Response to "Critique and Discussion of 'Multicenter Evaluation of Frequency and Impact of Activity Infiltration in PET Imaging, Including Microscale Modeling of Skin-Absorbed Dose'".

SPECT/CT-based dosimetry study is already in preparation. Consequently, we saw no need to report preliminary results, which soon will become obsolete anyway. The other reason is the challenging interpretation of dosimetry data in predicting clinical consequences of systemic radioligand therapy. The latter issue is now addressed by Pretze et al., who mention various physical and radiation biologic aspects of this theme. We appreciate receiving such an instructive letter, stimulating a fruitful academic discussion. First, Pretze et al. mention further radionuclides that could serve as alternatives to Lu. Indeed, the amount of Yb that is needed for the production of high-specific-activity Lu without Lu impurities (half-life, 161 d) is limited. Consequently, the costs for producing no-carrier-added Lu are relatively high. Routine availability of Tb and Cu is currently even worse than for Lu. In contrast, Re is readily available from a well-established generator system and, if generators are eluted regularly, converts it into reduced radionuclide costs by approximately a power of 10. The current shortage of Lu-PSMA-617 in the United States (Pluvicto from Novartis has been on the Food and Drug Administration’s shortage list since March 7, 2023) illustrates the logistic challenges of airfreight delivery even between well-developed countries. In regions with a lower airport density, just-in-time delivery of Lu (half-life, 6.7 d) radiopharmaceuticals is likely an illusion, and the same applies to Tb (half-life, 6.9 d) and Cu (half-life, 2.6 d). Hence, the 70-d half-life of the W/Re generator is the most reasonable option to have local access to PSMA radioligand therapy at all. Next, the letter addresses the challenge of projecting absorbed doses based on small-animal studies to human beings. Hence, studies in pigs would be required. However, beyond radiation geometry, the specific expression of PSMA in the proximal kidney tubules has to be considered. A study comparing human PSMA with its rat and pig orthologs exhibits different glutamate carboxypeptidase II expression levels among the species studied (2). Therefore, we considered the theranostic approach a more responsible way to continue clinical development; that is, innocuous Tc-PSMA-GCK01 imaging will be used to extrapolate the dosimetry of RePSMA-GCK01 therapy in men. Pretze et al. estimated that the same activities of Re-PSMA may convert into a 1.7 times higher kidney dose but only a 51% absorbed dose to tumors with a mass of 10g when compared with LuPSMA-617. This is not too much away from our own preliminary approximation. However, because of its higher b-energy, Re theoretically performs better against larger tumor lesions (23–32mm) than does Lu (3). In a tumor model that is very similar to the typical clustered PSMA expression pattern in prostate cancer, the increased cross-fire effect of Re improved its intercluster microdosimetry (4). Pretze et al. emphasized that the antitumor activity of RePSMA could be about 25% higher than identical absorbed doses delivered by Lu when dose-rate effects are considered. However, it is worth mentioning that the biologically effective dose concept, which is based on the linear-quadratic model, was developed for externalbeam radiotherapy, but its appropriateness for radioligand therapy still needs confirmation. In summary, considering the dosimetric consequences of replacing Lu with Re may not be enough because there are no validated tools to simulate the radiation biologic consequences. Even if physical absorbed doses—1 surrogate for the therapeutic index of a radiopharmaceutical—may be lower than for Lu-PSMA-617, we consider it warranted to assess the antitumor activity of RePSMA-GCK01 using a patient-related endpoint, for example, prostate-specific antigen response in a clinical trial.