Preclinical evaluation of [²²⁵Ac]Ac-DOTA-TATE for treatment of lung neuroendocrine neoplasms

Abstract

Purpose

There is significant interest in the development of targeted alpha-particle therapies (TATs) for treatment of solid tumors. The metal chelator-peptide conjugate, DOTA-TATE, loaded with the β-particle emitting radionuclide ¹⁷⁷Lu ([¹⁷⁷Lu]Lu-DOTA-TATE) is now standard care for neuroendocrine tumors that express the somatostatin receptor 2 (SSTR2) target. A recent clinical study demonstrated efficacy of the corresponding [²²⁵Ac]Ac-DOTA-TATE in patients that were refractory to [¹⁷⁷Lu]Lu-DOTA-TATE. Herein, we report the radiosynthesis, toxicity, biodistribution (BD), radiation dosimetry (RD), and efficacy of [²²⁵Ac]Ac-DOTA-TATE in small animal models of lung neuroendocrine neoplasms (NENs).

Methods

[²²⁵Ac]Ac-DOTA-TATE was synthesized and characterized for radiochemical yield, purity and stability. Non-tumor–bearing BALB/c mice were tested for toxicity and BD. Efficacy was determined by single intravenous injection of [²²⁵Ac]Ac-DOTA-TATE into SCID mice–bearing human SSTR2 positive H727 and H69 lung NENs. RD was calculated using the BD data.

Results

[²²⁵Ac]Ac-DOTA-TATE was synthesized with 98% yield, 99.8% purity, and displayed 97% stability after 2 days incubation in human serum at 37 °C. All animals in the toxicity study appeared healthy 5 months post injection with no indications of toxicity, except that animals that received ≥111 kBq of [²²⁵Ac]Ac-DOTA-TATE had chronic progressive nephropathy. BD studies revealed that the primary route of elimination is by the renal route. RD calculations determined pharmacokinetics parameters and absorbed α-emission dosages from ²²⁵Ac and its daughters. For both tumor models, a significant tumor growth delay and time to experimental endpoint were observed following a single administration of [²²⁵Ac]Ac-DOTA-TATE relative to controls.

Conclusions

These results suggest significant potential for the clinical translation of [²²⁵Ac]Ac-DOTA-TATE for lung NENs.

Introduction

Neuroendocrine tumors (NETs) arise from the neuroendocrine system in the lung, gastrointestinal tract, and pancreas, among other organs [1]. A large fraction (25%) of well-differentiated NETs are of bronchopulmonary origin. Well-differentiated and moderately differentiated lung NETs are referred to as lung carcinoids and are divided into typical and atypical classifications. Poorly differentiated lung neuroendocrine carcinomas are divided into large cell neuroendocrine carcinoma and small cell lung cancer (SCLC) [2, 3]. Few trials have investigated systemic treatments for patients with lung NETs [4–8]. The only medication approved by the US Food and Drug Administration (FDA) for progressive lung NETs is everolimus, an oral inhibitor of the mechanistic target of rapamycin (mTOR) [9].

Overexpression of somatostatin receptors in patients with neuroendocrine neoplasms (NEN) is utilized for both diagnosis and treatment. Somatostatin receptor 2 (SSTR2) is widely expressed in pulmonary NENs, including typical carcinoids, atypical carcinoids, large cell neuroendocrine carcinoma, and SCLC [10]. Therefore, somatostatin analogues (SSAs) with high affinity to SSTR2 can be used for targeting these types of tumors [10]. Peptide receptor radionuclide therapy (PRRT) with the [¹⁷⁷Lu]Lu-DOTA chelate conjugated to the octreotate SSA ([¹⁷⁷Lu]Lu-DOTA-TATE) has proved to be an effective therapy option for NETs [6, 7, 11].

The limitations of [¹⁷⁷Lu]Lu-DOTA-TATE therapy are that only 26–55% of patients achieve stable disease, 18–32% are treatment refractory, and half progress within 2 to 3 years after starting treatment [12–14].

¹⁷⁷Lu is a β-particle emitting radionuclide and there is evidence that a superior therapeutic index could be achieved using targeted α-particle therapy (TAT) as an alternative [15–20]. Compared to β-particles, α particles have a 200–400× greater linear energy transfer and a shorter range in solid tissue, <100 μm for α-emissions relative to a few mm for β-emissions. Hence, α-particle emissions are more likely to deposit energy within the boundaries of a tumor or metastasis relative to β-emissions, which can cause toxicity in surrounding normal tissues, and greater energy deposition leads to greater cell killing. Unlike β-emissions, it is hypothesized that α-emissions do not rely solely on generation of free radicals to generate DNA damage. Instead, the energy deposited is sufficient to directly cause DNA double-strand breaks [21]. This enables TATs to potentially evade a common mechanism of radiation resistance, i.e., free radical scavenging. To demonstrate these advantages, [²²⁵Ac]Ac-DOTA-TOC was compared to [¹⁷⁷Lu]Lu-DOTA-TOC in the treatment of rat pancreatic exocrine tumor xenografts in immunocompromised mice and TAT demonstrated greater efficacy and lower toxicity relative to the TBT [16]. In recently published case studies, significant efficacy of [²²⁵Ac]Ac-DOTA-TATE was demonstrated in patients’ refractory to [¹⁷⁷Lu]Lu-DOTA-TATE [22]. Furthermore, [¹⁷⁷Lu]Lu-DOTA-TATE has been compared with [¹⁷⁷Lu]Lu-DOTATOC in patients and a greater tumor residence time was observed for [¹⁷⁷Lu]Lu-DOTA-TATE, indicating that DOTA-TATE may be a superior targeting ligand for delivery of radiotherapy [23].

²²⁵Ac is an α-particle emitting isotope that has demonstrated significant utility in TAT and decays with a 10-day half-life via a complex decay scheme, emitting four α particles and depositing significantly higher energy levels in local tissue, 28 MeV, compared to single α-particle emitters, e.g., 7.8 MeV for ²¹²Bi (or ²¹²Pb), or 384 keV for the β-emitter ¹⁷⁷Lu [24, 25]. Hence, ²²⁵Ac has been described as an in vivo α-particle generator [26]. The main method for generating ²²⁵Ac for clinical studies is through the decay of ²²⁹Th which originates from ²³³U. There are three main sources of ²²⁹Th in the world: Oak Ridge National Laboratory (USA), The Institute of Physics and Power Engineering (Russia), and The Institute for Transuranium Elements (Germany). In addition, it has been shown that large-scale quantities can be produced through the decay of ²³²Th [27, 28]. To address the shortage of the clinical need of ²²⁵Ac, the US Department of Energy formed a Tri-lab collaboration of Los Alamos (LANL), Brookhaven (BNL), and Oak Ridge (ORNL) National Laboratories with the goal of developing an alternative route for production of ²²⁵Ac by use of a linear accelerator [29]. Due to the long half-life, ²²⁵Ac-based TATs can be produced at regional radiopharmaceutical facilities and distributed for use by Nuclear Medicine Departments within 1 to 3 days post-production.

Herein, we investigate the toxicity, biodistribution (BD), radiation dosimetry (RD) in mice, and efficacy of [²²⁵Ac]Ac-DOTA-TATE in xenograft tumor models of pulmonary NENs. These preclinical studies are needed to support the clinical translation of [²²⁵Ac]Ac-DOTA-TATE.

Materials and methods

Cell culture

NCI-H69 human SCLC, NCI-H727 human lung carcinoid, and HEK293 cells were purchased from American Type Culture Collection, expanded for 2 passages and cryopreserved. NCI-H69 and NCI-H727 cells were grown in RPMI-1640 Medium, (Life Technologies), 10% FBS (Life Technologies), 100 units/mL penicillin, 100 mg/mL streptomycin in 5% CO₂ at 37 °C. HEK293 cells were grown in DMEM/F12 medium (Life Technology). Cells were authenticated using short tandem repeat (STR) DNA typing according to ATCC’s guidelines [30], monitored by microscopy to assure maintenance of their original morphology, tested for mycoplasma contamination using the MycoAlert kit (Lonza) and experiments used cells of passage numbers <25.

SSTR2 transfection of HEK293 cells

HEK293 cells were transfected with 2 μg of SSTR2 (NM_001050) Human Tagged ORF Clone in pCMV6 vector with Neomycin selection marker (Origene) using the FuGENE HD Transfection Reagent (Promega) as described before [31]. SSTR2 expression was determined by quantitative real-time RT-PCR (qRT-PCR) using the primers described below. The HEK293/SSTR2 clone with the highest expression was maintained in medium containing 400 μg/mL of G418.

Chelation of lanthanum and europium

DOTA-(Tyr3)-octreotate (DOTA-TATE), ≥98% purity, was purchased (Bachem). DOTA-TATE was loaded with lanthanum (La³⁺) by addition of LaCl₃·7H₂O (3.9 mg, 10.5 μmol) to DOTA-TATE (5.0 mg, 3.5 μmol) in 1 mL of 0.1 M AcONa buffer (pH 6.0) and europium (Eu³⁺), by addition of EuCl₃·6H₂O (1.6 mg, 4.2 μmol) to DOTA-TATE (2 mg, 1.4 μmol) in 5 mL H₂O and 0.1 mL DMSO, and stirring 12 h at room temperature (RT). For both chelates, reaction completion was determined by HPLC (Agilent 1260 Infinity II HPLC system with a quaternary pump, a vial sampler, and a DAD detector) and a C18 column yielded the final product (~3.3 mg and 1.9 mg, respectively). A Phenomenex C18 column (Luna 5 μL C18(2) 100 Å, 4.6 × 250 mm) was used. The DAD detector was set to 214 nm. Phase A: 0.1% TEA/AcOH (TEAA) in water (pH 6.0) and Phase B: 90% acetonitrile in Phase A (pH adjusted to 6.0 with AcOH) using the following: 10–45% Phase B in Phase A in 50 min. Mass and purity were determined by HRMS (ESI) and HPLC, respectively.

Binding assay

The receptor number of HEK293/SSTR2 cells was determined using time-resolved fluorescence (TRF) saturation binding assay. Different concentrations of Eu³⁺-DOTA-TATE (0–2500 nM) were used as the test ligand and 5 μM of La³⁺-DOTA-TATE was used as the blocking ligand. Briefly, black 96-well plates (clear bottom, Corning, #3603) were coated with 0.25 mg/ml of PDL (Poly-D Lysine, Sigma). Hek293/SSTR2 cells were plated in the coated plates, at a density of 30,000 cells/well. On the day of the experiment (1 day after seeding the cells), the medium was aspirated and for the top half of the plate (total binding) 50 μL of the Eu³⁺-DOTA-TATE (2.5–2500 nM) test ligand was added to each well in a series of decreasing concentrations, followed by 50 μL of binding medium (DMEM, 1 mM 1,10-phenanthroline, 200 mg/L bacitracin, 0.5 mg/L leupeptin, 0.3% BSA). The bottom half of the plate (non-specific binding), was prepared as the top half, except that 5 μM of blocking La³⁺-DOTA-TATE ligand was added instead of binding medium. Cells were incubated with the ligands for 1 h at 37 °C in a cell culture incubator, followed by two washes with PBS. After washing, the cells were incubated with 50 μl of 2.0 M HCl for 2 h at 37 °C followed by neutralization with 55 μl of 2.0 M NaOH. Then, 115 μL of enhancement solution (PerkinElmer) was added to each well and cells were incubated for an additional 30 min at 37 °C prior to reading. The plates were read on a PerkinElmer VICTORx4 2030 multilabel reader using the standard Eu³⁺ time-resolved fluorescence (TRF) measurement (340-nm excitation, 400-μs delay, and emission collection for 400 μs at 615 nm). The standard curve was used to determine the amount of ligand present at the B_max obtained in the saturation binding assay.

The average number of cells per well at the end of the assay was calculated. To determine the receptor number, the following equation was used: (Eu³⁺ amount for B_max (moles)/average cell number per well) × 6.023 × 10²³ =mreceptor number per cell. This B_max value was then used to calculate ligand association and binding kinetics for each cell using the following equation: B = B₀ + (B_max-B₀)/(1−e^−kt), where B corresponds to a receptor saturation parameter, an analog of ligand-receptor complex formation [RL], with values between initial saturation B₀ and the maximum saturation B_max; k is the reaction rate constant, and t is time. NCI-H69 receptor number was determined as above except that ligand incubation step was reduced to 30 min with only × 1 PBS wash.

La³⁺-DOTA-TATE binding affinity was determined using a TRF competitive binding assay as previously [32], using HEK293/SSTR2 cells and 50 nM Eu³⁺-DOTA-TATE used as the competing ligand. Data points were acquired in quadruplicate and each assay was repeated 3 times.

Radiochemical synthesis and characterization

The ²²⁵Ac(NO₃)₃ (5.80 × 10⁴ Ci/g; carrier free; >10 mCi/ml concentration) was purchased from Oak Ridge National Laboratory (TN, USA). Complexation was achieved by reacting DOTA-TATE (5–10 μg in 5–10 μL water from 1.0 mg/mL solution) with ²²⁵Ac(NO₃)₃ (3.4 MBq) that was diluted in 100 μL of water containing 10 μL of 20% L-ascorbic acid. The pH was adjusted to 5.8 using 1 M Tris buffer (10–12 μL), and then incubated at 75 °C for 1 h (Scheme 1). Reaction progress and radiochemical purity of [²²⁵Ac]Ac-DOTA-TATE were measured without further purification 24 h after collection (to ensure secular equilibrium among ²²⁵Ac and its daughters) using ITLC with γ counting, radio-TLC, and γ counting of radio-HPLC fractions. Stability was determined by adding 50 μL of [²²⁵Ac]Ac-DOTA-TATE (1702 kBq) to 1 mL of human serum (n = 4), 37 °C incubation for 10 days, and TLC scanner (Bioscan) and γ-counter (PerkinElmer) quantification at time points [33].

Scheme 1.

Radiochemical synthesis of [²²⁵Ac]Ac-DOTA-TATE

Animal studies

Protocols were approved: University of South Florida IACUC protocol IS00006466. Animals were purchased from Charles River (Wilmington, MA). An equal number of males and females were used for each cohort.

Injected activity measurement

Syringes were prefilled with ± 10% of [²²⁵Ac]Ac-DOTA-TATE activity. Activities were measured using the Atomlab 500 Dose Calibrator (Biodex) as described before [34].

Toxicity

Cohorts (n = 6) of non-tumor-bearing BALB/c mice (6–8 weeks old) were given a single intravenous injection by tail-vein catheter of 200 μl of 55.5, 111 or 185 kBq of [²²⁵Ac]Ac-DOTA-TATE activity in 0.9% sterile saline (Cardinal Pharmaceuticals) in the syringe (45.8 ± 14.8, 94.7 ± 4.8, and 163.9 ± 9.6 mean injected activity), or 0.9% sterile saline alone as a control. These administration activities were selected based on results from a previous preclinical study using an ²²⁵Ac-based peptidic TAT [34]. Injected activity is determined for each animal by subtracting the amount of activity remaining in the syringe and catheter from the pre-injection amount of activity in the syringe. Mice were monitored for distressed behavior and weighed twice per week for 5 months, followed by euthanasia and tissue collection. Blood urea nitrogen (BUN), creatinine, alkaline phosphatase (ALKP), alanine aminotransferase (ALT), and aspartate aminotransferase (AST) in serum were analyzed using an automated biochemical analyzer (CLIP, Catalyst DX, IDEXX). Bone, bone marrow, brain, cecum, duodenum, kidney, liver, lymph nodes, muscle, pancreas, salivary gland, small intestine, and spleen were harvested and fixed in 10% formalin. Bone was decalcified in 14% ethylene diaminetetraacetic acid (EDTA) solution after fixation. Tissues were embedded in paraffin, sectioned (4–6 μm thickness), stained with hematoxylin and eosin (H&E), and examined blind by the veterinary pathologist (R.E.).

Biodistribution

Non-tumor-bearing BALB/c mice (6–8 weeks old) were administered by tail-vein catheter 200 μl of 74 kBq of [²²⁵Ac]Ac-DOTA-TATE activity in 0.9% sterile saline in the syringe (66.6 ± 2.9 kBq mean injected activity), euthanized at 24, 48, and 96 h p.i., n = 4 mice per cohort. This administration activity was selected based on results from a previous preclinical BD study using an ²²⁵Ac-based peptidic TAT [34]. Tissues were removed, weighed, and the percent injected activity per gram (%IA/g) calculated as before [34].

Radiation dosimetry

Dosimetry calculations were performed using BD data as previously described [34].

qRT-PCR and immunocytochemistry

SSTR2 expression was verified by qRT-PCR and ICC performed to determine protein. NCI-H69 and NCI-H727 RNA was extracted (RNeasy, Qiagen). Two sets of human SSTR2 primers were designed (Gene Runner for Windows 3.05): set 1: forward, 5’-AACCAGACAGAGCCGTACTA-3′ and reverse, 5′- GCATAGCGGAGGATGACA-3′; and set 2: forward, 5’-GCTGGCTTCCCTTCTACATATT-3′ and reverse, 5’-GAGGACCACCACAAAGTCAA-3′. B-actin was used for normalization as previously [35]. NCI-H69 and NCI-H727 ICC was performed using 30 μg/mL anti-SSTR2 antibody (Sigma, HPA007264) as before [34], except that H69 were grown on coverslips coated with 0.25 mg/ml poly-D-lysine (Sigma).

Anti-tumor efficacy

NCI-H69 and NCI-H727 subcutaneous xenograft tumor–bearing SCID mice (6–8 weeks old, n = 10/cohort) were injected by tail-vein catheter with 200 μl of 185 kBq [²²⁵Ac]Ac-DOTA-TATE activity in 0.9% sterile saline in the syringe (148.0 ± 17.8 and 144.3 ± 18.8 kBq mean injected activity), or 0.9% sterile saline alone as a control. This administration activity was selected based on preliminary results from the toxicity study herein. The efficacy study was initiated 4–1/2 months after initiation of the toxicity study and no toxicities were detected at this activity at that time. Animals underwent euthanasia after reaching 2000 mm³ tumor volume or if clinical endpoints were observed, e.g., 20% weight loss, tumor ulceration, hunched back, lack of grooming, and lethargy.

Histology and immunohistochemistry

Following euthanasia, organs (toxicity) and tumors (efficacy) were excised, formalin fixed, paraffin embedded, sectioned (5 μm), and stained with H&E (tumors and organs) and IHC (tumors) using rabbit SSTR2 polyclonal antibody (1:200 dilution, GTX70735, GeneTex) and a Discovery XT automated system (Ventana Medical Systems). Slides were scanned using a ScanScope XT digital slide scanner (Aperio, CA). To quantify SSTR2 expression in tumors, images from serial H&E and IHC central sections were analyzed using Visiopharm 2017.7 as before [34].

SSTR2 expression was evaluated in using gray-level co-occurrence texture analysis in three main steps: (1) tumor segmentation, (2) target expression extraction, and (3) texture analysis. The tumor region was segmented to limit the analysis to cells within the tumor. For target expression, a binary mask was created from the positive and negative stained cells with each pixel representing a cell with the value of 1 if unstained or 2 if stained. A co-occurrence texture analysis was applied to the binary mask to calculate the heterogeneity score for target expression. Specifically, the diagonal values in the gray-level co-occurrence matrix measure the closeness of the distribution of all gray levels. Therefore, heterogeneity is calculated from diagonal entries that are related, i.e., with target expression. The heterogeneity score ranges from 0 to 1, where 0 is completely homogeneous and 1 is completely heterogeneous.

Statistics

Cohort size was determined by power analyses. Analysis of variance (ANOVA) was used to assess differences in toxicity by activity and Dunnett’s test to determine differences among activities and saline control. Prior to ANOVA, the Anderson Darling test assessed whether a transformation was required. The unpaired t-test compared the ICC or IHC expression level of SSTR2 in NCI-H69 and NCI-H727 cells and tumors. Twoway ANOVA was used to compare tumor volume among treatment and saline groups and Kaplan-Meier analyses (GraphPad Prism 7 software) determined significance among the cohorts.

Results

Binding affinity of La³⁺-DOTA-TATE

There are no non-radioactive isotopes of actinium and lanthanum is a useful surrogate for Ac since both are trivalent ions in solution [36]. The HRMS (ESI) calculated mass for the prepared La³⁺-DOTA-TATE, C₆₅H₈₇LaN₁₄O₁₉S₂ (M + H)⁺, 1571.4855 MW, was 1571.4867 with 100% purity, t_R = 22.15 min (Supplemental Fig. 1). La³⁺-DOTA-TATE bound SSTR2 expressing HEK293/SSTR2 cells with high, 19.00 ± 9.2 nM K_i (Fig. 1).

Fig. 1.

Whole-cell competitive binding assay of La³⁺-DOTA-TATE. The binding affinity of La³⁺-DOTA-TATE was calculated to be 19.00 ± 9.2 nM K_i (n = 3 repeats). In this whole-cell lanthanide time-resolved fluorescence assay, HEK293/SSTR2 cells were used and 50 nM of Eu³⁺-DOTA-TATE was used as the competing ligand

The HRMS (ESI) calculated mass for the Eu³⁺-DOTA-TATE chelate, C₆₅H₈₇EuN₁₄O₁₉S₂ (M + H)⁺, 1585.5004 MW, was 1585.5031 with 100% purity, t_R = 22.14 min (Supplemental Fig. 2). Eu³⁺-DOTA-TATE binding affinity for SSTR2 was 22.12 nM K_d, and HEK293/SSTR2 cells were shown to express 1.87 × 10⁶ SSTR2 receptors on the cell-surface per cell (Supplemental Fig. 3A).

Radiosynthesis and characterization of [²²⁵Ac]Ac-DOTA-TATE

Radiosynthesis (Scheme 1) provided >98% yield with high radiochemical purity (≥99.8%) (Supplemental Fig. 4). Also, [²²⁵Ac]Ac-DOTA-TATE had an excellent in vitro stability, with 97% integrity remaining after 2 days in human serum at 37 °C (Table 1).

Table 1.

In vitro serum stability of [²²⁵Ac]Ac-DOTA-TATE

Day	% Intact TLC scanner
0	100
1	100.44±0.88
3	99.27±0.29
5	96.74±2.06
7	93.19±1.50
10	90.29±4.00

Toxicity

Weight gain was observed in all animals at 5 months post injection (p.i.), albeit less weight was gained by animals at the highest dose level relative to the lowest (Fig. 2a). Significant differences in weight change were observed among saline and 111 kBq and 185 kBq treatments (p < 0.01 and p < 0.001, respectively), but not 55.5 kBq (p = 0.50) (Supplemental Table 1). Blood BUN and creatinine were all in the normal range (Fig. 2b and c), p values = 0.12 and 0.33, respectively (Supplemental Table 2 and 3). No significant differences in ALT, AST, or ALKP liver enzymes were observed among the cohorts (Fig. 2d–f and Supplemental Table 4–6).

Fig. 2.

Toxicity of [²²⁵Ac]Ac-DOTA-TATE in BALB/c mice. a Percent weight gain, b BUN (reference range: 18–29 mg/dl), c blood creatinine (reference range: 0.2–0.8 mg/dL), d ALT (reference range: 28–132 U/L), e AST (reference range: 59–247 U/L), and f ALKP (reference range: 62–209 U/L)

No pathologic findings were observed for 55.5 and 111 kBq cohorts. In contrast, animals in cohorts receiving ≥111 kBq in the syringe (≥94.8 kBq actual injected activity) began to lose weight at ~100 days p.i. (Supplemental Fig. 5), and at necropsy chronic progressive nephropathy was observed (Fig. 3). No other organ (bone, bone marrow, brain, cecum, duodenum, liver, lymph nodes, muscle, pancreas, salivary gland, small intestine, and spleen) showed pathologic changes related to the treatment. Glycogen accumulation was observed in the livers of some animals (including the saline control group), but this was considered an incidental finding.

Fig. 3.

Histological appearance of kidney from a mouse administered a saline or b 179.0 kBq. (a) Kidney histology of a saline administered mouse, without significant abnormalities, where glomeruli are normocellular, with open capillary loops, and tubular epithelium is uniformly cuboidal, with round nuclei, and a prominent brush border. (b) Kidney histology of a 179.0-kBq administered mouse, with chronic progressive nephropathy. Extensive tubular cell regeneration, diffuse fibrosis, mild mononuclear inflammatory cell infiltrates, and thickened, hypercellular glomerular tufts are evident

Biodistribution

BALB/c mice (n = 4 per cohort) were injected intravenously with 74 kBq of [²²⁵Ac]Ac-DOTA-TATE α activity in the syringe. The activities from ²²⁵Ac and daughters ²²¹Fr and ²¹³Bi in equilibrium were determined for each organ and time point using γ spectroscopy (Fig. 4a–c and Supplementary Fig. 6). [34, 37] At 24 h p.i., the kidneys, liver, and stomach had 1.63 ± 0.72, 0.147 ± 0.07, and 0.31 ± 0.08 %IA/g, respectively, while only negligible activity was observed in the other tissues measured. Activity had largely cleared from the tissues by 96 h p.i. Since at the time of the BD measurement, ²²⁵Ac was in equilibrium with its daughters, whereas the actual distributed daughters had been decayed, ²²¹Fr and ²¹³Bi only represent the biodistribution of ²²⁵Ac.

Fig. 4.

Biodistribution of a ²²⁵Ac, b ²²¹Fr, and c ²¹³Bi after intravenous injection of [²²⁵Ac]Ac-DOTA-TATE in BALB/c mice

Radiation Dosimetry

Radiation dosimetry (RD) calculations were based on the data obtained from the BD studies at 24, 48 and 96 h p.i. The α-particle dose from ²²⁵Ac and each of its α-emitting daughters was calculated (Supplementary Fig. 6). The decay of ²²¹Fr to ²¹⁷At was assumed to take place in the same location as ²²¹Fr. Similarly, the decay of ²¹³Po was assumed to take place at the same location as ²¹³Bi, while accounting for its 98% branching ratio.

BD data for the different tissues were fitted using an exponential decay nonlinear regression, allowing the estimation of the initial activity due to uptake (A_o), the decay rate constant (λ_eff), decay half-life (T_eff), accumulated activity (Ã), and absorbed dose/injected activity (Gy/kBq) for each radionuclide in each tissue (Table 2). The total absorbed dose is the summation of the values for the five α-emitting radionuclides. The calculated total absorbed dose for [²²⁵Ac]Ac-DOTA-TATE was minimal in all tissues except kidney and liver, with 0.0068 and 0.0059 Gy/kBq, respectively. Figure 5 represents graphs of the absorbed dose from each radionuclide per tissue.

Table 2.

Radiation dosimetry and clearance kinetics parameters for [²²⁵Ac]Ac-DOTA-TATE in BALB/c mice

Parameter	Blood	Bone	Brain	Heart	Intestine	Kidney	Liver	Lung	Muscle	Pancreas	Salivary	Skin	Spleen	Stomach
225Ac
Initial activity/organ, A○ (kBq)	0.0003	0.0131	0.0019	0.0033	0.1527	0.3862	0.1200	0.0227	0.0024	0.0171	0.0028	0.0023	0.0039	0.0859
Decay rate constant, λeff(h−1)	0.0206	0.0030	−0.0103	−0.0004	0.0200	0.0250	−0.0053	0.0130	−0.0049	0.0060	−0.0077	−0.0092	0.0062	0.0190
Decay half-life, Teff (days)	1.4020	9.6270	2.8040	72.2028	1.4441	1.1552	5.4493	2.2216	5.8941	4.8135	3.7508	3.1393	4.6582	1.5201
Accumulated activity/organ,$\tilde{A}\left(\text{kBq}\times \text{h}\right)$	0.0080	1.4274	0.8190	0.4871	4.4589	8.2462	29.4458	1.0829	0.5613	1.4271	0.8906	0.8656	0.3228	2.6801
Absorbed dose/injected activity (Gy/kBq)	0.0000	0.0004	0.0001	0.0001	0.0001	0.0012	0.0012	0.0001	0.0001	0.0003	0.0002	0.0001	0.0001	0.0004
221Fr
Initial activity/organ, A○ (kBq)	0.0019	0.0170	0.0036	0.0067	0.01532	0.4011	0.1612	0.0251	0.0052	0.0169	0.0037	0.0039	0.0097	0.0852
Decay rate constant, λeff (h−1)	−0.0071	0.0100	−0.0044	0.0190	0.0130	0.0190	−0.0033	0.0080	0.0170	0.0080	−0.0028	−0.0015	0.0030	0.0170
Decay half-life, Teff (days)	4.0678	2.8881	6.5639	1.5201	2.2216	1.5201	8.7519	3.6101	1.6989	3.6101	10.3147	19.2541	9.6270	1.6989
Accumulated activity/organ, $\tilde{A}\left(\text{kBq}\times \text{h}\right)$	0.5594	1.0195	0.7934	0.2102	7.2991	12.5136	32.1637	1.7734	0.1848	1.1933	0.7036	0.6489	1.0604	3.0452
Absorbed dose/injected activity (Gy/kBq)	0.0001	0.0003	0.0001	0.0000	0.0002	0.0019	0.0014	0.0003	0.0000	0.0003	0.0002	0.0001	0.0004	0.0004
217At
Initial activity/organ, A○ (kBq)	0.0019	0.0170	0.0036	0.0067	0.1532	0.4011	0.1612	0.251	0.0052	0.0169	0.0037	0.0039	0.0097	0.852
Decay rate constant, λeff (h−1)	−0071	0.0100	−0.0044	0.0190	0.0130	0.0190	−0.0033	0.0080	0.0170	0.0080	−0.0028	−0.0015	0.0030	0.0170
Decay half-life, Teff (days)	4.0678	2.8881	6.5639	1.5201	2.2216	1.5201	8.7519	3.6101	1.6989	3.6101	10.3147	19.2541	9.6270	1.6989
Accumulated activity/organ, $\tilde{A}\left(\text{kBq}\times \text{h}\right)$	0.5594	1.0195	0.7934	0.2102	7.2991	12.5136	32.1637	1.7734	0.1848	1.1933	0.7036	0.6489	1.0604	3.0452
Absorbed dose/injected activity (Gy/kBq)	0.001	0.0003	0.0001	0.00001	0.0002	0.0021	0.0016	0.00003	0.00000	0.0003	0.0002	0.0001	0.0004	0.0005
213Bi
Initial activity/organ, A○ (kBq)	0.0042	0.0125	0.0066	0.0053	0.1340	0.3639	0.0948	0.0270	0.0034	0.0119	0.0056	0.0056	0.0068	0.0777
Decay rate constant, λeff (h−1)	0.01100	0.0070	0.0080	0.0030	0.0210	0.0250	−0.0073	0.0170	0.0050	0.0140	0.0030	0.0040	−0.0024	0.0190
Decay half-life, Teff (days)	2.6256	4.1259	3.6101	9.6270	1.3753	1.1552	3.9563	1.6989	5.7762	2.0629	9.6270	7.2203	12.0338	1.5201
Accumulated activity/organ, $\tilde{A}\left(\text{kBq}\times \text{h}\right)$	0.2345	0.9591	0.4653	0.5789	3.6681	7.7700	28.7854	0.9636	0.3058	0.5253	0.6047	0.5575	1.2470	2.4230
Absorbed dose/injected activity (Gy/kBq)	0.0000	0.0000	0.0000	0.0000	0.0000	0.0000	0.0000	0.0000	0.0000	0.0000	0.000	0.000	0.0000	0.0000
213Po
Initial activity/organ, A○ (kBq)	0.0042	0.0125	0.0066	0.0053	0.1340	0.3639	0.0948	0.0270	0.0034	0.0119	0.0056	0.0056	0.0068	0.0777
Decay rate constant, λeff (h−1)	0.0110	0.0070	0.0080	0.0030	0.0210	0.0250	−0.0073	0.0170	0.0050	0.0140	0.0030	0.0040	−0.0024	0.0190
Decay half-life, Teff (days)	2.6256	4.1259	3.6101	9.6270	1.3753	1.1152	3.9563	1.6989	5.7762	2.0629	9.6270	7.2203	12.0338	1.5201
Accumulated activity/organ, $\tilde{A}\left(\text{kBq}\times \text{h}\right)$	0.2345	0.9591	0.4653	0.5789	3.6681	7.7700	28.7854	0.9636	0.3058	0.5253	0.6047	0.5575	1.2470	2.4230
Absorbed dose/injected activity (Gy/kBq)	0.0000	0.0004	0.0001	0.0002	0.0001	0.0015	0.0016	0.0002	0.0001	0.0002	0.0002	0.0001	0.0006	0.0005
Total absorbed dose/injected activity (Gy/kBq)	0.0002	0.0014	0.0004	0.0004	0.0007	0.0068	0.0059	0.0009	0.0003	0.0011	0.0008	0.0004	0.0016	0.0017

Fig. 5.

Radiation dosimetry of ²²⁵Ac and daughters following administration of [²²⁵Ac]Ac-DOTA-TATE in BALB/c mice

The effective decay half-lives (T_eff) calculated for ²²⁵Ac in tissues that had significant uptake were shorter than the radioactive decay half-life of ²²⁵Ac (10 days), indicating biological clearance. For example, the calculated T_eff in kidney and liver were 1.16 and 5.45 days, respectively (Table 2). Hence, T_eff is a composite of radioactive decay and active biological clearance. The T_eff was only calculated to be longer in tissues with minimal uptake where instrument background likely interfered with the accuracy of measurement. It is notable that the kidneys had the highest initial activity, 0.39 kBq, relative to liver, 0.12 kBq, but the kidneys had a nearly 5-fold lower effective half-life compared to liver, leading to faster kidney clearance relative to liver and a higher but comparable total absorbed dose.

SSTR2 expression in NEN cells

NCI-H69 cells have significantly higher SSTR2 mRNA (p < 0.001) and protein expression (p < 0.05) relative to NCI-H727 (Fig. 6). The SSTR2 cell-surface receptor number per NCI-H69 cell was 1.60 × 10⁶ (Supplemental Fig. 3b). SSTR2 expression in NCI-H727 cells was not high enough to accurately calculate the receptor number.

Fig. 6.

SSTR2 expression on NCI-H69 and NCI-H727 cells. a qRT-PCR. Note the Log10 scale for the expression level. b ICC using DAPI (blue), WGA (green), and anti-SSTR2 antibody (red)

Anti-tumor efficacy

Eight days after tumor cell xenoengraftment, mice (n = 10 per cohort) were intravenously injected with a single bolus of 148.7 kBq or 144.3 kBq mean ²²⁵Ac activity for H69 and H727 cohorts, respectively, or saline solution. Figure 7a and b show representative images at 25 days p.i. Tumor volumes decreased significantly after treatment relative to saline controls (p < 0.001) prior to eventual regrowth (Fig. 7c and d). Saline control mice had significantly greater tumor volumes at the 25 d p.i. time point (p < 0.0001) compared to mice treated with [²²⁵Ac]Ac-DOTA-TATE. Fig. 7e and f show Kaplan-Meier analyses. Animals treated with [²²⁵Ac]Ac-DOTA-TATE had a significantly delayed time to experimental endpoint (p < 0.0001 for H69 and p = 0.0009 for H727) relative to the saline controls. The median time to endpoint was 93.6 ± 10.1 and 62.7 ± 6.7 days for the H69-treated and saline cohorts, respectively, and the median time to experimental endpoint was 62.7 ± 7.9 days and 51.4 ± 4.4 for the H727-treated and saline cohorts, respectively.

Fig. 7.

Efficacy of [²²⁵Ac]Ac-DOTA-TATE in SCID mice–bearing H69 and H727 tumors. The treated animals were intravenously injected with a single bolus of 148.7 kBq and 144.3 kBq mean ²²⁵Ac activity for H69 and H727 cohorts, respectively, or saline solution at 8 days after tumor cells inoculation. Representative images of mice-bearing a H69 and b H727 tumors (outlined), 25 days p.i. Tumor growth volumes of c H69 and d H727 tumors relative to saline controls (arrow indicates the day of injection). Kaplan-Meier plots of e % H69 and f H727 tumor–bearing mice that have reached the experimental endpoint. (g) Representative SSTR2 IHC images of treated and control g H69 tumors and h H727 tumors at experimental endpoint. i Quantified SSTR2 IHC expression j and quantified heterogeneity in treated and control H69 tumors and H727 tumors

IHC staining determined SSTR2 expression in tumors with and without treatment after reaching the experimental endpoint of 2000 mm³ tumor volume (Fig. 7 g–j). As seen in the in vitro ICC results, saline treated H69 tumors had significantly greater SSTR2 expression relative to saline treated H727, p < 0.0001 (Fig. 7i). Heterogeneity of SSTR2 expression was quantified and saline treated H727 tumors had significantly greater heterogeneity scores relative to saline treated H69 tumors, p < 0.0001 (Fig. 7j). Following a single administration of [²²⁵Ac]Ac-DOTA-TATE, SSTR2 expression significantly decreased by 10%, p < 0.01, in H69-treated tumors relative to saline treated tumors, with no significant difference observed in H727 tumors (Fig. 7i). Similarly, H69 tumors demonstrated increased heterogeneity by 22%, p < 0.05, following treatment, with no significant difference observed for H727 tumors (Fig. 7j). Hence, the decrease in expression in H69 tumors likely represents a population of cells with decreased SSTR2 expression rather than a general decrease in every cell within the tumor.

Discussion

Ballal, et al. recently presented the first clinical experience and early results on the efficacy and safety of [²²⁵Ac]Ac-DOTA-TATE in 32 patients with SSTR expressing metastatic GEP-NETs who were [¹⁷⁷Lu]Lu-DOTA-TATE refractory [22]. For these case studies, systemic TAT using [²²⁵Ac]Ac-DOTA-TATE was performed in all the patients with [²²⁵Ac]Ac-DOTA-TATE (100 kBq/kg body weight) at an interval of 8 weeks, to a cumulative dose of 55,500 kBq [22]. This pilot study was reported at an early stage, too soon to calculate overall and event-free survival. However, the short-term clinical results with a median follow-up duration of 8 months suggests that treatment with [²²⁵Ac]Ac-DOTA-TATE can overcome resistance to [¹⁷⁷Lu]Lu-DOTA-TATE and that [²²⁵Ac]Ac-DOTA-TATE can be used as a promising treatment option.

Despite this early clinical experience, preclinical animal studies are needed to better understand the biodistribution and radiation dosimetry of [²²⁵Ac]Ac-DOTA-TATE, and to support the development of novel methods of determining radiation dosimetry in patients. Furthermore, preclinical studies are required to support an investigational new drug (IND) application to the United States Food and Drug Administration (US FDA) to enable clinical trials in the USA and to enable development of advanced radiation dosimetry methods for clinical use. In addition, the method for synthesis and quality assurance of [²²⁵Ac]Ac-DOTA-TATE were not presented in the Ballal study.

Herein, we report the first preclinical study for [²²⁵Ac]Ac-DOTA-TATE. We have synthesized [²²⁵Ac]Ac-DOTA-TATE with high radiochemical yield and purity, and performed stability, BD, RD, toxicity, and efficacy studies in preclinical models of somatostatin receptor 2 (SSTR2) positive lung tumors. DOTA-TATE was chosen because [^{17 7}Lu]Lu-DOTA-TATE has been compared with [¹⁷⁷Lu]Lu-DOTA-TOC in patients and a greater tumor residence time was observed for [¹⁷⁷Lu]Lu-DOTA-TATE, indicating that DOTA-TATE may be a superior targeting ligand for delivery of radiotherapy [23].

Previous somatostatin receptor-targeted PRRT studies have shown that nephrotoxicity is a dose-limiting factor since reabsorption of radio-labeled SSAs by cells in the proximal tubule of the nephron occurs [38, 39]. A preclinical study on PRRT with [²²⁵Ac]Ac-DOTA-TOC conducted in nude mice bearing AR42J rat pancreatic NET xenograft tumors demonstrated significant kidney tubular necrosis at activities higher than 30 kBq [16]. Our BD results demonstrated that [²²⁵Ac]Ac-DOTA-TATE is primarily cleared by the renal route, with some hepatic clearance. Clearance kinetics parameters and RD were calculated, and the highest accumulated doses were observed in the kidneys and liver, with the highest uptake (initial accumulation, A_o) and fastest clearance (T_eff) in the kidney relative to the liver.

In the toxicity study, all animals survived to 5 months p.i., and there was no acute kidney damage observed by pathology at any dose level. BUN and creatinine levels were in the normal range for all animals in the study. However, chronic progressive nephropathy was observed in animals injected with ≥2.56 μCi (94.72 kBq) and, despite overall weight gain, animals that were injected with the highest activities, averaging 4.43 μCi (164 kBq) began losing weight at ~100 days p.i. It is notable that for clinical administrations, SSA PRRTs are co-infused with amino acids for renal protection by blocking peptide reabsorption in the proximal tubules, which significantly diminishes the radiation dose to the kidneys [6, 40]. Liver enzymes were normal and there was no treatment related damage observed in the liver or other tissues by pathology.

We confirmed a high and moderate endogenous expression of SSTR2 in H69 and H727 human lung cancer cells, respectively. Following a single treatment of [²²⁵Ac]Ac-DOTA-TATE, the in vivo efficacy study demonstrated significantly decreased tumor volume, increased tumor growth delay, and a prolonged time to experimental endpoint for animals bearing both tumor types. The responses were greater in H69 tumor–bearing animals compared to H727, which is likely due to the relatively higher and more homogeneous SSTR2 expression in H69 tumors. It is notable that H69 tumors that regrew following treatment had 10% lower SSTR2 expression and 22% increased heterogeneity of SSTR2 expression relative to controls, and these differences were significant, but the H727 tumors regrew with unaltered expression. Tumor regrowth with retained target expression suggests the potential for multiple dosing regimens to prolong survival. However, the loss of some expression in regrowth suggests the potential for development of treatment resistance over time.

In conclusion, our overall results suggest significant potential for the clinical translation of [²²⁵Ac]Ac-DOTA-TATE as a novel therapy for lung neuroendocrine neoplasms. Nephrotoxicity results suggest that renal protective measures will be needed, such as the co-injection of amino acid that is currently done during clinical administration of [¹⁷⁷Lu]Lu-DOTA-TATE [6, 40]. Furthermore, since it is not practical to directly image ²²⁵Ac in patients, personalized methods for determining radiation dosimetry need to be developed using companion imaging agents, e.g., [⁶⁸Ga]Ga-DOTA-TATE positron emission tomography and image voxel-based Monte Carlo simulations.

Supplementary Material

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s00259-021-05315-1.