Objectives

The ultimate goal is to generate clear and concise recommendations on in vivo and in vitro protocols/assays – in order to accelerate TB drug discovery and development

This two year proposal has four objectives: [1] Collect protocols, both historical and currently in use by our peers; [2] conduct comprehensive analyses of both in vitro and in vivo TB drug testing methods, leading to recommendations for the most robust, reliable and reproducible assays and the subsequent application of the best methods to test the efficacy and sterilizing activity of new drug regimens; [3] establish provision of guidance [advisory panel], as well as consensus on major recommendations by means of a Workshop; [4] dissemination of information by publications and by use of an easily accessed Web page.

Several years ago some basic recommendations were made in the "Blueprint" document issued (Grosset J, Lounis N, Cynamon MH and Orme IM. Animal models for testing new anti-tuberculosis drugs. Scientific Blueprint for Global Alliance, 2001), however, no serious attempts at assay optimization have been made since then, and for that matter the number of laboratories using such assays, especially in vivo protocols, has been relatively low. This is a timely moment therefore at which to consider developing the best assays, models and regimens. Tuberculosis continues as a global pandemic (www.who.int). New drugs in the pipeline give us a renewed sense of optimism for the future treatment of TB. There are six new compounds in various stages of clinical evaluation, and another eight at the lead optimization stage, according to the survey by the Working Group for TB Drugs (http://www.stoptb.org/wg/new_drugs/).

Objective 1

The goal is to retrieve and analyze historical and current methods from all interested parties and evaluate their potential usefulness. The starting point for the proposed studies is to collect protocols for testing TB drugs in animals. These go back to the late 1940’s, so this will involve library and database searches. The objective is to compare these historic methods with our modern day knowledge. In addition, we will contact the appropriate participants who attended the 2007 Gordon Conference as well as others invovled in preclinical testing; we will send them a comprehensive survey regarding their in-house in vitro and in vivo assays as well as questions regarding their willingness to participate in the effort.

We will collect a comprehensive data set on the following factors:

• In vitro studies for TB drug testing
  • Organism
  • Antimicrobial Agent
  • Experimental protocol design
  • Culture conditions
• In vivo animal studies for TB drug testing
  • Animal
  • Drug
  • Experimental protocol
  • TB strain

Timeline

1. Historical protocols: collecting of all pertinent protocols in the first months of the award. An analysis of merits/disadvantages of each will be done by early Spring 2008.
2. Literature review of contemporary protocols by early Spring 2008. An analysis and final synthesis of current protocols used by our peers will be done by Summer 2008.
3. Survey: Goals: Electronically prepared survey of questions will be disseminated Fall 2007. The results will be compiled by early Spring 2008.
4. Site visits to academic and industry laboratories will take place mid Spring 2008.
5. Summary of all assays and initial analysis will be compiled by Late Summer 2008.

Objective 2

We will begin head-to-head studies in specific assay conditions and in studies in M. tuberculosis infected animal models in which assays can be compared directly. A first set of in vivo experiments will commence at day one of the Project. A side by side evaluation is scheduled of the most widely used mouse models. Three models use aerosol infection, and the other intravenous inoculation (i.v.). In the i.v. model a high dose inoculum of M. tuberculosis is given intravenously, followed by initiation of therapy and monitoring changes in the bacterial load in the lungs. This will study whether the physiology of the model, from the actual site of deposition via the capilliary bed to the ensuing rapid generation of high levels of systemic immunity because of the inoculum size, may be different to human disease. The other models involve a more realistic low dose and a higher dose aerosol infection model. In the low dose aerosol model, peak bacterial loads are lower, and the emergence of immunity is slower. Because bacteria enter via the bronchial tree into the alveolar regions, sites of subsequent persistence and dissemination into lymphatic vessels may be more realistic. This difference in establishment of a memory immunity response may become important when studying the sterilizing properties of the compound and relapse of infection, but this important factor has never actually been extensively studied in drug testing. It may be that both models may turn out to be equally valuable.

We have consulted with our CSU statistician regarding sufficient statistical power to verify that the numbers of animals proposed is adequate; we will use groups of six animals per drug group and 10 mice to monitor relapse of infection. Our first study (STUDY A) will compare the four most widely used models head to head:

In all studies, mice will be purchased from Charles River Laboratories (Wilmington, MA) and acclimated for at least two weeks after arrival. Aerosol experiments will consist of infection with ~100 or ~5000 M. tuberculosis (strain Erdman was grown as pellicle to maintain virulence, using the Trudeau protocol) using a GlasCol aerosolizer (Terre Haute, IN). For intravenous infection we will inject 10⁷ bacteria into a lateral tail vein.

STUDY B in preparation.

Timeline

Both Study A and B will be completed by Late Fall 2008. Additional in vivo and in vitro studies will be planned based on input from Workshops and Advisory Panel in the second year of the award.

Objective 3

At the end of year 1 an obvious step will be to hold a Workshop to review the findings and seek recommendations for further studies in Year 2. The final objective of the program is to publish and disseminate our assessment of the relative merits as well as disadvantages of the assays and models of TB drug development. Ultimately, we will provide a series of recommendations for the standardization of in vitro and in vivo models and methods, but there is no reason why we cannot at least begin to draft this document as a result of this proposed workshop. This will be disseminated to our peers, as well as appearing on our Web-site, inviting peer criticism. We should reiterate here that it is our intention to deliver a flexible document highlighting the advantages of what we feel represents the best approaches as well as describing shortcomings of other methods.

The purpose of the draft recommendations will be to spur further debate regarding assay merits and pitfalls, and thus is timely and important. A plethora of different assay systems and conditions are used to study the effect of drug candidates on the growth of M. tuberculosis, making it difficult to compare data from one laboratory to another. Even simple procedures such as determining an MIC can include variables such as the particular M. tuberculosis strain that is used and whether results are presented in µg/ml or µM concentrations. Animal studies have similar problems in terms of variables; we need to delineate these, determine which of these are biologically significant, and converge on a rationally derived set of optimized assays for evaluating compounds.

Advisory Committee

At the beginning of the proposed program we will establish an advisory committee, selecting about five individuals with knowledge in the field and enthusiasm. Animal modelists are needed, and at least one person from the "industry" and one directly involved in clinical trials. A central role of this committee, in addition to general guidance, will be to ensure that the eventual recommendations are objective and conducted fairly.

Objective 4

Disseminate gathered information by means of publications and by use of this Webpage.