Objective: Recently a randomized placebo-controlled phase III trial of biodegradable polymers containing carmustine has demonstrated a significant survival benefit for patients treated with local chemotherapy. A local chemotherapy applied directly to the resection cavity may act directly on residual tumor cells in adjacent brain possibly leading to a local control of the tumor and increased survival.
Methods: We have analyzed the pattern of recurrence using serial MRI studies of 24 patients treated with GLIADEL Wafers or placebo wafers following resection of glioblastomas.
Results: Of 24 patients 11 received carmustine wafers and 13 placebo. The age distribution and Karnowsky performance scores of the two populations were not different. However, the median survival (14.7 versus 9.5 months; P = 0.007) and the time to neurological deterioration (12.9 +/- 4.85 vs. 9.4 +/- 2.73 months; P = 0.035) was significantly longer in the treatment group versus the placebo treated control. Preoperative and follow up MRI studies were evaluated in a blinded fashion. Out of 24 patients that entered the analysis 11 showed clearance of all contrast enhancement following resection of glioblastomas. Seventeen tumors progressed locally and 7 showed different patterns of distant failure. Within the carmustine treated group 8 patients showed a local treatment failure with recurrent tumors immediately adjacent to the resection cavity or progression form a residual tumor. Three patients showed a multifocal distant and local pattern of failure after complete or subtotal removal. In no case the local chemotherapy resulted in a distant recurrence only. However, the time to radiographic progression was 165.1 +/- 80.75 days for the GLIADEL Wafer group and 101.9 +/- 43.06 days for the placebo group (P = 0.023).
Conclusion: In this subgroup analysis of a phase III trial population both the clinical progression and radiological progression were significantly delayed in patients treated with local chemotherapy, resulting in an increased survival time. Local chemotherapy with carmustine containing wafer implants did not result in an altered pattern of recurrence and did not promote multifocal patterns of recurrence.