The present study was a retrospective comparison of the intra-patient tracer distribution between [68Ga]Ga-PSMA-11 and [18F]F-PSMA-1007. The study confirmed the expected significant higher hepatic uptake for [18F]F-PSMA-1007 compared to [68Ga]Ga-PSMA-11 due to its higher lipophilicity, leading to hepatic excretion. We also showed that within the same patient, the splenic uptake of [18F]F-PSMA-1007 is significantly higher than the hepatic uptake of [68Ga]Ga-PSMA-11.

Replacing the liver by the spleen as internal reference for [177Lu]Lu-PSMA-617 RLT enrollment could therefore lead to a shift in patient selection. Furthermore, the inter-individual range of splenic uptake is higher for both [18F]F-PSMA-1007 and [68Ga]Ga-PSMA-11 compared to hepatic PSMA-uptake. This may be one of the reasons why Seifert et al. excluded 42% of patients in a cohort with either [18F]F-PSMA-1007 or [68Ga]Ga-PSMA-11 scans based on VISION trial selection criteria for patients scanned with [18F]F-PSMA-1007 using the spleen as internal reference [12]. The higher SUVmean of the internal reference organ spleen in [18F]F-PSMA-1007 may shift patients away from [177Lu]Lu-PSMA-617 RLT compared to the drop-out rate of only 13% in the VISION study based on hepatic uptake in [68Ga]Ga-PSMA-11. Whether this leads to a similar patient selection as the cut-off used in [68Ga]Ga-PSMA-11 studies has never been tested systematically. The internal reference value for liver uptake with [18F]F-PSMA-1007 would exclude a large number of patients from RLT who would likely be included with [68Ga]Ga-PSMA-11. Other authors have compared physiological uptake in 14 individuals with PSA recurrence using both [68Ga]Ga-PSMA-11 and [18F]F-DCFPyL PET/CT, showing a marginal increase in liver parenchymal uptake with [18F]F-DCFPyL (SUVmean 6.2 vs. 5.1, p = .005) [6].

Some studies have suggested using the salivary glands as an internal reference, but also based on [68Ga]Ga-PSMA-11 images, reporting that patients in whom more than 80% of the lesions have uptake higher than the salivary glands benefit from treatment [13]. Notably, salivary gland uptake showed better overall comparability between the two tracers; however, it exhibited the highest intra- and inter-individual variability of all examined organs. This may limit the use of the proposed salivary gland score as a robust internal reference [13]. However, the use of [18F]F-PSMA-1007 in clinical practice for patient selection in [177Lu]Lu-PSMA-617 RLT is still under discussion. A comprehensive graphical representation of the main study results is shown in Fig. 7.

Graphical representation of SUVmean with mean and standard deviation for the parotid gland, spleen, liver, blood pool and bone for both [68Ga] Ga-PSMA-11 PET and [18F] F-PSMA-1007 PET.

To determine the potential influence of increased tumor volume on conventional tracer distribution in the present study, we quantified tumor volume. Only three patients were classified as having high-volume disease, defined as greater than 530 ml according to the criteria of Gafita et al. [11].

Although direct intra-lesion comparisons were not performed, the SUVmax of the most PSMA-active malignant PCA lesions showed comparability between the two scans. In particular, there was no evidence of generalized increased [18F]F-PSMA-1007 accumulation in primary tumors, local recurrences, lymph nodes or bone metastases. This observation differs from the prospective intra-individual analysis performed by Pattison et al., who reported a significantly higher uptake of [18F]F-PSMA-1007 in lymph nodes (mean SUVmax 11.1 vs. 8.7) [14]. However, our results are consistent with their observation that PSMA uptake in bone metastases was similar (mean SUVmax 30.9 vs. 30.7) [14].

Both [68Ga]Ga-PSMA-11 and [18F]F-PSMA-1007 PET tracers are commonly used, depending on considerations such as local accessibility, practicality, imaging quality, and the specific clinical context. In particular, [18F]F-PSMA-1007 may have diagnostic efficacy in assessing primary tumors [9, 14, 15] or identifying local recurrence [7, 10, 16] due to reduced urinary excretion. Conversely, the use of [18F]F-PSMA-1007 has been associated with an increased incidence of indeterminate findings, particularly in the bone, termed non-specific bone uptake, and celiac ganglia [8, 10].

This study has several limitations. First, due to the retrospective nature of the study the scans were performed at two different time points and in some cases for different indications (staging, biochemical recurrence, post-treatment response), this is precluding a direct intra-lesion analysis and therefore makes a comparison of uptake in malignant very limited. However, careful efforts were made to mitigate the temporal differences between the two scans, with no discernible differences in PSA levels at the time of imaging or in clinical indications. Second, a major limitation is the retrospective design, particularly given the heterogeneous nature of the patient cohort. Third, the current study cohort does not allow for a comprehensive assessment of patient eligibility for [177Lu]Lu-PSMA-617 RLT, as the majority of patients underwent imaging for staging or biochemical recurrence, limiting the scope of the study in this regard.