Biochemical Modulation of 5-FU in Systemic Treatment of Advanced Colorectal Cancer

January 2, 2001
Jean L. Grem, MD, FACP
Jean L. Grem, MD, FACP

Volume 15, Issue 1

Randomized studies have tested a variety of strategies to improve the activity of 5-fluorouracil (5-FU) in colorectal cancer patients. Results from 14 randomized trials comparing 5-FU administered via intravenous ( IV) bolus either

ABSTRACT: Randomized studies have tested a variety of strategies to improve theactivity of 5-fluorouracil (5-FU) in colorectal cancer patients. Results from 14randomized trials comparing 5-FU administered via intravenous ( IV) bolus eitheras a single agent or modulated by leucovorin indicate a significantly higherresponse rate with 5-FU/leucovorin (25% vs 13% of assessable patients).Sequential methotrexate followed by IV bolus 5-FU is associated with a higherresponse rate. Continuous infusion schedules also produce superior responserates compared to bolus 5-FU alone. Published meta-analyses indicate a small,but statistically significant, survival advantage for methotrexate/5-FU andinfusional 5-FU, but not for leucovorin-modulated 5-FU. Although the incidenceof hand-foot syndrome is higher with protracted infusional 5-FU, the incidenceof other toxicities including myelosuppression, diarrhea, and mucositis is low.Oral administration of 5-FU may simulate infusional schedules while avoidingcatheter-related complications. [ONCOLOGY 15(Suppl 2):13-19, 2001]

Introduction

Until irinotecan (Camptosar) becameavailable, 5-fluorouracil (5-FU) and 5-fluoro-2'-deoxy-uridine (FUDR) comprisedthe only commercially available treatment options that could be consideredstandard of care for advanced colorectal cancer. An understanding of theintracellular metabolism of 5-FU and its mechanism of action has suggestedpotential therapeutic strategies to improve its clinical activity. A variety ofapproaches to modulate 5-FU in patients with colorectal cancer have been testedin randomized studies over the past 15 years.

Leucovorin Modulation of 5-FU

Inhibition of thymidylate synthase by the 5-FU metabolite5-fluoro-2'-deoxyuridine monophosphate (FdUMP) is one of the principalmechanisms of 5-FU action.[1] Thymidylate synthase is an important therapeutictarget because it catalyzes the de novo formation of thymidylate (dTMP) fromdeoxyuridylate (dUMP). During this process, a methyl group is transferred fromthe reduced folate cofactor 5,10-methylenetetrahydrofolate to the carbon-5position of the uracil moiety. Thymidylate is subsequently converted tothymidine triphosphate, which is required for DNA synthesis and repair. Incontrast to the interaction of thymidylate synthase with its physiologicalsubstrate, the TS-FdUMP-folate complex is only slowly reversible.

The stability of the ternary complex is influenced by theintracellular level of 5,10-methylenetetrahydrofolate and the extent of itspolyglutamation. Pharmacologic concentrations of 5-formyltetrahydrofolate(leucovorin) expand the intracellular pools of 5,10-methylenetetrahydrofolate,thereby increasing the extent and duration of 5-FU-mediated thymidylate synthaseinhibition. This strategy for biochemical modulation of 5-FU by leucovorin hasproven successful in the clinical setting. The two most commonly employed5-FU/leucovorin regimens involve either weekly administration of IV bolus 5-FUdelivered at the midpoint of a 2-hour IV infusion of high-dose leucovorin, or amonthly (daily for 5 days every 4-5 weeks) schedule of IV bolus leucovorin and5-FU.

Intravenous Bolus Administration

Published results are currently available from 14 randomizedtrials that compared 5-FU administered by IV bolus either as a single agent ormodulated by leucovorin.[2-16] The response rates are based on the number ofeligible and assessable patients reported by each investigator, and thereforeare somewhat higher than they would be if all patients randomized to that armwere included in the denominator. Two trials included patients withnonmeasurable disease.[8,12] The combined data for all 14 trials indicate aresponse rate of 25.3% ± 2.4% (95% confidence interval) among 1,262 patientstreated with 5-FU/leucovorin compared to 13.8% ± 2.2% among 995 patientstreated with 5-FU alone. For time to progression, the median of the reportedvalues was 6.0 months with 5-FU/leucovorin and 4.2 months for 5-FU alone (Figure1). The median of the reported survival times was 12.4 months for5-FU/leucovorin and 11.0 months for 5-FU alone.

The original meta-analysis that assessed the value of bolus5-FU/leucovorin in patients with advanced colorectal cancer employed individualdata provided by the principal investigators from nine randomized trials.[6] All1,381 patients entered in these trials were included on an intent-to-treatbasis. There was no significant difference in median survival (11.5 vs 11.0months, odds ratio 0.97).

Finding the Optimal Schedule

A second generation of trials sought to establish the optimal5-FU/LV schedule. A randomized trial conducted by the North Central CancerTreatment Group (NCCTG) compared a monthly schedule of low-doseleucovorin-modulated 5-FU with a weekly schedule of high-doseleucovorin-modulated 5-FU.[17] In this trial, the overall response rate among102 assessable patients on the monthly schedule was 34% and among 98 patients onthe weekly schedule, it was 31%. Both the median time to tumor progression (~4.4 and ~ 6.3 months) and median survival (9.3 and 10.7 months) were somewhatlonger with the weekly schedule, but these differences were not statisticallysignificant. If the response rates for the monthly and weekly schedules for 5-FUalone or with leucovorin from the 14 randomized studies cited above are comparedside by side, the results are virtually identical (Figure2). Thus, these twoschedules of 5-FU/leucovorin can be considered therapeutically equivalent inadvanced colorectal cancer, although the spectrum of toxicity differs.

Dose-limiting diarrhea occurred more frequently with the weeklyschedule, while severe mucositis was uncommon. In contrast, mucositis tended tobe dose-limiting with the monthly schedule; however, severe diarrhea alsooccurred. The monthly schedule was considered the "gold-standard" forpivotal trials in colorectal cancer in the United States since an NCCTG studyshowed a survival advantage with leucovorin-modulated 5-FU.[12] Furthermore,because of the cost of leucovorin, the low-dose monthly regimen was also lessexpensive. However, the weekly schedule remains popular as it is more convenientfor some patients and is perceived as easier to manage because treatment can beinterrupted if early signs of toxicity occur. In contrast, with the monthlyschedule, clinical toxicity generally does not develop until after all 5 days oftreatment have been completed.

High- or Low-Dose Leucovorin

A number of randomized trials focused on ascertaining whetherhigh-dose leucovorin was necessary. Three randomized trials compared 20 mg/m2 ofracemic (d,l-) leucovorin or 10 mg/m2 of pure l-leucovorin with a ten-foldhigher dose administered with 5-FU (370-425 mg/m2) on the monthlyschedule.[18-20] None of the trials showed an advantage for high-dose leucovorinin terms of either response rate, time to progression, or survival. Two othertrials compared 20 or 25 mg/m2 leucovorin to 500mg/m2 leucovorin given as a2-hour infusion with 5-FU at 500 or 600 mg/m2 IV bolus at the midpoint.[3,21]Although the response rate was somewhat higher with high-dose leucovorin, it wasnot significant. Further, there was no difference in survival.

The impact of bolus 5-FU/leucovorin on survival has been greaterin the adjuvant setting. Randomized trials demonstrating a disease-free andoverall survival benefit with weekly bolus 5-FU/leucovorin as adjuvant therapyin patients with colorectal cancer used high-dose leucovorin (500 mg/m2) givenas a 2-hour infusion.[22,23] Because the benefit of low-dose leucovorin withweekly bolus 5-FU has not been tested in the adjuvant setting, caution should beexercised in extrapolating the data from advanced disease to the adjuvantsetting. In contrast, in adjuvant colon cancer trials evaluating the monthlyschedule, a benefit was demonstrated with both high-dose (200 mg/m2) andlow-dose (20 mg/m2) leucovorin.

Methotrexate/5-FU

Other strategies have been investigated to modulate the activityof 5-FU. Sequence-dependent synergy was noted in preclinical studies withmethotrexate preceded by 5-FU, while the opposite sequence wasantagonistic.[1,24] Inhibition of thymidylate synthase reduces consumption ofreduced folate cofactors that would otherwise be used in the conversion ofdeoxyuridine monophosphate to thymidine monophosphate. The availability of thereduced folates permits continued purine biosynthesis despitemethotrexate-mediated inhibition of dihydrofolate reductase. Administration ofmethotrexate prior to 5-FU establishes inhibition of purine biosynthesis. As aconsequence, intracellular levels of phosphoribosylpyrophosphate increase, andthis promotes the anabolism of 5-FU to FUMP by orotatephosphoribosyltransferase.

A number of phase III trials in advanced colorectal cancer havecompared methotrexate given at various intervals prior to 5-FU, with mixedresults. One pharmacodynamic study suggested that a 24-hour interval betweenmethotrexate and 5-FU was optimal for expansion of phosphorbosyl pyrophosphate(PRPP) pools.[25] A clinical study in advanced colorectal cancer that directlycompared a 24-hour interval between administration of methotrexate and 5-FU witha 1-hour interval demonstrated that the longer interval was superior.[26] Ameta-analysis of eight randomized trials involving 1,178 patients that compared5-FU alone with methotrexate-modulated 5-FU (± leucovorin rescue) revealed asignificantly higher response rate (19% vs 10%, odds ratio 0.51, P < .0001)and survival advantage (median, 10.7 vs 9.1 months, odds ratio 0.87, P = .024)in favor of methotrexate/5-FU.[27] However, because many of themethotrexate-containing arms included leucovorin rescue, at issue is whether theapparent benefit might have been due to leucovorin.

Administration of 5-FU by Continuous Infusion

Preclinical models indicate that both concentration and durationof exposure are important determinants of 5-FU-associated cytotoxicity.[1,28-31]The half-life of 5-FU in plasma is very short, and potentially cytotoxicconcentrations are maintained for only a few hours. Continuous infusion of 5-FUallows sustained plasma exposure. Several randomized trials have shown improvedactivity with infusional 5-FU over IV bolus 5-FU.[32-34] A randomized trialconducted by the Mid-Atlantic Oncology Program comparing 300 mg/m2/d as aprotracted infusion had a four-fold higher response rate compared with bolus5-FU (29.9% vs 6.9%).[32] The Eastern Cooperative Oncology Group (ECOG)confirmed a significantly higher response rate with protracted infusional 5-FUcompared to bolus 5-FU, 28.3% vs 16.3%.[33] However, none of the individualtrials showed a survival advantage.

Bolus vs Continuous Infusion

A meta-analysis of six randomized trials involving 1,219patients with colorectal cancer comparing bolus with infusional 5-FU found asignificantly higher response rate (22% vs 14%, odds ratio 0.55, P = .0002) andan advantage in median survival (12.1 vs 11.3 months, odds ratio 0.88, P = .04)in favor of infusional 5-FU.[34] Although hand-foot syndrome occurred morefrequently with infusional 5-FU, serious hematologic toxicity was seen in lessthan 5% of patients.[35,36] In contrast, grade 3-4 myelosuppression occurred in34% of patients treated with bolus 5-FU.

More recently, numerous intermittent, high-dose infusionalregimens have been evaluated, including weekly high-dose 5-FU infusions over 24or 48 hours, and an every 2-week schedule involving bolus and infusional5-FU.[37-40] A randomized phase II trial from the Southwest Oncology Group(SWOG) indicated that two arms featuring either 300 mg/m2 5-FU as a protractedinfusion or 2,600 mg/m2 as a weekly 24-hour infusion were at least as effectiveas IV bolus 5-FU ± leucovorin, but were associated with less toxicity.[8]Preliminary results from a five-armed phase III trial conducted jointly by ECOGand the Cancer and Leukemia Group B (CALGB) demonstrated that 2,600 mg/m2 5-FUgiven as a weekly 24-hour infusion was preferable to leucovorin-modulated bolus5-FU in view of comparable median survivals but a more favorable toxicityprofile. Grade 3-4 diarrhea occurred in only 13% of patients vs 32% to 34% ofpatients on two bolus 5-FU/leucovorin arms.[41]

De Gramont and colleagues developed a regimen that contains bothbolus and infusional 5-FU. The original de Gramont regimen entailed 5-FU 400mg/m2 bolus with leucovorin 200 mg/m2/2h followed by 600mg/m2 5-FU over 22hours days 1 and 2 every 2 weeks.[40] A randomized clinical trial demonstratedthat the de Gramont regimen was associated with both significantly higherresponse rates and a longer time to progression compared to the monthly MayoClinic 5-FU/leucovorin regimen.[40] Aranda et al found that a high-dose weekly48-hour 5-FU infusion (3500 mg/m2/48h) was superior to the Mayo bolus5-FU/leucovorin regimen in terms of response rate and time to progression.[39]

Leucovorin Modulation and the Infusional Regimens

The question of whether leucovorin modulation is beneficial inthe setting of infusional 5-FU regimens has been addressed in several trials.Two of the seven arms tested by the SWOG in advanced colorectal cancer patientsinvolved protracted infusion of 5-FU administered alone for 28 of 35 days (300mg/m2/d) or at a reduced dose (200 mg/m2/d) with weekly leucovorin 20mg/m2.[8]There were no differences in response rate (30% and 28%), median time toprogression (6 and 6 months), or survival (15 and 14 months).

A German Cooperative Oncology Group (AIO) added high-doseleucovorin 500 mg/m2 as a 2-hour infusion to the weekly high-dose 24-hour 5-FUinfusional regimen developed by Ardalan, which served as the control arm in aseries of randomized phase III trials.[38] An intergroup phase III trialconducted in Europe compared 2,600 mg/m2 5-FU as a weekly 24-hour infusion aloneor with a 2-hour infusion of high-dose leucovorin. Preliminary results suggestthat the leucovorin-modulated 24-hour infusion arm had a significantly higherresponse rate (20% vs 10%) and a longer time to progression (6.4 vs 4.4 months)but is associated with a higher incidence of grade 3-4 diarrhea.[42]

Thus, a variety of infusional 5-FU regimens are available thathave clinical activity in patients with colorectal cancer, and each appears tohave an advantage in terms of toxicity profile. A disadvantage of this treatmentapproach is catheter-related complications. About 15% to 20% of patients willrequire catheter removal for thrombosis, sepsis, malposition, or breakage, andan additional 10% to 15% of patients will have infections requiring antibiotictreatment only. These considerations have fueled interest in developing oral5-FU regimens that can simulate the IV infusional schedules while avoiding theinconvenience and morbidity associated with indwelling venous catheters and theneed for a pump.

Unsuccessful Modulation Strategies

Preclinical studies suggested that interferon enhances thecytotoxicity of 5-FU, and most of the data suggested increased DNA damage as thefinal common pathway.[43] Based on promising clinical results in phase IIstudies,[44] numerous randomized trials have compared interferon alpha-modulated5-FU with or without leucovorin on a variety of schedules. Preliminary resultshave been presented from two meta-analyses using primary data from 3,254patients entered into randomized trials evaluating interferon alpha as amodulator of 5-FU.

These trials followed three general designs.[45] The first twodesigns compared the same schedule of 5-FU ± interferon alpha or5-FU/leucovorin ± interferon alpha, with 5-FU given either as a bolus or bycontinuous infusion. The meta-analysis of 12 trials comparing the addition ofinterferon alpha to either 5-FU alone or 5-FU plus leucovorin involved 1,766patients. There was no difference in response rate (25% vs 24%) or mediansurvival (11.4 months vs 11.5 months) in the absence or presence of interferon,respectively.[45]

The third trial design compared 5-FU/leucovorin vs5-FU/interferon alpha, and the meta-analysis combined seven trials and 1,488patients. There was an advantage in favor of 5-FU plus leucovorin over 5-FU plusinterferon in terms of tumor response (23% vs 18%, P = .042), while the survivalbenefit was of borderline significance.[45] These results indicate thatinterferon alpha does not increase the efficacy of either 5-FU or5-FU/leucovorin.

Modulation by PALA

Interest in using N-phosphonacetyl-L-aspartate (PALA), aninhibitor of aspartate carbamoyltransferase, as a modulator of 5-FU resurfacedwith the suggestion that low doses of PALA capable of producing biochemicaleffects in patients could be combined with a full-dose of 5-FU.[1,37,46] Tworandomized trials have compared a high-dose 24-hour weekly infusion of 2,600mg/m2 5-FU alone or combined with PALA 250 mg/m2 given 24 hours prior to 5-FU.Leichman et al reported that the PALA-containing arm offered no advantage to5-FU alone in terms of response rate (14% vs 24%), time to progression (4 vs 6months), and survival (11 vs 15 months), respectively.[8] The multi-armedECOG/CALGB study included 430 patients randomized to either 2,600 mg/m2 5-FUalone or preceded by low-dose PALA. Survival seemed to be better with 5-FU alone(14.8 vs 11.9 months).[41] These results have dampened interest in usinglow-dose PALA as a modulator of 5-FU.

Other therapeutic approaches have also been explored. Oneproposed strategy involves inhibition of nucleoside transport as a means ofaugmenting 5-FU toxicity by preventing influx of potential rescue nucleosides(thymidine) and by trapping potentially cytotoxic nucleosides intracellularly(5-fluorodeoxyuridine and deoxyuridine).[1,47] A phase III trial that compared5-FU 600 mg/m2 with leucovorin 300 mg/m2 IV days 2 to 4 every 3 weeks alone orwith dipyridamole 75 mg po three times daily on days 1 to 5 showed no advantagewith the addition of dipyridamole. Response rates were 15% vs 13%, and survivalwas 11.6 vs 9.3 months, respectively.[48]

Hydroxyurea (Hydrea), an inhibitor of ribonucleotide reductase,has been proposed as a modulator of 5-FU based on two considerations: first, itsability to decrease deoxyribonucleotide triphosphate pools; and second, toprevent conversion of fluorodeoxyuridine diphosphate to its correspondingdeoxyribonucleotide derivative.[49] A recent phase III trial assessed the impactof hydroxyurea 35 mg/kg administered in three divided oral doses every 8 hoursstarting 6 hours after administration of 5-FU on the clinical activity of5-FU/leucovorin (600/500 mg/m2) given weekly for 6 of 8 weeks. No significantimprovements in response rate, time to progression, or survival were seen.[50]

Conclusions

In summary, leucovorin modulation of 5-FU by IV bolus hasclearly been documented to improve the response rate, and may also improve thetime to tumor progression. However, a meta-analysis of nine trials indicated nooverall improvement in survival. Methotrexate-modulation of 5-FU ± leucovorinhas also been demonstrated to improve the response rate and have a modest impacton median survival. Continuous infusion of 5-FU on a protracted schedule is alsosuperior to 5-FU in terms of response rate and a modest improvement in survival.Side by side comparisons of the response rates and median survival in themeta-analysis for 5-FU/leucovorin, methotrexate/5-FU and infusional 5-FU suggestthat overall, each strategy produced similar results (Figure3).

Administration of 5-FU by either protracted venous infusion andas a weekly 24-hr infusion appears to be at least as effective asleucovorin-modulated IV bolus 5-FU. The incidence of severe myelosuppression,mucositis, and diarrhea are reduced, but catheter-related complications remain aconcern. The emerging role of oral administration of 5-FU as either a"prodrug" or in conjunction with inhibitors of 5-FU catabolism holdpromise of an improved therapeutic index and increased patient convenience. Morerecent trials suggest that the original de Gramont regimen and a weekly infusionof 3,500 mg/m2 5-FU delivered over 48 hours are superior to the Mayo Clinicbolus 5-FU/low-dose leucovorin regimen.[39,40] Ironically, it may be difficultto simulate these high-dose intermittent infusional regimens with the currentlyavailable oral formulations of capecitabine, uracil/ftorafur, andeniluracil/5-FU.

Questions and Answers

Leonard Saltz, MD: In the laboratory, leucovorin seems towork so well. The data look so airtight. Any thought about why they look sonebulous in the clinical arena?

Jean Grem, MD: In advanced colon cancer, it is not really clearthat we have made a huge impact on median survival with any of the single agentstrategies, so even when the meta-analyses have shown a significant (P< 0.5) survival benefit, we are still talking about less than 1 or 1 ½months. Whether those data are really clinically meaningful is open to question.It appears that activity is increasing, but so is toxicity. It is a question ofhow we know which patients or which tumors are leucovorin deficient. In most ofthe trials, there really is not that much information about whether or notpatients are indeed folate depleted. Maybe only a small percentage of patientswould actually benefit from the leucovorin modulation and we just do not knowwho those people are.

Daniel Haller, MD: The more recent studies of 5-FU/leucovorinreally show much lower response rates and time to treatment failure than theearlier studies. Of course, one explanation could be that the recent studies arejust not as good as the first studies. Since the survival advantage stayed thesame, I wonder if the other differences are really artificial. This speaks towhat the FDA says about appropriate surrogates for patient benefit, wheresurvival is the only one that cannot be fudged and time to treatment failure andresponse rate are somewhat artificial.

I think the response rates and progression free survival in thenewer studies are lower because the studies were written differently. Hardlyanyone you treat, in real life, ever is on treatment for only 3 months. Patientstend to be treated longer. In terms of response rates, one looks at the globalview of a patient rather than the situation in which one 3-mm lesion has nowgrown to 4 mm and you must take the patient off study. So I really do worry thateven with overall survival, we do not have a very good surrogate for how anindividual drug might do in first-, second-, or third-line therapy.

Although I cannot prove it, I think one of the reasons forultimate treatment failure in some studies is that patients are on study for 3months or less. I think we are forced, given the definitions in trials, to takepatients off while they are still potentially benefiting. We are still not surehow long to keep people on treatment. Maybe the initial course of therapy, evenif truncated, was enough induction therapy to carry over into ultimately betterresults in terms of survivorship.

Dr. Saltz: You bring up a couple of interesting pointsconcerning how rigorously we have to adhere to specified guidelines and how thecriteria for evaluation have changed. Last year, Paul Chang and I published someobservations that the response rates for neuroendocrine tumors probably are notas high as a previous study had suggested, and one of our hypotheses was thatthe criteria for evaluation are different. We have become much more rigorous inour demand for identification of response, of treatment failure, and so on. Ithas gotten tougher to have a good response. It has gotten tougher to haveprogression free survival because our criteria have changed.

Dr. Haller: I think that, in a disease like colorectal cancer,where 80% to 90% of patients who start out in clinical trials have an ECOG(Eastern Cooperative Oncology Group) performance status of 0 or 1, stability maynot be such a bad word. The concept of not getting a response is not a repulsivethought; this is unlike pancreatic cancer, where the goal has to be somewhatreversed and we need to improve people’s overall performance status. I thinkthat colorectal cancer cries out for a better measure of success than the oneswe now have, which are a "one size fits all" approach.

Robert Diasio, MD: I want to follow-up on one of the pointsmentioned in terms of the role of leucovorin and in particular, the cell linedata vs the clinical situation. In many clinical situations, the tumors arereally not folate depleted, while many laboratory experiments show there is noexogenous availability of folate. We need enough 5, 10-methylenetetrahydrofolateto have the system work, so we need to use an exogenous source of folate. Ithink it really speaks to the nutritional state of the patient. It is probablynot as necessary in the clinical setting as previously thought.

The other open question concerns the actual activity thatreduced folates have and the selectivity of that activity—whether it is reallya better way to give 5-FU, or if it simply gives us confidence to tolerate alittle more toxicity because it is not "just 5-FU."

References:

1. Grem JL: 5-Fluoropyrimidines, in: Chabner BA, Longo DL,(eds): Cancer Chemotherapy and Biotherapy. Principles and Practice, 2nd ed, p149. Philadelphia, Lippincott-Raven Publishers, 1996.

2. Petrelli N, Herrera L, Rustum Y, et al: A prospectiverandomized trial of 5-fluorouracil vs 5-fluorouracil and high-dose leucovorin vs5-fluorouracil and methotrexate in previously untreated patients with advancedcolorectal carcinoma. J Clin Oncol 5:1559-1565, 1987.

3. Petrelli N, Douglass HD, Herrera L, et al: The modulation offluorouracil with leucovorin in metastatic colorectal carcinoma. A prospectiverandomized phase III trial. J Clin Oncol 7:1419-1426, 1991.

4. Martoni A, Cricca A, Guaraldi M, et al: Randomized clinicaltrial with a weekly regimen of 5-FU vs 5-FU plus intermediate-dose folinic acidin the treatment of advanced colorectal cancer. Ann Oncol 3:87-88, 1992.

5. Nobile MT, Vidili MG, Sobrero A, et al: 5-Fluorouracil aloneor combined with high dose folinic acid in advanced colorectal cancer patients.A randomized trial (abstract). Proc Am Soc Clin Oncol 7:97, 1988.

6. Piedbois P, Buyse M, Rustum Y, et al (The Advanced ColorectalCancer Meta-Analysis Project): Modulation of fluorouracil by leucovorin inpatients with advanced colorectal cancer: Evidence in terms of response rate. JClin Oncol 10:896-903, 1992.

7. Laufman LR, Bukowski R, Collier MA, et al: A randomized,double-blind trial of fluorouracil plus placebo vs fluorouracil plus oralleucovorin in patients with metastatic colorectal cancer. J Clin Oncol11:1888-1892, 1993.

8. Leichman CG, Fleming TR, Muggia FM, et al: Phase II study offluorouracil and its modulation in advanced colorectal cancer: A SouthwestOncology Group Study. J Clin Oncol 131:1303-1311, 1995.

9. Erlichman C, Fine S, Wong A, et al: A randomized trial offluorouracil and folinic acid in patients with metastatic colorectal carcinoma.J Clin Oncol 6:469-475, 1988.

10. Valone FH, Friedman MA, Wittlinger PS, et al: Treatment ofpatients with advanced colorectal carcinomas with fluorouracil alone, high-doseleucovorin plus fluorouracil, or sequential methotrexate, fluorouracil, andleucovorin: A randomized trial of the Northern California Oncology Group. J ClinOncol 7:1427-1436, 1989.

11. Doroshow JH, Multhauf P, Leong L, et al: Prospectiverandomized comparison of fluorouracil vs fluorouracil and high-dose continuousinfusion leucovorin calcium for the treatment of advanced measurable colorectalcancer in patients previously unexposed to chemotherapy. J Clin Oncol 8:491-501,1990.

12. Poon MA, O’Connell MJ, Moertel CG, et al: Biochemicalmodulation of fluorouracil: Evidence of significant improvement of survival andquality of life in patients with advanced colorectal carcinoma. J Clin Oncol7:1407-1418, 1989.

13. Labianca R, Pancera E, Aitini E, et al: Folinic acid +5-fluorouracil (5-FU) vs equidose 5-FU in advanced colorectal cancer. Phase IIIstudy of "GISCAD" (Italian Group for the Study of Digestive TractCancer). Ann Oncol 2:673-679, 1991.

14. Di Costanzo F, Bartolucci R, Sofra M, et al: 5-Fluorouracilalone vs high- dose folinic acid and 5-FU in advanced colorectal cancer: Arandomized trial of the Italian Oncology Group for Clinical Research (abstract410). Proc Am Soc Clin Oncol 8:106, 1989.

15. Petrioli R, Lorenzi M, Aquino A, et al: Treatment ofadvanced colorectal cancer with high-dose intensity folinic acid and5-fluorouracil plus supportive care. Eur J Cancer 31A: 2105-2108, 1995.

16. Borner MM, Castiglione M, Bacchi M, et al: The impact ofadding low-dose leucovorin to monthly 5-fluorouracil in advanced colorectalcarcinoma: Results of a phase III trial. Swiss Group for Clinical CancerResearch (SAKK). Ann Oncol 9:535-541, 1998.

17. Buroker TR, O’Connell MJ, Wieand HS, et al: Randomizedcomparison of two schedules of fluorouracil and leucovorin in the treatment ofadvanced colorectal cancer. J Clin Oncol 12:14-20, 1994.

18. Poon MA, O’Connell MJ, Wieand HS, et al: Biochemicalmodulation of fluorouracil with leucovorin: Confirmatory evidence of improvedtherapeutic efficacy in advanced colorectal cancer. J Clin Oncol 9:1967-1972,1991.

19. Labianca R, Cascinu S, Frontini L, et al: High- vs low-doselevo-leucovorin as a modulator of 5-fluorouracil in advanced colorectal cancer:A "GISCAD" phase III study. Ann Oncol 8:169-174, 1997.

20. Ychou M, Fabbbro-Peray P, Perney P, et al: A prospectiverandomized study comparing high- and low-dose leucovorin combined with same-dose5-fluorouracil in advanced colorectal cancer. Am J Clin Oncol 21:233-236, 1998.

21. Jäger E, Heike M, Bernhard H, et al: Weekly high-doseleucovorin vs low-dose leucovorin combined with fluorouracil in advancedcolorectal cancer: Results of a randomized multicenter trial. J Clin Oncol14:2274-2279, 1996.

22. Wolmark N, Rockette H, Fisher B, et al: The benefit ofleucovorin-modulated fluorouracil as postoperative adjuvant therapy for primarycolon cancer: Results from National Surgical Adjuvant Breast and Bowel ProjectProtocol C-03. J Clin Oncol 10:1879-1887, 1993.

23. Haller DG, Catalano PJ, Macdonald JS, et al: Fluorouracil,leucovorin and levamisole adjuvant therapy for colon cancer: Four-year resultsof INT-0089 (abstract 940). Proc Am Soc Clin Oncol 16:265a, 1997.

24. Bertino JR: Biomodulation of 5-fluorouracil withantifolates. Semin Oncol 24 (suppl 18): S18-52-S18-56, 1997.

25. Kemeny N, Ahmed T, Michaelson R, et al: Activity ofsequential low-dose methotrexate and fluorouracil in advanced colorectalcarcinoma. Attempt at correlation with tissue and blood levels ofphosphoribosylpyrophosphate. J Clin Oncol 2:311-315, 1984.

26. Marsh JC, Bertino JR, Katz KH, et al: The influence of druginterval on the effect of methotrexate and fluorouracil in the treatment ofadvanced colorectal cancer. J Clin Oncol 9:371-380, 1991.

27. Piedbois P, Buyse M, Blijham G, et al, for the AdvancedColorectal Cancer Meta-analysis Project: Meta-analysis of randomized trialstesting the biochemical modulation of fluorouracil by methotrexate in metastaticcolorectal cancer. J Clin Oncol 12:960-969, 1994.

28. Santelli G, Valeriote F: Schedule-dependent cytotoxicity of5-fluorouracil in mice. J Natl Cancer Inst 76:159-164, 1986.

29. Moran RG, Scanlon KL: Schedule-dependent enhancement of thecytotoxicity of fluoropyrimidines to human carcinoma cells in the presence offolinic acid. Cancer Res 51:4618-4623, 1991.

30. Aschele C, Sobrero A, Faderan MA, et al: Novel mechanisms ofresistance to 5-fluorouracil in human colon cancer (HCT-8) sublines followingexposure to two different clinically relevant dose schedules. Cancer Res52:1855-1864, 1992.

31. Sobrero AF, Aschele C, Bertino JB: Fluorouracil incolorectal cancer- a tale of two drugs: Implications for biochemical modulation.J Clin Oncol 15:368-381, 1997.

32. Lokich JJ, Ahlgren JD, Gullo JJ, et al: A prospectiverandomized comparison of continuous infusion fluorouracil with a conventionalbolus schedule in metastatic colorectal carcinoma: A Mid-Atlantic OncologyProgram Study. J Clin Oncol 7:425-432, 1989.

33. Hansen RM, Ryan L, Anderson T, et al: Phase III study ofbolus vs infusion fluorouracil with or without cisplatin in advanced colorectalcancer. J Natl Cancer Inst 88:668-674, 1996.

34. Rougier P, Paillot B, LaPlanche A, et al: 5-Fluorouracil(5-FU) continuous intravenous infusion compared with bolus administration. Finalresults of a randomised trial in metastatic colorectal cancer. Eur J Cancer33:1789-1793, 1997.

35. Meta-Analysis Group In Cancer: Efficacy of intravenouscontinuous infusion of fluorouracil compared with bolus administration inadvanced colorectal cancer. J Clin Oncol 16:301-308, 1998.

36. Meta-Analysis Group In Cancer: Toxicity of fluorouracil inpatients with advanced colorectal cancer. Effect of administration schedule andprognostic factors. J Clin Oncol 16: 3537-3541, 1998.

37. Ardalan B, Singh G, Silberman HA: Randomized phase I andphase II study of short-term infusion of high-dose fluorouracil with or withoutn-(phosphonacetyl)-L-aspartic acid in patients with advanced pancreatic andcolorectal cancer. J Clin Oncol 6:1053-1058, 1988.

38. Kohne CH, Schoffski P, Wilke H, et al: Effectivebiomodulation by leucovorin of high-dose infusion fluorouracil given as a weekly24-hour infusion: Results of a randomized trial in patients with advancedcolorectal cancer. J Clin Oncol 16:418-426, 1998.

39. Aranda E, Diaz-Rubio E, Cervantes A, et al: Randomized trialcomparing monthly low-dose leucovorin and fluorouracil bolus with weeklyhigh-dose 48-hour continuous-infusion fluorouracil for advanced colorectalcancer: A Spanish Cooperative Group for Gastrointestinal Tumor Therapy (TTD)study. Ann Oncol 9:727-731, 1998.

40. de Gramont A, Bosset JF, Milan C, et al: Randomized trialcomparing monthly low-dose leucovorin and fluorouracil bolus with bimonthlyhigh-dose leucovorin and fluorouracil bolus plus continuous infusion foradvanced colorectal cancer: A French intergroup study. J Clin Oncol 15:808-815,1997.

41. O’Dwyer PJ, Ryan LM, Valone FH, et al: Phase III trial ofbiochemical modulation of 5-fluorouracil by IV or oral leucovorin or byinterferon in advanced colorectal cancer: An ECOG/CALGB phase III trial(abstract 469). Proc Am Soc Clin Oncol 15:207, 1996.

42. Schmöll HJ, Köhne CH, Lorenz M, et al: Weekly 24-hinfusion of high-dose 5-fluorouracil with or without folinic acid vs bolus5-FU/FA (NCCTG/Mayo) in advanced colorectal cancer: A randomized phase III trialof the EORTC, GITCCG and the AIO (abstract 935). Proc Am Soc Clin Oncol 19:241a, 2000.

43. Grem JL, van Groeningen CJ, Ismail AA, et al: The role ofinterferon-a as a modulator of fluorouracil and leucovorin. Eur J Cancer 31A(Nos7/8):1316-1320, 1995.

44. Wadler S, Lembersky B, Atkins M, et al: Phase II trial offluorouracil and recombinant interferon alfa-2a in patients with advancedcolorectal carcinoma: An Eastern Cooperative Oncology Group study. J Clin Oncol9:1806-1810, 1991.

45. Thirion P, Cassano A, O’Dwyer P, et al: Meta-analysis ofalpha-interferon plus 5-fluorouracil in advanced colorectal cancer (abstract945). Proc Am Soc Clin Oncol 19: 244 A, 2000.

46. Grem JL, King SA, O’Dwyer PJ, et al:N-(Phosphonoacetyl)-L-Aspartate (PALA): A review of its biochemistry andclinical activity. Cancer Res 48: 4441-4454, 1988.

47. Grem JL: Biochemical modulation of 5-fluorouracil bydipyridamole: Preclinical and clinical experience. Semin Oncol 19 (suppl3):56-65, 1992.

48. Köhne CH, Hiddemann W, Schüller J, et al: Failure oforally administered dipyridamole to enhance the antineoplastic activity of5-fluorouracil in combination with folinic acid in patients with advancedcolorectal cancer. A prospective randomized trial. J Clin Oncol 13:1201-1208,1995.

49. Moran RG, Danenberg PV, Heidelberger C: Therapeutic responseof leukemic mice treated with fluorinated pyrimidines and inhibitors ofdeoxyuridylate synthesis. Biochem Pharmacol 31:2929-2935, 1982.

50. Di Costanzo F, Gasperoni S, Malacarne P, et al: High-dosefolinic acid and 5-fluorouracil alone or combined with hydroxyurea in advancedcolorectal cancer. A randomized trial of the Italian Oncology Group for ClinicalResearch. Am J Clin Oncol 21:369-375, 1998.