Human Bocavirus 2 in Children, South Korea

To the Editor: In 2009, Kapoor et al. and Arthur et al. published reports on the prevalence of the newly identified parvovirus, human bocavirus 2 (HBoV-2), in fecal samples (1,2). HBoV-1 had been discovered in 2005 (3), and reports indicate its possible role in respiratory diseases such as upper respiratory tract infections, lower respiratory tract infections (LRTIs), and in exacerbation of asthma (4); in these diseases, the virus co-infects with other respiratory viruses (5). Systemic infection with HBoV-1 and possible association of this virus with other diseases such as gastroenteritis, Kawasaki disease, and hepatitis have been reported (6–8). We looked for HBoV-2 in clinical samples from children with various diseases, including acute LRTIs, Kawasaki disease, Henoch-Schonlein purpura, and hepatitis. 
 
During September 2008–January 2009, a total of 212 nasopharyngeal aspirates were collected from 212 children (median age 8 months, range 1–59 months) hospitalized with acute LRTIs at Sanggyepaik Hospital in Seoul, South Korea. Previously, during January 2002–June 2006, a total of 173 serum samples had been obtained from children (age range 1 month–15 years) with hepatitis (hepatitis B, 20 samples; hepatitis C, 11 samples; unknown hepatitis, 31 samples), Kawasaki disease (12 samples), and Henoch-Schonlein purpura (18 samples) and from healthy children (same age range, 81 samples) (9). The study was approved by the internal review board of Sanggyepaik Hospital. 
 
DNA was extracted from serum samples, and RNA and DNA were extracted from nasopharyngeal aspirates by using a QIAamp Viral RNA Mini Kit (QIAGEN, Hilden, Germany) and a QIAamp DNA Blood Mini Kit (QIAGEN GmbH), respectively. All nasopharyngeal aspirates were tested by PCR for common respiratory viruses such as respiratory syncytial virus, influenza viruses A and B, parainfluenza virus, and adenovirus, as described previously (10). PCRs to detect HBoV-1 were performed by using primers for the nonstructural (NS) 1 and nucleocapsid protein (NP) 1 genes, as described previously (10). Additional PCRs for rhinovirus, human metapneumovirus, human coronavirus (hCoV)-NL63, hCoV-OC43, hCoV-229E, hCoV HKU-1, WU polyomavirus, and KU polyomavirus were performed, as described, for HBoV-2–positive samples (10). 
 
HBoV-2 was detected by performing first-round PCR with primers based on the NS gene, HBoV2-sf1, and HBoV2-sr1. Second-round PCR was performed by using primers HBoV2-sf2 and HBoVsr2, as described previously (1). The PCR products were sequenced by using an ABI 3730 XL autoanalyzer (Applied Biosystems, Foster City, CA, USA). The nucleotide sequences were aligned by using BioEdit 7.0 (www.mbio.ncsu.edu/BioEdit/BioEdit.html) and presented in a topology tree, prepared by using MEGA 4.1 (www.megasoftware.net). 
 
Of the 212 samples tested, the following viruses were detected: human respiratory syncytial virus (in 124 [58.4%] samples), human rhinovirus (24 [11.3%]), influenza virus A (18 [8.4%]), adenovirus (10 [4.7%]), and parainfluenza virus (8 [3.7%]). HBoV-1 was not detected in the study population. HBoV-2 DNA was found in 5 (2.3%) of the 212 samples collected; all positive samples had been obtained in October 2008. The age range of the children with HBoV-2–positive samples was 4–34 months (median 24 months), and all were male. The diagnoses were bronchiolitis for 3 children and bronchopneumonia for 2. The most frequently codetected virus was human respiratory syncytial virus, found in 4 (80%) of 5 samples. One sample that was negative for respiratory syncytial virus and positive for HBoV-2 was negative for all other respiratory viruses. 
 
Nucleotide sequences were determined for the NS-1 gene, and phylogenetic analyses, which included HBoV-3, a new lineage designated by Arthur et al. (2), showed that the NS-1 gene was relatively well conserved and that there were 2 major groups of the virus, the UK strain and the Pakistan strain. HBoV-2 strains isolated from South Korea belonged to the HBoV-PK2255 ({"type":"entrez-nucleotide","attrs":{"text":"FJ170279","term_id":"213012746","term_text":"FJ170279"}}FJ170279) cluster (Figure). 
 
 
 
Figure 
 
Phylogenetic analysis of nonstructural (NS) 1 gene sequences from human bocavirus 2 strains from Korea (KR), United Kingdom (UK), and Pakistan (PK), presented on a topology tree prepared by using MEGA 3.1 (www.megasoftware.net). Nucleotide alignment of ... 
 
 
 
Recent studies have detected HBoV-1 in serum samples of children with Kawasaki disease and of an immunocompromised child with hepatitis (7,8). However, neither HBoV-1 nor HBoV-2 was detected in the 172 serum samples from 61 patients with hepatitis, 12 with Kawasaki disease, 18 with Henoch-Schonlein purpura, and 81 healthy children. 
 
The absence of HBoV-1 in the samples examined was unexpected because HBoV-1 was detected in >10% of 558 respiratory samples collected from a demographically similar study population during the winter 2 years earlier (10). Future studies, with larger populations and over longer periods, are needed to delineate seasonal variations between HBoV-1 and HBoV-2. 
 
We demonstrated HBoV-2 DNA in the respiratory tract secretions of children with acute LRTIs. In most positive samples, the virus was found in addition to other respiratory viruses. A limitation is that the study did not consider health control measures and other clinical disease such as gastroenteritis and was conducted for a short time. The role of HBoV-2 in LRTIs remains unclear; further studies are needed to clarify whether this virus is only shed from the respiratory tract or whether it replicates in the gastrointestinal tract.


Human Bocavirus 2 in Children, South Korea
To the Editor: In 2009, Kapoor et al. and Arthur et al. published reports on the prevalence of the newly identified parvovirus, human bocavirus 2 (HBoV-2), in fecal samples (1,2). HBoV-1 had been discovered in 2005 (3), and reports indicate its possible role in respiratory diseases such as upper respiratory tract infections, lower respiratory tract infections (LRTIs), and in exacerbation of asthma (4); in these diseases, the virus co-infects with other respiratory viruses (5). Systemic infection with HBoV-1 and possible association of this virus with other diseases such as gastroenteritis, Kawasaki disease, and hepatitis have been reported (6)(7)(8).
We looked for HBoV-2 in clinical samples from children with various diseases, including acute LRTIs, Kawasaki disease, Henoch-Schönlein purpura, and hepatitis.
During September 2008-January 2009, a total of 212 nasopharyngeal aspirates were collected from 212 children (median age 8 months, range 1-59 months) hospitalized with acute LRTIs at Sanggyepaik Hospital in Seoul, South Korea. Previously, during January 2002-June 2006, a total of 173 serum samples had been obtained from children (age range 1 month-15 years) with hepatitis (hepatitis B, 20 samples; hepatitis C, 11 samples; unknown hepatitis, 31 samples), Kawasaki disease (12 samples), and Henoch-Schönlein purpura (18 samples) and from healthy children (same age range, 81 samples) (9). The study was approved by the internal review board of Sanggyepaik Hospital.
DNA was extracted from serum samples, and RNA and DNA were extracted from nasopharyngeal aspirates by using a QIAamp Viral RNA Mini Kit (QIAGEN, Hilden, Germany) and a QIAamp DNA Blood Mini Kit (QIAGEN GmbH), respectively. All nasopharyngeal aspirates were tested by PCR for common respiratory viruses such as respiratory syncytial virus, influenza viruses A and B, parainfluenza virus, and adenovirus, as described previously (10). PCRs to detect HBoV-1 were performed by using primers for the nonstructural (NS) 1 and nucleocapsid protein (NP) 1 genes, as described previously (10). Additional PCRs for rhinovirus, human metapneumovirus, human coronavirus (hCoV)-NL63, hCoV-OC43, hCoV-229E, hCoV HKU-1, WU polyomavirus, and KU polyomavirus were performed, as described, for HBoV-2 -positive samples (10).
Of the 212 samples tested, the following viruses were detected: human respiratory syncytial virus ( HBoV-1 was not detected in the study population. HBoV-2 DNA was found in 5 (2.3%) of the 212 samples collected; all positive samples had been obtained in October 2008. The age range of the children with HBoV-2-positive samples was 4-34 months (median 24 months), and all were male. The diagnoses were bronchiolitis for 3 children and bronchopneumonia for 2. The most frequently codetected virus was human respiratory syncytial virus, found in 4 (80%) of 5 samples. One sample that was negative for respiratory syncytial virus and positive for HBoV-2 was negative for all other respiratory viruses.
Nucleotide sequences were determined for the NS-1 gene, and phylogenetic analyses, which included HBoV-3, a new lineage designated by Arthur et al. (2), showed that the NS-1 gene was relatively well conserved and that there were 2 major groups of the virus, the UK strain and the Pakistan strain. HBoV-2 strains isolated from South Korea belonged to the HBoV-PK2255 (FJ170279) cluster (Figure).
Recent studies have detected HBoV-1 in serum samples of children with Kawasaki disease and of an immunocompromised child with hepatitis (7,8). However, neither HBoV-1 nor HBoV-2 was detected in the 172 serum samples from 61 patients with hepatitis, 12 with Kawasaki disease, 18 with Henoch-Schönlein purpura, and 81 healthy children.
The absence of HBoV-1 in the samples examined was unexpected because HBoV-1 was detected in >10% of 558 respiratory samples collected from a demographically similar study population during the winter 2 years earlier (10). Future studies, with larger populations and over longer periods, are needed to delineate seasonal variations between HBoV-1 and HBoV-2.
We demonstrated HBoV-2 DNA in the respiratory tract secretions of children with acute LRTIs. In most positive samples, the virus was found in addition to other respiratory viruses. A limitation is that the study did not consider health control measures and other clinical disease such as gastroenteritis and was conducted for a short time. The role of HBoV-2 in LRTIs remains unclear; further studies are needed to clarify whether this virus is only shed from the respiratory tract or whether it replicates in the gastrointestinal tract.
This study was partly supported by a research grant (2008) from Inje University.

Nontuberculous Mycobacterium Infection and Tumor Necrosis Factor-α Antagonists
To the Editor: Mycobacterium haemophilum is an aerobic, slowgrowing microorganism with optimal growth at 30°C to 32°C. It has a unique requirement for ferric iron-containing compounds (1), from which it acquired its name (i.e., haemophilum). Infections with M. haemophilum are rare, but cervicofacial lymphadenitis caused by M. haemophilum has been described in children (2). Besides cervicofacial lymphadenitis, extrapulmonary signs of M. haemophilum disease include subcutaneous noduli, arthritis, and osteomyelitis, which generally affect immunocompromised patients (3). Recently, 2 cases of cutaneous M. haemophilum infections after alemtuzumab treatment were reported (4). A small number of pulmonary M. haemophilum infections associated with AIDS or solid organ or bone marrow transplantation have been described (1). We report pulmonary M. haemophilum infection in a woman who had been immunosuppressed by tumor necrosis factor-α antagonist (TNF-αA) (adalimumab) treatment for rheumatoid arthritis.
A 72-year-old woman with a history of rheumatoid arthritis and obstructive sleep apnea syndrome had signs and symptoms of fatigue, mild fever episodes, and a nonproductive cough 9 months after treatment for rheumatoid arthritis had begun with methotrexate (MTX) and TNF-αA. Physical examination was unremarkable except for a body temperature of 38.9°C. Laboratory testing showed an increased erythrocyte sedimentation rate (ESR) (77 mm/h), an increased Creactive protein (CRP) level (60 mg/L), a normal leukocyte count (8,500 cells/ μL), and relative monocytosis (12%). HIV serologic testing results were negative. Chest radiograph showed an infiltrate in the right upper lobe. Chest computed tomography confirmed this finding and showed lymphadenopathy in the right hilus and mediastinum.
Notably, the tuberculin skin test result was negative at screening before she began the TNF-αA treatment, but was now positive (20 mm), suggesting mycobacterial infection. Auramine and Ziehl-Neelsen staining of sputum and bronchoalveolar liquids showed no acid-fast bacilli, and M. tuberculosis infection was not confirmed by PCR or culture. Eventually, a mediastinal lymph node biopsy was taken by endoscopic ultrasound guidance. Granulomatous inflammation and acid-fast bacilli were seen by microscopy. Corresponding cultures yielded a strain identified as M. haemophilum at the Netherlands National Institute for Public Health and the Environment (RIVM) by using the Inno-LiPA Mycobacteria v2 reverse line blot assay (Innogenetics, Ghent, Belgium). Strain identity was confirmed by sequencing of the complete 16S rDNA gene, which was identical to that of M. haemophilum available in the GenBank sequence database (National Center for Biotechnology Information; www.ncbi.nlm.nih.gov; accession no. X88923).
The RIVM performed drug susceptibility testing by using a modified agar dilution method (5). Middlebrook 7H10 media were enriched with 10% sheep blood hemolyzed by 1:1 dilution with water and subsequent freezing-thawing. Historic drug susceptibility data was reviewed (Table). Initially, adalimumab was discontinued, and our patient was treated with isoniazid, ethambutol, rifampin, and pyrazinimide because M. tuberculosis infection was suspected. After identification of M. haemophilum, our patient was treated with rifampin and azithromycin. A total treatment duration of 10 months resulted in complete resolution of the