Survival in patients with radiological diagnoses of glioblastoma: a retrospective study of 115 patients on a best supportive care pathway

Aims Glioblastoma multiforme (GBM) is a devastating disease with notoriously poor survival. Studies examining survival in patients given best supportive care (BSC) are few and far between. All patients harbouring brain tumours referred to the Neuro-oncology service at the Queen Elizabeth Hospital in Birmingham are recorded in the Somerset Cancer Registry. We set out to analyse survival times and identify patient and tumour-related factors significantly affecting prognosis. Method We identified 126 patients from 2015 to 2019 in our Somerset Cancer Registry with radiological diagnoses of glioblastoma for whom the Neuro-oncology MDT recommended BSC. We performed a retrospective analysis of clinical records and radiological images. 11 patients were excluded (8 due to insufficient imaging data, 2 who underwent subsequent surgery, 1 patient with brain metastases). Survival was measured in completed weeks since the index MDT decision. Associations between survival time and both patient- and tumour-related factors were assessed using Kaplan-Meier curves and log-rank tests. All analyses were performed using IBM SPSS 22 (IBM Corp. Armonk, NY), with p<0.05 deemed to be indicative of statistical significance throughout. Results Data were available for N=115 patients (69 males, 46 females), with a mean age of 79 ± 8 years. All patients died within 32 weeks of diagnosis, with a median survival time of 8 weeks. Only 8 patients survived for more than 20 weeks. Survival was significantly shorter in those with a greater number of main cerebral structures affected (p=0.044), with a median of 6 vs. 10 weeks for 3 or more vs. 1 structures affected (hazard ratio: 1.61, 95% CI: 0.99-2.62). Bilateral tumours involving the corpus callosum were also associated with shorter survival (p=0.039). None of the other factors considered were found to be significantly associated with survival, including age (p=0.193), gender (p=0.371), performance status (p=0.300) and tumour size (p=0.331). Conclusion With the exception of the number of main cerebral structures affected (frontal, parietal, temporal and occipital lobes, corpus callosum, insula, basal ganglia and brain stem), patient- and tumour-factors traditionally used by the MDT to prognosticate do not correlate with survival time in patients receiving BSC for radiological diagnoses of GBM. With 50% of the cohort dying within 8 weeks it is clear that we must reconsider the timing of referrals to palliative and hospice care. Finally, the fact that some patients survived for more than half a year with no surgical or oncological treatment suggests that the process of selecting patients for BSC vs aggressive treatments needs refinement.