- Top of page
- Patients and methods
Despite more than three decades of research after the identification of hepatitis B virus (HBV), chronic infection with the HBV still remains an unresolved health problem with 400 000 000 infected people worldwide and limited efficacy of existing treatment modalities.1–3 Interferon (IFN)-α introduced in the treatment of chronic hepatitis B (CHB) in 1986,4 has been associated with considerable intolerance and a sustained biochemical and virological remission rate of 18–30% depending on treatment dose and duration.3, 5, 6 Oral nucleoside and nucleotide analogues introduced in the treatment of CHB during the last decade tend to replace IFN-α because of their improved safety and tolerability profiles.3, 7, 8 However, this is offset by the need for long duration of treatment with these oral agents, which is associated with progressively increasing rates of viral resistance.3, 7, 9 Moreover, in hepatitis B e antigen (HBeAg)-negative CHB (CHBe−), the optimal duration of therapy with oral anti-viral agents remains unknown, while the hitherto reported sustained response rates after treatment discontinuation are minimal, if any.3, 7
In many parts of the world, individuals not expressing the HBeAg comprise the majority of patients with chronic HBV infection. In Greece, such patients represent over 90% of CHB cases with ≥95% of them being of genotype D.10, 11 In HBeAg-positive CHB (CHBe+) cases, HBeAg seroconversion is usually associated with biochemical and virological quiescence and represents the primary end point of therapy.3, 7, 12 However, this end point is not applicable in CHBe− patients. In this subset of patients, the appropriate duration of effective long-term nucleos(t)ide analogue therapy remains unknown and is hitherto decided rather arbitrarily on the basis of combined assessment of other variables.3, 12
Despite the side-effects and intolerance of IFN-α, 12- or 24-month courses with standard recombinant IFN-α in CHBe− have been found to be associated with off-treatment sustained response rates of 20–30%.13, 14 A 36% combined response rate has recently been reported after a 12-month course of pegylated IFN (peg-IFN)-α2a, in CHBe− patients.15 Concurrent administration of lamivudine (LAM) and IFN-α in HBeAg-positive or -negative patients has given, at the best, equivocal results so far.16–25
In this study, we investigated the efficacy of a partially overlapping scheme of LAM and conventional IFN-α2b combination therapy of 18-month duration in a group of HBeAg-negative patients followed up for 12 months after stopping treatment. Results were compared with those of a similar group of CHBe− controls treated at our center with IFN-α2b alone for 12 months. In the present study, we report the final results of this clinical trial.
- Top of page
- Patients and methods
Of the 36 patients assigned to the LAM–IFN-α treatment protocol, 31 (86%) completed the treatment and the observation period. Of the remaining five patients, one (3%) declined treatment with IFN-α and remained on LAM monotherapy and four (11%) discontinued IFN-α during the phase of LAM and IFN-α combination therapy because of IFN-α intolerance and opted to continue on LAM. All 36 IFN-α-treated controls completed 12 months of therapy. Mean follow-up from the beginning of therapy was 2.4 ± 0.9 years in the study group and 5.7 ± 2.6 years in the IFN-α-treated controls (Table 1).
Table 1. Demographic and baseline biochemical, virological and histological characteristics of patients with HBeAg-negative chronic hepatitis B treated with 12 month partially overlapping courses of LAM and IFN-α and of historical controls treated with a 12-month course of IFN-α alone
|Patient characteristics||LAM + IFN-α treated (N = 36)||IFN-α treated (N = 36)||P-value|
|Sex, male (%)||25 (69.4)||30 (83.3)||0.165|
|Age (years)*||55 (46–66)||52 (45–55)||0.068|
|ALT (IU/L)*||80 (52–139)||114 (67–214)||0.090|
|Serum HBV-DNA (log10 cp/mL)†||6.2 ± 1.1||6.5 ± 1.1||0.410|
|Histological grading†||7.9 ± 2.5||7.9 ± 2.6||0.953|
|Histological staging†||3.4 ± 1.4||3.3 ± 1.6||0.796|
|Histological cirrhosis, n (%)||9 (25)||12 (33.3)||0.391|
|Prior use of IFN-α, n (%)||12 (33)||10 (27.8)||0.609|
|Prior use of LAM, n (%)||2‡ (5.6)||0 (0)||0.151|
|Follow-up (years)†||2.4 ± 0.9||5.7 ± 2.6||<0.001|
The demographic and histological features of the study and the control group are shown in Table 1. Patients in the two groups were not significantly different in terms of age, sex, serum ALT and HBV-DNA levels, as well as, the histological grade, stage or presence of cirrhosis. Two patients in the study group had received a LAM course of 12 and 6 months duration, 2 and 2.6 years before entering the current protocol, respectively, but none had developed LAM resistance. A comparable percentage of patients in the study and control groups (33% vs. 28%, respectively) had received one or more IFN-α courses in the past.
At the end of treatment (EOT), 28 patients receiving the LAM–IFN-α combination demonstrated normal ALT (78%), as contrasted to 19 (59%) of those on IFN-α monotherapy (P = 0.026; Table 2). At the 12-month post-treatment the number of patients with ALT normalization were 14 (39%) in the LAM–IFN-α group and eight (22%) in the IFN-α group, (P = 0.125). After the 12-month post-treatment, biochemical relapses were observed in two more patients of the LAM–IFN-α group and one of the IFN-α control group making the long-term remission rates of 34% and 19% respectively (P = 0.181).
Table 2. Biochemical and virological response rates of patients with HBeAg-negative chronic hepatitis B treated with 12 months partially overlapping courses of LAM and IFN-α and of historical controls treated with a 12-month course of IFN-α alone
|Type of response||End of treatment||12-month post-treatment|
|LAM + IFN-α treated (N = 36)||IFN-α treated (N = 36)||P-value||LAM + IFN-α treated (N = 36)||IFN-α treated (N = 36)||P-value|
|Biochemical response, n (%)|
| Normal ALT||28 (77.8)||19 (52.8)||0.026||14 (38.9)||8 (22.2)||0.125|
|Virological response, n (%)|
| Serum HBV-DNA ≤30 000 cp/mL||20 (55.6)||14 (38.9)||0.157||10 (27.8)||7 (19.4)||0.405|
| Serum HBV-DNA <400 cp/mL||12 (33.3)|| 3 (8.3)||0.013|| 3 (8.3)||0||0.075|
|Median HBV-DNA change (log10 cp/mL)||−3.1||−2.2||0.698||−1.4||−1.8||0.278|
|Combined response, n (%)|
| Normal ALT and HBV-DNA ≤30 000 cp/mL||18 (50.0)||11 (30.6)||0.093|| 8 (22.2)||5 (13.9)||0.358|
| Normal ALT and HBV-DNA <400 cp/mL||11 (30.6)|| 3 (8.3)||0.017|| 3 (8.3)||0||0.076|
At EOT, 20 patients (56%) of the combined LAM–IFN-α treatment had serum HBV-DNA levels ≤30 000 cp/mL and 12 (33%) had HBV-DNA below the detectability level of the PCR assay (<400 cp/mL), as contrasted to 14 (39%) and three (8%) of the IFN-α control group respectively (P = 0.157 and P = 0.013, respectively). At the 12-month post-treatment, the respective rates were 28% and 8% (10 and 3 patients) for the LAM–IFN-α group and 19% and 0% (7 and 0 patients) for the IFN-α control group (P = 0.405 and P = 0.075, respectively).
At EOT, combined biochemical and virological response, defined as normalization of ALT and serum HBV-DNA levels ≤30 000 cp/mL, was observed in 18 (50%) patients of the LAM–IFN-α group and in 11 (31%) of the control group (P = 0.093) and at the 12-month post-treatment eight (22%) and five (14%) patients respectively (P = 0.358). Moreover, a combined response defined as both normal ALT and undetectable serum HBV-DNA (<400 cp/mL) was observed in 11 (31%) patients of the LAM–IFN-α and three (8%) of the IFN-α control group at EOT (P = 0.017) and in three (8%) and none (0%), respectively, at the 12-month post-treatment (P = 0.076).
During the 12 months of LAM therapy, four patients (11%) developed virological breakthroughs indicating development of HBV-resistant mutants. None of them experienced biochemical or virological response during the subsequent IFN-α administration.
- Top of page
- Patients and methods
Although IFN-α is generally considered to be a first-line drug in the treatment of HBeAg-negative CHB achieving sustained off-therapy response rates of 20–30%,3, 28, 29 its use in clinical practice has been rather limited over the last decade due to the introduction of nucleos(t)ide analogues. In contrast to the subcutaneous injections of IFN-α, the nucleos(t)ide analogues are administered orally, have a much better tolerability and safety profile and are highly effective as long as treatment continues and HBV remains sensitive.3, 7 However, these oral agents should be administered for long, perhaps indefinite, periods resulting in increasing rates of HBV resistance, which exceed 60% after the third year of LAM therapy.3, 9, 30 The combination of IFN-α and LAM is theoretically attractive, because it might combine the potent anti-viral activity of LAM with the anti-viral and immunomodulatory activities of IFN-α for a possible synergistic anti-HBV effect. However, the efficacy of this combination in the treatment of CHB has been controversial and reported to be superior,17, 19, 22–24 equal16, 18, 20, 21, 25, 31 or inferior32, 33 to LAM or to IFN-α monotherapy. In particular in HBeAg-negative CHB, the combination of standard IFN-α and LAM has been compared with standard IFN-α alone in only study, in which a superiority of the combination regimen was suggested.19
With a few exceptions,18, 23, 25, 31, 32 the two drugs have been administered simultaneously. However, comparison of different LAM–IFN-α combinations in chronically infected woodchucks has demonstrated an enhanced synergistic action of a staggered treatment scheme starting with a 12-week course of LAM, followed by a 12-week LAM–IFN-α combination therapy, followed by a 12-week IFN-α monotherapy.34 The staggered scheme was superior to simultaneous administration of LAM and IFN-α in suppressing viraemia and intrahepatic woodchuck hepatitis virus (WHV) replication. A similar staggered scheme administered in 14 French HBeAg-positive patients resulted in HBeAg seroconversion of 45% and HBsAg seroconversion of 21% of the treated patients.23 Therefore, a similar staggered LAM–IFN-α combination regimen of longer duration was selected to be evaluated in our study including patients with HBeAg-negative CHB.
Treatment of HBeAg-negative CHB patients, has been especially difficult, among other things, because of lack for a precise definition of ‘response’.3, 35 By inclusion criteria, HBeAg is negative in all patients; the definition of virological response at EOT is hampered by methodological difficulties concerning sensitivity of serum HBV-DNA assays, uncertainties in the clinical impact of low-grade viraemia and by the inherent property of the HBV to maintain a latent infection in the form of cccDNA and resume high-replication rates following a quiescent post-treatment period.1, 2 Since most post-treatment biochemical relapses occur within the 12 months following the end of IFN-α therapy,13, 36 we decided to use 12 rather than 6 months of follow-up to judge the final response to therapy. Because the end point of treatment in HBeAg-negative CHB cannot practically be the virological clearance, but a long-lasting, possibly permanent, alteration of the virus–host balance in favour of the host, we considered a level of serum HBV-DNA of ≤30 000 cp/mL at the EOT and of follow-up, as indicating that a treated patient had entered an inactive carrier state.37, 38 Our results would not have changed even if we had defined virological response as serum HBV-DNA levels ≤20 000 cp/mL, a cut-off level used in a recent clinical trial of peg-IFN-α,15 because none of our patients with virological or combined response had HBV-DNA levels at the EOT or follow-up period between 20 000 and 30 000 cp/mL.
This study was not a randomized trial, but it included a homogenous patient population as well as appropriate historical controls for comparison of the effect of study treatment. Controls were randomly selected from a historical cohort of patients treated with a 12-month course of IFN-α in the past and having similar characteristics with this study group patient. Although the biochemical response rate at the EOT was significantly higher in the LAM–IFN-α-treated group compared with IFN-α-treated controls (78% vs. 53%, P = 0.026), the difference became non-significant at the end of follow-up (39% vs. 22%, P =0.125). Similar findings were observed for the undetectability of serum HBV-DNA (<400 cp/mL) at the end of therapy (33% vs. 8%, P = 0.013) and the end of follow-up (8% vs. 0%, P = 0.075) or the combined response of ALT normalization and undetectable serum HBV-DNA. On the contrary, 56% and 28% of patients in the LAM–IFN-α group had low-viraemia levels (serum HBV-DNA ≤30 000 cp/mL) at the end of therapy and follow-up period, while the respective values in the IFN-α control group were lower (22% and 19%), but not significantly different. Findings similar to the above were also observed for the combined response of ALT normalization and serum HBV-DNA levels ≤30 000 cp/mL. It should be noted, however, that the comparison of this study group with the historical controls had the power to reject or detect an at least 25–30% higher difference in response rates because of the relatively small number of patients (n = 36). Thus, a difference in response of <25% from this LAM–IFN-α combination regimen cannot be definitely excluded.
Treatment of patients with HBeAg-positive or -negative CHB with peg-IFN-α has been reported to be associated with higher response rates compared with LAM monotherapy.5, 15 In a recent multicentre study of peg-IFN-α2a therapy in HBeAg-negative CHB, a 12-month course of peg-IFN-α monotherapy achieved almost identical response rates with a 12-month course of simultaneous administration of peg-IFN-α and LAM.15 It is also interesting to note that, at the 12-month post-treatment, the LAM–peg-IFN-α combination arm reported to achieve biochemical response in 52% and virological response with HBV-DNA ≤20 000 cp/mL in 41% of the patients,39 compared with 39% and 28%, respectively, in the present study. However, these data cannot be compared directly not only because they come from different patient populations, but also because different types of analysis were used (intention-to-treat analysis in our study and response rates in <60% of the initial patient population in the study of peg-IFN-α2a39). The recent availability of peg-IFN-α will probably result in the replacement of standard IFN-α in the treatment of CHB mostly because of its easier once weekly administration.7 However, our results are still useful because we evaluated the efficacy of a partially overlapping LAM and IFN-α combination therapy, which is different from the simultaneous combination regimens used in the peg-IFN-α trials.5, 15
During LAM therapy, four of our patients developed a virological breakthrough indicating emergence of resistant HBV strains. In most studies, administration of LAM together with conventional18, 20, 21 or peg-IFN-α15 has not been associated with emergence of resistant HBV strains so far, but such an event has already been observed.31 A possible explanation is that resistant strains may have developed during the LAM monotherapy phase preceding the LAM–IFN-α combination treatment period.
In conclusion, we demonstrated that approximately 22% of patients remain in biochemical and virological remission 12 months after the EOT with a staggered combination scheme of LAM and IFN-α of 18-month duration. The sustained off-therapy response rates after this combination scheme was slightly but not significantly better than the rates achieved with a 12-month course of IFN-α monotherapy in historical controls. However, because a superiority of the LAM–IFN-α combination scheme cannot be definitely excluded from the results of our small study, its efficacy may deserve further evaluation within a large, randomized-clinical trial.