Make your own free website on

The Second European Conference:

Perspectives in Colorectal Cancer, held in Barcelona, Spain in June, provided an opportunity for oncologists to share and discuss their latest data, experience, and opinions on treatment of colorectal cancer. On the first day of the conference, a panel of European experts gathered to discuss the optimization of patient care in colorectal cancer and review the progress we are making in improving treatment strategies for this disease. Colorectal cancer causes approximately 60,000 deaths in the United States per year. Nearly 190,000 new cases are diagnosed in Europe annually, with a corresponding figure of 130,000 in the United States. The median survival for patients with metastatic disease is 11 months.

Fluoropyrimidines: A Cornerstone in Colorectal Cancer Treatment

For more than 4 decades, the fluoropyrimidine 5-fluorouracil (5-FU) has been the basis of treatment for colorectal cancer. It has been established that the efficacy of fluoropyrimidine therapy is superior to best supportive care alone, and that the efficacy of 5-FU (at least in terms of response rates) can be improved through protracted administration schedules and biomodulation with leucovorin (LV).[1-6] However, no particular regimen is perceived as the "gold standard," and a wide range of "standard" regimens are used in clinical practice. Furthermore, although protracted or continuous infusion 5-FU is more effective than bolus 5-FU, particularly in terms of response rates, infusional regimens are inconvenient for patients, labor-intensive for medical staff, and are frequently associated with venous access-related complications. Therefore, there is a clear medical need for new, oral agents that avoid these problems and provide convenient, home-based therapy, which most patients prefer.[7,8]

Two oral fluoropyrimidine therapies -- capecitabine and UFT (uracil plus tegafur) plus LV -- are in advanced stages of development and have been investigated extensively in clinical trials. Capecitabine is already an established treatment option for anthracycline- and taxane-pretreated metastatic breast cancer. Professor Eric Van Cutsem of the Department of Internal Medicine, University Hospital Gasthuisberg, Leuven, Belgium, described how capecitabine is converted to 5-FU via a 3-step enzymatic pathway. The final step in the conversion involves the enzyme thymidine phosphorylase (TP), which is significantly more active in tumor tissue compared with healthy tissue. This results in the generation of 5-FU preferentially at the tumor site, thus potentially reducing systemic exposure to 5-FU.[9,10]

The components of UFT/LV therapy have all been used for the treatment of colorectal cancer. Uracil (a dihydropyrimidine dehydrogenase metabolite) and tegafur (a prodrug of 5-FU) are administered as the agent UFT, which is a fixed 4:1 molar ratio of the 2 compounds. UFT has been used for many years in Japan and Spain for the treatment of colorectal cancer.

A Critical Evaluation of Infusional and Oral Fluoropyrimidine Regimens

Selecting the most appropriate first-line treatment for patients with metastatic colorectal cancer is not a straightforward process. Dr. Twelves of the Department of Medical Oncology, Glasgow Royal Infirmary NHS Trust, United Kingdom, highlighted the wide range of fluoropyrimidine regimens used in clinical practice, many of which may need hospital admissions and require central venous access, whereas the newer therapies can be given orally as outpatient therapy. The availability of such a broad range of treatment options underlies the difficult task of assessing the relative strengths and weaknesses of these regimens.

While none of the most active fluoropyrimidine regimens stands out as superior in terms of efficacy, the bolus 5-FU/LV Mayo Clinic regimen is used in a large proportion of centers owing to its confirmed activity and ease of administration compared with infusional regimens. Treatment consists of bolus LV 20 mg/m2 followed by iv bolus 5-FU 425 mg/m2 on days 1 to 5, repeated every 28 days. A number of randomized, phase III trials have been conducted to compare new fluoropyrimidine regimens with the Mayo Clinic regimen as first-line treatment for colorectal cancer. Dr. Twelves described the 4 fluoropyrimidine regimens that he was going to discuss in his presentation -- 2 oral fluoropyrimidines and 2 of the most active infusional 5-FU/LV regimens:

Dr. Twelves went on to provide a critical evaluation of the data available from randomized, phase III trials of these regimens vs the Mayo Clinic regimen. The capecitabine data discussed by Dr. Twelves were the integrated data (prospective analysis) from more than 1,200 patients treated in 2 large trials first reported at ASCO 1999. One trial was conducted in the United States, Mexico, and Brazil[11] and the second was conducted in Europe and Australasia.[12] The UFT/LV data were from study 011 presented at ASCO 1999[13] and updated in September 1999 for regulatory filing to the FDA. The de Gramont data were published in 1997,[14] and the German AIO regimen data were from the EORTC/AIO study presented in May 2000 at the ASCO annual meeting in New Orleans.[15] The study by Schmoll and coworkers was a 3-arm trial with the infusional 5-FU regimen administered either with or without LV.

Before discussing the efficacy and safety data for the 4 regimens in detail, Dr. Twelves drew attention to some of the differences between the trials. He noted that in the de Gramont trial, which was not conducted for regulatory purposes, the analysis population was substantially smaller than the intent-to-treat population (448 patients randomized, 433 patients in the survival and time to disease progression (TTP) analyses, 348 patients in the response-rate analysis). He urged delegates to bear this in mind when comparing the trials. In the EORTC/AIO trial, only patients with measurable disease were included in the analysis of response rate. A notable difference in study design between the 4 studies was the longer interassessment intervals in the UFT/LV and de Gramont trials (8-10 and 12 weeks, respectively, compared with 6 and 4 weeks in the capecitabine and EORTC/AIO trials, respectively). The longer period between assessments results in cruder estimation of endpoints such as TTP. Another particularly important difference in the study designs in the context of this comparison, which was stressed later by Professor Jim Cassidy, is the different primary endpoints used in the 4 trials. For the UFT/LV, de Gramont, and EORTC/AIO trials, the primary endpoint was survival, whereas for the capecitabine analysis, the primary objective was to demonstrate at least equivalent efficacy in terms of response rate, with survival and TTP identified as secondary endpoints.

Impact on Survival

Dr. Twelves then compared the efficacy data from the 4 trials. In terms of overall survival, none of the regimens revealed a significant advantage over the Mayo Clinic regimen. Therefore, we must evaluate other endpoints such as response rate, TTP, tolerability, convenience, and medical resource utilization to assess the value of these regimens relative to one another.

Efficacy: Significant Differences in Response Rates and TTP

The primary endpoint of the capecitabine trial was response rate. Oral capecitabine achieved significantly higher response rates than the Mayo Clinic regimen (26% vs 17%, respectively; P < .0002). In addition, TTP was equivalent, with median TTP reported as 4.6 and 4.7 months, respectively (Table 1). In the UFT/LV trial, there was no statistically significant difference in response rates (12% with UFT/LV vs 15% with Mayo Clinic regimen), but the median TTP was significantly inferior with UFT/LV therapy compared with Mayo Clinic regimen (median 3.5 vs 3.8 months, respectively; P = .01).

Table 1. Median TTP (Study Drug vs the Mayo Clinic Regimen)

Median TTP (months)

Relation to Mayo Clinic Regimen

P value









de Gramont




(+ LV)
(- LV)




Both the de Gramont and EORTC/AIO infusional 5-FU/LV regimens offered modest but significant increases in TTP. Median TTP was 6.3 months with the de Gramont regimen compared with 5.0 months in the Mayo Clinic regimen arm (P = .001), and 6.3 vs 4.1 months with the EORTC/AIO (plus LV) regimen and Mayo Clinic regimen, respectively (P = .02). The de Gramont regimen also demonstrated a significant improvement in response rates compared with the Mayo Clinic regimen (33% vs 15%, P = .004), although Dr. Twelves reminded the audience that the response rate analysis was based on only three quarters of the randomized population. There was a large difference in response rates between the EORTC/AIO (20.5%) and the Mayo Clinic regimen (11.5%), but this difference did not reach statistical significance, possibly because of the relatively small number of patients with measurable disease.

Dr. Twelves also noted that in all 4 trials, the response rates achieved with Mayo Clinic regimen were consistently in the order of 12% to 15%, suggesting that the patient populations evaluated in the 4 trials were similar (Table 2).

Table 2. Overall Response Rate

Study Treatment

Mayo Clinic Regimen

P value




< .0002





de Gramont




(+ LV)
(- LV)







NS = not significant

Therefore, in terms of efficacy, concluded Dr. Twelves, none of the regimens provided a significant survival advantage over the Mayo Clinic regimen. However, capecitabine and the de Gramont regimen produced significant improvements in response rates, the infusional regimens provided significant improvements in TTP, and UFT/LV resulted in significantly inferior TTP.

Improved Tolerability

Dr. Twelves then focused on the differing safety profiles of the 4 regimens, and in particular the incidence of diarrhea, nausea/vomiting, stomatitis/mucositis, and neutropenia. All 4 regimens were associated with a significantly lower incidence of grade 3/4 neutropenia and less grade 3/4 stomatitis compared with the Mayo Clinic regimen. Both of the oral regimens resulted in significantly reduced incidences of diarrhea and nausea/vomiting. When Dr. Twelves showed data for grade 3/4 events, different trends in the safety profiles of the 4 regimens could be seen. For example, there was a trend toward a higher incidence of grade 3/4 nausea/vomiting in all regimens except for capecitabine. Grade 3/4 diarrhea was significantly less frequent with the de Gramont regimen compared with the Mayo Clinic regimen, whereas severe diarrhea may be problematic in patients treated with UFT/LV. A 26% incidence of grade 3/4 diarrhea was reported in patients receiving adjuvant UFT/LV in a phase III trial.[16] All 4 regimens showed a clear advantage over the Mayo Clinic regimen in terms of grade 3/4 neutropenia.

One of the more frequent toxic effects associated with capecitabine therapy is cutaneous hand-foot syndrome, which affects palms and soles. However, this adverse event is usually mild to moderate in intensity, emphasized Dr. Twelves, and resulted in the hospitalization of only 2 of more than 600 patients treated with capecitabine in the 2 phase III trials. If patients and oncologists learn to recognize the early signs of mild or moderate hand-foot syndrome, development of severe hand-foot syndrome can usually be avoided by interrupting therapy, he added.

Improving Patient Convenience

Clearly, efficacy and safety are important considerations when identifying the most suitable treatment option for patients. Another important factor is patient convenience and the impact of treatment on quality of life. Infusional regimens require regular inpatient treatment or the use of Hickman lines or port-a-cath systems, which are cumbersome for patients and restrict daily activities. Moreover, infusional therapy is frequently associated with venous access-related complications such as infection or blockage. In addition, most patients require regular hospital visits that disrupt their daily life. For example, during a typical 24-week treatment period, a patient would need only 5-8 hospital visits for oral fluoropyrimidine therapy compared with 24-36 visits with the infusional regimens such as de Gramont or EORTC/AIO. This is a particularly important issue when survival is limited and when patients are having palliative treatment, added Dr. Twelves.

Patient preference must also be taken into consideration when choosing the ideal therapy: a questionnaire-based survey of 103 patients revealed that most patients prefer oral chemotherapy, provided this is not at the expense of efficacy. These results were supported by data from a recently reported, randomized, crossover study, in which patients showed a strong preference for oral vs iv therapy (84% in favor of oral therapy).[17] The principal reasons stated for this preference were oral administration, home-based treatment, and ability to continue daily activities.

Combination Regimens

Developments in colorectal cancer treatment are not restricted to oral fluoropyrimidines; irinotecan and oxaliplatin have both demonstrated efficacy and tolerability in this setting and have had a major impact on the treatment of colorectal cancer. Based on the results of phase III clinical trials, irinotecan has been approved in Europe and the United States in combination with 5-FU/LV as first-line therapy and as monotherapy for second-line treatment for colorectal cancer. Oxaliplatin in combination with 5-FU/LV has been approved in Europe as first- and second-line treatment for colorectal cancer. More recently, a number of trials have been conducted to explore irinotecan and oxaliplatin in combination with new, oral fluoropyrimidines in an attempt to further improve efficacy.

During the conference, there were a number of presentations on new, phase I combination data, which are summarized below.

Irinotecan-Containing Combination Regimens

Irinotecan, a semisynthetic camptothecin derivative that inhibits topoisomerase I, is established as second-line therapy for colorectal cancer; and the combination of irinotecan, 5-FU, and LV is registered as first-line therapy. A randomized, phase III clinical trial has shown that the addition of irinotecan to infusional or bolus 5-FU/LV regimens results in superior efficacy in terms of response rate (35% vs 22%), TTP (median 6.7 vs 4.4 months), and survival (median 17.4 vs 14.1 months).[18] A 3-arm, randomized, phase III trial conducted in the United States also demonstrated that the addition of irinotecan to a bolus 5-FU/LV regimen resulted in superior response rates, time to disease progression, and survival.[19] An integrated analysis of both trials was presented at the ASCO annual meeting in May 2000 and supported these results.[20]

Irinotecan plus capecitabine. Not surprisingly, oncologists are keen to determine whether similar results can be achieved when irinotecan is administered in combination with oral fluoropyrimidines. New data from a phase 1/2 trial was initiated to identify the recommended dose for further clinical development and to assess the safety profile of the combination.

Capecitabine was administered on days 1-14 and 22-36 at doses of 1000 or 1250 mg/m2 twice daily, and irinotecan was administered as a weekly 30-minute iv infusion for 6 weeks at doses of 70 and 80 mg/m2. This cycle of treatment was repeated every 50 days. All patients had metastatic colorectal cancer, measurable disease, and had received no prior treatment for metastatic disease. The dose levels and patient recruitment to each level are shown in Table 3.

Table 3. Dose-Escalation Schedule for Capecitabine/Irinotecan Phase I Trial

Dose Level

Capecitabine (days 1-14 and 22-36, mg/m2 oral, twice daily)

Irinotecan (mg/m2 iv weekly x 6)

No. of patients

No. of cycles





















Dr. Vanhoefer presented data from 17 patients treated for 39 cycles, which show that the combination is feasible. Among these patients, there were only 2 cases of grade 3 or 4 neutropenia (grade 3 at the second-dose level, grade 4 at the third-dose level). Both of these events occurred in the first treatment cycle, and no further grade 3 or 4 neutropenia was observed in subsequent cycles. The principal grade 3 or 4 nonhematologic adverse event was diarrhea, which occurred in 3 patients (2 grade 3 at the third-dose level and 1 grade 4 at the second-dose level). Only 1 patient experienced hand-foot syndrome (grade 2 at the third-dose level). The maximum tolerated dose was defined as dose level 3. Preliminary efficacy data from the study are encouraging, with a high proportion of complete and partial responses. Data for 6 patients are not yet mature and therapy is ongoing.

Dr. Vanhoefer concluded that capecitabine plus irinotecan combination therapy is a feasible new treatment option. The recommended dose level for further evaluation is capecitabine 1250 mg/m2 twice daily, administered on days 1-14 and 22-36 in combination with weekly irinotecan 70 mg/m2 administered for 6 weeks, repeated every 50 days. This regimen is currently being evaluated in an extended phase I/II study, and further investigation in phase III trials is planned.

Irinotecan plus UFT/LV. Irinotecan has also been investigated in combination with UFT/LV as first-line treatment for advanced or metastatic colorectal cancer. Results of a phase I, dose-escalation trial involving 20 patients were summarized in a talk on the second day of the conference. A total of 97 treatment cycles have been administered, and the dose-limiting toxicities were diarrhea and febrile neutropenia (2 patients each).[22] The most common grade 3/4 toxicities were nausea (16%), fatigue/lethargy (19%), diarrhea (16%), and neutropenia (9%).

The recommended dose has been identified as UFT 250 mg/m2/day plus LV 90 mg/day (both administered as 3 divided doses, days 1-14) in combination with irinotecan 250 mg/m2 on day 1, every 3 weeks. Preliminary efficacy results from 14 patients show a response rate of 29% (1 complete response, 3 partial responses). Four additional patients were not evaluable for efficacy because of early toxicity. Following identification of the recommended dose for this combination, a phase II extension of the study is being conducted to further assess the efficacy and safety of this regimen. So far, 20 patients have been treated at the recommended dose.

Irinotecan plus raltitrexed. Another agent investigated in combination with a number of therapies and discussed by Dr. Twelves is raltitrexed, a thymidylate synthase inhibitor. However, results reported to date for this combination in a range of settings are disappointing. In a phase I combination study in heavily pretreated patients with colorectal cancer, a regimen of irinotecan 180 mg/m2 on day 1 followed by raltitrexed 2 mg/m2 on day 2, repeated every 2 weeks, demonstrated no antitumor activity.[23] There were no objective responses and the median TTP was just 13 weeks. The investigators suggested that further evaluation in less heavily pretreated patients was warranted.

Two additional trials of the combination as first-line treatment demonstrated promising antitumor activity (53% and 30%, respectively), but significant toxicity was observed in both trials.[24,25] Toxic deaths occurred in 5% of patients treated in the first study, with a further 38% of patients requiring dose reduction, and more than three quarters of patients in the second trial experienced grade 3/4 toxicity. Further evaluation using lower doses of these 2 drugs is planned.

Oxaliplatin as a Combination Partner

A second, new agent that featured heavily in the conference as a combination partner was oxaliplatin, a platinum analog. Oxaliplatin in combination with 5-FU/LV has demonstrated high activity as treatment for colorectal cancer, and the combination is registered in Europe for the first- and second-line treatment of colorectal cancer based on phase III clinical trial results.

Oxaliplatin plus capecitabine. Preclinical studies have shown that the combination of fluoropyrimidines and oxaliplatin can partially overcome clinical resistance to fluoropyrimidines. Therefore, a phase I, dose-escalation trial was conducted in 4 centers in the UK and Spain to investigate the combination of capecitabine and oxaliplatin.[26] The study included 23 patients with advanced/metastatic solid tumors resistant to standard chemotherapeutic regimens, and the latest results were presented by Professor Díaz-Rubio in Barcelona.

Oxaliplatin was administered at a fixed dose of 130 mg/m2 on day 1 of each 21-day cycle and capecitabine was administered at a starting dose of 500 mg/m2 twice daily, days 1-14. The dose-limiting toxicity was diarrhea (grade 3 in 2 patients [1 patient each at capecitabine 1000 and 1250 mg/m2 twice daily] and grade 4 in 1 patient [capecitabine 1250 mg/m2 twice daily]). The investigators identified a regimen of oxaliplatin 130 mg/m2 combined with capecitabine 1000 mg/m2 twice daily, days 1-14, administered as a 3-weekly cycle, as the most appropriate schedule for further development. The combination was well tolerated at this dose, and most toxic effects were of mild to moderate intensity.

Professor Díaz-Rubio highlighted the efficacy results obtained in a subgroup of 9 patients with pretreated colorectal cancer (including prior irinotecan in 4 patients). Five of these patients achieved partial responses and, in 3 of them, disease was stabilized. These promising results have led to the initiation of a phase II trial to further evaluate capecitabine/oxaliplatin combination therapy.

Oxaliplatin plus UFT/LV. Oxaliplatin has also been investigated in combination with UFT/LV. The study was conducted in Spain and used a slightly different schedule from the capecitabine trial. The treatment regimen consisted of oxaliplatin 100 mg/m2 on days 1 and 14, oral UFT 390 mg/m2 for 14 days, and LV 250 mg/m2 as a 2-hour infusion on day 1 followed by oral LV 7.5 mg every 12 hours for 14 days.[27] Treatment was repeated every 28 days. Grade 3/4 toxicities were quite common in the first 2 cycles (diarrhea 54%, nausea/vomiting 23%), and consequently the UFT dose was reduced from 390 mg/m2 to 300 mg/m2. At the lower dose, grade 3/4 diarrhea occurred in 18% of patients and grade 3/4 nausea/vomiting occurred in 9% of patients.

Oxaliplatin plus raltitrexed. Raltitrexed has shown more promise as a combination partner for oxaliplatin than irinotecan, and the most recent data were reviewed by Dr. Twelves. The efficacy of raltitrexed and oxaliplatin, administered at their full doses, has been evaluated in combination trials, with encouraging activity noted in patients with advanced colorectal cancer. In a phase II trial involving 71 patients with untreated metastatic colorectal cancer, 29 patients achieved objective tumor responses (1 complete response, 38 partial responses).[28] The median TTP was 6.3 months and median survival has not yet been reached.

Can Oral Fluoropyrimidines Replace 5-FU in Combination Regimens?

It is sometimes difficult to compare these phase I trials because of differences in study design and patient populations, and caution must be exercised when interpreting efficacy data because of the small sample sizes. Nevertheless, it is reasonable to say that combination regimens including oral fluoropyrimidines appear to be promising strategies for improving treatment for colorectal cancer patients.

It is still too early to select the optimal combination, commented Dr. Twelves, but the numerous phase I/II studies that have been reported recently show that several combination regimens are feasible and active. If 5-FU/irinotecan combination therapy becomes the gold standard for first-line treatment of colorectal cancer, it will be interesting to see if we can improve therapy in terms of efficacy, safety, and convenience by replacing 5-FU with an effective oral fluoropyrimidine.

Chemoradiation for Rectal Cancer

Combination trials using oral fluoropyrimidines have not been limited to chemotherapeutic agents, and the use of oral fluoropyrimidines as radiosensitizers has been explored in a number of early clinical trials. Chemoradiation using 5-FU as a radiosensitizer is widely accepted as standard therapy for rectal cancer in the adjuvant and neo-adjuvant setting. Protracted infusion of 5-FU has demonstrated superior efficacy to bolus 5-FU when used in combination with radiation therapy,[29] but is inconvenient for patients and labor intensive for medical staff.

Capecitabine as Radiosensitizer

Dr. Judith Tanner opened her presentation by explaining that there is a clear rationale for the use of capecitabine as a radiosensitizer, not only because it mimics continuous infusion 5-FU, but also because of the upregulation of TP in tumor tissue following radiotherapy. The tumor-selective activation of capecitabine is mediated by TP, and preclinical studies demonstrated that the antitumor activity of capecitabine plus radiotherapy combination treatment was more than additive, whereas irradiation combined with 5-FU showed no clear additive effects.[30]

Therefore, a phase I trial was initiated in Germany to determine the maximum tolerated dose of capecitabine plus radiotherapy as combination treatment for rectal cancer. Patients received a total irradiation dose of 50.4 Gy given over a period of approximately 6 weeks as 1.8 Gy daily fractions. Capecitabine was administered twice daily at doses of 250, 375, 500, 650, 825, and 1000 mg/m2 with the first dose administered 2 hours prior to radiotherapy. Dr. Tanner provided an update of the results presented at the ASCO annual meeting in New Orleans. A total of 30 patients have now been treated, although data for the 6 patients treated with capecitabine 1000 mg/m2 are not yet mature. Nineteen patients have been treated in the adjuvant setting, 9 in the neo-adjuvant setting, and 2 in the palliative setting.

Capecitabine/radiotherapy combination treatment has been well tolerated in this ongoing trial. There have been no grade 4 toxic effects and only 2 grade 3 toxic events (rash/itch at the 375 mg/m2 dose level, local skin toxicity at the 650 mg/m2 dose level). There was no relationship between the incidence or severity of toxicities and dose level. In addition, capecitabine/radiotherapy combination treatment has demonstrated antitumor activity: among 9 patients treated in the neo-adjuvant setting, tumors were downstaged in 8 patients, with 2 pathologically confirmed complete responses and 3 partial responses.

In addition to the favorable safety profile and promising efficacy demonstrated in this phase I trial, Dr. Tanner noted that it was particularly gratifying to treat her patients with a convenient, oral radiosensitizer instead of protracted infusion 5-FU. Capecitabine simplified chemoradiotherapy and was appealing to patients. Capecitabine therefore has the potential to replace 5-FU as the standard radiosensitizer for rectal cancer patients receiving chemoradiotherapy. When the recommended dose schedule has been identified, a phase II trial will be conducted in neo-adjuvant patients to evaluate this combination further.

UFT/LV as Radiosensitizer

UFT/LV has also been evaluated as a chemoradiosensitizer in rectal cancer patients, and data from a phase I/II trial were presented by Dr A. Jakobsen. Eighteen patients were enrolled in treatment with a total irradiation dose of 60 Gy administered in 30 fractions in combination with UFT 150-300 mg/m2/day plus a fixed dose of LV 22.5 mg/day.[31] All patients received the planned irradiation dose and all but 1 received the planned dose of UFT.

Grade 1 and 2 hematologic toxicities were relatively common, but only 1 patient experienced grade 3 diarrhea. Two patients achieved a partial response, and 9 of 12 patients whose tumors were operated had R0 and 3 had R1. Evaluation of preoperative UFT/LV/radiation combination therapy will continue in a phase II study.

Next Steps in Colorectal Cancer

The data presented at the meeting and reviewed in this activity indicate that we may be entering a new era in the therapy of colorectal cancer. 5-FU has been the mainstay of therapy of this disease for over 40 years. It seems likely that fluoropyrimidines will still be an integral part of therapy -- but 5-FU will be replaced by more convenient, less toxic, and possibly more active oral fluoropyrimidines. Capecitabine has the added advantage over UFT that it produces tumor-specific activation of 5-FU in the cancer site. It is for these reasons that, on current evidence, I would favor this agent as the optimal way of giving fluoropyrimidine in the future. The next major steps forward in this disease are, however, likely to be achieved by novel combinations of the active drugs that have been reviewed so far.

Phase III trials of irinotecan in combination with 5-FU have demonstrated that efficacy, including survival, can be enhanced by combining irinotecan with fluoropyrimidines. Similarly, the combination of oxaliplatin with 5-FU/LV has also been shown to produce dramatic improvements in response rate. A logical step in our attempts to further improve efficacy is to use highly active fluoropyrimidine agents such as capecitabine, which achieves superior response rates compared with a standard bolus 5-FU/LV schedule (Mayo Clinic regimen). A number of phase I trials exploring oral fluoropyrimidines in combination with irinotecan or oxaliplatin have been conducted and recommended doses for more extensive clinical development have been established. The next few years should see these combinations enter large-scale comparative studies.

In addition, as radiotherapy has been shown to upregulate TP activity, the combination of capecitabine with radiotherapy is being explored and the updated results presented by Dr. Tanner are encouraging. The results of phase II trials for all of these combination regimens are eagerly awaited.

It is also worth noting that these new agents are not necessarily restricted to colorectal cancer. The promising efficacy and safety shown by capecitabine has also been demonstrated in small scale pilot trials that have been performed in pancreatic, hepatocellular, gall bladder cancer patients, and a host of other solid tumor types. Recently presented phase II results of capecitabine monotherapy and phase I results of capecitabine/gemcitabine combination therapy in metastatic pancreatic cancer are encouraging, and phase II and III trials of capecitabine/gemcitabine combination therapy are planned following demonstration of a favorable safety profile and high antitumor activity.

In conclusion, the future of therapy in colorectal cancer is likely to involve the use of combinations and/or sequences of all of the new agents discussed above. The adjuvant use of such active combinations is being actively pursued and it seems likely that we will be able to cure a higher proportion of our patients in the near future by the judicious application of these new therapies.