Coxsackievirus A6 Recombinant Subclades D3/A and D3/H Were Predominant in Hand-Foot-And-Mouth Disease Outbreaks in the Paediatric Population, France, 2010-2018

Abstract

Coxsackievirus A6 (CVA6) emerged as the most common enterovirus of seasonal outbreaks of hand-foot-and-mouth disease (HFMD). We investigated CVA6 genetic diversity among the clinical phenotypes reported in the paediatric population during sentinel surveillance in France between 2010 and 2018. CVA6 infection was confirmed in 981 children (mean age 1.52 years [IQR 1.17-2.72]) of whom 564 (58%) were males. Atypical HFMD was reported in 705 (72%) children, followed by typical HFMD in 214 (22%) and herpangina in 57 (6%) children. Throat specimens of 245 children were processed with a target-enrichment new-generation sequencing approach, which generated 213 complete CVA6 genomes. The genomes grouped within the D1 and D3 clades (phylogeny inferred with the P1 genomic region). In total, 201 genomes were classified among the recombinant forms (RFs) A, B, F, G, H, and N, and 12 genomes were assigned to 5 previously unreported RFs (R-V). The most frequent RFs were A (58%), H (19%), G (6.1%), and F (5.2%). The yearly number of RFs ranged between 1 (in 2012 and 2013) and 6 (2018). The worldwide CVA6 epidemic transmission began between 2005 and 2007, which coincided with the global spread of the recombinant subclade D3/RF-A.

Keywords: atypical hand-foot-and-mouth disease; clinical epidemiology; enterovirus surveillance; molecular typing; recombination; third-generation sequencing; whole-genome sequencing.

Figure 1

Distribution of CVA6 and other EV types during ambulatory hand-foot-and-mouth disease surveillance. From April 2010 to March 2014, the data are from the surveillance of the population of the Clermont-Ferrand city area. From April 2015 to April 2017, the comprehensive national surveillance was interrupted. Colour code: red, CVA6; green, EV-A other than CVA6; black, other EV types.

Figure 2

Genetic diversity among the coxsackievirus A6 genomes determined in this study. The mean pairwise genetic diversity was calculated between the 213 CVA6 complete genomes assigned to subclades D1 and D3. The intra-clade nucleotide diversity (Pi) and inter-clade divergence (Div) were calculated as a mean proportion using a sliding window of 100 nts with a step size of 10 nts. Locations of genomic regions in the alignment: P1, nucleotide positions 748–3357; P2, nucleotide positions 3358–5349; P3, nucleotide positions 5350–7351.

Figure 3

Genetic pool of recombinant forms (RFs) identified among coxsackievirus A6 genomes. The 3Dpol gene was derived from a dataset of 760 (near full) genomes including the 213 CVA6 genomes determined in this study, and was analysed with a Bayesian method. Large full black circles at nodes indicate posterior probability values > 0.9 representing distinct RFs. The clades designated with letters A to Q correspond to the previously identified RFs. Open diamonds along branches indicate RFs previously unreported before this study. The 11 clades labelled with an asterisk include sequences reported in this study. For more clarity, the sequences within each clade were collapsed.

Figure 4

Phylogeny inferred from the P1 genomic region of coxsackievirus A6. A set of 760 P1 sequences was constructed from the 213 complete genomes from this study and 547 genomes available in GenBank. The phylogenetic analysis was performed with BEAST software [30]. Posterior probability is indicated by the size and intensity of the red colour of the circles at the different nodes. The inset table indicate the time to the most recent common ancestor (tMRCA) and the 95% highest probability density interval (95% HPD).

Figure 5

Chronogram and geographic distribution of CVA6 strains. Phylogenetic analysis was performed using the Nextstrain Augur toolkit [34].