1. Background
Hepatitis B virus (HBV) is a prototype member of the family Hepadnaviridae. It consists of a partially double - stranded circular DNA genome of approximately 3200 bases with 4 overlapping open reading frames (ORFs) (1). The 4 ORFs are the core/precore, polymerase (pol), envelope (env), and X. The circular nature of the DNA and the arrangement of the ORFs cause the env gene to completely overlap with the pol gene, and consequently result in important changes in hepatitis B surface antigen (HBsAg) and a considerable reduction in HBsAg - specific antibodies (anti - HBs) binding in vitro (2).
The high magnitude of HBV replication leads to the considerable production of virions (> 1012) during each replicative cycle of the virus. As the reverse transcriptase (rt) encoded by the pol gene lacks a proofreading activity, HBV replication is associated with a high mutation rate of 10 -5 substitutions/base/cycle. This results in the generation of all possible single - base changes in the genome, including single and double mutations, which in turn are responsible to develop nucleos(t)ide analogs (NUCs) resistance in patients before and during NUCs therapy (3).
The NUCs treatment strategies should be implemented as early as possible following the detection of drug - resistant HBV variants, especially before the virological and clinical breakthrough (4). Mutations may occur primary and secondary forms. Primary drug resistance mutations tend to reduce the susceptibility to an antiviral agent. However, secondary compensatory mutations repair replication defects related to primary drug resistance (5).
Turkey is one of the countries with a prevalence of 2% to 8% of intermediate endemicity for HBV infecion (6, 7). In a systematic review in Turkey, the estimated overall population with HBV prevalence was 4.57%, whereas the estimated total number of patients with chronic hepatitis B (CHB) was 3.3 million. However, the prevalence varied in different regions of the country; the prevalence was 3.47% and 6.72% in the Western and Eastern regions, respectively (8, 9).
With this background, the current study aimed at investigating the molecular characterization of HBV in Turkish patients with CHB.
2. Methods
2.1. Patients
The current study was conducted from 2010 to 2015 on 762 patients chronically infected with HBV from 7 regions and 35 clinics in 25 cities across Turkey. Due to the unsuccessful DNA sequencing reactions, 110 and 124 patients in the treatment - naive and treatment - experienced categories were excluded from the study, respectively. The clinical and demographic characteristics of the patients are mentioned in Table 1. The study consisted of 528 patients, treatment - naive (n = 271, 51.3%) and those under NUCs treatment (n = 257, 48.7%). All the patients were classified as HBV chronic carriers according to the European Association for the Study of the Liver (EASL) clinical practice guidelines (10). The study was also approved by the clinical research ethics committee of Kocaeli University (Project no. KKAEK 2009/24; date: November 24, 2009; approval no. 5/16), and written informed consent was obtained from each patient before entering the study. Blood samples were obtained before starting the therapy and during the viral breakthrough of the treatment. The samples were centrifuged immediately, and the sera were separated, aliquoted, and then, stored at - 20°C until use. Serological markers of HBV were measured by enzyme - linked immunosorbent assay (ELISA) in all local clinic units.
2.2. HBV DNA Detection
HBV DNA was isolated from the serum samples obtained from the patients on the BioRobot workstation using the magnetic particle technology (QIAsymphony SP; Qiagen GmbH, Hilden, Germany). HBV DNA was detected by a commercial real - time polymerase chain reaction (PCR) assay (Artus HBV QS - RGQ test; Qiagen GmbH, Hilden, Germany) on the real - time platform (Rotor - Gene Q; Qiagen GmbH, Hilden, Germany).
2.3. HBV pol Gene Sequencing
Specific primer pairs were constructed (forward: 5’ - TCGTGGTGGACTTCTCTCAATT - 3’ and reverse: 5, - CGTTGACAGACTTTCCAATCAAT - 3’) for the amplification of the HBV pol gene region (11). The PCR conditions were as follows: denaturation at 95°C for 10 minutes followed by 35 cycles consisting of an annealing step at 95°C for 45 seconds, extension at 60°C for 45 seconds, and a final step at 72°C for 45 seconds. The final concentration of the primers was 0.3 mM. The size of the derived amplicon in HBV was approximately 742 bp and included all the known NUCs resistance mutations in HBV. Phire Hot Start DNA polymerase (Finnzymes Oy, Vantaa, Finland) was utilized in the sequencing protocol. All PCR products were purified using the High Pure PCR product purification kit (Roche Diagnostics, Mannheim, Germany). Sequencing was performed using an ABI PRISM 3130 genetic analyzer (applied biosystems Inc., Foster City, California, United States). The BigDye Terminator version 3.1 Cycle Sequencing Kit (Amersham Pharmacia Biotech Inc., Piscataway, New Jersey, United States), 36 cm capillary array, and POP - 7TM polymer (applied biosystems Inc.) were used for sequencing. For cycle sequencing, the following thermal protocol was used: 35 cycles consisting of 95°C for 20 seconds, 50°C for 25 seconds, and finally 60°C for 2 minutes. The reverse and sequencing primers were used at a final concentration of 0.5 mM.
2.4. HBV pol/surface Gene Mutation Determination
The sequencing data were analyzed using the Genafor/Arevir - geno2pheno drug resistance tool (center of advanced European studies and research; Bonn, Germany [http://coreceptor. bioinf.mpi - inf.mpg.de]). The geno2pheno tool for HBV is a database specifically designed for the rapid computer - assisted virtual phenotyping of Hepatitis B and utilizes genome (nucleic acid) sequences as input. The program searches for homology between the input sequence and other DNA sequences already stored in its database, including relevant clinical data for drug resistance and surface gene mutations (12). The tool searches for HBV drug resistance mutations in the rt domain of the polymerase at amino acid positions 80 to 250 (13). Drug resistance mutations to the NUCs were categorized into primary, partial, and compensatory resistance groups (14).
The overlapping S - gene segment was obtained using the geno2pheno tool for amino acid substitutions at positions from 100 to 196 (15). The amino acid substitutions in HBsAg were categorized into antiviral drug - associated potential vaccine - escape mutants (ADAPVEMs) and typical HBsAg amino acid substitutions. The latter included hepatitis B hyperimmunoglobulin (HBIg) - selected escape mutation, vaccine escape mutation, hepatitis B misdiagnosis, and immune - selected amino acid substitutions (5, 14, 16-19).
Some mutations, especially ADAPVEMs, were not located in the “a” determinant of the HBsAg protein. Further, the important neutralizing domains of the HBsAg protein including the region outside the “a” determinant of the HBsAg protein for ADAPVEMs were analyzed.
2.5. Statistical Analysis
Statistical analysis was performed using IBM SPSS Statistics software, version 21.0 for Windows (SPSS; IBM corporation, New York, United States). The chi - square and t tests were utilized in the statistical analysis. A P value < 0.05 was considered significant.
3. Results
The current study consisted of a total of 528 patients, of which 325 (61.6%) were male with a mean age of 38 years (range: 18 - 69). Based on their HBeAg status, 351 (66.3%) patients were HBeAg - negative. Seven patients were anti - HIV antibody positive, 4 were anti - HDV IgG positive, and 1 patient was anti - HCV positive (Table 1).
Table 1.
Clinical and Demographic Characteristics of the Patients
| Variables | Value |
|---|---|
| Patients, no | 528 |
| Gender, M/F, (%) | 325/203 (61.6/38.4) |
| HBeAg positivity, n (%) | 178 (33.7) |
| ALT, median (range) U/L | 37 (9 - 2584) |
| HBV DNA load, median (range) IU/mL | 1.3 E+4 (6 E+2 - 9.8 E+6) |
| Participation from Regionsa, n (%) | |
| Marmara (Balikesir, Bursa, Edirne, Istanbul, Kocaeli) | 218 (41.3) |
| Central Anatolia (Ankara, Konya, Kayseri) | 104 (19.7) |
| South - eastern Anatolia (Adiyaman, Batman, Diyarbakir, Antep, Urfa) | 73 (13.8) |
| Black Sea (Bolu, Giresun, Sakarya, Samsun, Tokat) | 55 (10.4) |
| Aegean (Afyon, Denizli, Izmir) | 48 (9.1) |
| Mediterranean (Antalya, Maras, Mersin, Osmaniye) | 30 (5.7) |
| Clinical status, n (%) | |
| HBeAg negative CHB phase | 333 (63.1) |
| Immune reactive phase | 162 (30.7) |
| Inactive HBV carrier state phase | 18 (3.4) |
| Immune tolerant phase | 15 (2.8) |
| HBV genotype, n (%) | D: 526 (99.6) ; H: 2 (0.4) |
| HBV subgenotype of genotype D, n (%) | D1: 461 (87.7 ); D2: 29 (5.5); D3: 30 (5.7); D4: 6 (1.1) |
| Co - infection status, n (%) | |
| Anti - HCV positive | 1 (0.2) |
| Anti - HDV IgG positive | 4 (0.8) |
| Anti - HIV positive | 7 (1.3) |
| Treatment status, n (%) | |
| Treatment - naive | 271 (51.3) |
| Treatment - experienced | 257 (48.7) |
| Drug choice statusb, n(%) | |
| Lamivudine | 86 (33.5) |
| Combination | 71 (27.6) |
| Tenofovir | 36 (14.0) |
| Entecavir | 30 (11.7) |
| Interferon | 14 (5.4) |
| Telbivudin | 10 (3.9) |
| Adefovir | 10 (3.9) |
Abrrevations: CHB, chronic hepatitis B; F, female; M, male.
aPatients included 35 different clinics from 25 cities in Turkey.
bTherapy situation during rebound.
The number and percentage of male patients in the treatment - naive and treatment - experienced groups were 170 (62.7%) and 155 (60.3%) (P = 0.56), respectively. The HBeAg positivity in these groups was detected in 65 patients (24.0%) and 113 patients (44.0%) (P = 0.01), respectively.
The patients in the treatment - naive and treatment - experienced groups reported a mean ± standard deviation (SD) age of 39.32 ± 11.47 and 38.38 ± 12.72 years, mean HBV viral load of 8.5 + E7 (± 6.6 + E8) and 6.2 + E7 (± 6.5 + E9) IU/mL, mean alanine aminotransferase (ALT) levels of 69.99 ± 199.84 and 73.68 ± 107.92 U/L, and mean aspartate transaminase (AST) levels of 51.01 ± 110.44 and 50.18 ± 60.12 U/L, respectively. The differences in age, HBV viral load, ALT, and AST levels in the patients in the treatment - naive and treatment - experienced groups were not significant according to t test results (P = 0.43, 0.71, 0.81, and 0.92, respectively).
Among 50 (18.5%) and 112 (43.6%) patients in the treatment - naive and treatment - experienced groups, respectively, there were 191 (70.5%) and 142 (55.3%) patients in HBeAg - negative CHB phase and immune - reactive phase, respectively (P = 0.01).
The patients were administered lamivudine (LAM, 33.5%), combination therapy (27.6%), tenofovir (TDF, 14.0%), entecavir (ETV, 11.7%), PEGylated interferon (5.4%), telbivudine (Ldt) (3.9%), and adefovir (ADV, 3.9%) at the time of viral rebound in the treatment - experienced group.
Genotype D was identified in 526 patients (99.6%), and genotype H only in 2 naive patients (0.4%) (Table 1). However, subgenotypes D1, D2, D3, and D4 were determined in 461 (87.3%), 29 (5.5%), 30 (5.7%), and 6 (1.1%) patients, respectively.
In total, 174 (32.9%) Turkish patients with mutations in the HBV pol gene were primary, partial, or compensatory resistant to NUCs. Among them, 79 (45.4%) patients belonged to the treatment - naive group, whereas 95 (54.6%) belonged to the treatment - experienced group. The prevalence of the mutation in the gene encoding HBV polymerase in the treatment - experienced patients was statistically different from those of the patients in the treatment - naive group (P = 0.05). The most common primary resistance mutation both in the treatment - naive and treatment - experienced patients was rtM204I/V ± rtV173L ± rtL180M. The frequencies and patterns of mutations in the HBV pol gene are displayed in Table 2.
Table 2.
Characteristics of Genotypic Resistance Mutations to the Nucleos(t)ide Analogues
| Mutation Characteristic, No. (%) | Mutation Pattern | Nucleos(t)ide Analogue | Patient No. (%) | P Value | |
|---|---|---|---|---|---|
| Naive | Experienced | ||||
| Primary Resistance Mutation n = 84 (15.9) | rtA181G/S/T/V | LAM, LdT, L - FMAU, ADV, TDFa | 1 (0.18) | 8 (1.5) | 0.01 |
| rtT184A/I/L | ETVa | 0 | 9 (1.7) | ||
| rtM204I/V ± rtV173L ± rtL180M | LAM, LdT, L - FMAU, FTC | 9 (1.8) | 50 (9.4) | ||
| rtM204V + rtT184S | LAM, LdT, ETV | 0 | 1 (0.18) | ||
| rtI233V | ADV | 3 (0.5) | 1 (0.18) | ||
| rtN236T | ADV, TDF | 1 (0.18) | 1 (0.18) | ||
| Total | 14 (2.6) | 70 (13.2) | |||
| Partial Resistance Mutation n = 20 (3.7) | rtT184L | ETV | 0 | 1 (0.18) | 0.06 |
| rtA194S/T/X | TDF | 4 (0.7) | 6 (1.1) | ||
| rtS202G | ETV | 1 (0.18) | 6 (1.1) | ||
| rtM250V/R | ETV | 1 (0.18) | 1 (0.18) | ||
| Total | 6 (1.1) | 14 (2.6) | |||
| Compensatory Mutation n = 143 (27.0) | rtL91I | LdT | 23 (4.3) | 23 (4.3) | 0.09 |
| rtQ149K | ADV | 19 (3.5) | 15 (2.8) | ||
| rtV191I | TDF | 0 | 1 (0.18) | ||
| rtV214A | LAM, L - FMAU, FTC, TDF | 2 (0.3) | 3 (0.5) | ||
| rtQ215H/P/S | LAM, L - FMAU, FTC, TDF | 37 (7) | 18 (3.4) | ||
| rtN238D | ADV | 1 (0.18) | 1 (0.18) | ||
| Total | 82 (15.5) | 61 (11.5) | |||
aPropable resistance
The total prevalence of typical amino acid substitutions in HBsAg was 34.8% (n = 184). According to the treatment status, the prevalence was 33.5% (n = 91) and 36.1% (n = 93) in the treatment - naive and treatment - experienced patients, respectively. The HBIg escape mutation, vaccine escape mutation, misdiagnosis, and immune - selected escape mutations were 8.3%, 4.1%, 3.8%, and 18.6%, respectively. There was no statistically significant difference in the prevalence of HBsAg escape amino acid substitutions in the treatment - naive and treatment - experienced patients (P = 0.57). Typical patterns of amino acid substitutions in the HBsAg escape mutants are depicted in Table 3.
Table 3.
Typical HBsAG Escape Amino Acid Substitutions of the Study Patients
| HBsAg Amino Acid Substitution Category | Mutation Pattern | Patient No. (%) | Combined Pattern | Number of Patients | P Value | |
|---|---|---|---|---|---|---|
| Naive | Experienced | |||||
| HBIg escape | sT118A, sP120T, sT123A,sQ129R, sM133L, sY134N,sD144E, sG145K | 23 (8.4) | 21 (8.2) | sT123A + sG145K | 2a | 0.89 |
| Vaccine escape | sP120S, sM133L, sS143L,sD144E, sG145R, sS193L | 10 (3.7) | 12 (4.6) | 0.57 | ||
| Hepatitis B misdiagnosis | sP120S/T, sR122K, sT131I, sM133T, sS143L | 11 (4) | 9 (3.5) | 0.73 | ||
| Immune - selected amino acid substitution | sQ101H/R, sI110L, sG119I/R, sP120T,T123A/N, sP127T, sG130K/R, sT131N, sT140I, sS143T, sD144E, sG145R | 47 (17.3) | 51 (19.8) | sS143T + sD144E + sG145R | 1a | 0.46 |
| sI110L + sP120T | 2a | |||||
| sQ101H + sI110L | 1a | |||||
| sQ101H + sP127T | 1a | |||||
| sQ101H + sI110L + P120T | 1a | |||||
| Total | 91(33.5) | 93 (36.1) | ||||
aNumber of combined pattern patients included to the mutation pattern.
Six different ADAPVEM motifs were located both in the treatment - naive and treatment - experienced patients. They were sF161L/rtI169X, sE164D/rtV173L, sL172L/rtA181T, sL173F/rtA181V, sS195M/rtM204V, and sS196L/rtM204I. The total prevalence of ADAPVEMs was 2.9% (n = 8) and 7.7% (n = 20) in the treatment - naive and treatment - experienced groups, respectively. The prevalence of ADAPVEMs in the treatment - experienced patients was statistically different from that of the treatment - naive patients (P = 0.03). The ADAPVEM patterns were related to LAM, LdT, and ADV drugs. The ADAPVEMs motifs in the treatment - naive and treatment - experienced patients are displayed in Table 4.
Table 4.
ADAPVEM according to nucleos(t)ide analogues in treatment naive and experienced patients
| Mutation Characteristic | Mutation Pattern | Nucleos(t)ide Analogue | Patient, N (%) | P Value | |
|---|---|---|---|---|---|
| Treatment - Naive | Treatment - Experienced | ||||
| ADAPVEM N = 28a (5.3) | sF161L/rtI169X | ETV | 1 | - | 0.03 |
| sE164D/rtV173L + sS195M/rtM204V | LAM, LdT | - | 3 | ||
| rtA181T/sL172L | ADV | - | 1 | ||
| rtA181V/sL173F | ADV | - | 3 | ||
| sS195M/rtM204V | LAM, LdT | 4 | 7 | ||
| sS196L/rtM204I | LAM, LdT | 3 | 6 | ||
| Total | 8 (2.9) | 20 (7.7) | |||
Abbreviation: ADAPVEM, antiviral drug - associated potential vaccine - escape mutant; ADV, adefovir; LAM, lamivudine; LdT, telbivudine.
aSome of the patients had multiple mutations, however the percentage of mutation was calculated for 28 patients.
4. Discussion
The current study focused on analyzing the prevalence of mutations in the Turkish people chronically infected with HBV. Six different primary resistance mutation patterns were detected in 84 (48.2%) patients. Fourteen (8%) of them belonged to the treatment - naive, whereas 70 (40.2%) of them belonged to the treatment - experienced group. However, the most common primary resistance mutation was reported rtM204I/V ± rtV173L ± rtL180M. The prevalence of primary resistance mutations in the treatment - experienced patients was statistically different from that of the treatment - naive patients (P = 0.01), indicating that primary resistance mutations occurred mainly in patients with the use of NUCs during treatment. The viral rebounds during antiviral treatment should be analyzed in terms of drug resistance. Antiviral resistance can be detected prior to treatment; therefore, screening to detect resistance prior to the commencement of treatment may be regarded as a rational approach. The primary resistance mutations associated with amino acid positions 181, 204, 233, and 236 were detected in the treatment - naive patients. Sayan et al., reported 2 resistance mutations (rtI233V and rtN236T) associated with acyclic phosphonates (ADV); however, the group could not detect YIMM or YMDD mutations in patients with naive - CHB in Turkey (11). The current study findings suggested a possible accumulation of primary resistance mutations for NUCs, which may be attributed to their widespread use in the last 5 years in Turkey.
The available literature reports an overall prevalence of primary resistance mutations among treatment - naive patients to range from 1% to 30%. Such variability is likely to occur due to the differences in the methods applied to determine the mutation in the HBV pol gene, overall study design, and study population (20-29). However, Sayan et al., mentioned that the direct sequencing approach could limit the detection of primary resistance mutations (11). In the current study, the major primary resistance mutation, namely rtM204I/V, was usually found in combination with other mutation patterns both in the treatment - naive and treatment - experienced patients (30). The combination status regarding the rtM204V mutation may serve as a useful tool when selecting the study methodology.
The current study detected and characterized 4 different partial resistance mutations. They were mainly associated with ETV (rtT184L, rtS202G, and rtM250R/V). However, in half of the patients, the mutations in ADV gene (rtA194S/T/X) were detected. Patients carrying these mutations and undergoing a long - term therapy may require frequent monitoring for primary drug resistance against ETV and TDF (31, 32). The national insurance policy covered only LAM, Ldt, and interferons for first - line treatment (patients with HBV DNA < 2 E + 6 IU/mL). Until July 2015, the choice of treatment for CHB was limited; therefore, the current study did not discuss the type and duration of the therapy.
The most common compensatory mutations in the current study were rtQ215H/P/S, rtL91I, and rtQ149K in either single or combined profiles both in treatment - naive and treatment - experienced patients. The rtQ215H substitution was detected in patients receiving LAM or ADV therapy; however, its virological and clinical importance remained unclear (33). The primary function of compensatory mutations is to repair replication defects in viral polymerase activity related to the generation of primary drug resistance (11, 30, 34). Therefore, compensatory mutations can be detected during viral rebound, but without primary resistance during the NUCs therapy in patients with chronic hepatitis B. These, in turn, may assist to discriminate from primary drug resistance. Compensatory mutations without any primary or partial resistance mutations were detected in 27 patients in the treatment - naive and in 59 patients in the treatment - experienced groups. The substitution mutation, rtQ215, occurred even without exogenous selection pressures (35).
In the current study, six different types of ADAPVEMs were determined in Turkish patients with CHB predominantly associated with L - nucleosides (LAM and LdT) and ADV. However, ADAPVEMs were mainly observed in patients undergoing NUCs therapy (Table 4). The data gathered in the current study coincided with the findings of other studies (11, 36, 37), particularly, the frequency of rtM204I/V + sI195 M/sW196S/L mutations in LAM - resistant cases demonstrated a predominant presence (34). Thus, the ADAPVEMs can adversely affect the local or global immunization programs to control HBV with a potential to spread to individuals vaccinated against HBV infection (18, 19, 38, 39).
In the HBV genome, the HBV polymerase gene and envelope gene completely overlap (40); hence, the mutations in the pol ORF can cause alterations in the overlapping HBsAg. Typical HBsAg escape mutations were detected in the 4 categories of the patients. The immune - selected amino acid substitution category was the most common HBsAg amino acid substitution. However, some categories, such as HBIg escape and immune - selected mutations demonstrated certain patterns. Some typical HBsAg escape mutations, commonly detected in the patients with CHB, are sP120T, sM133I, sS143L, SD144A/E, sG145R, and sE164D (13, 14, 41). Typical HBsAg escape mutations may result as a consequence of a failure to control infection with vaccination or HBIg and misdiagnosis in the HBsAg testing stage (2). Patients with CHB undergoing NUCs treatments should be surveyed for typical HBsAg escape mutations for the benefit of public health (30, 41).
The determination of viral genotypes assists to analyze the progression of the disease, thereby aid to develop a suitable anti - viral therapy. Genotype D is widespread in Turkey and other Mediterranean countries (41, 42). D1 was more frequent with the presence of D1 - D4 subgenotypes in Turkey (43). The determination of genotypes and subgenotypes of HBV may contribute to accurate data generation associated with their circulation and transmissibility.
The coinfection data were very limited in Turkey; the coinfection of HBV/HIV, HBV/HDV, and HBV/HDV was 1.3%, 0.8%, and 0.2%, respectively. In a single large - scale study, Sayan et al., analyzed 1306 HIV - positive patients and coinfection of HIV/HBV was 2.7%. Sexual contact was reported as the acquisition route in 98.6% of the patients, whereas the use of injection drug was only 0.3% (44). Amiri et al., reported the coinfection of HBV/HIV as 1.8% and concluded that the use of injection drug severely affected the degree of coinfection (45). Both of the references 44 and 45 are regional (from Turkey and Iran).
In conclusion, the findings on drug resistance mutations in the treatment group require rational approaches such as prevention of unnecessary drug modifications due to compensatory mutations. The increasing incidence of drug resistance mutations warrants an analysis of drug resistance as an integral part to manage patients with CHB using NUCs. In addition, the surveillance of drug resistance mutations when treating HBV should be regarded as an area of supreme importance both for regional and global control of CHB.
