RESUMO
PURPOSE: Tumor-treating fields (TTFields) are an antimitotic treatment modality that interfere with glioblastoma (GBM) cell division and organelle assembly by delivering low-intensity, alternating electric fields to the tumor. A previous analysis from the pivotal EF-14 trial demonstrated a clear correlation between TTFields dose density at the tumor bed and survival in patients treated with TTFields. This study tests the hypothesis that the antimitotic effects of TTFields result in measurable changes in the location and patterns of progression of newly diagnosed GBM. METHODS AND MATERIALS: Magnetic resonance images of 428 newly diagnosed GBM patients who participated in the pivotal EF-14 trial were reviewed, and the rates at which distant progression occurred in the TTFields treatment and control arm were compared. Realistic head models of 252 TTFields-treated patients were created, and TTFields intensity distributions were calculated using a finite element method. The TTFields dose was calculated within regions of the tumor bed and normal brain, and its relationship with progression was determined. RESULTS: Distant progression was frequently observed in the TTFields-treated arm, and distant lesions in the TTFields-treated arm appeared at greater distances from the primary lesion than in the control arm. Distant progression correlated with improved clinical outcome in the TTFields patients, with no such correlation observed in the controls. Areas of normal brain that remained normal were exposed to higher TTFields doses compared with normal brain that subsequently exhibited neoplastic progression. Additionally, the average dose to areas of the enhancing tumor that returned to normal was significantly higher than in the areas of the normal brain that progressed to enhancing tumor. CONCLUSIONS: There was a direct correlation between TTFields dose distribution and tumor response, confirming the therapeutic activity of TTFields and the rationale for optimizing array placement to maximize the TTFields dose in areas at highest risk of progression, as well as array layout adaptation after progression.