Automated Synthesis of (R)-[¹⁸F]MH.MZ on the iPhase Flexlab Reaction Platform

Abstract

(R)-[¹⁸F]MH.MZ ([¹⁸F]MH.MZ) is a promising positron emission tomography (PET) radiotracer for in vivo study of the 5-HT_2A receptor. To facilitate clinical trials a fully automated radiosynthesis procedure for [¹⁸F]MH.MZ was developed using commercially available materials on the iPhase Flexlab module. The overall synthesis time was 100 minutes with a radiochemical yield of 7 ± 0.9% (n=3). The radiochemical purity was greater than 99% for [¹⁸F]MH.MZ with a molar activity of 361 ± 57 GBq/μmol (n=3). The protocol described herein reliably provides [¹⁸F]MH.MZ that meets all relevant release criteria for a GMP radiopharmaceutical.

1. Introduction

There have been several 5HT_2A receptor PET tracers developed over the years, with [¹⁸F]altanserin and [¹¹C]MDL-100907 being two of the most widely studied.^1–3 However, both tracers have their limitations. [¹⁸F]altanserin is metabolized into lipophilic metabolites that may cross the blood-brain barrier and bind uniformly and nonspecifically across all regions of the brain.^4–6 This complicates analysis of the PET images and requires a bolus/infusion approach that is both impractical and inconvenient. [¹¹C]MDL-100907 is one of the most 5HT_2A receptor-selective radioligands, but suffers from relatively low counts due to a combination of long scan duration and short half-life of carbon-11. [¹⁸F]MDL-100907 has also been prepared by several research groups, however its use as a clinical research tool has been limited by several factors including its radiosynthesis yield and chiral purity; as well as the commercial availability of the required precursors.^7–9 (R)-[¹⁸F]MH.MZ ([¹⁸F]MH.MZ) is a radiotracer that has been developed based upon the MDL-1000907 scaffold, combining the selectivity of [¹¹C]MDL-100907 with the favorable isotopic properties of fluorine-18.¹⁰ Additionally, complete displacement of MH.MZ was found with both 5HT_2A antagonist kentanserin and MDL-100907, indicating the high specificity of the radiotracer to 5HT_2A receptors.¹¹ One of the first human studies characterizing 5HT_2A receptor selectivity of [¹⁸F]MH.MZ in the human brain has shown that the affinity and selectivity of the radiotracer is comparable with that of [¹¹C]MDL-100907.¹² Uptake of the radiotracer reflects the reported distribution of 5HT_2A receptors in the human brain, with highest uptake in cortical regions and low uptake in the striatum, thalamus, and cerebellum. [¹⁸F]MH.MZ also displays no detectable metabolism in human plasma over a 180-min timeframe, which eliminates intensive metabolite analyses required for [¹⁸F]altanserin as well as [¹⁸F]setoperone.^{12, 13} Furthermore, the radiosynthesis of both [¹⁸F]altanserin and [¹⁸F]setoperone require somewhat complicated purification conditions.^{14, 15}

While the preparation of [¹⁸F]MH.MZ has been previously reported^{10, 12, 16}, the published methods required modification for their implementation at our core facility. In 2020 Kramer et. al. reported the successful clinical translation of MH.MZ using a series of automated modules to prepare [¹⁸F]MH.MZ in good yield.¹² Unfortunately, this method required the use of two synthesis and two HPLC purification modules. In an effort to reduce the number of required reaction modules, and apply the synthesis to our existing equipment, our efforts focused on adapting the radiosynthesis to a single radiosynthesis platform. Herein, we report the development and utilization of methodology for the automated production of [¹⁸F]MH.MZ that meets cGMP criteria for human use.

2. Procedure

2.1. General Methods.

All chemicals were obtained from commercial sources and were of analytical, ACS, or USP quality (Sigma-Aldrich, USA; Corning Life Sciences, USA; Pharmco-Aaper, USA) and were used without further purification. MDL-105725 and ethylene di-(p-toluenesulfonate) were purchased from Advanced Biochemical Compounds (ABX, Germany) in pre-aliquoted fractions. The Kryptofix/carbonate solution was prepared by mixing Kryptofix 222 (K₂₂₂) in acetonitrile (MeCN) (K₂₂₂, 250 mg, dissolved in 6 mL MeCN) with potassium carbonate (K₂CO₃) (K₂CO₃, 140 mg, dissolved in 6 mL 18 MΩ deionized water). The mixture was gently mixed to provide a clear and colorless solution, and stored at 0–10 °C. Solid phase extraction (SPE) cartridges (QMA Light, Sep-Pak® C18 Plus, C18 Light) were obtained from Waters (Waters®, MA, USA). The QMA light was purchased as the carbonate form, conditioned with 1 mL 18 MΩ deionized water and dried with filtered air. The C18 sep-paks were conditioned with 6 mL 200 proof ethanol (EtOH) followed by 20 mL of sterile H₂O prior to use. All radioactivity was measured using a CAPINTEC CRC-15 Dual PET, CRC-15 PET, or CRC-25 Dual PET dose calibrator. The radiochemical purity of the final product [¹⁸F]MH.MZ was determined by analytical radio-HPLC with a Phenomenex® Luna C18(2) column (250 × 4.6 mm,), mobile phase: 40% MeCN in 0.1M Ammonium Formate (pH=4.5), flow rate 1.0 mL/min, λ = 254 nm, RT of [¹⁸F]MH.MZ = 6.8 min.

Fluorine-18 was obtained in the chemical form of [¹⁸F]fluoride ion ([¹⁸F]F⁻) via the¹⁸O(p,n)¹⁸F nuclear reaction, by irradiating a cyclotron-target containing [¹⁸O]water with a proton beam of 40–65 μA for 30–60 min (PETtrace cyclotron, GE, Uppsala, Sweden).

2.2. Automated Radiosynthesis of [¹⁸F]MH.MZ.

The synthesis was carried out on a Flexlab automated reaction module (iPhase, Melbourne, Australia) (Figure 1) loaded as follows: vial 1: Kryptofix/carbonate Solution (0.6 mL); vial 3: Ethylene di-(p-toluenesulfonate) (12 mg dissolved in 1.2 mL of MeCN); vial 6: 3.8 mL 50% MeCN / 0.1M ammonium formate; vial 9: 0.6 mL DMSO; vial 11: 10 mL sterile water; vial 12: 1 mL ethanol; vial 13: 10 mL 0.9% saline; vial 15: 3.8 mL 40% MeCN / 0.1M ammonium formate (pH=4.5); vial 19: 3 mg MDL-105725 in 0.6 mL DMSO with 7.5 μL 1M tetrabutylammonium Hydroxide in methanol; Round Bottom Flask 1: 40 mL sterile water; Round Bottom Flask 2: 40 mL sterile water; QMA: QMA light sep-pak SPE B: C18 light sep-pak; SPE D: C18 plus sep-pak; HPLC A: 50% MeCN / 0.1M ammonium Formate; HPLC B: 40% MeCN / 0.1M ammonium Formate (pH=4.5). The Flexlab was used without any modification with the flowpath shown in Figure 1.

Figure 1:

Diagram of the iPhase Flexlab

After the end-of-bombardment (EOB) the [¹⁸F]F⁻ produced on the cyclotron was transferred to the target vial on the Flexlab module. The fluoride was then transferred through the QMA sep-pak to recover the [¹⁸O]H₂O and eluted into Reactor 1 with the Kryptofix/carbonate solution loaded in vial 1. The resulting solution was concentrated in vacuo and redissolved with the ethylene di-(p-toluenesulfonate) solution from vial 3. The mixture was reacted at 90 °C for 3 minutes, cooled via air cooling of the reactor, and diluted with the HPLC mobile phase from vial 6. The resulting mixture was transferred to the HPLC injection tube and loaded onto the first HPLC purification column (Phenomenex, Luna C18(2) 250 × 10 mm). This HPLC purification column was eluted with the HPLC A mobile phase and the desired radioactive peak was collected into round bottom flask 1, which was then transferred through the C18 light cartridge at SPE B to isolate the labeled intermediate.

This reaction intermediate was eluted from the sep-pak with 0.6 mL DMSO (vial 9) into reactor 2, to which the MDL-105725 solution (vial 19) had been previously added. The resulting mixture was heated to 110 °C for 10 minutes, at which time it was cooled to room temperature. This solution was diluted with the second HPLC mobile phase (vial 15), transferred to the second HPLC injection tube, and loaded onto the second HPLC purification column (Phenomenex, Luna C18(2) 250 × 10 mm). This HPLC purification column was eluted with the HPLC B mobile phase and the desired radioactive peak was collected into round bottom flask 2, which was then transferred through the C18 plus sep-pak at SPE D to isolate the product. The sep-pak was washed with sterile water (vial 11) and then the [¹⁸F]MH.MZ was eluted from the sep-pak with ethanol (vial 12) and diluted with saline (vial 13) into the product vial. The final drug solution was then passed through a 0.22 μm sterilizing filter into the final vial (Pall, USA, Part#: AEF1NTE), to provide the [¹⁸F]MH.MZ as a ready to inject solution. The radiosynthesis required approximately 100 minutes from the end of bombardment (EOB) to the drug product solution being deposited in the final vial.

2.3. Quality Control Procedures.

Quality control results from each of the three qualification batches are shown in Table 1. All quality control measurements were performed on GMP-qualified instruments unless otherwise stated. After completion of the synthesis an aliquot of the product was withdrawn for quality control, determining the appearance by visual inspection and radionuclidic identity by half-life measurement using a dose calibrator (Capintec CRC-15 Dual PET). Radiochemical and chemical purities were analyzed by analytical HPLC using a Phenomenex Luna C18(2) column (250 × 4.6 mm) with a guard column (Phenomenex SecurityGuard C18, 4 × 3 mm) ([¹⁸F]MH.MZ RT = 6.8 min)(40% MeCN in 0.1M Ammonium Formate_(aq) (pH=4.5), 1 mL/min, 254 nm). (Figure 2) Analysis for residual organic solvents was carried out using an Agilent 7890B GC (Agilent Technologies Inc.) with a capillary column (length 30 m, ID 0.520 mm, DB-WAX 1.0 μm, Agilent Technologies Inc.). Apyrogenicity tests were performed in-house using Endosafe-Nextgen PTS (Charles River Laboratories Inc.) to ensure that doses of [¹⁸F]MH.MZ contained < 8.75 endotoxin units (EU) per mL. ColorpHast® pH indicator strips (EMD Chemicals Inc.) were used to determine pH of the final product. Residual Kryptofix levels were determined by spot test.¹⁷ Final filter integrity testing was performed on the sterilizing filter by standard bubble-point along with sterility testing via direct inoculation into growth media. Stability testing of the [¹⁸F]MH.MZ product was performed at periodic times over six hours, testing for radiochemical purity by HPLC and pH.

Table 1:

Quality Control Results

QUALITY CONTROL TEST	REQUIREMENT FOR PASS	QUAL RUN 1	QUAL RUN 2	QUAL RUN 3
Appearance	Clear, colorless, no particulates	Clear, colorless, no particulates	Clear, colorless, no particulates	Clear, colorless, no particulates
Filter Integrity	Pass	Pass	Pass	Pass
pH	4 – 7	5.0	5.0	5.0
Molar Activity	≥ 18.5 GBq/μmol (≥ 500 Ci/mmol)	409.7 GBq/μmol (11073 Ci/mmol)	375.3 GBq/μmol (10144 Ci/mmol)	299.0 GBq/μmol (8081 Ci/mmol)
Radiochemical Purity (%) (HPLC)	≥ 90%	99.70 %	99.71 %	99.74 %
Maximum Dose Volume based on Chemical Impurities	≤ 50 μg Total Chemical Impurities	> 11 mL (Whole dose may be injected)	> 11 mL (Whole dose may be injected)	> 11 mL (Whole dose may be injected)
Maximum Dose Volume based on MH.MZ Cold Mass	≤10 μg of MH.MZ	> 11 mL (Whole dose may be injected)	10.66 mL	9.04 mL
Radiochemical Identity Relative Retention Time from Reference Std) by HPLC	0.9 – 1.1	0.9588	0.9765	0.9834
Residual Acetonitrile Level	≤ 410 ppm	151.43 ppm	124.21 ppm	139.45 ppm
Residual DMSO Level, ppm	≤ 5000 ppm	0.00 ppm	0.00 ppm	0.00 ppm
Residual Methanol Level	≤ 3000 ppm	0.00 ppm	0.00 ppm	0.00 ppm
Kryptofix-222 Level	< 50 μg/mL	< 50 μg/mL	< 50 μg/mL	< 50 μg/mL
Bacterial Endotoxin Levels (EU/mL)	< 8.75 EU/mL	<1.00 EU/mL	<1.00 EU/mL	<1.35 EU/mL
Radionuclidic Identity (t1/2)	105 – 115 min	111.35 min	105.07 min	109.90 min
Sterility (Observed growth after 14 d)	No Growth	No Growth	No Growth	No Growth
Starting Activity	Not Specified	102.7 GBq (2775 mCi)	120.7 GBq (3262 mCi)	103.6 GBq (2801 mCi)
Amount of Product	Not Specified	6416 MBq (173.4 mCi)	8261 MBq (223.3 mCi)	8214 MBq (222.0 mCi)
Activity Yield (Non-corrected)	Not Specified	6.25 %	6.85 %	7.93 %

Figure 2:

HPLC Chromatograms of reference standard and [¹⁸F]MH.MZ

3. Results and Discussion.

The radiosynthesis of [¹⁸F]MH.MZ is shown in Scheme 1. Our approach was based upon the previously published reports.^{10–12, 18} Previously, [¹⁸F]MH.MZ has been prepared using either multiple synthesis modules or a combination of manual and automated methods for the two chemical steps as well as for the two HPLC purifications. Previous efforts have been made to avoid the separate synthesis of [¹⁸F]Fluoroethyltosylate and subsequent alkylation, however the direct labelling approach did not offer any significant benefit over the two step process.¹⁶ The published direct labelling approach required a precursor that was unstable under the reaction conditions as well as requiring specialized storage conditions to prevent air exposure. While the direct labelling approach would have allowed for a simpler automated process, the benefit of using stable, commercially available reagents, let to the two-step approach described herein. The iPhase Flexlab module is unique in terms of commercially available automated radiosynthesis modules in that it possesses two separate HPLC purification sub-systems as well as two reformulation steps; making it the ideal module for the radiosynthesis of [¹⁸F]MH.MZ. Furthermore, it has two reactors capable of performing independent synthetic steps. We were able to design the automated radiosynthesis on the Flexlab to take advantage of these capabilities to accomplish the two-step production of [¹⁸F]MH.MZ using the basic flowpath setup on the module, enabling both ease of method sharing and allowing the module to continue to be used for other processes. Despite the existence of the two HPLC columns, care must be taken to ensure proper equilibration of the second column immediately at the conclusion of the first HPLC purification for the HPLC pump.

Scheme 1:

Synthesis of [¹⁸F]MH.MZ

One modification of the published radiosynthesis was the substitution of tetrabutylammonium hydroxide (TBAOH) for sodium hydroxide. In our hands the use of aqueous sodium hydroxide resulted in no alkylation of the MDL-105725 with [¹⁸F]fluoroethyltosylate, even at elevated temperatures. We determined that this was likely due to a slow deprotonation process of the phenol, as reports for other radiotracers required extended incubation time at elevated temperature for the base to take full effect.¹⁹ Use of commercially available TBAOH solution in methanol allowed for high conversion in the alkylation step in a reasonable time-frame, without specialized deprotonation steps.²⁰

Using the optimized conditions established, [¹⁸F]MH.MZ was obtained in a non-decay corrected radiochemical yield of 7.01 ± 0.85 %, providing in our hands 7.63 ± 1 GBq (206 ± 28 mCi) of [¹⁸F]MH.MZ, which is sufficient for clinical research use. From the radioactivity measurements in the automated module the first step to form [¹⁸F]fluoroethyltosylate is ~30% yield, with the subsequent alkylation with MDL-105725 showing a highly efficient reaction. (Supplementary Figure 2) Each qualification batch met all quality control release specifications as listed in Table 1. The retention time of the MH.MZ reference material was 6–7 minutes, and the relative retention time of the labeled [¹⁸F]MH.MZ was within a 10% difference (Figure 2).

4. Conclusion.

We have successfully developed an automated radiosynthesis of (R)-[¹⁸F]MH.MZ using a single commercially available reaction module and commercially available materials. Use of a single reaction platform allows for a simple, reliable process that gives good radiochemical yields and an excellent purity profile. Overall, the final (R)-[¹⁸F]MH.MZ passes all release criteria, reliably provides a sterile and pyrogen-free GMP-compliant final product, and has been approved as an Investigatory New Drug by the US Food and Drug Administration.