Patient information and clinical samples were derived from a multilocation observational study designed to evaluate the causes of community-acquired infection in Southeast Asia (6). We included all 493 samples (384 plasma, 92 pooled nasal and throat swabs, 10 stool, and 7 cerebrospinal fluid [CSF]) from 386 patients in Vietnam with community-acquired infection of unknown origin after extensive diagnostic workup for viral metagenomic analysis (7).

Analysis of metagenomic data revealed that, in 1 plasma sample, of 98,344 obtained reads, 5,342 reads were of HCV sequences, 430 of HIV sequences, and 273 of HPgV-2 sequences; we confirmed all reads by corresponding virus-specific reverse transcription PCR (RT-PCR). HPgV-2 sequence screening and HPgV-2 RT-PCR testing did not detect HPgV-2 in any of the remaining samples of the patients included in metagenomic analysis.

To explore the prevalence of HPgV-2 in HCV-infected patients in Vietnam, we used a reference-based mapping strategy to screen for HPgV-2 sequences in additional viral metagenomic datasets (Table 1). We detected HPgV-2 sequences in 5/79 HIV/HCV co-infected patients who participated in a trial evaluating the hepatic safety of raltegravir/efavirenz-based therapies in antiretroviral-naive HIV-infected subjects co-infected with HCV. We did not detect HPgV-2 sequences in 394 HCV-infected patients with clinically diagnosed hepatitis who participated in molecular epidemiologic studies of hepatitis viruses (Table 1).

We subsequently confirmed the result of this reference-mapping approach by HPgV-2 multiplex RT-PCR (8) testing of the extracted RNA from original samples. We conducted multiplex RT-PCR screening for HPgV-2 RNA in plasma samples of matched controls (78 HIV-infected patients and 80 healthy volunteers) of the 79 HCV/HIV co-infected patients; we found no evidence of HPgV-2 (Table 1). In addition, we did not detect HPgV-2 RNA in any plasma samples from patients with HAV (n = 71) and HBV (n = 103) infection (Table 1).

HPgV-2 RNA was detectable for <18 months in 3/5 patients with HCV/HIV co-infection (Table 2). We did not detect HPgV-2 RNA in the available follow-up serum sample collected 14 days after enrollment from the patient with community-acquired infection (Table 2).

All 5 HCV/HIV co-infected patients had CD4 counts >200 cells/µL at baseline and at 6-, 12-, and 18-month follow-up (Table 2), but none received specific anti-HCV drugs, which was attributed to drug unavailability or unaffordability during the study period. During follow-up, hepatitis and splenic abnormalities were detected in 4/5 patients, which were likely attributable to HCV infection (Table 2). The patient with community-acquired infection was recorded as surviving to 28 days of follow-up (Table 2).

Figure

Figure. Maximum-likelihood phylogenetic tree of amino acid sequences of coding sequences of human pegivirus 2 strains from Vietnam compared with global strains and other pegiviruses. We used the general matrix with empirical...

Using deep sequencing and a combination of overlapping PCRs and Sanger sequencing of PCR amplicons (primer sequences available upon request), we obtained 5 nearly complete genomes (coverage of >92%) and another partial genome (coverage of ≈69.1%). Pairwise comparison of HPgV-2 coding regions obtained in this study and previously reported HPgV-2 sequences showed overall sequence identities at the nucleotide level of >94.6% and at the amino acid level of >95.3% (data not shown). Phylogenetic analyses revealed a tight cluster between viruses from Vietnam and global strains (Figure). We submitted the HPgV-2 sequences we generated to GenBank (accession nos. MH194408–13).

Of the 5 HPgV-2 genome sequences we recovered, we generated 2 by deep sequencing. The results were above the proposed sequencing-depth threshold of >5 for sequences generated by next-generation sequencing (9) and sufficient for intrahost diversity investigation. One sequence we generated had mean coverage of 2,049 (range 12–9,912), with a total of 26 (10 [38%] nonsynonymous) positions carrying minor variations detected in the corresponding dataset (data not shown). For the other sequence, mean coverage was 32,531 (range 13–138,383), with a total of 37 (13 [35%] nonsynonymous) positions carrying minor variations in its dataset (data not shown).

We report the detection and genetic characterization of HPgV-2 in Vietnam and describe the observed demographic and clinical characteristics of patients with HPgV-2 infection. Together with reports from China, Iran, and the United States (1–4,8,10), our findings further emphasize the strong association between HPgV-2 and HCV, especially HCV/HIV co-infection. The absence of HPgV-2 in 394 HCV-infected patients may have been attributed to the small sample size and the fact that the reported prevalence of HIV among HCV-infected patients was <6.5% (11,12). Of note, HPgV-2 was detected in only 0.29% of HCV-monoinfected patients in China.

Previous reports showed that HPgV-2 viremia can be transient or persistent. Likewise, in our study, HPgV-2 RNA became undetectable after 14 days in a HCV/HIV co-infected patient with community-acquired infection of unknown origin, but remained detectable in other HCV/HIV co-infected patients through up to 18 months of follow-up.

The pathogenic potential of HPgV-2 remains unknown. Its role in HCV/HIV co-infection and response to treatment warrants further research, given its low detection rates in blood donors in the United States and China (1,3) and its absence in healthy persons (this study) but close association with HCV/HIV co-infection.

In the era of sequence-based virus discovery, a key question is whether the detected genome represents live virus or a non–replication competent genome. Addressing this question would require recovery of virus in cell culture. However, our detection of minor variations across 2 HPgV-2 genomes suggests that viral replication had occurred in the infected patients. Phylogenetically, the close relatedness between HPgV-2 strains from Vietnam and global strains suggests HPgV-2 has a wide geographic distribution.

Our study has some limitations. First, we only retrospectively tested available archived samples without formal sample size estimation, which may have explained the absence of HPgV-2 in the remaining 394 HCV patients. Second, we did not employ a serologic assay to screen for HPgV-2 antibodies in patients’ plasma. Third, we used only multiplex PCR with primers based on a limited number of available HPgV-2 sequences. Therefore, we may have missed genetically diverse HPgV-2 strains, and we may have underestimated the prevalence of HPgV-2 infections in Vietnam.

Our findings contribute expanded data about geographic distribution, demographics, and genetic diversity of HPgV-2. Because HCV and HIV infections are global public health issues, the extent to which HPgV-2 interacts with HCV and HIV in co-infected patients and the possible clinical consequences warrant further research.