The human enterovirus is a genus of positive-sense single-stranded RNA-viruses of the Picornaviridae family and includes enteroviruses (EVs) classified into species from A to D and rhinoviruses (RVs) classified into species from A to C, which collectively amount to over 250 different types.1
Up to 90% of EV infections are asymptomatic even if EVs can cause a wide spectrum of clinical manifestations ranging from mild respiratory/gastrointestinal symptoms to serious conditions, such as aseptic meningitis, encephalitis, neonatal sepsis and polio-like paralysis.1 EV infections are more common in children than in adults, and young children generally have more severe clinical presentations.1 In temperate countries, EVs usually show marked summertime seasonality1 even if EV circulation has also been reported in winter.2 Like other respiratory viruses, EV are transmitted via droplet or direct contact, causing mild to severe acute respiratory infections (SARIs) that are not distinguishable from those caused by other pathogens.1
Several emerging and reemerging respiratory EVs have recently been reported worldwide.3,4 The unexpected EV-D68 outbreaks in North America in 2014 and in Europe in 20165 caused great concern which prompted public health authorities and World Health Organization (WHO) to consider including emerging EVs in the Blueprint list of priority diseases.6
In Italy, the absence of an EV surveillance system makes EV epidemiology particularly challenging and EV circulation or newly emerging EVs impossible to monitor or determine.
The primary aim of this study was to evaluate the EV-positivity rate in respiratory samples collected from children ≤15 years hospitalized with SARIs at a university and research hospital in Milan from 2014 to 2017. The secondary aim was to describe the epidemiologic and molecular characteristics of EVs.
MATERIALS AND METHODS
Study Design, Population and SARI Case Definition
All of the respiratory samples (nasal/oropharyngeal swabs, nasopharyngeal aspirates, broncho-alveolar lavages) collected from SARI patients ≤15 years of age admitted to the “Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico,” a university and research hospital (a 900-bed tertiary center; 36000 hospitalizations/year) located in Milan (Italy) from September 1, 2014 to August 31, 2017 were included in the study.
The standard SARI case definition was used: acute respiratory infection with history of high fever or measured high fever (≥38°C) and cough with onset within the last 10 days and requiring hospitalization.7 The following data were collected for each SARI case: gender, age, hospital ward, symptom onset, date and presence of viral respiratory coinfection (defined as the simultaneous presence of at least one other respiratory virus, other than EV, in the same respiratory sample).
The study was divided into three 12-month study periods beginning on September 1, 2014 and ending on August 31, 2017.
This study was carried out in compliance with national legislation on data collection, data storage and reporting and patient confidentiality; the study complied with the requirements laid down by the 1975 Helsinki Declaration.
Molecular Detection of EVs
After DNA/RNA extraction (EZ1 DSP Virus kit, Qiagen GmbH, Hilden, Germany), the specimens were tested for the presence of 16 respiratory viruses—EV, RV, adenovirus, bocavirus, respiratory syncytial virus A and B, metapneumovirus, influenza viruses A and B, parainfluenza viruses 1–4, coronavirus 229E, NL63 and OC43—by a multiplex real-time one-step RT-PCR assay (Anyplex II, RV16-Detection, Seegene Inc., Seoul, Korea)8 on CFX-96 real-time PCR thermal cycler (Bio-Rad Inc., Hercules, CA). This multiplex real-time one-step RT-PCR assay allows the detection of EVs and RVs as specific targets.
The EV-positive samples were then analyzed with a one-step real-time RT-PCR assay targeting the 5′ untranslated region of EV-D68 as described by Poelman et al.9 A run was considered valid if it completed normally and it passed internal controls. A sample was considered positive when its cycle threshold value was below 39.
Molecular Characterization of EVs
EV-positive samples (excluding EV-D68) were molecularly characterized using a culture-independent method. All of the sequences were obtained using nested RT-PCR primers targeting the VP1 region of EVs (nt. 2602–2977).10 Amplicons were purified using a NucleoSpin gel and PCR clean-up kit (Macherey-Nagel Gmbh & Co. KG, Duren, Germany) and were sequenced with a big-dye terminator cycle-sequencing kit (Thermo Fisher Scientific, Inc., Waltham, MA) in an ABI Prism 3130xl genetic analyzer (Thermo Fisher Scientific, Inc., Waltham, MA). EV types were identified by carrying out nt. sequence similarity searches using both the Basic Local Alignment Search Tool from the National Center for Biotechnology Information website11 and the RIVM Enterovirus Genotyping Tool, version 0.1 from National Institute for Public Health and the Environment.12 A query sequence was assigned to a type if the percentage of query coverage was >70% and the percentage of nt. identity was at least 75%. The EV strains identified were aligned with EV prototype strains and other reference sequences retrieved from the GenBank database13 using ClustalW.14 The phylogenetic relationship between study and reference strains was assessed by computing sequence identity using BioEdit14 and the pairwise p-distance with MEGA, version 6.15
The statistical analysis was performed using Open Source Epidemiologic Statistics for Public Health OpenEpi, version 3.01.16 The data were summarized as medians and lower (Q1) and upper (Q3) quartiles and inter quartile ranges (IQR). The frequency was expressed as crude rate. Comparisons of proportions between all groups were carried out using the χ2 test based on binomial distribution; the unpaired t test was used to determine the differences between the continuous variables. The Mann-Whitney U test or analysis of variance was used to compare the means of the continuous variables. A P-value <0.05 was considered significant (two-tailed test).
Risk of infection was expressed as the number of individuals with a laboratory-confirmed infection of the total number of individuals with SARI and a specific characteristic. The odds ratio (OR) and exact confidence limits (95% CI) were calculated using the Mid-P exact test assuming a normal distribution.
The onset and the end of epidemic were defined as the first of 2 consecutive weeks with at least a 10% EV-positive rate and the last 2 consecutive weeks with <10% EV-positive rate, respectively.
Characteristics of SARI Cases
From September 1, 2014 to August 31, 2017, 2468 respiratory samples were collected from patients ≤15 years diagnosed with SARI. Case distribution defined by study season, gender, age group, hospital ward and symptom onset date is summarized in Table 1. The cases were distributed evenly across the 3 study periods: 33.6% between 2014 and 2015 period, 31.5% between 2015 and 2016 period and 34.9% between 2016 and 2017 (P > 0.05). Overall, 55.8% of the participants were males with a male-to-female ratio ranging from 1.2:1 in 2014 to 2015 to 1.38:1 in 2016 to 2017.
The median age of SARI patients was 10.2 months (IQR: 32.6 months). Of our SARI cases, 75.6% were children 0–3 years of age (Table 1), 22.5% (555/2468) of whom were children under 1 month of age, 32.4% (799/2468) were children 1–12 months of age and 20.9% (517/2468) were children 1–3 years of age. Of the SARI cases, 12.8% (317/2468) and 11.6% (286/2468) were children 4–6 years and 7–15 years of age, respectively (Table 1).
Of the respiratory samples collected, 93.4% (2304/2468) were from SARI patients on bed rest, 27.7% of whom came from neonatal/pediatric intensive care units (ICUs); the remaining 6.6% of the samples were collected from SARI patients attending the emergency room (ER). The risk of ICU admission was higher in 2016 to 2017 than in 2014 to 2015 (OR: 1.5; 95% CI: 1.2–1.9) and 2015 to 2016 (OR: 1.8; 95% CI: 1.5–2.3). The risk of ER admission was approximately 2- to 3-fold higher in 2016 to 2017 than in 2014 to 2015 (OR: 2.3; 95% CI: 1.6–3.5) and 2015 to 2016 (OR: 3.1; 95% CI: 2.1–4.9), respectively. 36.2% of SARI cases occurred from December to February and 14.5% from June to August (Table 1). During each study period, the number of SARI cases peaked at the end of December.
Detection of EVs and Epidemiologic Characteristics of EV-positive SARI
The EV-positivity rate was 9% with no statistical differences among study periods (Table 2).
No significant differences were observed in EV infection risk by gender (OR: 1.2; 95% CI: 0.8–1.5), even if 59.5% of EV-positive samples were collected from males (P = 0.02) (Table 2). The median age of EV-positive children was higher (18.9 months, IQR: 25.4 months) than that of EV-negative children (9 months, IQR: 32.7 months), but with no statistical significant difference (P = 0.23). Most (77%; 171/222) EV-positive cases were observed among children 0–3 years of age (Table 2), although the highest risk of EV infection was observed for children 4–6 years of age. In fact, the risk of EV infection in this age group was 1.4-fold (95% CI: 0.99–2.1) and 3.6-fold (95% CI: 1.8–7.4) higher than the risk of infection in 0–3 years children and in those >6 years, respectively. 18.1% (40/222) and 4.6% (11/222) of EV infections were identified in 4–6 years children and 7–15 years children, respectively (Table 1).
Most (84.2%) EV-positive SARI samples were collected from children admitted to general wards, while 12.7% of EV-positive cases required ICU admission (Table 2).
EVs were detected throughout the study period (Table 2); 27.5% of EV-positive samples were identified from December to February and 38.3% from June to September, when the EV transmission risk was almost 5-fold higher (OR: 4.7; 95% CI: 3.1–7.0) than in the other months. Overall, the highest EV-positive rate was observed from June to September and in December, when over half (55%) of all EV-positive cases occurred. During the 2016–2017 season, the highest EV-positive rate was observed in September (13%) and in December (18.2%) when 40% and 21.4% of EVs respectively, were identified in SARI patients requiring ICU admission. 57.3% of the EVs detected in samples from SARI cases attending the ER were identified during the 2016 to 2017 season (Table 2).
Throughout the study period, EVs came in 2 annual epidemic waves: the first wave began at the end of May, peaked in beginning of July and ended in mid-August, while the second wave started in mid-November, peaked in mid-December and ended at the end of December (Fig. 1).
Overall, other respiratory viruses were detected in 64.8% of EV-positive SARIs, and the risk of coinfection was 4-fold (OR: 4.32; 95% CI: 2.9–6.4) higher than a single EV infection. RV (42.3%), respiratory syncytial virus (16.2%) and bocavirus (15.5%) were the most frequently detected viruses. Children with single EV infections were significantly younger than the coinfected patients (median age: 10.2 versus 12.8 months; P = 0.007). The risk of coinfection was 2-fold (OR: 2.1; 95% CI: 1.1–4.4) higher than contracting a single EV infection during autumn (September–November), while the risk of a single EV infection was 3-fold (OR: 3.1; 95% CI: 1.7–5.6) higher than the risk of coinfection during summer (June–August).
Molecular Epidemiology of EVs
One hundred forty-two EV-positive samples of 222 (64%) were further characterized. Overall, 14.8% of EVs were EV-D68 and 90.5% of these were detected in 2016. The EV-D68 epidemic wave started in mid-June, peaked in the end of September and ended at the end of October 2016; during this period EV-D68 was identified in 8.5% of all-cause SARIs and in 47.1% (95% CI: 31.4–63.3%) of EV-positive SARIs, with an over 30-fold higher risk of EV-D68 infection than any other respiratory virus (OR: 34.0; 95% CI: 12.6–106.7) and any other EV type (OR: 31.0; 95% CI: 10.5–105). The median age of EV-D68-positive children was 20.6 months (IQR: 31.7 months) and 9.5% of them required intensive care. The risk of EV-D68 infection was similar for children admitted to ICUs and to general wards (OR: 1.3; 95% CI: 0.2–8.0).
Fifty-four EV-D68-negative samples (54/121; 44.6%) were successfully characterized by sequence analysis of the VP1 gene. After combining the results obtained by EV-D68 specific assay and EV-sequencing, 75 of 142 (52.8%) strains were assigned to a type. According to sequence analysis results, 22 different viral EV types were identified; EV-B species was the most frequently detected (49.3%), followed by D (28%), A (20%) and C (2.7%). As shown in Table 3, EV-B species encompassed the largest number of types (N = 12), followed by A (N = 7) and C (N = 2), while EV-D68 was the only type of species D (Table 3). Overall, 26.7% of EVs were coxsackieviruses (CV) A, 22.7% echovirus (Echo) and 20% CV-B (Table 3). EV-D68 was the type most frequently (28%) detected in our SARI series, followed by CV-B2 (9.3%) and Echo-25 (8%) (Table 3).
Twenty different EV types were identified in children 0–3 of age years while fewer EV types (N = 10) were detected in children >3 years.
Excluding EV-D68, no differences were observed in EV-type frequencies throughout the study period.
Except for EV-D68, 5 emerging/reemerging EV types were identified: namely, CV-A6, CV-A16, EV-A71, CV-A21 and EV-C105. The CV-A6 strains detected were genetically related (overall mean p-distance 0.063; range: 0.058–0.067), while the CV-A16 strains revealed a broader genetic diversity (overall mean p-distance: 0.106; range: 0.003–0.153) and both strains showed a high level of nt. similarity with available EV sequences (CV-A6: 90.8%–96.6%; CV-A16: 82%–95.5%). The identified EV-A71 belonged to the C2 sub-genogroup (nt. similarity: 92.2%–94.2%). The CV-A21 had a nt. similarity of 87.5%–99.1% with most of the available sequences and shared the highest similarity with the Dutch strains detected in 2013 and 2014. In this study, the EV-C105 strain showed high similarity with the other available sequences (overall mean p-distance: 0.08; range: 0.006–0.161) and a 99% nt. similarity with a strain detected in the same geographical area in 2011.
This study investigated the epidemiologic and molecular characteristics of EVs among pediatric SARI patients under 15 years of age during 3 consecutive seasons (from 2014 to 2017).
The results obtained by analyzing nearly 2500 pediatric SARIs in one of the largest Italian pediatric research hospitals demonstrated that EVs were frequently detected in respiratory samples collected from SARI cases; different EV types circulated throughout the study period and an EV-D68 outbreak was identified in 2016.
In this study, EVs were detected in 9% of all SARI cases as previously observed in other studies, in which the EV detection rate ranged between 4% and 8%.17–19 However, most studies targeting SARI-causing respiratory viruses used commercial multiplex-PCR assays which cannot distinguish EVs from RVs or totally exclude EVs, thus hindering the comprehensive evaluation of EV contribution to SARI.20 It is well known that the incidence rates of EV infections are highest in young children21; indeed most of the EV-positive SARIs (77%) were identified in children ≤3 years. In this study, the median age of EV cases was lower than that reported for children with influenza-like illness in the same Italian region (19 versus 31 months).2 However, although the highest number of EV-positive SARI cases was observed for children ≤3 years, children 4–6 years of age proved to be those at greatest risk of contracting EV infections, as reported by Li et al.22
Coinfection with more than one virus in samples from children hospitalized with respiratory infections has been widely reported23; in this study, the frequency of EV coinfection with at least one other respiratory virus (particularly RV) was statistically higher than the frequency of a single EV infection, as reported by Eifan et al,18 making it difficult to determine a causal relationship between SARI and EV. A systematic review and meta-analysis23 and a recent study20 have not found differences in clinical disease severity and in patient management in pediatric respiratory infections caused by EV/RV nor any differences between viral coinfections and single viral infections. In our study, the median age of children with single EV infections was lower than the median age of coinfected children. Twenty-five percent of our SARI cases required ICU admission while nearly 7% of the cases attended the ER, approximately 13% and 3% of EV-positive cases were admitted to the ICU and ER, respectively. As previously reported, the risk of EV infection among ICU and non-ICU cases was similar throughout the study period20—however, it increased significantly between September and December 2016.
Although respiratory viruses are more likely to spread during the winter months, EVs are usually more active in the summer and autumn.19,21 In our study, EVs circulated all-year-round with a well-defined seasonality during the 3-year surveillance period; in fact, the highest EV activity was observed between June and September and in December, when over 50% of the EV-positive cases were identified. Two EV epidemic waves occurred in each season: a first broader outbreak occurred from the end of May to mid-August and peaked in the beginning of July when there was an almost 5-fold greater risk of EV infection than in the other months and the risk of single EV infection was 3-fold higher than coinfection. The second smaller epidemic wave occurred from mid-November to end of December, when 27.5% of EV-positive SARI and approximately one-third of coinfections were reported. In a study on EV circulation in children with influenza-like illness carried out over 7 consecutive winter seasons, EVs showed an epidemic-like trend and an increased EV-positive rate in December 2016.2 These findings are in line with the European Centre for Disease Prevention and Control Rapid Risk Assessment24 and the US Centre for Disease Prevention and Control surveillance data,25 which have reported increases in EV activity and EV-related disease severity from April 2016 onwards, respectively.
From 2014 to 2016, the global emergence of EV-D68 causing SARI epidemics and cases of severe poliomyelitis-like syndromes in young children was widely reported5 and the risk of other emerging respiratory and neurotropic EVs remains a public health concern. EV-D68 is currently considered one of the most serious respiratory viruses, thus it has been promptly investigated; EV-D68 was identified in ~15% of EV-positive SARIs, which is a lower percentage than those reported in other studies, where detection rates ranged from 22% to 33%17,26 and may be biased and underestimated since residual respiratory samples were not available for all of the EV-positive SARIs. In this study, EV-D68 was identified almost exclusively in 2016 during a serious, time-limited outbreak that started in mid-June and ended at the end of October, with an over 30-fold higher risk of EV-D68 infection than the risk of infection with any other EV or respiratory virus. During this period, EV-D68 was identified in 8.5% of all SARIs and in almost 47% of EV-positive cases. The EV-D68 outbreak peaked in September 2016, when a significant increase in the number of EV-positive cases requiring intensive care was observed. EV-D68 infection can cause serious respiratory complications or severe neurologic impairments5; in our series, ICU admission was only required for ~10% of all EV-D68-positive SARIs.
The spread of EV types is generally unpredictable and different types of EVs can co-circulate in different years.21 Despite limitations due to the impracticability of sequencing all EVs, a remarkable heterogeneity of circulating EVs was observed; in fact, besides EV-D68, 21 EV types were identified during the study period and no significant changes were observed in their distribution over time as previously reported in France27 and Spain.28 As observed in other studies,17,19,27,28 EV-B was the most frequently identified species (~50%), followed by D (28%) represented by EV-D68 exclusively. Although in contrast with the findings of Hellferscee et al,17 the EV species A circulated more frequently than the EV species C (20% versus 2.7%) among patients with respiratory infections as previously reported.4,19 As regards the number of types, EV species A and B were the most heterogeneous, while EV-C only included EV-C105 and CV-A21. Some authors have shown that the polymorphism of the 5′ untranslated region region of several EV-C can bias their detection, thus underestimating their distribution in the population 3,4; moreover, Van Leer-Buter et al showed that EV-C species viruses mainly affect adult patients with respiratory infections.3
During the study period similar frequencies of CV-A, Echo and CV-B were observed. It is interesting to note that differences in frequencies of EV-type detection were shown according to age, with CV-A more frequently detected in children ≤3 years and CV-B in those 4–15 years of age. According to the findings of Cabrerizo et al,28 the most substantial EV-type heterogeneity was observed among the youngest children. Apart from EV-D68, CV-B2 and Echo-25 were the most frequently identified viruses in this study, while CV-B4, Echo-11, -6 and -30 were the most frequently detected in the Netherlands and France,3,19 thus emphasizing that spatial/temporal factors may influence the predominance of circulating EV types.
In our SARI series, EV-A71, CV-A6, CV-A16, CV-A21 and EV-C105—which are considered emerging/reemerging viruses that may endanger human health21—were seldom identified. Since 2016, large outbreaks of EV-A71 infection mainly associated with rhomboencephalitis-like neurologic complications have been reported throughout Europe,24 thus emphasizing the continuous genetic evolution of this virus and the importance of prompt detection and monitoring. CV-A6 had not widely circulated until recently when atypical and severe hand-foot-mouth disease outbreaks were reported worldwide.29 CV-A16 is known to frequently undergo genetic changes that may cause large and repeated outbreaks.30 EV-C105 and CV-A21 are considered emerging EVs associated with outbreaks of respiratory infections of particular concern to public health4 which were detected in this study and recently reported in Italy, the Netherlands and Denmark,3,4,31 thus indicating their arrival in Europe.
A limitation of this study is that the presence of bacterial coinfection was not assessed and that no data on preexisting morbidity conditions or risk factors associated with the development of SARI were available in our series.
In conclusion, the virologic monitoring of EVs in pediatric SARI cases enabled us to identify an EV-D68 outbreak in 2016, demonstrating a remarkable heterogeneity among circulating EVs and the circulation of emerging EV types with epidemic potential, and gave us a more complete molecular-epidemiologic picture of the EVs circulating in Italy. Molecular characterization of EVs is a valuable tool for gaining a better understanding of the molecular features of circulating viruses and for determining the introduction of emerging viruses. Therefore, combining the expertise of diagnostic and public health laboratories may be a successful strategy for monitoring the epidemiologic changes of viruses that endanger human health and should be considered part of the WHO epidemic preparedness program, as discussed in the recent WHO R&D blueprint initiative for emerging infectious diseases.
The authors would like to thank the children, parents and health-care personnel involved in the study.
1. Pallansch MA, Oberste MS, Whitton JL. Knipe DM, Howley PM, Cohen JI. Enteroviruses: polioviruses, coxsackieviruses, echoviruses, and newer enteroviruses. In: Fields Virology. 2013:vol 1. 6th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 490–530.
2. Pellegrinelli L, Bubba L, Galli C, et al. Epidemiology and molecular characterization
of influenza viruses, human parechoviruses and enteroviruses in children up to 5 years with influenza-like illness in Northern Italy during seven consecutive winter seasons (2010-2017). J Gen Virol. 2017;98:2699–2711.
3. Van Leer-Buter CC, Poelman R, Borger R, et al. Newly Identified Enterovirus
C Genotypes, Identified in the Netherlands through Routine Sequencing of All Enteroviruses Detected in Clinical Materials from 2008 to 2015. J Clin Microbiol. 2016;54:2306–2314.
4. Barnadas C, Midgley SE, Skov MN, et al. An enhanced Enterovirus
surveillance system allows identification and characterization of rare and emerging respiratory enteroviruses in Denmark, 2015-16. J Clin Virol. 2017;93:40–44.
5. Holm-Hansen CC, Midgley SE, Fischer TK. Global emergence of enterovirus
D68: a systematic review. Lancet Infect Dis. 2016;16:e64–e75.
8. Cho CH, Chulten B, Lee CK, et al. Evaluation of a novel real-time RT-PCR using TOCE technology compared with culture and Seeplex RV15 for simultaneous detection of respiratory viruses. J Clin Virol. 2013;57:338–342.
9. Poelman R, Schuffenecker I, Van Leer-Buter C, et al. European surveillance for enterovirus
D68 during the emerging North-American outbreak in 2014. J Clin Virol. 2015;71:1–9.
10. Nix WA, Oberste MS, Pallansch MA. Sensitive, seminested PCR amplification of VP1 sequences for direct identification of all enterovirus
serotypes from original clinical specimens. J Clin Microbiol. 2006;44:2698–2704.
14. Hall T. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp Ser. 1999;41:95–98.
15. Tamura K, Stecher G, Peterson D, et al. MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Mol Biol Evol. 2013;30:2725–2729.
16. Dean AG, Sullivan KM, Soe MM. Open Source Software for Public Health Epidemiology (OpenEpi). Available at: http://www.openepi.com
. Accessed June 26, 2019.
17. Hellferscee O, Treurnicht FK, Tempia S, et al. Enterovirus
D68 and other enterovirus
serotypes identified in South African patients with severe acute respiratory illness, 2009–2011. Influenza Other Respir Viruses. 2017;11:211–219.
18. Eifan SA, Hanif A, AlJohani SM, et al. Respiratory tract viral infections and coinfections identified by Anyplex™ II RV16 detection kit in pediatric patients at a riyadh tertiary care hospital. Biomed Res Int. 2017;2017:1928795.
19. Jacques J, Moret H, Minette D, et al. Epidemiological, molecular, and clinical features of enterovirus
respiratory infections in French children between 1999 and 2005. J Clin Microbiol. 2008;46:206–213.
20. Wishaupt JO, van der Ploeg T, de Groot R, et al. Single- and multiple viral respiratory infections in children: disease and management cannot be related to a specific pathogen. BMC Infect Dis. 2017;17:62.
21. Khetsuriani N, Lamonte-Fowlkes A, Oberst S, et al; Prevention CfDCa. Enterovirus
surveillance--United States, 1970–2005. MMWR Surveill Summ. 2006;55:1–20.
22. Li W, Zhang X, Chen X, et al. Epidemiology of childhood enterovirus
infections in Hangzhou, China. Virol J. 2015;12:58.
23. Asner SA, Science ME, Tran D, et al. Clinical disease severity of respiratory viral co-infection versus single viral infection: a systematic review and meta-analysis. PLoS One. 2014;9:e99392.
26. Savage TJ, Kuypers J, Chu HY, et al. Enterovirus
D-68 in children presenting for acute care in the hospital setting. Influenza Other Respir Viruses. 2018;12:522–528.
27. Molet L, Saloum K, Marque-Juillet S, et al. Enterovirus
infections in hospitals of Ile de France region over 2013. J Clin Virol. 2016;74:37–42.
28. Cabrerizo M, Díaz-Cerio M, Muñoz-Almagro C, et al. Molecular epidemiology of enterovirus
and parechovirus infections according to patient age over a 4-year period in Spain. J Med Virol. 2017;89:435–442.
29. Bian L, Wang Y, Yao X, et al. Coxsackievirus A6: a new emerging pathogen causing hand, foot and mouth disease outbreaks worldwide. Expert Rev Anti Infect Ther. 2015;13:1061–1071.
30. Hosoya M, Kawasaki Y, Sato M, et al. Genetic diversity of coxsackievirus A16 associated with hand, foot, and mouth disease epidemics in Japan from 1983 to 2003. J Clin Microbiol. 2007;45:112–120.
31. Piralla A, Daleno C, Girello A, et al. Circulation of two enterovirus
C105 (EV-C105) lineages in Europe and Africa. J Gen Virol. 2015;96pt 61374–1379.