Fibroblast activation protein expression in the tumor microenvironment is crucial in survival prediction and differentiation of recurrent gliomas: a head-to-head comparison of 68Ga-FAPI-04 and 18F-FET in PET/CT imaging

The precise differentiation of recurrent glioma from treatment-related changes, such as pseudoprogression or radiation necrosis, is essential for treatment planning and remains a significant challenge. Fibroblast activation protein (FAP) expressed by cancer-associated fibroblasts can be targeted with PET tracers for in vivo visualization and quantification. This study aims to evaluate the diagnostic and prognostic effectiveness of FAP expression in patients with potential recurrent glioma by directly comparing gallium-68 FAP inhibitor-04 and fluorine-18 fluoroethyl-L-tyrosine PET/CT imaging. Thirty glioma patients showing signs of possible recurrence during routine MRI follow-up after treatment were enrolled. PET-based semiquantitative parameters, clinical factors, and survival data were collected for analysis. Paired comparison of SUVmax, TBRmax, MTV, and TLU originating from two PET imaging studies indicated significant differences in TBRmax, MTV, and TLU, with P values of 0.000, 0.001, and 0.000, respectively. Univariate logistic regression analysis revealed a marginally non-significant difference in efficacy (P = 0.053) of the initial pathological diagnosis. In multivariate logistic regression analysis, PET parameters, initial pathological data, age, and gender were used to develop the predictive models step by step. Although trends towards significance were observed in the MTVFAPI:MTVFET ratio, no PET parameters reached statistical significance. The MTVFAPI:MTVFET ratio improved the area under the receiver operating characteristic curve (AUC). When PET parameters and initial pathological diagnosis were included, the MTVFAPI:MTVFET ratio significantly enhanced the model's AUC (P = 0.040) from 0.709 (0.465-0.953, 95% CI) to 0.847 (0.688-1.000, 95% CI). When replacing the initial diagnosis with initial WHO grade, the MTVFAPI:MTVFET ratio improved the AUC (P = 0.016) from 0.640 (0.400-0.880, 95% CI) to 0.852 (0.715-0.988, 95% CI). After factoring in age and gender in addition to the initial pathological diagnosis, the MTVFAPI:MTVFET ratio enhanced the AUC (P = 0.039) from 0.841 (0.677-1.000, 95% CI) to 0.963 (0.887-1.000, 95% CI). Similarly, after replacing the initial pathological diagnosis with the initial WHO grade, the MTVFAPI:MTVFET ratio significantly enhanced the AUC of the model (P = 0.046) from 0.762 (0.532-0.992, 95% CI) to 0.942 (0.850-1.000, 95% CI). The survival analysis revealed that the MTV-FAPI of the lesion has a significant impact on overall survival (P = 0.027, hazard ratio = 1.103, 95% CI: 1.011-1.204). This head-to-head exploratory study showed that glioma FAP expression volume is an independent risk factor that can significantly influence overall survival in patients with recurrent glioma. Although statistically insignificant, the MTVFAPI:MTVFET ratio, which represents the FAP expression volume percentage of the post-treatment glioma tissue, suggests trends toward significant differentiation between glioma recurrence and treatment-related changes.