Cardiotoxicity of 5-fluorouracil in combination with folinic acid in patients with gastrointestinal cancer

BACKGROUND: Cardiotoxicity related to the widely used cytotoxic compound 5-fluorouracil (5-FU) is rare compared with the frequency observed with the use of anthracyclines. More effective protocols incorporating active biomodulatory compounds like folinic acid (FA) or combination chemotherapy change type and severity of toxicity as well. The objective of the current study was to assess cardiotoxicity of the combination 5-FU and folinic acid. METHODS: The authors' multicenter experience with 390 patients treated for advanced gastrointestinal cancer with intermediate-dose folinic acid and 5-FU was reviewed. RESULTS: The overall risk of cardiotoxicity was 3%, which is not significantly higher than that reported with 5-FU alone. Eight of 53 patients with a history of cardiac disease reported cardiac symptoms (15.1%), compared with 5 of 337 patients (1.5%) with a no history of cardiac disease. Median time to symptoms was 3 days (range, 2-6). Nine patients had symptoms resembling myocardial ischemia, one patient died due to assumed myocardial infarction related closely to fluorouracil treatment, four patients had supraventricular arrhythmia, and one patient had congestive heart failure. A history of cardiac disease was the only risk factor associated with cardiotoxicity. Relapses were frequent on reinstitution of therapy despite cardiac symptoms in the preceding cycle. Therapeutically or prophylactically administered nitrates had no significant effect. CONCLUSION: Physicians should be aware of the cardiotoxic properties of active fluorouracil treatment. The combination of 5-FU and leucovorin does not differ from single-agent therapy in frequency or type of cardiotoxicity. Close monitoring of patients is mandatory, especially for those patients at high risk for cardiac side effects. Treatment should be discontinued if coronary symptoms develop, because neither effective treatment nor prophylaxis exists for such symptoms.     

Treatment of advanced colorectal cancer with folinic acid and 5-fluorouracil in combination with cisplatinum

51 patients with metastatic colorectal cancer (stage Dukes D) were treated with intravenous (i.v.) infusion on days 1, 3, 5, 8 and 16 with folinic acid (200 mg/m2) and 5-fluorouracil (600 mg/m2), and on days 1, 8 and 16 with cisplatinum (25 mg/m2 i.v.); cycles were repeated every 4 weeks. All 51 patients were evaluable for toxicity and response criteria. 26 patients had objective responses (3 complete responses, 5.9%; 23 partial responses, 45.1%), relative risk 51% (95% confidence intervals 36.7-65.0%). Response duration ranged from 4 to 28.0 months (median 16.8). Overall median survival of all patients included was 14.7 months (range 3.0-33.0). Toxicity of WHO grade III, requiring dose reduction, occurred in 9 (18%) patients. The regimen described here appears to be active, safe and well tolerated for treatment of patients with advanced colorectal cancer.     

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

An experimental study of Allergan's Oxysept Comfort system was performed by measuring the slight reddish hue that appears in the disinfecting solution, indicating to the users that their lenses are again ready to be worn. The temporal evolution of the color of the solution has been measured under standardized conditions and analyzed in the CIELAB system, from the perspective of the typical threshold discrimination of the human eye. Color differences between neutralized and non-neutralized solutions occurred in an appropriate direction of the color space to enhance discrimination and were clearly perceptible by normal observers (greater than 9.7 +/- 1.2 CIELAB units). Colorimetric analyses have been used to draw conclusions regarding observers with defective color vision. The color of the solution changes abruptly, approximately 25 min after the neutralization process begins, and remains nearly constant after about 60 min, this agreeing well with the temporal evolution of the hydrogen peroxide concentration.     

Oxaliplatin, folinic acid and 5-fluorouracil (folfox) in pretreated patients with metastatic advanced cancer. The GERCOD

Preliminary studies suggest synergy between oxaliplatin and fluorouracil (5-FU). To assess this issue, we performed a study in pretreated patients with advanced colorectal cancer (CRC) resistant to leucovorin and 5-FU. Regimen consisted of oxaliplatin day 1, 130 mg/m2 every two cycles (folfox 1) or 100 mg/m2/cycle (folfox 2) or 85 mg/m2/cycle (folfox 3) and leucovorin 500 mg/m2 as a 2-hour infusion, followed by 5-FU 22 h infusion 1.5-2 g/m2 for two consecutive days every 2 weeks. One hundred and thirteen patients have been treated. One complete response (CR) and 32 partial responses (PRs) were observed for an overall response rate of 29.2%. Sixty-seven patients had prior documented progression while receiving the same schedule of leucovorin and 5-FU than the one used in the folfox regimens, among them 18 had PRs (26.9%). The best response rate was observed in patients treated with the folfox 2 regimen: 41.7%. From start of folfox, median progression-free survival was 6 months and median survival 13 months. Limiting toxicities were peripheral neuropathy and neutropenia. Fifty-four percent of the patients experienced WHO toxicity > or = grade 3 with the folfox1 regimen, 45% with the folfox2 and 40% with the folfox3. The folfox regimens achieve a high response rate in pretreated patients with CRC. Further studies are needed to determine the best oxaliplatin dose-intensity.     

Response of primary colon cancer cells in hybrid spheroids to 5-fluorouracil

The specific role of sulfhydryl groups in cell-mediated cytotoxicity (CMC) is still unknown. Here we demonstrate that natural killer cells and lymphokine-activated killer cells, when incubated with phenylarsine oxide (PAO), an organoarsenic compound showed a dose- and time-dependent inhibition of CMC and antibody-dependent cellular cytotoxicity (ADCC). PAO interacted directly with the effector cells (EC) without affecting the target cells (TC) or EC:TC conjugate formation. The loss of cytotoxicity was not due to lack of degranulation or to inhibition of serine esterases in PAO-treated cells. However, PAO inhibited the target-induced down regulation of phosphatidylinositol (PI) level in NK cells indicating that PAO blocked the cytolytic cascade at an early stage, upstream of PI. In addition, PAO also did not affect the disulfide link of the zeta-chain dimers, implicated in signal transduction in cytotoxic lymphocytes but did cause the rapid phosphorylation of the zeta chain. Finally, the effect of PAO on CMC was competitively blocked by dithiothreitol, a dithiol, but not by beta-mercaptoethanol, a mono-thiol. Taken together, these results indicate for the first time how sulfyhydryl groups may regulate CMC and ADCC.     

Influence of intraperitoneal 5-fluorouracil plus folinic acid on the healing of colonic anastomoses in rats

We studied the effects of intraperitoneal 5-fluorouracil (5-FU) administration with or without the addition of folinic acid (FA) on the healing of colonic anastomoses in rats immediately after surgery. Sixty-three male Wistar rats were subjected to colonic anastomosis. During surgery, the rats were randomized into one of three groups. Therapy was administered as an intraoperative intraperitoneal injection which was repeated once daily for the first 2 postoperative days. A 0.9% NaCl solution was administered to the rats in the control group. In group 2, we injected 5-FU (20 mg/kg/day) and in group 3 5-FU (20 mg/kg/day) plus FA (2 mg/kg/day). The rats were sacrificed on postoperative days 3, 5 or 8. Rupture of the anastomosis was significantly higher in the rats of groups 2 and 3, compared with the control group (p < 0.05). There were, however, no differences between groups 2 and 3. Formation of adhesions and abscesses was more common in groups 2 and 3 than in the control group for all study days. A significant difference in the anastomotic bursting pressure was measured for the control group in comparison to groups 2 and 3 on days 5 and 8 (p < 0.05). Histologic evaluation also showed a more profound inflammatory reaction and delayed healing of the anastomoses in groups 2 and 3, compared to the control group. Therefore, the perioperative intraperitoneal administration of 5-FU can inhibit the healing of colonic anastomoses in rats. The addition of an intraperitoneal injection of FA does not aggravate this negative effect.     

Enhanced cytotoxicity with interleukin-1 alpha and 5-fluorouracil in HCT116 colon cancer cells

Recombinant human interleukin-1 alpha (rIL-1 alpha), at concentrations that were not growth-inhibitory when given alone (100-10,000 U/ml), enhanced the growth inhibition resulting from a 72-h fluorouracil (FUra) exposure in HCT116 colon cancer cells. Median-effect analysis of clonogenic assays indicated that rIL-1 alpha, given 24 h prior to and following a 24-h exposure to FUra, increased lethality in a more than additive fashion. rIL-1 alpha did not appear to significantly affect [3H]-FUra metabolism, total [3H]-FUra-RNA incorporation or RNA retention after drug removal, inhibition of thymidylate synthase, or thymidine triphosphate pool depletion. During continuous exposure to rIL-1 alpha, transient stimulation of RNA and DNA synthesis was observed at 72 h, with a return to normal by 96 h. A 24-h exposure to 10 microM FUra altered the elution profile of newly synthesized DNA as monitored by pH step alkaline elution. An accumulation of lower-MW single-stranded DNA species was noted with FUra compared to control, accompanied by a significantly decreased proportion of DNA retained on the polycarbonate filter: 10% retained vs. 32% for control (P = 0.01). A 48-h exposure to rIL-1 alpha alone did not affect the elution profile of nascent DNA species, nor did it enhance the effects of FUra. Although FUra did not appreciably affect pulse [3H]-uridine incorporation into RNA for the initial 8-24 h of FUra exposure, progressive inhibition of net RNA synthesis was observed thereafter. FUra prevented the stimulatory effect of rIL-1 alpha on RNA synthesis, and net RNA synthesis was significantly inhibited (by 64-79% after 72 and 96 h) with the combination compared to rIL-1 alpha alone. Continuous exposure to 10 microM thymidine did not rescue cells from the lethality of FUra alone or the combination of FUra plus rIL-1 alpha, suggesting that depletion of deoxythymidine triphosphate as a consequence of thymidylate synthase inhibition was not the most important component of FUra toxicity. In contrast, 1 mM uridine provided partial protection against the toxicity of FUra alone or with rIL-1 alpha. Although uridine did not affect FUra metabolism, it decreased FUra-RNA incorporation by 42-60%, presumably as a consequence of the 2-fold expansion of UTP pools. [125I]-rIL-1 alpha binding was nonspecific; with a 24-h exposure, however, internalized [125I]-rIL-1 alpha exceeded cell surface-bound material by 2-fold.(ABSTRACT TRUNCATED AT 250 WORDS)     

Effect of 5-fluorouracil and UFT on experimental liver metastasis model of colorectal cancer using mouse colon 26 cells

Effects of 5-fluorouracil (5-FU) and UFT on an experimental liver metastasis model were compared at equi-effective dosage levels against subcutaneous tumor of mouse colon 26. 5-FU at the dosage level of 40 mg/kg suppressed the subcutaneous tumor growth by 70.0% and 45.0% on day 13 and day 18, respectively, and UFT at 20 mg/kg provided almost equal suppression (63.0% and 48.0%). In the liver metastasis model, 5-FU at 40 mg/kg showed more potent prevention of the formation of metastatic foci (94.9%) than did UFT (60.4%) at 20 mg/kg. 5-FU at 40 mg/kg produced a much higher peak serum level of 5-FU than did UFT at 20 mg/kg and also showed a much higher AUC (area under the curve) level in the portal blood. These results suggest that oral administration of 5-FU might be useful in prevention of liver metastasis of colorectal cancer.     

Search for the optimal schedule for the oxaliplatin/5-fluorouracil association modulated or not by folinic acid: preclinical data

The combination of oxaliplatin (LOHP)-5-fluorouracil (FU)-folinic acid (FA) has provided high response rates in pretreated patients with advanced colorectal cancer that is resistant to FU-FA. However, the choice of the optimal schedule between LOHP, FU, and FA remains open. The purpose of the present study was to compare, at equivalent drug area under the curve, different schedules for the LOHP-FU +/- FA combinations on four human colorectal cancer cell lines. FU +/- FA was tested as a 2-h short exposure ("bolus"), a 118-h continuous exposure ("infusion"), or a 22-h mixed exposure ("De Gramont protocol"). LOHP was administered for 2 h before, during, or after FU +/- FA exposure. Isobologram analyses revealed that LOHP associated with FU +/- FA resulted in synergistic cytotoxic effects whatever the tested schedules (in > or = 75% of cases). For the FU-LOHP combination, cytotoxicity was significantly different according to the FU exposure type (short > mixed > continuous) and was independent of the LOHP position. In contrast, for the FU-FA-LOHP combination, neither the FU exposure type nor the LOHP position significantly influenced cytotoxicity. The presence of FA significantly enhanced the cytotoxicity of FU-LOHP (P < 0.001); this potentiation was independent of the FU exposure type and was significantly influenced by the LOHP position (LOHP after FU-FA > LOHP during FU-FA > LOHP before FU-FA). In conclusion, in contrast with the recognized superiority of continuous FU exposure over short exposure when the drug is given alone, the FU-LOHP combination is more cytotoxic when FU is given as a short exposure. This suggests the potential interest of such a schedule in the clinical setting.     

Colon carcinoma cells blocked in polarization exhibit increased expression of carcinoembryonic antigen

Carcinoembryonic antigen (CEA) is an oncofetal protein whose regulation is poorly understood, although CEA is commonly expressed on many carcinoma cell types and enhances experimental metastases. The abundance of membrane-associated CEA was increased 3-fold when HD6 colon carcinoma cells were prevented from polarizing by culture for 3 days in low calcium medium. Polarization is an early event in HD6 cell differentiation, with the polarized cells forming a tight, laterally adherent monolayer by culture in normal calcium medium. Lateral adherence can occur because 3 days of culture in normal calcium medium increases expression of calcium-dependent intercellular adhesion proteins: a 35-fold increase in membrane abundance of LCAM and a 16-fold increase in membrane abundance of the desmosomal protein desmoglein I. Polarized HD6 cells exhibit low levels of CEA only at their apical luminal surface. Rounded, unpolarized HD6 cells do not exhibit increases in either LCAM or desmoglein I membrane expression, but express increased levels of CEA molecules throughout their cell surface, where they act as intercellular adhesion molecules, allowing unpolarized cells to form random cell to cell contacts. Cells cultured in low calcium medium form calcium-independent cell aggregates whose formation can be blocked by Fab' fragments of anti-CEA monoclonal antibody col-1. The familiar pattern of random, multilayered associations of tumor cells both in vitro and in xenographs in vivo may be due to intercellular adhesion mediated by CEA which is up-regulated and expressed throughout the cell surface of unpolarized tumor cells.     

Cytarabine and cisplatin as salvage therapy in patients with metastatic colorectal cancer who failed 5-fluorouracil + folinic acid regimen. French Northern Oncology Group

The combination of CDDP and ARA-C has shown some clinical efficiency as first-line therapy in advanced colorectal cancer. Our study was aimed to evaluate the therapeutic activity of this combination in advanced colorectal cancer who failed 5-fluorouracil (FU) and folinic acid (LV) regimen. Seventeen patients with measureable metastatic colorectal cancer who failed 5FU-LV therapy as first line (n = 14) or second line treatment (n = 3), entered the study. Three patients who recurred during adjuvant treatment with 5FU and levamisol, were also included. Median age was 59.5 (40-69). Performance status was as follows: 0 (n = 5), 1 (n = 11), 2 (n = 3), 3 (n = 1). Site of metastases included liver (n = 16), lung (n = 7), abdomen (n = 2), pelvic recurrences (n = 2), cutaneous (n = 1). Seven patients had 2 metastatic sites and two 3. The treatment was given as follows: ARA-C 75 mg/m2/day, days 1-3, followed 1 hour later by CDDP 30 mg/m2/day, days 1-3, every 28 days. The median number of cycles was 3 (range: 1-6 cycles). All patients but one were evaluable for both response and toxicity. Of these patients, 50% experienced severe hematologic toxicity and nonhematologic toxicity mainly consisted of fatigue and/or vomiting. No objective response was observed, but there were 3 stabilizations and 16 progressive diseases. Median time to progression was 10 weeks. Thus, the CDDP/ARA-C regimen is not of clinical value as salvage therapy in advanced colorectal cancer because of its toxicity and its lack of efficiency.     

5-Fluorouracil, folinic acid, etoposide and cisplatin chemotherapy for locally advanced or metastatic carcinoma of the oesophagus

38 patients with advanced oesophageal carcinoma were treated with intravenous (i.v.) folinic acid (300 mg/m2), 5-fluorouracil (500 mg/m2), etoposide (100 mg/m2), and cisplatin (30 mg/m2) (FLEP), on days 1, 2 and 3, every 22-28 days. 26 patients had locally advanced disease (LAD) and 12 had metastatic disease (M1). Oesophagectomy was planned for patients with LAD in case of tumour regression after chemotherapy, while patients with M1 disease received chemotherapy only. The overall remission rate was 45% (17/38) including four clinical and two pathologically confirmed complete remissions. 16 patients underwent oesophagectomy, 12 after response to FLEP, and 4 after FLEP and subsequent irradiation +/- 5-fluorouracil/mitomycin. Toxicity was mainly haematological, with WHO grade 3 and 4 leukocytopenia in 50% and thrombocytopenia in 31% of the patients. Two treatment-related deaths were observed; one due to chemotherapy and one postoperatively. Median survival time of LAD patients was 13 months, and actuarial 2-year survival was 31%. Patients with complete tumour resection after FLEP had a median survival time of 18 months and a 2-year survival rate of 42%. Median survival of M1 patients was 6 months. FLEP is an active combination for oesophageal cancer, especially when used preoperatively in LAD.     

The action of 5-fluorouracil on human HT29 colon cancer cells grown in SCID mice: mitosis, apoptosis and cell differentiation

BACKGROUND: The effect of oxygen toxicity in human airways is still poorly documented. We prospectively evaluated the inflammatory reaction induced by nasal oxygen exposure in an experimental setting. METHODS: Healthy subjects without nasal symptoms were exposed to high FIO2 during 5 h. Oxygen was delivered from a tank at a flow of 4 l/min to one nostril of each subject and both nostrils were studied. Mucociliary clearance was measured as saccharine nasal transit time (SNTT). Nasal lavage was performed with 5 ml normal saline and the fluid recovered was processed for cytology and measurements of cytokines concentrations: TNF alpha, IL-6, IL-8 and soluble ICAM-1. Under local anaesthesia, biopsies were performed for immunochemistry and electron microscopy. RESULTS: After oxygen exposure mucociliary clearance decreased and SNTT increased from 16 [9-21] to 20.5 [14-32] min (median and extremes; P < 0.1). In the lavage fluid, concentration of IL-6 was higher in the oxygen-exposed nostril (40.5 [11-128] pg/ml) than in the non-exposed one (7 [0-34] pg/ml; P < 0.05). There was also a trend for a higher IL-8 in the exposed than in the non-exposed nostril, (respectively 501 [214-587] pg/ml and 214 [122-616] pg/ml, P < 0.08), and for a higher number of polymorphonuclear cells in exposed nostril. In the mucosal biopsies substance P was not found, but ICAM-1 expression was higher in the mucosa and submucosa of the exposed nostrils where mast cells were also more abundant and showed piecemeal degranulation. CONCLUSION: In summary, we found clinical, functional and biological evidence of ongoing nasal inflammation following high FIO2 inhalation for 5 h. Since the histology and behaviour of nasal and bronchi mucosa are very similar, the same inflammatory events are likely to be occurring in the bronchi upon high concentrations of inhaled oxygen.     

Comparison of 5-fluoro-2'-deoxyuridine with 5-fluorouracil and their role in the treatment of colorectal cancer

Despite more than 30 years of intensive studies on new drugs against advanced colorectal cancer, the fluoropyrimidines remain the drugs of choice for systemic treatment and for hepatic artery infusion (HAI). This overview describes new developments in advanced colorectal cancer chemotherapy, providing a rationale for more effective use of the fluoropyrimidines, with biochemical modulation, scheduling or by revealing biochemical mechanisms of action that correlate with antitumour activity. In human colorectal cancer cell lines and various animal tumour model systems 5-fluoro-2'-deoxyuridine (FdUrd) is more effective than 5-fluorouracil (5-FU). Comparably, FdUrd's modulation by leucovorin (LV) is more potent than 5-FU. In animal studies it is shown that intermittent high-bolus administration of FdUrd generates better antitumour activity, compared with equal toxic doses or any other schedule of 5-FU. These effects are related to prolonged-thymidylate synthase (TS) inhibition and the prevention of TS induction, rather than RNA incorporation. Preclinical studies with modulators such as N-phosphonacetyl-L-aspartate (PALA), WR-2721, mitomycin C and platinum derivatives provide a rationale for clinical use in the future. The first choice systemic chemotherapy of patients with advanced colorectal cancer remains 5-FU combined with LV. Some improvement in therapeutic efficacy has been achieved with locoregional HAI. In randomised studies HAI FdUrd improves the quality of life and survival as compared with optimal systemic therapy. Chronomodulation decreases toxicity, allowing dose intensification, while modulators such as LV or dexamethasone increase survival of patients treated with HAI FdUrd to 86% after 1 year. In conclusion, the clinical use of FdUrd has not been fully explored. Intermittent high-dose FdUrd, chronomodulation together with the use of modulators or drugs focused on prolonged TS inhibition, should be studied in large randomised studies.     

Successful treatment of recurrent multiple lung metastasis from colon cancer with combination chemotherapy using methotrexate, 5-fluorouracil, and high-dose leucovorin: a case report

A 66-year-old man, who had received sigmoidectomy for sigmoid cancer in 1985, was diagnosed as having multiple lung and liver tumors in September 1988. When celiac-angiography was performed, recurrent liver metastases from sigmoid cancer were suspected and he received a transarterial embolism with ADM 30 mg and MMC 20 mg. In addition, he was treated with a sequential chemotherapy with methotrexate (MTX), 1,200 mg intravenously (6 h-infusion) followed by 5-fluorouracil (5-FU), 600 mg/m2/day and leucovorin, 300 mg/body/day in continuous infusion for 5 days from day 2 with concomitant oral administration of dipyridamole (300 mg/day) over 14 days. Treatment was repeated every 28 days for two courses. For the third course, administration of only 5-FU, leucovorin and dipyridamole was performed. As a result, the size of pulmonary lesions was prominently reduced on computed tomography. Although mucositis, anal erosion, diarrhea and thrombocytopenia were noted, no severe side effects were observed. This sequential chemotherapy appears useful for metastatic lesions from colon cancer.     

Effect of adding the topoisomerase I poison 7-ethyl-10-hydroxycamptothecin (SN-38) to 5-fluorouracil and folinic acid in HCT-8 cells: elevated dTTP pools and enhanced cytotoxicity

Checkpoints maintain the interdependency of cell cycle events by permitting the onset of an event only after the completion of the preceding event. The DNA replication checkpoint induces a cell cycle arrest until the completion of the DNA replication. Similarly, the DNA damage checkpoint arrests cell cycle progression if DNA repair is incomplete. A number of genes that play a role in the two checkpoints have been identified through genetic studies in yeasts, and their homologues have been found in fly, mouse, and human. They form signaling cascades activated by a DNA replication block or DNA damage and subsequently generate the negative constraints on cell cycle regulators. The failure of these signaling cascades results in producing offspring that carry mutations or that lack a portion of the genome. In humans, defects in the checkpoints are often associated with cancer-prone diseases. Focusing mainly on the studies in budding and fission yeasts, we summarize the recent progress.     

Transcription and activity of 5-fluorouracil converting enzymes in fluoropyrimidine resistance in colon cancer in vitro

Cellular resistance to 5-fluorouracil (5-FU) is not completely understood. Since 5-FU shares the pyrimidine pathway with the physiological pyrimidines, we investigated the relationship between fluoropyrimidine metabolism, nucleic acid uptake and cytotoxicity of 5-FU in eight colon tumour cell lines including 5-FU-resistant subclones. The cytotoxicity of 5-FU was increased up to 423-fold when the anabolites 5-fluorouridine (FUrd), 5-fluorodeoxyuridine (FdUrd), and 5-fluorodeoxyuridine monophosphate (FdUMP) were compared with the parent drug in vitro. The enzymes uridine phosphorylase and thymidine phosphorylase were predictive for the cytotoxicity of 5-FU in 5/7 cell lines. Inhibition of uridine phosphorylase and thymidine phosphorylase by antisense strategies effectively antagonised 5-FU, abolishing 84% and 79% of its toxicity. The importance of thymidine phosphorylase was supported by a highly restricted enzyme activity in 5-FU-resistant cells. In 5-FU naive cells, a stimulating effect of 5-FU on thymidylate synthase mRNA and ribonucleotide reductase mRNA expression was observed. In these cells, antisense oligonucleotides to ribonucleotide reductase significantly reduced cell growth. Downregulation of ribonucleotide reductase mRNA in 5-FU-resistant subclones suggests different mechanisms in primary and secondary resistance to 5-FU. Most of the intracellular 5-FU was selectively incorporated into RNA (range: 45-91%) and generally spared DNA (range: 0.2-11%). In synthesising our data, we conclude that drug resistance could be overwhelmed through bypassing limiting steps in the activation of 5-FU. In the majority of colonic tumours, the activity of uridine phosphorylase and thymidine phosphorylase may have prognostic relevance for the cytotoxicity of 5-FU in vitro.