Experimental Animal Models of TB: Chemotherapeutics and Imaging

National Institute Of Allergy And Infectious Diseases

Investigators

Linked publications & trials

Abstract

This project encompasses approaches to understand how current anti-tubercular chemotherapy works using the most modern technologies and to develop new and improved therapies and therapeutic approaches. Individual projects within this framework are (1) developing structural and functional imaging techniques using PET/CT for use in live, M. tuberculosis (Mtb) infected animals, (2) development of advanced animal models for predicting drug efficacy under conditions that exactly mimic those experienced by TB patients, (3) understanding the activity of various drugs in animal models of tuberculosis therapy, (4) correlating responses seen in animal models with the pathology and response to therapy observed in human TB, and (5) developing techniques for assessing drug distribution, penetration, and pharmacokinetics in vivo. In 2021 we have developed moderate capacity to image mice in the Mediso PET/CT scanner using a chamber that holds 4 mice per imaging session. Most of our PET/CT studies have used 18F-2-fluoro-2-deoxyglucose (FDG) to image the metabolism of the eukaryotic cells in TB lesions in our animal models of tuberculosis. Yet, we would prefer a small molecule that could be used to label Mtb in vivo specifically and endogenously as a PET radiotracer. We focused on MTb antigen 85 enzymes that are expressed on the exterior of MTbs cell wall and can incorporate exogenous trehalose (a nonmammalian disaccharide consisting of a two 1-1,-linked glucose monomers) as either the mono- or dimycolate in the cell wall. We used these enzymes to chemically incorporate 18F trehalose (FDT) into bacteria in the lesions of infected rabbits and marmosets. FDT PET/CT scans seems to accurately reflect low and high bacterial burden in marmoset lesions assayed for bacterial load. Use of FDT in the imaging of Mtb in diverse models, including non-human primates, successfully co-opts Mtb-specific processing of trehalose to allow the specific imaging of TB-associated lesions and to monitor the effects of treatment in marmosets. This is a promising sign that the FDT will be able to give an earlier indication of treatment success or failure compared to FDG. We have continued developing a non-human primate (NHP) model for tuberculosis - the common marmoset. In the past we explored if the marmoset model accurately reflects the response to treatment by providing standard TB treatment (RIFR, INH, PZAZ, and EMBE) to infected symptomatic marmosets and demonstrated that marmosets show similar treatment results as humans. As a counterpart to an early bactericidal activity (EBA) and paired PET/CT clinical trial, NexGen EBA conducted in South Africa; NCT02371681, in 2018-2020 we have replicated the treatment groups and observations in randomized Mtb infected marmosets. In the study, the standard regimen was deconstructed and each drug was administered by itself or in pair-wise combinations to measure the effect of the drugs on the microbiological and radiographic markers as well as two 4 drug regimens HRZE and MRZE where M is moxifloxacin. We were looking for unique drug signatures in the radiologic features of the animals on treatment and comparing those to the histological presentation of the lesions upon necropsy. As a first assessment in 2021, there appeared to differences in treatment response between the MRZE and HRZE groups based on the simple disease volumetric analysis in both the CT and PET data. When the PET/CT image scoring system was applied, we found that the pathology in HRZE-treated marmosets resolved significantly faster but the reduction in uptake of FDG was not different. The increased resolution rate led to a significantly different volume of disease in the HRZE vs. MRZE groups after 2 months of treatment. We also observed synergy between PZA and INH reduction in CT volume and bacterial loads in both the humans and the marmosets. We hypothesize that understanding the specific contributions of each drug to the disease resolution will assist in the pairing of future agents into more successful and rapidly acting regimens. Follow-on EBA studies with PET/CTs of these new agents are in progress and we have marmoset data to use with a subset of them now and will be adding to dataset in 2022. We continue to study the oxazolidinone (OXA) antibiotics such as linezolid which have shown significant therapeutic effects in patients with extensively drug-resistant (XDR) TB despite modest effects in rodents and no EBA in human phase 2 trials. These new OXAs have different activities in the marmoset TB model that appear to be related to lesion type and physical distribution into the lesions. Together with the Gates Foundation's TB Drug Accelerator scientists we have engaged in developing novel OXAs that are TB-selective and less toxic than linezolid (LIN). Throughout 2021, we have been analyzing the PET/CT imaging data from marmosets from treated with other classes of antibiotics from partners in the TBDA program including diarylquinolines, quinolines, imidazopyridines, nitroimidazoles, among others. These classes of antibiotics are being explored as composing new regimens for treatment of MTB and understanding the specific contribution of each one to activity including consideration of spatial distribution and the kinetics of accumulation in lesions to avoid temporal and spatial black holes of monotherapy. With each new drug candidate, we test for in vivo efficacy with our academic and industry partners, we continue to assess the candidates penetration into granulomas and cavities in our model animals to correlate the information with any observed efficacy. Through samples collected in 2019-20, we now know that the diarylquinoline Bedaquiline (BDQ) accumulates significantly in both marmoset and rabbit TB lesions at steady state, including in cavity caseum, perhaps explaining its outstanding efficacy in hard to treat patients with XDR TB. In 2021, we tested a lower more human like dose of BDQ so that we could assist in predicting the activity of BDQ including regimens including the NIX regimen that includes the Pertominid (PA), Lin, and BDQ. The 3 drug regimen seems like it will be similar to HRZE in activity in the marmoset but not all the data are available at this time. We continue to explore host-directed therapy (HDT) as a method to increase drug efficacy by increasing agent delivery to the site of infection in the rabbit model of Mtb. We have been performing a series of experiments to determine if treatment with an agent that promotes normalization of blood vessel structure such that hypoxia is decreased and drug penetration increased could improve drug access to the lesion. In previous results, BDQ penetration is increased in lesions at steady state in both control and HDT-treated rabbits, so these experiments have been conducted again with moxifloxacin which does not accumulate at steady state in the rabbit model with results monitored by CT imaging, lesion histology, drug quantification and bacterial load. The host RNA has been isolated to determine what signaling is changed during the HDT treatment and if there is a change in the granuloma expression of matrix components as predicted.