gms | German Medical Science

Objective: Glioblastoma (GBM), the most frequent primary brain tumor in adults, has an exceptionally poor prognosis. Biologically, GBM critically depends on angiogenesis, which is primarily modulated by VEGF. Although several phase II studies have demonstrated efficacy of anti-VEGF strategies in recurrent GBM (rGBM) with regard to progression free survival, a recent phase III study failed to show beneficial effects of bevacizumab (BEV) on overall survival (OS). However, none of these studies have reflected the clinical experience that BEV can provide symptom control and improvement of focal neurological deficits. We therefore performed a retrospective study in patients treated for rGBM either with or without BEV focussing on functional status and neurological deficits.

Methods: Group A (n=124) did not receive BEV despite physician’s indication due to a negative decision of the health insurance; Group B (n=83) was treated with BEV after receiving a positive vote. The groups were balanced for gender, age, pretherapeutic Karnofsky Score (KPS), neurological performance score (NPS), salvage treatment strategy and IDH1 mutation status. However group B included significantly more patients with unmethylated MGMT promoter. The most frequent systemic treatment in group A was re-use of temozolomide (60.8%) followed by procabacine/CCNU (25.8%). In group B, BEV was most frequently given in combination with CCNU (43.4%). 54.2% of patients received BEV at first, 45.8% at second recurrence.

Results: Overall survival (OS) was not different between groups (A: 17.7 vs. B: 20.8 months; p=0.08). Progression free survival (PFS) after first recurrence was significantly longer in the BEV group (A: 2.5 vs. B: 6.2 months, p=0.01). Improvement rates of several clinically important dimensions were significantly higher in the BEV treated group compared to the non-BEV group: KPS (A: 4.1% vs. B: 21.6%, p=0.0001); NPS (A: 1.2% vs. B: 25.4%, p=0.0001); hemiparesis (A: 0% vs. B: 12.9%, p=0.04); aphasia (A: 7.8% vs. B: 40.6%, p=0.003); visual field defects (A: 11.1% vs. B: 40.0%, p=0.04) and cranial nerve dysfunction (A: 35% vs. B: 72.7%, p=0.01).

Conclusion: Although BEV did not prolong overall survival, it profoundly contributed to a prolonged PFS, reduced symptom burden, and improved focal neurological deficits.