2017 Volume 241 Issue 1 Pages 21-28
It is important to determine the genotypes or serotypes of hepatitis C virus (HCV) in patients before treatment with direct-acting antiviral agents (DAAs), because the effects of DAAs differ among genotypes. In Japan, two tests for HCV typing are available clinically, but only serotyping, not genotyping, is approved by the public health insurance. Although most serotype-1 Japanese patients are infected with genotype 1b HCV, it is known that a small proportion of patients show different results from two typing methods. This study focused on such patients and the effectiveness of treatment with daclatasvir plus asunaprevir (DCV/ASV) was evaluated. We analyzed 644 DCV/ASV-treated patients with serotype 1 or genotype 1b, and among them, 166 serotype-1 patients received a commercial-based direct sequencing (DS) test for resistant-associated variants of genotype 1b HCV. We found four patients (2.4%) with DS test failure, suggesting that the PCR primers targeting genotype 1b may not match. Importantly, none of the four patients achieved a sustained virological response. Our in-house DS test analyzing the 5′-untranslated region and coding regions for NS4 and NS5B of HCV showed that three of the four patients were infected with genotype 2 HCV, and one patient was infected with genotype 1a HCV. No recombinant virus of different genotypes was found. This study indicates that a subset of serotype-1 hepatitis C patients is infected with HCV of genotype 2 or 1a in Japan and that DCV/ASV is not effective for such patients. Thus, attention should be paid to DAA treatment without HCV genotyping.
Hepatitis C virus (HCV) infection is a world-wide health problem because it frequently leads to liver cirrhosis and hepatocellular carcinoma. Recently, direct-acting antiviral agents (DAAs) have been developed and HCV-infected patients who are not able to tolerate interferon (IFN)-based therapies have become to eradicate HCV effectively. In Japan, almost 70% of HCV patients are infected with genotype 1b HCV, while the rest are infected with genotypes 2a (20%) and 2b (10%) (Chung et al. 2010). Daclatasvir plus asunaprevir (DCV/ASV) combination therapy for 24 weeks is the firstly approved IFN-free DAA therapy in Japan for chronic hepatitis and compensated liver cirrhosis patients with serotype/genotype 1 HCV, and the overall sustained virological response (SVR) rate in IFN-ineligible/intolerant patients has been reported to be 87.4% (Kumada et al. 2014). Because the HCV serotype test is approved by the Japanese public health insurance but the HCV genotype test is not, only the serotype is checked in most patients with HCV infection in Japan. This was based on previous reports showing that the results of serotype and genotype agreed well (Tanaka et al. 1994; Mizoguchi et al. 1996) and the cost of the serotype test is lower.
The effects of DAAs, especially those of DCV/ASV, are affected by the presence of resistant-associated variants (RAVs). Because NS5A RAVs at Y93 and L31 have been known to reduce the SVR rate, it is not recommended to treat patients whose HCV has these RAVs, especially at Y93, with DCV/ASV. It was reported that the SVR rate in Japanese patients without NS5A RAVs was 95.4%, but that in patients with these RAVs was 39.7% (McPhee et al. 2015). Some commercial-based methods to detect NS5A RAVs have been developed, and a direct sequencing (DS) test, in which PCR primers target only genotype 1b HCV, is widely used in Japan (Kudoh et al. 2015). It is thought that patients whose DS tests were not successful could be infected with HCV other than genotype 1b in spite of serotype 1, but the details of such patients have been not documented. In this study, we focused on DCV/ASV-treated serotype 1 patients, whose DS test for NS5A RAVs failed, and analyzed the clinical and virological features of such patients.
In this study, a total of 644 patients were treated with DCV/ASV in Tohoku University Hospital and associated hospitals in the Study Group of Hepatitis Treatments for Clinicians from September 2014 to February 2016. This study was approved by the institutional ethics committees in Tohoku University Hospital and the associated hospitals. HCV serotype (Tanaka et al. 1994) and genotype (Okamoto et al. 1996; Ohno et al. 1997) were determined as described previously. The NS5A RAVs were tested with a commercial-based DS method targeting genotype 1b HCV (El-Shamy et al. 2008; Suzuki et al. 2012; Kudoh et al. 2015) or a commercial-based PCR-invader method (Tadokoro et al. 2014).
Extraction of RNA and RT-PCRTotal RNA was extracted from 50 μl of the serum sample, which was collected after obtaining informed consent, with a QIAamp Viral RNA Mini Kit (QIAGEN GmbH, Hilden, Germany). Reverse transcription (RT)-PCR was performed with PrimeScript II High Fidelity One Step RT-PCR Kit (TaKaRa Bio Inc., Shiga, Japan). To amplify 196 nucleotides (nt) excluding primers in 5′ untranslated region (UTR) (nt 92-287, the nt numbers were in accordance with a genotype 1b HCV isolate of 9448 nt [HC-J4, accession no. D13558]), primers KY80 (5′-GCA GAA AGC GTC TAG CCA TGG CGT-3′) and KY78 (5′-CTC GCA AGC ACC CTA TCA GGC AGT-3′) were used as described previously (Young et al. 1993). To amplify 346 nt in the NS5B region (nt 8276-8621), primers C016 (5′-GAY ACC CGM TGY TTT GAC TC-3′ [Y = T or C; M = A or C]) and C019 (5′-GCR GAR TAC CTR GTC ATA GC-3′ [R = A or G]) were used. To amplify 324 nt in the NS4 region of genotype 1 HCV (nt 5312-5635), primers C024 (5′-GCY GAC CTR GAG GTC GTC AC-3′) and C025 (5′-GTA TCC CRC TGA TGA ART TCC AC-3′) were used, and to amplify 346 nt in the NS4 region of genotype 2 HCV (nt 5301-5646, the nt numbers were in accordance with a genotype 2a HCV isolate of 9589 nt [HC-J6, accession no. D00944]), primers C026 (5′-RTA YAT CGC CAC VTG CAT GC-3′ [V = A, C, or G]) and C027 (5′-ATR CCR CTR ATR AAG TTC CAC-3′) were used.
Sequencing analysisThe amplified products were purified with QIAamp PCR Purification Kit (QIAGEN GmbH) and sequenced on both strands using the BigDye Terminator v3.1 Cycle Sequencing Kit on an ABI 3,500xL Genetic Analyzer (Applied Biosystems, Foster City, CA). The sequences were analyzed with Genetyx-Mac Ver.18 (Genetyx Corp., Tokyo, Japan). Phylogenetic trees were constructed using the neighbor-joining methods and bootstrap values were determined with 1,000 resamplings of the data sets.
Statistical analysesStatistical analyses were performed using chi-square test for the comparison of proportions between two groups and Mann-Whitney U test for comparison of continuous variables between two groups. Differences were considered to be statistically significant when P < 0.05.
Among 644 DCV/ASV-treated patients, 152 patients had genotype 1b HCV, and the remaining 492 serotype 1 patients had not been tested for the HCV genotype. The clinical characteristics of patients with genotype 1b and those with serotype 1 are shown in Table 1. The presence of liver cirrhosis was determined based on the result of abdominal ultrasound sonography, computed tomography, or FIB-4 index (Sterling et al. 2006). When FIB-4 index was higher than 5.83, the patients were considered as liver cirrhosis (Tawara et al. 2016). There were no significant differences between them. In total, 563 and 435 patients could be evaluated at 12 and 24 weeks after treatment, respectively, in October 2016. Although not significant, the SVR rate was higher in genotype 1b than serotype 1 at 12 weeks after treatment (91.1% vs. 86.9%) and at 24 weeks after treatment (89.1% vs. 82.5%).
There were 69 patients with treatment failure (non-SVR) containing breakthrough and relapse (Table 2). When patients with SVR24 and those with non-SVR were compared, the percentage of patients with NS5A Y93H mutations detected by DS was significantly higher in non-SVR patients (7.1% vs. 25.0%), as reported previously (Iio et al. 2016). The percentage of male was significantly higher in SVR24 patients (41.3% vs. 26.1%), and of note, the percentage of patients with DS test failure was significantly higher in non-SVR patients (0.9% vs. 16.7%).
Clinical characteristics of patients who were treated with DCV/ASV in this study.
DCV, daclatasvir; ASV, asunaprevir; ALT, alanine aminotransferase; PLT, platelet counts; LC, liver cirrhosis; IFN, interferon, PegIFN, pegylated interferon; RBV, ribavirin; PI, protease inhibitor; SVR12, sustained virological response at 12 weeks post-treatment; SVR 24, sustained virological response at 24 weeks post-treatment.
aPatients with test failure for NS5A mutations are excluded.
bPatients whose HCV had mixed type mutations are included in those with the mutation.
cPatients with more than 1% mutant HCV are included in those with the mutation.
Comparison of clinical characteristics of DCV/ASV-treated patients between SVR24 and treatment failure.
DCV, daclatasvir; ASV, asunaprevir; SVR 24, sustained virological response at 24 weeks post-treatment; ALT, alanine aminotransferase; PLT, platelet counts; LC, liver cirrhosis; IFN, interferon, PegIFN, pegylated interferon; RBV, ribavirin; PI, protease inhibitor; DS, direct sequencing.
aP values that are considered to be significant are shown in bold type.
Next, we focused on patients with DS test failure for NS5A RAVs. Among 492 serotype 1 patients who had not been tested for the HCV genotype, 166 and 195 patients were tested for NS5A RAVs with a commercial-based DS method and with a commercial-based PCR-invader method, respectively (Fig. 1). In the results of DS, HCV of 13 patients (7.8%) had mutations at L31 and/or Y93 including mixed type mutations, and of note, the DS test failed in four serotype 1 patients (2.4%). The clinical characteristics of the four patients are shown in Table 3. All four patients were female and the mean age was 71 years. The mean platelet counts were 14.7 × 104/μl and one patient had cirrhosis. Their time course of HCV RNA in the serum during DCV/ASV therapy is shown in Fig. 2. HCV RNA in Patient 3 had reduced to an undetectable level at week 4, but a breakthrough occurred at week 8 and the therapy was stopped at week 12. Three patients had not achieved HCV RNA negativity until week 8 and then the treatment was stopped. As a result, no serotype 1 patients whose NS5A RAVs tests had failed achieved SVR. There was a genotype 1b patient whose DS test failed (Fig. 1), but the patient achieved SVR successfully.
Flow-chart of study subjects.
The policies to use genotyping/serotyping for the determination of HCV type and those to use DS/PCR-invader for the determination of NS5A RAVs were different among institutes. The boxes with thick lines indicate the main targets of this study. DCV, daclatasvir; ASV, asunaprevir; RAVs, resistant-associated variants; DS, direct sequencing.
Clinical characteristics of the four patients with serotype 1 HCV whose DS tests to detect NS5A RAVs for genotype 1b HCV failed.
ALT, alanine aminotransferase; PLT, platelet counts; LC, liver cirrhosis; DCV, daclatasvir; ASV, asunaprevir; BT, breakthrough; NR, no response; UTR, untranslated region.
Time course of serum HCV RNA during the DCV/ASV therapy in 4 patients whose DS tests for NS5A RAVs were unsuccessful.
Patient 1 obtained HCV RNA negativity at 4 weeks of treatment but relapse occurred at 8 weeks. Patients 2-4 did not obtain HCV RNA negativity until 8 weeks of treatment and the therapy was stopped.
Using serum samples from the four non-SVR patients before the treatment (Patient 1) or after the treatment (Patient 2-4), the partial sequences in NS5B were determined. The phylogenetic trees showed that one patient had genotype 1a HCV, two patients had genotype 2a HCV, and one patient had genotype 2b HCV (Fig. 3). No patient appeared to have co-infection with different HCV strains. Because recombinant viruses of different genotypes were found in patients who showed a discrepancy between serotype and genotype (Hoshino et al. 2012), we also analyzed partial sequences in 5′ UTR. The genotypes of the four patients based on the 5′ UTR sequences were the same as those based on the NS5B sequences (Fig. 4A). Therefore, there was no recombinant virus in the three patients who had serotype 1 but genotype 2 HCV. The partial HCV sequence data in NS5B and 5′ UTR have been assigned to the DDBJ/GenBank/EMBL with the accession numbers LC150569-LC150572 and LC150573-LC150576, respectively.
Additionally, to investigate the reason why the three patients with genotype 2 HCV were judged as serotype 1, the partial sequences in NS4 were determined. In the enzyme-linked immunosorbent assay (ELISA) for serotyping, a group 1-specific peptide C14-1 and a group 2-specific peptide C14-2 (amino acids 1676-1760) were used (Tanaka et al. 1994), but there were no specific amino acid mutations or recombination in the region (Fig. 4B).
In the results of this study, all DCV/ASV-treated patients with HCV of genotypes other than 1b did not achieve SVR. Their HCV genotypes include 1a, 2a, and 2b, and DCV/ASV is not effective for HCV of these genotypes. There remains a clinical issue of how to treat these patients. After the unsuccessful DCV/ASV therapy, Patient 2 (genotype 2b) and Patient 3 (genotype 2a) were treated with sofosbuvir and ribavirin for 12 weeks and both have achieved SVR12. Patient 1 (genotype 2a) was treated with sofosbuvir/ledipasvir before the analysis of genotyping, and has achieved SVR12. Patient 4 with genotype 1a is waiting for next potent agents. Therefore, sofosubuvir-based therapies might be effective for such patients with serotype 1 but genotype 2 HCV.
Phylogenetic tree constructed based on HCV partial sequences in NS5B region.
This analysis contains 23 HCV partial sequences of 346 nucleotides including 4 sequences determined in this study. An isolate of genotype 6d (accession no. D84263) was used as an outgroup. The previously reported isolates, whose full-genome sequences and genotypes are known, are indicated as genotypes followed by isolate names and accession nos. The isolate names whose sequences were determined in this study are shown in bold type.
Comparison of HCV sequences determined in this study with previously reported isolates.
(A) Partial nucleotide sequences of HCV in 5′ UTR (nt 92-287, the nt numbers are in accordance with a genotype 1b HCV isolate of 9448 nt [HC-J4, accession no. D13558]). The HCV isolate names whose sequences were determined in this study are shown in bold type. The previously reported isolates, whose full-genome sequences and genotypes are known, are indicated as genotypes followed by isolate names and accession nos. (B) Partial amino acid sequences of HCV in NS4 (amino acids [aa] 1676-1760). C14-2 and C14-1 are sequences of group specific peptides of group 2 and 1, respectively, which are used in ELISA for serotyping.
HCV has a single-stranded and positive-sense RNA of approximately 9.5 kb in length. Based on the nucleotide sequence diversity, HCV is classified into six genotypes (Simmonds et al. 2005). It is important to determine the HCV genotype in patients with chronic HCV infection because it determines the effectiveness of therapy. For instance, it is well known that genotype 2 patients show higher SVR rates than genotype 1 with IFN-based therapies (Enomoto and Nishiguchi 2015). Because HCV genotyping has not been approved by the public insurance in Japan, serotyping is commonly used. The serotyping method detects antibodies against type-specific recombinant proteins in the NS4 region by ELISA (Tanaka et al. 1994). It distinguishes type 1 and 2 highly accurately and patients who show discrepancies between serotype and genotype have been very rare (Tanaka et al. 1994; Mizoguchi et al. 1996). However, in this study, at least 1.8% (3/166) of serotype 1 patients were infected with HCV of genotype 2 without specific amino acid mutations in the NS4 region. A previous report from Japan showed that, among 104 chronic hepatitis C patients, three patients showed discrepancies between serotype and genotype, and two of the three patients had recombinant virus of genotypes 2b/1b (Hoshino et al. 2012). Also, it was reported that among 252 serotype 1 patients in Japan, three patients were infected with HCV of genotype 2a or 2b, one patient had mixed infection of genotype 1b and 2a, and one patient had recombinant HCV of genotype 2b and 1b (Aikawa et al. 2016). In this study, there were three patients of serotype 1/genotype 2, but no patients had recombinant virus of different genotypes, because the genotypes based on the sequences in 5′ UTR, NS4 region, and NS5B region were consistent. Then why was the discrepancy found? Regretfully, serum samples before DCV/ASV therapy could not be obtained from Patient 2-4. Therefore, it is not completely denied that they had co-infection of HCV of genotype 1b and other genotypes as reported recently from Japan (Sohda et al. 2016), and that only genotype 1b HCV was cleared after DCV/ASV treatment. Also, there is another possibility that the patients were infected transiently with genotype 1 HCV and had acquired antibodies against the virus.
Because we could not obtain serum samples from a genotype 1b patient whose DS test failed, the reason why the test was unsuccessful was unknown. Her serum HCV RNA at the start of DCV/ASV was 6.9 log IU/ml and the sensitivity of PCR might not matter. It was considered that there were mutations or recombination in the NS5A region that does not match with the primers.
The suppressive effect of DCV/ASV on genotype 2 HCV has been tested only in vitro, and it was weaker than that on genotype 1b HCV. The EC50 of DCV against genotype 1b HCV replicon, Con1, was 9 pM, whereas that against genotype 2a, JFH-1, was 71 pM (Gao et al. 2010). Also, the EC50 of ASV against genotype 1b was 1.2 nM, whereas that against genotype 2a was 230 nM (McPhee et al. 2012). This weak suppression on genotype 2 was confirmed in this study, rather accidentally. Also, it is known that DCV is less effective on genotype 1a HCV; the EC50 against HCV replicon genotype 1a, H77, was 50 pM (Gao et al. 2010). Therefore, HCV genotyping is recommended before DAA therapies including DCV/ASV. Because the cost of HCV genotyping is fully paid by the hospitals in most cases including patients whose serotyping test failed, the approval of HCV genotyping by Japanese public insurance is required to be warranted.
In conclusion, we found HCV of genotype 2a, 2b, or 1a in DCV/ASV-unresponsive patients with serotype 1 hepatitis C, based on the failure of the DS test that detects NS5A RAVs of genotype 1b HCV. This study indicates that we have to pay attention to the administration of DAAs without HCV genotyping. Accurate genotyping is required for the selection of DAAs against HCV to further improve the SVR rate.
We are grateful to Drs. Osamu Abe (JR Sendai Hospital), Takehiro Akahane (Ishinomaki Red Cross Hospital), Osamu Kido (Iwate Prefectural Central Hospital), Osamu Kimura (South Miyagi Medical Center), Norihiro Kisara (JCHO Sendai South Hospital), Yutaka Mano (Sendai Medical Center), Yasunori Matsuda (Hachinohe Municipal City Hospital), Tatsuya Miura (Iwate Prefectural Chubu Hospital), Yutaka Miyazaki (Tohoku Kosai Hospital), Futoshi Nagasaki (Sendai City Hospital), Hiromasa Okamoto (Shirakawa Kosei General Hoapital), Hiroyoshi Onodera (Miyagi Cancer Center), Shuichi Sato (Kurihara Central Hospital), Toshihiro Sato (LC Clinic Sendai), and Takeshi Yamamoto (Tohoku Medical and Pharmaceutical University) for the collection of patient data.
The authors declare no conflict of interest.
