Infectious Disease Experts Propose Research Priorities to Refine Antibiotic Therapy

Antibiotic therapy could more precisely target pathogens without provoking resistance and be guided by more useful parameters than minimal inhibitory concentrations (MIC) if particular areas of research are prioritized, in the view of a work group drawn fromseveral infectious disease associations.

In a "Personal View" paper in The Lancet Microbe, work group members propose six priority areas of research which could, they write, "enhance our understanding of antibiotic pharmacology and facilitate more accurate predictions of treatment success..."

Representatives from the International Society of Anti-Infective Pharmacology (ISAP), and the European Society of Clinical Microbiology and Infectious Diseases (ESCMID), and from the Pharmacology Working Group of the International Societyof Antimicrobial Chemotherapy (ISAC) began a discussion on how their organizations could support and further antibiotic pharmacokinetic/pharmacodynamic (PK/PD) research, at the 2019 European Congress of Clinical Microbiology and Infectious Diseases in Amsterdam, Netherlands.

"We all proposed ideas that we considered to be critically important to the field moving forward, and these were consolidated to the 6 research priorities," lead author Zackery Bulman, PharmD, Department of Pharmacy Practice, University of Illinois Chicago, Chicago, IL, told Contagion.

Their priority areas for future antibiotic research are: antibiotic-pathogen interactions; targets for antibiotic combination therapies; models for the time-course of treatment response; models of host response to infection; personalized medicine through therapeutic drug management; and application of these in support of developing novel therapies.

Bulman and colleagues summarize these 6 priority areas:

• Define the pathogen's global response to antibiotic therapy (monotherapy or combination) to develop treatment regimens that provide optimal bacterial killing and prevent the amplification of resistance
• Generate new approaches and exposure targets for antibiotic combinations and better integrate them into the drug development process to potentiate their clinical utility
• Develop novel mathematical models that can rationally optimize antibiotic therapy by integrating new mechanistic findings and patient specific pharmacokinetics to predict the time-course of treatment response
• Characterize and model the complex interactions between the host immune system and pathogen to enhance our ability to successfully predict individual patient outcomes and clinical trial results
• Use integrative data analysis to link and translate mechanistic systems-level information to the clinical setting, allowing for personalization of antibiotic therapies
• Support clinical development and approval new and non-traditional therapeutics through innovative partnerships and the generation of novel laboratory and computational models.

In the area of antibiotic-pathogen interaction, Bulman and colleagues point out that very little is known about how bacterial cells respond to antibiotics at the level of gene regulation, expression, and metabolic perturbations. They advocate for integrative analysis of bacterial genomics, transciptomics, proteomics, and metabolomics data following antibiotic exposure to more fully elucidate cellular response to antibiotic exposure.

Predicting and achieving synergy with combinations of antibiotics have been hampered by several factors, according to the authors, including research models that have applied metrics from each antibiotic used separately, or not accounted for different pharmacodynamics of each antibiotic at a particular site of infection.

"Future research on development of optimal combinations to overcome antibiotic resistance should integrate the PKPD of multiple agents, mechanisms of synergy and resistance, and computational models," Bulman and colleagues recommend.

The authors call for greater emphasis on mechanism-based mathematical models to identify optimal time-course of treatment response, rather than rely only on traditional PKPD approaches. These models should account for the time-course of bacterial growth, killing, and resistance emergence, they indicate. They also advocate for increased investigations of specific patient populations with altered drug clearance, volume of distribution, or protein binding.

Studying host-related variables is as complicated as the components and dynamics of the intact, in-vivo immune system, and Bulman and colleagues propose that research into antibiotic-pathogen interaction go beyond the controlled conditions of in-vitro and murine models. They envision computational modeling that integrates data from multiple sources and accounts for time-related interactions between biomarkers to help identify how the immune system contributes to successful outcomes.

Personalized medicine should extend to therapeutic drug monitoring of antibiotics, and the authors call for developing more precise tools than comparison of antibiotic concentrations to an established therapeutic range.

"This process is imprecise and slow, as steady state must be reached before an individual dose can be derived, and, thus, substandard for antibiotics when optimal treatment should be achieved as early as possible," Bulman and colleagues observe.

They suggest that dvances in software should enable model-informed precision dosing (MIPD) that incorporates a range of metrics, such as time above MIC, and area under the curve (AUC) calculations of antibiotic concentration over time, without the need to wait for steady state.

Bulman commented that although procedures and methodologies for clinical trials were not highlighted in the 6 priority areas, well designed trials will be required to validate several of those areas.

"For example, new combination therapies that are rationally optimized in vitro and with the aid of in silico models will need to go through clinical studies prior to widespread implementation," Bulman said.