358P - Retrospective analysis to ascertain whether thromboembolic events, patient gender and tumour size have prognostic implications for glioblastoma mul...

Date 10 September 2017
Event ESMO 2017 Congress
Session Poster display session
Topics Central Nervous System Malignancies
Presenter Dominique Parslow
Citation Annals of Oncology (2017) 28 (suppl_5): v109-v121. 10.1093/annonc/mdx366
Authors D. Parslow, S. Pascoe
  • Oncology Department, Derriford Hospital Plymouth Hospitals NHS Trust, PL6 6DH - Plymouth/GB



Glioblastoma multiforme is a rare grade 4 incurable brain malignancy. Well established prognostic indicators for this disease include performance status, age and cognitive function at diagnosis. Presenting with a seizure is also known to predict a better prognosis. Patient gender, tumour size and thromboembolic events have not been previously known to have prognostic significance. The rationale of this study is to identify alternative features which could be used as additional prognostic indicators.


We conducted a retrospective analysis of all patients diagnosed with glioblastoma multiforme at Derriford Hospital, Plymouth, UK between 2009 and 2016. We analysed factors such as survival time since diagnosis, patient demographics, tumour size at diagnosis, performance status at diagnosis, presenting symptom, treatment undergone and the occurrence of venous thromboembolic events since diagnosis.


92 patients were included. The occurrence of venous thromboembolism had no impact on survival time (p = 0.386). Male sex appeared to predict a better prognosis than female sex, however, this did not quite achieve statistical significance with a p value of 0.09. Cox regression analysis revealed tumour size on diagnosis to be significantly negatively correlated with survival time, with a p value of 0.012. Our analysis agreed with previous findings that multifocal disease and increased age are poor prognostic indicators, and presenting with seizures is a good prognostic indicator. Patients who underwent radical debulking surgery followed by concomitant chemoradiation had a significantly longer survival time than patients who had best supportive care alone.


Our analysis has shown that increasing tumour size is negatively correlated with survival duration. This link has not been previously established. Female sex may also be a poor prognostic indicator, but our data did not achieve statistical significance so further research investigating this potential link may be warranted. Venous thromboembolic events had no impact on prognosis.

Clinical trial identification

Legal entity responsible for the study

Research Office, Plymouth Hospitals NHS Trust




All authors have declared no conflicts of interest.