Since the beneficial effects of vaccines are a result of changes in the immune system, it would not be surprising if some of the adverse effects were also. A classification of immunologic reactions that can cause disease has been proposed by Coombs and Gell (1968). Four reactions make up the classification: type I, immediate hypersensitivity, the most serious clinical manifestation of which is anaphylaxis; type II, reaction of antibody with tissue antigens; type III, Arthus-type reaction, caused by deposition of antigen-antibody complexes in tissues, leading to the tissue-damaging effects of complement and leukocytes; and type IV, delayed-type hypersensitivity, which is mediated largely by T lymphocytes and macrophages. In clinical reactions to foreign antigens, these categories frequently overlap. These reactions are a by-product of the body's capacity to reject foreign invasion, particularly by microorganisms. If these reactions are responsible for causing adverse events to vaccines, then these reactions would be extensions of the beneficial responses to vaccines, which are mediated by protective immunoglobulin G (IgG) antibodies and T-lymphocyte responses.
Anaphylaxis (a type I reaction) was described in some detail in the Institute of Medicine's report Adverse Effects of Pertussis and Rubella Vaccines (Institute of Medicine, 1991, Chapter 6). The discussion in that chapter applies equally to this report. The term anaphylaxis generally re-
fers to a sudden, potentially life-threatening, systemic condition mediated by highly reactive molecules released from mast cells and basophils. Mediators include histamine, platelet-activating factor, and products of arachidonic acid metabolism (Fisher, 1987). Release of mediators depends typically upon the interaction of antigen with specific antibodies of the IgE class that are bound to the mast cells and basophils. Antibodies of other immunoglobulin classes are thought to mediate anaphylaxis on occasion. By definition, the antibodies are formed by prior exposure to the same or a closely related antigen. Anaphylaxis results from widespread release of mediators that enter the circulation, and thus, anaphylaxis is an expression of allergy that is systemic. At a cellular level, the reaction begins within seconds of exposure to the inciting antigen. However, depending upon the degree of sensitization (IgE antibody formation), and presumably upon the rate with which the antigen enters the circulation, localized or systemic symptoms may not be expressed for minutes or a few hours (Dolovich et al., 1973; Pearlman and Bierman, 1989). In proposed changes to the Vaccine Injury Table, which is used by the Vaccine Injury Compensation Program to determine eligibility for compensation for vaccine-induced injuries, the time frame for the onset of anaphylaxis/anaphylactic shock following vaccination has been set at 4 hours (U.S. Department of Health and Human Services, 1992). Classic symptoms include pallor and then diffuse erythema, urticaria and itching, subcutaneous edema, edema and spasm of the larynx, wheezing, tachycardia, hypotension, and hypovolemic shock (Kniker, 1988; Pearlman and Bierman, 1989). These symptoms are due to leaking of fluid from blood vessels, constriction of smooth-muscle in certain viscera, and relaxation of vascular smooth muscle. If death occurs, it is most commonly from airway obstruction caused by laryngeal edema or bronchospasm, or from cardiovascular collapse from arterial smooth-muscle relaxation and transudation of fluids from the intravascular space (Pearlman and Bierman, 1989). Tissues at autopsy show primarily widespread edema.
Less severe manifestations of immediate hypersensitivity that do not qualify as anaphylaxis under the above definition occur commonly. These may be expressed as urticaria and generalized pruritus, wheezing, or more alarming symptoms such as facial and other edemas. However, hypotension, shock, and collapse do not occur either because the reactions are naturally less severe or because they are aborted by intervention with epinephrine or antihistamines.
The clinical presentation of anaphylaxis can also be produced by intravascular antigen-antibody reactions that activate the complement system. In this case, the antibodies may be of the IgG or IgM class. Peptides that are split from activated complement components act on mast cells and basophils to induce the release of the same mediators (Kniker, 1988). This reaction is recognized most clearly after intravenous administration of anti-
gen; it has been hypothesized to occur rarely after intramuscular or subcutaneous injection through rapid entry (within 1 to 5 minutes) of large amounts of the antigen into the venous circulation. This reaction in an infant presumably could be mediated by IgG antibody received transplacentally from the mother; such antibody would be expected to persist for the first 6 months of life and possibly longer (Benacerraf and Kabat, 1950; Cohen and Scadron, 1946). Anaphylaxis also can occur without an obvious cause (Wiggins et al., 1989).
INTERACTION OF ANTIBODY WITH NORMAL TISSUE ANTIGENS
In type II reactions, antibody combines with an antigen expressed on normal tissue cells, complement is activated, and the resultant inflammation damages the tissue. It is not clear whether this type of reaction is triggered by alteration in the expression of a tissue antigen or by the formation of an antibody to an antigen in food or an invading microorganism that then cross-reacts with a host antigen. Antigens in a vaccine could theoretically mimic a tissue antigen and elicit such a cross-reacting response, but this has not been shown. On first exposure to such an antigen, any resultant tissue reaction would be expected to develop in about 2 or 3 weeks; on reexposure, a tissue reaction might occur within a few days. (These estimates are hypothetical and are based on what is known about primary and secondary antibody responses to foreign antigens.) The basis for type II reactions is not understood.
The Arthus reaction (Arthus, 1903) is mediated differently from either anaphylaxis or type II reactions. Basic to this type III or Arthus reaction is the formation of antigen-antibody complexes, with a moderate excess of antigen, with deposition in the walls of blood vessels, and consequent organ damage. This is not an acute, immediately overwhelming condition. It generally develops over 6 to 12 hours if antibody levels are already high, or it can develop over several days (e.g., in serum sickness) as antibody levels increase and antigen persists. In this reaction, immune complexes in the walls of blood vessels initiate an inflammatory reaction involving complement and leukocytes, particularly neutrophils. Tissue sections show acute inflammation, and profound tissue destruction can occur.
Localized Arthus reactions have been reported to be common at the site of injection of some vaccines and occur when reimmunization is performed in the presence of high levels of circulating IgG antibody (Facktor et al., 1973). They are characterized by pain, swelling, induration, and edema
beginning several hours after immunization and usually reaching a peak 12 to 36 hours after immunization. They are self-limited, resolving over the course of a few days. Their frequency and severity can be lessened by spacing immunizations more widely, as has been recommended for tetanus-diphtheria toxoid booster injections.
Generalized Arthus reactions of a serum sickness-like character have also been invoked following vaccine administration. Such generalized serum sickness-like reactions were common in the era when horse serum was used to treat or prevent many infectious diseases and when very large quantities of immunogenic foreign protein were infused (sometimes repeatedly). These reactions require both IgG antibody and circulating excess antigen. Considering the small quantity of protein in present-day vaccines that is injected, it is not clear that such reactions could occur as a result of immunization. In animal models, symptoms and pathology tend to localize in the kidney, skin, joints, lung, and brain (Henson, 1982). The manifestations after vaccination most commonly ascribed to serum sickness-like mechanisms are arthritis and fever.
Delayed-type hypersensitivity (type IV reaction) results from the stimulation of antigen-specific lymphocytes with the resultant replication of these cells at the site of exposure to antigen. This stimulation induces the release of lymphokines, migration of macrophages to the site, further immunologic stimulation, and resultant tissue damage. As with IgG antibody responses, this form of hypersensitivity represents the normal immunologic response to certain types of foreign antigen, and it is seen commonly after recovery from natural infections. On first exposure, the response peaks after about 3 weeks; on reexposure, the response typically peaks after 24 to 48 hours.
Delayed-type hypersensitivity has been thought to be involved in the development of neurologic complications of vaccination, particularly the development of neurologic disease after receipt of the early rabies vaccines (no longer in use) because of the large quantities of contaminating nervous system antigens in the vaccines. The possible involvement of delayed-type hypersensitivity in reactions to contemporary vaccines is discussed in the chapters on the specific vaccines and adverse events.
EFFECT OF VACCINES ON THE IMMUNE SYSTEM
The capacity of the injection of capsular polysaccharide (PRP) from Haemophilus influenzae type b to transiently decrease antibody specific to (and only to) H. influenzae type b is discussed in Chapter 9. A different question has been raised, that is, the possibility of a generalized immuno-
logic suppression from simultaneous administration of more than one vaccine or vaccine component. Exposure to multiple foreign antigens is a common part of normal extrauterine life. During a single episode of upper respiratory viral infection, humans are exposed, depending on the particular virus involved, to between 4 and 10 foreign proteins, and during a routine ''strep throat'' infection, to between 25 and 50. Moreover, acquisition of a single new bacterium in the gastrointestinal tract, a common and normal event in consumption of everyday foods, or acquisition of one of the apparently harmless bacteria that inhabit the mouth and nose exposes the immune system to at least 50 potential antigens (Goldblatt et al., 1990). Each one of these foreign molecules typically contains numerous epitopes (antigenic determinants), each of which evokes a separate immune response. Moreover, each of the proteins is broken down in the body to expose still other epitopes, which may be antigenic depending on the genetic background of the host. The normal child may not respond to each of these proteins/epitopes, but in the case of Branhamella catarrhalis, a bacterium that inhabits the nasopharynxes of all normal children, antibodies to 17 different proteins can be detected after colonization (Goldblatt et al., 1990).
Infants, since they are born out of a germ-free environment into a world replete with microorganisms, undergo constant exposure to foreign antigens as their mucosal surfaces are populated with normal bacterial flora and as they are exposed to potentially more pathogenic microorganisms in the environment. During such encounters, the microorganisms would be expected to shed large amounts of their antigens for a period of days. The gradual rise in the levels of circulating immunoglobulins represents one part of the total immunologic response to this onslaught.
In the face of these normal events, it seems unlikely that the number of separate antigens contained in childhood vaccines, whether given orally or by injection, would represent an appreciable added burden on the immune system that would be immunosuppressive. Nevertheless, it is theoretically possible that some vaccine constituent might predispose an individual to infection through its action as an antigen or some other means. The combination of diphtheria-pertussis-tetanus vaccine has been the object of some research, in this regard, in part because pertussis toxin modulates certain immune functions in experimental animals. Fears that four deaths from bacterial infection after a trial of acellular pertussis vaccine in Sweden might have been due to the vaccine were allayed by a subsequent study (Storsaeter et al., 1988) that failed to find an increase in the number of patients hospitalized with bacterial infections after receipt of the vaccine. A report from Israel described an increase in minor infectious illnesses in the 30 days after administration of diphtheria and tetanus toxoids and pertussis vaccine (DPT) (Jaber et al., 1988). However, it is not possible to evaluate these results because of a combination of reporting bias, learning effect, and
modification of illness incidence by season. Subsequently, three investigations have reported a lower or unchanged incidence of both minor and major infections after immunizations (Black et al., 1991; Davidson et al., 1991; Joffe et al., 1992). All three studies included a case-control design to examine serious infections. Two studies were conducted in the Kaiser Permanente health maintenance organization, and in both of these, a statistically significant decrease in infections after immunization was demonstrated. There were, however, potential confounding covariants, such as breast-feeding, frequency of well-child visits, and attendance at day care, so that a protective effect could not be unequivocally assigned to DPT. The conclusion of all three studies was, nevertheless, that no association of increased susceptibility to infection could be demonstrated in the weeks following DPT and oral polio vaccine immunizations. Since several hundred individuals were examined in the three studies, an infrequent association between immunization and infectious disease was not excluded.
It is also well known that many natural vital infections, particularly measles, can temporarily suppress components of the immune system (Starr and Berkovich, 1964; Ward et al., 1991). and there have been concerns that live attenuated viral vaccines might have a similar effect. Soon after the live attenuated measles vaccine was developed, it was. shown that immunization temporarily suppressed the delayed-type hypersensitivity skin test response to purified protein derivative, an index of cell-mediated immunity to Mycobacterium tuberculosis (Brody and McAlister, 1964; Starr and Berkovich, 1964). The suppression was, however, less consistent and less prolonged than that following natural measles infection, presumably because of the attenuation of growth of the vaccine virus at all levels. Other viral vaccines, both live attenuated and inactivated, have been shown to have similar, although often mild and inconstant, effects on skin test responses to various antigens (Berkovich et al., 1972; Brody et al., 1964; Ganguly et al., 1976; Kupers et al., 1970). In addition, more recent studies have shown that, after measles immunization or reimmunization, certain lymphocyte functions, such as the ability to replicate when stimulated with phytohemagglutinin or to excrete certain chemotactic factors, are mildly but measurably depressed (Hirsch et al., 1981), and the number of CD8-positive lymphocytes falls slightly (Nicholson et al., 1992).
It may be asked, then, whether use of combination viral vaccines might exacerbate the potential problem of immune system suppression. The committee found no report of a systematic comparison of the effects of monovalent and polyvalent live attenuated vaccines on immunity. Combined measles-mumps-rubella vaccine (MMR) has been reported to have a temporary suppressive effect on neutrophil function (Toraldo et al., 1992) and on the ability of lymphocytes to proliferate in response to phytohemagglutinin or Candida albicans stimulation (Munyer et al., 1975), but neither study compared monovalent
and trivalent vaccines. MMR vaccination was included in one of the large health maintenance organization case-control studies of possible serious bacterial infection after immunization (Black et al., 1991); although there was a trend toward fewer infections in vaccinees than in controls, none of the differences was significant.
At present, the data are insufficient to answer with certainty whether immunosuppression in the form of laboratory and skin test abnormalities after the receipt of a vaccine does, in fact, indicate a decrease in the capacity to resist infection. Therefore, as new vaccines are developed and as the old ones are used at different dosages or in different combinations or are administered at different ages, this question should continue to be a concern. To date, studies of current vaccines suggest that if immunization leads to an infection, it must do so infrequently.
Arthus M. Injections répétées de sérum de cheval chez le lapin. Comptes Rendus des Séances de la Société de Biologie et ses Filiales (Paris) 1903;55:817-820.
Benacerraf B, Kabat EA. A quantitative study of the Arthus phenomenon induced passively in the guinea pig. Journal of Immunology 1950;64:1-19.
Berkovich S, Fikrig S, Brunell PA, Portugalaza C, Steiner M. Effect of live attenuated mumps vaccine virus on the expression of tuberculin sensitivity. Journal of Pediatrics 1972;80:84-87.
Black SB, Cherry JD, Shinefield HR, Fireman B, Christenson P. Lampert D. Apparent decreased risk of invasive bacterial disease after heterologous childhood immunization. American Journal of Diseases of Children 1991;145:746-749.
Brody JA, McAlister R. Depression of tuberculin sensitivity following measles vaccination. American Review of Respiratory Diseases 1964;90:611-617.
Brody JA, Overfield T, Hammes LM. Depression of the tuberculin reaction by vital vaccines. New England Journal of Medicine 1964;271:1294-1296.
Cohen P, Scadron SJ. Effects of active immunization of mother upon offspring. Journal of Pediatrics 1946;29:609-619.
Coombs RR, Gell PG. Classification of allergic reactions responsible for clinical hypersensitivity and disease. In: Gell PG, Coombs RR, eds. Clinical Aspects of Immunology, 2nd edition. Oxford: Blackwell: 1968.
Davidson M, Letson W, Ward JI, Ball A, Bulkow L, Christenson P, et al. DTP immunization and susceptibility to infectious diseases. American Journal of Diseases of Children 1991;145:750-754.
Dolovich J, Hargreave FE, Chalmeis R, Shier KJ, Cauldie J, Bienenstock J. Late cutaneous allergic responses in isolated-IgE-dependent reactions. Journal of Allergy and Clinical Immunology 1973;52:38-46.
Facktor MA, Bernstein RA, Fireman P. Hypersensitivity to tetanus toxoid. Journal of Allergy and Clinical Immunology 1973;52:1-12.
Fisher M. Anaphylaxis. Diseases of Man 1987;33:433-479.
Ganguly R, Cusumano CL, Waldman RH. Suppression of cell-mediated immunity after infection with attenuated rubella virus. Infection and Immunity 1976;13:464-469.
Goldblatt D, Turner MW, Levinsky RJ. Branhamella catarrhalis: antigenic determinants and
the development of the IgG subclass response in childhood. Journal of Infectious Diseases 1990;162:1128-1135.
Henson PM. Antibody and immune-complex-mediated allergic and inflammatory reactions. In: Lachmann PJ, Peters DK, eds. Clinical Aspects of Immunology, 4th edition. Oxford: Blackwell: 1982.
Hirsch RL, Mokhtarian F, Griffin DE, Brooks BR, Hess J, Johnson RT. Measles virus vaccination of measles seropositive individuals suppresses lymphocyte proliferation and chemotactic factor production. Clinical Immunology and Immunopathology 1981;21:341-350.
Institute of Medicine. Adverse Effects of Pertussis and Rubella Vaccines. Washington, DC: National Academy Press; 1991.
Jaber L, Shohat M, Mimouni M. Infectious episodes following diphtheria-pertussis-tetanus vaccination: a preliminary observation in infants. Clinical Pediatrics 1988;27:491-494.
Joffe LS, Glode MP, Gutierrez MK, Wiesenthal A, Luckey DW, Harken L. Diphtheria-tetanus toxoids-pertussis vaccination does not increase the risk of hospitalization with an infectious illness. Pediatric Infectious Disease Journal 1992;11:730-735.
Kniker WT. Anaphylaxis in children and adults. In: Bierman CW, Pearlman DW, eds. Allergic Diseases from Infancy to Adulthood. Philadelphia: W.B. Saunders Co.; 1988.
Kupers TA, Petrich JM, Holloway AW, St. Geme JW, Jr. Depression of tuberculin delayed hypersensitivity by live attenuated mumps virus. Journal of Pediatrics 1970;76:716-721.
Munyer TP, Mangi RJ, Dolan T, Kantor FS. Depressed lymphocyte function after measles-mumps-rubella vaccination. Journal of Infectious Diseases 1975;134:75-78.
Nicholson JKA, Holman RC, Jones BM, McDougal JS, Sprauer MA, Markowitz LE. The effect of measles-rubella vaccination on lymphocyte populations and subpopulations in HIV-infected and healthy individuals. Journal of Acquired Immune Deficiency Syndromes 1992;5:528-537.
Pearlman DS, Bierman CW. Allergic disorders. In: Stiehm ER, ed. Immunologic Disorders in Infants and Children, 3rd edition. Philadelphia: W.B. Saunders Co.; 1989.
Starr S, Berkovich S. Effects of measles, gamma-globulin-modified measles and vaccine measles on the tuberculin test. New England Journal of Medicine 1964;270:386-391.
Storsaeter J, Olin P, Renemar B, Lagergard T, Norberg R, Romanus V, Tiru M. Mortality and morbidity from invasive bacterial infections during a clinical trial of acellular pertussis vaccines in Sweden. Pediatric Infectious Disease Journal 1988;7:637-645.
Toraldo R, Tolone C, Catalanotti P, Ianniello R, D'Avanzo M, Canino G, et al. Effect of measles-mumps-rubella vaccination on polymorphonuclear neutrophil functions in children. Acta Paediatrica 1992;81:887-890.
U.S. Department of Health and Human Services. U.S. Public Health Service, National Vaccine Injury Compensation Program; Revision of the Vaccine Injury Table; proposed rule. Federal Register, 42 CFR Part 100, August 14, 1992;57(158):36877-36885.
Ward B J, Johnson RT, Vaisberg A, Jauregui E, Griffin DE. Cytokine production in vitro and the lymphoproliferative defect of natural measles virus infection. Clinical Immunology and Immunopathology 1991:61:236-248.
Wiggins CA, Dykewicz MS, Patterson R. Idiopathic anaphylaxis: a review. Annals of Allergy 1989;62:1-4.