Effectiveness of sofosbuvir based direct-acting antiviral regimens for chronic hepatitis C virus genotype 6 patients: Real-world experience in Vietnam

Background Hepatitis C virus (HCV) genotype 6 is the commonest cause of chronic hepatitis C infection in much of southeast Asia, but data on the effectiveness of direct-acting antiviral agents (DAAs) against this genotype are limited. We conducted a retrospective cohort study of patients attending the Hospital for Tropical Diseases (HTD), Ho Chi Minh City, Vietnam, to define the effectiveness of DAAs in the treatment of chronic HCV genotype 6 in actual practice. Methods We included all patients with genotype 6 infections attending our hospital between March 2016 and October 2017 who received treatment with sofosbuvir-based DAA treatment regimens, and compared their responses with those with genotype 1 infections. Results 1758 patients (1148 genotype 6, 65.4%; 610 genotype 1, 34.6%) were analyzed. The majority of patients (1480, 84.2%) received sofosbuvir/ledipasvir (SOF/LDV) ± ribavirin (RBV); 278 (15.8%) received sofosbuvir/Daclatasvir (SOF/DCV) ± RBV. The median age of the patients was 57 years, (interquartile range (IQR) 46–64 years) The baseline HCV viral load (log IU/ml) was significantly higher in patients infected with genotype 6 compared with those infected with genotype 1 (6.8, 5.3–6.6 versus 6.3, 5.3–6.5 log10 IU/ml, p = <0.001, Mann Whitney U test). A sustained virological response (SVR), defined as an undetectable viral load measured between 12 and 24 weeks after completing treatment, and indicating cure, was seen in 97.3% (1711/1758) of patients. Treatment failure, defined as HCV viral load ≥15 IU/ml ≥12 weeks after completing treatment appeared to be more frequent in patients infected with genotype 6 virus (3.2%, 37/1148) than in those infected with genotype 1 (1.7%, 10/610), p = 0.050 chi-squared test). We found no evidence that patient’s age, gender, liver cirrhosis, diabetes, HBV or HIV coinfection, prior treatment failure with pegylated interferon therapy, body mass index (BMI), aspartate aminotransferase to platelet ratio index (APRI), or fibrosis 4 (FIB-4) index were associated with treatment failure. Conclusions Our study suggests that patients with HCV genotype 6 infection in Vietnam may respond less well to treatment with sofosbuvir based DAAs than patients with genotype 1 infections. Further studies are needed to confirm this observation and to define whether it is driven by genotype-specific mutations.


Introduction
The Global Health Sector Strategy (GHSS) for viral hepatitis 2016-2021 calls for the elimination of viral hepatitis as a public health threat, reducing new infections by 90% and mortality by 65% by 2030 [1]. The WHO Western Pacific region including Vietnam bears the highest burden of Hepatitis C virus (HCV) globally, with approximately 19.2 million chronic infections [1]. The introduction of direct-acting antiviral agents (DAAs) has revolutionized HCV treatment and increasing numbers of patients are being treated. Several phase III clinical trials (Neutrino, Fission, and Valence) have demonstrated the efficacy, simplicity, and tolerability of DAAs [2][3][4] in the treatment of HCV in well-resourced setting. The sustained virologic response rate (SVR), defined as an undetectable viral load 12 weeks after completion of treatment and considered to represent cure, is consistently above 90% for most HCV-infected patient populations [5,6].
HCV is classified into 7 genotypes, and these have specific geographical distributions. HCV genotype 6 is largely confined to China and southeast Asia, including Vietnam, Laos, Cambodia, Myanmar, Taiwan and the southern Chinese provinces of Guangxi, Guangdong and Hainan [7,8]. Thirty-one subtypes of genotype 6 have been recognized in the region, indicating a local emergence and evolution [9]. In the south of Vietnam, up to 60% of HCV infections are caused by genotype 6 [10] and therefore the success of HCV elimination in the region depends upon the effectiveness of DAA combinations against this genotype. While a number of DAAs, such as Sofosbuvir (SOF), are believed to have antiviral effects that are independent of the virus genotype, there are limited data on the efficacy of DAA treatments for HCV genotype 6 infections. This reflects the limited numbers of patients with the genotype recruited into clinical trials [11,12]. While studies from New Zealand and Hong Kong that have included small numbers of genotype 6 infected patients suggest that SOF-based regimens, including SOF + Ledipasvir (LDV) and SOF+ Ribavirin (RBV), are likely to be effective for most cases [13,14], few data exist regarding the efficacy of treatment in resource poor settings.
The 'real world' effectiveness of medical treatments do not necessarily reflect their efficacy rates seen in clinical trials, and HCV infection is no exception [15,16]. These differences likely reflect heterogeneities in patient characteristics, clinical practice, resources, care coordination, treatment drug combinations, and treatment adherence and duration, and become apparent only when a drug is prescribed to a wider population [5]. Real-world data on the effectiveness of DAAs in HCV genotype 6 infections from the geographic locations where it is prevalent are particularly limited [17]. Understanding the effectiveness of DAAs in such settings and in normal practice is crucial to inform policymakers when designing HCV treatment programs.
Vietnam is among the top 20 countries with the highest HCV burdens, with a population seroprevalence of between 0.9% and 1.2% [18]. DAAs have been the recommended first line treatment for HCV infection in Vietnam since 2016 [19]. All provincial hospitals and selected referral HCV treatment centers, including the Hospital for Tropical Diseases (HTD), Ho Chi Minh City, have prescribed DAA treatment since then, although the cost of this has been met by patients [17]. We present here our experience of the use of two sofosbuvir-based DAA regimens (SOF/LDV ± RBV, and SOF/Daclatasvir (DCV) ± RBV) in treatment-naïve patients infected with HCV genotype 1 or 6.

Study description and ethical approval
We performed a retrospective, intent-to-treat cohort analysis of all chronic HCV (genotype 1 and 6) infected patients attending our hospital who began treatment with DAA combination therapy between March 2016 and October 2017. We include only DAA-inexperienced patients in the study; however, we did include patients who had previously failed to respond to treatment with non-DAA treatment history (i.e. Pegylated interferon (PegINF) and RBV). All DAA treatment was prescribed through the hospital pharmacy. To be included patients had to be age �18 years, infected with HCV genotype 1 or 6, and have initiated treatment with either SOF/LDV ± RBV, or SOF/DCV ± RBV. We excluded patients where the baseline and/or SVR HCV viral load data were unavailable including patients with incomplete treatment. The study received ethical approval from the Ethics Review Committee of the Hospital for Tropical Diseases (approval no CS/ND/16/02 date 23/11/2017).

Setting, patients and data extraction
The Hospital for Tropical Diseases (HTD), Ho Chi Minh City, is a 650-bed infectious disease hospital, and a designated specialized care provider and referral centre for patients with infectious hepatitis from the centre and south of Vietnam [10]. In 2015 HTD introduced an electronic record keeping system for the outpatients departments. These records include sociodemographic, clinical, imaging, prescribing, diagnostic and treatment outcome data for each patient under a unique identification number (ID). The HTD clinical laboratory maintains a separate database of all laboratory investigations conducted on patient samples; laboratory data are stored using a separate laboratory number linked to the unique patient ID. For this study, the hospital database was screened for the diagnosis of chronic HCV infection and treatment with DAAs. The hospital records management team extracted sociodemographic, clinical, laboratory and drug prescription and treatment outcome data for all eligible patients from the database according to the study proforma. All patient data were anonymized by replacing the patient identifier (unique ID) with a unique study number before transfer to the study investigators.

Treatment
Patients received treatment according to the Vietnamese national guidelines at the time, summarised in S1 Table [22]. In line with these guidelines, the treatment regimen was selected based upon the HCV genotype and the presence or absence of cirrhosis. For patients without cirrhosis, the guidelines recommended a treatment duration of 12 weeks. Patients with cirrhosis were treated with either SOF/LDV or SOF/DCV for 24 weeks, or either of these combinations together with RBV for 12 weeks. The choice regarding these latter regimens was made by the physician in discussion with the patient. SOF/LDV was given daily as a single oral fixed dose combination tablet consisting of 400 mg SOF and 90 mg LVD (sourced from any of Mylan Laboratories, India, Hetero Laboratories, India and M/s Natco Pharma, India). SOF/DCV was given as 400mg SOF and 30mg DCV once daily (sourced from Mylan Laboratories, India, Hetero Laboratories, India and M/s Natco Pharma, India). The treatment duration received by each patient was confirmed by review of prescriptions and the number of cumulative days' supply purchased by each patient, with purchase of medication for either 84 (12 weeks) or 168 days (24 weeks) from the hospital pharmacy being assumed to indicate the completion of 12 or 24 weeks of treatment respectively. We calculated the end of treatment (EOT) as the last day covered by the prescription related to the initial date of medication dispensing by the hospital pharmacy, cross-checked with the number of tablets bought. At HTD all patients receive advice on the importance of treatment adherence as per standard of care at each visit. Where doses are missed they are recommended to take the missed dose if within 16 hours of the due time. If more than 16 hours have elapsed, they are recommended to take the next dose at the due time.

Treatment outcome monitoring
HCV viral load was measured before IOT, at weeks 4, week 8 (if the HCV viral load was detectable at week 4), and either 12 or 24 weeks after the end of treatment (EOT, see S2 Table) [22]. Rapid virologic response (RVR) was defined as an HCV RNA <15 IU/mL 4 weeks after IOT. Treatment success-sustained virological response (SVR) was defined as unquantifiable HCV RNA (LOD <15 IU/mL) on all HCV RNA tests measured from 12 weeks or 24 weeks after the EOT or undetectable HCV RNA on last HCV RNA test 12 weeks or 24 weeks after EOT. Failure to achieve an SVR at 12 or 24 weeks after the EOT was defined as treatment failure. We defined breakthrough and relapse of infection as the achievement of an undetectable HCV RNA during treatment, followed by the detection of HCV RNA �15 IU/mL while on treatment (breakthrough), or after treatment completion (relapse).

Data analysis
Data analysis was performed using Statistical Package for Social Science (SPSS) software (IBM SPSS Statistics 23, NY USA). The main outcome of interest was the response to treatment. We analysed the success of the treatment on an intent to treat basis (n = 1758). Baseline descriptive statistics were summarized for the variables of interest. Comparisons between groups were performed using either the chi-squared or Fisher's exact tests for categorical variables; t-tests and the Mann-Whitney U-test were used for continuous variables. We used logistic regression to determine the baseline factors associated with SVR. A two-sided P value of �0.05 was considered statistically significant.
In patients with treatment failure (n = 47), the mean baseline viral load was higher than those who achieved SVR (6.7 ± 7.0 versus 6.4 ± 6.4 log IU/ml; p = 0.055 Mann Whitney U Test). However, the majority of these patients attained RVR with viral relapse occurring 12 weeks after EOT.
A bivariate analysis was conducted to examine the effects of HCV genotype and baseline viral load on SVR or cure. Patients with genotype 1 had a higher probability of achieving cure (OR = 1.99; 95% CI: 0.98-4.04; p = 0.054). Similarly, patients with lower viral loads had a higher probability of achieving SVR (OR = 1.8, 95% CI = 0.97-3.36, p = 0.061) Table 3.

Discussion
We performed a retrospective review of HCV in our hospital in order to understand the response in patients infected with HCV genotype 6 to DAAs. Genotype 6 is the most frequent cause in our patients. There are a few data on the response of genotype 6 infections to DAAs, particularly in low-income settings. Our study addresses this knowledge gap and adds to the real-world data on the effectiveness of DAAs in genotype 6 in clinical practice [11,12]. We compared treatment responses in patients infected with genotype 6 with those of patients infected with genotype 1 virus. We chose this comparison because i) current treatment guidelines recommend the same drug combinations can be used for each of these genotypes, and ii) the wealth of data from rich countries regarding the treatment response of genotype 1 infections allows us to set our experience in context. Our data add to the limited number of reports on treatment response that have emerged from Asian countries, including Vietnam [17,23]. Similar to earlier studies, we documented a high prevalence (54.7% of cases) of HCV genotype 6 among patients attending our hospital [24]. This might be due to lower rate of spontaneous clearance of HCV genotype 6 than other genotypes or genotype 6 infections respond poorly to historical (non-DAA) anti-HCV therapy (e.g. PegINF±RBV) [18]. HCV genotype 6 is unique in many respects including i) localized geographic epidemiology (Laos, Cambodia, Vietnam, Myanmar, and Southern China), ii) high genetic diversity [9], iii) high number of preexisting drug resistance mutations [25], and iv) variable in-vitro susceptibility to DAAs (e.g. LDV) [25]. Compared with genotype 1 infections, we observed a higher prevalence of genotype 6 infection in women compared with men. The reasons for this is unclear, but it may represent inequalities in health care access between men and women in Vietnam. Historically treatment for HCV has been expensive and funded by the patient; fewer women may have had access to funds for treatment. The median age of patients in our study was 57 years. We found that patients with genotype 1 infections were younger then genotype 6 infected patients, and had higher rates of HBV and HIV coinfection. This points to some separation in the epidemics of genotype 6 and 1 infections in Vietnam. However, this being a retrospective study, we were unable to interrogate this further. It is feasible that the differences in genotype epidemics are associated with different risk behaviors (e.g. injectable drug use, man sex with man and sexual risk behavior) in the Vietnamese population at specific times. We found that patients with genotype 6 infections tended to have higher baseline viral loads than patients with genotype 1 infections. The consistency of this finding with previous studies suggests that genotype 6 virus may have a higher replication rate than genotype 1 virus [24].
Vietnamese guidelines at the time of this study suggested patients should receive either 12 weeks of treatment with or without RBV or 24 weeks of treatment without RBV depending upon the degree of their underlying liver disease. The majority of patients in our study received a 12 week treatment course. This is probably because the 12 week treatment course is significantly cheaper than 24 weeks, although patient convenience and adherence may also have been part of clinical decision making [24]. The American Association for the Study of Liver Diseases (AASLD), the European Association for the Study of the Liver (EASL), and the Asian-Pacific Association for the Study of the Liver (APASL) recommend daily fixed-dose combination of i) glecaprevir/pibrentasvir for 8 weeks, ii) sofosbuvir/velpatasvir for 12 weeks, iii) SOF/LDV for 12 weeks, or iv) SOF/DAC for 12 weeks for treatment naïve HCV genotype 6 patients [26][27][28]. However, SOF/LDV is not currently recommended for treatment of subtype 6e infections [26]. There is a move towards recommending treatment durations for HCV infection (for example, as per AASLD guidelines); shorter durations of treatment can reduce costs and aid adherence. However, it is important that such recommendations for Asia are backed up by clinical trial evidence that includes patients with genotype 6 infections. Studies on treatment shortening are under evaluation in Vietnam.
We observed excellent cure rates (>95%) in both patients with genotype 1 and genotype 6 HCV infections in our cohort, and cure rates were similar for both SOF/LDV and SOF/DAC combination therapy. We did not observed statistically significant difference in treatment outcomes between genotype 1 and 6 for these durations of treatment, although larger numbers of patients need to be evaluated to ensure this is true. We found cure rates of 96.8% (1111/1148) in genotype 6 infections versus 98.4% (600/610) in Genotype 1 (p = 0.05) infections. The slightly lower response we observed in genotype 6 infections could be explained by i) the higher baseline viral load in genotype 6 disease, ii) presence of pre-existing drug resistance mutations in genotype 6, and iii) the genetic diversity of genotype 6. A study from Myanmar found unexpectedly low SVR rates with sofosbuvir/ledipasvir combination treatment in genotype 6 infected patients [29]. In-vitro susceptibility studies have shown that there is variability amongst sub-lineages of genotype 6 virus to some DAAs. For example, in vitro resistance selection studies with LDV identified the single Y93H or Q30E resistance-associated variants (RAVs) in the NS5A gene in HCV genotype 6e. Similar RAVs were also observed in patients after a 3-day monotherapy treatment with LDV in genotype 1b [30]. Subtype 6e is also the predominant subtype in southern Vietnam. [18]. We do not have highly resolved genotype data for the infections in our study; however, the overall excellent response rates that we found suggest that either the treatment combinations used in our cohort are in fact highly effective across the genotype 6 subtypes, or subtype 6e is not frequent in our patients.
In our study, treatment failure was higher in patients treated with a 24-week regimen. This is possible because patients with cirrhosis (Child pugh B or C) are often treated with a 24-weeks regimen. These patients were older, had higher liver stiffness, high APRI and FIB-4 index and respond poorly to DAAs.
The vast majority of patients in our study had rapid virological responses, with undetectable viral loads by 4 weeks after treatment initiation. There was no difference in the rates of RVR by genotype. However, eventual cure rates were similar between patients who did and did not achieve RVRs, suggesting that viral load measured at this time point has little clinical utility where at least 12 weeks of treatment is prescribed and the patient is adherent. However, our experience contrasts with that of others where RVR has appeared to have a predictive value [31].
We could not determine whether the treatment failure was a result of relapse (SVR failure after EOT) or breakthrough (SVR failure during treatment) as viral load data at end of treatment was not available. However, based on the fact that 89.3% (42/47) treatment failure had RVR, one might speculate that most of the treatment failure were due to viral relapse after EOT or viral breakthrough 4 weeks after IOT. This suggests a possible adaptation/mutation in the viral genome or selection of resistant variants during the course of treatment. It has been reported that RVR and very rapid virologic response (vRVR; undetectable serum HCV RNA level at week 2) has a high positive but low negative predictive value of SVR with dual sofosbuvir/ribavirin therapy [32].
Our study is not without limitations. Our centre is a tertiary care centre and therefore the patients and outcomes may not be representative of the wider patient population in Vietnam. During the period of the study, HCV treatment was available only to self-funded patients. Given that DAA treatment was costing around $2500/patient at the time, it is likely that most patients are wealthy and therefore patients from lower socioeconomic groups may not be represented. It was not possible to interrogate this with the available dataset. Our liver status data may be biased as we could only analyse the data from patients who could afford the test. Our study includes only patients who have completed the treatment as patients with incomplete treatment or discontinued treatment lacked SVR viral load data.
In conclusion, genotype 6 infection appears to be the predominant infecting HCV genotype in the south of Vietnam. Treatment outcomes in our tertiary referral centre were largely comparable to those in rich developed countries when treated for 12 weeks. It is possible that G6 outcomes are slightly worse than genotype 1, but any differences are small. However, there remains a need to generate evidence from randomized control trials on the best treatment combinations and options for patients in Asia infected with genotype 6.
Supporting information S1 Table. Treatment