discussed a selected number of these populations. The final section presents a preliminary research agenda of possible or putative links in these areas.
IMMUNE RESPONSE TO VACCINE ANTIGENS13
Infection-causing microbes and the vaccines designed to combat them have portions of proteins called antigens. These antigens stimulate a number of cells in the immune system, including macrophages, T cells, and B cells. An immune response begins when macrophages ingest antigens such as proteins entering the body and digest them into antigen fragments. A molecule called MHC (major histocompatibility complex) carries certain of these fragments to the surface of the cell, where they are displayed but they are still locked into the cleft of the MHC molecule. These displayed antigen fragments are recognized by T cells, which stimulate B cells to secrete antibodies to the fragments as well as prompt other immune defenses. According to Berkower, studies suggest that because T cells only recognize antigen fragments from proteins predigested by macrophages, they cannot distinguish between a specific antigen fragment that comes from an infecting microbe and the same antigen fragment that comes from a vaccine (Chisari and Ferrari, 1995).
Stimulated immune cells secrete a variety of chemical substances called cytokines, which determine which class of antibodies are generated. The cytokine interleukin 4, for example, can prompt B cells to secrete immunoglobin E (IgE) antibodies, which trigger allergic reactions. Other cytokines cause B cells to preferentially secrete IgG, which is mainly found in the blood, or IgA, most of which is found in body fluids.
Different MHC molecules bind to different antigen fragments; the set of MHC molecules and the genes that commandeer their production vary widely from one individual to another. Therefore, although the immune systems of two people may respond to the same protein in a vaccine, their T cells may respond to different portions of that protein. This diversity fosters differences in responses to vaccine antigens.
There are at least nine chemically distinct classes of immunoglobulins, and the balance of the various types of cytokines that stimulate antibody secretion determines the final response to a vaccine. Consequently, vaccine responses can differ between individuals because the same vaccine stimulates different individuals to generate different amounts of the various cytokines. Vaccine responses might not only differ in the short term, but they could also vary in the
long term if they trigger a predominantly IgE response. This response could trigger an allergic reaction to future immunizations with the same antigens (IOM, 1994a).
To prevent the body's immune system from destroying its own tissues in what is known as an autoimmune response, immature T cells that react against self-antigens are thought to be destroyed in the thymus gland, creating what is known as central tolerance. Peripheral tolerance might also occur, whereby those T cells that could potentially react to self-antigens and that are not destroyed in the thymus are somehow prevented from causing an autoimmune reaction. Although studies suggest that peripheral tolerance exists, at least in experimental animals, the mechanism for the process is not yet known. If peripheral tolerance exists in people, an autoimmune response might occur in response to vaccination if the vaccine somehow disrupts that peripheral tolerance or if peripheral tolerance is not strong on the day of vaccination (Miller et al., 1989).
Some people have suggested that vaccines can stimulate autoimmune reactions if some of the antigen fragments in vaccines resemble a person's self-antigens. However, it is unclear why an immune system that is tolerant of its own self-antigens would respond to a self-antigen mimic in a vaccine. Berkower speculated that vaccines might counter peripheral tolerance and foster an autoimmune reaction if they contain molecular mimics of self-antigens that are usually not exposed to T cells, because peripheral tolerance seems to depend on the continuous presence of an antigen.
Nakhasi suggested that an autoimmune response might be instigated by a vaccine or by natural infection if the microbial antigens bind to self-antigens in infected cells and change the antigens' shape such that they are no longer tolerated and can elicit an immune response.
According to McFarland, researchers suspect that molecular mimicry, which could possibly lead to an autoimmune disorder, might be occurring between self-antigens and antigens from microbes or vaccines if the two antigens share much of the same chemical structure. Recent studies suggest, however, that they need to have a similar structure only in the narrow region that binds to the T-cell receptor (Vogt et al., 1994; Wucherpfennig et al., 1994). In addition, the amino acids in this region do not have to be identical; rather studies suggest that they must have the same basic chemical and charge properties (Vogt et al., 1994; Wucherpfennig et al., 1994; Vergelli et al., 1996).
Some researchers have hypothesized that autoimmune diseases may be stimulated by viruses (Fujinami et al., 1985; Westall and Root-Bernstein, 1983). Westall and Root-Bernstein have postulated that this may occur if three criteria are met. The first one is that antigens are present that have molecular structure similar to self antigens found in certain human tissue (Westall and Root-
Bernstein, 1983). The second one is that an adjuvant derived from walls of the bacterial cells used in the vaccine's manufacture is present. The third one is that antigens exhibiting complimentarity are present. When these three criteria are present an antigenic challenge to the host evokes autoimmunity. This hypothesis has been named Multiple-Antigen-Mediated-Autoimmunity (MAMA) Syndrome. Root-Bernstein has presented evidence that it exists in AIDS patients who develop encephalomyelitis (Sela and Arnon, 1992). Waisbren suggested that the MAMA Syndrome might be a mechanism that induces autoimmunity in susceptible individuals after vaccination with virus components that are also present in human tissue. These antigens may be associated with other complimentary viral antigens and bacterial cell wall components present in the patient who has been vaccinated.
Each year about 300,000 premature infants are born in the United States, and most of these survive past infancy (Ventura et al., 1996). Preliminary studies suggest that premature large birthweight infants have the same or less risk of experiencing the usual adverse effects reported for full-term infants following vaccination with HBV or DPT (Bernbaum et al., 1989; Koblan et al., 1988; Ramsay et al., 1995; Losonsky, unpublished data). However, recent data with DPT immunization in very low birth weight infants suggest that moderate to severe adverse reactions can occur following routine vaccination with DPT in infants born less than 1,000 grams (Sudhakarow et al., 1996). Routine immunization of extremely premature infants is associated with significant adverse events. According to Losonsky, larger studies that cover the full range of vaccines given to infants need to be done to assess the risk of common adverse events in premature infants.
Current vaccine schedules are based largely on the immune responses seen in full-term rather than premature infants. The human immune system matures throughout fetal life and up until about 2 years of age, however, so even full-term infants do not develop an adequate protective immune response to most vaccines administered in the first weeks of life. Losonsky expressed concern that premature infants given vaccines at the standard times may not develop adequate immunity to the diseases against which the vaccines are designed to protect. There is also speculation that some premature infants might develop immunologic tolerance to the viral or bacterial antigens in vaccines such that
they do not develop an immune response to subsequent vaccines given later in life.
According to Losonsky, 98 percent of full-term infants achieve protective immune responses after receiving the third dose of HBV (Losonsky, unpublished data). A study of HBV given to premature infants, however, found that only 54 percent of infants with birth weights of less than 1,000 grams and 70 percent of infants with birth weights of 1,000 to 1,500 grams attained protective immune responses after receiving the third dose of the vaccine. Thirteen of the premature infants with little or no detectable antibody responses to the three doses of HBV were given a fourth dose of the vaccine at 9 to 12 months of age. Only 46 percent of these revaccinated infants achieved protective antibody levels (Losonsky, unpublished data).
These data suggest, Losonsky noted, the possibility that tolerance develops in some premature infants. Researchers are following those infants who did not respond to the fourth HBV dose to see if their lack of response is permanent. Other studies suggest that the immune systems of premature infants given DPT and oral polo vaccine at standard times respond as well as those of full-term infants (Koblin et al., 1988; Conway et al., 1993; D'Angio et al., 1995). However, a decreased immune response to some enhanced-potency inactivated polio-oral polio vaccine combinations and to certain Haemophilus influenzae type b conjugate vaccines was seen in premature infants. The response depended on gestational age and weight at the time of the first immunization (Munoz et al., 1995; D'Angio et al., 1995).
According to Losonsky, these findings suggest that vaccine schedules, dosages, and combinations for preterm infants may have to differ from those for full-term infants. She added that the best schedule and dosage for each vaccine cannot be predicted on theoretical grounds but can only be determined by further study. Care must be taken, she said, not to leave premature infants vaccinated but unprotected from disease. A representative of the National Institute of Allergy and Infectious Diseases (NIAID) pointed out, however, that the risk of inducing tolerance to vaccination for a specific disease must be balanced with the risk of the infant dying or suffering long-term consequences from natural infection. Premature infants, for example, are more likely to have respiratory abnormalities that may put them at greater risk for complications from the respiratory disease pertussis (whooping cough).