NOT FOR PUBLICATION
COMMISSION ON HUMAN MEDICINES
Title of paper:
Human papillomavirus (HPV) vaccine – update on safety
Type of paper:
Marketing Authorisation Holder
Sanofi Pasteur MSD
Human papillomavirus (HPV) vaccine
Redacted (section 40 personal
NOT FOR PUBLICATION
COMMISSION ON HUMAN MEDICINES
HUMAN PAPILLOMAVIRUS (HPV) VACCINE – UPDATE ON SAFETY
THE UK IMMUNISATION PROGRAMME
MHRA PHARMACOVIGILANCE STRATEGY
AND PREVIOUS SAFETY REVIEWS
CURRENT ASSESSMENT OF CERVARIX
CURRENT ASSESSMENT OF GARDASIL
RECENT UK MEDIA INTEREST IN HPV VACCINES
SPECIFIC EVENTS REPORTED
Chronic Fatigue Syndrome (CFS)
Guillain Barre Syndrome (GBS)
Acute disseminated encephalomyelitis (ADEM)
Primary ovarian failure (POF)
Complex Regional Pain Syndrome (CRPS)
Postural Orthostatic Tachycardia Syndrome (POTS)
Small Fibre Neuropathy
REVIEWS BY THE WORLD HEALTH ORGANISATION
AND US HEALTH AUTHORITIES
CHM July 2015
NOT FOR PUBLICATION
This paper provides an update on the safety of human papillomavirus (HPV) vaccine
in advance of the UK assuming EU lead on pharmacovigilance of Gardasil 9 HPV
2.1. Current HPV vaccines
Two vaccines that provide protection against HPV infection are currently authorised
in the EU; Gardasil (Sanofi Pasteur MSD) and Cervarix (GSK). Cervarix is a bivalent
vaccine, providing protection against HPV types 16 and 18. Gardasil is a
quadrivalent vaccine, providing protection against HPV types 16, 18, 6 and 11. The
Summary of Product Characteristics (SmPC) for Gardasil and Cervarix are at annex
1 and 2 respectively.
Both vaccines are authorised in many non-EU countries and it is estimated that more
than 50 million doses of Cervarix and more than 150 million doses of Gardasil have
been distributed worldwide.
Gardasil and Cervarix are ‘centrally-authorised’ in Europe, with Sweden and Belgium
the respective EU ‘Rapporteurs’.
2.2. Gardasil ‘9’ vaccine
Gardasil 9 received a positive CHMP opinion in March 2015 and a Commission
Decision on 3 July. Gardasil 9 provides protection against HPV types 16, 18, 6 and
11, as well as oncogenic types 31, 33, 45, 52 and 58.
The only differences between Gardasil 9 and Gardasil is the likely authorised dosage
schedule (3 doses, whereas Gardasil currently allows 2 doses) and the additional 5
antigens (virus-like particles; VLPs). During the EU licensing procedure for Gardasil
9, it was accepted that the safety profile was broadly the same as Gardasil and no
additional pharmacovigilance activities were imposed. However, as Gardasil and
Gardasil 9 may be initially co-marketed, there is a requirement for Sanofi Pasteur
MSD to minimise the risk of product confusion or inadvertent mixed-schedule (which
could impact immunogenicity for the additional 5 HPV types) through appropriate
As yet, Gardasil 9 is not marketed in the EU, although it was granted a licence in the
USA in December 2014. In the near future, it is possible that Gardasil 9 could
become the main HPV in use, not only in Europe but global y.
No announcement has been made on whether or not the UK will switch to using
Gardasil 9 in the near future. If it is used in the UK and given the similarity to
Gardasil, which has been in routine use in the UK since 2012, MHRA will take the
same approach to pharmacovigilance as it does for Gardasil. As with all vaccines,
this involves near real-time surveillance and a proactive approach to signal
detection, utilising CPRD data and ‘observed vs expected’ analysis of adverse
events of interest where necessary.
Sweden is CHMP Rapporteur for Gardasil 9. However, the UK is the ‘PRAC1
Rapporteur’ for Gardasil 9 meaning that, once authorised, MHRA will lead on all of
the routine EU post-authorisation pharmacovigilance and risk management activities
for Gardasil 9. Therefore, in advance of MHRA assuming the EU PRAC
Rapporteurship of Gardasil 9, advice from CHM is sought on the safety experience
with Gardasil, as well as an update on the UK safety experience with Cervarix (used
routinely in the UK until 2012).
3. THE UK IMMUNISATION PROGRAMME
HPV vaccine was introduced in the UK in September 2008 and is routinely offered to
girls aged 12-13 years (older girls were also offered the vaccine as part of an initial
‘catch-up’ campaign). Cervarix was purchased for use in the routine UK programme
from 2008 to 2012, and Gardasil replaced this (for contractual reasons) from
Gardasil is currently offered to all 12-13 year girls in the UK via school-based
programmes as a two dose schedule (0 and at least 6 months later). Cervarix and
Gardasil are available on the UK market and may still be used privately at other
1 The European Medicines Agency’s Pharmacovigilance Risk Assessment Committee (PRAC)
3.1. Expected benefits of the HPV vaccination programme in the UK
There are over 100 types of HPV, 40 of which can infect the genital tract. Of these,
13 have been associated with cervical cancer. Types 16 and 18 are linked with over
70% of cervical cancers in England. Infection with HPV is very common. Most of
these infections do not lead to cervical cancer, but in some people the infection can
persist and cause cervical abnormalities which, if left undetected and untreated, can
lead to cervical cancer. In most cases this develops over many years.
3.2. Cervical cancer
Cervical cancer is the most common cancer among women aged 15 to 34 years,
with around 2,500 cases of invasive cancer diagnosed in 2012. It is anticipated that
the HPV vaccine will eventually prevent several hundred deaths each year from
cervical cancer in the UK.
Public Health England introduced a set of surveillance activities when the
vaccination programme started. Although the expected decrease in cervical cancer
(a disease which peaks in women between 25 and 50) will be seen in some years’
time, a reduction in the circulation of HPV types 16 and 18 was expected before this.
A study3 comparing samples from sexually active young women, aged 16 to 24 in
2008, to similar samples from 2010-2012 found that high risk HPV infections (type 16
and 18) decreased from around 1 in 5 women prior to the immunisation programme,
to 1 in 15 following its introduction. The effect was most marked in the age group
with the highest vaccination coverage.
3.3. Genital warts and other cancers
HPV is also associated with other rarer types of anogenital cancers and some
cancers of the head and neck. Where HPV is linked to these cancers, types 16 and
18 (the vaccine strains) are the most commonly found. The vaccination programme
is expected to contribute to a decrease in these cancers over time.
Gardasil also protects against two low risk HPV types 6 and 11, which cause over
90% of genital warts. Genital warts are the most common viral sexually transmitted
infection in the UK, with over 71,000 new diagnoses in total in England in 2013.
Between 2008 and 2013, a significant decrease of 28% in genital warts diagnoses
2 Information taken from briefing prepared by Public Health England - Further details about HPV virus,
the disease and the vaccine, including references for the statistics and figures given above can be
found at the fol owing links;
were seen for females aged 15-19 and 17% for 15-19 year old males. Much of this
trend has been suggestive of an unexpected moderate protective effect from the
HPV 16/18 (Cervarix) vaccination against genital warts. It is anticipated that the
incidence of genital warts will continue to decline, with the introduction of Gardasil in
2012, as has been seen in other countries using this vaccine for a longer period of
time (Drolet et al. Lancet ID 2015).
3.4. Vaccine exposure in the UK
In England, approximately 3.3 mil ion girls were eligible to receive the vaccine either
routinely at 12-13 years or as part of the catch-up campaign. By the end of August
2014 approximately 2.3 million girls had received three doses of HPV vaccine since
the start of the programme (PHE, 2015). For the first six years of the programme full
course coverage for the routine cohorts averaged 86%, while for the older catch-up
cohorts it was 49%.
The following table relates to data from England only. Taking into account further
doses administered since August 2014, and all doses administered in Scotland,
Wales and Northern Ireland since 2008, (as well as any doses administered
privately) the total figure will be closer to (if not more than) 9 million doses and closer
to (if not more than) 3 million vaccinees
across the UK.
4. MHRA PHARMACOVIGILANCE STRATEGY AND PREVIOUS SAFETY
In 2008, as the UK was the first country in the world to launch a national
immunisation programme with the Cervarix brand of HPV vaccine, MHRA put in
place an ‘enhanced’, proactive pharmacovigilance strategy. The key elements of this
Daily assessment and categorisation of all suspected new ADRs
A proactive communication plan
o Issuing a letter to healthcare professionals involved in the immunisation
programme to encourage use of the Yellow Card Scheme
o Weekly Publication of ‘Suspected Adverse Reaction Analysis’ reports
which provided an ongoing assessment of all suspected adverse
events cumulatively reported via the Yellow Card Scheme
Weekly teleconference with the MAH (GSK) to discuss any new or emerging
safety issues identified from non-UK data sources
Modified disproportionality analysis methods (Empirical Bayes Geometric
Mean or EBGM) to conduct routine and modified signal detection analyses.
Proactive and near real-time ‘observed vs expected’ analysis of key ‘adverse
events of interest’. This approach involved using data from the Clinical
Practice Research Datalink (CPRD) to calculate age-specific background
incidence rates of adverse events of interest and vaccine usage estimates to
derive an ‘expected’ rate of events. On an ongoing basis, the number of
Yellow Card reports were compared against the ‘expected’ rate using
statistical methods (Maximised Sequential Probability Ratio Test; MaxSPRT)
and adjustments are made for possible levels of under-reporting. This
analysis allows possible new risks to be identified rapidly, and can give some
assurance when the ‘observed’ is well within the ‘expected’.
On 30 July 2010, at the end of the 2nd school year of the programme, MHRA
reverted to routine pharmacovigilance for Cervarix vaccine. However, ‘observed vs
expected’ analysis remained in place (and is now a standard and routine method that
MHRA has in place for all vaccines).
CHM was presented with four updates on Cervarix safety; two in 2009 (6 month and
1 year update), 2010 (two years into the routine UK programme) and in 2012 (at four
years and the end of routine UK use).
At the time of the 2012 update (paper at annex 3), over 6 million doses of Cervarix
had been administered in the UK and MHRA had received 6,213 Yel ow Card reports
(including 364 reports with unspecified brand). The overall reporting rate was around
1 Yel ow Card per 1,000 doses administered. CHM agreed that this reporting rate
was not unexpected for a newly marketed vaccine used within a new national
immunisation programme with such high vaccine exposure over a four year period.
Case reviews and/or ‘observed vs expected’ analysis were considered for reports of
Guillain Barre Syndrome (n=5), encephalitis [including 3 reported as N-methyl-D-
aspartate (NMDA) receptor encephalitis] (n=7), Bell’s palsy (n=9), convulsions
(n=97), and complex regional pain syndrome (CRPS, n=6). CHM agreed that the
available information did not support a causal association with such events given the
case details, high vaccine uptake and knowledge of background rates of such
Furthermore in June 2009 CHM specifically considered reports of chronic fatigue
syndrome (CFS) in association with Cervarix. The CHM agreed that the number of
cases of suspected CFS and other possible chronic fatigue like syndrome did not
represent a safety signal since many more cases would have been expected to
occur by chance amongst the number of girls immunised so far. Further to this,
MHRA has completed an epidemiological study which found no evidence of an
increased risk of CFS following Cervarix. The results of this study were published in
a peer-reviewed scientific journal in 2013 (Vaccine. 2013 Oct 9;31(43):4961-7). See
section 8 below.
Overall, CHM advised that no serious new risks had been identified in association
with Cervarix and the balance of risks and benefits of the vaccine remains positive.
As there was already a significant level of safety experience with Gardasil through
administration of tens of millions of doses in other countries at the time it replaced
Cervarix in the UK programme, MHRA did not put in the place the same ‘enhanced’,
pharmacovigilance strategy as for Cervarix. However, proactive pharmacovigilance
utilising near real-time ‘observed vs expected’ analysis was put in place (and
remains in place, as for all vaccines) and Gardasil safety has remained under close
evaluation by the MHRA since 2012.
5. CURRENT ASSESSMENT OF CERVARIX
5.1. UK data
Since the 2012 CHM paper and up to 30 June 2015, MHRA has received a further
419 Yellow Card reports (72 serious) in association with Cervarix (and 15
unspecified brand with date of vaccine administration before September 2012).
These are summarised in annexes 4 and 5. Given that Cervarix is no longer used in
the routine programme, this level of reporting after September 2012 will be due a
combination of delayed reporting and retrospective reporting.
These reports have remained under continual review and no new safety concerns
have been identified. Most reports related to the known side effects of vaccination
(as well as reports of fainting). For the one event with a fatal outcome, non-
Hodgkins’s lymphoma, there is no indication that this was due to vaccination.
We have received a further 11 reports of CFS and 1 report of complex regional pain
syndrome (CRPS) (this report had unspecified brand with a date of administration
prior to September 2012). CFS and CRPS are discussed below.
We have received two reports of encephalitis autoimmune. One included a diagnosis
of Postural Orthostatic Tachycardia Syndrome (POTS) and is included in annex 28.
The other was a patient report reported by GSK from the SaneVax website and with
no information on confirmation of diagnosis. It described a 13 yr old whose “life
turned into an extended nightmare” 5 days after vaccination with Cervarix. The
patient spent months in hospital and was diagnosed with anti-NMDA receptor
encephalitis. At the time of reporting the outcome of the events was unspecified.
Based on the details on the SaneVax website, this case is a likely duplicate of one of
the reports included in the 2012 review.
We have received 2 reports of narcolepsy with cataplexy. In one case a parent
reported that her daughter immediately “fell unwell with the symptoms of narcolepsy
and cataplexy”. It was reported that she was admitted to hospital on a number of
occasions, consulted a number of neurologists, and consulted specialists in
narcolepsy and cataplexy, and had undertook a sleep study to confirm diagnosis.
The second report was also a parent report with very vague details and no
confirmation of diagnosis, but with a reported onset of around 4 years after
vaccination. We have also received 3 reports of hypersomnia, but there is no
indication of narcolepsy in these reports.
The one report of ‘autoimmune disorder’ related to a report of lupus-like syndrome
occurring “less than one year after vaccination with Cervarix”. Diagnosis has not
Other specific adverse events reported in temporal association with either brand of
vaccine are discussed in section 8 below, including five reports of POTS (including
one autonomic nervous system imbalance) following Cervarix (one of the five reports
had unspecified brand with an administration date before September 2012).
5.6. Evaluation of spontaneous reports of autoimmune events
The MAH performed a detailed analysis of a very wide range of possible
autoimmune events as part of its routine pharmacovigilance activities. This is
discussed in detail in section 220.127.116.11 of annex 6.
This review compiled the cases of potential autoimmune disorders reported from
launch (17 May 2007) up to 31 August 2014 in 371 subjects who had received
vaccination with Cervarix (or an unspecified HPV vaccine). Cases were reported
most frequently from Japan (132/ 6,998,413 doses), The United Kingdom (70/
8,667,679 doses), Italy (26/2,228,149 doses) and Germany (25/668,037doses). The
distribution in number of case reports reflects the distribution of doses cumulatively
in these countries, with most of the doses that have been distributed in the United
Kingdom and Japan. Cumulatively, the most frequently reported MedDRA PTs were
SLE (35 cases), VIIth nerve paralysis/facial paresis (35 cases), Guillain-Barre
syndrome (30 cases), Juvenile arthritis (25 cases), Optic neuritis (20 cases), Immune
thrombocytopenic purpura (16 cases), ADEM (15 cases), Alopecia areata (14
cases), Erythema multiforme (12 cases), Multiple sclerosis, Demyelination and (10
cases) and Encephalitis/encephalitis autoimmune (n=16).
The analysis included ‘observed vs expected’ evaluation of optic neuritis, multiple
sclerosis, peripheral neuropathies, demyelination disorders, Bell’s palsy, ADEM/TM
(see below), Guillain Barre Syndrome (see below), amongst many more events.
None of these analyses identified a new safety signal.
Other than the risks that are currently included the SmPC for Cervarix (see annex 2)
and based on the most recent PSUR, the Rapporteur concluded that no additional
new risks have been confirmed during post-marketing use to date.
Specific adverse events reported in temporal association with either brand of vaccine
are discussed in section 8 below.
6. CURRENT ASSESSMENT OF GARDASIL
6.1. UK data
To date (24/06/2015), MHRA has received 1,483 Yellow Cards in association with
Gardasil. The MHRA has also received 151 reports since 31/07/2012 with
unspecified brand of which 61 reported the date of vaccine administration after
September 2012 (the 75 reports with unknown brand and date of administration are
included in this section). These are summarised in annex 7 and 8.
With at least 1.5 million doses administered, the overal reporting rate for Gardasil
and any brand of HPV vaccine remains around 1 Yellow Card per 1,000 doses
administered. The number of serious4 reports is 514, representing 35% of all reports.
This proportion is not too dissimilar to that for Cervarix (28% of total).
The overall reporting pattern is very similar to that of Cervarix, with the majority of
reports relating to the known side effects of vaccination (as well as reports of
fainting). The two events with a fatal outcome were due to leukaemia and a baby
born prematurely. There is no indication that these were due to vaccination. Isolated
reports of serious suspected ADRs remain under review, but this has not raised any
new safety signals.
We have received 8 reports of CFS (six of these had no brand or date of
administration), 2 of complex regional pain syndrome. CFS and CRPS are discussed
We have received 2 reports of encephalitis (one of these had no brand or date of
administration). One included a diagnosis of POTS and is included in annex 28. The
other case was identified from a magazine article in 2011 (and therefore likely
associated with Cervarix). It has been confirmed as duplicate of one of the reports
included in the 2012 review. On an unreported date post vaccination, the patient was
diagnosed with encephalitis.
We have received 2 reports of narcolepsy (one with cataplexy). One is most likely a
duplicate of one of the cases mentioned under Cervarix above. The other report was
also a parent report with very little detail or confirmation of diagnosis but within onset
reported as the same month as vaccination. The date of onset was 2011 so it is
likely that Cervarix was given. We have also received 6 reports of hypersomnia, but
there is no indication of narcolepsy in these reports.
Of the three reports of ‘autoimmune disorder’, one was concurrent with POTS
(described below), one was a parent reported with unconfirmed diagnosis of a ‘post
streptococcal disorders’, and one related to an unspecific dermatological disorder.
6 reports of ‘activities of daily living impaired’, 2 were concurrent with CFS,
one was with the above-mentioned dermatological disorder, one was associated with
an injection site reaction (leading to reduced limb mobility), one was with post-viral
gut motility disorder and one report related to muscle and joint disorder with chronic
tiredness (reported in the media).
4 The seriousness criteria for ADR reporting were determined by a working group of the Council for International Organizations
of Medical Sciences (CIOMS) and are de fined as 6 po ssible categories which are explained on the Yellow Card. MHRA asks
reporters to select one of the following criteria by ticking the appropriate box on the Yel ow Card: (1) patient died due to reaction
(2) life th reatening (3) resu lted i n h ospitalisation o r prolonged i npatient hospitalisation ( 4) co ngenital a bnormality an d (5)
involved persistent or significant disability or i ncapacity or ( 6) if t he reaction was deemed medically significant. In addition to
this, seriousness of reaction terms has also been defined by the MHRA in its medical dictionary. Therefore an ADR report can
be serious because the reporter considers the reaction to be serious or because the reaction term itself is considered serious in
the MHRA’s medical dictionary.
The one report of orthostatic hypotension was concurrent with POTS (see below).
Specific adverse events reported in temporal association with either brand of vaccine
are discussed in section 8 below, including six reports of POTS (including one
autonomic nervous system imbalance) following Gardasil (one of these reports of
POTS had no brand or date of administration).
7. RECENT UK MEDIA INTEREST IN HPV VACCINES
There have also been several cases of suspected side effects to HPV vaccine
brought to the attention of MHRA through the media and through correspondence
(including Parliamentary correspondence). Many of these relate to reports already
submitted via the Yellow Card Scheme. Many reports of alleged HPV vaccine side
effects, mostly from outside the UK, are also described in internet fora and social
media/patient and anti-vaccine networks (such as ‘SaneVax’ and ‘Arnica’).
Besides the specific issues discussed in section 8 below, recent articles in the media
have focused on the overall Yellow Card reporting rates for HPV as compared to
other routinely used vaccines.
On 31 May 2015, The Independent on Sunday ran a front page story with the
headline “Thousands of teenage girls enduring debilitating illnesses after routine
school cancer vaccination
” (the actual article focused on a case of postural
orthostatic tachycardia syndrome; POTS – see below). The headline was a
misrepresentation of the table below that was issued by MHRA in response to a
Freedom of Information request. Other newspapers picked up this headline, but
stating “Tens of thousands…..” based on an assumption that only 10% of suspected
ADRs are reported.
7.0. Overall safety profile and Yel ow Card reporting rates, and comparison
with similar UK vaccine programmes
The current ADR reporting rate for HPV vaccines in the UK is around 0.7 per 1,000
doses administered (since 2008, overall rate is around 1 per 1,000 doses, based on
just over 8,000 reports and at least 8 million doses given). This is currently greater
than the ADR reporting rate for other routinely used vaccines, which range from
0.04/1,000 doses for childhood pneumococcal vaccine to 0.2/1,000 doses for
rotavirus vaccine. As an illustration, the following tables are estimated ADR reporting
rates extracted from an update provided to JCVI in 2014:
In terms of relative Yellow Card reporting profiles (not relative safety profiles), a
suitable comparison to HPV vaccine is dT-IPV vaccine (Repevax) which is routinely
given as a single dose to a similar age group (around 14 years of age) via a school-
based programme. A reporting rate was not included in the table below as dT-IPV
vaccine is also used for tetanus boosting, and so the denominator is uncertain.
However, as most ADR reports for dT-IPV vaccine are in the adolescent age group,
we can reasonably assume a minimum annual exposure for 2013/14 in the table
below 530,000 vaccinees (annual school cohort ~660,000 with assumed 80%
vaccine uptake of a single dose).
Based on this, the estimated reporting rate for 2013/14 per 1,000 doses is 0.2/1,000
(0.15 for serious ADRs). Whilst the overall reporting rate for HPV vaccine is around
four-fold greater (two-fold for serious) than dT-IPV, this does not necessarily mean
that side effects are occurring at a greater rate and many factors can influence
different ADR reporting practice.
For instance, for the first few years of the HPV programme Cervarix had Black
Triangle status, which may have stimulated more reports of non-serious events. Of
particular note, we understand that certain school nurse teams have specific
processes in place to ask every vaccinee if they had any side effect after the
subsequent dose, and to report these to us. Indeed, the top 10 reporters (school
nurse teams) of Yellow Card reports following HPV vaccine have submitted over
40% of all reports (more than 3,500), 87% of which were non-serious.
Another factor in differential reporting behaviour is that most other vaccine
programmes in the UK are well-established, and reporting rates are generally lower
for older vaccines.
The proportion of all suspected ADRs for HPV vaccines that were reported as
serious (32%) is less than the proportion reported as serious for other routinely used
vaccines (68% overall) during the same time period. The vast majority of reports for
HPV vaccine relate to suspected side effects that we expect to see for most types of
vaccine given to adolescents and adults. These are most commonly dizziness,
headache, nausea, sore arms, vomiting, general malaise, tiredness, fever, and
rashes. These tend to be mild and transient. Many reports also relate to immediate
fainting, which is not a side effect of the vaccine but a response that any type of
needle insertion can provoke in some people.
Finally, the following four tables compare the overall relative proportion of serious
ADRs from each System Organ Class for vaccines routinely given to adolescents
and adults (HPV, dT-IPV, influenza and hepatitis B vaccine - extracted from the
Yellow Card update provided to JCVI in 2014). This shows a very similar pattern of
reporting that reflects adverse events reporting (regardless of causality) in
adolescent/adult vaccine programmes.
In summary, an overall reporting rate of 1 Yellow Card per 1,000 doses administered
in itself gives no cause for concern. The majority of these relate to the known
adverse events expected of any adolescent/adult immunisation which are mostly
non-serious, and the nature of reports (if not the current reporting rate) is broadly
similar across adolescent/adult vaccines. Local clinical practice in the solicitation and
reporting of suspected ADRs is a factor in differential reporting rates, and the rate is
still within the known frequency of common adverse events (e.g. the SmPC states
that injection site reactions, headaches, dizziness, nausea etc can occur at rates
around or above 10%).
To give some historical context, when meningitis C vaccine was introduced in 1999,
we received 13,000 Yellow Cards in the first year alone. The reporting rate per dose
at the time, was also 1/1,000 doses given, and that was long before patients and
carers were able to report suspected ADRs and before there was better access to
the Yellow Card Scheme via on-line reporting.
8. SPECIFIC EVENTS REPORTED IN TEMPORAL ASSOCIATION WITH HPV
8.1. Chronic Fatigue Syndrome (CFS)
CFS has not been raised as a signal in any of the Gardasil or Cervarix PSURs to
date. Prior to the start of the national HPV programme, MHRA undertook to evaluate
CFS following HPV vaccination proactively, not due to any existing safety concern,
but because we expected to receive many reports in temporal association with
vaccination given the natural incidence in the cohort eligible for vaccination. In
France and Canada, reports of CFS in temporal association with hepatitis B vaccines
in adolescents in the 1990s led to a fall in confidence in vaccine safety. The MHRA
strategy was therefore intended to gather robust data to evaluate any such case
reports in the UK.
As stated above, MHRA has completed an epidemiological study in 2013 which
found no evidence of an increased risk of CFS following Cervarix. The results of this
study were published in the journal Vaccine (2013 Oct 9;31(43):4961-7 - see annex
To date MHRA has received, 25 reports of CFS in association with Cervarix and 1 in
association with Gardasil and 9 brand not stated. As no safety signal has arisen in
relation to CFS, MHRA does not currently plan to extend the epidemiological study to
the time period when Gardasil has been used.
8.2. Guillain Barre Syndrome (GBS)
At the time of the 2012 safety update, MHRA had received 5 reports of GBS
following administration of more than 6 million doses of Cervarix. Given the expected
background incidence of GBS, the ‘observed’ vs ‘expected’ analyses suggested that
the reported cases are consistent with chance and there was no evidence of a safety
signal for GBS in the UK.
As of 30 June 2015, MHRA has received no additional
reports of GBS following either vaccine.
A publication by Souayah et al (www.ncbi.nlm.nih.gov/pubmed/20869467) discussed
cases of GBS following Gardasil reported to the US equivalent of the Yel ow Card
Scheme for vaccines (VAERS). However, a US epidemiological study found no
evidence of an association between Gardasil and GBS (see annex 12).
GBS was included as a possible side effect in the Gardasil SmPC several years ago,
mainly on the basis of evaluation of spontaneously reported cases worldwide.
Despite the epidemiological study that found no evidence of an association, this
remains in the Gardasil SmPC.
GBS is not included in the Cervarix SmPC, as the available evidence does not
support this. This is supported by the most recent ‘observed vs expected’ analysis in
the last Cervarix PSUR (see annex 6).
8.3. Acute disseminated encephalomyelitis (ADEM)
The MHRA has received no reports of ADEM in association with HPV vaccine.
In 2014, ADEM was included as a possible side effect in the Gardasil SmPC on the
basis of evaluation of spontaneously reported cases worldwide. The UK did not
support this, on the basis that the observed-expected analyses undertaken can only
be considered signal generating/strengthening, rather than supportive of a causal
relationship. The potential biological mechanism suggested by the Rapporteur
(molecular mimicry) was also speculative and should not be a significant factor in
drawing a conclusion on causality.
The last Cervarix PSUR included a detailed ‘observed vs expected’ analysis of
reports of ADEM (refer to section 18.104.22.168 of annex 6). The Rapporteur concluded that
the adverse event of ADEM [and TM] should not be included in the SmPC as there is
no proof of causality between ADEM and Cervarix exposure. However, the MAH
should continue to assess these events through PSURs.
8.4. Primary ovarian failure (POF)
There have been literature and non-UK spontaneous reports of primary ovarian
failure (POF) fol owing Gardasil. E.g. see annex 13.
Primary ovarian failure is a condition that occurs naturally amongst the age range of
girls who receive HPV vaccine. Little and Ward (see annex 13) highlight that the
annual natural incidence of premature ovarian failure is 10 per 100,000 person-years
between the ages of 15 and 29 years and that the cause of ovarian failure before the
age of 40 years is unknown in up to 90% of cases. It affects 1 in 10,000 women by
age 20, 1 in 1000 women by age 30, and 1 in 100 women by age 40 [Ferrarini E., et
Based on the background rate reported by Little and Ward, for every one million girls
vaccinated, around 25 cases of premature ovarian failure would be expected to
occur by coincidence within 3 months of vaccination.
The MHRA has received 1 report of primary ovarian failure following HPV vaccine to
date. With close to 3 million females vaccinated in the UK, there is no evidence that
menstrual cycle abnormalities were occurring following vaccination any more than
expected by chance amongst vaccinated girls.
The last Cervarix PSUR included an analysis of reports of POF (refer to section
22.214.171.124 of annex 6). Reports suggestive of primary ovarian failure remain under
review. The Rapporteur concluded that the available data is currently insufficient to
assess whether there is an increased risk of POF following vaccination with Cervarix,
and that this should remain under review.
The last Gardasil PSUR included an analysis of reports of POF (refer to page 37 of
annex 9). The query of the Company safety data base yielded 11 medically
confirmed reports which included the PT of ovarian failure
or premature menopause
received worldwide. 10 of the 11 cases were reported as non-serious.
Geographically, 4 reports were received from Australia, 3 reports were received from
Israel, 2 reports were received from the US, and 1 each was received from
Singapore and Germany. Age was reported in 9 reports as follows: 13, 13, 14, 14,
15, 16, 18, 21, and 43. The Rapporteur concluded that the reporting rate is
approximately 0.8 reports per every 10 million doses distributed, and does not
indicate a safety signal at present.
8.5. Complex Regional Pain Syndrome (CRPS)
Complex regional pain syndrome (CRPS) is characterised by severe pain, swelling
and changes in the skin temperature and colour of the arms or legs. The cause of
the syndrome is unknown but common predisposing conditions include trauma,
infection, surgery, cervical radiculopathy, soft tissue contusions, fractures, tendon
ruptures and myocardial infarction.
Data on the incidence of CRPS are scarce and mostly hospital based. A
retrospective cohort study conducted from 1996 to 2005 in the Integrated Primary
Care Information (IPCI), a general practice research database with electronic patient
record data from 600,000 patients across the Netherlands, estimates the overall
incidence rate of CRPS as 26.2 per 100,000 person years (95% CI: 23.0-29.7)5. In
this study the highest incidence occurred in females in the age category 61-70 years.
The upper extremity was affected more frequently than the lower extremity and a
fracture was the most common precipitating event (44%). Another study of the 1990
population (n=106,470) medical records from the Mayo Clinic and Olmsted Medical
5 De Mos M et al
. The incidence of complex regional pain syndrome: a population-based study. Pain
Group in Minnesota estimates the incidence rate as being much lower at 5.46 per
100,000 person years at risk with the most common trigger fracture reported at a
similar rate (46%)6 as the other study. The syndrome has been associated with
vaccines such as rubella and Hepatitis B vaccines. While there is a clear temporal
association in the majority of cases associated with the HPV vaccine CRPS may be
more likely attributed to needle trauma, as proposed by Genc et al.
than the vaccine constituents. It is also possible that such reports were coincidental.
At the time of the 2012 safety update, MHRA had received 6 reports of CRPS,
following administration of more than 6 million doses of Cervarix. In order to further
evaluate this signal the ‘snapshot’ Observed / Expected method was used for
analysis using the incidence rates 26.2 per 100,000 person years and 5.46 per
100,000 person years taken from the two published studies. Using these incidence
rates and usage data from September 2008 up to the end of May 2012 the Observed
/ Expected ratios were calculated as 0.03 (95% CI 0.01-0.07) and 0.16 (95% CI 0.06-
0.35) respectively. While this is a crude analysis using incidence rates not restricted
to 12-18 year old females, in both cases since the O/E ratios are wel below 1 these
data indicate that the observed number of events are well below the events expected
for this event in 12-18 year old girls who have received Cervarix.
As of 30 June 2015, MHRA has received 5 reports of CRPS following Cervarix, 2
following Gardasil and 3 with brand not stated (but most likely Cervarix based on the
8.5.1. Japanese suspension of HPV vaccine recommendation
In June 2013 and only 2 months after the start of their HPV vaccine programme, the
Japanese authorities temporarily stopped actively recommending immunisation with
the human papillomavirus (HPV) vaccines, Cervarix and Gardasil in girls and young
women. The vaccines, however, remain licensed for use in Japan and continue to
be available for girls and women who wish to receive them. Almost 9 million doses of
HPV vaccine (mostly Cervarix) had been given in Japan before this.
This action followed reports (around 40 at the time) of chronic pain in Japan,
including several (five at the time) cases of complex regional pain syndrome (CRPS;
a condition characterised by persistent burning pain in one or more limbs). The
Japanese action was a precautionary move whilst they gathered more data.
The symptoms were initially thought to be complex regional pain syndrome, which
manifests as chronic pain after a needle stick (e.g., after blood col ection or
vaccination), trauma, or other procedures/events resulting in pain. However, the
6 Sandroni P et al
. Complex regional pain syndrome type I: incidence and prevalence in Olmsted
county, a population-based study. Pain 2003;103
7 Genc H et al.
Complex regional pain syndrome type-I after rubella vaccine. Eur J Pain
characteristics of the reported clinical manifestations vary, especially time from
injection to onset of symptoms, which ranges from immediately after to 2 years after
Several other differential diagnoses were proposed, including macrophagic
myofasciitis, fibromyalgia, and chronic fatigue syndrome. After announcement of the
temporary withdrawal, the number of cases reported to the government increased to
121, as of September 2013. 8. In January, 2014, the Vaccine Adverse Reactions
Review Committee investigating these adverse events concluded that there was no
evidence to suggest a causal association between the HPV vaccine and the reported
adverse events after vaccination, but they still did not reinstate proactive
recommendations for its use9.
The latest information available to MHRA is from April 2014. This suggested that an
expert Committee in Japan had concluded that the cases may be psychosomatic in
origin. However, this remains in confidence as Japan has made no formal public
A letter to The Lancet in August 2013 was highly critical of the governance and
decision-making around the Japanese immunisation programme10. A further letter to
The Lancet on 27 June 2015 describes how HPV vaccine uptake has fallen from
over 70% to less than 1% due to this issue8.
A recent paper by Kinoshita et al (2014 - annex 14), described a series of cases of
suspected CRPS following HPV vaccine in Japanese girls, some of whom also had
POTS (see below).
9 Hanley et al. Vol 385 June 27, 2015
8.6. Postural Orthostatic Tachycardia Syndrome (POTS)
Annexes 15 to 22 include a range of background literature on POTS.
8.6.1. What is POTS?
POTS is characterised by an abnormally large increase in heart rate when changing
from a lying down to a standing up position, without any orthostatic hypotension. In
POTS, this excessive heart rate increase is usually accompanied by a range of
symptoms of orthostatic intolerance. These symptoms may include palpitations, light
headedness, weakness, ‘brain-fog’, peripheral coldness and purplish skin
discolouration and blurred vision. Some sufferers also experience fainting.
Although defined and diagnosed mainly by the tachycardia and orthostatic
symptoms, POTS is often associated with a wide range of other symptoms such as
migraine-like headaches, chronic aches and pains, gastrointestinal symptoms
(nausea, bloating, abdominal pain), sleep disturbance and shortness of breath. In
particular, fatigue is a very common feature. Many people diagnosed with POTS are
chronically fatigued, have sleep disturbance and are unable to perform daily
functions. There is now recognised to be a significant overlap between POTS and
Chronic Fatigue Syndrome (CFS).
There is no set pattern of symptoms in POTS sufferers and, aside from the defining
symptom of tachycardia, different people will show different symptoms. This wide
spectrum of symptoms probably reflects that the fact that the syndrome has several
distinct pathophysiological mechanisms.
In many people diagnosed with POTS, the range of symptoms can have a
detrimental impact on the overall quality of life. Anxiety, depression, and other
psychiatric disorders can also add to the complexity of the syndrome (Jarjour 2013 –
annex 15). Patients with POTS are sometimes clinically diagnosed as having anxiety
disorders such as panic disorder (Raj 2013 – annex 22).
8.6.2. How is POTS diagnosed?
Sufferers will generally present to their GP with gradual and chronic symptoms of
orthostatic intolerance and fatigue. Such symptoms can be non-specific and have
causes other than POTS, and so a diagnosis of POTS may not be clear cut.
If other causes are ruled out, POTS may be diagnosed if these chronic symptoms
are also associated with an excessive increase in heart rate when changing from
supine to a standing up position. For POTS to be diagnosed, patients need to
experience a sustained increase of 30 beats per minute or more within 10 minutes of
standing (or with tilt table test), without a fall in blood pressure. For those aged 12–
19 years this increase should be least 40 beats/min (Sheldon et al; 2015 - see annex
8.6.3. Is POTS new, and how well-recognised is it?
POTS began to be defined in adults in the mid-1990s (Schondrof and Low ; 1993),
but is likely to have existed long before then, with different names or diagnoses.
POTS has been described in adolescents since around 1999, and since then
awareness of the syndrome has increased. Over the past 15 years, a wide range of
medical literature has described POTS and there appears to be an increasing
awareness of the syndrome in the medical community.
POTS is still thought to be under-diagnosed and, according to Kizilbish et al (annex
16), some clinicians still do not recognise it as a legitimate diagnosis.
8.6.4. What are the features of POTS in adolescents?
A recent review by Kizilbash et al (2014 – annex 16), from the US Mayo Clinic,
summarises what is currently known about POTS in adolescence.
Whilst each patient has a unique set of symptoms, it is said that their stories often
sound familiar. Parents and adolescent patients with POTS often describe the long
and difficult process they experience from the moment they became ill, to decreased
school attendance with dropped extracurricular activities and poor academic
performance, to visiting a variety of doctors in order to find answers about their
child's illness. In this process, they often receive different diagnoses.
Acording to Kizilbash et al, there are several typical features retrospectively identified
in many adolescent POTS patients at presentation, including symptom onset during
early puberty, high achievers in school and athletics, joint hypermobility, and recent
illness or injury. Many adolescents with POTS have hyperextensibility, and some are
thought to have “benign joint hypermobility syndrome”. It is not clear whether the
elastic soft tissues actually predispose to the development of POTS or if the lax
tissues simply allow further increases in vasodilatation that make it more likely for
hyperextensible individuals to report more venous pooling and dizziness when they
8.6.5. Overlap with Chronic Fatigue Syndrome (CFS)
Many patients diagnosed with CFS also show symptoms of orthostatic intolerance
and signs of autonomic dysfunction, and several studies in UK patients have shown
that POTS can also be diagnosed in those with CFS, ranging from 13% (Lewis et al
2013 – annex 21) to 27% (Hoad et al; 2008 – annex 22). Some studies in the US
suggest the proportion of CFS patients with POTS is even higher. The level of
overlap will likely depend on the selection criteria within in each study, and it is
unclear how generalisable these estimates may be.
It is currently unclear whether POTS is a separate clinical entity distinct from CFS, or
whether patients with POTS form a subset of those with CFS with a specific group of
particularly marked symptoms (Lewis et al 2013).
As chronic fatigue is a common presenting feature in POTS, CFS may often be
diagnosed initially (or co-diagnosed) particularly in adolescents.With increasing
recognition of POTS in subsets of CFS patients, POTS may be a differential
diagnosis in many who are under evaluation for CFS. According to MacDonald et al
(2014 – annex 19), it is becoming increasingly clear that, historical y, many patients
with POTS were given a diagnosis of CFS.
8.6.6. How common is POTS, and who does it normally affect?
POTS can affect people of all ages, but the overwhelming majority of
women (80% to 85%) of child-bearing age (13–50 years). In adolescence, the
majority of affected individuals report symptoms beginning within a year or two of the
beginning of puberty, with worsening symptoms until the age of 16. About 80% of
female patients report an exacerbation of symptoms around menstruation (Raj; 2013
– annex 22). In 1999, the prevalence in the Unites States was estimated in adults at
170 per 100,000 [Low et al 2009 – annex 18].
Although POTS is thought to be under-diagnosed, there is a gradually increasing
awareness of POTS and diagnoses may be increasing. However, there is a lack of
reliable studies confirming the expected annual incidence of POTS in the UK in
different age groups. With increasing awareness of POTS, past estimates may
anyway be unreliable and changing over the time period since HPV vaccines were
introduced in the UK in 2008.
8.6.7. What causes POTS?
The causes and pathophysiology process of POTS are not well understood, and
there is no single precipitating factor. Although a decrease in return of blood to the
heart generally underlies the symptoms of POTS, the causes of this likely involve
multiple abnormal physiological processes that differ between sufferers. This is why
POTS is classed as a syndrome, rather than a specific disease.
Patients frequently report that their symptoms began after acute stressors such as
pregnancy, major surgery, or a presumed viral illness, but in others cases, symptoms
develop more insidiously. (Raj; 2013). A large number of patients initially become
symptomatic following a significant febrile illness, often mononucleosis or influenza
(Kizilbash et al).
It is generally thought that the underlying process in POTS is a dysfunction in the
autonomic nervous system. In POTS, it is thought autonomic reflexes can be
excessive or the nerves not functioning properly, leading to the symptoms of POTS.
This can vary between patients. Other illnesses can be associated with autonomic
dysfunction, so symptoms are not always specific to POTS. In some cases POTS
may also be caused by low blood volume.
It has been speculated that small fibre neuropathy may be a factor in some cases of
POTS (see section 8 below).
8.6.8. Can POTS be treated?
Many sufferers experience full recovery over time with or without treatment, but
some have persistent symptoms.
Treatment for POTS is generally tailored to the individual depending on the
symptoms they are experiencing and the underlying mechanisms for their illness.
This can range from measures such as controlling fluid intake and drugs to control
heart rate, which may help to correct the cardiovascular aspects of their symptoms.
Physical therapy and exercise may help to manage aspects of fatigue and exercise
intolerance. Given the impact on quality of life, there may also be behavioural
treatment, lifestyle changes and drugs to manage the psychological aspects of the
8.6.9. Reports of POTS following HPV vaccines
Since 2010, case reports of POTS following HPV vaccine have been reported in
several countries. The first report of which we are aware is a 20 year old girl from the
United States who developed symptoms of POTS (diagnosed several months later)
two weeks after her first dose of Gardasil (Blitshteyn 2010 – annex 23). Blitshteyn
(2013 – annex 24) subsequently reported a further five cases of POTS following
Gardasil in the US. There is likely to be have been a large number of people
immunised with Gardasil used in the US to date, and these few case reports need to
be viewed in that context.
A large series of POTS reports has now also emerged in Denmark, where it appears
a group of researchers from the Syncope Center at Frederiksberg Hospital are
actively seeking cases of POTS in adolescent girls, and questioning them specifically
on whether their symptom onset occurred following Gardasil vaccination. Brinth et al
(annexes 25 and 26) have published two papers in the literature that probably
involve the same patients in Denmark. In the latter paper, they have identified 53
patients from 2011 to 2013 with symptoms similar to POTS after Gardasil, although
only 28 had a POTS diagnosis. This research recently prompted with a Danish TV
documentary and significant media interest in the issue in Denmark.
Denmark is reported to have had very high uptake of its routine Gardasil programme
since 200912. After the first year, uptake among almost 34,000 eligible girls on a
national level was 80%, 75% and 62% respectively for the three HPV doses. Both of
the papers by Brinth et al. present findings from descriptive analyses that only
included patients referred to the syncope unit with orthostatic intolerance or other
symptoms consistent with autonomic dysfunction with onset of symptoms within 2
weeks following administration with Gardasil. There is no control group for either of
these studies and therefore no conclusions can be drawn as to whether Gardasil is
causally associated with POTS. These analyses simply provide a description of the
symptoms in these selected patients but it is unclear whether the symptoms are any
different to patients that did not receive the vaccine or in patients with symptom
onset longer than the 2 week period used in this study. Furthermore given that the
Danish researchers appear to be actively identifying cases of POTS that have
occurred within a short time after vaccination, and interviewed them “with a special
focus on symptoms and on the temporal association between vaccination and
” (a process which in itself can prompt bias in symptom recal ), the
cases in Denmark need to be interpreted with caution.
Kinoshita et al (2014 – annex 14) describe four cases of POTS following HPV
vaccine (brand not stated), who were under evaluation for complex regional pain
On an international level, it is likely that more case reports in temporal association
with HPV vaccines will continue to be reported given the recent publications and
8.6.11. UK reports of POTS
On 31 May 2015, The Independent on Sunday ran a front page story that focused on
an illness suffered by a teenager following her HPV vaccination given four years ago
[note – see section 9 below for wider recent media interest]. It states that she has
been diagnosed with Postural Orthostatic Tachycardia Syndrome (POTS). This case
has been reported to the MHRA via the Yellow Card Scheme as a suspected side
effect to the vaccine.
In terms of UK case reports, as of 30 June 2015 MHRA has received eleven reports
of POTS following HPV vaccine (5 following Cervarix and 5 following Gardasil and
one with no brand stated). [note: six cases were reported following the recent media
interest]. Two were reported in 2012, two in 2013, one in 2014 and six in 2015,
although symptom onset may have been before then (five have CFS initially/co-
diagnosed, and two have concurrent encephalitis).
Onset time was from 1 day to several weeks after the first dose, with symptoms
starting as a virus-like illness in some cases and some reporting worsening
symptoms after subsequent doses. The case details are included in annex 28. There
is no consistent clinical pattern in terms of plausible onset times from vaccination or
8.6.12. UK Observed vs expected analyses
There are no reliable data on the expected background incidence of POTS in those
eligible for HPV vaccine the UK (or indeed in any country).
Given that reliable
information exists on the natural incidence of CFS in the UK and the estimated
overlap with cases of CFS, this may be a reasonable starting to point to estimate a
minimum possible incidence of POTS.
Data from the UK (the same data that informed MHRA’s O/E analysis of CFS)
suggest that the background incidence of CFS ranges from 30 to 70 per 100,000 per
year in females aged 12 to 18 years. This rate is based on data in the 10 years prior
to 2008. The background incidence of POTS in adolescent girls in the UK may be
estimated from literature. A study by Lewis et al; 2013 stated that around 10% of
new CFS cases may have co-existing POTS. An additional study by McDonald et al;
2015 suggested that between 7-42% of POTS patients also have co-existing CFS.
Combining these pieces of information, 40% of the CFS incidence may give a
reasonable minimum estimate of the background incidence of POTS in adolescent
girls in the UK.
Around 3 million adolescent girls have been immunised across the UK since 2008.
Using this data and the estimated POTS incidence as detailed above, a range of
expected cases of POTS for different onset periods post-vaccination and different
spontaneous reporting scenarios are presented in the table below:
Table. O/E analysis of UK data
Assumed reporting rate
Expected number of
Within 1 year post-
Within 2 years post-
At any time post-
A total of 11 cases have been reported in the UK to date. An excess of observed
cases compared to the expected number is seen only at the 1% reporting level and
diagnosis within under 3 years post-vaccination. Since POTS is a relatively novel
condition and therefore likely to be under-diagnosed and with delays in diagnosis, it
seems unlikely that these scenarios reflect a true excess of cases. Additionally a
very conservative incidence rate for POTS has been used in this O/E analysis and it
is likely that the true incidence rate is higher which would increase the expected
8.6.13. Observed vs expected analysis based on Danish reports (analysis of last
In the last PSUR assessment, the MAH conducted an O/E analysis restricted to
Danish data only as the majority of cases are from Denmark (refer to section 2.2 and
page 48 onwards of annex 9).
In the absence of POTS incidence rates from the literature the MAH has followed the
suggestion by a POTS expert at the Mayo Clinic in Rochester, Minnesota to obtain a
range of estimates of the annual incidence of POTS in females aged 10-39 years
Using annual incidence data for CFS (60-150 per 100,000 females aged 10-39
years) and assuming 10-25% proportion of CFS cases have POTS and that 10% of
POTS cases have CFS.
Using annual incidence data for chronic fatigue in general (250-520 per 10,000
females aged 10-39 years) and assuming 25%-66% proportion of chronic fatigue
cases have POTS and 100% of POTS cases have fatigue as a symptom.
The annual incidence rate of POTS was estimated to be 60-150 per 100,000 females
under scenario (a) above and 2501-3302 per 100,000 females under scenario (b).
Other assumptions used in the analysis were:
75-100% doses distributed (1,351,593 doses) were administered
Spontaneous reporting rate of POTS post-vaccination 10-25%
Risk period post-vaccination of 1 month
Using these assumptions the MAH calculated the lowest potential expected number
of POTS cases within 1 month post-vaccination and the highest potential expected
number within 1 month as follows:
The MAH concludes that the observed counts of POTS are within the expected
range of counts for a 4-week period after vaccination. The Rapporteur, however,
concludes that due to the observed cases exceeding the lowest expected estimate a
causal association cannot be ruled out.
The MAH O/E analysis is however not particularly informative due to the large range
of expected cases (5-929). Additionally some of the assumptions made are
The exposure data should be expressed as the number of vaccinees and not
the number of doses. A conservative approach would be to assume that al
girls received 3 doses of vaccine and therefore the total exposure would be
450,531 vaccinees. The impact of this would have been to overestimate the
expected number of cases.
The highest expected estimate assumes that 100% of doses distributed have
been administered. This is highly unlikely and the 75% assumption used in
the lowest estimate is much more plausible assumption.
A risk period for developing POTS of 4 weeks post-vaccination was chosen
with no rationale given. Time to onset for all reports received globally was 1-
715 days post-vaccination so to restrict the analysis to 4 weeks seems
inappropriate. Additionally only 16/28 of the Danish cases met this 4 week
criteria thus excluding nearly half of the cases from the analysis.
Under-reporting scenarios for spontaneous reports include 10-25% reporting.
Given the current interest in POTS following HPV vaccination in Denmark and
the active identification of cases it seems likely that the reporting rate is higher
A revised O/E analysis addressing some of the issues above would give an expected
number of cases of POTS under different scenarios as follows:
Table. Revised O/E analysis of Danish data (1)
Expected number of
Within 1 year post-
Within 2 years post-
At any time post-
O/E assumptions: all vaccines given 3 doses, 75% doses distributed are administered, scenario (a)
above for incidence rate of POTS – chosen as this is the more conservative estimate.
The time to onset of all cases globally was within 2 years so following this risk period
post-vaccination the observed number of POTS cases in Denmark of 28 (in the study
by Brinth et al) are well within the expected rate for all reporting scenarios. Even
within the more restrictive risk period of within 1 year post-vaccination the observed
number of cases is very similar to or less than expected.
A further more conservative O/E analysis of the Danish data is shown below using
the same POTS incidence rate as calculated for the UK O/E analysis above. This
rate assumes that 10% of patients with CFS also have POTS and that 40% of POTS
patients also have CFS.
Table. Revised O/E analysis using Danish data (2)
Expected number of
Within 1 year post-
Within 2 years post-
At any time post-
Using this more conservative estimate of POTS incidence inevitably gives a lower
expected numbers of cases. In this scenario an excess of observed is seen for the 1
year post-vaccination risk period but the total number of Danish cases (n=28) will
also reduce under this assumption. An excess is also seen for the 25-50% reporting
scenarios. However, with higher awareness and active identification of cases in
Denmark by Brinth et al it is likely that the reporting rate could be high and even
In conclusion the O/E analysis using Danish data only provides evidence of an
excess of observed cases compared to expected when a restricted risk period and a
low reporting rate are used. These scenarios are not considered to be the most
8.7. Small Fibre Neuropathy
POTS, CRPS and fibromyalgia can have a variety of overlapping symptoms, that
may relate to some dysfunction of the autonomic nervous system in some patients.
However, a recent publication by Manuel Martínez-Lavín13 (annex 27 - referred to in
the headline article in The Independent on Sunday) has suggested that small fibre
neuropathy is a common underlying mechanism in these three disorders.
Furthermore, as these conditions have been reported in temporal association with
HPV vaccine, Martínez-Lavín has also suggested that HPV-induced small fibre
neuropathy as a common mechanism through which it may cause these conditions,
and has coined the phrase ‘HPV syndrome’.
The author goes on to suggest that aluminium in HPV vaccines is a possible cause
of this. Finally, Martínez-Lavín questions the benefits of HPV vaccine, arguing that
most HPV infections will be cleared by cell-mediated immune responses within 3
years of infection.
The article by Martínez-Lavín is purely speculative, and selective in the articles it
uses in support of the theory. Aluminium has been used in numerous vaccines for
decades, and no other vaccines have been associated with POTS, fibromyalgia or
CRPS (in the context of a chronic condition as opposed to possible needle stick
trauma) to date. The author fails to acknowledge this, and fails to offer the high
natural prevalence of these conditions in adolescent females as a plausible
alternative explanation for the emerging case reports. The author offers an
association between aluminium in vaccines and other autoimmune disorders in
support of the theory, but also fails to state that these too have been discredited14.
As with any postulated adverse effects of the vaccines, small fibre neuropathy as a
possible adverse effect of HPV vaccine should be taken seriously and kept under
review. However, Manuel Martínez-Lavín provides no new data in support of the
theory and this article should be treated with caution.
In the last PSUR for Gardasil and based a speculated aetiology for POTS and CRPS
of small fibre neuropathy, the Rapporteur asked the MAH to discuss the possible
biological plausibility of a common pathophysiology for vaccine-induced POTS and
CRPS (pages 79-81 of annex 9). The MAH argues that there is little evidence for
this, that POTS and CRPS are heterogeneous with likely multi-factorial
pathophysiology and the theory of small fibre neuropathy as a cause of vaccine-
induced POTS and CRPS remains speculative.
9. HPV VACCINE SAFETY REVIEWS BY THE WORLD HEALTH
ORGANISATION AND US HEALTH AUTHORITIES
Outside of the EU, other health authorities are closely evaluating the safety of HPV
as part of their own routine pharmacovigilance and risk evaluation activities.
In the United States, the Centre for Disease Control and Prevention (CDC) has
continually evaluated the safety of HPV vaccine (mainly Gardasil used in US). In
2011, CDC conducted a large study following more than 600,000 doses of HPV
vaccine. Adverse events in the HPV vaccinated population were compared to
another appropriate population (such as adolescents vaccinated with vaccines other
than HPV) and the conditions studied included Guillain–Barré syndrome (GBS),
stroke, venous thromboembolism (VTE), appendicitis, seizures, syncope (fainting),
and allergic reactions. None of these adverse events were found to be any more
common after HPV vaccination than among their comparison groups. CDC’s
analysis of ADR data continues to show no new or unusual patterns of adverse
events to suggest a HPV vaccine safety concern15.
The World Health Organization’s Global Advisory Committee on Vaccine Safety
(GACVS) has also reviewed the safety of HPV vaccine on several occasions16.
GACVS’s most recent advice is that multiple studies have demonstrated no increase
in risk of autoimmune diseases, among girls who have received HPV vaccine
compared to those who have not, and concluded that it remains reassured by the
safety profile of HPV vaccine. GACVS issued a statement in March 2014 supporting
the continued use of the vaccine (annex 28).
Reports of suspected side effects following HPV vaccines have remained under
close and continual review by the MHRA since the national immunisation programme
began in 2008. This not only includes UK reports via the Yellow Card Scheme, but
also the international experience with HPV vaccines. MHRA works very closely with
its European counterparts in reviewing safety.
The overall profile and pattern of reporting of suspected side effects to HPV vaccines
via the Yellow Card Scheme is not unusual. Reported events are not necessarily
causally-associated with the vaccine, and the pattern is anyway broadly what would
be expected of any vaccine given to teenagers and adults. Although Yellow Card
reporting rates do not equate directly to the frequency of side effects actually
occurring, and cannot be used to compare the relative safety of different vaccines,
the reporting rate is still within the expected, known frequency of common side
effects based on experience in clinical trials.
The fact that the absolute number of Yellow Card reports for HPV vaccines and the
reporting rate is greater than other routinely-used vaccines over the same time
period is not a cause for concern. This does not necessarily mean that more side
effects are occurring and can most likely be explained partly by differential local
practices in the way suspected side effects following prior doses are collected and
reported when vaccinees present for subsequent doses.
It remains the case that no serious new side effects to HPV vaccines have been
confirmed based on reports to the Yellow Card Scheme to date. Around 3 million
eligible girls have received HPV vaccine so far, and reports of serious suspected
side effects remain consistent with chance. Such reports will remain under continual
In relation to Postural Orthostatic Tachycardia Syndrome (POTS), available
published data suggest that a proportion those diagnosed with Chronic Fatigue
Syndrome (CFS) also have features/diagnoses of POTS. It is therefore reasonable
to assume that the minimum expected background annual incidence of POTS in
those eligible for vaccination is around 40% of the known incidence of CFS. Although
as POTS can occur independently of CFS, the expected background frequency
could be significantly greater.
Given the number of girls vaccinated to date in the UK, and the lack of a consistent
clinical pattern in terms of symptoms or plausible onset times from vaccination , UK
reports of POTS following vaccination to date are consistent with a chance
association. Such reports will remain under close review, taking account of the
experience from other countries.
The HPV vaccine programme is expected to have significant future benefits
preventing hundreds of deaths every year from cervical cancer, as well as reducing
genital warts and other cancers also related to HPV infection. The early signs are
very encouraging, with evidence from Public Health England that the UK HPV
immunisation programme is already preventing high risk HPV infections.
Based on the data currently available, the known risks of HPV vaccines are
adequately described in the Summaries of Product Characteristics, and these
known, possible risks are far outweighed by the benefits in reducing mortality and
morbidity from HPV-related infections.
11. ADVICE SOUGHT
The Commission is asked if it agrees with the conclusions above
Annex 1: Gardasil Summary of Product Characteristics
Annex 2: Cervarix Summary of Product Characteristics
Annex 3: Cervarix HPV vaccine – end of routine programme
safety review – Redacted under section 40 (personal
information) and section 43 (commercial interest)
Annex 4: Cervarix Product Analysis Print (all cases post
Annex 5: HPV brand unspecified Drug Analysis Print (vaccine
administration before September 2012)
Annex 6: Cervarix PRAC Rapporteur PSUR preliminary
assessment report – withheld under section 27 (international
Annex 7: Gardasil Product Analysis Print
Annex 8: HPV brand unspecified Drug Analysis Print (vaccine
administration after September 2012)
Annex 9: Gardasil PRAC Rapporteur PSUR preliminary
assessment report – withheld under section 27 (international
Annex 10: Arnheim-Dahlstrom, Pasternak B, Svanstrom H,
Sparen P and Hviid A. Autoimmune, neurological, and venous
thromboembolic adverse events after immunisation of
adolescent girls with quadrivalent human papillomavirus
vaccine in Denmark and Sweden: cohort study BMJ (347):
Annex 11: Bryan P et al, Bivalent human papillomavirus
vaccine and the risk of fatigue syndromes in girls in the UK:
Vaccine (2013 Oct 9;31(43):4961-7)
Annex 12: Gee J, et al, Monitoring the safety of quadrivalent
human papillomavirus vaccine: Findings from the Vaccine
Safety Datalink: Vaccine (2011 8279-8284)
Annex 13: Little D T, Rodrick H, Ward G, Premature ovarian
failure 3 years after menarche in a 16-year-old girl following
human papillomavirus vaccination: BMJ Case Reports 2012
Annex 14: Kinoshita T, et al Peripheral Sympathetic Nerve
Dysfunction in Adolescent Japanese Girls Following
Immunization with the Human Papillomavirus Vaccine. Intern
Med 53: 2185-2200, 2014
Annex 15: Jarjour I T, Postural Tachycardia Syndrome in
Children and Adolescents. Semin PediatrNeurol20:18-26 C
Annex 16: Kizilbash S J et al, Adolescent Fatigue, POTS, and
Recovery: A Guide for Clinicians. Curr Probl Pediatr Adolesc
Health Care 2014;44:108-133
Annex 17: Sheldon et al, 2015 Heart Rhythm Society Expert
Consensus Statement on the Diagnosis and Treatment of
Postural Tachycardia Syndrome, Inappropriate Sinus
Tachycardia, and Vasovagal Syncope. Heart Rhythm, Vol 12,
No 6, June 2015
Annex 18:Low P et al, Postural Tachycardia Syndrome
(POTS) J Cardiovasc Electrophysiol
. 2009 March ; 20(3): 352–
Annex 19: McDonald C, Koshi S, Busner L, et al. Postural
tachycardia syndrome is associated withsignificant symptoms
and functional impairment predominantly affecting young
women: a UK perspective. BMJ Open 2014;4:e004127.
Annex 20: Hoad A et al, Postural orthostatic tachycardia
syndrome is an under-recognized condition in chronic fatigue
syndrome. Q J Med 2008; 101:961–965
Annex 21: Lewis I et al, Clinical characteristics of a novel
subgroup of chronic fatigue syndrome patients with postural
orthostatic tachycardia syndrome. Intern Med
2013; 273: 501–510.
Annex 22: Raj S R, Postural Tachycardia Syndrome (POTS) Circulation.
Annex 23: S. Blitshteyn, Postural tachycardia syndrome after
vaccination with Gardasil. European Journal of Neurology
Annex 24: S. Blitshteyn, Postural tachycardia syndrome
following human papillomavirus vaccination. European Journal
of Neurology 2010 doi:10.1111/ene.12272
Annex 25: Brinth L et al, Orthostatic intolerance and postural
tachycardia syndrome as suspected adverse effects of
vaccination against human papillomavirus: Vaccine 2015
Annex 26: Brint L et al, Suspected side effects to the
quadrivalent human papilloma vaccine. Dan Med
Annex 27: Martinez-Lavin M, Hypothesis: Human
papillomavirus vaccination syndrome—small fiber neuropathy
and dysautonomia could be its underlying pathogenesis.
Clinical Rheumatology 2015 DOI 10.1007/s10067-015-2969-z
Annex 28: Line listing- UK spontaneous suspected ADR
reports in association with HPV vaccine and Postural
Orthostatic Tachycardia Syndrome (POTS) – redacted under
section 40 (personal information)