PRIMARY OBJECTIVES: I. Determine the maximum tolerated dose (MTD) of FR901228 (depsipeptide) in patients with recurrent malignant gliomas who are taking enzyme-inducing antiepileptic drugs (EIAEDs). (Phase I) II. Determine the safety profile of this drug in these patients. (Phase I) III. Determine the pharmacokinetics and pharmacodynamics of this drug in these patients. (Phase I) IV. Determine the clinical efficacy of this drug, as measured by 6-month progression-free survival and objective tumor response, in these patients. (Phase II) V. Determine the safety profile of this drug when administered at the phase I MTD concurrently with or without EIAEDs in these patients. (Phase II) OUTLINE: This is a multicenter, phase I, dose-escalation study followed by a phase II study. Patients are stratified according to study phase (I vs II), concurrent use of enzyme-inducing anti-epileptic drugs (EIAEDs) (yes vs no), histology (recurrent glioblastoma multiforme/gliosarcoma vs recurrent anaplastic glioma), pre-operative candidacy (yes vs no), and measurable/evaluable disease (yes vs no). Patients are assigned to 1 of 2 treatment groups (group A: no EIAEDs or group B: concurrent use of EIAEDs). Phase I (group B only): Patients receive FR901228 (depsipeptide) IV over 4 hours on days 1, 8, and 15. Courses repeat every 28 days in the absence of disease progression or unacceptable toxicity. Cohorts of 3-6 patients receive escalating doses of FR901228 until the maximum tolerated dose (MTD) is determined. The MTD is defined as the dose preceding that at which 2 of up to 6 patients experience dose-limiting toxicity. Phase II (groups A and B): Group A (phase II): Patients receive FR901228 as in phase I at dose level 1. Group B (phase II): Patients receive FR901228 as in phase I at the MTD.
Inclusion Criteria: * Phase I and phase II: * Histologically confirmed recurrent intracranial malignant glioma, including any of the following: * Glioblastoma multiforme * Gliosarcoma * Anaplastic astrocytoma * Anaplastic oligodendroglioma * Anaplastic mixed oligoastrocytoma * Malignant astrocytoma not otherwise specified * Unequivocal evidence of tumor progression by MRI or CT scan while on a steroid dosage that has been stable for at least 5 days * Patients previously treated with interstitial brachytherapy or stereotactic radiosurgerymust have confirmation of true progressive disease (rather than radiation necrosis) by positron-emission tomography, thallium scan, magnetic resonance spectroscopy, or surgical documentation * Must have failed prior radiotherapy that was completed at least 6 weeks ago * No more than 2 prior therapies (initial treatment and treatment for 1 relapse)\* * Surgical resection for relapsed disease with no anticancer therapy for up to 12 weeks, followed by a second surgical resection, is considered treatment for 1 relapse * Patients in group B must have been receiving enzyme-inducing antiepileptic drugs (EIAEDs) for at least the past 2 weeks * Performance status - Karnofsky 60-100% * More than 8 weeks * WBC ≥ 3,000/mm\^3 * Absolute neutrophil count ≥ 1,500/mm\^3 * Platelet count ≥ 100,000/mm\^3 * Hemoglobin ≥ 10 g/dL (transfusions allowed) * SGOT \< 2 times upper limit of normal (ULN) * Bilirubin \< 2 times ULN * Creatinine \< 1.5 mg/dL * No congestive heart failure (i.e., New York Heart Association class II-IV, ejection fraction \< 40% by MUGA scan or \< 50% by echocardiogram and/or MRI) * No myocardial infarction within the past year * No uncontrolled dysrhythmias * No poorly controlled angina * No significant left ventricular hypertrophy by EKG * No cardiac ischemia (ST depression of 2 mm) by EKG * No hypertrophic or restrictive cardiomyopathy from prior treatment or other causes * No uncontrolled hypertension (i.e., blood pressure ≥ 160/95 mm Hg) * No cardiac arrhythmia requiring antiarrhythmic medication * No known cardiac abnormalities (e.g., congenital long QT syndrome and QTc interval \> 480 milliseconds) * No history of sustained ventricular tachycardia, ventricular fibrillation, Torsade de Pointes, or cardiac arrest unless controlled with concurrent automatic implantable cardioverter defibrillator * No known history of coronary artery disease (e.g., Canadian class II-IV angina) * No other significant cardiac disease * No other malignancy within the past 3 years except nonmelanoma skin cancer or carcinoma in situ of the cervix * No active infection * No significant uncontrolled medical illness that would preclude study participation * No disease that would obscure toxicity or dangerously alter drug metabolism * Not pregnant or nursing * Negative pregnancy test * Fertile patients must use effective barrier contraception during and for at least 2 weeks after study participation * Fertile male patients must continue barrier contraception for 3 months after study participation * At least 1 week since prior interferon or thalidomide * No concurrent prophylactic filgrastim (G-CSF) * No concurrent anticancer immunotherapy * At least 2 weeks since prior vincristine * At least 6 weeks since prior nitrosoureas * At least 3 weeks since prior procarbazine * No prior FR901228 (depsipeptide) * No other concurrent anticancer chemotherapy * See Disease Characteristics * At least 1 week since prior tamoxifen * No concurrent anticancer hormonal therapy * See Disease Characteristics * No concurrent anticancer radiotherapy * See Disease Characteristics * Prior recent resection of recurrent or progressive tumor allowed if patient has recovered * Recovered from all prior therapy * At least 2 weeks since prior EIAEDs (patients in Group A only) * At least 4 weeks since prior cytotoxic therapy * At least 4 weeks since prior investigational agents * At least 1 week since prior isotretinoin * At least 1 week since other prior non-cytotoxic therapy (except radiosensitizers) * No concurrent valproic acid * No concurrent hydrochlorothiazide * No concurrent medication that causes QTc prolongation * No other concurrent anticancer therapy * No other concurrent investigational drugs
are designated in this study