A large-scale outbreak of hand, foot and mouth disease, France, as at 28 September 2021

We report a large-scale outbreak of hand, foot and mouth disease (HFMD) in France. As at 28 September 2021, 3,403 cases have been reported (47% higher than in 2018–19). We prospectively analysed 210 clinical samples; 190 (90.5%) were enterovirus-positive. Most children presented with atypical HFMD. Coxsackievirus (CV)A6 (49.5%; 94/190) was predominant; no enterovirus A71 was detected. Dermatological and neurological complications of HFMD justify prospective syndromic and virological surveillance for early detection of HFMD outbreaks and identification of associated types.

Since September 2017, PARI has been conducting prospective surveillance of the most common paediatric infectious diseases seen in outpatient settings. The EV-associated mucocutaneous illnesses HFMD/HA are included in the PARI surveillance panel [4]. The PARI network includes 113 paediatricians (as at September 2021; ca 4% of all French paediatricians) located throughout all French metropolitan regions except Corsica and is coordinated by the Association Clinique et Thérapeutique du Val de Marne (ACTIV). Each paediatrician uses the same medical software (AxiSanté 5-InfanSoft, CompuGroup Medical, Nanterre, France) for patient record management. The system enables paediatricians to report cases of HFMD/HA based on diagnoses according to the International Statistical Classification of Diseases, 10th revision, including B08.4 (enteroviral vesicular stomatitis with exanthem), B08.5 (enteroviral vesicular pharyngitis), B97.1 (enterovirus as the cause of diseases classified elsewhere), and R21 (rashes and other nonspecific skin eruptions suggesting enterovirus infection) [5]. Automated data extraction from computerised medical records tracks the clinical epidemiology of HFMD/HA. A weekly newsletter reporting the number of cases or each monitored infectious disease is sent to all paediatricians and the AL-NRC by ACTIV.
A rapid increase in the number of mucocutaneous manifestations suggestive of EV infection in children was observed at week 24 in 2021 by the PARI network. The number of cases continued to increase until week 27, during which 291 cases were recorded (Figure 1). This is the highest number of weekly cases reported since syndromic surveillance began in 2017. A second HFMD/HA upsurge began in week 35 and peaked in week 37. As at week 38 in 2021, 3,403 cases were reported, compared to respective case numbers in 2020 (n = 1,201), 2019 (n = 1,996) and 2018 (n = 2,616) for the same period.

Clinical and virological surveillance
Syndromic surveillance was complemented by virological surveillance of HFMD/HA. Virological surveillance was first established in April 2014. After a suspension in 2020 because of the COVID pandemic, the surveillance was reactivated in 2021 during week 24, following the PARI network alert. Briefly, volunteer paediatricians in the PARI network collected, at discretion, throat or buccal swab specimens from children clinically diagnosed with HFMD/HA and sent samples with a standardised clinical report form to the AL-NRC for pan-EV testing (Enterovirus R-GENE, bioMérieux, Marcy-l'Étoile, France) and EV genotyping [6]. Based on the clinical items checked on the form, typical HFMD was defined by the presence of at least one typical localisation of HFMD, e.g. oral ulcerations, rash on palms, soles, buttocks, elbows or knees excluding any other site, whereas atypical HFMD was defined as HFMD with the presence of a rash at atypical sites or a generalised rash, or an atypical presentation such as Gianotti-Crosti syndrome (including papulo-vesicular eruption on buttocks and limbs) or eczema coxsackium. Herpangina was defined as localised oral ulcerations in the posterior part of the oral cavity [6].
The phylogenetic analysis of 71 partial VP1 sequences of the 94 CVA6 strains sampled in France in 2021 showed that 69 sequences were grouped in two main distinct lineages within the genogroup D. The two remaining sequences were dispersed within the genogroup D.
All sequences displayed close genetic relationships with viruses recovered in France in 2017-18 ( Figure  4). However, whole genome analysis of viral strains is needed to determine whether the strains involved in the 2021 epidemic could represent new variant forms.

Ethical statement
Informed consent was obtained from all parents or guardians of children. The syndromic surveillance was approved by the ethical committee of CHI Créteil Hospital, France [4]. The virological surveillance was approved by the review committee of the University Hospital of Clermont-Ferrand, France (reference AU1098).

Discussion
The PARI network revealed a large-scale outbreak of HFMD/HA in France in 2021. The number of cases was 47% higher than in years 2018-19 during the same time period. Virological surveillance from week 24 showed that this epidemic was mainly associated with the CVA6 type. The biannual pattern in summer and autumn was  similar to that usually expected for these diseases in Europe or Asia [2,4,7]. However, the 2021 epidemic was characterised by a sharp rise in the summer and an earlier autumn peak compared with previous years. This pattern may reflect that a higher proportion of children was susceptible to viral infections including EV. The year 2020 was, in fact, marked by a very low circulation of EV, both among hospitalised patients [8] and among children seen in primary care. In response to the COVID-19 pandemic, hygiene reinforcement may have reduced infectious contacts and immune stimulation, consequently leading to a greater susceptibility to infections in children [9]. A similar phenomenon was illustrated by the recently reported outbreak of respiratory syncytial virus (RSV) infections in Japan in July 2021 [10], which was substantially larger and earlier compared with previous years. The same situation could be observed in Europe for RSV infections [11].
In our study, we did not observe a shift in the age of disease onset; the mean age of children with HFMD/ HA was similar to that of children in the 2014 epidemic (2.09 vs 2.10 years) [6]. Although we have not yet found available epidemiological data worldwide on HFMD/HA for the year 2021, it seems unlikely that the epidemic is limited to France. Signals of an HFMD/HA outbreak may have been missed elsewhere in Europe because most existing surveillance systems are hospital-based and focus on neurological EV infections [12].
We found that nearly two thirds of the children with confirmed EV-associated HFMD/HA in our study presented with atypical disease, a proportion usually observed in epidemics associated with CVA6 [6,13]. Since its re-emergence in 2008, this type has been commonly associated with atypical forms of the disease [6,13],  The phylogenetic tree was constructed by the neighbour-joining method and evaluated with 1,000 bootstrap pseudoreplicates, using Molecular Evolutionary Genetics Analysis (MEGA) version 6.0 software [19]. Only bootstrap values ≥ 70% are indicated by a black star. Genetic distances were calculated with Tamura-Nei's model of evolution and branch length is drawn to the indicated scale (proportion of nt substitutions per site). Sequences were 495 bp long and started at nt 1 relative to the 1D gene of the Gdula prototype strain. The analysis was performed with 71 partial VP1 sequences of 94 CVA6 strains. The remaining 23 sequences were less than 495 bp long. Taxon names of all sequences are provided in Supplement. The strains collected in 2021 are labelled with a filled circle. Geographical origin and time of isolation of strains are indicated by the ISO-code abbreviation followed by the year of isolation. Genogroups were designed as previously described [20].
sometimes affecting the whole body or associated with late-onset symptoms such as onychomadesis [14]. In addition, CVA16 was more frequently associated with atypical disease forms in 2021 than in 2014 (61.9%; n = 26/42 vs 34.1%; n = 42/123) [6]. Moreover, while paediatricians are very familiar with HMFD/HA, other clinicians should also be aware of these diseases, as they are not uncommon in adults living with children.
In a CVA6-associated outbreak described in Finland in 2008, one third of the patients with HFMD/HA were adults [15]. As in children, presentation of the disease in adults can be atypical, which can make the diagnosis difficult and, in some cases, lead to hospitalisation [16].
In our primary care-based study, no data on hospitalisation were collected. As some HFMD/HA cases may be complicated by extensive skin manifestations or severe neurological conditions, this information would be valuable, as it could allow the close monitoring of hospitalised cases of HFMD/HA to detect a change in the clinical presentation or in the frequency of complications associated with these diseases. Data on hospitalised cases of HFMD/HA could be collected through hospital-based surveillance of EV infections as proposed by the European Non-Poliovirus Enterovirus Network (ENPEN), which has established standardised protocols for surveillance for HFMD, respiratory and neurological infections caused by EV and parechoviruses [17]. In France, no increase in the number of hospitalised cases of HFMD/HA has been observed through the EV hospital-based surveillance [8].

Conclusions
The widespread circulation of non-polio EV, the epidemic pattern of EV infections and the recent emergence of EV-D68 and EV-A71 involved in severe neurological conditions warrant reinforcement of the surveillance of EV infections [18]. The PARI ambulatory syndromic surveillance proved to be effective for the early tracking of the 2021 outbreak. Detection of future HFMD/HA outbreaks caused by EV-A71 or an emergent EV could be achieved in combination with virological investigations. This study shows that a paediatric ambulatory network is of value for timely detection of diseases mostly seen within primary care settings.