Volume 88, Issue 10 pp. 1659-1671
Review
Full Access

Clinical impact of the hepatitis C virus mutations in the era of directly acting antivirals

Nicola Coppola

Corresponding Author

Nicola Coppola

Department of Mental Health and Public Medicine, Section of Infectious Diseases, Second University of Naples, Naples, Italy

Correspondence to: Nicola Coppola, Department of Mental and Public Health, Section of Infectious Diseases, Second University of Naples, Via L. Armanni 5, 80133 Naples, Italy. E-mail: [email protected]

Search for more papers by this author
Carmine Minichini

Carmine Minichini

Department of Mental Health and Public Medicine, Section of Infectious Diseases, Second University of Naples, Naples, Italy

Search for more papers by this author
Mario Starace

Mario Starace

Department of Mental Health and Public Medicine, Section of Infectious Diseases, Second University of Naples, Naples, Italy

Search for more papers by this author
Caterina Sagnelli

Caterina Sagnelli

Department of Clinical and Experimental Medicine and Surgery, Second University of Naples, Naples, Italy

Search for more papers by this author
Evangelista Sagnelli

Evangelista Sagnelli

Department of Mental Health and Public Medicine, Section of Infectious Diseases, Second University of Naples, Naples, Italy

Search for more papers by this author
First published: 18 March 2016
Citations: 22

Abstract

Introduced in 2013–2014, the second- and third-wave directly acting antivirals (DAAs) have strongly enhanced the efficacy and tolerability of anti-HCV treatment, with a sustained virological response (SVR) in 90–95% of cases treated. The majority of patients who did not achieve an SVR were found to be infected with HCV strains with a reduced susceptibility to these drugs. Indeed, the high error rate of the viral polymerase and a fast virion production (100-fold higher than the human immunodeficiency virus) result in a mixture of viral genetic populations (quasi-species) pre-existing treatment initiation. These mutants occur frequently in the NS5A region, with a moderate frequency in the NS3/4A region and rarely in the NS5B region. Treatment-induced resistant mutants to NS5A DAAs persist for years after treatment discontinuation, whereas those resistant to the NS3 DAAs have a shorter duration. This review focuses on the type and prevalence of viral strains with a reduced sensitivity to DAAs, their clinical impact and influence on the response to treatment and, consequently, on treatment choice for DAA-experienced patients. J. Med. Virol. 88:1659–1671, 2016. © 2016 Wiley Periodicals, Inc.

INTRODUCTION

The World Health Organization (WHO) has estimated that 130–170 million people are infected with hepatitis C virus (HCV) and that more than 350,000 people die each year of HCV-related liver diseasesin the world [Mohd Hanafiah et al., 2013].

Acute hepatitis by HCV is frequently asymptomatic and when symptomatic it is characterized form by nausea, malaise and jaundice. Acute HCV infection shows a chronicization in about two-thirds of the cases [Orland et al., 2001; Sagnelli et al., 2014a], and usually is characterized by an indolent course or a slow progression to liver cirrhosis and hepatocellular carcinoma [Sagnelli et al., 2013]. In some cases, however, a spontaneous acute exacerbation [Sheen et al., 1996; Rumi et al., 2005; Sagnelli et al., 2005, 2014b], induces disease progression. Host factors (sex, age at in ection, immune response, routes of transmission, and genetic background), viral factors (HCV genotype and viral quasi-species), co-morbidities (viral co-infection, insulin-resistance, liver steatosis, and an immunosuppressive clinical condition) and lifestyle factors (alcohol intake) may speed the progression of the disease to liver cirrhosis and HCC [Seeff, 2002; Sagnelli et al., 2014; 2014a, 2015a,2015b].

The combination of pegylated interferon (Peg-IFN) and ribavirin (RBV) was the treatment of choice for chronic hepatitis C (CHC) for nearly a decade [Ghany et al., 2009; AISF, 2010; EASL, 2011; Coppola et al., 2012; Sagnelli et al., 2014c]. This combination therapy provided a sustained sustained viral response (SVR) in half of the patients infected with HCV-genotype 1, in around 80% of those with HCV-genotype 2 and around 65% of those with HCV-genotype 3.

In 2011, the directly acting antivirals (DAAs) NS3/4A protease inhibitors telaprevir and boceprevir were approved to treat HCV-genotype-1 infection, each in triple combination with Peg-IFN and RBV [Jacobson et al., 2011; Poordad et al., 2011; Coppola et al., 2014b]. This triple therapy achieved a 70% SVR in CHC patients with HCV-genotype 1, but was characterized by a low tolerability and a high pill burden.

The second- and third-wave DAAs introduced in 2013–2014 enhanced the efficacy and tolerability of anti-HCV treatment [Gentile et al., 2013, 2014a,2014b; Jacobson et al., 2013; EASL, 2014; Coppola et al., 2015c], with an SVR in 90–95% of cases treated. Most of the patients who did not respond to DAA-based treatment were infected with HCV strains with drug resistance. In fact, the high error rate of the viral polymerase and a very fast virion production (100-fold higher than the human immunodeficiency virus) result in a mixture of viral genetic populations (quasi-species) pre-existing treatment initiation [Ogata et al., 1991; Neumann et al., 1998; Hutchison et al., 2014; Lontok et al., 2015]. Although investigated in several studies, some questions on HCV quasi-species with a reduced susceptibility to drugs remain answered, namely their involvement in treatment failure, their relevance in re-treatment, and the real usefulness of testing patients selected for DAA-based treatments for baseline resistance.

This review article evaluates the type and prevalence of viral strains with a reduced sensitivity to DAAs, their clinical impact and influence on the response to treatment and treatment choice in naïve and DAA-experienced patients. The article is addressed in particular to physicians who have in care patients with chronic hepatitis C in their everyday clinical practice.

HCV VIROLOGY

HCV is a positive-sense, single-stranded RNA virus with a single open reading frame (ORF) of 9.6 kb, classified in the Flaviviridae family [Choo et al., 1989]. The ORF codes for a single polyprotein of approximately 3,000 residues that cellular and viral proteases split into mature structural and non-structural proteins. Non-structural proteins, essential for the replicative viral cycle, have been chosen as targets for DAA therapy [Ruta and Cernescu, 2015].

Non-structural protein (NS) 3 is a multifunctional enzyme that possesses serine protease and RNA unwinding activities required for HCV replication. Non-structural protein 4A (NS4A) binds to the N-terminal NS3 domain to stimulate NS3 serine protease activity. Thus, the NS3/4A domains are interdependent [Rudolf et al., 2008]. The NS3/4A serine protease cleaves the HCV polyprotein, generating the NS3, NS4A, NS4B, NS5A, and NS5B proteins.

The NS5A protein is involved in viral replication, assembly and release of HCV particles, but by a virtually unknown mechanism [Hanoulle et al., 2009].

NS5B is an RNA-dependent RNA polymerase and it is structurally organized in a characteristic “right hand motif” containing palm and thumb domains. The error-prone nature of the HCV NS5B polymerase and the accumulation of mutations in a small hypervariable region in the envelope-encoding genes generate a high level of variability. This variability translates into the existence of seven major HCV genotypes (with 30–35% variation at the nucleotide level), 67 subtypes (with less than 15% difference at the nucleotide level) each composed of a myriad of viralquasi-species, and nine recombinant forms (e.g., the most frequently reported genotype 2 k/1b, which has multiple isolates) [Smith et al., 2014]. Each genotype exhibits a different degree of variability: 7 subtypes in genotype 1; 11 in genotype 2; 6 in genotype 3; 17 in genotype 4; 24 in genotype 6; and only 1 subtype in genotypes 5 and 7 [Gower et al., 2014]. There are multiple consequences related to this enormous viral heterogeneity, such as the possibility of re-infections with different genotypes because of the very limited cross-antigenicity, the emergence of immune-escape mutants, which accounts for the high rate of chronic infections, the genotype- and subtype-specific response to treatment and the spontaneous or drug-induced selection of viral-resistant strains entailing the use of combination therapies.

DIRECTLY ACTING ANTIVIRALS

The DAAs used in the treatment of CHC are the NS3/4A protease inhibitors(PIs), the NS5A inhibitors and the NS5B inhibitors. Table I shows the main characteristics of these classes of DAAs.

Table I. Characteristics of DAAs According to the Drug Class
NS3/4A inhibitors NS5A inhibitors NS5B inhibitors
Drugs telaprevir, boceprevir, simeprevir, asunaprevir, paritaprevir, grazoprevir, vaniprevir daclatasvir, ledipasvir, ombitasvir, elbasvir velpatasvir sofosbuvir, dasabuvir, beclabuvir
HCV-genotype coverage Genotype 1 and 4 Pan-genotype Pan-genotype
Genetic barrier Low Very low High
Cross-resistance of class Yes Yes No
  • * Only for simeprevir, asunaprevir, paritaprevir, grazoprevir, and vaniprevir.

NS3/4A Protease Inhibitors

The NS3/4A PIs interact with the protease substrate-binding site and prevent the cleavage of the viral polyprotein. The HCV protease requires small molecule inhibitors to promote specific interactions to achieve tight binding with the enzyme. Thus, the first generation PIs (boceprevir and telaprevir), drugs acting only against HCV-genotype 1, had a low genetic barrier to resistance and considerable cross-resistance. The second and third generation PIs (simeprevir, asunaprevir, paritaprevir, grazoprevir, and vaniprevir) have higher genetic barriers and an antiviral activity against also 4.

NS5A Inhibitors

The mechanism by which NS5A is involved in viral packaging and assembly and regulates HCV replication remains poorly understood, as does the mechanism of action of the NS5A inhibitors. NS5A comprises three distinct structural domains [Hanoulle et al., 2009]. Domain I seems to be important for viral replication and has been crystallized as a homodimer. Domain II is involved in binding to cyclophilin A and has been postulated as playing a role in antagonizing the innate immune response to HCV. Domain III is involved in the assembly of infectious viral particles. Since NS5A is important in multiple steps of HCV replication, the NS5A inhibitors have potent anti-viral activity against all HCV genotypes; however, they have a relatively low barrier to resistance. Currently used NS5A inhibitors are daclatasvir, ledipasvir, ombitasvir, and elbasvir.

NS5B Inhibitors

Both nucleos(t) ide or non-nucleoside analog NS5B inhibitors have been developed. By mimicking polymerase nucleotide substrates, sofosbuvir, a nucleotide inhibitor with a broad activity and high resistance barrier against different HCV genotypes, is incorporated into the nascent RNA chain, causing chain termination.

The NS5B non-nucleoside analog inhibitors, dasabuvir and beclabuvir, bind outside the polymerase active site.

HCV MUTATIONS WITH A REDUCED SENSITIVITY TO DAAs

Tables II–IV show the mutations, according to the HCV genotype, in NS3/4A, NS5A and NS5B and the amino acid substitutions associated with a reduced sensitivity to DAAs.

NS3/4A Protease Inhibitors

Among patients with HCV-genotype-1, the NS3/4A mutations are detected more frequently in those with genotype 1a than with genotype 1b [Costantino et al., 2015]. The NS3/4A mutations with a reduced sensitivity to the first generation PIs, boceprevir, and telaprevir,involve the amino acid positions V36, T54, R155, A156, and D168 and cause cross-resistance. Although the more recently developed PIs have a higher genetic barrier, their antiviral activity is substantially impaired by the mutations R155 K and D168E/V/A/Y emerging during treatment, as documented in patients who did not achieve an SVR (Table II).

Table II. HCV Mutations in the NS3 Region Associated With Reduced Sensitivity to NS3 Inhibitors
Mutation Codon Reduced sensitivity to Genotype Mean fold change in resistance compared to wild-type replicon Reference
V36 M/A WT:GTG MT:ATG boceprevir telaprevir paritaprevir asunaprevir 1a–1b boceprevir = 3 (genotype 1a) paritaprevir = 2 (genotype 1a) asunaprevir = 2 (genotype 1a) Bartolini et al. [2015]; Beloukas et al. [2015]; Boglione et al. [2015]; Lontok et al. [2015]
V36L WT:GTG MT:CTT boceprevir telaprevir 1a–1b Vallet et al. [2011]; Cuypers et al. [2015]
T54S/A WT:ATC MT:TCC MT:GCT boceprevir telaprevir 1a–1b boceprevir = 2 (genotype 1a), 3 (genotype 1b) Boglione et al. [2015]; Larrat et al. [2015]; Lontok et al. [2015]; Nagpal et al. [2015]; Ruggiero et al. [2015]
T54S WT:ATC MT:TCC asunaprevir 1b Karino et al. [2013]
V55A WT:GTC MT:GCT boceprevir paritaprevir asunaprevir 1a–1b boceprevir = 3 (genotype 1b) asunaprevir = 1 (genotype 1b) Bartolini et al. [2015]; Beloukas et al. [2015]; Larrat et al. [2015]; Lontok et al. [2015]
Y56H WT:TAT MT:CAC asunaprevir paritaprevir 1b Lontok et al. [2015]
Y56H WT:TAT MT:CAC paritaprevir 1a–4 Lontok et al. [2015]
Y56L WT:TAT MT:CTT asunaprevir 1b Lontok et al. [2015]
Q80K WT:CAA MT:AAA(K) MT:CGG(R) simeprevir asunaprevir 1a–1b simeprevir = 11 (genotype 1a)–8 (genotype 1b) asunaprevir = 3 (genotype 1a)–1(genotype 1b) Paolucci et al. [2013]; Beloukas et al. [2015]; Boglione et al. [2015]; Lontok et al. [2015]
Q80R WT:CAA MT:CGG(R) simeprevir, asunaprevir 1b simeprevir = 6 asunaprevir = 1 Lontok et al. [2015]
V107I WT:GTT MT:ATC boceprevir 1a–1b Lontok et al. [2015]
S122G WT:TCC MT:GGT(G) simeprevir 1b simeprevir = 0.5 Boglione et al. [2015]; Cuypers et al. [2015]; Lenz et al. [2015]; Ruggiero et al. [2015]
S122 R WT:TCC MT:CGG(R) simeprevir 1b simeprevir = 21 Lontok et al. [2015]
S122A/I/T WT:TCC MT:GCT(A) MT:ATC(I) MT:ACC(T) simeprevir 1b Lenz et al. [2015]; Lontok et al. [2015]; Ruggiero et al. [2015]
S122D/G/I/N/T WT:TCC MT:GAT(D) MT:GGT(G) MT:ATC(I) MT:AAT(N) MT:ACC(T) asunaprevir 1b Cuypers et al. [2015]; Ruggiero et al. [2015]
I132V WT:ATC MT:GTC telaprevir simeprevir 1a telaprevir = 2 Boglione et al. [2015]; Ruggiero et al. [2015]
R155K WT:CGC MT:AAG boceprevir telaprevir simeprevir asunaprevir paritaprevir vaniprevir 1a boceprevir = 5 simeprevir = 86 asunaprevir = 21 paritaprevir = 37 vaniprevir = >1,000 Bartolini et al. [2015]; Boglione et al. [2015]; Larrat et al. [2015]; Lontok et al. [2015]
R155K WT:CGC MT:AAG Boceprevir Telaprevir 1b telaprevir = 7 simeprevir = 32 Bartolini et al. [2015]; Boglione et al. [2015]; Larrat et al. [2015]; Lontok et al. [2015]
R155M WT:CGC MT:GGC(G) telaprevir 1b telaprevir = 5 Lontok et al. [2015]
R155C/Q WT:CGC MT:TCG(C) MT:CAG(Q) boceprevir simeprevir asunaprevir vaniprevir 1b Cento et al. [2012]; Akuta et al. [2015]; Lenz et al. [2015]
R155G/M WT:CGC MT:GGC(G) MT:ATG(M) MT:ACC(T) telaprevir 1a Vallet et al. [2011]; Cento et al. [2012]; Zhouy, [2013]; Dierynck et al. [2014]
R155T WT:CGC MT:ACC(T) telaprevir 1b telaprevir = 20 Lontok et al. [2015]
R155S/F/T/V WT:CGC MT:TTT(F) MT:ACC(T) MT:GTG(V) telaprevir 1b Vallet et al. [2011]; Akuta et al. [2015]
A156S WT.GCG MT:AGC telaprevir boceprevir vaniprevir 1a Zhouy, [2013]; Dierynck et al. [2014]; Lontok et al. [2015]
A156S WT.GCG MT:AGC telaprevir boceprevir 1b boceprevir = 13 telaprevir = 9.6 Lontok et al. [2015]
A156T WT.GCG MT:ACC telaprevir boceprevir 1b telaprevir = >62 simeprevir = 37 Aloia et al. [2015]; Buti et al. [2016]; Paolucci et al. [2015]
A156V WT.GCG MT:GTG telaprevir 1a Aissa Larousse et al. [2014]; Akuta et al. [2014]; Ezat et al. [2014]; Lontok et al. [2015]
A156T/V WT.GCG MT:ACC(T) MT:GTG(V) telaprevir boceprevir 1b Aissa Larousse et al. [2014]; Akuta et al. [2014]; Ezat et al. [2014]; Cao et al. [2015]; Larrat et al. [2015]
A156F WT.GCG MT:TTC telaprevir 1b telaprevir = >62 Jiang et al. [2013]
A156N WT.GCG telaprevir 1b Telaprevir = >93 Jiang et al. [2013]
A156G WT.GCG MT:GGA simeprevir 1b Simeprevir = 19 Cento et al. [2012]; Cao et al. [2015]; Jensen et al. [2015]
V156F WT:GTG MT:TTT boceprevir 1a–1b Boglione et al. [2015]
V158I WT: GTG MT:ATA boceprevir 1a–1b Qiu et al. [2009]; Lontok et al. [2015]; Sarrazin, [2015]
D168N WT:GAT MT:AAT telaprevir boceprevir vaniprevir 1a–1b telaprevir = 0.6 (genotype 1b) simeprevir = 5.5 (genotype 1b) Jiang et al. [2013]; Lontok et al. [2015]
D168E WT:GAT MT:GAA simeprevir paritaprevir asunaprevir 1a–1b asunaprevir = 58 (genotype 1a)–78 (genotype 1b) paritaprevir = 14 (genotype 1a) simeprevir = 38 (genotype 1b) Cento et al. [2012]; Lenz et al. [2015]; Ruggiero et al. [2015]; Sarrazin, [2015]
D168V WT:GAT MT:GTC vaniprevir simeprevir paritaprevir 1a asunaprevir = 373 paritaprevir = 96 Cento et al. [2012]; Akuta et al. [2014]; Poordad et al. [2014]; Ruggiero et al. [2015]; Schnell et al. [2015]
D168V WT:GAT MT:GTC asunaprevir paritaprevir vaniprevir simeprevir 1b simeprevir = 3,100 asunaprevir = 280 paritaprevir = 159 vaniprevir = 725 Cento et al. [2012]; Akuta et al. [2014]; Lontok et al. [2015]; Ruggiero et al. [2015]; Schnell et al. [2015]
D168V WT:GAT MT:GTA paritaprevir 4 Cento et al. [2012]; Lontok et al. [2015]
D168V WT:GAT MT:GTA paritaprevir 4d Lontok et al. [2015]; Schnell et al. [2015]
D168A WT:GAT simeprevir asunaprevir paritaprevir vaniprevir 1a asunaprevir = 23 Paritaprevir = 50 Lenz et al. [2015]
D168A WT:GAT MT:GCG paritaprevir simeprevir vaniprevir 1b simeprevir = 784 asunaprevir = 127 varitaprevir = 27 Akuta et al. [2014]; Cao et al. [2015]
D168T WT:GAT MT:ACC asunaprevir saniprevir 1a Akuta et al. [2014]; Lontok et al. [2015]
D168T WT:GAT MT:ACC simeprevir vaniprevir 1b simeprevir = 334 McPhee et al. [2013]; Soumana et al. [2014]; Lenz et al. [2015]
D168H WT:GAT MT:CAC simeprevir vaniprevir 1a Cento et al. [2012]; Lenz et al. [2015]
D168H WT:GAT MT:CAC asunaprevir simeprevir 1b simeprevir = 401 asunaprevir = 98 Cento et al. [2012]
D168F WT:GAT MT:TTC simeprevir asunaprevir vaniprevir 1b Lenz et al. [2015]
D168K WT:GAT MT:AAA paritaprevir 1b paritaprevir = 882 Lontok et al. [2015]
D168Y WT:GAT MT:TAC asunaprevir paritaprevir vaniprevir 1a asunaprevir = 622 paritaprevir = 219 McPhee et al. [2013]; Herzer et al. [2015]; Lontok et al. [2015]; Pilot-Matias et al. [2015]
D168Y WT:GAT MT:TAC asunaprevir 1b asunaprevir = 238 McPhee et al. [2012]; McPhee et al. [2013]; Herzer et al. [2015]; Lontok et al. [2015]; Pilot-Matias et al. [2015]
D168E WT:GAT MT:GAC paritaprevir simeprevir asunaprevir 1a asunaprevir = 58 paritaprevir = 14 Shindo et al. [2012]; Akuta et al. [2014]; Itakura et al. [2015]; Lontok et al. [2015]
D168E WT:GAT MT:GAA simeprevir asunaprevir 1b simeprevir = 38 asunaprevir = 78 McPhee et al. [2012]; Shindo et al. [2012]; Akuta et al. [2014]; Lontok et al. [2015]
D168G WT:GAT MT:GGT asunaprevir 1b Lenz et al. [2010]; Akuta et al. [2014]; Lontok et al. [2015]
D168N/S WT:GAT MT:AAC(N) MT:AGC(S) vaniprevir telaprevir simeprevir 1b telaprevir = 0.6 (D168N) simeprevir = 5.5 (D168N) Lenz et al. [2010]; Jiang et al. [2013]; Lontok et al. [2015]
Q168A/E/H/N WT:CAA MT:GAA MT:CAC MT:AAC telaprevir boceprevir simeprevir asunaprevir vaniprevir 4 Jensen et al. [2015]
I/V170A WT:ATT(I) WT:GTG(V) MT:GCG boceprevir asunaprevir 1b boceprevir = 7.9 Cento et al. [2012]; Boglione et al. [2015]; Cao et al. [2015]
I/V170T WT:ATT(I) WT:GTG(V) MT:ACC boceprevir simeprevir 1a Cento et al. [2012]; Boglione et al. [2015]
I/V170T WT:ATT(I) WT:GTG(V) MT:ACC boceprevir 1b Vallet et al. [2011]; Cento et al. [2012]; Boglione et al. [2015]
I/V170T WT:ATT(I) WT:GTG(V) MT:ACC boceprevir 1a Cento et al. [2012]; Boglione et al. [2015]
V170M WT:GTG MT:ATG asunaprevir 1b Karino et al. [2013]

The spontaneous NS3/4A mutations are relatively frequent but do not affect the efficacy of treatment, except for the Q80 K substitution. This substitution, more frequent in patients with HCV-genotype 1a than with genotype 1b, has been associated with a high treatment failure in the simeprevir/Peg-IFN-based regimens; in the IFN-free regimen including simeprevir this effect is less pronounced, causing treatment failure only in a relatively small percentage of patients with genotype 1a and cirrhosis. In fact, in 116 CHC patients with HCV-genotype-1a without cirrhosis who received simeprevir + sofosbuvir ± ribavirin, an SVR was achieved with similar frequency by the 46 with the baseline Q80 K variant and by the 70 without (96% vs. 97%) [Kwo et al., 2015]. Instead, of the 72 patients with cirrhosis harboring HCV-genotype 1a, an SVR was less frequently achieved by the 34 with Q80 K at baseline than by the 38 without (74% vs. 92%) [Lawitz et al., 2015a]. Therefore, testing for the Q80 K baseline variant is useful only for cirrhotic patients with genotype 1a for whom a treatment regimen including simeprevir is under consideration.

No gold standard therapy to re-treat patients who have failed previous PI-based treatment has as yet been identified, but some of the IFN-free regimens have been shown to be highly effective in this setting. In fact, around 95% of non-responders to a combined therapy with Peg-IFN, ribavirin plus boceprevir or telaprevir achieved an SVR when re-treated with sofosbuvir + ledipasvir ± ribavirin for 24 months [Afdhal et al., 2014].

NS5A Inhibitors

The HCV mutations involving the amino acid positions M28, Q30, L31, H58, and Y93 reduce the sensitivity to the NS5A inhibitors, which have a very low genetic barrier (Table III).

Table III. HCV Mutations in the NS5A Region Associated With Reduced Sensitivity to NS5A Inhibitors
Mutation Codon Reduced sensitivity to Genotype Mean fold change in resistance compared to wild-type replicon Reference
M28T WT:ATG MT:ACG daclatasvir ledipasvir ombitasvir elbasvir 1a daclatasvir = 205 ledipasvir = 61 ombitasvir = 8,965 Liu et al. [2015]; Premoli and Aghemo, [2015]; Sarrazin, [2015]
M28V WT:ATGb MT:GTA daclatasvir ombitasvir 1a–4 ombitasvir = 58 (genotype 1a) Fridell et al. [2011]; Lawitz et al. [2012]; Bartels et al. [2013]; Paolucci et al. [2013]; Cuypers et al. [2015]; Premoli and Aghemo, [2015]; Sarrazin, [2015]
M28A/S WT:ATG MT:GCG(A) MT:TCT(S) daclatasvir 1a Lontok et al. [2015]; Premoli and Aghemo, [2015]; Sarrazin, [2015]; Smith et al. [2016]
L28M/T WT:CTC MT:ATG(M) MT:ACG(T) daclatasvir 1b Fridell et al. [2011]; Lawitz et al. [2012; Bartels et al. [2013]; Karino et al. [2013]; Paolucci et al. [2013]; Cuypers et al. [2015]
L28M WT:CTC MT:ATG daclatasvir 4 McPhee et al. [2014]; Cuypers et al. [2015]; Lontok et al. [2015]
L28V/S WT:CTC MT:GTG MT:AGC ombitasvir 4 Paolucci et al. [2013]; Schnell et al. [2015]
P29S WT:CCA MT:AGC daclatasvir 1b Lontok et al. [2015]
P29 DELETION WT:CCA daclatasvir 1b Lontok et al. [2015]
Q30E/H WT:CAA MT:GAG(E) MT:CAT(H) daclatasvir ledipasvir 1a McPhee et al. [2014]; Lindström et al. [2015]; Lontok et al. [2015]; Premoli and Aghemo, [2015]; Sarrazin, [2015]
Q30R WT:CAA MT:CGG daclatasvir ombitasvir ledipasvir elbasvir 1a daclatasvir = 365 ombitasvir = 800 ledipasvir = 632 Poordad et al. [2014]; Liu et al. [2015]; Premoli and Aghemo, [2015]; Sarrazin, [2015]
Q30/D/G/K/T WT:CAA MT:GAC(D) MT:GGC(G) MT:AAG(K) MT:ACG(T) daclatasvir 1a Lontok et al. [2015]; Premoli and Aghemo, [2015]; Sarrazin, [2015]
R30G/H/P/Q WT:CGG MT:GGA(G) MT:CAT(H) MT:CAA(Q) daclatasvir 1b Cuypers et al. [2015]; Lontok et al. [2015]
L30H/S WT:TTA MT:CAT(H) MT:AGC(S) daclatasvir 4 Lontok et al. [2015]
L31M WT:CTG MT:ATG daclatasvir ledipasvir 1a, 1b daclatasvir = 105 (genotype 1a) ledipasvir = 554 (genotype 1a) daclatasvir = 3 (genotype 1b) McPhee et al. [2012]; Plaza et al. [2012]; Bartels et al. [2013]; Karino et al. [2013]; Bartolini et al. [2015]; Beloukas et al. [2015]; Cuypers et al. [2015]; Nagpal et al. [2015]; Premoli and Aghemo, [2015]; Sarrazin, [2015]
A30K WT:GCC MT:GCA elbasvir 3a Liu et al. [2015]
L31I/V WT:CTG MT:ATA(I) MT:GTG(V) daclatasvir 1a daclatasvir = 100 (L31V) McPhee et al. [2012]; Kai et al. [2015]; Liu et al. [2015]
L31V/F/I WT:TTA MT:GTG(V) MT:TTC(F) MT:ATC(I) daclatasvir 1b daclatasvir = 15 (L31V) McPhee et al. [2012, 2014]; Liu et al. [2015]; Sarrazin, [2015]
L31/F WT:TTA MT:TTT elbasvir 3a Liu et al. [2015]
P32L WT:CCT MT:CTG daclatasvir 1b Plaza et al. [2012]; Wang et al. [2014]
H58D WT:CAC MT:GAC daclatasvir ledipasvir ombitasvir 1a ledipasvir = 1,127 ombitasvir = 243 Lontok et al. [2015]
P58S WT:CCG MT:TCA daclatasvir 1b Aissa Larousse et al. [2015]
A92K WT:GCC MT:AAA daclatasvir 1b Vazquez-Chacon et al. [2014]; Lontok et al. [2015]
Y93H/N/C WT:TAC MT:CAC MT:AAT MT:TGT daclatasvir ledipasvir ombitasvir elbasvir 1a daclatasvir = 1,600 (Y93H) ledipasvir = 1,677 (Y93H) ombitasvir = 41,383 (Y93H) Suzuki et al. [2012]; McPhee et al. [2012, 2014]; Itakura et al. [2015]; Kai et al. [2015]; Lindström et al. [2015]; Liu et al. [2015]; Sarrazin, [2015]
Y93S WT:TAC MT:AGC ledipasvir 1a Sarrazin, [2015]
Y93H WT:TAC MT:CAC daclatasvir ledipasvir ombitasvirm elbasvir 1b daclatasvir = 12 ombitasvir = 77 Karino et al. [2013]; McPhee et al. [2014]; Poordad et al. [2014]; Itakura et al. [2015]; Kai et al. [2015]; Lindström et al. [2015]; Liu et al. [2015]; Sarrazin, [2015]; Wang et al. [2015]
Y93C WT:TAC MT:TGC ledipasvir 1b Nakamoto et al. [2014]; Lontok et al. [2015]
Y93H WT:TAC MT:CAC daclatasvir elbasvir 3a Liu et al. [2015]; Lontok et al. [2015]; Sarrazin, [2015]

The NS5A mutations may be detectable as natural baseline polymorphisms, and in these cases the degree of resistance is of clinical importance [Krishnan et al., 2015a]. For instance, the Y93H/N mutation has a very high fold-change in EC-50 (>100) for all NS5A inhibitors in genotype 1a, but only for ledipasvir in genotype 1b. In genotype 1a, the Q30R and L31M/V mutations are associated with a high resistance to all NS5A inhibitors, while the M28T is associated with resistance only to daclatasvir and ombitasvir. Instead, in genotype 1b the L31V mutation is associated with a low resistance to all NS5A inhibitors.

Although scanty, the data suggest that only highly resistant mutations at baseline have a clinical relevance in therapy-naïve patients treated with NS5A inhibitors. In fact, in 123 patients treated with sobosbuvir plus ledipasvir, the presence of NS5A mutations with a fold-change in EC-50 of more than 100 was associated with a reduced SVR prevalence in the 46 receiving a short-term regimen (8 weeks for naïve patients and 12 weeks for non-responders) compared with the 77 treated for a longer period (12 weeks in naïve and 24 in experienced) (76% vs. 96.1%), whereas for the 1,020 patients in the same study with mutations with a fold-change in EC-50 of less than 100 or no mutation, the SVR was high in all treatment regimens (95–100%) [Sarrazin et al., 2014].

Recently, Zeuzem evaluated the impact of baseline NS5A resistance mutations by ultra-deep sequencing in 1,566 patients treated with ledipasvir-sofosbuvir-based regimens and found their presence associated with a less frequent SVR only in patients with cirrhosis or in non-responders to a previous antiviral treatment [Zeuzem et al., 2015]. In addition, by population sequencing Jacobson evaluated the prevalence and clinical impact of baseline NS5A resistant mutants in patients with genotype 1a and 1b treated with elbasvir/grazoprevir-based regimens. The baseline prevalence of NS5A resistant mutants was 15–42% and theelbasvir-specific prevalence was 2–32%. The impact of these resistant mutants on the SVR was negligible in genotype 1b patients regardless of the presence of elbasvir-resistant mutants(SVR in 98% and 100% of cases, respectively), whereas the impact was strong in those with genotype 1a with elbasvir- (SVR 58%) or NS5A- (SVR 86%) resistant mutants [Jacobson et al., 2015].

The type and number of NS5A mutations also have an important role in terms of frequency of SVR at the time of re-treatment of patients with previous failure to NS5A inhibitors. For instance, in 41 patients with genotype one with a failed sofosbuvir/ledipasvir regimen administered for 8 or 12 weeks, retreatment with the same drugs for 24 weeks achieved an SVR in 100% of 11 patients without NS5A mutations, in 69% of 16 with one mutation and in 50% of 14 with 2 or more mutations. Of the 16 patients with only 1 NS5A mutation, the sofosbuvir/ledipasvir retreatment regimen achieved an SVR in all 5 with Q30R or M28T, in 4 of 5 with L31M and in only 2 of 6 with Y93H/N [Lawitz et al., 2015b].

Another important point is the durability of HCV mutations. HCV mutations with reduced sensitivity to NS5A inhibitors may persist for years. Dvory-Sobol et al. [2015] showed that in 86% of 58 patients with ledipasvir failure, the NS5A mutations persisted for 96 weeks after this antiviral treatment was discontinued. Krishnan et al. [2015b] confirmed these data and suggested a different durability of the NS5A and NS3/4A mutations. In patients with therapy failure for the combination of paritaprevir, ombitasvir, and dasabuvir, the authors detected NS3/4A mutations in 46% of 67 patients tested 48 weeks after treatment discontinuation and in 9% of 57 tested after 96 weeks, whereas the NS5A mutations were still present in 97% of 70 patients and in 96% of 51 tested respectively at weeks 48 and 96 after treatment discontinuation.

The information available on NS3/4A and NS5A mutations should guide treatment decisions for therapy-experienced patients. In patients with mild or moderate chronic hepatitis with a previous failed treatment regimen including any NS5A inhibitors (e.g., daclatasvir or ledipasvir plus sofosbuvir or ombitasvir/parataprevir-ritonavir plus dasabuvir) watchful waiting for further information and/or for a more effective oncoming antiviral treatment may be a rational choice. Instead, in patients with cirrhosis or with other urgencies requiring early treatment, the tests to detect NS3 and NS5A mutations should be performed quickly to identify an effective treatment. If only NS5A mutations are detected, simeprevir plus sofosbuvir, or ledipasvir plus sofosbuvir or ombitasvir/paritaprevir-ritonavir plus dasabuvir regimens, possibly associated with ribavirin [Reddy et al., 2015], may be considered. If only NS3A mutations are present, ledipasvir or daclatasvir plus sofosbuvir, possibly associated with ribavirin, should be considered the best choice [Reddy et al., 2015]. If both NS3A and NS5A mutations are present, this unfortunate patient should be treated in specific trials. Recently, 22 patients not responding to different DAA-based regimens (ombitasvir/paritaprevir-ritonavir/dasabuvir, or telaprevir + peg-IFN + ribavirin, or sofosbuvir + ribavirin, or sofosbuvir + Peg-IFN + ribavirin, or simeprevir + sofosbuvir or simeprevir + samatasvir + ribavirin) were enrolled in a pilot study. These patients were treated with ombitasvir/paritaprevir-ritonavir/dasabuvir plus sofosbuvir ± ribavirin for 12 weeks if infected with genotype 1b (regardless of the presence of cirrhosis) or genotype 1a without cirrhosis, and for 24 weeks if infected with genotype 1a with cirrhosis. The SVR rate was very high, ranging from 92% to 100% in these three groups [Poordad et al., 2015].

NS5B Inhibitors

Owing to different mechanisms of action, no cross-resistance is observed across currently approved nucleotide and non-nucleoside polymerase inhibitors (Table IV).

Table IV. HCV Mutations in the NS5B Region Associated With Reduced Sensitivity to NS5B Inhibitors
Mutation Codon Reduced sensitivity to Genotype Mean fold change in resistance compared to wild-type replicon Reference
L159F WT:CTT MT:TTT sofosbuvir 1a, 1b, 2, 3a sofosbuvir = 1 Chopp et al. [2015]; Svarovskaia et al. [2016]
S282T WT:AGT MT:ACC sofosbuvir 2b sofosbuvir = 16 Chopp et al. [2015]
V321A WT:GTG MT:GCG sofosbuvir 1a, 3a Chopp et al. [2015]; Svarovskaia et al. [2016]
C316Y WT:TGT MT:TAC dasabuvir 1a–1b dasabuvir = 1,472 (genotype 1a) dasabuvir = 1,569 (genotype 1b) Poordad et al. [2014]; Sarrazin, [2015]
C316N WT:TGT MTAAT sofosbuvir 1b Wang et al. [2015]
M414T WT:ATG MT:ACA dasabuvir 1a–1b dasabuvir = 32 (genotype 1b) Poordad et al. [2014]; Sarrazin, [2015]
Y448H WT:TAC MT:CAC dasabuvir 1a dasabuvir = 975 (genotype 1a) Lontok et al. [2015]
A553T WT:GCT MT:ACG dasabuvir 1a dasabuvir = 152 (genotype 1a) Lontok et al. [2015]
S556G WT:AGC MT:GGC dasabuvir 1a–1b dasabuvir = 30 (genotype 1a) dasabuvir = 11 (genotype 1b) Cuypers et al. [2015]
G554S WT:CGA MT:TCA dasabuvir 1a dasabuvir = 198 (genotype 1a) Lontok et al. [2015]
D559G/N WT:GAC MT:GGA MTAAT dasabuvir 1a Lontok et al. [2015]
A421V WT:GCG MT:GTG beclabuvir 1a Cuypers et al. [2015]
P495L/S WT:CCG MT:TTG MT:AGC beclabuvir 1a Lontok et al. [2015]

Sofosbuvir has a high genetic barrier, as confirmed by the observation that of the patients who failed sofosbuvir therapy, both in the trials and in clinical practice, only a few showed a development of resistant mutants. For instance, a baseline C316N/H/F was found in 6HCV-genotype-1b patients with treatment failure and in only one patient with HCV-genotype 1a who had experienced a relapse. In the ELECTRON study using sofosbuvir as monotherapy, the S282T substitution was detected in a HCV-genotype 2 patient who relapsed 4 weeks after the treatment [Gane et al., 2013]. In a analysis of sofosbuvir phase-3 trials involving HCV-genotype 3 patients with non-response, the mutations L159F and V321A were more frequently detected. These variants showed a 1.2- to 1.6-fold reduced susceptibility to sofosbuvir only in vitro, an observation warranting further investigation to define any possible association with viral resistance [Donaldson et al., 2014; Svarovskaia et al., 2014].

Recently, the emergence of L159 and V321 substitutions in the NS5B region in 403 patients with failed sofosbuvir-based treatment was analyzed by ultra-deep sequencing [Svarovskaia et al., 2016]. L159F was detected in 15% (53 of 353) and V321A in 5% (17 of 353) of patients with virological failure, while a sofosbuvir plus ledipasvir combination reduced the emergence of L159F or V321A to 2% (1 each of 50) at virological failure. L159F and V321A did not jeopardize the outcome of retreatment with sofosbuvir, ribavirin, and pegylated interferon.

Regarding dasabuvir, a non-nucleoside analog inhibitor, the M414T and S556G mutations were frequently detected in patients infected with HCV-genotype 1a who did not achieve an SVR, while the S556G was frequently observed also in genotype-1b-infected patients [Poordad et al., 2014; Zeuzem et al., 2014].

As regards beclabuvir, another non-nucleoside analog inhibitor, the NS5B amino acid substitutions A421V and P495L/S were frequently observed in genotype-1a—infected patients who did not achieve an SVR [Sims et al., 2014], whereas no patient infected with genotype 1b has so far experienced a virological failure.

CONCLUSIONS

More than 90% of patients treated with the recently introduced IFN-free DAA therapies achieve an SVR. However, due to the high variability of the HCV genome, numerous resistant mutants, both spontaneous, and treatment-induced, are produced during HCV replication. These mutants arise frequently in the NS5A region, with a moderate frequency in the NS3/4A region and rarely in the NS5B region. Treatment-induced resistant mutants to NS5A DAAs persist for years after treatment discontinuation, whereas those resistant to the NS3 DAAs have a shorter duration.

Despite the impressive amount of information on viral resistant mutants that has become available in recent years, the clinical impact of the reduced sensitivity to DAAs of the resistant strains has not been fully elucidated. The information at hand, however, suggests that the identification of HCV mutations is not necessary for DAA-naïve patients, with the exception of patients not responding to previous antiviral treatment not including DAAs or those with genotype 1a and cirrhosis, for whom sequencing in the NS3/4A and in NS5A regions is plausible. Instead, before starting a second-line DAA treatment, the identification of HCV mutants in the NS3/4A, NS5A, and NS5B regions is necessary for all patients failing to respond to a previous DAA-based regimen.

    The full text of this article hosted at iucr.org is unavailable due to technical difficulties.