Brain tumor clinical trials: pitfalls and promise for the future.

Abstract

Despite intensive effort in laboratory and clinical investigations, the prognosis for patients with malignant gliomas remains grim. Currently, patients with glioblastoma multiforme (GBM) have a median survival of less than 1 year, similar to that reported in 1980 [1]. The prognosis for patients with lower-grade tumors, such as anaplastic astrocytoma (AA), has shown some improvement with combined modality treatment, although the improvement has not been dramatic. One of the most significant advances in clinical research in brain tumor treatment has been the development of a better understanding of difficulties in interpreting clinical trials. This has led to a critical reappraisal of the standard of care of patients with malignant gliomas. The recent publication of a retrospective analysis of a large clinical database of patients with anaplastic astrocytoma by Prados et al. [2] underscores these concerns about the established treatment standards. These investigators reviewed the Radiation Therapy Oncology Group (RTOG) database and analyzed patients with newly diagnosed AA who were treated in a protocol that included either procarbazine, CCNU, and vincristine (PCV; n=175) or BCNU (n=257) as adjuvant therapy. Extensive statistical analysis was performed to balance known prognostic factors (age, extent of initial tumor resection, and Karnofsky performance status). Cox proportional hazards modeling and analysis using matched cases between the groups failed to show any statistically significant improvement in survival between the two adjuvant treatments. Prior to this analysis, adjuvant PCV was considered the standard of care for patients with newly diagnosed AA. This treatment recommendation was based on the publication of a subset analysis of 73 patients with anaplastic glioma ([AG] anaplastic astrocytoma, anaplastic oligodendroglioma, and anaplastic mixed glioma) extracted from a large randomized phase III trial comparing PCV with BCNU in patients with newly diagnosed malignant gliomas (GBM and AG) [3]. Although PCV has been considered the standard treatment for anaplastic glioma for many years, the analysis by Prados et al [2] has called the role of PCV into question. Criticisms of this earlier study include the small number of patients evaluated, the post-hoc nature of the analysis, and the inclusion of patients with anaplastic oligodendrogliomas, who have a better prognosis and response to PCV [4]. These problems highlight some of the difficulties in establishing a treatment standard. Recent critical and insightful analyses by experts in the field have elucidated many other issues in brain tumor clinical trials that must be addressed as new treatments are tested and new standards are developed [5‐8].