Evaluation of ADC mapping as an early predictor for tumor response to chemotherapy in recurrent glioma treated with bevacizumab/irinotecan. Proof of principle

Purpose: The assessment of the radiological response of recurrent glioma is based on the Macdonald or RECIST criteria 8 to 10 weeks from the start of treatment. Magnetic resonance imaging using an apparent diffusion coefficient map may provide an earlier measure for predicting the response to therapy of recurrent glioma. Materials and Methods: Twelve patients with recurrent high-grade glioma were enrolled in a feasibility study of pretreatment MRI on day 1, intra-treatment MRI in week 3, and post-treatment MRI in week 12. Prognostically relevant ADC values (ADCprog) of each recurrent glioma at 3 weeks were calculated as a function of their pre- and intra-therapy ADC values (ADCpre - ADCintra = ADCprog). Because we hypothesized that smaller ADC values correlate with less Brownian motion of water molecules in the extracellular space and that a higher cell density may restrain this water diffusion, we set smaller ADC values at a second time point as 'progressive disease' (PD) and higher ADC values as 'partial response' (PR). A change in ADCprog of less than 10 x 106 mm2/sec was set as 'stable disease' (SD). The ADCprog values were always calculated before the final scan after 3 months was performed. The readers were blinded to the future development of the tumor. Results: In 10 of the 12 patients we could correctly predict the tumor response to chemotherapy. One patient died before the three-month control, and one recurrent glioma did not develop as predicted. ADC mapping is found to predict patient response at 3 weeks from the start of treatment, revealing that early changes in tumor diffusion values could be used as a prognostic indicator also for chemotherapeutically treated recurrences of high-grade glioma. (orig.)