Using the PET/CT radiotracer [⁶⁸Ga]Ga-DOTA-mDesmo to target V1b receptors and localize corticotropinoma in Cushing’s disease

Abstract

Background

Cushing’s disease, the most common cause of Cushing’s syndrome, is driven by pituitary tumors (corticotropinoma) and characterized by the overexpression of CRH R1 and V1b receptors. Accurate detection of these tumors remains challenging. This study aims to develop and evaluate a corticotropinoma specific radiopharmaceutical, [⁶⁸Ga]Ga-DOTA-mDesmo, that targets the V1b receptor for both anatomical and functional identification of corticotropinomas.

Methods

Molecular docking was used to validate the binding affinity of [⁶⁸Ga]Ga-DOTA-mDesmo to the V1b receptor. Radiolabeling was optimized with gallium-68, followed by quality controls and physicochemical characterization. ACTH-release assay was performed using primary-cultured corticotropinoma cells. Receptor specificity was confirmed via radioimmunoassay using recombinant human V1b receptor. Ex vivo biodistribution studies were performed in healthy male Wistar rats at 30-, 60-, and 120-min post-injection (8 ± 1.1 MBq).

Results

Here we show that [⁶⁸Ga]Ga-DOTA-mDesmo binds effectively to the V1b receptor, with a binding energy of −13.98 kcal/mol and the key interacting residues of V1b are GLN301, SER304, ASN309, and ASP314. Radiolabeling achieves high yield (~96%) and purity (>99%), with human serum stability for up to 4 h. In vitro studies confirm that DOTA-mDesmo acts as an agonist in corticotropinoma cells. Excess cold DOTA-mDesmo results in a 50% blocking in binding. Biodistribution in rats indicates renal clearance, with high %ID/g in the kidneys (7.37 ± 0.58) and urinary bladder (4.32 ± 0.48), and negligible uptake in the pituitary gland, our organ of interest.

Conclusions

These finding support [⁶⁸Ga]Ga-DOTA-mDesmo as a promising radiotracer for non-invasive, receptor-targeted PET/CT imaging of corticotropinomas in patients with Cushing’s disease.

Shukla et al. develop a PET/CT agent, [68Ga]Ga-DOTA-mDesmo, targeting vasopressin V1b receptors overexpressed in ACTH-secreting pituitary tumors causing Cushing’s disease. Their preclinical findings support its potential use for functional and anatomical identification of tumors that are otherwise difficult to detect.

Plain language summary

Cushing’s disease is a condition caused by tumors in the pituitary gland, the area of the body from which some hormones are released. Cushing’s disease leads to overproduction of adrenocorticotropic hormone and cortisol, which results in serious health problems. The tumors are often difficult to find, especially when they are very small. We developed a radioactive molecule called [⁶⁸Ga]Ga-DOTA-mDesmo to help clinicians to localize these tumors. We tested this agent using computer models, lab experiments, and animal studies. We find that this molecule can attach to the tumor cells and may help doctors see them more clearly on scans. This could allow earlier and more accurate diagnosis of Cushing’s disease, which may improve treatment and outcomes for patients in the future.

Introduction

Cushing’s syndrome (CS) is a rare and potentially life-threatening disorder characterized by excessive cortisol secretion, with an incidence rate of ~0.7–2.4 cases per million individuals annually^1–4. It can result from both exogenous (e.g., corticosteroid medication) or endogenous overproduction. Endogenous CS is further classified as either adrenocorticotropic hormone (ACTH)-dependent comprising pituitary adenoma (Cushing’s disease, CD) and ectopic ACTH-secreting tumors (Ectopic Cushing’s Syndrome, ECS), or ACTH-independent, such as an adrenal tumor^3–7.

Among these, CD, caused by ACTH-secreting pituitary adenomas (corticotropinomas), poses major diagnostic challenges. Differentiating CD from ECS is particularly difficult, as most corticotropinomas are microadenomas, often less than 6 mm in size. These small tumors are often undetectable through conventional imaging. Contrast-enhanced magnetic resonance imaging (CE-MRI) has a sensitivity of only 50–60%. Additionally, CE-MRI cannot differentiate a functional corticotropinoma and an incidental nonfunctioning pituitary lesion, which can be present in up to 10% of ECS cases^8,9.

Bilateral inferior petrosal sinus sampling (BIPSS) has sensitivity and specificity of 90–95% for distinguishing CD from ECS. However, it is invasive, costly, requires high technical skills, and has a low positive predictive value for corticotropinoma lateralization (i.e., whether the lesion is located on the right side or left side of the pituitary)^9,10. [⁶⁸Ga]Ga-DOTANOC positron emission tomography-computed tomography (PET/CT) is useful in identifying ectopic tumors expressing somatostatin receptors, but it does not aid in diagnosing CD^10,11.

These challenges contribute to the fact that in ~30% cases of CD remains occult, meaning the source of ACTH excess cannot be confidently localized. The lack of information on accurate localization and lateralization of corticotropinoma leads to incomplete surgical resection and reduced remission rates.

Desmopressin is a synthetic analog of arginine vasopressin (AVP), known to selectively stimulate ACTH release in CD patients via the AVP receptor 1b (V1b), a transmembrane receptor, which is overexpressed in corticotropinomas¹². Several studies have established the role of the V1b receptor in the pathophysiology of CD. Keyzer et al. reported intense V1b receptor expression in corticotropinomas, in contrast to basal expression in normal pituitaries and negligible expression in other pituitary tumors¹³. Sakai et al. further confirmed that desmopressin-induced ACTH secretion was specific to CD and mediated via the V1b receptor, with no response observed in healthy individuals and other conditions¹⁴. Luque et al. further confirmed this by showing that ACTH release in corticotropinoma cultures was blocked by V1b antagonists¹⁵. These findings suggest the key role of the V1b receptor in pathophysiology and desmopressin-induced ACTH response in CD.

We have extended the application of desmopressin and designed a targeted molecule called DOTA-mDesmo to facilitate the chelation of radiometals such as gallium-68 or ⁶⁸Ga (t_1/2 = 68 min and E_β⁺_max = 1.9 MeV) and subsequent PET/CT imaging. We hypothesize that [⁶⁸Ga]Ga-DOTA-mDesmo will selectively bind to the overexpressed V1b receptors in the pituitary. PET/CT imaging will enable the localization of radioactivity that will simultaneously confirm the functionality of the tumor as corticotropinoma and as well as accurately localizes corticotropinomas.

In this study, we report the designing and synthesis of a targeted radiopharmaceutical, [⁶⁸Ga]Ga-DOTA-mDesmo followed by quality control assessments and physicochemical characterization. In-vitro receptor binding affinity, ACTH stimulation assay, and biodistribution in healthy rats were performed. A ready-to-use kit was also developed to enable the convenient production of [⁶⁸Ga]Ga-DOTA-mDesmo at any Nuclear Medicine center equipped with a ⁶⁸Ge/⁶⁸Ga generator.

The [⁶⁸Ga]Ga-DOTA-mDesmo binds effectively to the V1b receptor, with a binding energy of −13.98 kcal/mol and the key interacting residues of V1b are GLN301, SER304, ASN309, and ASP314. Radiolabeling achieves high yield (~96%) and purity (>99%), with human serum stability for up to 4 h. The mDesmo kit demonstrates a radiochemical purity of more than 99% for [⁶⁸Ga]Ga-DOTA-mDesmo. In vitro studies confirm that DOTA-mDesmo acts as an agonist in corticotropinoma cells. Excess cold DOTA-mDesmo results in a 50% blocking in binding. Biodistribution in rats indicates renal clearance, with high %ID/g in the kidneys (7.37 ± 0.58) and urinary bladder (4.32 ± 0.48), and negligible uptake in the pituitary gland, our organ of interest.

Methods and materials

Designing of mDesmo

Desmopressin, a synthetic analog of AVP was modified. Phenylalanine (Phe) was added at position 1 and glycinamide (Gly-NH₂) was replaced with threoninol (Thr-ol) at position 9. The modified peptide, now a 10 amino acid sequence, was chelated with 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA) at the N-terminal to facilitate the radiometal chelation. The modified desmopressin conjugated with chelator was given the name “DOTA-mDesmo” and patented¹⁶.

Structure preparation and molecular docking

The amino acid sequence of the human V1b receptor (UniProtKB: P47901.1) was used to generate its 3D model via AlphaFold2^17,18. Ligand structures, including natural AVP, desmopressin, and [⁶⁸Ga]Ga-DOTA-mDesmo, were modeled and refined using PyMOL and GROMACS for energy minimization^19–21. Molecular docking was performed using AutoDock-GPU to evaluate ligand binding with the V1b receptor^22,23. Key interactions were visualized using PyMol, and hydrogen bonds were analyzed using LigPlot+²⁴.

Synthesis of DOTA-mDesmo

Following a successful in-silico study, the ligand DOTA-mDesmo was custom-synthesized. The compound was characterized using electron spray ionization mass spectrometry (ESI-MS). The purity was assessed using a high-performance liquid chromatography (HPLC) machine equipped with S 3245 UV/VIS and S 3700 Gamma detectors. A detailed methodological description can be found in Supplementary Methods.

Radiolabeling with [⁶⁸Ga]GaCl₃

The ⁶⁸Ge/⁶⁸Ga generator was milked using 4 mL HCl (0.05 M) to obtain radioactive gallium-68 as [⁶⁸Ga]GaCl₃. The [⁶⁸Ga]GaCl₃ (740 MBq) was incubated with DOTA-mDesmo (1.2–12 nmole) at 100 °C for 10 min. The reaction pH varied from 3.5 to 5.5 using sodium acetate. The radiochemical purity of [⁶⁸Ga]Ga-DOTA-mDesmo was assessed using radio HPLC and radio-instant thin-layer chromatography (radio-ITLC, COMECER, TLC 204). For the radio-ITLC analysis: sodium citrate (0.5 M, pH 3.0) and acetonitrile: water (1:1, v/v) were used as the mobile phases; silica gel-coated TLC plates (ITLC-SG) and Whatman paper 3 (W#3) were used as a stationary phase. The final product solution was purified using the C-18 cartridge to remove the unchelated [⁶⁸Ga]GaCl₃ and unwanted reaction excipients. The detailed protocol is summarized in Supplementary Methods.

DOTA-mDesmo kit preparation

A 15-vial batch of lyophilized kits was prepared under aseptic conditions. The kit components were DOTA-mDesmo (5 µg), sodium acetate buffer, and D-mannitol. The contents of each vial were freeze-dried and sealed under the vacuum. The lyophilized kits were stored at −20 °C²⁵.

Preparation of [⁶⁸Ga]Ga-DOTA-mDesmo from kits

The radiolabeling was as follows. The [⁶⁸Ga]GaCl₃ (370–555 MBq or 10–15 mCi) was added to the DOTA-mDesmo lyophilized kit and incubated at 100 °C for 10 min. The radiochemical yield was determined using a radio HPLC and radio-ITLC scanner. The final mixture was purified using a C18 cartridge and passed through a 0.22-micron membrane filter.

Quality control tests

The radionuclide identity and radionuclide purity were assessed using gamma spectrometry (NaI (Tl) based gamma-ray spectrometer, CAPTUS 4000e). The shelf-life of [⁶⁸Ga]Ga-DOTA-mDesmo was assessed for up to 4 h in normal saline at room temperature. The stability in human serum was also assessed for up to 4 h at 37 °C. The sterility test was performed following USP 71. The [⁶⁸Ga]Ga-DOTA-mDesmo culture was incubated in tryptic soy broth (30 g/L) at 25 °C and fluid thioglycolate medium (29.8 g/L) at 30 °C for 14 days and visualized for turbidity daily. The endotoxin levels were determined following USP 85 using the Chromogenic Endotoxin Quant Kit. The maximum permissible endotoxin limit is 175 EU/V (EU—endotoxin unit; V—volume per injection).

Determination of partition coefficient (Log P)

The Log P value of [⁶⁸Ga]Ga-DOTA-mDesmo was determined using the n-octanol/saline model. Equal volumes of the radioactive drug solution and n-octanol solution were mixed vigorously for 10 min, followed by centrifugation at 1509 RCF for 10 min to separate the phases. Radioactivity in each fraction was measured, and Log P values were calculated using the following formula:

$$\mathrm{Log}\mathrm{Po}/\mathrm{w}=\mathrm{Log}\left(\mathrm{Radioactivity}\mathrm{in}\mathrm{Octanol}/\mathrm{Radioactivity}\mathrm{in}\mathrm{Water}\right)$$ 1

Plasma protein binding

To assess plasma protein binding, trichloroacetic acid (TCA) precipitation method was employed. Briefly, 0.1 mL of [⁶⁸Ga]Ga-DOTA-mDesmo was mixed with 0.9 mL of plasma, incubated at 37 °C for an hour, treated with 10% TCA for 2–3 min, and then centrifuged at 671 RCF for 5 min. The radioactivity in the pellet (plasma protein bound [⁶⁸Ga]Ga-DOTA-mDesmo) and supernatant (unbound [⁶⁸Ga]Ga-DOTA-mDesmo) were measured to calculate % PPB.

$$\%\mathrm{PPB}=\mathrm{Radioactivity}\mathrm{in}\mathrm{Pellet}/\mathrm{Total}\mathrm{Radioactivity}$$ 2

In-vitro receptor binding assay

A radioimmunoassay was performed to determine the binding affinity of [⁶⁸Ga]Ga-DOTA-mDesmo for the recombinant human V1b receptor procured from CUSABIO ((Cat No: CSB-CF002469HU(A4)). The [⁶⁸Ga]Ga-DOTA-mDesmo (0.3 µM) was incubated with 15.6−250 nM V1b receptor (MW: 53 kDa) in phosphate-buffered saline (PBS) for 1 h at 37 °C. An excess of cold DOTA-mDesmo (5 and 20 µM) was added to the reaction mixture to block the binding. The reaction mixtures were transferred to a pre-rinsed 30 kDa cutoff Amicon Ultra-15 centrifugal filter device (Merck; 2021-09). Volume makeup was done up to 4 mL using PBS and the filter device was centrifuged at 4800 RCF for 10 min followed by washing of the retentate with an additional 4 mL PBS buffer. The counts of the permeate (unbound [⁶⁸Ga]Ga-DOTA-mDesmo) and retentate (V1b bound [⁶⁸Ga]Ga-DOTA-mDesmo) were recorded using NaI(Tl) based gamma-ray spectrometer (CAPTUS 4000). The percentage binding was calculated by dividing the retentate activity by the total (retentate and permeate activity).

Primary culture and in-vitro ACTH release assay

Surgically excised tissue was obtained from the Department of Neurosurgery, PGIMER, Chandigarh, after taking approval (INT/IEC/2021/SPL-422) from the institutional ethical committee of PGIMER Chandigarh and written informed consent from patients. The tissue was immediately transferred into the PBS buffer supplemented with pen-strep antibiotic solution. Blood and fatty tissue were removed and explants less than 2 mm were prepared. Cells were disaggregated using collagenase treatment (DMEM, 10% FBS, 1% Pen-strep, and 3% collagenase). A 24-well plate was seeded with ~1 × 10⁵ cells and incubated at 37 °C under a 5% CO₂ environment for 24 h in adherent media (DMEM, 10% FBS, and 1% Pen-strep). The media was replaced with growth media (DMEM, 10% FBS, 1% Pen-strep, 20 ng/mL EGF, 20 ng/mL FGF, and B27), and cells were cultured for 48 h. The growth media was discarded from the cells and replaced with only DMEM and pen-strep. The test treatment was given as follows: (1) 5 μM AVP, (2) 5 μM DOTA-mDesmo, (3) 10 μM  DOTA-mDesmo, (4) control cells with only growth media. ACTH release was measured by electro-chemiluminescence-immuno-assay (ELECSYS-2010, Roche Diagnostics, Germany).

Biodistribution in healthy Wistar rats

Animal biodistribution studies were performed after obtaining ethical clearance from the Institutional Animal Ethics Committee (722/IAEC/111/109/105). The uptake and distribution of [⁶⁸Ga]Ga-DOTA-mDesmo depends upon the expression of the V1b receptor, which is not specific to gender and sex. Therefore, the study was performed in only one sex i.e., male. The selection of male rats was due to the ease of handling and availability. A total of nine healthy male Wistar rats (16 weeks old), weighing 274 ± 13 g were issued from the animal house PGIMER, Chandigarh. A mean radioactivity of 8 ± 1.1 MBq of [⁶⁸Ga]Ga-DOTA-mDesmo was injected intravenously via the penile vein. All animals were primarily anesthetized using a mixture of ketamine (40 mg/kg) and xylazine (4 mg/kg), followed by sacrifice through cervical dislocation. Animals (n = 3 at each time point) were sacrificed at 30, 60, and 120 min post-injection of [⁶⁸Ga]Ga-DOTA-mDesmo. Various organs were collected and weighed in weight-corrected tubes. Counts in each organ were recorded using NaI(Tl) based gamma-ray spectrometer (CAPTUS 4000). The percent injected dose per gram of tissue (%ID)/g was calculated using the formula:

$$\%\mathrm{ID}/\mathrm{g}=\frac{\mathrm{Counts}\mathrm{in}\mathrm{tissue}}{\mathrm{Injected}\mathrm{Activity}\mathrm{Counts}\times \mathrm{Weight}\mathrm{of}\mathrm{tissue}}\times 100$$ 3

Where,

$$\mathrm{Injected}\mathrm{Activity}\mathrm{Counts}=\mathrm{Total}\mathrm{injected}\mathrm{counts}-\left(\right)\mathrm{Injection}\mathrm{site}\mathrm{residual}\mathrm{counts}+\mathrm{Syringe}\mathrm{residual}\mathrm{counts}\left)\right)$$

For the estimation of total blood pool activity, the blood volume was assumed to be 7% of the total body weight of each rat.

Statistics and reproducibility

Statistical analyses were primarily descriptive and expressed as means. Radiolabeling reactions, radioligand binding assays, and ACTH-release assays were conducted in triplicates, with data presented as mean ± standard deviation (SD). Replicates refer to independent experimental repeats performed under identical conditions. For the animal studies, due to limited sample size and the exploratory nature of the investigation, statistical evaluation was restricted to the calculation of mean ± SD.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Results

In-silico study

Docking studies showed favorable binding energies for desmopressin (−15.49 kcal/mol) and [⁶⁸Ga]Ga-DOTA-mDesmo (−13.98 kcal/mol). Desmopressin formed nine hydrogen bonds and [⁶⁸Ga]Ga-DOTA-mDesmo formed 14 hydrogen bonds with the V1b receptor. Key interacting residues of V1b receptor were GLN301, SER304, ASN309, and ASP314. The docked pose and interactions are provided in Fig. 1.

Fig. 1. Molecular docking and visualization of ligand interactions with the V1b receptor.

Docked complexes of V1b receptor with ligands Desmopressin and [⁶⁸Ga]Ga-DOTA-mDesmo visualized using PyMOL. a, c Surface representation of the V1b receptor displaying the binding pockets with Desmopressin and [⁶⁸Ga]Ga-DOTA-mDesmo shown in ball-and-stick models. b, d Cartoon representation of the V1b receptor highlighting key interacting amino acid residues with the ligands (ball-and-stick models).

Synthesis and radiolabeling

The mass of DOTA-mDesmo was determined as 1648.9 Da (Supplementary Fig. S1). HPLC analysis showed a purity of more than 99%, with a retention time (tR) of 7.87 min (Supplementary Fig. S2). The radiochemical yield of 96 ± 2.5% was achieved when 740 MBq of [⁶⁸Ga]GaCl₃ was incubated with 9 nmole of DOTA-mDesmo at 100 °C for 10 min, with a reaction pH of 4.5. The radio-HPLC analysis showed two peaks for [⁶⁸Ga]Ga-DOTA-mDesmo at tR 2.74 and 8.33 (Fig. 2a). The minor peak at tR 2.74 corresponded to Ga-68. Similarly, the radio-ITLC chromatogram developed using sodium citrate with ITLC-SG and acetonitrile: water with W#3 showed two peaks for [⁶⁸Ga]Ga-DOTA-mDesmo (Supplementary Fig. S3). The minor peak in both chromatograms corresponded to Ga-68 (Supplementary Fig. S3). The radiochemical purity was further enhanced to more than 99% using C-18 cartridge purification validated by both radio-HPLC and radio-ITLC (Fig. 2a–c).

Fig. 2. Radiochemical evaluation and optimization of [⁶⁸Ga]Ga-DOTA-mDesmo synthesis.

a Radiochemical yield assessment of [⁶⁸Ga]Ga-DOTA-mDesmo using radio-HPLC and radio-ITLC. Radio-HPLC chromatogram showing retention times of Ga-68, [⁶⁸Ga]Ga-DOTA-mDesmo (pre and post-purification), and kit produced [⁶⁸Ga]Ga-DOTA-mDesmo. b, c Radio-ITLC chromatogram of purified [⁶⁸Ga]Ga-DOTA-mDesmo, developed using ITLC-SG plate on sodium citrate buffer and Whatman paper 3 on acetonitrile: water (1:1). d Effect of reaction pH on radiochemical yield. e Effect of DOTA-mDesmo concentration on radiochemical yield. The experiment was performed in triplicates and the data is presented as mean ± SD.

The reaction pH played a crucial role in radiochemical yield. The highest yield (98 ± 1.7) was achieved at a reaction pH of 4.5, while a slight decrease towards 4.0 led to a 16% reduction (82 ± 2.4) in radiochemical yield. The radiochemical yield was reduced to 16% and 5 % at pH 3.5 and 5.5, respectively (Fig. 2d). The effect of DOTA-mDesmo concentration on RCY was also evaluated. A 50% yield was attained when DOTA-mDesmo and [⁶⁸Ga]GaCl₃ were incubated at a ratio of 0.25:37 (µg: MBq). The maximum yield was observed at the ratios of 0.75:37 and 1:37 (Fig. 2e). The specific activity for [⁶⁸Ga]Ga-DOTA-mDesmo was achieved up to 51 ± 2.1 MBq/µg.

Moreover, the mDesmo kit demonstrated a radiochemical purity of more than 99% for [⁶⁸Ga]Ga-DOTA-mDesmo, ensuring the consistent and efficient radiolabeling of DOTA-mDesmo (Fig.2a).

Quality control

Gamma-ray spectra of [⁶⁸Ga]Ga-DOTA-mDesmo showed energy peaks at 511 KeV and 1022 KeV, indicating >99% radionuclide purity (Supplementary Fig. S4). No photopeak was observed after 10 half-lives of Ga-68, ruling out the presence of Ge-68 breakthrough (Supplementary Fig. S4).

A single radio peak at R_f 0.0 was observed at each time point for up to 4 h, indicating the optimum shelf-life of [⁶⁸Ga]Ga-DOTA-mDesmo at room temperature for imaging (Supplementary Table S1). Similarly, in human serum and PBS at 37 °C, a single radio peak at R_f 0.0 was obtained for up to 4 h, demonstrating the stability of [⁶⁸Ga]Ga-DOTA-mDesmo under physiological conditions (Supplementary Table S1). The in-house synthesized [⁶⁸Ga]Ga-DOTA-mDesmo was found to be sterile, as no visible growth or turbidity was observed for up to 14 days. Furthermore, the calculated endotoxin value for the in-house synthesized [⁶⁸Ga]Ga-DOTA-mDesmo was measured to be 8.6 ± 1.2 EU/V, which was well below the maximum permissible limit of <175 EU/V (Supplementary Fig. S5 and Table S2).

Plasma protein binding and partition coefficient

The TCA precipitation method yielded a plasma protein binding value of 65.64 ± 2.2 for [⁶⁸Ga]Ga-DOTA-mDesmo. The Log P value of [⁶⁸Ga]Ga-DOTA-mDesmo was found to be −2.7 ± 0.06.

In-vitro receptor binding study

As the concentration of V1b was increased from 0.02 µM to 0.25 µM, the binding was also increased from 5 to 44%. A maximum of 44 ± 2.2 % binding was achieved for 0.3 µM [⁶⁸Ga]Ga-DOTA-mDesmo and 0.25 µM V1b receptor (Table 1). Moreover, the presence of cold mDesmo from 5 µM to 20 µM resulted in a reduction of the binding from 44 to 22%. A 50% blocking effect was observed in the presence of 20 µM of cold DOTA-mDesmo (Table 1).

Table 1.

Binding of [⁶⁸Ga]Ga-DOTA-mDesmo to purified human V1b receptor

	Binding (%) (Mean ± SD)
V1b (15.6 nM) + [68Ga]Ga-DOTA-mDesmo	5.1 ± 1.3
V1b (31. 25 nM) + [68Ga]Ga-DOTA-mDesmo	21.8 ± 2.1
V1b (62.5 nM) + [68Ga]Ga-DOTA-mDesmo	25.7 ± 2.4
V1b (125 nM) + [68Ga]Ga-DOTA-mDesmo	30.7 ± 1.9
V1b (250 nM) + [68Ga]Ga-DOTA-mDesmo	44 ± 2.2
V1b (250 nM) + [68Ga]Ga-DOTA-mDesmo + DOTA-mDesmo (5 µM)	38.8 ± 1.7
V1b (250 nM) + [68Ga]Ga-DOTA-mDesmo + DOTA-mDesmo (20 µM)	22.1 ± 2.1

Primary culture and in vitro ACTH release assay

The primary cultured corticotropinoma cells from surgically excised tissue are shown in Fig. 3a. Corticotropinoma cells showed autonomous ACTH production (Fig. 3b). Cells treated with AVP and DOTA-mDesmo demonstrated more than 200% stimulation of ACTH release compared to autonomous levels (Fig. 3c).

Fig. 3. Primary cultured corticotropinoma cells and ACTH stimulation by DOTA-mDesmo.

a Primary cultured corticotropinoma cells from tumor tissue observed under a light microscope at 10× magnification, and at 20× magnification (inset), with 1 × 10⁵ cells per well on day 1. b Autonomous ACTH production in tumor cultures over 3 days. c Variation in ACTH stimulation in response to incubation with 5 µM AVP and 5–10 µM DOTA-mDesmo for 1 h. Control ACTH secretion (573 ± 98 pg/mL) at 1 h was set to 100%. The experiment was performed in triplicates and the data is presented as mean ± SD.

Ex vivo biodistribution in healthy Wistar rats

High radioactivity levels in the kidneys (%ID/g: 7.37 ± 0.58) and urinary bladder (%ID/g: 4.32 ± 0.48) indicated that [⁶⁸Ga]Ga-DOTA-mDesmo was primarily excreted via the renal route (Fig. 4a). The pituitary gland, our organ of interest, showed 0.5 ± 0.03 %ID/g (Fig. 4a). The relatively high %ID/g value was due to the denominator effect resulting from the small mass of the pituitary gland (~0.008 g). However, the absolute activity in the pituitary was negligible (%ID: 0.005 ± 0.001) at 30 min (Fig. 4b) and decreased with time. At 30 min, the blood showed a %ID/g of 1.61 ± 0.22, with the total blood pool accounting for ~31% of the injected activity. The blood pool activity reduced sharply to ~3% at 120 min (Fig. 4e), due to renal excretion of the tracer and as well as physical decay of ⁶⁸Ga (t_1/2 = 68 min). Mild physiological uptake was also noted in various organs, including muscles, lungs, adrenals, liver, pancreas, spleen, and heart. However, clearance of [⁶⁸Ga]Ga-DOTA-mDesmo from these organs occurred within 120 min following administration (Fig. 4a).

Fig. 4. Biodistribution of [⁶⁸Ga]Ga-DOTA-mDesmo and anatomical localization of pituitary in rats.

a Histogram showing percent injected dose per gram of tissue (%ID/g) of [⁶⁸Ga]Ga-DOTA-mDesmo in various organs of Wistar rats at different time points. b Total blood pool (%ID) activity. c Total pituitary (%ID) activity. d Anatomical location of pituitary. e Hematoxylin and Eosin (H&E) staining of extracted pituitary tissue Three animals were assessed (n = 3) at each time point and the %ID/g is presented as mean ± SD.

Discussion

In this study, we present a targeted radiopharmaceutical, [⁶⁸Ga]Ga-DOTA-mDesmo, that holds promise for both delineating and confirming the functional status of corticotropinomas.

Direct radiolabeling of desmopressin with ⁶⁸Ga was not feasible due to structural limitations, hence, few modifications were made to make the peptide radiolabeling-compatible while preserving its receptor binding affinity. The rationale for the modifications were as follows: (1) To facilitate DOTA conjugation at N-terminus: native desmopressin contains 3-mercaptopropionic acid at position 1, which lacks a free amine group for DOTA conjugation. Therefore, we substituted position 1 with phenylalanine (Phe), which reintroduces a primary amine at the N-terminus, thereby enabling conjugation with the DOTA chelator. (2) To preserve the disulfide bridge: the disulfide bond between Cys1 and Cys6 in vasopressin/desmopressin is essential for receptor-binding and peptide stability. By conjugating bulky DOTA in cysteine itself might have caused steric hinderances and thereby might have hampered disulfide bridge. By introducing Phe at position 1 (instead of modifying the cysteine as in desmopressin), we preserved the native disulfide bridge to maintain receptor specificity and structural integrity. (3) To prevent dual conjugation of DOTA: desmopressin has glycinamide (Gly-NH₂) at position 9, which contains a free amine group that could lead to unwanted conjugation of DOTA chelator at both N- and C-termini. To avoid this, we replaced Gly-NH₂ with threoninol (Thr-ol), thereby removing the extra amine and ensuring site-specific conjugation at the N-terminus only. This also prevented potential steric hindrance near Arg8, which is critical for receptor binding. (4) To maintain metabolic stability: native AVP has a plasma half-life of ~3 min due to enzymatic degradation. Zaoral et al. demonstrated that modifying the N- and C-termini improved peptide activity and synthesized desmopressin, which has a plasma half-life of ~55 min²⁶. Although desmopressin already contains a modified C-terminus (Gly-NH²), but this residue was replaced with Thr-ol, not only to prevent double conjugation, but also to protect the peptide from enzymatic degradation. These modifications resulted in a radiolabeling-compatible analog, DOTA-mDesmo, which maintains the receptor specificity for the V1b receptor, validated by molecular docking studies.

Molecular docking was conducted to evaluate the binding affinity of [⁶⁸Ga]Ga-DOTA-mDesmo toward the V1b receptor. Desmopressin, the structural scaffold of DOTA-mDesmo, was included as reference ligand. The binding energy of [⁶⁸Ga]Ga-DOTA-mDesmo was found to be −13.98 kcal/mol, forming 14 hydrogen bonds with key receptor residues including GLN301, SER304, ASN309, and ASP314, indicating strong and stable interaction. The binding energies of all the ligands were optimum for stable interactions. Therefore, DOTA-mDesmo could be a suitable carrier of Ga-68 in PET-imaging of corticotropinoma with overexpressed V1b receptors.

Following the encouraging outcomes obtained from in-silico investigations, the DOTA-mDesmo peptide was custom-synthesized with a 99% purity as confirmed by HPLC. Chelation of [⁶⁸Ga]GaCl₃ to DOTA-mDesmo resulted in >98% radiochemical yield and specific activity of 51 ± 2.1 MBq/µg. The reaction pH drastically affected the radiochemical yield. The highest yield (98 ± 1.7%) was achieved at a reaction pH of 4.5 (Fig. 2). The effect of DOTA-mDesmo concentration on radiochemical yield indicates the judicious use of the precursor, especially with a limited number of patients. The quality control tests adhered to the established guidelines of the US Pharmacopeia for pharmaceuticals. The radiochemical purity of [⁶⁸Ga]Ga-DOTA-mDesmo was more than 99%. The prepared formulation demonstrated sterile preparation, good serum stability, and endotoxin levels within permissible limits. Overall, the results from the quality control tests uphold the suitability of [⁶⁸Ga]Ga-DOTA-mDesmo for intravenous administration.

[⁶⁸Ga]Ga-DOTA-mDesmo bound to V1b receptor at varying degrees. The highest binding (44 ± 2.2%) was achieved when 0.3 µM [⁶⁸Ga]Ga-DOTA-mDesmo incubated with 0.25 µM V1b receptor. In addition, binding was blocked by ~50% with an excess cold DOTA-mDesmo, indicating specificity of [⁶⁸Ga]Ga-DOTA-mDesmo towards the V1b receptor. Also, the ability of DOTA-mDesmo to stimulate corticotropinoma cells to release excess ACTH confirms that, after modification, DOTA-mDesmo retained the agnostic effect for the V1b receptor. A concern might arise that DOTA-mDesmo stimulates ACTH release in corticotropinoma cells, which could transiently elevate cortisol levels of the patients during imaging. However, it is to be noted that ACTH and cortisol stimulation tests using Desmopressin and CRH are standard clinical practices widely employed in the differential diagnosis of Cushing’s syndrome. Additionally, desmopressin stimulation test requires intravenous administration of 10 µg of desmopressin¹². In contrast, radiolabeled formulation of DOTA-mDesmo would require a maximum of 2.5 µg for patient imaging. Moreover, ACTH and cortisol stimulation tests do not provide anatomical localization of corticotropinoma, whereas [⁶⁸Ga]Ga-DOTA-mDesmo PET/CT image may offer a huge advantage in functional as well as anatomical delineation of corticotropinoma. Therefore, stimulation of V1b receptor and release of ACTH by DOTA-mDesmo would not be a concern.

Understanding the plasma protein binding characteristics of a drug is crucial for evaluating its pharmacological behavior. Generally, a plasma binding value above 80% is considered high²⁷. It is important to note that high plasma binding can result in increased blood pool activity, which may impact the distribution and clearance of the drug within the body. The TCA precipitation method revealed a moderate level of plasma binding for [⁶⁸Ga]Ga-DOTA-mDesmo (~65%).

The preclinical biodistribution studies in Wistar rats revealed that [⁶⁸Ga]Ga-DOTA-mDesmo was excreted via the renal route with high tracer uptake observed in the kidneys even at 120 min of post-injection (Fig. 4). The increased counts in the kidney may be partially attributed to the low water intake by rats. The blood pool activity was observed for up to 60 min post-injection, but it was subsequently decreased at 120 min. The prolonged blood pool activity could be attributed to the moderate plasma protein binding of [⁶⁸Ga]Ga-DOTA-mDesmo. Notably, our investigation focused on the pituitary gland, which served as our primary organ of interest. The study showed that [⁶⁸Ga]Ga-DOTA-mDesmo did not accumulate in the normal functioning pituitary gland. It is important to mention that the pituitary is a tiny but important organ. The observed uptake value in the pituitary is due to the denominator effect resulting from the small mass (~0.008 g) of the pituitary gland (Fig. 4). This lack of activity in the pituitary can be attributed to the minimal V1b receptor expression in the normal pituitary gland. The main goal of this experiment was to qualitatively assess the physiological distribution of the tracer. Given the consistent biodistribution pattern at each time point, we made an ethically appropriate decision to limit the number of animals (n = 3). However, we acknowledge that a small sample size may affect the statistical power and generalizability of the quantitative conclusions.

Accurate localization and lateralization of microadenoma within the 10 mm pituitary is crucial for the effective surgical therapy of Cushing’s disease. Trans-sphenoid surgery is the procedure to remove pituitary tumors. The low sensitivity (<60%) of CE-MRI limits the ability to detect small corticotropinomas. A sizable portion of pituitary microcorticotropinomas are often not visualized on CE-MRI⁹. Furthermore, it does not give any information regarding the functionality of the tumor. BIPSS is an invasive technique and is limited by the requirement for high technical skills and cost. The primary purpose of this test is to pinpoint the pituitary or extra-pituitary region as the source of excess ACTH. The lateralization ability of BIPSS for corticotropinoma is ~70% and thereby surgeries done based on BIPSS lead to a low remission rate^8,9. Additionally, the prevalence of pituitary incidentalomas in the cases of ECS may further complicate the matter. While anatomical imaging can confirm the presence of a pituitary lesion in these cases, it cannot determine its functional status, and the existence of an ectopic lesion may go unnoticed^5,6.

This study provides a targeted PET radiotracer, that can overcome these challenges in the localization of ACTH-secreting corticotropinomas. Although various nuclear medicine modalities such as [⁶⁸Ga]Ga-labeled somatostatin analogs, [¹⁸F]FDG, [⁶⁸Ga]Ga-Pentixafor, [¹¹C]C-methionine, and [⁶⁸Ga]Ga-CRH have been explored for the localization of corticotropinomas²⁸. However, each has specific limitations in the context of corticotropinoma localization.

Normal pituitary demonstrates high somatostatin receptor density that leads to very high physiological uptake of [⁶⁸Ga]Ga-labeled somatostatin analogs. It would be difficult to visualize small pituitary adenomas with [⁶⁸Ga]Ga-labeled somatostatin analogs. However, the ectopic ACTH-secreting tumors are easily visualized with [⁶⁸Ga]Ga-labeled somatostatin analogs²⁸. Although [¹¹C]C-methionine PET is a sensitive method in detecting pituitary adenomas, however, it is not useful in the assessment of the functionality of the tumor²⁹. Similarly, [⁶⁸Ga]Ga-Pentixafor, which targets CXCR4 receptors, has shown promise with a reported diagnostic accuracy of 88.6% for PET/MRI³⁰. However, CXCR4 receptor overexpression is not exclusive for corticotropinomas, other pituitary tumors, such as growth hormone-secreting and nonfunctioning pituitary adenomas also express CXCR4 receptors³¹. Hence, the lesions detected on [⁶⁸Ga]Ga-Pentixafor are questionable in their functionality. [¹⁸F]FDG PET/CT reported sensitivity for corticotropinomas is only 40%³². Recently, the utility of [⁶⁸Ga]Ga-CRH PET/CT in the diagnosis of CD has been reported^33,34. [⁶⁸Ga]Ga-CRH delineates corticotropinoma via CRH receptors (CRH R1 and R2) and it also showed diffused uptake in the whole pituitary due to the physiological expression of CRH R1/ R2 receptors.

In contrast, several studies have reported the overexpression of the V1b receptor in corticotropinomas^12,15. The current study demonstrated that [⁶⁸Ga]Ga-DOTA-mDesmo selectively binds to V1b receptors. This study presents substantial evidence for the receptor-specific targeting of corticotropinomas using [⁶⁸Ga]Ga-DOTA-mDesmo. [⁶⁸Ga]Ga-DOTA-mDesmo PET/CT has the potential to emerge as a key diagnostic modality for corticotropinomas.

Encouraged by the promising preclinical results, a clinical trial was conducted and registered with the Clinical Trial Registry of India (CTRI/2022/08/044615) to evaluate the diagnostic utility of [⁶⁸Ga]Ga-DOTA-mDesmo PET/CT in the localization of corticotropinoma in Cushing’s disease. The cohort comprised 30 patients (10 males and 20 females), with 24 patients of CD (17 microadenomas and 7 macroadenoma), 4 patients of ECS, and 2 patients of ACTH-independent CS. [⁶⁸Ga]Ga-DOTA-mDesmo was able to localize the pituitary tumor in all 24 patients of CD with diagnostic accuracy of 100% (CI:86–100%), in contrast to CE-MRI, which was able to localize lesions in only 16 patients with a diagnostic accuracy of 67% (CI:48–86%). Additionally, there was no uptake of [⁶⁸Ga]Ga-DOTA-mDesmo in the pituitary of patients with ECS. The outcome of the clinical trial demonstrated the potential of [⁶⁸Ga]Ga-mDesmo PET/CT as a non-invasive molecular imaging modality to help neurosurgeons in intraoperative neuro-navigation during trans-sphenoidal surgery is published separately^35,36.

Conclusion

The present study aimed to tackle the diagnostic challenge in ACTH-dependent Cushing’s syndrome (Cushing’s disease). The study offers a targeted PET tracer, [⁶⁸Ga]Ga-DOTA-mDesmo, for the precise localization and lateralization of corticotropinomas. [⁶⁸Ga]Ga-DOTA-mDesmo PET/CT may provide the operating surgeon with the required information for tumor delineation while safeguarding the normal functioning pituitary. The successful synthesis and preclinical validation establish the feasibility of this tracer. However, clinical studies are warranted to thoroughly evaluate its diagnostic accuracy and clinical utility in the management of Cushing’s disease.

Limitations of the study

This study is limited by the absence of small animal PET imaging, which restricts real-time visualization of tracer kinetics and dynamic biodistribution. Additionally, the unavailability of corticotropinoma-bearing animal models precluded the direct assessment of tumor-targeting capability and in vivo specificity of [⁶⁸Ga]Ga-DOTA-mDesmo. This represents a key limitation, as the physiological behavior of the tracer in healthy animals may not fully reflect its performance in disease conditions. Without tumor-bearing models, it is difficult to assess tracer uptake in actual V1b receptor–overexpressing tissues or tumors, which limits the extrapolation of the results to human pathological states. However, this limitation is partly mitigated by subsequent clinical evaluation of the tracer, the results of which have been published separately in EJNMMI, demonstrating the potential of [⁶⁸Ga]Ga-DOTA-mDesmo in Cushing’s disease patients³⁶.

Author contributions

Somit Pandey.: In silico studies, radiolabeling optimization, quality control assessment, physicochemical characterization, cell culture, animal studies, data analysis, and manuscript writing; Rama Walia: Study conception, design, data interpretation, manuscript review, and proofreading; Gurvinder Kaur.: radiopharmaceutical preparation, cell culture and animal studies; Kumud Pandav: In silico studies; Imran Rather: Animal Studies; Nivedita Rana: Data interpretation and manuscript review; Bhagwant Rai Mittal: Facility, resources, and data interpretation; Jaya Shukla: study conceptualization, design, funding, resources, data interpretation, manuscript review, and proofreading. The first draft of the manuscript was written by Somit Pandey and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Peer review

Peer review information

Communications Medicine thanks Alejandro Ibáñez-Costa, Jie Ding, and the other anonymous reviewer(s) for their contribution to the peer review of this work. Peer review reports are available.

Data availability

The source data for Figs. 2d, e, 3b, c, 4a, and Table 1 can be found in the Supplementary Data File.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s43856-025-01033-z.