Intracranial hemorrhage in glioblastoma: Incidence, risk factors, and outcome.

Abstract

2074 Background: The incidence and clinical implications of intracranial hemorrhage (ICH) in patients with glioblastoma have not been well studied. Risk factors for ICH and associations with therapy and outcome remain unknown. It has been speculated that prior radiotherapy, chemotherapy-associated thrombocytopenia, steroid medications as well as age, other comedications and comorbidities contribute to the risk of ICH in this patient population. Methods: We identified consecutive patients with glioblastoma per WHO 2021 definition from the electronic charts of the University Hospital in Zurich, Switzerland and explored the incidence, putative risk factors, management and outcome of ICH in patients with glioblastoma. ICH was identified on central MRI and imaging report review. The study has been approved by Swissethics (2023-00532). Results: Among 387 patients diagnosed between 2010 and 2019, 6 patients presented with ICH, and 22 patients experienced ICH in the further disease course, after a median time of 153 days (IQR 79-323 days). Of the 6 patients with ICH at diagnosis, 2 patients were on anti-platelet agents and 1 was on anticoagulation. ICH occurred in the tumor region in all these patients. There were no specific features in their history compared with patients that did not present with ICH at diagnosis. Among 22 patients with ICH after diagnosis, 9 patients (41%) experienced ICH prior to first progression. ICH were not restricted to the tumor region, but included subdural hematomas in 7 patients and an epidural hematoma in one patient. Fifteen patients (68%) were hospitalized for ICH, and 5 patients (23%) had a craniotomy. Thirteen patients (59%) had no change in management, 4 patients (18%) started a new intervention, and 5 patients (23%) went to palliative care. Compared with patients without any ICH ever documented, the 28 patients with ICH were more often male, but not older, did not have a history of arterial hypertension, diabetes, chronic kidney disease or smoking, and did not have larger tumors at diagnosis. At the timepoint of intracranial hemorrhage, six patients (27%) had been preexposed to bevacizumab, yet, 188 patients (53%) of the reference cohort received bevacizumab across the disease trajectory without experiencing an ICH. Three patients (14%) had platelet counts below 100,000/µl. Overall survival in the 28 patients who experienced an ICH was not inferior to survival in the population of patients without an ICH. Conclusions: ICH at diagnosis is rare in glioblastoma and its etiology remains to be identified. It may be related to intrinsic features of the tumor and its blood vessels since no other patient-related risk factors were identified. Overall, male sex was the only risk factor associated with ICH in this study, but ICH does not signify a poor outcome overall.