RATIONALE: Drugs used in chemotherapy, such as floxuridine and dexamethasone, work in different ways to stop the growth of tumor cells, either by killing the cells or by stopping them from dividing. Monoclonal antibodies, such as bevacizumab, can block tumor growth in different ways. Some block the ability of tumor cells to grow and spread. Others find tumor cells and help kill them or carry tumor-killing substances to them. Bevacizumab may also stop the growth of tumor cells by blocking blood flow to the tumor. Giving chemotherapy directly into the arteries around the tumor together with bevacizumab may kill more tumor cells.

PURPOSE: This phase II trial is studying how well giving floxuridine and dexamethasone as a hepatic arterial infusion together with bevacizumab works in treating patients with unresectable primary liver cancer.

OBJECTIVES:

Primary

• Determine the median time to progression in patients with unresectable primary hepatic malignancy treated with hepatic arterial infusion comprising floxuridine and dexamethasone in combination with systemic bevacizumab.

Secondary

• Determine the utility of dynamic contrast-enhanced MRI (DCE-MRI) for assessing changes in tumor perfusion before and during treatment.
• Correlate DCE-MRI findings with radiographic tumor response.

Tertiary

• Correlate the expression patterns of vascular endothelial growth factor (VEGF) receptor (VEGFR)-1, VEGFR-2, and VEGFR-3 and their cognate ligands (including VEGF-A, VEGF-B, VEGF-C, VEGF-D, and placenta growth factor [PlGF]) with disease progression and survival after therapy.
• Assess the pro-angiogenic activity of peripheral blood before and during treatment.
• Assess tumors for immunohistochemical markers of hypoxia (e.g., hypoxia-inducible factor [HIF-1α], carbonic anhydrase IX [CA IX], and glucose transporters [Glut-1 and Glut-3]) for correlation with initial and treatment-related changes in perfusion and permeability, as determined by DCE-MRI.

OUTLINE: This is an open-label, nonrandomized study.

Patients undergo placement of the hepatic arterial infusion (HAI) pump and a cholecystectomy. Approximately 2 weeks later, patients receive floxuridine and dexamethasone by HAI continuously on days 1-14 and bevacizumab IV over 30-90 minutes on day 15 of course 1 and on days 1 and 15 of all subsequent courses. Courses repeat every 4 weeks in the absence of disease progression or unacceptable toxicity.

Patients undergo dynamic contrast-enhanced MRI (DCE-MRI) on days 1 and 15 of course 1 and then every 8 weeks thereafter.

Tumor and nontumor tissue is collected at the time of the HAI pump placement. Tissue is examined for the expression of vascular endothelial growth factor (VEGF)-A, -B, -C, and -D, placenta growth factor (PlGF), and VEGF receptor (VEGFR)-1, -2, and -3 by immunohistochemistry. Peripheral blood is collected at baseline and on day 1 of each course. Plasma levels of VEGF-A, -B, -C, and -D are measured by immunoenzyme techniques. Blood is also examined by flow cytometry and immunological methods and by protein extraction and analysis of VEGF and VEGFR expression (by western blot). Immunohistochemical markers of hypoxia in tissue, including hypoxia-inducible factor (HIF-1α), carbonic anhydrase IX (CA IX), glucose transporters (Glut-1 and Glut-3), and Ki-67 are assessed.

After completion of study treatment, patients are followed periodically.

PROJECTED ACCRUAL: A total of 55 patients will be accrued for this study.

DISEASE CHARACTERISTICS:

• Histologically confirmed hepatocellular carcinoma (HCC) or intrahepatic cholangiocarcinoma (ICC)

  • Peripheral, cholangiolar, or cholangiocellular types
  • Mixed HCC/ICC disease allowed
  • Unresectable disease
  • Less than 70% liver involvement

• Radiographically bidimensionally measurable disease, defined as lesion ≥ 2 cm in the greatest diameter

• May have failed prior systemic chemotherapy or ablative therapy

• No radiographic evidence of esophageal varices

• No history of variceal hemorrhage

• No occlusion of the main portal vein or the right and left portal branches

• No clinical ascites

• Patients ineligible for first-line MSKCC protocols for HCC are eligible for this study provided there is no clinical or radiographic evidence of extrahepatic disease

• No metastatic disease, including CNS metastases

PATIENT CHARACTERISTICS:

• Life expectancy ≥ 12 weeks

• Karnofsky performance status 60-100%

• Considered a candidate for general anesthesia and hepatic artery pump placement

• Platelet count > 100,000/mm³

• Albumin > 2.5 g/dL

• Bilirubin < 1.8 mg/dL

• WBC > 3,500/mm³

• PTT < 1.5 times upper limit of normal

• INR < 1.5 OR in-range INR (usually 2.0-3.0) for patients on a stable dose of therapeutic warfarin

• Urine protein < 1+ by dipstick or urine analysis OR urine protein:creatinine ratio < 1.0

  • If proteinuria ≥ 2+ at baseline, patient must have < 1 g protein/24-hour collection

• No concurrent disease or illness that would preclude study participation, including any of the following:

  • Hepatic encephalopathy
  • Sclerosing cholangitis
  • Gilbert's disease
  • Active infection

• No known CNS disease

• No history of allergic reactions attributed to compounds of similar chemical or biologic composition to bevacizumab

• No psychiatric illness or social situation that would preclude study compliance

• No abdominal fistula, gastrointestinal perforation, or intra-abdominal abscess within the past 6 months

• No serious or nonhealing active wound, ulcer, or bone fracture

• No bleeding diathesis or coagulopathy

• No clinically significant cardiovascular disease, including any of the following:

  • Uncontrolled hypertension, defined as systolic blood pressure (BP) > 150 mm Hg or diastolic BP > 100 mm Hg on antihypertensive medications
  • New York Heart Association class II-IV congestive heart failure
  • Vascular disease (e.g., aortic aneurysm, aortic dissection)
  • Myocardial infarction within the past 6 months
  • Symptomatic peripheral vascular disease
  • Unstable angina within the past 6 months
  • History of hypertensive crisis
  • Transient ischemic attack
  • Stroke

• No other concurrent malignancy except localized basal cell or squamous cell skin cancer

• Chronic hepatitis and/or cirrhosis allowed provided it is Child-Pugh class A disease

• Not pregnant or nursing

• Negative pregnancy test

• Fertile patients must use effective contraception

PRIOR CONCURRENT THERAPY:

• See Disease Characteristics
• More than 4 weeks since prior and no other concurrent experimental therapy except on a Genentech-sponsored bevacizumab cancer study
• More than 4 weeks since prior major surgical procedure or open biopsy
• More than 1 week since prior minor surgical procedure (e.g., core biopsy), excluding placement of a vascular access device
• No prior external-beam radiation therapy to the liver
• No prior floxuridine
• No chronic daily treatment with nonsteroidal anti-inflammatory medications known to inhibit platelet function
• No chronic daily treatment with aspirin (> 325 mg/day)
• No concurrent or recent use of a thrombolytic agent
• No concurrent major surgery