Introduction

An estimated 257 million people globally are living with chronic hepatitis B (CHB) infection, according to the World Health Organization in 2018 [1]. Treatment’s main goals in CHB are to halt disease progression and prevent disease-related complications, achieved by suppression of hepatitis B virus (HBV) DNA replication [2]. To the present date, CHB treatment is either based on nucleos(t)ide analogue (NA) or on interferon IFNα, currently pegylated (PegIFNα) [3, 4]. NAs that have been approved for HBV treatment in humans include lamivudine (LAM), adefovir dipivoxil (ADV), entecavir (ETV), telbivudine (LdT), tenofovir disoproxil fumarate (TDF) and tenofovir alafenamide (TAF), and can be classified into those associated with low barrier against HBV resistance (LAM, ADV, LdT) and those with high barrier to HBV resistance (ETV, TDF, TAF) [3,4,5]. The main advantage of treatment with a potent NA with high barrier to resistance (i.e., ETV, TDF, TAF), considered to be the first-line treatment for CHB, is its predictable high long-term antiviral efficacy leading to undetectable HBV DNA levels in the vast majority of compliant patients as well as its good safety profile [3,4,5]. Moreover, it has been shown that NAs can improve the liver fibrosis and reduce the hepatocarcinogenesis in patients with CHB [6,7,8].

A significant number of patients has been treated with NAs to date, having increased the experience with their efficacy, resistance and safety profile over the years. Despite the fact that NAs have a favourable safety profile [3, 4], undesired adverse events (AEs) may occur during the treatment of CHB infection. Given the eminent number of patients currently receiving NAs, even a small risk of any of these toxicities can be translated into a major medical issue.

The main objective of this systematic review is to analyse available information in published studies on AEs associated with Nas’ use in adults.

Methods

Eligibility criteria

The following research questions were addressed:

  • What are the most common AEs with the use of NAs in the CHB treatment?

  • Is there any difference in the incidence of AEs between the different NAs?

  • Do patients receiving TAF have fewer AEs compared to TDF?

A PICO model was constructed (participants, interventions, control and outcome):

Participants

  • Adults > 18 years old diagnosed with HBV infection.

Interventions

  • Antiviral therapy with NAs (LAM, ADV, LdT, ETV, TDF or TAF).

Control

  • We used only the data for the currently approved dose, i.e. LAM 100 or 150 mg; ADV 10 mg; LdT 600 mg; ETV 0,5 or 1,0 mg; TDF 300 mg; TAF 25 mg.

  • Studies based only on drug-combination regimens were excluded due to difficulties in evaluating cause–effect relationship.

  • Studies with both single drug arm and drug-combinations arm; only the single drug arm were included in the analysis.

Outcome measure

  • Adverse events (AEs).

Exclusion criteria

We excluded studies whose patients presented acute HBV infection, acute liver failure, decompensated cirrhosis, pregnancy, hepatitis C or D or human immunodeficiency virus (HIV) coinfection, schistosomiasis infection; patients receiving corticosteroids, chemotherapy, or immunosuppressive therapy; transplant recipients; and hemodialysis patients. Likewise, studies that did not report AEs or stated "no serious adverse events" or “no significant difference in side effects between groups” with no further AEs description were excluded.

Literature search strategy

This study was performed according to the PRISMA statement [9]. We chose the following MesH terms: chronic hepatitis B, side effects and treatment. We reviewed all articles published from 1 January 1990 up to 19 February 2018 in MEDLINE of PubMed, EMBASE, the Cochrane Library and LILACS databases and included studies published in English language. Since the NAs have a good safety profile with a small percentage of AEs, we enrolled both observational (i.e. cohort, case–control and cases series) and randomized controlled trials (RCTs) as a search strategy for maximizing AEs sensitivity. All the references identified were managed by Endnote. The flowchart in Fig. 1 shows the process of review of publications. We followed an established protocol which had been registered in PROSPERO (International prospective register of systematic reviews) [10], and the record is available on https://www.crd.york.ac.uk/prospero/ (Registration No. CRD42018086471).

Figure1
figure 1

Flowchart of study selection. MEDLINE Medical Literature Analysis and Retrieval System Online, EMBASE Excerpta Medica Database, Cochrane The Cochrane Library, LILACS Literatura Latino–Americana e do Caribe em Ciências da Saúde, AEs adverse events, TDF tenofovir disoproxil fumarate, LdT telbivudine

Full size image

Data collection and quality assessment

The following data were extracted from included studies: study design, country where the study was conducted, first author, publication year, number of participants, inclusion and exclusion criteria, drug dosing regimens and AEs reported. Two reviewers independently performed data extraction (RSF and VVV) and discrepancies were discussed during a consensus meeting.

To facilitate data analysis, AEs were divided into groups, similar to those found in the VigiAccess™ database, as follows: blood and lymphatic system disorders; cardiac disorders; ear and labyrinth disorders; endocrine disorders; eye disorders; gastrointestinal disorders; general disorders; hepatobiliary disorders; infections and infestations; laboratory abnormalities; metabolism and nutrition disorders; musculoskeletal and connective tissue disorders; neoplasms; nervous system disorders; psychiatric disorders; renal and urinary disorders, reproductive system disorders; respiratory disorders; skin and subcutaneous tissue disorders.

Results

Studies

A total of 1698 articles were retrieved. Two authors conducted an initial screening and 1483 studies were excluded after reading titles and abstracts. Following the removal of duplicates, 157 full-text articles were assessed for eligibility. Thirty-seven studies were excluded for the following reasons: non-detailed adverse events (26), only patients with decompensated cirrhosis (4), patients with liver failure (1), dose of TDF not approved (1), dose of LdT not approved (1) and duplicated population studied (4). Finally, 120 articles were selected for analysis. The listing of the included reporters and their characteristics are shown in Table 1.

Table 1 Studies reported in this review
Full size table

There were 6419 patients treated with LAM, 5947 treated with ETV, 3566 treated with TDF, 3096 treated with LdT, 1178 treated with ADV and 876 treated with TAF.

Table 2 contains the AEs described in the studies, depending on the drugs used.

Table 2 Frequency of AEs reported according to the drug
Full size table

Neoplasms were documented in 39 patients, with hepatocellular carcinoma being the most frequent– 67% (n. 26: LAM-8; ETV-9; TDF-3; LdT-6) [Table 2]. None of the cases were related to NAs use.

Lamivudine (LAM)

In studies using 100 mg of LAM, a total of 5554 AEs were reported (0.87 AE/patient treated) [Fig. 2, Table 2].

Fig. 2
figure 2

Number of patients treated and absolute value of adverse events reported for each drug. LAM lamivudine, ETV entecavir, TDF tenofovir disoproxil fumarate, LdT telbivudine, ADV adefovir dipivoxil, TAF tenofovir alafenamide

Full size image

The most frequent AEs reported were gastrointestinal disorders (20.1%), infections and infestations (15.7%), general disorders (14.6%), respiratory disorders (12.5%) and nervous system disorders (12%) [Table 2].

Among gastrointestinal events, the most reported were abdominal pain or discomfort (n = 411).

The most commonly described infections were upper respiratory tract infection (n = 413).

General disorders included nonspecific symptoms, with asthenia/fatigue being the most reported (n = 672).

The most noticed neurological event was headache (n = 509). Regarding respiratory problems, viral respiratory infections were the most reported (n = 177).

Hepatic enzyme increase was the most documented laboratory abnormality (n = 473).

Although rhabdomyolysis was not described, 70 cases of elevated creatine kinase (CK) were documented, but did not lead to drug withdrawal.

Entecavir (ETV)

In the studies using ETV (0.5 or 1.0 mg), a total of 1086 AEs were reported (0.18 AE/patient treated) [Fig. 2, Table 2].

The most frequent AEs reported were infections and infestations (21.3%), laboratory abnormalities (20.1%), nervous system disorders (17.8%), gastrointestinal disorders (9.4%) and general disorders (7.1%) [Table 2].

Nasopharyngitis was the most frequent infection (n = 210). Regarding laboratory abnormalities, ALT elevation was the most reported (n = 117).

Headache corresponded to 95% of the nervous system disorders (n = 185). Among gastrointestinal disorders, diarrhea was the most common (n = 62). Of the general disorders, fatigue was the most reported (n = 71). CK elevation has been described in 49 patients.

Telbivudine (LdT)

In the studies using LdT (600 mg), a total of 2302 AEs were reported (0.74 AE/patient treated) [Fig. 2, Table 2].

The most frequent AEs reported were infections and infestations (28.2%), gastrointestinal disorders (17.6%), laboratory abnormalities (15.1%), nervous system disorders (11%) and general disorders (9.3%) [Table 2].

Nasopharyngitis was the most frequent infection (n = 295). Among gastrointestinal disorders, diarrhea was the most common (n = 114). Regarding laboratory abnormalities, CK elevation was the most reported (n = 211).

Headache corresponded to 72% of the nervous system disorders (n = 183). Of the general disorders, fatigue was the most reported (n = 71).

Adefovir dipivoxil (ADV)

A total of 1426 AEs was documented in studies with 10 mg of ADV (1.2 AE/treated patient) [Fig. 2, Table 2].

The most frequent AEs reported were infections and infestations (18.2%), gastrointestinal disorders (17.1%), nervous system disorders (15.1%), respiratory disorders (14%) and laboratory abnormalities (12.6%).

Nasopharyngitis was the most frequently described infection (n = 181). Regarding gastrointestinal disorders, the most common was abdominal pain (n = 115). Among neurological alterations, headache was the most described (n = 185). Regarding respiratory problems, flu syndrome was the most reported (n = 96).

ALT elevation was the most frequently described laboratory abnormality (n = 90).

CK elevation has been described in 22 patients.

Tenofovir disoproxil fumarate (TDF)

A total of 837 AEs were documented in studies with 300 mg of TDF (0.23 AE/treated patient) [Fig. 2, Table 2].

The most frequent AEs reported were laboratory abnormalities (18.8%), gastrointestinal disorders (15.3%), infections and infestations (13.1%), general disorders (9.8%) and nervous system disorders (7.9%).

Creatinine elevation was the most frequently described laboratory abnormality (n = 30). Regarding gastrointestinal disorders, the most common was nausea (n = 44). Nasopharyngitis was the most frequently described infection (n = 51). Fatigue was the most reported symptom in the general disorders section (n = 44).

Among neurological alterations, headache was the most described (n = 54). CK elevation has been described in 13 patients.

When evaluating renal and urinary disorders, 24 cases of urine erythrocytes and 2 cases of urine glucose were reported.

Tenofovir alafenamide (TAF)

A total of 998 AEs were reported in studies with 25 mg TAF (1.14 AE/treated patient) [Fig. 2, Table 2].

The most frequent AEs reported were gastrointestinal disorders (22.7%), laboratory abnormalities (20.2%), infections and infestations (17.5%), renal and urinary disorders (11.1%) and nervous system disorders (8.6%).

Regarding gastrointestinal disorders, the most common finding was occult blood in stool (n = 63). Among the laboratory abnormalities, the most reported were elevated ALT (n = 75) and elevated LDL cholesterol (n = 35). Nasopharyngitis was the main infections described (n = 89). Concerning the renal/urinary changes, urine erythrocytes (n = 68) and urine glucose (n = 43) were reported.

Headache was the most reported neurological disorder (n = 84). Elevation of CK has been described in 23 patients.

TDF versus TAF

Tables 3 and 4 summarized data on AEs on bone density and renal disorders, respectively, from two studies comparing TDF and TAF.

Table 3 Mean percentage decrease in hip and spine bone mineral density with TDF and TAF in studies comparing the two drugs
Full size table
Table 4 Mean increase in serum creatinine (Cr) from baseline and the median decrease in estimated glomerular filtration rate (eGFR) with TDF and TAF in studies comparing the two drugs
Full size table

With regard to bone density, TDF caused greater bone loss in both hip and spine compared to TAF [Table 3].

On the other hand, when analysing the renal AEs, there was no clinically significant difference between the two drugs regarding the elevation of serum creatinine, but there was a greater reduction in the glomerular filtration rate in patients who received TDF [Table 4].

Drug discontinuation due adverse events

In the studies analysed, the percentage of drug discontinuation with LAM, ETV, TDF, LdT, ADV and TAF were, respectively, 1% (n. 62), 0.6% (n. 33), 1.3% (n. 47), 0.4% (n. 11), 0.2% (n. 2) and 1% (n. 9) [Fig. 3].

Fig. 3
figure 3

Percentage of drug discontinuation due adverse events for each nucleos(t)ide analogue. LAM lamivudine, ETV entecavir, TDF tenofovir disoproxil fumarate, LdT telbivudine, ADV adefovir dipivoxil, TAF tenofovir alafenamide

Full size image

Discussion

The aim of CHB treatment was to control viral replication, thereby reducing the risk of complications such as liver failure, cirrhosis and hepatocellular carcinoma. CHB treatment is often based on long-term NAs use, with the following drugs being approved: LAM, ETV, LdT, ADV, TDF and TAF, of which ETV, TDF and TAF are considered to be first-line drugs, due to its potency and high genetic barrier to resistance. Identification of potential associated AEs, even if with low incidence, might be a key factor in improving adherence and outcomes. We performed a systematic literature review of studies that included LAM, ETV, LdT, ADV, TDF and TAF since 1990 and extracted all of the reported AEs from them.

One must be aware upon reading this review, there is no necessarily causation between documented AEs and pharmacological treatment [11]. As hepatitis B infection itself may lead to extrahepatic organ involvement [12], it might be difficult to determine whether extrahepatic manifestations/symptoms are treatment-related or a disease manifestation.

Data collected in this systematic review corroborate with the understanding that serious AEs are rare within the use of NAs. The most common AEs in all NAs assessed were abdominal pain/discomfort, nasopharyngitis/upper respiratory tract infections, fatigue and headache. These symptoms are not uncommon in the general population and, perhaps, these findings are related to their high prevalence among the general population rather than to drug treatment itself.

Extrahepatic AEs may result from mitochondrial toxic effect of NAs [13]. They suppress viral replication by the inhibition of the HBV polymerase enzyme. As NA structures are similar to natural nucleosides, some of these agents can also inhibit human mitochondrial polymerase-γ and cause mitochondrial toxicity [12, 14, 15]. Mitochondrial toxicity was first noticed during HIV treatment with highly active antiretroviral therapy. Because NAs lead to a minimal mitochondrial polymerase-γ inhibition, NAs associated mitochondrial toxicity cases have been rarely reported. All NAs carry a warning of mitochondrial toxicity as part of their prescribing information [12, 14, 15]. Clinical manifestations of mitochondrial toxicity include hematologic disorders, peripheral neuropathy, skeletal and cardiac myopathy, pancreatitis, hepatic failure and lactic acidosis [13, 14, 16]. Among the few AE reported in studies of this systematic review, those that could be correlated to mitochondrial toxicity are CK elevation (70 cases with LAM, 49 with ETV, 211 with LdT, 22 with ADV, 13 with TDF and 23 with TAF)—but without clinical repercussion that required drug suspension; and one case of ETV-related pancreatitis.

Tenofovir disoproxil fumarate (TDF) is a prodrug of tenofovir that was approved as a NA by the United States FDA for use in CHB infection in 2008. TDF is converted to tenofovir by hydrolysis and then phosphorylated by cellular enzymes to tenofovir diphosphate [13]. It is a highly potent inhibitor of HBV DNA replication and recommended as a first-line treatment choice in CHB by the current clinical guidelines due to the absence of drug resistance [17] [18]. Tenofovir has been shown to have a potential nephrotoxic effect in patients with HIV infection who were treated for an especially extended period. However, in clinical trials, nephrotoxicity does not seem to be a major problem in HBV monoinfection [15]. Increases in serum creatinine of > 0.5 mg/dL were reported to be detected in 1% of patient [19]. Another AE concern within TDF use is the bone mass reduction. In randomized clinical trials, a great loss of bone mineral density (BMD) had been well-described in patients with HIV infection treated with TDF [20] [21, 22]. However, tenofovir-related bone fractures were not reported in patients with HBV monoinfection [20]. The exact mechanism of bone toxicity in CHB is not clear. For example, the prevalence of BMD loss in patients receiving tenofovir was similar to those who were not exposed to tenofovir. Tenofovir was reported to be associated with a lower T score only in the hips. Additionally, in this study there was no significant correlation between duration of exposure to tenofovir and reduction in BMD at any side. Additionally, a large retrospective study in Hong Kong demonstrated that BMD reduction remains stable on a plateau from year 4 through year 7 of tenofovir treatment, for both hip and lumbar spine [23]. These data indicate that loss of bone mass is not a progressive event with the use of TDF.

A pro-drug formulation, tenofovir alafenamide (TAF), was recently launched in North America and Europe, being approved for the treatment of CHB in 2016 by the FDA (Food and Drug Administration). The pharmacokinetics of TAF leads to a 6.5-times higher intracellular concentration of the phosphorylated moiety tenofovir diphosphate, and 91% lower serum concentration of tenofovir, compared to TDF [24,25,26]. Given these pharmacokinetic differences, TAF dose can be far lower: a 25-mg once-daily dose of TAF is bioequivalent to TDF at 300-mg once daily, in terms of tenofovir plasma. Pharmacodynamic studies suggest that the lower tenofovir concentrations in plasma produced by TAF translate to reduced off-target drug exposure, for example, in the kidneys and bones, with implications regarding AEs [27]. TAF is, therefore, predicted to confer the same clinical efficacy as TDF, with potential improvements in its tolerability [27, 28].

In Tables 1 and 2 of this review, we report the results of two studies comparing TDF and TAF (Buti et al. and Chan et al.) [29, 30] concerning renal and bone alteration. The study by Buti et al. had a follow-up of 3 years and Chan et al. had a follow-up of 48 weeks. Both studies suggest that the bone density reduction was greater with the use of TDF, although no drug-related fractures were described. The same occurred with glomerular filtration rate, also with a greater reduction in the groups that received TDF. With these data, we raised two main questions: (1) what is the exact clinical repercussion of these findings? (2) Will such changes remain stable or continue to progress over the years?

Interestingly, renal/urinary changes were the 4th most reported group of AEs among patients on TAF, while the 6th for TDF and the 8th and 12th for ETV and LAM, respectively. Regarding reports for TAF in this group of AEs, there were 43 cases of glycosuria (versus 2 cases with TDF) and 68 cases of urine erythrocytes (versus 24 with TDF). At this time, we do not know the clinical relevance of these findings and whether they may represent any indication of renal tubular damage. Also, the number of patients treated with TAF is markedly lower than the number of patients who received the other NAs. Yet, TAF displays the highest proportion of AEs per patient treated among NAs.

These data fortify the idea that perhaps the greater safety of TAF in relation to TDF may have been overestimated, as already mentioned in the Hill et al. meta-analysis, which compared both drugs in HIV and CHB therapy [31].

It is known that susceptibility to AEs may vary by population. Previously, cases of Fanconi syndrome due to long-term use of adefovir have been reported using adefovir, with a higher incidence in East Asian populations [32]. However, in this review, the incidence of AEs according to ethnicity could not be differentiated. We believe that the low incidence of AEs from NAs makes this differentiation difficult.

Another important point to highlight is that the efficacy of treatments for CHB can be affected by a number of factors, including the development of AEs and poor patient compliance. In fact, a significant number of virological breakthrough may be related to medication nonadherence [33]. Hongthanakorn et al. analysed 148 patients with CHB and demonstrated that 38% of patients who experienced virological breakthrough were not confirmed to have antiviral resistance mutations, suggesting that medication nonadherence may be the cause of the virological breakthrough in these patients [34].

In this review, all drugs had a small percentage of discontinuation due to AEs, which is consistent with the literature. For example, Suzuki et al. reported that 1.3% of patients who were treated with ETV discontinued NA therapy because of AEs. Another study that evaluated LAM, LdT and ETV during the 3-year period found that patients with ETV had the best adherence [35]. This result strengthens the idea of ETV as one of the first-line agents in the treatment of CHB. Nevertheless, it should be emphasized that poor adherence, often still neglected, can have a negative effect on the treatment of chronic hepatitis B, with inadequate viral suppression, increased incidence of cirrhosis and hepatocellular carcinoma, and potential emergence of NAs-resistant [36, 37]. The situation of HBV resistance to NAs in some countries is severe and, to prevent emergence of resistance, NAs with the lowest rate of genotypic resistance should be administered (TDF, TAF or ETV) and adherence reinforced [33, 36,37,38].

Conclusion

Treatment of CHB with NAs is safe, with a low incidence of adverse events. The most common AEs with all drugs are abdominal pain/discomfort, nasopharyngitis/upper respiratory tract infections, fatigue and headache. TDF demonstrated a greater reduction in the glomerular filtration rate and bone density of the lumbar spine and hips when compared to TAF. Currently, the number of patients treated with TAF still is too small to consolidate that TAF is really safer than TDF.