This is an HTML version of an attachment to the Freedom of Information request 'Summary Minutes for CHM meeting on 16th & 17th July 2015'.

RESTRICTED COMMERCIAL 
CHM 2012/8th MEETING 
 BVEAG 
2012/5th MEETING 
PEAG 2012/8th MEETING 
NOT FOR PUBLICATION 
     
 
COMMISSION ON HUMAN MEDICINES 

 
BIOLOGICALS AND VACCINES EXPERT ADVISORY GROUP 
 
PHARMACOVIGILANCE EXPERT ADVISORY GROUP 
 
Title of paper: 
Cervarix HPV vaccine – end of routine programme safety review 
 
Type of paper: For advice 
  
Product:
  
Marketing Authorisation Holder (MAH): 
Cervarix  
GlaxoSmithKline (GSK) Biologicals S.A. 
Active Constituent: 
Pharmaceutical form: 
Human Papillomavirus virus-like particles  
Suspension for injection 
(HPV types 16 and 18) 
Therapeutic area: 
Legal status: 
Papillomavirus infections 
POM 
Uterine cervical dysplasia 
Immunisation  
ATC code: J07BM02 
Previous assessments: 
Assessors:  
CHM 08/08; 02/09; 06/09; 09/09, 09/10 
 
 
INDEX                                                                                                         Page 
 
1.  ISSUE                                                                                                       2 
 
 
 
2.  BACKGROUND                                                                                        2 
 
3.  USAGE DATA                                                                                           6 
 
4.  SPONTANEOUS ADVERSE REACTION (ADR) DATA 
               7 
 
 
5.  OBSERVED/EXPECTED ANALYSES                                                    26 
 
6.  CPRD STUDY 
                                                                               30 
 
 
 
7.  DISCUSSION                                                                                          38 
 
8.  CONCLUSIONS 
 
                                                                  41 
             
 
9.  ADVICE SOUGHT   
                                                                  42 
 
 
 
 
 
ANNEXES 
                                                                                            43 
 
 
 
 
 
 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  


link to page 2 link to page 2 RESTRICTED COMMERCIAL      NOT 
FOR PUBLICATION 
 
COMMISSION ON HUMAN MEDICINES 

 
BIOLOGICALS AND VACCINES EXPERT ADVISORY GROUP 

 
PHARMACOVIGILANCE EXPERT ADVISORY GROUP 
 
 
Cervarix HPV vaccine – end of routine programme safety review 
 
 
 
1. ISSUE 
 
Over 99% of cervical cancer is attributable to human papillomavirus (HPV) infection with 
greater than 70% of cases related to HPV types 16 and 18. To reduce the burden of disease 
in the UK, the HPV vaccine Cervarix, containing HPV types 16 and 18, was introduced into 
the national immunisation programme in September 2008 for adolescent girls aged 12 to 13 
years (school year 8) and for a short time period to older girls aged 13 to 18 years as part of a 
catch-up programme.  
 
From September 2012 Cervarix will be replaced with Gardasil in the routine national HPV 
immunisation programme. From this date all eligible girls are expected to start and complete 
the vaccination course using Gardasil. With routine use of Cervarix in the national HPV 
immunisation programme drawing to a close this report summarises the safety experience of 
Cervarix within the national HPV immunisation programme up to 31st July 2012.  
 
The advice of the Commission on Human Medicines (CHM) and its Pharmacovigilance EAG 
(PEAG) and Biologicals and Vaccines EAG (BVEAG) is sought on the conclusions of the 
report.  
 
 
 
2. B

ACKGROUND 
 
2.1 

Cervical Cancer and HPV 
 
Worldwide more than 273,000 deaths are from cervical cancer each year and these account 
for 9% of all female cancer deaths. Increased screening activity in recent years has reduced 
cervical cancer mortality rates in the UK by nearly 70% in 2008 (2.4 per 100,000 females) 
compared to 30 years earlier (7.1 per 100,000 females in 1979)1. However the cost to 
patients afflicted is still high with 957 deaths from cervical cancer in 2008 in the UK with an 
estimated crude rate of 3.1 per 100,000 population. 
 
Over 99% of cervical cancers are caused by human papillomavirus (HPV) infection and of the 
estimated100 types of HPV there are about 40 genital HPV types that infect the genital area2
                                                      
http://info.cancerresearchuk.org/cancerstats/types/cervix/mortality/uk-cervical-cancer-mortality-
statistics#source1 accessed August 2012 
2 McCance DJ. Papillomaviruses. In: Zuckerman AJ, Banatvala JE, Pattison JR, Griffiths P and Schoub 
B (eds) Principles and practice of clinical virology. 2004 5th edition. Wiley & Sons Ltd. 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  


link to page 3 link to page 3 link to page 3 link to page 3 link to page 3 link to page 3 link to page 3 link to page 3 While some HPV infections can resolve of their own accord, genital HPVs can cause cancer, 
genital warts, anogenital cancers and cancers of the head and neck3 4.  
 
HPV is believed to cause cervical cancer by changing infected epithelial cells. HPV DNA can 
integrate into human DNA in the cervical epithelial cells at the site of infection and it is this 
process that is likely to lead to cancer progression5. However the exact nature of this process 
and the role of other factors are not fully understood.  
 
Genital HPVs are classified as ‘high-risk’ or oncogenic types which cause cervical cancer and 
early cervical changes and ‘low-risk’ types, which lead to the development of benign genital 
warts. In Europe the two main high-risk types, HPV 16 and HPV 18, are together responsible 
for over 70% of all cervical cancers6.  While the majority of cases of high-risk HPV infection 
may not lead to cervical cancer, HPV can cause abnormalities of the cervix which in turn can 
lead to cervical cancer. Often the time between infection by a high-risk HPV and development 
of cervical cancer is several years7.  
 
It is estimated by the Centers for Disease Control and Prevention that at least 50% of all 
sexually active women and men are infected by genital HPV at some point in their lives8. A 
study of antibodies to four types of HPV infection (6, 11, 16 and 18) showed that HPV 
infection in females increases rapidly from 14 to 24 years of age9 with infection more likely to 
occur in late teens and early twenties. Abstinence from any sexual activity greatly reduces the 
risk of genital HPV infection and while condoms also reduce the risk they are not 100% 
effective10.  
 
 
2.2 HPV 
Vaccines 
 
In the UK there are two HPV vaccines currently authorised, Cervarix manufactured by GSK 
and Gardasil manufactured by Sanofi Pasteur MSD.  
 
2.2.1 Cerv arix 
 
Cervarix was granted a marketing authorisation by the European Commission on 20th 
September 2007 through the centralised procedure with Belgium as Rapporteur (lead 
Member State). Cervarix is a suspension for injection that contains purified proteins (L1 
proteins) for HPV types 16 and 18. It is available in vials or prefilled syringes and indicated for 
use from the age of 9 years for the prevention of premalignant cervical lesions and cervical 
cancer causally related to certain oncogenic HPV types. The indication is based on the 
                                                      
3 Parkin DM & Bray F. Chapter 2: The burden of HPV-related cancers. Vaccine 2006; 24 S3 S11-25. 
4 Stanley M. Prophylactic HPV vaccines: prospects for eliminating ano-genital cancer. Br J Cancer 2007; 
96(9): 1320-3.
5 Woodman CB et al. The natural history of cervical HPV infection: unresolved issues. Nat Rev Cancer 
2007; 7(1): 11-22. 
6 Smith JS et al. Human papillomavirus type distribution in invasive cervical cancer and high-grade 
cervical lesions: a meta-analysis update. Int J Cancer 2007; 121(3): 621-32. 
7 Moscicki AB et al. Chapter 5: Updating the natural history of HPV and anogenital cancer. Vaccine 
2006; 24 S3 S42-51. 
8 Genital HPV Infection – Fact Sheet. Centers for Disease Control and Prevention (CDC) website 
http://www.cdc.gov/std/HPV/STDFact-HPV.htm accessed August 2012  
9 Jit M et al. Prevalence of human papillomavirus antibodies in young female subjects in England. Br J 
Cancer 2007; 97(7): 989-91. 
10 Koutsky L. Epidemiology of genital human papillomavirus infection. Am J Med 1997; 102(5A): 3-8. 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  


link to page 4 link to page 4 link to page 4 link to page 4 link to page 4 demonstration of vaccine efficacy in women aged 15-25 years and on the immunogenicity of 
the vaccine in girls and women aged 9-25 years. The clinical development programme to 
support licensure of the Cervarix consisted of 9 clinical studies11.  
 
Cervarix is manufactured by recombinant DNA technology using highly purified virus-like 
particles (VLPs) of the major capsid L1 protein of oncogenic HPV types 16 and 18. VLPs 
mimic the structure of the native virus but do not contain any viral DNA. Immunisation with 
Cervarix elicits an immune response to the L1 proteins assembled in VLPs. Production of 
antibodies is more rapid following exposure to HPV and thus the risk of diseases caused by 
HPV types 16 and 18 is reduced. Since the VLPs contain no viral DNA they cannot infect 
cells, reproduce or cause HPV infection. 
 
Cervarix is made using an adjuvant system that consists of aluminium hydroxide and 
monophosphoryl lipid A (MPL) known as AS0412. MPL is a purified lipid extracted from 
bacteria, which enhances the response of the immune system to the vaccine.  
 
Cervarix has been registered in 123 countries and since launch up to November 2011, over 
28 million doses have been distributed worldwide.  
 
The HPV vaccine has been demonstrated to be over 99% effective in preventing pre-
cancerous lesions associated with HPV types 16 and 18 in women who have not already 
been infected by these types13  14. However the vaccine does not protect against disease if 
HPV infection is already established in the individual. While the vaccine does not protect 
against all HPV types that cause cervical cancer, there is evidence that Cervarix also 
provides cross-protection against HPV types 31, 33 and 45, the three most common cancer-
causing virus types following types 16 and 1815
 
In the UK the Joint Committee on Vaccination and Immunisation (JCVI) recommended a 
vaccine schedule for Cervarix in the national HPV immunisation programme of three doses 
(0, 1-2 and 6 months) by intramuscular injection for girls in school year 8 (aged 12 to 13 
years) over a period of 12 months. When the vaccine was first introduced in September 2008 
this was followed by a catch-up programme for older girls between 13 and 18 years of age. 
Cervarix is considered to be suitable for the majority of adolescent girls and young women 
although there are a very few individuals who cannot receive the vaccine including those who 
have had a confirmed anaphylactic reaction to a previous dose of HPV vaccine or to any 
components of the vaccine. Vaccination is also delayed in those who are pregnant, breast-
feeding or those who have a high temperature. Individuals with immunosuppression should 
be considered for vaccination although these individuals may not develop a full antibody 
                                                      
11 Further details of the clinical trials can be found on the European Medicines Agency (EMA) website 
(www.ema.europa.eu), and the pivotal trials have also been published in the medical literature (e.g. 
Paavonen et al. The Lancet. Vol 369. June 30. 2007). 
12 Adjuvant System 04, a trade name for combination of adjuvants used in various vaccine products by 
GSK 
13 Ault KA and FUTURE II Study Group. Effect of prophylactic human papillomavirus L1 virus-like-
particle vaccine on risk of cervical intraepithelial neoplasia grade 2, grade 3, and adenocarcinoma in 
situ: a combined analysis of four randomised clinical trials. Lancet 2007; 369(9576):1861-8. 
14 Harper DM et al. Sustained efficacy up to 4.5 years of a bivalent L1 virus-like particle vaccine against 
human papillomavirus types 16 and 18: follow-up from a randomised control trial. Lancet 2006; 
367(9518): 1247-55. 
15 Paavonen J et al. Efficacy of the HPV-16/18 AS04-adjuvanted vaccine against cervical infection and 
pre-cancer caused by oncogenic HPV types: final event-driven analysis in young women (the PATRICIA 
trial). The Lancet. 2009 Jul 25;374(9686):301-14. 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  


link to page 5 link to page 5 response. Prescribing guidance is outlined in the Cervarix Summary of Product 
Characteristics (annex 1).  
 
In the Cervarix clinical trials involving over 19,000 females the most common adverse 
reactions observed after vaccine administration were injection site reactions including pain, 
redness, swelling; fatigue; myalgia and headache with a frequency of very common (≥1/10) 
(annex 1). Injection site pain occurred the most frequently (after 78% of all doses). The 
majority of these reactions were of mild to moderate severity and not long lasting. 
 
Common (≥1/100 to <1/10) reactions observed include gastrointestinal symptoms such as 
nausea, vomiting, diarrhoea and abdominal pain; itching/pruritus; rash; urticaria; arthralgia; 
and fever (≥38°C), while reactions identified within the clinical trials as uncommon (≥1/1000 to 
<1/100) include upper respiratory tract infection; dizziness; and other injection site reactions 
such as induration and local paraesthesia. During the post marketing period 
lymphadenopathy; allergic reactions (including anaphylactic and anaphylactoid reactions); 
angioedema; and syncope or vasovagal responses to injection sometimes accompanied by 
tonic-clonic movements were associated with Cervarix administration and are also listed in 
the SPC (annex 1).  
 
 
2.2.2 Garda sil 
 
The second HPV vaccine marketed in the UK is Gardasil and this vaccine was authorised 
before Cervarix on 20th September 2006 by the European Commission. In addition to HPV 
types 16 and 16 Gardasil contains two extra HPV types (types 6 and 11). HPV types 6 and 11 
are responsible for approximately 90% of genital wart cases and Gardasil has been 
demonstrated to be 99% effective at preventing genital warts caused by specific HPV types16
Gardasil is indicated for the prevention of premalignant genital lesions (cervical, vulvar and 
vaginal) and cervical cancer caused by specific HPV types and for the prevention of genital 
warts (condyloma acuminate). 
 
To date Gardasil has been used to a limited extent in the UK as previously the vaccine had 
not been procured by the Department of Health for use within the routine national HPV 
immunisation programme. However from September 2012 Cervarix will be replaced by 
Gardasil in the national HPV immunisation programme for girls in school year 8 (aged 12 to 
13 years). Gardasil has been extensively used in other European countries and in the US with 
more than 26 million people vaccinated worldwide. 
 
 
2.3 
Previous assessments of safety data 
 
In September 2008 the CHM considered the MHRA’s enhanced Pharmacovigilance strategy 
for Cervarix prior to the start of the national HPV immunisation programme. This included 
signal evaluation and risk assessment involving daily assessment and categorisation of all 
suspected new adverse reactions (ADRs); a proactive communication plan (encouraging use 
of the Yellow Card Scheme and weekly publication of ‘Suspected Adverse Reaction Analysis’ 
reports on the MHRA website17); safety updates in the Drug Safety Update (DSU) bulletin; 
                                                      
16 Barr E and Tamms G. Quadrivalent human papillomavirus vaccine. Clin Infect Dis 2007; 45(5): 609-7. 
17 www.mhra.gsi.gov.uk/HPVvaccine 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  


link to page 6 link to page 6 weekly teleconferences with the MAH (GSK) to discuss ADRs reported to the Yellow Card 
Scheme and 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; and proactive and real-time 
‘observed vs expected’ analysis using a statistical sequential test method (Maximised 
Sequential Probability Ratio Test (MaxSPRT)) of key adverse events of interest to identify 
possible new risks associated with HPV vaccines (annex 2).  
 
In February 2009 and again in September 2009 (1-year safety review) the CHM reviewed 
suspected ADRs reported in association with Cervarix but did not identify any new or serious 
risks. The CHM concluded that the balance of benefits and risks of Cervarix remained 
positive. 
 
Furthermore in June 2009 the Commission 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. The Commission advised that the available evidence did 
not support a causal association between Cervarix and CFS but recommended that the issue 
should be kept under close review. Further analyses of CFS is presented in the 
Observed/Expected analyses (section 5) and in a study using the UK Clinical Practice 
Research Datalink (CPRD)18 (section 6). 
 
In September 2010 a 2-year safety review was presented to the CHM that summarised the 
UK safety experience of the HPV vaccine following 2 years of use. The review concluded that 
after 4.5 million doses of Cervarix in the UK the vast majority of suspected ADRs were related 
either to the signs and symptoms of recognised side effects listed in the product information 
or were due to the injection process and not the vaccine itself (i.e. ‘psychogenic’ in nature). 
For the cases of other medical conditions reported, the available evidence and 
observed/expected analyses did not suggest that the vaccine caused chronic fatigue 
syndrome, facial palsy, Guillain Barré Syndrome and encephalitis and these may have been 
coincidental events. No new risks were identified in association with Cervarix despite 
significant exposure in the UK. The CHM endorsed the conclusions of the safety review and 
agreed that the balance of risks and benefits of Cervarix was positive. 
 
 
 
3. USAGE 
DATA 
 
Over 6 million doses of Cervarix have been administered in the UK since the programme 
started in 2008. While there is some variability in uptake across England, Scotland, Wales 
and Northern Ireland, the HPV immunisation programme has been considered a success with 
uptake figures among the highest in the world19.  During the first three years of the HPV 
programme, over 65% of all females born between 1 September 1990 and 31 August 1998 
completed the three-dose course, with the highest coverage achieved in the youngest 
cohorts. The proportion of UK females aged 12-13 years receiving doses one, two and three 
                                                      
18 Formerly the General Practice Research Database (GPRD) 
19 http://immunisation.dh.gov.uk/annual-hpv-vaccine-coverage-in-england-in-201011-report 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  


link to page 7 link to page 7 link to page 7 link to page 7 ranged from 85%, 83.1% and 77.5% in 2009/2010 and 88.4%, 86.6% and 80.9%20. More 
recent provisional data for this past year (1st September 2011 to 30th June 2012) show 
uptakes of 90.4%, 89.0% and 82.6% for doses one to three respectively from a cohort of 
293,721 year 8 girls aged 12-13 years of age21.  
 
Since 2010 the vast majority of vaccine has been administered to girls 12-13 years of age. 
For example in Wales 41,095 doses of a total of 44,453 doses administered were to those 
who turned 13 years of age in 2011-201222. This is reflected in the ADR data with the majority 
of cases reported in females aged 12-13 years (Table 3; section 4.1). 
 
 
 
4. 
SPONTANEOUS ADVERSE REACTION (ADR) DATA 
 
All suspected adverse reactions (ADRs) associated with Cervarix and HPV brand unspecified 
received by the Medicines and Healthcare products Regulatory Agency (MHRA) through its 
spontaneous reporting scheme, the Yellow Card Scheme, up to 31st July 2012 were included 
within the review. This included UK reports received from healthcare professionals, patients 
and marketing authorisation holders (MAHs). Given that the vast majority of HPV vaccine 
currently used in the UK is Cervarix provided by the Department of Health within the national 
HPV immunisation programme, ADR reports that were HPV brand unspecified were also 
included within the analysis. 
 
 
4.1 
Total ADR Reports 
 
Since Cervarix was first authorised in the UK the MHRA has received a total of 6213 reports 
including 14,300 reactions. The total number of reports considered to be serious23 was 1906 
which equates to 31% of the total number of ADR reports. The proportion of serious reports is 
not unexpected given that a large majority of these reports were psychogenic in nature and 
due to the injection process and not due to the vaccine per se   but classified as serious 
according to the MedDRA (Medical Dictionary for Regulatory Activities) dictionary (e.g. 
syncope). Section 4.4 of the SPC warns about psychogenic responses to vaccination (annex 
1). 
 
The total numbers of reports received by the MHRA by year and by calendar month are 
presented in Table 1 and Figure 1 respectively.  
 
 
 
                                                      
20 Annual HPV vaccine coverage in England: 2009/10 Routine programme for year 8 females (12-13 
years old) and catch-up campaign for year 10-13 females (14 -18 years old) DH HPA 22 December 
2010 
21 http://immunisation.dh.gov.uk/hpv-vac-uptake-jun12/ 
22 Data provided by Public Health Wales VPDP (data source National Community Child Health 
Database of Wales (data extracted July 2012) 
23 Seriousness defined by Council for International Organizations of Medical Sciences (CIOMS) serious 
flags (fatal, life-threatening, cause hospitalisation, result in persistent or significant disability or 
incapacity, require intervention to prevent permanent damage, or cause congenital anomalies) or the 
reporter considered the reaction to be serious. On Sentinel specific ‘alert terms’ are also considered to 
be serious  
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  


Table 1. Total number of ADR reports associated with Cervarix and  
HPV vaccine brand unspecified received by the MHRA by year 
 
 
Year received 
Number of Reports 
 
2008 1292 
2009 1912 
2010 1794 
2011 1069 
2012* 146 
 Total 
6213 
*Data are up to 31st July 2012 
 
 
 
With over 6 million doses administered the overall reporting rate is estimated to be about 1 
report per 1,000 doses administered. This reporting rate is not unexpected for a newly 
marketed vaccine used within a new national immunisation programme with such high 
vaccine exposure over a four year period.  
 
 
Figure 1. Number of ADR reports associated with Cervarix and HPV 
vaccine brand unspecified (April 2008 to July 2012)
600
500
eports 400
DR R 300
200
ber of A
100
Nu
0
1
r-08
l-08
09
11
12
n-
-09
l-09
t-09
n-10
r-10
l-10
t-10
n-
r-12
l-12
Ap
Ju
Oct-08 Ja
Apr
Ju
Oc
Ja
Ap
Ju
Oc
Jan-
Apr-11 Jul-11 Oct-1 Ja
Ap
Ju
Month
 
 
 
At the time of initial use of Cervarix within the national HPV immunisation programme 
healthcare professionals involved in the immunisation programme were contacted to 
encourage reporting of ADRs via the Yellow Card Scheme. As a new Black Triangle ▼ 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  


link to page 9 vaccine24 for use in adolescents the initial high level of ADR reporting was not unexpected. 
Exposure was greatest during the initial two years of the immunisation programme when the 
catch-up programme was used to vaccinate females 14 to 18 years in addition to those 12-13 
years of age. In recent years ADR reporting levels have fallen with only 146 reports thus far in 
2012 despite vaccination uptake remaining consistently over 80% in the routine cohort,. 
Overall the reporting levels are not considered unusually high for a new vaccine used at a 
national level.  
 
The number of ADR reports by reporter source (Table 2) highlights the large contribution 
made by nurses to the Yellow Card Scheme throughout this vaccination programme with 
nurses contributing to two thirds of all reports received by the Agency. As the main 
administrators of the vaccine in schools they were well placed to observe and then report 
ADRs to the Agency and their valuable contribution is recognised.  
 
 
Table 2. The number and percentage of ADR reports by reporter source 
 
 
 
Reporter Qualification 
Number of Reports 
Percentage 
 
of Reports 
 
UNKNOWN 

0.05 % 
HOSPITAL PHARMACIST 

0.08 % 
CARER 
10 
0.16 % 
COMMUNITY PHARMACIST 
10 
0.16 % 
PHARMACIST 
38 
0.59 % 
PATIENT 
41 
0.64 % 
HOSPITAL DOCTOR 
91 
1.42 % 
PHYSICIAN 
97 
1.52 % 
HOSPITAL HEALTHCARE 
100 
1.56 % 
PROFESSIONAL 
CONSUMER OR OTHER NON 
155 
2.42 % 
HEALTH PROFESSIONAL 
PARENT 
194 
3.03 % 
HOSPITAL NURSE 
299 
4.67 % 
GP 
399 
6.23 % 
OTHER HEALTHCARE 
681 
10.6 % 
PROFESSIONAL 
NURSE 
4280 
66.8 % 
Total: 
6403* 
100 % 
*The total number of reports is higher than 6213 as some reports contain more  
than one reporter (e.g. a parent and a healthcare professional) 
 
 
Figure 2 shows a change in trend from the 2-year review in which a greater proportion of 
electronic reports were received while recent data up to 31st July 2012 show a greater 
proportion of paper reports than electronic reports for Cervarix and HPV brand unspecified.  
 
                                                      
24http://www.mhra.gov.uk/Safetyinformation/Howwemonitorthesafetyofproducts/Medicines/BlackTria
ngleproducts/index.htm 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  


Figure 2. Number of ADR electronic and paper reports
Number of Electronic
Reports (n = 2511)
Number of Paper
Reports (n = 3702)
 
Reports received broken down by patient age are presented in Table 3. As expected the 
majority of cases were reported in individuals of the appropriate age specified within the 
national HPV vaccination programme which is currently year 8 (aged 12-13 year olds) and 
when the catch up programme was initially implemented in those <18 years of age. There 
have been 20 reports in patients <10 years of age and the vast majority of these were 
reported in those less than 2 years of age (parent-child reports). The safety of Cervarix during 
pregnancy is discussed in section 4.8 below. A single case was reported in a 7 year-old but 
this is likely to be a reporting error on behalf of the reporter.  
 
 
       Table 3. The number and percentage of ADR reports by age of the patient 
 
 
Ages 
Number of 
Percentage of 
 
Reports 
Reports  
<10 years 
20 
0.32 % 
10 Years 

0.03 % 
11 Years 
13 
0.21 % 
12 Years 
2283 
36.75 % 
13 Years 
1128 
18.16 % 
14 Years 
458 
7.37 % 
15 Years 
645 
10.38 % 
16 Years 
454 
7.31 % 
17 Years 
622 
10.01 % 
18 Years 
277 
4.46 % 
19 Years 
23 
0.37 % 
20 Years 

0.05  % 
21 Years 

0.03% 
22 Years 

0.02 % 
23 Years 

0.02 % 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
10 

24 Years 

0.02 % 
25 Years 

0.02 % 
26 Years 

0.03 % 
27 Years 

0.02 % 
30 Years 

0.02 % 
31 Years 

0.02 % 
32 Years 

0.02 % 
37 Years 

0.02 % 
53 Years 

0.02 % 
Unknown 
271 
4.36 % 
Total  
6213 
100 % 
 
 
 
The number of reports broken down by MedDRA (Medical Dictionary for Regulatory Activities) 
terminology System Organ Class is provided in Table 4. A further breakdown of the number of 
reactions by high-level term (HLT) and preferred term (PT) in each SOC are at annex 3. 
Events of interest in specific SOCs and those terms that are likely to be of interest to those 
administering or using the vaccine are described below. 
 
 
Table 4. Total number and percentage of adverse reactions by MedDRA System Organ Class 
 
 
 
 
MedDRA System Organ Class (SOC) 
No. of 
Percentage 
Adverse 
of Adverse 
Reactions 
Reactions 
 
 
Nervous system disorders 
4263 
29.81 % 
General disorders and administration site conditions 
2940 
20.56 % 
Gastrointestinal disorders 
2100 
14.69 % 
Musculoskeletal and connective tissue disorders 
1455 
10.17 % 
Skin and subcutaneous tissue disorders 
1301 
9.10 % 
Vascular disorders 
436 
3.05 % 
Respiratory, thoracic and mediastinal disorders 
370 
2.59 % 
Eye disorders 
281 
1.97 % 
Psychiatric disorders 
232 
1.62 % 
Investigations 185 
1.29 

Immune system disorders 
123 
0.86 % 
Infections and infestations 
118 
0.83 % 
Injury, poisoning and procedural complications 
95 
0.66 % 
Reproductive system and breast disorders 
72 
0.50 % 
Cardiac disorders 
68 
0.48 % 
Metabolism and nutrition disorders 
59 
0.41 % 
Ear and labyrinth disorders 
57 
0.40 % 
Blood and lymphatic system disorders 
36 
0.25 % 
Congenital, familial and genetic disorders 
33 
0.23 % 
Pregnancy, puerperium and perinatal conditions 
31 
0.22 % 
Renal and urinary disorders 
16 
0.11 % 
Surgical and medical procedures 
11 
0.08 % 
Neoplasms benign, malignant and unspecified (including cysts and polyps) 

0.06 % 
Endocrine disorders 

0.03 % 
Social circumstances 

0.02 % 
Hepatobiliary disorders 

0.01 % 
Total 
14300 
100 % 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
11 

4.2 
Reports with a fatal outcome  
 
There have been two cases with a fatal outcome associated with Cervarix since authorisation. 
Both these cases were presented in the 2-year review from September 2010. There was no 
indication that Cervarix caused or contributed to the unfortunate fatal events. No additional 
cases with fatal outcomes have been reported since the period of the review.  Case details 
are provided below. 
  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4.3 
Nervous System Disorders 
 
The SOC with the greatest number of reports (4263) was the Nervous System Disorders SOC 
(29.81%). Within this SOC the highest numbers of suspected adverse reactions include 
headache (1128), dizziness (1367), syncope (501), hypoaesthesia (251), paraesthesia (148) 
and tremor (110).These reactions or related terms are included within the product information 
(annex 1).  
 
Within the 2-year review a total of 892 reports containing 2236 reactions (19% of total reports) 
were classified as psychogenic events based on signs / symptoms of a fear or anticipatory 
response to the needle injection. This includes terms such as dizziness and nausea in the 
context of syncope, loss of consciousness/altered state of consciousness, vision disturbance 
(including transient blindness), injury, limb jerking (often misinterpreted/reported as a 
seizure/convulsion), limb numbness or tingling and difficulty in breathing. A warning to 
highlight these psychogenic reactions has been added to section 4.4 of the Cervarix SPC 
(annex 1) given that this population is particularly prone to these type of reactions.  
 
Other events of interest within this SOC include Guillain-Barré Syndrome, encephalitis and 
Bell’s palsy (VIIth nerve paralysis/facial palsy) and convulsions.  
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
12 

 
4.3.1 Guillain-Barré  Syndrome 
 
At the time of the 2-year review the MHRA had received 5 reports of Guillain-Barré Syndrome 
(GBS). These reports are detailed as follows: 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
•  A female of unspecified age was reported as developing suspected GBS and at 
unspecified time after vaccination. Despite follow-up, no further information was 
available. 
 
All 5 cases of GBS were included in the Observed/Expected analysis using MaxSPRT test 
method (annex 2). Two cases were included in the first year’s analysis (2008-2009) and three 
in the second year’s analysis (2009-2010). Given the expected background incidence of GBS 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
13 

in the vaccinated population, 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. 
 
Since the 2-year review there have been no additional cases of GBS associated with Cervarix 
or HPV unspecified. The absence of additional cases within the past two years despite a 
further 1.5 million doses administered supports the previous position that GBS is unlikely to 
be causally related with Cervarix. 
 
4.3.2 Encephalitis  
 
There have been 6 reports of encephalitis and 1 report of encephalitis lethargica reported to 
date. The first 5 cases were previously reported within the 2-year review and described 
below.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
•  A female of unknown age was reported by a pharmacist as developing encephalitis at 
an unspecified time after vaccination with Cervarix. No further details are available 
despite follow-up. 
 
•  A female of unknown age was reported by a practice nurse as developing 
encephalitis at an unspecified time after vaccination with a 3rd dose of Cervarix. No 
further details are available despite follow-up. 
 
Since then 2 additional cases have been reported: 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
14 

link to page 15  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Within the 2-year review the MaxSPRT analysis suggested the cases reported were 
consistent with chance (annex 2). Three of the cases previously reviewed refer to NMDA 
receptor antibody encephalitis. This is a form of acute encephalitis caused by an autoimmune 
reaction against NR1- and NR2- subunits of the glutamate NMDA receptor that is responsible 
for controlling memory function and synaptic plasticity. The disease is generally associated 
with tumours25 although not all cases are found to have evidence of tumours. Although the 
background incidence of anti-NMDA encephalitis, and the prevalence of anti-NMDA 
antibodies in the healthy population, is not known the available medical literature appears to 
indicate a preponderance of the condition in adolescent females (even without ovarian 
tumours).  
 
Neither of the two new cases have been directly medically confirmed as the parents of the 
children reported encephalitis in both cases. In the first case the reaction was initially 
classified as chronic fatigue syndrome and the parent provided follow-up information 
 
 
 Nevertheless the first case has 
been added for the Observed/Expected Analyses (see section 5). The second case occurred 
post May 2012 (the time cut-off for which usage data used in the analysis) and therefore was 
not included within the analysis. However details of this case are sparse and the diagnosis of 
encephalitis is questionable as the case refers to “encephalitic reaction” and this has not been 
medically confirmed. 
 
 
4.3.3  Bell’s palsy (VIIth nerve paralysis / Facial palsy) 
 
A total of 9 reports of Bell’s palsy (VIIth nerve paralysis / Facial palsy) have been reported to 
date. This includes 5 unique reports facial palsy reported in the previous 2-year review:  
 
 
 
 
                                                      
25 Dalmau J et al. Paraneoplastic Anti–N-methyl-D-aspartate Receptor Encephalitis Associated with 
Ovarian Teratoma. Ann. Neurol. Jan 2007; 61(1):25–36.
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
15 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
An additional case was not reported as facial palsy but the case details were suggestive of 
possible facial palsy and was therefore included in the 2-year analysis: 
 
 
All 5 confirmed cases of facial palsy and the one additional possible case were included in the 
previous MaxSPRT analysis (annex 2). Two cases were included in the first year’s analysis 
and four in the second year’s analysis. The results for both years did not indicate a signal for 
any reporting level in either year. 
 
Since 28th July 2010 (the cut-off date for the 2-year review) the MHRA has received an 
additional three reports: 
 
 
 
 
 
 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
16 

Two of these cases occurred within the defined time period and are included within the 
Observed/Expected analysis (section 5). 
 
Following a review of the Rapporteur’s assessment of the 7th Periodic Safety Update Report 
(PSUR), covering the period 18th November 2010 to 17th November 2011, in April 2012 the 
MHRA requested the submission of a cumulative review by the MAH of all cases of Bell’s 
palsy / VIIth nerve paralysis / facial paresis within a 3 month period from the time of the 
request. All case details including temporal associations and patient past medical histories 
were requested where available. This review is pending.  
 
 
4.3.4 Conv ulsions 
 
There have been a total of 97 seizures and seizure disorders reported including convulsion 
(74), grand mal convulsion (11), clonic convulsion (1) and other reactions. At the time of the 
2-year review there were 31 reports of convulsion and 7 reports of grand mal convulsion plus 
other reactions. The majority of cases of convulsion were reported in association with other 
reactions such as loss of consciousness / syncope or with additional conditions. Tonic clonic 
movements are listed in the SPC in the context of psychogenic events associated with 
syncope (annex 1). There is currently insufficient evidence of a causal association of Cervarix 
with convulsions that are not psychogenic in nature. Convulsions will continue to be 
monitored.  
 
 
4.3.5  Complex regional pain syndrome  
 
Up to 31st July 2012 six cases of complex regional pain syndrome have been reported: 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
17 

• 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
 
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 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
18 

link to page 19 link to page 19 link to page 19 link to page 19 person years (95% CI: 23.0-29.7)26. 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 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%)27 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.
200528, rather than the vaccine constituents. It is also possible that such reports were 
coincidental.   
 
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 well 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. Cases of CRPS will continue to be monitored. 
 
 
4.4 
General disorders and administration site conditions 
 
This was the SOC with the second highest number of reactions reported (2940 ADRs; 
20.56%) with the vast majority relating to injection site reactions (652 in HLT), fatigue (378), 
malaise (499), pyrexia (319) / feeling hot (147), peripheral oedema (229) and pain (128). The 
majority of these reactions are listed in the SPC (annex 1).  
 
Events of interest include Chronic Fatigue Syndrome (CFS) / Myalgic encephalomyelitis (ME) 
 
 
4.4.1  Chronic Fatigue Syndrome (CFS) / Myalgic encephalomyelitis (ME) 
 
Chronic Fatigue Syndrome (CFS) is diagnosed when other possible diagnoses have been 
excluded. In June 2009, CHM considered a specific safety paper which assessed reports of 
CFS and chronic fatigue in temporal association with Cervarix vaccine. At the time MHRA had 
received 1 report of CFS, 4 reports of post viral fatigue syndrome29, and two reports of 
‘chronic’ fatigue. An additional 3 cases of Chronic Fatigue-like Syndrome separately reported 
in the media were also assessed. The CHM advised that the available evidence did not 
support a causal association between Cervarix and CFS. 
                                                      
26 De Mos M et al. The incidence of complex regional pain syndrome: a population-based study. Pain 
2007; 129(1-2):12-20. 
27 Sandroni P et al. Complex regional pain syndrome type I: incidence and prevalence in Olmsted 
county, a population-based study. Pain 2003;103(1-2):199-207. 
28 Genc H et al. Complex regional pain syndrome type-I after rubella vaccine. Eur J Pain 2005; 9(5): 517-520.  
29 Myalgic encephalomyelitis codes to post viral fatigue syndrome in MedDRA 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
19 

 
At the time of the 2-year review (up to 28th July 2010) the MHRA had received 4 reports of 
CFS and 6 reports of post-viral fatigue syndrome (see annex 2). Results of the MaxSPRT 
analysis for both 2008-2009 and 2009-2010 suggested that the reported cases were 
consistent with chance and this was supported by the lack of consistent temporal association 
and clinical characteristics of the reported cases of CFS/ME. Since that time period the 
MHRA has received an additional 10 cases of CFS within the past two years presented 
below.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
20 

 
 
 
 
 
 
 
 
  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
21 

 
A total of eight cases of CFS/ME and post viral fatigue syndrome (section 4.5.1) have been 
included within the Observed / Expected Analyses as these reaction onset dates occurred 
within the defined time period (section 5).  
 
 
 
4.5 
Infections and Infestations 
 
4.5.1  Post Viral Fatigue Syndrome 
 
Post viral fatigue syndrome (PVFS) codes to the Infections and Infestations SOC. Up to 31st 
July 2012 there have been 16 reports of PVFS. Up to 28th July 2010 the MHRA had received 
7 reports of post-viral fatigue syndrome (see annex 2). 
 
As 3 of the PVFS cases describe CFS/ME (see section 4.4.1), only the remaining 6 cases are 
detailed below. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
22 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
A total of eight cases of CFS/ME (section 4.4.1) and post viral fatigue syndrome have been 
included within the Observed / Expected Analyses (section 5).  
 
 
4.6 
Immune System Disorders 
 
Within the Immune System Disorders SOC there have been a total of 123 reactions of which 
63 are reported as an anaphylactic /anaphylactoid reaction. Compared with the 2-year review 
47 cases of anaphylaxis /anaphylactoid reactions were reported at that time point with the 
majority of cases containing insufficient details to suggest true anaphylaxis as defined by the 
Brighton criteria. Even in the unlikely event that all reported cases were true anaphylaxis, a 
reporting rate of 63 cases per 6,000,000 doses administered would be consistent with broad 
estimates of vaccine-induced anaphylaxis. Allergic reactions including anaphylactic and 
anaphylactoid reactions are listed in the SPC and patients known to be hypersensitive to the 
active substance or the excipients are contraindicated from vaccine administration (annex 1).  
 
 
4.7 Neopla
sms  
 
There have been a total of 9 neoplastic reactions within the Neoplasms Benign, Malignant 
and Unspecified (Incl Cysts and Polyps) SOC including medulloblastoma, acute lymphocytic 
leukaemia, acute myeloid leukaemia, chronic myeloid leukaemia, neoplasm malignant, benign 
hydatidiform mole, salivary gland cancer stage 1, anogenital warts and skin papilloma. This 
includes a case with a fatal outcome of 1a malignant tumour of the chest (see section 4.2).  
These cases do not suggest that Cervarix is associated with an increased neoplastic risk.  
 
 
4.8 
Pregnancy and Congenital Anomalies 
 
Cervarix is not recommended during pregnancy.  During the Cervarix clinical development 
program, a total of 3,993 pregnancies were reported including 2,009 in women who had 
received Cervarix (annex 1). Overall, the proportions of pregnant subjects who experienced 
specific outcomes (e.g., normal infant, abnormal infants including congenital anomalies, 
premature birth, and spontaneous abortion) were similar between treatment groups. Animal 
studies do not indicate direct or indirect harmful effects with respect to fertility, pregnancy, 
embryonal/foetal development, parturition or post-natal development.  
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
23 

 
Within the Pregnancy, Puerperium and Perinatal conditions SOC there have been a total of 
31 ADRs reported (Table 5) and a further 33 ADRs within the Congenital, Familial and 
Genetic disorders SOC (Table 6). 
 
 

Table 5. Pregnancy, Puerperium and Perinatal conditions SOC 
 
 
 
Reaction Preferred Terms 
Number 
 
Abortion spontaneous 
14 
Foetal distress syndrome 

Foetal growth restriction 

Premature baby 

Premature labour 

Labour complication 

Live birth 

Foetal death 

Pregnancy with injectable contraceptive  

Total 
31 
 
 
Table 6. Congenital, Familial and Genetic Disorders SOC 
 
 
 
Reaction Preferred Terms 
Number 
 
PRETERNATURAL ANUS 

ATRIOVENTRICULAR SEPTAL DEFECT 

CEREBRAL PALSY 

CONGENITAL ANOMALY 

VACTERL SYNDROME 

DOUBLE OUTLET RIGHT VENTRICLE 

HYPOSPADIAS 

LIMB HYPOPLASIA CONGENITAL 

LIMB REDUCTION DEFECT 

TALIPES 

SPINE MALFORMATION 

CLEFT PALATE 

HIGH ARCHED PALATE 

PULMONARY APLASIA 

PULMONARY HYPOPLASIA 

CONGENITAL CYSTIC KIDNEY DISEASE 

BIRTH MARK 

ICHTHYOSIS 

PILONIDAL CYST CONGENITAL 

SEBACEOUS NAEVUS 

ANKYLOGLOSSIA CONGENITAL 

ARTERIOVENOUS MALFORMATION 

Total 
33 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
24 

link to page 25  
There is no clear pattern to suggest that Cervarix use during pregnancy results in a specific 
congenital abnormality or is associated with a high risk of spontaneous abortion. This is 
supported by pregnancy outcome data30 collated by both the Health Protection Agency (HPA) 
using their Vaccines in Pregnancy (VIP) registry and the UK Teratology Information Service 
(UKTIS). 
 
In the recent renewal procedure the MAH presented their worldwide pregnancy outcome data 
during post-marketing (Table 7).  
 
                                                      
30 Unpublished data  
31 MAH data extracted from Rapporteurs’ Joint Preliminary Assessment Report – Cervarix 
EMEA/H/C/721/R/35 dated 23rd May 2012 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
25 

 
GSK will be conducting a post-marketing safety study to assess the risk of spontaneous 
abortions during the first 23 weeks of gestation in women aged 15 to 25 years exposed to 
Cervarix in the UK using the CPRD.  
 
The inadvertent use of Cervarix during pregnancy will continue to be closely monitored.  
 
 
5. 
OBSERVED / EXPECTED ANALYSES 
 
As part of the enhanced pharmacovigilance strategy and presented within the 2-year review, 
age-specific and gender-specific background incidence rates for a range of ‘events of interest’ 
were calculated on a weekly basis using 10 years of historical data from GPRD. Incidence 
rates were used to estimate the expected number of reports on a continuous cumulative 
basis. A cumulative relative risk was calculated each week based on the cumulative sum of 
certain ADR reports received as the ‘observed’ and the cumulative incidence rate and the 
cumulative exposure deriving the ‘expected’. A sensitivity analysis adjusted the expected 
number of reports for varying degrees of under-reporting. 
 
The statistical sequential test method, the Maximised Sequential Probability Ratio Test 
(MaxSPRT), was used to compare the number of reported cases of suspected side effects of 
interest against the normal rates of such illnesses that are expected to occur by chance in the 
vaccinated age groups to determine if the vaccine may carry any excess risks. These 
analyses adjust for various levels of possible under-reporting through the Yellow Card 
Scheme. The method signals when the observed number of reports exceeds the expected 
based on a critical value derived from the poisson distribution. Sequential methods are useful 
to adjust for the multiple testing that occurs with weekly surveillance. ‘Observed vs expected’ 
analyses were conducted weekly over two 1 year periods 2008/2009 and 2009/2010 (see 
annex 2). 
 
5.1 
Observed vs. Expected analyses for 2010/11 and 2011/12 
 
Analyses were conducted for Guillain-Barré syndrome, Bell’s / facial palsy, chronic fatigue 
syndrome / post viral fatigue syndrome and encephalitis for the academic years 2010/11 and 
2011/12. The latest available exposure data (number of girls vaccinated) was available up to 
31st May 2012 for England, March 2012 for Wales and between June-Aug 2011 for Scotland 
and Northern Ireland. 
 
Cases received between September 2010 and 31st July 2012 were included within the 
analyses if the reaction onset date was between these dates. 
 
There were a few retrospective cases of chronic fatigue syndrome and Bell’s palsy for which 
the reaction dates were in the timeframes that have been documented before i.e. 2008-09 
and 2009-10. However these cases were not included in the analysis. 
 
5.1.1  Chronic fatigue syndrome/ME and Post viral fatigue syndrome 
 
From September 2010 to August 2011 there were 5 cases of CFS reported in girls aged 12- 
13 years. There were 3 cases reported during September 2011 and May 2012. No signals 
were observed for any of the reporting levels (Figures 3 and 4). 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
26 

 
Figure 3 
Maximised SPRT for ME/Chronic Fatigue Syndrome for girls aged 12-13 years (2010-2011)
4.5
4
3.5
3
io
 Rat 2.5

od
Critical value
iho
10% events reported
ikel
2
25% events reported
g L
50% events reported
Lo
75% events reported
1.5
100% events reported
1
0.5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
Week of Surveillance
 
 
 
 
Figure 4 

Maximised SPRT for ME/Chronic Fatigue Syndrome for girls aged 12-13 years (2011-2012)
4.5
4
3.5
3
tio
2.5
Critical value
10% events reported
kelihood Ra 2
25% events reported
 Li
50% events reported
Log
75% events reported
1.5
100% events reported
1
0.5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
Week of Surveillance
 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
27 

5.1.2  Bell’s palsy/VIIth nerve paralysis 
 
There were two cases of Bell’s palsy / VIIth nerve paralysis reported during Sept 2011 and 
May 2012 and the analysis does not indicate a signal for any reporting level (Figure 5). 
 
Figure 5 
Maximised SPRT for Facial palsy for girls aged 12-13 years (2011-2012)
4.5
4
3.5
3
atio
R 2.5

Critical value
ihood 
10% events reported
kel
2
25% events reported
50% events reported
Log Li
75% events reported
1.5
100% events reported
1
0.5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
Week of Surveillance
 
 
 
 
 
5.1.3 Encephalitis

 
 
There was one case of encephalitis reported during September 2010 and August 2011 and 
the analysis does not indicate a signal for any reporting level (figure 6). No events were 
reported between September 2011 and May 2012. 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
28 

Figure 6.  
Maximised SPRT for Encephalitis for girls aged 12-13 years (2010-2011)
4
3.5
3
Critical value
10% events reported
25% events reported
50% events reported
2.5
75% events reported
tio
100% events reported
 Ra
od 2
keliho
g Li
Lo 1.5

1
0.5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
Week of Surveillance
 
 
 
Conclusion on observed vs expected analyses 
 
Previous incidence rates calculated for the 2-year review were used to estimate the expected 
number of reports. Unlike the 2-year review which calculated cumulative relative risks each 
week for a 2 year period these data were calculated retrospectively to provide ‘snapshot’ 
Observed / Expected analyses. While the conclusions that can be drawn from these data are 
limited as a number of assumptions have been made, the data do not suggest that any of the 
events of interest are greater than the critical threshold. Therefore the number of observed 
cases is not greater than expected and no new safety signals have been identified for the 
events of interest Guillain-Barré syndrome, Bell’s / facial palsy, chronic fatigue syndrome / 
post viral fatigue syndrome and encephalitis using this methodology during the periods 
2010/2011 and 2011/2012.   
 
 
 
 

Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
29 

 
6. 

UK  Clinical Practice Research Datalin k (CPRD)  Study: Hum an Papillomavirus 
vaccine and the risk of fatigue syndrome in children in the UK  

 
6.1 Purpose 
 
This section of the paper considers the results of a study which uses the UK Clinical Practice 
Research Datalink (CPRD, formerly the General Practice Research Database or GPRD) to 
investigate the potential association between human papillomavirus (HPV) vaccination and 
the occurrence of fatigue syndrome in girls in the UK. The aim is to provide information on 
any potential association prior to the switch in the UK from the use of Cervarix in the national 
vaccination program to the alternative product Gardasil.  
 
6.2 Bac
kground 
 
The UK HPV immunisation programme commenced in September 2008, with routine 
immunisation of 12-13 yr old females and an initial catch-up of females aged up to 18 years. 
The vaccine programme has been very successful. During the first three years of the 
programme, over 65% of all females born between 1 September 1990 and 31 August 1998 
completed the three-dose course. The highest coverage has been achieved in the youngest 
cohorts, with more than 84% of 12- to 13-year-old females completing the three-dose course. 
More than 5 million doses were administered up to December 2011. 
 
6.3 CPRD 

study 
 
6.3.1 Objectiv es 
 
The aim of this study is to investigate the potential association between HPV vaccination and 
the occurrence of fatigue syndrome in females in the UK using two different approaches: 
 
1. To describe trends in the recording of fatigue diagnoses in 12 to 20 years old girls in the 
CPRD before and after the introduction of the vaccination campaign (Ecological study). 
 
2. To carry out a self-controlled case series (SCCS) study to investigate the temporal 
relationship between HPV vaccination and fatigue diagnoses in females in the CPRD.  
 
6.3.2 Metho ds 
 
Ecological study 
 
Patients with a clinical diagnosis of fatigue syndrome (annex 4 - Appendix A) were extracted 
from the Clinical Practice Research Datalink (CPRD) general practice database, in March 
2012. Data up to 31 December 2011 were included. An incident diagnosis was defined as the 
first recorded clinical diagnosis per acceptable patient during active follow up in an up-to-
standard practice. 
 
The incidence of fatigue syndromes, in girls aged 12-20 years, was calculated overall, by 
category of first diagnosis (chronic fatigue syndrome, post viral fatigue syndrome, 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
30 

fibromyalgia, and neuralgia). Where missing, month of birth was randomly assigned. The 
incidence rate of diagnosis is presented by year 2000-2011.  
 
Poisson regression was used to compare trends in incidence rates before (2006-2007) and 
after (2009-2011) the introduction of the HPV vaccination campaign.  
 
Sensitivity analyses examining the incidence in adults aged 21+ years and boys aged 12-20 
years were conducted using the same inclusion criteria as before. In addition, a further 
sensitivity analysis, conducted only in girls aged 12-20 years, including referrals for fatigue 
syndromes and symptoms (annex 4 - Appendix B), as well as records of clinical diagnoses, 
was conducted. Again only records occurring in acceptable patients during active follow up in 
an up-to-standard practice were included.  
 
Self-controlled case series study 
 
Self-controlled case series (SCCS) methodology was developed to investigate associations 
between acute outcomes and transient exposures, using only data on cases. In a case series 
analysis only cases are sampled with the likelihood of an event occurring in a pre-defined risk 
period compared to other non-risk periods within the same individual. All fixed covariates are 
therefore implicitly controlled for while age or other temporal variation can be adjusted for in 
the model. In brief, an overall study time-window, usually defined by age and calendar time 
boundaries (but also in terms of vaccination date), is chosen. Then, all individuals with an 
event within this study time-window are identified. The vaccination dates of each case are 
then used to define one or more risk periods, during which individuals are hypothesized to be 
at increased (or reduced) risk of the event of interest after or before vaccination. All other time 
within an individual’s observation period, that does not fall within a risk period, is included in 
that individual’s control period, which together form the study baseline. Estimation of 
parameters in the SCCS method is achieved by fitting a conditional Poisson regression 
model. 
 
Girls with a record of a first vaccination (as a clinical or immunisation record, see Appendix C 
of annex 4) followed by an incident diagnosis of a fatigue syndrome, occurring when in active 
follow up in an up-to-standard practice,  were included in the main analysis. Note that 
vaccination is, in general, conducted by school nurses and then reported to the GP. In each 
SCCS analysis, only girls with a recorded first vaccination (as recorded explicitly using the 
correct clinical/immunisation code in Appendix C of annex 4) were included with this date 
used as the index date for the main analysis. Patients with any diagnosis (including symptoms 
and referrals) prior to their first vaccination were excluded. Patients with less than one year of 
active follow up, in their most recent registration period, in an up-to-standard practice prior to 
their first vaccination were also excluded to try and ensure identification of incident cases. 
The risk window was defined to start the day after the date of first vaccination and continue 
for one year. The index date for the primary analysis was the date of the first clinical diagnosis 
of fatigue syndrome. Time before vaccination was not considered unexposed and was 
excluded from the analysis, as the occurrence of fatigue syndrome may affect the likelihood of 
vaccination, so the time from one year post first vaccination to the date of last follow up was 
considered as the not at risk period. Age, in years, was adjusted for as a discrete time varying 
covariate. All cases identified in the main analysis were investigated in further detail. 
Summaries of the time from vaccination to first diagnosis are presented as well as the dates 
of diagnosis.  
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
31 

 
The first sensitivity analyses included cases with a prescription record for a HPV vaccination 
(annex 4 - Appendix C) and not an immunisation date, assuming the prescription date is the 
same as the vaccination date. Given that HPV vaccination would not be prescribed in the 
normal sense the accuracy of this assumption may be doubted hence they are not included in 
the main analysis. The index date of first clinical diagnosis remained the same. A second 
sensitivity analysis, including first referrals for, and symptoms of, fatigue syndromes as well 
as clinical diagnoses was also considered. The index date in this second sensitivity was 
therefore defined as the first record of symptoms, referral, or diagnosis occurring when in 
active follow up in an up-to-standard practice. Again, patients with any diagnosis (including 
symptoms and referrals) prior to their first vaccination were excluded. Finally, a further 
analysis using a shorter risk window of 6 months was considered defining the not at risk 
period as all time from 6 months post date of vaccination to the date of last follow up as the 
not at risk period.  
 
6.3.3 Resul ts 
 
Ecological study 
 
1,294 first clinical diagnoses of fatigue syndromes were identified in girls aged 12-20 years 
between 2000 and 2011 in the CPRD. Figure 6.1 shows the rate of first diagnoses of fatigue 
syndromes in girls aged 12-20 years by year.  
 
 
Figure 6.1: Rate of first diagnoses of fatigue syndromes in girls aged 12-20 years 2000-
2011. 

80
70
60
s
yr 
50
000 p
40
e per 100, 30
at
R

20
10
0
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Year
 
 
From 2003 to 2005 there was a steep decline in the rate of diagnosis of fatigue syndromes in 
this group. From 2006 onwards the rate of diagnosis appears to have plateaued with perhaps 
a small further temporary decline in 2010. Comparing the incidence rate of diagnoses in 
2009-2011 to 2006-2007 results in an incident rate ratio of IRR=0.94 (95% confidence 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
32 

interval: 0.78-1.14, two-sided p=0.52) suggesting that there has been no change in the 
incidence of fatigue syndromes in girls aged 12-20 years since the introduction of the HPV 
vaccination. 
 
Figure 6.2 shows the rate of incident diagnoses of fatigue syndromes in girls aged 12-20 
years by year stratified by the type of diagnosis. It can be seen that the decrease in diagnosis 
of fatigue syndromes between 2003 and 2005 was driven by a steep decline in the diagnosis 
of, firstly, post viral fatigue syndrome (PVFS) and, secondly, neurasthenia. There was no 
obvious trend in the diagnosis of chronic fatigue syndrome or fibromyalgia during the same 
period. Again, when comparing the rate of diagnosis for each category of fatigue syndrome 
2006-2007 with 2009-2011, there has been no change in the rate of diagnosis of chronic 
fatigue syndrome, fibromyalgia, or PVFS (p>0.2). However, there has been a very significant 
decrease in the diagnosis of neurasthenia (IRR 0.08, 95% CI: 0.02-0.24, p<0.001). 
 
 
Figure 6.2: Rate of first diagnoses of fatigue syndromes in girls aged 12-20 years 2000-
2011 by type of diagnosis. 

40
35
30
s
yr 
25
000 p
20
 100,
er 
15
e p
at
10
5
0
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Year
CFS
Fibromyalgia
PVFS
Neurasthenia
 
 
It is clear that changes in recording practices have occurred over the study period, therefore, 
it is important to examine these trends in both boys aged 12-20 years and adults aged 21+ 
years to check that a decrease in the recording or making of a diagnosis is not masking an 
increase in the incidence of true diagnoses.  
 
Figure 6.3, showing the analysis from the first sensitivity analysis, shows that the decrease in 
recording seen in girls aged 12-20 years between 2003 and 2005 was also seen across the 
rest of the patient population. As with the girls, a non-significant decrease in the incidence of 
fatigue syndromes has been seen in adults (0.96, 0.93-1.01, p=0.11). However, in boys aged 
12-20 there has been a significant decrease in the diagnosis of fatigue syndromes (2009/11 
vs. 2006/7, IRR: 0.66, 95% CI: 0.50-0.87, p=0.002). 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
33 

Figure 6.3: Sensitivity analysis I - Rate of first diagnoses of fatigue syndromes 2000-
2011 by age and gender.
 
100
90
80
70
60
00,000 pyrs 50
er 1 40
 p
te
Ra 
30
20
10
0
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Year
Girls aged 12-20
Boys aged 12-20
Adults 21+ years 
 
 
Figure 6.4 shows the results from the second sensitivity analysis, in girls 12-20 years, where 
referrals for fatigue syndromes and symptoms are also included. 
 
 
Figure 6.4: Sensitivity analysis II - Rate of first diagnoses of fatigue syndromes 2000-
2011 including symptoms and referrals. 

120
100
s
yr 
80
000 p
60
per 100,

at 
40
R
20
0
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Year
Clinical diagnoses only 
Sensitivity analysis including symptoms and referrals
 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
34 

The decrease in clinical diagnoses 2003 to 2005 was also seen when including symptoms 
and referrals (Figure 6.4). However, beyond that a further reduction has been seen (2009/11 
vs. 2006/7 IRR: 0.77, 95% CI: 0.66-0.91, p<0.001) not seen when only considering clinical 
diagnoses. 
 
Self-controlled case series study 
 
In the main analysis 89 girls with an incident clinical diagnosis of a fatigue syndrome after 
their recorded first vaccination with the HPV vaccine were identified according to the stated 
inclusion criteria. 78 (93%) and 57 (78%) were after a further second and third vaccinations 
respectively although dates of second and third vaccinations were recorded in only 84 and 73 
of the girls.  
 
The distributions of time from vaccination to and diagnosis and dates of diagnosis are shown 
in Figures 6.5 and 6.6 respectively.  
 
 
Figure 6.5: Time from first vaccination to diagnosis for 89 cases 
 
15
10
 cases
 of
ber
um
N
5
0
0
3
6
9
12
15
18
21
24
27
30
33
36
39
Months from first vaccination to diagnosis
 
 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
35 

Figure 6.6: Dates of diagnosis of fatigue syndrome for 89 cases 
15
10
 cases
 of
ber
um
N 5
0
08
09
09
09
09
10
10
10
10
11
11
11
11
 20
 20
 20
 20
 20
 20
 20
 20
 20
 20
 20
 20
 20
4
1
2
3
4
1
2
3
4
1
2
3
4
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Date of diagnosis
 
 
The median (IQR) time, from date of first vaccination to diagnosis of fatigue syndrome was 
13.9 (6.9-22.2) months. 38 cases were diagnosed in the one year risk window. There was a 
peak in the number of cases identified in Q4 2010 and Q1 2011 although this could be due to 
either an increase in fatigue syndrome or an increase in recording of vaccination at that time. 
Patients were followed for a median 2.3 (IQR: 2.0-3.1) years following the date of first 
vaccination.  
 
59 (66%) cases had no recorded symptoms or a referral to a specialist prior to the first 
diagnosis of fatigue syndrome. Only 29 had symptoms of tiredness recorded prior to their first 
diagnosis with a median (IQR) time of 5.9 (1.9-8.6) months between first symptoms and 
diagnosis. Nine had a specialist referral a median (IQR) 8.3 (3.4-22.9) weeks prior to 
diagnosis. Also, of note, a further 10 girls had a first referral a median (IQR) 3.4 (0.1-25.1) 
weeks after their first recorded diagnosis date. 
 
The conditional Poisson regression model estimated a non-significant decreased risk of 
fatigue syndrome following HPV vaccination (IRR: 0.87, 95% CI: 0.43-1.72, p=0.68). As 
expected, an increasing risk with age was also seen (full results not presented).  
 
Several sensitivity analyses were conducted. The first, including prescription records for HPV 
vaccination (95 cases in total), also showed a non-significant decreased risk (IQR: 0.82, 95% 
CI: 0.42-1.59, p=0.55).   
 
Similarly, the second, where the index date was defined as the first symptoms of, referral for, 
or diagnosis of a fatigue syndrome also estimated a non-significant decreased risk of fatigue 
syndrome within one year of first HPV vaccination (IRR: 0.77, 95% CI: 0.39-1.51, p=0.45). 92 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
36 

patients were included in this analysis (3 had no clinical diagnosis recorded so were not 
included in the main analysis) with a median 14.0 (IQR: 7.1-23.0) months between 
vaccination and event date.  
 
Finally, a non-significant increased risk is observed in the main analysis if the risk period is 
shortened to 6 months (IQR: 1.13, 95% CI: 0.56-2.25 p=0.74). In this model, 21 events 
occurred in the risk window. 
 
6.3.4 Discus sion 
 
The first section of this study, the ecological analysis, showed no difference in the incidence 
of fatigue syndrome in girls aged 12-20 years in 2009-2011 compared to the years 
immediately before the introduction of the HPV vaccination (2006-2007). However, a trend 
was observed of decreasing diagnosis in the earlier years 2003-2005. A similar pattern was 
seen in adults aged 21+ years. However, in boys aged 12-20 this trend was seen to continue 
with a further decrease in diagnosis in 2009-2011 compared to 2006-2007. This ecological 
analysis does not suggest the possibility of a large increased risk but it does indicate a need 
for further controlled analyses.  
 
Before commencing with the self controlled case series the quality of the data in the CPRD to 
examine the association of HPV vaccine with an increased risk of fatigue syndromes was 
considered. The GP data in the CPRD has been extensively used for epidemiological studies 
of all kinds and several self controlled case studies. However, to date, there are no 
publications using the database to examine HPV vaccination. The completeness of the HPV 
vaccination data is of some concern as only 30-40% of girls have any record of a HPV 
vaccination (either first, second, or third dose) despite accurate data collected by the NHS 
showing generally more than 80% of girls receive the full three doses (and an even greater 
percentage receiving at least one dose). In general, we might assume that the data is missing 
completely at random, meaning that results from the SCCS are unbiased, although it may be 
dependent on other factors such as medical history, geographical differences in how the 
vaccine programme is delivered, or media attention which could lead to selection biases. 
Additionally, there may also be a concern around the accuracy of the recorded vaccination 
dates given the delay in reporting by the school nurse to the GP.  
 
Several complications arise in the studying of fatigue syndromes. The insidious onset of the 
conditions can lead to recall bias regarding the onset of symptoms. Moreover, the diagnosis 
of fatigue syndromes is usually one of elimination meaning that the diagnoses seen here, 
prior to referral to a specialist, may be premature and unconfirmed. However, the self 
controlled case series provided no evidence of an increased risk of fatigue syndrome.   
 
As with all observational studies of this type, an increased risk of fatigue syndrome with HPV 
vaccination cannot be ruled out, particularly due to the high level of missing exposure data 
and the difficult nature of the diagnosis. However, it can be concluded that this study found no 
evidence of an association between HPV vaccine and fatigue syndromes based on the 
available data within the CPRD. 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
37 

link to page 38  
7. DISCUSSIO

 
In September 2008 Cervarix was first routinely used in the UK for the prevention of 
premalignant cervical lesions and cervical cancer in female adolescents in a national HPV 
immunisation programme. The vaccine is expected to prevent up to 400 deaths due to 
cervical cancer each year in the UK. Since Cervarix was introduced its safety has been 
closely monitored through an enhanced pharmacovigilance strategy and recently in a more 
routine manner. When a vaccine is administered to so many people over a relatively short 
time period, it is inevitable that some vaccine recipients will develop medical conditions not 
long after vaccination. For such conditions that also occur naturally in the absence of 
vaccination, their occurrence shortly after vaccination does not necessarily mean that the 
vaccine caused the condition. Previous reviews of the safety of Cervarix concluded that the 
balance of benefits and risks of the HPV vaccine were positive.  
 
With the imminent switch from Cervarix to Gardasil within the HPV immunisation programme 
this report summarises the safety experience of Cervarix in the UK including consideration of 
usage data, suspected ADRs received through the Yellow Card Scheme with a closer review 
of events of interest using Observed/Expected analyses. These data are supplemented with a 
CPRD study that specifically investigated the potential association between HPV vaccination 
and fatigue syndromes in girls in the UK.  
 
It is estimated that over 6 million doses of Cervarix have been administered in the UK since 
authorisation in 2007. While there is some variability in uptake across the UK, the HPV 
immunisation programme has been considered a success with uptake figures generally >80% 
across all doses administered and these are among the highest uptake figures in the world32.  
During the first couple of years of the programme there was greater usage of the vaccine with 
a catch-up campaign for older girls between 13 and 18 years of age in addition to school year 
8 (aged 12 to 13 years). In recent years the vast majority of vaccine is administered to girls 
12-13 years of age. 
 
Since Cervarix was first authorised in the UK up to 31st July 2012 the MHRA has received a 
total of 6213 reports including 14,300 reactions. With over 6 million doses administered the 
overall reporting rate is estimated to be about 1 report per 1,000 doses administered. Overall 
the reporting rate is not unexpected for a newly marketed vaccine used within a novel national 
immunisation programme. In recent years ADR reporting levels have fallen despite continued 
high use. Two thirds of the reports originated from nurses and as the main administrators of 
the vaccine their valuable contribution to the Yellow Card Scheme is recognised.  
 
Approximately 31% of all reports received were coded as serious. The proportion of serious 
reports is not unexpected given that a large majority of these reports were psychogenic in 
nature and due to the injection process and not due to the vaccine per se  but classified as 
serious according to the MedDRA dictionary. Section 4.4 of the SPC warns about 
psychogenic responses to vaccination as this population is particularly prone to these events 
(annex 1). Reporting of serious ADRs may also have been influenced by the MHRA’s 
communication to healthcare professionals at the start of the HPV vaccine programme to 
encourage the use of the Yellow Card Scheme.  
 
                                                      
32 http://immunisation.dh.gov.uk/annual-hpv-vaccine-coverage-in-england-in-201011-report 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
38 

Since authorisation there have been two reports with fatal outcomes associated with Cervarix 
and both of these were presented previously in the 2-year review from September 2010. 
There was no indication that the vaccine caused or contributed to the Streptococcal sepsis or 
malignant neoplasm reported in these cases.  
 
Nervous system disorders was the SOC with the greatest number of ADRs (n=4263, 29.81%) 
with the majority of reactions listed in the SPC or considered psychogenic events based on 
signs / symptoms of a fear or anticipatory response to the needle injection.  
 
As part of the enhanced pharmacovigilance strategy and presented ‘events of interest’ 
including Guillain-Barré Syndrome (GBS), encephalitis, Bell’s palsy (VIIth nerve 
paralysis/facial palsy) and chronic fatigue syndrome / post viral fatigue syndrome were 
analysed using Observed / Expected analyses and MasSPRT within the 2-year review (annex 
2). New ‘snapshot’ Observed / Expected analyses were conducted for these events of interest 
for the academic years 2010/11 and 2011/12 (section 5).  
 
Guillain-Barré Syndrome 
At the time of the 2-year review the MHRA had received 5 reports of GBS. Since then there 
have not been any additional cases of GBS associated with Cervarix or HPV brand 
unspecified. The absence of additional cases within the past two years despite a further 1.5 
million doses administered supports the previous position that GBS is unlikely to be causally 
related with Cervarix. 
 
Bell’s palsy (VIIth nerve paralysis / Facial palsy) 
A total of 9 reports of Bell’s palsy (VIIth nerve paralysis / Facial palsy) have been reported to 
date. This includes 5 unique reports facial palsy reported in the previous 2-year review. There 
were two cases of Bell’s palsy / VIIth nerve paralysis reported during Sept 2011 and May 
2012 and the MaxSPRT analysis does not indicate a signal for any reporting level (Figure 5). 
In April 2012 the MHRA requested the submission of a cumulative review by the MAH of all 
cases of Bell’s palsy / VIIth nerve paralysis / facial paresis within a 3 month period from the 
time of the request. All case details including temporal associations and patient past medical 
histories were requested where available. This review is pending.  
 
Encephalitis 
There have been 6 reports of encephalitis and 1 report of encephalitis lethargica reported to 
date. The first 5 cases were previously reported within the 2-year review and the MaxSPRT 
analysis suggested the cases reported were consistent with chance (annex 2). There was one 
case of encephalitis reported during September 2010 and August 2011 and the analysis does 
not indicate a signal for any reporting level (Figure 6). Three of the cases previously reviewed 
refer to NMDA receptor antibody encephalitis a form of acute encephalitis caused by an 
autoimmune reaction against NR1- and NR2- subunits of the glutamate NMDA receptor. 
Although the background incidence of anti-NMDA encephalitis, and the prevalence of anti-
NMDA antibodies in the healthy population, is not known the available medical literature 
appears to indicate a preponderance of the condition in adolescent females (even without 
ovarian tumours).  
 
Chronic Fatigue Syndrome / Myalgic encephalomyelitis / Post viral fatigue syndrome 
Chronic Fatigue Syndrome (CFS) / Myalgic encephalomyelitis (ME) has been extensively 
reviewed. In June 2009, CHM considered the risk of CFS and chronic fatigue with Cervarix. 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
39 

link to page 40 link to page 40 The CHM advised that the available evidence did not support a causal association between 
Cervarix and CFS. At the time of the 2-year review (up to 28th July 2010) the MHRA had 
received 4 reports of CFS and 6 reports of post-viral fatigue syndrome. Results of the 
MaxSPRT analysis for both 2008-2009 and 2009-2010 suggested that the reported cases 
were consistent with chance and this was supported by the lack of consistent temporal 
association and clinical characteristics of the reported cases of CFS/ME (annex 2). Using the 
same conservative approach as within the 2-year review, cases of both CFS/ME and post 
viral fatigue syndrome were included within the Observed / Expected Analyses (section 5). 
Since the time of the 2-year review the MHRA has received an additional 10 cases of CFS 
and 9 cases of PVFS within the past two years. With 3 of the cases describing both CFS/ME 
and PVFS, a total of eight cases of CFS/ME and post viral fatigue syndrome were included 
within the Observed / Expected Analyses with reaction onset dates occurring within the 
defined time period (section 5). During September 2010 to August 2011 there were 5 cases of 
CFS reported in girls aged 12-13 years (Figure 3). There were 3 cases reported during 
September 2011 and May 2012 (Figure 4). No signals were observed for any of the reporting 
levels. 
 
CFS is a fairly common condition that can occur naturally among adolescent girls. It is 
estimated that 250,000 people in Britain are affected by CFS33. There is no diagnostic 
laboratory test or biomarker for CFS and it is often diagnosed when other possible diagnoses 
have been excluded. A recent study34 in children aged 11 to 16 years (n=2855) enrolled in 
three English state secondary schools (two mixed gender and one girls-only) found that 1% of 
enrolled children (28 of 2855 pupils) missed ≥ 20% of school because of CFS. The study was 
conducted prior to initiation of the HPV immunisation programme from September 2007 to 
February 2008 and therefore provides evidence of CFS occurrence in adolescents in the 
absence of HPV vaccination. CFS will continue to be closely monitored.  
 
CPRD Study 
The MHRA recently conducted a study using the CPRD (section 6). The first aim of this study 
was to investigate the potential association between HPV vaccination and the occurrence of 
fatigue syndrome in females in the UK with an ecological study that described trends in the 
recording of fatigue diagnoses in 12 to 20 years old girls in the CPRD before and after the 
introduction of the vaccination campaign. The ecological analysis showed no difference in the 
incidence of fatigue syndrome in girls aged 12-20 years in 2009-2011 compared to the years 
immediately before the introduction of the HPV vaccination (2006-2007). A trend was 
observed of decreasing diagnosis in the earlier years 2003-2005 and a similar pattern was 
seen in adults aged 21+ years. However, in boys aged 12-20 this trend was seen to continue 
with a further decrease in diagnosis in 2009-2011 compared to 2006-2007. The ecological 
analysis does not suggest the possibility of a large increased risk but it does indicate a need 
for further controlled analyses.  
 
The second part of the study was a self-controlled case series (SCCS) study that investigated 
the temporal relationship between HPV vaccination and fatigue diagnoses in females in the 
CPRD. As with any similar type of study, the SCCS could not rule out an increased risk of 
fatigue syndrome with HPV vaccination, mainly to the high level of missing exposure data with 
                                                      
33 http://www.meassociation.org.uk/?p=1001 
34 Crawley EM et al. Unidentified Chronic Fatigue Syndrome/myalgic encephalomyelitis (CFS/ME) is a 
major cause of school absence: surveillance outcomes from school-based clinics. BMJ Open 
2011;1(2):e000252 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
40 

only 30-40% of girls having a record of HPV vaccination while the uptake rates are known to 
be generally >80% (section 3) and the difficult nature of the diagnosis. Overall there was no 
evidence of an association between HPV vaccine and fatigue syndromes based on the 
available data within CRPD.  
 
Complex regional pain syndrome 
Up to 31st July 2012 six cases of complex regional pain syndrome have been reported. 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 and CRPS 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 rather 
than the vaccine constituents. A ‘snapshot’ Observed / Expected method was used to calculate 
Observed / Expected ratios (section 4.3.5) and while this crude analysis is subject to limitations, 
the observed number of events in 12-18 year old girls administered Cervarix was found to be well 
below the expected number of events based on published incidence rates not restricted to 12-18 
year olds. Cases of CRPS will continue to be monitored. 
 
Use during pregnancy 
Cervarix is not recommended during pregnancy. During the Cervarix clinical development 
program, a total of 3,993 pregnancies were reported including 2,009 in women who had 
received Cervarix. However the proportions of pregnant subjects who experienced specific 
outcomes (e.g., normal infant, abnormal infants including congenital anomalies, premature 
birth, and spontaneous abortion) were similar between treatment groups. There have been a 
total of 31 ADRs reported within the Pregnancy, Puerperium and Perinatal conditions SOC 
(Table 5) and a further 33 ADRs reported within the Congenital, Familial and Genetic 
disorders SOC (Table 6). There is no clear pattern to suggest that Cervarix use during 
pregnancy results in a specific congenital abnormality or is associated with a high risk of 
spontaneous abortion. This is supported by pregnancy outcome data collated by both the 
Health Protection Agency (HPA) using their Vaccines in Pregnancy (VIP) registry and the UK 
Teratology Information Service (UKTIS). To further investigate the risk of spontaneous 
abortion GSK is conducting a post-marketing safety study to assess this risk during the first 
23 weeks of gestation in women aged 15 to 25 years exposed to Cervarix in the UK using the 
CPRD. The inadvertent use of Cervarix during pregnancy will continue to be closely 
monitored.  
 
 
8. CO
NCLUSIONS 
 
Cervarix has been subject to enhanced pharmacovigilance since it was first introduced for 
routine use in the national HPV immunisation programme in September 2008. The need to 
quickly identify potential new safety signals, and to minimise the risk of coincidental events 
being attributed to the vaccine, was recognised prior to use. To take this forward the MHRA 
introduced an enhanced pharmacovigilance strategy with intensive monitoring and effective 
communications. Using the Observed/Expected analyses and MaxSPRT events of interest 
including Guillain-Barré Syndrome (GBS), encephalitis, Bell’s palsy (VIIth nerve 
paralysis/facial palsy) and chronic fatigue syndrome / post viral fatigue syndrome were 
previously shown to occur at a level less than expected within the background population. 
Overall the enhanced pharmacovigilance strategy endorsed by the CHM has been a success.  
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
41 

 
Despite significant usage, with over 6 million doses administered in the UK, the number of 
suspected ADRs received by the MHRA is not unexpected with the majority recognised 
adverse reactions of Cervarix or psychogenic in nature and related to the vaccination 
procedure rather than the vaccine. Using the Observed/Expected analyses no new signal has 
been detected for the events of interest in the past two years. A recent CPRD study found no 
evidence of an association between HPV vaccine and fatigue syndromes based on the 
available data within CPRD. No other new signals have been identified that warrants 
regulatory action. 
 
The safety experience of Cervarix up to 31st July 2012 supports the previous conclusion that 
the benefit/risk balance of Cervarix remains positive. As with all medicines and vaccines the 
safety of Cervarix will continue to be monitored by the MHRA. 
 
 
 
9. ADVICE 

SOUGHT 
 
The advice of the CHM, PEAG and BVEAG is sought on the conclusions of the report.  
 
 
 
Cervarix Safety Review PEAG Aug 2012; BVEAG, CHM Sept 2012  
42 

Document Outline