Stimulation of 5-fluorouracil metabolic activation by interferon-alpha in human colon carcinoma cells

The influence of elevated levels of platelet-derived endothelial cell growth factor/thymidine phosphorylase on tumourigenicity, tumour growth, and oxygenation

PURPOSE

Investigation of the effect of platelet-derived endothelial cell growth factor/thymidine phosphorylase (PD-ECGF/TP) on various aspects of tumour growth in a xenograft model, including growth rate, tumourigenicity and oxygenation levels.

METHODS AND MATERIALS

MDA 231 breast cancer cells overexpressing PD-ECGF/TP protein were made by retroviral transduction. These cells were grown in vitro and in vivo as xenografts. Direct measurement of tumours was used to record growth parameters, while the comet assay with the bioreductive drug RSU 1069 was used to assess tumour cell oxygenation.

RESULTS

We report that MDA 231 breast tumour cell lines expressing an increased range of levels of PD-ECGF/TP have increased tumourigenicity positively related to the level of PD-ECGF/TP when implanted in nude mice. As previously reported, tumours grown from these overexpressing cell lines grew faster than the parental line. These tumours expressed higher levels of TP activity and showed increased immunocytochemical staining for PD-ECGF. In addition, the rate of growth was found to be positively related to the level of PD-ECGF/TP expressed by the tumour cells. When the comet assay was used to compare the oxygenation status of cells between the parental and PD-ECGF/TP overexpressing tumours, the latter were found to have a larger proportion of well oxygenated cells. This is consistent with these tumours having an increased and functionally competent vascular supply in response to the expression of PD-ECGF/TP.

CONCLUSION

PD-ECGF/TP appears to be capable of influencing tumourigenicity, angiogenesis and tumour growth in a proportional manner and can directly influence tumour oxygenation levels via its role in formation of functional vasculature.

Structural and theoretical studies suggest domain movement produces an active conformation of thymidine phosphorylase

Two new crystal forms of Escherichia coli thymidine phosphorylase (EC 2.4.2.4) have been found; a monoclinic form (space group P21) and an orthorhombic form (space group I222). These structures have been solved and compared to the previously determined tetragonal form (space group P43212). This comparison provides evidence of domain movement of the alpha (residues 1 to 65, 163 to 193) and alpha/beta (residues 80 to 154, 197 to 440) domains, which is thought to be critical for enzymatic activity by closing the active site cleft. Three hinge regions apparently allow the alpha and alpha/beta-domains to move relative to each other. The monoclinic model is the most open of the three models while the tetragonal model is the most closed. Phosphate binding induces formation of a hydrogen bond between His119 and Gly208, which helps to order the 115 to 120 loop that is disordered prior to phosphate binding. The formation of this hydrogen bond also appears to play a key role in the domain movement. The alpha-domain moves as a rigid body, while the alpha/beta-domain has some non-rigid body movement that is associated with the formation of the His119-Gly208 hydrogen bond. The 8 A distance between the two substrates reported for the tetragonal form indicates that it is probably not in an active conformation. However, the structural data for these two new crystal forms suggest that closing the interdomain cleft around the substrates may generate a functional active site. Molecular modeling and dynamics simulation techniques have been used to generate a hypothetical closed conformation of the enzyme. Analysis of this model suggests several residues of possible catalytic importance. The model explains observed kinetic results and satisfies requirements for efficient enzyme catalysis, most notably through the exclusion of water from the enzyme's active site.

A pharmacokinetic and pharmacodynamic study in vivo of human HT29 tumours using 19F and 31P magnetic resonance spectroscopy

19F-MRS (magnetic resonance spectroscopy) was used to study the pharmacokinetics of 5-fluorouracil (5-FU) in human (HT29) tumour xenografts, with and without pretreatment of the mice using either thymidine (40 min) or interferon-alpha (2 and 24 h). A 200 mg/kg i.p. bolus dose of 5-FU was eliminated from control tumours with a t1/2 of 25.4 +/- 2 min (mean +/- SEM, n = 11), while both thymidine (500 mg/kg) and interferon (50,000 IU/mouse) significantly increased t1/2 to 36.5 +/- 6.1 (n = 5) and 48.1 +/- 13.6 min (n = 4), respectively (P = 0.04, Gabriel's ANOVA). Thymidine increased 5-FU anabolism to cytotoxic 5-fluoronucleotides, and decreased the amount of tumour catabolites; the latter probably recirculated from liver since isolated HT29 cells did not catabolize 5-FU. These in vivo observations were confirmed by 19F-MRS quantification of tumour extracts. Interferon did not significantly affect 5-FU metabolism in the tumour or liver, nor the 5-FU t1/2 in liver. Treatment of tumours with 5-FU or interferon had no effect on tumour growth, whereas the combination strongly inhibited growth. 31P-MRS of HT29 tumours showed that 2 and 24 h after i.p. injections of interferon there was a significant increase in the pHint of 0.3 +/- 0.04 units (P = 0.002), while pHext and the tumour NTP/Pi ratio were unchanged. The large increase in the negative pH gradient (-delta pH) across the tumour plasma membrane caused by interferon suggest the delta pH may be a factor in tumour retention of 5-FU, as recently shown in isolated tumour cells.

Apoptosis in cultured human colon cancer cells induced by combined treatments with 5-fluorouracil, tumor necrosis factor-alpha and interferon-alpha

Biochemical analysis using nick end-labeling was performed to investigate the effect of various combinations of 5-fluorouracil, natural human tumor necrosis factor-alpha and natural human interferon-alpha on the induction of apoptosis in RPMI 4788 human colon cancer cells. After treatment with 5-fluorouracil (1 mM) for 48 h, the number of nick end-positive cells was significantly increased in comparison to the situation without treatment. When tumor cells were treated with 1 mM 5-fluorouracil, 2.86 Japan Reference Units (JRU)/ml natural human tumor necrosis factor-alpha and 1 x 10(3) IU/ml natural human interferon-alpha in combination for 48 h, the number of nick end-positive cells was significantly higher than that after treatment with 5-fluorouracil alone. MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) assay revealed a significant decrease of relative viability, as compared to treatment with 5-fluorouracil (1 mM), 5-fluorouracil + natural human tumor necrosis factor-alpha, or 5-fluorouracil + natural human interferon-alpha for 48 h. Pretreatment with 5-fluorouracil (1 mM) for 24 h prior to treatment with natural human tumor necrosis factor-alpha (2.86 JRU/ml) and natural human interferon-alpha (10(3) IU/ml) for 24 h resulted in a significant increase of nick end-positive cells compared to pretreatment with natural human tumor necrosis factor-alpha and natural human interferon-alpha prior to treatment with 5-fluorouracil for 24 h (p < 0.05). These results suggest that 5-fluorouracil alone can induce apoptosis in RPMI 4788 tumor cells and that this effect can be enhanced by combination with natural human tumor necrosis factor-alpha and natural human interferon-alpha.

Bimonthly high dose leucovorin and 5-fluorouracil 48-hour infusion with interferon-alpha-2a in patients with advanced colorectal carcinoma. Groupe d'Etude et de Recherche sur les Cancers de l'Ovaire et Digestifs (GERCOD)

BACKGROUND; The rationale for the modulation of 5-fluorouracil (5-FU) with interferon-alpha (IFN) is inhibition of 5-FU catabolism and 5-FU resistance. Clinical trials have shown debatable results when IFN is given in high doses with 5-FU used as a bolus alone or in combination with leucovorin (LV). A first-line Phase II study was performed in 50 patients with metastatic colorectal carcinoma who were given a bimonthly combination of high dose LV, a high dose 48-hour infusion of 5-FU, and a low dose of IFN.

METHODS

The regimen was comprised of a 2-hour infusion of LV, 500 mg/m2, on each of 2 consecutive days, and a 48-hour infusion of 5-FU, 1.5 to 2 g/m2/24 hours, starting after Day 1 of LV treatment every 2 weeks until there was evidence of disease progression. IFN was administered subcutaneously three times weekly at a dose of 3 MU (body surface area [BSA] < 1.75 m2) or 4.5 MU (BSA > OR = 1.75 m2).

RESULTS

World Health Organization toxicity Grade 3-4 occurred in 21 patients (42%): diarrhea in 6%, mucositis in 12%, neutropenia in 30%, and alopecia in 8%. The overall response rate was 44%; 1 patient had a complete response (2%), 21 had partial responses (42%), 23 had stable disease (46%), and 5 had disease progression (10%). The median progression free survival was 9 months, and median survival was 25 months.

CONCLUSIONS

Bimonthly high dose LV, a high dose 48-hour infusion of 5-FU, and a low dose of IFN had good activity in patients with advanced colorectal carcinoma. However, as in other schedules of LV and 5-FU, IFN induces high grade toxicity.

Biochemical mechanisms of interferon modulation of 5-fluorouracil activity in colon cancer cells

The antiproliferative effect of 5-fluorouracil (5-FU) in colon cancer can be enhanced by interferons (IFN-alpha and IFN-gamma). The mechanisms by which IFNs modulate 5-FU activity are not completely elucidated. IFN-alpha may elevate the levels of the active 5-FU metabolite 5-fluoro-2'-deoxyuridine-5'-monophosphate (FdUMP) in the cell, possibly leading to increased inhibition of the target enzyme thymidylate synthase (TS), which might enhance DNA damage. It has been shown that IFN-gamma can prevent 5-FU induced overexpression of TS. We studied IFN modulation in three colon cancer cell lines (SW948, WiDr, human; C26-10, murine) and the sublines WiDr/F and C26-10/F, which were adapted to low folate levels. A 1.5-fold increase in 5-FU sensitivity was observed in C26-10 and C26-10/F (by murine IFN-alpha, beta); in SW948, WiDr and WiDr/F (by human IFN-gamma) and in SW948 and WiDr/ F (by human IFN-alpha). In none of the cell lines did human IFN-alpha, IFN-gamma or murine IFN-alpha, beta increase FdUMP levels after exposure to 5-FU. TS activity, indirectly measured by incorporation of [6-3H]-deoxyuridine into DNA, was inhibited by 5-FU, but the IFNs did not enhance inhibition. DNA damage was measured as a drug-induced decrease of double-stranded (dss) DNA compared to control cells. After 5-FU exposure, dss DNA decreased to 60-75% in WiDr, WiDr/F and SW948 cells. Human IFN-alpha alone caused minimal DNA damage (95% dss DNA), but increased 5-FU-induced effects to 35-50% dss DNA. IFN-gamma did not cause DNA damage and did not enhance 5-FU-mediated DNA damage. Expression of TS protein, analysed by ELISA, was increased after 5-FU exposure of SW948 cells, but this increase was not affected by addition of either IFN-alpha or IFN-gamma. It is concluded that one of the mechanisms involved in modulation of 5-FU activity is the effect of IFN-alpha on 5-FU-mediated DNA damage, but for IFN-gamma no mechanism of action was found.

Human recombinant interferon alpha-2a plus 3'-azido-3'-deoxythymidine. Synergistic growth inhibition with evidence of impaired DNA repair in human colon adenocarcinoma cells

We reported that 3'-azidothymidine-3'-deoxythymidine (AZT) plus 5-fluorouracil or methotrexate produces additive cytotoxicity in HCT-8 cells: a reflection of increased AZT metabolism when de novo thymidylate (dTMP) synthesis was inhibited. We now report that AZT plus human recombinant interferon alpha-2a (rIFN-alpha 2a) produces synergistic growth inhibition in these cells. Evaluation of the effect of rIFN-alpha 2a on dTMP metabolism revealed that exposure to rIFN-alpha 2a (+/-AZT) did not affect dTMP synthase activity significantly but increased thymidine (dThd) kinase activity significantly. Consequently, AZT nucleotide production and incorporation into DNA were increased by coexposure to rIFN-alpha 2a. This alone, however, cannot explain the observed synergism. Therefore, the effect of these agents on DNA excision/repair processes was assessed. Isotope clearance studies demonstrated that rIFN-alpha 2a did not alter the rate of [3H]AZT excision from DNA. In contrast, filter-elution studies revealed that rIFN-alpha 2a (+/-AZT) produced more DNA damage and delayed repair compared with the effects produced by AZT alone. Since DNA polymerases alpha and beta are directly involved in gap-filling repair synthesis, experiments next assessed the effect of rIFN-alpha 2a and/or 3'- azido-3'-deoxythymidine-5'-triphosphate (AZTTP) on their activities. Polymerase alpha was inhibited slightly by AZTTP but not by rIFN-alpha 2a. Polymerase beta activity, however, was inhibited dramatically by rIFN-alpha 2a + AZTTP. Finally, western analysis revealed that a 24-hr exposure to 5000 IU/mL rIFN-alpha 2a (+/-20 microM AZT) significantly reduced wild-type p53 expression compared with AZT-exposed cells. We conclude that rIFN-alpha 2a enhances AZT-induced tumor cell growth inhibition by (i) increasing AZT metabolism, and (ii) inhibiting DNA repair and p53-mediated cell cycle control processes.

The treatment of disseminated malignant melanoma with special reference to the role of interferons, vinca alkaloids and tamoxifen

Malignant melanoma continues to increase in incidence. While early melanoma is highly curable by surgical means, the prognosis of patients with more advanced lesions and/or metastatic disease remains poor. Conventional chemotherapy with dacarbazine has a low frequency and short duration of response. Alternative drugs with single-agent activity include vinca alkaloids, nitrosoureas, procarbazine and platinum compounds. The addition of tamoxifen to chemotherapy, particularly cisplatin-based chemotherapy, appears to be beneficial. Recent studies suggest that combination chemotherapy may give better outcomes than single-agent treatment. Significant clinical activity has also been demonstrated with the use of interferons, particularly interferon alpha, and also with IL-2. Two recent studies suggest that the addition of interferon to chemotherapy may be beneficial. In addition, specific active immunotherapy with tumour vaccines has shown promise. The optimal methods of combining these treatment methods, such as chemotherapy and biological response modifiers/immunotherapy, however, remain to be defined.

Platelet-derived endothelial cell growth factor thymidine phosphorylase in tumour growth and response to therapy

Angiogenesis plays an important role in the growth and metastasis of solid tumours. Platelet-derived endothelial cell growth factor (PD-ECGF) is known to be chemotactic for endothelial cells in vitro and angiogenic in vivo. It is also known as gliostatin, a factor promoting neuronal survival, and thymidine phosphorylase (dThdPase), which catalyses the reversible phosphorylation of thymidine to thymine and 2-deoxyribose-1-phosphate. This enzymatic activity is critical for angiogenic activity. PD-ECGF protein is highly expressed in tumours compared with most normal tissues and has been correlated with tumour growth, invasion and metastasis in clinical studies. In addition, dThdPase activity (by inference PD-ECGF) has been found to be a major determinant of the toxicity of 5-fluorouracil and its prodrugs, which are extensively studied clinically as anti-cancer agents. This review attempts to summarize recent gains in understanding the nature, location and action of PD-ECGF and its specific relevance to tumour biology.

A phase II trial of interferon-alpha and 5-fluorouracil in patients with advanced renal cell carcinoma. A Southwest Oncology Group study

BACKGROUND

Renal cell carcinoma is a common neoplasm that is often refractory to treatment. It is occasionally responsive to immunomodulating agents including interferon-alpha, which enhances the effects of 5-fluorouracil upon cells. Combinations of these two drugs have been most frequently tested in patients with gastrointestinal cancers, with some promising results. Because interferon-alpha has activity for renal cell carcinoma, a trial of this combination in patients with this malignancy was undertaken.

METHODS

The Southwest Oncology Group performed a Phase II clinical trial of the combination of 5-fluorouracil and interferon-alpha for recurrent or metastatic renal cell carcinoma. Eligibility criteria included no prior treatment with medications for cancer, a performance status of 2 or better, and bidimensionally measurable disease. The regimen studied consisted of 5-fluorouracil, 750 mg/M2/day, by continuous intravenous infusion on Days 1-5, and interferon-alpha-2b (Intron A), 5 x 10(6)U/M2/day, subcutaneously on Days 1, 3, and 5, repeated every 21 days.

RESULTS

Forty eligible patients were treated; twenty of the 40 underwent a nephrectomy. The regimen was tolerable: 3 patients had Grade 4, and 17 had Grade 3 toxicity. There were 5 partial responses (13% with 95% confidence limits of 4-27%). Median progression free survival for all 40 patients was 4 months and median overall survival was 15 months from the time of registration.

CONCLUSIONS

The combination of 5-fluorouracil and interferon-alpha given by this schedule, although tolerable and occasionally yielding responses, is not an improvement over existing therapies.

Phase II clinical trial with 5-fluorouracil, recombinant interferon-alpha-2b, and cisplatin for patients with metastatic or regionally advanced carcinoma of the esophagus

BACKGROUND

Recombinant interferon-alpha (IFN) augments the cytotoxicity of both 5-fluorouracil (5-FU) and cisplatin in vitro. A phase II study of 5-FU and IFN resulted in response rates of 25-27% in patients with metastatic esophageal carcinoma.

METHODS

A Phase II trial was initiated to determine the clinical utility of a three-drug combination (FIP) in patients with regionally advanced or metastatic esophageal carcinoma. Eligibility included biopsy-proven Stage III or IV squamous cell carcinoma or adenocarcinoma of the esophagus with no prior chemotherapy, adequate performance status, nutritional status, bone marrow, hepatic and renal function, and signed informed consent. Patients were treated in the exact sequence of IFN==>cisplatin==>5-FU. Patients received 5-FU, 750 mg/m2/day for 5 days followed by weekly bolus therapy at the same dose; cisplatin, 100 mg/m2 on Day 1, followed by weekly therapy, 25 mg/m2 over the course of 1 hour; and IFN, 10 MU subcutaneously 3 times/week beginning on Day 1. All patients received sargramostim (granulocyte-macrophage colony-stimulating factor, Escherichia coli-derived), 5 micrograms/kg subcutaneously 5 times/week. No patients received radiotherapy.

RESULTS

Twenty-four patients were enrolled; 23 were eligible, and 1 was excluded on pathology review (patient was found to have a leiomyoblastoma). The demographics of the population were: median age, 63 years (range, 43-73 years); 18 male patients; squamous cell carcinoma: adenocarcinoma ratio, 22:1, and Stage III:IV ratio, 10:13. Grade 3-4 National Cancer Institute Common Toxicity Criteria toxicities included: leukopenia (13), thrombocytopenia (14), and infection (9). Grade 3 diarrhea, mucositis, and vomiting occurred in 6 patients, 4 patients, and 1 patient, respectively. There were two instances of sudden death, likely related to tumor progression. Major responses occurred in 15 of 23 patients (65%; 95% confidence interval, 43%, 85%) (1 complete response, 14 partial responses). The median survival was 8.6 months; with a median follow-up of 26 months, estimated 30-month survival was 31%.

CONCLUSIONS

This regimen, although moderately toxic, has substantial activity in metastatic and regionally advanced squamous cell carcinoma of the esophagus. Further investigations should be conducted to determine the role of IFN in the treatment of esophageal carcinoma.

The modulating effect of interferon alpha-2a on the antitumor activity of UFT against a human gastric carcinoma xenograft, SC-1-NU, in nude mice

The modulating effect of recombinant human interferon alpha-2a (IFN) on the antitumor activity of UFT, a mixed compound of tegafur and uracil at a molar ratio of 1:4, was investigated against SC-1-NU, a human gastric cancer xenograft serially transplanted in nude mice. IFN was administered subcutaneously at a dose of 60,000 IU/mouse daily for 14 days, and UFT was given at a dose of 15 mg/kg as tegafur daily, except on Sundays, for 3 weeks. The agents were administered either alone or simultaneously. Synergistic antitumor activity on SC-1-NU was produced by the combination of IFN and UFT without any increment of side effects, and the combination therapy also increased intratumoral thymidylate synthetase (TS) inhibition and the amount of 5-fluorouracil (5-FU) in the intratumoral RNA. Thus, IFN seems to modulate the antitumor activity of UFT against SC-1-NU through an inhibition of DNA synthesis and RNA distortion, and therefore this combination could be useful for clinical application.

Increased sensitivity to the prodrug 5'-deoxy-5-fluorouridine and modulation of 5-fluoro-2'-deoxyuridine sensitivity in MCF-7 cells transfected with thymidine phosphorylase

Platelet-derived endothelial cell growth factor (PD-ECGF) is identical to human thymidine phosphorylase (dThdPase). The human MCF-7 breast cancer cell line was transfected with the dThdPase cDNA and expressed a 45 kDa protein that was detected with anti-dThdPase antibody. Cell lysates possessed elevated dThdPase activity and cells had up to 165-fold increased sensitivity to the prodrug 5'-deoxy-5-fluorouridine (5'-DFUR) in vitro. Sensitivity to 5-fluorouracil (5-FU) and 5-fluoro-2'-deoxyuridine (5-FUdR) was unchanged. Recombinant dThdPase was shown to catalyse directly the phosphorolytic cleavage of 5'-DFUR to 5-FU. Exogenous thymidine (dThd) reversed the toxicity of 5-FUdR on the parental line (1 microM dThd increased the IC50 value 1000-fold), but the dThd rescue was substantially modulated in the dThdPase-expressing clone 4 (1 microM dThd raised the IC50 value 3-fold). We observed a substantial 'bystander' killing effect when small proportions of dThdPase-expressing cells were mixed with parental MCF-7 cells. dThdPase activity was on average 27-fold higher in breast tumours than in normal breast. The levels of wild-type MCF-7 are similar to the low end of the tumour expression. Thus, in some tumours resistance to 5'-DFUR therapy could be due to low dThdPase activity, and transfection to raise the dThdPase levels within the broad tumour range or above it should markedly enhance sensitivity to the prodrug. These results confirm that dThdPase is a major pathway in the metabolic activation of 5'-DFUR, and the bystander effect suggests that this may be a suitable enzyme for gene therapy-directed enzyme/prodrug activation therapy.

Thymidine phosphorylase mediates the sensitivity of human colon carcinoma cells to 5-fluorouracil

Interferon-alpha (IFN alpha) potentiates the antitumor activity of 5-fluorouracil (FUra) in colon cancer in vitro, in vivo, and clinically. A likely mechanism for this action is the induction by IFN alpha of thymidine phosphorylase (TP), the first enzyme in one pathway for the metabolic activation of FUra to fluorodeoxyribonucleotides. To test this hypothesis, an expression vector containing the TP cDNA was transfected into HT-29 human colon carcinoma cells. Five stable transfectants were selected and analyzed. All showed increased sensitivity to FUra cytotoxicity, ranging from a 2-fold to a 19-fold decrease in the IC50 for FUra, compared to wild-type cells. Levels of TP mRNA, protein, and enzyme activity were elevated in the transfectants, and there was a significant correlation between the relative increase in sensitivity to FUra and both the increase in both TP mRNA levels and TP activity. Transfected cells exhibited increased formation of FdUMP, but not the ribonucleotides FUDP and FUTP, from FUra when compared to wild-type cells. The changes in TP activity, FdUMP formation, and FUra sensitivity in the transfected cells were comparable with those seen after treatment of wild-type cells with IFN alpha. These studies provide direct evidence for the role of TP in mediating the sensitivity of colon carcinoma cells to FUra, and further support the importance of the induction of TP in the biomodulating action of IFN alpha on FUra chemosensitivity.

The role of interferon-alpha as a modulator of fluorouracil and leucovorin

Several preclinical studies have demonstrated that interferon-alpha (IFN-alpha) may enhance the cytotoxicity of fluoropyrimidines in a greater-than-additive manner in a variety of human cancer cell lines. The underlying mechanism(s) have varied in different cancer cell lines, and include increased fluorouracil anabolism to fluorodeoxyuridine monophosphate, further inhibition of thymidylate synthase, stimulation of thymidine and uridine phosphorylase activities, greater DNA damage, and enhanced natural killer cell-mediated lysis of tumour targets. These preclinical studies stimulated clinical evaluation of IFN-alpha in combination with 5-fluorouracil (5-FU) with and without leucovorin (LV), and the initial clinical results appeared promising. We summarise preclinical research concerning the interaction of 5-FU and IFN-alpha. The rationale for combining 5-FU with IFN-alpha and LV is discussed, and we describe our clinical experience with the combination of 5-FU, LV and IFN-alpha-2a. The insights and unresolved questions concerning the clinical application of this combination are also discussed.

Treatment of advanced colorectal cancer with 5-fluorouracil and interferon-alpha: an overview of clinical trials

5-Fluorouracil (5-FU) is the most active single agent for treatment of advanced colorectal cancer, although objective responses occur in only 20% of patients, and there seems to be no impact on overall survival. Experimental findings suggesting that interferon-alpha (IFN-alpha) enhances 5-FU cytotoxicity have stimulated an increasing number of clinical trials to evaluate the therapeutic potential of this combination. This article summarises the possible mechanisms of interaction of 5-FU and IFN-alpha, and provides an overview of the current status of this approach in advanced colorectal cancer. A computerised (Medline) and manual search were performed to identify all trials using 5-FU and IFN-alpha for the treatment of advanced colorectal cancer published in the English literature between 1960 and 1994. Information abstracted included treatment regimen, number of patients, pretreatment status, complete and partial remissions, remission duration, overall survival, and toxicity. A total of 417 patients were enrolled in 16 trials using different regimens of 5-FU and IFN-alpha, and double modulation of 5-FU with leucovorin (LV) and IFN-alpha was investigated in nine trials involving 332 patients. The mean overall response rate in these phase II trials was only 31% (range 3-76) and 35% (range 0-54), respectively. Early results of six prospectively randomised studies of 5-FU or 5-FU/LV +/- IFN-alpha also did not suggest a significant enhancement of the antitumour effectiveness with the addition of IFN-alpha. There is increasing evidence, however, that administration of IFN-alpha along with 5-FU enhances toxicity. Because of their modest therapeutic index, currently employed regimens of 5-FU +/- LV plus IFN-alpha cannot be recommended for routine use at the present time. The combination of 5-FU plus LV represents an equally effective and less expensive alternative. Nevertheless, there is still hope that further attempts to elucidate the complex mechanisms of this potentially synergistic drug combination will allow the rational design of regimens with a superior therapeutic index.

Purification, cloning, and expression of murine uridine phosphorylase

Uridine phosphorylase was purified 10,300-fold from tumors of the murine colorectal adenocarcinoma cell line, Colon-26. Degenerate DNA probes were synthesized corresponding to partial amino acid sequences and used to screen a Colon-26 cDNA library. A cDNA clone of 1327 base pairs that contains a 5' untranslated region, a coding region of 933 base pairs, and a 3' nontranslated region with a polyadenylated tail was identified. The cDNA was confirmed to be uridine phosphorylase by 1) sequence comparison to uridine phosphorylase of Escherichia coli, 2) substrate specificity studies with recombinant protein expressed in COS-7 cells that demonstrated relatively high enzyme activity with uridine as substrate compared low levels when thymidine was used, and 3) inhibition of enzyme activity by the competitive inhibitor 2,2'-anhydro-5-ethyluridine. Northern blot analysis using the cDNA as a probe, demonstrated high levels of mRNA expression in Colon-26. Expression was low in NIH3T3 cells, but high in DMBA-3 and PH-1 cells, which are NIH3T3-derived cells that have been transformed with mutated murine Ha-ras and viral Ha-ras, respectively. Expression of uridine phosphorylase mRNA in these cell lines was further enhanced by treating the cells with the inflammatory cytokines, tumor necrosis factor-alpha, interleukin 1 alpha, and interferon gamma.

Double modulation of 5-fluorouracil with interferon alpha 2a and high-dose leucovorin: a phase I and II study

Twenty-nine patients with adenocarcinomas of gastrointestinal or unknown primary, and three with advanced neuroendocrine tumours, were entered into a study of bolus plus infusional 5-fluorouracil (FUra) modulated with high-dose leucovorin (LV) and recombinant interferon alpha 2a (IFN-alpha). Successive cohorts of > or = 4 patients received IFN-alpha at 1.5, 3, 4.5, 6 and 9 MU on alternate days throughout the treatment period. The FUra/LV regimen consisted of: LV 200 mg m-2 i.v. infusion over 2 h, FUra 400 mg m-2 i.v. bolus then FUra 400 mg m-2 i.v. infusion over 22 h, all repeated on day 2, on a 14-day cycle. FUra was given at 75% dose for the first course, increasing (in the absence of WHO grade > or = 2 toxicity) to 87.5% for the second and 100% for subsequent courses up to a maximum of 12. The maximum tolerated dose (MTD) of IFN-alpha was 6 MU on alternate days, with 7/8 patients at 9 MU requiring dose reductions. At 6 MU IFN-alpha, the MTD of FUra was not exceeded at 100% (i.e. 400 mg m-2 bolus and infusion, days 1 and 2), and FUra-related toxicities (mucosal, haematological, dermatological) were extremely mild. Twenty-nine patients were assessable for tumour response, among whom WHO criteria partial responses were seen in 7/14 with colorectal, 1/4 with gastric, 0/1 with pancreatic, 1/3 with neuroendocrine and 3/6 with unknown primaries. Median response duration was 51 weeks. Minor responses and stable disease were seen in a further six patients. Median survival of patients with advanced adenocarcinomas was 9 months, with 33% surviving beyond 18 months. This schedule offers a safe way of co-administering FUra, LV and IFN-alpha. The addition of IFN-alpha, while causing significant independent toxicity, does not significantly increase the dose-limiting mucosal toxicities of FUra/LV. Further investigation is required to determine the contribution of IFN-alpha to the anti-tumour activity of the combination.

Lack of effect of interferon alpha 2a upon fluorouracil pharmacokinetics

The disposition of 5-fluorouracil (FUra) was studied in 19 colorectal cancer patients during treatment with FUra and high-dose leucovorin (LV) with or without interferon alpha 2a (IFN-alpha). All received LV 200 mg m-2 over 2 h, then FUra 400 mg m-2 over 5 min then FUra 400 mg m-2 over 22 h, repeated on day 2, on a 14 day cycle. Nine patients also received IFN-alpha 6 MU every 48 h, starting at least 2 weeks before the study. Series of 14 blood samples were assayed for FUra by reversed-phase high-performance liquid chromatography (HPLC). Minimum Akaike information criterion estimation was used to determine the simplest effective pharmacokinetic model. This consisted of a single compartment with first-order (linear) and Michaelis-Menten (non-linear) components to drug elimination. This model gave r2 > 0.98 in 19/20 data sets. With the Michaelis constant (KM) set at 15 microM, values were derived for the volume of distribution (Vd), the maximum rate of non-linear elimination (Vmax) and the first-order elimination rate constant (K1.e). Mean (+/- s.d.) values in control (no IFN-alpha) patients were: Vd 10.4 (+/- 1.9) l m-2, Vmax 182 (+/- 59) mumol l-1 h-1 and k1.e 4.35 (+/- 0.58) h-1. No significant differences were detected in patients receiving IFN-alpha, in whom the equivalent mean values were Vd 10.0 (+/- 0.9) l m-2, Vmax 141 (+/- 27) mumol l-1 h-1 and k1.e 3.96 (+/- 0.5) h-1. Mean trapezoidal AUC0-22 h was similar in the two groups (control patients 116 microM h, IFN-alpha patients 125 microM h). No significant correlations with renal or hepatic function were detected. These results, while not inconsistent with previous reports of a reduced rate of FUra elimination at higher IFN-alpha doses, suggest that any clinical effect of this moderate dose of IFN-alpha on FUra toxicity or activity is due to modulation at target cells, not to pharmacokinetic interaction.

New targets for pyrimidine antimetabolites in the treatment of solid tumours. 1: Thymidylate synthase

Thymidylate synthase forms the target for anticancer therapy with fluoropyrimidines. Anticancer activity can be increased by the use of different modulators of fluoropyrimidine metabolism, which lead to an enhanced inhibition of thymidylate synthase. In vitro and in vivo studies with fluoropyrimidines and two of these modulators, folinic acid (leucovorin) and interferon, are summarized. The promise of these preclinical results is reflected by the response data of several clinical trials. The biochemical effects of these modulators are described and illustrated by the fluoropyrimidine-mediated inhibition of thymidylate synthase in tumour samples, which is clearly enhanced by folinic acid. The regulation of thymidylate synthase synthesis may also be crucial for total blockade of thymidylate synthase activity. This regulation may be influenced by interferon-gamma. Although the addition of modulators increases the activity of fluoropyrimidines at the level of thymidylate synthase, most solid tumours, especially colorectal carcinomas, are resistant to these combinations. For this reason, new, more potent inhibitors of thymidylate synthase have been developed, the antifolates. Preclinical data show that some of these compounds have good antitumour activity, but they still have to prove their value in the clinic. These two approaches, the use of modulators and new compounds, have shown activity preclinically and the extension of these findings to clinical studies stresses the importance of thymidylate synthase as a target in fluoropyrimidine therapy of solid tumours.

Preclinical and clinical aspects of biomodulation of 5-fluorouracil

Although single agent 5-FU has for many years been the standard therapy for advanced colorectal malignancies, a number of recent clinical trials show higher response rates with biomodulation of 5-FU by several different agents. In general, trials of leucovorin, methotrexate, interferon, and PALA given in biomodulatory doses and sequences with 5-FU have demonstrated comparable response rates over a broad range. However, in the absence of controlled direct comparative phase III trials, final judgement on clinical superiority of a particular regimen must be reserved. Nevertheless, on the basis of current data, certain approaches appear promising and warrant further investigation. Compared to single agent 5-FU, survival benefit has been demonstrated with both low and high dose leucovorin/5-FU regimens and response rates in the 20-50% range appear reproducibly higher than those of 5-FU alone. Low dose and either continuous infusion or repetitive dosing of leucovorin, as well as the effect of treatment sequence and intervals between drugs, require additional investigation. When given 20-24 h before 5-FU, methotrexate achieves response rates similar to leucovorin modulated 5-FU, but the potential role of rescue leucovorin used in many of the trials makes definitive interpretation difficult. Interferon/5-FU regimens attaining response rates of 30-40% are promising but need to be carefully and rationally designed. Low dose PALA with effective doses of 5-FU achieving responses in 35-45% of patients represent a marked improvement in earlier trials of high dose PALA, but additional studies with higher doses not compromising 5-FU dose intensity should be considered. Certainly, the concomitant use of multiple modulating agents also needs further investigation. While many such trials already performed attained results no better than single agent biomodulation, the preliminary results obtained by Grem and colleagues with IFN/LV/5-FU in untreated patients, and by Conti et al. using TMTX/LV/5-FU in previously treated patients are encouraging. Further understanding of the mechanisms of action and interaction of modulating agents should allow additional rational combinations to be explored clinically. Cisplatin biomodulation of 5-FU has been studied in gastrointestinal and head and neck malignancies achieving excellent results in the latter group. Preclinical evidence exists which suggests, however, that 5-FU modulation of cisplatin may be more effective, especially when 5-FU is administered 24 h or more before cisplatin. Clinical investigation of this sequence is currently lacking. Data to support the clinical promise of AZT, IdUrd, uridine, and the benzylacyclouridines are not yet available, although preclinical and preliminary clinical studies are promising.

Biologic agents as biochemical modulators: pharmacologic basis for the interaction of cytotoxic chemotherapeutic drugs and interferon

Biochemical modulation of cytotoxic cancer chemotherapeutic agents is one means of enhancing the activity and selectivity of antitumor drugs. Traditionally this approach has utilized detailed information regarding a particular enzymatic reaction or biochemical pathway to develop potential modulating agents. In contrast, the reported clinical therapeutic activity of IFN in combination with cytotoxic agents has prompted a reexamination of the biochemical actions of the cytokine. Interferon elicits a number of cellular actions that might contribute to its pharmacologic activity, including both direct antitumor effects and host-mediated actions. The best understood are those related to the cytotoxicity of the fluoropyrimidine antimetabolites and include enzymatic reactions involved in fluoropyrimidine metabolic activation, catabolism, and interaction with its target enzyme. However, even in this instance, a mechanistic association of a specific pharmacologic action with therapeutic activity remains to be determined. These studies demonstrate that cytokines and other biologic agents may exert specific biochemical modulations that augment (or potentially attenuate) the activity of the cytotoxic chemotherapeutic agents.

Interferon alpha and 5'-deoxy-5-fluorouridine in colon cancer: effects as single agents and in combination on growth of xenograft tumours

Interferon-alpha (IFN-alpha) enhances the activity of the 5-fluorouracil (5-FU) prodrug 5'-deoxy-5-fluorouridine (5'-dFUrd) in colorectal cancer cells in vitro by upregulating the enzyme pyrimidine nucleoside phosphorylase (PNPase), which is responsible for converting 5'-dFUrd to 5-FU. We examined whether such enhancement also occurs in vivo using human colorectal xenografts in nude mice. The Co-115 line has high basal levels of PNPase and the enzyme level was increased in tumours from mice treated for 3 weeks with 50,000 IU/day (5 days/week) of IFN-alpha A/D. The prodrug 5'-dFUrd (200 mg/day, 5 days/week) had a much greater antitumour activity than 5-FU had when it was used at an approximately equitoxic dose (20 mg/day, 5 days/week). However, because of the high activity of 5'/dFUrd as a single agent, no enhancement by IFN-alpha A/D was observed. Studies on xenografts of WiDr cells indicated that this line is much less sensitive to 5'-dFUrd. However, treatment of animals with IFN-alpha A/D at doses of 75,000 IU/day or 150,000 IU/day resulted in significant inhibition of WiDr tumour growth. Combination treatment with 75 mg/kg/day or 150 mg/kg/day of 5'-dFUrd resulted in enhanced antitumour activity, particularly at the higher dose of IFN-alpha A/D. Synergy of this drug combination was confirmed by isobologram analysis. Analysis of PNPase levels in WiDr tumours, excised from mice treated with IFN-alpha A/D, demonstrated that the enzyme activity was increased by IFN-alpha in a dose-dependent manner. Slight increases were also seen in normal liver and intestine from the same animals. Our results indicate that modulation of converting enzymes for anticancer prodrugs by cytokines could be a novel therapeutic strategy for combination therapy of colorectal cancer.

Synergistic cytotoxicity of AZT plus alpha and gamma interferon in chronic myeloid leukemia cell line K562

We have previously reported that the antineoplastic activity of 3'-azido 3' deoxythymidine (AZT) can be increased by drugs that inhibit "de novo" thymidylate synthesis, such as 5-fluorouracil, methotrexate and hydroxyurea. In the present study we tested the combinations AZT+alpha interferon (IFN) and AZT+gamma IFN on in vitro growth of the human acute-phase chronic myeloid leukemia (CML) cell line K562. After 72 hours incubation, not only AZT+alpha-IFN but also AZT+gamma-IFN were synergistic in inhibiting K562 growth, as demonstrated by isobologram analysis of the data. This enhanced cytotoxicity was confirmed by the evaluation of [3H]AZT incorporation into cellular DNA, that was increased by 50% and 222% in the presence of alpha- and gamma-IFN, respectively. The addition of 50 mumol/l thymidine to the culture medium was able to reduce the cytotoxicity of the drug combinations to the degree observed with each compound alone; furthermore, the increased incorporation of AZT into DNA was completely reversed. These data indicate the existence of a biochemical interaction between AZT and IFNs that results in an increased cytotoxic effect. While the combination AZT+alpha-IFN is currently being tested in HIV-related malignancies, AZT+gamma-IFN is new and deserves further study in human CML acute and chronic phase models, in view of possible clinical applications.

Treatment of carcinoma of the esophagus with 5-fluorouracil and recombinant alfa-2a-interferon

BACKGROUND

Combinations of 5-fluorouracil (5FU) and recombinant alfa-2a-interferon (IFN) are synergistic in vitro and have demonstrated activity in colorectal carcinoma, renal cell carcinoma, and urothelial tumors.

METHODS

A Phase II trial of the combination of 5FU, 750 mg/m2 daily x 5 followed by weekly bolus therapy, and IFN, 9 MU subcutaneously three times per week, was initiated in patients with esophageal carcinomas. Patients were required to have biopsy-proven squamous cell or adenocarcinoma of the esophagus, locally advanced or metastatic disease beyond the scope of surgical resection, and adequate performance status, renal, hepatic, and bone marrow function.

RESULTS

Twenty-one patients were enrolled; one patient was inevaluable for response because he had received prior chemotherapy, but was evaluated for toxicity. Eleven patients had metastatic disease, and 10 had locally advanced disease. Thirteen patients had squamous cell carcinoma and 8 adenocarcinoma. Toxicities were acceptable with no serious diarrhea and only two cases of serious stomatitis, although a greater than expected incidence of neurologic toxicity was observed. There were five responders (25%) including two patients with advanced or locally advanced disease rendered pathologically free of disease. One patient, initially considered surgically unresectable, was able to undergo a total thoracic esophagectomy after responding to treatment with 5FU/IFN, at which time only a single microscopic focus of carcinoma in situ was found. She remains alive and free of disease at 18+ months. A second patient who presented with metastatic disease and nearly complete obstruction of the esophagus regained normal swallowing function after treatment with 5FU/IFN; rebiopsy of all lesions revealed the patient to be pathologically free of disease. He survived over 2 years.

CONCLUSIONS

This regimen employing a single cytotoxic agent has activity in esophageal carcinoma. Strategies employing biochemical modulation deserve additional investigation in the treatment of esophageal carcinoma.

Cytokines induce uridine phosphorylase in mouse colon 26 carcinoma cells and make the cells more susceptible to 5'-deoxy-5-fluorouridine

The antiproliferative activity of 5-fluorouracil (5-FUra) and 5'-deoxy-5-fluorouridine (5'-dFUrd), used in combination with typical cytokines and growth factors, was investigated in mouse colon 26 carcinoma cells. Tumor necrosis factor alpha (TNF alpha), interleukin-1 alpha (IL-1 alpha), and interferon gamma (IFN gamma) at low doses showing < 50% inhibition of cell growth by themselves enhanced the susceptibility of the cells to the activity of 5'-dFUrd. In particular, a mixture of these cytokines greatly enhanced the activity of 5'-dFUrd and 5-FUra by up to 12.4- and 2.7-fold, respectively, whereas the activity of other cytostatics was only slightly changed (< 1.5-fold). Basic fibroblast growth factor also increased the susceptibility, but only to 5'-dFUrd. This preferential enhancement of the activity of 5'-dFUrd would be due to induction by the cytokines of uridine phosphorylase (Urd Pase), by which 5'-dFUrd is converted to 5-FUra. TNF alpha, IL-1 alpha, IFN gamma, and a mixture of these factors increased the enzyme activity by up to 3.7-fold in colon 26 cells. Consequently, the anabolism of 5'-dFUrd to fluoronucleotides and the incorporation of 5-FUra into RNA in colon 26 cells were increased by TNF alpha treatment. In addition, the increase by the cytokine mixture in the susceptibility to 5'-dFUrd was abolished by an inhibitor of Urd Pase, 2,2'-anhydro-5-ethyluridine. These results indicate that induction of Urd Pase activity by cytokines is a critical event that increases the susceptibility to 5'-dFUrd.

Mechanisms by which interferon potentiates chemotherapy

IFNs potentiate the activity of multiple chemotherapy drugs in vitro; improved activity has been noted with 5-FU and IFN-alpha against colon carcinoma and with IFN-alpha and DTIC against melanoma. The mechanisms whereby interferon potentiates the activity of chemotherapeutic drugs are complex, and different mechanisms may be operative with different drugs and against different tumors. Both indirect immunoaugmentative effects of interferon and direct biochemical modulation of the drugs (as documented in detail with 5-FU) may play a role; further elucidation of these mechanisms will enable us to optimally utilize the combination of interferon and antineoplastic agents.

Five-day infusional fluorodeoxyuridine with oral leucovorin and escalating doses of interferon alpha-2b: a phase I study

In a previous phase I study we identified the maximally tolerated dose (MTD) of a continuous intravenous infusion of fluorodeoxyuridine (FUdR) to be 0.3 mg/kg daily for 5 days when combined with oral leucovorin (LV) given at 100 mg q4h. In an attempt to modulate FUdR further, we added escalating doses of interferon alpha-2b (IFN) to FUdR/LV in a phase I cohort study. A total of 36 patients with refractory solid tumor were treated at two dose levels of FUdR and five dose levels of IFN. Although the initial patient cohort was treated with a dose of FUdR lower than that previously identified as the MTD [FUdR at 0.2 mg/kg daily with LV at 100 mg q4h and IFN at 2 million units (MU)/m2 daily], three of six patients developed grade 3 mucositis, indicating that the toxicity of FUdR/LV was increased in the presence of low doses of IFN. After decreasing the FUdR dose to 0.1 mg/kg daily, we could increase the dose of IFN from 2 to 30 MU/m2 daily in five additional cohorts of patients. With increasing IFN doses, no increase in mucositis or dermatitis was observed, indicating no further potentiation of FUdR/LV toxicity with higher IFN doses. However, known toxicities of IFN, including transient myelosuppression and hepatic transaminase elevation, were observed more frequently at IFN doses of 15 and 30 MU/m2 daily, where they became dose-limiting. We conclude that IFN modulates FUdR/LV at low doses, resulting in increased FUdR toxicity. When the dose of IFN is increased, this FUdR/LV toxicity does not appear to be potentiated further and IFN-related toxicities become dose-limiting.

Cytokines induce thymidine phosphorylase expression in tumor cells and make them more susceptible to 5'-deoxy-5-fluorouridine

The present study shows that various cytokines such as tumor necrosis factor (TNF alpha), interleukin-1 alpha (IL-1 alpha), and interferon-gamma (IFN gamma) make tumor cells much more susceptible to the cytostatic 5'-deoxy-5-fluorouridine (5'-dFUrd) than to 5-fluorouracil (5-FUra) and other cytostatics. These three cytokines increased the susceptibility of human cancer cell lines (COLO201, MKN45 and WiDr) but did not affect that of normal fibroblast WI38 cells. The cytokine mixture induced a 50-fold increase in the susceptibility of COLO201 to 5'-dFUrd, whereas a 12-fold increase and a less than 5-fold enhancement in the susceptibility to 5-FUra and other cytostatics, respectively, were observed. The increased susceptibility would be a result of the induction of thymidine phosphorylase (TdR Pase), which is the essential enzyme for the conversion of 5'-dFUrd to 5-FUra. The cytokine mixture increased TdR Pase activity by up to 47 times and greatly induced its mRNA expression in the cancer cell lines. These results suggest that the therapeutic benefit of 5'-dFUrd would be improved by its use in combination with the cytokines.