Subclade B3 and D3 enterovirus D68 among children with severe community-acquired pneumonia in China, 2021 to 2024 Authors: Yuanyuan Zhou, Jiayi Lin, Xuejie Zhang, Yi Zhu, Liting Wang, Linfeng Cao, Yingyan Shi, Guangrong Sun, Junfeng Wu, Suqing Liu, Shiyong Zhao, Weibang Huo, Xiaohui Zou, Wenjie Tan, Roujian Lu, Jun Shen Published: 2025 ABSTRACT AbstractEnterovirus D68 (EV-D68) is an emerging pathogen that causes severe respiratory infections in children worldwide and has the potential to re-outbreak. The aim of this study was to assess EV-D68 detection and genetic variability in pediatric patients with community-acquired pneumonia (CAP) during and after the COVID-19 pandemic in China. Respiratory samples, including bronchoalveolar lavage fluid (BALF), were collected from hospitalized children with CAP in Shanghai, Shandong, Fujian, Zhejiang and Jiangsu province between January 1, 2021 and December 31, 2024. Samples were screened for EV-D68 using real-time RT-PCR, and clinical characteristics of positive cases and virus strain sequences were analyzed. Of 2,842 cases, 0.7% (21/2,842) positive with EV-D68. The median age of these patients was 5.0 years (IQR: 3.6-8.0 years), with 42.9% (9/21) presented severe CAP. Of the 21 identified virus strains, 15 belonged to subclade B3 and 6 to subclade D3. In 2024, a total of 14 strains were detected, including 6 from the newly emerged subclade D3. Seven strains, 5 of which were subclade D3, were found in BALF samples from severe CAP patients. One child with severe CAP exhibited subclade D3 EV-D68 viremia. Notable mutations in the VP1 protein included H99D, K168E, and E169K. Our research showed that the EV-D68 continued to spread among children in eastern China before and after the COVID-19 pandemic. From 2021 to 2024, subclades B3 and D3 predominated, with increase detections in 2024 and association with pediatric severe CAP.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-30467-7.Keywords: Enterovirus D68, Enterovirus, Phylogenetic analysisSubject terms: Diseases, Medical research, Microbiology Introduction Since the global outbreak of enterovirus D68 (EV-D68) in 2014, surveillance data show a biennial prevalence pattern1–3. The anticipated 2020 epidemic peak was averted, likely due to COVID-19 preventive measures3,4. However, EV-D68 continued circulating globally, causing outbreaks in some regions4–6. Based on the analysis of the VP1 sequence, the EV-D68 virus has evolved into four distinct genotypes: A, B, C, and D. Since 2014, subclade B3 has replaced D1 and B2, becoming a major cause of severe respiratory infections and acute flaccid myelitis in children6,7. In 2016 and 2018, the A2/D3 strain emerged in China and Spain, linked to severe respiratory infections8,9. Our prior surveillance identified one subclade D3 strain in a pediatric community-acquired pneumonia (CAP) case10. We hypothesized that subclade D3 could trigger a new epidemic, similar to B37,11. Ongoing EV-D68 evolution raises concerns about compromised herd immunity and potential severe respiratory or central nervous system (CNS) infections, particularly in children5,12. Hospital-based viral surveillance is critical for early detection and prevention of outbreaks13. In 2024, we observed a high frequency of subclade D3 EV-D68 in BALF samples from children with severe CAP in China. Methods Study design and participants From January 1, 2021, to December 31 2024, respiratory samples, including nasopharyngeal swash (NPS), sputum, and BALF from hospitalized children across six hospitals were collected as previous reported10. One sample per patient was obtained within 48 h of admission. In each monitoring site located in Shandong, Fujian, Zhejiang, and Jiangsu provinces in the East China region, 50–100 samples are collected annually. For patients with positive respiratory EV-D68 detection hospitalized in Children’s Hospital of Fudan University, EV-D68 testing was also conducted on cerebrospinal fluid (CSF), serum, and stool samples when these specimens were concurrently available. Clinical data, including intensive care unit (ICU) admissions, mechanical ventilation requirements, lung CT imaging, bacterial culture results, and multiplex PCR findings for respiratory pathogens, were systematically collected. CAP severity was evaluated using the 2011 Infectious Diseases Society of America (IDSA) guideline, and CNS infection diagnostic criteria followed the 2018 IDSA guideline14,15. EV-D68 detecting and sequencing Viral RNA extracted from 200 µL of sample using the QIAamp MinElute Virus Spin kits (Qiagen, Germany). Real-time RT-PCR targeting the VP1 region of EV-D68 genome was performed using the following primers and probe: forward primer (5’- CACYGAACCAGAAGARGCCA − 3’), reverse primer (5’- CCAAAGCTGCTCTACTGAGRAA − 3’), and probe (5’-FAM- TCGCACAGTGATAAATCAGCAYGG –BHQ1-3’). The reaction conditions were as described previously10. VP1 gene amplification was conducted using OneStep RT-PCR with the forward primer (5’- TCCCACTACAGGATGTTYAA − 3’) and reverse primer (5’-GACTGTCTYTCCTCTGTGGT − 3’). The 50 µL reaction mixture contained 5 µL RNA extract, 10 µL 5×QIAGEN OneStep RT-PCR Buffer, 2µL dNTP mix (25 mM each), 2 µL of each primer (10 mM), 2 µL QIAGEN OneStep RT PCR Enzyme Mix (containing Omniscript and Sensiscript reverse transcriptases and HotStarTaq DNA Polymerase), and 27 µL RNase-/DNase-free water. Thermal cycling consisted of reverse transcription step at 50 °C for 30 min, PCR activation at 95 °C for 15 min, 40 cycles of amplification (denaturation at 94 °C for 30 s, annealing at 52 °C for 30 s, extension at 72 °C for 30 s), and a final extension at 72 °C for 10 minutes. The VP1 sequence was determined by Sanger sequencing. Selected positive specimens underwent whole-genome sequencing employing second-generation sequencing technology. Following sequencing, raw reads were subjected to quality control and adapter trimming using fastp (v0.24.1). Subsequently, the high-quality reads were aligned to the reference genome with BWA (v0.7.17-r1188). Genome coverage and sequencing depth at each position were determined using samtools depth (v1.19.2). Single nucleotide polymorphisms (SNPs) were called using Freebayes (v1.3.8) and subjected to quality filtering with bcftools (v1.21), which excluded variants supported by a read depth of less than 30×. A consensus genome sequence was then generated from the reference sequence incorporating the high-confidence SNPs using the bcftools consensus module. The final outputs comprised the assembled consensus sequence and a filtered SNP dataset. Phylogenetic analysis The nucleotide sequences of the VP1 gene and the complete genomes from a total of 68 EV-D68 viruses, including the newly identified isolates, were aligned using MAFFT (v7.487). Two maximum-likelihood phylogenetic trees were constructed with IQ-TREE2 (version 2.3.6) using the general time-reversible (GTR) model and 1000 bootstrap replicates. The resulting phylogenetic trees were visualized using iTOL (Interactive Tree of Life, https://itol.embl.de/). Both trees were rooted with the Fermon strain collected in 1962. Virus identification The clinical EV-D68 strains were inoculated onto rhabdomyosarcoma (RD) cells to monitor cytopathic effects (CPEs) and characterized by transmission electron microscopy for morphology. Following cell resuscitation and passaging, the virus was inoculated into the cell line at multiplicities of infection (MOI) of 0.1, 1, 2, 5, and 10. CPE was monitored at 24 and 48 h of post-inoculation. Sequence analysis To investigate the potential antigenic variation of the enterovirus isolates in this study, we focused on the structure of the major antigenic protein, VP1. It is known that the surface-exposed BC, DE, and GH loops of the VP1 protein constitute the primary antigenic s recognized by neutralizing antibodies. We performed a sequence alignment of the VP1 sequences obtained in this study with those of reference strains. By manually inspecting the sequence of the VP1 protein, we systematically scanned every amino acid residue including within the BC, DE, and GH loops. Statistical analysis Descriptive analysis was conducted to examine the clinical characteristics of children positive for EV-D68. Continuous variables with normal distribution were presented as mean ± standard deviation, and those with non-normal distributions as median and interquartile range (IQR: 25th- 75th percentiles). Categorical variables were summarized as frequencies and percentages. All analyses were performed using Stata/MP 17.0. Results EV-D68 detection Of the 2,842 patients, 2.5% (719/2,842) were diagnosed with severe CAP. The nucleic acid of EV-D68 was detected in 21 samples collected from 21 patients, including 6 NPS, 8 sputum samples, and 7 BALF specimens. The median age the 21 EV-D68 positive patients was 5.0 years (IQR: 3.6-8.0 years), and the average hospitalization duration was 6.0 days (IQR: 5.0–8.0 days). The overall detection rate was 0.7% (21/2842). The annual detection rates from 2021 to 2024 were 0.0% (0/576), 1.2% (6/479), 0.2% (1/508), and 1.1% (14/1279), respectively. Elevated detection rates were observed in 2022 and 2024 compared to the other years. Most EV-D68 cases occurred in autumn (13 cases, 1.3% of seasonal samples), with no cases in winter. Among the 21 positive cases, all were 11 years old or younger. EV-D68 was detected in 9 severe CAP cases, with rates of 1.2% (1/83) in 2022 and 2.1% (8/381) in 2024. In 2024, 8 cases of severe pneumonia were identified, with 7 hospitalized at Children’s Hospital of Fudan University in Shanghai and one in Jiangsu Province. BALF samples were collected from those 7 patients in Shanghai, and a sputum sample was obtained from a patient in Jiangsu Province. Co-detection with other respiratory viruses occurred in 85.7% (18/21) of EV-D68 positive samples, and 4 samples also tested positive for bacterial or fungal pathogens (Table 1). Table 1Characteristics of 2,842 pediatric patients and 21 EV-D68 positive patients.CharacteristicsN (%)Age, median (IQR), yrYear (n)2021202220232024Severepneumonia719 (25.3)5.4 (2.0, 8.0)6183194381Non-severepneumonia2123 (74.7)1.2 (0.2, 5.4)51539631489821 EV-D68 positive patients Severepneumonia9 (42.9)7.0 (4.3, 8.0)0108 Non-severepneumonia12 (57.1)5.0 (1.6, 7.5)0516 Viralencephalitis1 (4.8)7.00001 Viremia1 (4.8)10.00001 ICUadmission3 (14.3)9.0 (3.6, 10.0)0102 Generalward18 (85.7)5.0 (3.2, 7.0)05112 Mechanicalventilation1 (4.8)3.60001Symptoms Fever15 (71.4)5.0 (3.6, 7.0)0519 Cough21 (100.0)5.0 (3.6, 8.0)06114 Wheeze8 (38.1)7.5 (4.3, 9.0)0305 Dyspnea3 (14.3)9.0 (3.6, 10.0)0102 Vomiting2 (9.5)5.5 (1.0, 10.0)0101 Rash1 (4.8)1.00100 Characteristics of 2,842 pediatric patients and 21 EV-D68 positive patients. Severe pneumonia Non-severe pneumonia Severe pneumonia Non-severe pneumonia Viral encephalitis ICU admission General ward Mechanical ventilation Description of severe EV-D68 cases in 2024 It is noteworthy that among the 9 severe pneumonia cases with positive EV-D68 test results, 8 occurred in 2024, including one case complicated by central nervous system infection and one case associated with viremia. A 7-year-old boy with severe CAP and EV-D68 positive was diagnosed with CNS infection (Fig. 1A). He presented with fever, headache, and lethargy but no seizures or altered consciousness. Cranial magnetic resonance imaging and electroencephalogram results were normal, and symptoms were consistent with meningitis. Two CSF samples collected 6 days apart showed white blood cell counts of 15 × 106/L and 20 × 106/L, respectively, suggesting viral meningitis. However, EV-D68 RNA was not detected in the CSF by real-time RT-PCR and metagenomic next-generation sequencing (mNGS). In another case, EV-D68 was identified in BALF, serum, and stool samples from a 10-year-old girl with respiratory failure requiring high-flow nasal cannula oxygen therapy (Fig. 1B). Real-time RT-PCR cycle threshold (Ct) values for EV-D68 in BALF, serum, and stool were 19.1, 36.8, and 39.0, respectively, and positive in mNGS. Lung CT imaging of two EV-D68-positive cases. (A) A 7-year-old boy with severe CAP and CNS infection showed consolidation and atelectasis in the lower lobe of his right lung (indicated by the arrow). (B) A 10-year-old girl with respiratory failure due to severe pneumonia and with EV-D68 viremia. Consolidation and atelectasis were observed in her lower lobe of the left lung (indicated by the arrow). Virus genotype The 21 virus strains were sequenced and submitted to GenBank (accession numbers PQ766584-PQ766593, PQ826414- PQ826419, PQ857715-PQ857717, PQ895337 and PV018745), and the dataset included 17 partial VP1 gene sequences and 4 complete genomes. The viral strains designated SH2024-25870 and SH2024-22990 have been assigned the VP1 gene sequences PQ766594 and PQ826413, and the complete genome sequences PV018745 and PQ895337, respectively. Figures 2 and 3 show phylogenetic analyses based on the VP1 gene of 20 strains and 4 strains using complete genomes. The complete genome sequences of our four strains are consistent with the phylogenetic lineages inferred from the VP1 sequence. Among the 21 strains, 15 belong to subclade B3 and 6 to subclade D3. B3 strains were detected in 2022 (6 strains), 2023 (1 strain), and 2024 (8 strains, including one form the 7-year-old boy). All 6 subclade D3 strains were detected in 2024, including one form the 10-year-old girl, with 5 from BALF samples of severe CAP patients and 1 from a sputum sample. Phylogenetic analysis of EV-D68 based on the VP1 gene. Sequences from this study uploaded to GenBank are marked with stars. The PQ857716 strain (subclade B3), containing only a partial VP1 sequence (604 bp), was excluded from the analysis. Phylogenetic analysis of EV-D68 based on the complete genome. Sequences from this study uploaded to GenBank are marked with stars. The genotyping results based on the complete gene sequences of the four strains were concordant with those based on the VP1. Virus identification On RD cells, the CPEs were evident 72 h post-inoculation, showing cell rounding, detachment, and death (Fig. 4A-B). Electron microscopy revealed non-enveloped spherical virus particles approximately 20 to 30 nm in diameter. Ultrastructural analysis confirmed features typical of enteroviruses (Fig. 4C-D). Characterization of EV-D68 strains in RD cells. (A) CPE in RD cells. (B) Normal cell morphology. (C) Virus particles observed by Transmission Electron Microscopy with negative staining. (D) Virus particles in ultrathin sections of RD cells. Sequence analysis The 20 EV-D68 VP1 sequences generated in this study exhibited 84.9–99.8% nucleotide identity and 90.9–100.0% amino acid (AA) identity. Comparison with the Fermon reference sequence (AY426531) revealed AA substitutions at 21 positions in the VP1 protein, primarily in the BC loop (8 substitutions), DE loop (5 substitutions), and GH loop (1 substitution) (Fig. 5). No distinct abnormalities were observed in the viral strain full sequences from severe cases, and comparative analysis of the sequences revealed no unique or suspicious mutation sites associated with virulence in this study. Compared with the Fermon strain, the newly determined virus strain was found to harbor amino acid substitutions at specific positions in the VP1 sequence, including H99D, K168E, and E169K. The subclade D3 strains exhibited notable amino acid differences relative to the subclade B3 strains at positions 95–99 within the BC-loop region. Amino acid sequence alignment of capsid protein VP1. Compared to the original Fermon strain, mutations in the newly identified EV-D68 occur mainly in the BC-loop, DE-loop, and GH-loop regions of VP1. Notably, these include point mutations such as H99D, K168E, and E169K. Discussion The newly prevalent EV-D68 differs in tissue tropism from the original strains16,17. Since 2014, global surveillance for EV-D68 have been established due to concerns about its re-emergence2,6. The appearance of new virus subclades may overcome herd immunity. During our 12-year surveillance of EV-D68 in hospitalized CAP children in China, we observed biennial peak prevalence in 2022 and 2024, similar to patterns seen in 2016 and 201810. Detection rates were 1.2% in 2022 and 1.1% in 2024, higher than 0% in 2021 and 0.2% in 2023, and the average detection rate was 0.20% between 2013 and 2020. Additionally, more cases were detected in autumn in Shanghai. We observed that positive cases were more common among preschool-age children, with an average age of 5.0 years (IQR: 3.6-8.0 years), compared to our previous report of an average age of 1.3 years (IQR: 0.4–5.2 years) from 2013 to 2020. This shift may be linked to strict COVID-19 control measures in China, which limited exposure to EV-D68 for younger children. In this study, 14.3% (3/21) of positive patients required ICU admission, including one on mechanical ventilation, indicating a potential increase in severity of respiratory infections caused by the current EV-D68 strains. In 2024, 7 strains EV-D68 strains were detected in BALF samples, compared to none in 2021 and 2023. We identified two pediatric cases: one with severe pneumonia and viremia, and another with severe pneumonia and viral meningitis. The CT images clearly reveal the extent of pulmonary inflammation in the patients. Subclade D3 and B3 of EV-D68 were detected in these cases. EV-D68 has been shown to induce viremia in both human and animal models, suggesting systemic infection potential18,19. EV-D68 primarily causes respiratory symptoms and is transmitted via the respiratory tract, unlike other HEV strains. Its optimal replication temperature of 33–37 °C contributes to this characteristic20. Recent studies indicate that emerging EV-D68 variants now replicate efficiently at 37 °C, correlating with deeper infections observed since 2014, including pulmonary and CNS involvement20–23. EV-D68 uses sialic acid receptors to infect respiratory epithelial cells, which are also present in gastrointestinal epithelial cells, potentially explaining simultaneous respiratory and gastrointestinal infections24. In this study, EV-D68 was detected in a stool sample from a child with severe pneumonia, likely due to viral shedding from respiratory tract infection or ingestion of respiratory secretions, or with gastrointestinal infection. Given that EV-D68 particles and RNA may not survive the entire digestive process, it is more plausible that the virus originated from digestive tract epithelial cells, possibly during viremia. Further research is needed to confirm whether subclade D3 strains can infect both tracts simultaneously. Globally, since 2014, the B3 subclade of EV-D68 has replaced D1 and B2, becoming the main cause of severe respiratory infections and acute flaccid myelitis in children5,8. In Shanghai, our previous study identified B3 emerging in children starting in 2014. Subclade D3 was first detected in 2019 and became the dominant subtype causing severe CAP in children by 2024, consistent with its association with severe pneumonia in adults in Spain in 20188. Notably, D3 prevalence increased among children with severe pneumonia in 2024, supporting concerns that D3 may become a globally predominant strain with enhanced pathogenicity. Based on prior analyses, we identified amino acid mutations in the VP1 sequence, specifically in the BC, DE, and C-terminal epitopes. Genome recombination may drive genome diversification. Among the 21 newly discovered virus sequences, changes were observed in the VP1 region, particularly in the BC, DE, and GH loops. These findings suggest that the major antigenic sites and receptor binding sites within the VP1 region represent critical determinants in the evolution of viral virulence and may be associated with alterations in the cell receptors utilized25,26. However, comparative analysis of the viral VP1 amino acid and nucleotide sequences in this study did not reveal any association between these individual mutation sites and clinical virulence. Whether these mutations are associated with functional changes remains to be confirmed, particularly when compared to recent clinical isolates from severe infections. Summary, subclade D3 of EV-D68 may cause severe pneumonia in children in China in 2024. This variant can cause severe respiratory and viremia. Close monitoring is needed due to the potential risk of severe infections. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Author contributions Jun Shen, Roujian Lu, and Wenjie Tan conceived the study and reviewed the final manuscript. Yuanyuan Zhou, Jiayi Lin, Xuejie Zhang, Yi Zhu, and Xiaohui Zou conducted the experiments, analyzed the results, and wrote the initial draft. Liting Wang, Linfeng Cao, Yingyan Shi, and Weibang Huo collected the samples and clinical data. All authors reviewed the final manuscript. Funding This study was supported by the National Key Research and Development Program of China (2022YFC2704901), the Nantong Municipal Key Medical Discipline, and the Beijing Natural Science Foundation (L242052). Data availability The datasets generated during the current study are available in the GenBank (accession numbers PQ766584-PQ766593, PQ826414- PQ826419, PQ857715-PQ857717, PQ895337 and PV018745). Competing interests The authors declare no competing interests. Ethics approval and consent to participate This study was conducted in accordance with in accordance with the Declaration of Helsinki. Ethical approval was obtained from the Children’s Hospital of Fudan University ethical committee (approval number: 2020 − 208 and 2024-33). Written informed consent was obtained from the legal guardians of the participants after explaining the study objectives. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Yuanyuan Zhou, Jiayi Lin and Xuejie Zhang contributed equally to this work. Contributor Information Wenjie Tan, Email: tanwj@ivdc.chinacdc.cn. Roujian Lu, Email: lurj@ivdc.chinacdc.cn. Jun Shen, Email: shenjun_@fudan.edu.cn. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement The datasets generated during the current study are available in the GenBank (accession numbers PQ766584-PQ766593, PQ826414- PQ826419, PQ857715-PQ857717, PQ895337 and PV018745).