RATIONALE: Drugs used in chemotherapy, such as oxaliplatin, capecitabine, and floxuridine, work in different ways to stop the growth of tumor cells, either by killing the cells or by stopping them from dividing. Hepatic arterial infusion uses a catheter to carry cancer-killing substances directly into the liver. Giving the drugs in different ways may kill more tumor cells. It is not yet known whether giving oxaliplatin and capecitabine together with floxuridine is more effective than giving oxaliplatin and capecitabine alone in treating patients who are undergoing surgery and/or ablation for liver metastases from colorectal cancer.

PURPOSE: This randomized phase III trial is studying how well oxaliplatin, capecitabine, and hepatic arterial infusion of floxuridine works compared to oxaliplatin and capecitabine in treating patients who are undergoing surgery and/or ablation for liver metastases from colorectal cancer.

OBJECTIVES:

Primary

• Compare progression-free interval (PFI) in patients undergoing surgical resection and/or ablation for hepatic metastases from colorectal cancer treated with adjuvant therapy comprising oxaliplatin and capecitabine with vs without hepatic arterial infusion of floxuridine.

Secondary

• Compare overall survival and liver PFI between the two treatment groups.
• Assess toxicity in each of the treatment regimens.
• Compare self-reported symptoms between two treatment groups.
• Compare quality of life in each of the treatment regimens.

Tertiary

• Examine the prognostic worth, in terms of PFI, of specific molecular markers in hepatic metastases.

OUTLINE: This is a randomized study. Patients are stratified according to intended surgical technique (surgical resection alone vs cryoablation or radiofrequency ablation [RFA] alone vs combination of resection and ablation) and prior adjuvant chemotherapy regimen (chemotherapy with vs without oxaliplatin vs no chemotherapy). Patients are randomized to 1 of 2 treatment arms.

All patients undergo surgical resection and/or hepatic cryoablation or RFA to remove colorectal hepatic metastases. Patients randomized to arm II also undergo intra-arterial pump and catheter placement.

• Arm I (oxaliplatin and capecitabine): Within 8 weeks after surgery and/or ablation, patients receive oxaliplatin IV over 2 hour on day 1 and oral capecitabine twice daily on days 1-14. Treatment repeats every 21 days for 8 courses in the absence of disease progression or unacceptable toxicity.
• Arm II (oxaliplatin, capecitabine, and hepatic arterial infusion of floxuridine): Within 8 weeks after surgery and/or ablation, patients receive a continuous hepatic arterial infusion of floxuridine on days 1-14, oxaliplatin IV over 2 hours on day 22, and oral capecitabine twice daily on days 22-35. Treatment repeats every 42 days for 4 courses in the absence of unacceptable toxicity. Beginning with course 5, patients receive oxaliplatin IV over 2 hours on day 1 and oral capecitabine twice daily on days 1-14. Treatment with oxaliplatin and capecitabine repeats every 21 days for 4 courses in the absence of disease progression or unacceptable toxicity.

Quality of life is assessed at baseline, 4-6 weeks after surgery or ablation, 18 weeks after beginning of chemotherapy, and 4-6 weeks after completion of chemotherapy.

After completion of study treatment, patients are followed periodically for 5 years.

PROJECTED ACCRUAL: A total of 400 patients will be accrued for this study.

DISEASE CHARACTERISTICS:

• Histologically* or cytologically confirmed colorectal adenocarcinoma
• No other cellular type (e.g., sarcoma, lymphoma, or carcinoid) NOTE: *If the primary colorectal tumor and the hepatic lesions have been identified at the same time and it is possible to biopsy the colorectal lesion, the patient will be eligible without histiologic confirmation of the colorectal primary cancer as long as other radiographic studies or scans document the characteristics of a colorectal cancer
• Synchronous or metachronous metastatic disease confined to the liver
• No more than 6 hepatic metastatic lesions that can potentially be resected or ablated
• For patients presenting with synchronous lesion(s) in the colon and/or rectum, the primary tumors must, in the opinion of the investigator, appear to be completely resectable
• Must be able to undergo surgery and/or ablation withing 28 days following randomization
• No evidence of extrahepatic metastases
• No prior colorectal metastases
• No recurrent colorectal cancer concurrent with hepatic metastases

PATIENT CHARACTERISTICS:

• Life expectancy ≥ 5 years, excluding their colorectal cancer
• ECOG (Zubrod) performance status 0-1
• No other malignancy within the past 5 years except carcinoma in situ of the cervix, melanoma in situ, or basal cell or squamous cell skin cancer
• Absolute granulocyte count ≥ 1,200/mm^3
• Platelet count ≥ 100,000/mm^3
• PT/INR ≤ 1.5
• Direct bilirubin normal
• Alkaline phosphatase ≤ 2.5 times upper limit of normal (ULN)
• AST ≤ 2.5 times ULN
• Creatinine clearance > 50 mL/min
• Not pregnant or lactating
• Negative pregnancy test
• Patients with child bearing potential must agree to use adequate contraception
• Able to swallow oral medication
• No preexisting chronic hepatic disease (e.g., chronic active hepatitis or cirrhosis)
• No grade 3 or 4 anorexia or nausea
• No vomiting ≥ grade 2
• No clinically significant peripheral neuropathy defined as ≥ grade 2 neurosensory or neuromotor toxicity
• No psychiatric or addictive disorders or other condition that, in the opinion of the investigator, would preclude study participation

PRIOR CONCURRENT THERAPY:

• Prior adjuvant fluorouracil alone or in combination with levamisole, leucovorin calcium, irinotecan hydrochloride, or oxaliplatin allowed if these regimens were completed > 6 months ago
• No prior resection/ablation, hepatic arterial infusion therapy or any systemic chemotherapy for metastatic disease
• Prior excisional biopsy allowed
• No prior radiotherapy to the liver
• No concurrent bevacizumab in patients receiving hepatic arterial infusion of floxuridine
• No concurrent halogenated antiviral agents such as sorivudine or brivudine in patients receiving fluorouracil, floxuridine, or capecitabine
• No concurrent filgrastim (G-CSF), pegfilgrastim, or sargramostim (GM-CSF) as primary prophylaxis for neutropenia
• Following neutropenic events, these drugs may be used at the physician''''s discretion during subsequent cycles
• No other concurrent cancer therapy
• No other concurrent investigational agents