The rise and rise of microbial resistance to currently available antibiotics is driving the development of new generation anti-infective products. John Carter reports.
Unlike high profile challenges such as the fight against cancer, where breakthroughs arising from the human genome project can be expected in coming years, antibiotic resistance is a moving target driven to a great extent by our own mismanagement of antibiotic therapies and misuse in agriculture. The notorious methicillin resistant Staphylococcus aureus (MRSA), scourge of post-operative environs, is increasingly showing resistance to asuper antibiotics' such as vancomycin (VISA and GISA strains) and indeed has been joined by vancomycin resistant Enterococci (VRE) and multiple resistant Streptococcus pneumoniae (MRSP) to name but a few.
European guidelines for the clinical evaluation of anti-infective drug products (Beam et al., 1993(1)) specifically place experimental evaluation of new compounds in animal models as pre-requisites to clinical trials. Determination of efficacy in vivo, combination therapies and synergies, possible drug-host-microbe interactions, pharmacodynamics and metabolism may be served by such models. While the limitations of extrapolation from such studies are recognised, the use of animals as models of disease infection has become extensive with more than 1000 different models described.
Depending on their nature, purpose and predictive value Zak et al. 1999(2) assigned these to various categories, each having justification at some point in the development and therapeutic strategy for an anti-infective product. For example, development of the novel anti-infective Daptomycin, a lipopeptide antibiotic showing significant activity against vancomycin resistant MRSA and VRE, has been facilitated by the use of animal disease models. In particular, the neutropenic mouse S.aureus thigh model used by Louie et al., 2001(3) to evaluate the pharmacodynamics and dose range activity of daptomycin in vivo, provided data for establishing therapeutic doses in human clinical trials.
Drug discovery
The pre-clinical development of new anti-infective agents is a costly business. At Huntingdon Life Sciences, in vitro and in vivo PLOT(s) (Pre-clinical lead optimization technologies) are increasingly being utilised to focus the drug discovery pipeline. Sources of novel anti-infectives include combinatorial libraries, biopharmaceuticals, natural products, short peptides and proteins. For all candidates, in vitro primary screening using simple agar diffusion techniques against standard laboratory strains is performed. Selection of active agents for definitive Minimum Inhibitory Concentration (MIC) tests against clinically relevant bacterial/fungal pathogens utilises 96-well micro-titre plate broth-dilution techniques to minimise compound usage and enable rapid screening against large numbers of clinical isolates.
A comprehensive range of microbiological bioassays has been utilised for many years at HLS to support antibiotic bioavailability and pharmacokinetic phases of veterinary and human clinical trials. However, it has not been until recently that HLS has addressed the important in vitro: in vivo anti-infective interface by establishing a range of commonly used anti-infective animal disease models to bridge the gap between discovery of in vitro activity and demonstration of clinical efficacy, essential to continue the development pathway to new drug approval.
Preliminary infection studies
Infectious disease agents are sourced from the UK Public Health Laboratory Service or UK Veterinary Investigations Centres as appropriate or may be supplied by Sponsors. Prior to the in vivo disease model study, in vitro MIC studies are performed against the infectious agent selected for the disease model to confirm that the anti-infective product shows effective activity in vitro. Pre-liminary in vivo studies to determine optimal infectious concentrations and dose regime of selected bacterial/fungal pathogens are usually performed over a range of five challenge concentrations and are designed to produce acute systemic infection within 24 48 hours or topical infection (eg wound healing models), sufficient to evaluate the potential anti-infective aprotective' efficacy of the candidate product.
Mouse peritonitis model
Progress through the drug discovery pathway will influence model selection. The first step into in vivo testing can be served by a basic mouse peritonitis model. In a working example of a validated Streptococcus pneumoniae mouse peritonitis model, HLS has successfully demonstrated dose-related efficacy of candidate anti-infectives administered by oral and subcutaneous routes (Fig. 1) in comparison with a commercially available reference antibiotic, clarithromycin. Protective efficacy was measured in terms of survival to the end of a seven-day observation period post intra-peritoneal injection on Day 1 with 4 x 105 colony forming units Streptococcus pneumoniae prepared in 5 per cent mucin saline beef broth. Fig. 1 shows that mice treated with the candidate anti-infective products at selected dose/concentrations showed no adverse clinical signs over the study period in contrast to the low degree of protective efficacy shown by clarithromycin in this model. Pathogens such as Pseudomonas aeruginosa and Candida albicans may be employed in this model, however our experience has shown that an alternative intravenous infection model is more appropriate for evaluation of anti-infective efficacy against systemic infection with MRSA.
Murine thigh infection model
Models such as the amurine thigh infection model' can offer greater versatility by providing sensitive and reproducible experimental infections permitting pharmacokinetic and dose range evaluations. Commonly, this model employs immuno-compromised mice pre-treated with cyclophosphamide to induce neutropenia. Transient depletion of neutrophil and leukocyte numbers permits the study of direct interactions between the pathogenic micro-organism and the anti-infective agent in the absence of a competent immune response. Non-neutropenic mice may also be used. Determination of viable bacterial counts in infected/target tissues is a universal parameter associated with this model type and this quantitative procedure is routinely used at HLS to plot infection processes in similar models of infection.
Efficacy of anti-infectives as topical applications are evaluated at HLS using a rat cutaneous wound-healing model. Wound infections using MRSA are established and then wound-healing potential of topical products is assessed using visual parameters and by recovery and enumeration of MRSA from the infected wound site using direct swabbing and surface rinse culture techniques.
A wider perspective
HLS has established expertise in the wider realm of disease modelling to keep pace with the growing range of small molecule biotech therapies arising from anti-tumour research, the increasing interest in cellular therapies based on stem cell research and the availability of transgenic animals for use in disease models.
* Anti-tumour and Xenograft models Evaluation of anti-tumour compounds in vivo using human tumour cells in immune deficient nude mice is carried out for single tumour line xenografts and in the near future, for multiple tumour line assessment (hollow fibre technology). Research is also ongoing with xenografts in surgical models of disease such as renal insufficiency and other organ failures.
* Cellular therapy models Cellular therapies for diabetes, haemopoesis and Parkinson's for example, cannot readily be tested for efficacy in standard animal models due to natural immunogenic interference. Research is ongoing to transfer such models into immuno-deficient or immuno-suppressed animals. Currently available models include Streptozocin-induced diabetes, MPTP-induced Parkinson's and Chemical depletion of bone marrow.
* Transgenic/Knockout mouse models New developments in the generation of transgenic and knock-out mice have led to the availability of a range of genetically defined disease models. HLS has commercial experience with the p53 model and participated in the ILSI project working with p53, the results from which provided the basis of EMEA(4) acceptance of the p53 model and TgrasH2 model for regulatory carcinogenicity assessment.
Expansion of disease modelling expertise at HLS in the fields of anti-infectives, anti-tumours and modelling of cellular based therapies is continually evolving to help meet the modern challenges of drug discovery and development faced by the pharmaceutical industry today.
Enquiry No 24
John Carter is with Huntingdon Life Sciences, Huntingdon, UK. www.huntingdon.com
