Introduction

Hepatitis C virus (HCV) infection is an ongoing global health problem with 58 million people living with chronic HCV infection and an estimated 290,000 deaths related to HCV in 2019 according to the World Health Organization [1]. In Taiwan, the estimated prevalence of anti-HCV seropositivity was around 3–4% among the general population with an estimated 40,000 of them living with active viremic infection [2]. HCV has been a leading cause of cirrhosis, end-stage liver disease, and liver-related deaths worldwide [3]. Extrahepatic manifestations in multiple systems including but not limited to renal (membranoproliferative vasculitis), neurologic (peripheral neuropathies), rheumatologic (polyarthritis), immune (B-cell lymphoma and cryoglobulinemic vasculitis) and cutaneous organs (purpura and necrotizing vasculitis) can also occur in up to 70% of patients with HCV infection [4, 5]. These hepatic and extra-hepatic manifestations have contributed to the huge burden of disease that negatively impacts patient well-being and quality of life. Treatment and cure of HCV not only can prevent and even reverse both the hepatic and extra-hepatic morbidity and mortality in the individual-level but also can reduce the reservoir of those infected with HCV and thus halt HCV transmission to others in the population-level [5].

The availability of highly effective interferon-free all-oral direct-acting antivirals (DAAs) has completely revolutionized the treatments for HCV and has marked a new era in the treatment of HCV [6, 7]. This change in HCV treatment landscape also spurred the World Health Organization to make a call for the elimination of HCV as a public health threat in 2016 and subsequently laid out a roadmap to achieve this goal by the year 2030 [8]. Taiwan has accelerated its efforts to eliminate HCV since 2018 by committing to achieve the World Health Organization's 2030 goal of treating 80% of eligible patients by 2025 [2]. The introduction of pangenotypic DAAs can help simplify HCV treatment pathways and decentralize testing at the primary care level and these regimens have been recommended by the World Health Organization for all eligible patients with chronic HCV infection [1, 9, 10].

Of the two first-line pangenotypic DAAs, sofosbuvir/velpatasvir is recommended for a duration of 12 weeks for HCV patients with or without decompensated cirrhosis, with the addition of ribavirin recommended in the presence of decompensated cirrhosis and glecaprevir/pibrentasvir (GLE/PIB) is recommended for a duration of 8, 12, or 16 weeks for HCV patients without liver decompensation [11,12,13]. Short-duration therapy for HCV may help reduce the burden on health care resources, overcome elimination barriers, increase treatment adherence and allow more patients to be treated [14,15,16]. GLE/PIB, a ribavirin-free, fixed-dose, once-daily, all-oral combination pangenotypic DAA regimen, is the only DAA regimen that has been approved for a short treatment duration of 8 weeks for patients with HCV of all genotypes and without cirrhosis [17]. For patients with compensated cirrhosis, the phase 3 EXPEDITION-8 trial demonstrated that the 8-week GLE/PIB regimen had a high SVR12 rate of 99.7% for treatment-naïve HCV patients with compensated cirrhosis and this 8-week regimen was subsequently approved by the US Food and Drug Administration (FDA) in September 2019 and by the Taiwan FDA in April 2020 [18]. A handful of real-world reports also showed high effectiveness and safety profiles of the 8-week GLE/PIB regimen in small-scale treatment-naïve patients with HCV and compensated cirrhosis [16, 19,20,21,22,23,24,25]. However, large-scale real-world data on this topic are scarce, especially in the Asian population. In the present study, we reported the real-world effectiveness and safety of the 8-week GLE/PIB regimen in treatment-naïve Taiwanese patients with HCV and compensated cirrhosis using the nationwide data from Taiwan Association for the Study of the Liver (TASL) HCV Registry (TACR) program [16, 26, 27].

Materials and methods

Patients and study design

The TACR is an ongoing non-interventional, observational nationwide registry program organized and supervised by TASL, which aims to set up a database and biobank of HCV patients who receive all-oral DAA therapy in Taiwan. By July 2022, 53 study sites, including 21 medical centres, 27 regional hospitals and five primary clinics from all around Taiwan, were participating in the program. Enrolling criteria for this registry program include HCV patients aged ≥ 20 years with detectable HCV RNA and received all-oral DAA-containing regimens. The program protocol was approved by the Institutional Review Board at the individual study site, which conformed to the guidelines of the International Conference on Harmonization for Good Clinical Practice. Informed consent was obtained from all patients before being enrolled. All patient records were reviewed, and data were extracted and validated at each participating site using a standardized case report form and a unified coding dictionary for the pre-defined patient (including demographics, laboratory profiles, HCV treatment data, comorbidities and cirrhotic status) and virologic characteristics (including HCV viral loads, genotypes and treatment outcomes) before and after antiviral treatment. The choice of DAA regimen was granted at the discretion of the treating physician based on the labels approved by the Taiwan FDA, which were in compliance with the local and major international guidelines for HCV infection, [12,13,14] and the regulations of the Ministry of Health and Welfare in Taiwan. The diagnosis of cirrhosis was based on any of the following modalities: liver histology; transient elastography (FibroScan®; Echosens, Paris, France) (> 12 kPa); acoustic radiation force impulse (> 1.98 m/s); fibrosis-4 (FIB-4) index (> 6.5); or the presence of clinical, radiological, endoscopic, or laboratory evidence of cirrhosis and/or portal hypertension. Inactive hepatocellular carcinoma (HCC) was defined as those who were subjected to curative therapies including surgical resection, local ablation or liver transplantation and without imaging evidence of viable HCC within 3 months prior to the commence of DAA prescription. Chronic kidney disease (CKD) was defined as a decreased estimated glomerular filtration rate (eGFR, < 60 ml/min/1.73 m2) or kidney function impairment (e.g., presence of proteinura) for more than 3 months.

Assessments and outcomes

For this study, data were analyzed as of 31 October 2021 for treatment-naïve patients with HCV and compensated cirrhosis who received 8 weeks of GLE/PIB. Effectiveness was reported as sustained virologic response at post-treatment week 12 (SVR12) which was defined as undetectable HCV RNA level (< lower limit of quantification) at off-therapy week 12. The intention-to-treat (ITT) population included all patients who received 1 or more doses of GLE/PIB and the modified ITT (mITT) population excluded patients with nonvirologic failure (ie, patients who were lost to follow-up or discontinued treatment without experiencing virologic failure). SVR12 rates were assessed in the ITT and the mITT populations. SVR12 rates were also evaluated in prespecified patient subgroups categorized according to baseline characteristics including age, gender, HCV genotype, and comorbidities such as HBV or HIV co-infection, HCC, cardiovascular disease, cerebrovascular accident, diabetes mellitus, hypertension, hyperlipidemia, people who injected drugs (PWID), and CKD and in the following subgroups of interest including patients with HCV genotype 3, platelet count < 150,000/µl, FibroScan > 20 kPa, or both platelet count < 150,000/µl and FibroScan > 20 kPa. Drug adherence, defined as the percentage of actual dosage being taken divided by the anticipated DAA dosage throughout the treatment course in each subject, was evaluated. Safety outcomes expressed as the percentages of patients with adverse events (AEs) was assessed in the ITT population. Healthcare resource utilization (HCRU) was defined as the number of clinic visits from GLE/PIB initiation to the SVR12 survey visit.

Statistical analyses

Descriptive statistics including mean (± standard deviation, SD), number or frequency (percentage) were used to summarize baseline demographics, clinical characteristics, drug adherence, AE and HCRU. The overall and stratified SVR12 rates were shown in numbers and percentages with 95% confidence interval (CI). Frequencies were compared between groups using the χ2 test with the Yates correction or Fisher’s exact test. Group means were compared using analysis of variance and Student’s t-test or the nonparametric Mann–Whitney U test when appropriate. Stepwise logistic regression analysis was performed to determine factors associated with treatment failure by analyzing the covariates with a p value < 0.1 in the univariate analysis. All statistical analyses were performed using the SPSS 12.0 statistical package (SPSS, Chicago, IL, USA). A two-tailed p-value of < 0.05 was considered significant.

Results

Patient characteristics

As shown in Fig. 1, there were a total of 8,305 treatment-naïve HCV patients with compensated cirrhosis in the TACR platform as of 31 October 2021. GLE/PIB was prescribed for 806 patients and 301 of them received the 8-week regimen for at least 1 dose (ITT). SVR12 data were not available for 26 patients and finally 275 patients were analyzed in the mITT population. As shown in Table 1, among the 301 patients of the ITT population, the average age was 63.0 ± 13.2 years, 162 or 53.8% patients were male, 142 or 47.2% patients were older than 65 years, the most common genotype was genotype 2 (166 or 55.1%) followed by genotypes 1a/1b (88 or 29.2%), genotypes 4/5/6 (21 or 7%), mixed/unclassified (18 or 6%) and genotype 3 (8 or 2.7%). The average HCV RNA was 5.8 ± 1.2 log10 IU/mL with 55 or 18.3% of patients had an HCV RNA level of > 6,000,000 IU/ml. The average alanine aminotransferase (ALT) was 77.9 ± 66.9 IU/L, aspartate aminotransferase (AST) 76.1 ± 67.3 IU/L, platelet count 154.5 ± 78.7 103 U/L, albumin 4.1 ± 0.4 g/dL, total bilirubin 0.9 ± 0.4 mg/dL, creatinine 1.0 ± 0.7 mg/dL, and FIB-4 score 5.0 ± 4.6. HBV coinfection was noted in 19 or 6.3% of patients and HIV coinfection in 7 or 2.3% of patients. Twenty five or 8.3% had HCC (11 or 3.7% of them had active HCC) and 18 or 6% of patients were people who inject drugs (PWID). Patients with Child–Pugh score of A5 and A6 were 251 (83.4%) and 41 (13.6%), respectively. The remaining 9 (3%) patients had no complete parameters for Child–Pugh classification. The most common comorbidities included hypertension (124 or 41.2% patients), CKD (82 or 27.2% patients), diabetes mellitus (73 or 24.3% patients), hyperlipidemia (39 or 13.0% patients) and cardiovascular disease (30 or 10.0% patients). The values or rates of baseline characteristics and clinical features of the mITT population were numerically comparable to those of the ITT population (Supplementary Table 1).

Fig. 1
figure 1

Study flow diagram. DAA direct-acting antiviral; GLE glecaprevir; ITT intention-to-treat; mITT modified intention-to-treat; PIB pibrentasvir

Full size image
Table 1 Baseline characteristics and clinical features (ITT) (n = 301)
Full size table

Treatment responses

The overall SVR12 rate was 98.2% (270/275) in the mITT population and 89.7% (270/301) in the ITT population. Of the 5 patients with virologic failure, four were non-responders, one was a relapser, and none had a virologic breakthrough. The characteristics of the 5 patients with virologic failure were described in detail in Supplementary Table 2. For the 26 patients with non-virologic failure, 18 were lost to follow-up, three had no available data and five died. The reasons of the five deaths included one from lung cancer with pneumonia and sepsis, two from sepsis, one from respiratory failure and one with unknown reasons (Table 2). The SVR rates for genotypes 1, 2, 3, 4/5/6, and mixed/unclassified were 97.5% (78/80), 98.1% (154/157), 100% (6/6), 100% (17/17), 100% (15/15), respectively in the mITT population (Fig. 2). For patient subgroups of interest including FibroScan > 20 kPa, platelet count < 150,000/µl, both FibroScan > 20 kPa and platelet count < 150,000/µl, APRI > 1.09, FIB4 > 3.25, active HCC and albumin < 3.5 g/dL, the SVR rates were 100% (12/12), 98.0% (144/147), 100% (7/7), 100% (125/125), 98% (145/148), 100% (9/9) and 95.8% (23/24), respectively (Fig. 3). The Fig. 4 demonstrated the SVR12 rates of patients stratified by specific comorbidities including CKD (73/77, 94.8%), diabetes mellitus (66/68, 97.1%), cardiovascular disease (27/28, 96.4%), cerebrovascular accident (7/7, 100%), hypertension (114/117, 97.4%), and hyperlipidemia (35/35, 100%).

Table 2 Overall SVR12 and reasons for non-SVR
Full size table
Fig. 2
figure 2

Rate of SVR12 stratified by different genotypes. mITT modified intention-to-treat

Full size image
Fig. 3
figure 3

Rate of SVR12 stratified by patient subgroups of interest. mITT modified intention-to-treat

Full size image
Fig. 4
figure 4

Rate of SVR12 stratified by patient subgroups of co-morbidities. mITT modified intention-to-treat; CAD coronary artery disease; CKD chronic kidney disease; CVA cerebrovascular accident; DM diabetes mellitus; HLD, hyperlipidemia

Full size image

Subgroup analysis for SVR12

As shown in Supplementary Table 3, the SVR12 rates in the mITT population were all > 95% when stratified by baseline patient characteristics including age, gender, HBV or HIV coinfection, history of HCC, HCV genotype, platelet count, albumin, PWID and fibrosis degree as depicted by Fibroscan, APRI or FIB-4 index except for HCV RNA > 6,000,000 IU/ml (SVR12 rate 92.2%, 95% CI 81.5–96.9) and the presence of CKD (SVR12 rate 94.8%, 95% CI 87.4–98.0).

Univariate and subsequent multivariate logistic regression analyses showed that factors independently associated with treatment failure included baseline HCV RNA level > 6,000,000 IU/ml (adjusted odds ratio [OR]: 20.58, 95% CI 1.98–214.10, p = 0.01), and drug adherence < 80% (adjusted OR 42.66, 95% CI 1.54–1185.30, p = 0.03). Age, gender, HCV genotype, or comorbidities such as HBV or HIV co-infection, HCC, PWID, and CKD were not associated with treatment outcomes (Table 3).

Table 3 Univariate and multivariate logistic regression analysis of factors associated with treatment failure (mITT)
Full size table

Changes of laboratory profiles

Table 4 showed the changes of laboratory profiles at week 4, end of treatment and 12 weeks off-therapy compared to their counterparts at the baseline. The laboratory profiles that changed at week 4 included total bilirubin (from 0.9 ± 0.4 to 1.0 ± 0.7 mg/dL, p < 0.01), ALT (from 77.9 ± 66.9 to 25.2 ± 20.3 IU/L, p < 0.01), and AST (from 76.1 ± 67.3 to 32.5 ± 33.4 IU/L, p < 0.01). The laboratory profiles that changed at end of treatment included INR (from 1.1 ± 0.1 to 1.03 ± 0.07 s, p < 0.01), total bilirubin (from 0.9 ± 0.4 to 1.0 ± 0.6 mg/dL, p < 0.01), albumin (from 4.1 ± 0.4 to 4.2 ± 0.4 g/dL, p = 0.03), ALT (from 77.9 ± 66.9 to 23.9 ± 18.56 IU/L, p < 0.01), and AST (from 76.1 ± 67.3 to 29.4 ± 17.6 IU/L, p < 0.01). The laboratory profiles that changed 12 weeks off-therapy included INR (from 1.1 ± 0.1 to 1.03 ± 0.08 s, p < 0.01), total bilirubin (from 0.9 ± 0.4 to 0.8 ± 0.5 mg/dL, p < 0.01), albumin (from 4.1 ± 0.4 to 4.2 ± 0.4 g/dL, p = 0.03), ALT (from 77.9 ± 66.9 to 24.1 ± 16.9 IU/L, p < 0.01), and AST (from 76.1 ± 67.3 to 29.3 ± 15.9 IU/L, p < 0.01). Notably, the serum creatinine levels did not significantly change throughout the treatment course.

Table 4 Laboratory findings before, during, and after treatment
Full size table

Safety

Of the 301 patients, 75 or 24.9% reported at least one AE as demonstrated in Table 5. Of them, 7 or 2.3% patients had serious adverse events but none was considered DAA-related. Drug discontinuation due to AEs was observed in four patients, two temporarily due to acute pancreatitis and hospitalization for unknown reasons, respectively and two permanently due to patient noncompliance and hyperbilirubinemia, respectively. Common AEs that occurred in ≥ 1% of patients included fatigue in 27 (9.0%), pruritus in 21 (7.0%), insomnia in 5 (1.7%), skin rash in 4 (1.3%), and headache in 3 (1.0%) patients.

Table 5 Safety summary (DAA, direct-acting antiviral)
Full size table

Drug adherence

Of all the 301 patients, drug adherence for 80–100%, <80%, 60–80%, 40–60%, 20–40% and < 20% was 285 (94.7%), 16 (5.3%), 4 (1.3%), 6 (2.0%), 3 (1.0%), and 3 (1.0%), respectively (Supplementary Table 4).

Healthcare resource utilization

The healthcare resource utilization (number of clinic visits from DAA initiation to SVR12) for the 301 ITT population was 5.3 ± 1.3 as shown in Supplementary Table 5.

Discussion

The high efficacy and safety profiles of the latest generation pangenotypic DAAs, GLE/PIB and sofosbuvir/velpatasvir, have resulted in a high SVR12 rate for HCV patients of all genotypes and specific populations in the individual patient level [9, 10]. To unleash the true population-level effectiveness potential of these potent oral DAA regimens, multilevel barriers including complex pre-treatment assessments, patient compliance and high HCRU will need to be addressed to help achieve the World Health Organization’s goal of eliminating HCV as a major public health threat by 2030 [8]. The universal short-duration of 8 weeks for all treatment-naïve patients, regardless of the presence of cirrhosis or not, could help minimize the burden of pre-treatment assessments and improve the accessibility to DAA therapies [18]. This benefit is of especial relevance as the epidemiology studies indicated that HCV infection has been shifting towards a predominantly treatment-naïve population [28]. The present study provided complementary results to the EXPEDITION-8 trial that 8-week GLE/PIB regimen was as effective and safe in routine clinical practice as in the registered trial for treatment-naïve HCV patients with compensated cirrhosis.

Another benefit of short treatment duration is regarding reduced HCRU. Analyzing the cost-effectiveness of healthcare costs with various treatment durations is beyond the scope of our present study. However, simplified clinic visit monitoring can enhance the capacity to scale-up DAA therapy and the COVID-19 pandemic has further highlighted the need for fewer clinic visits that minimize face-to-face contact. The mean number (mean ± SD) of clinic visits during the GLE/PIB treatment (from GLE/PIB initiation to SVR12 survey visit) was 5.3 ± 1.3 in the present study (Supplementary Table 5). This result is consistent with our previous TACR study that demonstrated the need of 5.94 ± 0.88 visits for patients receiving 8-week regimen, 6.90 ± 1.16 visits for patients receiving 12-week regimen, and 6.77 ± 1.07 visits for patients receiving 16-week regimen in a cohort of 3,144 patients receiving GLE/PIB (p < 0.0001 for 8-week regimen vs. 12-week regimen) [16].

The other potential benefit of short treatment duration is regarding the concern that treatment adherence can decline with longer treatment regimens [29]. In the present real-world study, drug adherence for all the 301 patients were 80–100% in 285 (94.7%) patients, 60–80% in 4 (1.3%) patients, 40–60% in 6 (2.0%) patients, 20–40% in 3 (1.0%), and < 20% in 3 (1%) patients. By mITT analysis, the treatment failure rates for the ≥ 80% and < 80% adherence groups were 1.5% (4/272) and 33.3% (1/3), respectively with an adjusted OR of 42.66 (95% CI 1.54–1185.30) for treatment failure for the < 80% adherence group (p = 0.03). Shortening treatment duration without compromising SVR12 rates remains an important issue in HCV therapy, especially for the special populations [29].

In addition to drug adherence, another factor associated with a higher risk of treatment failure in the present study is HCV RNA level ≥ 6,000,000 IU/ml with an adjusted OR of 20.58 (95% CI 1.98–214.10, p = 0.01) compared to HCV RNA level < 6,000,000 IU/ml. Our result is in concordance with another real-world study which demonstrated that high viral load (≥ 107 IU/ml) was a significant predictor associated with virologic failure in non-cirrhotic genotype 2 patients receiving 8-week GLE/PIB therapy [30]. These results suggest that more caution is needed with a short-course GLE/PIB regimen when treating HCV patients with a high viral load. In a meta-analysis of real-world data, the mITT SVR12 rate in treatment-naïve patients with compensated cirrhosis who received GLE/PIB treatment for 12 weeks (n = 362; 7 cohorts) was 99.0% [31]. As a result, in order to ensure that a maximal percentage of patients can attain SVR12, a 12-week GLE/PIB regimen may be considered for this subgroup of patients with a high viral load of > 6,000,000 IU/ml. Alternatively, as the virologic failure rate is low, those who failed the 8-week GLE/PIB can be treated with the 2nd line DAA. Since reports from the West did not identify high HCV viral load as a factor of GLE/PIB treatment failure for treatment-naïve patients with compensated cirrhosis, further research is needed to prove whether this is a phenomenon unique to Asian populations as well as the cost-effectiveness of management for this subgroup of patients.

Our study evaluated the changes in common laboratory parameters during and after GLE/PIB therapy. The ALT and AST rapidly declined after therapy initiation (p < 0.01). The albumin (p = 0.03) and international ratio (INR) of prothrombin time (p < 0.01) improved less rapidly at the end of treatment. The total bilirubin level slightly increased during treatment (p < 0.01) and declined to below the baseline level after achieving SVR12 (p < 0.01). All these changes in liver-related laboratory parameters were sustained after cessation of GLE/PIB treatment. This might indicate resolution of liver inflammation and improvement of liver function in accordance with the achievement of SVR12. Intriguingly, there was no significant change of creatinine levels throughout the course of GLE/PIB treatment as both GLE and PIB metabolism and clearance occur primarily in the biliary system and that renal excretion of each of the two medications is negligible.

Our study revealed a high overall SVR12 rate of 98.2% (270/275) in the mITT population. For selected special patient groups including patients with HCV genotype 3, platelet count < 150,000 /µl, FibroScan > 20 kPa, or both platelet counts < 150,000/µl and FibroScan > 20 kPa, the SVR12 rates were 100% (6/6), 98.0% (144/147), 100.0% (12/12) and 100.0% (7/7), respectively (Table 3). Among the five patients with virologic failure, one was the relapser, four were the non-responders and none had a breakthrough. Most non-virologic failure resulted from lost to follow-up or liver-unrelated mortalities (Table 2).

In line with other real-world studies, our large-scale real-world data demonstrated that 8-week GLE/PIB was well tolerated by HCV patients with cirrhosis [17, 21,22,23,24,25,26]. Although AEs were reported in as high as 24.9% (75/301) of the patients in the ITT population, almost all AEs were mild and had no clinical impacts. The most frequent AEs included fatigue (9.0%), pruritus (7.0%), insomnia (1.7%), skin rash (1.3%) and headache (1.0%) (Table 5). Seven patients (2.3%) experienced serious AEs, including five liver-unrelated deaths, one pancreatitis and one hospitalization for unknown reasons. Two patients (0.7%) permanently discontinued GLE/PIB due to either poor drug compliance or hyperbilirubinemia. None of them were considered as GLE/PIB related.

As a real-world observational study, there existed some inherent limitations for this study. First, because of the low prevalence rate of genotype 3–6 infection in Taiwan, the majority of the patient subgroups were genotypes 1 and 2. Second, because of inconsistent evaluations and incomplete assessments in real-world clinical practice, the information on AEs was subject to reporting biases, and the causal relationships between the AEs and the treatment could not be well-defined. Finally, as clinical diagnosis of liver cirrhosis is challenging, many different tools were employed in this study for diagnosis of cirrhosis.

Conclusion

This large-scale real-world Taiwanese cohort study demonstrated that 8-week GLE/PIB therapy was effective and well tolerated for treatment-naïve HCV patients with compensated cirrhosis. SVR12 rates were similarly high as in the clinical trials, including those with characteristics of advanced liver disease.