Natural history of untreated HDV patients: Always a progressive disease?

A severe course has been described in early studies on chronic hepatitis D (CHD), with faster pace towards liver cirrhosis with subsequent high liver‐related morbidity and mortality in the majority of patients. Earlier studies have included risk groups as people using intravenous drugs (PWID) or those with multiple co‐morbidities. During the last decade, the epidemiological landscape of CHD has changed with domestic cases decreasing while increasing cases of CHD consisting of younger persons immigrating from endemic regions to low‐endemic regions. Recently, further insights into the spectrum of the disease with an indolent disease course in a substantial proportion of persons with CHD have been gained. At diagnosis, ≥30%–50% had already established liver cirrhosis. Older age, liver cirrhosis, co‐infection with HIV and lack of interferon (IFN) therapy are the main predictors of worse clinical outcome. The newly introduced and upcoming antivirals against CHD are highly anticipated, considering the historically low virological response rates to antiviral therapy. Further knowledge is needed to fully comprehend the natural course and the spectrum of this severe form of viral hepatitis. This is also to be able to evaluate the long‐term effects of the new antivirals on disease progression.


| INTRODUC TI ON
Global estimates of hepatitis D virus (HDV) infection suggest that 5%-10% of chronic hepatitis virus (HBV) carriers are co-infected, translating to 15-25 million individuals. 1-3 Certain high endemic regions or "hot spots" of HDV infection exist in Eastern Europe, West-and Central Asia, Sub-Saharan Africa and the Amazon basin of South America. 4 A declining trend of HDV prevalence was reported in several countries including Italy, Spain, Turkey and Taiwan in the 1990s, owing to hepatitis B virus (HBV) vaccination as well as improvement of other protective measures. 5 The epidemiological landscape of chronic hepatitis D (CHD) has also changed in the last decade, with shift of the high-risk groups in high-income countries. [6][7][8] Previously, the highest risk for CHD was found among people using intravenous drugs (PWID), followed by other risk groups such as recipients of haemodialysis or multiple blood transfusions, men who have sex with men and commercial sex workers. 1,2 Concomitant of declining domestic populations of CHD, increasing dominant cases by immigrants with often younger ages have been observed. 6,9 Less virulent viral strains than previously observed have been shown to be circulating among these new risk groups which could influence the clinical course. 10,11 Intrahousehold clusters of HDV infection remain though prevalent in some endemic regions. 12

Hepatitis D virus (HDV)-hepatitis B virus (HBV) co-infection is
considered to be the most severe form of viral hepatitis, with rapid progression towards liver cirrhosis, hepatocellular carcinoma and liver-related death. 13 Recent data have though challenged this estimated severe disease course involving majority of infected persons, with possible further insight into the spectrum of the CHD disease. 14,15 Comprehension of the natural history of HDV infection with a wider perspective is important, especially considering the recently introduced or new antiviral treatments in pipeline against HDV and the need to assess the impact of these drugs on the natural course of the disease. 16,17

| CHALLENG E S IN A SS E SS ING THE NATUR AL HIS TORY OF HDV INFEC TI ON
Before describing the published data on the natural course of CHD with progression rate of the disease over time, the several challenges associated with assessing the general course through reviewing the results of the literature are addressed here to put the data in context. First of all, the number of longitudinal cohort studies has been relatively limited (shown in Table 1), not being so extensively studied as mono-HBV or hepatitis C virus (HCV) infection. Studies with longitudinal cohort design, rather than cross-sectional ones are needed to assess the risk for the outcomes over time. Furthermore, the sample size of patients with anti-HDV positivity has been generally small, with often even smaller sample size or non-available data of patients with HDV RNA positivity in the studies. The latter group is the real relevant population to be studied for the natural course of CHD, and also the target group for HDV antiviral treatments.
Furthermore, high heterogeneity within or between the cohorts has been present. [18][19][20] Varying proportion of patients with advanced liver disease at baseline or varying follow-up times, patients consisting of different proportion of genotypes or ethnicities, many studies based on patients diagnosed through anti-HDV positivity rather than confirmed HDV RNA viraemic cases with risk of including those with resolved infection, and inclusion of patients with different proportion on antiviral therapy which can alter the course of the disease have all contributed to the high heterogeneity of the cohort studies with difficulty to draw general conclusion from the studies.
As also mentioned above, the risk groups for CHD have shifted in the last decade in high-income countries, with recent data of cohorts with different characteristics. 5,10 Although the majority of persons with CHD reside in high endemic regions, the data of longitudinal cohort studies have been almost exclusively collected from low-endemic countries. Another aspect is that the measured endpoints have also varied in different studies, making the assessment cumbersome. 14,19 A possible selection bias could be that newly diagnosed HBsAgpositive individuals do not often undergo universal testing for anti-HDV around the world, but risk-based according to guidelines or clinician-driven decisions. 21 Also, selection bias towards patients with rapid progression to advanced liver disease could be present in studies from tertiary centres, with concern of overestimating the risk for the outcomes. 14,[18][19][20] Ideally large cohort studies of CHD cases in the community with less selection bias would be needed, but obstacles such as CHD being a largely undiagnosed disease and low prevalence make such studies hard to achieve. 22

| CO -INFEC TI ON OR SUPER-INFEC TI ON AT TR ANS MISS ION
Infection with HDV can occur through co-infection with HBV to individuals without prior HBV infection, or by super-infection to chronic HBV carriers. 23 While simultaneous co-infection causes an acute selflimited hepatitis with recovery in the majority of cases, super-infection is associated with a chronic (>6 months) and a more aggressive disease course. 24,25 Chronic HDV infection is therefore more because of super-infection than co-infection. Though, there is an increased risk of fulminant hepatitis, regardless of pattern of HDV infection. [26][27][28]

| HI G H PRE VALEN CE OF ADVAN CED FIB ROS IS ALRE ADY AT D IAG NOS IS
Observational studies have demonstrated that approximately 25% of patients with CHD have significant fibrosis and 30%-50% have liver cirrhosis at diagnosis, with subsequent high risks for liver complications. 29 This high prevalence of cirrhosis already at baseline was also consistently found in several cohorts, regardless of geographical region or level of care. 14,19,30 In a large nationwide French cohort of 1112 patients with HDV co-infection with median age of 36.5 years and the majority from

Lay summary
• Chronic hepatitis D (CHD) is considered to be the most severe form of viral hepatitis, with ≥30%-50% having already established liver cirrhosis at diagnosis.
• Assessing the natural history of CHD is though challenging, because of limited published data and high heterogeneity.
• The landscape of CHD has shifted in several lowendemic countries in recent years, because of increased migration.
• Recent estimates of liver-related morbidity and mortality imply that CHD has less aggressive course in a substantial proportion of infected individuals.
• More knowledge about the natural course of CHD is needed, especially considering the newly introduced or upcoming new antivirals.
Sub-Saharan Africa, 312 (28%) had evidence of cirrhosis at baseline. 11 In a cohort from the United States comprising of 1191 patients with CHB and 42 patients co-infected with HDV, 73% of co-infected presented with liver cirrhosis at diagnosis. 31 Analysis of 1576 patients enrolled from centres in Europe, Asia, North-and South America, revealed that 48% of patients with positive anti-HDV had liver cirrhosis including 13% with decompensated cirrhosis at study enrolment. In a Swedish cohort of patients with CHD, cirrhosis was detected in 28% of patients within the first 6 months of healthcare encounter. 14 A study from Germany showed that 46% of patients had advanced disease stage at the time of diagnosis. 32 High prevalence of cirrhosis was also noted in cohorts of patients with CHD from Spain, England, USA and Turkey. [33][34][35][36] This emphasizes the urgent need for better screening and treatment strategies for both HBV and HDV infections around the world, to avoid establishment of cirrhosis in affected individuals.

| E ARLY S TUD IE S ON CHD: AN AG G RE SS IVE COUR S E IN MA JORIT Y OF INFEC TED PER SONS
Several longitudinal cohort studies in the 1980s following the discovery of HDV have reported an aggressive course of HDV infection with majority of patients progressing rapidly to advanced liver disease. 37,38 In a sub-analysis of 71 patients with chronic active HDV, 41% progressed to liver cirrhosis within 3.7 years. 37 Severe necroinflammatory changes were observed in liver biopsies from 23 patients with acute HDV, with 60% progressing to cirrhosis in a mean of 29 months. 38 These early studies constituted of very small sample size of patients, and other risk groups than those included in more recently published studies. 39 Patients in these earlier studies could include patients needing poly-blood transfusions with multiple morbidities or PWID constituting a majority of patients with possibly different co-morbidities and/or more pathogenic strains of HDV circulating among this risk group. 28,40,41 Most of the studies used anti-HDV positivity as diagnosis for chronic HDV infection, which cannot distinguish between an existing and a past infection. 37 In an analysis of 18 patients with anti-HDV positivity, with mean age of 26 years and 72% men, progression to cirrhosis was shown for 70% (7/10) within follow-up time of only 2 years. 42 An aggressive course was also described in another study with 15% of 22 patients progressing to cirrhosis within 2 years of diagnosis, with 36% (n = 8) of patients being PWID. 43 In one cohort study containing 137 patients with HDV co-infection from Southern Italy, development of liver cirrhosis was seen in 41% of patients after a short follow-up time of mean 3.7 years, and 31% of patients with liver cirrhosis died. 37 Another analysis of a series of liver specimens from 23 patients with HDV super-infection showed rapid progression to severe liver disease stage within 2 years. 38 In an analysis with predominant Northern Italian patients, 77% of patients co-infected with HDV experienced a stable course, liver cirrhosis developed in 41% of 75 patients followed for (2-6) years, while 14% deteriorated with no difference in progression between both groups. 37 The severe course of HDV was demonstrated in geographic regions outside Europe as well, in cohort studies from late 1980s to early 1990s. In a study from Taiwan, 80% of CHB super-infected with HDV showed chronic active hepatitis, with 9.4% annual incidence of cirrhosis. 44 In a prospective cohort from Venezuela, 42% progressed to endstage diseases or death within 10 years. 45 Among 74 persons with chronic HDV infection (CHD), the annual mortality was 6.9 per 100 person-year compared to 1.42 in HBV mono-infected, with 89% of deaths being liver-related in the HDV group. 46 In a study from Jordan, CHB populations with different disease stages were analysed, with the result of HDV infection being more frequent among patients with chronic hepatitis (23%) or HCC (67%) than in asymptomatic CHB (2%). 47

| COHORT S TUD IE S IN 1990 S AND ONWARDS
Further longitudinal studies of 1990s to early 2000s showed that 15%-50% of patients with CHD developed liver cirrhosis, while 10%-15% developed HCC within 10 years after HDV diagnosis. 48,49 Other studies suggested a 4% annual risk of progression to cirrhosis. 50 In a prospective analysis of 200 patients with CHB compensated cirrhosis, HDV infection was associated with increased risks of liver decompensation, HCC development and mortality by a relative risk of 2.2, 3.2 and 2.0, respectively. [18][19][20] In a cohort of predominantly Caucasian patients with CHB and cirrhosis, 20% of patients experienced liver decompensation events whereas HCC developed in 9% after a median follow-up time of 3.5 years. 51 In another study of 159 patients, also predominantly Caucasians with a mean age of 34 years, the cumulative transplantation-free survival at 5 and 10 years was 81%, 58% for patients with asymptomatic cirrhosis, respectively. 52 The corresponding figures were 49% and 40%, respectively, for patients with advanced cirrhosis. 52 Of note, 35% of the cohort were treated with IFNα leading to achieved virological response in 21%, which had most likely affected this outcome. 52 Of note, the data of HDV RNA positivity among anti-HDV positive patients were lacking in some studies, and the sensitivity for detection of HDV viremia in laboratory techniques has varied between assays and over time, which could have impacted some results. 35   and Sub-Saharan Africa, are relying on cross-sectional reports with no possibility to analyse the course of the disease over time. [57][58][59] Further studies to analyse the natural course in high-endemic countries are therefore needed.

| FAC TOR S A SSO CIATED WITH S E VERE DISE A SE COURSE
Studies assessing the predictors for increased risks for liver-related outcomes have shown different parameters such as, e.g. liver cirrhosis and ethnicity or genotype, with some being still debated ( Figure 1). 49,66 Predictors were found to be Asian origin, higher

| HDV AND HBV G ENOT YPE S OR E THNI CIT Y
Eight HDV genotypes have been identified to date, exhibiting high degree of genetic heterogeneity. 68 Studies on the effect of HDV genotypes on the clinical outcomes are still scarce and whether certain genotypes are associated with worsened outcomes is debated. 11,15 The most frequent HDV genotype is genotype 1 prevailing worldwide especially in North America and Western Europe. 68 HDV-genotypes 2 and 4 were isolated in East Asia and Russia, genotype 3 in the Amazon region, genotypes 5-8 were isolated in Africa. 69,70 Because of migration, genotypes 5, 6, 7 and 8 have been identified in several European countries. 70 HDV genotype 1 was associated with the severer liver disease compared to genotype 2 and 5 in cohorts from Asia and Europe. 71 Genotype 3 has been associated with acute fulminant liver disease in Brazil, 72 while genotype 5 showed higher risk for progression to cirrhosis. 15 Still knowledge is scarce on the contribution of genotypes 5-8 identified in Sub-Saharan Africa to the significant burden of HBV-related HCC in this region. 73 Despite the heterogeneity of the data, some studies have suggested that HBV genotypes can play a role in the outcome of HDV co-infection. 71 Studies reported that HBV genotype C might lead to a progressive liver disease in patients with CHD. 71 Co-existence of HBV genotype F and HDV genotype 3 was associated with a severer course compared to HBV genotype B and HDV genotype 2. 74 Patients with CHD from Central, Eastern Europe and South America were likely to have disturbed liver function in cross-sectional analyses, but whether this finding is related to viral rather than other confounders is unknown. 72,75 It is difficult to distinguish the impact of genotype from the potential role of ethnicity on the natural course of HDV infection since certain genotypes are more prevalent in some regions. Eastern European ethnicity has been associated with faster progression, while African origin has been associated with less incidence of liver cirrhosis. 15

| ROLE OF HDV RNA VIREMIA AND VIR AL LOAD
It is important to identify patients with detectable HDV RNA among anti-HDV-positive patients to assess the prognosis and to identify those in need of antiviral treatment, but studies have shown substantially poor rate of HDV RNA testing. 7 It is estimated that two-thirds of patients with HDV have replicating HDV RNA at the time of diagnosis with ~30% of patients with viremia having liver cirrhosis. 14,19 Studies have consistently reported that HDV RNA replication is associated with higher propensity towards liver-related complications. 11,14,15,76 There has been though lack of standardization of confirmatory molecular diagnostic techniques historically with concerns about comparability of HDV RNA detection assays, 77 with risk of influencing the results of some studies. 77  In our analysis of 233 patients with CHD and detectable viremia, HDV RNA quantity was similar regardless of cirrhosis, with predictability of worse outcome mostly attributed to detection of HDV RNA rather than a certain cutoff value. 14 HDV RNA levels were associated with progression to cirrhosis in an Italian study of 105 patients with CHD without cirrhosis (adjusted odds ratio, aOR = 1.60, 95% CI 1.20-2.12, p = .007), and with development of HCC (multivariable analysis: aOR = 1.88, 95% CI 1. 11-3.19). 30 Also in the same study, a cut-off of 600 000 HDV RNA copies/ml to predict transition to cirrhosis was suggested. 30 In our sub-analysis of 164 patients with HDV RNA replication and absence of cirrhosis at the start of follow-up, 96% and 78% remained free of any liver-related events or liver cirrhosis, respectively, during a mean of 6.3 years of follow-up. 14 This is in line with other studies indicating that underlying cirrhosis is a stronger factor than HDV RNA replication to forecast worse outcome. 30 Patients with detectable HDV RNA without liver cirrhosis were still more likely to develop cirrhosis and decompensation in a Spanish cohort of 151 patients compared to their peers with resolved infection. 33 Whether HDV RNA viral load, i.e. levels in blood, is a prognostic factor for the disease progression has not been fully studied. The lack of standardized HDV RNA quantification assays adds to the complexity to explore this issue. There is though some indication that high HDV viral load can be associated with a worsened outcome.
In an Italian study, a cut-off of value of HDV RNA level ≥ 600 000 copies/ml was found to be predictive of disease progression. 30 In this study, HDV RNA level was though not found to be a determinant of liver disease progression in patients with established liver cirrhosis. In our Swedish cohort study, a specific cut-off could not be determined to predict liver disease progression (data not published). 14 There are however some data showing that HDV RNA level decrease after interferon treatment is associated with a clinical benefit, even if not reaching HDV RNA negativity at the end of the treatment. 19,78,79 HDV RNA decrease by ≥2 log10 decline together with ALT normalization have therefore been suggested by the Hepatitis Delta International Network (HDIN) as surrogate markers for clinical outcomes in treated patients, when evaluating the efficacy of the HDV drugs in pipeline. 80 These surrogate markers are also recommended to be used by U.S. Food and Drug Administration (FDA) and European Medical Agency in clinical trials (EMA) in HDV drug clinical trials. 81,82 Whether the data of interferon treated can be extrapolated to treatment-naïve patients is not known, considering the immune-modulating properties of interferon. 83

| HBV AND HDV PAT TERN OF REPLI C ATI ON
Patients with CHD are mostly HBeAg negative and HBV DNA replication is usually attenuated by HDV. 52 HDV is usually the predominant virus in 70%-75% of co-infected patients, although repeated longitudinal measurements pointed that viral dominance could fluctuate along the course of CHD. 84,85 HDV viral dominance was associated with delay to IFN response compared to patients with HBV dominance or non-dominance suggesting that IFN therapy duration can be influenced by the pattern of viral dominance. 86 Some studies have reported that at advanced CHD stage, a decline in HDV RNA levels occurs in contrast to an increase in HBV DNA levels. 87 It is however difficult to draw a firm conclusion about which virus is more dominant in later disease stages, as earlier studies included analyses with less sensitive assays for viral detection, and now usually advanced CHB carriers with HBV and HDV-coinfection are treated with NUCs to suppress HBV DNA replication. In a study of 120 patients from Spain, Italy, Portugal and England, the rate of detection of HDV RNA declined in the stages of liver cirrhosis and hepatocellular carcinoma compared to active hepatitis phase, but HBV DNA detection rate was similar. 88 An early Taiwanese study of 185 patients compared the detection rates of HBV DNA and HDV RNA replication across HDV disease stages cross-sectionally, suggested that in late stages replication of either viruses can occur with a trend of HBV DNA dominance. 87 The frequency of HBV DNA detection was 66%, 70%, 80% in chronic hepatitis, cirrhosis stage and HCC, respectively. 87 This contrasts with 74%, 74% and 63% detection rate of HDV RNA in chronic hepatitis, cirrhosis stage and HCC, respectively. 87 In another study with 80 patients with CHD, HBsAg levels correlated with HDV RNA levels, however both did not correlate with liver disease stage or grade. 89 In a study with a long-term follow-up of 159 patients, HDV RNA replication was more likely to be present in severe hepatitis (76%) and early cirrhosis (71%) rather than in decompensated cirrhosis (33%). 52 This may indicate that HDV replication drives progression to liver cirrhosis, and once cirrhosis is established HDV RNA replication attenuates and HBV replication resumes. 52 In our Swedish cohort study, HDV RNA levels were numerically higher than HBV DNA levels regardless of presence of cirrhosis, when analysing the data cross-sectionally at the study enrolment (data not published). 14

| CO -INFEC TI ON WITH HIV AND H C V
It is estimated that 15% of patients with HIV/HBV are coinfected with HDV with possible higher estimates in regions with high HBV and HIV endemicity. 90 HDV tends to be the dominant virus in triple infection with HBV and HCV. 91 The interaction of HDV with HIV leads to possible immune dysregulation that accelerates the progression to cirrhosis and development of HCC. 92 In two recent meta-analyses triple infection with HIV/HBV/HDV increased the risk of HCC occurrence by 5-7 times than HIV/HBV infection. 60,61 In a retrospective analysis from Taiwan of 104 patients with HIV with median age of 35  years, patients with HDV co-infection experienced higher frequency of liver cirrhosis, decompensation events and death. 93 In a cohort of predominantly Northern European patients with HIV/ HBV, HDV infection was associated with double the risk of death, especially higher risks for HCC and liver-related events. 64

| ROLE OF HOS T IMMUNE RE S P ON S E OR G ENE TI C S
There are studies showing disturbances of the immune system in patients with CHD, with premature ageing of immune cells and impaired CD8+ T cells. 94,95 Data are absent though about if a specific host immune response is associated with rapid disease progression.
There are however data suggesting that viral escape together with CD8+ T cell failure play a role in establishment of chronic infection. 95 In one cross-sectional study, the frequency of both activated liver NK cells and degranulating blood NK cells inversely correlated with HDV RNA viremia and liver fibrosis, respectively. 94 Decreased frequency of NK and MAIT cells were seen in the livers of CHD compared to healthy liver. 94 Premature ageing of immune cells was noted, in which senescence marker of CD 57 was upregulated in CD8+ T cells in HDV infection compared to HBV mono-infection and healthy controls. 96 Data about host genetics and association with disease progression are also lacking. Further studies are therefore needed to explore the role of the host immune response or genetics to the disease progression.

| EFFEC T OF ANTIVIR AL THER APY ON THE D IS E A S E PROG RE SS I ON
Pegylated (PEG)-interferon (IFN)α or the recently approved entry inhibitor Bulevirtide are recommended treatments for patients with CHD. Historically, IFNα therapy has been associated with poor virological response of 20%-30%, with high relapse rates with 48 weeks of therapy. 97,98 In a Turkish cohort of 99 patients with CHD, a maintained virological response at 5 years was reached in 50% with prolonged courses of IFN therapy (median 24 months). 79 The virological response to IFN therapy was achieved regardless of age, sex, quantitative HDV RNA and HBV DNA at the start of therapy. 79 Cumulative evidence demonstrated that IFN therapy decreases the risks of liver decompensation events, and possibly HCC in patients with CHD. 19 Responders to IFN therapy were less likely to experience liver-related events compared to non-responders regardless of presence of cirrhosis. 14,19 Patients with CHD who cleared HBV DNA and HDV RNA upon treatment with IFN had stable liver disease, with possibility of liver fibrosis regression. 78 Disparity in access to antiviral therapy in different countries and populations can result in differences in the overall prognosis, when following cohorts of patients with CHD. A recent analysis from 22 centres in Italy, reported that 82% of native patients with CHD have received prior therapy compared to 38% among immigrants, despite the younger age of the latter group. 18 Considering the recently introduced or new antivirals against HDV in pipeline, 99 it is crucial that the full picture of the natural course with the spectrum is known in the general population of individuals with CHD. This is to ensure accurate assessment of how these new antivirals intervene on the natural history of the disease on short and longterms, and to be able to possibly aid in decision-making of access to treatments in healthcare settings with limited economical resources. This is especially important when considering the anticipated high costs and potential side-effects of these new drugs, weighing the costs and benefits for new treatments of this severe disease.

| S EROLOG IC AL B IOMARKER S CORREL ATING WITH CHD D IS E A S E AC TI V IT Y
In 120 patients with CHD, levels of anti-HDV IgM correlated with increased transaminases and histological inflammation. 100 In the same study, a second independent cohort of 78 patients was used to study the occurrence of liver-related outcomes. These outcomes occurred higher in anti-HDV IgM positive patients, compared to anti-HDV IgM negative patients (39% vs. 9%; p = .05). Hence, detection of anti-HDV IgM might not only be useful in differentiating acute co-infection from super-infection but may also be used to identify a subgroup of patients with very mild clinical course. 100,101 In another study, anti-HDV IgM was correlated with histological inflammation and response to IFN therapy. 102 In one study, HBsAg levels correlated with HDV RNA levels, but not with liver disease stage in an analysis of 91 patients with CHD -genotype 1. 89 Only the parameter of gamma glutamyl transferase (GGT) was associated with liver cirrhosis. 89 Another study reported that HBeAg status does not affect the clinical course of CHD in an analysis of 534 patients with anti-HDV positive. 103 In a crosssectional analysis enrolling 233 patients with CHD, serum cholinesterase levels were correlated with parameters of advanced liver disease in patients with liver cirrhosis. 104 Table 2 presents the current scores developed for patients with CHD to predict the development of clinical events or identify advanced fibrosi. A widely used score is the baseline anticipation score (BEA score) which is constituted of six baseline parameters as age, sex, bilirubin level, international normalized ratio (INR) level, thrombocytes count and being from an Eastern Mediterranean origin. 105   Delta fibrosis score (DFS) was developed in 100 patients with HDV RNA viremia, using an equation including serum cholinesterase, serum gamma glutamyl transferase and age to predict liver fibrosis. 107 DFS had an AUROC of 0.87, a positive predictive value (PPV) of 93% and a sensitivity of 85%. 107 The main limitations of the published scores were small sample sizes, lacking HDV RNA detection, which is an important predictor of the natural course of HDV, as a parameter in the algorithm, and absence of external validation in larger cohorts with different CHD populations. However, for determination of which patients should be prioritized for antiviral therapy, especially if high cost of antiviral therapies and in a setting with limited resources, an algorithm as BEA score could though be very useful.

| CON CLUS IONS
Chronic hepatitis D remains a challenge for care among viral hepatides, with higher risks for fast progression to liver-related complications than HBV mono-infected or HCV-infected individuals. The natural history with wide spectrum of CHD may not be fully known, because of limited published data and high heterogeneity. Recent estimates of liver-related morbidity and mortality have however implied that CHD has less aggressive course in a substantial proportion of infected individuals than previously thought with further insight into the spectrum of the disease, but larger cohort studies are needed to confirm these findings. Such studies are highly anticipated also considering the newly introduced or in pipeline antivirals, to be able to fully assess the short-and long-term effects of these drugs on the natural course of the disease.

ACK N OWLED G EM ENT
We would like to thank Leila Relander for her help in the figure.

FU N D I N G I N FO R M ATI O N
This study has been funded by grants from Region Stockholm (ALF grant).

CO N FLI C T O F I NTE R E S T
Habiba Kamal has received research grant from Gilead. Soo Aleman has received honoraria for lectures and educational events from Gilead, AbbVie, MSD, Biogen, and research grants from Gilead and AbbVie.

E TH I C A L A PPROVA L S TATE M E NT
Not applicable.

PE R M I SS I O N TO R E PRO D U CE M ATE R I A L FRO M OTH E R S O U RCE S
Not applicable.