How do naturally occurring YMDD‐motif mutants influence the clinical course of lamivudine‐naïve patients with chronic hepatitis B virus infection?

See article in J. Gastroenterol. Hepatol. 2006; 21: 1783–1788. Chronic hepatitis B virus (HBV) infection is a critical public health problem worldwide. According to a World Health Organization report, HBV infection is a common disease that affects approximately 350 million persons, many of whom develop cirrhosis and hepatocellular carcinoma, accounting for 0.5–0.75 million deaths per year. Lamivudine therapy has been approved as a first-line treatment for chronic HBV infection, based on clinically proven efficacy in reducing viral load and disease activity.1–8 Lamivudine, the (–) enantiomer of 2′-deoxy-3′-thiacytidine (3TC), is a cytosine nucleoside analog that inhibits reverse transcription by inducing chain termination of nascent viral DNA in human immunodeficiency virus (HIV) and HBV in vivo.9 HBV is a DNA virus, but it replicates via reverse transcription using its own RNA-dependent DNA polymerase in a process similar to that used by retroviruses such as HIV. The HBV DNA polymerase has reverse transcription activity and a highly conservative tyrosine-methionine-aspartate-aspartate (YMDD) motif in the catalytic domain ‘C’, which is both the binding and the functional site of lamivudine.9 Prolonged administration of this potent antiviral agent can unfortunately be limited in many cases by the emergence of lamivudine-resistant HBV; the resistance mechanism is closely associated with mutations in the YMDD motif (YMDD mutants), with the most common substitutions being methionine at amino acid position 204 in the reverse transcriptase domain to either isoleucine (rtM204I, YIDD mutant) or valine (rtM204V, YVDD mutant).10–12 The emergence of YMDD mutants considerably reduces the viral susceptibility to lamivudine and this loss of efficacy generally increases over time from 14-47% at 1 year to 26–71% at 2 years, 49–57% at 3 years, and 67% by 4 years.3,6–8,13,14 More recent and sensitive detection techniques for YMDD mutants, however, have revealed the emergence of YMDD mutants in some patients soon after the commencement of lamivudine therapy, suggesting that these resistant strains may occur naturally. Evidence now increasingly indicates that YMDD mutants do occur naturally and can be found in HBV carriers who have never received lamivudine or other antiviral therapies.15–22 The reported incidence of YMDD mutants varies, ranging from 0-27.7% for asymptomatic carriers or patients with chronic hepatitis. These reports also suggest that lamivudine potently inhibits wild-type viruses, resulting in a selective increase of pre-existing YMDD mutants, which would have constituted only a minor portion of coexisting populations prior to lamivudine therapy. The selectively increased mutants might therefore dominate wild-type populations subsequent to treatment with lamivudine, causing an increased incidence of breakthrough hepatitis. According to a recent study,23 a five-fold or more increase in YMDD mutants over wild-type viruses in viral loads, but not the mere presence of the mutants, appears to significantly precede viral breakthrough. Of course, there is also the likely possibility that de novo induction of nucleotide substitution(s) by lamivudine activity will induce further evolution of wild-type viruses into YMDD mutants. It is therefore a significant concern as to whether lamivudine is non-effective or contraindicated in HBV patients with naturally occurring YMDD mutants. In an article published in this issue of the Journal, Akarsu et al.22 (from Turkey) reported on the presence and incidence of YMDD mutants in inactive HBV carriers positive for antibodies to hepatitis B e antigen (anti-HBe) who have never received antiviral therapy. All of the carriers had normal liver function and a low viral load (<105 copies/mL) persistently for at least 1 year (mean 3.5 years). Line-probe assays detected YMDD mutants in 13 of the 71 carriers (18.3%). Akarsu et al.22 therefore concluded that naturally occurring YMDD mutants exist at a high rate, even in a group of lamivudine-naïve and inactive anti-HBe-positive HBV carriers. Patients' HBeAg/anti-HBe status, viral load and HBV genotype could be considered as candidate virus-related factors affecting the incidence of naturally occurring YMDD mutants. Kobayashi et al.15 and Kirishima et al.16 (both from Japan) found these mutants only in anti-HBe-positive asymptomatic and chronic hepatitis carriers, respectively. When the analyses were limited to anti-HBe-positive patients, 33.3% of asymptomatic carriers (five of 15) and 22.2% of chronic hepatitis patients (four of 18), respectively, showed the YMDD mutants. In contrast, Matsuda et al.17,18 and Ohishi et al.19 (both from Japan) each identified only one chronic hepatitis patient with the YMDD mutants among their anti-HBe-positive groups (79 and 55 patients, respectively), even when highly sensitive detection techniques were used. Of these studies, Matsuda et al.17,18 also detected only one YMDD mutant patient among the HBeAg-positive group (42 patients). Heo et al.20 (from Korea) found one HBeAg-positive and two anti-HBe-positive patients with mutant viral strains among 40 patients(7.5%), while Huang et al.21 (from China) reported that these mutants were detected at a high rate even in HBeAg-positive patients: 16 of 59 HBeAg-positive patients (27.1%) and 12 of 45 anti-HBe-positive patients (26.7%, not significant). The cause of the marked variability among these studies remains unclear. Akarsu et al.22 detected naturally occurring YMDD mutants in none of seven inactive anti-HBe-positive carriers with 104−105 copies/mL, in two of 10 with 103−104 copies/mL, and in 11 of 54 with <103 copies/mL (not significant). In the report by Huang et al.,21 the mutants were identified in 12 of 48 patients with >105 copies/mL (25.0%) and in 16 of 56 patients with ≤105 copies/mL (28.6%, not significant). In other studies, the mutants were found in patients with low viral loads,15,16 while patients with high viral loads also had the mutants.18,20 No significant difference was therefore apparent between mutant copies and stratification of viral load. However, two paradoxical phenomena have emerged. One is that the pretreatment viral load is an independent, significant contributor to the emergence of YMDD mutants during lamivudine therapy,24–27 in contrast with the pre-existence of YMDD mutants in naïve patients with a low viral load. The other is that viral load is constantly low in naïve patients with pre-existing mutants, in contrast with high viral loads of YMDD mutants during viral breakthrough against lamivudine. These YMDD mutants may each have different properties in nature, such as replication competence and pathogenic potential. In fact, previous reports documented that lamivudine therapy was effective for patients with naturally pre-existing YMDD mutants.18,19,21 Of note, pre-existing YMDD mutant strains differed from post-treatment or post-breakthrough YMDD mutant strains,18 suggesting that the former does not prevail but rather is cleared away, and that the latter might replace the former, but would not be selected by lamivudine pressure. Abundant quasispecies are produced in the viral duplication process, because HBV with reverse transcriptase lacks proofreading activity, thus lowering the fidelity of DNA replication. Of the variety of pre-existing viral strains in nature, the YMDD mutant strains are likely to contain further anonymous mutations comprising amino acid substitutions outside the YMDD motif,18,19 thus conferring low or no replication competence and comprising only a small subset of HBV populations. It is doubtful if such mutants replicate robustly enough to become a predominant population against lamivudine. To determine if pre-existing YMDD mutants contribute to viral breakthrough, the mutants should first be identified, specifically in patients with high pretreatment viral loads and subsequent viral breakthrough. Next, it should be ascertained whether or not these mutants are identical strains; pre-existing YMDD mutants should ideally be evaluated for their replication-competence and lamivudine-resistance activity using in vivo or in vitro exogenous expression experiments. A number of studies have investigated if the response to lamivudine and the emergence rate of YMDD mutants during therapy differ among various HBV genotypes. However, only a few studies showed an association between naturally occurring YMDD mutants and HBV genotypes and these involved only a relatively small number of patients. For instance, Matsuda et al.18 (from Japan) found only two of 71 naïve patients with naturally occurring YMDD mutants and both were infected with genotype B (Ba and Bj, respectively). The study comprised 61 (86%) patients infected with genotype C, six (8%) with genotype B and four (6%) with genotype A. Considering the preponderance of genotype C in Japan, the incidence of mutants in genotype B seemed high. In contrast, Huang et al.21 (from China) compared the mutation rates among 104 naïve patients infected with genotype B, C or D and demonstrated the opposite results; YMDD mutants were frequently found with genotype C infections (6/15, 40%) and with C + B (6/16, 38%) or C + D (8/25, 32%) coinfections, but were absent or infrequent with genotype B (0/9, 0%), D (2/9, 22%) or coinfections of B + D (0/12, 0%). Akarsu et al.22 (from Turkey) reported the incidence of YMDD mutants in Turkish HBV carriers to be 18% (13/71) in patients who were considered to have genotype D. Because the geographic distribution of HBV genotypes (from A to H) is divergent worldwide and characteristic of each location, extensive multinational studies would be valuable in clarifying the association between the incidence of mutants and HBV genotypes. The variable results between studies may be attributed to host-related factors, including patient age, disease duration, observation period, disease stage and infection route. Differences could also arise from different study designs, patient populations and study scales. Patient age does not appear to be significant, with naturally occurring mutants detected in patients aged between 20 and 60 years,15–22 although anti-HBe-positive asymptomatic carriers with mutants tended to be relatively older.15 The disease stage at which mutants were detected also varied, ranging from an anti-HBe-positive asymptomatic carrier state to chronic inactive or active hepatitis, including severe acute exacerbation of the disease accompanied by hepatic failure, to cirrhosis.15–22 Because HBV reverse transcriptase has no proofreading activity for the duplicating process, it is possible that mutations could naturally and accidentally occur, even at the YMDD motif location, due to random nucleotide misincorporation. Naturally occurring mutants have been reported mostly from the East and Far East,15–21 which contain high- and intermediate-endemic areas for HBV infection. The present article22 relates to populations from the Middle East. Geographical areas appear to be closely related to the geographic distribution of HBV genotypes. Alternatively, there may be unidentified virus- or host-related backgrounds underlying the incidence of YMDD mutant appearance. According to Huang et al.,21 another two reports from China showed high rates of mutant occurrence, at 17% (19 of 110 chronic hepatitis patients) and 26% (32 of 122 chronic hepatitis patients), respectively. It would be interesting to gain similar statistics on other areas, such as Western and African countries. Finally, the reason for the different YMDD mutant rates among studies may reflect the different sensitivities and specificities of methods used for detecting YMDD mutants, particularly among viral quasispecies or predominantly major populations. A highly sensitive assay might identify the presence of minor viral species among complex quasispecies that would be otherwise missed by conventional methods, providing valuable information on the dynamics of spontaneous viral genome variability and selection of mutants by antiviral therapy. The rate of YMDD mutants might increase further if more sensitive assays were routinely used. The availability of standard samples should improve the sensitivity and specificity of various assays, by setting a point for comparison between detection methods. In conclusion, a percentage of HBV populations spontaneously develop YMDD-motif mutants in lamivudine-naïve HBV carriers, and the emergence of YMDD mutants might be an accidental phenomenon as viral quasispecies mature. The pre-existence of YMDD mutants by itself does not ensure clinical resistance to lamivudine or cause breakthrough hepatitis following lamivudine therapy. Perhaps, pre-existing and post-treatment or post-breakthrough YMDD mutant strains might each have different biological properties. Multinational, large-scale population studies using highly sensitive standard detection assays are required to clarify the significance of naturally occurring YMDD-motif mutants in HBV infection and progression.