Despite a decline in cases of acute hepatitis B and the low hepatitis B virus (HBV) chronicity rates in adults, still some patients progress to HBV-related fulminant liver failure. In this review, we discuss treatment options that may prevent the progression of severe acute hepatitis B to fulminant liver failure and death. In severe acute HBV with prolonged prothrombin time and increased bilirubin, interferon failed to be effective while antiviral treatment, particularly with lamivudine, appears to improve survival (mean survival almost 80%). Outcome without antiviral therapy has remained considerably poor, whereas there is no convincing evidence of amelioration of HBV-targeted immunity. Of note, most patients who died or required transplantation despite lamivudine therapy, were started on lamivudine at advanced stages compared with those survived. This suggests that prompt and timely antiviral therapy is crucial. Owing to the abovementioned results the design of randomized placebo-control trials in the setting of severe acute hepatitis B seems unethical. On the contrary, the design of multicentre double-blind randomized trials to compare the efficacy between lamivudine and entecavir or even tenofovir in acute severe HBV cases is ideally needed, but these studies appear to be very difficult to perform considering that these cases are not frequent and therefore, it is almost impossible to have two arms adequately numerous and homogenous for statistical evaluation. Thus, in the absence of solid evidence based data, the hepatologists could treat their patients with severe acute hepatitis B with lamivudine or the most potent antivirals entecavir or tenofovir.
Hepatitis B virus (HBV) is the smallest DNA virus known to infect and cause disease in humans. About 350 million people are believed to be chronically infected although about 2 billion have been exposed to the virus. This indicates that approximately 80% of patients clear the HBV surface antigen (HBsAg) spontaneously (Table 1). Clearance rates are exceeding 90% in those patients with clinically evident acute hepatitis with considerable higher rates in icteric vs. anicteric cases (Fig. 1; Table 1) .
|Chronicity rate (%)||Population studied, 95% Confidence intervals, additional comments||Reference|
|13.3||Alaskan natives, 25/188 (CI: 8.5–18.2%) age dependent (7.7% in 30 years of age or older vs. 28.6% in less than 4 years of age)||22|
|0.5||US veterans, 1/392 (CI: 0.6–1.4%)||28|
|0.2||Greek patients 1/507 (CI: 0.04–1.1%)||29|
|8.5||Shanghai, 25/294 (CI: 5.31–11.69%), genotype dependent [C2 20/50 (40%, CI: 26.4–53.6%), B2 5/36 (13.9%, CI: 2.6–25.2%)]||23|
|1||Japan 3/301 (CI: 0–2.1%)||33|
|13.4||Japan 11/82 (CI: 6–20.8%, genotype dependent: 7/31 genotype A, 1/9 genotype B, 3/42 genotype C)||24|
|2.6||Taiwanese students 1/39 (CI: 0–7.5%)||27|
|6.3||Germany 5/80 (CI: 2.5–14.4%)||25|
|4.6||Turkey 11/240, (CI: 2.4–12.1%)||26|
Overall, because of the vaccination success, most countries see declines in the reported incidences of acute hepatitis B [2-6]. However, still acute hepatitis B is a cause of death, i.e. the most recent report on the incidence of hepatitis B in Germany reported three deaths in 748 patients (0.4%) with acute hepatitis B . The mortality rate is substantially higher among hospitalized patients, with 36 of 890 (4%) Japanese and 90 of 2,169 Spanish patients died [8, 9]. This rate was 1% in those without fulminant hepatitis vs. 52.8% among the 53 patients with fulminant hepatitis. Thus, there could be good reason to treat acute hepatitis B, but there might likewise be arguments against treatment of acute hepatitis B.
Treatment to prevent evolution to chronic hepatitis B virus infection?
The HBsAg to anti-HBc ratio ranges between 1 in 4 in high endemic regions (with HBV infection early in life in the absence of vaccination) to 1 in 12 in low endemic regions [10-16]. The higher HBsAg to anti-HBc ratio usually occurs in population where mother to child transmission or early childhood infection is the dominant route of infection, which would be unlikely to be amendable to antiviral strategies other than vaccination and potentially antiviral therapy in late pregnancy [17-21].
These HBsAg to anti-HBc ratios would suggest 75–92% spontaneous clearance rates, fitting with the frequently cited 90–95% spontaneous clearance rates in adults. This fits with a study in 188 Eskimo patients with acute hepatitis B, where HBV chronicity was age dependent (7.7% in 30 years of age or older vs. 28.6% in less than 4 years of age; Table 1) .
Clearance rates of acute hepatitis B are generally believed to be between 90% and 97.5% for adult population (Table 1) [22-27]. However, clearance of HBsAg after acute infection in adults might occur in as many as 99.5% reported from a study of 392 US Army personnel  and 99.8% from a study of 507 Greek patients . This variation could potentially be because of the difference in HBV e antigen (HBeAg) status and precore mutants, respectively, as well as viral inoculum. Indeed, higher HBV chronicity rates have long been associated with the HBeAg status as attested for instance by the increase rate of vertical transmission of the infection in infants born from HBeAg-positive compared with HBeAg-negative mothers [30-32]. A recent study from Shanghai reporting only a 92% spontaneous clearance rate actually suggested a genotype effect in regard to chronicity . This genotype effect is in line with some cross-sectional observations of genotype difference in relation with acute and chronic hepatitis B from the US , as well as a 22-year multicentre cross-sectional Japanese study where chronicity evolved in 2 of 23 (9%) genotype Ae vs. only 1/187 (0.53%) genotype Bj patients . This later report is similar to a smaller report also from Japan reporting higher persistence of HBV in genotype A vs. genotype B or C .
Certainly, aiming at preventing HBV chronicity as in acute hepatitis C virus (HCV) infection [35, 36], would be extremely unlikely successful, given the low chronicity rate of acute hepatitis B cases particularly in adults (see Table 1). Indeed, in an attempt to perform a solid acute hepatitis B study to proof prevention of chronic HBV infection through treatment intervention, we need two arms with numerous and homogenous groups of patients which is practically impossible and also very cost intensive. However, shortening of disease duration and preventing death because of progression of severe acute HBV cases to HBV-related fulminant liver failure could be a reasonable target for antiviral therapy in acute hepatitis B.
Another reason for treating acute hepatitis B is the higher chance of being HBV-DNA negative, in case the acute hepatitis progresses to liver failure and a liver transplantation is required as the lower the HBV-DNA at the time of transplantation, the lower the risk of reinfection after transplantation. However, given the very effective antiviral therapies today, starting at the time of transplantation might still be sufficient. In addition, a potential risk of the treatment of acute hepatitis B to actually increase persistence comes from a Japanese study in 57 acute HBV cases . In this study, a 12% (7 of 57 patients) persistence rate was reported and the role of prednisolone or glycyrrhizin administration during the acute illness in HBV persistence cannot be ruled out (the infection persisted in 86% of patients who received treatment vs. only 2% in those who did not; P = 0.01) .
Antivirals in the setting of severe acute to fulminant hepatitis B
In the setting of fulminant hepatitis liver transplantation remains an option where the graft is endangered by reinfection. For patients with chronic hepatitis B, pretreatment with antivirals plus HBIg starting at time of transplantation has become standard of care in most centres about 10 years ago. When reports about the safety of lamivudine in the setting of severe life threatening reactivation emerged, we started using lamivudine to also decrease risk of reinfection in the setting of fulminant hepatitis B . However, when we initiated lamivudine therapy to patients with fulminant hepatitis B to prepare them for liver transplantation, we observed immediate improvement in alanine aminotransferase (ALT), prothrombin time and bilirubin in most patients (see an example in Fig. 2a), only few patients showed prolonged bilirubin elevation (see Fig. 2b) despite immediate clinical improvement and normalization of prothrombin time. In a large series of 37 patients treated with lamivudine for severe/acute to fulminant hepatitis B, a significantly better survival rate was observed compared with historical controls (78.3% vs. 20%, P < 0.001) . The definition of severe acute hepatitis B in the later study , but also in other similar studies [39, 40], was according to the presence of at least two of the following three parameters: bilirubin levels more or equal to 10 mg/dl, International Normalized Ratio (INR) more or equal to 1.6 and the presence of hepatic encephalopathy.
Similar to our data , very good survival rates (mean survival rate almost 80%) in patients receiving lamivudine have been reported by others [39, 41-52], whereas outcome without antiviral therapy has remained poor [53-62] (see also Table 2).
|Transplant-free survival with lamivudine||Transplant-free survival without lamivudine|
|29/37 (78.3%)38||4/20 (20% historical control, P < 0.001)38|
|7/10 (70.0%)41||6/23 (26.1% historical control, P < 0.001)41|
|3/3 (100%)42||1/3 (33.3%)38|
|5/6 (83.3%)43||0/10 (0%) [HCV-coinfected]55|
|13a/15 (86.6%)39||0/5 (0%) [Dialysis population]56|
|5/5 (100%)45||0/5 (0%)58|
|9/10 (90%)46||1/1 (100%)59|
|16/17 (94.1%)47||7/9 (77.7%)60|
|3 cases 3/3 (100%)48–50||23/40 (57.5%)61|
|b14/38 (37%)51||b6/39 (15.3%, P = 0.03)51|
|37/40 (92.5%)52||30/40 (75%, P = 0.03)52|
|Total: 143/183 (78.2%)||Total: 75/164 (45.7%)|
Importantly, in studies using lamivudine, patients who continue to require a liver transplantation or those who died were started on lamivudine therapy later at the advance stages of their disease with clinically overt encephalopathy, where there might already have been too much liver tissue lost for recovery compared with those surviving. This was clearly demonstrated in two studies from China [51, 52]. In the first study, the mortality of patients with fulminant hepatitis B in the lamivudine group was significantly lower than that of the control group, but this difference was observed only for patients without systemic inflammatory response (SIRS) . For patients with fulminant hepatitis and SIRS, which in general is developed in the advanced stages of a disease course, there was no difference  (Table 2). In the second study, 80 patients with severe acute hepatitis B were randomized for lamivudine or placebo . Significant clinical benefit was observed, with only 3/40 (7.5%) dying in the lamivudine arm compared with 10/40 (25%, P = 0.03) in the placebo arm  (Table 2). As patients were receiving active lamivudine in case of failure to improve the authors could evaluate the early administration of lamivudine was crucial to patients not developing liver failure, independent of death .
Contrary to all previous mentioned studies [38, 39, 41-52], an article from India  recruiting 71 patients with 31 (22/31 with severe acute HBV infection) receiving 100 mg lamivudine daily and 40 in the placebo group reported that lamivudine is of no use in severe acute hepatitis B. However, patients on lamivudine had significantly lower levels of HBV-DNA (median: 3.7 log copies/ml vs. 4.3 log copies/ml; P = 0.04) at the week 4 visit while, tended to improve their serum bilirubin, ALT and INR values faster, although failed to reach level of significance . Based on our own calculations, 140 patients with severe acute HBV infection would be required to proove superiority for faster improvement. For non-superiority, an even higher number of patients are required compared with a superiority trial. Thus, it seems that this Indian study  was under powered. Therefore, we think it is reasonable to reinforce that the recent EASL recommendations  on the management of acute severe hepatitis B should probably be currently followed until proven incorrect.
Whether or not other antivirals will be as beneficial in acute severe hepatitis B will remain to be seen. There is no doubt that entecavir and tenofovir are superior to lamivudine in chronic hepatitis B , but in the setting of severe acute hepatitis B there is limited data compared with lamivudine as the most potent antivirals entecavir and tenofovir are clinically available only for a few years. Actually, a prospective study in six patients with HBV-related acute liver failure receiving 1 mg/day of entecavir within 1–18 days after admission showed normalization of prothrombin time, ALT and bilirubin within 3 months, although one patient with late onset of treatment went on to develop chronic hepatitis B . For completeness, one additional patient with severe acute hepatitis B and successful course using entacavir has been reported from Italy , whereas two patients from Germany with severe acute hepatitis B and a favourable outcome were recently published after switching lamivudine to entecavir or to an add-on treatment with adefovir .
On the contrary, more data concerning the two newer antivirals entecavir and tenofovir has been published in cases with severe acute reactivation of chronic or occult HBV infection after immunosuppression or chemotherapy, a situation which is similar enough with that of the “naïve” acute severe to fulminant hepatitis B [65-78]. In these studies the treatment was started as rescue therapy after exacerbation of chronic or occult hepatitis B and particularly in those who had received rituximab-based regimen, with favourable outcome in most of the affected cases [65-78]. It should be noted, however, that not all of these patients with HBV reactivation suffered from severe hepatic failure as defined above [38-40], while several difficulties of managing severe HBV reactivation are still present (e.g. in terms of the optimal duration of prophylaxis) . Furthermore, one article described an impaired outcome when entecavir was compared with lamivudine in the setting of severe reactivations of chronic hepatitis B , whereas drug-induced liver injury or lactic acidosis have also been described in cases of HBV reactivation treated with the newer potent antivirals [81, 82]. Thus, it might be that lamivudine also exhibits a yet unexplained immediate benefit for the liver unrelated to its antiviral potency, given the mostly immediate response observed in most of our patients with severe acute hepatitis B, with improved prothrombin time after only 1 day of lamivudine therapy, whereas 11/14 patients who did not require a liver transplantation showed normalization of prothrombin time from less than 40% within a week of lamivudine therapy . As in “naïve” severe acute to fulminant hepatitis B cases, the potential antiviral treatment is not for a long time as in cases of chronic hepatitis B or in cases with either overt or occult HBV infection who suffered from severe acute reactivation, the use of lamivudine seems to be logical as it has the advantage of the lower cost compared with the newer antivirals whereas the possibility of the emergence of YMDD mutants is rather very low.
Interferon in patients with fulminant hepatitis B
Interferon was explored for treating chronic hepatitis B in the 1980s , initially as lymphoblastoid and later as recombinant interferon . An early study in 1982 suggested low interferon levels in patients with fulminant hepatitis B similar to healthy volunteers and significantly lower than in chronic hepatitis B or other viral infection . Thus, subsequently, studies evaluated the role of interferon in fulminant hepatitis B and acute hepatitis B (see Table 3a and b), but failed to demonstrate any significant benefit [85-89].
|(A) Setting of fulminant hepatitis|
|16 (10 HDV) vs. 70 historical controls||FHF, grade 3 or 4 coma||25% survival (20% in HDV, 33.3% in HBV) vs. 27.4% (historical controls; 25.9% in HDV, 33.3% in HBV)||85|
|11 (7 HDV coinfection, 2 HDV superinfection)||FHF, grade 1–3 coma||No improvement (2 of 9 HDV survived)||86|
|(B) Setting of non-fulminant hepatititis|
|20 vs. 34||Prolonged hepatitis B (HBsAg positive 12 weeks after dx)|
16/20 (80%) HBsAg clearance IFN 3 × 5 MU
18/34 (53%) HBsAg clearance no IFN
|100 patients (ALT > 10 ULN, Bil> 5 mg/dl|
Placebo (n = 33)
3 MU rIFN-a2b (n = 34)
10 MU rIFN-a2b (n = 33)
Earlier HBsAg clearance on 3 MU
Higher anti-HBs with 10 MU
Overall inconsistent results
|309 acute HBV||Only abstract in English||States beneficial outcome with IFN||89|
However, once patients have progressed to fulminant liver failure with clinically overt encephalopathy, there might already be a too advanced disease with poor chance for liver regeneration. As noted above, this could be also the case for patients with fulminant hepatitis B being treated at the advance stages of their disease with lamivudine or other nucleosides, since the American fulminant hepatitis network reported lack of benefit with lamivudine use . Indeed, lamivudine was received by the older and more severely ill patients in that study . Therefore, evaluating interferon therapy for patients with acute hepatitis, but not yet in fulminant liver failure, might be an option. Although interferon is associated with potential adverse immune activation leading to more severe acute hepatitis, none of the studies performed so far employing interferon found evidence for increased severe adverse events, although side effects were more frequent with interferon than the placebo groups (see Table 3a and b) .
Three studies have been published [87-89], one of which was only available as English abstract of the otherwise Chinese article, which stated that interferon was beneficial . One study found inconsistent results with opposing direction for 3 million vs. 10 million units of interferon three times weekly  whereas, the third study took into account an interesting approach evaluating only those patients who were still HBsAg positive at week 12 . This is a minority of all patients with acute infection, in whom, however, an immune activating approach might be more promising than the antiviral therapy.
In overall, while interferon failed to be effective in fulminant hepatitis B, lamivudine might be efficient especially when given early enough before the development of the advanced stages of the disease with clinically overt encephalopathy.
Do other therapies work too?
Treatment to abbreviate the duration of being ill
Direct antivirals in the setting of acute hepatitis B could have the advantage of shortening the replication and thereby reduce disease duration without the side effects of interferon. However, a fast reduction of HBV proteins might lead to lower presentation to the immune system and subsequently might negatively impact immunity against HBV. Indeed, the fast decline of HBV-DNA in most patients with acute hepatitis B (Fig. 1) serves as an additional argument against using direct antiviral in the setting of acute hepatitis B as early studies on HBV viral markers suggested the more severe the acute hepatitis, the lower the viral load [91, 92]. In a study of 110 patients with acute icteric hepatitis, only 19% had detectable HBV-DNA with hybridization assay and a limit of detection of 0.045 pg/ml (corresponding to approximately 12600 copies/ml or 2500 IU/ml). Interestingly the detection of HBV-DNA was less likely if patients were already beyond the peak of their ALT values with 5/80 (6.25%) vs. 16/30 (53.3%) in patients who could be tested for HBV-DNA prior to the peak of their maximal ALT value .
Subsequent studies with more sensitive assays using the chimpanzee model and evaluating an outbreak of acute hepatitis B in an “autohemotherapy unit” confirmed in more detail the course of HBV-DNA increase and decline during acute hepatitis B in relation with ALT and symptoms (Fig. 1) [94-96].
Other treatment(s) of acute hepatitis B
Randomized trials in the setting of acute hepatitis B
Given the observed differences between lamivudine treated [39, 41-52] and untreated patients [53-62], a randomized trial in the setting of severe acute hepatitis B seems unethical. However, in the past, several small trials have evaluated several different options, none of which was further developed. The likelihood to see a study solidly proving superiority of treatment intervention for acute hepatitis B is unlikely, taking into account the low chronicity rates in adults, the low frequency of severe acute to fulminant HBV cases, and the need of large and homogenous cohorts of patients that should be included. In this context, an exception could be persons with documented acute hepatitis B and prolonged persistence of HBsAg or prolonged disease activity .
As stated above in severe acute hepatitis B with prolonged prothrombin time and increased bilirubin, antiviral treatment intervention seems to improve survival, while there is no convincing evidence that this kind of therapy in that setting would ameliorate HBV-targeted immunity [38, 39, 41-52]. The most solid data are indeed published for lamivudine whereas the same level of evidence to demonstrate safety in severe acute hepatitis B is not yet available for entecavir or tenofovir as these newer more potent antivirals have become available more recently. However, existing evidence shows that entecavir and tenofovir are efficient in a similar situations as the acute, severe reactivation of overt or occult HBV infection [65-78], but still toxicity with entecavir or tenofovir cannot be excluded at the same level of certainty [81, 82]. Overall, we can suggest that lamivudine or the more potent antivirals could be used by the hepatologists if they must treat cases with severe acute hepatitis B.
Thymosin: In a placebo controlled trial, thymomodulin as oral agent (Leucotrofina) did improved ALT normalization, but non-significantly the HBsAg clearance .
Ursodeoxycholic acid (UDCA): One randomized study evaluated the role of UDCA in acute hepatitis B and found some benefit on ALT normalization, but persistence of HBsAg at 12 months was seen in only 1/33 UDCA-treated patients vs. 5/26 placebo treated patients (P = 0.02). However, the reported frequency of chronicity in the placebo group is rather unusual and despite randomization the two groups differed in some characteristics . In addition, another study in 43 HBV patients did not find a difference on the clinical outcome in relation with UDCA use . Of interest, however, none of the patients developed persistent HBsAg .
Cyanidanol: In the 1980s, a study reported fast recovery of acute hepatitis when Cyanidanol was given vs. placebo . However, chronicity rates in both arms were similar with 4/42 and 3/41 patients respectively . Thus, this trial was also in line with earlier studies. Seven controlled trials using Cyanidanol in the setting of acute viral hepatitis had been reported at a conference, concluding that benefits were not consistent and the clinical relevance was judged to be small, although it was believed that Cyanidanol might have immune modulating potential, thereby accelerating recovery .
N-Acetylcystein (NAC): NAC was found to inhibit HBV replication in a cell culture model , and whereas NAC might be beneficial in the setting of acute liver failure , no significant benefit was found in acute viral hepatitis .
Stop hepatitis B virus infections in outbreak situations
Treating acute HBV infection to stop outbreaks of HBV is not quite reasonable. As indicated in several studies ([91-93]; Fig. 1), by the time patients become symptomatic and would therefore seek medical attention and be amendable to treatment, the viral load is already low or undetectable. Furthermore, outbreaks of HBV infection have usually been tracked to inadequate safety precautions with medical procedures rather than person to person transmission. Ideally, no such outbreaks shall occur nowadays. However, 15 outbreaks have already been identified in the last 10 years in the US, all because of unsafe glucose monitoring practices . Of note, mortality was found to be higher (4.1%) in this older population than what is generally observed in patients with acute hepatitis B, which is estimated to lead towards fulminant hepatitis in less than 1%.
Therefore, it might be worthwhile exploring immediate screening for HBsAg and treatment with antiviral in the setting of HBV outbreaks in nursing homes or assisted living facilities to prevent death by intervention with antiviral therapy pre-emptively in case of detectable HBV viremia prior to onset of hepatitis. However, realistically it is unlikely a study be large enough to achieve 80% power for demonstrating efficacy, also taking into consideration the ethical dilemma of withholding antiviral therapy in such settings. It shall be stressed, however, that HBV-infection can be prevented by adequate precautions and vaccination of persons potentially exposed to HBV .
- Aiming at preventing HBV chronicity by treatment of acute cases would be extremely unlikely to proove success, given the low chronicity rates in adults, the low frequency of severe acute to fulminant HBV cases and the need for large and homogenous cohorts of patients that should be included.
- For those patients with acute severe to HBV-related fulminant liver failure, interferon failed to be effective while antiviral treatment particularly with lamivudine, to prevent further deterioration of liver function and subsequently to prevent HBV reinfection in the setting of potentially needed transplantation, appears to improve survival (mean survival approaching almost 80%) especially when lamivudine is started before the development of the advance stages of the disease with clinically overt encephalopathy suggesting that a prompt and timely initiation of antiviral therapy in patients with significantly prolonged prothrombin time as marker of impaired liver function is crucial.
- Other antivirals than lamivudine have been investigated so far, in small case series of acute severe hepatitis B with some promising preliminary results mostly with the use of entecavir. Ideally, the design of a multicentre double blind randomized trial to compare the efficacy of lamivudine and entecavir or even tenofovir in patients suffering from acute severe to fulminant hepatitis B would be helpful, but such a trial appears to be very difficult to be performed as these cases are not frequent and considering that two arms with adequately numerous and homogenous population are needed for statistical analysis.
- The outcome without antiviral therapy has remained considerably poor while there is no convincing evidence of amelioration of HBV-targeted immunity indicating that a randomized placebo-control trial in the setting of severe acute hepatitis B is rather unethical and therefore, at present the EASL recommendations  on the management of acute severe hepatitis B with the use of nucleos(t)ide analogues should be followed until proven incorrect.
- Last but not least! Severe reactivations of chronic or occult hepatitis B during or after cessation of intensive immunosuppression or chemotherapy should be treated immediately with potent antivirals as there is already much evidence showing that entecavir and tenofovir are very efficient in these cases [17, 65, 66, 79].
- “A million dollar answer”: Although difficult to be performed further research is worthwhile in an attempt firstly to investigate the predictive pathogenetic pathways which can lead to the progression of acute hepatitis B to severe acute hepatitis and fulminant liver failure in an index patient and secondly, to promptly identify those patients who will develop severe acute to fulminant hepatitis B and therefore, would be candidates for treatment intervention with oral nucleos(t)ides.
HLT received support from BMBF via DLR project 01KG0507; KZ and GND have partially been supported by the Research Committee of the University of Thessaly (Code No: 2466).