Animal and pharmacokinetic studies

Tamargo and colleagues published the first paper using BCNU-embedded wafers in 1993. Rats with intracranial 9L gliosarcomas implanted with the wafers had 2- to 3-fold longer survivals than those treated with intraperitoneal BCNU injections. They tested 2 different polymer bases using this model, ultimately choosing polifeprosan 20 (copolymer of 1,3-bis( p -carboxyphenoxy)propane and sebacic acid in a 20:80 M ratio) because it better protects BCNU from degradation before it is released.

The initial pharmacokinetic studies on BCNU wafers were performed in rabbits, using radiolabeled BCNU. Grossman and colleagues demonstrated that BCNU was distributed widely in the brain ipsilateral to the implants, at distances up to 12 mm from the wafers. Follow-up studies in monkeys measured the drug concentration at the site of implantation on day 1 at 3.5 mmol/L, decreasing to approximately 1.0 mmol/L at a distance of 3 mm from implantation. Seven days postimplant, only the area within 0.5 mm had a concentration greater than 1 mmol/L. The inhibitory concentration of BCNU on human glioma lines in vitro has been reported as from 15 to 300 μmol/L, implying that the area adjacent to the implant receives well in excess of the therapeutic drug levels for at least 7 days. Further studies in monkeys demonstrated the safety of BCNU wafers with radiation therapy, and allowed for human trials to begin.

Human trials with single-agent BCNU wafers

The initial trials of single-agent BCNU wafers are summarized in Table 1 .

A phase I to II trial in patients with recurrent glioma identified the 3.85% BCNU wafer as the most effective, based on survival from the time of implantation. The higher-dose group (6.35%) did not have more side effects, but overall survival was lower, possibly because it contained 100% glioblastoma (GB) patients. In the 2 lower-dose cohorts only 60% of patients had GB. The 3.85% wafer was chosen for use in subsequent trials, although there was no apparent difference in toxicity. The phase III trial enrolled 222 patients with recurrent glioma, randomizing them to active or placebo wafers. This trial demonstrated an increased overall survival of 8 weeks ( P = .061) and led the Food and Drug Administration (FDA) to approve Gliadel (3.85% BCNU wafer) for use in recurrent malignant glioma in 1996.

A phase I safety study of the BCNU wafer in combination with radiation therapy for newly diagnosed glioma was run concurrently with the phase III trial for recurrence. The study demonstrated an increased median survival compared with historical controls, but also showed a higher rate of severe adverse events compared with earlier trials, which included seizures, intracranial hypertension, and neurologic decline in the postoperative period. The phase III trial for newly diagnosed glioma had a lower complication rate than observed in the Phase I trial, but remained concerning. Five percent of patients in the BCNU group had cerebrospinal fluid leaks and 9.1% of patients had intracranial hypertension. Although there was a statistically significant improvement in survival in both the recurrent and up-front trials, it is important to note that patients in the placebo arms did not receive treatment after radiation therapy until they had recurrent tumor Standard practice at the time would likely have followed radiation with chemotherapy. This makes the modest results somewhat difficult to interpret (see Table 1 ).

Human trials with single-agent BCNU wafers

The initial trials of single-agent BCNU wafers are summarized in Table 1 .

A phase I to II trial in patients with recurrent glioma identified the 3.85% BCNU wafer as the most effective, based on survival from the time of implantation. The higher-dose group (6.35%) did not have more side effects, but overall survival was lower, possibly because it contained 100% glioblastoma (GB) patients. In the 2 lower-dose cohorts only 60% of patients had GB. The 3.85% wafer was chosen for use in subsequent trials, although there was no apparent difference in toxicity. The phase III trial enrolled 222 patients with recurrent glioma, randomizing them to active or placebo wafers. This trial demonstrated an increased overall survival of 8 weeks ( P = .061) and led the Food and Drug Administration (FDA) to approve Gliadel (3.85% BCNU wafer) for use in recurrent malignant glioma in 1996.

A phase I safety study of the BCNU wafer in combination with radiation therapy for newly diagnosed glioma was run concurrently with the phase III trial for recurrence. The study demonstrated an increased median survival compared with historical controls, but also showed a higher rate of severe adverse events compared with earlier trials, which included seizures, intracranial hypertension, and neurologic decline in the postoperative period. The phase III trial for newly diagnosed glioma had a lower complication rate than observed in the Phase I trial, but remained concerning. Five percent of patients in the BCNU group had cerebrospinal fluid leaks and 9.1% of patients had intracranial hypertension. Although there was a statistically significant improvement in survival in both the recurrent and up-front trials, it is important to note that patients in the placebo arms did not receive treatment after radiation therapy until they had recurrent tumor Standard practice at the time would likely have followed radiation with chemotherapy. This makes the modest results somewhat difficult to interpret (see Table 1 ).