427P - Prognostic Factors for Post-Progression Survival in Patients Receiving Radical Treatment for Glioblastoma Multiforme

Abstract

ABSTRACT Aim: Glioblastoma multiforme (GBM) is the most common and most aggressive primary brain tumour in adults. Current optimal treatment is maximal surgical resection, followed by chemo-radiotherapy and adjuvant chemotherapy using temozolomide. Even with this regimen, median overall survival (OS) is 14.6 months. Multiple studies have confirmed the prognostic significance of age, performance status, extent of surgical resection and O6-methylguanin-DNA-methyltransferase (MGMT) status at diagnosis. However, there is relatively little data on the prognostic factors influencing survival after first progression. Methods: We performed a retrospective case review of hospital surgery, radiology, oncology and radiotherapy records to identify patients who had received radical chemoradiotherapy for a histologically confirmed GBM between 2007 and 2013. We collected data on patient demographics, extent of surgical resection (biopsy, sub-total resection, gross-total resection), subsequent treatment and survival. The date of diagnosis was taken as the date of surgery/ biopsy. Date of progression and date of death were obtained from radiology reports and clinic letters. We conducted exploratory analyses to investigate possible prognostic factors for post-progression survival. Results: We identified 196 patients. For this study, the analytical cohort consisted of 115 patients for whom we had data on progression-free, post-progression and overall survival All patients were ECOG PS 0 or 1 at time of treatment. The median age was 59. 45 underwent biopsy, and 8 gross total resection at time of initial diagnosis. The median PFS and OS were 5.2 and 14.1 months respectively. On progression, 52 patients had subsequent treatment, usually with temozolomide, temozolomide & bevacizumab or procarbazine, lomustine (CCNU) and vincristine (PCV) chemotherapy. The median post-progression survival (PPS) was 4.7 months. On univariate analysis, PPS was influenced by age at diagnosis (4 vs. 7.2 months), initial extent of surgery (4.5 vs. 7.2 months) and MGMT status (4.3 vs. 7.4 months), but not the length of PFS, although none of these factors was statistically significant. Conclusions: Post-progression survival is poor, and is not influenced by progression-free survival. However, there was a trend for the known prognostic factors at baseline to be of prognostic benefit in predicting post-progression survival. Disclosure: All authors have declared no conflicts of interest.