Local changes in rates of group A Streptococcus disease and antibiotic resistance are associated with geographically widespread strain turnover events

• Published in: Virulence Vol. 1; no. 4; pp. 247 - 253
• Main Authors: Metzgar, David, McDonough, Erin A, Hansen, Christian J, Blaesing, Carl R, Baynes, Darcie, Hawksworth, Anthony W, Blair, Patrick J, Faix, Dennis J, Russell, Kevin L
• Format: Journal Article
• Language: English
• Published: United States 01.07.2010
• Subjects: Anti-Bacterial Agents - pharmacology; Anti-Bacterial Agents - therapeutic use; Binding; Biology; Bioscience; Calcium; Cancer; Cell; Cycle; Drug Resistance, Bacterial; Humans; Landes; Macrolides - pharmacology; Microbial Sensitivity Tests; Military Facilities - statistics & numerical data; Military Personnel; Molecular Epidemiology; Organogenesis; Pharyngitis - epidemiology; Pharyngitis - microbiology; Proteins; Species Specificity; Streptococcal Infections - epidemiology; Streptococcal Infections - microbiology; Streptococcus pyogenes - classification; Streptococcus pyogenes - drug effects; Streptococcus pyogenes - genetics; Streptococcus pyogenes - pathogenicity; Virulence
• ISSN: 2150-5594; 2150-5608
• DOI: 10.4161/viru.1.4.11979

This study addresses the effects of dynamic strain turnover and antibiotic prophylaxis on rates of group A Streptococcus (GAS) antibiotic resistance and disease. The authors analyzed the strain distributions, disease rates, and patterns of antibiotic resistance of 802 GAS isolates collected from 2002 through 2007. These samples were collected from patients with GAS infection symptoms at 10 military facilities. Macrolide resistance peaked at 25% during 2004, due to the geographically widespread dominance of a single resistant strain (M75). The resistant strain was not retained regardless of local patterns of macrolide use, and resistance rates decreased upon replacement of M75 with macrolide-susceptible strains. Disease rates were similarly correlated with dominance of specific M types. Statistical analysis revealed temporal correlations between strain distributions at multiple locations. Only the most common strains yielded enough data at multiple sites for statistically significant comparison of temporal fluctuations in dominance, but these (including M44, M3, M18, M118, and M6) all yielded highly significant temporal correlations of 90% or greater on yearly scales. As expected given the complexity and variability of strain distributions on shorter time scales, analysis on a monthly scale yielded lower degrees of positive correlation (31-62%), but in this case all significant correlations were still positive. Shifts in antibiotic resistance profiles and disease rates at specific sites appear to be associated with strain replacements happening on larger scales, independent of antibiotic use at individual sites.