18F-FDG and 18F-FLT
18F-FDG (CYCLODX, CIUSSS de l’Estrie - Centre Hospitalier Universitaire de Sherbrooke, Canada) and 18F-FLT were prepared by the Sherbrooke Molecular Imaging Center (CIMS, Sherbrooke, Quebec, Canada). 18F-FLT was produced using the protected nosylate precursor and the method of Yun et al. [22].
Cell culture
The HCT116 human colorectal carcinoma cell line obtained from ATCC was routinely cultured in modified Eagle’s medium (Sigma-Aldrich, Oakville, Canada) supplemented with 10% fetal bovine serum, 2 mM glutamine,1 mM sodium pyruvate, 100 units/ml penicillin, and 100 μM streptomycin in a fully humidified incubator at 37 °C in an atmosphere containing 5% CO2.
Human colorectal cancer xenograft mouse model
Experiments were performed with outbred male nude mice at 4–6 weeks of age (Charles River Laboratories, Saint-Constant, QC, Canada). The animals were maintained in an animal facility, under specific pathogen-free conditions. Housing and all procedures involving animals were performed according to the protocol approved by the Université de Sherbrooke Animal Care and Use Committee (protocol number 235-14B). Human colorectal HCT116 tumor cells (2 × 106, 0.1 mL) were inoculated subcutaneously (s.c.) into each rear thigh and one on the right shoulder. During each animal handling implantation, the animals were anesthetized with an intraperitoneal injection of ketamine/xylazine (87/13 mg/mL) at 1 mL/kg. Tumor size measurements began 1-week post-injection and continued biweekly. Tumor volumes were calculated with the following formula: V (mm3) = π/6 × a (mm) × b2 (mm2), where a and b were the largest and smallest perpendicular tumor diameters, respectively. All experiments began when tumor volumes reached a diameter of about 5–7 mm. The tumor-bearing animals were randomized into different groups of two to four animals each.
Distribution kinetics and tumor clearance of i.t. 18F-FLT and 18F-FDG
The animals were anesthetized by inhalation of 1.5% isoflurane and 1.5 L/min oxygen during i.t. injection and PET imaging procedures. The single i.t. infusion of 5 MBq of 18F-FLT or 18F-FDG solution was applied into the tumor on one side of the rear thigh, whereas the contralateral tumors were not treated. The solution was introduced into the central section area of the tumor. For each injection, the needle tip placement was at approximately one-third depth in the tumor along with the needle insertion direction. The leakage of the radiolabeled compound is the main concern for i.t. injection. To avoid this complication, the i.t. infusion was performed at a slow infusion rate (10 μL/min) over 10 min and the needle was left in place within the tumor for about 5 min following completion of the i.t. infusion to reduce any backflow of the 18F-FLT or 18F-FDG solution. Total infusion volume for each tumor was limited to about 30–50% of the tumor volume from caliper measurements, which were determined on the day of the study.
The administration of 18F-FLT or 18F-FDG by i.t. injection was performed with the animal placed inside the scanner, at the start of data acquisition at time 0. Dynamic PET data were acquired in list mode from time 0 to 120 min post-injection using the Triumph/LabPET8™ platform (Gamma Medica, Northridge, CA) at the CIMS.
PET images were reconstructed on a 120 × 120 × 128 matrix with a 0.5 × 0.5 × 0.6 mm3 voxel size using the standard LabPET 3D maximum likelihood expectation maximization algorithm implementing a 3D model of the physical detector response. Frame durations for the reconstructed images were 10 × 1 min, 10 × 5 min, and 4 × 15 min. All PET images were corrected for physical radionuclide decay, dead time, and differences in crystal detection efficiency.
To quantify the radiotracer uptake, regions of interest (ROI) were drawn around tumors, organs, and whole body in the last image frame using the Amide software [23]. These ROIs were then applied to all frames to obtain time-activity curves (TAC) for each organ. The ROI activity was expressed as percent injected dose per gram of tissue (%IA/g) with the whole body radioactivity measured by PET. The residency time (in hours) for each organ was calculated using decay-uncorrected TAC as follows [24]:
$$ {\tau}_{\mathrm{h}}=\frac{\underset{0}{\overset{2\mathrm{h}}{\int }}\mathrm{TAC}(t) dt+\mathrm{TAC}\left(2\mathrm{h}\right)\underset{2\mathrm{h}}{\overset{\infty }{\int }}{e}^{-\lambda t} dt}{A_0} $$
where TAC(t) is the activity in the organ at time t, TAC(2 h) is the activity in the organ at the last time point of measurement (2 h), λ is the physical decay constant of 18F, and A0 is the injected activity in the main tumor. Trapezoidal rule was used to numerically integrate the organ TAC over the measurement time, while the analytical integration was performed on the exponential decay term.
In order to assess the radiation burden of 18F-FLT to bone and bone marrow, the uptake of 18F-FLT following i.t. (n = 3) and i.v. (n = 1) administration of 18F-FLT was compared from PET acquisitions by tracing ROI on the forepaw long bones and on the spine. The relative exposure of the bone marrow was estimated from the area under the non-decay-corrected time-activity curves (AUC) extrapolated to infinity with the physical decay of 18F.
Determination of the tumor response after i.t. infusion of 18F-FLT and 18F-FDG radiotherapy and EBRT
In order to determine the dose dependence of tumor response, single i.t. injections of 15 and 25 MBq 18F-FLT or 18F-FDG were administered into the tumor on one side of the rear thigh, whereas the contralateral tumors were left untreated. All experiments began when tumor volumes reached a diameter of about 5–7 mm.
External beam gamma radiation was performed with a 4C Gamma Knife (Elekta Instruments AB, Stockholm, Sweden). A single i.t. injection of 0.9% saline was administered into the tumor to emulate the radiotracer administration. Mice were anesthetized and positioned in our in-house stereotactic frame designed for the 4C Gamma Knife [25]. The radiation treatment (15 Gy, dose rate of 3.6 Gy/min) using 8-mm collimators was delivered at predetermined coordinates targeting the tumor. Radiation was applied to the tumor located on one side of the rear thigh, whereas the other side was kept as the non-irradiated control tumor.
Tumor growth was measured after treatment twice a week. Tumor volumes were calculated as described in the mouse model section. Fivefold growth delay (5Td) was considered to be the time required for the tumor volume to increase by a factor of 5, compared to the initial volume at the beginning of treatment. Tumor growth delay (TGD) was calculated by subtracting the 5Td value of the treated group from the 5Td of the control group. An enhancement factor (EF) was also calculated by dividing the 5Td of the treated group by the 5Td of the non-treated control group.
Assessment of absorbed dose from PET imaging by Fricke dosimetry
The mean 18F-FDG activity in different tissues measured in our mouse model by PET imaging was correlated to the absorbed dose assessed in vitro by Fricke dosimetry [26, 27], as described by Tippayamontri et al. [28]. Briefly, the dose-response of the Fricke dosimeter and total activity measured by PET were determined at different times for a 3 mL Fricke solution and a 3 mL of deionized water that contained 60 MBq of 18F-FDG. The total absorbed dose in the Fricke solution was assessed at 1450 min. The dose was correlated to the activity measured by PET for the same solution. The characteristics of the LabPET8™ scanner (Gamma Medica) was previously described in [29]. During the calibration procedure, PET imaging was performed at time 0, 0.5, 1, 2, 3, and 4 h after adding 18F-FDG into the Fricke solution, with scanning times of 3, 5.12, 9.50, 16.24, and 30.06 min, respectively. The optical density and radioactivity were measured prior to and after the PET scans. CT imaging of the vials was performed for attenuation correction of the emission data. The raw data were reconstructed and corrected relative to the reconstructed resolution of the PET scanner. The decay, dead time, random subtraction, and differences in crystal detection efficiencies were also included in the correction factor. A region of interest (ROIs) analysis was carried out with the built-in function in the LabPET image analysis software. The radioactivity in the subject vial (Fisherbrand 15 × 45 mm, 1DR, Fisher Scientific) was obtained as cps/mL from reconstructed PET images. The relationship of absorbed dose (Gy) and time-integrated activity (MBq.h) with administered activity (MBq) is shown in the Additional file 1: Figure S1 and Additional file 2: Figure S2, respectively.
To obtain quantitative radioactivity data with mice, the PET system was calibrated by acquiring data from a mouse phantom filled with an 18F-FDG solution of known radioactivity. Thus, the pixel counts of the PET image in cps/mL could be converted into the activity concentration (MBq/mL) by multiplying the ROIs with known added activity of 18F-FDG. Total accumulated absorbed dose in the tumor tissue and normal organs can be calculated by following Eq. 1.
$$ D\left(\mathrm{Gy}\right)=\breve{A}\left(\mathrm{MBq}.\mathrm{h}/\mathrm{g}\right)\times M\ \left(\mathrm{g}\right)\times C\left(\mathrm{Gy}/\mathrm{MBq}.\mathrm{h}\right) $$
(1)
where:
Ă is the time-integrated activity per gram of tissue (MBq.h/g)
M is the tissue mass (g)
D is the absorbed dose (Gy)
C is the conversion factor of 0.09 Gy/MBq.h, derived from the relationship between Fricke dosimetry and PET imaging (Additional file 3: Conversion factor for absorbed dose estimated by the Fricke chemical primary standard dosimeter).
Prostaglandin E2 quantification by liquid chromatography/tandem mass spectrometry
PGE2 has been quantified to assess inflammation [30]. Muscle tissues nearby the irradiated area were extracted and snapped frozen with liquid nitrogen after 4 h of either i.t. injection of 5 MBq 18F-FLT or 15 Gy gamma irradiation. Tissues were homogenized with a Dounce homogenizer in 2 mL of acetone-saline solution (2:1), containing 10 ng of the internal standard prostaglandin E2d4 (PGE2-d4), which contains four deuterium atoms at the 3, 3′,4, and 4′ positions (internal standard, Cayman Chemical, Ann Arbor, MI, USA) and 0.05% butylated hydroxytoluene to prevent the oxidation of prostanoids. The homogenate was transferred to a screw-top tube, vortexed for 1 min, and centrifuged (10 min, 1800 g, room temperature). The supernatant was transferred to another tube and mixed with 2 mL hexane by vortexing for 1 min. After centrifugation (10 min, 1800 g, room temperature), the upper phase containing lipids was discarded. The lower phase was acidified with 30 μL of 2 M formic acid and then 2 mL of chloroform containing 0.05% butylated hydroxytoluene were added. The mixture was vortexed and again centrifuged (10 min, 1800 g, room temperature) to separate the two phases. The lower phase containing chloroform was transferred to a conical centrifuge tube for evaporation with a SpeedVac Concentrator (Sarant, Nepean, ON, Canada). Samples were reconstituted in 100 μL methanol:10 mM ammonium acetate buffer, pH 8.5 (70:30), and filtered with Spin-X centrifuge tube filter 0.45 μm (10 min, 1300 g, room temperature). Samples were stored at − 20 °C for later analyses.
PGE2 was quantified by LC/MS/MS using the same procedure as reported by Desmarais et al. [31]. Briefly, the apparatus consisted of an API 3000 mass spectrometer (Applied Biosystem, Streesville, ON, Canada) equipped with a Sciex turbo ion spray (AB Sciex, Concord, ON, Canada) and a Shimadzu pump and controller (Columbia, MD, USA). Prostaglandins were chromatographically resolved using a Kromasil column 100-3.5C18, 150 × 2.1 mm (EKa Chemicals, Valleyfields, QC, Canada). A linear acetonitrile gradient from 45 to 90% during 12 min at a flow rate of 200 μL/min was used. The mobile phase consisted of water buffered with 0.05% acetic acid and acetonitrile 90% with acetic acid 0.05%. The injection volume was 10 μL per sample, which were kept at 4 °C during analysis. Individual products were detected using negative ionization and the monitoring of the transition m/z 351 ➔ 271 for PGE2 and 355 ➔ 275 for PGE2d4 with a collision voltage of − 25 V. For quantification of specific ions, the area under the curves was measured.
Mitotic activity assessed by immunohistochemistry of Ki67
Animals were euthanized 4 h after combined i.t. treatment with 5FU and 5 MBq 18F-FLT or 15 Gy gamma irradiation. Tumor samples were removed and fixed in 10% buffered formalin. The 5-μm sections from paraffin-embedded blocks were stained with conventional hematoxylin-eosin.
For Ki67 staining, 5-μm sections from paraffin-embedded blocks were deparaffinized in xylene, rehydrated using graded alcohol, and washed with PBS buffer (pH 7.4). For antigen retrieval, sections were placed in 0.01 M sodium citrate buffer (pH 6.0) for 10 min inside a steamer cooker. Sections were cooled to room temperature and washed with PBS buffer. Endogenous peroxide was blocked by 3% H2O2 for 15 min. Sections were incubated in 10% bovine serum albumin (BSA) for 1 h at room temperature. Thereafter, sections were incubated in overnight at 4 °C with the primary mouse monoclonal antibody Ki67 diluted in 0.5% BSA (PM375 AA Biocare Medical, Concord, California, USA). Sections were treated with the second anti-rabbit antibody (PM375 AA Biocare Medical, Concord, California, USA) diluted in 0.5% BSA for 1 h at room temperature. Diaminobenzidine tetrahydrochloride (0.6 mg/mL in Tris buffer saline, pH 7.6 containing 0.04% hydrogen peroxide) was used to develop the brown color. Methyl green was used to counterstain the slides. A negative control (with primary antibody omitted) was taken along with each batch. Counting of Ki67-positive cells was carried out in ten consecutive fields of 20×. The Ki67 index was estimated by the percentage of Ki67-positive cells in all the counted tumor cells.
Statistical analysis
All statistical analyses were performed using Prism 7.03 for Windows (GraphPad software). All results are reported as mean ± SD. The number of animals ranged from 2 to 4: (control untreated, i.t. 5FU, 15 Gy EBRT, i.t. 5FU + 15 Gy, i.t. 5FU + i.t. 18F-FLT 15 MBq, i.t. 5FU + i.t. 18F-FDG 15 MBq, n = 4), (PET i.t. 18F-FLT, i.t. FLT, i.t. 18F-FDG 15 MBq, untreated distant tumor (i.t. 18F-FLT 10 MBq), n = 3), and (PET i.t. 18F-FDG, i.t. 18F-FLT 15 MBq, i.t. 18F-FLT 25 MBq, n = 2). Statistical analyses were performed as described in the figures and table legends. Ordinary one-way ANOVA with a Dunnett’s multiple comparisons test was used to compare the residence time of 18F-FLT in the infused tumor to that of non-target tissues and to compare the 5Td of the non-treated animal to that of the different experimental groups. Ordinary one-way ANOVA with a Tukey’s multiple comparison test was used to compare the 5TD values of the (i.t. 5FU + 15 Gy EBRT) to those of the (i.t. 5FU + i.t. 18F-FLT 15 MBq) and (i.t. 5FU + i.t. 18F-FDG 15 MBq) groups. Differences were considered statistically significant at p ≤ 0.05.