Professional Documents
Culture Documents
Background: The Centers for Disease Control and Prevention recommend immunizing susceptible
high-risk groups, such as hemodialysis patients, against hepatitis B virus. However, hemodialysis
patients may not develop seroprotective antibodies despite receiving high doses of the vaccine. Recent
reports indicate that combined vaccination against hepatitis B and hepatitis A viruses may improve the
immunogenicity of hepatitis B vaccine in healthy individuals, but the effectiveness of this strategy in
hemodialysis patients is unknown.
Study Design: Prospective randomized controlled trial.
Setting & Participants: Hepatitis B virus–seronegative hemodialysis patients with undetectable
antibody levels at baseline.
Intervention: Intramuscular administration of Twinrix (inactivated hepatitis A virus [720 ELISA units]
and purified hepatitis B virus surface antigen [20 g]; GlaxoSmithKline) and Engerix-B (purified hepatitis
B virus surface antigen [20 g]) at 0, 1, and 6 months plus Engerix-B, 40 g, at month 2 (intervention
arm) or Engerix-B, 40 g, at 0, 1, 2, and 6 months (control arm). Both groups received a total dose of 160
g of hepatitis B antigen.
Outcomes: The primary outcome was the difference in seroprotection rates at 7 months, defined by
antibody titers ⬎10 mIU/mL. The secondary outcome was frequency of adverse events.
Measurements: Antibody response at months 3 and 7.
Results: 96 patients were enrolled, and 73 completed the investigation. At 3 months, there was no
difference in the groups’ seroprotection rates (25% vs 27%; P ⫽ 0.4). At the completion of the vaccination
series, using per-protocol analysis, 27 of 40 (68%) and 16 of 33 (49%) had antibody titers ⬎10 mIU/mL in the
treatment and control groups, respectively (P ⫽ 0.05; RR, 1.4; absolute abatement, 19%). Intention-to-treat
analysis showed 58% and 38% seroprotection rates in the treatment and control groups, respectively (P ⫽
0.02; RR, 1.5; absolute abatement, 20%). There was no difference in adverse events.
Limitations: Lack of evidence of long-term protection.
Conclusion: Vaccination of hemodialysis patients with a combined hepatitis A and hepatitis B
regimen resulted in a statistically significant and clinically important improvement in seroprotection
against hepatitis B virus compared with hepatitis B monovalent vaccine.
Am J Kidney Dis 56:713-719. © 2010 by the National Kidney Foundation, Inc.
American Journal of Kidney Diseases, Vol 56, No 4 (October), 2010: pp 713-719 713
714 Tung et al
are well established: direct percutaneous inocula- Twinrix (GlaxoSmithKline) is the only combined
tion of virus through exchange of contaminated hepatitis A and hepatitis B vaccine available (the
blood, blood products, body fluids, and hemodi- hepatitis B components are the same as used in
alysis. the Engerix-B vaccine). Both monovalent and
The current recommendation by the Centers combined hepatitis A and B vaccines have simi-
for Disease Control and Prevention is to immu- lar adverse-event profiles.
nize susceptible high-risk groups, which in- Although promising results of combined hepa-
cludes patients undergoing hemodialysis. After a titis A and B vaccine have been shown in healthy
primary series of hepatitis B vaccination, 90%- adults, the effect of combined vaccination on
95% of healthy immunocompetent adults de- anti-HBsAg antibody response in hemodialysis
velop protective anti–hepatitis B surface antigen patients is unknown. The hemodialysis program
(HBsAg) antibodies, defined as ⬎10 mIU/mL. at St. Joseph’s Healthcare in Hamilton, Ontario,
However, the overall efficacy of the vaccine in Canada, has a current enrollment of approxi-
hemodialysis patients is found to be much lower, mately 500 patients, most of whom are suscep-
with a median of 64% (range, 34%-88%) and tible to hepatitis B virus infection. We designed
86% (range, 40%-98%) developing seroprotec- this randomized controlled trial to investigate the
tive antibody even after receiving higher doses difference in efficacy in developing protective
of the monovalent vaccine with 3- or 4-dose antibody response and adverse reactions be-
regimens, respectively.8 tween the combination hepatitis A and B vaccine
As a result, numerous approaches to improve versus monovalent hepatitis B vaccine alone in
immunization have been attempted, including varia- hemodialysis patients attending the hemodialysis
tions in dose and frequency,9,10 use of novel adju- clinic.
vants,11,12 and coadministration of immunomodula- The primary objective is to determine whether
tors, such as interleukin 2,13 interferons,14,15 vaccinating hemodialysis patients with com-
levamisole,16 glycophosphopeptical (AM3; mar- bined hepatitis A and B vaccine resulted in a
keted as Inmunoferon),17 and thymopentin,8 all statistically significant difference in anti-HBsAg
with varying results. Intradermal administration antibody response compared with monovalent
also has been studied, and a recent study using hepatitis B vaccine. (Clinically important anti-
smaller more frequent doses of intradermal vaccina- body response is defined as the development of
tion has proved to be successful.18 at least a 10-mIU/mL concentration of anti-
To date, there are limited data about hepatitis HBsAg antibodies.) The secondary objective is
A vaccine immunogenicity in hemodialysis pa- to determine the frequency of adverse events
tients, although findings have indicated that vac- associated with the vaccine administration.
cination is feasible, well tolerated, and as effec-
tive as in healthy individuals, with seroconversion METHODS
rates of up to 100% in hemodialysis patients with
Setting and Participants
intramuscular vaccination.19
Recent reports suggest that combined vaccina- This was a prospective randomized controlled trial in
outpatient hemodialysis patients. Patients were identified
tion of hepatitis B and hepatitis A may improve through the Infection Prevention and Control Department at
immunogenicity to hepatitis B in healthy indi- St. Joseph’s Healthcare. There was a program-wide screen-
viduals. In 1 study that compared the geometric ing for hepatitis B virus infection susceptibility and immu-
mean of anti-HBsAg titers at month 6, patients nity at the time of this study. Hepatitis B virus–susceptible
receiving the combined vaccine showed a statis- patients were given an option of participating in the study.
Hemodialysis patients presenting to the St. Joseph’s
tically significant higher response than with Healthcare Centre hemodialysis outpatient department were
monovalent vaccines.20 Other studies also reflect eligible provided they met the following inclusion and
the same trend at varying points in the vaccina- exclusion criteria. Inclusion criteria were receiving hemodi-
tion series.21-24 alysis treatment, 18 years or older, undetectable antibody to
There currently are 2 monovalent recombinant HBsAg, and able and willing to give informed consent.
Exclusion criteria were the presence of HBsAg and antibody
hepatitis B vaccines on the market in Canada: to hepatitis B core antigen, treatment with intravenous
Recombivax HB (Merck Frosst, www.merckfrosst.ca) immune globulin within the last 6 months, hypersensitivity
and Engerix-B (GlaxoSmithKline; www.gsk.com).25 to components of either vaccine, and contraindication to
Hepatitis A and B Versus Hepatitis B for Hepatitis B Seroprotection in Hemodialysis Patients 715
intramuscular injections. Patients were not excluded based tered. The laboratory personnel performing tests for anti-
on pre-existing antibodies to hepatitis A. body levels and the independent statistician assessing clini-
Consents were obtained and patients were randomly as- cal outcomes were blinded.
signed starting in February 2005. Vaccination began in Adverse events were recorded after each dose. Patients
March 2005, and the vaccination schedules were completed were interviewed to obtain ratings for pain and swelling,
by November 2005. Ethics approval was obtained from the which were evaluated using a visual analogue scale. Tempera-
St. Joseph’s Healthcare Research Ethics Board. Consent was tures were measured 1 hour after vaccination and on days 2
obtained initially by the nephrology research nurse, then and 4. Fever was defined as temperature ⬎37.5°C. Adverse
later in the study by a physician and study coordinator. events also were documented in the patient chart by the
hemodialysis nurses. When there was a language barrier and
Study Design no translators were available, the chart was consulted.
Patients were given 4 doses at 0, 1, 2, and 6 months
according to the vaccine schedule and were randomly as- Statistical Analysis
signed to receive one of the following regimens (Fig 1):
In planning this study, calculations were based on a 50%
Engerix-B, 20 g (1 mL), and Twinrix (720 ELISA units of
seroconversion rate with attrition of about 20% due to age
inactivated hepatitis A virus and 20 g of recombinant
and other considerations in the population base. A 30%
HBsAg protein) at 0, 1, and 6 months plus Engerix-B, 40 g,
difference in seroconversion rates between treatment groups
at month 2 or Engerix-B, 40 g (2 mL), at 0, 1, 2, and 6
was considered effective. This determined a sample size of
months. Both groups received a total dose of 160 g of
48 per group, yielding 80% power and 5% effect size.
hepatitis B vaccine. The route of administration was intramus-
Descriptive statistics are expressed in terms of mean ⫾
cular injection.
standard deviation for continuous variables and percentages
The primary outcome was the difference in proportion of
for categorical variables. Standard statistical tests for com-
patients achieving seroprotection against hepatitis B virus
paring mean values, unpaired t test, and unpaired proportion
between the 2 arms, with seroprotection defined as an
anti-HBsAg antibody titer ⬎10 mIU/mL at month 7. The tests were used to assess treatment differences.
secondary end point was the frequency of adverse events
associated with vaccine administration. RESULTS
Consents were obtained by a research nurse, and partici-
pants were randomly assigned to receive either the combina- Patient Characteristics
tion vaccination (ie, Twinrix and Engerix-B) or monovalent
hepatitis B vaccination (ie, Engerix-B). Blood samples were
The flow diagram (Fig 2) outlines the events
drawn at baseline and months 3 and 7 to determine antibod- of this study. In particular, 454 patients were
ies to HBsAg. screened for eligibility. Three hundred fifty-eight
For practical reasons, neither patients nor clinical study patients were excluded based on the following:
personnel were blinded with respect to the vaccine adminis- 226 did not meet inclusion criteria, and others
did not wish to participate, already received a
dose of hepatitis B vaccine, or needed time to
discuss with family member/own nephrologist.
A total of 96 hemodialysis patients were en-
rolled. Twelve patients died (including 1 who
died before the first dose), 3 withdrew from the
study, 6 were lost to follow-up, and 2 were
excluded for protocol deviation. Of 96 patients
originally enrolled, 73 completed the study, for
which per-protocol analysis was performed. For
robustness, we also performed an intention-to-
treat (ITT) analysis. As mentioned, several pa-
tients had incomplete data, with some having no
seroprotection measurements performed, and oth-
ers having only a month-3 seroprotection mea-
surement recorded. In the first case (no measure-
ments), we assumed that the patient did not reach
seroconversion at the end of month 3 and month
7. For patients obtaining seroprotection measure-
ment at the end of month 3 only, we assumed that
Figure 1. Vaccination and blood work regimen. value for the month-7 measurement as well.
716 Tung et al
Treatment Control P
Pain ratings
Dose 1 0.65 ⫾ 1.40 (n⫽34) 0.31 ⫾ 0.82 (n⫽34) 0.2
Dose 2 1.00 ⫾ 2.53 (n⫽38) 0.47 ⫾ 1.34 (n⫽34) 0.3
Dose 3 0.35 ⫾ 1.53 (n⫽39) 0.39 ⫾ 1.84 (n⫽33) 0.9
Swelling ratings
Dose 1 0.13 ⫾ 0.65 (n⫽33) 0.04 ⫾ 0.18 (n⫽35) 0.5
Dose 2 0.13 ⫾ 0.84 (n⫽39) 0.00 ⫾ 0.00 (n⫽35) 0.4
Dose 3 0.00 ⫾ 0.00 (n⫽41) 0.29 ⫾ 1.69 (n⫽35) 0.3
Note: Treatment consists of administration of Twinrix and Engerix-B; control is administration of Engerix-B only. Ratings
are from a 12-point (cm) scale.
718 Tung et al
our population, the seroprotection rate in our duration of a patient’s seroprotection or determine
control group also was lower compared with the need for future booster shots or revaccination.
previous studies of hemodialysis patients, in Concurrent to the study was an initiative for
which median seroprotection rates of 64%-86% program-wide hepatitis B vaccination. Initially, there
were reported.8 There are a number of factors was a designated registered nurse to administer all
that decrease response to hepatitis B vaccination, hepatitis vaccines. Midway through the study, there
including sex, age, HLA type, and nutritional was a change in policy for bedside nurses to admin-
status.28 Of these, older age and male sex may ister vaccines and during the period when nurses
have had a significant impact in decreasing sero- were trained, some doses in the vaccination sched-
conversion rates in this study.28 We have not had ule were delayed. It is unclear whether this would
the opportunity to examine other possible factors have an impact on results.
that led to low seroprotection rates in our study. Laboratory personnel were blinded. Because
At the beginning of the study, there were 454 the primary outcome is objective, not blinding
patients in St. Joseph’s hemodialysis program, of the patients and clinical study personnel would
whom 131 were immune to hepatitis B virus by not be expected to bias the outcome.
vaccination or past infection. There may be a Vaccination against hepatitis A virus in dialy-
potential benefit in providing this regimen in sis patients is not routinely performed because it
place of the monovalent vaccine, especially for is neither associated with hemodialysis therapy
patients who do not respond to revaccination nor transmitted through parenteral mechanisms.30
after failure with a primary series. There also is no documented evidence for ad-
Identifying a vaccine regimen that is most effec- verse events associated with vaccination of pa-
tive in preventing hepatitis B virus infection in tients previously immune to hepatitis A virus,
patients undergoing hemodialysis will be important although these patients may experience a postvac-
in reducing the risk of morbidity and mortality cination increase in the geometric mean titer of
secondary to this infection. Furthermore, this will anti–hepatitis A virus antibody. We therefore did
have a positive impact on health care costs by not assess pre-existing antibodies for hepatitis A
reducing additional laboratory investigations and virus at the beginning of the study.19
treatment and shorten the number of inpatient hos- Since the start of this study, there have been
pital days. Nevertheless, the benefit of adding Twin- new strategies to improve immunization re-
rix to the hepatitis B immunization regimen must sponse. Although not yet approved in Canada
be weighed against the added costs of the drug. and the United States, HBV-AS04 (Fendrix;
Using average wholesale prices, the combined vac- GlaxoSmithKline) has been widely used. This
cination regimen costs US $488.90, whereas the new hepatitis B vaccine uses an adjuvant system
monovalent vaccine regimen costs US $292.64.29 consisting of an aluminum salt and MPL (3-O-
However, at our institution, a patient who has not desacyl-4=-monophosphoryl lipid A) that has been
achieved seroprotection at month 7 will receive a shown to elicit an earlier immunogenic response,
second vaccination series; thus, successful vaccina- higher antibody titers, and higher persistence of
tion with the combined regimen may result in drug seroprotection at 42 months compared with the
cost savings compared with 2 consecutive monova- standard hepatitis B vaccine (Engerix), which
lent vaccination series (which would cost $585.28). uses aluminum hydroxide.11,12 Because HBV-
Limitations in this study include its small sample AS04 may become standard therapy, future stud-
size. Recruitment was limited by the short duration ies comparing combination vaccination with this
of the study. In addition, at the time of the study, a new formulation should be considered.
program-wide hepatitis B immunization was initi- In conclusion, there was a statistically significant
ated in hemodialysis patients at the hospital. Be- difference in seroprotection between the 2 groups
cause of the delay in initiating the study, many at the completion of the vaccination series. Vaccina-
originally susceptible patients were already in the tion of hemodialysis patients with a combined
process of receiving the vaccine within this pro- hepatitis A and hepatitis B regimen may be more
gram before the study had begun. The short study effective than hepatitis B monovalent vaccine in
period also did not allow for evaluation of the providing seroprotection against hepatitis B virus.
Hepatitis A and B Versus Hepatitis B for Hepatitis B Seroprotection in Hemodialysis Patients 719