Phase II trial of gefitinib in recurrent glioblastoma.

PURPOSE: To evaluate the efficacy and tolerability of gefitinib (ZD1839, Iressa; AstraZeneca, Wilmington, DE), a novel epidermal growth factor receptor tyrosine kinase inhibitor, in patients with recurrent glioblastoma. PATIENTS AND METHODS: This was an open-label, single-center phase II trial. Fifty-seven patients with first recurrence of a glioblastoma who were previously treated with surgical resection, radiation, and usually chemotherapy underwent an open biopsy or resection at evaluation for confirmation of tumor recurrence. Each patient initially received 500 mg of gefitinib orally once daily; dose escalation to 750 mg then 1,000 mg, if a patient received enzyme-inducing antiepileptic drugs or dexamethasone, was allowed within each patient. RESULTS: Although no objective tumor responses were seen among the 53 assessable patients, only 21% of patients (11 of 53 patients) had measurable disease at treatment initiation. Seventeen percent of patients (nine of 53 patients) underwent at least six 4-week cycles, and the 6-month event-free survival (EFS) was 13% (seven of 53 patients). The median EFS time was 8.1 weeks, and the median overall survival (OS) time from treatment initiation was 39.4 weeks. Adverse events were generally mild (grade 1 or 2) and consisted mainly of skin reactions and diarrhea. Drug-related toxicities were more frequent at higher doses. Withdrawal caused by drug-related adverse events occurred in 6% of patients (three of 53 patients). Although the presence of diarrhea positively predicted favorable OS from treatment initiation, epidermal growth factor receptor expression did not correlate with either EFS or OS. CONCLUSION: Gefitinib is well tolerated and has activity in patients with recurrent glioblastoma. Further study of this agent at higher doses is warranted.

Production and characterization of two ependymoma xenografts.

Childhood ependymomas exhibit epidemiologic, anatomic, histologic, and biologic features and distinguish them from other gliomas. Because of their propensity to grow in functionally sensitive regions of the brain, adequate tumor sampling for basic and therapeutic research is limited. We have established xenografts in both subcutaneous and intracranial nude mouse systems (D528 EP-X, D612 EP-X) from the ependymomas of two nonrelated children. Median subcutaneous growth rates (reported in days to grow from 200 mm3 to 1000 mm3) are 82 days for D528 EP-X (n = 10) and 50 days for D612 EP-X (n = 10). D528 EP-X grows intracranially with a median postimplantation survival of 85 days (n = 10); D612 EP produces a median postimplantation survival of 72.5 days (n = 10). Both xenografts grow as well-formed masses with no evidence of infiltration into either brain or subcutaneous tissues. While perivascular pseudopalisading is found in both xenografts, true ependymal rosette formation is absent. Ultrastructurally, neither xenograft exhibits cilia, but both produce abundant intermediate filaments. By light microscopy, the neoplastic cells are immunoreactive for the intermediate filaments glial fibrillary acidic protein, vimentin, and nestin. Karyotypically D528 EP exhibits 46,XX,del(6)(q22q26)/46,XX while D612 EP exhibits 50,XX, +X,t(1;8)(p11;q11),t(1;8)(p11;q11), +1,-4, der(5)t(4;5)(q12;q35), +der(5)t(4;5)(q12;q35),-6, +9, +9,-16, +der(17)t(6;17)(p11;p11), +mar. Restriction fragment length polymorphism studies comparing the primary brain tumor tissue from each patient against multiple passages of the resulting xenografts confirm the origin of both xenografts. These xenografts represent models on which future studies into the oncogenesis, progression and therapy of ependymomas can be performed.