Madison
Frazier‡
,
Jay S.
Wright‡
*,
David M.
Raffel
,
Jenelle
Stauff
,
Wade P.
Winton
,
Peter J. H.
Scott
* and
Allen F.
Brooks
*
Department of Radiology, University of Michigan, Ann Arbor, Michigan 48109, USA. E-mail: jawr@med.umich.edu; pjhscott@med.umich.edu; afb@med.umich.edu
First published on 31st July 2024
The most prominent myocardial voltage-gated sodium channel, NaV1.5, is a major drug target for treating cardiovascular disease. However, treatment determination and therapeutic development are complicated partly by an inadequate understanding of how the density of SCN5A, the gene that encodes NaV1.5, relates to treatment response and disease prognosis. To address these challenges, imaging agents derived from NaV1.5 blocking therapeutics have been employed in positron emission tomography (PET) imaging to infer how SCN5A expression relates to human disease in vivo. Herein, we describe the preparation of a novel fluorine-18 labelled analogue of lidocaine, a known NaV1.5 inhibitor, and compare this agent to a previously described analogue. Evidence from rodent and non-human primate PET imaging experiments suggests that the imaging utility of these agents may be limited by rapid metabolism and clearance.
Owing to the reported efficacy of this tracer, and since we have recently reported facile radiolabelling of the lidocaine scaffold via new chemistry developed in our group, we opted to conduct further preclinical investigations to establish the metabolic properties of fluorine-18 labelled lidocaine derivatives. Notably, lidocaine undergoes a well-understood metabolic pathway beginning with oxidative N-dealkylation of the tertiary amine by cytochrome-p450 enzymes (CYP). Radiocaine, containing a sp3 C–18F label at the terminus of an ethyl chain (Fig. 1), may be susceptible to an analogous dealkylation, thereby eliminating the key fluorine-18 label. Therefore, it is currently unclear whether the myocardial uptake of radiocaine originates solely from the parent imaging agent or from metabolites that exhibit non-specific binding, which could confound Na channel evaluation via PET. We sought to test this hypothesis by investigating 1-18F, an analogue of radiocaine labelled on the aromatic ring and previously synthesised by our laboratories by sequential C–H radiofluorination.5
Herein, we describe the development of a fully automated radiosynthesis of lidocaine derivative 1-18F, labelled as an aromatic radiofluoride via sequential iridium-catalysed C–H borylation and copper-mediated radiofluorination. This radiosynthetic protocol enables rapid access to sufficient 1-18F for preclinical evaluation. It was hypothesised that CYP dealkylation of 1-18F, in analogy to the metabolic dealkylation of 1-H, would afford an alternative distribution of radiolabelled metabolites to radiocaine, which could be visualised in vivo via PET. Therefore, we evaluated the preclinical efficacy of this agent in rodents and non-human primates (NHP). These imaging studies elucidate the utility of fluorine-18 labelled lidocaine scaffolds as potential radiopharmaceuticals for evaluating myocardial Na channel density and monitoring cardiovascular disease progression and treatment.
Therefore, conditions that would more efficiently activate the C–Hb bond of 1-H were sought. It was hypothesised that elevated temperature could improve the solubility of 2-Bpin and allow for C–H borylation in improved conversions. To our delight, doubling the loading of the precatalyst components, increasing the temperature from 80 °C to 100 °C, and switching HBpin with the more reactive B2pin2 (2.20 equivalents), CHB Conditions B, afforded 3-Bpin as the major product in >90% conversion as determined via1H NMR spectroscopy. Notably, precipitation was not observed over the course of this reaction. Furthermore, the aromatic protons in 3-Bpin (Ha′′) are, as expected, shifted downfield relative to the aromatic protons of 1-H (Ha) in the crude 1H NMR spectrum, originating from the ortho-deshielding effect of the aromatic boryl group. Furthermore, by analysing the crude reaction mixture under inert conditions, we could also clearly observe diborylated 3-(Bpin)2 as an additional product of this reaction. Treatment of this crude reaction mixture with EtOH decomposed 3-(Bpin)2, leaving 3-Bpin as the sole product, corresponding to N–B alcoholysis.
The automated labelling of 1-H was next investigated with the optimised CHB conditions. Pleasingly, 1-18F could be prepared for preclinical imaging studies using a modified labelling protocol (Scheme 2). This first involved independently preparing crude 3-Bpin from 1-H under manual Ir-catalysed CHB conditions B. Next, [18F]TBAF was prepared from cyclotron-produced 18F−via an azeotropic dry-down in a commercial radiosynthesis module (GE TRACERlab FXFN). [Cu(OTf)2py4] dissolved in DMA, followed by an aliquot of the crude CHB mixture containing boronate 3-Bpin and nBuOH dissolved in DMA, were added successively to the reactor.8 The reaction mixture was heated at 120 °C for 20 min, followed by semi-preparative purification with a reverse phase column (Kinetex F5 5 μm, 100 Å, 250 × 10 mm) using 10 mM NH4HCO3 buffer in 30% MeCN/H2O at pH 10. Reformulation with a C18 Sep-Pak cartridge afforded 17 ± 10 mCi (629 ± 370 mBq) 1-18F in 0.91 ± 0.55% (n = 4) isolated non-decay-corrected (ndc) radiochemical yield (RCY) and >99% radiochemical purity (RCP), determined by radio-high-performance liquid chromatography. 1-18F was also obtained in 3.76 Ci μmol−1 ndc molar activity (Am) with a 141 ± 19 min synthesis time (n = 4, see ESI† for full details).
We next imaged Sprague–Dawley rodents using 1-18F (Fig. 2 and 3). To rationalise the binding and clearance profiles of 1-18F and radiocaine, we considered the probable metabolic pathways of each imaging agent to predict the fate of each fluorine-18 label based on the known metabolism of 1-H (Scheme 3). Critically, 1-H is rapidly distributed in vivo with an initial biological half-life of 30 min. Clearance occurs at a rate of 1.44 L min−1, and >95% of the parent compound is converted to various metabolites, including monoethylglycinexylidide (MEGX) and glycinexylidide (GX), by CYP. Deethylation produces acetaldehyde, which can undergo oxidation by alcohol dehydrogenase to form acetate, which is converted into acetyl CoA for entry into the TCA cycle.9 In analogy, the [18F]fluoroethyl chain in radiocaine may be susceptible to this process, forming [18F]fluoroacetaldehyde ([18F]FCH2CHO) followed by [18F]fluoroacetate ([18F]FCH2CO2−), which would instead be trapped in the TCA cycle owing to the presence of the C–F bond. Indeed, cardiac uptake has been documented for [18F]fluoroacetate.10,11 Conversely, 1-18F should produce the fluorine-18 labelled analogues of monoethylglycinexylidide and glycylxylidide, [18F]FMEGX and [18F]FGX, respectively.
![]() | ||
Fig. 2 Selected rodent PET imaging data obtained using 1-18F. Upper: Rodent total summed image frames with rainbow colour table maximum intensity projection baseline control study. Lower: Rodent total summed image frames with rainbow scale table maximum intensity projection – blocking study with 1-H. Very rapid clearance of 1-18F was observed from the heart (ca. 2 min post-injection) and a significant uptake in the kidneys and bladder, which is consistent with the generation of metabolites such as [18F]FMEGX and [18F]FGX. Two of the rats received an injection of 2 mg mL−1 of lidocaine immediately before imaging for blocking studies. See ESI† for further details. |
![]() | ||
Fig. 3 Time activity curves displaying kinetic data of rodent baseline control study (upper) and 1-H blocking study (lower) (see ESI† for further rodent and NHP PET data). |
![]() | ||
Scheme 3 Known metabolic pathway for lidocaine 1-H and analogous metabolic pathways predicted for radiocaine and 1-18F. |
The rodent PET data shows that 1-18F indeed exhibits uptake in the myocardium (4% ID g−1, Fig. 2) followed by rapid elimination, with most of the signal detected within 15 min from the start of the scan. Lung uptake was also recorded, which is consistent with the pharmacokinetics of 1-H.12 For cardiothoracic regions of interest (e.g., heart wall and lungs), 1-18F has fast clearance (ca. 2 min), whereas the liver, kidneys, and bladder show either much slower clearance (>90 min in the liver, see ESI† for further time activity curves) or retention. However, an experimentally significant difference exists between the baseline and lidocaine-blocking images. Therefore, to account for the possible differences in metabolic rates and mechanisms between species, primate PET imaging studies were subsequently conducted using 1-18F (Fig. 4). Very rapid clearance of 1-18F was observed from the heart (ca. 2 min post-injection) and a significant uptake in the kidneys and bladder, which is consistent with the generation and renal uptake of metabolites such as [18F]FMEGX and [18F]FGX (see ESI† for further rodent and NHP PET data). Compared to radiocaine, this low myocardial uptake may be explained by the absence of the labelled metabolite [18F]fluoroacetate, which likely confounds NaV1.5 occupancy evaluation via PET. Therefore, the apparent retention of radiocaine may not originate from NaV1.5 binding alone. Combined, these preliminary observations call into question the Na channel target specificity of radiocaine and 1-18F and their suitability as agents for assessing NaV1.5 occupancy.
![]() | ||
Fig. 4 Upper: Non-human primate total summed image frames with fire table maximum intensity projection (baseline control study). Lower: Time activity curves displaying kinetic data of NHP baseline control study. No blocking study was performed owing to low cardiac uptake in the baseline study (see ESI† for further rodent and NHP PET data). |
Footnotes |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4md00293h |
‡ These authors contributed equally. |
This journal is © The Royal Society of Chemistry 2024 |