Purpose

The apparent diffusion coefficient (ADC), the native longitudinal relaxation rate (R1_N), and the extravascular extracellular volume fraction (v_e) are three related parameters, which are affected by tumor cell density and water content in the tissue.^1,2 Investigating their relationship and treatment-induced changes in their relationship will lead to better understanding the three parameters as imaging biomarkers for cancer treatment response. In this study, we compare the three parameters for patients with vestibular schwannomas (VS) treated with bevacizumab.

Methods

12 patients with type 2 neurofibromatosis (NF2) related VS were imaged on a 1.5-T Philips Achieva scanner. Patients were treated with the anti-vascular endothelial growth factor antibody, bevacizumab, 5 mg/kg every 2 weeks. The patients received scans before and three-month after treatment. Dynamic contrast-enhanced (DCE) MRI was performed using a low dose (0.02 mmol/kg) high temporal resolution (LDHT) technique.³ Data were analyzed with the extended Tofts model to allow measurement of contrast transfer coefficient (K^trans), blood plasma fraction (v_p), and extravascular extracellular space fraction (v_e). R1_N was measured using a variable flip angle technique. ADC was calculated from diffusion-weighted imaging. Longitudinal co-registration was performed for all scans for each participant. The ADC volume maps, 3D-VFA series and images from the LDHT-DCE 4D dynamic series from each visit were co-registered to one pre-contrast high spatial resolution DCE 3D data set acquired on day zero. VSs were segmented using a semiautomatic segmentation program (SX plugin for Osirix, Segmentix, Fr). The results of segmentation were reviewed and, where necessary, edited by an experienced neuroradiologist (A.J.). Average values were then derived from the 3D region of interest (ROI). The changes in the tumor mean values in ADC, R1_N, and v_e over the course of bevacizumab treatment were compared. The 20 VSs were divided into two groups: responders (Res.) and nonresponders (Non.). Response was defined as a volume reduction exceeding 0.125 cm³ or a relative volume decrease exceeding 5%.

Results

20 VSs from the 12 patients were analyzed. Fig. 1 demonstrates linear regression analysis of the relationship between each pair of the ADC, R1_N and v_e mean tumor values pre- and post-therapy. The results from the linear regression analysis (R² and P-value) are listed in Table 1. Pre-treatment mean ADC and R1_N showed a strong correlation (R² = 0.648, P < 0.0001) between the 2 parameters. ADC and v_e, and R1_N and v_e showed a weak correlation respectively. However, these correlations were lost following bevacizumab therapy. The difference in mean parameter values pre- and three month post-therapy was tested with the paired t-test (Table 2). Only R1_N showed significant difference pre- and post-therapy in the responder group.

Discussion

The tumor mean values of ADC, R1_N and v_e showed correlation to each other before bevacizumab treatment. The correlation lost three-month after treatment. Both R1_N and ADC displayed power in prediction of treatment.³ R1_N showed more sensitive to anti-angiogenic therapy induced changes in VS than ADC and v_e.

Conclusion

This study may provide useful information for better understanding of R1_N, v_e and ADC as imaging biomarkers for cancer treatment response in vestibular schwannomas over the course of bevacizumab treatment.

Acknowledgements

No acknowledgement found.