Human Bocavirus Infection in Children With Respiratory Tract Disease

Abstract

Background: 

Human bocavirus (HBoV) is a ubiquitous, newly described member of the Parvoviridae family frequently detected in the respiratory tract of children, but only few reports provide data proving the link between HBoV and respiratory tract disease (RTD).

Objectives: 

To evaluate the incidence of HBoV infection in children with RTD; to analyze the clinical features of HBoV infection; and to clinically compare HBoV, respiratory syncytial virus (RSV), and human metapneumovirus (HMPV) infections.

Study Design: 

A prospective 1-year study was conducted in children <5 years of age hospitalized with RTD and in asymptomatic control children.

Results: 

Human bocavirus was detected in 55 (10.8%) of the 507 children tested and in none of the 68 asymptomatic control children (P = 0.01). About 80% of these infections occurred between November and March. Coinfection with another virus was observed in 22 (40%) of the HBoV-positive children. HBoV viral load was significantly higher in samples from children with HBoV monoinfection than in those with coinfection. Subsequent detection of HBoV more than 2 months after the initial detection could be documented in 3 children. Clinical features associated with HBoV infection were similar to those observed with either RSV or HMPV infections, but HBoV infections were less severe than RSV infections.

Conclusions: 

The difference observed in HBoV prevalence between children with RTD and controls provides support for a role of this virus in RTD. The frequent associations of HBoV with other respiratory viruses might be explained by the persistence of HBoV in the respiratory tract. The significance of HBoV viral load in nasopharyngeal secretions as a marker of pathogenicity merits further investigation.

Respiratory tract disease (RTD) is one of the leading causes of morbidity and mortality in young children, and a great variety of viruses, including respiratory syncytial virus (RSV), human metapneumovirus (HMPV), influenza viruses, parainfluenza viruses, picornaviruses (rhinoviruses or enteroviruses), adenoviruses, and coronaviruses account for the majority of RTD in this age group. Recently, a new virus, a member of the Parvoviridae family, has been identified in respiratory tract secretions from children with RTD. This virus has been classified in the Bocavirus genus and named human bocavirus (HBoV).¹ The ubiquitous distribution of HBoV was subsequently demonstrated by detection of the virus in children worldwide.^2–9 HBoV has been mainly detected in children at a frequency ranging from about 2% to 18%.^2,3,10–12 Despite the high rate of coinfection with other respiratory pathogens and the few studies including asymptomatic controls, there is increasing evidence for a causal role of HBoV in respiratory illness.

In the present study, the clinical features of children <5 years of age hospitalized with RTD and positive for HBoV detection were analyzed and compared with those of children positive for either RSV or HMPV detection.

MATERIALS AND METHODS

Patients and Specimens

Patients included in this study were children <5 years of age, admitted for RTD to a pediatric unit of the University Hospital of Montpellier (France) between November 1, 2006 and October 31, 2007. Five hundred fifty nasopharyngeal aspirates were obtained from 507 children with RTD. There were 279 (55%) males and 228 (45%) females (M/F sex ratio, 1:2) with a median age of 5.2 months (range, 10 days–60 months). Asymptomatic controls were 68 children without any respiratory sign or fever; most of these control children attended the pediatric emergency unit for traumatic injuries, or were seen either for elective surgery or for routine well child visits. There were 38 males and 30 females (M/F sex ratio, 1:3) with a median age of 7.8 months (range, 15 days–59 months). These differences between children with RTD and controls were not significant. Nasopharyngeal aspirates were collected from patients and controls and immediately tested for respiratory viruses.

Virologic Investigation

RSV; influenza viruses types A and B; parainfluenza viruses types 1, 2, and 3; and adenoviruses were detected using direct immunofluorescence assays. Samples were tested for HPMV antigen by a commercial enzyme immunoassay (human metapneumovirus EIA kit, Biotrin International, Lyon, France). Samples were also inoculated onto MRC5 cell monolayers for virus isolation. Aliquots of each sample were stored at −80°C.

DNA Extraction and Real-time PCR Amplification

DNA was extracted from 200 μL of nasopharyngeal aspirate using the QIAmp Viral DNA Mini Kit (Qiagen, Courtaboeuf, France), according to the manufacturer's instructions.

The nucleic acid extracts were tested for HBoV DNA by real-time PCR as previously described.¹⁰ Standard curves for the quantification of HBoV were constructed using serial dilutions of a plasmid containing an overlapping fragment of the PCR target in the NP-1 region of the HBoV genome. Sensitivity of the assay was estimated to be 2.7 log HBoV DNA copies per milliliter of sample. Total DNA level in HBoV-positive sample extracts was measured using the LightCycler control DNA kit (Roche Diagnostics, Meylan, France). HBoV DNA levels were expressed as log HBoV DNA copies/ng of total DNA as previously described.¹⁰

Data Analysis

Categorical variables between groups were compared with the χ² test or the 2-tailed Fisher exact test when the expected cell size was below 5. For continuous variables, comparisons were based on the nonparametric Mann-Whitney U test. All statistical tests were done by Statgraphics Plus software version 5.1 (Manugistics, Rockville, MD). P < 0.05 was considered to be significant. Among children monoinfected with HBoV, clinical data were stratified according to viral load. A threshold for defining a high viral load was arbitrarily set at 5.0 log HBoV DNA copies/ng of DNA.

RESULTS

Detection of Respiratory Viruses

A respiratory virus was identified in 261 (51.4%) children with RTD. Direct immunofluorescence assays, enzyme immunoassay, and viral culture documented that 165 (32.5%) children were infected with RSV; 50 (9.9%), with HMPV; 11 (2.2%), with a picornavirus; 8 (1.6%), with an influenza virus type A; 6 (1.2%), with an adenovirus; and 4 (0.8%), with a parainfluenza virus type 3. Coronaviruses were not tested, whereas the frequency of picornaviruses was obviously underestimated, cell culture being less sensitive than molecular assays.

HBoV Detection and Viral Load

Real-time PCR for HBoV DNA was positive in 59 (10.7%) samples from 55 (10.8%) children and in no sample from asymptomatic controls (P = 0.01). Dual infections were detected in 22 (40%) of the HBoV-positive children. Twenty of these coinfections involved RSV, one involved HMPV, and yet another one involved parainfluenza virus type 3.

HBoV viral load ranged from <2.7 to 9.0 log HBoV DNA copies/mL of sample. Because nasopharyngeal aspirates are expected to yield varying quantities of cellular DNA, HBoV viral load was expressed as the log number of HBoV DNA copies/ng of total DNA. The median HBoV viral load was 2.48 log HBoV DNA copies/ng of DNA (range, 1.12–7.95 log HBoV DNA copies/ng of DNA). HBoV viral load measured in samples containing only HBoV (median, 3.98 log HBoV DNA copies/ng of DNA; range, 1.12–7.95 log HBoV DNA copies/ng of DNA) was significantly higher (P = 0.007) than that measured in samples containing a copathogen (median, 1.84 log HBoV DNA copies/ng of DNA; range, 1.12–6.38 log HBoV DNA copies/ng of DNA). A viral load >3.0 log HBoV DNA/ng of DNA was observed in 21 of 33 (63.6%) samples containing HBoV alone and in 3 of 22 (13.6%) samples containing a copathogen (P < 0.001). However, in children monoinfected with HBoV, a relationship was not observed between HBoV viral load and either clinical symptoms or severity criteria (Table 1).

TABLE 1: Clinical Characteristics of Children Infected With Human Bocavirus (HBoV)

Multiple samples were available several weeks to several months apart for 6 HBoV-positive children, and 3 of them showed persistent HBoV PCR positivity (Fig. 1). For 2 children, HBoV shedding was detected in samples collected more than 2 months apart, whereas for the third child, HBoV DNA could be detected in samples collected in February and the following September.

Figure 1.: HBoV viral load changes over time in nasopharyngeal secretions from 6 children.

Clinical Features Associated With HBoV Detection

The clinical characteristics at presentation of children positive for HBoV detection are shown in Table 1. Cough, rhinorrhea, dyspnea, and wheezing were the most frequently reported clinical signs. Gastrointestinal disorder was observed in 7 (12.7%) children.

The final diagnosis is shown in Table 1. Typical lower respiratory tract infections like bronchiolitis, pneumonia, and asthma were the leading diagnoses, whereas upper respiratory pathologies (pharyngitis, tracheitis, laryngotracheitis) or acute otitis media were less common.

Preexisting medical conditions were reported for 12 HBoV-infected children (21.8%). Furthermore, more than half of the HBoV-infected children had previously experienced an episode of asthma or bronchiolitis.

In children with HBoV monoinfection, stratification of clinical data on the basis of HBoV viral load (Table 1) did not demonstrate significant differences between children with a viral load above 5.0 log HBoV DNA copies/ng of DNA and those with a lower viral load.

Comparison of HBoV, RSV, and HMPV Infections

Age Distribution.

As shown in Table 2, the median age of the children positive with HBoV was significantly higher than that of the children infected with RSV or HMPV. RSV infections occurred mainly in children <12 months of age. Indeed, RSV was detected in 141 of 361 (39.1%) children <12 months of age and in 24 of 146 (16.4%) children >12 months of age (P < 0.001). On the contrary, there was no significant difference in the frequency of either HBoV or HMPV among children less than or more than 12 months of age: HBoV was detected in 35 (9.7%) children <12 months of age and in 20 (13.7%) children >12 months of age (P = 0.19), whereas HMPV was detected in 40 (11.1%) children <12 months of age and in 10 (6.8%) children >12 months of age (P = 0.15).

TABLE 2: Comparison of Features Associated With HBoV, HMPV, and RSV, Excluding Co-Infections

Seasonal Distribution.

The seasonal distribution of HBoV, HMPV, and RSV is presented in Figure 2, https://links.lww.com/A529. During winter months of 2006 and 2007, the outbreak of RSV infection peaked in December and January. HBoV was detected all the year round but the highest incidence was observed during winter months (80% between November and March). A similar pattern was observed for HMPV, except for its absence during the summer months. All the HBoV coinfections, mainly involving RSV, were detected during the RSV outbreak, in December and January.

Clinical Features.

As compared with children infected with RSV, those infected with HBoV had a shorter duration of hospital stay, a lower frequency of chest radiographic abnormalities, less frequent oxygen supplementation, and a shorter duration of oxygen therapy (Table 2). Considering the same criteria, no obvious difference was observed between children coinfected with RSV and HBoV and those infected with RSV alone.

Clinical symptoms and final diagnostics associated with HBoV, HMPV, or RSV detections appeared very similar, except that HMPV was more frequently associated with upper respiratory infections. Underlying conditions as well as history of prematurity were observed at similar frequencies for the 3 viruses. Previous bronchiolitis or asthma episodes were more frequently observed among HBoV-infected children (Table 2).

DISCUSSION

In our study, HBoV was detected in 10.8% of children with RTD aged <5 years. Focusing on this limited age group is supported by previous studies, showing that median age of HBoV-infected children is <2 years of age,^5,8–10 and by a seroepidemiologic study in Japan, showing that nearly all individuals have serologic evidence of previous HBoV infection by the age of 5 years.¹³ Our results, in agreement with previous reports, indicate that HBoV is one of the leading viruses detected in respiratory samples of young children with RTD.^2,6,7,12 In our study, HBoV was the second most frequent virus identified after RSV, with a similar rate of detection to that of HMPV. By contrast, HBoV was not detected in children free of respiratory symptoms. This observation, which is in line with the few available studies that included a control population and reported none or few HBoV-positive samples in asymptomatic children,^14–16 argues for a role of HBoV in RTD. However, a recent study reported a 43% prevalence of HBoV in asymptomatic control children versus a 13.8% prevalence in symptomatic children,¹⁷ but this unusual observation needs to be confirmed.

As previously reported, HBoV was frequently associated with another respiratory pathogen. However, it was observed that HBoV viral loads were significantly higher in samples collected from children positive for HBoV alone than in those collected from coinfected children. This observation was also reported by Allander et al¹⁴ who suggested that high viral loads are indicative of a causative role of HBoV in RTD, whereas low viral loads indicate asymptomatic shedding. In the present study, however, there were no significant differences in clinical symptoms, final diagnosis, and disease severity between children with HBoV DNA levels higher or lower than 5 log HBoV DNA copies/ng of DNA. The small number of children in each arm might explain the lack of significance of the observed difference. Another explanation is that all the children were included on the basis of symptomatic illness and this represents a bias for analyzing clinical data. On the other hand, because coronaviruses and rhinoviruses were not detected, some coinfections were probably unrecognized. Furthermore, competition or interference between viruses in the respiratory tract could alternatively explain the lower viral loads observed in coinfections. More extensive studies are therefore needed to investigate whether a high viral load is indicative of a causative role of HBoV in RTD.

A persistent shedding of HBoV was observed in 3 children. In 2 cases, a large decrease in viral load was observed after the initial detection, suggesting that a persistent low-level shedding of HBoV in respiratory secretions may follow the HBoV primary infection. In the third patient, the observation of similar viral load levels at a 7-month interval, may suggest a protracted viral shedding, but the hypothesis of a reinfection cannot be ruled out. A long-lasting shedding, which is a feature shared by other human parvoviruses,¹⁸ might explain both the absence of obvious seasonality for the HBoV spread and the high frequency of codetection of HBoV with other respiratory pathogens. However, in the hypothesis of a long-term asymptomatic HBoV carriage, some asymptomatic control children should have been positive for HBoV DNA detection as reported in a previous study.¹⁵ The absence of HBoV DNA detection among the asymptomatic children could result from the small sample size of our control population. Furthermore, because DNA yield was not assayed in asymptomatic samples, we cannot rule out the occurrence of false negative results because of insufficient amounts of DNA in some of these samples.

The incidence of HBoV infection was lower than that of both HMPV and RSV infections in children aged <1 year, and children infected with HBoV were older than those infected with HMPV or RSV. This difference is surprising if one considers a similar respiratory route of transmission for the 3 viruses. There are currently no obvious data related to the mode of transmission of HBoV and, even if a respiratory route may be suspected, an alternative oro-fecal route cannot be ruled out because several studies reported the presence of HBoV in stool samples.^19–21 Thus, the later acquisition of HBoV in comparison with other respiratory viruses might be partly explained by an oral route of transmission, at the time of introducing various foods in the infant's diet.

The medical history of more frequent previous episodes of bronchiolitis or asthma in HBoV-positive children is noteworthy. This observation could be explained by the higher age of these children, which could imply that some had a previous respiratory tract infection and a subsequent HBoV-associated RTD. On the other hand, HBoV was detected in 27.3% of the children admitted for asthma exacerbation. This observation is in agreement with previous studies reporting a possible association of HBoV with asthma.^14,22

Taken together, our results indicate that clinical signs associated with HBoV detection cannot be distinguished from those related to other respiratory viruses. HBoV infection seems to be less severe than RSV infection. On the other hand, the lack of difference in severity between children coinfected with HBoV and RSV and those infected with RSV alone suggests that coinfection with HBoV does not worsen RSV infection.