Hepatitis B Vaccination Increases Multiple Sclerosis Risk (original) (raw)

http://www.medscape.com/viewarticle/489057

Hepatitis B Vaccination Increases Multiple Sclerosis Risk

Laurie Barclay, MD

Sept. 13, 2004

Hepatitis B virus (HBV) vaccination increases the risk of developing
multiple sclerosis (MS), according to the results of a nested, case-control
study published in the Sept. 14 issue of Neurology.

"Our analyses include 163 cases of MS and 1,604 controls," lead author
Miguel A. Hern�n, MD, DrPH, from the Harvard School of Public Health in
Boston, Massachusetts, says in a news release. "We estimated that
immunization against hepatitis B was associated with a threefold increase
in the incidence of MS within the three years following vaccination."

More than 140 countries have followed the World Health Organization
recommendation to integrate the HBV vaccine into national immunization
programs. In 1996, the French government suspended routine immunization of
pre-adolescents in schools because about 200 cases of MS and other central
nervous system demyelinating disorders were reported after HBV vaccination.
However, studies evaluating the potential link between HBV vaccination and
increased risk of MS have been inconclusive.

Using the General Practice Research Database (GPRD) in the U.K., the
investigators identified patients with a first MS diagnosis recorded
between January 1993 and December 2000. Cases were patients with the
diagnosis of MS confirmed by examination of medical records, and with at
least three years of continuous recording in the GPRD before their date of
first symptoms (index date). Each of the 163 cases was matched with up to
10 randomly selected controls (n = 1,604), matched for age, sex, practice,
and date of joining the practice.

Compared with patients who were not vaccinated, the odds ratio of
developing MS in patients who received the HBV vaccination within three
years before the index date was 3.1 (95% confidence interval,
1.5 - 6.3). Tetanus and influenza vaccinations were not associated with
increased risk of MS.

"Our study cannot distinguish whether the vaccine hastens the onset of MS
in persons destined to develop the disease years later, or whether it
causes new cases of MS in susceptible individuals," Dr. Hern�n says. "It is
also important to stress that 93% of the MS cases in our study had not been
vaccinated."

Study limitations include potential confounding. The investigators note
that any considerations regarding the administration of HBV vaccine must
reflect the large benefits derived from the prevention of a common and
potentially lethal infection, and they recommend additional research to
elucidate the association between HBV vaccine and MS.

The National Multiple Sclerosis Society supported this study.

In an accompanying editorial, Robert T. Naismith, MD, and Anne H. Cross,
MD, from Washington University in Saint Louis, Missouri, describe the
"bumpy" road to universal immunization against HBV and praise the sound
methodology of this study. However, they note that the selection process
for cases, which was thought to be necessary to properly perform the study,
might have led to some unrecognized bias.

"No data are presented to change the current Level C recommendation from
the Immunization Panel of the MS Council for Clinical Practice Guidelines,
which was approved by the AAN," the editorialists write.

"This committee determined that the HBV [vaccine] had not been found to be
detrimental in those with established MS, and recommended that patients
follow the Centers for Disease Control immunization guidelines.... The
indisputable benefit that the HBV [vaccine] provides against an infection
that kills 5,000 per year in the United States must be weighed against any
uncommon risks that remain in dispute."

Neurology. 2004;63:772-773, 838-842

Reviewed by Gary D. Vogin, MD

------------------------------

Hernan MA, Jick SS, Olek MJ, Jick H. Recombinant hepatitis B vaccine and
the risk of multiple sclerosis: A prospective study. Neurology 2004 Sep
14;63(5):838-842.

Department of Epidemiology (Dr. Hernan), Harvard School of Public Health,
Boston; Boston Collaborative Drug Surveillance Program (Drs.
Susan S. Jick and Hershel Jick), Boston University, Lexington, MA; and
Department of Neurology (Dr. Olek), College of Medicine, University of
California, Irvine. Email: miguel_hernan@post.harvard.edu

Abstract: Background: A potential link between the recombinant hepatitis B
vaccine and an increased risk of multiple sclerosis (MS) has been evaluated
in several studies, but some of them have substantial methodologic
limitations.

Methods: The authors conducted a nested case-control study within the
General Practice Research Database (GPRD) in the United Kingdom. The
authors identified patients who had a first MS diagnosis recorded in the
GPRD between January 1993 and December 2000. Cases were patients with a
diagnosis of MS confirmed through examination of medical records, and with
at least 3 years of continuous recording in the GPRD before their date of
first symptoms (index date). Up to 10 controls per case were randomly
selected, matched on age, sex, practice, and date of joining the practice.
Information on receipt of immunizations was obtained from the computer
records.

Results: The analyses include 163 cases of MS and 1,604 controls. The OR of
MS for vaccination within 3 years before the index date compared to no
vaccination was 3.1 (95% CI 1.5, 6.3). No increased risk of MS was
associated with tetanus and influenza vaccinations.

Conclusions: These findings are consistent with the hypothesis that
immunization with the recombinant hepatitis B vaccine is associated with an
increased risk of MS, and challenge the idea that the relation between
hepatitis B vaccination and risk of MS is well understood.

Discussion.

We estimated that immunization against hepatitis B was associated with a
threefold increase in the incidence of MS within the 3 years following
vaccination. Other common immunizations were not associated with an
increased risk of MS.

Our study cannot distinguish whether the hepatitis B vaccine hastens the
onset of MS in persons destined to develop the disease years later, or
whether it causes new cases of MS in susceptible individuals. However, the
similar age at first symptoms between vaccinated and unvaccinated cases
does not support the former explanation.

Elucidating the reasons for the association between hepatitis B vaccine and
MS may eventually contribute to a better understanding of the etiology of
MS, but any decision concerning hepatitis B vaccination needs to take into
account the large benefits derived from the prevention of a common and
potentially lethal infection. It is also important to stress that 93% of
the MS cases in our study had not been vaccinated.

The use of a nested case-control study minimized the bias due to
inappropriate selection of controls, and the use of prospectively recorded
computerized vaccination records prevented recall bias. Other types of
differential misclassification of vaccination history are also unlikely
because exposure information was gathered prospectively before the first
symptoms of the disease. A certain degree of non-differential
(random) misclassification of vaccination history is possible (e.g., a
small proportion of persons might have been vaccinated without their GP's
knowledge), but its practical consequence would be an attenuation of the
association between vaccination and MS. As always in observational
research, confounding is a theoretical explanation for the association. Our
analyses are therefore matched on and adjusted for various known or
suspected risk factors for MS.

The use of computerized medical records is an efficient strategy to
identify individuals with a diagnosis of MS. However, our approach was to
combine the use of computerized medical records to identify individuals
with a diagnosis of MS with the retrieval and review of paper medical
records to determine their date of first symptoms, since we found that the
computer records did not provide sufficient information to determine the
subjects' clinical history, including date of first MS symptoms. An
accurate determination of the date of first symptoms is important because,
as we observed, the probability of hepatitis B vaccination decreases after
clinical onset of MS. Thus, the use of dates that are posterior to the true
date of first symptoms may cause a downward bias of the OR for acute
exposures such as vaccinations.

Several case-control studies have evaluated the association between
hepatitis B vaccination and risk of MS or demyelination (table 4). Two
French studies found about a 1.5-fold increase in the risk of a first
episode of CNS demyelination during the 2 months following hepatitis B
vaccination.5,6 In both studies, the date of first symptoms was ascertained
by review of medical records, and dates of vaccination were obtained
retrospectively by questionnaire and phone interview of the participants.
Concurrently with the first reporting of results from the French studies, a
preliminary assessment of the association between hepatitis B vaccination
and MS in the GPRD found a 1.6-fold increase (95% CI 0.6, 4.0) in the risk
of MS or demyelination during the 12 months following hepatitis B
vaccination.7 MS diagnoses and dates of first symptoms were ascertained by
review of computerized records only.

More recently, a case-control study in three North American health
maintenance organizations (HMOs) found a nonsignificant increase in the
risk of MS or optic neuritis 1 to 5 years after vaccination against
hepatitis B, and no increase before 1 year or after 5 years.8 The date of
first symptoms was retrieved from medical records and telephone interviews,
and vaccination histories included both vaccinations recorded in HMO
records and those reported in telephone interviews.

A case-control study nested in the Nurses' Health Studies did not find an
increased risk of MS associated with hepatitis B vaccination in
women.9 The vaccination status was obtained retrospectively and the
analysis included only women who self-reported never having been vaccinated
in a questionnaire, and those who self-reported having been vaccinated and
for whom vaccination certificates were available. This design may cause
selection bias leading to a downwardly biased OR.18-20 Perhaps more
important, the date of first symptoms of the disease was retrospectively
assessed by questionnaires sent to each case and the current treating
neurologist or internist.

Two other studies did not find an increased risk of MS after immunization
against hepatitis B. A study conducted in a database consisting of
integrated pharmacy and medical claims from six HMOs in the United States
found no difference in the 3-year risk of diagnosis of demyelinating
diseases between subjects vaccinated and non-vaccinated for hepatitis B.10
This null finding is consistent with our null finding in the GPRD when we
used date of diagnosis, rather than date of first symptoms, of MS to define
the period of risk. An ecologic study compared the number of adolescents
who developed MS before (1986 to 1992) and after (1992 to 1998) a
school-based hepatitis B vaccination program was implemented in British
Columbia, Canada.11 Nine out of 288,657 unvaccinated teenagers and 5 out of
289,651 vaccinated teenagers had first symptoms of MS, but the unvaccinated
had up to 13 years of follow- up, while the vaccinated had only up to 7
years of follow-up and therefore less opportunity to be diagnosed with MS.

The recombinant hepatitis B vaccine is a non-infectious viral vaccine
derived from hepatitis B surface antigen (HBsAg) produced in genetically
engineered yeast (Saccharomyces cerevisiae) cells. Although several viruses
(e.g., Epstein-Barr virus) have been postulated to cause MS, the hepatitis
B virus has not been prominent in the discussions of viral triggers of
MS.21 It is therefore unclear how a recombinant vaccine that contains
purified HbsAg, a portion of the hepatitis B virus, could trigger the
immunologic processes that lead to MS. The vaccine also contains an
adjuvant (aluminum hydroxyphosphate sulfate), a mercury-based preservative
(thimerosal, eliminated from recent formulations), and yeast proteins (up
to 5%), but these components have not been separately studied in relation
to the risk of MS.

-------------------------------------------------------------------
Dawn Richardson
PROVE(Parents Requesting Open Vaccine Education)
prove@vaccineinfo.net (email)
http://vaccineinfo.net/ (web site)
-------------------------------------------------------------------
PROVE provides information on vaccines, and immunization policies and
practices that affect the children and adults of Texas. Our mission is to
prevent vaccine injury and death and to promote and protect the right of
every person to make informed independent vaccination decisions for
themselves and their family.
-------------------------------------------------------------------
This information is not to be construed as medical OR legal advice.
-------------------------------------------------------------------
Subscribe to PROVE Email Updates:
http://vaccineinfo.net/subscribe.htm
Tell a Friend about PROVE:
http://vaccineinfo.net/subscribe/friends.shtml
-------------------------------------------------------------------
Removal from PROVE Email Updates:
Click here: http://vaccineinfo.net/unsubscribe.htm
You are currently subscribed as pwatson@utdallas.edu
.

-------------------------------------------------------------------