Phase I therapy trials in children with cancer.
PURPOSE: This study examined the response and toxicity rates of antineoplastic drugs evaluated in phase I clinical trials in children to identify trends in response and toxicity over time. PATIENTS AND METHODS: Full length, peer-reviewed articles describing the results of single agent phase I therapy trials in children younger than 21 years with cancer were reviewed. Tumor-specific response data and doses of drugs that resulted in objective responses were noted. Deaths that occurred on study caused by drug toxicity, progressive disease (PD), or complications of marrow aplasia were identified, along with drug doses that resulted in toxic death. Temporal trends in response rates, toxicity, and number of patients entered in trials were examined. RESULTS: A total of 1,606 patients with cancer were enrolled in 56 single-agent pediatric phase I therapy trials published between 1978 and 1996. Of these, 1,257 were evaluated for response by tumor type. The overall objective response rate was 7.9%. Response rates were highest for patients with neuroblastoma (17.7%) and acute myelogenous leukemia (11.6%). Patients with osteosarcoma and rhabdomyosarcoma had response rates of < 3%. Sixty percent of responses in patients with solid tumors occurred at 81 to 100% of the maximum tolerated dose (MTD), although 42% of responses in patients with leukemia occurred at > 100% of the MTD. Death on study was noted in 7.0% of all patients entered in trials. Only 0.7% of patients experienced a death related to drug toxicity. PD accounted for the death of 5.6% of study participants. A trend of increasing response rate despite smaller trial size was noted over the last 7 years of this period. CONCLUSION: Phase I trials in children with cancer represent a safe mechanism to determine the MTD, toxicity profile, and pharmacokinetics of new agents for use in children with cancer.