Dynamic 18F-Pretomanid PET imaging reveals excellent brain penetration in animal models of TB meningitis and first-in-human studies

doi:10.21203/rs.3.rs-1999411/v1

Download PDF

Article

Dynamic 18F-Pretomanid PET imaging reveals excellent brain penetration in animal models of TB meningitis and first-in-human studies

https://doi.org/10.21203/rs.3.rs-1999411/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 29 Dec, 2022

Read the published version in Nature Communications →

Version 1

posted

You are reading this latest preprint version

Pretomanid is a nitroimidazole antimicrobial with potent activity against drug-resistant Mycobacterium tuberculosis and approved in combination with bedaquiline and linezolid (BPaL) for the treatment of multidrug-resistant (MDR) pulmonary tuberculosis (TB). However, the penetration of pretomanid into privileged sites such as the brain, as well as the efficacy of the BPaL regimen for TB meningitis, remain unknown. Importantly, there are no well accepted treatments for TB meningitis due to MDR strains, resulting in high mortality. We developed a new synthetic route to obtain ¹⁸F-pretomanid (chemically identical to the parent antibiotic) and performed cross-species positron emission tomography (PET) imaging to noninvasively measure intralesional pretomanid concentration-time profiles in the central nervous system (CNS). Dynamic PET in mouse and rabbit models of TB meningitis demonstrated excellent CNS penetration of pretomanid. Antibiotic levels were spatially compartmentalized and cerebrospinal fluid (CSF) levels did not correlate with those in the brain parenchyma. Post-mortem autoradiography and mass spectrometry corroborated the imaging findings. The bactericidal activity of the BPaL regimen in the mouse model of TB meningitis was substantially inferior to the standard TB regimen, likely due to restricted penetration of bedaquiline and linezolid into the brain parenchyma. Finally, first-in-human dynamic ¹⁸F-pretomanid PET in six healthy volunteers demonstrated excellent penetration of pretomanid into the CNS, with significantly higher levels in the brain parenchyma versus CSF (P < 0.01). These data have important implications for developing new antibiotic treatments for TB meningitis.

Tuberculosis (TB) remains one of the leading killers from a single infectious agent¹ and TB meningitis is the most devastating extrapulmonary form, especially in the young and immunocompromised^{2, 3, 4}. Multidrug-resistant (MDR)-TB, caused by Mycobacterium tuberculosis resistant to first-line antibiotics (i.e., isoniazid and rifampin), is on the rise. TB meningitis due to MDR strains is associated with high mortality^{5, 6, 7}, and drug-resistance is an independent predictor of death⁸. In a recent retrospective cohort study among 237 patients with TB meningitis, mortality was significantly higher among patients with drug-resistant (67%) versus drug-susceptible disease (24%, P < 0.001)⁹. Moreover, mortality was significantly higher (adjusted hazards ratio of 7.2) in patients with drug-resistant TB meningitis after 90 days of initiation of treatment (P < 0.001). New drugs and more efficacious treatments against MDR-TB are therefore urgently needed to combat this public health threat. Pretomanid (formerly PA-824) is a small molecule belonging to the nitroimidazole class of antimicrobial agents, approved by the U.S. Food and Drug Administration (FDA) in 2019 for the treatment of pulmonary MDR-TB, in combination with bedaquiline and linezolid (BPaL - bedaquiline, pretomanid, linezolid)¹⁰. Pretomanid is active against both replicating and non-replicating M. tuberculosis, which contributes to its excellent bactericidal activity^{11, 12, 13, 14}.

With few exceptions, current antibiotic dosing recommendations are based on plasma concentrations, without information on drug concentrations at the site of infection. Since inappropriate antibiotic levels in target tissues can lead to selection of resistant organisms, toxicity or organ injury, and ultimately treatment failure, a growing number of studies and the U.S. FDA increasingly support measuring antibiotic concentrations in infected tissues¹⁵. Therefore, we have developed positron emission tomography (PET)-based clinically-translatable technologies for noninvasive, simultaneous, and unbiased, multi-compartment in situ measurements of antibiotic concentration-time curves in animals and humans^{16, 17, 18}. In this study, we report the development of ¹⁸F-pretomanid as a molecular imaging tool to noninvasively assess whole-body drug biodistribution (Fig. 1). Detailed animal studies in mouse and rabbit models of TB meningitis were utilized^{17, 19, 20}. In brief, infected animals underwent dynamic PET/computed tomography (CT) with ¹⁸F-pretomanid to obtain time-activity curves (TACs) and areas under the curve (AUCs) by quantifying the PET signal in volumes of interest (VOI). Post-mortem autoradiography and histology were also performed in all animal models. Given the unknown potential of pretomanid-containing regimens for TB meningitis, the bedaquiline (B), pretomanid (P) and linezolid (L) regimen was tested in a mouse model of TB meningitis²⁰. Mass spectrometry and traditional microbiology techniques were employed to evaluate intraparenchymal drug levels and bactericidal efficacy (bacterial burden quantified with colony-forming units [CFU]) longitudinally. Radiosynthesis of ¹⁸F-pretomanid under current Good Manufacturing Practices (cGMP) facilitated translation into humans and first-in-human dynamic ¹⁸F-pretomanid PET studies were performed in accordance with U.S. FDA guidelines.Fig. 1. Schematic representation of the experimental design. Mouse and rabbit models of TB meningitis were used for radiotracer validation, pharmacokinetic characterization and pretomanid-containing regimen testing. ¹⁸F-Pretomanid synthesis was optimized for clinical translation under cGMP and first-in-human PET imaging was performed. ¹⁸F-Pretomanid was administered intravenously. Following dynamic PET/CT imaging, several volumes of interest (VOI) were drawn to obtain PET-derived time-activity curves (TACs). TACs were used to calculate the area under the curve (AUC) and AUC tissue/plasma ratios.

Radiosynthesis and characterization of ¹⁸F-pretomanid

¹⁸F-Pretomanid was obtained through halogen exchange ¹⁸F-fluorination (Fig. 2) of an aryl-bromodifluoromethoxyl precursor, which was obtained from a five-step synthetic route (Fig. S1-2). The radiosynthetic approach employed a silver-catalyzed nucleophilic displacement, following an adapted protocol by Khotavivattana et al²¹. We obtained a radiochemical yield (RCY) of 5 ± 2% (non-decay corrected [n.d.c.]), from starting ¹⁸F-fluoride to formulated ¹⁸F-pretomanid, and found that reducing the amount of precursor or changing the solvent had a detrimental effect on the RCY. High-performance liquid chromatography (HPLC) analysis on the isolated product revealed a radiochemical purity > 98%. While this method allowed the radiosynthesis of ¹⁸F-pretomanid for animal studies, the use of dichloroethane precluded its clinical translation. Therefore, we tested alternative reaction solvents to translate the synthesis of ¹⁸F-pretomanid under cGMP conditions which unfortunately significantly reduced RCY. However, automated cGMP synthesis of ¹⁸F-pretomanid was successful under microwave irradiation at 100 watts for 10 min (reaching 120°C) in dimethylformamide in the absence of silver salts, despite previous reports that catalyst-free thermal activation had been unsuccessful for the preparation of ¹⁸F-labeled aryl-OCF₃ compounds. Under these conditions, ¹⁸F-pretomanid was obtained in 5.7 ± 0.3% n.d.c. yield and a specific activity of 68 ± 2 GBq/µmol. HPLC analysis showed ≥ 95% radiochemical purity and a single peak corresponding to the ¹⁹F-reference pretomanid in the UV chromatogram (Fig. S3).

In vitro, unbound ¹⁸F-pretomanid remained stable (> 90%) in mouse, rabbit, and human serum at 37°C for three hours. Defluorination was not observed. Pretomanid is known to be highly protein bound (~ 86%) in human plasma²². When incubated with mouse, rabbit, and human serum at 37°C, the protein binding level of ¹⁸F-pretomanid was 75–77% in healthy human, 78–80% in healthy rabbit, 80–83% in M. tuberculosis-infected rabbit, 75–80% in healthy mouse, and 74–78% in M. tuberculosis-infected mouse sera (table S1). Overall, ¹⁸F-pretomanid had similar protein binding (74–83%) to unlabeled pretomanid, and no significant differences were found between species over time. ¹⁸F-Pretomanid experimental LogD_7.4, which represents its distribution coefficient at physiological pH, was 1.9 ± 0.1, which is only a 0.4 Log decrease when compared to unlabeled pretomanid²³. Thus, ¹⁸F-labeled and unlabeled pretomanid are expected to have similar tissue partitioning. Whole-body biodistribution of ¹⁸F-pretomanid was measured in mice with experimentally-induced pulmonary TB utilizing PET/CT and gamma counting (Fig. S4). Upon intravenous injection, ¹⁸F-pretomanid rapidly distributed to all major organs, which was also confirmed by post-mortem biodistribution quantification by gamma-counting (Fig. S4A-B). The activity in the bone was low and did not substantially increase over time which indicates that defluorination did not occur in vivo (Fig. S4C). Similar to the parent drug, ¹⁸F-pretomanid underwent both renal and hepatobiliary excretion (Fig. S4D). Low uptake was observed in muscle and high uptake was found in brown adipose tissue (BAT), which cleared within hours (Fig. S4E). Spatial distribution with ex vivo autoradiography in the mouse model of pulmonary TB showed reduced uptake of ¹⁸F-pretomanid in lung lesions compared to unaffected lung (P < 0.0001; Fig. S4F). The upper-body biodistribution of ¹⁸F-pretomanid was also measured in rabbits showing similar findings as in mice (Fig. S5).

¹⁸F-Pretomanid has excellent central nervous system (CNS) penetration

Our initial studies characterizing the whole body biodistribution of ¹⁸F-pretomanid in mouse models of pulmonary TB revealed that the penetration into the brain parenchyma was high, with a median AUC ratio (brain/plasma) of 1.73 (IQR, 1.41–2.04), even in healthy brains. This prompted detailed investigation of ¹⁸F-pretomanid penetration into infected brains in two mammalian models of TB meningitis also showing excellent brain penetration of ¹⁸F-pretomanid [AUC ratios (brain/plasma) > 1]. However, in vivo 3D PET/CT and 2D ex vivo autoradiography in the mouse model of TB meningitis showed reduced uptake of ¹⁸F-pretomanid with filling defects at the center of the brain lesion (visible in live animals with ¹⁸F-FDG PET/CT and ex vivo histopathology, respectively) (Fig. 3A-B). The AUC ratio (brain/plasma) was lower in the brain lesions (median, 1.35; IQR, 0.81–1.52) compared to the unaffected brain (median, 1.56; IQR, 1.22–1.69; P = 0.0004; Fig. 3C-D). Similar findings were noted in a rabbit model of TB meningitis, with median AUC ratio (brain/plasma) of 1.87 (IQR, 1.66–4.63) into brain lesions (Fig. 3E, F) versus 2.75 (IQR, 1.64–5.73) into the unaffected brain (Fig. 3G-H).

Assessment of the BPaL regimen in the mouse model of TB meningitis

We measured the efficacy of the pretomanid-containing BPaL regimen, currently the only U.S. FDA approved regimen for MDR-TB, in the mouse model of TB meningitis and compared it with the first-line, standard TB treatment (standard-dose rifampin, isoniazid, and pyrazinamide – HR₁₀Z) (Fig. 4A). Adjuvant dexamethasone was administered with both regimens per current TB meningitis treatment guidelines^{2, 26}. While both regimens decreased the bacterial burden compared to untreated mice, the bactericidal activity of the BPaL regimen was substantially inferior to the standard TB regimen (P < 0.001) (Fig. 4B-D and table S2).

In previous studies^{24, 25}, we have reported that the AUC ratios (brain/plasma) of ⁷⁶Br-bedaquiline and ¹⁸F-linezolid (both chemically identical to the parent drugs) into uninfected brain tissues were 0.15 and 0.34 respectively (Fig. 4E). However, our current study demonstrated that ¹⁸F-pretomanid AUC ratios (brain/plasma) were high (> 1.5) at the start of treatment (Fig. 4E) and remained high (> 1) after two weeks of treatment (Fig. S6). Brain parenchymal and CSF drug and metabolite levels were also measured by mass spectrometry, which demonstrated discordant penetration into the brain parenchyma and CSF compartments (P < 0.01, Fig. 4F-H and table S3). While linezolid levels were higher in the CSF compared to the brain parenchyma, both pretomanid and bedaquiline levels were higher in the brain parenchyma compared to the CSF. Bedaquiline rapidly undergoes N-demethylation in vivo to form a metabolite (M2), which is also active against M. tuberculosis. We found that M2 levels (albeit still low) were higher in the brain parenchyma than the parent drug (table S3).

It has been hypothesized that outcomes in TB meningitis are also associated with changes in intracerebral inflammation^{2, 5}. Therefore, we also assessed levels for select inflammatory cytokines in brain lysates, which did not show significant differences between the two regimens (Fig. S7).

First-in-human ¹⁸F-Pretomanid PET studies

Six healthy volunteers (three male and three female), with an age range of 20–53 years were recruited in an ongoing first-in-human ¹⁸F-pretomanid PET study at the Johns Hopkins Hospital (table S4). The procedures were safe, no adverse or clinically detectable pharmacological effects were noted and no anatomic abnormalities were detected in any subject. Similar to the animal studies, after intravenous administration, ¹⁸F-pretomanid rapidly distributed to all major organs, and underwent both renal and hepatobiliary excretion. Importantly, ¹⁸F-pretomanid demonstrated excellent penetration into the brain parenchyma and the CSF, with an AUC (tissue/plasma) ratio > 1 (Fig. 5 and Fig. S8). Additionally, and similar to the findings in the animal models, ¹⁸F-pretomanid exposures were compartmentalized with significantly lower penetration noted in the CSF (ventricles), compared to the brain parenchyma (Fig. 5; P < 0.05).

While optimal drug dosing can shorten current treatment regimens, suboptimal dosing is a major factor promoting antibiotic resistance, which the World Health Organization declared as one of the top ten threats to human health²⁷. Notably, there are no well-accepted treatments for TB meningitis due to MDR strains, which are associated with 40–100% mortality with current regimens^{5, 6, 7, 9}. Therefore, there is an urgent need to optimize antibiotic treatments for TB meningitis, especially for those caused by MDR strains.

Pretomanid has been studied for pulmonary TB, and exhibits time-dependent bactericidal activity in the lungs²². However, there is limited data on its penetration into privileged sites such as the CNS². Additionally, TB treatments can be particularly challenging because of the simultaneous co-existence of heterogeneous lesions with a complex microenvironment in the same host^{16, 17, 20, 28}. For instance, the blood-brain barrier (BBB) limits the access of several antimicrobials into the CNS^{2, 20}. Although mass spectrometry studies have demonstrated good CNS penetration of pretomanid in uninfected, healthy rats²⁹, these data provide measures only at a single time point and do not provide detailed spatiotemporal concentration-time measures in infected tissues. In addition to the unique microenvironments at the infection site, the physicochemical properties of specific drugs affect pharmacokinetic (PK) parameters such as drug penetration and clearance³⁰. Our studies demonstrate that pretomanid has excellent penetration into the brain, including infected brain tissues in both mice and rabbits. While prior studies have shown that rifampin penetration into the brain decreases dramatically as early as two weeks after initiation of TB treatments^{17, 20}, we show that CNS penetration of pretomanid remains high [AUC (tissue/plasma) ratio > 1] even after initiation of treatment with dexamethasone-containing regimens, which decreases BBB permeability.

Given the high brain penetration of ¹⁸F-pretomanid, we measured the efficacy of the BPaL regimen, currently, the only U.S. FDA approved regimen for MDR-TB, in the mouse model of TB meningitis and compared it with the first-line, standard TB treatment (HR₁₀Z). Adjuvant dexamethasone was administered with both regimens per current TB meningitis treatment guidelines. While the BPaL regimen substantially decreased the bacterial burden in the brain (compared to untreated mice), surprisingly and contrary to what is noted for pulmonary TB³¹, the bactericidal activity of the BPaL regimen was substantially inferior to the standard TB regimen, even six weeks after treatment initiation.

To understand the lower than expected efficacy of the BPaL regimen, we measured the penetration of each antibiotic into the CNS using PET (AUC measures in live animals) and single time-point post-mortem levels using mass spectrometry. As noted, pretomanid has excellent penetration into the CNS. However, linezolid levels were lower than anticipated based on other published literature³², and bedaquiline levels were substantially lower (tissue/plasma ratio < 0.1), but in line with the low CNS levels noted in published studies^{33, 34}. Importantly, we also measured the antibiotic levels by mass spectrometry, which demonstrated discordant concentrations in the brain parenchyma and CSF compartments, and were influenced by drug properties. For example, linezolid, an oxazolidinone with low protein-binding that was found to be beneficial in retrospective studies for the treatment of TB meningitis³⁵, had substantially higher CSF concentration, but lower levels were achieved in the brain tissues. Conversely, both bedaquiline (and M2 metabolite) as well as pretomanid, which are lipophilic and with higher protein-binding, showed higher concentration in the brain tissue compared with the CSF. This is consistent with differential partitioning of drugs into the protein and lipid rich brain parenchyma versus CSF which is hydrophilic and low in protein. We have noted similar discordance between the CSF and the brain tissue compartments in antibiotic levels^{17, 20, 36}, and markers of inflammation²⁰. The current studies extend this concept demonstrating that drug properties also influence discordant penetration into the CNS.

The ability to perform cross-species and translational clinical studies with ¹⁸F-pretomanid PET is a major advantage of this technology. Therefore, we developed a cGMP-compliant radiosynthesis scheme for ¹⁸F-pretomanid to facilitate clinical translation and perform first-in-human imaging studies. First-in-human, dynamic ¹⁸F-pretomanid PET in healthy controls confirmed the excellent but heterogenous brain penetration visualized in our mammalian models of TB meningitis, validating ¹⁸F-pretomanid PET as a powerful translational tool in drug development and treatment optimization. Additionally, the ¹⁸F-pretomanid PET signal in the CSF (brain ventricles), was significantly lower than in other regions of the brain parenchyma except for the corpus callosum. These results confirm compartmentalized and discordant antibiotic penetration into the CNS and highlight the importance of using animal models and imaging technology in humans that can measure antibiotic exposures in the brain parenchyma as well as the CSF.

Our studies have some limitations which we aimed to overcome. Since even minor modifications in the chemical structure of antimicrobials can lead to drastic changes in the physicochemical properties and biological activity of small molecule drugs, we developed a radiosynthetic approach using radiofluorination chemistry that retains the chemical identity of pretomanid. Additionally, the PET studies administered microdoses (ng-µg) of ¹⁸F-pretomanid per subject for the PET studies and current evidence suggests that microdosing is a reliable predictor of the drug biodistribution at therapeutic doses^{37, 38}. Another limitation of the study is that mass spectrometry drug levels were measured at Tmax (time to reach maximum concentration) and represent a single time point, and not time-concentration curves (AUC) as measured by PET. This difference in measurement could explain why the brain/plasma ratios were found to be higher by mass spectrometry compared to PET-derived AUC ratios. Finally, while rabbit studies utilized both males and females, only female mice were used, and future studies with both males and females would be required to investigate how gender may affect pretomanid PK.

In summary, this study highlights the value added by imaging technologies in PK characterization of novel drugs and treatment optimization. The use of radiolabeled antimicrobials, combined with conventional microbiology techniques and novel animal models, offers the unmatched potential to noninvasively assess PK profiles in infected tissues, characterized by complex, heterogeneous lesions existing simultaneously in the same host²⁸. Importantly, our studies demonstrate that pretomanid has excellent penetration into the CNS, confirmed in first-in-human ¹⁸F-pretomanid studies, suggesting that novel pretomanid-based regimens in combination with other antibiotics active against MDR strains and with excellent brain penetration should be considered for the treatment of MDR-TB meningitis. Finally, our studies reaffirm the compartmentalized and discordant antibiotic penetration into the CNS, which are related to the physiochemical properties of the antibiotic. Overall, this is an important finding as CSF studies are commonly utilized in many clinical trials, but CSF may not be an adequate surrogate of disease or drug levels in the brain tissues.

Study design

The overall goal of this study was to develop radiolabeled-pretomanid for dynamic, concentration-time profiles using PET to characterize pretomanid brain PK (in animal and human studies) and assess the efficacy of a pretomanid-containing regimen for TB meningitis. We performed cross-species analyses for preclinical validation of ¹⁸F-pretomanid as a molecular imaging tool to noninvasively assess whole-body drug penetration and then optimized the radiosynthesis of ¹⁸F-pretomanid under cGMP to facilitate the execution of first-in-human studies. Dynamic ¹⁸F-pretomanid PET/CT was used to obtain TACs and AUCs by quantifying the PET signal in VOI in animal models and human studies. In animal models, post-mortem autoradiography/histology and mass spectrometry were also performed to characterize pretomanid's heterogenous brain penetration, both spatially and temporally. Given the unknown potential of pretomanid-containing regimens for TB meningitis, BPaL, the only FDA approved pretomanid-containing regimen, was tested in the mouse model of TB meningitis to evaluate intraparenchymal drug levels and bactericidal activity longitudinally. All protocols were approved by the Johns Hopkins University Biosafety, Radiation Safety, Animal Care and Use (RB19M417, MO19M382) and Institutional Review Board Committees (IRB00303845). ¹⁸F-Pretomanid was used according to the FDA Radioactive Drug Research Committee guidelines for investigational drugs.

Radiosynthesis of ¹⁸F-pretomanid

Radiosynthesis of ¹⁸F-pretomanid was first developed as a manual synthesis using silver-catalyzed nucleophilic displacement for preclinical animal studies. For clinical translation, the radiosynthesis was adapted to an automated cGMP synthesis of ¹⁸F-pretomanid under microwave irradiation in dimethylformamide in the absence of silver salts. Details of the synthesis of ¹⁸F-pretomanid, including the radiolabeling precursor and intermediates and in vitro characterization of ¹⁸F-pretomanid, is available in supplementary materials.

Animal studies

Mouse model of Pulmonary TB

Four to six-week-old female C3HeB/FeJ mice (Jackson Laboratory) were aerosol-infected with frozen stocks of M. tuberculosis (H37Rv), using the Middlebrook Inhalation Exposure System (Glas-Col)³⁹.

Mouse model of TB meningitis

Studies were performed as described by us previously²⁰. Briefly, 2.7 µL of M. tuberculosis H37Rv (from titrated frozen stocks) was injected intraventricularly in female C3HeB/FeJ mice (7–8 weeks old) using a stereotactic device, and infection was allowed to incubate for 14 days.

Rabbit model of TB meningitis

Studies were performed as described by us previously^{17, 19, 20}. Briefly, male and female New Zealand White rabbits (5–7 day old, Robinson Services Inc.) were infected using a direct intracranial inoculation of 6.67 ± 0.67 CFU/g (Log₁₀) modified for intraventricular injection. M. tuberculosis infection was allowed to incubate for at least 14 days before PET imaging was performed.

Ex vivo biodistribution

Anesthetized mice (n = 4 per time point) were injected with ¹⁸F-pretomanid (2.6 ± 1.9 MBq) via the tail vein and sacrificed 1, 2, and 4 h post-injection. Organs and tissues of interest were harvested, weighed, and the radioactivity gamma-counted. Data are represented as %ID/g.

Autoradiography

Following intravenous injection of ¹⁸F-pretomanid, animals were sacrificed, perfused with saline, and brains collected. Tissues were embedded in OCT and 20 µm sections placed on slides. The radioactive slides were placed inside an exposure cassette (GE, code no. 29175523) and exposed for 2–5 half-lives before being developed by phosphor storage imaging (GE Typhoon). Following radioactive decay, the tissues used for autoradiography were fixed in formalin and underwent hematoxylin and eosin staining.

Antimicrobial treatments and efficacy: Drug stocks were prepared and administered five days a week as previously described^{31, 40}. Rifampin (10 mg/kg/day), isoniazid (10 mg/kg/day), pyrazinamide (150 mg/kg/day), pretomanid (50 mg/kg/day divided twice daily) and linezolid (100 mg/kg/day divided twice daily) were prepared for oral administration via gavage and dexamethasone was prepared for intraperitoneal injection and stored at 4°C. Mouse dosing was utilized to match the standard human equipotent dosing: rifampin (10 mg/kg/day), isoniazid (10 mg/kg/day), pyrazinamide (25 mg/kg/day), pretomanid (50 mg/kg/ day) and linezolid (100 mg/kg/day). Treatment efficacy was determined by whole brain bacterial burden (CFU) at 2 and 6 weeks after initiation of treatment using 7H11 plates supplemented with activated charcoal.

Imaging: M. tuberculosis-infected mice and rabbits were imaged inside transparent and sealed biocontainment containers (Minerve), compliant with biosafety level (BSL)-3 containment and capable of delivering O₂-anesthetic mixture to sustain live animals during imaging as previously described⁴¹. PET acquisition and subsequent CT were performed using the nanoScan PET/CT (Mediso, Arlington, VA) and images automatically co-registered. ¹⁸F-Pretomanid PET/CT: Mice were injected with ¹⁸F-pretomanid (pulmonary TB, 3.77 ± 1.78 MBq; TB meningitis, 3.88 ± 1.44 MBq) via the tail vein. Anesthetized rabbits were injected with ¹⁸F-pretomanid (4.65 ± 0.75 MBq) through the ear vein. The injection time coincided with the start of the dynamic PET acquisition. ¹⁸F-FDG PET/CT: Animals were imaged as previously described²⁰.

Mass spectrometry

Terminal samples (blood, brain and CSF) were collected at the appropriate plasma T_max for each antimicrobial drug^{42, 43, 44}. Blood was collected in EDTA tubes (BD Microtainer, Fisher Scientific) for plasma separation. Bedaquiline, bedaquiline M2 metabolite, pretomanid and linezolid in plasma, CSF and brain tissues from M. tuberculosis-infected mice were quantified using validated ultra-high-performance liquid chromatography (UPLC) and tandem mass spectrometry (LC–MS/MS) at the Infectious Diseases Pharmacokinetics Laboratory of the University of Florida. Calibration standard curves ranges were bedaquiline (and M2 metabolite) 2.00 to 0.01 µg/mL, pretomanid 30.00 to 0.01 µg/mL, and linezolid 30.00 to 0.03 µg/mL.

Cytokine analysis

Samples for cytokine analysis were collected from supernatants after centrifuged brain homogenization and stored at -80°C until analysis. IL-1α, vascular endothelial growth factor A (VEGFA) and matrix metalloproteinase-8 (MMP-8) were analyzed on Luminex Multiplex assays by the Oncology Human Immunology Core of Johns Hopkins University.

Human studies

¹⁸F-Pretomanid was synthesized as a sterile solution with high specific activity (40.27 ± 12.82 GBq/µmol) and high radiochemical purity (> 95%) using cGMP by the Johns Hopkins PET Radiotracer Center. Written informed consent was obtained from all healthy volunteers and deidentified images were analyzed. All subjects had a physical exam by a trained physician and screening laboratory tests before imaging to confirm eligibility. Each subject received an intravenous bolus of 359.52 ± 2.79 MBq of ¹⁸F-pretomanid followed by dynamic PET utilizing a multi-bed protocol immediately after tracer injection (0–60 min) and 180 min after tracer injection (180–210 min) using Siemens Biograph mCT 128-slice scanner.

Image analysis

PET/CT images were analyzed using VivoQuant 2020 (Invicro) or PMOD (3.402) for animal and human studies, respectively. In humans, the brain segmentation tool (Hammers N30R83) was used to draw VOIs and OsiriX MD 11.0 DICOM Viewer (Pixmeo SARL) was used to create 3D MIP images. Data for blood were obtained by placing a VOI in the left heart ventricle and then correcting to plasma using individual subject hematocrit values or standard hematocrit⁴⁵ and red blood cell partition coefficient of pretomanid⁴⁶ (table S5). PET data were adjusted for mass using the density of each organ obtained from the CT (Hounsfield units). Data are expressed as percent injected dose per weight of tissue (%ID/g) or standard uptake values (SUV) in animal or human studies, respectively. Heatmap overlays were created using RStudio.

Statistical analysis

Data were analyzed using Prism 8 version 8.1.0 (GraphPad). AUCs were calculated using the linear trapezoidal rule. Comparisons were made using a two-tailed, Mann-Whitney-Wilcoxon test or ANOVA followed by Bonferroni's multiple-comparison test. P values ≤ 0.05 were considered statistically significant.

Acknowledgments

The authors would like to thank Eric Nuermberger (Johns Hopkins Hospitals) for assistance with antibiotic dosing, Mariah Klunk for assistance with PET/CT imaging, and Axia Chemicals Ltd. for developing the precursor for ¹⁸F-pretomanid. We also want to thank all the study subjects as well as Amanda Henderson, Ergi Shapiro and Jeff Leal (Johns Hopkins Hospitals) for help with recruiting the human subjects and curating the human imaging data respectively. Funding: This work was funded by the US National Institutes of Health R01-AI145435-A1, R01-AI153349, and R01-HL131829, and NIAID K08-AI139371.

Author contributions

FM, CAR-B and SKJ conceptualized and designed the studies. FM and PDJ performed the manual radiosyntheses. DPH and RFD developed the cGMP synthesis. FM, CAR-B, PDJ, MB and KF performed the mouse studies. XC performed the cytokine analysis. FM, CE, JK and EWT performed the rabbit studies. EWT supervised the rabbit studies. AAO and SKJ wrote the protocol for the human studies. AAO, CAR-B, EWT and MKB recruited and consented the human subjects and analyzed the human imaging data. CAP performed mass spectrometry analysis. SKJ provided funding and supervised the project. FM, CAR-B, EWT and SKJ wrote the manuscript with substantial input from all co-authors

Competing interests: All authors declare that they have no competing interests

Data and materials availability: All data are available in the main text or the supplementary materials.

1. WHO. Global tuberculosis report. Geneva: World Health Organization (2021).

2. Jain SK, et al. Tuberculous meningitis: a roadmap for advancing basic and translational research. Nat Immunol 19, 521-525 (2018).

3. Wilkinson RJ, et al. Tuberculous meningitis. Nat Rev Neurol 13, 581-598 (2017).

4. Chiang SS, et al. Treatment outcomes of childhood tuberculous meningitis: a systematic review and meta-analysis. Lancet Infect Dis 14, 947-957 (2014).

5. Thwaites GE, et al. Effect of antituberculosis drug resistance on response to treatment and outcome in adults with tuberculous meningitis. J Infect Dis 192, 79-88 (2005).

6. Senbayrak S, et al. Antituberculosis drug resistance patterns in adults with tuberculous meningitis: results of haydarpasa-iv study. Ann Clin Microbiol Antimicrob 14, 47 (2015).

7. Vinnard C, et al. Long-term Mortality of Patients With Tuberculous Meningitis in New York City: A Cohort Study. Clin Infect Dis 64, 401-407 (2017).

8. Heemskerk AD, et al. Clinical Outcomes of Patients With Drug-Resistant Tuberculous Meningitis Treated With an Intensified Antituberculosis Regimen. Clin Infect Dis 65, 20-28 (2017).

9. Evans EE, et al. Long term outcomes of patients with tuberculous meningitis: The impact of drug resistance. PLoS One 17, e0270201 (2022).

10. Keam SJ. Pretomanid: First Approval. Drugs 79, 1797-1803 (2019).

11. Stover CK, et al. A small-molecule nitroimidazopyran drug candidate for the treatment of tuberculosis. Nature 405, 962-966 (2000).

12. Singh R, et al. PA-824 kills nonreplicating Mycobacterium tuberculosis by intracellular NO release. Science 322, 1392-1395 (2008).

13. Dogra M, et al. Comparative bioactivation of the novel anti-tuberculosis agent PA-824 in Mycobacteria and a subcellular fraction of human liver. Br J Pharmacol 162, 226-236 (2011).

14. Lenaerts AJ, et al. Preclinical testing of the nitroimidazopyran PA-824 for activity against Mycobacterium tuberculosis in a series of in vitro and in vivo models. Antimicrob Agents Chemother 49, 2294-2301 (2005).

15. Muller M, dela Pena A, Derendorf H. Issues in pharmacokinetics and pharmacodynamics of anti-infective agents: distribution in tissue. Antimicrob Agents Chemother 48, 1441-1453 (2004).

16. Ordonez AA, et al. Dynamic imaging in patients with tuberculosis reveals heterogeneous drug exposures in pulmonary lesions. Nat Med 26, 529-534 (2020).

17. Tucker EW, et al. Noninvasive (11)C-rifampin positron emission tomography reveals drug biodistribution in tuberculous meningitis. Sci Transl Med 10, eaau0965 (2018).

18. Mota F, et al. Radiosynthesis and Biodistribution of (18)F-Linezolid in Mycobacterium tuberculosis-Infected Mice Using Positron Emission Tomography. ACS Infect Dis 6, 916-921 (2020).

19. Tucker EW, et al. Microglia activation in a pediatric rabbit model of tuberculous meningitis. Dis Model Mech 9, 1497-1506 (2016).

20. Ruiz-Bedoya CA, et al. High-dose rifampin improves bactericidal activity without increased intracerebral inflammation in animal models of tuberculous meningitis. J Clin Invest 132, (2022).

21. Khotavivattana T, et al. 18F‐Labeling of Aryl‐SCF3,‐OCF3 and‐OCHF2 with [18F] Fluoride. Angewandte Chemie 127, 10129-10133 (2015).

22. Ahmad Z, et al. PA-824 exhibits time-dependent activity in a murine model of tuberculosis. Antimicrob Agents Chemother 55, 239-245 (2011).

23. European Medicines Agency. Pretomanid FGK, Assessment report. European Medicines Agency (2020).

24. Radiosynthesis and Biodistribution of 18 F-Linezolid in Mycobacterium tuberculosis-Infected Mice Using Positron Emission Tomography. ACS Infect Dis (2020).

25. Ordonez AA, et al. Radiosynthesis and PET Bioimaging of (76)Br-Bedaquiline in a Murine Model of Tuberculosis. ACS Infect Dis 5, 1996-2002 (2019).

26. Thwaites GE, et al. Dexamethasone for the treatment of tuberculous meningitis in adolescents and adults. N Engl J Med 351, 1741-1751 (2004).

27. WHO. Ten threats to global health (2019).

28. Ordonez AA, et al. Visualizing the dynamics of tuberculosis pathology using molecular imaging. J Clin Invest 131, (2021).

29. Bratkowska D, et al. Determination of the antitubercular drug PA-824 in rat plasma, lung and brain tissues by liquid chromatography tandem mass spectrometry: application to a pharmacokinetic study. J Chromatogr B Analyt Technol Biomed Life Sci 988, 187-194 (2015).

30. Bjarnsholt T, Whiteley M, Rumbaugh KP, Stewart PS, Jensen PO, Frimodt-Moller N. The importance of understanding the infectious microenvironment. Lancet Infect Dis 22, e88-e92 (2022).

31. Xu J, et al. Contribution of Pretomanid to Novel Regimens Containing Bedaquiline with either Linezolid or Moxifloxacin and Pyrazinamide in Murine Models of Tuberculosis. Antimicrob Agents Chemother 63, (2019).

32. Beer R, et al. Pharmacokinetics of intravenous linezolid in cerebrospinal fluid and plasma in neurointensive care patients with staphylococcal ventriculitis associated with external ventricular drains. Antimicrobial agents and chemotherapy 51, 379-382 (2007).

33. Akkerman OW, et al. Pharmacokinetics of Bedaquiline in Cerebrospinal Fluid and Serum in Multidrug-Resistant Tuberculous Meningitis. Clin Infect Dis 62, 523-524 (2016).

34. Upton CM, Steele CI, Maartens G, Diacon AH, Wiesner L, Dooley KE. Pharmacokinetics of bedaquiline in cerebrospinal fluid (CSF) in patients with pulmonary tuberculosis (TB). J Antimicrob Chemother 77, 1720-1724 (2022).

35. Sun F, et al. Linezolid manifests a rapid and dramatic therapeutic effect for patients with life-threatening tuberculous meningitis. Antimicrob Agents Chemother 58, 6297-6301 (2014).

36. Tucker EW, et al. Delamanid Central Nervous System Pharmacokinetics in Tuberculous Meningitis in Rabbits and Humans. Antimicrob Agents Chemother 63, (2019).

37. Lappin G, Noveck R, Burt T. Microdosing and drug development: past, present and future. Expert Opin Drug Metab Toxicol 9, 817-834 (2013).

38. Burt T, et al. Phase 0/microdosing approaches: time for mainstream application in drug development? Nat Rev Drug Discov 19, 801-818 (2020).

39. Weinstein EA, et al. Noninvasive determination of 2-[18F]-fluoroisonicotinic acid hydrazide pharmacokinetics by positron emission tomography in Mycobacterium tuberculosis-infected mice. Antimicrob Agents Chemother 56, 6284-6290 (2012).

40. Rosenthal IM, et al. Dose-ranging comparison of rifampin and rifapentine in two pathologically distinct murine models of tuberculosis. Antimicrob Agents Chemother 56, 4331-4340 (2012).

41. Ordonez AA, et al. Radioiodinated DPA-713 imaging correlates with bactericidal activity of tuberculosis treatments in mice. Antimicrob Agents Chemother 59, 642-649 (2015).

42. Irwin SM, et al. Bedaquiline and Pyrazinamide Treatment Responses Are Affected by Pulmonary Lesion Heterogeneity in Mycobacterium tuberculosis Infected C3HeB/FeJ Mice. ACS Infect Dis 2, 251-267 (2016).

43. Nuermberger E, et al. Combination chemotherapy with the nitroimidazopyran PA-824 and first-line drugs in a murine model of tuberculosis. Antimicrob Agents Chemother 50, 2621-2625 (2006).

44. Bigelow KM, et al. Pharmacodynamic Correlates of Linezolid Activity and Toxicity in Murine Models of Tuberculosis. J Infect Dis 223, 1855-1864 (2021).

45. O'Connell KE, et al. Practical murine hematopathology: a comparative review and implications for research. Comp Med 65, 96-113 (2015).

46. Stancil SL, Mirzayev F, Abdel-Rahman SM. Profiling Pretomanid as a Therapeutic Option for TB Infection: Evidence to Date. Drug Des Dev Ther15, 2815-2830 (2021).

There is NO Competing Interest.

20220825SuppMaterials.docx
Supplementary Materials
20220824Reportingsummary.pdf
Reporting Summary

Download PDF

Journal Publication

published 29 Dec, 2022

Read the published version in Nature Communications →

Version 1

posted

You are reading this latest preprint version

Dynamic 18F-Pretomanid PET imaging reveals excellent brain penetration in animal models of TB meningitis and first-in-human studies

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

Radiosynthesis and characterization of ¹⁸F-pretomanid

¹⁸F-Pretomanid has excellent central nervous system (CNS) penetration

Assessment of the BPaL regimen in the mouse model of TB meningitis

First-in-human ¹⁸F-Pretomanid PET studies

Discussion

Materials And Methods

Study design

Radiosynthesis of ¹⁸F-pretomanid

Animal studies

Human studies

Image analysis

Statistical analysis

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1

Dynamic 18F-Pretomanid PET imaging reveals excellent brain penetration in animal models of TB meningitis and first-in-human studies

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

Radiosynthesis and characterization of 18F-pretomanid

18F-Pretomanid has excellent central nervous system (CNS) penetration

Assessment of the BPaL regimen in the mouse model of TB meningitis

First-in-human 18F-Pretomanid PET studies

Discussion

Materials And Methods

Study design

Radiosynthesis of 18F-pretomanid

Animal studies

Human studies

Image analysis

Statistical analysis

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1

Radiosynthesis and characterization of ¹⁸F-pretomanid

¹⁸F-Pretomanid has excellent central nervous system (CNS) penetration

First-in-human ¹⁸F-Pretomanid PET studies

Radiosynthesis of ¹⁸F-pretomanid