To the Editor: For nearly 3 decades, human parvovirus B19 (B19V) was considered to be the only pathogenic parvovirus found in humans. Since 2005, several new human parvoviruses have been found, including human bocaviruses (HBoV1–4) and human parvovirus 4 (PARV4) (1–5), and during 2012, metagenomic analysis of fecal samples from children in Burkina Faso with acute diarrhea showed a highly divergent parvovirus, which was named bufavirus (BuV) (6). Its sequence in the coding region showed <31% similarity with known parvoviruses, the closest genera being Protoparvovirus and Amdoparvovirus. Subsequent studies, on the basis of PCR results, showed that 4% of fecal samples from Burkina Faso (n = 98) and 1.6% from Tunisia (n = 63) harbored either of 2 genotypes of this new virus, which belongs to the species Primate protoparvovirus 1 of the genus Protoparvovirus (6,7; http://ictvonline.org).

To assess the occurrence of BuV in northern Europe, we analyzed 629 fecal samples from patients of all ages (median 51.5 years, range 0–99) in Finland who had gastroenteritis. To gain a more complete representation of BuV occurrence, we obtained samples retrospectively from routine diagnostics for bacterial and viral gastroenteritis-inducing pathogens (HUSLAB, Helsinki University Central Hospital Laboratory Division, Helsinki, Finland) and analyzed all samples available during the collection periods.

The samples originally sent to HUSLAB for bacterial diagnosis (bacterial cohort, n = 243) had been analyzed during October 2012–March 2013 for Salmonella spp., Shigella spp., Campylobacter spp., Yersinia spp., Vibrio cholerae, and Escherichia coli (subtypes enterohemorraghica, enteropatogena, enterotoxigenic and enteroagregativa) by using culture or PCR (8). In 81 (33.3%) of the samples, >1 bacterial pathogen was found.

The samples originally sent for viral diagnosis (viral cohort, n = 386) had been tested in HUSLAB for norovirus during April–May, 2013 by using reverse transcription quantitative PCE (RT-qPCR)(HUSLAB in-house). Further diagnosis for rotavirus and adenovirus had been requested by physicians from 105 (27.2%) of 386 samples (Diarlex MB antigen detection assay, Orion Diagnostica, Espoo, Finland), and for astrovirus from 33 (8.6%) samples (RT-PCR, HUSLAB in-house). A viral pathogen was discovered in 141 (36.5%) samples; in 139, the pathogen was norovirus.

The samples had been sent from diverse locations within Finland, and thus were not from a few isolated outbreaks. No further information on patients and samples was available for either cohort, and not enough samples were left for retrospective analysis of additional pathogens. The Ethics Committee of the Hospital District of Helsinki and Uusimaa approved the study.

BuV DNA was detected by using a new real-time qPCR with the following primers and probe: BuV forward, 5′-ACAGTGTAGACAGTGGATTCAAACTT-3′; BuV reverse, 5′-GTTGTGGTTGGATTGTGGTTAGTTC-3′; BuV qPCR probe, 5′-FAM-CGGAAGAGATTTTGACAGTGCYTAGCAA-BHQ1–3′. The detailed qPCR protocol is shown in the Technical Appendix. The analytical sensitivity of the RT-qPCR assay was 5–10 copies per reaction.

Of the 629 fecal samples, 7 (1.1%) were positive for BuV DNA, of which 4 were from the bacterial cohort and 3 from the viral cohort. BuV DNA quantity was low in all samples, ranging from 1.9 × 10³ to 3.2 × 10⁴ copies per milliliter of fecal supernatant (Table). In contrast to the original discovery of the virus in children with diarrhea (6), all positive samples were from adults (median age 53 years, range 21–89 years). All BuV DNA–positive results were confirmed by repeated BuV qPCR, by amplifying and sequencing another area of the virus, or by both methods (Table): all sequenced amplicons were more similar to the BuV genotype 1 (Figure) (6). Two of the BuV-positive samples were from the same patient, taken 4 days apart, and the latter sample also harbored norovirus. The additional 6 BuV-positive samples were negative for the other viral or bacterial pathogens tested.

Seven fecal samples collected from adults in Finland contained BuV DNA, indicating that circulation of the virus is restricted neither to children nor to Africa. However, the low DNA loads in all the positive samples suggest that BuV might not be the primary cause of these cases of gastroenteritis. A known gastroenteritis-inducing pathogen (norovirus) was found in 1 of the 7 BuV-positive samples. We did not observe any clustering of the 7 positive samples into a specific period (Table).

Although the association with gastroenteritis seems weak, BuV might cause symptoms of other types. We did not include feces from healthy subjects for comparison. The identified BuV DNA in our samples could originate from previous or current infections unrelated to gastroenteritis, or be associated with prolonged virus secretion in the respiratory or digestive tracts, a phenomenon shown, e.g., for (9,10). Acquisition of the virus from a food source cannot be ruled out, although 1 patient harbored the DNA for at least 4 days, during which a 10-fold increase in viral load was observed.

Overall, this study shows that BuV circulates in northern Europe and can be found in the feces of patients with gastroenteritis. Despite the absence of known pathogens among 6 of 7 BuVs-shedding patients, the causative role of BuV in gastroenteritis remains uncertain. Serologic studies will help clarify a possible association between BuVs and diarrhea or other diseases.

Top

Related Links

• More Letter Articles