Global Vaccine Safety

Global Advisory Committee on Vaccine Safety, 16-17 June 2010

Published in the WHO Weekly Epidemiological Record on 23 July 2010

The Global Advisory Committee on Vaccine Safety (GACVS), an expert clinical and scientific body, was established by WHO to provide independent, scientifically rigorous advice on issues of vaccine safety that potentially have global importance.1 The Committee held its 22nd meeting in Geneva (Switzerland) on 16–17 June 2010.2 The Committee reviewed: (i) the safety of pandemic A (H1N1) influenza vaccines; (ii) an apparent increase in febrile reactions following administration of a seasonal influenza vaccine in Australia; (iii) the finding of DNA from porcine circoviruses in rotavirus vaccines; (iv) the safety of live attenuated hepatitis A vaccines; (v) the safety profile of a recently prequalified meningococcal A conjugate vaccine; and (vi) new data on yellow fever vaccine-related risks.

Safety of pandemic A (H1N1) influenza vaccines

Since its emergence in March 2009, pandemic influenza A (H1N1) 2009 virus has caused significant morbidity and mortality globally. For example, in the United States the pandemic A (H1N1) 2009 virus is estimated to have caused 61 million cases of illness, 274 000 hospitalizations, and 12 470 deaths between April 2009 and April 2010.3 In response to the pandemic, over 30 pandemic A (H1N1) 2009 vaccines were licensed worldwide: these included live attenuated vaccines; inactivated unadjuvanted vaccines (split, subunit virion or whole virion); and inactivated adjuvanted vaccines (split or subunit virion). Between September 2009 and June 2010, >350 million doses of vaccine were administered, targeting various populations.

As of 6 June 2010, more than 214 countries had reported laboratory-confirmed cases of pandemic A (H1N1) 2009 influenza, including at least 18 156 deaths. Younger age groups have been disproportionately affected by the pandemic virus and have had higher mortality from the disease than from seasonal influenza. Pregnant women have been disproportionately represented in admissions to intensive care units. Although older age groups have had lower rates of infection than with seasonal influenza, their hospitalization rates and mortality rates have been high.

Most of the safety information about pandemic influenza vaccines has been derived from passive surveillance, but there has been some active surveillance for specific conditions or circumstances for which it was thought, a priori, that there might be an increased risk (such as Guillain–Barré syndrome) or when the vaccine has been used for specific groups of patients (for example, in pregnant women or people who are immunocompromised).

Since the initiation of vaccination campaigns, WHO has coordinated an unprecedented and continuing exchange of safety information among regulatory and public health authorities from many countries. Follow-up of vaccinated populations continues, and additional data on safety are expected later in 2010. In order to establish reliably the risk–benefit balance of the vaccines, these findings should be considered along with additional observations on the impact of the pandemic disease itself. An interim assessment of the potential risks (or signals) of adverse reactions evaluated in different countries and geographical areas for several different products is given below.

In the United States, where an estimated 65 million people were vaccinated with inactivated unadjuvanted vaccines and 17 million individuals were vaccinated with live attenuated vaccine, no signals of unexpected side-effects were detected through passive surveillance. The occurrence of Guillain–Barré syndrome in people who had been vaccinated was evaluated through several different active surveillance systems, one of which yielded a weak signal; further analyses are under way. Preliminary analysis suggests that if an increase in the risk of Guillain–Barré in those vaccinated is confirmed, the risk may be approximately 1 case/1 000 000 doses, which is similar to the risk reported during some years with seasonal trivalent, inactivated, unadjuvanted vaccine. In addition, there were weak signals of thrombocytopenia and Bell’s palsy; these are also being investigated.

In Japan, approximately 18 million individuals were vaccinated with 21 million doses of inactivated unadjuvanted vaccine. Safety was monitored primarily through passive surveillance. Cases of interstitial lung disease, thrombocytopenia, idiopathic thrombocytopenic purpura and allergic purpura following immunization were evaluated, but were not considered to represent new safety issues.

As of April 2010, China reported that 97 million people, representing 7% of the population, had been vaccinated with locally produced, inactivated, unadjuvanted vaccines. Safety has been monitored through passive surveillance. The rate of adverse events following immunization was about 8.6/100 000 doses administered. The reported rate for serious reactions was around 1/100 000 doses administered; the majority of these reactions were Schönlein–Henoch purpura, anaphylactic shock, laryngeal oedema and febrile convulsions. The number of cases of Guillain–Barré reported was not higher than expected. Although deaths following vaccination were reported, these were found to result from underlying conditions.

In Canada, an estimated 12.5 million people were vaccinated with inactivated vaccine adjuvanted with AS03. Passive surveillance detected a signal for allergic events; this is being investigated.

In the European Economic Area, at least 38.5 million people were vaccinated with 1 of 3 authorized vaccines marketed in the area. These vaccines included: inactivated unadjuvanted whole virion vaccine; inactivated subunit vaccine adjuvanted with MF59; and inactivated split-virus vaccine adjuvanted with AS03. Additional vaccines were authorized nationally, and when these are included >46 million people are estimated to have been vaccinated. The safety of vaccines has been monitored at the European level using passive surveillance; adverse events following immunization have been reported to EudraVigilance (which covers products authorized by the European Union). After thorough evaluation of spontaneous reports, no new safety issues have been identified.

GACVS reached the following conclusions:

  • Reporting mechanisms for adverse events following immunization with pandemic A (H1N1) 2009 influenza vaccines have been enhanced in many countries. Continuing to monitor vaccine safety (maintaining pharmacovigilance) is critical; monitoring should include regular information-sharing with WHO by national regulatory and health authorities. Most of the safety information has been derived from passive surveillance. Data from active surveillance will be assessed as they become available.
  • The safety profile of the pandemic A (H1N1) 2009 influenza vaccines noted above is reassuring.
  • Most of the adverse events that have been reported after immunization have not been serious. To date, no unexpected safety concerns have been identified.
  • Active surveillance for Guillain–Barré syndrome is under way in a number of countries, and analyses are pending in many of these. So far, the risk of Guillain–Barré syndrome, if any, appears to be no greater than has been reported previously for some seasonal, trivalent, inactivated influenza vaccines.
  • Active surveillance of pregnancy outcomes also continues. So far, available data on the safety of the vaccines are reassuring.
  • It is critical to strengthen adverse event reporting following immunization, which relies on existing infrastructure for ongoing pharmacovigilance. Prospectively agreed-upon case definitions for adverse events (e.g. for anaphylaxis, Guillain–Barré syndrome, convulsions) are also important because they facilitate global comparisons of safety profiles of vaccines used in different countries.

Febrile seizures after seasonal influenza vaccine in Australia

Australia’s regulatory authority provided a summary of the increased number of reports of fever and febrile convulsions in children aged <5 years following administration of the 2010 seasonal inactivated influenza vaccine (Fluvax) made by CSL (Parkville, Victoria, Australia). No clinical or epidemiological factors have yet been identified that explain the increase. Vaccine testing has shown no abnormalities, but testing continues. The Chief Medical Officer of Australia advised suspending the use of all 2010 seasonal influenza vaccines in healthy children aged <5 years. For children with medical risk factors, the Chief Medical Officer advised using monovalent A (H1N1) 2009 vaccine or seasonal 2010 vaccine made by other manufacturers. Australia will apprise GACVS of any additional information. GACVS has not been made aware of reports of increased fever or febrile convulsions from other 2010 seasonal vaccines but noted the importance of reviewing data on the use of the vaccines elsewhere in the southern hemisphere, as well as data that will be collected during the use of 2010 seasonal vaccines in the northern hemisphere.

Porcine circoviruses and rotavirus vaccines

GACVS reviewed new data on the finding of porcine circovirus DNA in Rotarix (GlaxoSmithKline) and RotaTeq (Merck & Co.), two oral vaccines for preventing rotavirus gastroenteritis. In March 2010, GlaxoSmithKline confirmed a report from academic investigators that Rotarix contained DNA from porcine circovirus type 1 (PCV1). GACVS met by teleconference on 25 March 2010 to discuss this finding, and issued interim advice regarding the safety of the vaccine. In May 2010, Merck reported that DNA from porcine circovirus type 2 (PCV2) had been found in RotaTeq.

Neither PCV1 nor PCV2 is known to infect or cause disease in humans. GlaxoSmithKline reported that PCV1 DNA has been identified in both the master cell bank and master viral seed used for vaccine production, and thus has been present in the vaccine throughout its clinical development, including in the vaccine used in prelicensure clinical trials. Initial data reported by GlaxoSmithKline suggest that Rotarix contains infectious PCV1, however PCV1 does not result in productive infection in human cell lines. Analysis of prevaccination and postvaccination serum samples from 40 infants who participated in clinical trials of Rotarix revealed no evidence of serological response to PCV1. DNA from PCV1 was detected in initial postvaccination stool samples (on days 3 or 7) in 4/40 infants; results were inconclusive on a stool sample from 1 additional infant. None of the infants had positive stool samples later.

Merck reported finding low levels of PCV1 and PCV2 DNA in bulk lots and final-container lots of RotaTeq. The amount of PCV DNA found in RotaTeq is consistent with introduction from irradiated trypsin, and infectivity assays are under way.

Additional information on both products is expected from both manufacturers and from other investigators.

The safety of Rotarix and RotaTeq is supported by large prelicensure clinical trials and extensive postlicensure safety experience. Worldwide, over 69 million doses of Rotarix and 37 million doses of RotaTeq have been distributed. Rotavirus gastroenteritis is the most common cause of severe diarrhoeal disease in young children throughout the world, causing an estimated 527 000 deaths annually among children aged <5 years. Given the extensive clinical data supporting the safety of both Rotarix and RotaTeq, and the benefits of rotavirus vaccination, GACVS considers that the benefits of vaccination far outweigh any currently known risk associated with the use of either vaccine. GACVS will continue to review data as they become available and will update this statement as necessary.

Live attenuated hepatitis A vaccine

Live attenuated hepatitis A virus vaccines (H2 and L-A-1 strains) have been developed, manufactured and licensed in China. Since 2007, hepatitis A vaccines (both inactivated and live) have been integrated into the national immunization programme. Managers of regional immunization programmes may opt to use any of the vaccines. More than 10 million doses of H2 and L-A-1 vaccine are used annually in China. The only other country where the live vaccine is registered is India, where the H2-strain vaccine was licensed in 2005 and is used only on the private market.

An overview of the development and manufacture of hepatitis A vaccines in China was presented to the Committee. The live attenuated H2-strain and L-A-1-strain vaccines were developed from strains isolated from 2 children with hepatitis; they were attenuated with multiple (>25) cell-culture passages. Vaccine lots are made in locally developed human diploid embryonic lung fibroblast cells (termed 2BS). A special additive (comprising trehalose, sodium glutamate, arginine and dextran) is used to stabilize virus titre on lyophilization.

A representative from the Chinese Centre for Disease Control and Prevention presented a brief account of the epidemiology of hepatitis A in China. There is a temporal relationship between the widespread use of hepatitis A vaccines and a decrease in disease incidence. The overall experience of hepatitis A control in China is good, although the parts played by vaccines and socioeconomic or other factors has not been assessed.

Limited data about reactions to H2-strain vaccine during clinical trials and through passive surveillance for both vaccines did not identify any significant safety issues. However, it will be essential to conduct rigorous high-quality postmarketing surveillance in selected communities to measure and monitor safety and adverse reactions. Studies of children vaccinated with live vaccines have shown shedding of the vaccine virus and secondary infection among contacts, so postmarketing surveillance may provide a context in which to conduct specific studies to examine the outcome of secondary infection and virus circulation if it occurs.

In view of the volume of use of live hepatitis A vaccines in China and their potential usefulness outside China, carefully collected and validated data on safety and efficacy will be valuable. Data of particular interest will be molecular markers of attenuation, the genetic stability of attenuation markers after human passage, the safety of orally ingested vaccine virus, and clinical safety and efficacy as demonstrated in well-conducted and sufficiently large clinical trials. The role of the National Regulatory Authority will be of critical importance in providing oversight on the safety of vaccines used in China.

Updated safety of meningococcal A conjugate vaccine

The Committee was updated on clinical safety data from studies of a new meningococcal A conjugate vaccine (MenAfriVac, Serum Institute of India, Pune, India) and of plans for mass vaccination campaigns to be launched imminently. The vaccine is a lyophilized meningitis A conjugate vaccine developed by the Meningitis Vaccine Project. Safety data presented to the Committee in December 2009 did not raise any particular concerns, but the Committee highlighted the need for additional data from larger numbers of participants to better assess the safety profile.

The reactogenicity and safety of the vaccine have been evaluated in 7 clinical trials (3 of which are ongoing) in 5 countries (Gambia, Ghana, India, Mali and Senegal). These included one study of children aged 14 weeks to 18 months; the others have been conducted in volunteers aged 1–29 years, the target age group for the initial mass vaccination campaigns. As of 31 May 2010, a total of 4614 vaccinated participants and 2040 controls had been followed to collect data on immunogenicity and adverse events for at least 1 month after vaccination and for serious adverse events for at least 1 year. A total of 237 serious adverse events have been reported (including 16 deaths); after causality assessment, 235 of these events were considered to be unrelated to the vaccine. One of the serious adverse events that was considered to be probably vaccine-related was a hypersensitivity reaction with facial oedema in an infant aged 10 months; this is a well-documented hypersensitivity reaction occurring with meningococcal and other conjugate vaccines. The other vaccine-related serious adverse event was a case of simple febrile convulsion in a child aged 17 months following administration of the meningococcal A vaccine along with a pentavalent diphtheria–tetanus–pertussis plus Haemophilus influenzae type b plus hepatitis B vaccines. Both children recovered without sequelae. In all completed studies, the rates of other serious adverse events were similar in vaccinated and control participants. Most of the serious adverse events reported (189/237) accrued from an ongoing study of children aged 14 weeks–18 months; 74 (39%) of these were attributed to malaria, 66 (35%) to acute gastroenteritis, and 33 (14%) to acute respiratory tract infection. The occurrence of these events was consistent with the seasonal age-specific morbidity in the areas where the study is being conducted.

MenAfriVac has not been administered to pregnant or lactating women, but 15 women are known to have become pregnant shortly after administration. Of the 14 completed pregnancies studied, 13 resulted in a normal live-born infant and 1 in a stillbirth, which occurred 13 months after vaccination following obstructed labour in a 26-year-old woman with a history of stillbirth. The Committee recommended that studies be conducted to evaluate vaccine safety in pregnancy because of the likelihood of unintentional administration of the vaccine to pregnant women during mass vaccination campaigns.

The Committee also highlighted the importance of gaining additional information on MenAfriVac with respect to the duration of protection, its effect on carriage of Neisseria meningitidis, interactions with vaccines delivered by the Expanded Programme on Immunization, any possible effect on the prevalence of other serotypes (serotype replacement), and the safety and immunogenicity of the vaccine in groups considered to be potentially at high risk, such as people infected with HIV and those who are severely malnourished.

The Committee concluded that the data accumulated in clinical trials do not indicate that there are any significant safety issues with the vaccine. Since the vaccine will soon be used in mass campaigns, the Committee reiterated its previous advice that, where possible, phased introduction of the vaccine would be desirable so that additional safety data may be accumulated through careful postmarketing surveillance; the Committee was pleased to note that such studies are being planned.

Yellow fever vaccine and breastfeeding

The Committee reviewed recent data suggesting that 3 neonates (aged 10 days, 23 days and 5 weeks) developed encephalitis as a result of infection with yellow fever vaccine virus transmitted to them from their recently-vaccinated mothers. All 3 infants were being breastfed, but the mode of transmission has not been established. All 3 mothers had received the vaccine for the first time during the infant’s first month of life. Further research is needed to quantify the potential risk of transmission of yellow fever vaccine virus from mothers to infants, including the possibility of transmission through breastmilk.

Mass vaccination campaigns being conducted in West Africa provide an opportunity to conduct studies that will clarify these issues. Such studies might test breastmilk from vaccinated mothers for the presence of vaccine virus, and test infants for evidence of seroconversion to the vaccine virus. The potential risk of transmission may vary depending on whether mothers are vaccinated for the first time or have been previously vaccinated.

In areas where yellow fever is endemic, or during outbreaks, the Committee believes that the benefits of vaccinating nursing mothers are likely to far outweigh the risk of potential transmission of vaccine virus to infants; the Committee also believes that the benefits of breastfeeding far outweigh the alternatives for infant feeding. Nursing mothers who are considering travel to endemic areas should be counselled regarding the benefits and potential risks of vaccination. Vaccination is recommended if vaccination is indicated for a breastfeeding woman and travel cannot be avoided or postponed.

  • See No. 41, 1999, pp. 337–338.
  • GACVS invited additional experts to present and discuss evidence related to particular topics. These experts included people affiliated with: Therapeutic Goods Administration in Woden (Australia), who presented information on seasonal influenza vaccines; Ohio State University, Columbus (United States), who presented information on circoviruses; the Ministry of Health, Beijing (China) and Hadassah–Hebrew University Hospital in Jerusalem (Israel), who discussed live attenuated hepatitis A vaccines; and the Centers for Disease Control and Prevention, Atlanta (United States), who discussed yellow fever vaccines.
  • Updated CDC estimates of 2009 H1N1 influenza cases, hospitalizations and deaths in the United States, April 2009 – April 10, 2010. United States Centers for Disease Control and Prevention. (