EF-19, a post-approval registry study of tumor treating fields (TTFields) in recurrent glioblastoma (rGBM).

Abstract

e14536 Background: Tumor Treating Fields (TTFields) are an anti-mitotic non-invasive therapy of low intensity alternating electric fields delivered to the tumor via a portable medical device. In phase 3 studies that led to FDA approvals, TTFields plus maintenance temozolomide significantly extended survival in newly diagnosed GBM, and achieved comparable survival outcome to best standard of care (BSC) as monotherapy in recurrent GBM (rGBM). The EF-19 study aimed to confirm efficacy of TTFields vs BSC in rGBM in post-approval real-life setting. Methods: This non-inferiority, prospective, non-randomized, post approval registry trial enrolled 192 rGBM patients (>21 yrs, KPS > 70). Patients were treated with TTFields (200 kHz, > 18h/day). Eligibility criteria: histologic GBM, past treatment per Stupp protocol for primary disease and radiological evidence of progression in the supratentorial region. Primary endpoint was overall survival (OS); secondary endpoints were OS in the per protocol (PP) population ( > 1 course of TTFields or BSC in each respective arm), time to treatment failure and adverse events. The TTFields patient registry data were compared to OS of all 117 rGBM patients in the BSC group of the EF-11 trial (Stupp R, EJC 2012). The sample size of 192 patients (10% loss to follow up) was based on non-inferiority log-rank test with a two-sided alpha level of 0.05 and a power of 80%, comparing time to event (i.e., death) between patients treated with TTFields and BSC. The analysis was based on true hazard ratio (HR) of 1.0 comparing TTFields to control with an upper one-sided 95% confidence bound of HR not exceeding 1.375. Results: Median OS with TTFields versus EF-11 BSC was 7.4 vs. 6.4 months, p = 0.053; HR = 0.64 (95%CI 0.46-0.91, Cox-test P = 0.012). Median OS (PP population) with TTFields versus EF-11 BSC was 8.1 months vs. 6.5 months; p = 0.045; HR 0.65. The results showed a significant superiority in HR of OS between the 2 groups, as the 95% confident interval upper limit of the HR was lower than the pre-defined threshold for non-inferiority for interval bound of 1.375. The overall incidence of adverse event was lower with TTFields than EF-11 BSC (67% vs. 95%). The median time to treatment failure was longer in the TTFields arm (3.3 months (95% CI 2.6, 3.9) versus BSC arm (1.6 months; 95% CI 1.1, 1.9); HR = 0.53 (95% CI 0.41, 0.68, p < 0.0001). Skin AE was the most frequently reported AEs in TTFields arm; no unexpected adverse events were reported with TTFields. Conclusions: The results of the EF-19 registry study confirm the effectiveness and safety of TTFields monotherapy in rGBM.