Parametric mapping of dynamic 68Ga FAPI-46-PET data of 42 patients with pancreatic lesions

Le résumé fourni par la source

Abstract Purpose Differential diagnoses of primary pancreatic lesions include pancreatic ductal adenocarcinomas (PDAC) and inflammatory lesions of the pancreas (ILP). Post-pancreatic surgery, differentiation of postoperative reactive tissue (PRT) and PDAC-recurrence challenges oncological imaging. Static 68Ga-FAPI-PET/CT uptake is increased in all of these lesions with marked overlap in signal intensity, hampering their FAPI-PET-based assessment. Here, we evaluated static and parametric imaging parameters for discrimination of pancreatic lesions in primary and post-pancreatic surgery scenarios. Methods 55 Patients with pancreatic lesions (36 primary, 19 post-pancreatic surgery) underwent static and dynamic 68Ga-FAPI-46-PET/CT. Primary lesions were classified either by histology following PET/CT or follow-up (> 6 months). Post-surgery, PRT and PDAC-recurrence were classified by CT- and clinical course (> 18 months). Parametric maps (1 tissue compartment (1TC), 2TC and Logan plot (LP)) from dynamic PET-data were generated via image-based aortic input function using PMOD-software. Pancreatic lesions (PDAC, ILP, PRT, PDAC-recurrence) were then delineated using VOI-technique (30–70% isocontour) and signal intensities were analyzed. SPSS was used to detect outliers, unpaired t-tests was applied for comparison of static and parametric imaging parameters. Receiver-operating-characteristic curves for differentiating PDAC/ILP or recurrent PDAC/PRT were generated. Results 42 patients were included in the final analysis: in primary setting, 16 PDAC and 10 ILP; in post-surgery setting 9 PDAC-recurrences and 7 PRT. In the primary setting, although PDAC showed higher SUVmax/mean than ILP, no significant differences in maximum/mean signal values neither in static imaging nor in parametric maps were detected. With regard to the differentiation of PDAC-recurrences versus postoperative tissue, LPmax were significantly higher in PDAC-recurrences compared to PRT (4.74 vs. 2.40, p-value 0.020) with AUC 82.5% (95-CI 0.62-1.0) and a possible diagnostic threshold at > 3,49 (LR + 5.44), while differences in static imaging or other parametric maps were not statistically significant. Conclusion Differentiating pancreatic lesions remains challenging. While LPmax significantly distinguished PDAC-recurrence from PRT, other parametric mapping parameters yielded no significant results. Larger studies and additional dynamic data analysis methods should be explored.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.