Global Vaccine Safety

Global Advisory Committee on Vaccine Safety, report of meeting held on 11-12 December 2013

Published in the WHO Weekly Epidemiological Record on 14 February 2014

The Global Advisory Committee on Vaccine Safety (GACVS), an expert clinical and scientific advisory body, was established by WHO to provide independent, scientifically rigorous advice on vaccine safety issues of potential global importance.1 GACVS held its 29th meeting in Geneva, Switzerland, on 11–12 December 2013.2 The Committee reviewed the following topics:

  • the safety profiles of chimeric Japanese encephalitis, inactivated poliovirus, and rotavirus vaccines;
  • allegations related to the safety of human papillomavirus vaccine (HPV);
  • investigations related to increased pyrogenicity of seasonal influenza vaccine;
  • the development of a global vaccine safety surveillance manual addressing basic concepts for immunization programme managers and regulatory staff.

Safety profile of Japanese encephalitis (JE) chimeric vaccine

During the June 2013 meeting of GACVS the safety profiles of 1 live attenuated and 2 inactivated Japanese encephalitis (JE) vaccines based on the SA 14-14-2 strain were considered, and the Committee concluded that there were no significant concerns regarding the safety profile of these vaccines.3 During the December 2013 meeting GACVS considered the safety profile of a novel chimeric JE vaccine (Imojev). This vaccine is a live vaccine construct using the yellow fever (YF) 17D and JE SA-14-14-2 vaccines strains. Construction of the vaccine involves insertion of the nucleic acid sequences encoding the envelope proteins (prM and E) of the JE SA 14-14-2 strain into the YF17D backbone, resulting in a chimeric vaccine virus which is attenuated and lacks neurotropic properties.

Pre-licensure and post-licensure safety and immunogenicity data for Imojev were presented. This vaccine is currently licensed in Australia, Malaysia, the Philippines and Thailand. Pre-licensure data are available for 2486 adults and 2248 children (9–18 months, at first dose). The vaccine is immunogenic and immunogenicity does not appear to be affected by concomitant administration of the measles/mumps/rubella (MMR) vaccine. Short-term safety data for injection site and systemic reactions (reported by >10% of vaccine recipients) were presented and showed that in the adult population, adverse reaction rates were significantly lower with Imojev than with a mouse brain-derived vaccine.4 There is limited post-licensure safety experience with Imojev, with approximately 49 000 doses administered to date, and a larger safety database will be necessary to evaluate the risk of rare adverse events.

GACVS expressed interest in receiving additional information about potential environmental safety issues relative to the use of a chimeric vaccine. These include the theoretical risk of reversion or genetic reassortment with wild-type JE viruses or other circulating flaviviruses that could result in the vaccine virus acquiring neurotropic and/or infectivity properties, and vaccine virus transmission through mosquito hosts. However the biological plausibility of this is remote, given the short duration of viraemia post vaccination and the limited potential for virus vaccine replication and dissemination within the mosquito.

Post-licensure studies are essential in countries were widespread use of a JE chimeric vaccine is planned or is currently implemented. In particular, post-licensure studies and surveillance should include active surveillance of cases of encephalitis along with a laboratory determination of the aetiology of the encephalitis. Safety data on JE vaccines (including but not limited to the chimeric vaccine) administered to immunocompromised persons and pregnant and lactating women are limited.

Safety of inactivated poliovirus vaccines (IPV)

With several countries on the threshold of adopting the use of IPV, in line with the Global Vaccine Action Plan (GVAP) strategy for polio eradication, the GACVS session on IPV focussed on visiting: i) the safety record of IPV, as determined in controlled clinical trials during development of the currently available stand-alone and IPV-containing combination vaccines; ii) Adverse events following immunization (AEFI) reports related to IPV from the Vaccine Adverse Events Reporting System (VAERS) of the USA; and iii) issues related to the manufacturing process for IPV.

The first polio vaccine was developed by Jonas Salk, from formaldehyde-inactivated wild polio viruses. Salk’s IPV was tested and proved highly efficacious against paralytic poliomyelitis in a large clinical trial conducted in US schoolchildren in 1954, which was rapidly followed by licensure of the product and implementation of mass vaccination campaigns in children in the USA, Canada and Western Europe. In less than a year, however, this first IPV was the centre of one the most serious vaccine safety events recorded, the Cutter incident, in which inadequate inactivation of the polio viruses during the manufacturing process resulted in 61 cases of vaccine-associated paralytic poliomyelitis (VAPP), 80 family contact cases, 17 community contact cases and 11 deaths. Following this incident, IPV manufacturing techniques were modified to ensure complete inactivation and avoid any potential risk of injecting live polio viruses. This also resulted in a reduction of the immunogenicity of IPV preparations. In the 1970s an enhanced-potency IPV, similar in immunogenicity to the original product, replaced the second generation IPV. Currently IPV is offered as an individual vaccine as well as in vaccine combinations for primary immunization and for boosters. The available data indicate that known adverse events following IPV administered alone are limited to non-serious reactions. Local reactions, as may occur with any inactivated vaccine, are most common. Adverse events due to IPV administered as a combination with other vaccines are difficult to differentiate from those induced by the other vaccines, e.g. diphtheria+ tetanus+ whole cell pertussis (DTwP). Reviews have not documented any serious adverse events causally related to IPV. Further, a dose of IPV administered prior to a course of oral poliovirus vaccine (OPV) reduces the risk of VAPP compared with an exclusively OPV series.

IPV was introduced in the childhood immunization schedule in the USA in 1997, replacing OPV. Currently, there are 4 licensed vaccines of which 1 is IPV stand-alone vaccine and 3 are in combination with other vaccines. An assessment of AEFI in all ages indicated that most adverse events in VAERS reported from 1 January 1999 to 31 December 2012 were non-serious. Less than 1% of reports were for IPV given alone. The vaccines most commonly co-administered with IPV are pneumococcal conjugate, Haemophilus influenzae type b (Hib), hepatitis B, diphtheria+ tetanus+ acellular pertussis (DTaP), and rotavirus vaccines. Although sudden infant death syndrome (SIDS) is the most commonly coded term for deaths in infants for all IPV-containing vaccines, the Institute of Medicine review (2003) rejected a causal relationship between SIDS and multiple vaccines.5 Based on available data, GACVS is reassured that IPV and IPV-containing vaccines have an excellent safety profile.

GACVS was also presented with an overview of the manufacturing process of IPV by a licensed vaccine manufacturer. The complexities of the manufacturing process were noted, in particular the methods used to ensure virus inactivation and containment to prevent accidental environmental contamination. WHO discussed plans for IPV vaccine technology transfer to emerging country vaccine manufacturers. GACVS noted, given the complexities of the IPV manufacturing process, the importance of ensuring appropriate technical support, training and regulatory oversight to IPV vaccine manufacturers.

Increased occurrence of febrile seizures with a seasonal influenza vaccine

GACVS reviewed progress by the Australian Therapeutic Goods Administration (TGA) and the company that manufactures Fluvax (CSL, Parkville, Victoria, Australia) a trivalent influenza vaccine (TIV) vaccine which, in 2010, was associated with an increased risk of fever and febrile seizures, particularly in children aged <5 years.6 This resulted in a 3-month suspension of the Australian influenza vaccination programme for children. Subsequent investigations confirmed that no other TIVs were associated with this increased risk. Fluvax is now contraindicated in children aged <5 years and avoided in those aged <9 years.

The manufacturer has conducted several analyses in order to clarify the etiological mechanism of increased pyrogenicity of this specific vaccine product. Initial findings identified several possible contributing factors that may have triggered the reaction to the vaccine. These factors include, in particular, the presence of large RNA fragments, as well as characteristics of the B Brisbane seed virus strain used in seasonal influenza vaccines in 2010 that has a greater ability to maintain RNA fragments during the manufacturing process. GACVS noted that the virus splitting process used by CSL differs from that used by other manufacturers. The company informed GACVS of a planned modification to the vaccine manufacturing process that will be implemented in 2014. It is expected that this modification will lead to a reduction or elimination of the possible contributing factors and therefore a reduction in the additional pyrogenicity.

GACVS recommended that further studies be undertaken in healthy adult (non-pregnant) subjects to ascertain the impact of the new manufacturing process, particularly on safety of the product. Once shown to be safe in this group, the safety of Fluvax in pregnant women will need to be assessed. GACVS concurred with TGA’s decision to contraindicate the use of the present CSL vaccine in children aged <5 years. The Committee also took note of measures mandated by TGA to reduce inadvertent vaccination, which have included a number of programmatic measures including package labelling. GACVS noted these events illustrate the importance of post-licensure brand-specific safety surveillance which presents particular challenges with seasonal influenza vaccines.

Update on intussusception following rotavirus vaccine administration

GACVS last reviewed the safety profile of Rotateq and Rotarix vaccines during its December 2011 meeting.7 At that time, the Committee concluded that both vaccines had a good safety profile, but that they may be associated with an increased (up to 6-fold) risk of intussusception after the first dose of vaccine in some populations. During the current meeting, new data from Australia and the USA were reviewed in order to update the assessment of intussusception risk related to both vaccines.

In Australia, a recently published study of intussusception cases identified from national hospitalization databases, supplemented by active hospital-based surveillance from July 2007 through June 2010, was reviewed.8 As both vaccines are available in the country, the study allowed estimation of product-specific risks of intussusception. Findings were similar for both vaccines, suggesting that a significant risk of intussusception exists after the first and second dose of both vaccines. The average vaccine-attributable risk for intussusception, based on the estimated relative incidence in the 1–21 days after dose 1 and the 1–7 days after dose 2, was estimated to be 5.6 additional cases per 100 000 vaccinated infants.

In the USA, data are available from the spontaneous reporting system for vaccine safety (VAERS), as well as from 2 distinct vaccine safety monitoring systems that allow for cohort study designs: the Vaccine Safety Datalink (VSD) and the Post-licensure Rapid Immunization Safety Monitoring system (PRISM). VSD is a network of linked databases, involving 9 integrated health-care delivery institutions, whilst PRISM is a sentinel-like system using claims data from national health insurance companies. VAERS data showed that for Rotateq, from 2006 to 2012, 584 confirmed cases of intussusception were reported for 47 million doses distributed. A cluster of cases was observed between days 3 and 6 after doses 1 and 2. For Rotarix, 66 confirmed intussusception cases were reported for 7.4 million doses distributed. The VSD analyses identified a small cluster of cases following Rotarix, with 6 cases of intussusception for 200 000 doses administered. In contrast, no such cluster was found with Rotateq, with 8 intussusception cases identified (4 each after dose 1 and dose 3) for 1.3 million doses administered. The PRISM data suggest that Rotateq is also associated with clusters of intussusception cases with an attributable risk of approximately 1 case per 100 000 doses whilst the number of cases is currently too small to allow calculation of an attributable risk for Rotarix.

GACVS acknowledged that the findings from both countries tend to confirm a risk of intussusception following administration of both vaccines, in particular during the first 7 days following a first dose. The Committee noted that attributable risk estimates vary across studies. This might reflect differences in the background rate of intussusception (estimated to be double in Australia compared to the USA) but could also reflect sampling uncertainty in all available estimates and limitations of the surveillance systems that lead to some uncontrolled biases (e.g. differences in diagnostic tests and case definitions in different settings). Overall, the findings remain reassuring that the risk of intussusception following current rotavirus vaccines remains small compared to the benefits of preventing the impact of severe diarrhoea. Given possible population differences in risk of intussusception, it is important that rotavirus vaccine introduction in other parts of the world be accompanied by similar active intussusception surveillance studies together with rotaviral disease surveillance so that the benefits and risks can be ascertained with relevant evidence.

Human papillomavirus vaccines safety (HPV)

GACVS reviewed evidence related to autoimmune disease and the HPV, with a focus on multiple sclerosis (MS). The last review was conducted in June 2013, when the Committee reviewed updated data from the USA, Australia, Japan, and the manufacturers of Cervarix (GlaxoSmithKline) and Gardasil (Merck). With >175 million doses distributed worldwide and more countries offering the vaccine through national immunization programmes, the Committee continued to be reassured by the safety profile of the available products. Serious adverse events that have been reported as potential signals have been investigated in more detail and were not confirmed, including Guillain-Barré syndrome, seizures, stroke, venous thromboembolism, anaphylaxis and other allergic reactions. Surveillance of pregnancy outcomes among women inadvertently vaccinated during pregnancy through spontaneous reports and registries has not detected any adverse outcomes above expected rates.

While surveillance data and epidemiologic studies on HPV vaccine have remained reassuring, allegations have continued to surface in the media and elsewhere about the safety of the vaccine. Epidemiologic studies before and after licensure showed no increased risk of autoimmune disease, including MS. Since the introduction of HPV vaccines, such diseases have been under particularly careful investigation given their correspondingly high age-specific background incidence.

Examples of such studies include a register-based cohort study in Sweden and Finland that included almost 1 million girls aged 10–17 years, among whom almost 300 000 were vaccinated against HPV.12 The study investigated whether vaccination was associated with an increased risk of autoimmune, neurological or thromboembolic events. The study results did not show evidence of any association between exposure to HPV vaccine and autoimmune, neurological, and venous thromboembolic adverse events.

In the USA, an observational study involving almost 200 000 girls and young women who had received at least 1 dose of HPV vaccine found no increased incidence of 16 investigated autoimmune diseases in the vaccinated compared to the non-vaccinated group.13 The incidence of MS in the vaccinated cohort, for example, was not significantly higher than the non-vaccinated cohort (incidence rate ratio 1.37, 95% confidence interval 0.74–3.20). In a third study, a pooled analysis of data from 11 clinical trials involving nearly 30 000 participants aged >10 years, of which 16 142 received at least 1 dose of Cervarix and 13 811 received either a placebo containing aluminium hydroxide or 1 of 2 different hepatitis A vaccines. No increased risk for the onset of autoimmune diseases after administration of Cervarix was observed in comparison to the control group.14

The Committee was provided with an overview of cases that were the subject of concern in France. These included one case of MS that had been ajudicated by a French Regional Commission for Conciliation and Compensation. Another 14 cases of MS were reported through regional pharmacovigilance centres and/or the manufacturers to the European Medicines Agency. All 15 cases had been classified as being of “doubtful” causality, according to the French grading system. In addition, the overview from France included results of a cohort study involving 2 million girls aged 12–16 showing a lack of increase in hospitalization rates for autoimmune diseases among those who received the HPV vaccine (2.1/10 000 patients/year) compared to those who did not (2.09/10 000 patients/year).

In summary, GACVS was presented with a series of cases of adverse events following administration of the HPV vaccine. Multiple studies have demonstrated no increase in risk of autoimmune diseases, including MS, among girls who have received HPV vaccine compared to those who have not. The Committee remains reassured by the safety profile of the vaccine, but noted the importance of continued surveillance and epidemiological investigation with an emphasis on the collection of high quality data; such data are essential for interpretion of any adverse events which may occur following vaccination. Allegations of harm due to vaccination based on incomplete information may lead to unnecessary harm when effective vaccines are not used.

Vaccine safety monitoring manual

There is a need for a global manual that addresses the basic concepts of vaccine safety surveillance. GACVS reviewed a draft document based on a recent publication15 by the WHO Regional Office for the Western Pacific. GACVS advised that the global version should be designed primarily for immunization programme managers (at all levels) and for regulatory authority staff. Such a manual should focus on the general principles of immunization, AEFI detection, reporting, investigation, analysis and follow-up activities. It is essential to provide guidance on the systems and functions required of AEFI surveillance, and the channels of communication of safety data, as well as including sample forms that can be adapted by individual countries.

It is important to note that the manual should not be exhaustive with respect to all vaccine safety monitoring principles and methods. It should, however, provide links to the appropriate references and materials. For example, the new causality assessment classification endorsed by GACVS is described in another manual16 that, although designed as a reference for expert review committee members, can serve to supplement the chapter in the general document. The new manual should, therefore, limit its content to an explanation of the need for and basic principles of causality assessment and what is the purpose, general principles and outcomes of causality assessment, without advanced technical discussion. Links to periodically updated e-documents such as the AEFI rate sheets17 will be more helpful than incorporating them into the main text of the manual which could quickly become outdated.

The Committee also advised on some important aspects to be addressed in the manual. It should describe the general structure of an AEFI surveillance system. This includes in particular a focus on relationships between immunization programmes and regulatory agencies. It should also stress the importance of vaccine safety communication to the community, to decision makers and to all levels of immunization services. With respect to clinical interventions or recommended diagnostic methods for specific AEFI, GACVS advised that the manual should focus on general principles, given the diversity of current clinical practices and health-care resources. However, the manual should include generic forms for use in surveillance of vaccine safety and investigation of serious AEFI. Likewise, it should provide access to the content of the aide-memoires on AEFI investigation18 and causality assessment.19

  • See No. 41, 1999, pp. 337–338.
  • GACVS invited additional experts to present and discuss evidence related to particular topics. These experts included persons affiliated with: Center for Biologics Evaluation and Research (U.S. F.D.A), Rockville MD, USA; Centers for Disease Control and Prevention, Atlanta GA, USA; CSL, Parkville, Australia; Emory University, Atlanta GA, USA; National Institute of Infectious Diseases, Tokyo, Japan; National Institute for Public Health and the Environment, Bilthoven, The Netherlands; Royal Children’s Hospital, Melbourne, Australia; Sanofi Pasteur, Lyon, France; Therapeutic Goods Administration, Symonston, Australia.
  • See No. 88, 2013, pp. 301–312.
  • Torresi J, McCarthy K, Feroldi E, et al. Immunogenicity, safety and tolerability in adults of a new single-dose, live-attenuated vaccine against Japanese encephalitis: randomised controlled phase 3 trials. Vaccine. 2010; 28:7993-8000.
  • Stratton K et al. Immunization safety review: vaccinations and sudden unexpected death in infancy. Institute of Medicine (IOM), 2003.
  • See No. 29, 2013, pp. 301–312.
  • See No. 6, 2012, pp. 54–56.
  • Carlin JB et al. Clin Infect Dis. 2013; 57:1427-34.
  • Siegrist CA. Autoimmune diseases after adolescent or adult immunization: what should we expect? CMAJ. 2007 Nov 20;177(11):1352-4.
  • Siegrist CA, Lewis EM, Eskola J, Evans SJ, Black SB. Human papilloma virus immunization in adolescent and young adults: a cohort study to illustrate what events might be mistaken for adverse reactions. Pediatr Infect Dis J. 2007 Nov;26(11):979-84.
  • Callréus T, et al. Human papillomavirus immunisation of adolescent girls and anticipated reporting of immune-mediated adverse events. Vaccine. 2009 May 14;27(22):2954-8.
  • Arnheim-Dahlström L, et al. Autoimmune, neurological, and venous thromboembolic adverse events after immunisation of adolescent girls with quadrivalent human papillomavirus vaccine in Denmark and Sweden: cohort study. BMJ. 2013 Oct 9; 347.
  • Chao C et al. Surveillance of autoimmune conditions following routine use of quadrivalent human papillomavirus vaccine. J Intern Med. 2012 Feb;271(2):193-203.
  • Descamps D, et al. Safety of human papillomavirus (HPV)-16/18 AS04-adjuvanted vaccine for cervical cancer prevention: a pooled analysis of 11 clinical trials. Hum Vaccin. 2009 May;5(5):332-40.
  • Immunization safety surveillance guidelines for immunization programme managers on surveillance of adverse events following immunization (Second Edition). World Health Organization, Western Pacific Region, 2013.
  • See
  • See
  • See
  • See