Journal of Clinical Oncology · 2009 · 10 citations · 0 references
MedicineOncologyPharmacotherapyMetastasisNeuro-oncologySolid Tumor PharmacotherapyBrain MetastasesBrain MetastasisBrain TumorsCentral Nervous System Tumor PharmacotherapyMetastatic Brain TumorsNeuro-oncology Clinical TrialsBrain Metastasis TrialsCentral Nervous System MetastasisDrug Safety InformationHemorrhagic Brain MetastasesHepatocellular Carcinoma TreatmentOpen-label TrialsPhase Iii Oncology Trials
2007 Background:. The background rate of CNS hemorrhage (CH) in cancer patients with brain metastases not receiving bevacizumab (BV) treatment is 5%-29%, depending on tumor type. Patients with CNS metastases were routinely excluded from most BV late-stage clinical trials, following the occurrence of CH in one hepatocellular carcinoma patient with occult CNS metastasis in a BV phase I trial. In this particular tumour type, up to 87.5% has been reported as a background rate of CH without BV use. Available safety information in BV-treated patients with CNS metastases has been reviewed to determine whether the exclusion of these patients from trials is still justified. Methods: A retrospective exploratory analysis was conducted using safety data from three datasets: A.) 13 randomized controlled phase II or III trials (RCTs); B.) Two ongoing open-label trials MO19391 (ATHENA) and MO19390 (SAIL). While patients with known CNS metastases, based on imaging or clinical signs and symptoms, were excluded from trial in A and B datasets, patients who were found to have CNS lesions had either unrecognized CNS metastases at study entry or had developed these during the trial.; C.) 2 ongoing open-label studies AVF3752g and AVF3671g in patients with NSCLC, where inclusion of patients with treated CNS metastases was allowed. Incidence of CH in patients with brain metastases was quantified in each dataset. Results: A.) In 13 RCTs, with a total of 8443 patients with locally advanced, unresectable, or metastatic breast, renal, pancreatic, colorectal cancer, or NSCLC, brain metastases were identified in 187 patients (91 in BV-arms and 96 in control arm). Three of 91 (3.29%) BV-treated patients developed grade 4 CH, and one in 96 patients in the control arm (1.04%) developed grade 5 CH; B.) In open-label studies no CH was reported in 68 patients with CNS metastases, out of 3,252 patients enrolled. C.) In two studies in patients with treated CNS metastases, one subject in 83 (1.2%) BV treated patients developed a grade 2 CH. Conclusions: In this retrospective review, the rates of CH in BV-treated patients with CNS metastases is low, and appears consistent with historical rates of CH in these patient populations. Ongoing trials are expected to provide additional data regarding the risk of CH in patients with primary and metastatic CNS tumors. [Table: see text]