatigue, grade 1 constipation, and grade 1 elevation in alkaline phosphatase; none of these toxicities were observed to be dose dependent. The maximum tolerated dose and dose-limiting toxicity were not reached at the doses given. BAU plasma concentrations and area under the curve correlated linearly with the oral dose level. The pharmacokinetics of BAU were consistent with a first-order clearance, with average peak concentrations ranging from 19 microM (200 mg/m2) to 99 microM (1600 mg/m2) and tbeta1/2 ranging from 3.0 to 3.9 h at the four dose levels. Compared with baseline plasma uridine, treatment of patients with 200, 400, 800, and 1600 mg/m2 BAU increased peak uridine concentrations by 120, 150, 250, and 175%, respectively. On the basis of this clinical study, the suggested Phase II starting dose of BAU in combination with 5-FU is 800 mg/m2. Studies combining BAU with 5-FU and incorporating appropriate molecular and biochemical end points to assess the effects of this drug combination on tumor and/or surrogate tumor tissue are under way.

Use of L-plastin promoter to develop an adenoviral system that confers transgene expression in ovarian cancer cells but not in normal mesothelial cells.

The objective of this study was to develop an adenoviral vector system that would generate a pattern of expression of exogenous therapeutic genes appropriate for the treatment of ovarian cancer. For this purpose, we have generated a replication-deficient recombinant adenoviral vector, AdLPLacZ, which contains the human L-plastin (LP) promoter (LP-P) driving the Escherichia coli LacZ gene. LP is constitutively expressed at high levels in malignant epithelial cells but is not expressed in normal tissues, except at low levels in mature hematopoietic cells. Because adenoviral vectors infect early hematopoietic multilineage precursor cells only poorly or not at all, this vector would be of use in the peritoneal cavity and in vitro for marrow purging. We first analyzed the expression of the LacZ reporter gene in ovarian and breast cancer cell lines, normal fibroblasts, and leukemia cell lines using the adenoviral vector in which the LacZ gene is governed by the LP-P promoter (AdLPLacZ) or in which the LacZ gene is governed by the cytomegalovirus (CMV) promoter (AdCMVLacZ). We found equivalent and high levels of expression of beta-galactosidase (beta-gal) by AdLPLacZ and AdCMVLacZ vectors in the breast or ovarian cancer cell lines as well as in a fibrosarcoma cell line, indicating that the adenoviral vectors infected these cells and expressed their transgenes equally with the LP and CMV promoters. Expression of the LacZ gene with the CMV vector but not with the LP-P vector was observed in experiments with normal fibroblasts, indicating that the vectors infected the cells, but that the LP-P was not active within them. In hematopoietic cells such as U937 cells, no measurable beta-gal activity was detected in cells infected either by AdLPLacZ or by AdCMVLacZ, indicating that the adenoviral vectors were not infecting the cells. Although beta-gal activity was observed in fresh ascitic ovarian cancer cells after infection with adenoviral vectors containing CMV or the LP promoters, beta-gal activity was detected in a portion of a biopsy of normal peritoneum when the tissues were exposed to the AdCMVLacZ vector, but not when tissues were exposed to the AdLPLacZ vector. These results suggest that the transcription of therapeutic genes in cells infected by the AdLP vectors would be restricted to LP expression-positive ovarian carcinoma cells but would not be seen in the normal mesothelial cells of the peritoneal cavity. This possibility implies that adenoviral vectors carrying therapeutic genes driven by the LP-P would be of use for the intracavitary treatment ovarian cancer.

Toxoplasma gondii tachyzoites possess an unusual plasma membrane adenosine transporter.

Nucleoside transport may play a critical role in successful intracellular parasitism by Toxoplasma gondii. This protozoan is incapable of de novo purine synthesis, and must salvage purines from the host cell. We characterized purine transport by extracellular T. gondii tachyzoites, focusing on adenosine, the preferred salvage substrate. Although wild-type RH tachyzoites concentrated [3H]adenosine 1.8-fold within 30 s, approx. half of the [3H]adenosine was converted to nucleotide, consistent with the known high parasite adenosine kinase activity. Studies using an adenosine kinase deficient mutant confirmed that adenosine transport was non-concentrative. [14C]Inosine, [14C]hypoxanthine and [3H]adenine transport was also rapid and non-concentrative. Adenosine transport was inhibited by dipyridamole (IC50 approx. 0.7 microM), but not nitrobenzylthioinosine (15 microM). Transport of inosine, hypoxanthine and adenine was minimally inhibited by 10 microM dipyridamole, however. Competition experiments using unlabeled nucleosides and bases demonstrated distinct inhibitor profiles for [3H]adenosine and [14C]inosine transport. These results are most consistent with a single, dipyridamole-sensitive, adenosine transporter located in the T. gondii plasma membrane. Additional permeation pathways for inosine, hypoxanthine, adenine and other purines may also be present.

Effect of clinically modeled regimens on the growth response and development of resistance in human colon carcinoma cell lines.

Two human colon cell lines, HCT-8 and HT-29, were exposed to 5-fluorouracil (FUra) under conditions similar to the human plasma pharmacokinetic profile achieved by a single bolus dose or a sustained i.v. infusion. The bolus treatment for 5 days caused a substantial cell kill; however, only a moderate inhibition in cell growth was obtained with sustained exposure to the clinically relevant level of 2 microM. To achieve a cell kill equivalent to the bolus method, a sustained concentration of 10 microM was required. This would constitute a 60% increase in the total area under the curve (AUC) compared with the bolus treatment. After three courses of therapy with each of the schedules, emerging cell lines displayed a similar degree of resistance. HT-29 resistant cell lines returned to the original sensitivity within a few weeks, and most of the enzymes involved in the metabolic activation of FUra returned to their pretreatment activities. However, resistance and enzymatic modifications remained in the HCT-8 line for at least 3 months. In the HCT-8 cell line derived from bolus treatment, resistance was associated with a 50-60% reduction in uridine kinase activity. In the line derived from continuous exposure, there was a 35-40% reduction in uridine kinase in addition to a greater reduction in the activity of orotate phosphoribosyltransferase. These changes in both resistant cell lines resulted in a decreased incorporation of [3H]FUra into nucleic acids and a reduced formation of di- and triphosphate nucleotides of FUra.

Aberrant cell cycle inhibition pattern in human colon carcinoma cell lines after exposure to 5-fluorouracil.

In this report, we describe the use of two human colon carcinoma cell lines, HCT-8 and HT-29, as potential models to study DNA- and RNA-directed cytotoxicity due to 5-fluorouracil (FUra) exposure by flow microfluorimetric analysis of DNA cell content. The sensitivity of the HT-29 line (EC50 = 0.9 microM) to FUra was somewhat greater than that of the HCT-8 line (EC50 = 4 microM), but each presented a dramatically different DNA histogram after exposure to FUra. In HCT-8, an unexpected and nearly complete disappearance of cells in S-phase occurred, whereas in HT-29 the expected accumulation of cells at the G1-S border was observed. The absence of HCT-8 cells in S-phase also occurred as a result of two RNA polymerase inhibitors: actinomycin D and dichloro-D-ribofuranosylbenzimidazole. However, an accumulation of cells in S-phase was observed in the presence of 5-fluorodeoxyuridine. These results suggest that in the HCT-8 cell line, FUra predominantly causes an RNA-related toxicity. By comparison, the rate of formation of 5-fluorodeoxyuridine monophosphate, the increased dUMP pool size, and low thymidylate synthase activity in the HT-29 line are consistent with its greater susceptibility to DNA-directed toxicity. Further evidence was seen in the prevention of FUra cytotoxicity by thymidine in HT-29, but not in HCT-8 cells. Similarly, Leucovorin synergized the action of FUra in HT-29 but not in HCT-8. Enzymatic correlates supporting these observations are seen in the greater activity of uridine kinase than thymidine kinase (20:1) in HCT-8 cells compared with that in HT-29 cells (4:1).

Enhancement of antineoplastic activity of 5-fluorouracil in mice bearing colon 38 tumor by (6R)5,10-dideazatetrahydrofolic acid.

(6R)5,10-Dideazatetrahydrofolic acid (DDATHF, Lometrexol), a potent antitumor drug in vivo and in vitro, is an inhibitor of the two folate-dependent enzymes in the de novo purine biosynthesis pathway: glycinamide ribonucleotide (GAR) and amino imidazole carboxamide (AICAR) transformylases. A single dose of DDATHF (50 mg/kg, i.p.) in C57/BL6 mice caused a prolonged depletion of purine nucleotides (ATP and GTP) in colon 38 tumor and only a temporary effect in liver. GAR transformylase activity was higher in colon 38 tumor than in liver, but a kinetic analysis on the purified enzyme showed no differences in Ki values for DDATHF or Km values for the folate substrate. As a consequence of de novo purine synthesis inhibition, there was a 2- to 3-fold elevation of 5-phosphoribosyl-1-pyrophosphate pools in colon 38 tumor between 4 and 12 hr after DDATHF administration. When DDATHF (50 mg/kg) was administered 4 or 8 hr prior to 5-fluorouracil (5-FU; 85 mg/kg, i.p., weekly), these biochemical effects significantly increased the antitumor activity of 5-FU, with a modest increase in toxicity. Lower doses of DDATHF (25 and 37.5 mg/kg) when combined with 5-FU also resulted in an improved antitumor activity without additional toxicity. The two different schedules of administration for DDATHF, 4 and 8 hr prior to 5-FU, showed no differences in antitumor activity or toxicity.