The blog is a place to post your thoughts and ideas.

A number of Extending the Cure studies have highlighted the changing epidemiology of methicillin-resistant S. aureus (MRSA): first, we noted the emergence of community strains (CA-MRSA) and their move into hospitals, where they added to the existing burden of healthcare-associated strains (HA-MRSA). Then we visualized the prevalence of methicillin resistance stabilizing and even decreasing in the late 2000s.

Most recently, Klein et al found that CA and HA infections show different seasonal and age patterns: the community-associated strains that cause skin infections tend to peak in the summer months, and primarily affect younger populations. Hospital strains, particularly pneumonia, have a winter peak and affect older populations. 

What has precipitated the emergence of CA-MRSA? How can we reconcile seasonal fluctuations in hospital and community MRSA infections with the seasonality of antibiotic use? Understanding the dynamics of the two strains can help us attribute the success of interventions and predict the future trajectory of Staph resistance.

The continuing co-existence of the two MRSA phenotypes is puzzling from an ecological standpoint. Resistance comes as a trade-off for bacteria. While it grants their survival against antibiotics, it also imposes a “fitness cost”: cells limit their ability to reproduce and infect new hosts, as they have to spare extra resources to ward off drugs. CA-MRSA is resistant to fewer drugs, so it has a much lower fitness cost. This makes community strains more infectious, and, in theory, they should eventually replace the less fit HA strain in the human population.

On the other hand, the distinct seasonal and age pattern and the very names of HA and CA-MRSA infections suggest the two strains may not be competing for the same turf. Hospitalized subpopulations may be thought of as their own ecological niche, a refuge where HA-MRSA’s reduced fitness cost is not a competitive disadvantage, given the strain’s ability to survive the more aggressive antibiotics prescribed to hospital patients. But that still would not explain the lasting presence of community strains in hospitals and vice versa. To complicate the picture, there is a constant flow of individuals between hospitals and surrounding communities. 

How is co-existence between the two strains maintained, given the highly variable rates of antibiotic usage and healthcare exposure across the US population? A recent paper by Kouyos, Klein et al in PLoS Pathogens offers a fascinating theoretical framework to address this question. The model describes how community and hospital populations transition from being susceptible (i.e. non-infected) to carrying (i.e. being colonized or infected with) either strain of MRSA. Equations are calibrated with empirical data reflecting US population age structure, hospitalization rates, community and hospital antibiotic usage, and other variables. 

Comparing several models of increasing complexity, the authors note that “heterogeneities increase the range of co-existence”, i.e., outcomes where community and hospital strains of MRSA co-exist over the long run become more plausible once we acknowledge that:

  • Older people spend more time in hospitals, where HA-MRSA fares better, and have fewer contacts in the community, making the competitive disadvantage of the HA-MRSA strain matter less for that age group. 
  • Hospital patients are more likely to receive more powerful antibiotics, and thus select for the HA-MRSA strain in that environment.  


But even if theoretically plausible, how long will it take to reach this equilibrium? The constructed models become stable in unrealistic time frames (>50 years), suggesting that the emergence of CA-MRSA in the 1990s and the current co-existence may be just a transient phase. Furthermore, co-existence becomes even less plausible when the model takes into account competition between the two MRSA strains and methicillin-susceptible S aureus. 

All of these kinks suggest future models should account for additional heterogeneities, such as spatial variations, long-term care populations, the evolution of new strains, demographic changes, etc. However, the study still offers valuable insights: in all models, decreasing the transmissibility of HA-MRSA strain, for instance, through enhanced infection control, leads to its eventual replacement with CA-MRSA. Such an outcome would have profound effects on further infection control measures: CA-MRSA is actually a more virulent and transmissible pathogen and, in authors’ words, “prevention efforts that focus currently on hand-hygiene among health-care workers could lose their effectiveness in reducing the spread of MRSA”.