cannot be assessed until studies firmly establish the incidence of these disorders in the general population. LLDB can be used to help establish such background rates for acute events.
ADDITIONAL AREAS FOR RESEARCH
A number of additional areas in detecting and responding to adverse events following vaccination warrant further research, according to workshop participants. One question is whether the age at which a vaccine is given is a factor in the adverse events experienced. For example, experience with vaccination with the pertussis vaccine in Japan suggested that very young infants were more susceptible to neurologic complications from the whole-cell pertussis vaccine than slightly older infants. However, younger infants have a higher background rate of neurologic diseases than older infants, making assessment of the role of vaccines difficult. Two examples discussed at the workshop suggested to some that later immunization is possibly associated with an increased risk of adverse reactions. A participant presented preliminary findings that earlier administration of BCG vaccine12 was associated with a decreased risk of developing diabetes, a condition known to have a complex, multifactorial etiology (Classen, 1996). A vaccine manufacturer's representative mentioned evidence of an association (temporal but not necessarily causal) of hepatitis B vaccine with multiple sclerosis (another condition for which etiology is not well understood) in adults but not children.
Related to whether age is a factor in developing adverse events from vaccines is whether some groups of individuals are more predisposed to experiencing such adverse events than others. Immunodeficient individuals are clearly more susceptible to adverse events following vaccination with live viruses. Consequently, these vaccines are not given to them. Some studies also suggest that influenza virus or pneumococcal vaccines given to HIV-infected individuals with advanced disease can increase the amount of HIV in their bodies (Janoff et al., 1995; O'Brien et al., 1995; Tasker et al., 1995). In addition, one study that used LLDBs reported that neurologically impaired children were at a much greater risk of having seizures following childhood vaccinations than a healthy control group (Davis, unpublished data). Such children are at increased risk of having seizures even in the absence of vaccination.
An FDA representative suggested that large clinical trials of a new vaccine be performed in a heterogeneous population before the vaccine is licensed for
general use to attempt to detect groups that might be at greater risk of experiencing adverse events. Others commented that this would be difficult and costly, however, because of the increased sample sizes required. Currently, because of the need for close monitoring, accessibility to health care, and data analysis, as well as exclusion criteria for the study of experimental vaccines, clinical trials of vaccines are usually done with a homogeneous population of healthy individuals least likely to experience adverse reactions to the vaccines. Because 2 subpopulations can differ both in their risk of experiencing adverse reactions to a vaccine and in their risk of the disease that the vaccine is designed to prevent, studies to evaluate risks in specific subpopulations could provide needed information.
A consumer representative suggested that VAERS data be used to identify categories of children at higher risk of experiencing adverse events following vaccination. Once identified, these children could be screened out of vaccination programs. She also suggested using the database of the federal Vaccine Injury Compensation Program to detect common denominators among individuals who were injured or who died following vaccination. Analysis of that information could be useful for researchers to identify high-risk children who should not be vaccinated in the future. A CDC representative suggested that the large linked databases are best suited to identifying those at risk of adverse events following vaccination, once the association has been established. This individual believed that VAERS does not have the data necessary for a complete epidemiologic study to assess risk factors (i.e., information on vaccinated children who did not experience the event and on children who experienced the event but who were not vaccinated). The use of Vaccine Injury Compensation Program data would be too expensive if laboratory data were required and too imprecise if history alone were used.
An additional question that was addressed is the extent to which vaccines can trigger disorders through an immune reaction. It is known, for example, that some vaccines can trigger the Guillain-Barré syndrome, which is immune mediated. There is thus justification in studying other immune disorders to evaluate whether they could be related to vaccines. A full assessment of the safety of vaccines requires long-term studies that include controls at low risk for developing the disease prevented by the vaccine being tested, a physician suggested.
A final important question is whether there are long-latency adverse events following vaccinations. A vaccine researcher has studied several adults who have experienced deteriorating brain function over a period of years that in some cases has left them permanently comatose (Martin et al., 1994, 1995). Tests of their brain tissues, blood, and the fluid circulating in their brains and spinal cords have ruled out known causes of brain deterioration. According to
the speaker, genetic studies suggest that their illness is caused by a newly identified cytomegalovirus closely related to a type that infects African green monkeys. Although there is no evidence of the possible source of the genetic material identified, because the kidney cells of these monkeys are used to culture the viral strains used in the live-virus polio virus vaccine, the speaker hypothesized that at least some lots of this vaccine might harbor a strain of cytomegalovirus that can cause severe brain deterioration decades after vaccination. Although the monkey kidney cells currently used in oral polio virus vaccine production come from laboratory monkeys that are not thought to be infected with viruses and that are screened for viral contamination, previously unsuspected (and therefore not screened for) viruses or viruses present at very low levels could theoretically contaminate the vaccine. A consumer organization representative suggested that FDA regulations for vaccine safety be reviewed and updated to reflect the state of the art in scientific technology so that vaccines are more thoroughly tested for adventitious agents. This testing should include evaluations of the possible roles of additives such as aluminum and formaldehyde in adverse events.
A CDC representative noted that a large number of possible long-term adverse effects could be considered. Given the limited resources at hand, he suggested limiting studies to those effects that seem to be biologically plausible. According to workshop participants, these findings reinforce the need for long-term studies to determine late-onset adverse effects of vaccines. They also suggest that newly devised laboratory tools for virus detection such as the polymerase chain reaction should perhaps be applied even to those vaccines that have historically been accepted as safe to detect adventitious viruses that might cause long-term adverse effects. Such research findings might help to narrow the search for the long-term adverse effects of vaccines.