Early Assessment of Antiangiogenic Effects of Sorafenib using IVIM in Mouse Model with Hepatocellular Carcinoma
Yong Zhang1, Bing Wu1, Xin Chen2, and Zaiyi Liu2

1GE Healthcare MR Research China, Beijing, China, People's Republic of, 2Radiology, Guangdong General Hospital, Guangzhou, China, People's Republic of

Synopsis

Antiangiogenic therapy is efficient to treat hypervascular tumor such as hepatocelluar carcinoma (HCC). Unlike chemotherapy and radiation therapy, tumor dimension doesn’t change in its early phase. Hence traditional reponse criteria based on morphological change fails to early assess the therapeutic response of antiangiogenic treatment. This study used intravoxel incoherent motion (IVIM) theory to separate perfusion and diffusion characteristics in HCC over sereval time points after antiangiogenic Sorafenib administration. It was found that IVIM-derived pure diffusivity and pseudo-diffusivity were able to detect the microvascular collapse and cellular edema in HCC at early phase of antiangiogenic medication.

Purpose

Tumor angiogenesis is crucial for the progression of cancer, especially in highly vascularized tumors such as hepatocellular carcinoma (HCC) [1]. Sorafenib is an anti-angiogeneic medication already in clinical stage, However its curing effectiveness is case dependent. Hence evaluation of the short term antiangiogenic effects of Sorafenib is necessary for optimizing the treatment plan. Conventional CT and MR fails for such purpose [2] as the functional change takes place at an earlier stage than dimensional changes. Apparent diffusion coefficient (ADC) from DWI was reported to be a sensitive means for early stage effects, however the reported findings have been conflicting [3], perhaps due to the mixed diffusion and perfusion effects conveyed by traditional ADC. IVIM, on the other hand, separately measures the microscopic diffusion and perfusion process, and we evaluate its use in monitoring the early term antiangiogenic effects in this work.

Methods

This study was approved by the local institutional review board for animal care and use. Experiments were performed in 12 female BALB/c-nu nude mice (32-40 days old), each weighing 14-17 g. MHCC-97 Hepatocellular carcinoma cells were planted subcutaneously in each nude mouse. After implantation of tumor, nude mice were maintained for three weeks. Prior to treatment, all nude mice were scanned on a GE 1.5 T whole body scanner equipped with a phased array wrist coil. For each mouse, 10 axial slices covering whole tumor were scanned using spin-echo diffusion-weighted echo-planar imaging with 11 b-values (0, 25, 50, 75, 100, 150, 200, 400, 600, 800, 1000 s/mm2). Other imaging parameters were: FOV=10 mm, slice thickness=2.5 mm, no interslice gap, acquisition matrix=128 × 128, TR/TE=4000/85.5 ms. The same IVIM acquisition was repeated on all nude mice at 1 hour, 3 hours, 6 hours, 12 hours after the administration of Sorafenib at a dose of 15 mg/kg. One randomly selected mouse was sacrificed for histologic H&E staining at each time point. For IVIM scan at each time point, all DWIs were coregistered to the b0 image and then fitted to the IVIM bi-exponential model S/S0 = f·exp(-b·D*) + (1-f)·exp(-b·D). Three metrics including true diffusivity D, perfusion-related diffusivity D* and perfusion fraction f were derived [4]. The edge of tumor was manually delineated on b0 image and transferred to the three metrics for measurement. Paired t test was used to compare the mean value of IVIM metrics between every two consecutive time points using SPSS (Chicago, IL, USA). P-values of 0.05 were considered to indicate significant difference.

Results

A typical set of IVIM parametric maps before Sorafenib administration are illustrated in Figure 1, along with the placement of ROI that encompasses the lesion region. Figure 2 showed the change of D, D* and f over time after Sorafenib administration, along with the histology staining at different time points; the mean values of IVIM metrics can be seen in Table 1. Different behaviours of IVIM metrics can be observed: D rapidly dropped by 42.5% (p < 0.05) at 1 hour after Sorafenib administration, but then stayed relatively stable afterwards; D* slightly decreased at 1 hour after treatment with no statistical significance and further decreased by 12.8% at 3 hours after treatment (p < 0.05), it then stayed relatively constant afterwards; f showed steady decrease over time.

Discussions and Conclusion

Compared to traditional DWI, IVIM offers several parameters that are associated with different microscopic processes. From the data, it can be seen that D was most sensitive to the initial effects of antiangiogenic treatment, potentially due to the formation of cellular edema that restricts the microscopic diffusion, however the level of diffusivity did not further change after a sharp drop. D* is affected by the blood velocity and caliber of the vessels [5], its delayed response may reflect the initial maintaining of the blood flow within the lesion. f directly maps the density of the vascular structures and the gradual drop agrees with the hypothesis. Although the sample size limits the statistical power in this study, it can be seen that IVIM-derived metrics are sensitive biomarkers for monitoring the early stage effects of antiangiogenic treatment; As compared with conventional ADC, different aspects of the micrio-strucutre change can be separately assessed that may lead to better understanding of the medication effects after administration of Sorafenib.

Acknowledgements

No acknowledgement found.

References

[1] Yu-Quan Xiong et. al., J Cancer Res Clin Oncol, 2011. [2] Koh DM et. al., Br J Radiol, 2006. [3] Koh DM et. al., Am J Roentgenol, 2007. [4] Le Bihan et. al., Radiology, 2008. [5] Boris Guiu et. al., Radiology, 2012.

Figures

Representative b0 image, ADC and IVIM-derived metrics before Sorafenib administration. a-e: b0, ADC, D, D* and f

Change of IVIM metrics over time after administration of the Sorafenib, along with the histology staining at corresponding time points.

Mean value of ADC and IVIM metrics at different time points



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
2799