Normal-organ distribution of PSMA-targeting PET radiopharmaceutical ¹⁸F-flotufolastat: a post hoc analysis of the LIGHTHOUSE and SPOTLIGHT studies

Abstract

Background: High-affinity radiohybrid PSMA-targeting radiopharmaceutical ¹⁸F-flotufolastat (¹⁸F-rhPSMA-7.3) is newly approved for diagnostic imaging of prostate cancer. Here, we conduct a post hoc analysis of two phase 3 studies to quantify ¹⁸F-flotufolastat uptake in a range of normal organs. Methods: All 718 evaluable ¹⁸F-flotufolastat scans from LIGHTHOUSE and SPOTLIGHT were re-evaluated. Additionally, patients’ medical records were reviewed and any patients with high tumor burden (PSA>20 ng/mL), altered biodistribution (e.g., chronic kidney disease), major anatomical changes to normal organs (e.g., nephrectomy), or any other history of cancer were excluded. A medical physicist defined volumes of interest over specific organs for evaluation of SUV_mean and SUV_peak per PERCIST 1.0 criteria. Normally distributed data are reported as mean (SD) and non-normally distributed data as median (IQR). The co-efficient of variation (CoV; calculated as SD/mean for normally distributed data and IQR/median for non-normally distributed data) was used to quantify variability of SUV metrics. Results: In total, scans from 546 patients (244 primary, 302 recurrent) were eligible for this analysis. All organs were considered to be normally distributed except for the bladder and spleen. In the liver, the mean SUV_mean was 6.7 (SD 1.7), CoV 26%, while the bladder median SUV_mean was 10.6 (IQR 11.9), CoV 112%. The mean SUV_peak in the liver was 8.2 (SD 2.1), CoV 26% and median SUV_peak in the bladder was 16.0 (IQR 18.5), CoV 116%. Conclusions: Physiological ¹⁸F-flotufolastat uptake in normal organs was broadly consistent with other renally-cleared radiopharmaceuticals, which may have clinically significant implications when considering patient selection for radioligand therapy. Additionally, the bladder median SUV_peak for ¹⁸F-flotufolastat was lower than that previously reported for ⁶⁸Ga-PSMA-11 and ¹⁸F-DCFPyL.

Introduction

Following approval of the first prostate-specific membrane antigen (PSMA) targeting positron emission tomography (PET) radiopharmaceutical by the US Food and Drug Administration (FDA) in 2020 [1], PSMA-PET has become the mainstay for diagnostic imaging in patients with suspected or recurrent prostate cancer, with three radiopharmaceuticals now approved for use in the USA [2-6].

The most recently approved radiopharmaceutical, ¹⁸F-flotufolastat (formerly ¹⁸F-rhPSMA-7.3), is a high-affinity radiohybrid PSMA-ligand [7]. Data from the Phase 3 LIGHTHOUSE and SPOTLIGHT studies show ¹⁸F-flotufolastat to be well tolerated and to provide clinically useful information regarding the presence of N1 and M1 disease prior to surgery in newly diagnosed prostate cancer, and for the localization of recurrent disease, particularly among patients with prostate specific antigen (PSA) levels ≤0.5 ng/mL [8-10]. Moreover, a post hoc analysis of over 700 scans from across the two studies confirmed early clinical data suggesting ¹⁸F-flotufolastat has low average urinary excretion and showed that in 96% of patients ¹⁸F-flotufolastat image assessment is not impacted by urinary activity (majority read data) [11,12]. Within professional body guidelines ¹⁸F-flotufolastat is considered among a common class of radiopharmaceuticals for PET/computed tomography (CT), collectively referred to as PSMA-ligands or PSMA-PET [2,13].

Further to the role of PSMA-PET radiopharmaceuticals in diagnostic imaging, the March 2022 approval of ¹⁷⁷Lu-PSMA-617 (¹⁷⁷Lu-vipivotide tetraxetan) radioligand therapy for patients with metastatic castration resistant prostate cancer (mCRPC) has prompted the additional clinical use of PSMA-PET for the selection of patients with mCRPC who may be suitable for PSMA radioligand therapy (RLT) [2,14]. Following approval of ¹⁷⁷Lu-PSMA-617, the FDA included in the indication for Locametz (kit for preparation of gallium ⁶⁸Ga-PSMA-11) the selection of patients with mCRPC for whom ¹⁷⁷Lu-PSMA-617 directed therapy is indicated [15]. Data from the Vision trial supported approval of Locametz for this indication, where PSMA-positive disease was generally defined as the presence of at least one tumor lesion with ⁶⁸Ga-PSMA-11 uptake greater than normal liver [15,16].

Subsequently, the Society of Nuclear Medicine and Molecular Imaging (SNMMI) PSMA PET Appropriate Use Criteria (AUC) Working Group suggested that ¹⁸F-DCFPyL could be considered equivalent to ⁶⁸Ga-PSMA-11 for ¹⁷⁷Lu-PSMA-617 patient selection based on comparable biodistribution data for ¹⁸F-DCFPyL and ⁶⁸Ga-PSMA-11, particularly in the liver, given its common use as a reference organ [4,5,14,17]. To date, only two studies (N=34 and N=11 subjects) of note have been published that compare normal-organ biodistribution of ⁶⁸Ga-PSMA-11 and ¹⁸F-DCFPyL using either absolute SUV metrics or tumor-to-liver ratios [18,19].

Most recently, following FDA approval of ¹⁸F-flotufolastat, the latest National Comprehensive Cancer Network (NCCN) prostate cancer guidelines state that owing to the equivalency of normal-organ distribution of the three FDA-approved PSMA-PET radiopharmaceuticals, ¹⁸F-DCFPyL and ¹⁸F-flotufolastat may also be used with ¹⁷⁷Lu-PSMA-617 [2]. The 2024 SNMMI AUC update also recommends that ¹⁸F-flotufolastat can be used for ¹⁷⁷Lu-PSMA-617 patient selection but notes there to be limited data on the biodistribution of ¹⁸F-flotufolastat and how it compares with ¹⁸F-DCFPyL and ⁶⁸Ga-PSMA-11 [20].

Given the important relationship of normal-organ biodistribution and PSMA-PET interpretation, especially in the liver due to its use as a reference organ for radioligand therapy patient selection, we conducted a post hoc analysis of a large population of patients to quantify the uptake of ¹⁸F-flotufolastat in the liver and other normal organs in all evaluable patients who underwent ¹⁸F-flotufolastat PET/CT as part of either the LIGHTHOUSE or SPOTLIGHT study.

Materials and methods

Patient population

As previously reported, the Phase 3 LIGHTHOUSE (NCT04186819) and SPOTLIGHT (NCT04186845) studies assessed the diagnostic performance of ¹⁸F-flotufolastat in patients with newly diagnosed, unfavorable intermediate to very high-risk prostate cancer, or in patients experiencing biochemical recurrence, respectively [8,9]. Both studies were performed in line with the principles of the Declaration of Helsinki and approved by the relevant Ethics Committees. All patients provided written informed consent [8,9].

All 718 evaluable scans from the efficacy populations (352 from the LIGHTHOUSE and 366 from the SPOTLIGHT study) were reviewed for this post hoc analysis. In addition, the medical history data from the original studies were reviewed for each patient and any patients meeting the following criteria were excluded from the present analysis: high tumor burden (defined as baseline PSA>20 ng/mL); altered biodistribution (e.g., chronic kidney disease, hepatic cysts); major anatomical changes in the normal organs under study (e.g., nephrectomy, cystectomy); or any other history of cancer or additional primary cancer.

¹⁸F-flotufolastat PET/CT

Patients were asked to continue taking any prescribed medications; to arrive for their appointment well hydrated; and to void immediately prior to entering the scanning room. The target administered activity of ¹⁸F-flotufolastat was 296 MBq (median [range] activity 307.3 MBq [213.5-397.8 MBq] in LIGHTHOUSE; 306.0 MBq [230.1-355.2 MBq] in SPOTLIGHT) [8,9]. At 50-70 minutes post-injection, the PET/CT acquisition was started with a time per bed position of 3 minutes, in a caudal-cranial direction from mid-thigh to skull base. No co-administration of x-ray contrast agent was allowed, and no late imaging was performed. Images were iteratively reconstructed, utilizing time-of-flight if available; no point-spread-function correction algorithms were used. All PET systems in these multi-site studies were approved by the image core lab prior to scanning the first patient.

Image analysis

The present post hoc analysis utilized a comparable methodology to Ferreira et al to facilitate general comparison of the peak standardized uptake value (SUV_peak) for ¹⁸F-flotufolastat to those previously reported for ⁶⁸Ga-PSMA-11 and ¹⁸F-DCFPyL [18].

A professionally-accredited medical physicist used image review software (MIM Encore™, version 7.2, MIM Software Inc., Cleveland, USA) to define and outline the volumes of interest (VOIs) over the following organs using either a semi-automated tool (PET Edge^®), or spheres: lacrimal glands (20 mm sphere); parotid glands and sub-mandibular glands; lumen of the descending thoracic aorta (20 mm sphere); parenchyma of the right hepatic lobe (60 mm sphere); kidneys; third portion of the duodenum; urinary bladder contents; and gluteal muscle (20 mm sphere), as illustrated in Figure 1. In cases where the physiological activity was similar to adjacent local background and the semi-automated tool could not be used, the low dose CT scan was used to define or amend the VOI boundary. The placement of each outline was reviewed by a qualified nuclear medicine physician to ensure accuracy and standardization, as well as to ensure that pathological disease was excluded from the results.

Figure 1.

Maximum intensity projection ¹⁸F-flotufolastat PET scan of a patient with prostate cancer (SUV scale 0-10). Representative volumes of interest (VOIs) are shown on each of the normal organs measured in this analysis (left). Sagittal image (right lower) showing care was taken to ensure the bladder VOI did not include ¹⁸F-flotufolastat avid local disease, also afforded by the comparatively low degree of urinary activity within the bladder in this case.

Mean standardized uptake value (SUV_mean) and peak standardized uptake value (SUV_peak), as defined in the PERCIST 1.0 criteria [21], were computed for each VOI using the native semi-automated tool in the image review software. For paired organs, the arithmetic mean SUV metric is given. In patients where either the lacrimal glands, or both the lacrimal glands and parotid glands were not included in the scan field of view, if all other eligibility criteria were met then it was assumed that biodistribution was not affected, and the SUV metrics for other organs were included in the dataset. In patients where only a single side of the paired organ was visible, the SUV metrics for that organ were excluded.

This method also builds upon a previously published analysis conducted in the same patient population [11] by extending their assessment of ¹⁸F-flotufolastat urinary activity in a single slice over the maximum radioactive bladder diameter to a quantitative assessment of the entire radioactive bladder volume.

Statistics

Statistical analysis was performed using SAS Version 9.4 (SAS Institute Inc., Cary, NC). Violin and box-and-whisker plots were created in R Stats (R Foundation for Statistical Computing, Vienna, Austria). Data were assessed for normality via review of the histogram and Q-Q plots. Shapiro-Wilk, as used in Ferreira et al [18], was not used due to the large sample size and the known oversensitivity in large datasets. Normally distributed data are reported as the arithmetic mean and standard deviation (mean (SD)). Non-normally distributed data are presented as the median and interquartile range (median (IQR)). Variability of SUV metrics for each organ was quantified using the co-efficient of variation (CoV), calculated as the SD/mean for normally distributed data and IQR/median for non-normally distributed data.

Results

A total of 718 subjects were available for analysis in this post hoc study (Figure 2); of these, 546 (244 primary, 302 recurrent) met the additional eligibility criteria and were included in the final cohort.

Figure 2.

Patient flow chart. *Patients may have met ≥1 exclusion criteria.

For the subset of paired organs at the extreme axial field of view, a total of 362 lacrimal, 540 parotid and 543 sub-mandibular locations were found to be sufficiently within the field of view to be included in the results.

Biodistribution

Quantitative data for all organs are presented in Table 1. SUV data for all organs were considered to be normally distributed except for the bladder and spleen. Figure 3 presents SUV distribution data for the key organs analyzed in this report (liver and bladder).

Table 1.

Summaries of SUV_mean (A) and SUV_peak (B) ¹⁸F-flotufolastat physiological activity in normal organs

A. SUVmean physiological activity
Organ	n	Statistic	SUVmean	SD/IQR	CoV
Aorta	546	Mean	1.9	0.4	22%
Bladder	546	Median	10.6	11.9	112%
Duodenum	546	Mean	7.4	2.5	33%
Gluteal	546	Mean	0.6	0.1	19%
Kidney	546	Mean	22.4	5.5	25%
Lacrimals	362	Mean	5.1	2.2	42%
Liver	546	Mean	6.7	1.7	26%
Parotid	540	Mean	11.1	3.3	30%
Spleen	546	Median	6.9	3.1	46%
Sub-mandibular	543	Mean	12.1	3.4	28%
B. SUVpeak physiological activity
Organ	n	Statistic	SUVpeak	SD/IQR	CoV
Aorta	546	Mean	2.0	0.5	22%
Bladder	546	Median	16.0	18.5	116%
Duodenum	546	Mean	11.1	3.8	34%
Gluteal	546	Mean	0.6	0.1	21%
Kidney	546	Mean	37.8	9.0	24%
Lacrimals	360	Mean	6.7	2.5	37%
Liver	546	Mean	8.2	2.1	26%
Parotid	540	Mean	16.8	5.1	30%
Spleen	546	Median	9.9	4.6	47%
Sub-mandibular	543	Mean	17.8	5.0	28%

Standard deviation is presented for normally distributed metrics while interquartile range is presented for non-normally distributed metrics. CoV, Coefficient of variation; IQR, interquartile range; SD, standard deviation.

Figure 3.

Histogram and Q-Q plots of ¹⁸F-flotufolastat SUV_mean and SUV_peak in the liver and bladder.

In the key organs of liver and bladder, the SUV_mean in the liver was a mean of 6.7 (SD 1.7), CoV 26%; and in the bladder a median of 10.6 (IQR 11.9), CoV 112%. In the same key organs, SUV_peak in the liver was a mean of 8.2 (SD 2.1), CoV 26%; and in the bladder a median of 16.0 (IQR 18.5), CoV 116%.

The population distribution of SUV_mean and SUV_peak values across the key organs of interest is displayed graphically in Figure 4.

Figure 4.

Box-whisker and Violin plots of SUV_mean (A and B) and SUV_peak (C and D) in the key organs of liver and bladder. Data that are non-normally distributed (B and D) are represented by median SUV metric, interquartile range, and outliers. Data that are considered normally distributed (A and C) are represented by mean SUV metric and standard deviation.

Discussion

¹⁸F-Flotufolastat is a newly FDA-approved PSMA-targeted PET radiopharmaceutical for diagnostic imaging in patients with prostate cancer. Further to their role in diagnostic imaging, PSMA-targeted PET radiopharmaceuticals are now being utilized to help select patients with mCRPC who may benefit from PSMA-targeted RLT based on the presence of uptake in metastatic lesions greater than normal liver on PET images. This clinical use is supported by data from the Vision Trial and recommendations by the NCCN and SNMMI AUC Working Group [2,14,16,20], though until now there has been limited information published around ¹⁸F-flotufolastat normal organ biodistribution. We quantitatively evaluated ¹⁸F-flotufolastat uptake in the liver and other normal organs to further describe its physiologic biodistribution and better inform clinicians around its use in evaluating patients who may be eligible for RLT.

This post hoc quantitative analysis of ¹⁸F-flotufolastat uptake in normal organs in 718 men who underwent PET/CT in two phase 3 clinical studies shows the range of normal organs with physiological ¹⁸F-flotufolastat activity to be consistent with previously published phase 1 data,¹² as well as more broadly with other PSMA radiopharmaceuticals that are predominantly excreted via the urinary tract [17,18,22,23]. The highest SUV metric, mean SUV_mean 22.4 (SD 5.5) (mean SUV_peak 37.8 (SD 9.0)) was seen in the kidney, which is typical for a PSMA radiopharmaceutical. However, bladder activity median SUV_mean was 10.6 (IQR 11.9) (median SUV_peak 16.0 (IQR 18.5)); this SUV_peak is lower than that reported for either ⁶⁸Ga-PSMA-11 or ¹⁸F-DCFPyL by Ferreira et al (43.1 and 57.3, respectively) [18], offering further support to the early clinical data, and data previously reported from this population, that show ¹⁸F-flotufolastat has low average urinary excretion compared with values reported for other renally-cleared PSMA-PET radiopharmaceuticals [11,12]. The radiohybrid technology platform from which ¹⁸F-flotufolastat is developed supports optimal kidney clearance through high PSMA-binding affinity, high internalization by PSMA-expressing cells, medium-to-low lipophilicity, and high human serum albumin binding [7,9,24,25]. This likely contributes to the lower average urinary activity observed for ¹⁸F-flotufolastat at the time of PET imaging as illustrated by the patient image in Figure 1. A visual inspection of the population distribution (Figure 4) shows that while many subjects have low bladder SUV metrics (SUV_mean < circa 20, SUV_peak < circa 25), there are a smaller number of higher values and a few extremely high outliers which likely affected the CoV reported here. Nevertheless, as previously reported, qualitative image assessment of over 700 ¹⁸F-flotufolastat PET/CT images by three blinded nuclear medicine physicians show that ¹⁸F-flotufolastat urinary activity does not impact disease assessment for the vast majority (96%) of patients [11].

To date, ¹⁸F-flotufolastat has been shown to offer clinically useful information regarding identification of both N1 and M1 disease prior to surgery in patients newly diagnosed with unfavourable intermediate-to-very high-risk prostate cancer [8]. In patients with biochemical recurrence, studies in the literature show ¹⁸F-flotufolastat to compare favorably with the other FDA-approved PSMA-targeted PET radiopharmaceuticals, offering high patient-level detection rates in the range of 73-83% [9,26]. Moreover, among 121 patients with very low PSA levels (<0.5 ng/mL), ¹⁸F-flotufolastat has a detection rate of 64% [10], while DCFPyL and ⁶⁸Ga-PSMA-11 are reported to have rates of 36% and 38%, respectively [27,28]. When comparing patient-level positive predictive values for histopathologically-verified lesions, all of the FDA-approved PSMA-targeting radiopharmaceuticals show similar values: 84% for ⁶⁸Ga-PSMA-11 [28], 79-83% for ¹⁸F-DCFPyL [27], and 82% for ¹⁸F-flotufolastat [9].

Further to this established favorable diagnostic performance, we show here that the range of organs with physiological ¹⁸F-flotufolastat uptake is consistent with early phase data [12], and the population median SUV metric point estimates are broadly consistent with other predominantly renally-cleared PSMA-targeted radiopharmaceuticals, but with lower median bladder values as discussed above [18].

The comparison of lesion uptake with reference organs (e.g., the liver) has been successfully used with other radiopharmaceuticals and in other cancers [29]. This approach is now playing a role in RLT selection, having been used in several recent large clinical studies leading to its inclusion in the ⁶⁸Ga-PSMA-11 prescribing information [5,16,30], as well as being recommended by joint European Association of Nuclear Medicine (EANM)/SNMMI prostate cancer guidance [13]. The mean liver SUV_peak seen in the present study, 8.2 (SD 2.1), CoV 26%, is comparable to values reported for ⁶⁸Ga-PSMA-11 and ¹⁸F-DCFPyL by Ferreira et al (6.7 (SD 1.5), CoV 21.9%, and 7.5 (SD 1.7), CoV 22.5%, respectively), further supporting their conclusion of its suitability for use as a reference physiological organ [18]. Moreover, given the mean liver SUV_peak for ¹⁸F-flotufolastat is comparable to values reported for ⁶⁸Ga-PSMA-11 and ¹⁸F-DCFPyL [18], these data support using the liver as a reference physiological organ for PSMA-positive lesion identification with ¹⁸F-flotufolastat, per existing practice guidance recommendations around RLT patient selection [2].

The large number of patients included in this analysis is a particular strength, allowing a robust set of data, however there are some limitations of note. While the assessment methodology in Ferreira et al [18] was followed as closely as possible, there are methodological differences in the data acquisition and patient populations. The data presented here are not an intra-patient comparison, and there will be variation on a per-patient basis. Additionally, there are likely differences in the hydration status of the patients evaluated in this post hoc analysis, as this was not prospectively controlled for in the original studies. Finally, this work is not intended to draw conclusions on the diagnostic efficacy of any particular PSMA radiopharmaceutical, for which a formal head-to-head study would be required.

In conclusion, these data from a large post hoc analysis show ¹⁸F-flotufolastat to have comparable biodistribution to both ⁶⁸Ga-PSMA-11 and ¹⁸F-DCFPyL in the liver, suggesting that these radiopharmaceuticals can be used interchangeably in local site protocols for evaluating patients with prostate cancer who may be eligible for RLT. Although not a head-to-head study, the newly measured ¹⁸F-flotufolastat SUV_peak in the bladder appears lower than the reported values for other renally cleared PSMA agents.

Disclosure of conflict of interest

RP and JS are employees of Blue Earth Diagnostics Ltd., Oxford, UK. BF is an employee of Blue Earth Therapeutics Ltd., Oxford, UK. PD is an employee of Blue Earth Diagnostics Inc., Monroe Township, New Jersey, USA.