Vaccine Safety—Achieving the Proper Balance

Influenza is an acute respiratory disease that causes illness in individuals of all ages. Influenza causes repeated infections throughout life, is highly communicable, and is responsible for annual epidemics of varying severity. While influenza leads to a self-limited respiratory disease in the majority of individuals, it is deadly in others. Influenza is the leading cause of vaccine-preventable hospitalizations and deaths in both children and adults in the United States.^{1 - 2} The seriousness of influenza in young children, older adults, and in persons of all ages with certain underlying medical conditions is the reason these individuals and their close contacts are targeted for influenza vaccination each year.²

Although hospitalization and death rates among healthy older children and adults are low, a substantial number of such persons miss work or school, visit a health care professional, or receive antibiotics for influenza-associated illnesses. In most years, 10% to 30% of healthy unvaccinated school-aged children and 5% to 15% of healthy unvaccinated young and middle-aged adults experience symptomatic influenza illness.² Population-based studies of healthy children aged 5 to 15 years estimate that 5% to 7% of such children have an outpatient visit related to influenza, and 6% receive an antibiotic prescription.^{3 - 4} Morbidity appears to be similar in healthy adults. While over 70 million healthy persons younger than 50 years are contacts of persons in high-risk groups and are recommended for annual vaccination to reduce influenza transmission to vulnerable contacts, only 18% of such persons receive influenza vaccine.⁵

The live, attenuated influenza vaccine (LAIV), licensed in 2003, represents a new approach to influenza vaccination and is licensed for healthy individuals 5 to 49 years of age.² Potential advantages of LAIV over the trivalent inactivated influenza vaccine, or “flu shot,” include its intranasal route of administration and induction of both mucosal and systemic immune responses. Evidence suggests that LAIV is efficacious during years in which mismatches between the vaccine and circulating influenza strains occur.⁶ However, since LAIV is licensed only for individuals at low risk for influenza complications, the criteria for an acceptable level of adverse events must be stringent.

All health interventions require an assessment of risks and benefits. For vaccines given to predominantly healthy populations to prevent future disease, safety standards must be high and should be weighed against benefits likely to be derived from disease prevention. Many childhood diseases, like measles, affected nearly all children in the United States prior to vaccine development, and thus widespread vaccination for these diseases benefited nearly every child. In contrast, other vaccines, such as the vaccines for poliomyelitis, targeted less common diseases but eliminated the devastating consequences. When childhood diseases are common, the benefits of vaccines seem clear. However, with the dramatic declines in many childhood diseases due to universal vaccination policies, the focus on vaccine safety has intensified. Safety considerations have become paramount for vaccines against diseases that are less common or that are unlikely to cause death or permanent disability.

Prior to licensure of vaccines, clinical trials sufficiently powered to be informative regarding common adverse events are performed. However, rare adverse events are unlikely to be detected in such trials because of their low frequency and the limited number and careful selection of enrolled participants. In these situations, postmarketing surveillance is essential to identify rare adverse events once the vaccine is used widely.

The Vaccine Adverse Event Reporting System (VAERS) was established in 1990, under joint sponsorship of the Centers for Disease Control and Prevention and the Food and Drug Administration, to collect and review reports of suspected adverse events after administration of any licensed vaccine.⁷ The strength of VAERS is its size, representing the cumulative US population that receives the vaccine. However, since VAERS is a passive surveillance system that relies on voluntary reporting, the extent of underreporting and the difficulty in assigning causal attribution of an event that is temporally related to vaccine administration is problematic. Thus, the greatest asset of VAERS is to identify “signals” that require confirmation of causality with additional studies. The utility of signal detection in VAERS is demonstrated by the lack of reports on vaccine-associated poliomyelitis after the introduction of inactivated polio vaccines and the identification of the temporal association of intussusception with a rhesus-rotavirus vaccine licensed in 1999.⁷

In this issue of JAMA, Izurieta et al⁸ report on the first 2 years of postmarketing experience of LAIV and add to the growing body of evidence supporting its safety. Since licensure, 2.5 million doses of LAIV have been administered, compared with approximately 20 000 doses administered during clinical trials.⁸ While rare, the most concerning adverse event reported to VAERS was anaphylaxis.⁸ Since the ease of administration of LAIV is expected to facilitate vaccination at schools and other nontraditional health care sites, awareness of and preparation for the possibility of serious allergic reactions are important.

An additional concern about the safety of LAIV has been the potential for transmission of the attenuated vaccine virus to unvaccinated contacts, particularly in health care settings. Twenty-two cases of “secondary transmission” to unvaccinated contacts were reported through VAERS and none resulted in hospitalization.⁸ When diagnostic laboratory studies were performed in one case of secondary transmission, wild-type influenza virus, not the vaccine strain, was identified.⁸ Ascertainment of whether the other 21 cases of respiratory illness in contacts of LAIV vaccinees were temporally or causally related to vaccine is impossible to determine with the information provided.

More appropriate studies to assess the ability of LAIV to cause disease after contact with a vaccine recipient provide major reassurances that the risk of such an event is low. While a significant proportion of vaccinees, particularly children, shed the current and prior versions of the vaccine strain, the peak titer of the shed virus is significantly below the estimated infective dose of the virus.⁹ A meticulous study of vaccine virus transmission in a day care population confirmed only 1 transmission that did not lead to disease and estimated the transmission rate between 0.6% and 2.5%.¹⁰ Repeated studies have shown that the live, attenuated virus is genetically stable and retains its phenotypic properties of attenuation and restriction of efficient replication at or above baseline human lower respiratory tract temperatures.^{10 - 11}

The major concern about secondary transmission is that vaccine virus will be transmitted to high-risk contacts. In clinical trials, LAIV has been safely administered to small numbers of human immunodeficiency virus–infected individuals.¹² These and other studies substantiate the current recommendation that LAIV is safe for close contacts of high-risk patients except the most highly immunocompromised, such as hematopoetic stem cell transplant recipients receiving care in protected environments.²

The VAERS data give no cause for concern with regard to triggering asthma in healthy persons in the indicated age groups, consistent with prelicensing clinical trials.⁸ However, VAERS is much less helpful in evaluating associations in groups for whom the vaccine is not yet licensed, since the number of such persons who receive vaccine are too small to be useful. After a large prelicensure placebo-controlled safety study found a “statistically significant” association between receipt of LAIV and reactive airways disease among children 18 to 35 months of age, LAIV was not licensed for children younger than 5 years pending further safety data.¹³ That study was limited by lack of adjustment for multiple comparisons and a lack of temporal relationship with vaccination over the 42-day observation period. Recently, Piedra et al¹⁴ compared rates of medically attended asthma in periods before and after receipt of LAIV in 4529 children aged 18 months to 4 years (including those with mild asthma) over 4 seasons and provided reassurance about the lack of association between asthma and LAIV.

In an era in which influenza vaccine delays and shortages have become the norm, LAIV represents an important option to increase vaccination rates among healthy persons for their own protection and for the protection of those with whom they have close contact. Currently, LAIV is licensed for healthy persons 5 to 49 years of age, although expanded indications may be forthcoming (based on the intent of the company to file for broader licensure). As with any licensed product, continued monitoring of the safety of LAIV will be important. However, the cumulative evidence, including the VAERS experience, should reassure clinicians and patients of the safety of both licensed influenza vaccines. Decisions regarding which influenza vaccine to choose in the healthy 5- to 49-year-old age group should be based on availability, patient preference, and cost.