Molecular Methods Enhance the Detection of Pyoderma-Related Streptococcus pyogenes and emm-Type Distribution in Children

Abstract

Background: Streptococcus pyogenes-related skin infections are increasingly implicated in the development of rheumatic heart disease (RHD) in lower-resource settings, where they are often associated with scabies. The true prevalence of S pyogenes-related pyoderma may be underestimated by bacterial culture.

Methods: A multiplex quantitative polymerase chain reaction (qPCR) assay for S pyogenes, Staphylococcus aureus, and Sarcoptes scabiei was applied to 250 pyoderma swabs from a cross-sectional study of children aged <5 years in The Gambia. Direct PCR-based emm-typing was used to supplement previous whole genome sequencing (WGS) of cultured isolates.

Results: Pyoderma lesions with S pyogenes increased from 51% (127/250) using culture to 80% (199/250) with qPCR. Compared to qPCR, the sensitivity of culture was 95.4% for S pyogenes (95% confidence interval {CI}, 77.2%-99.9%) in samples with S pyogenes alone (22/250 [9%]), but 59.9% (95% CI, 52.3%-67.2%) for samples with S aureus coinfection (177/250 [71%]). Direct PCR-based emm-typing was successful in 50% (46/92) of cases, identifying 27 emm-types, including 6 not identified by WGS (total 52 emm-types).

Conclusions: Bacterial culture significantly underestimates the burden of S pyogenes in pyoderma, particularly with S aureus coinfection. Molecular methods should be used to enhance the detection of S pyogenes in surveillance studies and clinical trials of preventive measures in RHD-endemic settings.

Keywords: Streptococcus pyogenes; emm-typing; PCR; skin infection; strain diversity.

Figure 1.

Bacterial quantity in polymerase chain reaction–positive samples for Streptococcus pyogenes (A) and Staphylococcus aureus (B) by bacteriological culture status, and for S pyogenes (C) and S aureus (D) in coinfected samples compared to samples with a single identified bacterial pathogen. Statistical differences determined using a 2-tailed Mann Whitney U test.

Figure 2.

Participant infection status by age, as determined by quantitative polymerase chain reaction (qPCR) result. Those with infection status “Negative” had a clinical diagnosis of pyoderma but neither Streptococcus pyogenes nor Staphylococcus aureus was detected by qPCR.

Figure 3.

Bacterial etiology and load of pyoderma cases according to season. Participant infection status by season, as determined by quantitative polymerase chain reaction (qPCR) result, for all study participants (A) and participants with pyoderma (B). Those with infection status “Negative” had a clinical diagnosis of pyoderma but neither Streptococcus pyogenes nor Staphylococcus aureus was detected by qPCR, while those with status “No pyoderma” did not have swabs taken. Sample bacterial quantity for S pyogenes (C) and S aureus (D) in samples taken before the onset of the rainy season compared to during the rainy season. Statistical differences determined using a 2-tailed Mann Whitney U test.

Figure 4.

Distribution of Streptococcus pyogenes emm-types. emm-types from 153 of 199 S pyogenes quantitative polymerase chain reaction (PCR)–positive samples were available. A, emm-types by typing method. Direct PCR-based emm-typing increased both the number of samples of 21 emm-types in the whole genome sequencing (WGS)–defined dataset, as well as detection of 6 emm-types not in the WGS dataset. B, S pyogenes emm-type by geographical cluster, labeled with the cluster number designated during our previous study. Clusters are ordered by detected number of pyoderma cases from high to low. Black boxes and dots in the emm-type legend denote emm-types identified by PCR but not WGS.