Prognostic factors in glioblastoma multiforme. 10 years experience of a single institution.
BACKGROUND: To analyze prognostic factors in patients with a glioblastoma multiforme treated in an academic institute over the last 10 years. PATIENTS AND METHOD: From 1988 to 1998, 198 patients with pathologically confirmed glioblastoma multiforme were analyzed. Five radiation schedules were used mainly based on pretreatment selection criteria: 1. 60 Gy in 30 fractions followed by an interstitial iridium-192 (Ir-192) boost for selected patients with a good performance and a small circumscribed tumor, 2. 66 Gy in 33 fractions for good performance patients, 3. 40 Gy in eight fractions or 4. 28 Gy in four fractions for poor prognostic patients and 5. no irradiation. RESULTS: Median survival was 16 months, 7 months, 5.6 months, 6.6 months and 1.8 months for the groups treated with Ir-192, 66 Gy, 40 Gy, 28 Gy and the group without treatment, respectively. No significant improvement in survival was encountered over the last 10 years. At multivariate analysis patients treated with a hypofractionated scheme showed a similar survival probability and duration of palliative effect compared to the conventionally fractionated group. The poor prognostic groups receiving radiotherapy had a highly significant better survival compared to the no-treatment group. Patients treated with an Ir-192 boost had a better median survival compared to a historical group matched on selection criteria but without boost treatment (16 vs 9.7 months, n.s.). However, survival at 2 years was similar. Analysis on pretreatment characteristics at multivariate analysis revealed age, neurological performance, addition of radiotherapy, total resection, tumor size post surgery and deterioration before start of radiotherapy (borderline) as significant prognostic factors for survival. CONCLUSION: Despite technical developments in surgery and radiotherapy over the last 10 years, survival of patients with a glioblastoma multiforme has not improved in our institution. The analysis of prognostic factors corresponded well with data from the literature. A short hypofractionated scheme seems to be a more appropriate treatment for patients with intermediate or poor prognosis as compared to a conventional scheme. The benefit in median survival for patients treated with an interstitial boost is partly explained by patient selection. Since there were no long-term survivors with this boost treatment, its clinical value, if there is one, is still limited.