Chemical Exchange Saturation Transfer (CEST) of endogenous agents is growing in popularity as a pre-clinical MR technique and has recently been used to evaluate treatment efficacy. Recently, CEST (in particular, amide proton transfer) has been used to assess several treatments including distinguishing tumour necrosis from radiation necrosis [1] in rat gliomas and high-intensity focused ultrasound (HIFU) treatments in murine colon carcinomas [2]. Others have quantified the repeatability of in vivo CEST by measuring the amide, amine, and aliphatic peaks in a fast-growing murine tumour cell line [3]. In this study, we evaluate the efficacy of two treatments that result in rapid loss of perfusion ultimately resulting in necrosis: 1) the vascular targeting drug combretastatin A-4 phosphate (CA4P) and 2) a large 10Gy dose of whole-body radiation.

Animals: Fifteen (15) NOD/SCID mice were implanted with a murine squamous cell carcinoma (SCC VII) on the left flank and tumours were allowed to grow until they reached 300mm$$$^3$$$. Animals were imaged daily for three days with no interventions to assess physiological variability in the amine, amide, and aliphatic peaks. Following the third imaging session, 5 of the tumours were injected i.p. with 120 $$$\mu$$$L of CA4P at a dose of 80 mg/kg and imaged 24 hours later. 5 other mice received a 10Gy dose of radiation.

MRI: Imaging was performed using a 7T scanner (Bruker Biospec 70/30, Germany) with an 86-cm coil for transmit and a custom built surface receive coil. A high-resolution, T2-weighted anatomic image was acquired using a 2D multi slice RARE inversion recovery sequence (TR/TE = 7700/11 ms, acceleration factor 8). CEST scans were acquired using an EPI-based imaging scheme and a continuous-wave saturation at B1=1.0$$$\mu$$$T for 10s at each saturation frequency offset. Spatial resolution of the CEST scan was 0.5 mm in-plane with 1.5 mm slice thickness. 80 offset frequencies were acquired with a range from -20 to +20 ppm with higher spectral resolution near the peaks of interest (2.5ppm, 3.5ppm, -2ppm, -3ppm). Total scan time for the CEST sequence was $$$\approx$$$ 27 minutes. The offset frequencies were ordered in an alternating positive/negative ranked fashion.

Analysis: First, the water peak was fit using a Lorentzian line shape function. Individual voxels were corrected for B0 inhomogeneities by referencing the central water peak to 0 ppm. A superposition of four Lorentzians (water, amide, amide, and aliphatic) was used to fit the data using non-linear curve-fitting techniques in iPython (scipy). Constrains were placed on individual peak locations and width but peak amplitudes were allowed to vary freely.

Histology: Following the last imaging session, animals were injected with 50 $$$\mu$$$L of the perfusion dye carbocyanine 5 minutes prior to sacrifice and the tumours were immediately excised and frozen. Sequential sections 10$$$\mu$$$m thick were obtained every 0.5 mm and analyzed to identify the fraction of perfused vessels and necrosis. Sections were stained with TUNEL to mark apoptosis and CD31 to mark blood vessels, and imaged using a robotic microscope and camera to obtain tiled images of whole tumour sections [5].

Figure 1A shows that decomposing the raw data into four Lorentzian functions describes the data well. Locations of the peaks are as previously described in the literature [3]. Preliminary analysis indicates that the effects of CA4P on the tumour 24 hours after i.p. administration were comparable to normal tumour progression as all three peaks were not significantly different between the controls and the treated tumours. As hypothesized, changes in the amide and aliphatic peaks following radiation treatment were found. Further investigation with histopathology is required to fully understand the effects of radiation damage in the tumours and the mechanism of action that causes such a large shift in the aliphatic peak.

This study characterizes the evolution of the CEST signal in three signal regions as tumours progress towards larger amount of necrosis through three different mechanisms. Firstly the quickly growing control tumours develop increasing levels of necrosis through natural tumour progression. Secondly, CA4P induces vascular shutdown and triggers the onset of necrosis in its wake. Lastly, 10Gy of external radiation causes widespread cell death in the tumour and displays yet another CEST signature in response. Ultimately, while acquisition of the full CEST spectrum leads to a more complex analysis, there is a wealth of data that can be explored for a more rich analysis compared to basic asymmetry analysis.