Comparison of cytotoxic and anti-angiogenic treatment responses using functional diffusion maps in FLAIR abnormal regions

4 . Furthermore, recent evidence suggests that recurrence during treatment with anti-angiogenic drugs acting along the VEGF signaling pathway may result in diffuse tumors with a predominantly infiltrative phenotype 5 . Therefore, the overall purpose of the current study was to compare cellularity metrics extracted from fDMs applied to FLAIR abnormal regions between progressive, stable, or responsive patients treated with either standard therapies (chemotherapy & radiation therapy) or anti-angiogenic therapy combined with chemotherapy (bevacizumab & irinotecan). This will enable us to examine the utility of fDMs for the evaluation of anti-angiogenic treatment regimens, and determine whether these treatments result in significant increases in infiltrative tumor volume compared to standard treatment. Methods A total of 20 patients with intracranial tumors, scanned at least 3x, were enrolled in the current study for a total of 112 scan sessions and 92 unique fDM datasets. Eight patients were treated with bevacizumab (10mg/kg) and irinotecan (125mg/m 2 ) for a total of 60 MRI scan sessions (52 fDMs). Twelve patients were given temozolomide (200mg/m 2 /day) concurrently with standard radiation therapy for a total of 52 MRI scan sessions (40 fDMs). Each scan session was classified as responsive disease (RD), stable disease (SD), or progressive disease (PD) after each scan session based on clinical MRI scans (SPGR, FLAIR, and pre-/post-contrast T1-w MRI) and neurological assessment. Functional diffusion maps (fDMs) were calculated from registered sequential diffusion weighted images and then applied to regions of FLAIR abnormality with ADC thresholds of 0.55 x 10 -3 mm 2 /s, as per the fDM protocol 2 . Voxels were stratified as not changing (green), increasing ADC (red), or decreasing ADC (blue) according to this threshold. The volume of hypercellularity (physical volume of blue voxels, in mL), volume of hypocellularity (physical volume of red voxels, in mL), and the time rate of change in hypercellular volume (in uL/day; based on the change in volume and time between each subsequent scan days), and the rate of change in hypocellular volume (in uL/day) were calculated from resulting fDMs. Results Representative fDMs are shown in Figure 1 for both treatment groups and response categories. Overall, trends in the physical volume of hypercellularity suggest that patients treated with bevacizumab/irinotecan had a significantly higher volume of hypercellularity compared with patients treated with temozolomide (Mann-Whitney, P 0.1). The time rate of change in hypocellular volume, however, was significantly different between treatments, tumor status, and the interaction between treatment and tumor status (Fig. 2D; ANOVA, P<0.05). Specifically, the rate of hypercellular volume change was higher in RD compared to SD (Tukey, P=0.025), and PD during temozolomide treatment resulted in a significantly faster decrease in hypocellular volume compared with bevacizumab and irinotecan treatment (Tukey, P<0.001). Discussion The results of the current study support the hypothesis that recurrence after treatment with bevacizumab results in growth of an infiltrative (i.e. non-enhancing) tumor type, since the overall volume of hypercellularity was significantly higher in patients treated with bevacizumab compared with temozolomide. However, it must also be noted that patients on the anti-angiogenic therapy regimen had recurrent tumors and failed the standard regimen of chemotherapy plus radiation therapy. Consequently, future studies are needed to determine if increased cellularity remote from the primary tumor is due to an infiltrating tumor phenotype promoted by anti- angiogenic therapies or if it is a late response to standard therapies. Still, in general, this study suggests the volume of hypercellularity and the time rate of change in hypercellular volume are sensitive metrics for detection of tumor progression in both cytotoxic and anti-angiogenic treatment paradigms.