Altering fractionation during radiation overcomes radio-resistance in patient-derived glioblastoma cells assessed using a novel longitudinal radiation cytotoxicity assay.

Abstract

Current radiotherapy (RT) in glioblastoma (GBM) is delivered as constant dose fractions (CDF), which do not account for intratumoral-heterogeneity and radio-selection in GBM. These factors contribute to differential treatment response complicating the therapeutic efficacy of this principle. Our study aims to investigate an alternative dosing strategy to overcome radio-resistance using a novel longitudinal radiation cytotoxicity assay. Theoretical In-silico mathematical assumptions were combined with an in-vitro experimental strategy to investigate alternative radiation regimens. Patient-derived xenograft (PDX) brain tumor-initiating cells (BTICs) with differential radiation-sensitivities were tested individually with sham control and three regimens of the same nominal and average dose of 16 Gy (over four fractions), but with altered doses per fraction. Fractions were delivered conventionally (CDF: 4, 4, 4, 4 Gy), or as dynamic dose fractions (DDF) "ramped down" (RD: 7, 5, 3, 1 Gy), or DDF "ramped up" (RU: 1, 3, 5, 7 Gy), every 4 days. Interfraction-longitudinal data were collected by imaging cells every 5 days, and endpoint viability was taken on day 20. The proposed method of radiosensitivity assessment allows for longitudinal-interfraction investigation in addition to endpoint analysis. Delivering four-fraction doses in an RD manner proves to be most effective at overcoming acquired radiation resistance in BTICs (Relative cell viability: CDF vs. RD: P < 0.0001; Surviving fraction: CDF: vs. RD: P < 0.0001). Using in-silico cytotoxicity prediction modeling and an altered radiosensitivity assessment, we show DDF-RD is effective at inducing cytotoxicity in three BTIC lines with differential radiosensitivity.