This study showed that pretargeted immuno-PET is a very sensitive imaging modality to detect CEA-expressing tumor lesions in an orthotopic mouse model. The intraperitoneal tumors were clearly delineated with a high tumor-to-background contrast, providing high sensitivity: all tumor lesions ≥ 10 μL were detected with this method at a very good confidence rate. The smallest lesions that were detected had a volume as low as 5 to 8 μL, which is in the same range of the spatial resolution of the dedicated animal PET scanner.
The animal model used in this study was well characterized by Koppe et al . The human colon carcinoma cell line LS174T has a reproducible growth pattern in BALB/c nude mice after intraperitoneal injection. Three weeks after tumor cell inoculation, small tumor nodules were observed in the rectovesical pouch, the mesentery, and the subhepatic, -splenic, and -phrenic spaces. The preclinical model mimics peritoneal disease of patients with metastasized colorectal cancer [35, 36].
In a previous imaging study, we demonstrated the feasibility of pretargeted immuno-PET using 68Ga- or 18F-labeled di-HSG peptides in mice with subcutaneous tumors . In the current study, the activity concentration of the 68Ga-labeled IMP288 in the intraperitoneal tumors was similar to that in the subcutaneous tumors [32, 37]. In our intraperitoneal tumor model, the variation in tumor size was much wider than that in the subcutaneous model. Our biodistribution results showed an inverse relationship between tumor weight and activity concentration. This correlation corresponds with the findings of other investigators [38–41]. Sharkey et al. showed specific uptake of 124I-labeled peptide after pretargeting with TF2 in microdisseminated human colon cancer colonies in the lungs of nude mice. In that model, high tumor-to-non-tumor ratios were obtained, illustrating the excellent tumor targeting potential of the pretargeting strategy .
FDG-PET/CT has shown high sensitivity and negative predictive value in diagnosing CRC [43, 44]. Therefore, it was used in the present study as a reference method. The imaging quality of FDG-PET in this preclinical study was optimized by minimizing uptake of FDG in other organs by anesthesia, fasting, and warming of the animals . Its uptake in the myocardium, brain, intestines, and liver is comparable to the clinical situation. The ratios between normal and tumor tissues might have appeared to be less favorable than in patients, which might have compromised the detection of the tumors.
Based on our preclinical results, we feel that pretargeted immuno-PET can be of additive value in the clinical setting. When staging patients with primary tumors in the detection of eventual metastases, a highly sensitive and specific imaging method is required. Furthermore, in patients to be screened prior to curative liver metastasectomy, the disclosure of occult extrahepatic lesions will prevent useless operations. More so, immuno-PET can help select patients who could undergo radioimmunotherapy. As the pretargeting system with the DOTA-conjugated peptides is very flexible, it can be labeled with a broad variety of radionuclides, such as 90Y and 177Lu for pretargeted radioimmunotherapy, or with 111In and 99 mTc for SPECT imaging. Our preclinical results show similar biodistribution of the 111In/177Lu- or 68Ga-labeled peptide . Images about targeting known, non-biopsied lesions can confirm antigen expression and accessibility of the therapeutic dose. Information on the biodistribution and pharmacokinetics can help adjust treatment regimes by providing dosimetry data. This could be used to optimize dosing and to avoid toxicities.
For clinical application, 68Ga has some major advantages. It is readily available in a nearly carrier-free state from an in-house 68Ge/68Ga generator. IMP288-DOTA can be stably and rapidly labeled with 68Ga. Its half-life matches the pharmacokinetics of the peptide. In the present study, the positron range of 68Ga (median range, 3.5 mm) might have limited image resolution. Visser et al.  showed that with the intrinsic spatial resolution (approximately 1.5 mm) of our state-of-the-art, small-animal PET scanner, the finite positron range has become the limiting factor for the overall spatial resolution and activity recovery in small structures imaged with 68Ga. Combined with the partial volume effect, this could explain the lower detection rate of the smallest tumor lesions with pretargeted immuno-PET despite the higher radioactivity concentration of TF2 and 68Ga-IMP288 in the smaller tumors.
Due to the flexibility of the di-HSG peptides, the use of other PET radionuclides for this pretargeting system can be explored. 18F, the most widely used positron-emitting radioisotope, would be suitable due to its short positron range in the tissue (0.62 mm), which might increase the image resolution. McBride and, subsequently, Laverman et al. developed an innovative and rapid method for labeling peptides with 18F based on a metal chelator [46, 47]. The biodistribution and PET images in the subcutaneous LS174T tumors in the nude mice showed the feasibility of this approach . Translation of this preclinical imaging method to the clinical situation will show the effect of the intrinsic resolution of the clinical PET scanner in combination with the spatial resolution of the radionuclide.