P14.125 Retrospective analysis of long-term response to bevacizumab in combination with irinotecan for recurrent glioblastoma: identification of prognostic factors. (6th September 2019)

• Record Type: Journal Article
• Authors: Besson, C; Morisse, M; Brut, H; Waissi, W; Noel, G; Chauffert, B; Prébay, D; Schott, R; Etienne-Selloum, N
• Is Part Of: Neuro-oncology. Volume 21(2019)Supplement 3
• Page Start: iii98
• Page End: iii98
• Publication Date: 2019-09-06
• Subjects: Brain Neoplasms -- Periodicals; Brain -- Tumors -- Periodicals; Brain -- Cancer -- Periodicals; Nervous system -- Cancer -- Periodicals; 616.99481
• DOI: 10.1093/neuonc/noz126.360
• Languages: English
• ISSNs: 1522-8517

Abstract

Abstract: BACKGROUND: In absence of standard treatment for recurrent glioblastoma (rGBM), numerous prospective and retrospective studies have evaluated the off-label combination of bevacizumab (BEV) with irinotecan (IRI) in patients with rGBM. We report here our single center experience with this combination and we investigated prognostic factors for long-term response. MATERIAL AND METHODS: We performed a retrospective analysis of consecutive patients treated initially by Stupp protocol and with BEV-IRI for a rGBM between 2007 and 2017. Times to progression and overall survival, as well as toxicities, were investigated and analysed. Patients without progression at least 12 month after the first administration of BEV-IRI were considered as long-term responders. The primary end-point was overall survival post-BEV-IRI (OS-BEV-IRI). RESULTS: One-hundred eleven patients were eligible for the analysis. Median age at the diagnosis was 57 years and the value of WHO Performance Status (PS) at the recurrence was 0 to 1 for 67, 5% of patients. Kaplan-Meier median progression-free survival (PFS-BEV-IRI) and overall survival (OS-BEV-IRI) at recurrence estimates (calculated from start of BEV-IRI) were 6.51 and 10.41 months, respectively. The median OS (calculated from diagnosis) was 22, 4 months. Twenty-Three patients (20, 7%) were long-term responders to BEV-IRI regimen. This subgroup was not significantly different than the short-term responders according to age or PS distribution, but the relative proportion of biopsy in comparison to other surgery modalities was significantly increased in long-term responders (p<0, 0001). Univariate analysis showed that PS 0–1 (p=0, 007), biopsy (p=0, 0022) are significantly associated with a better prognosis, but not age. Eighty three patients (75%) had toxicities, mainly grade 1 and 2 (92%), such as hypertension, proteinuria, haemorrhage, thrombosis, nausea, diarrhoea, fatigue or neutropenia. Most of the grade 3 and grade 4 toxicities were related to BEV treatment. Adverse events were significantly more frequent in long-term responders (p=0, 0096). CONCLUSION: BEV-IRI Combination is well tolerated and may offer some clinical benefits in recurrent GBM patients, more particularly if only biopsy was performed instead of surgery. Our results strengthened the role of these agents for the treatment of recurrent GBM.