• Knowledgebase: Immunization Questions

    Questions about immunizations, side effects, and not vaccinating.

    3. How Vaccine Safety is Monitored Before and After Licensure - Top

    How Vaccine Safety is Monitored Before and After Licensure

    Pediatric Annals, 30:7, July 2001

    Kristine K. Macartney, MD; and Paul A. Offit, MD
    Pediatric Annals 30:7 July 2001

    Vaccination is among the most effective and widely used of public health interventions. However, when immunization programs have high rates of vaccination and low incidences of vaccine-preventable diseases (such as in the United States), adverse vaccine events receive increased attention from the medical community and the public. Concerns regarding these events (whether true associations or coincidental) can lead to a loss of public confidence in immunization, decreased vaccine coverage, and the return of epidemic diseases.34 All vaccines have possible side effects, so recommendations for their use involve balancing risks and benefits.

    Because vaccines are given to healthy individuals„Ÿmost commonly infants„Ÿand are required for school entry, public tolerance of adverse reactions is appropriately low.35 Thus, vaccines are held to a higher standard of safety than are other medical interventions or pharmaceutical products. This article discusses the strengths and weaknesses of the mechanisms in place to determine vaccine safety, both before and after licensure, and how pediatricians can keep apprised of vaccine safety.

    ASSESSING VACCINE SAFETY BEFORE LICENSURE

    At the foundation of vaccine safety are standards that ensure purity and consistency of these
    products. In the United States, this is regulated by the Centers for Biologics Evaluation and Research (CBER) within the Food and Drug Administration (FDA), using statutes specified in the Code of Federal Regulations.6 In addition, the World Health Organization provides guidance for products that are used internationally.

    Laboratory testing of vaccine purity and consistency is required before and after licensure. This includes extensive laboratory and animal tests tailored to each specific vaccine. For all vaccines, an intensive search for contaminating agents is required. Testing is done for all known viral, bacterial, or fungal agents. Next, potency tests are applied to each product.

    Furthermore, manufacturers are required to conform to good manufacturing practices regarding personnel, equipment, and record keeping. At least six large lots of vaccine, each containing tens of thousands of doses, must be produced with identical potencies to demonstrate that the manufacturing process is consistent and reliable. In addition, manufacturers are required to provide information that describes appropriate vaccine storage and handling, and also safe injection practices.

    Fulfilling the requirements of the CBER often takes between 5 and 10 years and is extraordinarily expensive. The average cost to develop a vaccine before licensure is between $300 and $500 million. This is one of the reasons why most vaccines can be manufactured only by large corporations.

    Once a candidate vaccine has undergone extensive laboratory testing and been approved as an investigational new drug for use in clinical trials, human studies of safety, immunogenicity, and efficacy generally proceed in three stages or phases.

    Phase 1 trials usually involve only 20 to 80 par ticipants; frequently these subjects do not belong to the intended target population for the vaccine (eg, testing for pediatric vaccines may first be done in adults). These studies provide information on the most basic aspects of safety and tolerability, and can detect only extremely common adverse events.

    Phase 2 trials enroll 100 to 200 subjects and provide information about the vaccine's immunogenicity and common side effects. These studies are performed in the proposed target group (eg, infants). They may also provide some information on the vaccine's efficacy for disease prevention.

    Phase 3 trials enroll several thousand to tens of thousands of participants. The number of subjects chosen is primarily based on how many will be necessary to determine vaccine efficacy. The quality of the information regarding adverse effects depends on the sample size studied and the duration of time the subjects are observed after vaccination (usually not more than 42 days). The incidence of relatively common side effects, such as injection site reactions, fussiness, and fever, can be determined. These large trials also address the consistency of the response and examine concomitant use with other vaccines normally given at the same time.

    The experimental design of these clinical trials has two strengths: placebo groups are included and investigators are required to detect adverse events in a blinded fashion. This makes it relatively simple to determine how often the vaccine causes a particular adverse event. However, despite their comprehensive design, pre-licensure trials of a vaccine can miss rare side effects, or adverse events with a delayed onset, or potential reactions that are a function of culturally or ethnically diverse subpopulations.3 Unpredictable adverse events may occur as infrequently as once every 100,000 to 1 million doses, and the difficulty and cost of studying them increases with their rarity. To detect such events, continuing and comprehensive assessment of vaccine safety must be done after licensure.

    MONITORING VACCINE SAFETY AFTER LICENSURE

    These inherent limitations of pre-licensure studies mean that post-licensure (also known as postmarketing) studies of vaccine safety are essential. While a new vaccine is being administered to millions of individuals, adverse events may be detected that were not evident in earlier clinical trials. Two types of post-licensure surveillance systems are used„Ÿpassive and active. Passive surveillance is more common and relies on spontaneous reporting of adverse events, such as that for the Vaccine Adverse Event Reporting System (VAERS) in the United States.7 3 In active surveillance systems, adverse events following vaccination are prospectively assessed using preexisting large linked databases (eg, the Vaccine Safety DataLink [VSD3). These have been developed more recently.

    Vaccine manufacturers are now required to perform formal post-licensure assessment of vaccine
    safety (called Phase 4 studies). These also usually rely on large linked databases, and definitely
    enhance the ability to detect adverse events not apparent in pre-licensure trials. In addition to formal post-licensure surveillance, epidemiologic studies have played an important role in assessing the risk of an adverse event in vaccinated groups compared with unvaccinated groups.

    In contrast to the regulated environment of randomized pre-licensure clinical trials, many factors
    make it difficult to detect true adverse events related to a vaccine after licensure. These include difficulties in determining the specifics of exposure to vaccines (eg, multiple vaccines or the high mobility of patients between immunization providers), assessing outcomes (eg, adverse events that lackdiagnostic criteria or a lack of knowledge of the biologic mechanism of events such as chronic fatigue syndrome or autism), and performing analyses (eg, uncertain sample size or for confounding variables that influence vaccination status).33 Unfortunately, many individuals, including some pediatricians, incorrectly assume that causality can be determined solely on temporal associational Although an adverse event occurred after vaccination, this does not mean that the vaccine caused it.

    Establishing a causal relationship between a specific vaccine and an adverse event is a complex
    process that involves consideration of many factors. These include biologic plausibility, strength of the association, previous known toxicity of the agent, characteristics of the event and the vaccinee, alternative causes, statistical significance of available epidemiologic studies, and analytical bias.

    THE VAERS

    The VAERS is a product of the National Childhood Vaccine Injury Act of 1986. This national passive surveillance system began in 1990.3 It is supervised by the Centers for Disease Control and Prevention (CDC) and the FDA and receives reports of any adverse events that occur following vaccination. VAERS forms (Figure) can be submitted by anyone (including pediatricians, pharmaceutical companies, and parents) and there is no restriction for reporting based on the interval between vaccination and the onset of illness, or whether medical care was required. There is substantial protection regarding disclosure of the identity of the forms and the patients. Information from VAERS forms is submitted to a database from which selected serious events and deaths can be compiled and analyzed by the FDA and the CDC. Approximately 10,000 VAERS forms are received annually.

    A significant limitation of the VAERS is that it receives only information regarding vaccinated individuals with adverse events. To calculate the rate at which an adverse event occurs and to determine whether vaccines cause a particular adverse event requires the four pieces of information identified in Table 1 as a, b, c, and d. (not shown here). The VAERS does not receive information regarding the number of vaccine doses administered (although the number distributed by the manufacturer can be used as a surrogate), and the occurrence or absence of the adverse event in unvaccinated individuals is not reported. This can be further confounded by underreporting, deficient quality of the data received, incomplete reports, lack of consistent diagnostic criteria for a suspected adverse effect, the simultaneous administration of multiple vaccines, and reporting bias. The latter includes increased reporting when a vaccine is first licensed, increased reporting with vaccines that are "in the news," and differences in reporting rates between the public sector and the private sector. All of these problems limit the information that can be derived from the VAERS.3 Unfortunately, VAERS data have been misunderstood by the media and misused by antivaccine groups, and have been presented as if a reported adverse event following vaccination is caused by the vaccine.

    The strengths of the VAERS are that it is the only surveillance system that covers the entire U.S. population and that it includes the largest number of case reports temporarily associated with vaccination. Although a crude tool, the VAERS can generate a signal of potential problems, including new, rare, or unusual adverse events. This can then trigger further investigation. For example, VAERS data contributed to the investigation of idiopathic thrombocytopenic purpura occurring after measles-mumps-rubella vaccinell and intussusception after the rotavirus vaccine.

    The VSD

    The VSD, created by the CDC in 1990, is an active surveillance system using a large linked database. Information regarding vaccine and medical outcome histories is collected prospectively
    from the computerized clinical databases of four health maintenance organizations (HMOs) in Seattle, Washington; Portland, Oregon; Oakland, Califomia; and Los Angeles, Califomia. It can provide information from approximately 6 million people (2% of the U.S. population) and has now been expanded to include all age groups.

    The strengths of the VSD are that underreporting and recall bias are reduced, information regarding the total number of vaccinated individuals is available, and unvaccinated control subjects are available for comparison. Also, because of the large populations studied, relatively rare adverse events can be detected over time.

    The limitations of the VSD are that it is not fully representative of the United States geographically or socioeconomically, and that adverse events may not be captured in the HMO database. Another important caveat is that, because of the high vaccination rate in HMOs, relatively few unvaccinated control subjects are available for study. This means that VSD data are examined predominantly by "risk-interval" analysis: adverse events are monitored after vaccination for a biologically plausible risk interval. Then the incidence of the adverse event occurring in the risk period is compared with the incidence of the event in nonrisk observation periods, as shown in Table 2 (not included here).

    SAFETY CONCERNS REGARDING RECENTLY INTRODUCED VACCINES

    Two recently introduced vaccines demonstrate specific issues regarding assessment of vaccine
    safety before and after licensure.

    Rotavirus Vaccine

    In August 1998, a tetravalent rhesus-human reassortant rotavirus vaccine was licensed for use in the United States and recommended for the routine vaccination of healthy infants. In July 1999, use of the vaccine was suspended following reports to the VAERS of intussusception in infants who had recently received the vaccine. In addition, a preliminary review of post-licensure data from the VSD and active post-licensure surveillance in Minnesota indicated an increased risk of intussusception from the vaccine.l2 In October 1999, following further review of post-licensure data, the Advisory Committee on Immunization Practices (ACIP) withdrew its recommendation that rotavirus vaccine be administered to children in the United States, and the vaccine was withdrawn.

    The experience with the rotavirus vaccine demonstrated that the mechanisms to monitor vaccine safety after licensure in the United States can function effectively. In pre-licensure studies of rotavirus vaccine, intussusception occurred in 5 of 10,054 vaccine recipients, and in 1 of 4,633 control subjects, a difference that was not statistically significant. Intussusception was considered likely to be a temporal rather than a causal association because of this and numerous other observations.

    However, the occurrence of intussusception in the pre-licensure studies prompted the ACIP to recommend post-licensure surveillance for this specific adverse event, and led to the analysis of VAERS data early in the post-licensure period.

    During the first 10 months after licensure, and after approximately 1.5 million doses had been
    given, 15 cases of intussusception had been reported to the VAERS. This number was within the range of the number of cases expected from known background rates of intussusception. However, because underreporting is inherent in the VAERS, the actual number of cases was suspected to be greater. In response to the VAERS reports, the CDC initiated an investigation across 19 states. This determined that intussusception occurred with significantly increased frequency during the first 2 weeks after vaccination with the rotavirus vaccine, particularly following the first dose. The withdrawal of the rotavirus vaccine raised questions as to how future vaccine candidates should be evaluated for safety, particularly regarding the number of participants required in clinical trials. The number of participants needed to detect rare events depends on the particular adverse event and the rate at which it occurs in vaccine recipients and control subjects. Because such events are, by nature, unanticipated, how large should clinical trials be?

    Requiring large sample sizes for vaccine studies before licensure may detect unanticipated rare adverse events. However, such large studies also delay the availability of a vaccine that could be
    preventing disease and death, and make the cost of developing vaccines higher and perhaps prohibitive. In practical terms, it is not feasible for pre-licensure studies to be large enough to detect all rare adverse events. A system of thorough and timely post-licensure assessment of vaccine safety, such as that which worked effectively with the rotavirus vaccine, remains crucial to ensuring vaccine safety.

    Pneumococcal Conjugate Vaccine

    In February 2000, a 7-valent pneumococcal polysaccharide protein conjugate vaccine was licensed for use in the United States. It is now recommended for the routine immunization of all
    children 2 to 23 months old and for certain children 24 to 59 months old.ls The efficacy and safety of this vaccine was evaluated in a trial of 37,830 healthy children in an HMO in Northern California: half received the pneumococcal vaccine and half received a conjugated meningococcal group C (control) vaccine.

    The design of the trial allowed an interim analysis. In August 1998, 17 cases of invasive pneumococcal disease had occurred in the group vaccinated with the control vaccine, whereas no cases had occurred in the group fully vaccinated with the pneumococcal vaccine. This difference was statistically significant. Because of the high efficacy demonstrated at this interim evaluation, an independent Study Advisory Group recommended termination of enrollment. However, blinded fol low-up and blinded vaccination per protocol of both study groups was continued for an additional 8 months. By this time, an additional 23 cases of invasive disease had occurred among those who had completed the series of control and pneumococcal vaccinations. All but 1 case was in the control group.

    There is no doubt that the blinded follow-up period allowed generation of important additional efficacy and safety data regarding the vaccine. However, because the vaccine's efficacy in preventing invasive pneumococcal disease appeared to have been determined at the interim analysis, was continuation of the trial beyond this period required? Earlier termination of the study may have made an effective vaccine available to children more promptly and could have prevented invasive bacterial disease in those control children observed for an additional 8 months. Herein lies the difficulty in striking a balance when determining vaccine safety and efficacy in pre-licensure studies.

    HOW DO PEDIATRICIANS LEARN ABOUT VACCINE SAFETY?

    Pediatricians are increasingly required to provide advice on the risks and benefits of immunizations, respond to questions about vaccine safety, and address misconceptions about vaccination.

    Parents with questions regarding vaccination are likely to rely most on the information and opinions provided by their child's pediatrician.l7 In contrast, misinformation regarding vaccine safety is found in the media, on the Internet, and through materials disseminated by antivaccine groups. In addition, the presence of conflicting information in the medical literature can be confusing for both patients and pediatricians. What information should the public and pediatricians trust, and how can pediatricians keep abreast of issues in vaccine safety?

    The answer to interpreting conflicting scientific studies lies in the hands of expert committees that thoroughly and systematically review the scientific information regarding vaccines, and present recommendations regarding their benefits and risks. The risks associated with routine vaccines were extensively reviewed by the Institute of Medicine in the early 1990sls and by the ACIP in 1996. Extensive review of other routinely used and new vaccines should continue as further evidence regarding their use becomes available. The ACIP also periodically modifies recommendations regarding vaccine use depending on the dynamic balance of benefits and risks. For example, in 1998 the recommendation was made to change to an all inactivated polio vaccination schedule in the United States because of the lack of circulating wild-type polio and the continued risk of vaccine associated paralytic polio from oral polio vaccine.

    Several professional organizations have also developed materials that provide comprehensive, up-to-date information for both pediatricians and the public regarding vaccine safety and that counter the misconceptions and misinformation about immunizations. Many of these groups can be contacted via the Internet and are listed in Table 3 (not shown here). Information about the risks and benefits of immunizations is also provided in pamphlets that have been developed by the CDC and in Vaccine Information Statements, which must be provided to patients at the time of vaccination. Copies of the information sheets can be obtained from the web site of the National Immunization Program.

    Vaccine package inserts also contain information regarding vaccine safety. However, they are designed more as a legal document than as medical information. They do not effectively communicate risks, but rather list virtually all adverse events that have been reported following vaccination, whether they are causally related or coincidental. For example, the package inserts for both brands of hepatitis B vaccine state that adverse reactions reported with the use of the marketed vaccine have included "nervous system disease multiple sclerosis." However, studies have shown that the incidence of multiple sclerosis is not increased among individuals who have received the hepatitis B vaccine.

    CONCLUSION

    Concerns regarding vaccine safety are now particularly prevalent in the United States. Although no vaccine is 100% harmless, the benefits of vaccination continue to outweigh the risks for our pediatric population. It is a challenging task for pediatricians to stay informed regarding issues in vaccine safety and to communicate these risks to parents. This task is especially difficult because parents are increasingly skeptical of vaccines, have little familiarity with the terrible effects of vaccine preventable diseases, and are occasionally influenced by well-organized antivaccine groups and by media eager for controversy. The truth is that vaccines have a remarkable record of safety. Comprehensive mechanisms are in place to optimize vaccine safety beginning in the pre-licensure phase and continuing throughout a vaccine's use. Pediatricians have an important role in monitoring vaccine safety after licensure by reporting adverse events to the VAERS. Developments in biotechnology promise to offer safer vaccines, as does a more systematic approach to assessing vaccine safety (especially in the post-licensure phase) via computerization and centralization of health care services. Finally, organizations that promote vaccine use and counter allegations about vaccine safety offer much to keep pediatricians and the public informed about the benefits of vaccination. - Updated: July 11, 2001

    [e-Mail me the Knowledgebase]-   [Search our Knowledgebase] -  [Question Not Answered?]    


    • James L Horwitz, MD

    David C. Thomas, MD

    Barbara Lindberg, PNP