<sup>18</sup>F-AlF-NOTA-Octreotide Outperforms<sup>68</sup>Ga-DOTATATE/NOC PET in Neuroendocrine Tumor Patients: Results from a Prospective, Multicenter Study
Elin Pauwels, Frederik Cleeren, Térence Tshibangu, Michel Koole, Kim Serdons, Lennert Boeckxstaens, Jeroen Dekervel, Timon Vandamme, Willem Lybaert, Bliede Van den Broeck, Annouschka Laenen, Paul M. Clément, Karen Geboes, Eric Van Cutsem, Sigrid Stroobants, Chris Verslype, Guy Bormans, Christophe M. Deroose
Abstract
<sup>18</sup>F-labeled somatostatin analogs (SSAs) could represent a valid alternative to the current gold standard, <sup>68</sup>Ga-labeled SSAs, for somatostatin receptor imaging in patients with neuroendocrine tumors (NETs), given their logistic advantages. Recently, <sup>18</sup>F-AlF-NOTA-octreotide (<sup>18</sup>F-AlF-OC) has emerged as a promising candidate, but a thorough comparison with <sup>68</sup>Ga-DOTA-SSA in large patient groups is needed. This prospective, multicenter trial aims to demonstrate noninferiority of <sup>18</sup>F-AlF-OC compared with <sup>68</sup>Ga-DOTA-SSA PET in NET patients (ClinicalTrials.gov, NCT04552847). <b>Methods:</b> Seventy-five patients with histologically confirmed NET and routine clinical <sup>68</sup>Ga-DOTATATE (<i>n</i> = 56) or <sup>68</sup>Ga-DOTANOC (<i>n</i> = 19) PET, performed within a 3-mo interval of the study scan (median, 7 d; range, −30 to +32 d), were included. Patients underwent a whole-body PET 2 h after intravenous injection of 4 MBq/kg of <sup>18</sup>F-AlF-OC. A randomized, masked consensus read was performed by 2 experienced readers to count tumor lesions. After unmasking, the detection ratio (DR) was determined for each scan, that is, the fraction of lesions detected on a scan compared with the union of lesions of both scans. The differential DR (DDR; difference in DR between <sup>18</sup>F-AlF-OC and <sup>68</sup>Ga-DOTATATE/NOC) per patient was calculated. Tracer uptake was evaluated by comparing SUV<sub>max</sub> and tumor-to-background ratios in concordant lesions. <b>Results:</b> In total, 4,709 different tumor lesions were detected: 3,454 with <sup>68</sup>Ga-DOTATATE/NOC and 4,278 with <sup>18</sup>F-AlF-OC. The mean DR with <sup>18</sup>F-AlF-OC was significantly higher than with <sup>68</sup>Ga-DOTATATE/NOC (91.1% vs. 75.3%; <i>P</i> < 10<sup>−5</sup>). The resulting mean DDR was 15.8%, with a lower margin of the 95% CI (95% CI, 9.6%–22.0%) higher than −15%, which is the prespecified boundary for noninferiority. The mean DDRs for the <sup>68</sup>Ga-DOTATATE and <sup>68</sup>Ga-DOTANOC subgroups were 11.8% (95% CI, 4.3–19.3) and 27.5% (95% CI, 17.8–37.1), respectively. The mean DDR for most organs was higher than zero, except for bone lesions (mean DDR, −2.8%; 95% CI, −17.8 to 12.2). No significant differences in mean SUV<sub>max</sub> were observed (<i>P</i> = 0.067), but mean tumor-to-background ratio was significantly higher with <sup>18</sup>F-AlF-OC than with <sup>68</sup>Ga-DOTATATE/NOC (31.7 ± 36.5 vs. 25.1 ± 32.7; <i>P</i> = 0.001). <b>Conclusion:</b><sup>18</sup>F-AlF-OC is noninferior and even superior to <sup>68</sup>Ga-DOTATATE/NOC PET in NET patients. This validates <sup>18</sup>F-AlF-OC as an option for clinical practice somatostatin receptor PET.