153
Immunology Today, vol. 6, No. 5, 1985
New
directions in research
Pathogenesis of milk hypersensitivity Because cow's milk has been represented as an ideal food, particularly for human babies, its consumption has increased and become more widespread. However, concern has grown about health hazards related to milk, including hypersensitivities to milk proteins, milk-borne infections, a variety of metabolic and cardiovascular diseases, and possibly other disorders. These hazards have been discussed in a constant flow of clinical reports, research investigations and at least three independent review books 1-s. Milk hypersensitivity has provoked particular interest and while some questions have been clarified, many others have emerged and stimulated several provocative investigations. As a foreign antigen ingested by normal subjects, bovine milk evokes certain immunologic responses that in most instances cause no untoward effect. Bottle-fed infants have circulating antibodies to milk proteins that rise to a peak at 3 to 6 months, gradually decline during late childhood, and reach low levels in adults 4-6. These antibodies are mainly of the three major immunoglobulin classes, mostly IgG, to a lesscr extent IgA, and least IgM. IgD and IgE antibodies may be detected as well. The nature and intensity of immunologic responses to milk, or other foods, vary widely from subject to subject and seem to be influenced by several factors, including the genetic coding, the quantity and nature of ingested allergen, the degree of digestion, and the histologic and immunologic integrity of the gut mucosa. In what way does the immunologic response in milksensitive subjects differ from that in normal subjects? A wide variety of 'abnormal' immunologic findings, encompassing all components of the immune system, has been connected with milk hypersensitivity 1'7. The corollary is that, underlying clinical hypersensitivity reactions to milk, there are heterogeneous immunologic mechanisms. Clinical and laboratory observations indicate that any of the four basic types of hypersensitivity reactions delineated by Gell and Coombs may be involved and that more than one type of reaction may be seen in a particular patient, even when there is a single clinical manifestation. However, in many instances the immunologic mechanism causing the disorder cannot be defined by current techniques. Type I anaphylactic or immediate hypersensitivity is primarily IgE-mediated and is usually provoked by a small quantity of allergen. The antigen-antibody reaction occurs at the surface of mast cells or basophils and results in the release of histamine and other mediators. IgG4 antibody (short-term reagin) was thought to have a role in this type of response in some patients 8 but this observation could not be confirmed 9'1°. The longstanding belief that an increased allergen load enhances sensitization has been recently challenged by Firer et al. lJ and Gerrard and Shenassa lz, who noted that infants ingesting small quantities of bovine milk had higher titers of IgE antibodies to milk than did those who ingested
large quantities. This observation needs further investigation and, if substantiated, should be considered when planning the feeding pattern of infants at high risk of atopy. Type I reactions may result in vomiting, diarrhea, colic, urticaria, angioedema, eczema, rhinorrhea, bronchospasm, or systemic anaphylaxis. A modified anaphylactic reaction has been proposed as a cause of cot death in some infants after tracheal aspiration of milk Is, but the available data have not been convincing. Diagnostic studies indicative of type I reactions are tests for IgE antibodies, including immediate-type hypersensitivity skin testing, serum-specific IgE antibodies by radioallergosorbent testing (RAST), enzyme-linked immunosorbent assay, radioimmunoelectrophoresis, or radioimmunodiffusion, leukocyte histamine release, passive cutaneous transfer, and immunofluorescence studies of affected tissue. Type II, cytotoxic or cytolytic hypersensitivity, involves antigen-antibody reaction at the cell surface and results in cell damage. The antibody is usually of the IgG or IgM class, and the reaction activates the complement cascade and attracts mononuclear cells. The type II reaction has not been clearly elucidated in milk hypersensitivity. Perhaps it explains the reduction in platelet count that occurs in certain food-sensitive patients after an oral challenge with the offending food 14. It is likely also to be the operative mechanism in cases of severe thrombocytopenia noted in connection with milk ingestion 15'16. Apart from cytolysis, activation of complement results in the cleavage of C3a and C5a that have anaphylatoxic activities and thus can degranulate mast cells and result in reactions similar to those mediated by IgE antibodies. Type II reactions are often associated with a drop in the serum hemolytic complement activity or of the concentrations of individual complement components. Type III, Arthus-type or immune-complex hypersensitivity, involves fixation of complement to antigenantibody complexes formed in slight-to-moderate degrees of antigen excess. The antibody is usually of the IgG class (except IgG4 subclass, which does not fix complement) but can also be of the IgM or IgA class. In occasional cases, igE complexes may be formed ~7. When sufficien~ quantities of the circulating complexes become lodged in the postcapillary venules of the target organ, they initiate vasculitis and attract mononuclear cells that further enhance tissue damage. IgE antibodies are not essential in this reaction, but a concomitant IgE-mediated reaction would increase the vascular permeability and enhance local deposition of the complexes. Type III reaction seems to play a role in the pathogenesis of several milk hypersensitivity disorders, such as gastro-intestinal bleeding in infants, gastroenteropathy, cutaneous vasculitis, and chronic pulmonary disease (Heiner syndrome). There is evidence, at least in experimental animals, that milk hypersensitivity can cause rheumatoid-like joint lesions ~8. Laboratory findings that support type III reaction include © 1985, Elsevier Science Pubfshers B.V., Amsterdam
0167
q919/85/$02.00
154
increased titers of precipitating, hemagglutinating or complement fixing antibodies, and decreased serum hemolytic complement activity or individual complement components. More specific, however, is the demonstration of milk-specific immune complexes in the circulation 19.20or in the shock organ 21. Type IV, cell-mediated or delayed hypersensitivity responses require sensitized T lymphocytes that respond to the specific antigen by proliferation, release oflymphokines, and attraction of mononuclear cells and neutrophils. The end result is tissue damage. The little information available suggests that type IV reactions might have a role in some milk hypersensitivity disorders such as certain gastro-intestinal manifestations 22'23. Laboratory tests for type IV reactivity include in-vitro lymphoblast transformation or release of lymphokines when the patient's lymphocytes are exposed to the specific milk proteins. The application of the classification above to individual patients is sometimes difficult. In many instances of obvious clinical allergic reactions to milk, as verified by the elimination-challenge test, the currently available laboratory tests fail to clarify the underlying immunologic mechanism. On the other hand, abnormal laboratory findings may be noted in individuals drinking milk ad libitum without any apparent ill effect. Several factors are responsible for this discrepancy, including our lack of understanding of the mechanisms of milk hypersensitivity. Apart from selecting the inappropriate test, the problem may be in using the wrong antigen or a nonpurified preparation. More than 30 antigenic components have been identified in bovine milk2~; each is capable of inducing specific immune responses. A test result may sometimes be positive when a purified antigen is used but negative when whole milk is used. Furthermore, in certain instances, the actual antigen may be a new antigen formed by processing, digestion, or other modification of the native protein. For example, eight new antigens could be generated by successive pepsin hydrolysis offl-lactoglobulin 25. Most diagnostic procedures presently available are not well standardized and require substantial experience in doing the test and interpreting the results. Skin testing and serum IgE antibody determination by R A S T are the most common procedures in screening for food allergy and would be the most appropriate for type I hypersensitivities, but their reliability is far from optimal. Recently we compared the results of these two tests with the result of double-blind oral challenge tests with various foods 26'27. Skin testing was done first by the scratch technique and if the reaction was equivocal, the test was repeated intradermally. R A S T was determined by a paper-disc radioimmunoassay. In case of milk, agreement with the challenge result (whether positive or negative) was 53 % for skin testing and 60 % for RAST. The test sensitivity tended to be higher for R A S T (56 % vs. 44 %), but the two tests had similar specificity and predictive accuracy (Table I). In the positive challenge group, skin testing and
Immunology Today, ool. 6, No. 5, 1985
R A S T were both positive in only 33% and both negative in 11%. In the negative challenge group, tests were both negative in only 50% and both positive in 16%. Various degrees of discrepancy were also noted when food IgE T A B L E I. Skin testing and R A S T compared with double-blind oral challenge in milk hypersensitivity Test characteristic Sensitivity (true positive test) Specificity (true negative test) Positive predictive accuracy Negative predictive accuracy
Skin testing
RAST
44 % 67 % 67 % 44 %
56 % 67 % 71% 50 %
antibodies were measured in the serum by different radioimmunoassay techniques 28. In summary, therefore, our understanding of the mechanisms of milk hypersensitivity leaves a lot to be desired. The most urgent need at present is the development of more reliable diagnostic tests. []-]
Acknowledgement I thank Miss Virginia H o w a r d and Miss AngelaJones for assistance in preparation of this manuscript. SAMI L. BAHNA Section of Allergy and Immunology, Department of Pediatrics, Louisiana State, University School of Medicine, New Orleans, LA 70112, USA.
References 1 Bahna, S. L. and Heiner, D. C. (1980) Allergies to Milk. Grune and Stratton, New York 2 Delmont, J. (ed.) (1983) Milk Intolerances and Rejection. Karger, Basel 3 Freed, D. L. J. (ed.) (1984) Health Hazards offVIilk. Bailliere Tindall, London. 4 Rothberg, R. M. and Farr, R. S. (1965) Pediatrics 35, 571-588 5 Kletter, B., Gery, I., Freier, S., and Davies, A. M. (1971) Int. Arch. Allergy Appl. Immunol 40, 656-666 6 May, C. D., Foman, S.J. and Remigio, L. (1982) Acla. Paediat. Scand 71, 43-51 7 Bahna, S. L. and Gandhi, M. D. (1983) Ann. Allergy 50, 218-224 8 Parish, W. E. (1971) Clin. Allergy 1, 369-380 9 Bj6rksten, B., Ahlstedt, S., Bj6rksten, F., Carlsson, B., Fallstrom, S. P., Juntunen, K., Kajosaari, M. and Kober, A. (1983) Allergy 38, 119-124 I0 Shakib, F. (1984)Int. Arch. Allergy Appl. Immunol. 75, 107-112 11 Firer, M. A., Hosking, C. S. and Hill, D. J. (1981) Br. Med. J. 283, 693-696 12 Gerrard, J. W. and Shenassa, M. (1983) Ann. Allergy 50, 375-379 13 Coombs, R: R. A. and MeLaughlin, P. (1982) Lancet i, 1388-1389 14 Osvath, P. and Markus, M. (1968) Acta Paediat. Acad Sci. Hung. 9, 279-284 15 Whitfield, M. F. and Barr, D. G. D. (1976)Arch. Dis. Child. 51,337 343 16 Jones, R. H. T. (1977)Arch. Dis. Child. 52, 744-745 17 Brostoff, J., Carini, C., Wraith, D. G. and Johns, P. (1979) Lancet i, 1268-1270 18 Coombs, R. R. A. and Oldham, G. (1981) Int. Arch. Allergy Appl. Immunol. 64, 287-292 19 Frick, O. L. (1982)J. Allergy Clin. Immunol. 69 (supplement), 157 20 Paganelli, R., Matricardi, P. M. and Aiuti, F. (1984) Clin. Rev. Allergy 2, 69 21 Shiner, M., Ballard, J. and Smith, M. E. (1975) Lancet i, 136-140 22 Endr6, L. and Osvath, P. (1975)Acta Allergol. 30, 34-42 23 Scheinmann, P., Gendrel, S., Charlas, J. and Paupe, J. (1976) Clin. Allergy 6, 515-521 24 Gjesing, B. and L~wenstein, H. (1984) Ann. Allergy 53, 602-608 25 Spies, J. R., Stevan, M. A., Stein, W. J. and Coulson, E. J. (1970) j. Allergy 45, 208-219 26 Gandhi, M. D. and Bahna, S. L. (1985)J. Allergy Clin. Immunol. 75,205 27 Bahna, S. L. and Gandhi, M. D. (1985)J. Allergy Clin. ImmunoL 75,204 28 Firestone, J. E., Jr. and Bahna, S. L. (1985)J. Allergy Clin. Immunol. 75, 2O4
Immunology Today, vol. 6, No. 5, 1985
New
directions in research
Pathogenesis of milk hypersensitivity Because cow's milk has been represented as an ideal food, particularly for human babies, its consumption has increased and become more widespread. However, concern has grown about health hazards related to milk, including hypersensitivities to milk proteins, milk-borne infections, a variety of metabolic and cardiovascular diseases, and possibly other disorders. These hazards have been discussed in a constant flow of clinical reports, research investigations and at least three independent review books 1-s. Milk hypersensitivity has provoked particular interest and while some questions have been clarified, many others have emerged and stimulated several provocative investigations. As a foreign antigen ingested by normal subjects, bovine milk evokes certain immunologic responses that in most instances cause no untoward effect. Bottle-fed infants have circulating antibodies to milk proteins that rise to a peak at 3 to 6 months, gradually decline during late childhood, and reach low levels in adults 4-6. These antibodies are mainly of the three major immunoglobulin classes, mostly IgG, to a lesscr extent IgA, and least IgM. IgD and IgE antibodies may be detected as well. The nature and intensity of immunologic responses to milk, or other foods, vary widely from subject to subject and seem to be influenced by several factors, including the genetic coding, the quantity and nature of ingested allergen, the degree of digestion, and the histologic and immunologic integrity of the gut mucosa. In what way does the immunologic response in milksensitive subjects differ from that in normal subjects? A wide variety of 'abnormal' immunologic findings, encompassing all components of the immune system, has been connected with milk hypersensitivity 1'7. The corollary is that, underlying clinical hypersensitivity reactions to milk, there are heterogeneous immunologic mechanisms. Clinical and laboratory observations indicate that any of the four basic types of hypersensitivity reactions delineated by Gell and Coombs may be involved and that more than one type of reaction may be seen in a particular patient, even when there is a single clinical manifestation. However, in many instances the immunologic mechanism causing the disorder cannot be defined by current techniques. Type I anaphylactic or immediate hypersensitivity is primarily IgE-mediated and is usually provoked by a small quantity of allergen. The antigen-antibody reaction occurs at the surface of mast cells or basophils and results in the release of histamine and other mediators. IgG4 antibody (short-term reagin) was thought to have a role in this type of response in some patients 8 but this observation could not be confirmed 9'1°. The longstanding belief that an increased allergen load enhances sensitization has been recently challenged by Firer et al. lJ and Gerrard and Shenassa lz, who noted that infants ingesting small quantities of bovine milk had higher titers of IgE antibodies to milk than did those who ingested
large quantities. This observation needs further investigation and, if substantiated, should be considered when planning the feeding pattern of infants at high risk of atopy. Type I reactions may result in vomiting, diarrhea, colic, urticaria, angioedema, eczema, rhinorrhea, bronchospasm, or systemic anaphylaxis. A modified anaphylactic reaction has been proposed as a cause of cot death in some infants after tracheal aspiration of milk Is, but the available data have not been convincing. Diagnostic studies indicative of type I reactions are tests for IgE antibodies, including immediate-type hypersensitivity skin testing, serum-specific IgE antibodies by radioallergosorbent testing (RAST), enzyme-linked immunosorbent assay, radioimmunoelectrophoresis, or radioimmunodiffusion, leukocyte histamine release, passive cutaneous transfer, and immunofluorescence studies of affected tissue. Type II, cytotoxic or cytolytic hypersensitivity, involves antigen-antibody reaction at the cell surface and results in cell damage. The antibody is usually of the IgG or IgM class, and the reaction activates the complement cascade and attracts mononuclear cells. The type II reaction has not been clearly elucidated in milk hypersensitivity. Perhaps it explains the reduction in platelet count that occurs in certain food-sensitive patients after an oral challenge with the offending food 14. It is likely also to be the operative mechanism in cases of severe thrombocytopenia noted in connection with milk ingestion 15'16. Apart from cytolysis, activation of complement results in the cleavage of C3a and C5a that have anaphylatoxic activities and thus can degranulate mast cells and result in reactions similar to those mediated by IgE antibodies. Type II reactions are often associated with a drop in the serum hemolytic complement activity or of the concentrations of individual complement components. Type III, Arthus-type or immune-complex hypersensitivity, involves fixation of complement to antigenantibody complexes formed in slight-to-moderate degrees of antigen excess. The antibody is usually of the IgG class (except IgG4 subclass, which does not fix complement) but can also be of the IgM or IgA class. In occasional cases, igE complexes may be formed ~7. When sufficien~ quantities of the circulating complexes become lodged in the postcapillary venules of the target organ, they initiate vasculitis and attract mononuclear cells that further enhance tissue damage. IgE antibodies are not essential in this reaction, but a concomitant IgE-mediated reaction would increase the vascular permeability and enhance local deposition of the complexes. Type III reaction seems to play a role in the pathogenesis of several milk hypersensitivity disorders, such as gastro-intestinal bleeding in infants, gastroenteropathy, cutaneous vasculitis, and chronic pulmonary disease (Heiner syndrome). There is evidence, at least in experimental animals, that milk hypersensitivity can cause rheumatoid-like joint lesions ~8. Laboratory findings that support type III reaction include © 1985, Elsevier Science Pubfshers B.V., Amsterdam
0167
q919/85/$02.00
154
increased titers of precipitating, hemagglutinating or complement fixing antibodies, and decreased serum hemolytic complement activity or individual complement components. More specific, however, is the demonstration of milk-specific immune complexes in the circulation 19.20or in the shock organ 21. Type IV, cell-mediated or delayed hypersensitivity responses require sensitized T lymphocytes that respond to the specific antigen by proliferation, release oflymphokines, and attraction of mononuclear cells and neutrophils. The end result is tissue damage. The little information available suggests that type IV reactions might have a role in some milk hypersensitivity disorders such as certain gastro-intestinal manifestations 22'23. Laboratory tests for type IV reactivity include in-vitro lymphoblast transformation or release of lymphokines when the patient's lymphocytes are exposed to the specific milk proteins. The application of the classification above to individual patients is sometimes difficult. In many instances of obvious clinical allergic reactions to milk, as verified by the elimination-challenge test, the currently available laboratory tests fail to clarify the underlying immunologic mechanism. On the other hand, abnormal laboratory findings may be noted in individuals drinking milk ad libitum without any apparent ill effect. Several factors are responsible for this discrepancy, including our lack of understanding of the mechanisms of milk hypersensitivity. Apart from selecting the inappropriate test, the problem may be in using the wrong antigen or a nonpurified preparation. More than 30 antigenic components have been identified in bovine milk2~; each is capable of inducing specific immune responses. A test result may sometimes be positive when a purified antigen is used but negative when whole milk is used. Furthermore, in certain instances, the actual antigen may be a new antigen formed by processing, digestion, or other modification of the native protein. For example, eight new antigens could be generated by successive pepsin hydrolysis offl-lactoglobulin 25. Most diagnostic procedures presently available are not well standardized and require substantial experience in doing the test and interpreting the results. Skin testing and serum IgE antibody determination by R A S T are the most common procedures in screening for food allergy and would be the most appropriate for type I hypersensitivities, but their reliability is far from optimal. Recently we compared the results of these two tests with the result of double-blind oral challenge tests with various foods 26'27. Skin testing was done first by the scratch technique and if the reaction was equivocal, the test was repeated intradermally. R A S T was determined by a paper-disc radioimmunoassay. In case of milk, agreement with the challenge result (whether positive or negative) was 53 % for skin testing and 60 % for RAST. The test sensitivity tended to be higher for R A S T (56 % vs. 44 %), but the two tests had similar specificity and predictive accuracy (Table I). In the positive challenge group, skin testing and
Immunology Today, ool. 6, No. 5, 1985
R A S T were both positive in only 33% and both negative in 11%. In the negative challenge group, tests were both negative in only 50% and both positive in 16%. Various degrees of discrepancy were also noted when food IgE T A B L E I. Skin testing and R A S T compared with double-blind oral challenge in milk hypersensitivity Test characteristic Sensitivity (true positive test) Specificity (true negative test) Positive predictive accuracy Negative predictive accuracy
Skin testing
RAST
44 % 67 % 67 % 44 %
56 % 67 % 71% 50 %
antibodies were measured in the serum by different radioimmunoassay techniques 28. In summary, therefore, our understanding of the mechanisms of milk hypersensitivity leaves a lot to be desired. The most urgent need at present is the development of more reliable diagnostic tests. []-]
Acknowledgement I thank Miss Virginia H o w a r d and Miss AngelaJones for assistance in preparation of this manuscript. SAMI L. BAHNA Section of Allergy and Immunology, Department of Pediatrics, Louisiana State, University School of Medicine, New Orleans, LA 70112, USA.
References 1 Bahna, S. L. and Heiner, D. C. (1980) Allergies to Milk. Grune and Stratton, New York 2 Delmont, J. (ed.) (1983) Milk Intolerances and Rejection. Karger, Basel 3 Freed, D. L. J. (ed.) (1984) Health Hazards offVIilk. Bailliere Tindall, London. 4 Rothberg, R. M. and Farr, R. S. (1965) Pediatrics 35, 571-588 5 Kletter, B., Gery, I., Freier, S., and Davies, A. M. (1971) Int. Arch. Allergy Appl. Immunol 40, 656-666 6 May, C. D., Foman, S.J. and Remigio, L. (1982) Acla. Paediat. Scand 71, 43-51 7 Bahna, S. L. and Gandhi, M. D. (1983) Ann. Allergy 50, 218-224 8 Parish, W. E. (1971) Clin. Allergy 1, 369-380 9 Bj6rksten, B., Ahlstedt, S., Bj6rksten, F., Carlsson, B., Fallstrom, S. P., Juntunen, K., Kajosaari, M. and Kober, A. (1983) Allergy 38, 119-124 I0 Shakib, F. (1984)Int. Arch. Allergy Appl. Immunol. 75, 107-112 11 Firer, M. A., Hosking, C. S. and Hill, D. J. (1981) Br. Med. J. 283, 693-696 12 Gerrard, J. W. and Shenassa, M. (1983) Ann. Allergy 50, 375-379 13 Coombs, R: R. A. and MeLaughlin, P. (1982) Lancet i, 1388-1389 14 Osvath, P. and Markus, M. (1968) Acta Paediat. Acad Sci. Hung. 9, 279-284 15 Whitfield, M. F. and Barr, D. G. D. (1976)Arch. Dis. Child. 51,337 343 16 Jones, R. H. T. (1977)Arch. Dis. Child. 52, 744-745 17 Brostoff, J., Carini, C., Wraith, D. G. and Johns, P. (1979) Lancet i, 1268-1270 18 Coombs, R. R. A. and Oldham, G. (1981) Int. Arch. Allergy Appl. Immunol. 64, 287-292 19 Frick, O. L. (1982)J. Allergy Clin. Immunol. 69 (supplement), 157 20 Paganelli, R., Matricardi, P. M. and Aiuti, F. (1984) Clin. Rev. Allergy 2, 69 21 Shiner, M., Ballard, J. and Smith, M. E. (1975) Lancet i, 136-140 22 Endr6, L. and Osvath, P. (1975)Acta Allergol. 30, 34-42 23 Scheinmann, P., Gendrel, S., Charlas, J. and Paupe, J. (1976) Clin. Allergy 6, 515-521 24 Gjesing, B. and L~wenstein, H. (1984) Ann. Allergy 53, 602-608 25 Spies, J. R., Stevan, M. A., Stein, W. J. and Coulson, E. J. (1970) j. Allergy 45, 208-219 26 Gandhi, M. D. and Bahna, S. L. (1985)J. Allergy Clin. Immunol. 75,205 27 Bahna, S. L. and Gandhi, M. D. (1985)J. Allergy Clin. ImmunoL 75,204 28 Firestone, J. E., Jr. and Bahna, S. L. (1985)J. Allergy Clin. Immunol. 75, 2O4
