Reversibility of ⁶⁸Ga-FAPI-2 Trapping Might Prove an Asset for PET Quantitative Imaging

TO THE EDITOR: In a recent paper, Giesel et al. analyzed the tissue biodistribution and preliminary dosimetry of 2 quinoline-based PET tracers that act as fibroblast activation protein (FAP) inhibitors, namely, ⁶⁸Ga-FAPI-2 and ⁶⁸Ga-FAPI-4 (1). The authors reported a fast clearance via the kidneys, a low tracer uptake in normal organs, equal tumor-to-background contrast ratios at 1 h after injection, and an almost equal uptake in comparison with ¹⁸F-FDG. However, from 1 to 3 h after injection, in contrast to ⁶⁸Ga-FAPI-4, which displayed a prolonged tumor retention (25% washout), ⁶⁸Ga-FAPI-2 tumor uptake decreased by 75%, thus reflecting release of the tracer from the malignant tissue. This difference explains why a companion paper by Kratochwil et al. used ⁶⁸Ga-FAPI-4 for identifying the most promising indications for future application (2).

We suggest that ⁶⁸Ga-FAPI-2 trapping reversibility, evidenced by a decrease in tumor uptake observed at late imaging, might prove an asset for PET quantitative imaging. Figure 2, by Giesel et al., shows ⁶⁸Ga-FAPI-2 and ⁶⁸Ga-FAPI-4 maximal SUV (SUV_max) at 10, 60, and 180 min after injection in 2 patients with metastasized breast cancer, respectively (1). Because the 2 tracers have rapid clearance from blood, we assume their input function (IF) has become negligible at 60, and, a fortiori, at 180 min after injection. Thus, a previously published method designed for ¹⁸F-FDG may be adapted to ⁶⁸Ga-FAPI-2 and ⁶⁸Ga-FAPI-4 for assessing their release rate k_B (in min⁻¹; Eq. 3 in Laffon et al. (3)). For the sake of clarity, let us assume an IF monoexponential decay with decay-corrected time constant α and initial amplitude A_P(t = 0) (in min⁻¹ and kBq.mL⁻¹, respectively). The decay-corrected tissue activity concentration related to trapped tracer (in kBq/mL), which is proportional to SUV_max, can be approximated from 60 to 180 min after injection, by:Eq. 1where Ki is the uptake rate constant of the tracer (in mL.min⁻¹.mL⁻¹). Fitting the outer extreme metastasis data (extracted with the WebPlotDigitizer software) at 60 and 180 min after injection in Figure 2 with a monoexponentially decaying function leads to the following range for k_B: 0.01435–0.01439 and 0.00129–0.00212 min⁻¹ for ⁶⁸Ga-FAPI-2 and ⁶⁸Ga-FAPI-4, respectively. For comparison, k_B for ¹⁸F-FDG trapping in the normal human liver has been estimated to be 0.00650 min⁻¹ on average (4). It is noteworthy that, because only 2 time points were analyzed and only 1 patient per tracer was examined in Figure 2 by Giesel et al., the assessment of k_B measurement uncertainty is out of the scope of the current paper (1). Therefore, additionally to SUV, we suggest that one could take advantage of the significant ⁶⁸Ga-FAPI-2 trapping reversibility to better characterize tumors by means of calculating k_B. Furthermore, the above-proposed fitting of k_B might be easily performed at the voxel level, thus allowing parametric imaging of tracer release. Finally, let us note that a multiexponentially decaying IF does not alter the current line of argument.

To conclude, ⁶⁸Ga-FAPI PET/CT is a promising new diagnostic method for imaging various cancers that overexpress FAP (1). We suggest that the choice between ⁶⁸Ga-FAPI-2 and ⁶⁸Ga-FAPI-4 should not be only based on the criterion of reversible versus irreversible (or nearly) trapping of the tracer, even if the latter is an indubitable advantage for a theranostic purpose. Indeed, one could also take advantage of the significant trapping reversibility of ⁶⁸Ga-FAPI-2 to better characterize malignant tissues. Furthermore, we suggest that performing both uptake and release quantitation of ⁶⁸Ga-FAPI-2 trapping might be an innovative tool for assessing the response to treatment.

• © 2020 by the Society of Nuclear Medicine and Molecular Imaging.