Diagnostic performance of [¹⁸F]-FDG or [⁶⁸Ga]-FAPI PET/MRI for nodal assessment in colorectal cancer: a retrospective comparative analysis.

Abstract

PurposeThe evaluation of lymph node metastasis in colorectal cancer (CRC) remains a critical challenge, as it significantly impacts prognosis and treatment decisions. PET/MRI with various tracers may take advantages of MRI together with the functional imaging of PET. This study aims to evaluate the diagnostic performance of PET/MRI for lymph node metastasis of CRC in a real-world clinical cohort and to explore the imaging characteristics of PET/MRI using [¹⁸F]-FDG or [⁶⁸Ga]-FAPI tracers.MethodsThis retrospective study included 51 patients with confirmed colorectal cancer who underwent PET/MRI imaging ([¹⁸F]-FDG or [⁶⁸Ga]-FAPI) between March 2020 and December 2022. The choice of PET tracer was determined according to clinical indication and tracer availability. The N stage and lesion status (N0 vs. N + and positive node vs. negative node) were independently evaluated by two nuclear medicine physicians. ROC curve analysis was performed using histopathological results and clinical follow-up as the reference standards.ResultsThe accuracy for nodal (N) status assessment was 91.4% for PET/MRI and 85.7% for MRI. PET/MRI showed numerical higher diagnostic performance than MRI for detecting lymph node metastases (AUC 0.88 vs. 0.81; p = 0.19). At the region-based lymph node level, PET/MRI demonstrated significantly higher accuracy (93.6% vs. 87.8%, p = 0.002), specificity (98.9% vs. 92.5%, p = 0.01), and positive predictive value (85.7% vs. 40.0%, p = 0.002). In the exploratory analysis of [⁶⁸Ga]-FAPI subgroup, no false-positive regional findings were observed, corresponding to a specificity of 100% (50/50; 95% CI, 92.9%-100%). It also showed numerical higher SUVmax and TBR compared to [¹⁸F]-FDG PET/MRI, although the differences were not statistically significant (4.35 vs. 2.25, p = 0.64; 17.4 vs. 5.65, p = 0.29). Among 11 patients who underwent a watch-and-wait or local excision strategy under PET/MRI surveillance, 1 experienced local regrowth and 1 developed lung metastasis. No nodal recurrence was observed during follow-up.ConclusionPET/MRI demonstrated encouraging diagnostic accuracy and high specificity in evaluating lymph node metastasis in colorectal cancer. These findings support further investigation of PET/MRI as part of multimodal imaging strategies for CRC management, particularly in patients considered for organ preservation.