In vivo bioluminescence tomography-guided radiation research platform for pancreatic cancer: an initial study using subcutaneous and orthotopic pancreatic tumor models

Abstract

Genetically engineered mouse model(GEMM) that develops pancreatic ductal adenocarcinoma(PDAC) offers an experimental system to advance our understanding of radiotherapy(RT) for pancreatic cancer. Cone beam CT(CBCT)-guided small animal radiation research platform(SARRP) has been developed to mimic the RT used for human. However, we recognized that CBCT is inadequate to localize the PDAC growing in low image contrast environment. We innovated bioluminescence tomography(BLT) to guide SARRP irradiation for in vivo PDAC. Before working on the complex PDAC-GEMM, we first validated our BLT target localization using subcutaneous and orthotopic pancreatic tumor models. Our BLT process involves the animal transport between the BLT system and SARRP. We inserted a titanium wire into the orthotopic tumor as the fiducial marker to track the tumor location and to validate the BLT reconstruction accuracy. Our data shows that with careful animal handling, minimum disturbance for target position was introduced during our BLT imaging procedure(<0.5mm). However, from longitudinal 2D bioluminescence image(BLI) study, the day-to-day location variation for an abdominal tumor can be significant. We also showed that the 2D BLI in single projection setting cannot accurately capture the abdominal tumor location. It renders that 3D BLT with multiple-projection is needed to quantify the tumor volume and location for precise radiation research. Our initial results show the BLT can retrieve the location at 2mm accuracy for both tumor models, and the tumor volume can be delineated within 25% accuracy. The study for the subcutaneous and orthotopic models will provide us valuable knowledge for BLT-guided PDAC-GEMM radiation research.

Keywords: bioluminescence tomography; image-guided radiation therapy; pancreatic cancer; small animal irradiator.

Figure 1.

The (a) photo and (b) schematic of MuriGlo; the BL signal emitted from mouse is directed by a motorized 3-mirror and then a stationary mirror to the filter and lens-camera setup. The filters enable multi-spectral BLI acquisition. The 3-mirror system can rotate 360° around animal and capture multi-projection images. (c) The photo of transportable mouse bed; the grid opening is designed for acquisition of BL signal at 180°. Abbreviation: S = superior, P = posterior, R = right for a mouse at supine position placed on the mouse bed.

Figure 2.

The surgical steps of orthotopic pancreatic tumor implantation; (a) a fraction of tumor tissue at approximately 3 mm in diameter was cut from a subcutaneous tumor. (b) The openings of skin and peritoneum were made on the left flank of mouse in sequence. (c) The spleen and pancreas were gently exteriorized, and the pancreas was spread out on the skin. (d) A small pocket was created in the pancreatic parenchyma. (e) The tumor fraction was inserted into the pancreas pocket. (f) A 7-0 suture was used for stitching the tumor fraction firmly inside the pancreas pocket and closing the pocket. (g) The spleen and pancreas were placed back into the abdominal cavity. (h) The incisions on the peritoneum and skin were closed in sequence.

Figure 3.

Mouse was proceeded with BLT at the 3^rd week after cell injection. We used four wavelengths (590, 610, 630, and 650 nm) for the reconstruction, and applied 10% threshold to determine the BLT volume. The BLT-reconstructed volume was overlapped on SARRP CBCT image at (a) coronal, (b) sagittal, and (c) transverse views.

Figure 4.

SARRP CBCT images at VD view (a) before and (b) after the BLI acquisition; the titanium wire (orange arrow) centered at the orthotopic pancreatic tumor was used as the marker to identify the tumor location.

Figure 5.

The unfiltered BLIs of three mice (a1-8, #1; b1-8, #2; c1-8, #3) at 0° projection; The images are arranged in the order of (1) 0.5, (2) 1, (3) 1.5, (4) 2, (5) 3, (6) 4, (7) 5, and (8) 7 weeks after the tumor implantation. The intensity of BL signal increased and the area of BL pattern expanded overtime. Because the pancreatic tumor could move to different positions around the abdomen region, it would affect the intensity and the pattern of BL signal at the 0° projection image (b5-b6, and c3-c6). The artifact due to strong BLI signal is seen in Figure 5c8.

Figure 6.

The unfiltered BLIs of Mouse #2 at (a) 3 and (b) 4 weeks after the tumor implantation at (a1 and b1) 0 and (a2 and b2) 180° projections; it clearly shows that the tumor can move to different locations and multi-projection BLT is necessary to quantify the tumor location in 3D.

Figure 7.

Initial study of BLT reconstruction for the orthotopic pancreatic tumor model (mouse #1 1.5 weeks after the tumor implementation); (a) multi-projection BLI at 630 nm mapped on the mesh of mouse body; (b) VD view of the mouse body mesh overlapped with the BLT-reconstructed tumor volume and titanium wire; we used the data of three wavelengths (610, 630, and 650 nm) for the reconstruction, and applied a threshold of 60% maximum to determine the BLT volume. The titanium wire was used for identifying the tumor location.