In the current coronavirus disease 2019 (COVID-19) pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), children are less severely affected than adults.1–3 Children may even be infected with SARS-CoV-2 asymptomatically, and it remains unclear what drives (severity of) respiratory symptomatology in pediatric COVID-19. In seasonal infections with common human coronaviruses (HCoV) in childhood, increased severity is associated with younger age, chronic illness, and codetection of respiratory viruses known to cause severe respiratory tract infections.4,5 Until now, associations between bacterial coinfections and presence or severity of HCoV-associated respiratory symptoms have not been investigated. Understanding the determinants of the symptomatology of common HCoV infections may incite ideas for research into the determinants of COVID-19 severity in children. Therefore, we evaluated respiratory viruses and the nasopharyngeal bacterial microbiome in seasonal HCoV-associated lower respiratory tract infections (LRTI) in young children.6
PATIENTS AND METHODS
Details on study design, laboratory methods, and bioinformatics were previously published.6 In short, we enrolled 154 children between 4 weeks and 5 years old, who were admitted to a Dutch teaching hospital for a physician-diagnosed LRTI between 2013 and 2015, and 307 community-dwelling age-, gender-, and season-matched healthy controls. Children with an underlying medical condition were excluded from the study. The study protocol was approved by the Dutch National Ethics Committee. Written informed parental consent was obtained from all participants. Nasopharyngeal swabs were obtained from cases at hospital admission and from asymptomatic controls during a home visit and were tested for the presence of respiratory viruses, including HCoV serotypes OC43, HKU1, NL63, and 229E using qualitative multiplex Real Time-PCR (RespiFinder® SMARTfast 22, Maastricht, The Netherlands). From the same samples, bacterial DNA was isolated, and the V4 hypervariable region of the 16S rRNA gene was amplified. Amplicon pools were sequenced using the Illumina MiSeq platform (San Diego, CA) and preprocessed in our previously described bioinformatics pipeline.7 Operational taxonomic units that could be confidently annotated to 2 different species were assigned a double species annotation. Quantitative polymerase chain reaction targeting the autolysin (lytA) gene was used to identify Streptococcus pneumoniae,8 and multiplex real-time polymerase chain reaction was used to identify Haemophilus influenzae, Staphylococcus aureus, and Moraxella catarrhalis (FTD Bacterial pneumonia_CAP, Fast Track Diagnostics, Esch-sur-Alzette, Luxembourg). In the present post hoc analysis, we investigated whether HCoV-associated LRTI was related to the presence of other respiratory viruses and bacterial abundances. To this end, cases and controls were unmatched and grouped according to detection of HCoV in nasopharyngeal swabs. HCoV-negative controls were not included in this post hoc analysis. HCoV-positive LRTI cases were compared with HCoV-positive healthy controls and with HCoV-negative LRTI cases. A P value below 0.050 was considered statistically significant. Differences in age and gender were assessed with t tests and χ2 tests. Differences in viral (co)detection and number of viruses were assessed using age-adjusted logistic and linear regression where required. Relative abundances of the 8 most highly abundant operational taxonomic units were compared using Wilcoxon rank-sum tests. Statistical analysis was performed in R version 3.6.1.
Viral data were available from 150 LRTI cases and 303 healthy controls. HCoV was detected at a similar rate in 24 (16.0%) LRTI cases and in 58 (19.1%) healthy controls. HCoV disease and carriage were highly seasonal, with 23 HCoV-positive LRTI cases (95.8%) and 49 HCoV-positive healthy controls (84.5%) detected between December and March. Age was comparable between HCoV-positive LRTI cases (mean 16.96 months [SD, 15.91]) and HCoV-positive healthy controls (mean 16.39 months [SD 14.94]) as well as HCoV-negative LRTI cases (mean 17.50 months [SD 14.99]). Gender was also not significantly different between HCoV-positive LRTI cases (12 [50.0%] female), HCoV-positive healthy controls (24 [41.4%] female), and HCoV-negative LRTI cases (45 [35.7%] female).
Codetection of other respiratory viruses with HCoV occurred significantly more often in LRTI cases (23, 95.8%) than in healthy controls (40, 69.0%; P = 0.026). The total number of viruses detected in HCoV-positive LRTI cases amounted to maximum 4 (mean 2.54 [SD 0.88]) and was significantly higher than in HCoV-positive healthy controls (mean 2.09 [SD 0.94]; P = 0.043), as well as HCoV-negative LRTI cases (mean 1.37 [SD 0.66]; P < 0.001). Respiratory syncytial virus (RSV) was significantly more often detected in HCoV-positive LRTI cases (17, 70.8%) than in HCoV-positive healthy controls (2, 3.4%; P < 0.001) and in HCoV-negative LRTI cases (54, 42.9%; P = 0.014). By contrast, rhinovirus was less often detected in HCoV-positive LRTI cases (9, 37.5%) compared with HCoV-positive healthy controls (36, 62.1%; P = 0.042) and HCoV-negative LRTI cases (65, 51.6%), although the latter difference was not significant (P = 0.17). Other respiratory viruses were relatively rare and not significantly different between groups (Fig. 1A).
Bacterial data were available from 24 HCoV-positive cases, 123 HCoV-negative cases, and 58 HCoV-positive controls. HCoV-positive LRTI cases had significantly higher relative abundances of Haemophilus influenzae/haemolyticus (P = 0.011) and Corynebacterium macginleyi/accolens (P = 0.036), as well as significantly lower relative abundances of Moraxella catarrhalis/nonliquefaciens (P = 0.011) and Moraxella lincolnii (P = 0.002) compared with HCoV-positive healthy controls (Fig. 1B). Notably, the relative abundance of H. influenzae/haemolyticus was even higher in HCoV-positive LRTI cases (median 32.7% [interquartile range (IQR) 0.3%–78.3%]) than in HCoV-negative LRTI cases (median 6.1% [IQR 0.1%–49.4%]), but this difference was not significant (P = 0.080). The relative abundance of respiratory pathogen S. pneumoniae was higher in HCoV-negative LRTI cases (median 5.6% [IQR 0.5%–23.8%]) than in HCoV-positive LRTI cases (median 1.4% [IQR 0.2%–14.8%]), but this difference was not significant, either (P = 0.32). The relative abundances of Staphylococcus aureus/epidermidis, Corynebacterium pseudodiphtheriticum/propinquum, and Dolosigranulum pigrum were also not significantly different between HCoV-positive LRTI cases, HCoV-positive healthy controls, and HCoV-negative LRTI cases.
In summary, we have shown that a higher rate of RSV codetection and a higher abundance of H. influenzae/haemolyticus distinguished children with HCoV-associated LRTI from asymptomatic HCoV carriers and from children with a non-HCoV-associated LRTI. Both RSV and H. influenzae/haemolyticus are common respiratory pathogens, but their increased presence in children with HCoV-associated LRTI compared with children with non-HCoV-associated LRTI is notable. In line with previous findings, single HCoV infections were very rare.9 By contrast, we observed a lower rate of rhinovirus codetection and lower relative abundances of Moraxella species in children with an HCoV-associated LRTI compared with asymptomatic HCoV carriers, which is in line with previous observations of cases and controls within this cohort.6 Our findings may imply that when children are colonized with HCoV in their upper respiratory tract, the local viral, and bacterial community (ie, the microbiota) may either aggravate or provide resistance against symptomatic infections. An alternative explanation could be that HCoV may be an innocent bystander during respiratory infections in children, while known pathogens like RSV and H. influenzae are more causally involved.
Although our results cannot be directly extrapolated to COVID-19, bacterial and viral codetection may also be associated with (severity of) respiratory symptomatology in SARS-CoV-2 infections. One study demonstrated that 20% of COVID-19 patients were coinfected with common respiratory viruses, including RSV and rhinovirus.10 Furthermore, a case series of 20 children hospitalized for COVID-19 found a higher rate of (mostly viral) coinfections in 40% of cases.11 Moreover, they observed an elevated procalcitonin in 80% of cases, which is suggestive for bacterial (co)infections.12 Together, these findings suggested that particularly in children, viral and bacterial coinfections may aggravate the clinical presentation of COVID-19. In contrast, viruses and bacteria that commonly inhabit the respiratory tract of young children are also known to interfere with viral invasion and limit symptomatic illness,13,14 but this has not yet been studied for SARS-CoV-2.
Mechanisms by which the residential flora, including viruses and bacteria, may influence the severity of viral infections remain to be elucidated but may include direct interspecies interactions and indirect effects of the microbiota on host immunity and metabolism. Future studies in, for example, animal models and human cell lines or challenge models are required to unravel the exact pathogenetic mechanisms. In conclusion, we show that further studies unraveling interactions between HCoV, including SARS-CoV-2, and the respiratory but also gut microbiota may be crucial to understand why the clinical presentation of HCoV-associated respiratory infections, including COVID-19, is highly variable, especially in children.
1. Dong Y, Mo X, Hu Y, et al. Epidemiology of COVID-19 among children in China. Pediatrics. 2020;145:e20200702.
2. Lu X, Zhang L, Du H, et al. SARS-CoV-2 infection in children. N Engl J Med. 2020;382:1172–1173.
3. Castagnoli R, Votto M, Licari A, et al. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection in children and adolescents: a systematic review. JAMA Pediatr. 2020;2:882–889.
4. Varghese L, Zachariah P, Vargas C, et al. Epidemiology and clinical features of human coronaviruses in the pediatric population. J Pediatric Infect Dis Soc. 2018;7:151–158.
5. Heimdal I, Moe N, Krokstad S, et al. Human coronavirus in hospitalized children with respiratory tract infections: a 9-Year Population-Based Study from Norway. J Infect Dis. 2019;219:1198–1206.
6. Man WH, van Houten MA, Mérelle ME, et al. Bacterial and viral respiratory tract microbiota and host characteristics in children with lower respiratory tract infections: a matched case-control study. Lancet Respir Med. 2019;7:417–426.
7. Bosch AATM, de Steenhuijsen Piters WAA, van Houten MA, et al. Maturation of the infant respiratory microbiota, environmental drivers, and health consequences. A Prospective Cohort Study. Am J Respir Crit Care Med. 2017;196:1582–1590.
8. Da M, Carvalho GS, Tondella ML, et al. Evaluation and improvement of real-time PCR assays targeting lytA, ply, and psaA genes for detection of pneumococcal DNA. J Clin Microbiol. 2007;45:2460–2466.
9. Calvo C, Alcolea S, Casas I, et al. A 14-year prospective study of human coronavirus infections in hospitalized children. Pediatr Infect Dis J. 2020;39:653–657.
10. Kim D, Quinn J, Pinsky B, et al. Rates of co-infection between SARS-CoV-2 and other respiratory pathogens. JAMA. 2020;24:4–5.
11. Xia W, Shao J, Guo Y, et al. Clinical and CT features in pediatric patients with COVID-19 infection: different points from adults. Pediatr Pulmonol. 2020;55:1169–1174.
12. Simon L, Gauvin F, Amre DK, et al. Serum procalcitonin and C-reactive protein levels as markers of bacterial infection: a systematic review and meta-analysis. Clin Infect Dis. 2004;39:206–217.
13. Nickbakhsh S, Mair C, Matthews L, et al. Virus-virus interactions impact the population dynamics of influenza and the common cold. Proc Natl Acad Sci U S A. 2019;116:27142–27150.
14. Man WH, de Steenhuijsen Piters WA, Bogaert D. The microbiota of the respiratory tract: gatekeeper to respiratory health. Nat Rev Microbiol. 2017;15:259–270.