Although it is possible to cure bowel cancer when it is detected at an early stage, in many cases it may spread to involve other organs and in these cases is generally incurable. Chemotherapy prolongs survival and improves quality of life in such patients, but standard chemotherapy for this disease has not been defined.

There are several possible chemotherapy treatments for patients with bowel cancer, which has spread to other organs. However, these treatments are only partly effective and only work for a limited period of time. Most treatments are associated with a number of possible side effects which may have a detrimental effect on quality of life. Thus, it is imperative that more effective treatments with the lowest possible risk of side effects are developed.

Previous studies have shown that the addition of a new type of antibody treatment (bevacizumab) to an intensive combination chemotherapy regimen improved survival in patients with advanced bowel cancer and extended the time before tumours began to grow. However, intensive chemotherapy is likely to only be a suitable treatment for a proportion of patients with bowel cancer, because intensive chemotherapy causes a high rate of side effects.

This study compares a gentle chemotherapy treatment (capecitabine chemotherapy tablets given by mouth) with the combination of capecitabine and bevacizumab and the combination of capecitabine, bevacizumab and intravenous mitomycin C.

It is expected that a gentle chemotherapy treatment or a gentle chemotherapy treatment combined with bevacizumab would be an appropriate treatment for both young and fit patients as well as older and less fit patients who would not easily tolerate intensive chemotherapy.

Aims - The phase II stage of the study aims to determine the relative toxicity of the combination of capecitabine and bevacizumab and the combination of capecitabine, mitomycin C (MMC) and bevacizumab with that of capecitabine monotherapy and to assess tumour response rate (RECIST criteria) for each arm

For Phase III stage the primary objective is to compare progression-free survival (PFS) on the three arms. Secondary objectives are to determine treatment related toxicity; to determine tumour response rates (RECIST criteria); to determine overall survival for each treatment arm; to compare disease related symptoms and Quality of life and to determine cost effectiveness of bevacizumab containing treatments.

Research Plan Synopsis - Trial Design: Randomised, stratified multicentre phase II/III study. The study will proceed in 2 phases, initially a randomised phase II stage evaluating safety after 60 patients (approx 20 per arm) and 150 patients (approx 50 per arm) have completed at least 6 weeks' treatment. This will continue with a randomised phase III stage evaluating activity, toxicity and quality of life measures.

Treatments: Patients will be randomised to treatment in either one of the three arms: I) Capecitabine as monotherapy ; 2) Capecitabine and bevacizumab; or 3) Capecitabine and bevacizumab and MMC.

Drug administration: Arm 1: Capecitabine 2500mg/m2/d (in 2 divided doses) d1-14 q3weekly. Arm 2: Capecitabine administered as per Arm 1 plus Bevacizumab 7.5 mg/kg q3weekly. Arm 3: Capecitabine and Bevacizumab administered as per Arm 2 plus Mitomycin C 7 mg/m2 q 6weekly (maximum dose 14 mg, maximum 4 treatments).

Analysis: A total sample size of 333 patients (111 per group) will be required to detect an improvement of at least 3.1 months in progression free survival from 5.5 to 8.6 months using a 2-tailed comparison, 2.5% level of significance, 3-year accrual and 1-year follow-up. The 12-month survival rate for patients on capecitabine alone is 50%. A sample size of 111 per arm will have 80% power to detect an increase of 17% in the 1-year rate from 50% to 67% based on a significance level of 2.5%, 3-year accrual and 1-year follow-up. For both endpoints (PFS and survival) the difference between capecitabine alone and the regimen containing MMC is expected to be greater (in the order of 4.5 months) which will yield > 80% power to detect the difference in PFS.

Whilst not a primary comparison, the study will nevertheless still have 80% power to detect a 5.5 month difference between the two experimental arms as a secondary comparison based on a level of significance of 1.7%. Secondary endpoints include treatment related toxicity. Toxicity analyses will include treatment-received population, which includes all patients who received at least 1 dose of study treatment. Toxicity will be described by tabulating the proportions of patients with a worst toxicity grade of 0, 1, 2, 3, or 4 for each of the relevant NCI CTC AE scales.

Phase II: Confidence intervals for the difference in the incidence of bevacizumab and MMC associated grade 3/4 toxicities on the three arms will be calculated. If the incidence of these toxicities in the triple combination regimen (Capecitabine/MMC/ bevacizumab) exceeds the rate of toxicity experienced in the capecitabine/bevacizumab combination by more than 20% then consideration will be given to dose adjustment or stopping recruitment into the triple combination arm.

Phase III: The PFS for capecitabine chemotherapy alone is expected to be about 5.5 months and this is expected to increase to 9 months with the addition of bevacizumab, which is considered to be clinically meaningful. Using an overall 95% confidence level and 80% power and a 2.5% significance level for each comparison 111 patients per arm are required to detect differences based on a 36-month accrual and 12-month follow-up.

Outcomes and Significance - This randomised phase II/III study aims to compare capecitabine monotherapy with capecitabine plus bevacizumab and capecitabine plus bevacizumab plus MMC in patients with previously untreated metastatic colorectal cancer.

The use of either MMC or bevacizumab to 5FU based chemotherapy appears to result in improved activity without substantial increases in toxicity. Thus regimens incorporating these agents could have significant activity and be well tolerated. These regimens could be suitable as a low toxicity palliative regimen for a broad range of the population of patients with metastatic colorectal cancer including older patients with co-morbidities.

As it is anticipated that rates of acute toxicity with each regimen will be lower than those observed with oxaliplatin or CPT-11-based regimens, the target population may be more broad ranging than most other studies. Thus, it may include older patients, patients with limited performance status (PS2), patients with co-morbidities or patients in whom there are concerns relating to toxicity with oxaliplatin or CPT-11-based combination chemotherapy. However, it is not restricted to this population and younger, fitter patients may also be enrolled in the study as a lower rate of side effects is likely to be associated with improved quality of life.

Inclusion Criteria:

• Histological diagnosis of colorectal cancer
• Metastatic disease that is not resectable
• Age > 18 years
• Any patient in whom the investigator considers capecitabine monotherapy appropriate
• Measurable and/or non-measurable disease as assessed by CT scan
• ECOG performance status 0, 1 or 2. Patients with PS2 should have serum albumin >30 g/L
• No prior chemotherapy except for adjuvant chemotherapy given in association with (i) complete resection of primary colon or rectal cancer provided there is no clinical, radiological or biochemical evidence of relapse for at least 6 months after completion of adjuvant treatment and/or (ii) complete resection of limited colorectal metastases to liver and/or lung provided there is no clinical, radiological or biochemical evidence of relapse for at least 6 months after completion of adjuvant treatment
• Adequate bone marrow function with platelets > 100 X 109/l; neutrophils > 1.5 X 109/l i) Adequate renal function, with calculated creatinine clearance >30 ml/min (Cockcroft and Gault). For patients with creatinine clearance <50 ml/min the starting dose of capecitabine may not be greater than 2000 mg/m2/d (see Section 7.1)
• Adequate hepatic function with serum total bilirubin < 1.5 X upper limit of normal range
• Life expectancy of at least 12 weeks
• No other concurrent uncontrolled medical conditions
• No other malignant disease apart from non-melanotic skin cancer or carcinoma in situ of the uterine cervix or any other cancer treated with curative intent >2 years previously without evidence of relapse
• Women and partners of women of childbearing potential must agree to use adequate contraception
• Written informed consent

Exclusion Criteria:

• Medical or psychiatric conditions that compromise the patient's ability to give informed consent or to complete the protocol
• Patients with a lack of physical integrity of the upper gastrointestinal tract, or known malabsorption syndromes.
• Uncontrolled hypertension
• Active bleeding disorders within the last 3 months
• Patients on full anticoagulation with warfarin. (Patients who require full anticoagulation and who wish to participate in the study should be converted to low molecular weight heparin). (Note: patients receiving full anticoagulation with low molecular weight heparin should have no evidence of tumour invading or abutting major blood vessels on any prior CT scan)
• Participation in any investigational drug study within the previous 8 weeks
• Patients with uncontrolled clinically significant cardiac disease, arrhythmias or angina pectoris
• Patients with a history of acute myocardial infarction or cerebrovascular accident within the last 12 months
• Regular use of aspirin (>325mg/day) or NSAIDs (low dose aspirin (<325 mg/d), or occasional use of NSAIDs is acceptable)
• CNS metastases
• Major surgical procedure within the last 28 days
• Serious non-healing wound, ulcer or bone fracture
• 24 hour urinary protein > 2g/ 24 hours ( performed if urine dipstick > 1+ )
• Pregnancy or lactation