Randomized, placebo-controlled trial of tenofovir disoproxil fumarate in adolescents with chronic hepatitis B§


  • Potential conflict of interest: K. M. is a consultant to Gilead, Roche, and Merck and holds stock ownership in Merck; S. R. is an employee of Gilead and holds stock ownership; A. C. is an employee of Gilead and holds stock ownership; P. D. is an employee of Gilead and holds stock ownership; J. M. is an employee of Gilead and holds stock ownership; P. P. is an employee of Gilead and holds stock ownership.

  • In addition to the authors, the following investigators participated in the study: H. Zhelev, M. Bosheva (Bulgaria); F. Gottrand, A. Lachaux (France); J. Kuydowicz, D. Lebensztejn, M. Marczynska, W. Slużewski (Poland); A. Constantinescu, D. Farcau (Romania); P. Jara Vega, C. Ribes-Koninckx (Spain); F. Ozgenc (Turkey); P. Harmatz, G. Subbarao (United States).

  • §

    See Editorial on Page 2016


Tenofovir disoproxil fumarate (DF) is highly effective for the suppression of hepatitis B virus (HBV) in chronically infected adults. This study evaluated the safety and efficacy of tenofovir DF in adolescents with chronic hepatitis B (CHB). In this double-blind, placebo-controlled trial, adolescents 12 to <18 years of age with CHB were randomized to tenofovir DF 300 mg (n = 52) or placebo (n = 54) once daily for 72 weeks. The primary endpoint was virologic response (HBV DNA <400 copies/mL) at week 72. One hundred six patients were enrolled; 101 patients completed 72 weeks of treatment. At baseline, 91% of patients were hepatitis B e antigen–positive and 85% had prior exposure to HBV therapy. A virologic response was observed in 89% (46/52) of patients who received tenofovir DF and 0% (0/54) of patients who received placebo (P < 0.001). Treatment response was not affected by prior HBV treatment. Furthermore, no resistance to tenofovir DF developed through week 72. Among patients with an alanine aminotransferase (ALT) level greater than the upper limit of normal at baseline, normalization of ALT occurred in 74% of patients receiving tenofovir DF and 31% of patients receiving placebo (P < 0.001). The rate of grade 3/4 adverse events was higher among patients treated with placebo (24%) than patients treated with tenofovir DF (10%). No patients met the safety endpoint of a 6% decrease in spine bone mineral density at week 72. Conclusion: Tenofovir DF therapy in HBV-infected adolescents was well tolerated and highly effective at suppressing HBV DNA and normalizing ALT values in both treatment-naïve adolescents and those with prior exposure to HBV therapy. (HEPATOLOGY 2012;56:2018–2026)

Despite the success of recent vaccination efforts, chronic hepatitis B (CHB) remains a serious global health care problem and is a major cause of serious liver disease.1 It is estimated that approximately 350 million people live with CHB infection and approximately 600,000 die each year due to the acute or chronic consequences of hepatitis B.1 Mathematical modeling suggests that over 80% of these deaths are from infections contracted during childhood.2 This is most likely because approximately 90% of those infected as infants and 30%-50% of those infected from 1 to 4 years of age develop chronic infection.1 These chronically infected children are at high risk of developing severe liver disease and dying from hepatitis B virus (HBV)-related sequelae. Approximately 25% of individuals who become chronically infected in childhood later develop cirrhosis or cancer of the liver.1

Studies in adults with CHB have shown that the risk of developing liver complications is correlated with serum HBV DNA levels, which can be mitigated by effective viral suppression.3, 4 It is not yet known whether effective control of viral load during childhood may also affect disease progression, but the risk of life-threatening liver disease among those infected during childhood makes effective long-term management of their disease a high priority. Unfortunately, effective management of children and adolescents with CHB is challenging, and there is still debate about when treatment should begin as well as the best antiviral therapy to use in this patient population.5, 6 Only a few therapies have been approved for the treatment of HBV infection in pediatric patients (interferon alpha-2b, lamivudine, and adefovir) and the use of these drugs is often limited by adverse effects, low potency, or the development of treatment-resistant mutations.5, 6 The development of treatment-resistant viral mutations is of particular concern in young patients because of the potential need for extended duration of therapy and the long-term consequences.6

Tenofovir disoproxil fumarate (DF) is an oral prodrug of tenofovir, an acyclic nucleoside phosphonate (nucleotide) analogue of adenosine 5′-monophosphate, with an excellent safety profile and potent anti-HBV efficacy in adults.7, 8 A recent meta-analysis indicated that the antiviral efficacy of tenofovir DF is effective in reducing HBV DNA levels and normalizing alanine aminotransferase (ALT) levels as well as promoting hepatitis B e antigen (HBeAg) seroconversion and hepatitis B surface antigen (HBsAg) loss in adults with CHB.7 In adults, tenofovir DF is also associated with reversal of cirrhosis.9 Finally, during long-term studies of up to 5 years duration, no treatment-resistant mutations emerged.9 The present study evaluated the efficacy, safety, and tolerability of tenofovir DF 300 mg once daily compared with placebo in adolescents aged 12 to <18 years with CHB infection.


ALT, alanine aminotransferase; BMD, bone mineral density; CHB, chronic hepatitis B; DF, disoproxil fumarate; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; LLOQ, lower limit of quantification; PCR, polymerase chain reaction; pol/RT, polymerase/reverse-transcriptase; ULN, upper limit of normal.

Patients and Methods

Study Design.

This was a randomized, double-blind, placebo-controlled, phase 3, 72-week clinical trial followed by a 120-week open-label extension; only the 72-week double-blind phase is reported here. The study was conducted in compliance with all regulatory obligations and the institutional review board and informed consent regulations at each investigational site (ClinicalTrials.gov NCT00734162).

Patient Population.

Adolescents aged 12 to <18 years with chronic HBV infection, defined as documented positive serum HBsAg for at least 6 months prior to enrollment, were included. They could be either positive or negative for HBeAg. Key inclusion criteria at screening included HBV DNA ≥105 copies/mL (measured by polymerase chain reaction [PCR; COBAS TaqMan HBV test]) and either ALT ≥2 times the upper limit of normal (ULN) or any history of ALT ≥2 times the ULN within the past 24 months. The ULN was defined as 43 U/L for males and 34 U/L for females. Patients also had to weigh at least 35 kg and be able to swallow oral tablets. Patients must have discontinued any oral anti-HBV nucleoside/nucleotide therapy ≥16 weeks prior to screening and any interferon therapy ≥6 months prior to screening. Investigational sites in Poland required patients to have had a history of treatment for HBV or a contraindication for treatment with existing drugs. Exclusion criteria included previous therapy with tenofovir DF; serological evidence of coinfection with human immunodeficiency virus, hepatitis C virus, or hepatitis D virus; history of significant bone disease, decompensated liver disease, or renal disease; or evidence of hepatocellular carcinoma.


Randomization was centralized and stratified by age (12-14 years or 15-17 years) and geographical region (North America or Europe). Patients were randomized in a 1:1 ratio to receive oral tenofovir DF 300 mg or a matching placebo once daily for 72 weeks. Randomization was accomplished via an interactive voice or web response system, using an allocation sequence generated with SAS software by an independent party. Patients were also required to take a daily multivitamin containing 100% of the recommended daily allowance for vitamin D.

All investigators, patients, and clinical research personnel remained blinded to treatment and HBV efficacy outcomes. Unblinding was possible if there was concern about the patient's welfare or if any patient had a sustained ALT elevation of grade 4 for ≥16 weeks or an ALT flare. An ALT flare was defined as either an ALT measurement >2 times the baseline level and >10 times the ULN (with or without associated symptoms) or an ALT elevation of one grade or to twice a previous value that was associated with abnormalities in other laboratory parameters suggestive of worsening hepatic function. If breaking the blind for an ALT flare revealed that the patient was in the placebo group, the patient could be offered open-label treatment with tenofovir DF.

Outcome Measures.

Patients were evaluated at baseline, weeks 4 and 8, and every 8 weeks thereafter through week 72.

The primary efficacy measure was the percentage of patients with HBV DNA <400 copies/mL (virologic response) at week 72. Secondary efficacy outcomes included the percentage of patients with HBV DNA <169 copies/mL (below the lower limit of quantification [LLOQ]), the percentage of patients with normal or normalized ALT levels, HBV DNA and ALT levels over time, HBeAg loss and seroconversion, HBsAg loss and seroconversion, the percentage of patients with both HBV DNA below 400 copies/mL and normalized ALT, and any evidence of HBV polymerase/reverse-transcriptase (pol/RT) resistance mutations.

A central laboratory (Covance Central Laboratory Services S.A., Geneva, Switzerland, and Covance Central Laboratory Services, Inc., Indianapolis, IN) evaluated all laboratory samples. HBV DNA was quantified using the Roche COBAS TaqMan HBV test (Roche Diagnostics, Indianapolis, IN), which provides fully automated real-time PCR HBV viral load quantitation in serum and plasma. Analysis of ALT was performed using a Roche Modular Analyzer, which was calibrated daily.

The pol/RT domain of the HBV polymerase region (amino acids 1-344) was sequenced in patients with HBV DNA ≥400 copies/mL at week 72, patients who discontinued from the study early with HBV DNA ≥400 copies/mL after 24 weeks of treatment, and patients who experienced virologic breakthrough. Genotypic analysis was conducted by DDL Diagnostic Laboratories (Rijswijk, the Netherlands). Briefly, DNA was isolated from 200 mL of serum using the Roche MagNA Pure instrument, and the HBV pol/RT domain was amplified via PCR and nested PCR using the Expand high-fidelity PCR kit (Roche Molecular Systems). Di-deoxy sequencing of the amplified product was conducted using the ABI Big Dye terminator cycle sequencing kit employing a selection of forward and reverse primers, and analysis of the raw sequence data used ABI Seqscape software.

Virologic breakthrough was defined as HBV DNA measurements of ≥400 copies/mL (after an earlier value <400 copies/mL) or a 10-fold increase in HBV DNA levels over the patient's lowest value. When conserved site changes were identified and/or when patients experienced virologic breakthrough, the HBV pol/RT was isolated from patients' serum for phenotypic sensitivity testing to inhibition by tenofovir DF.10 If the conserved site change of interest occurred as a mixture with wild-type virus, then a clone containing the appropriate amino acid substitution was tested. Average values for 50% of effective concentration obtained for the post-baseline sample were compared with those obtained for the patient's baseline isolate to determine the fold-change to tenofovir DF.

Safety assessments included patient signs and symptoms as well as radiographic and laboratory findings. The primary safety endpoint was the cumulative incidence of at least a 6% decrease from baseline in lumbar spine bone mineral density (BMD) through week 72. Any clinical manifestation of hepatic decompensation or worsening hepatic function was considered a serious adverse event; this included any case of serum ALT that was more than twice the baseline level and more than 10 times the ULN, regardless of the presence of symptoms. Adverse events were graded by investigators according to severity based on the Gilead-modified National Institute of Allergy and Infectious Disease severity scale; clinical laboratory abnormalities were graded by the central laboratory using this severity scale.

Data Analysis and Statistical Methods.

Intention-to-treat analyses included all patients who were randomized and received at least one dose of study drug.

A Cochran-Mantel-Haenszel test controlling for age strata (12 to 14 years or 15 to <18 years at randomization) was used to evaluate between-group differences in the proportion of patients achieving the primary efficacy endpoint (HBV DNA <400 copies/mL at week 72). All continuous endpoints were summarized using descriptive statistics. All categorical endpoints were summarized by number and percentage of patients who met the endpoint.

A sample size of 50 patients in each treatment group was determined to provide at least an 80% power to detect a 30% between-group difference in the proportion of patients achieving the primary efficacy endpoint (based on a two-sided Fisher's exact test with a significance level of 0.05).


Patient Population

During the 9-month enrollment period, a total of 106 patients were enrolled at 21 centers in Europe and the United States: 52 in the tenofovir DF group and 54 in the placebo group.

A total of 101 patients (51 in the tenofovir DF group and 50 in the placebo group) completed the 72-week double-blind period (Fig. 1). Only one patient in the tenofovir DF group discontinued the study prior to week 72 (due to a syncopal event considered not related to the study drug). Of the four patients in the placebo group who did not complete the double-blind period, two patients were discontinued at the investigator's discretion and entered into treatment-free follow-up, and two were discontinued due to persistently elevated ALT and were enrolled in the open-label treatment phase (per protocol).

Figure 1.

Patient disposition.

The demographic and baseline characteristics of the study population were similar between the two treatment groups (Table 1). Patients in this study were predominantly Caucasian (93%) and enrolled at investigational sites in Europe (95%). Most patients (69%) were male, the mean age was 15 years, and the mean number of years of documented HBV positivity was 10.5. The modes of acquisition of HBV infection were classified into intravenous drug use, blood product transfusion, contact with infected individual, vertical transmission, other factors, and unknown. More than one mode of acquisition was possible. Thirty-nine (37%) patients reported an unknown mode of acquisition. Vertical transmission was cited by 22 (21%) patients. Thirty-two (30%) patients reported other factors; among them, hospitalization and medical procedures were the most prominent (27 [26%] patients).

Table 1. Patient Demographics and Baseline Characteristics
 Tenofovir DF (n = 52)Placebo (n = 54)Total (N = 106)
  • *

    At or below the ULN (≥43 U/L for males, ≥34 U/L for females).

Age, years, mean ± SD (minimum, maximum)15.5 ± 1.34 (12, 17)15.3 ± 1.43 (12, 17)15.4 ± 1.38 (12, 17)
Sex, no. (%)   
 Male38 (73.1)35 (64.8)73 (68.9)
 Female14 (26.9)19 (35.2)33 (31.1)
Race, no. (%)   
 White49 (94.2)49 (90.7)98 (92.5)
 Asian1 (1.9)1 (1.9)2 (1.9)
 Black1 (1.9)01 (0.9)
 Other1 (1.9)4 (7.4)5 (4.7)
Region, no. (%)   
 Europe50 (96.2)51 (94.4)101 (95.3)
 North America2 (3.8)3 (5.6)5 (4.7)
Genotype, no. (%)   
 A35 (67.3)34 (63.0)69 (65.1)
 B1 (1.9)2 (3.7)2 (2.8)
 C1 (1.9)01 (0.9)
 D15 (28.8)18 (33.3)33 (31.1)
Baseline HBeAg-positive, no. (%)48 (92.3)48 (88.9)96 (90.6)
Baseline HBsAg-positive, no. (%)52 (100)54 (100)106 (100)
Years positive for HBV, mean10.210.810.5
Prior treatment, no. (%)43 (82.7)47 (87.0)90 (84.9)
 Interferon37 (71.2)44 (81.5)81 (76.4)
 Lamivudine31 (59.6)31 (57.4)62 (58.5)
 Adefovir5 (9.6)7 (13.0)12 (11.3)
 Entecavir02 (3.7)2 (1.9)
Years since first HBV treatment, median8.317.998.1
Baseline HBV DNA, log10 copies/mL, mean ± SD (minimum, maximum)8.01 ± 1.4 (4.91, 10.11)8.24 ± 1.4 (4.79, 10.08)8.13 ± 1.4 (4.8, 10.1)
Baseline ALT, U/L, mean ± SD (minimum, maximum)101 ± 107.5 (19, 563)101 ± 90 (16, 501)101 ± 98.5 (16, 563)
Normal ALT,* no. (%)17 (32.7)12 (22.2)29 (27.4)

All patients were positive for HBsAg at baseline, 91% were positive for HBeAg, and most patients had HBV genotype A (65%) or D (31%). Overall, 85% of patients had been treated for HBV previously, 61% had been treated with an oral medication previously, 44% and 9% had received two or three prior HBV medications, respectively, and the median time since first treatment was 8 years. The most commonly used prior treatments were interferon (76%) and lamivudine (59%).

The majority of demographic and clinical characteristics did not differ between patients who were from Poland, the country with the greatest number of enrolled patients (n = 74), compared with the other countries (n = 32). Differences were observed only in the distribution of race (all patients from Poland were white, whereas white patients comprised 75% of the population from all other countries), HBV DNA genotype, and prior treatment. Furthermore, except for the distribution of race, all characteristics were similar between the site that enrolled the largest number of patients (n = 23) and all other sites (n = 84).

Overall adherence to the study drug was measured by pill count and was summarized by treatment and age group. The mean adherence was high and similar in the tenofovir DF and placebo groups (99% and 98%, respectively) and across all age groups.


In the tenofovir DF group, the primary endpoint of HBV DNA <400 copies/mL was achieved by 89% (46/52) of patients by week 72. By comparison, no patients in the placebo group achieved this endpoint by week 72 (P < 0.001) (Fig. 2A). Among patients treated with tenofovir DF, HBV DNA <169 copies/mL (below the LLOQ) was achieved by 85% (44/52) of patients by week 72. The difference between the tenofovir DF and placebo groups in the proportion of patients achieving either of these levels of viral suppression was statistically significant (P ≤ 0.001).

Figure 2.

(A) Virologic response: percentage of patients who achieved HBV DNA levels <400 copies/mL or <169 copies/mL. (B) Mean ± SD log10 HBV DNA throughout the study. (C) Mean ALT levels throughout the study. (D) Percentage of all patients with normal ALT levels.

Kinetics of Viral Load Decline.

Mean HBV DNA at baseline was approximately 8 log10 copies/mL in both study groups (Table 1). Mean HBV DNA concentrations rapidly declined in the tenofovir DF group while remaining near baseline levels in the placebo group (Fig. 2B). As early as week 4, mean HBV DNA in the tenofovir DF group had decreased more than 3 log10 copies/mL to approximately 5 log10 copies/mL. By week 40, mean HBV DNA in the tenofovir DF group had decreased 5.6 log10 copies/mL to approximately the LLOQ (2.2 log10 copies/mL), where it remained through week 72. The same degree of viral load reduction was observed irrespective of the presence (n = 6) or absence (n = 46) of baseline lamivudine-resistant mutations.

Virologic Breakthrough.

Virologic breakthrough was defined as HBV DNA measurements of ≥400 copies/mL or a 10-fold increase in HBV DNA levels over the patient's HBV DNA nadir. At week 72, among patients treated with tenofovir DF, four patients had virologic breakthrough, and one patient never achieved an HBV DNA level of <400 copies/mL (i.e., no breakthrough). All four instances of virologic breakthrough were associated with tenofovir DF plasma levels below the limit of detection, suggesting nonadherence with tenofovir DF dosing. Consistent with this observation, sequence analysis of the HBV pol/RT and subsequent phenotypic analysis of patient isolates from week 72 samples did not identify any tenofovir DF resistance–associated mutations in the HBV pol/RT of any patients evaluated.

Alanine Aminotransferase.

Consistent with the inclusion criteria, baseline ALT was elevated in the majority (73%) of patients; the mean ALT level was 101 U/L (Table 1). Consistent with the declining HBV DNA levels, mean ALT levels quickly decreased in the tenofovir DF group; in the placebo group, they remained elevated (Fig. 2C). As early as week 16, the mean ALT level in the tenofovir DF group had declined to approximately 44 U/L. Mean ALT levels remained near or below this value through week 72. Among the patients with an ALT level greater than the ULN at baseline, the percentage of patients whose ALT normalized was 74% (26/35) in the tenofovir DF group and 31% (13/42) in the placebo group (P < 0.001).

At baseline, 33% (17/52) of patients in the tenofovir DF group and 22% (12/54) of patients in the placebo group had ALT levels within the normal range (Table 1). In the tenofovir DF group, the percentage of patients with a normal ALT level increased steadily throughout the study (Fig. 2D). In the placebo group, there was a steady but much smaller increase in the percentage of all patients with normal ALT levels. By week 72, the percentage of all patients with normal ALT levels was 77% (40/52) in the tenofovir DF group and 39% (21/54) in the placebo group (P < 0.001).


Among patients who were HBeAg-positive at baseline, 21% (10/48) of patients in the tenofovir DF group and 15% (7/48) in the placebo group experienced HBeAg loss by week 72, a difference that was not statistically significant. Only one patient in the tenofovir DF group experienced HBsAg loss (week 64) and seroconversion (week 72); one other tenofovir DF–treated patient experienced a transitory HBsAg loss at week 32 that did not persist thereafter.

Composite Endpoints.

In total, 71% of patients in the tenofovir DF group achieved HBV DNA <400 copies/mL and normal ALT level at week 72 compared with no patients in the placebo group (P < 0.001). The composite endpoint of HBV DNA <400 copies/mL, normalized ALT, and HBeAg loss was achieved at 72 weeks by 21.2% of patients in the tenofovir DF group compared with no patients in the placebo group (P < 0.05). A total of 14.6% of patients in the tenofovir DF group versus no patients in the placebo group attained DNA <400 copies/mL, normal ALT, and HBeAg loss (P < 0.05).

Subgroup Analyses.

Substantial viral suppression occurred regardless of baseline ALT, HBeAg status, prior use of interferon or oral HBV medication, genotype (A or D), or age (Table 2). As shown in Table 2, an ALT level greater than the ULN at baseline was associated with a higher rate of HBV DNA suppression; ad hoc analysis suggested no difference in the likelihood of HBV DNA suppression if elevated ALT is further divided into 1-2 times and >2 times the ULN. These analyses, however, lacked sufficient numbers for rigorous statistical comparison.

Table 2. Percentage of Patients with HBV DNA Below Thresholds of Interest Within Relevant Patient Subgroups
HBV DNABaseline ALTBaseline HBeAgInterferon and/or Oral HBV MedicationGenotypeAge
NormalALT > ULNNegativePositiveAny Prior TreatmentPrior Treatment-NaïveAD12 to 14 Years15 to <18 Years
Tenofovir DF (n = 17)Placebo (n = 12)Tenofovir DF (n = 35)Placebo (n = 42)Tenofovir DF (n = 4)Placebo (n = 6)Tenofovir DF (n = 48)Placebo (n = 48)Tenofovir DF (n = 9)Placebo (n = 7)Tenofovir DF (n = 43)Placebo (n = 47)Tenofovir DF (n = 35)Placebo (n = 34)Tenofovir DF (n = 15)Placebo (n = 18)Tenofovir DF (n = 10)Placebo (n = 13)Tenofovir DF (n = 42)Placebo (n = 41)
  • Data are presented as no. (%).

  • *

    n = 41.

  • n = 47.

  • n = 40.

Week 24                    
 <400 copies/mL10 (58.8)0 (0.0)22 (62.9)0 (0.0)4 (100.0)0 (0.0)28 (58.3)0 (0.0)4 (44.4)0 (0.0)28 (65.2)0 (0.0)23 (65.7)0 (0.0)7 (46.7)0 (0.0)6 (60.0)0 (0.0)26 (61.9)0 (0.0)
 <169 copies/mL10 (58.8)0 (0.0)15 (42.9)0 (0.0)3 (75.0)0 (0.0)22 (45.8)0 (0.0)2 (22.2)0 (0.0)23 (53.5)0 (0.0)18 (51.4)0 (0.0)5 (33.3)0 (0.0)4 (40.0)0 (0.0)21 (50.0)0 (0.0)
Week 48                    
 <400 copies/mL12 (70.6)0 (0.0)33 (94.3)0 (0.0)*4 (100.0)0 (0.0)41 (85.4)0 (0.0)7 (77.8)0 (0.0)38 (88.4)0 (0.0)31 (88.6)0 (0.0)12 (80.0)0 (0.0)9 (90.0)0 (0.0)36 (85.7)0 (0.0)
 <169 copies/mL11 (64.7)0 (0.0)31 (88.6)0 (0.0)*4 (100.0)0 (0.0)38 (79.2)0 (0.0)7 (78.8)0 (0.0)35 (81.4)0 (0.0)29 (82.9)0 (0.0)11 (73.3)0 (0.0)8 (80.0)0 (0.0)34 (81.0)0 (0.0)
Week 72                    
 <400 copies/mL12 (70.6)0 (0.0)34 (97.1)0 (0.0)4 (100.0)0 (0.0)42 (87.5)0 (0.0)8 (88.9)0 (0.0)38 (88.4)0 (0.0)31 (88.6)0 (0.0)13 (86.7)0 (0.0)9 (90.0)0 (0.0)37 (88.1)0 (0.0)
 <169 copies/mL11 (64.7)0 (0.0)33 (94.3)0 (0.0)4 (100.0)0 (0.0)40 (83.3)0 (0.0)8 (88.9)0 (0.0)36 (83.7)0 (0.0)30 (85.7)0 (0.0)12 (80.0)0 (0.0)9 (90.0)0 (0.0)35 (83.3)0 (0.0)


Adverse events occurred in 44 of 52 (85%) patients in the tenofovir DF group and 48 of 54 (89%) patients in the placebo group. The rate of grade 3 or 4 adverse events was higher among patients treated with placebo (24%) than patients treated with tenofovir DF (10%) (Table 3). This difference was driven largely by the increased incidence of hepatic flares in the placebo group. Serious adverse events that were considered by the investigator to be related to study medication occurred in four patients, one in the tenofovir DF group (hepatitis) and three in the placebo group (two with increased ALT and one with abdominal pain). No deaths occurred during the study.

Table 3. Summary of Adverse Events
Adverse EventsTenofovir DF (n = 52)Placebo (n = 54)P Value
  1. All data are presented as no. (%).

  2. Boldface P values = statistically significant.

Any events44 (84.6)48 (88.9)
Any treatment-related events8 (15.4)9 (16.7)
Grade 3 or 4 adverse events5 (10.0)13 (24.0)
Treatment-related grade 3 or 4 events1 (1.9)4 (7.4)
Serious events6 (11.5)12 (22.2)
Treatment-related serious events1 (1.9)3 (5.6)
Adverse events occurring in ≥5% of patients in either treatment group   
Pharyngitis15 (28.8)11 (20.4)0.37
Nasopharyngitis5 (9.6)12 (22.2)0.11
Increased ALT3 (5.8)12 (22.2)0.024
Acne2 (3.8)10 (18.5)0.029
Upper respiratory tract infection5 (9.6)7 (13.0)0.76
Abdominal pain3 (5.8)7 (13.0)0.32
Headache2 (3.8)8 (14.8)0.093
Hepatitis3 (5.8)7 (13.0)0.32
Rhinitis5 (9.6)3 (5.6)0.48
Tonsillitis2 (3.8)5 (9.3)0.44
Increased blood creatine phosphokinase4 (7.7)2 (3.7)0.43
Cough3 (5.8)3 (5.6)1.00
Lymphadenopathy06 (11.1)0.027
Diarrhea4 (7.7)1 (1.9)0.20
Epistaxis3 (5.8)2 (3.7)0.68
Contusion1 (1.9)3 (5.6)0.62
Dysmenorrhea04 (7.4)0.12
Pharyngolaryngeal pain04 (7.4)0.12
Pyrexia3 (5.8)1 (1.9)0.36
Toothache1 (1.9)3 (5.6)0.62
Increased aspartate aminotransferase03 (5.6)0.24
Herpes zoster03 (5.6)0.24
Nail disorder3 (5.8)00.11
Vomiting03 (5.6)0.24

All adverse events occurring in ≥5% of patients are listed in Table 3. The most common adverse events were pharyngitis, nasopharyngitis, increased ALT, acne, and upper respiratory tract infection. The only adverse events for which there was a statistically significant between-group difference (all higher in the placebo group) were increased ALT (tenofovir DF, 6%; placebo, 22%; P = 0.024), acne (tenofovir DF, 4%; placebo, 19%; P = 0.029), and lymphadenopathy (tenofovir DF, 0%; placebo, 11%; P = 0.027).

One patient in the tenofovir DF group discontinued the study due to syncope. This patient had a history of syncope, and this adverse event was not considered to be related to study medication. Two patients in the placebo group were discontinued due to sustained grade 4 ALT elevation for ≥16 weeks and were enrolled in the open-label phase of the study at week 40 (Fig. 1).

No patients in either group experienced a ≥6% decrease in lumbar spine BMD at any time during the study. Five patients, three in the tenofovir DF group and two in the placebo group, had a decrease of >4% in lumbar spine BMD. None of these patients experienced a bone fracture or other bone-related adverse event. The mean change in lumbar spine BMD z score from baseline to week 72 was −0.05 in the tenofovir DF group and 0.07 in the placebo group. Corresponding mean change in whole-body BMD z scores from baseline to week 72 were −0.15 and 0.06, respectively.

Both treatment groups experienced an overall increase in mean lumbar spine BMD. There was a greater increase in mean BMD in the placebo group than the tenofovir DF group at all visits at which BMD was measured: weeks 24 (tenofovir DF, 2%; placebo, 3%; P = 0.005), 48 (tenofovir DF, 4%; placebo, 6%; P = 0.046), and 72 (tenofovir DF, 5%; placebo, 8%; P = 0.053).

There were no observed grade 3 or 4 increases in serum creatinine or decrease in serum phosphorus, and no patient had a confirmed increase from baseline creatinine of ≥0.5 mg/dL. Eight patients, six in the tenofovir DF group and two in the placebo group, had a confirmed increase in serum creatinine of 0.3 mg/dL. All of these elevations were transient or within the normal range. The mean change in creatinine from baseline to week 72 was 0.1 mg/dL in both treatment groups.

Hepatobiliary adverse events were reported in three patients in the tenofovir DF group (all cases of hepatitis) and 10 patients in the placebo group (eight cases of hypertransaminasemia and two cases of hepatomegaly). Two patients in the tenofovir DF group and 10 patients in the placebo group had hepatic flares. Both hepatic flares in the tenofovir DF group resolved without interruption of treatment.


This study demonstrates that tenofovir DF is well tolerated and highly effective at suppressing HBV DNA in adolescents with CHB. The significant decrease in HBV DNA levels in patients treated with tenofovir DF was accompanied by a decline in ALT levels. A lower incidence of hepatic flares was observed in the tenofovir DF group compared with the placebo group, further illustrating the drug's potent ability to suppress viral replication. In addition, subgroup analyses suggested that antiviral efficacy was high regardless of baseline ALT, HBeAg status, age, or prior HBV therapy. Treatment with tenofovir DF was not associated with a statistically significant change in HBeAg serologic status during the first 72 weeks. This may have been due to the relatively short time frame of the study.6

The antiviral efficacy observed in the first 18 months in this adolescent population was consistent with what has been observed with tenofovir DF treatment in adults with CHB.7, 8 In the present study, from week 24 onward, HBV DNA levels were <400 copies/mL (virologic response) in the majority of patients treated with tenofovir DF, but in none of the placebo-treated patients.

The high rate of antiviral efficacy observed in the present study is notable given the nature of the population enrolled. The majority of patients enrolled had both high HBV DNA levels at baseline and a history of previous treatment. Also, nearly 60% of patients had been treated with lamivudine previously. A clinical study in adults showed that tenofovir DF had potent antiviral efficacy even in patients who had experienced treatment failure on lamivudine,11 and the present study suggests that tenofovir DF is similarly effective in younger patients who have failed on lamivudine.

In contrast, the use of lamivudine in children and adolescents has been greatly limited by its tendency to promote treatment-resistant viral mutations. Evidence of treatment-resistant mutations was observed in 19% of children (aged 2-17 years) treated with lamivudine for 1 year12 and in 64% of those treated for up to 3 years.6, 13

All cases of virologic breakthrough that occurred at week 72 were associated with evidence of nonadherence to tenofovir DF dosing, with no genotypic or phenotypic evidence of drug resistance. Furthermore, despite concerns that adherence to treatment among adolescents may be suboptimal,14, 15 adherence in this adolescent population was high, and consequently the response rates were similarly high. Tenofovir DF has a pharmacokinetic profile that enables simplified, once-daily dosing, which may facilitate adherence in this patient population. Furthermore, this study revealed that treatment was associated with a good safety profile, with a relative lack of adverse events.

Suppression of HBV DNA may be, however, only a limited surrogate endpoint for the development of long-term sequelae of chronic HBV infection, such as hepatocellular carcinoma. Few patients in this relatively short-term study achieved seroconversion and loss of HBeAg and HBsAg, although a positive trend was observed. Long-term follow-up is ongoing to further characterize the impact of treatment on seroconversion.

As PCR technology becomes more sensitive and more efficacious treatments are investigated, evaluations of HBV DNA levels below the present cutoff of <400 copies/mL will be of interest. At the end of the double-blind phase of this study, HBV DNA levels were <169 copies/mL (below the LLOQ) in the large majority of tenofovir DF–treated patients but in none of the placebo-treated patients.

Treatment-related adverse events were also less common in the tenofovir DF group than in the placebo group, as was the frequency of hepatic flares. No significant bone, renal, or hepatic complications were identified, although the increases in BMD were slightly lower in patients in the tenofovir DF group than in the placebo group.

Interpretation of the data concerning effects of tenofovir DF on BMD is complex due to the potential impact of underlying disease and concomitant treatments.16 In the present study, no patients experienced a ≥6% decrease in lumbar spine BMD at any time during the study. Overall, the differences in lumbar spine and whole-body BMD z scores between the tenofovir DF and placebo groups were small and suggest that the impact of tenofovir DF on bone health is unlikely to be of clinical significance, at least over the first 72 weeks of therapy. Nevertheless, because the increase in BMD in the present study was significantly less in patients treated with tenofovir DF than in those treated with placebo, bone safety issues will continue to be monitored regularly in the open-label phase of this study.

One potential limitation of the present study is that the patient population was mostly Caucasian and European and, therefore, infected with HBV genotypes A or D. It is not yet clear whether race or ethnicity, per se, have any effect on the likelihood of disease progression or response to treatment,3 although there is evidence suggesting that the HBV genotype may influence disease progression and treatment efficacy.3 We also examined the response based on the predominant genotypes, genotypes A and D. The response appeared to be similar regardless of genotype. A further limitation of this study is the lack of histologic data. A relatively mild HBV histology was expected in this population at baseline, and limited interval improvement was anticipated due to the study duration of only 72 weeks. Furthermore, because this was a placebo-controlled study, half of the patients would have been exposed to two biopsies but would not have been receiving active treatment. Taken together, the scientific benefit of obtaining biopsies was not thought to outweigh the potential patient risk. Further follow-up will occur in the 2½-year open-label phase of this study and will address the durability of response and safety of tenofovir DF.

In conclusion, tenofovir DF therapy in HBV-infected adolescents was well tolerated and highly effective at suppressing HBV DNA and normalizing ALT values in both treatment-naïve patients and those with prior exposure to oral HBV therapy. No resistance to tenofovir DF was observed through week 72, and lamivudine-associated mutations at baseline appeared to have no effect on virologic response. Tenofovir DF is, therefore, a valuable treatment option for the management of CHB in adolescents.


We thank Amy Lindsay, Ph.D., and Evelyn Albu, Ph.D., of Percolation Communications LLC for providing editorial assistance.