Acta Neurochirurgica

, Volume 149, Issue 3, pp 245–253

Prognosis factors of survival time in patients with glioblastoma multiforme: a multivariate analysis of 340 patients


  • J.-F. Mineo
    • Department of Neurosurgery, Hospital Roger SalengroUniversity Medical Centre
  • A. Bordron
    • Department of Oncology and Hematology, Hospital MorvanUniversity Medical Centre
  • M. Baroncini
    • Department of Neurosurgery, Hospital Roger SalengroUniversity Medical Centre
  • C. Ramirez
    • Department of Neurosurgery, Hospital Roger SalengroUniversity Medical Centre
  • C.-A. Maurage
    • Laboratory of PathoanatomyEurasante, University Medical Centre
  • S. Blond
    • Department of Neurosurgery, Hospital Roger SalengroUniversity Medical Centre
  • P. Dam-Hieu
    • Department of Neurosurgery, Hospital La Cavale BlancheUniversity Medical Centre
Clinical Article

DOI: 10.1007/s00701-006-1092-y

Cite this article as:
Mineo, J., Bordron, A., Baroncini, M. et al. Acta Neurochir (Wien) (2007) 149: 245. doi:10.1007/s00701-006-1092-y


Background. The prognosis of glioblastoma multiforme remains poor despite recent therapeutic advances. Several clinical and therapeutic factors as well as tumour characteristics have been reported as significant to survival. A more efficient determination of the prognostic factors is required to optimize individual therapeutic management. The aim of our study was to evaluate by univariate then multivariate analysis the factors that influence prognosis and particularly survival.

Methods. Data of 340 patients with newly-diagnosed GBM were retrospectively analyzed. Univariate analysis of prognosis factors of survival time was performed. Factors that seemed determinant were evaluated by Kaplan–Meier survival curves. Finally, the significant factors found in univariate analysis were tested in multivariate analysis using the COX regression method.

Findings. Using multivariate analysis, the following factors were found to influence survival: radiotherapy was the predominant factor followed by radical surgery, tumour location, age and chemotherapy. Patients treated with temozolomide had a markedly better survival rate than patients treated with other chemotherapies (Log-rank test P < 0.005). The values of GBM type (de novo or secondary), as well as repeated surgery and partial surgery (vs. simple biopsy) were suggested by univariate analysis but not confirmed by the COX regression method. After radical surgery, progression-free survival was correlated to overall survival (r = 0.87, P < 10e-5).

Conclusions. The influence of radiotherapy on survival was greater than the influence of age, an argument supporting the proposition of radiotherapy for patients until at least age 70. In the case of recurrence, the correlation between overall survival and progression-free survival is an important factor when considering the therapeutic options. Initial radical surgery and repeated procedures dramatically influence survival. The benefit of partial surgery remains difficult to evaluate. Partial surgery could be used to decrease intracranial pressure and to minimize residual tumours in order to enable treatment by chemotherapy and radiotherapy. The value of temozolomide treatment was confirmed.

Keywords: Glioblastoma multiforme (GBM); prognosis; surgery; radiotherapy; temozolomide (TMZ); secondary glioblastoma.

Copyright information

© Springer-Verlag 2007