Scientists in Germany call for a reassessment of the HPV vaccination
and an end to misleading information

 

Translated from German: 23.12. 2008

German version: 25.11. 2008

Scientists in Germany call for a reassessment of the HPV vaccination
and an end to misleading information

As of autumn 2006, adolescent girls and women in Germany can be vaccinated against human
papilloma virus (HPV). Ever since then there have been intense discussions about possible
side effects and the costs of the vaccines, as well as the partly misleading promotional
information which is given out to the public. The question of how effective the vaccines really
are is hardly ever asked. Yet this crucial issue of efficacy – i.e., to what degree the vaccine
actually lowers the rate of new cervical cancer cases – has not been sufficiently evaluated, and
is the object of misleading information.

STIKO made its recommendations before the relevant studies were published
In March 2007 the “Ständige Impfkommission” (STIKO) [Standing Vaccination Committee]
of the Robert Koch-Institute recommended the HPV vaccine for all girls aged 12-17 years to
decrease the burden of disease due to cervical cancer. However, at the time the
recommendation was made, the results of the decisive studies had not yet been published.
It was not until May 2007 that the most important studies on the Gardasil® vaccine, FUTURE
I and FUTURE II, were published in The New England Journal of Medicine (NEJM). The
central message of an editorial published in this journal more than one year later was: “The
bad news is that the overall effect of the vaccines on cervical cancer remains unknown” (Haug
2008)1. The most important study on the second vaccine, Cervarix®, was published in June
2007. Cervarix® has not yet been approved in the USA.

What do studies and other documents say about the efficacy of the HPV vaccines?
Cervical cancer is closely associated with HPV infection. Of the approximately 100 known
strains of HPV, at least 13 can trigger cervical cancer. Strains 16 and 18, which are the target
of these two vaccines, are assumed to be responsible for 70% of all cervical cancer cases.
However, rather than assessing the effect of the vaccine against cervical cancer, the studies
examined the incidence of high-grade cervical lesions (a potential precursor of cervical
cancer) in 15-26 year-old women.

The vaccine did achieve a 98%2 decrease in the precancerous stages associated with HPV 16
or 18 in women who had not yet been infected with these two strains. This gave rise to much
optimism, and it was widely proclaimed that the 70% of cervical cancer cases associated with
HPV strains 16 and 18 could be almost completely prevented (thus achieving a decline in all
cervical cancer cases of almost 70%). This assumption, however, has so far not been
confirmed by studies.

In analyses which included all women enrolled, FUTURE I found a decline of 7.8%3 in the
incidence of all high-grade cervical lesions (number extracted from EMEA [European
Medicines Agency] 2008), and FUTURE II a decrease of 17%.4 These evaluations have not
yet been considered by STIKO. Regarding the second vaccine, Cervarix®, STIKO based its
recommendations exclusively on data on the prevention of persistent infections. Data
regarding the efficacy of Cervarix® against precancerous stages or cancer were not yet
available.

The efficacy of Gardasil®, which has been described as “modest”, was partially explained by
the fact that some of the women examined had already been infected with HPV 16 or 18. It is
well known that the vaccine does not work once women have been infected. Hence it was
Contact: Dr. Ansgar Gerhardus, Universität Bielefeld, ansgar.gerhardus@uni-bielefeld.de
recommended that women should be vaccinated before they could be infected with HPV, i.e.,
before they become sexually active. STIKO, in its recommendation, set this age at 12-17
years. Data on the efficacy against early stages of cervical cancer, however, are only available
for females aged 15-17 years, not for 12-14 year-olds.

To make up for this lack of data, the FUTURE studies conducted statistical analyses including
only girls and women who tested negative for HPV 16 or 18 at the beginning of the study.
This group was expected to approximate 12-year old girls. However, it remains unclear how
Gardasil® affects the total number of high-grade cervical lesions in this group. Here, the only
available data comes from a two-year follow-up of the FUTURE studies published by the
FDA in 2006 which found an efficacy of 16.9%.5 The FUTURE II study gives an efficacy
estimate of 27%.6 EMEA provides estimates of 37.9%7 for 2006 and 46.1%8 for 2008 for
varying populations. However, in order to derive the figure of 46.1%, about half of the
women enrolled were excluded post hoc.

When asked for additional data, Sanofi-Pasteur MSD Germany replied: “Unpublished figures
and tables are only available to the colleagues who were immediately involved in evaluating
the results, that is to say, at the headquarters in the USA. We do not have this data, and we
will not get it, either.”

The STIKO recommendations for the HPV vaccine must be reassessed now
The STIKO recommendations made in March 2007 were not based on explicit data on
efficacy. Instead, STIKO mentions that “lifelong immunity” was 92.5%9, apparently based on
its own extrapolations. No explanation was given for how this number was arrived at, nor was
– or is – there any data on “lifelong” immunity. No study indicated an efficacy of this
magnitude.

The STIKO recommendations must be reassessed immediately. STIKO should consider more
recent study results and ask manufacturers for the missing data, which must then be included
in a new evaluation. This evaluation should indicate clearly and precisely what efficacy
STIKO expects of the vaccine, and what assumptions and data these expectations rest on.
Adolescent girls and women must be informed adequately The results of the studies clearly
contradict many overly optimistic pronouncements.

Adolescent girls and women are entitled to be adequately informed. We strongly object to
stirring up fear regarding the risk of cervical cancer and feelings of guilt by disseminating
incorrrect information. We demand that gaps in the data be discussed openly. Assertions that
a vaccine reduces the risk of cervical cancer by 70% or even 98% should simply not be made
at this point in time. Instead, data should be used which is supported by sound research and
which gives all those involved an opportunity to make an informed choice.

Prof. Martina Dören Dr. Ansgar Gerhardus Prof. Ferdinand M. Gerlach Prof. Claudia Hornberg
Charité, Berlin University of Bielefeld University of Frankfurt University of Bielefeld
Prof. Michael M. Kochen Prof. Petra Kolip Prof. Wolf-Dieter Ludwig
University of Göttingen University of Bremen Charité, Berlin
Prof. Ingrid Mühlhauser Prof. Oliver Razum Prof. Rolf Rosenbrock
University of Hamburg University of Bielefeld WZB, Berlin
Corinna Schach Prof. Norbert Schmacke Prof. Jürgen Windeler
University of Bremen University of Bremen MDS, Essen
References with exact positions of quotations and numbers
1 Haug, CJ (2008) Human Papillomavirus Vaccination — Reasons for Caution. N Engl J Med
2008;359:861-62.

http://content.nejm.org/cgi/reprint/359/8/861.pdf
Quotation from p. 861, first paragraph 2 Number: 98%
The FUTURE II Study Group. Quadrivalent vaccine against human papillomavirus to prevent high-grade
cervical lesions. N Engl J Med 2007;356:1915-1927.
http://content.nejm.org/cgi/reprint/356/19/1915.pdf?ijkey=24cc07867ea2faad48e9e69c3463c8905ab010bf
Abstract and page 1920. Table 3. Last Column
3 Number: 7.8% (for FUTURE I study = Protocol 013)
EMEA (2008): Gardasil: European Public Assessment Report. Scientific Discussion (May 2008):
http://www.emea.europa.eu/humandocs/PDFs/EPAR/gardasil/EMEA-H-703-II-13-AR.pdf
Page 11. Table 6. 6th column, 4th row from below.
4 Number: 17%

The FUTURE II Study Group. Quadrivalent vaccine against human papillomavirus to prevent high-grade
cervical lesions. N Engl J Med 2007;356:1915-1927.
http://content.nejm.org/cgi/reprint/356/19/1915.pdf?ijkey=24cc07867ea2faad48e9e69c3463c8905ab010bf
Abstract and page 1921. Table 3. Last Column
5 Number: 16.9%

FDA (2006): Vaccines and Related Biological Products Advisory Committee. (VRBPAC). Background
Document, May 2006

http://www.fda.gov/ohrms/dockets/ac/06/briefing/2006-4222B3.pdf
Page 17. Table 25. 10th column
6 Number: 27%

The FUTURE II Study Group. Quadrivalent vaccine against human papillomavirus to prevent high-grade
cervical lesions. N Engl J Med 2007;356:1915-1927.
http://content.nejm.org/cgi/reprint/356/19/1915.pdf?ijkey=24cc07867ea2faad48e9e69c3463c8905ab010bf
Page 1922. 2nd paragraph, 6th line
7 Number: 37.9%

EMEA (2006): Gardasil: European Public Assessment Report. Scientific Discussion (October 2006):
http://www.emea.europa.eu/humandocs/PDFs/EPAR/gardasil/070306en6.pdf
Page 28, 2nd paragraf following the heading “Population benefit integrated summary of efficacy”. Number
refers to R-MITT-2-Population
8 Number: 46.1%

EMEA (2008): Gardasil: European Public Assessment Report. Scientific Discussion (May 2008):
http://www.emea.europa.eu/humandocs/PDFs/EPAR/gardasil/EMEA-H-703-II-13-AR.pdf
Page 22. Table 15. R-MITT-2-population, 6th column.
9 Number: 92.5%

Ständige Impfkommission (STIKO) am Robert Koch-Institut (2007): Impfung gegen humane
Papillomaviren (HPV) für Mädchen von 12 bis 17 Jahren – Empfehlung und Begründung.
Epidemiologisches Bulletin 12/2007: 97-103.
http://www.rki.de/cln_091/nn_195848/DE/Content/Infekt/EpidBull/Archiv/2007/Ausschnitte/HPV__STIK
O__12__07,templateId=raw,property=publicationFile.pdf/HPV_STIKO_12_07.pdf
Page 101. 2nd paragraph following the heading “Mögliche Auswirkungen einer Impfung gegen HPV