Evidence That Recurrence Patterns Of Ttfields Treated Patients Affect Patient Outcome: Post-Hoc Analysis Of The Randomized Phase 3 Ef-14 Trial

A recent post-hoc analysis of the EF-14 phase III trial in newly diagnosed glioblastoma patients showed that the patterns of recurrence were different in patients treated with TTFields+Temozolimde vs. patients treated with Temozolomide alone. There was greater local control in the TTFields-treated arm, as areas of progression were located further from the primary lesion than in the control arm. Additionally, infratentorial progression was only observed in the TTFields-treated arm. These results led us to examine the hypothesis that patterns of recurrence may affect patient outcome (OS/PFS). Participants in the EF-14 trial who exhibited radiological progression were included in this study (treatment: N = 306/466, control: N = 122/229). Enhancing tumor was segmented on T1c MRIs at baseline and at progression. Areas disconnected from the original lesion were defined as distant progression. Kaplan-Meier curves were created both for the treatment arm and the control arm to evaluate the relationship between progression location vs PFS and OS. Additionally, hazard ratios were calculated using Cox proportional hazards models accounting for prognostic factors (sex, age, KPS, region, MGMT methylation status). Patients that exhibited distant progression in the TTFields group had a significantly longer PFS than patients in the TTFields treated group who exhibited local progression (distant: 7.9 month vs. local: 5.2 month, HR = 0.78, p = 0.012). Patients in the control group did not show any difference in their PFS (distant: 3.8 month vs. local: 3.7 month, HR = 0.78, p = 0.26). For OS there was a trend towards a higher OS in TTFields treated patients who progressed distantly progression compared to those who progressed locally (distant: 23.9 month vs. local: 18.8 month, HR = 0.81, p = 0.085). HRs calculated using Cox models accounting for prognostic factors confirmed that within the TTFields-treated arm the site of progression was independently associated with a prolonged PFS. Distal progression was associated with extended PFS in the TTFields treated arm, with no such association observed in the control arm. This observation strengthens the hypothesis that TTFields is altering the natural history of GBM through local control of the tumor.