The first step in the tuberculosis drug development process is to consider the desired activity profile and properties we are aiming for, as this will determine the direction of the research programme. The 'ultimate' tuberculosis drug would possess both rapid bactericidal activity and sterilizing activity, thus killing all M. tuberculosis populations, and the spectrum of activity would include all multidrug-resistant tuberculosis isolates. The drug must be orally bioavailable, have low toxicity, good tissue distribution to maximize activity against intracellular organisms, a long elimination half-life so that intermittent chemotherapy may be considered and, last but not least, it must be inexpensive to produce given the constraints on pricing that exist. Clearly, it will be difficult to build all these properties into a single molecule and compromises will be made. Critically, any new drug must offer a significant advantage over the drugs already available, such as activity against multidrug-resistant tuberculosis, better tolerability, intermittent dosing or shorter overall therapy duration.
It is most unlikely that M. tuberculosis has a single Achilles' Heel, and that more than one agent will always be needed to bring about a complete cure. The most achievable goal is to generate a novel bactericidal agent. Such an agent may be evaluated by the minimal inhibitory concentration (MIC) for M. tuberculosis growth in culture, and in short-term acute in vivo models of infection. Clinical testing in humans is straightforward, with early bactericidal activity (Mitchison 1996) being a relatively good predictor of final clinical efficacy. On the other hand, disadvantages such as emergence of resistance, fuelled by ineffective or inappropriate combinations, or reservation of the new drug for use in cases where treatment is already failing would make it uneconomic to produce such a drug, unless some other property such as a long half-life would provide an advantage.
Targeting the persisting organisms offers the tantalizing prospect of reducing the duration of therapy. However, specific molecular targets are not so obvious. Significant research effort is required to validate such targets and new in vitro and in vivo models must be developed if such agents are to be fully evaluated. It will be more difficult to optimize the properties of any agent that has no MIC in culture, and surrogate markers of sterilization must be identified. Clinical trials will be lengthy and require large patient numbers, increasing development costs.
We may therefore envisage intervening at several steps in the disease process. The targets may be visualized on a spectrum, ranging from those functions essential for survival of the bacterium in culture through host—pathogen interaction to host immunity (Fig. 2). The effort required to identify and validate targets varies considerably, with proportionally greater effort being expended on targets that have the highest novelty.
Was this article helpful?