Sofosbuvir-based therapies achieved satisfactory virological response in Chinese with genotypes 3 and 6 infection: a real-world experience from a centre

As previously shown by others, sofosbuvir-based regimens yield high sustained virological response rates in patients with HCV infection except for genotype 3b complicating with cirrhosis. The real-world study aims to explore ecacy and safety of sofosbuvir-based regimens in genotypes 3 and 6 infected patients, especially the impact of ribavirin co-administration on sustained virological response in cirrhotic patients with genotype 3b infection. Methods A retrospective cohort study included 173 patients initiated on sofosbuvir-based regimens. Main endpoint of treatment was sustained virological response at post-treatment week 12 (SVR12). our knowledge, only a few studies in real-world practice explored the ecacy of SOF-based regimens particularly SOF/VEL in patients with genotype 3 and 6 infection. One study conducted by Tang H provided the data on the ecacy of SOF-based regimens in Chinese patients infected with genotype 3(n = 60), the SVR24 rate was over 90% in SOF + DCV + RBV and SOF /VEL regimens except of SOF + DCV regimens. Neither sub-genotype information was provided nor the effect of ribavirin on SVR12 was analyzed. 27 Another real-world study from China showed that genotype 3a infected patients yield high and similar SVR12 rate in SOF/VEL and SOF + DCV ± RBV regimens. 28 However, the number of patients enrolled were small and the SVR12 rate on 3b sub-type with cirrhosis wasn’t shown. In present study, patients with genotype 3 infection achieved SVR12 rates over 90% even high to 100% except for SOF + RBV and SOF/VEL regimens. Among cirrhotic patients infected with genotype 3b(n = 13), 84.6% was treated with RBV co-administration regimens, and nobody failed to achieve SVR12.

Europe but limited in China, especially regimens of SOF/VEL with or without ribavirin (RBV) . 19,21−23 Therefore, there is an urgent need for more real-world research from China to be conducted.
The purpose of present real-world study was aimed to explore the e cacy and safety of SOF ± RBV, SOF + daclatasvir (DCV) ± RBV and SOF/velpatasvir (VEL) ± RBV regimens in Chinese patients with genotypes 3 and 6 infection, especially whether RBV co-administration have impact on SVR in patients with GT3b infection and cirrhosis.

Methods study design and participants selection
This is a retrospective cohort study revolving a total of 173 patients initiated on SOF-based regimens in routine practice from the Second A liated Hospital of Chongqing Medical University between July 2017 and January 2020. Patients infected with genotypes 3 and 6 were initiated on 6 kind of SOF-based regimens, including SOF, SOF+DCV and SOF/VEL regimens with RBV or not for regular duration according to the guideline, availability of DAAs and the physician. Inclusion criteria for screening participants was 18 years or older with hepatitis C virus infection, regardless of hepatitis B virus (HBV) co-infection, cirrhosis status and treatment experience.

Data collection
The following information were collected at baseline: general demographic feature(gender, age), viral related data (HCVgenotype and level of serum HCV-RNA), clinical laboratory examinations(the level of aspartate aminotransferase (AST), alanine aminotransaminase (ALT), total bilirubin, total bile-acid, albumin, hemoglobin, platelet count and AFP), assessment of cirrhosis status and comorbidity (diabetes, chronic renal diseases, fatty liver) and co-infection with HBV. The serum HCV-RNA was determined by Roche Cobas Ampliprep/ Cobas TaqMan or Kehua. The liver cirrhosis was diagnosed by APRI index over 2 or FIB-4 index over 3.25 or liver stiffness over 14.6 kPa, or ultrasonography imaging and examination indicated.

Outcome
The e cacy of treatment was determined by the sustained virological response at post-treatment week 12 (SVR12) de ned as the main endpoint of treatment, the improvement of brosis evaluated by FIB-4 ((Age × AST)/ (PLT × (square root of ALT)), and APRI (AST/PLT*100). The clinical symptoms and adverse events were monitored during the treatment and follow-up for safety and tolerance assessment.

Statistic analysis
The deference of continuous variables between pre-treatment and post-treatment were compared by paired t-test. The Chisquare test was performed in comparison of categorical data. Variables that were associated with liver brosis progression with a p-value < 0.1 were included in multivariate logistic regression models. Two sided P values were calculated for all tests. And P value < 0.05 was considered statistically signi cant. Data were analyzed via SPSS version 24.0 (IBM Corp.,Armonk, NY, USA).

Baseline characteristics
A total of 173 patients infected with genotypes 3 (n=93) and 6 (n=80) received SOF-based regimens were enrolled at initiation, whose baseline demographic and clinical characteristics were shown in Table1. 24.9% patients with cirrhosis, 7.7% were treatment experienced. The percentage of patients received the SOF±RBV, SOF+DCV±RBV and SOF/VEL±RBV regimens for regular duration accounting for 22.5%, 13.9% and 63.6%. There was no statistic difference in age, gender, level of ALT, total bilirubin, total bile-acid, albumin, hemoglobin, AFP, serum RNA, HBV co-infection and treatment experience except for higher level of AST, lower level of platelet, higher proportion of cirrhosis and RBV co-administration in patients with genotype 3 infection.
We further analyzed the impact of factors such as, gender, age, cirrhotic status, viral load, treatment history, HBV co-infection and level of baseline ALT on SVR12 in patients with genotypes 3 and 6 infection. Data on Fig.4 shew above-mentioned factors had no impact on SVR12 in both genotypes 3 and 6.
High SVR12 rate in sub-type 3b Data on Fig.5 showed that patients infected with 3b sub-type achieved a SVR12 rate of 97.1% (34/35), accounting for similar and high SVR12 rates irrespective of gender, age, cirrhosis status, treatment history, viral load, HBV co-infection and level of baseline ALT. Moreover, there was no signi cant difference in SVR12 rates between 3b and non-3b genotype, which was 96.5% (55/57). Among cirrhotic patients infected with 3b sub-type, the SVR12 rate was 100% (13/13) , 85 % of these patients were treated with RBV to SOF-based regimens. The detailed clinical process of those patients (n=12) were described in Fig.6. One patient treated with SOF/VEL+RBV for 12 weeks was not be presented because the time of undetectable HCV-RNA was not sure.
Patients failed to achieve SVR12 Totally, three patients with good compliance failed to achieve SVR12, all of whom are treated naive and with elevated level of AST at baseline, including 2 non-cirrhotic patients with genotypes 3b and 6a infection treated with SOF+RBV regimen for 24 weeks and SOF/VEL regimen for 12 weeks, one cirrhotic patient with genotype 3k infection treated with SOF/VEL regimen for 12 weeks with alcohol intake during the period of medication. Unfortunately, we didn't take further action, such as testing for drug resistance or occult HCV infection, to identify the cause of treatment failure.
Higher SVR12 rate in RBV co-administration regimen The baseline characteristics in patients treated with RBV or not were presented in Table 2. Patients treated with RBV coadministration to SOF-based regimens achieved the SVR12 rate of 98.4% (61/62), numerically higher than 93.3% (28/30) in RBV free regimen. We compared the SVR12 rates in patients treated with or without RBV in diverse characteristics such as genotype, viral load, cirrhosis status HBV co-infection and treatment history, nally RBV co-administration mildly increase SVR12 rate but no statistic difference was found. (Fig.7) Of note, SVR12 rate was achieved in 96.6% among patients with cirrhosis (28/29), 100% with RBV co-administration regimens and 87.5% without RBV, comparing to 97.6% and 95.5% in patients without cirrhosis (61/63), respectively.

Antiviral treatment contributes to brosis improvement
The data on Fig.8 showed the APRI was signi cantly reduced after treatment as compared to that at baseline (2.04 versus 0.56, p<0.0001) and the similar result was found in FIB-4(3.43 versus 2.15, p<0.0001). We further compared the scores of APRI and FIB-4 between pre-treatment and post-treatment among patients with different characteristics, which indicated that patients with HBV co-infection, treatment experienced and normal ALT level at baseline gained no signi cant decrease in APRI or FIB-4 scores ( Table 3 and 4). Also we found that APRI or FIB-4 scores were increased in some patients (n=15). The multivariate logistic regression (Table 5) showed that patients with normal ALT level at baseline was identi ed as a negative predictor of brosis improvement (p=0.01, OR=6.668).
SOF-based regimens were well-tolerated SOF-based treatment was well-tolerated by patients. Adverse events (AEs) occurred in 13.29% of patients, and no severe AEs, death occurred and no discontinuation due to AEs. And there were no signi cant differences in 6 treatment regimens (Table 6).

Discussion
In present study, we assessed that SOF-based regimens achieved satisfactory virological response and were well-tolerated in Chinese patients infected with genotypes 3 and 6 in real-world practice. Although higher proportion of cirrhosis, higher level of AST and lower platelet count at baseline were found in patients infected with genotype 3 comparing to genotype 6, similar and high SVR rates were estimated in two genotypes. The SVR12 rates are similar and high in diverse SOF-based regimens. And the addition of RBV treatments yield numerically superior SVR12 rates than RBV free regimens irrespective of genotype, cirrhosis status and treatment experience.
SOF (an NS5B nucleotide polymerase inhibitor), DCV (an NS5A replication complex inhibitor, which is not FDA-approved but has been studied widely), and SOF/VEL(a new pangenotypic drug, which is the combination of NS5A and NS5B nucleotide polymerase inhibitor), highly effective in HCV infection and approved in treatment among patients with genotypes 3 and 6 infection according to the guideline and clinical trials. 15,24 However, it's reported that SVR12 rates in real-world practice were not same satisfactory as comparing to clinical trials. 25 And a phase III clinical trial from Asia showed unsatisfactory SVR12 rate of SOF/VEL treatment on patients infected with genotype 3b, signi cant inferior than other genotypes, especially in cirrhotic patients. 26 In China, to our knowledge, only a few studies in real-world practice explored the e cacy of SOF-based regimens particularly SOF/VEL in patients with genotype 3 and 6 infection. One study conducted by Tang H provided the data on the e cacy of SOF-based regimens in Chinese patients infected with genotype 3(n = 60), the SVR24 rate was over 90% in SOF + DCV + RBV and SOF /VEL regimens except of SOF + DCV regimens. Neither sub-genotype information was provided nor the effect of ribavirin on SVR12 was analyzed. 27 Another real-world study from China showed that genotype 3a infected patients yield high and similar SVR12 rate in SOF/VEL and SOF + DCV ± RBV regimens. 28 However, the number of patients enrolled were small and the SVR12 rate on 3b sub-type with cirrhosis wasn't shown. In present study, patients with genotype 3 infection achieved SVR12 rates over 90% even high to 100% except for SOF + RBV and SOF/VEL regimens. Among cirrhotic patients infected with genotype 3b(n = 13), 84.6% was treated with RBV co-administration regimens, and nobody failed to achieve SVR12.
RBV, has been shown to improvement in e cacy of pegylated interferon, which was recommended in patients with decompensated cirrhosis in the time of DAA therapy according to guideline. 24,29 It's still controversial about the role of RBV on antiviral e cacy in DAA therapy. Data from a systematic review showed that the addition of RBV to SOF/VEL and SOF + DCV signi cantly increased the e cacy and side effect meanwhile. 10 And another data from several studies reported that RBV coadministration had no impact on antiviral e cacy. 18,19 In present study, RBV were more frequently used in patients with genotype 3 infection than genotype 6, which obtained similar and high SVR12 rates and side effects. Additionally, RBV coadministration to three SOF-based regimens achieved numerically superior SVR12 rate in genotype 3 and 6 infected patients with cirrhotic or non-cirrhotic, comparing to RBV free regimens. Moreover, there was evidence for increased occurrence of HCC in patients treated with RBV free SOF-based regimens. 30 Accordingly, this may provide some of evidence for the treatment for genotype 3 and 6 in China, that is RBV co-administration to SOF-based regimens should be applied for more satisfactory SVR12 and lower incidence of HCC. Likewise, it's contributes to achieving the ambitious goal of eliminating HCV by 2030.
It's well known that antiviral treatment contributes to brosis improvement. And data from a long-term follow-up study showed that high ALT at baseline predicts the brosis progress. 31 A real-world study from China rst reported that liver brosis evaluated by APRI and FIB-4 were signi cantly improved in patients with genotype 3 infection. 27 In present study, we also found that brosis improvement evaluated by APRI and FIB-4 occurred in most of patients, including patients with cirrhotic or high viral load but not patients with HBV co-infection, treatment experienced or normal level of ALT at baseline, which was diametrically opposed to that in above study has been mentioned. In our opinion, it's an issue of great concern to monitoring of cirrhotic status for patients with following characteristics, treatment experienced, HBV co-infection or normal level of ALT at baseline. And more research should be conducted to identify the predictors like baseline ALT of brosis progress.
Although the analysis on the percentage of RVR suggest that patients treated with SOF/VEL + RBV regimens achieved signi cantly lower RVR rate. The SVR12 rate wasn't signi cantly different in diverse regimens. which indicated that the RVR had no impact on SVR12. This may provide some implications for the surveillance during treatment, that is there is no necessary to detect the viral load in serum during early stage of antiviral therapy. Maybe twice tests are feasible, before antiviral treatment and post-treatment 12 weeks, when multiple detection of viral load is unavailable and can't be affordable.
There are some other insu ciency in present retrospective study. Only 82% patients detected serum viral load with Roche Cobas Ampliprep/ Cobas TaqMan, whose lowest limit of quanti cation (LLOQ) is 15 IU/mL at post-treatment 12 weeks or with Kehua detection whose LLOQ is 1000 IU/mL at post-treatment 24 weeks. And for patients who failed to achieve SVR12, we didn't take further action to con rm the reasons, such as alcohol intake, drug resistance, and others. We didn't use the liver biopsy to assess the degree of liver brosis improvement, which may affect the results in some way.

Conclusion
Real-world data show that SOF-based regimens are highly effective and safe for patients with HCV genotypes 3 and 6 infection. Availability of data and materials: The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

Competing interests:
The authors report no con icts of interest in this work.

Funding:
This work was prepared independently without nancial support.

Authors' contributions:
PH conceived the study, LW and XQL retrieved and analyzed the data, and all authors contributed to the writing of the manuscript.   in patients with divers characteristics between pre-treatment and post-treatment. Notes: Continuous variables are expressed as mean and SD. We compared the APRI (AST/PLT*100) in patients with divers characteristics between pre-treatment and post-treatment. Notes: Continuous variables are expressed as mean and SD (standard deviation), and categorical variables as n. The variables that were associated with liver brosis progression with a p-value < 0.1 were included in multivariate logistic regression models. a: aspartate aminotransferase to platelet ratio index; b: brosis-4.  The RVR rates in patients treated with diverse regimens. We compared the rapid virological response (RVR) rates between 6 treatment regimens, including sofosbuvir (SOF), SOF+ribavirin (RBV), SOF+daclatasvir (DCV), SOF+DCV+RBV, SOF/velpatasvir (VEL) and SOF/VEL +RBV regimens. Difference between SOF and SOF/VEL+RBV regimens: p=0.015. Difference between SOF+DCV+RBV and SOF/VEL+RBV regimens : p=0.028. Difference between other treatment regimens: P>0.05.  The SVR12 rates in patients infected with genotypes 3 and 6 in diverse regimens. Overall, the difference in SVR12 rate between six treatment regimens: P>0.05; For patients infected with genotype 3, the difference in SVR12 rate between six treatment regimens: P>0.05; For patients infected with genotype 6, the difference in SVR12 rate between six treatment regimens: P>0.05 The SVR12 rates in patients infected with genotype 3 and 6 in different characteristics. We analyzed the impact of factors such as, gender, age, cirrhotic status, viral load, treatment history, HBV co-infection and level of baseline ALT on SVR12 in patients with genotypes 3 and 6 infection. For genotype 3, no statistic difference was found between above factors: p>0.05. The similar results were found in patients infected wiith genotype 6. The SVR rates in patients infected with sub-type 3b with different characteristics. We analyzed the impact of factors such as, gender, age, cirrhotic status, viral load, treatment history, HBV co-infection and level of baseline ALT on SVR12 in patients with sub-type 3b infection, no statistic difference was found between above factors: p>0.05.

Figure 7
The SVR12 rates in patients treated with or without RBV in different characteristics. We compared the SVR12 rates in cirrhotic, non-cirrhotic, treatment-experienced (treated with DAAs or interferon and RBV regimens), treatment-naive, high viral load ( level of serum HCV-RNA> 800000IU/mL), low viral load (HCV-RNA≤ 800000IU/mL), and two genotypes. In patients with above characteristics, no statistic difference was found between groups with RBV and without RBV: p>0.05.

Figure 8
The improvement in brosis evaluated by APRI and FIB. We compared the APRI (AST/PLT*100) and FIB-4 ((Age × AST)/ (PLT × (square root of ALT)) between pre-treatment and post-treatment. Difference in APRI between pre-treatment and post-treatment: p<0.0001, similar result was found in FIB-4.