Global Vaccine Safety

Global Advisory Committee on Vaccine Safety, report of meeting held 5 to 6 December 2012

Published in the WHO Weekly Epidemiological Record on 8 February 2013

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 27th meeting in Geneva, Switzerland, on 5–6 December 2012.2 The Committee reviewed: the safety profile of varicella vaccines; the risk of narcolepsy related to use of Pandemrix® and that of Guillain-Barré syndrome (GBS) with multiple influenza A(H1N1)pdm09 vaccine use; and safety aspects of development of dengue vaccines. GACVS also reviewed progress with implementation of the Global Vaccine Safety Blueprint through the Global Vaccine Safety Initiative.

Varicella vaccines

A systematic post-licensure review of the varicella vaccine Varivax® (Merck) safety in the United States of America (USA) was presented in preparation for an update of the WHO position paper on varicella vaccines.3 A summary of the 2011 US Institute of Medicine (IOM) report,4 a literature review from December 2010 to October 2012, and review of key post-licensure observational studies from the US Centers for Disease Control and Prevention and Merck were included. The focus of the review was to update the safety profile of the varicella vaccine, especially for events considered significant. The IOM committee addressed 15 potential adverse events by a comprehensive review of the literature from 1950 to December 2010. Five events were assessed as having convincing evidence in support of a causal relationship with the vaccine: disseminated varicella infection (widespread chickenpox rash shortly after vaccination); disseminated varicella infection with subsequent infection resulting in pneumonia, meningitis or hepatitis; vaccine strain viral reactivation (appearance of chickenpox rash months to years after vaccination); vaccine strain viral reactivation with subsequent infection resulting in meningitis or encephalitis; and anaphylaxis. While the risks for these adverse events were not quantified in the IOM review, GACVS reviewed evidence from case series and other studies that demonstrated them to be rare events. Ten other adverse events were assessed to have insufficient evidence to support causality: encephalopathy, seizure, cerebellar ataxia, acute disseminated encephalomyelopathy, transverse myelitis, GBS, small fibre neuropathy, new onset arthropathy, stroke, and thrombocytopenia.

A review of more recent post-licensure safety studies of the combination measles, mumps, rubella (MMR) and varicella vaccine, which contains the same Oka strain as Varivax® (ProQuad®), identified a new risk of febrile seizures after vaccination among children aged 12–23 months, compared with children receiving separate MMR and varicella vaccination. In addition, Merck’s pregnancy registry for Varivax® revealed no cases of congenital varicella syndrome during 16 years of vaccine use and the data do not support a signal of an increased risk of spontaneous abortion or birth defects. Finally, a comprehensive literature review from 2010 to 2012 revealed no additional safety concerns.

GACVS raised several questions not covered by this review. These included: (1) whether varicella vaccination increases the risk of shifting varicella disease to older age groups, where disease is generally more serious, and whether this potential risk depends on the number of vaccine doses administered (i.e. would a single dose lead to greater risk than 2 doses and would additional booster doses be required?); (2) whether risks from currently available varicella vaccines remain similar to those described earlier; and (3) what the risk–benefit ratio of varicella vaccine use would be in low and middle income countries (LMICs) with a high proportion of undetected immunocompromised people, especially children with HIV, cancer and other immunodeficiencies. GACVS recommended that additional data are needed to determine the full safety profile of varicella vaccine if it is to be deployed in LMICs. GACVS recommended conducting surveillance for varicella disease to assess the effectiveness, as well as enhanced vaccine adverse event monitoring if varicella vaccine is introduced in LMICs.

Noting that substantial new safety evidence has accumulated since the last WHO report in 1998, GACVS concluded that additional data should be gathered and reviewed to complete the full benefit–risk assessment of varicella vaccine globally.

Pandemic Influenza vaccines

GACVS reviewed 2 safety updates on influenza A(H1N1)pdm09 vaccines, which included associations with narcolepsy and with GBS. The association between use of the adjuvanted pandemic vaccine Pandemrix® (GlaxoSmithKline) and abrupt juvenile narcolepsy has thus far been confirmed in 4 countries (Finland, Ireland, Norway and Sweden) with high uptake of vaccine among children and adolescents. In all these countries the absolute risk was low but the relative risk was significantly raised, ranging from 6.6 (95% confidence interval [CI]: 3.1–14.5) in Sweden to 13.0 per 100 000 (95% CI: 4.8– 34.7) in Ireland. An association in adults has so far been observed only in France. Additional studies are also being finalized in the United Kingdom (UK) and Canada. Although this vaccine is no longer being used and all lots of Pandemrix® (2009 H1N1) have now expired, GACVS considered that research should continue to better characterize the possible underlying biological mechanisms of this association. Most cases of narcolepsy, with or without exposure to Pandemrix®, occur in subjects who carry the HLA DQB1*0602 allele. The importance of understanding the triggers and causes of this association will be crucial, especially since new vaccines will be required to protect against future pandemics.

The association between GBS and influenza vaccine first emerged following swine influenza vaccination in the USA in 1976 (attributable risk: around 1 case of GBS per 100 000 vaccinations). GBS is a relatively rare (1–2 cases per 100 000 persons annually) acute peripheral immune-mediated neuropathy. In up to two-thirds of cases, GBS is preceded by an infectious illness, particularly a gastrointestinal or respiratory infection. The most frequently identified pathogen associated with subsequent GBS is Campylobacter jejuni (estimated at 1 GBS case per 3000 infectious episodes).

After 1976, several studies demonstrated no increased or a slightly increased risk of GBS after use of human seasonal influenza vaccines but vigilance remains high and GBS was carefully monitored during the influenza A(H1N1)pdm09 pandemic vaccination campaign.

GACVS has reviewed published and unpublished active surveillance studies that monitored GBS cases during influenza A(H1N1)pdm09 pandemic vaccination. The data are from single countries such as Canada, France, Germany, Sweden, the UK and the USA as well as a multinational European Union study and a global study. Some but not all of these studies have shown a relative incidence of GBS of 2.28 to 3.76 following both unadjuvanted and adjuvanted influenza A(H1N1)pdm09 pandemic vaccines. Overall, the data available are compatible with a small increased risk of GBS after influenza A(H1N1)pdm09 vaccination that is substantially lower than that observed following the 1976 swine influenza vaccination campaign in the USA.

Live attenuated dengue vaccines

GACVS reviewed progress with the development of tetravalent recombinant live dengue virus vaccines, of which at least one chimeric candidate with a yellow fever virus genetic backbone is undergoing phase III studies. The objective was to appraise the safety assessment plans proactively in order to determine the data critical to safety should the vaccine attain authorization for use in populations where the burden of dengue is significant.

To date, no serious vaccine-related events have been documented in the 41 700 subjects who have participated in different phases of the dengue vaccine trials. In addition, among vaccine recipients, no excess cases of dengue fever or severe dengue attributable to the vaccine virus have been demonstrated compared with control groups. Published results from a phase 2b study conducted among school-age children in Thailand indicated an overall efficacy of 30% but there was evidence of greater protection against 3 of the 4 serotypes.5 GACVS agreed that the safety profile observed up to date is encouraging but efficacy to protect against dengue remains a critical factor to be confirmed.

GACVS recognized that several issues will remain challenges for the evaluation of the safety of dengue vaccines if phase III studies indicate efficacy against clinical disease, in particular the lack of harmonization of dengue case classifications and the lack of consensus on the follow-up time or exposures to different dengue virus types needed to monitor the theorical risk of vaccine-mediated enhanced severe disease outcomes. Safety evaluation of dengue vaccines is also complicated by the rarity of suspected adverse events that could be readily attributable to the vaccine, i.e. neurotropic or viscerotropic disease from the yellow fever vaccine virus backbone, and severe dengue from natural infection potentially induced by incomplete vaccine protection.

In 2008, WHO issued technical recommendations to guide the development of dengue vaccines. GACVS concurs that long-term follow-up of vaccinated and unvaccinated cohorts, including those in randomized double-blind controlled trials, should be sought to help determine the safety of dengue vaccines when exposed to several natural infection cycles. In addition, storage of adequate samples of sera and peripheral blood mononuclear cells will be important for future studies on the immune mechanisms of protection or of sensitization conferred by dengue vaccines. Post-licensure studies will also provide estimates of the long-term effectiveness of immunization against multiple dengue virus serotypes in large populations; assess the risk of vaccine virus escape and any herd effects of vaccination; help establish whether booster immunization is needed; and indicate any potential age shifts in dengue presentations. GACVS identified several approaches that should be considered in designing those studies: collection of background dengue surveillance data; use of phased introduction (e.g. stepped wedge designs); use of case-control and case-only methods (to detect rare early adverse events); and epidemiological exploration of signals as the most appropriate way to establish causality for potentially related adverse events following immunization.

Global Vaccine Safety Initiative

In 2011, WHO and a group of partners developed the Global Vaccine Safety (GVS) Blueprint, a strategic document with a vision of establishing effective vaccine pharmacovigilance systems in all countries.6 The GVS Blueprint is a part of the Global Vaccine Action Plan, which was endorsed by the World Health Assembly in May 2012. The Global Vaccine Safety Initiative (GVSI) was set up to implement the Blueprint and is advised by the GVSI planning group. The Blueprint strategic goals, which GVSI is expected to achieve during 2012–2020, include: (1) ensuring minimal capacity in vaccine safety for all LMICs; (2) promoting enhanced vaccine pharmacovigilance activities in countries with specific needs; and (3) establishing a global technical support structure.

GACVS reviewed the development of the GVSI workplan and the display of its products through its website.7 The GVSI identified activities to cover the 8 strategic goals of the Blueprint through a broad network of stakeholders engaged in global vaccine pharmacovigilance. An activity portfolio has been developed as a management tool for implementing the Blueprint. In the portfolio, activities are prioritized based on their expected impact, feasibility and desirability. The portfolio provides initiators, managers and donors of each activity with due recognition for their respective roles. It is also a resource for all stakeholders in global pharmacovigilance to help identify ongoing efforts, allow for better synergies, minimize duplications and enable resource mobilization. The WHO regional offices lead country support in capacity building and in addressing vaccine safety concerns. Currently, the portfolio includes 94 activities, 31 of which have been identified in the first priority category.

GACVS examined areas of interactions between its mandate to advise on vaccine safety issues of global importance and that of the GVSI to support and strengthen global vaccine pharmacovigilance capacity. GACVS recognizes that the GVSI strategies are required to improve vaccine safety systems and develop networks to strengthen the monitoring, evaluation and response to vaccine safety issues. The rapid development of activities expected to be associated with the GVSI will also generate a complex network of stakeholders where roles and responsibilities should be defined as clearly as possible. The GVSI addresses many aspects that intersect with the work of other established groups. The complementarity of advice from vaccine safety bodies with that of other immunization and public health advisory groups requires particular attention, both at global and regional levels. GACVS identified several areas where the GVSI should clearly outline the leading role of WHO as a global convener in health monitoring and systems. In addition, efforts should include: involving regional and national immunization technical advisory groups with vaccine safety assessment and communication efforts; use of existing academic and educational institutions to expand vaccine safety training resources, engagement of vaccine producers in promoting information exchange; and active development of a roster of vaccine pharmacovigilance experts with appropriate cultural awareness and geographical proximity to support country demands.

The GVSI proposes solutions for a number of unmet needs in vaccine safety. The Blueprint framework has the potential to involve many new players in a broad collaborative effort. Yet the increased volume of activities and attention will generate competition and competency issues, requiring clear accountability and quality assurance. GACVS therefore proposed a thorough analysis of how the complex needs for vaccine safety can best be addressed and development of an accountability framework for all stakeholders participating in the GVSI. The role of GACVS with respect to the GVSI will remain at the consultancy and advisory level, providing independent evaluation of the evidence for the global vaccine safety issues identified by and for WHO. The rapidly increasing number of proposed activities will also require a more comprehensive system of prioritization with appropriate criteria. The committee is well positioned to help set those criteria, with a focus on ensuring that technical partners’ activities meet the needs of LMICs in strengthening their vaccine safety capacities.

  • 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 (US Food and Drug Administration), Rockville, MD, USA; Centers for Diseases Control and Prevention Dengue Branch, San Juan, Puerto Rico; London School of Hygiene and Tropical Medicine, London, United Kingdom; Merck & Co, Upper Gwynedd, PA, USA; National Institute for Health and Welfare, Helsinki, Finland; Sanofi Pasteur, Lyon, France; Shantha Biotechnics Limited, Hyderabad, India; University of California, Los Angeles, CA, USA; University of Cincinnati, OH, USA; University of Laval, Quebec, Canada.
  • See No. 32, 1998, pp. 241–248.
  • Stratton K et al., eds. Adverse events of vaccines: evidence and causality. Washington, DC, Institute of Medicine of the National Academies. August 2011.
  • Sabchareon A et al. Protective efficacy of the recombinant, live-attenuated, CYD tetravalent dengue vaccine in Thai schoolchildren: a randomised, controlled phase 2b trial. Lancet, 2012, 380:1559–1567.
  • See
  • See