P14.66 TTFields dose distribution and tumor growth patterns confirm clinical activity of TTFields: MRI analysis of the randomized phase 3 EF-14 trial

Abstract

BACKGROUND: The randomized phase 3 EF‐14 trial showed that the addition of tumor treating fields (TTFields) to temozolomide (TMZ) treatment improved survival in newly diagnosed glioblastoma patients over TMZ alone. As TTFields delivery is designed to optimize dose at the tumor bed, we hypothesized that tumor recurrence is likely to occur in distant sites and we investigated the pattern of tumor growth at relapse in due consideration of TTFields dose intensity distribution. METHODS: Patients on therapy for more than 2 months who exhibited radiological progression were included in the study (treatment: N=280/466, control: N=122/229) . Contrast‐enhanced T1‐weighted scans at baseline and at relapse served for segmentation of enhancing tumor and necrosis. Lesions outside a proximal boundary zone of 20mm surrounding lesions identified at baseline were defined as distal. In addition, infratentorial progression was assessed. Moreover, we identified patients treated with TTFields with sufficient MRI data for dose calculations and radiographically confirmed tumor relapse (N=225/466). TTFields intensity distributions were computed in a finite element method applied in a realistic head model with virtual transducer arrays. TTFields dose density (mW/cm3) was defined as TTFields power loss density multiplied by average patient compliance during the first 6 months of therapy. At baseline, regions of residual tumor and regions of normal brain were identified to calculate TTFields dose within each of these regions and noted its relationship with subsequent recurrence. RESULTS: Distal progressions were more common in the TTFields arm (18% vs. 8%). Infratentorial progression was seen in 4% of the treatment arm vs. 0 patients in the control. Lesions at progression were more distant from the original lesion in the TTFields arm (57.0 + 26.2 mm vs. 46.6 + 14.8 mm). Local tumors grew at a significantly lower pace in the TTFields arm as compared to the control arm (3.6 + 14.5 ml vs. 8.3 + 17.9 ml). The average dose density in areas of enhancing tumor that regressed to normal was higher than in the areas of normal brain that progressed to enhancing tumor (0.84 mW/cm3 vs. 0.75 mW/cm3). Regardless of the expansion margin used for analysis, TTFields dose density in areas of normal brain that remained normal at the time of progression was higher than in the areas that progressed to enhancing tumor. CONCLUSIONS: TTFields appear to induce distant tumor growth pattern and are associated with a lower tumor growth rate. In addition, the importance of dose density at the tumor bed is stressed as higher dose densities were seen in areas of normal brain that remain normal at the time of progression providing further rationale for carefully planning array placement to maximize delivery to areas at higher risk of recurrence. Taken together, these findings confirm the clinical activity of TTFields.