Chronic hepatitis C is a major reason for development of cirrhosis and hepatocellular carcinoma and a leading cause for liver transplantation. The development of direct-acting antiviral agents lead to (pegylated) interferon-alfa free antiviral therapy regimens with a remarkable increase in sustained virologic response (SVR) rates and opened therapeutic options for patients with advanced cirrhosis and liver graft recipients. This concise review gives an overview about most current prospective trials and cohort analyses for treatment of patients with liver cirrhosis and liver graft recipients. In patients with compensated cirrhosis Child-Pugh-Turcotte (CTP) class A, all approved agents are safe and SVR rates do not significantly differ from patients without cirrhosis in general. In patients with decompensated cirrhosis CTP class B or C, daclastasvir, ledipasvir, velpatasvir, and sofosbuvir are approved, and SVR rates higher than 90% can be achieved. Especially for patients with a model of end stage liver disease score higher than 15 and therefore eligible for liver transplantation, data is scarce. Reported SVR rates in patients with cirrhosis CTP class C are lower compared to patients with a less severe liver disease. In liver transplant recipients with a maximum of CTP class A, SVR rates are comparable to patients without LT. Patients with decompensated graft cirrhosis should be treated on an individual basis.

The World Health Organization estimates that approximately 150 million people worldwide are currently infected with the hepatitis C virus (HCV)[1]. Chronic hepatitis C is a major reason for development of cirrhosis and hepatocellular carcinoma (HCC) in Eastern Asia, Europe, and North America and a leading cause for liver transplantation (LT)[2-4]. In patients with HCV associated cirrhosis, the annual incidence of HCC ranges from 1% to 7%[5-7].

Sustained virologic response to antiviral therapy, defined as undetectable HCV-RNA 12 wk (SVR₁₂) after end of treatment (EOT), is associated with improved survival and reduced risk of HCC development in patients without and with cirrhosis[3,8,9]. The former standard treatment for HCV infection based on the combination of (pegylated) interferon-alfa [(peg)-IFN] and ribavirin (RBV) for 12 to 72 wk was associated with limited SVR rates and burdened by therapy associated adverse events[10,11]. In patients with HCV associated cirrhosis registered for liver transplantation (LT), SVR rates after (peg)-IFN ± RBV range between 0% and 50%, depending inter alia on HCV genotype and severity of liver disease[12]. In patients with decompensated cirrhosis, (peg)-IFN based antiviral therapy is contraindicated[12]. In liver transplant recipients, SVR rates to (peg)-IFN based antiviral therapy are low, and interferon-alfa-associated, immune-mediated graft dysfunction is a major concern[12,13].

The development of direct-acting antiviral agents (DAA) for (peg)-IFN free antiviral therapy regimens lead to a remarkable increase in SVR rates and opened therapeutic options for patients with contraindications or low SVR rates to (peg)-IFN based antiviral therapy regimens[14,15]. Currently available DAA can be classified according to the viral target protein as NS3/4A protease inhibitors (PI), NS5B polymerase inhibitors (including non-nucleoside and nucleoside inhibitors) and NS5A-inhibitors[16]. Antiviral regimens containing elbasvir (EBR)/grazoprevir (GZR), simeprevir (SMV), and ombitasvir (OMV)/paritaprevir/r (PTV/r) - with or without dasabuvir (DSV) - are approved by the European Medicine Agency (EMA) and/or the Federal Drug Administration (FDA) with or without RBV for treatment of patients without and with cirrhosis maximum Child-Pugh-Turcotte (CPT) class A. Regimens containing daclatasvir (DCV), ledipasvir (LDV), velpatasvir, and sofosbuvir (SOF) are approved with or without ribavirin (RBV) for patients with all stages of liver disease including cirrhosis CPT class B and C (Table 1). As SOF is not approved for patients with severe renal impairment, there is still no interferon-free therapy regimen approved for patients with decompensated cirrhosis and severe kidney disease. This article gives a concise review on current data of DAA-based antiviral therapy in patients with advanced HCV associated liver disease.

A PubMed database research using the terms “hepatitis C”, “cirrhosis” and “direct acting antiviral” was performed by the date of 21^th of February 2016 to identify relevant clinical studies as well as national and international guidelines. The systematic research resulted in 226 hits, from those we identified 16 original articles, 10 case reports/case series, 137 reviews, 3 national guidelines, 8 articles presenting data from DAA trials without participation of cirrhotic patients, and 51 articles investigating other topics than DAA therapy. One article was listed twice. The PubMed research was amended by studies not fully published but known to the authors, and references listed in systematically identified articles. Totally, 27 trials were identified, which reported safety and efficacy of DAA-based antiviral therapy in patients with compensated and decompensated cirrhosis (Table 2)[17-40].

Additionally, a second PubMed database research using the terms “hepatitis C”, “liver transplantation” and “direct acting antiviral” was performed to identify relevant clinical studies as well as national and international guidelines dealing with patients in the liver transplant setting. This systematic PubMed research revealed 72 publications, from those we identified 2 original articles, 3 case reports/series, 45 reviews, one national guideline and 21 articles investigating other topics than DAA therapy or animal model studies. The PubMed research was amended by studies not fully published but known to the authors, and references listed in systematically identified articles. In total, 6 trials were identified including also studies known to the authors as congress proceedings and not yet fully published.

The majority of prospective phase II and III trials included only a limited number of patients with cirrhosis[17,22,24-26,28,29,32-34,37-40], and only few trials investigated especially patients with (decompensated) cirrhosis[18-20,23,27,30,31,35]. Data of patient subgroups with cirrhosis were not reported discretely in the majority of studies, including, but not focusing on cirrhotic patients. Additionally to prospective, controlled trials, safety and efficacy of DAA regimens were recorded in “real life” cohort studies and compassionate use or early access programs[41-51]. Data obtained from early access and compassionate use programs must be interpreted with caution, because treatment duration and regimens, e.g., use of RBV, are mostly not controlled by respective protocols. Nevertheless, patients with decompensated cirrhosis or high MELD score (≥ 16) were enrolled in a substantial number in these trials, and therefore, these data are of interest and will be presented in the respective sections of this review.

All currently approved agents can be administered safely in patients with CPT class A cirrhosis in interferon-free antiviral regimens[18,31,35,52-60]. However, with respect to genotype and DAA regimen, some details have to be considered.

Randomized, prospective trials of approved DAA regimens were performed in patients with advanced liver disease [(decompensated) cirrhosis, after LT] and mainly HCV genotype 1, 3, and 4 infection. Although some studies investigated patients with “decompensated” cirrhosis, data are scarce in patients with a MELD score higher than 15[69]. From the limited data may be concluded that patients with CPT class C cirrhosis have lower SVR₁₂ rates than patients without or with compensated cirrhosis. Improvements of MELD score as well as albumin levels indicate an improvement in liver function in patients with DAA induced SVR[31,70]. However, liver function stayed unchanged in a significant proportion of patients or even worsened despite of SVR. Moreover, serious adverse events including fatal courses have been reported in patients with decompensated cirrhosis and DAA based antiviral therapy ranging from 18% to 34%[19,20]. Most trials did not report a benefit in SVR rates for 24 wk of treatment over 12-wk duration[18,22]. However, some data indicate similar SVR₁₂ rates for a treatment of 12 or 24 wk only with additional RBV in the 12 wk arm[20]. The relevant studies are discussed in detail below.

The randomized open-label phase II studies SOLAR-1 and SOLAR-2 evaluated the combined treatment with LDV/SOF and RBV for 12 or 24 wk in patients with HCV GT 1 or 4 and advanced cirrhosis or post liver transplantation[19,31]. Totally 337 patients were enrolled in the SOLAR-1 study, 89% GT 1 and 11% GT 4. Data of 108 patients with decompensated cirrhosis have been reported supplementary prior to the final publication[19,70]. Patients had CPT class B (n = 55) or C (n = 53) cirrhosis, and 26% of patients (28) had a MELD score > 15. At therapy baseline, 96% of patients with cirrhosis CPT class C had ascites. The overall SVR₁₂ rate was 87%, and did not differ significantly in patients with CPT class B cirrhosis between the 12 wk (SVR, 87%) and the 24 wk (89%) duration arm. In patients with cirrhosis CPT class C, SVR₁₂ rates were 86% and 87% in the respective study arms. Of note, 4 patients with cirrhosis CPT class B/C received a liver graft, 5 patients discontinued treatment because of adverse events, and one patient died. The transplant cohorts of the SOLAR-1 study and the SOLAR-2 study are discussed in the respective section below.

The combination of DCV/SOF and RBV for 12 wk was evaluated in the open label, phase III ALLY-1 study in patients with advanced cirrhosis or after LT[36]. Although all genotypes were allowed, mainly patients with genotype 1 infection were included in the cirrhosis cohort (45/60, 75%). Patients transplanted during treatment were eligible for additional 12 wk of treatment immediately post-transplant. In the cirrhosis cohort, CPT class distribution was 20% class A, 53% class B, and 27% class C, and MELD score ranged between 8 and 27. The overall SVR₁₂ rate in the cirrhosis cohort was 83% (50/60 patients with all GT) and did not differ with respect to prior treatment status or general baseline characteristics. However, SVR rates were higher in patients with CPT class A (10/11 GT 1 patients) or B (22/24 GT 1 patients) than in patients with class C (5/10 GT 1 patients, for other GT not separately displayed).

The combination of VEL/SOF has also been evaluated in 267 patients with decompensated cirrhosis in the phase III, open-label ASTRAL-4 study[20]. Although designed as a pan-genotype study, mainly patients with genotype 1 (78%) and 3 (15%) infection were enrolled, while patients with genotypes 2, 4, 5, and 6 were included only to a low percentage of 4%, 3%, 0%, and < 1%, respectively. Patients were randomized to 12 wk of SOF/VEL with or without RBV or 24 wk with SOF/VEL without RBV. Enrolled patients had a median baseline CPT score of 8 (range 5 to 10) and a median baseline MELD score of 10 (range 6 to 24). However, 95% of patients had a baseline MELD score of 15 or less. At screening all patients had CPT class B, but 7% of patients had CPT class A at treatment baseline. This has to be kept in mind, when transferring scientific data to clinical practice in patients with “truly” decompensated cirrhosis. Overall SVR₁₂ rates were 83% for 12 wk of SOF/VEL, 94% for 12 wk of SOF/VEL/RBV and 86% for 24 wk of SOF/VEL. SVR₁₂ rates for CPT class or subgroups were not reported separately. An improvement in CPT class was achieved in 47% of patients, 42% had no change and 11% patients showed a worsened CPT score. Overall improvement of MELD score was observed. This effect was stronger in patients with an initial MELD score > 15. In detail, 51% of patients with a baseline MELD score ≤ 15 showed an improvement of MELD score, while MELD score stayed unchanged or worsened in 22% and 27% of patients, respectively. Patients with a baseline MELD > 15 showed an improved MELD score in 81%, while MELD score stayed unchanged or worsened in 11% and 7%, respectively. Although these are encouraging data, it has to be considered, that (severity of) portal hypertension, a major risk factor of death in patients with cirrhosis, is not adequately reflected by MELD. The fixed combination of VEL/SOF is approved by the EMA and the FDA for 12 wk in combination with RBV.

As mentioned above, data from early access and compassionate use programs, as well as data from “real life” cohorts are useful supplements to prospective controlled trials, because inclusion criteria often allowed to include a substantial number of patients with advanced liver disease, e.g., decompensated cirrhosis. Final data are currently available from the National Health Service England Expanded Access Programme, which included 467 patients with hepatic decompensation or life-threatening extrahepatic manifestations[49]. In this prospective, observational cohort study, patients were treated for a maximum of 12 wk with DCV/SOF ± RBV or LDV/SOF ± RBV by clinician’s discretion. The majority (409/467, 88%) of patients had past or current symptoms of hepatic decompensation, defined by presence of ascites, variceal bleeding or encephalopathy. At baseline, 319 patients were classified as CPT class B, and 43 as CPT class C, and median (range) MELD score was 11 (6-32). Overall SVR₁₂ was 82% (381/467); 4% (17/467) of patients died and 3% (16/467) of patients were lost to follow-up. In detail, SVR₁₂ rates were 85% (39/46) for patients with genotype 1 infection treated with DCV/SOF ± RBV and 92% (170/185) for patients treated with LDV/SOF ± RBV, respectively. Patients with genotype 3 infection achieved SVR₁₂ in 73% (91/125) when treated with DCV/SOF ± RBV and in 61% (41/67) treated with LDV/SOF ± RBV[49]. Of note, DCV is currently approved in patients with cirrhosis and HCV genotype 3 infection in combination with SOF and RBV by EMA for 24 wk and by FDA for 12 wk treatment[62,63].

It is unquestionable, that efficacy and safety has improved with DAA based IFN-free therapies compared to (peg)-IFN based therapy regimen in patients with decompensated cirrhosis. For a subgroup of patients with decompensated cirrhosis and DAA induced SVR, an improvement in liver function has been reported. Nevertheless, severe adverse events were reported in a substantial percentage of patients, and ascites or hepatic encephalopathy did not resolve completely in all patients[43,44,49,71]. Currently, no biomarkers are available to predict the individual clinical course in patients with decompensated cirrhosis to decide whether treatment should be initiated before or after LT[49,72].

SMV is not approved by the EMA or the FDA for patients with cirrhosis CPT class B or C. Nevertheless, treatment of patients with advanced cirrhosis with SMV/SOF with or without RBV has been reported[43,44]. Saxena et al[43] presented data of a multicenter cohort of 106 GT 1 patients with cirrhosis treated with SMV/SOF with or without RBV for 12 wk. Patients were analyzed according to CPT class A versus CPT class B/C and compared to matched untreated controls. The median (range) baseline MELD score in the overall study population was 9 (8-11), and cirrhosis was classified as CPT class A in 64% and CPT class B/C in 35%. Overall, SVR rates were 91% in patients with CPT class A and 73% in patients with CPT class B/C cirrhosis, respectively[43]. Two deaths occurred, one in each group, one liver related and one not. Shiffman et al[44] reported retrospectively analyzed data from 120 patients with cirrhosis and HCV genotype 1 infection treated with SMV/SOF. From those 67%, 21% and 12% had CPT class A, B, and C cirrhosis, respectively. Hepatic decompensation or portal hypertension was present in 30% of patients. The overall SVR₁₂ rate was 87%, 77%, and 67% in CPT class A, B, and C patients, respectively. Serious adverse events were reported in 11% of patients, sepsis (n = 2, 1/2 fatal outcome), variceal bleeding (n = 2), hepatocellular carcinoma (n = 2), and increase in bilirubin (n = 8). Although these data indicate that SMV might be safe in patients with decompensated cirrhosis, there is general concern because the simeprevir mean steady state area under the curve is increased in patients with CPT class B and C cirrhosis by 2.4-fold and 5.2-fold, respectively (summary of product characteristics).

Curry et al[21] conducted an open-label phase II study for patients on the waiting list to LT for HCC of any GT. Endpoint of the study was undetectable HCV RNA 12 wk after LT. Patients received SOF and RBV up to 48 wk before LT, 61 patients with CPT ≤ 7 were included. LT was performed in 46 patients, 43 from those had undetectable HCV RNA at transplant and were included in efficacy analysis. Outcome of those 43 patients was: 30 patients (70%) had SVR12 after LT, 10 patients (23%) suffered from relapse, 3 patients (7%) died (primary graft non-function, n = 2; hepatic artery thrombosis, n = 1). Overall SVR post LT from all 61 patients was 49%. The risk of HCV graft infection was negatively correlated with the time interval before LT, when HCV RNA was undetectable. Safety and efficacy of LDV/SOF are currently investigated in patients with genotype 1 or 4 infection in a peri-transplant setting[73].

In patients with detectable HCV RNA at transplantation, HCV graft infection is almost inevitable, and HCV infection of the liver graft often shows an aggravated course with development of graft cirrhosis in up to 30% percent of patients within 5 years after transplantation[74,75]. In a minority of patients, HCV graft infection presents as fibrosing cholestatic hepatitis (FCH), a severe form of hepatitis C, leading to graft loss and death within months up to 2 years in the majority of patients[76]. Therefore, HCV graft infection is associated with a decrease in patient and graft survival[77].

Patients with fibrosis or cirrhosis after LT have a poor tolerance and low efficacy to (peg)-IFN based antiviral therapy[78]. Moreover, plasma cell hepatitis is a rare but feared complication of (peg)-IFN therapy[76,79]. First generation PI - in combination with (peg)-IFN and RBV - were associated with a slight increase in SVR rates, but a high rate of serious adverse events[80,81]. The introduction of DAA-based IFN-free antiviral therapy widened therapeutic options in patients after liver transplantation and a proof of concept study using SOF with RBV for 24 wk resulted in an overall SVR rate of 70%[82]. Currently safety and efficacy data of DAA based antiviral therapy in liver graft recipients are available from prospective trials[19,31,36,83,84] and cohort studies[77,85-90]. Table 3 summarizes available prospective and controlled trials. The most important prospective trials are the SOLAR-1 and -2 studies (LDV/SOF/RBV), the CORAL-I study (3D/RBV), and the ALLY-1 trial (DCV/SOF/RBV)[19,31,36,84].

The combination of LDV and SOF with or without RBV was investigated in several studies[19,31,83]. Reddy et al[83] performed a prospective, randomized multicenter study to evaluate treatment with LDV/SOF with RBV for 12 wk vs 24 wk in patients with GT1 und GT4 in liver graft recipients at median (range) 4.4 (0.4-23.3) years after LT. Two-hundred-twenty-three patients were enrolled, 83% had prior HCV treatment, 47% of all patients were PI treatment experienced. Fifty percent of patients had F0-F3 fibrosis, while 23%, 22% and 4% had CPT class A, B, and C cirrhosis, respectively. SVR rates were high in patients without (12 wk, 96%; 24 wk, 94%) as well as in patients with graft cirrhosis (12 wk, 92%; 24 wk, 84%). Overall safety was excellent, however, 5 patients with graft cirrhosis died during the study period due to gastrointestinal bleeding, multiorgan failure, intestinal perforation, cardiac problems, complications of cirrhosis and progressive multifocal leukoencephalitis. These early results implicated no difference between 12 and 24 wk of LDV/ SOF plus RBV.

The SOLAR-1 and -2 studies demonstrated high efficacy and an excellent safety profile of LDV/SOF for 12 or 24 wk in patients without and with graft cirrhosis. Both studies have to be highlighted, because cirrhosis stage was well classified, and all stages of severe liver disease after LT namely graft cirrhosis CPT class C and FCH were included. The overall SVR₁₂ rates were > 90%. A longer treatment of 24 wk was not superior over 12 wk with respect to SVR in both studies. Nevertheless, lower SVR rates were observed in patients with decompensated graft cirrhosis[19,31].

Thirty-four patients after liver transplantation without and with mild fibrosis (≤ F2) were enrolled in an open label, phase II study (CORAL-I) by Kwo et al[84] receiving OBV/PTV/r/DSV plus RBV for 24 wk. Patients had a calcineurin inhibitor based immunosuppression. The overall efficacy assessed by SVR₁₂ was high with 33/34 (97%). Reported adverse events were mostly mild, fatigue, namely headache, and cough. Of note, substantial dosage modifications of immunosuppressants were required to maintain respective therapeutic levels.

The post-transplant cohort of the ALLY-1 open label, phase III study included 53 HCV-infected patients of any genotype to the treatment of 12 wk with DCV/SOF and RBV[36]. Seventy-seven percent of patients had HCV genotype 1 infection. No patients with FCH or decompensated graft cirrhosis were included. The efficacy was high with an overall SVR₁₂ rate of 94%. In the compassionate use program of DCV/SOF also patients with FCH or decompensated graft cirrhosis were included, however, the total number (n = 12) of respective patients included was low[87]. In this program, patients were treated with DCV/SOF in equal parts with or without RBV for 24 wk. Nine of twelve completed 24 wk of treatment, while three patients died before end of treatment. Preliminary post-treatment data were available for five patients, and so far no viral relapse was reported. Of note, dose adjustment of immunosuppressants was not necessary during treatment.

Robust data for the use of SMV/SOF in liver graft recipients are available from cohort studies, only[77,86,90]. The largest study enrolled 132 patients with HCV genotype 1 infection at median (range) 32 (2-317) months after LT, who were treated with SMF/SOF with and without RBV for 12 wk[86]. Overall, 60% of patients were infected with genotype 1a, 30% had METAVIR F3-F4, 4% had decompensated graft cirrhosis, and 11% had FCH. Furthermore, 7% of patients had also a kidney transplant, and 82% had previously failed (peg)-IFN/RBV-based regimens. Immunosuppression contained tacrolimus in 91%. Overall SVR₁₂ rate was 90%. However, patients with genotype 1a infection and advanced fibrosis (METAVIR F3-F4) had significantly lower SVR rates (71%) than those with F0-F2 (91%). Twenty-five patients received RBV (20%) with no significant impact on SVR. However, 72% patients developed clinical relevant anemia. One death - possibly due to drug induced lung injury - occurred, all other adverse events were classified mild. Minimal dose adjustments in immunosuppression were necessary[86]. A further, but smaller study (n = 42) is of interest, because 14% of patients with decompensated graft cirrhosis were included. Overall 95% of patients achieved SVR₁₂, 97% of patients without and 88% of patients with cirrhosis, respectively (P = NS).

The calcineurin inhibitors ciclosporin and tacrolimus are substrates of CYP3A[91-93]. Clinical significant interactions with ciclosporin and tacrolimus have been described for SMV and OMV/PTV/r. For co-administration of SMV and tacrolimus monitoring of tacrolimus trough blood levels is recommended, while again the co-administration of SMV and ciclosporin is not recommended, because SMV levels may raise (summary of product characteristics). Dose reduction and respective drug monitoring of ciclosporin (20% of daily dose) and tacrolimus (fixed 0.5 mg weekly dose) is recommended for co-administration with OBV/PTV/r ± DSV. Clinical significant drug-drug interactions with DAA have not been reported for mycophenolate mofetil. However, persistent anemia with need of blood cell transfusions was reported due to combined treatment with mycophenolate mofetil and peg-IFN/RBV/SMV[94].

There are increasing data reporting DAA based, (peg)-IFN free treatment of patients with HCV associated cirrhosis. In patients with compensated cirrhosis CPT class A, all approved agents are safe and SVR rates do not significantly differ from patients without cirrhosis[95]. In patients with decompensated CPT class B/C cirrhosis, DCV, LDV, VEL and SOF alone or in combination with RBV are safe, and SVR rates > 90% can be achieved. For most patients, a treatment course of 12 wk with or without RBV is considered sufficient. In patients with severest cirrhosis (CPT class C, MELD > 15), data from randomized trials are scarce. However, SVR rates seem lower compared to patients with less severe liver disease, and yet no (bio)markers are available to predict the further clinical outcome.

In patients with HCV graft infection after LT mostly open label trials and cohort analyses, and only few randomized trials are available. Data are conclusive that SVR rates are not different to patients without LT and maximum CPT class A cirrhosis. Patients with decompensated graft cirrhosis should be treated on an individual basis. Moreover, DAA based therapy is relatively safe in patients after LT, and therapy discontinuations due to therapy side effects are rare. Nevertheless, some challenges are to overcome. Potential drug-drug interactions - especially with immunosuppression - and concomitant impaired renal function have to be considered. A cautious surveillance during antiviral therapy is advisable to identify infections and immediately administer antibiotic treatment.

In summary, while all approved agents are eligible for patients with CPT class A cirrhosis, only 4 agents - DCV, LDV, VEL and SOF - are currently approved for patients with all severity of liver disease including CPT class B and C cirrhosis. Liver graft recipients with compensated liver disease can be treated according to patients without prior LT. The standard treatment duration for the majority of patients is 12 wk.