Medullary thyroid carcinoma is a neuroendocrine cancer originating from calcitonin-producing parafollicular cells in the thyroid gland. Although surgical removal of thyroid gland is the best treatment, recurrence is common. In addition, since current therapeutic drugs are limited and development of resistance is commonly observed, further drug development is important¹. We have demonstrated that the amide proton transfer (APT) imaging detects early tumor response to therapy and predicted the treatment outcomes prior to the anatomical changes in the glioma². Our purpose was to investigate advantage of APT imaging in evaluation of anticancer effect of two new drug treatments, administered individually or in combination, in a mouse model of medullary thyroid carcinoma.

Animal Protocol: We used a novel orthotopic mouse model (NSE/p25-gfp bi-transgenic mouse line) which develops medullary thyroid carcinoma upon overexpression of p25 which is a cyclin-dependent kinase 5 activator³. Seventeen mice were subjected to the first MRI session (Day0) at 14-15 weeks after transgene induction and were divided into three groups treated as follows; monotherapy group: 0.74mg/kg/day Romidepsin which is a histone deacetylase inhibitor (HDAC: N=4); combination group: 35mg/kg/day Nintedanib which is a tyrosine kinase inhibitor (TKI) + 0.37mg/kg/day Romidepisin (N=8); and control group: vehicle (phosphate buffered saline: PBS, N=5). Drugs were intraperitoneally administrated daily (Nintedanib) or every 3days (Romidepsin). The same MRI session was repeated at Day14 and Day21. The tumors were harvested after the final MRI session for histological analysis.

MRI: All MRI sessions were conducted in a 7T animal MRI (Varian, Inc, Palo Alto, CA). High-resolution axial multislice T2-weighted images (T2WI) were obtained on entire tumor using a fast spin-echo sequence (TR/TE = 2500/60 msec; FOV = 32×32 mm; matrix size = 256×256; thickness = 1 mm; gapless; NEX = 8). On a single 1-mm-slice delineating the maximum diameter of the tumor, APT imaging was performed as follows: Gradient echo images were obtained following a presaturation pulse (continuous-wave: CW block pulse, B₁ = 2.3 μT, duration = 5 s) which was applied at 29 frequency offsets from 7 to -7 ppm with an interval of 0.5 ppm. Other imaging parameters were: TR/TE = 5.32/2.64 ms, flip angle = 20°, FOV = 32×32 mm, matrix = 128×64 (reconstructed to 128×128), NEX = 8. A control image was obtained with the presaturation pulse at 300 pm. Water saturation shift referencing (WASSR) images were collected for B₀ inhomogeneity correction with a CW pulse (B₁ = 0.2 μT, duration = 200 ms) which was applied at 31 frequency offsets from 1.5 to -1.5 ppm every 0.1 ppm. Total acquisition time for each animal was approximately 50 min.

Image Analysis: Tumor volumes were manually measured on the T2WI. The z-spectra were fitted through all offsets on a pixel-by-pixel basis followed by the correction for B₀ inhomogeneity as previously reported⁴. MTR asymmetry (MTR_asym) was defined as: MTR_asym = S_sat(−offset)/S₀−S_sat(+offset)/S₀, where S_sat and S₀ are signal intensities on the images with presaturation pulse at 7 to -7 ppm and control (300 ppm), respectively. The calculated MTR_asym map at the offset of 3.5 ppm is called the APT-weighted image (APTWI). Region-of-interests (ROIs) were carefully placed in the tumors on APTWI, and the average of both lobes of the thyroid was calculated as the APT ratio (APTR) of the tumor. APTR was also measured in the brain stem for a reference in each mouse. Corrected tumor APTR was calculated as the difference between these two APTRs (tumor – brain).

Tumor growth in volume was inhibited in the treated groups and a significant difference was found in the combination group compared to the control group at Day21 (Fig.1). Figure 2 shows the typical APTWIs in all groups at Day0 and Day14 where the APTR obviously changes in the combination group. The corrected APTR relative to baseline (Day0) showed substantial reduction in the combination group at Day14 and Day21 while it stayed almost constant in the other two groups through the protocol (Fig.3). The present study indicated that APT imaging could detect different treatment responses of the thyroid cancer in different treatment strategies. The temporal change in APT signals was occurred earlier in the time course of chemotherapy and preceded morphologic changes, which was consistent with previous results in the brain tumor². APT imaging might provide more sensitive molecular metabolite changes in the course of anticancer treatment and may have prognostic potential to provide a definitive diagnosis without additional follow-ups on medullary thyroid carcinoma.