Radiation dosimetry and pharmacokinetics of the tau PET tracer florzolotau (18F) in healthy Japanese subjects

The sample size of this study was not based on a statistical power calculation. This study was intended to assess the safety and pharmacokinetics of florzolotau in healthy male subjects and to estimate absorbed doses and effective doses. The absorbed doses and the effective doses of florzolotau in healthy subjects had been evaluated in the previous phase 1 study in the US (Study 4216, n = 2, [19]) and the clinical research in Taiwan (Study 16083111, n = 12, [20]). The number of subjects in this study was determined to be 3, a minimum number to compare the absorbed doses and effective doses between different ethnic populations.

This study was conducted at the Shiga Medical Center Research Institute in Moriyama, Japan. The clinical study protocol including amendments, informed consent form (ICF), and all other study-related documents were reviewed and approved by an Institutional Review Board of Shiga General Hospital, and the study was conducted in compliance with the study protocol, Good Clinical Practice (GCP) as set forth in the International Conference on Harmonization (ICH) guidelines on GCP (ICH E6(R2)) and registered with the Japan Pharmaceutical Information Center (JapicCTI-205302). The subject inclusion criteria included being male Japanese, age between 20 and 64, the ability to provide his written informed consent, and not diagnosed with any disease affecting participation in the study and being normal at the screening tests of vital signs, electrocardiogram, hematology, and blood biochemical tests. Four subjects were screened, and 3 subjects with height of 175.8 ± 8.7 cm and body weight of 75.7 ± 8.7 kg were enrolled. The fourth subject was enrolled as a spare subject and did not undergo the studies according to the prior arrangement in the study protocol.

Florzolotau was produced by Ibaraki laboratory of Fujifilm Toyama Chemical Company (currently PDRadiopharma Inc., Tokyo, Japan). Florzolotau was synthesized by the NEPTIS perform synthesizer (Ora, Neuville, Belgium) by the reaction of tosylate precursor of florzolotau, (3-((2-((1E,3E)-4-(6-(methylamino)pyridin-3-yl)buta-1,3-dien-1-yl)benzo[d]thiazol-6-yl)oxy)-2-((tetrahydro-2H-pyran-2-yl)oxy)propyl4-methylbenzenesulfonate (APRINOIA Therapeutics Inc., Tokyo, Japan). The cyclotron program sequence was initiated, bombarding the ¹⁸O target water according to a preset beam current and time, and producing ¹⁸F-fluoride ions by the ¹⁸O(p,n)¹⁸F nuclear reaction. After the bombardment, the target water¹⁸F-fluoride ion was transferred to the NEPTIS synthesizer. ¹⁸F-fluoride ion dissolved in¹⁸O enriched target water was stored in a conical tube (¹⁸F tube) and then passed through a QMA ion exchange column [pre-activated with water for injection (5 mL), K₂CO₃ solution (5 mL), and water for injection (5 mL)]. The ¹⁸F-fluoride ion in the target water was trapped onto the ion exchange column. The ¹⁸F-fluoride ion was eluted from QMA ion exchange column with 1.0 mL of K222-K₂CO₃ solution and added to the reaction flask. The eluted radioactivity was heated to 95 °C and dried under a compressed nitrogen purge as well as under negative pressure. Then, acetonitrile was added to the reaction tube several times, and ¹⁸F-fluoride ion was further dried by azeotropic evaporation of water–acetonitrile (95 °C under a compressed nitrogen purge and reduced pressure) for 7 min. After completion, the heating temperature was raised to 100 °C under reduced pressure to obtain the dried K222-¹⁸F-potassium fluoride. The precursor solution (about 2 mg dissolved in 1.5 mL dimethyl sulfoxide, Sigma-Aldrich, Saint Louis, MO, USA) was added into the reaction flask containing dry K222-¹⁸F-potassium fluoride, and the resulting reaction mixture was heated to 110 °C. Compressed nitrogen was slowly introduced into the reaction mixture while maintaining the heating temperature at 110 °C for at least 5 min. 3 N HCl solution (0.8 mL) was added to the reaction mixture, and the reaction mixture was heated to 120 °C for 0.5 min. After completion, the reaction mixture was cooled to 60 °C, and then 5.0 mL of a 0.5 N NaOH solution was added. Next, the resulting solution was then aspirated into a syringe and diluted with 3% sodium ascorbate. The solution in the syringe was passed through a tC18 column (Agilent Technologies, Santa Clara, CA, USA), and then the cartridge was washed with 3% w/v sodium ascorbate water for injection solution. The crude florzolotau product was eluted from the tC18 column with ethanol. The obtained florzolotau solution was loaded onto a semi-preparative high-performance liquid chromatography (HPLC, Agilent 1260, Agilent Technologies, Santa Clara, CA, USA) column for purification. The synthesis of florzolotau was done 3 times, and the yield was 43.7 ± 4.1% and the activity was 199.7 ± 0.9 MBq/mL at the end of synthesis.

The study procedure and schedule are outlined in Table 1. A subject was tentatively enrolled in the study between Day-28 and Day-14 after vital signs, ECG, and laboratory test, and the expected date of PET scan was scheduled. Florzolotau (185 MBq at the calibration time) was administered as a single intravenous injection. Actual injection doses for subject 1001, 1002, and 1004 were 194.9, 195.7, and 194.4 MBq, respectively (5.27 ± 0.02 mCi). Whole-body PET scans were acquired at Session 1 to Session 6. Venous blood and urine were also collected for pharmacokinetic measurements, and those collected after image acquisition of Session 6 were also used for the laboratory test including hematological parameters, blood biochemical test, coagulation test, and urinalysis.

Serial PET scans of the whole body (top of head to mid-thighs) were acquired PET/computed tomography (CT) scanner (Biograph Truepoint 16, Siemens Healthcare, Erlangen, Germany) to assess the bio-distribution and radiation dosimetry of florzolotau in healthy male subjects. Four whole-body PET scans for 90 s/bed position were acquired after completion of vital signs and ECG examinations immediately after the florzolotau administration, the scan was started 15–16 min after the injection (Session 1). Following an approximately 30 min break, two whole-body PET scans for 2 min/bed position were acquired (Session 2). Since then, one whole-body PET scan for 4 min/bed position was repeated at an approximately 30 min break (Session 3–6). A transmission CT scan for positioning and attenuation correction was performed before commencing the Session 1. After that, when the subject moved, a transmission CT scan was performed before or after the imaging session as needed. The acquired emission data were reconstructed using the 3-dimensional ordered subset expectation maximization method with attenuation, random, and scatter correction and without the point spread function correction. The reconstruction parameters were matrix size of 256 × 256 with 21 subsets and 2 iterations.

Reconstructed data were analyzed using PMOD software (version 4.0, PMOD Technologies, Zurich, Switzerland). Three-dimensional volumes of interest (VOIs) were constructed on the PET images to include all organ activities. The VOIs of the target organs including the brain, lungs, heart wall, stomach wall, liver, gallbladder, spleen, pancreas, upper large intestine, lower large intestine, rectum, kidneys, and urinary bladder wall were placed using the HERMES Data Analysis Application Hybrid Viewer Dosimetry version 4.0 (Hermes Medical Solutions, Stockholm, Sweden). Regarding the upper large intestine and the lower large intestine, the VOIs were placed from the ileocecal region to the transverse colon and from the descending colon until the sigmoid colon, respectively. The anatomical information from the CT data and the human whole-body reference atlas was used to delineate subject specific VOIs. Average counts for the VOIs were determined at each whole-body scan using the PET emission data.

Decay-corrected activity in each organ was expressed as a percentage of the injected dose (%ID) and plotted against time to generate time–activity curves (TACs). The numbers of disintegrations, formerly referred to as residence time, were calculated from the area under the non-decay-corrected TACs from time zero to infinity with trapezoid method (or model-fitting), multiplied by the volume of the organ. The residence times were entered in OLINDA/EXM software (version 2.1, Hermes Medical Solutions, Stockholm, Sweden), and the absorbed doses in each organ and whole-body effective dose were calculated using the following formulas.

S-factors are implemented within the OLINDA/EXM software. The International Commission on Radiological Protection (ICRP) adult male phantom model (70 kg and 170 cm tall, ICRP-89) was used to calculate the absorbed doses, and the ICRP-103 tissue weighting factors were used to calculate the effective dose. Calculations were performed without modeling of urinary bladder voiding.

Before the tracer administration, another venous route was secured on the other side forearm than the injection side. A total of 10 blood samples were taken from the venous route: immediately, approximately 10, 20, 30, and 40 min after the injection, and after each following session. Radioactivity in whole blood was measured using a gamma-counter (ARC-8001, Hitachi Aloka Medical, currently Fujifilm Healthcare Manufacturing, Minamiashigara, Japan). Subjects were urged to urinate before administration of the study drug. Then, urine was collected in principle after image acquisition of Session 6, and urination after image acquisition of Sessions 1, 2, 3, 4, and 5 was also allowed when a subject felt a need to urinate. Urine was collected after Sessions 2 and 6 for subject 1001 and after Sessions 4 and 6 for subject 1002 and 1004. Urine radioactivity was measured using the gamma-counter as well.

Vital signs (systolic/diastolic blood pressure, heart rate, and respiratory rate) and ECG were taken at screening, prior to and immediately after administration of florzolotau and after image acquisition of Session 6. The laboratory tests including hematology, blood biochemical test, coagulation test, urinalysis, and serological tests were also performed at screening and after imaging acquisition of Session 6.

Serial whole-body PET images of individual subjects are presented in Fig. 1. Uptake of the radiotracer in the brain, lung, heart, liver, kidney, and pancreas was high at 10–25 min after the florzolotau injection, then the radioactivity gradually decreased with time. The radioactivity in gallbladder wall and intestine was high at 10–25 min, and the radioactivity increased with time, the uptake peaked 1–2 h and then gradually declined.

The time–activity curves of the decay-corrected %ID of various organs for 3 subjects are shown in Fig. 2. Organ showing the highest mean initial ¹⁸F activity uptake at the initial imaging time point of 0.25–0.42 h was the liver (29.0 ± 4.0%). The %ID was also high in the intestine (4.69 ± 1.65%) and the brain (2.13 ± 0.18%). The organs with a fast wash-in and washout were brain, lungs, heart, and kidneys. The radioactivity in the intestine and gallbladder was increased with time, and the radioactivity in the rectum and urinary bladder was higher at 2–3 h after the injection. The residence time for each organ is shown in Table 2. Among the mean residence times of organs, the largest average retention time was the liver of 0.47 ± 0.05 h, followed by the small intestine of 0.43 ± 0.06 h and the gallbladder wall of 0.29 ± 0.09 h. The bladder and the bowel showed substantial uptake from 15 min until the end of the study due to urine accumulation and radioactivity elimination through the gastrointestinal tract. The residence time was short in brain, lungs, heart wall, kidneys, spleen, pancreas, upper large intestine, lower large intestine, rectum, and urinary bladder wall.

The decay-corrected radioactivity in blood samples of three subjects is shown in Fig. 3. The mean radioactivity taken immediately after florzolotau injection (1–2 min after the injection) was 0.0097 ± 0.0013 MBq/mL, and the level was steeply decreased at 10 min after the injection to 0.0018 ± 0.0006 MBq/mL and remained until at the end of Session 6 (0.0018 ± 0.0010 MBq/mL). The blood half-life of the compound was 16.4 ± 5.4 h. The mean total decay-correct radioactivity in urine until the end of Session 6 (for 270 min after the injection) was 7.49 ± 0.73 MBq. Excretion of ¹⁸F activity by renal pathway was 3.84% of ID.

Absorbed doses in each organ and whole-body effective doses were calculated using OLINDA/EXM (version 2.0). The results are summarized in Table 3. The highest absorbed dose was 508 μGy/MBq of the gallbladder wall, followed by the liver of 79.4 μGy/MBq, pancreas of 42.5 μGy/MBq, upper large intestine of 34.2 μGy/MBq, and lower large intestine of 24.1 μGy/MBq. The effective dose was calculated as 19.7 μSv/MBq according to tissue weighting factor reported by ICRP-103. Based on the results of this experiment, the subject will receive and effective dose of approximately 3.61 mSv when 185 MBq florzolotau was given to a male subject.

There have been no deaths or discontinuations due to adverse events. No serious adverse event was noted in this study. Evaluation of vital signs and ECG revealed no statistically significant findings nor result in clinically meaningful individual changes.

Vital signs (systolic/diastolic blood pressure, heart rate, and respiratory rate) and ECG were examined at baseline (screening period), prior to and immediately after administration of the study drug and after the final PET imaging session on Day 1. There were no clinically significant changes in any vital signs (body temperature, Systolic/diastolic blood pressure, heart rate, and respiratory rate) and ECG in any subjects after study drug administration.

Blood for clinical laboratory tests was collected at baseline (screening period) and after the final PET imaging session on Day 1. Only subject No. 1004 experienced 6 adverse events. All the adverse events were mild, which were urine protein, increases in aspartate aminotransferase, alanine aminotransferase, gamma-glutamyl-transferase, dyslipidemia, and increase in blood bilirubin, which were recovered at the biochemical test conducted 19 days after the injection. These adverse events were judged as not related to the study drug.