Prevalence of HCV genotypes and subtypes in Southeast Asia: A systematic review and meta-analysis

Known for its high genetic diversity and variation in genotypic presence in different regions of the world, hepatitis C virus (HCV) is estimated to infect about 71 million people globally. Selection of an appropriate therapeutic regimen largely depends on the identification of the genotype responsible for the infection. This systematic review and meta-analysis was conducted to provide a comprehensive view of HCV genotype and subtype distribution in Southeast Asia (SEA). The review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA). We searched five databases without year and language restrictions. Data from 90 eligible studies involving 15,089 genotypes and 9,646 subtypes representing 10 SEA countries were analyzed. The pooled estimates showed that genotype 1 (46.8%) [95% CI, 43.2–50.4; I2 = 92.77%; p < 0.001] was the most dominant HCV genotype in the region, followed by genotype 3 (23.1%) [95% CI, 19.4–27.2; I2 = 93.03%; p < 0.001], genotype 6 (16.5%) [95% CI, 13.8–19.6], genotype 2 (4.6%) [95% CI, 3.5–5.9], genotype 4 (1.1%) [95% CI, 0.7–1.5] and genotype 5 (0.8%) [95% CI, 0.4–1.3]. Philippines had the highest prevalence of genotypes 1 and 2. Genotype 6 became more prevalent after year 2000. Over 40 different subtypes were identified, with subtypes 1b (26.3%), 1a (21.3%), and 3a (14.3%) being the most prevalent of all the reported subtypes. Although on a global scale, genotype 6 is considered highly prevalent in SEA, evidence from this study reveals that it is the third most prevalent genotype within the region.


Introduction
Hepatitis C virus (HCV) infection constitutes a major health challenge globally. HCV is a principal cause of hepatocellular carcinoma (HCC), liver cirrhosis and liver failure, and it is thought to infect about 71 million people worldwide [1]. The virus, which belongs to the Flaviviridae family, consists of a positive-sense single-stranded RNA genome that spans approximately 9.6 kb [2,3]. Flanked by 5' and 3' untranslated regions (UTRs), its long single open reading frame codes for structural (core [C] and envelope [E1 and E2]) and non-structural (P7, NS2, NS3, NS4A, NS4B, NS5A and NS5B) proteins [2,3]. HCV exhibits a profound genetic diversity-an occurrence partly attributed to the absence of proofreading nuclease activity of its ribonucleic acid (RNA)-dependent RNA polymerase [4,5]. Similar to human immunodeficiency virus (HIV), multiple quasispecies of HCV emerge quite frequently in vivo, posing significant threat to drug efficacy and vaccine development. Mutations that occur in the least conserved regions (i.e., hypervariable regions of the NS5A and envelope genes) of the HCV viral genome are thought to play a crucial role in immune evasion and the establishment of chronic infection [5].
Sequence and phylogenetic analysis of HCV genome has revealed at least six major genotypes (designated as genotype 1 to 6) and numerous subtypes (connoted with lowercase alphabets; e.g. 1a, 1b, 2a, etc.) [6]. Although HCV exhibits an extraordinary sequence diversity, all genotypes possess identical complement of co-linear genes of similar size within the open reading frame (ORF), and the known variants contain considerably similar sequence across the genome [7]. This makes it possible to classify them based on partial sequences from specific regions (e.g. core/E1 or NS5B) within the genome [7]. The distribution of HCV genotypes is largely dependent on geographical locations. Although genotypes 1, 2 and 3 seem to be distributed globally, 'endemic' strains of genotype 1 and 2 are mainly in West Africa while genotype 3 is dominant in South Asia [4]. Genotypes 4, 5, and 6 on the other hand, are predominant in the Middle East and Central Africa, Southern Africa, and Southeast Asia, respectively [4].
Distinct genotypes and subtypes of HCV vary in their degrees of resistance to some of the currently available antiviral drugs, making treatment decision genotype oriented [6,8]. Determination of infecting genotype is thus fundamental to HCV therapy as it helps ascertain which antiviral drug to administer, its dosage, as well as treatment duration [9]. While there are several HCV genotyping methods, full genomic analysis or sequencing and phylogenetic analysis of informative and conserved regions (e.g. core/E1 or NS5B) of the genome remains the gold standard [5,6,10].
In the past, attempts have been made to provide global and regional distribution of HCV genotypes [4,[11][12][13]. However, there has been no detailed report on the actual distribution of the virus' genotypes in Southeast Asia (SEA)-a multiethnic and socioculturally diverse Asian region spanning Brunei, Cambodia, East Timor, Indonesia, Laos, Malaysia, Myanmar, Philippines, Singapore, Thailand and Vietnam. This is further compounded by the fact that newer variants of the virus are continually being identified and characterized. In this article, we conducted a Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) compliant review of published articles reporting HCV genotypes in SEA since the virus was first identified in 1989 to provide an in-depth and up-to-date information on the distribution of HCV genotypes in the region.

Search strategy and selection criteria
We accessed five electronic databases (PubMed, Scopus, ScienceDirect, Google Scholar, and Web of Science) for studies reporting HCV genotypes in eleven SEA countries (Brunei, Cambodia, East Timor, Indonesia, Laos, Malaysia, Myanmar, Philippines, Singapore, Thailand, Vietnam). The databases were searched using a combination of terms (e.g., "hepatitis c", "genotype") related to the distribution of HCV genotypes in SEA without year or language restrictions. Details of the search strategy used is provided in S1 File. A final search was made on April 30, 2020.
Studies that conducted HCV genotyping and whose samples or analyzed data were collected in at least one of the SEA countries were included. We excluded studies (1) that are reviews, (2) with unknown sample origins (3) with only serological findings, (4) whose full text could not be obtained, or (5) contained duplicated data.

Quality assessment
The methodological quality of all the included studies was assessed using the Joanna Briggs Institute (JBI) critical appraisal checklist for prevalence data [14] (S2 File). A score of '1' for 'yes' and '0' for other parameters were assigned to attain a total quality score that ranged from 0 to 9. Studies with overall score of 7-9 were considered to be of sufficient quality. Two authors (A.A.I and Y.W.) independently assessed the studies. Studies were included if there was a consensus between the two reviewers. The quality assessment of the 90 included studies is provided in S1 Table. Data extraction and analysis From each of the included studies, three authors (A.A.I., N.A.M., and Y.W.) independently extracted information regarding country, year of publication, sample size, year of sample collection as well as reported HCV genotypes and subtypes. Data was also extracted for HCV genotypes reported without details of the subtypes. Disagreements were resolved by discussion and adjudication including a fourth author (R.H.S.). We converted HCV genotypes and variants published prior to 1994 to the consensus nomenclature proposed by Simmonds et al. [15] since a standardized genotype classification system was not available in the earlier years of HCV research. Where more than one article reported data from the same sample, record, or patient cohort, only one was counted and selected. Similarly, studies that genotyped and analyzed samples from more than one SEA country were categorized as "multi-country" rather than the individual countries included, and the data were extracted and analyzed in that form to avoid confusion. Data on 'inconclusive' or 'untyped/undetermined' genotypes and/or subtypes were neither extracted nor analyzed. Cases of more than one genotype and/or subtype from a single patient were tagged 'mixed.' Although identified, these mixed genotype data were excluded from the genotype and subtype analysis. In the case of immigrant workers, irrespective of their original countries, their data were pooled to reflect the countries they were working in when their samples were collected.
Data analysis was conducted using OpenMeta Analyst and meta (version 4.15-1) and metafor (version 2.4-0) packages of R (version 4.0.3) and RStudio (version 1.3.1093) [16]. The pooled prevalence of HCV genotypes was calculated, and subgroup analysis was done according to country and period of data collection. Random-effect model using the DerSimonian-Laird method of meta-analysis was employed to determine the pooled estimates of the reported HCV genotype and subtype proportions. Data were transformed using the logit transformation. A forest plot was subsequently generated to visually summarize details of the individual studies alongside the estimated common effect and degree of heterogeneity. Publication bias was examined using funnel plots (visual aid for detecting bias) and Egger's regression test. The heterogeneities (i.e., variation in study outcomes between studies) of study-level estimates were evaluated by Cochran's Q test and quantified using I 2 statistics. I 2 values of 25%, 50%, and 75% were considered low, moderate, and high heterogeneity, respectively [17]. Subgroup meta-analysis was used to analyze sources of heterogeneity. Sensitivity test was conducted using the leave-one-out analysis. P value of < 0.001 was considered to be statistically significant for all tests. A protocol was not lodged for this study.

Search results and eligible studies
The article selection process for this study is shown in Fig 1. In brief, our search of five databases returned 556 records. After duplicate removal and exclusion of studies that did not meet the inclusion criteria, a total of 90 studies were eligible and thus included in quantitative synthesis (Fig 1).

Characteristics of the eligible studies
All the eligible studies were of high methodological quality (S1 Table). Majority of the 90 studies included in this meta-analysis were reported from Thailand (n = 27). A total of 15,089 HCV genotypes were reported across the 90 studies and ranged from 6 (in Indonesia) to 3,020 (in Cambodia) ( Table 1). Six HCV genotypes (genotype 1 to 6) were basically reported. The most targeted regions for genotyping analysis were the 5'UTR, core and NS5B regions, meanwhile, sequencing and phylogeny were the major genotyping methods utilized. While HCV genotype information was available for all included studies, only 69 studies provided data on HCV subtypes. No study met our search criteria for East Timor. As such, genotype data analyzed included studies from ten SEA countries. The reported genotypes and subtypes varied across countries.

Subgroup meta-analysis
Subgroup analysis was conducted to assess genotype distribution across the Southeast Asian countries and to identify possible source of heterogeneity among the studies. The result of subgroup meta-analysis by the distribution of HCV genotypes across countries revealed different degrees of variability in the studies ( Table 2 and S11-S16 Figs). Overall, high genotype prevalence estimates were observed for genotypes 1, 3, and 6. For genotype 1, Philippines had the highest estimate (79.5%) although with fewer studies (n = 2) while Laos had the lowest (4.4%) with one study ( Table 2). Studies from Malaysia and Cambodia showed low heterogeneity (I 2 = 30.11% and 31.91%, respectively). For genotype 3, Malaysia had the highest estimate (63.1%) while Cambodia had the lowest (0.7%) ( Table 2). Also, studies from Malaysia had low heterogeneity (I 2 = 27.85%). For genotype 6, the highest estimate (95.6%) was from Laos, although with a single study, meanwhile, Indonesia and Malaysia had the lowest estimates (1.3% each). With the exception of Philippines and Malaysia, heterogeneity was of moderate to high in studies from other countries ( Table 2).
Mixed genotypes and/or subtypes were identified in reports from some of the countries.

Discussion
The genotypic diversity of HCV has for several years been a major hurdle to drug and vaccine development. While there are effective antiviral drugs against HCV, there is yet to be a licensed vaccine. Thus, antiviral therapy remains the mainstay of managing HCV infection. Numerous studies have demonstrated that different antiviral compounds and drugs (including the more recent interferon-free direct-acting antiviral agents such as telaprevir, sofosbuvir, etc.) display variable antiviral activities against the infecting genotypes and subtypes of HCV [108][109][110][111][112][113][114], a concern largely attributed to resistance associated substitutions in HCV [115,116]. A clear knowledge of the actual distribution of HCV genotype and subtypes in a region is pivotal not only to the treatment of the infection and the potential selection of candidate genotypic vaccine target for the region, but also to the development of appropriate government policies, interventions and programs.
In this study, we present a detailed and comprehensive view of HCV genotype distribution in SEA from available published regional data since the discovery of the virus in 1989. Apart from East Timor, genotype data was available for all the countries in this region. Most of the studies were from Thailand, Vietnam and Indonesia; possibly because of the burden of HCV infection [117][118][119] in relation to their high population. Similar to many other nations across the globe, we identified genotype 1 as the most dominant genotype in SEA. While this finding does not contradict the popular notion that genotype 6 is prevalent in SEA as supported by the extensive global study conducted by Messina et al. [4], it underscores the need for a cautious interpretation of results of studies conducted at global and regional levels. It is pertinent to clarify that genotype 6 is more prevalent in SEA than any other region in the world when  discussed according to global prevalence. However, within SEA, genotype 1 is the most prevalent, followed by genotypes 3 and 6, respectively ( Table 2). Combination of the pooled prevalence of these genotypes (1,3 and 6) accounted for more than 85% of the genotypes reported in the region, indicating that the three genotypes are the major HCV genotypes of concern. Notably, we found that the highest prevalence for genotype 1 (79.5%; 95% CI, 45.6-94.7), and genotype 2 (20.5; 95% CI, 5.3-54.4) occurred in Philippines and with an even distribution of the other genotypes. In addition, we observed that genotype 6 became more prevalent from the first decade of the twenty-first century. As suggested by previous reports [4,118], our study

PLOS ONE
showed that genotypes 4 and 5 were almost inexistent in SEA as only few isolated cases have been reported. Although the prevalence of HCV subtypes appears to vary in this study, we identified more than forty different subtypes ( Table 3), indicating the presence of a marked diversity of the virus' subtypes in the region. Overall, subtypes 1b (26.3%; 95% CI, 22.1-30.9), 1a (21.3%; 95% CI, 17.0-26.4), and 3a (14.3%; 95% CI, 11.0-18.5) constitute the major subtypes in SEA identified by this study. The predominance of subtype 1b corroborates earlier reports on the subtype as the predominant HCV subtype across the globe [4,11]. In addition to the cases of dual 'mixed' genotype infections identified in this study, several cases involving a mixture of more than two genotypes and/or subtypes in patients were reported in Indonesia, Vietnam, and Thailand [28, 80,93]. Expectedly, such infected individuals would require more than the usual combinations of the available HCV drugs for optimal treatment outcomes. Given the high cost of treatment, the occurrence of 'mixed' genotypes and subtypes revealed in this study unveils

PLOS ONE
an additional hurdle that should be considered in the current race for a vaccine, as agents capable of generating cross-reactive immunity would be highly invaluable. This study however has its limitations. There was paucity of studies in some of the included SEA countries. For example, the estimates for Brunei and Laos were derived from one study each. This could amount to an underestimation of genotypes and subtypes. Another concern is the variations in the methods of genotyping, as well as the targeted region for genotyping. Furthermore, because we prioritized peer-reviewed publications in an attempt to limit the tendency of including studies with unreliable genotype reports, we could have missed some relevant studies. Nevertheless, a broad search of multiple databases was done to access as many studies as possible in order to curtail the impact of overlooked studies. Despite these limitations, we believe this study provides a comprehensive and up-to-date situation of HCV genotype distribution in SEA.