Biodistribution and post-therapy dosimetric analysis of [¹⁷⁷Lu]Lu-DOTA^ZOL in patients with osteoblastic metastases: first results

In this retrospective study, we analyzed four patients with metastatic skeletal disease secondary to mCRPC or bronchial carcinoma that were treated between July 2016 and September 2017 with [¹⁷⁷Lu]Lu-DOTA^ZOL. All treatments were performed in the context of an individual treatment attempt as no other treatment options were left for these patients. Written informed consent was obtained from all patients. The local ethical committee waived the ethical statement due to the retrospective character of the study. All procedures were followed in accordance with ethical standards of the institutional review board and therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and all subsequent revisions.

After confirming sufficient uptake in the bone metastases with [⁶⁸Ga]Ga-DOTA^ZOL PET/CT shown in Fig. 1e, patients were hospitalized in our treatment unit. All patients had normal kidney function confirmed by renal function tests and renal scintigraphy. Patients were administered a mean activity of 5968 MBq (161.3 mCi) (5873–6000 MBq) of [¹⁷⁷Lu]Lu-DOTA^ZOL intravenously. Table 1 shows patient details along with a history of previous treatments.

DOTA^ZOL was radiolabeled in 0.8 ml ascorbic buffer (210 mg Na-L-ascorbat +42 mg gentisic acid in 1 ml 0.05 N HCl) with non-carrier-added lutetium-177, both obtained from ITG Isotope Technologies Garching GmbH, Garching, Germany. Manual synthesis was carried out on a thermoshaker at a temperature of 95 °C for 30 min. For quality control, an aliquot was retained from the final formulation. Quality control was performed with silica gel–coated aluminum TLC plates (silica 60 F254.5 × 4.5 cm, Merck, Darmstadt, Germany). Analysis was performed with a single trace radio-TLC scanner (PET-miniGITA, Elysia-Raytest, Straubenhardt, Germany) and evaluation software (GinaStar TLC, Elysia-Raytest, Straubenhardt, Germany). Development of silica TLC plates was conducted in 0.1 M citrate buffer (pH 4) mobile phase, whether [¹⁷⁷Lu]Lu-DOTA^ZOL is found at R/F: 0–0.1 and DOTA at R/F: 0.5. Second, developed in 1 M acetylaceton/aceton mixture (1:1) as mobile phase, where [¹⁷⁷Lu]Lu-DOTA-ZOL is found at 0–0.1 and unlabelled lutetium-177 at 0.8–1. At last, a phosphate buffer with MeOH 19:1 as a mobile phase, where lutetium-177 colloide is found at 0–0.2 and [¹⁷⁷Lu]Lu-DOTA^ZOL at 0.8–1. Radio-HPLC was used to determine the radiochemical purity, especially the content of radiolysis products as well as unlabelled gallium-68. RadioHPLC was performed using Agilent 1260 Infinity II reverse phase HPLC system (Agilent Technologies, Santa Clara, California) equipped with GABI γ-HPLC flow detector (Elysia-raytest, Straubenhardt, Germany) and a PC interface running Gina Star software (Elysia-raytest, Straubenhardt, Germany). A Nucleodur 100-3 C18 ec 250/4 column (Macherey-Nagel GmbH & Co. KG, Düren, Germany) was used. The gradient elution system utilized mobile phase A (Puffer) and mobile phase B (MeOH) at a flow rate of 0.5 mL/min isocratic with 90% phase A and 10% phase B. Radioactivity was measured with a dose calibrator (ISOMED 2010, MED Nuklear-Medizintechnik Dresden GmbH, Dresden, Germany). A radiochemical yield of ≥ 95% and a radiochemical purity ≥ 98% was obtained. A comparison of the chemical structures of [¹⁷⁷Lu]Lu-DOTA^ZOL and [¹⁷⁷Lu]Lu-EDTMP can be found in Fig. 2.

Serial whole body planar scintigraphy (anterior and posterior views) was performed with dual head Symbia SPECT/CT system (Symbia T, Siemens Healthineers, Erlangen, Germany) at 20 min and 3, 24, and 167 h post injection (p.i). Acquisition was done in the supine position at a speed of 10 cm/min using medium energy parallel hole collimators with 20% energy window centered at a photopeak of 208 keV. Images were processed using an iterative ordered subset maximization algorithm provided by the manufacturer into a matrix of 256 × 1024. The first data set obtained at 20 min (prior to voiding of the bladder) was considered as a reference with 100% of administered activity. A standard source of known activity was placed between the legs in all images at the time of acquisition. For conversion of counts/min to activity, the gamma camera was pre-calibrated using a known activity of [¹⁷⁷Lu]Lu-DOTA^ZOL and imaging it at the same speed and distance of 10 cm/min.

One to two milliliters of blood samples was drawn at 20 min and 3, 8, 24, and 167 h p.i. Due to high counts that may lead to errors in measurement, 0.2-ml blood samples were prepared and measured along with 0.2-ml sample from the known standard activity of [¹⁷⁷Lu]Lu-DOTA^ZOL using a 1480 WIZARD^TM 3n Gamma counter. The calibration factor determined from the standard activity measurement was used to determine the activity in blood samples at the respective data points.

The images were qualitatively analyzed to assess the biodistribution of [¹⁷⁷Lu]Lu-DOTA^ZOL in the whole body and organs. All organs with uptake equal to or more than that of the kidneys were considered as source organs. Kidneys and urinary bladder showed high uptake in the 24-h image data sets and were considered as source organs that included kidneys and urinary bladder. Adductor muscle was measured as soft tissue reference. Whole body ROI’s were drawn. A rectangular ROI was drawn near the head region above the shoulder for background measurement and an elliptical ROI was used for measurement of the standard source placed between the legs. Same sized ROI’s were replicated on serial images (kidneys ROi’s up to the 24-h data set and all remaining ROI’s in all subsequent image data sets).

Background corrected counts in the right and left kidney, soft tissue, urinary bladder, and whole body were determined on anterior and posterior images. The geometric mean counts/min in all source organs at all data time points was determined. Using EANM dosimetry committee guidelines [31], whole body activity at subsequent time points (T) was determined by multiplying the injected activity with the normalized geometric mean whole body counts at the respective time points as given in Eq. 1.

where t = 20 min, T = subsequent time points, and A₀ = initial injected activity. Likewise, activity in the urinary bladder was also determined by multiplying the injected activity with the normalized geometric mean counts in the urinary bladder.

For calculation of activity in the right and left kidneys at all data points, a conjugate view method with a simple geometrically based subtraction technique given in Eq. 2 [32] was used.

where I_A = anterior count rate, I_P = posterior count rate, f_j represents source organ self-attenuation correction which was calculated from the source region linear attenuation coefficient μ_j, and source thickness t_j using Eq. 3 [32]. Factor μ_et represents the transmission factor across the patient thickness t in the area of the ROI with a linear attenuation coefficient μ_e calculated using Eq. 4 [32]. From [⁶⁸Ga]Ga-DOTA^ZOL- PET/CT of a respective patient, CT-based measurements of source organ thickness as well as whole body thickness and thickness anterior and posterior to source organs at the same level were used. C is the calibration factor determined for gamma camera with a known standard source and was the same in all the studies. For the measurements of μ_j and μ_i (linear attenuation coefficients for whole thickness), we applied a CT-based Hounsfield unit method described by Kabasakal et al. [33] for [¹⁷⁷Lu]Lu-PSMA-617 dosimetric analysis.

A simple geometric based background subtraction technique using Eq. 5 [32] was used.

where I_ADJ is the count rate through the patient from a soft tissue area of the same size as that of the organ ROI. I_A, I_P, t_j, and t are the same as previously defined.

Percent injected activity in the whole body, urinary bladder, and kidneys at all time points was determined. To calculate percent injected activity in the skeleton, from percent whole body activity, percent blood, urinary bladder, and kidneys activities were subtracted.

Percent injected activity in the whole body, kidneys, and skeletal system at all data time points was used to determine residence times (MBq-h/MBq) by fitting biexponential kinetic analysis using OLINDA/EXM version 2.0 (Hermes Medical Solutions, Stockholm, Sweden) software in these organs. The residence times for the skeletal system were assumed to be distributed equally between trabecular and cortical bone. An indirect blood-based method using patient-based red marrow-to-blood ratio (RMBLR) and bone marrow mass was used to determine bone marrow self-dose [34, 35].

Urinary excretion fraction at all time points was determined by applying the function A₀(1 − e^−ƛ,T). With a logarithmic function fit on the urinary excretion curve, effective excretion half-life was obtained. Using the total urinary excretion fraction, effective excretion half-life and 4-h voiding interval as input in Cloutier’s dynamic urinary bladder model, residence time for urinary bladder contents was obtained. By subtracting residence times for kidneys, bone marrow, and skeletal system from whole body residence time, the remainder of body residence time was calculated.

Residence time for kidneys, cortical and trabecular bone, urinary bladder contents, red marrow self-dose, and remainder of body were used as an input in OLINDA/EXM version 2.0 (Hermes Medical Solutions, Stockholm, Sweden) software for calculation of organ absorbed doses and effective doses after adjusting the weight of patient organs by multiplying the reference adult male weight with factor obtained by dividing patient weight with the reference adult male weight. The mean of residence times and organ absorbed doses (mSv/MBq) were calculated.

Figures 1 and 3 show the biodistribution of [¹⁷⁷Lu]Lu-DOTA^ZOL in one patient with bronchial carcinoma and one patient with mCRPC, respectively. In the initial 20 min p.i. image data set, highest uptake was seen in the urinary bladder followed by kidneys and soft tissue with minimal uptake in the skeletal system. The kidneys showed a rapid decrease in activity at 3 h with minimum to no uptake after 24 h p.i. The intense uptake was seen in the skeletal system from 3 h onwards. Blood and soft tissue clearance and lesion to normal bone contrast increased in later images up to 168 h. The mean ± SD 24-h whole body retention was found to be 31.25 ± 6.5.

In this small patient study, we observed fast uptake and clearance kinetics of kidneys in patients with bronchial carcinoma compared with mCRPC patient, which resulted in better skeletal to soft tissue contrast as early as 3 h p.i. in the bronchial carcinoma patient compared with 24 h p.i. in mCRPC patient.

Mean residence times (MBq-h/MBq) (Table 2) was found to be highest in trabecular and cortical bone (31.9 h) followed by the remainder of the body (11.7 h), kidneys (1.84 h), urinary bladder (1.52 h), and bone marrow (0.03 h). In patient no. 1, residence times for the skeletal system and the kidney were lower compared with the other patients.

Mean organ absorbed doses (Table 3) were found highest (3.33 ± 0.35 mSv/MBq) for osteogenic cells, followed by kidneys (0.49 ± 0.16 mSv/MBq), red marrow (0.461 ± 0.064 mSv/MBq), and urinary bladder wall (0.322 ± 0.022 mSv/MBq). Kidney and osteogenic cell absorbed doses were lowest in patient no. 1. The mean total body dose was 0.092 ± 0.033 mSv/MBq.