Study design

The study was a randomized cross-over interventional study (Fig. 2). Each subject underwent three 8-minute dynamic 82Rb PET/CT scans over two days: Day A and Day B (mean interval 9.7 ± 5.0 days between examination days). On Day A, a single scan was performed; on Day B, a scan was performed before (baseline) and after a 2-hour sustained amino acid-infusion. The order of examination days was allocated by computer-generated randomization in blocks of four.

Fig. 2 Participants

Healthy subjects were recruited in the period September 2018 to February 2019 by advertising in the local newspaper. Each participant completed a screening program prior to enrollment consisting of a medical history, clinical examination, electrocardiography, office blood pressure measurement, urine dip stick and the following blood samples: sodium, potassium, creatinine, albumin, alanine aminotransferase, hemoglobin, erythrocyte volume fraction, leucocytes, and thrombocytes. Pregnancy was ruled out in fertile female subjects.

Inclusion criteria were men and women aged 50 to 80 years with a body mass index (BMI) in the range 18.5 to 30.0 kg/m2. Fertile women had to use either hormonal contraceptives or exert sexual abstinence during the entire study period. Exclusion criteria were current use of medicine except contraception, pregnancy or breastfeeding, alcohol intake > 14 units per week for men and > 7 units per week for women, smoking, substance abuse, arterial hypertension, current malignant disease, or signs of clinically relevant renal disease, heart disease, pulmonary disease, liver disease, endocrine disease or neurological disease in the medical history, clinical examination or in the screening tests. Withdrawal criteria were development of one or more exclusion criteria or withdrawal of consent.

Number of subjects

To ensure a power of 80%, significance level of 5%, minimal relevant difference in RBF of 0.11 ml/min/cm3 and a standard deviation (SD) of the difference between two values for the same subject of 0.12 ml/min/cm3, 12 subjects were required for the study. To account for expected dropout, 20 subjects were included.

Pre-scan procedure

In the 24 h preceding each examination day, standardization of fluid intake to 35 ml/kg body weight plain water was implemented, no dietary restrictions were imposed (none of the subjects were on a vegetarian diet) and subjects were instructed to avoid strenuous exercise. On scan days, subjects arrived at 8 am at the Department of Nuclear Medicine, Regional Hospital Herning, Denmark, after a fasting period of 12 h. No fluid intake was allowed during the fast.

Radiopharmaceutical (study drug)

82Rb is acquired for each individual dynamic study scan by elution of a Strontium-82 (82Sr)/82Rb generator (Cardiogen-82; Bracco Diagnostics Inc., Monroe Township, NJ, USA). Necessary quality control procedures were performed on each examination day according to approved guidelines (Bracco Diagnostics Inc.), including tests for 82Sr and 85Sr breakthrough. The system was calibrated to deliver a dose of 555 Megabecquerel (MBq) 82Rb for each injection.

PET/CT-scan

PET/CT scans were carried out on a single PET/CT scanner (Siemens Biograph mCT; 64 slice-4R) with an axial field-of-view (FOV) of 22 cm. The PET/CT scanner was quality controlled and calibrated according to necessary system procedures on each examination day.

On Day A, a peripheral venous catheter was inserted into a cubital vein for 82Rb injection. On Day B, an additional catheter was placed in a cubital vein in the other arm for amino acid-infusion. Subjects rested in a sitting position for about 30 min before voiding. They were then placed in the PET/CT scanner in the supine position with arms above the head and the generator infusion system coupled to the catheter on the left side, after which the subjects underwent a single bed-position PET/CT scan with the AA and the entire kidneys in the same FOV. To position the scanner over the required FOV, an initial scout image was performed followed by a low-dose CT scan for attenuation correction purposes only (25 mAs, 100 kV). Immediately hereafter, a 555 MBq 82Rb bolus injection was administered and a dynamic list-mode PET scan acquired for 8 min [18, 19]. On Day A, the cubital catheter was removed after completion of the scan. On Day B, an intravenous infusion of amino acids was started after the first scan. Subjects rested in a sitting position for the duration of the infusion. The cubital catheters were removed after completion of a second, identical, PET/CT scan which was performed after 120 min of amino acid infusion.

List mode data were rebinned using 37 frames (20 × 3 s, 10 × 12 s, 4 × 30 s and 3 × 60 s) and iteratively reconstructed (21 subsets, 2 iterations) using Siemens TrueX and time-of-flight reconstruction in a 256 × 256 matrix (3.2 × 3.2 × 3.0 mm) and post-filtered with a 5.0 mm Gaussian filter, resulting in attenuation-corrected and decay-corrected dynamic sequences. We did not find it necessary to adjust for motion of the kidneys.

Each participant received a total effective radiation dose < 3.5 milli-Sievert (mSv): Low-dose CT scans contributing 1.2 mSv and each 82Rb bolus injection contributing 1.26 µSv/MBq.

Amino acid-infusion

The amino acid-solution consisted of a mixture of essential and non-essential amino acids (Vamin 18 g N/l, electrolyte free, 1130 mosm/L, Fresenius Kabi AB, Sweden), as described in Table 1. Amino acids were infused with a rate of 0.029 ml/kg/min for a duration of 120 min [26].

Table 1 Composition of amino acids (g/L) in Vamin 18 g N/l, electrolyte free, 1130 mosm/L Analysis of 82Rb PET/CT studies

PMOD software (PMOD Technologies Ltd., Zurich, Switzerland, version 4.102) was used to perform pharmacokinetic modelling.

Data analysis was performed as described in our preceding study [19] with the following minor adjustments: Volumes of interest (VOIs) were defined for the AA and the kidneys (Fig. 3) with AA box-dimensions 10 × 10 × 40 mm3. To define the aortic background and kidney-VOIs, limiting boxes were placed around the organs-of-interest and the cold contour tool (typical cut-off 20–35%) and hot contour tool (typical cut-off 50–55%) used for the respective VOIs. The deliminated kidney-VOIs represent the total renal volume, including both the cortex and the medulla but excluding the renal pelvis; discrimination of the cortex and medulla was not possible.

Fig. 3

VOIs placed in the AA (orange) and aortic background (purple) (a) and contouring the kidneys in one study subject (b). R = right. L = left

Time activity curves for each VOI were obtained to define the AA-derived input function for kinetic modelling and kidney tissue data. Necessary correction for partial volume effects and spill-over in the AA activity was performed using:

$$}_}}\left( } \right) = \beta \times }_}}\left( } \right) + (1 - \beta ) \times }_}}}\left( } \right),$$

(1)

where RA(t) is measured activity in the AA, CA(t) the corrected AA activity, β the required recovery coefficient (RC) (β = 0.612 as determined in phantom studies [19]), and CBg(t) the measured aortic background activity. In contrast, kidneys-VOIs are so large that partial volume and spill-over effects are negligible, obviating the need for background correction. Kidney activity was however corrected for 82Rb count efficiency using a scanner specific RC of 0.659 determined by phantom studies (unpublished), and a blood volume fraction of 10% was assumed to account for activity from the vascular space within the kidney VOIs [27, 28].

For each kidney, K1 values (ml/min/cm3) were calculated using the 1-tissue compartment model (Fig. 1). The uptake rate K1 represents the renal 82Rb clearance (ml/min/cm3) [18, 29] and can, as previously described, be used as an estimate of flow [19] due to high first pass 82Rb extraction [16].

Total flow estimation

Total renal 82Rb clearance can be estimated from the measured 82Rb clearance (K1) and the total kidney volume (VTotal); the estimation of which, is given by the applied kidney-VOIs:

$$}\,}\,} = }_1} \times }_}}}$$

(2)

Total 82Rb clearance was normalized to body surface area (BSA) using the Dubois formula [30]:

$$\text\text\text=0.007184\times\text\text\text\text\text\text^\times\text\text\text\text\text\text^$$

(3)

Units of BSA are m2, with height (cm) and weight (kg).

Statistical analysis

Statistical analysis was performed in SPSS Statistics ver. 20 (IBM Corp., Armonk, NY, USA).

A paired sample t-test was used for comparison of unstimulated K1 values on Day A and B and between K1 values for baseline and response to stimulation on Day B. Day-to-day variation was calculated based on the difference between the unstimulated K1 values on Day A and the unstimulated K1 values on Day B. Statistical significance was defined as p < 0.05.