S. Serfling, Y. Zhi, A. Schirbel, T. Lindner, T. Meyer, E. Gerhard-Hartmann, C. Lappa, R. Hagen, S. Hackenberg, A. K. Buck, A. Scherzad

• Purpose In cancer of unknown primary (CUP), positron emission tomography/computed tomography (PET/CT) with the glucose analog [\(^{18}\)F]FDG represents the standard imaging approach for localization of the malignant primary. Frequently, however, [\(^{18}\)F]FDG PET/CT cannot precisely distinguish between small occult tumors and chronic inflammation, especially in Waldeyer’s tonsillar ring. To improve the accuracy for detecting primary tumors in the Waldeyer’s tonsillar ring, the novel PET tracer [\(^{68}\)Ga]Ga-FAPI-4 for specific imaging ofPurpose In cancer of unknown primary (CUP), positron emission tomography/computed tomography (PET/CT) with the glucose analog [\(^{18}\)F]FDG represents the standard imaging approach for localization of the malignant primary. Frequently, however, [\(^{18}\)F]FDG PET/CT cannot precisely distinguish between small occult tumors and chronic inflammation, especially in Waldeyer’s tonsillar ring. To improve the accuracy for detecting primary tumors in the Waldeyer’s tonsillar ring, the novel PET tracer [\(^{68}\)Ga]Ga-FAPI-4 for specific imaging of fibroblast activation protein (FAP) expression was used as a more specific target for cancer imaging. Methods Eight patients with suspicion of a malignant tumor in Waldeyer’s tonsillar ring or a CUP syndrome were examined. PET/CT scans with [\(^{18}\)F]-FDG and [\(^{68}\)Ga]Ga-FAPI-4 were performed for pre-operative tumor localization. After surgical resection, histopathological and immunohistochemical results were compared to PET/CT findings. Results Histopathology revealed a palatine or lingual tonsil carcinoma in all patients. In case of lymph node metastases smaller than 7 mm in size, the [\(^{18}\)F]FDG PET/CT detection rate of cervical lymph node metastases was higher than that of [\(^{68}\)Ga]FAPI PET/CT, while both tracers identified the primary tumors in all eight cases. The size of the primary and the lymph node metastases was directly correlated to the respective FAP expression, as detected by immunohistochemistry. The mean SUVmax for the primary tumors was 21.29 ± 7.97 for \(^{18}\)F-FDG and 16.06 ± 6.29 for \(^{68}\)Ga-FAPI, respectively (p = 0.2). The mean SUVmax for the healthy contralateral tonsils was 8.38 ± 2.45 for [\(^{18}\)F]FDG and 3.55 ± 0.47 for [\(^{68}\)Ga]FAPI (p < 0.001). The SUVmax ratio of [68Ga]FAPI was significantly different from [\(^{18}\)F] FDG (p = 0.03). Mean TBRmax for the [\(^{68}\)Ga]Ga-FAPI-4 tracer was markedly higher in comparison to [\(^{18}\)F]FDG (10.90 vs. 4.11). Conclusion Non-invasive imaging of FAP expression by [\(^{68}\)Ga]FAPI PET/CT resulted in a better visual detection of the malignant primary in CUP, as compared to [\(^{18}\)F]FDG imaging. However, the detection rate of lymph node metastases was inferior, presumably due to low FAP expression in small metastases. Nevertheless, by offering a detection method for primary tumors with the potential of lower false positive rates and thus avoiding biopsies, patients with CUP syndrome may benefit from [\(^{68}\)Ga]FAPI PET/CT imaging.…