Preparation of 64Cu-labeled cetuximab
Production of 64Cu was performed as previously described . Cetuximab obtained from Merck Serono (Darmstadt, Germany) was 64Cu-labeled with 3,6,9,15-tetraazabicyclo[9.3.1]pentadeca-1(15),11,13-triene-3,6,9-triacetic acid (PCTA; Macrocyclics, Plano, TX, USA) as the chelator, which had previously been found to result in high radiolabeling yield and in vitro serum stability [22, 23]. 64Cu-PCTA-cetuximab was synthesized using previously reported methods  with a specific activity of 1.7 GBq/mg. The injected protein dose of 64Cu-PCTA-cetuximab was adjusted to 20 µg per mouse by adding an unlabeled antibody as reported previously .
Cell culture and mouse model
Human pancreatic cancer xPA-1 cells expressing red fluorescent protein (RFP) in the cytoplasm and green fluorescent protein in the nucleus (xPA-1-dual-color [xPA-1-DC]; AntiCancer, San Diego, CA, USA) with EGFR overexpression  were used in this study. xPA-1-DC cells were cultured in RPMI-1640 medium (Wako, Osaka, Japan) supplemented with 10% fetal bovine serum in a humidified atmosphere of 95% air and 5% CO2 at 37°C.
All animal experimental procedures were approved by the Animal Ethics Committee of the National Institutes for Quantum and Radiological Science and Technology (QST, Chiba, Japan) and conducted in accordance with the institutional guidelines. Six-week-old female BALB/c nude mice were obtained from Japan SLC (Shizuoka, Japan) and used in this study. Before the experiments, the mice were acclimated for at least 1 week. This study used a mouse model with intrapancreatic cancer xenografts generated by inoculating xPA-1-DC cells into the pancreas. After an abdominal incision, 5 × 106 xPA-1-DC cells in 25 µL RPMI-1640 medium mixed with 25 µL ice-cold Matrigel (BD Biosciences, Franklin Lakes, NJ, USA) were slowly injected into the pancreas. One week later, the model mice were used for the experiments. Prior to surgery experiments, the tumor location in the established model was confirmed using a stereoscopic fluorescence microscope (MZ16F; Leica, Wentzler, Germany).
OpenPET-guided surgery and its efficacy
The model mice were randomized into two groups: OpenPET-guided surgery group and control group (n = 7 per group). For the OpenPET-guided surgery group, 64Cu-PCTA-cetuximab (7.4 MBq) were ip administered 24 h before surgery. The OpenPET-guided surgery (i.e., day 0) was performed according to previously reported procedures and settings . Briefly, this study used a prototype of the OpenPET system, which was developed previously for use in small animal experiments [11, 12], containing 32 detector blocks (4-layer depth-of-interaction detectors with 16 × 16 × 4 crystals each) in a cylinder with a diameter of 25 cm. The detectors are axially shifted incrementally to form an accessible open space (14-cm wide) for surgical procedures in mice. In this system, the field of view with its cylindrical shape defined by two parallel planes slanted at 45° relative to the axial direction was 11.4 cm in diameter with a 10.2-cm axial length, and a spatial resolution is approximately 2 mm [11, 12]. This system used a 1-pass list-mode dynamic row-action maximum-likelihood algorithm with a graphics processing unit for high-speed reconstruction that enabled image updating in cycles of < 1 s while accumulating list-mode data. Reconstructed OpenPET images were represented as sliced images of the transaxial, coronal, and sagittal planes on the screen in front of the surgeon. These images were displayed as radioactivity-density values (kBq/mL) based on the calibration with standards having known radioactivities. During surgery, the mice remained under 2% isoflurane anesthesia, and their body temperatures were maintained with a heater. Real-time OpenPET images were acquired to identify the tumor location before surgery. The laparotomy was conducted by cutting the abdominal wall and skin, and the pancreas with the primary tumor was exteriorized and observed with OpenPET. While monitoring the position of the tumor within the pancreas in real-time using the OpenPET, detected tumors were resected after ligation with a clamp to prevent bleeding. The presence or absence of remaining signals was intraoperatively checked using real-time OpenPET imaging. If signals were detected, additional resections for residual tumor tissue were reconducted until the signals disappeared. After confirming the absence of signals from residual tumors, the pancreas was returned to the abdomen, and the peritoneum and skin were closed using surgical sutures. After the OpenPET-guided surgery, resected tissues were examined with a stereoscopic fluorescence microscope to verify tumor RFP signals. In the control group (on day 0), similar surgical procedures with incisions and anesthesia were conducted to exclude any effects of the operation itself. To examine the benefits of OpenPET guidance over the conventional surgical method, tumor resection without OpenPET guidance, i.e., with the naked eye alone, was also attempted. Mice of both groups were weighed and observed for 50 days after the surgery. Mice were also sacrificed when reaching a humane endpoint defined as a noticeable extension of the abdomen, development of ascites, or body weight loss (> 20%). On day 7 after the surgery, tumor growth was observed using in vivo fluorescent imaging with an IVIS Lumina imaging system (PerkinElmer, MA, US) to detect tumor RFP signals.
Uptake of 64Cu-PCTA-cetuximab into resected tumor specimens
The accumulation of 64Cu-PCTA-cetuximab into the resected specimens was evaluated after the surgery. The resected tumor specimens were weighed, and radioactivity levels were measured with a γ-counter (1480 Automatic gamma counter Wizard 3; PerkinElmer). The values of the percentage of injected dose per gram (%ID/g) were calculated.
Data are expressed as the mean and standard deviation. Differences in survival were evaluated using the log-rank test. P values < 0.05 were considered statistically significant.