RATIONALE: Imatinib mesylate may stop the growth of tumor cells by blocking some of the enzymes needed for cell growth. It is not yet known which dose of imatinib mesylate is more effective in treating gastrointestinal stromal tumor.

PURPOSE: This randomized phase III trial is studying two different doses of imatinib mesylate to compare how well they work in treating patients with unresectable or metastatic gastrointestinal stromal tumor.

OBJECTIVES:

Primary

• To compare outcomes of patients with unresectable or metastatic gastrointestinal stromal tumor that expresses KIT (CD117) treated with low-dose imatinib mesylate vs high-dose imatinib mesylate.

Secondary

• To assess response rates in patients treated with two different doses of imatinib mesylate.
• To assess the toxicities of two different doses of imatinib mesylate in these patients.

OUTLINE: This is a multicenter study. Patients are stratified according to participating center, measurability of disease (measurable vs non-measurable), and WHO performance status (0-2 vs 3). Patients are randomized to 1 of 2 treatment arms.

• Arm I: Patients receive low-dose oral imatinib mesylate once daily in the absence of disease progression or unacceptable toxicity.
• Arm II: Patients receive high-dose oral imatinib mesylate twice daily in the absence of disease progression or unacceptable toxicity.

In the event of disease progression, patients on arm I may cross over to arm II and receive high-dose imatinib mesylate. Patients who continue to progress despite treatment with high-dose imatinib mesylate are removed from the study.

After completion of study therapy, patients are followed periodically.

DISEASE CHARACTERISTICS:

• Histologically confirmed gastrointestinal stromal tumor (GIST)

  • Metastatic or unresectable disease

• Immunohistochemical confirmation of KIT (CD117) expression by tumor as documented by DAKO antibody staining

• Measurable or non-measurable disease by conventional imaging (CT scan or MRI) or physical examination

  • If a target lesion has been previously embolized or irradiated, there must be objective evidence of progression to be considered for response assessment

• No known brain metastasis

PATIENT CHARACTERISTICS:

• WHO performance status 0-3
• Bilirubin ≤ 1.5 times upper limit of normal (ULN)
• AST and ALT ≤ 2.5 times ULN (≤ 5 times ULN if hepatic metastases are present)
• Creatinine ≤ 1.5 times ULN
• ANC ≥ 1,000/mm³
• Platelet count ≥ 100,000/mm³
• Hemoglobin ≥ 9 g/dL (transfusions allowed)
• Not pregnant or nursing
• Negative pregnancy test
• Fertile patients must use effective barrier contraception during and for up to 3 months after completion of study therapy
• No NYHA class III-IV cardiac disease
• No congestive heart failure or myocardial infarction within the past 2 months
• No severe and/or uncontrolled concurrent medical disease (e.g., uncontrolled diabetes, uncontrolled chronic renal or liver disease, or active uncontrolled infection [e.g., HIV])
• No other prior malignancy except adequately treated basal cell or squamous cell skin cancer, carcinoma in situ of the cervix, adequately treated stage I or II cancer from which the patient is currently in complete remission, or any other cancer from which the patient has been disease-free for 5 years
• No medical, psychological, familial, sociological, or geographical condition that, in the opinion of the investigator, may preclude the patient's ability to tolerate or complete study treatment, comply with study protocol and follow-up schedule, or give reliable informed consent

PRIOR CONCURRENT THERAPY:

• Recovered from all prior therapy

• More than 28 days since prior chemotherapy, biologic therapy, or any other investigational drug

• More than 14 days since prior major surgery

• No concurrent therapeutic anticoagulation with coumarin derivatives

  • Concurrent therapeutic anticoagulation with low-molecular weight heparin allowed
  • Concurrent mini-dose coumarin derivatives (i.e., equivalent to 1 mg of oral warfarin daily) as prophylaxis allowed

• No concurrent cytokines (i.e., filgrastim [G-CSF] or sargramostim [GM-CSF]) to support blood counts

• No other concurrent investigational drugs

• No other concurrent anticancer agents, including chemotherapy, radiotherapy, or anticancer biologic therapy