Hepatitis C virus (HCV) infection remains a key worldwide epidemiological issue with recent estimates indicating ∼80 million viremic infections on a global level.1 So far, 7 HCV genotypes and more than 80 subtypes have been identified with subtypes 1a (G1a) and 1b (G1b) being the most common, irrespective of the HCV monoinfection or HIV/HCV coinfection status.2 The HCV G1 distribution varies significantly according to geographical region, with the highest prevalence of G1a reported in North America and Oceania, whereas G1b is the major subtype in Europe, however, particularly predominant in East European countries.3,4
Recent advances in HCV treatment achieve virus elimination in ∼95%–100% of cases with the use of direct-acting antiviral (DAA) combinations.1 Treatment outcomes are affected not only by the HCV genotype a patient is infected with, but also by the presence of pre-existing resistance-associated variants (RAVs).5 Lower treatment response rates were observed, especially in the setting of the NS3 Q80K polymorphism,6 and for high-level resistance conferring NS5A variants located at codon positions 28, 30, 31, 58, and 93.5,7
Many RAVs are not natural variants but selected under DAA-selective pressure, which continue to evolve after treatment failure, often reverting back to the wild-type variant after a few months or only after years in case of NS5A.8,9 This delay in reversion may result in transmission of drug-resistant strains, however, so far infrequently reported.10,11 This may become especially important in the light of the increasing frequency of diagnoses of recent HCV infection among men-who-have-sex-with-men (MSM) with and without HIV infection, with the latter reported in candidates for HIV pre-exposure prophylaxis.12,13
In Europe, Australia, and North America, acute hepatitis C (AHC) transmission networks have been documented, associated with sex-related HCV acquisitions.14–16 In addition, among HIV-positive MSM, alarmingly high incidence rates of HCV reinfection after viral cure have been reported, ranging from 7.3 to 7.4 to 15.2/100 person-years.17 This raises concerns for the possible introduction of RAVs within a risk population that is actively fueling HCV transmission, potentially resulting in reduced first-line DAA therapy efficacy and/or limited treatment options after DAA failure.
In Poland, around 160 thousand cases are considered HCV-RNA positive,18 with ∼30% of people living with HIV being coinfected with HCV. In the last decade, the HIV-predominant transmission route shifted from injecting drug use (90% decrease) to MSM (345% increase), with approximately 65% of new infections with known transmission route linked to the group of MSM. A stable increase in the number of new HIV diagnoses per year is observed across the country, with >22.000 cases diagnosed as of December 2017.19 In the group of HIV/HCV coinfected patients, a decrease of 17.3 years in life expectancy is reported as well as a low likelihood of survival beyond the age of 65, associated with an increased risk of cardiovascular diseases, diabetes, neurocognitive performance, and kidney disease, compared with those that are HCV monoinfected.20 Although availability of DAA therapies has recently increased, access to treatment is not universal yet and is still prioritized for patients with advanced liver fibrosis, at least in Poland. Moreover, data on pre-existing variants conferring resistance to the most used anti-HCV drug classes still remain sparse, with only one Polish study investigating variants located in the NS3 region published so far,21 whereas no data are available for Polish patients carrying NS5A RAVs.
In this study, we aimed to map the genetic variability of viral proteins NS3 and NS5A in a cohort of HCV G1 patients coinfected with HIV, including patients diagnosed with acute hepatitis C. Moreover, phylogenetic inference was applied to identify clustering patterns within the Polish epidemic and to investigate the possibility of natural RAVs circulating in the identified networks.
MATERIAL AND METHODS
Samples from 112 patients, all white origin and coinfected with HIV-1 and HCV genotype 1 (G1a or G1b), the most prevalent genotype in Poland, were analyzed. Also, this genotype was the most prevalent among patients with acute hepatitis C, with other genotypes accounting for 30.1%.22 The samples were linked to care in 5 Polish HIV treatment centers (Warsaw, Wrocław, Cracow, Szczecin, Zielona Góra) and sequenced at the Pomeranian Medical University in Szczecin, Poland. The study protocol was approved by the Bioethical committee of the Pomeranian Medical University, approval number KB-0012/26/17. All samples were gathered from patients with chronic hepatitis C detected at HIV diagnosis (78/112, 66.07%) or from patients with AHC observed after HIV diagnosis (38/112, 33.92%). Patients with AHC diagnosed with genotype 1 were consecutively enrolled at the clinics participating in the study, whereas individuals with chronic hepatitis C were selected based on availability of stored leftover samples. AHC was defined as hepatitis C antibody seroconversion from negative to positive with concomitant increase of aminotransferases and detection of HCV-RNA during follow-up. Most (36/38, 94.74%) of the included patients with AHC were on stable antiretroviral therapy regimens suppressing their HIV viral load to <50 copies per mL. None of the patients were treated before with pegylated-interferon and ribavirin, nor with any DAA at the time of sample collection.
The clinical data collected for all patients included the following parameters: age, sex, HCV transmission route, lymphocyte CD4 count and HIV-RNA viral load at HIV diagnosis, year of HCV infection for the AHC cases, hepatitis B virus coinfection status (defined as hepatitis B surface antigen positive), alanine aminotransferase activity, and HCV-RNA level (last available measurement before or at the date of sampling, with a median time range between the date of sampling and the HCV-RNA level measurement being 3.7 months). History of syphilis was assumed positive if any serological test (VDRL, FTA-ABS, and TPHA) was ever recorded reactive.
NS3 and NS5A genetic sequencing was performed using the Sanger methodology as previously published.23 Amplicons obtained by nested polymerase chain reaction were used for sequencing with the BigDye technology on an ABI 3500 platform (Applied Biosystems, Foster City, CA). NS3 and NS5A sequence assembly was performed with the Recall online tool, and variants were scored above a threshold of 15%. The final data set included 112 patients and consisted of 108 NS3 (108/112 = 96.4%) and 102 NS5A sequences (102/112 = 91.07%), as sequencing experiments, either targeting NS3 or NS5A, repetitively failed for the remaining samples. Of the successfully amplified sequences, 98 cases (87.5%) were characterized by the presence of paired NS3/NS5A sequences. All sequences generated have been submitted to GenBank (pending sequence IDs).
Genotypic Drug Resistance and Phylogenetic Analyses
The G1 subtype (a or b), clade I or II (in case of G1a), and NS3 and NS5A RAVs were identified using the rules-based algorithm geno2pheno HCV v.0.92. For interpretation of drug resistance, any variant scored as “reduced susceptibility” or “resistant” to known NS3 or NS5A inhibitors was included in the analysis. HCV genotype and subtype were confirmed using the Oxford HCV Automated Subtyping Tool and the COMET subtyping tool.
For phylogenetic analysis, only the data set of 98 paired NS3/NS5A sequences was used, as short fragments are often associated with poor phylogenetic signal. First, joint NS3/NS5A sequences were aligned with the software Clustal X 2.0.10, separately for G1a (n = 40) and G1b (n = 58), using 2 reference sequences (G1a HQ850279 to root G1b tree and G1b D90208 to root G1a tree) as outgroup. Alignments were edited in MEGA 7.0 to improve their quality, whereas sequences containing stop codons were removed. The alignments covered NS3 codons 1-181 for both G1a and G1b, and NS5A codons 6-103 for G1a and 1-103 for G1b.
To investigate clustering patterns, data sets were supplemented with a selection of highly similar sequences obtained from the public database GenBank (see Supplemental Digital Content, http://links.lww.com/QAI/B110 for the Genbank IDs), using the software tool BLAST. More in detail, for each sequence from the initial data set, 10 sequences with the highest similarity score and with 100% NS3 and NS5A sequence coverage were selected. After removal of duplicates, the final data sets consisted of 136 G1a and 151 G1b sequences. To construct phylogenetic trees, a maximum likelihood approach with an approximate likelihood ratio test (aLRT) (online version PHYMLv.3.0) and the use of smart model selection, indicating the general time reversible nucleotide substitution model with 4 gamma categories as the optimal model for both data sets. A neighbor-joining tree, supported by 1000 bootstrap replicates, was constructed in Clustal (results not shown), showing a similar topology as the maximum likelihood tree. For the identification of clusters (≥3 sequences) and transmission pairs, the software Cluster Picker was used. The maximum genetic distance for both data sets was set at 0.08, combined with a clade support of aLRT >0.85, to call a group of sequences a cluster, similar to a previous analysis.24 The selection of these parameters was in agreement with the clinical information related to exposure risk (disclosed epidemiological information in physician provided clinical notes on the shared sexual partners/parties). All trees were visualized in Figtree v.1.4.3.
Statistical comparisons were performed using Fisher exact and χ2 tests for nominal variables, whereas for continuous variables, the Mann–Whitney U test for nonparametric statistics was used, with P values ≤0.05 considered statistically significant. Confidence intervals and interquartile ranges were indicated where appropriate. The commercial software Statistica (13 PL, Statasoft Polska, Warsaw, Poland) was used for all statistical calculations.
Clinical Group Characteristics and HCV Genotype Distribution
Initially, by the geno2pheno algorithm, 47 sequences were assigned as G1a and 65 as G1b. However, in 2 cases, the subtype needed to be corrected, according to the concordant results from the 2 well-known subtyping tools used, so as a final result, infection with G1a was found in 45 patients (40.18%), of whom 37 were infected with clade I (82.2%) and 8 with clade II (17.8%), whereas G1b infection was observed in 67 individuals (59.82%). All genotype-related differences in group characteristics are outlined in Table 1. Notably, G1a infection was more common among men (41/45, 91.11%) compared with G1b (51/67, 76.12%, P = 0.042), with a significantly higher frequency of G1a among MSM (27/45, 60.0%) compared with injection drug users (IDUs) and their respective partners (18/45, 40%, P = 0.003). In addition, G1a was enriched among AHC cases (21/45, 46.67% vs. 17/67, 25.37% for G1b) and individuals with a history of syphilis (18/45, 52.24% for G1a vs. 14/67, 24.45% for G1b, P = 0.009), whereas the opposite was true for patients diagnosed with chronic HCV (24/45, 32.43% for G1a vs. 50/67, 67.57% for G1b, P = 0.019). Acute HCV infection was exclusively observed among MSM and associated with a history of syphilis (27/35, 84.38% with a positive syphilis diagnosis vs. 8/35, 14.04% with a negative syphilis serology, P < 0.001).
Prevalence of Resistance-Associated Variants
Overall frequencies of HCV RAVs were calculated jointly for the paired NS3/NS5A sequences (n = 98, 40 G1a and 58 G1b) and separately for the available NS3 (G1a, n = 42 vs. G1b, n = 66) and NS5A (G1a, n = 43 G1b, n = 59) sequences. The overall NS3/NS5A RAVs frequency was 14.3% (14/98 strains), and this frequency was significantly higher among G1a-infected cases (n = 11/40, 27.0%) compared with G1b (n = 3/58, 5.2%, P = 0.005) (Fig. 1A).
In total, NS3 RAVs were found in 15.7% of the patients (17/108) while, for NS5A, only in 2.9%. The most common variant in the data set was located in protein NS3, more particularly Q80K, found in 11.1% (12/108) of the analyzed NS3 sequences. One patient infected with G1a had RAVs both in the NS3 and NS5A region. A notably higher frequency of NS3 RAVs among G1a-infected individuals (n = 13/42, 31.0%) compared with G1b (n = 4/66, 6.1% P = 0.0005) was demonstrated. The most predominant NS3 was Q80K, exclusive to G1a, observed in 12/42 (28.6%) G1a patients, and especially more prevalent among MSM (n = 9, 19.57%) compared with patients with a history of IDU and their partners (n = 3, 4.84%, P = 0.016) (Fig. 1B). Other observed NS3 RAVs were 54S, 55A, 117H, and 168E. Overall, NS3 RAV frequencies (n = 7/37, 18.92%), including Q80K polymorphism (n = 6/37, 16.22%), were similar among HIV cases with AHC compared with HCV chronically infected patients (n = 10/71, 14.08% and n = 6/71, 8.45%, respectively).
NS5A RAVs were only observed in 3 sequences (2.9%) across the entire data set, including 2 G1a (4.7%) and 1 G1b strain (1.7%), more particularly NS5A variants 28T/V and 31M (Fig. 1C). All 3 patients were identified to be chronically infected with HCV.
No further significant differences in the distribution of NS3 and/or NS5A RAVs were noted for sex, transmission route, history of syphilis, diagnosis of AHC, when analyzing both the paired NS3/NS5A sequences or separately for NS3 and NS5A, neither for G1a or G1b.
In total, 72 of 98 NS3/NS5A-paired sequences (73.47%) formed 10 clusters and 3 pairs of sequences. For G1a, 32 of 40 Polish strains (80.0%) formed 4 clusters and 2 pairs, including 3 clusters consisting exclusively of sequences derived from MSM and 1 cluster consisting of 10 sequences, which were all associated with IDU transmission (Fig. 2A). For G1b, 6 clusters and 1 pair including 40 sequences (69.0%) were counted, of which both the pair and 4 clusters consisted of sequences obtained from IDU-infected individuals and their heterosexual partners, whereas a large cluster of 14 sequences was found to be associated with MSM transmission (Fig. 2B).
In G1a, the occurrence of both pairing and clustering among sequences derived from MSM patients was notably more common compared with G1b (n = 20/40, 50.0% for G1a-infected MSM vs. n = 15/58, 25.86% for G1b MSM, P = 0.02). Six sequences harboring the NS3 Q80K polymorphism (50% of all sequences with Q80K in the study) clustered within a G1a-MSM cluster, all were derived from patients with documented AHC infection during HIV follow-up and were diagnosed in Warsaw and Wrocław (Table 2). Five transmission clusters included cases diagnosed at a variety of clinical centers with a mean intercity distance of 370 km. Patients characterized by NS5A RAVs were not shown to cluster closely together in the trees, nor for G1a or G1b.
In the current study, we analyzed HCV G1a and G1b NS3 and NS5A genetic sequences obtained from individuals coinfected with HIV, to investigate resistance patterns associated to DAA treatment failure. Transmission networks were constructed to reflect clustering patterns and transmissibility of natural RAVs in association with specific clinical characteristics. A high frequency of infections with G1a was observed among MSM with documented AHC after diagnosis of HIV and after being virologically suppressed with antiretroviral treatment. In this group of patients, sequences characterized by the NS3 Q80K polymorphism clustered together. However, NS5A RAVs were absent among patients with AHC and also rarely observed (<5%) in the overall study group, supporting to start treatment with NS5A inhibitors at an early stage. Phylogenetic inference revealed that the HIV/HCV coinfected population in Poland is facing 2 separate epidemics, dependent on the HCV genotype 1 subtype. Although the more isolated G1b epidemic is probably related to the spread of autochthonous Polish strains, the HCV subtype 1a epidemic in the MSM population may be fueled because of mixing with strains from other European epidemics, transmitted through sexual contact between MSM.
In our data set, RAVs in the NS3 region were found in 31% and 6.1% of all G1a and G1b sequences, respectively. The high frequency in G1a patients can be explained by the presence of NS3 polymorphism Q80K, which is known to be virtually absent in G1b.25 This finding was consistent with previously published cohorts, which reported prevalence rates ranging from <10% to ∼50%, depending on the geographical region.26 The Q80K polymorphism is specifically associated to G1a clade I and is only of key importance for patients treated with second-wave protease inhibitor simeprevir, due to significant reduction of response rates in combinations containing this agent and pegylated-interferon/ribavirin or sofosbuvir,27 compared with a small or even no impact for other PIs.28 In our study, other NS3 variants associated with protease inhibitor resistance were detected on amino acid positions 54, 55, 117, and 168, with variants on position 117 only affecting susceptibility to boceprevir or telaprevir, agents that are nowadays no longer used.5 Variant D168E reduces susceptibility to grazoprevir and paritaprevir and provides full resistance to simeprevir.29 A RAV on NS3 position 168, reported for <1% of the DAA-naive population in general, was previously reported to be common among G1b-infected patients experiencing virologic failure on treatment with faldaprevir or asunaprevir.30 In general, the presence of naturally occurring NS3 non-Q80K variants was infrequent in our study (<10%), which is in line with reports of other European and Asian cohorts.3,31
Among the analyzed NS5A sequences, RAVs were rarely detected, more particularly a prevalence of 4.7% for G1a and 1.7% for G1b. NS5A variants M28T/V, which are associated with antiviral resistance to daclatasvir, elbasvir, ledipasvir, and ombitasvir32,33 and L31M conferring resistance to daclatasvir and elbasvir,34,35 were observed in the data set. The frequency of these RAVs was in agreement with previous reports, where pre-existing M28T/V variants were found among 6% of G1a patients and in 7% of G1b-infected patients.3,28 However, these prevalence rates are still lower compared with the population of HIV-infected patients diagnosed with acute HCV infection (24%) or compared with a large European HCV database (28%).36,37 Variant L31M was previously reported to be even less prevalent, more particularly present in 1.2% of Japanese patients and 8% according to a study using public sequences.36,38 The NS5A variant Y93H, related to increased therapy failure among individuals treated with NS5A inhibitor containing regimens, including asunaprevir and daclatasvir, or grazoprevir and elbasvir was not detected, which is consistent with a Dutch report presenting a low prevalence for this variant in patients recently infected with HCV,37 however, in general higher for HCV1b-infected patients.39
Clustering of more than 3 sequences, supported by a high aLRT value and a short evolutionary distance, was commonly observed in the studied data set, with 5 clusters among IDUs and their respective heterosexual partners and 4 MSM-transmission clades identified. Only in one cluster, 2 cases with the history of IDU were linked to MSM risk behavior, indicating that, in general, HCV epidemics are restricted to particular transmission groups. Also, as G1a predominated among MSM, clustering was more common for this subtype. In addition, in the analyzed cohort of MSM, no IDU was reported; however, use of oral or intranasal chem-sex agents was common, in agreement with the reported broad use of recreational drugs in the MSM community, previously associated with HCV acquisition.40 It should be also noted that other sexual practices, such as the use of sexual toys, anal douches, or lubricants might additionally facilitate HCV transmission.40,41 Of note, similarly to another European cohort, association between acute hepatitis C and previous diagnosis of syphilis among HIV-infected MSM was observed in our study.42
Clustering among HCV sequences harboring the NS3 Q80K polymorphism was reported previously for various countries such as the United Kingdom and the Netherlands.10,41 It should be noted that, in our study, one of the transmission clusters (cluster 2) consisted of 8 G1a clade I sequences, all acquired from individuals with AHC and of which 6 sequences (75%) harbored the NS3 Q80K polymorphism. It may be hypothesized that, for the 2 remaining sequences, reversion to the wild-type variant occurred, similarly as observed by Newsum et al,10 where probable reversions of Q80K to the wild-type Q80Q were noted among DAA unexposed cases. However, such reversions were impossible to confirm due to the lack of longitudinal sampling in our study. It should be noted that, sexual transmission events of RAVs, including Q80K, have been rarely reported for individual cases.11,15 However, the founder effect that dominates the history of Q80K strains reflects the occurrence of many former transmission events, as all sequences characterized by polymorphism Q80K cluster together in a large clade.10,26 Here, we show that transmission clusters, including the one's harboring Q80K, were not limited to sequences from patients located in a single city but were in half of the cases collected in clinical centers distant from each other (eg, >350-km distance between Warsaw and Wrocław and even >500 km between Warsaw and Szczecin), indicating countrywide span of the identified transmission networks. The key limitation of this study was the nonrandom sampling of the sequences because they were selected based on sample availability, as well as on the patients' history of an acute or recently acquired HCV infection. Moreover, at the time of the study completion, no systematic sequencing of HCV was performed, resulting in a rather low number of viral sequences. It should be stated, although, that the increase in the number of acute HCV infections among MSM living with HIV is a recent phenomenon, and so far, there is no systematic surveillance for DAA resistance in this group. Also, the use of deep sequencing methodologies might have allowed the identification of variants present at lower thresholds and to obtain longer fragments characterized by higher phylogenetic signal. However, this technology was not available and, moreover, the clinical impact of minor variants on treatment response is still being debated,43 hence why our findings resulted from the use of Sanger population sequencing are still significant without using of next-generation sequencing.
Finally, it should be stressed that G1a, present in 40.2% of the samples and including 46.7% of the AHC cases, has only been identified infrequently in Poland. More particularly, G1a has been reported in only 2.5% of the HCV monoinfections and in 17.8% among the HIV/HCV G1–infected cases20,44 and was virtually absent in a former data set focusing on the global epidemiology of HCV genotype distribution, where only G1b infections were observed in the country. In Poland, G1a cases were previously observed among young individuals, however, not linked to IDU transmission, in contrast to what has been reported for other regions. Current analysis indicates enrichment of G1a among MSM with possible expanding epidemics of these strains. In Western Europe, G1a was the most prevalent subtype among MSM enrolled in HIV pre-exposure prophylaxis programs, both in HIV-positive and HIV-negative men.41
The identification of transmission clusters among HIV/HCV-infected patients, especially for the ones that recently acquired HCV through MSM contact, suggests that sexual transmission may not only fuel the HCV epidemic but also promote the spread of DAA resistance–associated variants. Such a phenomenon is of primary importance for long persisting or highly prevalent variants such as polymorphism Q80K in the NS3 region and several NS5A variants. Efforts for continued surveillance of variants affecting susceptibility to DAAs should be extended, especially for MSM, as transmission of RAVs known to confer resistance to the currently used drug regimens, may adversely affect HCV treatment options in the future.
1. Hepatitis C virus prevalence and level of intervention required to achieve the WHO targets for elimination in the European Union by 2030: a modelling study. Lancet Gastroenterol Hepatol. 2017;2:325–336.
2. Simmonds P, Becher P, Bukh J, et al. ICTV virus taxonomy profile: Flaviviridae. J Gen Virol. 2017;98:2–3.
3. Welzel T, Bhardwaj N, Hedskog C, et al. Global epidemiology of HCV subtypes and resistance-associated substitutions evaluated by sequencing-based subtype analyses. J Hepatol. 2017;23:30148–30144.
4. Blach S, Zeuzem S, Manns M, et al. Global prevalence and genotype distribution of hepatitis C virus infection in 2015: a modelling study. Lancet Gastroenterol Hepatol. 2017;2:161–176.
5. EASL recommendations on treatment of hepatitis C 2016. J Hepatol. 2017;66:153–194.
6. Nguyen LT, Gray E, Dean J, et al. Baseline prevalence and emergence of protease inhibitor resistance mutations following treatment in chronic HCV genotype-1-infected individuals. Antivir Ther. 2015;20:865–869.
7. Sarrazin C, Dvory-Sobol H, Svarovskaia ES, et al. Prevalence of resistance-associated substitutions in HCV NS5A, NS5B, or NS3 and outcomes of treatment with ledipasvir and sofosbuvir. Gastroenterology. 2016;151:501–512.
8. Krishnan P, Tripathi R, Schnell G, et al. O057: long-term follow-up of treatment-emergent resistance-associated variants in NS3, NS5A and NS5B with paritaprevir/r-, ombitasvir- and dasabuvir-based regimens. J Hepatol. 2015;62:S220.
9. Yoshimi S, Imamura M, Murakami E, et al. Long term persistence of NS5A inhibitor-resistant hepatitis C virus in patients who failed daclatasvir and asunaprevir therapy. J Med Virol. 2015;87:1913–1920.
10. Newsum AM, Ho CK, Lieveld FI, et al. The hepatitis C virus nonstructural protein 3 Q80K polymorphism is frequently detected and transmitted among HIV-infected MSM in The Netherlands. AIDS. 2017;31:105–112.
11. Franco S, Tural C, Nevot M, et al. Detection of a sexually transmitted hepatitis C virus protease inhibitor-resistance variant in a human immunodeficiency virus-infected homosexual man. Gastroenterology. 2014;147:599–601.
12. Hullegie SJ, van den Berk GE, Leyten EM, et al. Acute hepatitis C in The Netherlands: characteristics of the epidemic in 2014. Clin Microbiol Infect. 2016;22:17.
13. Boesecke C, Grint D, Soriano V, et al. Hepatitis C seroconversions in HIV infection across Europe: which regions and patient groups are affected? Liver Int. 2015;35:2384–2391.
14. Matthews GV, Pham ST, Hellard M, et al. Patterns and characteristics of hepatitis C transmission clusters among HIV-positive and HIV-negative individuals in the Australian trial in acute hepatitis C. Clin Infect Dis. 2011;52:803–811.
15. Bartlett SR, Jacka B, Bull RA, et al. HIV infection and hepatitis C virus genotype 1a are associated with phylogenetic clustering among people with recently acquired hepatitis C virus infection. Infect Genet Evol. 2016;37:252–258.
16. Centers for Disease Control and Prevention. Sexual transmission of hepatitis C virus among HIV-infected men who have sex with men–New York City, 2005-2010. MMWR Morb Mortal Wkly Rep. 2011;60:945–950.
17. Ingiliz P, Martin TC, Rodger A, et al. HCV reinfection incidence and spontaneous clearance rates in HIV-positive men who have sex with men in Western Europe. J Hepatol. 2017;66:282–287.
18. Walewska-Zielecka B, Religioni U, Juszczyk G, et al. Anti-hepatitis C virus seroprevalence in the working age population in Poland, 2004 to 2014. Euro Surveill. 2017;22:1560–7917.
19. Rosinska M, Janiec J, Niedzwiedzka-Stadnik M. Increase of new HIV diagnoses among men who have sex with men in Poland, 2000 to 2011. Euro Surveill. 2013;18:20642.
20. Leszczyszyn-Pynka M, Ciejak P, Maciejewska K, et al. Hepatitis C coinfection adversely affects the life expectancy of people living with HIV in northwestern Poland. Arch Med Sci. 2016;12.
21. Zabek P, Opoka-Kegler J, Baka M, et al. Prevalence of hepatitis C virus mutants resistant to protease inhibitors among polish HCV genotype 1-infected patients. Przegl Epidemiol. 2013;67:411–413.
22. Cielniak I, Siwak E, Firląg-Burkacka E, et al. Characteristics of new dynamic epidemic of acute hepatitis C among HIV infected individuals from warsaw outpatient clinic. 16th European AIDS Clinical Society Conference, Milan, 25–27 October 2017, abstract number PE16/11 2017.
23. Walker A, Bergmann M, Camdereli J, et al. A genotype independent, full-genome reverse-transcription protocol for HCV genotyping and resistance testing. J Clin Virol. 2017;91:42–48.
24. Jacka B, Applegate T, Krajden M, et al. Phylogenetic clustering of hepatitis C virus among people who inject drugs in Vancouver, Canada. Hepatology. 2014;60:1571–1580.
25. Cuypers L, Li G, Libin P, et al. In hepatitis C virus genotypes 1-6: significance for direct-acting antiviral treatment and drug resistance. Viruses. 2015;7:5018–5039.
26. Cuypers L, Vrancken B, Fabeni L, et al. Implications of hepatitis C virus subtype 1a migration patterns for virus genetic sequencing policies in Italy. BMC Evol Biol. 2017;17:017–0913.
27. D'Offizi G, Camma C, Taibi C, et al. Clinical and virological predictors of sustained response with an interferon-based simeprevir regimen for patients with chronic genotype 1 hepatitis C virus infection. New Microbiol. 2017;40:19–26.
28. Poveda E, Wyles DL, Mena A, et al. Update on hepatitis C virus resistance to direct-acting antiviral agents. Antivir Res. 2014;108:181–191.
29. Jensen SB, Serre SB, Humes DG, et al. Substitutions at NS3 residue 155, 156, or 168 of hepatitis C virus genotypes 2 to 6 Induce Complex patterns of protease inhibitor resistance. Antimicrob Agents Chemother. 2015;59:7426–7436.
30. Berger KL, Scherer J, Ranga M, et al. Baseline polymorphisms and emergence of drug resistance in the NS3/4A protease of hepatitis C virus genotype 1 following treatment with faldaprevir and pegylated interferon alpha 2a/ribavirin in phase 2 and phase 3 Studies. Antimicrob Agents Chemother. 2015;59:6017–6025.
31. Cao Y, Bao Y, Xia W, et al. Resistance-associated mutations to HCV protease inhibitors naturally pre-existed in HIV/HCV coinfected, treatment-naive patients. Clin Res Hepatol Gastroenterol. 2016;40:597–604.
32. Liu R, Curry S, McMonagle P, et al. Susceptibilities of genotype 1a, 1b, and 3 hepatitis C virus variants to the NS5A inhibitor elbasvir. Antimicrob Agents Chemother. 2015;59:6922–6929.
33. Krishnan P, Beyer J, Mistry N, et al. In vitro and in vivo antiviral activity and resistance profile of ombitasvir, an inhibitor of hepatitis C virus NS5A. Antimicrob Agents Chemother. 2015;59:979–987.
34. Sulkowski M, Hezode C, Gerstoft J, et al. Efficacy and safety of 8 weeks versus 12 weeks of treatment with grazoprevir (MK-5172) and elbasvir (MK-8742) with or without ribavirin in patients with hepatitis C virus genotype 1 mono-infection and HIV/hepatitis C virus co-infection (C-WORTHY): a randomised, open-label phase 2 trial. Lancet. 2015;385:1087–1097.
35. Kosaka K, Imamura M, Hayes CN, et al. Emergence of resistant variants detected by ultra-deep sequencing after asunaprevir and daclatasvir combination therapy in patients infected with hepatitis C virus genotype 1. J Viral Hepat. 2015;22:158–165.
36. Patino-Galindo JA, Salvatierra K, Gonzalez-Candelas F, et al. Comprehensive screening for naturally-occurring Hepatitis C virus resistance to direct-acting antivirals in the NS3, NS5A and NS5B genes in worldwide isolates from viral genotypes 1-6. Antimicrob Agents Chemother. 2016;8:02776–02715.
37. Christiansen MT, Hullegie SJ, Schutten M, et al. Use of whole genome sequencing in the Dutch acute HCV in HIV study: focus on transmitted antiviral resistance. Clin Microbiol Infect. 2017;23:28.
38. Iio E, Shimada N, Abe H, et al. Efficacy of daclatasvir/asunaprevir according to resistance-associated variants in chronic hepatitis C with genotype 1. J Gastroenterol. 2017;52:94–103.
39. Cuypers L, Li G, Neumann-Haefelin C, et al. Mapping the genomic diversity of HCV subtypes 1a and 1b: implications of structural and immunological constraints for vaccine and drug development. Virus Evol. 2016;2:1–11.
40. Ireland G, Higgins S, Goorney B, et al. Evaluation of hepatitis C testing in men who have sex with men, and associated risk behaviours, in Manchester, United Kingdom. Sex Transm Infect. 2017;93:404–409.
41. Hoornenborg E, Achterbergh RCA, Schim Van Der Loeff MF, et al. Men who have sex with men starting pre-exposure prophylaxis (PrEP) are at risk of HCV infection: evidence from the Amsterdam PrEP study. AIDS. 2017;31:1603–1610.
42. Sanchez C, Plaza Z, Vispo E, et al. Scaling up epidemics of acute hepatitis C and syphilis in HIV-infected men who have sex with men in Spain. Liver Int. 2013;33:1357–1362.
43. McPhee F, Hernandez D, Zhou N. Effect of minor populations of NS5A and NS5B resistance-associated variants on HCV genotype-3 response to daclatasvir plus sofosbuvir, with or without ribavirin. Antivir Ther. 2016:23.
44. Panasiuk A, Flisiak R, Mozer-Lisewska I, et al. Distribution of HCV genotypes in Poland. Przegl Epidemiol. 2013;67:11–16.