Seasonal, Geographic, and Temporal Trends of emm Clusters Associated With Invasive Group A Streptococcal Infections in US Multistate Surveillance

PR Smeesters, D Laho, B Beall, AC Steer… - Clinical Infectious …, 2017 - academic.oup.com
PR Smeesters, D Laho, B Beall, AC Steer, CA Van Beneden
Clinical Infectious Diseases, 2017academic.oup.com
To the Editor—Previous reports have demonstrated that 90% of invasive group A
Streptococcus (GAS) strains in the United States represent approximately 25 different emm
types, with most disease occurring in winter and early spring [1–3]. However, little is
published about more complex geographical and temporal trends for these infections. An
emm-cluster system that classifies the numerous GAS emm types [4] into 48 functional emm
clusters sharing structural and binding properties has been established [5]. This clustering …
To the Editor—Previous reports have demonstrated that 90% of invasive group A Streptococcus (GAS) strains in the United States represent approximately 25 different emm types, with most disease occurring in winter and early spring [1–3]. However, little is published about more complex geographical and temporal trends for these infections. An emm-cluster system that classifies the numerous GAS emm types [4] into 48 functional emm clusters sharing structural and binding properties has been established [5]. This clustering scheme complements an emm patterning system devised 20 years ago, which associates specific markers within emm and neighboring related genes with tissue tropisms exhibited by GAS strains [6]. The emm clusters are useful in predicting basic genetic features of GAS isolates because they are markers that correlate with M protein functions and antigens as well as the genetic determinants of other features such extracellular matrix binding proteins [5, 7–9]. The emm cluster typing is easy to use, and cluster type can be predicted from emm typing results [4, 5, 10]. We report emmclusters associated with invasive GAS infections in surveillance areas in 9 US states from 1 January 2003 through 31 December 2013, identified through the Centers for Disease Control and Prevention (CDC)’s Active Bacterial Core surveillance (ABCs), and evaluate overall and state-specific seasonal and temporal distribution of emm clusters. During 11 years of surveillance, we identified 114 different emm types among 10 175 isolates causing invasive GAS disease. An emm cluster was attributed to 10 136 (99.6%) isolates, with a total of 24 different emm clusters represented. Almost all (99.8%) of the cluster-associated isolates belonged to 11 emm clusters (in decreasing order of frequency: E4, AC3, E3, AC4, E6, AC5, E1, E2, D4, M6, M5).
When analyzing the proportions of emm clusters over time, we observed that isolates belonging to AC clusters (AC3, AC4, and AC5) peaked in winter (January to March) while those belonging to E clusters (E1, E2, E3, E4, and E6) were the most common during summer (Figure 1). During winter, AC cluster and E cluster isolates were associated with a similar average number of invasive GAS infections (147 and 145, respectively), with each accounting for 45% of all infections. In contrast, during summer, E cluster isolates were associated with an average of 100 infections (66% of all infections), compared with an average of
Oxford University Press