Quantitative Bio - imaging of Gadolinium -157 in Tissues Through Laser - ablation ICP - MS for Neutron Capture Therapy

Objective : Gadolinium-157 ( 157 Gd) has attracted interest for its use in the preparation containing metal elements of the neutron capture therapy (NCT), one of the radiation therapies. The study, however, has not developed because of the difficulty to quantify the concentration of 157 Gd in the tissue. Therefore, we established a quantitative imaging technique for 157 Gd in bio-tissue employing laser-ablation inductively coupled mass spectrometry (LA-ICP-MS). Materials and Methods : 4 female New Zealand white rabbits, which were inoculated with rabbit VX-2 cells participated in this study. 157 Gd in water-in-oil-in-water (WOW) emulsion was injected via the proper hepatic artery into the rabbits and, after 24 or 72 hours, the rabbits were killed, and the liver tissues were harvested. We also prepared 7 standard tissues which were mixed with gadoteridol solutions, the final amount of 157 Gd was 0, 4.4, 22, 44, 220, 440 and 660 μ g/g, respectively. The harvested livers and standards were sectioned on a cryostat at 5 μ m intervals and they were analyzed by LA-ICP-MS. Result : In an experiment on animal cancer tissue, 157 Gd was observed to accumulate around the cancer. Conclusion : The 157 Gd concentration in bio tissue can be quantitatively assessed through LA-ICP-MS imaging and it was expected to contribute the progress of NCT study.


Introduction
Neutron capture therapy (NCT) uses secondary radiation particles emitted by the nuclear neutron capture reaction to kill cancer cells. Locher introduced NCT soon after the discovery of the neutron 1) . For therapeutic application of NCT to malignant melanoma and gliomas, boron-10 ( 10 B) compounds have been used as short-range alphaparticle-producing agents 2) 3) .
Recently, gadolinium-157 ( 157 Gd) has attracted attention as an alternative NCT agent because it has the largest thermal-neutron capture cross section among all stable nuclides (255,000 barns, 66 1 times as large as that of 10 B) and gamma-rays and Auger electrons are released by the neutron capture reaction 4) . This property makes it possible to reduce the total neutron fluence needed for the same number of thermal neutron absorptions with 10 B. In addition, because gadolinium has been used as a contrast agent of magnetic resonance imaging diagnosis, 157 Gd-NCT is expected to be used in combination with magnetic resonance imaging.
To increase the therapeutic effect of NCT, it is important to enhance the accumulation of 10 B or 157 Gd in a tumor. Although NCT with 157 Gd was first established in the 1980s 5) , 157 Gd-NCT has not progressed in contrast with 10 B-NCT, because it is difficult to quantitatively measure the amount of 157 Gd accumulated in the tissue.
Laser-ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) is an analytical technology that combines the laser ablation sampling technique and ICP-MS. A laser beam is focused on the sample surface to generate fine particles, which are then transported to the ICP-MS instrument for digestion and ionization. The LA-ICP-MS technique is sensitive enough to determine the abundance of trace elements 6)-12) . Gadolinium imaging employing LA-ICP-MS has thus been developed recently 13)-17) . Although Kostiv et al. has reported the quantitative imaging of nanoparticles containing Gd 18) , the quantitativity was not enough because they did not match the matrix of trace elements in calibration standards with the samples analyzed.
The present study developed a method of quantitative imaging 157 Gd in cancer tissue using the LA-ICP-MS technique, by matching the matrix of trace elements in calibration standards to those of tissue samples.

Preparation of standard liver tissue for calibration
To examine the quantitative relationship between the signal intensity obtained by LA-ICP-MS and the 157 Gd concentration in the liver, we prepared a standard tissue in accordance with Hareʼs procedure with minor modification 19) : Frozen livers of intact New Zealand white rabbit were purchased from Sankyo Labo Service Corporation and defrosted at 4℃ and then washed three times with deionized H2O to remove residual blood. Segments of 3-5 cm 3 were sectioned, and blood vessels, fluids and connective tissues were removed using surgical scissors. After homogenization of the liver using a handheld blender (TK-210, Tescom, Tokyo), 53.6 μl of gadoteridol standard solution in saline (0, 1.88, 9.35, 18.8, 93.5, 186 and 280 gadoteridol-mg/ml) was added to the homogenate of 1 g each to obtain nominal 157 Gd amounts of 0, 4.4, 22, 44, 220, 440 and 660 μg/g in the homogenized tissue, respectively. A portion of the mixture at each gadolinium concentration was packed into a plastic histology mold, frozen in liquid nitrogen, and stored at -80℃ for LA-ICP-MS analysis. The different point to Hareʼs method was the matching of the matrix (i.e., the composition of substances in samples) for LA-ICP-MS analysis 20) .
A portion of the above mixture at each concentration of 157 Gd was weighed and put into a perfluoroethylene bottle with 0.4 ml of 68% HNO3, left overnight at room temperature (25℃), and then digested with 0.2 ml of H2O2 in a microwave oven (ETHOS PLUS, Milestone General, Bergamo, Italy). These digested liver samples were stored at room temperature until conventional ICP-MS analysis was performed.

Preparation of VX-2 cancer liver from rabbits administrated with 157 Gd
All animal experiments in this study were conducted in accordance with the guidelines of Meiji Pharmaceutical Universityʼs animal ethics committee (Approval number: 2612) and the Declaration of Helsinki. To examine the distribution in vivo, 157 Gd in water-in-oil-in-water (WOW) emulsion was prepared as reported previously 21) ; gadoteridol solution (1,396.5 mg/5 ml) was filtrated using a controlled porous glass membrane and added to 5 ml of iodized poppy-seed oil (lipiodol) containing surfactant to form a water-inoil (WO) emulsion. The WO emulsion was emulsified again with aqueous phase containing 5 ml of saline and surfactant (to form another WOW emulsion). The 157 Gd WOW emulsion was then injected via the proper hepatic artery into female New Zealand white rabbits, which were inoculated with rabbit VX-2 cells (Shope virus induced squamous carcinoma cell line, skin origin) to the left lobe of the liver two weeks before 22) . At 24 or 72 hours after injection, the rabbits were killed and the liver tissues harvested, frozen in liquid nitrogen, and stored at -80℃ until use.

Analysis of standard liver and VX-2 cancer in liver tissues
The standard liver and VX-2 cancer in liver tissues were sectioned on a CM3050S cryostat (Leica Microsystems, Bensheim, Germany) at 5 μm intervals. Cut sections were mounted on glass microscope slides, air-dried and stored at -80℃. The sections were inserted into a cell and ablated line by line using a commercial laser ablation instrument, an NWR213 (ESI New Wave Research, Oregon, USA), coupled to an Agilent 8800 Triple Quad ICP mass spectrometer (Agilent Technologies, Australia). The distribution of 157 Gd and 63 Cu were visualized using with the iQuant2 software 23) . Additionally, 157 Gd concentrations in the digested liver samples were measured using the conventional ICP-MS method as previously reported 24)-33) .
As evidence has been given that copper concentrations are higher in the liver than in the skin 34)-38) , we expected that the distribution of copper could be distinguished between the liver and the VX2 inoculation cells. LA-ICP-MS analysis was thus also conducted on the copper-63 ( 63 Cu) distribution in the VX-2 cancer in the liver sample to visualize the location of cancer tissue. Overlaying the distributions of 63 Cu and 157 Gd could help confirm the distribution of 157 Gd in the liver and cancer tissues, because 157 Gd accumulated in tissues is washed out in usual histological techniques, such as hematoxylin and eosin staining. Instrumental parameters and analytical conditions are summarized in Table-1.

Results
Figure-1 shows the standard curves calculated from the signal intensity of LA-ICP-MS on sections of the standard tissues and 157 Gd concentrations in the corresponding digested liver obtained by conventional ICP-MS measurement. The homogeneity of the 157 Gd distribution in the standard section was assessed for 10 repeated measurements; the percentage relative standard deviation was less than 30% for concentrations over 5.5 μg/g. The linearity of calibration curve was good, ranging from 0-760 μg/g 157 Gd concentration (r 2 = 0.9977). Calculated limits of detection and limits of quantification of 157 Gd determined from the blank tissue were respectively 0.0135 and 0.0409 μg/g, which were calculated as Matsukawa et al 39) .
The distribution of 157 Gd in the VX-2 cancer in

Discussion
We showed the linear correlation between the signal intensity of imaging and actual tissue concentration of 157 Gd in the liver. Even 24 hours after the injection of 157 Gd in WOW emulsion, almost all data points (99.95%) of the 157 Gd signal intensity fell within the range of the standard curve. This method can therefore be used to estimate the amount of 157 Gd accumulated in the tissue by varying the dose administered for NCT. As expected, the location of cancer tissue was identified by the 63 Cu distribution. In this study, we utilized the low concentration of copper in VX-2 cells to distinguish the normal and cancer tissues. This observation, together with those of higher concentration of copper in lymphoma, breast cancer and gastrointestinal tract cancer 40) , as well as lower concentration of zinc in liver and pancreatic carcinomas 41) , suggests that copper and zinc distributions are available for the determination of the cancer area.
As identified from the imaging results, 157 Gd was located at the tumor surface lesion including feeding vessels 24 hours after administration. Meanwhile, gadolinium was found in deeper lesions in the cancer tissue and not only border lesions with the feeding vessels 72 hours after administration. In NCT, differences in 157 Gd concentrations between the surrounding normal tissue and cancer tissue were important for the evaluation of the preparation. For a therapeutic option to be viable, the radiation dose delivered to the cancer must exceed the background radiation that normal tissue receives from nonspecific neutron absorption. Generally, the selective tumor/normal tissue concentration ratios must be above unity and preferably 3: 1 or higher 42) . The results suggest that the concentration of 157 Gd can be estimated by considering the distribution over time.

Conclusions
There were two device points in the present study. First, because the matching of the matrix (i.e., the composition of substances in samples) was important for LA-ICP-MS analysis 20) , we prepared standard samples and liver tissue samples following the method of Hare et al 19) . The volume of liquid added was the same for all analyzed samples, providing an excellent decision coefficient. Second, we imaged 63 Cu as a marker to determine the area of cancer in the sample, overlaying the 63 Cu image on the 157 Gd image. The present study thus showed that the 157 Gd concentration in bio tissue can be quantitatively assessed through LA-ICP-MS imaging. We thus expect rapid progress in NCT using 157 Gd agents.

Conflicts of interest
There are no conflicts to declare.

Figure-2
Quantitative gadolinium imaging of rabbit VX-2 cancer in liver Quantitative images of 157 Gd (left panels). Images of copper distribution in rabbit hepatic cancer (middle panels) and merged images 157 Gd and copper distributions (right panels).