Addition of either irinotecan or methotrexate to bolus 5-fluorouracil and high-dose folinic acid every 2 weeks in advanced colorectal carcinoma: a randomised study by the Southern Italy Cooperative Oncology Group

P. Comella+, E. Crucitta, F. De Vita, L. De Lucia, A. Farris, F. Del Gaizo, S. Palmeri, A. Iannelli, S. Mancarella, S. Tafuto, L. Maiorino, F. Buzzi and G. De Cataldis§

Division of Medical Oncology A, National Tumour Institute, Naples, Italy

Received 15 October 2001; revised 16 November 2001; accepted 11 December 2001.


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Purpose

The purpose of this study was to compare the activity and toxicity of the combination of irinotecan (IRI) plus folinic acid (FA)-modulated 5-fluorouracil (5-FU) i.v. bolus with a regimen of double modulation of 5-FU with methotrexate (MTX) and FA in patients with advanced colorectal carcinoma.

Patients and methods

Two-hundred and thirty-four patients were enrolled: 118 patients received IRI 200 mg/m2 (90-min i.v. infusion) on day 1, followed by levo-FA 250 mg/m2 (2-h i.v. infusion) and 5-FU 850 mg/m2 (i.v. bolus) on day 2 (IRIFAFU), and 116 patients received MTX 750 mg/m2 (2-h i.v. infusion) on day 1, followed by levo-FA 250 mg/m2 (2-h i.v. infusion) and FU 800 mg/m2 (i.v. bolus) on day 2 (MTXFAFU). Both cycles were repeated every 2 weeks until progression or to a maximum of 16 cycles. Response rate (RR) was the main end point of the study; responses were assessed every four cycles and confirmed after 2 additional months of treatment.

Results

RR was significantly greater with IRIFAFU (36%) than with MTXFAFU (20%) (P <0.001). Multivariate analysis showed that IRIFAFU was significantly associated with a greater activity (P = 0.028). Median progression-free survival was longer with IRIFAFU than with MTXFAFU (7.2 months compared with 4.8 months; P = 0.048). Median survival time (MST) did not differ between the two arms (14.7 months compared with 14.8 months, respectively). Patients not receiving second-line chemotherapy, however, lived longer when treated in the first-line with IRIFAFU (MST 11.9 months compared with 6.4 months; P = 0.038). IRIFAFU caused a significantly greater occurrence of grade 3 or 4 neutropenia (40% compared with 9%; P = 0.001) and diarrhoea (13% compared with 4%; P = 0.024), but a significantly lower incidence of stomatitis (3% compared with 12%; P = 0.007), than the comparative regimen.

Conclusions

IRIFAFU appeared comparable in terms of activity and toxicity with other weekly or biweekly bolus or infusional combination regimens. IRIFAFU, however, seems easier to administer, because it does not require infusional catheter or pump devices, and it is less expensive. It may represent a new option for treating advanced colorectal carcinoma.

Key words: advanced colorectal carcinoma, combination regimen, irinotecan plus modulated 5-fluorouracil, randomised trial


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Over the past 20 years, several efforts have been made in the attempt to improve the activity of 5-fluorouracil (5-FU) by means of its biochemical modulation [1]. A meta-analysis of nine randomised phase III studies, however, comparing 5-FU plus folinic acid (FA) with 5-FU alone in advanced colorectal carcinoma, although demonstrating a significant advantage in terms of response rate (RR), was unable to show a survival improvement [2]. Methotrexate (MTX) has also been used to potentiate the activity of 5-FU. A meta-analysis of eight randomised trials comparing MTX and 5-FU with 5-FU alone not only showed an advantage for the combination in terms of RR, but also demonstrated a small but statistically significant prolongation of survival [3]. Furthermore, the double biochemical modulation of 5-FU by MTX and FA in metastatic colorectal cancer has been addressed by several authors. In our experience, a regimen including MTX followed 24 h later by levo-FA (LFA) and 5-FU every 2 weeks obtained a 31% RR and a 62-week median survival [4]. In addition, four randomised studies reported a higher RR with the double in comparison with the single 5-FU modulation [58].

Irinotecan (IRI or CPT-11) has been initially tested in 5-FU-refractory patients, obtaining a moderate activity and a survival benefit, suggesting a lack of cross-resistance between the two drugs [9, 10]. Furthermore, CPT-11 with 5-FU has been compared with 5-FU alone as first-line management of advanced colorectal cancer patients in two randomised trials [11, 12]. In one of these studies, CPT-11 125 mg/m2, FA 20 mg/m2 and 5-FU 500 mg/m2 were given together i.v. weekly for 4 of every 6 weeks [11]. In another study, patients were randomised to receive either one of two 5-FU infusional regimens (24-h weekly or 48-h every other week), or the same regimen with the addition of CPT-11 (80 mg/m2 or 180 mg/m2, respectively) [12]. In both trials, a statistically significant survival advantage was obtained with the combination regimens. We have recently explored a new scheduling of CPT-11 and FA-modulated 5-FU, administering the two cytotoxic drugs in close sequence on days 1 and 2 (CPT-11 200 mg/m2 as short i.v. infusion and 5-FU 850 mg/m2 as i.v. bolus, respectively) in a biweekly regimen, obtaining a very encouraging RR [13].

Therefore, we planned to assess the activity and toxicity of this new regimen in a randomised study, keeping our double modulation of 5-FU as reference treatment. The interim analysis of this trial has been already published [14]. Here we report the final analysis of the this trial.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Eligibility as well as exclusion criteria have been previously reported [14]. Eligible patients had to have at least one measurable indicator lesion. The target lesions were measured by computed tomography (CT) scan or magnetic resonance imaging (MRI), with ultrasonography if necessary. Patients gave informed consent to participate in this trial, which was approved by the Independent Ethical Committee for Clinical Research of the National Tumour Institute of Naples.

The primary end points of this phase II/III randomised trial were: (i) to evaluate the RR and the acute toxicity profile of two new regimens in which 5-FU and LFA were combined with either CPT-11 (IRIFAFU) or Tomudex® (TOM; TOMFAFU) in patients with advanced colorectal carcinoma, and (ii) to estimate the difference in RR between each of these two experimental regimens and the reference regimen of MTX followed by LFA + 5-FU (MTXFAFU) every 2 weeks. Secondary end points were to compare the time to progression (TTP) and the overall survival of treated patients.

Patients who meet the inclusion criteria, after stratification for site of primary cancer (colon compared with rectum) and performance status (PS) [0 compared with 1 or 2 of the Eastern Cooperative Oncology Group (ECOG) scale], were randomly allocated to receive one of the three following treatments: CPT-11 200 mg/m2 as 90-min i.v. infusion on day 1, LFA 250 mg/m2 as 2-h i.v. infusion and 5-FU 850 mg/m2 i.v. bolus on day 2 every 2 weeks (IRIFAFU); TOM 3.0 mg/m2 as 15-min i.v. infusion on day 1, LFA 250 mg/m2 as 2-h i.v. infusion and 5-FU 1050 mg/m2 i.v. bolus on day 2 every 2 weeks (TOMFAFU); or MTX 750 mg/m2 as 2-h i.v. infusion on day 1, LFA 250 mg/m2 as 2-h i.v. infusion and 5-FU 800 mg/m2 i.v. bolus on day 2 every 2 weeks (MTXFAFU).

Treatment was administered in each arm of the trial every 14 days until a complete response (CR) was achieved (after which four additional cycles were administered), or up to a maximum of 6 months of treatment in the case of partial response (PR) or stable disease. Therapy was discontinued earlier in the presence of documented progression of disease, unacceptable toxicity, patients refusal, or when the attending physician judged that it was in the patients best interest. After the discontinuation of treatment, patients were controlled every 2 months to assess relapse of disease (for responder patients) or progression of disease (for patients with stable disease).

In each arm of the trial, doses were adjusted according to blood cell count performed at day 8: if a grade 4 neutropenia was detected at that time, or a febrile neutropenia occurred before recycling, the dose of cytotoxic drugs was subsequently reduced by 25%. If a grade 4 neutropenia occurred again despite this reduction, a further reduction by 25% was adopted. In case of neutropenia [absolute neutrophil count (ANC) <1500/mm3] on day 15, recycling was delayed until recovery. If a 2-week delay was required to reach an ANC value >=1500/mm3, a 25% dose reduction was then adopted. If a >2-week delay was required for bone marrow recovery, the patient was discharged from the study.

Dose modifications were also planned for severe non-haematological toxicities: in case of severe diarrhoea or other toxicities grade 3 or more (except alopecia and anaemia), therapy was delayed for 1–2 weeks until complete recovery or grade 1 or less, then re-instituted with a 25% reduction of cytotoxic drugs.

During treatment, blood cell count was performed weekly (or more often in the presence of neutropenia), while biochemistry was repeated at each cycle. Physical status and evaluation of toxicity was assessed at each cycle. Toxicity was scored according to WHO criteria [15], and the worst toxicity suffered by each patient during the whole treatment was recorded. The cholinergic syndrome was arbitrarily scored as previously reported. Measurements of the disease with CT or MRI scan was repeated after every four cycles and at the end of treatment. Response to treatment was defined according to WHO criteria [16]. Responses were radiologically reassessed 8 weeks after their first documentation, and only confirmed responses were computed in each arm of the trial.

Duration of response was calculated from the time of the start of treatment, or from the date it was first documented (in the case of CR) to the date of documented tumour progression. TTP was calculated from the date of registration to the date of documented tumour progression, or death. Patients who discontinued the treatment early because of toxicity, refusal or reasons other than progression were considered as censored at that time interval. Overall survival time was calculated for all patients from the date of registration to the date of death for any cause, or to patients’ last follow-up.

To define the sample size, the two-stage design for phase III randomised trials [17] has been utilised: assuming a 15% difference in RR between the control and each experimental regimen (20% compared with 35%), the comparison on the first 48 patients per arm has an 80% power to reveal at least a 4% difference in response. If this is not the case, the study could be terminated early. This was the case for the TOMFAFU arm, which was close to accrual after interim analysis. On the contrary, accrual continued in the IRIFAFU and MTXFAFU arms to reach a total of 116 patients per arm: this final sample size gave also a sufficient power for a comparison of TTP. Indeed, anticipating a median TTP of 4 months for the MTXFAFU arm, this sample size had an 80% power to detect a 50% increase in TTP (i.e. 6 months) for IRIFAFU patients, with a type I error of 0.05 (two-sided test).

All curves were estimated with the Kaplan–Meier method [18], and compared with either the Wilcoxon test (which gives more weight to the earlier events) or the log-rank test [19]. All comparison were made with a two-sided test.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Patient population
For this final analysis, a total of 238 patients were registered up to December 2000. Four patients, however, did not met the eligibility criteria, leaving 234 eligible patients: 118 patients were allocated to the IRIFAFU and 116 to the MTXFAFU arm. Main patient characteristics are listed in Table 1. In summary, 65 patients (28%) had received previous adjuvant chemotherapy, 61 patients (26%) had suffered a recent loss of body weight, and slightly more than half of the patients had only one site of disease, the remaining having two (33%) or three sites (11%). All baseline characteristics were evenly distributed between the two arms of the trial, with only a small (not significant) imbalance in previous weight loss, presence of local disease or disease-related symptoms.


View this table:
[in this window]
[in a new window]
 
Table 1. Characteristics of eligible patients in the two arms of the trial
 
Patients in the IRIFAFU arm received a total of 980 courses, with a median number of eight cycles (range 1–20) per patient. In the MTXFAFU arm, a total of 843 cycles were delivered, with a median number of seven (range 1–16) cycles per patient. A similar proportion of patients in both arms were treated with at least 12 cycles (31% compared with 28%, respectively) or 16 cycles (8% compared with 3%, respectively). At the date of this analysis, all but three patients (one in the IRIFAFU and two in the MTXFAFU arm) had discontinued their treatment.

Off-study management was not planned in the protocol, and it was left to the individual choice of attending physicians. About one-third of patients received further treatment in both arms: second-line chemotherapy (40% compared with 36%), surgery (4% compared with 3%) or radiotherapy (2% in both arms), and a few patients (five in all) received a third-line chemotherapy. In the IRIFAFU arm, most second-line treatments were oxaliplatin-based (38 patients), while in the MTXFAFU arm 20 patients were treated with CPT-11, usually combined with other cytotoxic drugs, and six patients received oxaliplatin-based chemotherapy, alone or in combination; nine patients underwent a combination of CPT-11 and oxaliplatin.

Activity analysis
In the IRIFAFU arm, nine CRs and 33 PRs were registered, giving an overall RR of 36% [95% confidence interval (CI) = 28% to 44%]. Twelve patients obtained a minor response, and 16 patients showed a stabilisation lasting at least 6 months, giving an overall tumour growth control in 70 of 118 patients (59%) (Table 2). All major responses were registered after a median of 3.3 months of treatment (range 2.0–7.3 months), and lasted a median of 10.2 months (range 4.0–25.2). In the MTXFAFU arm, four patients achieved a CR, and 19 patients were considered in PR, for an overall RR of 20% (95% CI = 13% to 27%), while an overall control of tumour growth was achieved in 55 patients (47%). Responses were registered after a median of 3.5 months (range 1.5–7.3) of treatment, and lasted a median of 9.3 months (range 2.7–24.8).


View this table:
[in this window]
[in a new window]
 
Table 2. Activity reported in the two arms of the trial
 
Univariate analysis of baseline characteristics, regardless of treatment employed, showed that RR was adversely affected by previous loss of body weight (P = 0.0041), ECOG PS >=1 (P = 0.02), and the presence of more than one site of disease (P = 0.05), while age >=63 years (P = 0.06), carcinoembryonic antigen value >=26 ng/ml (P = 0.08), presence of symptoms (P = 0.08) and a >=25% liver involvement (P = 0.10) resulted in borderline significance. No difference in RR was observed according to previous adjuvant chemotherapy, or between the presence or absence of synchronous metastases. The RR produced by the IRIFAFU regimen (36%) was significantly greater than that obtained with MTXFAFU (20%) (P = 0.011). The overall growth inhibitory activity also favoured the IRIFAFU arm (59% compared with 47%; P = 0.09). The multivariate analysis of factors affecting RR revealed that the type of initial treatment (P = 0.028), the recent loss of body weight (P = 0.004), the number of disease sites (P = 0.030), and the age of patient (P = 0.033) retained a significant effect, while the other covariates lost their significance.

Safety analysis
Comparison of toxicity (Table 3) showed that neutropenia of any type, and severe neutropenia, significantly affected more patients treated with IRIFAFU (P = 0.001); no difference, however, in occurrence of complicated neutropenia was observed. Any grade of thrombocytopenia was more frequent with MTXFAFU (P = 0.005). As for non-haematological toxicity, the proportion of patients suffering severe diarrhoea was three-fold greater in the IRIFAFU than in the MTXFAFU arm (P = 0.024), but occurrence of stomatitis of any grade in the MTXFAFU arm was twice that in the IRIFAFU arm (P = 0.002); moreover, the proportion of patients suffering from severe stomatitis was four-fold greater in the MTXFAFU arm (P = 0.007). Derangement of liver enzymes occurred in twice as many patients in the MTXFAFU arm (P = 0.05).


View this table:
[in this window]
[in a new window]
 
Table 3. Acute side-effect (WHO toxicity grade) according to arm of treatment (per cent of patients)
 
Follow-up analysis
At the time of this report (31 October 2001), the median observation time for the whole series was 23 months, ranging between 11 and 39 months. At this time, 173 had showed a progression of disease, 90 in the IRIFAFU arm and 83 in the MTXFAFU arm. Figure 1 shows the estimated probabilities of treatment failure according to arm of treatment: median TTP was 7.2 months for IRIFAFU and 4.8 months for MTXFAFU (Wilcoxon test, P = 0.048; log-rank test, P = 0.17). At 6-month interval from initial therapy, probability of progression-free survival was 55% compared with 44% in the two arms, respectively (Fisher’s test, P = 0.039).



View larger version (19K):
[in this window]
[in a new window]
 
Figure 1. Actuarial progression-free survival curves according to arm of treatment. Bold line, IRIFAFU (5-fluorouracil and folinic acid combined with irinotecan); fine line, MTXFAFU (methotrexate followed by folinic acid + 5-fluorouracil).

 
One hundred and forty-one patients (60%) have died: 72 patients in the IRIFAFU arm and 69 patients in the MTXFAFU arm. Probability of survival did not differ between the two arms: median survival time (MST) was 14.7 months in the IRIFAFU arm and 14.8 months in the MTXFAFU; estimated survival rates were 60 and 57% at 1 year, 23 and 25% at 2 years, and 11 and 9% at 3 years, respectively. These results are probably explained by the effective second-line treatments administered after the induction failure. Indeed, comparison of survival of patients in whom first-line chemotherapy failed, and not receiving salvage chemotherapy for whatever reason (deterioration of clinical condition, patient refusal, etc.), showed that patients treated up-front with IRIFAFU lived about twice as long as patients receiving MTXFAFU: MST was 11.9 months compared with 6.4 months (Wilcoxon test, P = 0.038; log-rank test, P = 0.088).


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
This study was designed to compare the activity and toxicity of a new regimen including CPT-11 combined with FA-modulated 5-FU administered as i.v. bolus (IRIFAFU), with a double modulation of 5-FU by previous exposure to MTX and FA (MTXFAFU). The new regimen was devised on the grounds of preclinical in vitro and in vivo evidence suggesting that pre-exposure to CPT-11 followed by 5-FU, and not the simultaneous combination, might better exploit the synergism between the two cytotoxic drugs [16, 1922].

In this study, we have demonstrated that the proportion of complete and partial responses was significantly greater with IRIFAFU. Taking into account also the patients showing a tumour shrinkage that did not qualify for a major response, we are confident that the RR of the experimental regimen was not overestimated in this trial. It is also worth noting the activity of this regimen in hepatic metastases (40%), regardless of the extent of liver involvement. Multivariate analysis, including possible confounding factors, confirmed the significantly higher activity of the IRIFAFU regimen.

This greater activity was obtained at a price of acceptable toxicity. Despite the frequent occurrence of severe neutropenia, neutropenic fever/infections were seldom reported. Other haematological toxicities were negligible. Diarrhoea was troublesome in some patients, and it was responsible for an early death in this study. The whole occurrence of severe diarrhoea (<14%), although greater than that produced by the MTXFAFU regimen, did not, however, seem higher than reported with other standard weekly or monthly 5-FU i.v. bolus regimens. Occurrence of severe stomatitis was substantially lower with the IRIFAFU than with the MTXFAFU regimen. Excluding hair loss, less than one-third of patients treated with IRIFAFU suffered an adverse event of grade 3 or greater, and the proportion of patients dying early (within 2 months from initial therapy) from possible treatment-related side-effects was <3%.

Our study was also powered to detect a 50% increase in median TTP (from 4 to 6 months) with the experimental treatment. Actually, a 50% increase in TTP was produced by the IRIFAFU treatment. This difference was significant when assessed with the Wilcoxon test, but not with the log-rank test. Indeed, TTP curves diverged from 2 to 8 months from initial therapy, than gradually approached again to overlap at 12 months and thereafter.

We have tried to report our results in the context of two other large trials comparing a combination regimen of CPT-11 plus modulated 5-FU with modulated 5-FU alone in advanced colorectal cancer [11, 12] (Table 4). In the Saltz trial, the combination regimen produced a significantly greater RR (39% compared with 21%), and a significantly longer TTP (7.0 months compared with 4.3 months) and MST (14.8 months compared with 12.6 months) than the Mayo regimen. This regimen was characterised by much more severe diarrhoea (23% compared with 13%), but significantly less stomatitis (2% compared with 17%) in comparison with the reference regimen. Severe neutropenia affected about 54% of patients, but grade 4 incidence was lower than that reported with the monthly regimen (24% compared with 42.5%). In this study, drug-related deaths occurred in <1% of treated patients. An alert has recently been raised, however, on the basis of unexpectedly high proportions of early deaths reported when using this regimen in ongoing trials in the adjuvant (4.8%) as well in the metastatic setting (2.2%) [23].


View this table:
[in this window]
[in a new window]
 
Table 4. 5-Fluorouracil and folinic acid combined with irinotecan regimens in advanced colorectal carcinoma: randomised trials
 
Another trial compared the addition of CPT-11 to the weekly or biweekly 5-FU infusional regimens [12]. The addition of CPT-11 to the weekly 24-h infusional regimen was more active than the comparative regimen (RR 40% compared with 25%); however, it also produced a significantly greater occurrence of severe diarrhoea (44%) and neutropenia (29%). On the contrary, the biweekly 48-h combination regimen caused more severe diarrhoea (13%) than the single-agent treatment (6%), but it appeared better tolerated than the weekly combination regimen. Severe neutropenia occurred, however, in 46% of treated patients, and neutropenic fever or infections affected 5.5% of them. In this trial, 1 of 145 patients treated with the combination regimen died after the first cycle. Both these infusional regimens needed the insertion of an indwelling transcutaneous or totally implanted central venous catheter. This may represent a drawback, because the treatment may be complicated by the occurrence of deep venous thrombosis, infections, or catheter occlusion. In addition, a disposable external pump is required, increasing the cost of the whole treatment. Therefore, our IRIFAFU regimen compared well with these weekly or biweekly combination treatments. Occurrence of severe neutropenia or diarrhoea seemed similar, as were the RR, the median TTP and the MST. Moreover, we believe that our regimen is more easily administered in the outpatient clinic; indeed, it is short-lasting and, since it does not require infusional devices or pumps, it is more safe; it may be also less expensive, which is relevant in countries were there are budgetary restrictions on health care.

In conclusion, the IRIFAFU biweekly i.v. bolus regimen was more active than the MTXFAFU regimen, because it produced a significantly greater RR, and a longer median time to treatment failure and TTP. The use of a more active regimen as a front-line approach seemed to improve the outcome of patients who, for any reason, were not submitted to second-line chemotherapy. This regimen caused more diarrhoea and neutropenia, but less thrombocytopenia and stomatitis, than the comparative regimen, and it was generally well tolerated by patients. It may represent a new option for treating advanced colorectal cancer patients, and it is now the reference regimen for our Cooperative Group.


    Acknowledgements
 
We wish to thank Maurizio Montella and Anna Crispo (Epidemiology and Biostatistic Unit, National Tumour Institute, Naples) for contributing to the statistical analysis of this trial.

The following investigators (and Institutions) are co-authors of this trial: P. Comella, R. Casaretti, A. Avallone, M. Montella, G. Comella (National Tumour Institute, Naples); F. De Vita, M. Orditura, G. Catalano (2nd University Medical School, Naples); L. De Lucia (City Hospital, Caserta); E. Crucitta, V. Lorusso, M. De Lena (Oncology Institute, Bari); M. Biglietto (Cardarelli Hospital, Naples); A. Farris, G. Sanna, M. G. Sarobba (University Medical School, Sassari); F. Del Gaizo, C. Belli (City Hospital, Avellino); S. Palmeri, A. Russo, N. Pizzardi, V. Accurso (University Medical School, Palermo); A. Iannelli (City Hospital, Siderno); S. Mancarella, F. Musca (City Hospital, Campi Salentino); S. Tafuto, A. Gravina (City Hospital, Pozzuoli); L. Maiorino, G. Leopaldi (San Gennaro Hospital, Naples); G. P. Ianniello (City Hospital, Benevento); F. Buzzi (City Hospital, Terni); C. Brunetti (City Hospital, Manduria); D. Muci (City Hospital, Nardò); S. Leo (IRCCS, Castellana Grotte); G. De Cataldis (Da Procida Hospital, Salerno); F. Avino, L. De Luca (Pellegrini Hospital, Naples); E. Greco, C. Aiello (City Hospital, Lametia).


    Footnotes
 
+ Correspondence to: Dr P. Comella, Division of Medical Oncology A, National Tumour Institute, Naples, Italy. Tel: +39-081-590-3227; Fax: +39-081-590-3821; E-mail: pcomella@sirio-oncology.it Back

§ Listed in Acknowledgements. Back


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. Sobrero A, Kerr D, Glimelius B et al. New directions in the treatment of colorectal cancer: a look to the future. Eur J Cancer 2000; 36: 559–566.[ISI][Medline]

2. Advanced Colorectal Cancer Meta-Analysis Project. Modulation of fluorouracil by leucovorin in patients with advanced colorectal cancer: evidence in terms of response rate. J Clin Oncol 1992; 10: 896–903.[Abstract]

3. Advanced Colorectal Cancer Meta-Analysis Project. Meta-analysis of randomized trials testing the biochemical modulation of fluorouracil by methotrexate in metastatic colorectal cancer. J Clin Oncol 1994; 12: 960–969.[Abstract]

4. Comella P, Lorusso V, Casaretti R et al. Concurrent modulation of 5-fluorouracil with methotrexate and l-leucovorin: An effective and moderately toxic regimen for the treatment of advanced colorectal carcinoma. A multicenter phase II study of the Southern Italy Cooperative Oncology Group. Tumori 1999; 85: 465–472. [ISI][Medline]

5. Abad A, Garcia P, Gravalos C et al. Sequential methotrexate, 5-fluorouracil (5-FU), and high-dose leucovorin versus 5-FU and high-dose leucovorin versus 5-FU alone for advanced colorectal cancer. Cancer 1995; 75: 1238–1244.[ISI][Medline]

6. Glimelius B. Biochemical modulation of 5-fluorouracil: A randomized comparison of sequential methotrexate, 5-fluorouracil and leucovorin versus sequential 5-fluorouracil and leucovorin in patients with advanced symptomatic colorectal cancer. Ann Oncol 1993; 4: 235–250.[Abstract]

7. Polyzos A, Tsavaris N, Giannopoulos C et al. Biochemical modulation of fluorouracil: comparison of methotrexate, folinic acid, and fluorouracil versus folinic acid and fluorouracil in advanced colorectal cancer: a randomized trial. Cancer Chemother Pharmacol 1996; 38: 292–297.[ISI][Medline]

8. Sobrero A, Zaniboni A, Frassineti GL et al. Schedule specific biochemical modulation of 5-fluorouracil in advanced colorectal cancer: a randomized study. Ann Oncol 2000; 11: 1413–1420.[Abstract]

9. Rougier P, Van Cutsem E, Bajetta E et al. Randomized trial of irinotecan versus fluorouracil by continuous infusion after fluorouracil failure in patients with metastatic colorectal cancer. Lancet 1998; 352: 1407–1412.[ISI][Medline]

10. Cunningham D, Pyrhonen S, James RD et al. Randomized trial of irinotecan plus supportive care versus supportive care alone after fluorouracil failure for patients with metastatic colorectal cancer. Lancet 1998; 352: 1413–1418.[ISI][Medline]

11. Saltz LB, Cox JV, Blanke C et al. Irinotecan plus fluorouracil and leucovorin for metastatic colorectal cancer. N Engl J Med 2000; 343: 905–914.[Abstract/Free Full Text]

12. Douillard JY, Cunningham D, Roth AD et al. Irinotecan combined with fluorouracil compared with fluorouracil alone as first-line treatment for metastatic colorectal cancer. A multicentre randomised trial. Lancet 2000; 355: 1041–1047.[ISI][Medline]

13. Comella P, Casaretti R, De Vita F et al. Concurrent irinotecan and 5-fluorouracil plus levo-folinic acid given every other week in the first-line management of advanced colorectal carcinoma: a phase I study of the Southern Italy Cooperative Oncology Group. Ann Oncol 1999; 10: 915–921.[Abstract]

14. Comella P, De Vita F, Mancarella S et al. Biweekly irinotecan or raltitrexed plus 6S-leucovorin and bolus 5-fluorouracil in advanced colorectal carcinoma: a Southern Italy Cooperative Oncology Group phase II–III randomized study. Ann Oncol 2000; 11: 1323–1333.[Abstract]

15. Miller AB, Hoogstraten B, Staquet M, Winkler V. Reporting results of cancer treatment. Cancer 1981; 47: 207–214.[ISI][Medline]

16. Guichard S, Cussac D, Hennebelle I et al. Sequence-dependent activity of the irinotecan-5FU combination in human colon-cancer model HT29 in vitro and in vivo. Int J Cancer 1997; 27: 729–734.

17. Thall PF, Simon R, Ellenberg SS. A two-stage design for choosing among experimental treatments and a control in clinical trials. Biometrics 1989; 45: 537–547.[ISI][Medline]

18. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457–481.[ISI]

19. Haybittle JL. Significance testing in comparison of survival curves from clinical trials of cancer treatment. Eur J Cancer Clin Oncol 1986; 11: 1279–1283.

20. Mullany S, Svingen PA, Kaufmann SH et al. Effect of adding topoisomerase I poison 7-ethyl-10-hydrossycamtpthecin (SN-38) to 5-fluorouracil and folinic acid in HCT8 cells: elevated dTTP pools and enhanced cytotoxicity. Cancer Chemother Pharmacol 1998; 42: 391–399. [ISI][Medline]

21. Mans DRA, Grivicich I, Peters GJ et al. Sequence-dependent growth inhibition and DNA damage formation by the irinotecan-5-fluorouracil combination in human colon carcinoma cell lines. Eur J Cancer 1999; 35: 1851–1861.[ISI][Medline]

22. Cao S, Rustum YM. Synergistic antitumor activity of irinotecan in combination with 5-fluorouracil in rats bearing advanced colorectal cancer: role of drug sequence and dose. Cancer Res 2000; 60: 3717–3721.[Abstract/Free Full Text]

23. Sargent DJ, Niedzwiecki D, O’Connel MJ et al. Recommendation for caution with irinotecan, fluorouracil, and leucovorin for colorectal cancer. N Engl J Med 2001; 345: 144–145.[Free Full Text]