Human Bocavirus: Passenger or Pathogen in Acute Respiratory Tract Infections?

Abstract

Summary: Human bocavirus (HBoV) is a newly identified virus tentatively assigned to the family Parvoviridae, subfamily Parvovirinae, genus Bocavirus. HBoV was first described in 2005 and has since been detected in respiratory tract secretions worldwide. Herein we review the literature on HBoV and discuss the biology and potential clinical impact of this virus. Most studies have been PCR based and performed on patients with acute respiratory symptoms, from whom HBoV was detected in 2 to 19% of the samples. HBoV-positive samples have been derived mainly from infants and young children. HBoV DNA has also been detected in the blood of patients with respiratory tract infection and in fecal samples of patients with diarrhea with or without concomitant respiratory symptoms. A characteristic feature of HBoV studies is the high frequency of coinciding detections, or codetections, with other viruses. Available data nevertheless indicate a statistical association between HBoV and acute respiratory tract disease. We present a model incorporating these somewhat contradictory findings and suggest that primary HBoV infection causes respiratory tract symptoms which can be followed by prolonged low-level virus shedding in the respiratory tract. Detection of the virus in this phase will be facilitated by other infections, either simply via increased sample cell count or via reactivation of HBoV, leading to an increased detection frequency of HBoV during other virus infections. We conclude that the majority of available HBoV studies are limited by the sole use of PCR diagnostics on respiratory tract secretions, addressing virus prevalence but not disease association. The ability to detect primary infection through the development of improved diagnostic methods will be of great importance for future studies seeking to assign a role for HBoV in causing respiratory illnesses.

INTRODUCTION

Human bocavirus (HBoV) was first described in September 2005 by Tobias Allander and coworkers at the Karolinska University Hospital, Stockholm, Sweden (2). The finding resulted from the intensive investigation of two chronologically distinct pools of nasopharyngeal aspirates (NPAs) obtained from mostly pediatric patients with suspected acute respiratory tract infections (ARTIs). Thus, HBoV joined the ranks of viruses colloquially termed “respiratory viruses,” which are detected predominantly in patients with infection of the respiratory tract. A random PCR-cloning-sequencing approach was employed. In the original study, HBoV DNA was subsequently identified in 17 out of 540 NPAs (3.1%). Coincident detection of another virus occurred for three patients (17.6% of positive patients), including two instances of human respiratory syncytial virus (RSV) and one detection of human adenovirus (AdV) (2). No other viruses were detected in 14 of 17 HBoV-positive symptomatic patients, at a glance suggesting a high occurrence of sole detections. However, common respiratory viruses were not sought using PCR, and several other known respiratory pathogens, including human rhinoviruses (HRVs) and human coronaviruses (HCoVs), were not sought by any means. The fact that HBoV was not detected randomly in the material but was detected significantly more often in the absence of other detected viruses nevertheless suggested that HBoV may be a causative agent of previously unexplained respiratory tract disease. All 14 children without codetection had been admitted to an inpatient medical treatment center after presenting with symptoms of cough and fever during the previous 1 to 4 days.

Since the first report, the worldwide presence of HBoV in children with ARTI has been confirmed by over 40 studies. However, most published studies describe virus prevalence and were not designed to address the issue of disease association. Thus, to date, the evidence for an association between HBoV and respiratory tract disease is incomplete. The many prevalence studies have found an unusually high number of coinfections where HBoV occurs simultaneously with other viruses, making the association of HBoV with disease more complex. Moreover, Koch's revised postulates cannot be applied to HBoV, since neither a method for HBoV culture nor an animal model of infection has been established (26). This situation applies to most newly identified viruses, including HCoV-NL63 (72) and HCoV-HKU1 (82), polyomaviruses KI (2) and WU (29), and the HRVs, HRV-QPM, NAT-001, and NAT-045 (51). Many newly identified viruses have probably remained undetected until now exactly because of their inability to replicate in vitro under standard conditions and may therefore never fulfill Koch's postulates. Well-designed clinical studies will be needed to confirm the causative role of a virus for a disease, as proposed by Fredericks and Relman (26). A number of these studies will be required before a causative role for HBoV in respiratory tract disease can be established.

This review includes all HBoV studies published online up to early 2007, includes prospectively collected data from the winter season from 2005 to 2006 (74), and discusses virological and clinical aspects of this newly identified virus.

TAXONOMY

HBoV is a putative member of the family Parvoviridae (subfamily Parvovirinae, genus Bocavirus). Until the identification of HBoV, human parvovirus B19 (B19V) (subfamily Parvovirinae, genus Erythrovirus) had been the only human pathogen in the family. B19V is the causative agent of fifth disease, hydrops fetalis (53), and aplastic anemia, in particular in patients with preexisting hematopoietic disease (20, 21, 23, 30, 42, 76). HBoV was classified as a bocavirus based on genomic structure and amino acid sequence similarity shared with the namesake members of the genus, bovine parvovirus (13) and canine minute virus (9, 65). Consequently, the first human member of this virus genus has been provisionally termed human bocavirus (2, 52). Other human parvoviruses of interest include the newly identified human parvovirus 4 (PARV4), which is currently unclassified, and the five current species of human adeno-associated viruses (AAV), which reside in the genus Dependovirus. PARV4 is detected in human plasma used in the manufacture of medicinal products, but no pathogenic roles have as yet been demonstrated (28). The AAVs rely on another “helper” virus to replicate, usually an AdV, but in their absence AAVs integrate in a site-specific manner into the human genome.

The International Committee on Virus Taxonomy defines species within the genus Bocavirus as probably antigenically distinct, with natural infection confined to a single host species. Species are <95% related by nonstructural gene DNA sequence. To date, studies of HBoV have addressed only the molecular criterion. This is indeed the main criterion, since there have been no comparative antigenic studies among any of the species of this genus. Although humans are assumed to be the natural host of HBoV, it should be noted that no studies have investigated lower animals for the presence of HBoV.

BIOLOGY OF BOCAVIRUS

The members of the family Parvoviridae are small, nonenveloped viruses. They have isometric nucleocapsids with diameters of 18 to 26 nm that contain a single molecule of linear, negative-sense or positive-sense, single-stranded DNA. The complete genome has a length of approximately 4,000 to 6,000 nucleotides (nt) (1, 2).

The complete genome length of HBoV has not been determined, but at least 5,299 nt were identified in one of the reference strains. It can be assumed from the genome structure of other parvoviruses that the genomic DNA of bocavirus is flanked by hairpin structures. These structures cannot be deciphered by sequencing methods alone; thus, the complete sequence of the entire genome will not be available until the flanking structures are elucidated (2).

The genome contains three proposed open reading frames, with two open reading frames putatively encoding the nonstructural proteins (NS1 and NP-1) and one encoding two viral capsid proteins, VP1 and VP2; the VP2 sequence is nested within VP1 (2). The function of the HBoV NS1 protein is unknown, but one could speculate on its role in HBoV DNA replication, since the related protein in other parvoviruses is likely to be involved in the binding and hydrolysis of nucleoside triphosphates and to have helicase activity (85). NP-1 is absent from other parvoviruses and its function is unknown (2, 65). Phylogenetic analyses have shown that two genetically distinct but very closely related clusters cocirculate in the United States, Sweden, Canada, and France (7, 35). As expected, the deduced coding sequence for the structural proteins VP1 and VP2 from different isolates showed high variability compared to the coding sequences for the nonstructural NS1 and NP-1 proteins, reflecting the more immunogenic character of the virion-associated proteins.

The cells hosting HBoV replication have not been determined. Parvoviruses in general require proliferating cells for their replication. Studies of animal bocaviruses suggest infection of respiratory and gut epithelium and lymphatic organs (18, 19). HBoV DNA is present in patients with ARTI and sometimes reaches high copy numbers in respiratory tract secretions, consistent with infection of the respiratory epithelium (1). HBoV DNA has also been detected in the sera of patients with ARTI and in the feces of patients with ARTI and/or gastroenteritis, suggesting the possibility that a range of cells may support HBoV replication in vivo (1, 27, 50, 56, 73).

Until recently, HBoV infection could be identified only by the detection of its nucleotide sequence. In a recent report, Brieu et al. described parvovirus-like particles in HBoV DNA-positive NPAs by electron microscopy (12). Confirmation of these findings by immunoelectron microscopy with a (hitherto unavailable) HBoV-specific antibody would support the assumption that HBoV DNA, at least at high copy numbers, is virion associated. Antibodies elicited in humans against HBoV structural proteins have also recently been demonstrated (22, 33).

LABORATORY DIAGNOSIS

To date, the detection of HBoV has been performed predominantly on NPAs and swabs and has been possible only with PCR-based methods, since no virus culture method, animal model of infection, or antibody preparation for antigen detection has been available (2). No comparative studies to identify an optimal sampling site have been reported, and the selection of a sampling site is also hindered by a lack of knowledge about the site of HBoV replication. Specimen handling and storage is infrequently detailed in the published studies. However, the most frequent approach is certainly immediate or batched column-based nucleic acid extraction and PCR testing of convenient populations by use of patient material that has been previously stored after routine microbial testing. Oligonucleotide sequences from PCR methods described to date are summarized in Table 1, but as of yet no comparative studies have identified an optimal gene target or oligonucleotide set(s). For diagnostic purposes, more-conserved genetic regions are preferred; thus, primers directed toward the NS1 gene should yield the most robust assays. However, the limited genetic variability of HBoV allows multiple suitable PCR targets, including the frequently targeted NP1 gene. The use of real-time PCR serves to minimize the risk of amplicon carryover contamination, reduce the result turnaround time, and add an extra layer of specificity (if an oligoprobe-based approach is employed) and can prove less costly overall. Different research groups have described real-time PCR assays that permit some degree of quantification of the viral load in respiratory secretions (1, 37, 45, 62). Since there is no way to standardize respiratory tract specimen collection, respiratory virus quantification by PCR is better described as being semiquantitative (47).

TABLE 1.

Overview of those published protocols describing oligonucleotide sequences used for PCR detection of HBoV

Study (reference)	Detection method	Primer name (sequence)	Probe sequence and labelinga	Target region
Allander et al., 2007 (1)	Real-time PCR (LightCycler)	Boca-forward (GGAAGAGACACTGGCAGACAA); Boca-reverse (GGGTGTTCCTGATGATATGAGC)	FAM-CTGCGGCTCCTGCTCCTGTGAT-TAMRA	NP-1
Allander et al., 2005 (2)	PCR	188F (GAGCTCTGTAAGTACTATTAC); 542R (CTCTGTGTTGACTGAATACAG)	None	NP-1
Arden et al., 2006 (5); Sloots et al., 2006 (68)	PCR	HBoV 01.2 (TATGGCCAAGGCAATCGTCCAAG); HBoV 02.2 (GCCGCGTGAACATGAGAAACAGA)	None	NP1
Bastien et al., 2006 (7)	PCR	VP1/VP2F (GCAAACCCATCACTCTCAATGC); VP1/VP2R (GCTCTCTCCTCCCAGTGACAT)	None	VP1/2
Bastien et al., 2007 (8)	PCR	VP/VP2-1017F (GTGACCACCAAGTACTTAGAACTGG); VP/VP2-1020R (GCTCTCTCCTCCCAGTGACAT)	None	VP1/2
Foulonge et al., 2006 (24)	Real-time PCR (LightCycler)	BocaRT1 (CGAAGATGAGCTCAGGGAAT); BocaRT2 (GCTGATTGGGTGTTCCTGAT)	FAM-CACAGGAGCAGGAGCCGCAG-TAMRA	NP1
	Sequencing	BocaSEQ1 (AAAATGAACTAGCAGATCTTGATG); BocaSEQ4 (GAACTTGTAAGCAGAAGCAAAA); BocaSEQ2 (GTCTGGTTTCCTTTGTATAGGAGT); BocaSEQ3 ( GACCCAACTCCTATACAAAGGAAAC)	None
Kesebir et al., 2006 (35)	PCR	GGACCACAGTCATCAGACCCACTACCATCGGGCTG	None	VP1
Kleines et al., 2007 (37)	Real-time PCR (LightCycler)	Same as that found in the work of Allander et al., 2005 (2)	GGAAGAGACACTGGCAGACAAC-fluorescein; LC-Red 640-CATCACAGGAGCAGGAGCCG	NP1
Kupfer et al., 2006 (39); Simon et al., 2007 (66)	PCR	OS1 (CCCAAGAAACGTCGTCTAAC); OS2 (GTGTTGACTGAATACAGTGT)	None	NP1
Lin et al., 2007 (43)	Real-time PCR (TaqMan)	Forward primer (AGCTTTTGTTGATTCAAGGCTATAATC); reverse primer (TGTTTCCCGAATTGTTTGTTCA)	FAM-TCTAGCCGTTGGTCACGCCCTGTG-TAMRA	NS
Lu et al., 2006 (45)	Real-time PCR (iCycler iQ real-time detection system [Bio-Rad])	Primer, fwd (TGCAGACAACGCYTAGTTGTTT); Primer, rev (CTGTCCCGCCCAAGATACA)	FAM-CCAGGATTGGGTGGAACCTGCAAA-Black_Hole_Quencher	NS1
		Primer, fwd (AGAGGCTCGGGCTCATATCA); Primer, rev (AGAGGCTCGGGCTCATATCA)	FAM-AGGAACACCCAATCARCCACCTATCGTCT-Black_Hole_Quencher	2478-2497; 2558-2537
Manning et al., 2006 (49)	Nested PCR	Outer sense primer (TATGGGTGTGTTAATCATTTGAAYA); outer antisense primer (GTAGATATCGTGRTTRGTKGATAT); inner sense primer (AACAAAGGATTTGTWTTYAATGAYTG); inner antisense primer (CCCAAGATACACTTTGCWKGTTCCACCC)	None	NS
		Outer primers used were CCAGCAAGTCCTCCAAACTCACCTGC and GGAGCTTCAGGATTGGAAGCTCTGTG; inner primers followed the sequence of the primer sequences 188F and 542R	None	NP
Qu et al., 2007 (59)	Real-time PCR (TaqMan)	TAATGACTGCAGACAACGCCTAG; TGTCCCGCCCAAGATACACT	FAM-TTCCACCCAATCCTGGT-MGB
Neske et al., 2007 (56)	Real-time PCR (Light Cycler) and phylogenetic analysis	BoV2466a (TGGACTCCCTTTTCTTTTGTAGGA) targeting NP1 2466-2443 (real-time PCR); BoV3885s (ACAATGACCTCACAGCTGGCGT) (phylogenetic analysis); BoV4287s (CAGCCAGCACAGGCAGAATT) (phylogenetic analysis); BoV4456a (TCCAAATCCTGCAGCACCTGTG) (phylogenetic analysis); BoV4939a (TGCAGTATGTCTTCTTTCTGGACG) (phylogenetic analysis)	FAM-TGAGCTCAGGGAATATGAAAGACAAGCATCG-TAMRA	NP1; VP2
Regamey et al., 2007 (60)	Real-time PCR (TaqMan)	Primer forward (CACTGGCAGACAACTCATCACA); primer reverse (GATATGAGCCCGAGCCTCTCT)	AGCAGGAGCCGCAGCCCGA	NS1
Schenk et al., 2007 (62)	Real-time PCR (LightCycler)	HBoV-UP (AGGAGCAGGAGCCGCAGCC); HBoV-DP (CAGTGCAAGACGATAGGTGGC)	HBoV-P: FAM-ATGAGCCCGAGCCTCT-TAMRA	NP1
Smuts and Hardie, 2006 (69)	Seminested PCR	NP-1 s1 (TAACTGCTCCAGCAAGTCCTCCA); NP-1 as1 (GGAAGCTCTGTGTTGACTGAAT); NP-1 as1 and NP-1 s2 (CTCACCTGCGAGCTCTGTAAGTA)	None	NP1
		VP s1 (GCACTTCTGTATCAGATGCCTT); VP as1 (CGTGGTATGTAGGCGTGTAG); VP s2 (CTTAGAACTGGTGAGAGCACTG)	None	VP1/2

^aFAM, 6-carboxyfluorescein; TAMRA, 6-carboxytetramethylrhodamine; MGB, minor groove binder (Applied Biosystems).

Nevertheless, recent results obtained by “quantitative” real-time PCR suggest that high HBoV viral loads (defined as >10⁴ copies/ml) are frequently present as the sole viral finding for children admitted for acute wheezing, while the clinical significance of low to moderate viral loads is uncertain. High viral load in the respiratory tract was frequently associated with the detection of HBoV DNA in the blood (1). HBoV DNAemia declined after the resolution of symptoms, suggesting that high viral load may represent primary infection. Fry et al. (27) compared patients with pneumonia to healthy control subjects. Quantitative results were not reported in detail but, importantly, they found that the healthy controls exclusively had low numbers of HBoV DNA copies in respiratory specimens, while both high and low HBoV DNA loads were found among those cases with pneumonia. Low to moderate viral load is relatively commonplace among studies (62, 38, 45), suggesting that a large proportion of HBoV detection by PCR may represent virus shedding of uncertain clinical relevance. Thus, standardized diagnostic tests that can more accurately identify primary infections, which are more likely the true symptomatic cases, are a top priority for future HBoV research. The predictive value of high virus copy numbers as well as the diagnostic value of PCR testing of blood samples should be further investigated, but if its limitations are kept in mind, quantitative PCR is a useful interim tool for understanding the course of HBoV infection. Preliminary serological studies have recently been published (22, 33) and it is expected that serology will be a very useful diagnostic addition to the study of HBoV infection, as it has been for B19V (83).

HBoV AND RESPIRATORY TRACT DISEASE

The fact that HBoV is prevalent in samples from patients with ARTI does not guarantee a causative role for the symptoms. For example, many viruses are transmitted via the respiratory tract without triggering substantial respiratory symptoms. Establishing the causative role of a specific agent in disease is a thorough process requiring multiple studies (26). It is particularly difficult to do so in the absence of culture systems and/or animal models, which is a problem common to studies of other newly identified viruses. Nevertheless, the pathogenicity of newly identified HCoV or HRV strains has not been a major issue of debate, most likely because of their genetic relatedness to other established respiratory pathogens. With HBoV, the situation is different and confounded by several facts. First, HBoV is not related to a known human respiratory pathogen. Second, HBoV may be shed persistently, since other human parvoviruses (B19V, PARV4, and the AAVs) have the capacity for asymptomatic persistence (41, 44, 50). Third, HBoV is commonly detected in association with other respiratory viruses which have an established pathogenic potential. These facts raise the possibility that HBoV detection in respiratory tract samples simply reflects asymptomatic persistence or prolonged viral shedding. Another hypothesis is that HBoV is reactivated or produces a transient asymptomatic superinfection that is triggered by the presence of another replicating respiratory agent. A few studies providing data relevant to these issues have been published (1, 2, 27, 33, 35, 48, 49).

The first description of HBoV by Allander et al. (2) did, as mentioned earlier, include a study indicating a statistical association between the detection of HBoV on one hand and the patient suffering from otherwise unexplained ARTI on the other hand. Diagnostics for other viruses was incomplete. However, simple asymptomatic shedding of HBoV would still not result in the observed skewed distribution of HBoV findings.

Manning et al. (49) identified HBoV in stored respiratory tract samples and compared the frequencies of reported symptoms associated with each of the different agents sought. Of the 21 HBoV-positive patients, 20 children had symptoms of ARTI versus 1 asymptomatic child, a situation similar to that for RSV but different from what was found for AdVs. The most common clinical diagnosis was “lower RTI,” made for 15 patients (72%).

To date, five studies have included control groups of asymptomatic children (1, 27, 35, 48, 49). All studies found highly significant prevalence differences between individuals with ARTI and asymptomatic individuals. Kesebir and coworkers detected HBoV DNA from 22 of 425 NPAs of symptomatic children, while none of the 96 asymptomatic children tested positive for HBoV (35). Allander et al. found no positives among 64 asymptomatic children compared to 49 of 259 (19%) positive samples from children with acute wheezing (1). Unfortunately, in these studies the type of specimen varies between both groups, with an unknown impact on the efficiency of collection, nucleic acid extraction, and PCR sensitivity. In the study of Allander et al., the asymptomatic children were slightly older than the children with ARTI (1).

Maggi et al. tested 335 children with ARTI and 51 asymptomatic children (30 healthy infants and 21 preadolescent healthy children) and detected 4.5% positives among the nasal swabs of ARTI cases and no HBoV in nasal swabs of asymptomatic children (48). However, the main weakness of this study is that cases and controls were sampled during different years, which may have falsely lowered the detection of virus in the asymptomatic group.

One recently published study by Fry et al. (27), performed in Thailand but coordinated by the Centers for Disease Control and Prevention (Atlanta, GA), included nasopharyngeal swabs from 1,168 patients with community-acquired pneumonia, 512 patients with “influenza-like illness,” and 280 asymptomatic individuals. HBoV DNA was detected in only 3 asymptomatic individuals (1%), whereas 20 out of 512 (3.9%) outpatients with “influenza-like illness” (according to the WHO definition) and 53 (4.5%) out of 1,168 hospitalized patients with the diagnosis “pneumonia” tested positive for HBoV (27). For children aged from 0 to 4 years, the HBoV prevalence was 12% among pneumonia cases and 2% among asymptomatic controls. To date, this is the only study from which all groups have been sampled in the same way, making these data more robust. Among hospitalized children of <5 years of age with the diagnosis “pneumonia,” HBoV was the third most commonly detected virus (12%). Higher prevalence could be confirmed only for RSV and HRV (27). Viral loads of this study were reported separately (45). Most positive samples among all groups had low viral loads but, importantly, high loads were seen only among cases and not among asymptomatic controls.

A main concern regarding studies comparing virus prevalences in respiratory tract samples of symptomatic and asymptomatic individuals is the risk for bias related to respiratory tract sampling. An inflammatory process, regardless of cause, will produce a cell-rich mucoid secretion, easily available for sampling, while asymptomatic individuals have very little nasopharyngeal secretion at all. Thus, the detection of, e.g., an intracellular persisting virus could very well be enhanced by an inflammatory process regardless of its cause. The main alternative hypothesis to HBoV being a pathogen is that the virus is persisting or being shed for long periods from the respiratory tract at copy numbers near the lower limit of PCR detection. Because of these possibilities, comparisons of prevalence among symptomatic versus asymptomatic subjects must be interpreted with great care unless viral loads are reported.

It is also possible to establish a statistical association between HBoV and disease without using asymptomatic controls. Allander et al. (1) studied patients hospitalized for acute wheezing in Finland and found that the occurrence of HBoV in blood was linked in time with an acute infectious episode and normally disappeared after recovery. In another statistical analysis of the data, HBoV-positive patients with and without other pathogens detected in the respiratory tract were compared. HBoV was significantly more prevalent in patients where no other virus explaining the symptoms was detected. Interestingly, only the cases with high HBoV loads showed this association. Thus, in two ways, internal symptomatic controls could be used to support a statistical association between HBoV and disease in this study. Results were highly significant and at the same time 76% of HBoV-cases were codetections with other viruses, showing that frequent codetections are not necessarily an argument against disease association. The study suggested that high-load and viremic HBoV infection is associated with respiratory tract symptoms, while detection of a low viral load in the nasopharynx alone has uncertain relevance. It was hypothesized that these two entities represent primary infection and persistence, respectively, each accounting for approximately half of the HBoV findings in this particular material. This hypothesis has recently been confirmed by applying serology to the same material (33). Further studies are needed in order to determine the length of possible viral shedding or persistence. Regamey et al. detected HBoV DNA in one patient's respiratory specimen 3 weeks after the acute phase of infection (60).

In summary, several studies have found a statistical association between HBoV and acute respiratory symptoms, in a way that is consistent with a causal role. However, accurately establishing a causal relationship will require further studies, since current data also indicate that HBoV does not have a causal role for many of the ARTI cases in which it is detected. The diagnostic value in the individual case of detecting HBoV DNA in the respiratory tract therefore remains unclear.

Possible Role of Coinfections

While the main hypothesis explaining the frequently observed HBoV codetections involves some kind of innocuous persistence or prolonged shedding, a possible role for HBoV as a true copathogen remains uncertain and uninvestigated. Its frequent presence alongside other viruses cannot be questioned (Table 2) . The results of our University of Bonn study reported a codetection frequency of 36%. Such high percentages have been reported by most studies which have looked for coinfections, with codetection frequencies of 18% to 90% being reported (2, 27). Manning and coworkers detected one or more additional viruses in 43% (23/53) of HBoV-positive samples (49). The overall frequency of codetection among HBoV-negative samples that were positive for other viral pathogens in the same study was 17% (47/271) (P < 0.001). One explanation for the wide range of results is the nonstandardized diagnostic panel common to published studies. In addition, differences in test sensitivity have to be considered, in particular in light of the high proportion of low-load infections (14, 77). Codetection of another virus with HBoV, usually when the latter is at low viral load, occurs frequently from patients with ARTI, but HBoV is still rare in asymptomatic individuals. One hypothesis for this is that detection of innocuous HBoV shedding is enhanced by airway inflammation caused by another virus, as discussed above. Another is that HBoV is involved in the pathogenesis and, in some way, the aggravation of symptoms, so that it is frequently observed in hospitalized patients. Yet another possibility is that HBoV is a helper virus which aids other viruses or itself requires the aid of another ongoing infection for activation or reactivation of replication. There are currently no data defining a mechanism by which HBoV could be described as either a pathogen or a passenger. To date, it remains uncertain whether codetection with any respiratory viruses results in more-serious clinical outcomes. This is not a question unique to HBoV infections but an important facet of ARTIs in general that must be addressed in the future, perhaps with the aid of animal models.

TABLE 2.

Basic data from 23 analyzed studies

Study (reference)	Study region	No. of NPAs testeda	No. (%) of HBoV-positive samples	Median age (range) (mo)b	% Hospitalized	No. (%) of patients with relevant comorbiditiesg	Patients with coinfection (%)	Other viral copathogens (no.; %)
Allander et al., 2005 (2)	Stockholm, Sweden	540	17 (3.1)	HBoV pos only, 13.5 (8-48)	—d	9 (64)	18	RSV (2/17; 12%); AdV (1/17; 6%)
Allander et al., 2007 (1)	Turku, Finland	259	49 (19)	HBoV pos only (n = 12), 15.6 (9.6-38.4)	—	NA	76	Rhinovirus, enterovirus (43%); AdV (16%); RSV (14%)
Arnold et al., 2006 (6)	San Diego, CA	1,474	82 (5.6)	12.0 (10 days to 16 yr)	57	21 (31; 19% prematurity)	12	RSV (9/82; 11%)
Bastien et al., 2006 (7)	Saskatchewan, Canada	1,209	18 (1.5)	138.0 (9 mo-60 yr)	50	NA	*f	*
Choi et al., 2006 (14)	Seoul, South Korea	515	58 (11.3)	NA	—	†e	38	AdV (7/58; 12%); RSV (5/58; 9%); HMPV (5/58; 9%); Parainfluenza viurs 3 (3/58; 5%)
Chung et al., 2006 (15)	Seoul, South Korea	336 (225 “virus negative”; 111 “virus positive”)	27 (8); 17/225 (7.5) of the “virus negative”	14 (1-69); 15 (1-83) of the “virus negative”	—	NA	37	RSV (5/27;19%); HMPV (4/27;15%); AdV (1/27;4%)
Foulongne et al., 2006 (24)	Montpellier, France	589	26 (4.4)	13.0 (4-43)	—	16 (62)	35	RSV (5/26; 19%); AdV (2/26; 8%); HMPV (2/26; 8%)
Fry et al., 2007 (27)	Nonthaburi, Thailand	792	53 (4.5)	NA (1 mo-≥65 yr)	—	NA	83	RSV (23%); human parainfluenza virus (23%); AdV (2%); influenza virus A/B (9%); rhinovirus (42%)
Kaplan et al., 2006 (34)	Amman, Jordan	312	57 (18.3)	8.0	—	NA	72	AdV (25/57; 44%); RSV (23/57; 40%); HMPV (1/57; 2%), influenza virus A (1/57; 2%)
Kleines et al., 2007 (37)	Aachen, Germany	94	12 (12.8)	7.8 (1-30)	—	5/12 (41)	42	RSV (42%)
Kesebir et al., 2006 (35)	New Haven, CT	425	22 (5.2)	12.5 (1-24)	75	12 (60)	5*	*
Lin et al., 2007 (43)	Zhejiang Province, China	257	7 (2.7)	“Infants and children”	—	NA	NA	NA
Ma et al., 2006 (46)	Sapporo, Japan	318	18 (5.7)	15.0 (9-31)	89	NA	‡h	‡
Maggi et al., 2007 (48)	Pisa, Italy	Total of 335: A (infants), n = 200; B (adults), n = 84; C (asmptom. children), n = 51	A, 9 (4.5); B, no HboV detected; C, no HboV detected	A, 16 ± 13	100%	1 (11) neurological	44	RSV (3/9; 33%); rhinovirus, influenza virus A, HMPV (1/9; 11%)
Manning et al., 2006 (49)	Edinburgh, Scotland	924i	4 (1.6)	1.0-2.3 yrc	0	0	43	Rhinovirus, RSV, parainfluenza virus, AdV
Monteny et al., 2007 (54)	Rotterdam, The Netherlands	257 children with fever	(4)	1.0-2.3 yrc	0	0	33	Rhinovirus, RSV, parainfluenza virus, AdV
Naghipour et al., 2007 (55)	Rasht, Guilan, Iran	261	21 (8.3)	13 ± 2	—	5 (24) asthma	33	RSV, AdV, or influenza virus A
Qu 2007 et al., (59)	Beijing, China	252	5 (5.5)	9 (3.4-11.3)	—	0	10	HCoV 229E
Regamey et al., 2006 (60)	Berne, Switzerland	112	5 (4.2)	NA	0	0	56	RSV, AdV, rhinovirus (2×), HCoV NL63
Sloots et al., 2006 (68)	Queensland, Australia	324	18 (5.6)	NA	—	NA	56	RSV (2/38; 5%); AdV (1/18; 6%); HMPV (1/18; 6%)
Smuts and Hardie, 2006 (69)	Cape Town, South Africa	341	38 (11.0)	<36 mo	—	NA	37	RSV (14/40; 35%); rhinovirus (4/40;10%); influenza virus (5/40;13%); HCoV (3/40;8%); AdV (1/40; 3%)
Vicente et al., 2007 (73)	San Sebastián, Spain	527 (stool)	48 (9.1)	<36 mo				40 HBoV-positive respiratory samples: 25 (62.5%) coinfections with other viruses (13 RSV, 3 rhinovirus, 3 influenza virus A, 2 HCoV-OC43, 1 AdV, 1 influenza virus B); 48 HBoV-positive fecal samples: 28 (58.3%) coinfections with another intestinal pathogenj
Völz et al., 2007 (74)	Bonn, Germany	389	11 (2.8)	9 (3 mo to 17 mo)	—	NA	36	RSV (3/11; 27%); norovirus (1/11; 9%)
Weissbrich et al., 2006 (77)	Würzburg, Germany	835	87 (10.3)	22 (18 days to 8 yr)	—	NA	39	RSV (14/87; 16%); influenza virus A (9/87; 10%); AdV (9/87; 10%); parainfluenza virus 1/2/3 (1/87; 1%); influenza virus B (1/87; 1%)

^aasmptom., asymptomatic.

^bpos, positive; NA, data not available.

^cThis study included children with fever from 3 months to 6 years of age.

^d-, this study included NPA samples from hospitalized patients.

^e†, this study included patients without predisposing risk factors, such as an underlying disease.

^f*, this study included NPA samples that had tested negative for influenza virus A/B, parainfluenza virus 1/2/3, AdV, and RSV.

^gRelevant comorbidities are those underlying diseases or conditions that may have, or are known to have, any influence on the clinical severity of a viral respiratory infection.

^h‡, this study included NPA samples that had tested negative for influenza virus A/B, RSV, and HMPV.

ⁱStudy included 924 NPA samples from 574 individuals.

^jSalmonella enterica serovar Enteritidis (n = 1), Campylobacter jejuni (n = 5), rotavirus (n = 14), norovirus (n = 7), C. jejuni and norovirus (n = 1).

EPIDEMIOLOGY

Reports suggest that HBoV has worldwide endemicity. It has been detected over several years in many countries, including Sweden, Australia, the United States, Japan, Germany, South Africa, Jordan, France, Canada, Iran, Spain, The Netherlands, Korea, Thailand, Switzerland, and China (1, 2, 4-7, 14, 15, 24, 25, 27, 34-37, 39, 43, 45, 46, 48-50, 52, 54, 55, 59, 60, 62, 66, 68, 69, 73, 74, 77).

Based on phylogenetic analysis of predicted amino acid alignments, HBoV exists worldwide as a single lineage composed of two subtly different genotypes, as shown in Fig. 1. The greatest variability was observed within the 285-bp portion of VP1/VP2 (Fig. 1C), whereas significantly lower genetic variability was seen for viral sequences within the NS1 and the NP-1 genes (Fig. 1A and B).

FIG. 1.

Phylogenetic analysis of HBoV. Phylogenetic trees are based on 61 partial NS1 genes (245 nt, corresponding to nt positions 1509 to 1753 in the ST1 isolate [accession number DQ000495]) (A), 167 partial NP-1 genes (242 nt, corresponding to nt positions 2340 to 2581 in the ST1 isolate) (B), and 133 partial VP1/VP2 genes (285 nt, corresponding to nt positions 4547 to 4831 in the ST1 isolate) (C). Sequences were aligned using the program CLUSTAL_X, version 1.83 (71). Phylogenetic relationships of the aligned sequences were inferred from the generated alignment by the neighbor-joining method (61). The reliability of the tree topology was evaluated by 500 replicates of bootstrap resampling (84). Phylogenetic trees were visualized using the TREEVIEW software tool (58). Trees show the HBoV sequences used for the analysis and their individual geographical origins. The numbers in parentheses indicate the numbers of isolates for the respective locations. For reasons of clarity, this figure does not include GenBank accession numbers. Detailed information on GenBank accession numbers of sequences used for the phylogenetic analysis is available upon request. (D) Phylogenetic placement of HBoV and other members of the genus Parvovirinae. Bootstrapped (n = 1,000) neighbor-joining tree based on 80% of the complete genomic nucleotide sequence. Bootstrap values (%) are indicated at each branching point. B19, erythrovirus B19; PTMPV, pig-tailed macaque parvovirus; RMPV, rhesus macaque parvovirus; SPV, simian parvovirus; ChPV, chipmunk parvovirus; BPV2 and BPV3, bovine parvovirus 2 and 3; GPV, goose parvovirus; MDPV, Muscovy duck parvovirus; AMDV, Aleutian mink disease virus; PPV, porcine parvovirus; CPV, canine parvovirus; RPV-1a, rat parvovirus-1a; KRPV, Kilham rat parvovirus; MPV1, mouse parvovirus 1; MVM, minute virus of mice; MVC, minute virus of canines.

Prevalence of HBoV

The proportion of respiratory specimens from symptomatic hospitalized children that contain HBoV sequences has ranged from 1.5% to 19% (1, 7) Most children infected with HBoV have been younger than 24 months (2, 15, 55, 59, 68, 69), but older children may also be affected (7, 14). For example, 4 of 27 (15%) positive specimens in the study of Chung et al. were obtained from children of >36 months of age (15). Thus, it seems reasonable to include older children into prospective surveillance studies. As expected, studies which included only hospitalized children and children with wheezing (14) presented a higher illness severity than those studies that analyzed respiratory specimens from outpatients as well (6, 59, 60). In order to determine a more realistic prevalence of HBoV in respiratory tract samples, future prospective studies should also include appropriately age-matched children and adults embodying a clinical description of “asymptomatic” as well as patients presenting with mild illness.

There are few systematic studies including adults, but available studies indicate a very low virus prevalence by PCR in the respiratory tract of adults (2, 7, 27, 49).

Recently, seroepidemiological data were published from Japan by Endo et al. (22). Anti-HBoV antibodies were detected in 145 of 204 (71.1%) serum samples from people aged from 0 month to 41 years from the Hokkaido Prefecture. The seroprevalence was lowest (5.6%) in the age group from 6 to 8 months and highest in the age groups older than 6 years (94.1 to 100%). The findings of high antibody prevalence and low virus prevalence among individuals older than 6 years are consistent with each other and suggest that there may be protective immunity after past infection. Positive antibody titers were also detected in the age group younger than 6 months, but this phenomenon is explained by the antibody transfer via the placenta to the fetus predominantly in the third trimester of pregnancy (22).

Seasonal Distribution of HBoV Detection

According to the literature, HBoV DNA-positive ARTIs occur in children across a range of months. The peak “respiratory season” varies from year to year (79). Therefore, it is not feasible to draw conclusions concerning the epidemiology of a newly identified virus based on snapshot analyses of single seasons or even multiple respiratory seasons. In accordance with the University of Bonn's data, most authors reporting from regions with temperate climates have observed a higher occurrence of HBoV detections during the winter and spring months (2, 69). Choi et al. (Korea 2000 to 2005) reported a relatively high occurrence of HBoV in the late spring and early summer. They did not reveal any obvious correlation to changes in the parallel RSV season (14). Maggi et al. from Italy could not confirm a seasonal distribution of the HBoV infections in their study of hospitalized infants with RTI (48), but they found significant differences between years, with no HBoV detected in any specimen from 2000 to 2002 (n = 43, including 30 specimens from symptomatic infants). The weakness of most retrospective studies is that more specimens are collected during the winter months, because that is the epidemic season for most viral RTIs. Both more-active sampling and enhanced detection of HBoV by other infections, as discussed above, could therefore lead to false observations of seasonal patterns. It must also be kept in mind that the numbers reported in most studies probably reflect a mix of incidence and carrier prevalence. The true incidence and seasonality of primary HBoV infection remain unknown.

Transmission

Nothing is known about the routes of HBoV transmission. Because of its sometimes very high copy numbers in respiratory tract secretions, aerosol and contact transmission are likely effective, as they are for other respiratory viruses. Hand-to-hand, hand-to-surface, and self-inoculation routes have certainly proven to be efficient steps in the transmission of the “common cold.” Since we know that HBoV DNA exists in some capacity within feces, the possibility of fecal-oral transmission must also be considered. Further studies should include more testing of stool samples for HBoV to confirm the extent and nature of virus DNA shedding and the capacity of the virus to survive disinfectants (10, 11, 17, 31, 38) and permit broader investigations of a possible role for HBoV as an enteric pathogen. So far there have been no studies on the tenacity of the virus or about the effect of commonly used hospital-grade disinfectants. Since other parvoviruses are known to be highly resistant to disinfectants (10, 11), such investigations will be important but will require an HBoV culture system or animal model of infection.

Kesebir and coworkers reported 3 infants (14%) of 22 with presumed nosocomial HBoV infection (35). The infected infants were 1, 4, and 6 months of age at the time their NPAs were sampled and had been hospitalized since birth. Two of the three patients had HBoV-positive NPAs within a period of 4 days and had been cared for by the same medical personnel on the same ward. Phylogenetic analysis of the two positives showed identical nucleotide sequences in both the NP1 and VP1/VP2 gene region; however, because of the low genetic variability of HBoV, the significance of such a finding should not be exaggerated. Notably, vertical transmission could not be excluded. Three of 12 HBoV-positive children reported in the study of Kleines et al. developed symptoms of ARTI after at least 4 weeks of hospitalization (37). Since the incubation period of HBoV infection is unknown, it is not possible to state that this was nosocomial transmission.

The presence of HBoV DNA in the blood combined with suspected persistence could have implications for transfusion medicine, since organs or blood products derived from acutely infected donors could be contaminated and serve as a source of infection (1, 59). However, unlike PARV4, HBoV was not detected in plasma pools (28).

CLINICAL OBSERVATIONS

While the role for HBoV in causing any symptoms remains unclear, studies of the symptoms reported for HBoV-positive patients nevertheless provide an important starting point. Besides some case reports (39, 62, 66), 23 research study publications were included in this review that contained data about symptoms and outcomes for and radiological findings and laboratory results from HBoV-positive hospitalized children (Tables 2 and 3) (1, 2, 6, 7, 14, 15, 24, 25, 27, 34, 35, 37, 43, 46, 49, 54, 55, 59, 60, 68, 69, 73, 77). We also added our data, which were collected in the winter of 2005/2006 (74).

TABLE 3.

Diagnosis at discharge from hospital (%)

Study (reference)	% Patients from indicated study with indicated diagnosis upon dischargea:
	Upper RTI	Croup	Bronchitis	Bronchiolitis	Pneumonia	Asthma exacerbation	Gastrointestinal symptoms	Febrile seizures
Allander et al., 2005 (2)	NA	NA	NA	NA	NA	NA	NA	NA
Allander et al., 2007 (1)	0b	0	25	67	75	8	0	0
Arnold et al., 2006 (6)	24	4	NA	26	24	24	16	NA
Bastien et al., 2006 (7)	22	NA	NA	11	17	NA	NA	NA
Choi et al., 2006 (14)	5	8	NA	25	56	11	NA	NA
Chung et al., 2006 (15)	24	0	76	0	0	0	12	NA
Foulongne et al., 2006 (24)	15	NA	NA	46	11	27	NA	NA
Kaplan et al., 2006 (34)	NA	NA	NA	NA	NA	NA	NA	NA
Kesebir et al., 2006 (35)	NA	NA	NA	NA	NA	NA	25	NA
Ma et al., 2006 (46)	—c	NA	44	11	33	5	NA	NA
Maggi et al., 2007 (48)	0	0	0	55	45	0	11	0
Manning et al., 2006 (49)	29	NA	NA	NA	NA	NA	NA	NA
Naghipour et al., 2007 (55)	0	0	14	NA	46	10	NA	NA
Qu et al., 2007 (59)	0	0	29	NA	62	NA	9	5d
Völz et al., 2007 (74)	9	9	45	3	64	NA	9	9
Weissbrich et al., 2006 (77)	40	NA	16	3	17	NA	NA	9

^aData are percentages for all investigated patients unless otherwise indicated. NA, data not available.

^bAllander et al. (1) reported the diagnosis of acute otitis media for 42% of HBoV-positive children with bronchial obstruction without viral coinfection.

^cStudy included only patients with lower RTIs.

^dQu et al. reported one patient (10 months of age) with an acute life-threatening event.

Limitations of Available Studies

Only a few studies of HBoV have collected clinical data prospectively (the University of Bonn study presented here [Germany] and the studies of Regamey et al. [Switzerland] [60], Monteny et al. [The Netherlands] [54], Allander et al. [Finland] [1], and Fry et al. [Thailand] [27]). In the remaining studies we cite, laboratory, clinical, and radiological findings have been acquired retrospectively, similar to many studies of human metapneumovirus (HMPV) infection (78, 81). Only nonstandardized, research-only, in-house PCR diagnostics have been employed to date. Because of the obligate use of PCR, one cannot truly talk about “infection”; rather, each HBoV DNA-positive specimen should be described as a virus “detection.”

Considering that prolonged shedding of HBoV or reactivation by other infections may account for a remarkable number of the HBoV detections discussed above, it is a severe limitation that diagnostic assays separating these cases from primary infections are not yet available. Most published studies have not taken this into consideration.

A lack of international consensus about the definition of certain respiratory diseases is another obstacle to accurately characterizing the clinical outcomes of HBoV infection, just as it is for any respiratory infection. There is no agreement about the definition of obstructive bronchitis, recurrent obstructive bronchitis in infants, bronchiolitis, bronchopneumonia, or lobar pneumonia (3, 70). Six out of the 23 analyzed studies used the diagnosis “bronchiolitis” (6, 7, 14, 24, 25, 46, 77). The percentages of “bronchiolitis” within the diagnostic spectrum ranged from 3.2% to 46% (24, 77). Two studies provided differing definitions (6, 14), whereas the remaining publications did not even comment on the clinical criteria. Only 7 of 23 studies (1, 2, 6, 27, 35, 37, 46) definitively stipulated a radiological confirmation of the clinical diagnosis “pneumonia.” Most studies did not make a distinction between (central) bronchopneumonia and segmental or lobar pneumonia.

Symptoms Presumably Associated with HBoV Infection

Clinical symptoms most frequently reported in individuals where HBoV is the only detected virus include cough, rhinorrhea, and fever, which are also the most common nonspecific symptoms leading to respiratory viral testing in children. The most common clinical diagnoses given to HBoV-positive patients, with or without coinfections, include upper RTI, bronchitis, bronchiolitis, pneumonia, and acute exacerbation of asthma. This clinical spectrum is in accordance with other viral ARTIs, similar to the situation with RSV infections (75) and with HMPV infections (78, 80). There are no described distinct clinical signs differentiating HBoV-positive infections from those ascribed to other viruses (2, 37). This could imply that HBoV indeed has a clinical picture similar to those seen for other ARTIs or simply that because of the mentioned diagnostic problems with HBoV many of the studied patients were actually suffering from other infections. Symptoms seem to persist for 1 to 2 weeks on average (range, 2 days to 3 weeks) (1, 60); Monteny et al. reported a prolonged course of fever (>7 days or recurring) in HBoV-infected patients (54). HBoV has also been detected in individuals with skin rash, although no causal association has been identified (6, 15, 54). Allander at al. reported a 42% incidence of acute otitis media in solely HBoV-positive patients (2). Except for this report, there are few data on bacterial coinfections.

The possibility that HBoV, like the closely related bovine and canine bocaviruses (18, 19), could cause gastroenteritis was raised in the first report on HBoV (2). Gastrointestinal symptoms have been described for up to 25% of all patients (6, 35, 54). Maggi et al. (48) detected HBoV DNA in stools of a 6-month-old boy followed for neurological problems who had presented with diarrhea and bronchopneumonia. Both respiratory and stool specimens were positive for HBoV and antigen negative for rotaviruses, AdVs, astroviruses, and calicivirus 1 and 2. Vicente at al. investigated the presence of HBoV DNA in 527 stool samples from ambulatory patients with gastroenteritis (<36 months of age) with or without additional respiratory symptoms (73). Of these, 48 (9.1%) were positive for HBoV DNA. Other enteric pathogens were found in 58% of all HBoV-positive fecal samples (Table 2). In contrast, Lee et al. detected HBoV DNA in only 0.8% of 942 hospitalized children with gastroenteritis (40). Neske and coworkers reported a high frequency of HBoV DNA in stool samples derived from children who were also positive for HBoV DNA in NPAs (56).

Chest Radiography Findings

In the University of Bonn study, the majority of HBoV-positive patients (10/11) showed symptoms severe enough for physicians dealing with the patients to perform a chest radiograph, and 8 of 10 (80%) patients with a chest radiograph had visible abnormalities (74). In these 11 patients no coinfections were observed. This high percentage of pathology is in accordance with the results of other clinical research groups, which found similar pathology in 43% to 83% of cases (2, 24, 25).

The most common diagnosis in this study was (central) bronchopneumonia; in 18% a segmental/lobar pneumonia was diagnosed (74). Cases of HBoV-positive central pneumonia as well as interstitial and lobar pneumonia, especially in newborns and infants, have been described.

HBoV and Acute Wheezing

ARTIs have frequently been detected in infants with recurrent airway obstruction (“wheezing”) and in older children and adults with asthma exacerbations (32, 63, 64, 80). In fact, the highest frequency of laboratory confirmations are described by studies of children with acute expiratory wheezing, usually attributed to viruses. Recently published clinical studies report asthma exacerbation as a clinical entity in up to 27% of HBoV-positive patients (6, 24, 25), but several of the analyzed studies did not explicitly exclude relevant viral pathogens such as RSV and the HRVs. Half of the patients in the original study by Allander and coworkers presented with asthma as an underlying disease (2). Allander et al. subsequently reported that 49 of 259 (19%) children hospitalized for acute wheezing in Finland were HBoV positive and speculated that this unusually high percentage could imply that wheezing is the main manifestation of HBoV infection. Fry et al. (27) found a statistical association between HBoV detection and reporting wheezing among patients with pneumonia in Thailand. Naghipour et al. found that 5 HBoV-infected patients (24%) had a history of asthma (55), while Maggi investigated respiratory specimens from 22 adult patients with acute asthma exacerbation and did not detect HBoV (48). Chung et al. investigated nasopharyngeal specimens from 231 children (1 month to 5 years of age) hospitalized with acute wheezing (16). Besides RSV (13.8%), HBoV was the most frequently detected virus (13.8%) in 5.6% without coinfection; HMPV and HCoV-NL63 were detected in 7.8% and 1.3% of wheezing children, respectively.

HBoV in Immunocompromised Patients

Several clinical research groups have reported HBoV-positive immunosuppressed/immunodeficient patients (6, 49, 69). Arnold and coworkers described two pediatric patients positive for HBoV after organ transplantation (6). Smuts and coworkers reported HBoV detections for eight human immunodeficiency virus-infected pediatric patients (69), and Manning and coworkers described two HBoV-positive immunosuppressed adult patients (49). Kupfer et al. have recently published the clinical case of a severe infection in a 28-year-old HBoV-positive female patient with malignant B-cell lymphoma (39). On admission, the patient had a pancytopenia, high fever, and clinical and radiological signs of pneumonia (reticulonodular infiltrates in the computed tomography scan of the thorax). Despite the application of antibiotics, antifungals, and the antiviral ganciclovir, fever continued for 14 days. HBoV DNA was detected retrospectively, suggesting HBoV as the sole potential pathogen in NPAs. Since unexplained pulmonary disease is common in this group of patients, the role of HBoV in causing the symptoms is unclear. However, in studies to date, most symptomatic adults positive for HBoV DNA have fallen into a category of immunosuppression. On the other hand, this may be the result of a selection bias, since these are the adult or elderly patients in which respiratory diagnostic specimens are taken in case of an infection. HBoV has not been identified in lymphoid tissue and in bone marrow and brain, respectively, from human immunodeficiency virus type 1-infected and uninfected adults upon autopsy (50).

Laboratory Results for HBoV-Infected Patients

Very few investigators have been able to document the course of markers of inflammation such as C-reactive protein (CRP) or white blood cell (WBC) count for HBoV-positive patients. None of the first 11 HBoV-positive children treated in our center (University of Bonn) in 2005/2006 fulfilled the laboratory criteria of a suspected bacterial coinfection (WBC, >15 × 10⁹/liter; CRP, >40 mg/liter) (67). The median WBC count was 11.3 × 10⁹/liter (range, 6.7 × 10⁹ to 16.7 × 10⁹) and the median CRP concentration was 12.5 mg/liter (range, <0.03 to 114) (74). Others reported median WBC and CRP concentrations of similar magnitudes (46). In Allander's recent report on 12 children with HBoV infection (no viral coinfection) (1), 9 displayed a radiologically confirmed pneumonia. The median WBC was 9.1 (6.3 to 16.3) × 10⁹/liter and the median CRP was 18 (0 to 78) mg/liter.

TREATMENT AND PREVENTION

The clinical impact of HBoV infection is uncertain, and to date there have been no studies on the benefits of particular therapeutic approaches for HBoV-infected individuals and no proposals for novel therapeutics. Prospective controlled studies evaluating specific treatment approaches for patients infected by HBoV will require analytically sensitive, specific, and clinically relevant diagnostic procedures. In our center (University of Bonn), the high frequency of radiographically confirmed pneumonia (70%) might explain the use of antimicrobial chemotherapy for 82% of HBoV-positive patients.

A rapid HBoV testing method capable of identifying clinically relevant cases could reduce the unjustified and generally ineffective use of antibiotics in these patients.

If HBoV turns out to contribute significantly to the disease burden on children, possibilities for vaccine development will likely be investigated, like it has been for RSV, HMPV, and parainfluenza viruses. To date, no such studies have been reported.

CONCLUSIONS AND FURTHER RESEARCH

Our current knowledge of HBoV infection suggests that the virus is sometimes a passenger and sometimes a pathogen in acute respiratory tract disease. A better understanding of the natural course of HBoV infection and an expanded arsenal of diagnostic tests capable of discriminating carriage from infection will be necessary before any clinical questions can be comprehensively addressed. Detection of HBoV DNA in blood and serological assays have shown promising preliminary results. To date, retrospective studies report a peak of HBoV detections during the first and second years of life. Some case reports have raised concerns about serious clinical outcomes among immunosuppressed individuals. To date there is neither a method for virus culture nor an animal model of infection, but hopefully the introduction of serological detection of specific antibodies will permit us some insight into the pathogenesis and natural course of HBoV infection.

Many additional questions cannot yet be answered by the studies that have been reported and should be addressed by future studies. How is HBoV transmitted, and is HBoV a causative agent of gastrointestinal diseases? Does the virus persist in the human host? Could coinfection with HBoV increase the severity of concurrent viral infections? What is the immune response to HBoV infection? Can HBoV cause exacerbations of asthma and chronic obstructive pulmonary disease?

HBoV might be one of the most recently identified respiratory viruses, but its nature has attracted as much interest and raised as many questions as many of its better-characterized relatives. After decades of research, the most widespread and frequent causes of human infections, the respiratory viruses, are still as confounding as ever.