Recursive Partitioning Analysis. Historical benchmark for median survival in GBM cohorts.
GBM Survivorship Probe
Determine the RPA Prognostic Class and estimated median survival for patients with Glioblastoma (GBM) based on age and functional status.
Guidelines & Evidence
Clinical Details
Section 1
When to Use
When to Use
Counseling patients regarding survival expectations following a glioblastoma diagnosis.
Deciding between aggressive trimodality therapy (Surgery + RT + TMZ) versus Best Supportive Care/Biopsy alone.
Section 2
Literature
Development
The RTOG (Radiation Therapy Oncology Group) analyzed three randomized trials spanning huge cohorts of high-grade glioma patients. Using statistical "recursive partitioning," they allowed an algorithm to split the patients by the most defining survival characteristics. The result was indisputable: Age <50 and KPS >70 trumped almost all other variables in prolonging survival.
Section 3
Pearls/Pitfalls
Class V and VI Reality
Patients in Class VI (Age ≥50, KPS <70) have a median survival of less than 5 months. Operating aggressively on these patients to achieve a Gross Total Resection often paradoxically hastens their death by introducing morbidity they cannot tolerate. Biopsy and short-course palliative radiation is more appropriate.
Section 4
Evidence Appraisal
Primary Reference
Recursive partitioning analysis of prognostic factors in three Radiation Therapy Oncology Group malignant glioma trials
Curran WJ Jr et al. • J Natl Cancer Inst. 1993;85(9):704-10
