First clinical investigation to predict lymphovascular and/or perineural invasion in gastric cancer using ¹⁸F-FAPI-42 PET/CT parameters.

Abstract

This study was conducted to explore the predictive value of PET parameters derived from ¹⁸F-FAPI-42 PET/CT in assessing lymphovascular and/or perineural invasion (LVI/PNI) in gastric cancer (GC) patients. 72 GC patients who underwent ¹⁸F-FAPI-42 PET/CT prior to surgical resection were included. Clinicopathological factors and PET parameters were collected and analyzed in LVI/PNI-negative and LVI/PNI-positive groups. The predictive value of PET parameters for LVI/PNI status was evaluated using the receiver operating characteristic (ROC) curve. A nomogram was developed using significant predictors from multivariate stepwise regression analysis and its performance was assessed by decision curve analysis (DCA). Univariate analysis indicated a significant association between LVI/PNI status and PET parameters (SUVmax, SUVmean, and TBR) (all p < 0.001). The area under the ROC curve (AUC) values for predicting LVI/PNI were 0.932 [95% CI (0.877-0.987)] for SUVmax, 0.923 [95% CI (0.861-0.984)] for SUVmean, and 0.925 [95% CI (0.865-0.985)] for TBR. The optimal cutoff values for prediction, along with their corresponding sensitivity and specificity, were 3.86 (93.3% and 81.5%) for SUVmax, 2.04 (93.3% and 81.5%) for SUVmean, and 9.75 (91.1% and 81.5%) for TBR. Multivariate analysis identified histological grade and SUVmax as independent risk factors for LVI/PNI prediction. Our nomogram had good discriminatory ability (AUC = 0.934) and offered net benefits in predicting LVI/PNI status by DCA. This study demonstrates that FAPI uptake parameters exhibit an exceptionally high capacity and serve as a noninvasive preoperative tool for predicting LVI/PNI status in GC, with SUVmax emerging as the most suitable predictive indicator.