Two patterns of alanine aminotransferase increase to predict long-term viral response in chronic hepatitis B patients: virus- or host-induced?Hong You1,2,*, Xiaoning Wu1,2, Xiaojuan Ou1,2, Hong Ma1,2, Qianyi Wang1,2, Tianhui Liu1,2, Min Cong1,2, Ping Wang1,2, Baoen Wang1,2, Jidong Jia1,2
Date accepted: 16 August 2010
Date published online: 18 March 2011
Copyright (c) 2011 International Medical Press, all rights reserved.
Background: Serum alanine aminotransferase (ALT) increase is a well-known phenomenon during interferon treatment for chronic hepatitis B. However, little is known about these increases during nucleoside/nucleotide treatment and the effects on long-term clinical outcomes.
Methods: A total of 170 treatment-naive hepatitis B e antigen (HBeAg)-positive chronic hepatitis B patients were treated with a nucleoside/nucleotide analogue for at least 2 years and followed up for 1 more year post-treatment. Clinical characteristics were detected and analysed at baseline and at every 3-month interval.
Results: Two patterns of ALT increase, virus- and host-induced, were detected. Virus-induced increases were characterized by a rapid increase in serum ALT and HBV DNA typically after 2 years of treatment, and were more common than host-induced ALT increases (15.9% versus 6.5%; P<0.05) with a median ALT increase of 5.7-fold the upper limit of normal (ULN). Host-induced ALT increases were characterized by moderately increased ALT (median 2.5-fold ULN) with a slow decrease in HBV DNA that occurred mainly in the first year of treatment (63.6%). Most importantly, host-induced increases were associated with favourable long-term treatment outcomes in HBV DNA undetectable rate (82% versus 0%), HBeAg seroconversion (82% versus 7%) and histological improvement. Moreover, interferon-γ-expressing T-helper cells were increased in patients with host-induced ALT increases.
Conclusions: Two patterns of ALT increases may occur during nucleoside/nucleotide analogue treatment. Host induced ALT increases, accompanied by decreased HBV DNA, lead to better long-term clinical outcomes.
Serum alanine aminotransferase (ALT) and HBV DNA are two main clinical markers of chronic hepatitis. These are commonly used to assess the viral response, adjust the treatment strategy and predict the clinical outcomes after anti-HBV treatment with either interferon (IFN) or nucleoside/nucleotide analogues [1–3].
Being one of the most easily monitored clinical markers, ALT has been considered as a good indicator of inflammatory activity. During IFN-based therapy, approximately 25–40% of patients exhibit an ALT flare that can lead to hepatic decompensation but that is also associated with an increased response rate to therapy . Flink et al.  and Janssen et al.  identified two types of ALT flares. Virus-induced flares occured after increased HBV DNA level with low treatment response. By contrast, host-induced flares were followed by decreased HBV DNA that was associated with high treatment response. This favourable ALT flare was recognized by the immunostimulatory effects of IFN . However, most of the host-induced ALT flares that were considered only occurred after IFN treatment. In a study of hepatitis B e antigen (HBeAg)-positive chronic hepatitis B patients treated with pegylated IFN, pegylated IFN plus lamivudine or lamivudine alone, patients treated with pegylated IFN showed a clear association of a higher ALT increase (flares) with a higher HBeAg seroconversion rate . Patients receiving lamivudine alone had no such association of the level of ALT increase and seroconversion. The combination group was in between .
Until now, few studies have investigated the incidence of host-induced ALT increases during nucleoside/nucleotide analogue treatment. Our recently published study identifying two types of HBeAg seroconversion in chronic hepatitis B patients after nucleoside/nucleotide analogue treatment found that the seroconversion with ALT increase had a sustained favourable outcome and lower reversion rate than for those with no ALT increases (50% verseus 23%; P<0.01) . To further identify ALT increases during nucleoside/nucleotide treatment and the effects on long-term clinical outcomes, the current study investigated HBeAg-positive chronic hepatitis B patients who were treated with nucleoside/nucleotide analogues (either lamivudine, adefovir dipivoxil, entecavir or telbivudine) for at least 2 years, with a 1 year follow-up period. Clinical characteristics including serum ALT levels and other diagnostic profiles were examined at baseline and at 3-month intervals thereafter.
A total of 170 treatment-naive chronic hepatitis B HBeAg-positive patients from Beijing Friendship Hospital (Capital Medical University, Beijing, China) were treated with nucleoside or nucleotide analogues. Treatment regimens were either 100 mg per day lamivudine, 10 mg per day adefovir dipivoxil, 0.5 mg per day entecavir or 600 mg per day telbivudine. All patients were persistently positive for hepatitis B surface antigen (HBsAg) for at least 6 months, negative for antibodies to HCV and hepatitis delta virus, and showed no serological markers suggestive of autoimmune disease. None of the patients had a history of alcohol abuse, drug use or hepatotoxin exposure. For these HBeAg-positive HBV patients, either HBV DNA levels >20,000 IU/ml with increased ALT levels >2-fold the upper limit of normal (ULN), or liver biopsy showing moderate/severe inflammation or significant fibrosis with detectable HBV DNA levels [10,11]. Among all 170 treatment-naive HBeAg-positive chronic hepatitis B patients, 78 patients underwent liver biopsy before treatment and 15 had biopsy before and after treatment. Grade 0–4 indicated inflammation/necrosis and stage 0–4 indicated fibrosis/cirrhosis. The definition of improvement in histology was one grade or one stage improvement. Liver function tests including serum ALT, HBV markers and HBV DNA were detected before treatment and every 3 months following treatment. HBeAg seroconversion was defined as the persistent loss of serum HBeAg and the development of anti-HBe.
Serum ALT levels were measured by routine automated methods according to the manufacturer’s instructions with the definition of ULN ALT level as 40 U/ml. ALT increases (flares) were defined as ALT>2-fold baseline [5,12]. Serum HBV DNA levels were determined by a real-time TaqMan PCR assay, as previously described . Serum HBsAg, anti-HBs antibody, HBeAg, anti-HBe antibody and anti-HBc antibody were examined by electrochemiluminescence immunoassay using an Abbott Architect i2000 (Abbott, Wiesbaden, Germany) according to the manufacturer’s instructions.
The peripheral blood cells from 21 patients (15 patients with virus-induced increase and 6 with host-induced increase) before treatment and the time of ALT increase were investigated in intracellular cytokine expression of IFN-γ and interleukin (IL)-4. Briefly, 200 μl of blood was incubated at 37°C for 4 h in PMA (50 ng/ml), ionomycin (500 ng/ml) and brefeldin A (10 ng/ml) in RPMI media. After the cells were fixed with 2% paraformaldehyde for 20 min, cells were washed and permeabilized with 1% BSA, 0.5% saponin in phosphate-buffered saline for 10 min. These cells were stained with anti-CD3, anti-CD8, anti-IFN-γ and anti-IL-4, and were gated for CD8+ cells; the population was analysed for IFN-γ and IL-4 expression by FACS .
Statistical analyses were performed using the Statistical Programme for Social Sciences (version 11.5; SPSS, Inc., Chicago, IL, USA). Continuous variables, such as ALT, HBV DNA and time were presented as medians (range) and categorical variables as frequencies. Categorical variables were tested by the χ2 or Fisher’s exact test. P-values of <0.05 were considered statistically significant.
The demographic features of the patients are shown in Table 1. Generally, a total of 170 HBeAg-positive treatment-naive chronic hepatitis B patients were mostly male (male:female =112:58) with a median age of 37 years, median ALT of 75 U/l and median HBV DNA of 6.5 log IU/ml . These patients were treated with one of the four available oral anti-HBV medicines, lamivudine, adefovir dipivoxil, entecavir or telbivudine, for at least 2 years and followed up for 1 more year. Clinical characteristics were examined at baseline and at 3-month intervals.
ALT, alanine aminotransferase; HBeAg, hepatitis B e antigen.
Each patient was analysed individually by on-treatment ALT and HBV DNA. Two patterns of ALT increase were found. Virus-induced ALT increase was characterized by a rapid increase of ALT accompanied by HBV DNA breakthrough (an increase by at least 0.8–1.0 log IU/ml; Figure 1A). Most of these increases were induced by drug resistance through HBV mutants after 1 year of nucleoside/nucleotide treatment, as seen for lamivudine treatment (28.9%), adefovir dipivoxil treatment (18.4%) and telbivudine treatment (20.0%), but low in entecavir-treated cases (1.7%). After add-on rescue therapy, both increased ALT and HBV DNA of most of the patients decreased.-
Host-induced ALT increase was signified by a moderately increased serum ALT with decreased HBV DNA (Figure 1B). These increases were less common but were still observed for lamivudine (11.5%), adefovir dipivoxil (7.9%) or telbivudine (10.0%) treatment groups and were very rare in the entecavir group. In this study, ALT and HBV DNA were detected before treatment and every 3 months after treatment. Although ALT and HBV DNA may change separately during the 3-month interval, in our study most host-induced increases showed a decrease in HBV DNA at the same time that ALT was rising.
There was no significant difference in gender, age, ALT or HBV DNA before treatment when comparing the baseline characteristics of virus-induced ALT increases and host-induced ALT increases (Table 1). However, during nucleoside/nucleotide analogue treatment, these two patterns of ALT increase had observable differences in frequency, length of occurrence, increased ALT fold and HBV DNA response (Table 2 and Figure 2). Virus-induced ALT increases were more common than host-induced ALT increases (15.9% versus 6.5%). During nucleoside/nucleotide analogue treatment, virus-induced ALT increases most often occurred in the second and third years (92.6%) due to the development of anti-HBV drug resistance, with a median time of increases of 96 weeks after treatment. Meanwhile, host-induced ALT increases could be seen mainly in the first year (63.6%) with the median time of increases 36 weeks after treatment and may be induced by immune system involvement. Serum ALT level was typically higher during virus-induced increases than during host-induced increases (5.7- versus 2.5-fold ULN). More importantly, HBV DNA level increased by 2 log during virus-induced increases compared with a 2 log decrease during host-induced increases.
ALT, alanine aminotransferase; HBeAg, hepatitis B e antigen; ULN, upper limit of normal.
Multivariate analysis was performed to predict ALT increases and revealed that gender, age, baseline ALT or baseline HBV virus load were not significantly associated with ALT increase.
The three most valuable clinical long-term outcomes were HBV DNA undetectable rate, HBeAg seroconversion and histological improvement. Although the HBV DNA undetectable rate was similar after 1 year of treatment, it was significantly higher for those with host-induced increases than virus-induced increases in the second year (82% versus 11%; Figure 3) and the third year (82% versus 0%). HBeAg seroconversion rate was also higher for host-induced increase patients with seroconversion in approximately 82% of patients after 3 years, which was sustained (Figure 3). Virus-induced increase patients had only 7% HBeAg seroconversion after 3 years. Histological data also showed similar tendency to viral response and HBeAg seroconversion with both inflammation and fibrosis improved in the host-induced increase group, but the limited case number did not reach the requirement for common statistical analysis.-
Three different definitions for ALT increase were used: ALT≥2×ULN, ALT≥3×ULN and ALT≥5×ULN. As shown in Table 2, ALT≥3×ULN had a similar clinical outcome compared with ALT≥2×ULN, that is, host-induced increases were associated with favourable long-term treatment outcomes in HBV DNA undetectable rate and HBeAg seroconversion. Because ALT≥5×ULN only had 3 cases of host-induced ALT increase, it is difficult to do this comparison. In addition, ALT increase with treatment response of each nucleoside/nucleotide analogue was analysed (Table 3).
Values shown are n/total n unless indicated otherwise. Responders were defined as patients with a decrease in serum HBV DNA to undetectable levels with alanine aminotransferase in the normal range. Partial responders were patients with a decrease in serum HBV DNA by >2 log IU/ml but still detectable. Non-responders were patients with a decrease in serum HBV DNA by <2 log IU/ml.
IFN-γ-expressing T-cells in peripheral blood were highly associated with elimination of HBV and anti-HBV treatment response. Immune clearance of HBV, as with most viral infections, depends on the activation of specific CD8+ cytotoxic T-lymphocytes and appropriated CD4+ T-helper cells. As shown in Figure 4B, host-induced ALT increases had higher numbers of IFN-γ-expressing T-helper cells than the virus-induced ALT increase group and healthy control (25.3% versus 15.1% and 15.8%; P<0.05).-
ALT increase is a very common issue in chronic hepatitis B patients. During nucleoside/nucleotide analogue treatment, increased serum ALT is usually considered an indicator of virus resistance that may lead to severe inflammation and even liver failure. However, increased ALT has also been found to be a favourable marker for higher HBV DNA undetectable rate and HBeAg seroconversion [14–16] from either the natural history of spontaneous ALT increases  or from the experience of anti-HBV treatment. In the present study we did not only find good (host-induced) and bad (virus-induced) ALT increases during nucleoside/nucleotide analogue treatment of chronic hepatitis B patients, but also investigated these two patterns of increases in association with clinical outcomes. Most importantly, we attempted to investigate the mechanisms involved in production of these increases.
First, identifying good (host-induced) and bad (virus-induced) ALT increases is important during anti-HBV treatment in real clinical practice. When a significant increase of ALT is accompanied by a decrease in HBV DNA, the doctor should encourage the patient to continue treatment and expect a positive outcome. On the contrary, when ALT increases with HBV DNA breakthrough is observed in a patient, it is likely due to the development of resistance and highlights the need for add-on rescue therapy or switch to another treatment.
Second, it seemed ALT increases (especially host-induced increases) were different between nucleoside/nucleotide analogue and IFN treatment. Based on the peak increases, our study found the median ALT of host-induced increase by nucleoside/nucleotide analogue-treated patients was 2.5×ULN (range 2.0–10.6), and is significantly lower than that reported by Flink et al. , which shows the ALT median at 13.8×ULN (range 5.3–60) with IFN-treated patients. From the frequencies of ALT increases, our study found only 6.5% of nucleoside/nucleotide analogue-treated patients had host-induced increases, which are significantly less common than that of IFN-treated patients (36%) . This difference in ALT increases between nucleoside/nucleotide analogue and IFN is likely related to the immune system. Although previous reports using lamivudine [17,18] and telbivudine  treatments showed significant effects on T-cell subsets and PD-1 expression, IFN still had a much stronger immune modulation than the oral nucleoside/nucleotide analogues.
Finally, we would like to mention that our study was retrospective and heterogeneous with four oral antiviral medicines. Further prospective clinical trials involving increased numbers of patients with each nucleoside/nucleotide analogue, analysis of HBV genotypes, modes of transmission, histology of liver biopsies, early viral response, HBV mutations and the immune response of individual patients will likely strengthen our understanding of ALT increases in association with disease and prognosis.
This work was supported by the National Natural Science Foundation of China (30972602), Major Science and Technology Special Project of China Eleventh Five-year Plan (2008ZX10002-004 and 2009ZX10005-016), Program for New Century Excellent Talents in University to HY, National High Technology Research and Development Program of China (2006AA02A410) and Major State Basic Research Development Program (2007CB512802).
The authors declare no competing interests.