Here, we conducted both a multi-reader qualitative analysis and a single reader quantitative analysis to determine the potential impact of any urinary activity of 18F-flotufolastat in the bladder and ureters on the assessment of 18F-flotufolastat-PET images from patients with newly diagnosed and recurrent prostate cancer.

Quantitative findings show the overall urinary activity of 18F-flotufolastat to be relatively low, and, while this study was not designed as a head-to-head comparison, the median bladder SUVmax (17.1) and SUVmean (12.5) recorded here compare favorably with values reported in the literature for other renally excreted PSMA-targeted radiopharmaceuticals (Supplemental Table 2).

Physiological bladder activity is a common feature of PET radiopharmaceuticals, and since activity in the bladder could potentially obscure lesions adjacent to the bladder, various approaches have been used in order to minimize the impact on image interpretation. These include administering intravenous fluids to dilute accumulating activity in the bladder [5], imaging the pelvic region at an early timepoint before bladder accumulation becomes substantial [5, 8], and employing later imaging timepoints after forced diuresis [10, 11]. If clinically necessary, co-administration of a diuretic is suggested in the 68Ga-PSMA-11 prescribing information as a method to potentially decrease artifact from 68Ga-PSMA-11 accumulation in the urinary bladder and ureters [15]. However, administration of a diuretic may cause discomfort to a patient because of the increased urge to void, which could be of particular significance in prostate cancer patients who may have micturition difficulties due to enlarged intact prostates, or who may suffer from incontinence following surgery. In the LIGHTHOUSE and SPOTLIGHT studies, no patients were co-administered diuretics for scanning purposes.

Qualitative findings further demonstrate the minimal impact of 18F-flotufolastat urinary activity in both the bladder and ureters on 18F-flotufolastat-PET-only image interpretation. By majority read, 96% of scans were considered to be unaffected without consulting other modalities than PET by urinary activity in the bladder (score 0 or 1). However, we note that PET-only assessment of the scans could be affected by urinary activity in up to 13% of the cases, dependent on the reader (3% by majority read). Moreover, in those cases where urinary activity did influence the PET assessment, the median bladder SUVmax and SUVmean were markedly higher. However, the data acquired did not specify if it was urinary activity in the bladder, in the ureters, or the urethra that interfered with the assessment of the scan. Therefore, the higher median SUVs observed in the bladder do not directly infer that the higher bladder activity obscured a lesion in close proximity to the bladder; rather, it can be taken to be representative of higher urinary tract activity in general. Ureteric activity was absent in more than half of the patients and, as shown in Fig. 2, even when present, had little impact on disease assessment. As shown in Fig. 5, the ureters depicted here do not interfere with identifying the adjacent LN metastases. One of the factors that might interfere with assessment of the scan is a transient focus of urine in the ureter, which could be mistaken for a LN metastasis. In this particular analysis, that focus might not be classified as “stasis of urine” and therefore unclassified as a ureter. However, readers were instructed to assign a qualitative score 2 if they were not able to distinguish between urinary activity and disease. Those cases are therefore captured in the qualitative analysis and, as stated above, are part of the small group of patients (24/712, 3.4%) in which urinary activity interfered with scan assessment. We therefore conclude that the ureteric activity reported here did not interfere with PET-only assessment in the vast majority of cases.

Patient with visible ureters and two pelvic lymph nodes. The figure presents the MIP image (SUV scale 0–10) of a patient with two visible ureters and two separately identifiable left pelvic lymph node metastases

Halo artifacts are a scatter correction artifact which derive from high organ-to-background activity ratios such as between the bladder and surrounding soft tissue, and a positive correlation between organ-to-background ratio and halo size has been reported [6]. Halo artifacts have been reported to occur with other renally cleared radiopharmaceuticals, particularly with PET/MRI [6, 16,17,18]. Use of diuresis has been reported to assist in reducing this phenomenon [16]. Although not strictly radiopharmaceutical-dependent, their occurrence is due to the physiologic biodistribution of the radiopharmaceuticals, and those with higher bladder accumulation are more likely to be affected [6]. Here, we show that the mean bladder SUVs of 18F-flotufolastat are lower than those reported for both 18F-DCFPyL and 68Ga-PSMA-11, but also that halo artifacts were extremely rare, occurring in fewer than 1% of patients (majority read) undergoing 18F-flotufolastat-PET/CT.

There are some limitations to the present analysis. First, the aim of the study was not to identify any lesions that were obscured by the bladder through comparison of urinary activity in the urinary tract to activity in potential lesions, rather to qualitatively assess if urinary activity might interfere with reading. Thus, SUVs in the urinary tract were not compared with those of metastatic lesions, and this should be the focus of future studies. Second, while reader agreement was broadly high across all measured variables, there was some level of disagreement on the presence of halo artifacts (ranged from 0 to 27 across readers). While one reader had a stricter definition of a halo, i.e., a reconstruction artifact deriving from the high activity in the bladder (and finding none), another reader actively searched for more subtle photopenic rim around the bladder that could potentially be called a halo artifact (identifying 27 cases). However, even with the latter, most sensitive definition used, only in 4% of cases a potential halo could be found. Furthermore, the two patients who had a halo artifact as per majority read had a qualitative score of 1 (by all three readers) indicating that the halo artifact did not interfere with image interpretation. While reader agreement was not formally tested in the present analysis, inter-reader comparison data have previously been reported for the blinded reads in the prostate/bed, pelvic LN, and extra-pelvic regions in the LIGHTHOUSE and SPOTLIGHT studies, which show a greater than ≥ 95% and ≥ 75% patient-level inter-reader agreement, respectively [19, 20]. Third, this post hoc analysis was not designed as a head-to-head comparison with other PSMA-PET radiopharmaceuticals, and thus, any comparisons of the median SUV reported here with data from other radiopharmaceuticals reported in the literature should be made with caution. Finally, as stated above, this study did not categorize the level of the urinary tract at which urinary activity interfered with assessment in patients who scored 2 on the qualitative scale. Therefore, it remains unknown to which extent bladder activity interfered with assessing local recurrence, and ureteric activity interfered with assessing pelvic and retroperitoneal LN.

Correctly staging patients is essential to guide patients to the most effective treatment options, and the addition of further accurate tools to the armamentarium of diagnostic options is vital in order to provide for patients across the prostate cancer spectrum. This large dataset from two phase 3 prospective trials suggests 18F-flotufolastat-PET urinary activity is relatively low and rarely impacts disease assessment, building on early clinical data [13] that show 18F-flotufolastat-PET to have lower average urinary excretion than reported for other renally cleared PSMA-PET radiopharmaceuticals to indicate a potential benefit of this novel radiopharmaceutical.