- Email: [email protected]
Immunologic changes associated with the development of tolerance in children with cow milk allergy
Immunologic changes associated with the development of tolerance in children with cow milk allergy John M. James, MD, a n d Hugh A. Sampson, MD From the Divisionof Pediatric Allergy/Immunology, Johns HopkinsUniversitySchool of Medicine, Baltimore, Maryland
The purpose of this study was to determine whether cow milk-specific antibody responses correlated with the development of clinical tolerance in cow milkallergic children. Double-blind, placebo-controlled food challenges were performed annually in 29 patients with cow milk allergy. Clinical reactivity was lost in 11 (38%) of 29 patients. The median a g e for all patients at the time of diagnosis by these food challenges was 3 years; more than 80% of patients in each group had atopic dermatitis as part of their presenting symptoms. Casein-specific and ~-lactoglobulin-specific IgE, IgG, IgG1, and IgG4 antibody concentrations were analyzed in all patients at regular intervals. In the patients becoming clinically tolerant to cow milk, the IgE-specific antibody concentrations and IgE/IgG-specific ratios for both milk proteins were lower initially and decreased significantly with time, in comparison with those in the group who retained clinical sensitivity. The concentrations of IgG1, and IgG4-specific antibody to casein and the IgE/IgG1 and IgE/IgG4 ratios for both casein and ~-lactoglobulin were significantly less in the patients losing clinical reactivity. No differences in the IgG-specific concentrations were observed in either group at any of the evaluation times noted above. Monitoring similar casein-specific and ~-lactoglobulin-specific IgE concentrations and IgE/IgG ratios may help predict which patients will ultimately lose their clinical reactivity to cow milk. (J PEDIATR1992;121: 371-7)
Cow milk, one of the first foreign proteins encountered during infancy, has been implicated in gastrointestinal, cutaneous, and respiratory hypersensitivity reactions. 1 Reactions to cow milk also are one of the most common food allergies in children, occurring in 2.0% to 2.5% of children 3 years of age and younger. 24 Both genetic and environmental factors have been implicated in the pathogenesis of this
disorder. 58 Because cow milk is so ubiquitous in the Western food supply, the milk-allergic individual faces an overwhelming task in identifying and restricting these proteins from the diet. 9 DBPCFC ELISA
Supported in part by a grant from the Nutrasweet Co. and the International Life Sciences Institute and in part by grants from the National Institute of Allergy and Infectious Diseases (No. AI24439) and from the General Clinical Research Center Program, Division of Research Resources, National Institutes of Health (No. RR00052). Submitted for publication Jan. 29, 1992; accepted April 13, 1992. Reprint requests: Hugh A. Sampson, MD, Johns Hopkins Hospital, CMSC-1103, 600 N. Wolfe St., Baltimore, MD 21205.
9/20/38681
I I
Double-blind, placebo-controlled food challenges Enzyme-linked immunosorbent assay
I
Approximately half of the hypersensitivity reactions to cow milk in a group of infants less than 1 year of age have been reported to be IgE mediated. 2 At least 20 protein components, any of which may stimulate antibody production, have been identified in cow milk. l~ Specific IgE antibodies to casein and whey, which constitute 76% to 86% and 14% to 24% of milk proteins, respectively, account for the majority of IgE-mediated reactions to cow milk. Cellular
371
372
James and Sampson
immune responses to milk proteins and complement activation resulting from milk antigen-specific antibody complexes also have been implicated in the pathogenesis of cow milk hypersensitivity, but convincing evidence is lacking.5, z1 Although 25% of infants with IgE-mediated cow milk allergy retain their clinical reactivity,2 the majority of infants "outgrow" or "lose" their milk allergy by their third birthday. Recent studies have documented the remission of cow milk allergy in one third of older milk-allergic children 1 to 2 years after the institution of a strict milk-elimination diet.12, 13 Despite the available information, the specific immunologic mechanisms responsible for the development of this clinical tolerance remain to be elucidated. This report summarizes the natural history of 29 patients placed on an elimination diet after a diagnosis of cow milk allergy. These patients have undergone repeated doubleblind, placebo-controlled food challenges on an annual basis to reassess their clinical reactivity. Milk-specifiC antibody concentrations in the patients in whom clinical tolerance developed were compared with those in children remaining allergic to milk. METHODS During the past 10 years, cow milk hypersensitivity has been evaluated in 181 patients. Seven patients had a convincing history of anaphylaxis (laryngeal edema, severe wheezing, hypotension) and were not subjected to challenges. The remaining 174 patients underwent DBPCFC to cow milk in our clinical research unit. Results of initial challenges were positive in 80 patients; 94 had negative results. Of the 80 milk-allergic patients, 29 have undergone subsequent annual rechallenges. Sera were collected and stored at - 2 0 ~ C each challenge day. As previously described] 4 all patients were required to discontinue antihistamine therapy for 7 to 14 days before admission; none had received systemically administered corticosteroids for at least 2 months before the evaluation. Glycerinated food extracts (1:10 or 1:20 wt/vol, Greer Laboratories, Inc., Lenoir, N.C.) were applied to the patient's back or to the Volar surface of the patient's forearm by a clinical research nurse. The mean diameter of the wheal-and-flare response was recorded as previously described.15, 16 Skin test results were considered positive when a food antigen elicited a wheal of at least 3 mm greater than the control prick. The DBPCFC was done by admi.~tering up to 8 to 10 gm of dehydrated, powdered milk disguised in juice or in capsules (500 rag) during a 60- to 90-minute period. 14, 17 The initial test dose was usually 500 mg but was reduced to 250 mg if a potentially severe reaction was suspected. Two
The Journal of Pediatrics September 1992
challenges, one of which consisted of a placebo (sucrose or corn starch), were performed each day, at 8 AM and at 12 noon. Children were observed for at least 24 hours after the challenge, although most reactions occurred within 90 minutes and no reaction occurred later than 4 hours after the challenge. Positive reactions were established when objective cutaneous, nasal, pulmonary, or gastrointestinal tract symptoms occurred within 2 hours of the DBPCFC. Cutaneous symptoms consisted of a pruritic, erythematous morbilliform rash primarily confined to atopic dermatitis predilection sites or, rarely, urticaria. Gastrointestinal tract symptoms included abdominal pain with or without nausea, and vomiting with or without diarrhea. Respiratory tract symptoms were characterized as nasal, laryngeal, or pulmonary and included nasal congestion, rhinorrhea, nasal pruritus and sneezing, throat itching, cough and hoarseness, dyspnea, and wheezing. All negative DBPCFC results were confirmed by open challenge. Equivocal reactions or those Which occurred in individuals with negative skin test results were repeated at least once. After a diagnosis of cow milk allergy was confirmed, all study patients were placed on a strict milk-elimination diet until clinical reactivity was demonstrated by negative DBPCFC results. ELISA to determine serum-specific lgE and IgG concentrations to cow milk proteins. Sensitive ELISAs were used to determine Specific concentrations of IgE and IgG antibody to both Casein and B-lactoglobulin in all the patients. As previously described, 18 standard curves were generated in the upper two rows of 96-well mierotiter plates (Dynatech LaboratorieS, Inc., Chantilly, Va.) with either monoc|onal antihuman IgG antibody (ICN Biomedicals, Inc., Costa Mesa, Calif.) or antihuman IgE antibody (kindly provided by Andrew Saxon, University of California at Lo~ Angeles). Lower wells were coated with 500 ng of either casein or ~-lactoglobulin (Sigma Chemical Co., St. Louis, Mo.). Standard curves for IgG and IgE were calibrated against secondary World Health Organization reference standards (ranges: IgG, 0.78 to 200 ng/ml; IgE, 0.078 to 20.0 ng/ml). Patient sera were incubated in lower wells simultaneously with standard curves in upper wells. For IgG determinations, afffinity-purified, alkaline phosphatase-conjugated goat antihuman IgG (Jackson Immunoresearch Labs, Inc., West Grove, Pa.) was added, and for IgE quantitation, biotinylated, affinity-purified goat antihuman IgE (Kirkegard and Perry Laboratories, Inc., Gaithersburg, Md.) was used, followed by an incubation with an avidin-alkaline phosphatase conjugate. Plates were developed with p-nitrophenylphosphate; absorbance was read at 405 to 490 nm with an automated microtiter plate reader (Molecular Devices Corp., Menlo Park, Calif.). After standard curves were generated, patient antibody concentrations specific for
Volume 121 Number 3
Development o f tolerance to cow milk
373
T a b l e . Clinical presentations and demographics of the two patient groups
Gender Median age (yr)* Median age (too) at positive DBPCFC result* Prick skin test reaction to cow milk: median wheal size (mm) Initial Final Mean time to final DBPCFC (yr) Atopic disorders? Atopic dermatitis Allergic rhinitis Asthma
Development of tolerance
Persistent sensitivity
6M, 5 F 7 (3-13) 36 (1-132)
12M, 6 F 6 (4-14) 36 (15-93)
6.5 6.0 3.0
7.0 9.0 3.2
9 (82) 7 (64) 5 (45)
16 (89) 9 (50) 10 (56)
*Ranges are notedin parentheses. tPercentages are noted in parentheses. casein and/3-1actoglobulin were determined in nanogramequivalent units. The intraassay coefficient of variation was less than 10%. ELISA to determine serum IgG1 and IgG4 antibody concentrations specific to casein and/%lactoglobulin. Genetically engineered human-mouse chimeric IgG1 and IgG4 antibodies specific for 4-hydroxy-3-nitrophenylacetyl and nitrophenyl-o-caproic acid-succinimide, coupled to human serum albumin (NP-HSA) (kindly provided by Dr. Robert Hamilton), were prepared and used to generate standard curves as described previously.19-22 Mouse biotinylated monoclonal antibodies specific for human IgG 1 Fc (Hybridoma Reagent Lab, Baltimore, Md.) and human IgG4 Fc (ICN Biomedicals) were used in subclass assays. Respective working ranges for subclass assays were 2 to 1000 ng/ml for IgG1, and 0.36 to 182 ng/ml for IgG4. Standard curves were generated in the upper two rows of 96-well microtiter plates with NP-NSA; lower wells were coated with either casein or/3-1actoglobulin. The remaining steps were similar to the ELISA described above, except that biotinylated monoclonal antibody conjugates, either anti-IgG1 or antiIgG4, were added to appropriate plates. Nonparametric tests were used to analyze specific antibody concentrations, median skin test wheals, and milk protein-specific IgE/IgG, IgE/IgG 1, and IgE/IgG4 ratios. The Wilcoxon signed rank was utilized for intragroup comparisons and the Wilcoxon rank sum for intergroup comparisons. RESULTS Of 80 children documented to have allergy to cow milk, 29 have returned for at least three annual milk challenges. Six boys and five girls (aged 3 to 13 years; median age, 7 years) lost their clinical reactivity to cow milk. The median time, until clinical tolerance was documented by DBPCFC,
was 3 years. In contrast, 12 boys and 6 girls (aged 4 to 12 years; median age, 6.0 years) retained their reactivity to cow milk; the average number of additional annual challenges was 3.2. The clinical presentations and demographics of the two patient groups are depicted in the Table. No significant differences in clinical presentation or age at diagnosis were evident between patient groups achieving clinical tolerance and those retaining milk allergy. Results of prick skin tests were similar in both groups. The patients who ultimately became clinically tolerant had no significant change in their mean wheal diameter from the initial evaluation to the time when tolerance was documented (median, 6.5 to 6.0 ram). Likewise, in those children remaining reactive to milk, there was no significant difference in the mean wheal diameter at the time of the initial admission and the most recent one (median, 7.0 to 9.0 mm). In addition, there were no significant differences between the two groups when the median skin test wheal diameters were compared initially and at the final challenge. During the initial DBPCFC, 11 patients losing reactivity had symptoms that included the following: cutaneous, 10 patients; respiratory, 6 patients; and gastrointestinal, 2 patients. All these symptoms were absent during their final challenges. In the group of 18 patients retaining reactivity, initial challenge symptoms included the following: cutaneous, 17 patients; respiratory, 11; and gastrointestinal, 5; and (in their final challenges) cutaneous, 10; respiratory, 14; and gastrointestinal, 8. There was a decrease in cutaneous symptoms in this group from the initial to the final challenge, but increases in respiratory and gastrointestinal symptoms. No overwhelming pattern of initial target organ response correlated with loss of cow milk hypersensitivity in this select group of patients. Initial milk-specific IgE antibody concentrations were
374
James and Sampson
The Journal of Pediatrics September 1992
10000~1 1000
P < .01
P < .01
A
E
100
v
W
0'I
10
9
.11
~
~r
CAS
Fig.4.
BLAC
CAS
BLAC
Initial and final casein-specific and ~-lactoglobulin-specific IgE concentrations in patients ultimately achieving clinical tolerance to cow milk (dark circles) and in those remaining allergic (clear circles). CAS, Casein; BLAC,/3-1actoglobulin. Medians are depicted by asterisks.
100000
1
P<.01
10000 i
O X
1
1000I
P < .01
P < .05
.2 (.0
1001
~
*
*
*
.1 ~
Fig.2.
CAS
BLAC
CAS
BLAC
Initial and final casein-specific and 13-1actoglobulin-specific IgE/IgG ratios in patients ultimately achieving clinical tolerance to cow milk (dark circles) and in those remaining allergic (clear circles). CAS, Casein; BLAC, 13-1actoglobulin. Medians are depicted by asterisks.
Volume 121 Number 3
Development o f tolerance to cow m i l k
375
10 7
10 6 ID
'0
10 5
0
10 4
)<
lk
10 3 O~
O)
10 2
101
10 0 CAS
BLAC
, CAS
BLAC
Fig. 3. Initial and final casein-specificand/3-1actoglobulin-specificIgE/IgG4 ratios in patients ultimately achieving clinical tolerance to cow milk (dark circles) and in those remaining allergic (clear circles). CAS, casein; BLAC,/3-1actoglobulin. Medians are depicted by asterisks.
not significantly lower in patients ultimately becoming milk tolerant, but final casein-specific and ~-lactoglobulin-specific IgE concentrations were significantly lower (p <0.01) in patients achieving clinical tolerance than in patients remaining clinically reactive (Fig. 1). Furthermore, in patients becoming tolerant, casein-specific and/3-1actoglobulin-specific IgE concentrations decreased significantly (,/9 <0.01) from the initial to the final evaluation (Fig. 1). No differences in the IgG concentrations specific to the milk proteins were observed in either patient group at any time or between evaluations. There were slight but significant decreases from the initial to the final (negative DBPCFC results) IgG1 and IgG4 casein-specific antibody concentrations in the group of patients achieving clinical tolerance (0.01 < p < 0.05). All other comparisons of these two IgG subclass antibodies, both within and between groups, revealed no significant differences in antibody concentrations. The patients ultimately achieving clinical tolerance had initial IgE/IgG ratios that were significantly less for casein (p <0.01) and/3-1actoglobulin (0.01 < p < 0.05); at final evaluation, both milk protein-specific I g E / I g G ratios were significantly less (p <0.01) than the same ratios in patients remaining allergic (Fig. 2). Finally, the IgE/IgG ratios for
both milk proteins also decreased significantly (p <0.01) from the initial to the final evaluation in the clinically tolerant patients. The patients ultimately losing clinical reactivity to cow milk initially had signifcantly lower ~-lactoglobulin-specific IgE/IgG1 ratios (0.01 < p < 0.05) and IgE/IgG4 ratios for both casein (0.01 < p < 0.05) and/3-1actoglobulin (p <0.01) (Fig. 3). Final IgE/IgG1 and IgE/IgG4 ratios for casein and /3-1actoglobulin were both significantly less (p <0.01) in patients achieving clinical tolerance. DISCUSSION In children with IgE-mediated cow milk hypersensitivity, milk protein-specific IgE concentrations were significantly lower in patients achieving tolerance than in the group retaining milk allergy. These findings are in contrast to reports of other investigators. Hill et al. 23 studied 47 pediatric patients with cow milk allergy. Patients were classified into one of three groups on the basis of the time of onset of symptoms after initiation of open milk challenges. Fifteen patients reacted within 45 minutes (immediate reactors), 24 patients within 1 to 20 hours (intermediate reactors), and 8 patients beyond 20 hours (late reactors). Cow milk-specific IgE antibodies were not found consistently in any
376
James and Sampson
group. Tolerance to cow milk developed in 6 immediate reactors, 10 intermediate reactors, and 2 late reactors. No consistent changes in serum concentrations of cow milkspecific IgG, IgA, IgM, or IgE antibodies were associated with loss of clinical reactivity. Other investigators have shown that shrimp-specific IgE and IgG antibodies are significantly elevated in patients with positive DBPCFC reac, tions to shrimp, but no significant long-term changes occur in shrimp-specific IgE, IgG, IgA, or IgM antibody concentrations. 24 Furthermore, shrimp-allergic patients were shown to have significantly elevated shrimp-specific IgG2 and IgG4 concentrations in comparison with normal control subjects, although none of these changes were diagnostic of clinical shrimp hypersensitivity.25 In contrast to the Hill study, all our patients were well characterized as having IgE-mediated cow milk allergy confirmed by DBPCFC. Utilizing this more homogeneous population of milk-allergic children, we were able to detect subtle immunologic changes associated with the development of oral tolerance in cow milk-allergic children. The studies of Daul et al. 24, 2s involved adults reactive to an allergen (shrimp) to which patients do not generally lose their clinical reactivity. Whether the natural history of cow milk allergy and changes in antibody-specific responses may be extrapolated to other IgE-mediated food allergies remains to be determined. In a recent study of cow milk-allergic children,26 the gut immune response to cow milk antigens was indirectly assessed by means of a solid-phase, enzyme-linked immunoassay (ELISPOT). As in our study, cow milk-specific IgG antibodies were found in all children, and no significant changes were seen--whether or not children achieved clinical tolerance. Cow milk-allergic children retaining clinical sensitivity had a significant increase in circulating immunoglobulin-secreting cells after milk challenges, whereas those achieving clinical tolerance had no increase in cell numbers. The investigators concluded that "outgrowing" cow milk allergy is associated with maturation of the gutassociated immune system and loss of responsiveness to ingested food antigens. This maturational process also may be associated with the decrease in cow milk-specific IgE antibodies seen in our patients. We conclude that significant decreases in milk-specific IgE antibody concentrations are evident by the time clinical tolerance is documented in children with IgE-mediated cow milk allergy. Individual IgE milk-specific antibody concentrations at either the time o ~ e negative DBPCFC results or the final DBPCFC results are significantly lower in children losing clinical reactivity than in those remaining allergic. The milk-specific IgE/IgG ratios reflect similar differences in the two patient groups, but do not appear to
The Journal of Pediatrics September 1992
discriminate any further the differences in individual patients achieving clinical tolerance from those retaining clinical reactivity. Monitoring IgG-, IgGl-, and IgG4-specific antibodies to casein and/3-1actoglobulin is not helpful, but monitoring serum-specific IgE concentrations and calculating a ratio of IgE to IgG antibody to common milk protein allergens may be useful in predicting which patients will ultimately lose their clinical reactivity to cow milk, and may also help determine when patients should undergo additional challenges. REFERENCES
l. Savilahti E. Cow's milk allergy. Allergy 1981;36:73-88. 2. Host A, Halken S. A prospective study of cow milk allergy in Danish infants during the first 3 years of life. Allergy 1990; 45:587-96. 3. Jakobsson I, Lindeberg TA. A prospectivestudy of cow's milk protein intolerance in Swedish infants. Acta Pediatr Scand 1979;68:853-9. 4. Hide DW, Guyer BM. Cows milk intolerance in Isle of Wight infants. Br J Clin Pract 1983;37:285-7. 5. Hill DH, Firer MA, Shelton MJ, Hosking CS. Manifestations of milk allergy in infancy: clinical and immunologicfindings. J PEDIATR1986;109:270-6. 6. Cow's milk allergy in the first year of life. Acta Paediatr Scand Suppl 1988;348:2-13. 7. Vandenplas Y, Sacre L. Influences of neonatal serum IgE concentration,family history and diet on the incidenceof cow's milk allergy. Eur J Pediatr 1986;145:493-5. 8. Ford RPK, Fergusson DM. Egg and cow's milk allergy in children. Arch Dis Child 1980;55:608-10. 9. Yunginger JW. Food antigens. In: Metcalfe DO, Sampson HA, Simon RA, eds. Food allergy: adverse reactions to foods and food additives. Boston: Blackwell, 1991:38-9. 10. BleurninkE, Young E. Identification of the atopic allergen in cow's milk. Int Arch Allergy Appl Immunol 1968;34:521-43. 11. Sampson HA. Immunologicmechanisrns in adverse reactions to foods. Immunol Allergy Clin North Am 1991;11:701-16. 12. Sampson HA, Scanlon SM. Natural history of food hypersensitivity in children with atopic dermatitis. J PED1ATR1989; 115:23-7. 13. WilsonNW~.I~lamburgerRN. Allergy to cow's milk in the first year of life and its prevention. Ann Allergy 1988;61:323-7. 14. Sampson HA. Role of immediate food hypersensitivityin the pathogenesis of atopic dermatitis. J Allergy Clin Immunol 1983;71:473-80. 15. Voorhorst R, van Krieken H. Atopic skin test reevaluated. II. Variability in results of skin testing done in octuplicate. Ann Allergy 1973;31:195-204. 16. Sampson HA. Comparative study of commercial food antigen extracts for the diagnosis of food hypersensitivity. J Allergy Clin Immnol 1988;82:718-26. 17. Leinhas .IL, McCaskill C, Sampson HA. Food allergy challenges: guidelinesand implications.J Am Diet Assoc 1987;87: 604-8. 18. Fasano MB Wood RA, Cooke SK, Sampson HA. Egg hypersensitivity and adverse reactions to measles, mumps, and rubella vaccine. J PEDIATR1992;120:878-81.
Volume 121 Number 3
Development o f tolerance to cow milk
19: Neuberger MS, Williams GT, Mitchell EB, Joubal SS. Flanogan JG, Rabbitts TH. A hapten-specific chimeric IgE antibody with human physiologic effeetor function. Nature 1985;314:268-70. 20. Bruggemann M; Williams GT, Bindon CL, et al. Comparison of the effector functions of human immunoglobulins using a matched set of chimeric antibodies. J Exp Med 1987;166:135161. 21. Hamilton RG. Engineered human antibodies as immunologic quality control reagents. Ann Biol Clin (Paris) 1990;48:473-7. 22. Hamilton RG. Application of engineered chimeric antibodies to the calibration of human antibody standards. Ann Biol Clin (Paris) 1991;49:242-8. 23. Hill D J, Firer MA, Ball G, Hosking CS. Recovery from milk
377
allergy in early childhood: antibody studies. J PEDIATR 1989; 114:761-6. 24. Daul CB, Morgan JE. Lehrer SB. The natural history of shrimp-specific immunity. J Allergy Clin Immunol 1990;86:8893. 25. Morgan JE, Daul CB, Lehrer SB. The relations among shrimp-specific IgG subclass antibodies and immediate adverse reactions to shrimp challenge. J Allergy Clin Immunol 1990;86:387-92. 26. Isolauri E, Suomalainen H, Kaila M, et al. Local immune response in patients with cow milk allergy: follow-up of patients retaining allergy or becoming tolerant. J PEDIATRi992; 120:915.
BOUND VOLUMES AVAILABLE TO SUBSCRIBERS Bound volumes of the 1992 issues of THE JOURNAL OF PEDIATRICS are available to subscribers (only) from the Publisher, at a cost of $58.00 for domestic, $78.06 for Canadian, and $74.00 for international subscribers, for Vol. 120 (January-June) and Vol. 12l (July-December), shipping charges included. Each bound volume contains subject and author indexes, and all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durable buckram, with the Journal name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact M o s b y - Y e a r Book, Inc., Subscription Services, 11830 Westline Industrial Dr., St. Louis, M O 63146-3318, USA/800-325-4177, ext. 4351, or 314-453-4351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular Journal subscriptio n .
The purpose of this study was to determine whether cow milk-specific antibody responses correlated with the development of clinical tolerance in cow milkallergic children. Double-blind, placebo-controlled food challenges were performed annually in 29 patients with cow milk allergy. Clinical reactivity was lost in 11 (38%) of 29 patients. The median a g e for all patients at the time of diagnosis by these food challenges was 3 years; more than 80% of patients in each group had atopic dermatitis as part of their presenting symptoms. Casein-specific and ~-lactoglobulin-specific IgE, IgG, IgG1, and IgG4 antibody concentrations were analyzed in all patients at regular intervals. In the patients becoming clinically tolerant to cow milk, the IgE-specific antibody concentrations and IgE/IgG-specific ratios for both milk proteins were lower initially and decreased significantly with time, in comparison with those in the group who retained clinical sensitivity. The concentrations of IgG1, and IgG4-specific antibody to casein and the IgE/IgG1 and IgE/IgG4 ratios for both casein and ~-lactoglobulin were significantly less in the patients losing clinical reactivity. No differences in the IgG-specific concentrations were observed in either group at any of the evaluation times noted above. Monitoring similar casein-specific and ~-lactoglobulin-specific IgE concentrations and IgE/IgG ratios may help predict which patients will ultimately lose their clinical reactivity to cow milk. (J PEDIATR1992;121: 371-7)
Cow milk, one of the first foreign proteins encountered during infancy, has been implicated in gastrointestinal, cutaneous, and respiratory hypersensitivity reactions. 1 Reactions to cow milk also are one of the most common food allergies in children, occurring in 2.0% to 2.5% of children 3 years of age and younger. 24 Both genetic and environmental factors have been implicated in the pathogenesis of this
disorder. 58 Because cow milk is so ubiquitous in the Western food supply, the milk-allergic individual faces an overwhelming task in identifying and restricting these proteins from the diet. 9 DBPCFC ELISA
Supported in part by a grant from the Nutrasweet Co. and the International Life Sciences Institute and in part by grants from the National Institute of Allergy and Infectious Diseases (No. AI24439) and from the General Clinical Research Center Program, Division of Research Resources, National Institutes of Health (No. RR00052). Submitted for publication Jan. 29, 1992; accepted April 13, 1992. Reprint requests: Hugh A. Sampson, MD, Johns Hopkins Hospital, CMSC-1103, 600 N. Wolfe St., Baltimore, MD 21205.
9/20/38681
I I
Double-blind, placebo-controlled food challenges Enzyme-linked immunosorbent assay
I
Approximately half of the hypersensitivity reactions to cow milk in a group of infants less than 1 year of age have been reported to be IgE mediated. 2 At least 20 protein components, any of which may stimulate antibody production, have been identified in cow milk. l~ Specific IgE antibodies to casein and whey, which constitute 76% to 86% and 14% to 24% of milk proteins, respectively, account for the majority of IgE-mediated reactions to cow milk. Cellular
371
372
James and Sampson
immune responses to milk proteins and complement activation resulting from milk antigen-specific antibody complexes also have been implicated in the pathogenesis of cow milk hypersensitivity, but convincing evidence is lacking.5, z1 Although 25% of infants with IgE-mediated cow milk allergy retain their clinical reactivity,2 the majority of infants "outgrow" or "lose" their milk allergy by their third birthday. Recent studies have documented the remission of cow milk allergy in one third of older milk-allergic children 1 to 2 years after the institution of a strict milk-elimination diet.12, 13 Despite the available information, the specific immunologic mechanisms responsible for the development of this clinical tolerance remain to be elucidated. This report summarizes the natural history of 29 patients placed on an elimination diet after a diagnosis of cow milk allergy. These patients have undergone repeated doubleblind, placebo-controlled food challenges on an annual basis to reassess their clinical reactivity. Milk-specifiC antibody concentrations in the patients in whom clinical tolerance developed were compared with those in children remaining allergic to milk. METHODS During the past 10 years, cow milk hypersensitivity has been evaluated in 181 patients. Seven patients had a convincing history of anaphylaxis (laryngeal edema, severe wheezing, hypotension) and were not subjected to challenges. The remaining 174 patients underwent DBPCFC to cow milk in our clinical research unit. Results of initial challenges were positive in 80 patients; 94 had negative results. Of the 80 milk-allergic patients, 29 have undergone subsequent annual rechallenges. Sera were collected and stored at - 2 0 ~ C each challenge day. As previously described] 4 all patients were required to discontinue antihistamine therapy for 7 to 14 days before admission; none had received systemically administered corticosteroids for at least 2 months before the evaluation. Glycerinated food extracts (1:10 or 1:20 wt/vol, Greer Laboratories, Inc., Lenoir, N.C.) were applied to the patient's back or to the Volar surface of the patient's forearm by a clinical research nurse. The mean diameter of the wheal-and-flare response was recorded as previously described.15, 16 Skin test results were considered positive when a food antigen elicited a wheal of at least 3 mm greater than the control prick. The DBPCFC was done by admi.~tering up to 8 to 10 gm of dehydrated, powdered milk disguised in juice or in capsules (500 rag) during a 60- to 90-minute period. 14, 17 The initial test dose was usually 500 mg but was reduced to 250 mg if a potentially severe reaction was suspected. Two
The Journal of Pediatrics September 1992
challenges, one of which consisted of a placebo (sucrose or corn starch), were performed each day, at 8 AM and at 12 noon. Children were observed for at least 24 hours after the challenge, although most reactions occurred within 90 minutes and no reaction occurred later than 4 hours after the challenge. Positive reactions were established when objective cutaneous, nasal, pulmonary, or gastrointestinal tract symptoms occurred within 2 hours of the DBPCFC. Cutaneous symptoms consisted of a pruritic, erythematous morbilliform rash primarily confined to atopic dermatitis predilection sites or, rarely, urticaria. Gastrointestinal tract symptoms included abdominal pain with or without nausea, and vomiting with or without diarrhea. Respiratory tract symptoms were characterized as nasal, laryngeal, or pulmonary and included nasal congestion, rhinorrhea, nasal pruritus and sneezing, throat itching, cough and hoarseness, dyspnea, and wheezing. All negative DBPCFC results were confirmed by open challenge. Equivocal reactions or those Which occurred in individuals with negative skin test results were repeated at least once. After a diagnosis of cow milk allergy was confirmed, all study patients were placed on a strict milk-elimination diet until clinical reactivity was demonstrated by negative DBPCFC results. ELISA to determine serum-specific lgE and IgG concentrations to cow milk proteins. Sensitive ELISAs were used to determine Specific concentrations of IgE and IgG antibody to both Casein and B-lactoglobulin in all the patients. As previously described, 18 standard curves were generated in the upper two rows of 96-well mierotiter plates (Dynatech LaboratorieS, Inc., Chantilly, Va.) with either monoc|onal antihuman IgG antibody (ICN Biomedicals, Inc., Costa Mesa, Calif.) or antihuman IgE antibody (kindly provided by Andrew Saxon, University of California at Lo~ Angeles). Lower wells were coated with 500 ng of either casein or ~-lactoglobulin (Sigma Chemical Co., St. Louis, Mo.). Standard curves for IgG and IgE were calibrated against secondary World Health Organization reference standards (ranges: IgG, 0.78 to 200 ng/ml; IgE, 0.078 to 20.0 ng/ml). Patient sera were incubated in lower wells simultaneously with standard curves in upper wells. For IgG determinations, afffinity-purified, alkaline phosphatase-conjugated goat antihuman IgG (Jackson Immunoresearch Labs, Inc., West Grove, Pa.) was added, and for IgE quantitation, biotinylated, affinity-purified goat antihuman IgE (Kirkegard and Perry Laboratories, Inc., Gaithersburg, Md.) was used, followed by an incubation with an avidin-alkaline phosphatase conjugate. Plates were developed with p-nitrophenylphosphate; absorbance was read at 405 to 490 nm with an automated microtiter plate reader (Molecular Devices Corp., Menlo Park, Calif.). After standard curves were generated, patient antibody concentrations specific for
Volume 121 Number 3
Development o f tolerance to cow milk
373
T a b l e . Clinical presentations and demographics of the two patient groups
Gender Median age (yr)* Median age (too) at positive DBPCFC result* Prick skin test reaction to cow milk: median wheal size (mm) Initial Final Mean time to final DBPCFC (yr) Atopic disorders? Atopic dermatitis Allergic rhinitis Asthma
Development of tolerance
Persistent sensitivity
6M, 5 F 7 (3-13) 36 (1-132)
12M, 6 F 6 (4-14) 36 (15-93)
6.5 6.0 3.0
7.0 9.0 3.2
9 (82) 7 (64) 5 (45)
16 (89) 9 (50) 10 (56)
*Ranges are notedin parentheses. tPercentages are noted in parentheses. casein and/3-1actoglobulin were determined in nanogramequivalent units. The intraassay coefficient of variation was less than 10%. ELISA to determine serum IgG1 and IgG4 antibody concentrations specific to casein and/%lactoglobulin. Genetically engineered human-mouse chimeric IgG1 and IgG4 antibodies specific for 4-hydroxy-3-nitrophenylacetyl and nitrophenyl-o-caproic acid-succinimide, coupled to human serum albumin (NP-HSA) (kindly provided by Dr. Robert Hamilton), were prepared and used to generate standard curves as described previously.19-22 Mouse biotinylated monoclonal antibodies specific for human IgG 1 Fc (Hybridoma Reagent Lab, Baltimore, Md.) and human IgG4 Fc (ICN Biomedicals) were used in subclass assays. Respective working ranges for subclass assays were 2 to 1000 ng/ml for IgG1, and 0.36 to 182 ng/ml for IgG4. Standard curves were generated in the upper two rows of 96-well microtiter plates with NP-NSA; lower wells were coated with either casein or/3-1actoglobulin. The remaining steps were similar to the ELISA described above, except that biotinylated monoclonal antibody conjugates, either anti-IgG1 or antiIgG4, were added to appropriate plates. Nonparametric tests were used to analyze specific antibody concentrations, median skin test wheals, and milk protein-specific IgE/IgG, IgE/IgG 1, and IgE/IgG4 ratios. The Wilcoxon signed rank was utilized for intragroup comparisons and the Wilcoxon rank sum for intergroup comparisons. RESULTS Of 80 children documented to have allergy to cow milk, 29 have returned for at least three annual milk challenges. Six boys and five girls (aged 3 to 13 years; median age, 7 years) lost their clinical reactivity to cow milk. The median time, until clinical tolerance was documented by DBPCFC,
was 3 years. In contrast, 12 boys and 6 girls (aged 4 to 12 years; median age, 6.0 years) retained their reactivity to cow milk; the average number of additional annual challenges was 3.2. The clinical presentations and demographics of the two patient groups are depicted in the Table. No significant differences in clinical presentation or age at diagnosis were evident between patient groups achieving clinical tolerance and those retaining milk allergy. Results of prick skin tests were similar in both groups. The patients who ultimately became clinically tolerant had no significant change in their mean wheal diameter from the initial evaluation to the time when tolerance was documented (median, 6.5 to 6.0 ram). Likewise, in those children remaining reactive to milk, there was no significant difference in the mean wheal diameter at the time of the initial admission and the most recent one (median, 7.0 to 9.0 mm). In addition, there were no significant differences between the two groups when the median skin test wheal diameters were compared initially and at the final challenge. During the initial DBPCFC, 11 patients losing reactivity had symptoms that included the following: cutaneous, 10 patients; respiratory, 6 patients; and gastrointestinal, 2 patients. All these symptoms were absent during their final challenges. In the group of 18 patients retaining reactivity, initial challenge symptoms included the following: cutaneous, 17 patients; respiratory, 11; and gastrointestinal, 5; and (in their final challenges) cutaneous, 10; respiratory, 14; and gastrointestinal, 8. There was a decrease in cutaneous symptoms in this group from the initial to the final challenge, but increases in respiratory and gastrointestinal symptoms. No overwhelming pattern of initial target organ response correlated with loss of cow milk hypersensitivity in this select group of patients. Initial milk-specific IgE antibody concentrations were
374
James and Sampson
The Journal of Pediatrics September 1992
10000~1 1000
P < .01
P < .01
A
E
100
v
W
0'I
10
9
.11
~
~r
CAS
Fig.4.
BLAC
CAS
BLAC
Initial and final casein-specific and ~-lactoglobulin-specific IgE concentrations in patients ultimately achieving clinical tolerance to cow milk (dark circles) and in those remaining allergic (clear circles). CAS, Casein; BLAC,/3-1actoglobulin. Medians are depicted by asterisks.
100000
1
P<.01
10000 i
O X
1
1000I
P < .01
P < .05
.2 (.0
1001
~
*
*
*
.1 ~
Fig.2.
CAS
BLAC
CAS
BLAC
Initial and final casein-specific and 13-1actoglobulin-specific IgE/IgG ratios in patients ultimately achieving clinical tolerance to cow milk (dark circles) and in those remaining allergic (clear circles). CAS, Casein; BLAC, 13-1actoglobulin. Medians are depicted by asterisks.
Volume 121 Number 3
Development o f tolerance to cow m i l k
375
10 7
10 6 ID
'0
10 5
0
10 4
)<
lk
10 3 O~
O)
10 2
101
10 0 CAS
BLAC
, CAS
BLAC
Fig. 3. Initial and final casein-specificand/3-1actoglobulin-specificIgE/IgG4 ratios in patients ultimately achieving clinical tolerance to cow milk (dark circles) and in those remaining allergic (clear circles). CAS, casein; BLAC,/3-1actoglobulin. Medians are depicted by asterisks.
not significantly lower in patients ultimately becoming milk tolerant, but final casein-specific and ~-lactoglobulin-specific IgE concentrations were significantly lower (p <0.01) in patients achieving clinical tolerance than in patients remaining clinically reactive (Fig. 1). Furthermore, in patients becoming tolerant, casein-specific and/3-1actoglobulin-specific IgE concentrations decreased significantly (,/9 <0.01) from the initial to the final evaluation (Fig. 1). No differences in the IgG concentrations specific to the milk proteins were observed in either patient group at any time or between evaluations. There were slight but significant decreases from the initial to the final (negative DBPCFC results) IgG1 and IgG4 casein-specific antibody concentrations in the group of patients achieving clinical tolerance (0.01 < p < 0.05). All other comparisons of these two IgG subclass antibodies, both within and between groups, revealed no significant differences in antibody concentrations. The patients ultimately achieving clinical tolerance had initial IgE/IgG ratios that were significantly less for casein (p <0.01) and/3-1actoglobulin (0.01 < p < 0.05); at final evaluation, both milk protein-specific I g E / I g G ratios were significantly less (p <0.01) than the same ratios in patients remaining allergic (Fig. 2). Finally, the IgE/IgG ratios for
both milk proteins also decreased significantly (p <0.01) from the initial to the final evaluation in the clinically tolerant patients. The patients ultimately losing clinical reactivity to cow milk initially had signifcantly lower ~-lactoglobulin-specific IgE/IgG1 ratios (0.01 < p < 0.05) and IgE/IgG4 ratios for both casein (0.01 < p < 0.05) and/3-1actoglobulin (p <0.01) (Fig. 3). Final IgE/IgG1 and IgE/IgG4 ratios for casein and /3-1actoglobulin were both significantly less (p <0.01) in patients achieving clinical tolerance. DISCUSSION In children with IgE-mediated cow milk hypersensitivity, milk protein-specific IgE concentrations were significantly lower in patients achieving tolerance than in the group retaining milk allergy. These findings are in contrast to reports of other investigators. Hill et al. 23 studied 47 pediatric patients with cow milk allergy. Patients were classified into one of three groups on the basis of the time of onset of symptoms after initiation of open milk challenges. Fifteen patients reacted within 45 minutes (immediate reactors), 24 patients within 1 to 20 hours (intermediate reactors), and 8 patients beyond 20 hours (late reactors). Cow milk-specific IgE antibodies were not found consistently in any
376
James and Sampson
group. Tolerance to cow milk developed in 6 immediate reactors, 10 intermediate reactors, and 2 late reactors. No consistent changes in serum concentrations of cow milkspecific IgG, IgA, IgM, or IgE antibodies were associated with loss of clinical reactivity. Other investigators have shown that shrimp-specific IgE and IgG antibodies are significantly elevated in patients with positive DBPCFC reac, tions to shrimp, but no significant long-term changes occur in shrimp-specific IgE, IgG, IgA, or IgM antibody concentrations. 24 Furthermore, shrimp-allergic patients were shown to have significantly elevated shrimp-specific IgG2 and IgG4 concentrations in comparison with normal control subjects, although none of these changes were diagnostic of clinical shrimp hypersensitivity.25 In contrast to the Hill study, all our patients were well characterized as having IgE-mediated cow milk allergy confirmed by DBPCFC. Utilizing this more homogeneous population of milk-allergic children, we were able to detect subtle immunologic changes associated with the development of oral tolerance in cow milk-allergic children. The studies of Daul et al. 24, 2s involved adults reactive to an allergen (shrimp) to which patients do not generally lose their clinical reactivity. Whether the natural history of cow milk allergy and changes in antibody-specific responses may be extrapolated to other IgE-mediated food allergies remains to be determined. In a recent study of cow milk-allergic children,26 the gut immune response to cow milk antigens was indirectly assessed by means of a solid-phase, enzyme-linked immunoassay (ELISPOT). As in our study, cow milk-specific IgG antibodies were found in all children, and no significant changes were seen--whether or not children achieved clinical tolerance. Cow milk-allergic children retaining clinical sensitivity had a significant increase in circulating immunoglobulin-secreting cells after milk challenges, whereas those achieving clinical tolerance had no increase in cell numbers. The investigators concluded that "outgrowing" cow milk allergy is associated with maturation of the gutassociated immune system and loss of responsiveness to ingested food antigens. This maturational process also may be associated with the decrease in cow milk-specific IgE antibodies seen in our patients. We conclude that significant decreases in milk-specific IgE antibody concentrations are evident by the time clinical tolerance is documented in children with IgE-mediated cow milk allergy. Individual IgE milk-specific antibody concentrations at either the time o ~ e negative DBPCFC results or the final DBPCFC results are significantly lower in children losing clinical reactivity than in those remaining allergic. The milk-specific IgE/IgG ratios reflect similar differences in the two patient groups, but do not appear to
The Journal of Pediatrics September 1992
discriminate any further the differences in individual patients achieving clinical tolerance from those retaining clinical reactivity. Monitoring IgG-, IgGl-, and IgG4-specific antibodies to casein and/3-1actoglobulin is not helpful, but monitoring serum-specific IgE concentrations and calculating a ratio of IgE to IgG antibody to common milk protein allergens may be useful in predicting which patients will ultimately lose their clinical reactivity to cow milk, and may also help determine when patients should undergo additional challenges. REFERENCES
l. Savilahti E. Cow's milk allergy. Allergy 1981;36:73-88. 2. Host A, Halken S. A prospective study of cow milk allergy in Danish infants during the first 3 years of life. Allergy 1990; 45:587-96. 3. Jakobsson I, Lindeberg TA. A prospectivestudy of cow's milk protein intolerance in Swedish infants. Acta Pediatr Scand 1979;68:853-9. 4. Hide DW, Guyer BM. Cows milk intolerance in Isle of Wight infants. Br J Clin Pract 1983;37:285-7. 5. Hill DH, Firer MA, Shelton MJ, Hosking CS. Manifestations of milk allergy in infancy: clinical and immunologicfindings. J PEDIATR1986;109:270-6. 6. Cow's milk allergy in the first year of life. Acta Paediatr Scand Suppl 1988;348:2-13. 7. Vandenplas Y, Sacre L. Influences of neonatal serum IgE concentration,family history and diet on the incidenceof cow's milk allergy. Eur J Pediatr 1986;145:493-5. 8. Ford RPK, Fergusson DM. Egg and cow's milk allergy in children. Arch Dis Child 1980;55:608-10. 9. Yunginger JW. Food antigens. In: Metcalfe DO, Sampson HA, Simon RA, eds. Food allergy: adverse reactions to foods and food additives. Boston: Blackwell, 1991:38-9. 10. BleurninkE, Young E. Identification of the atopic allergen in cow's milk. Int Arch Allergy Appl Immunol 1968;34:521-43. 11. Sampson HA. Immunologicmechanisrns in adverse reactions to foods. Immunol Allergy Clin North Am 1991;11:701-16. 12. Sampson HA, Scanlon SM. Natural history of food hypersensitivity in children with atopic dermatitis. J PED1ATR1989; 115:23-7. 13. WilsonNW~.I~lamburgerRN. Allergy to cow's milk in the first year of life and its prevention. Ann Allergy 1988;61:323-7. 14. Sampson HA. Role of immediate food hypersensitivityin the pathogenesis of atopic dermatitis. J Allergy Clin Immunol 1983;71:473-80. 15. Voorhorst R, van Krieken H. Atopic skin test reevaluated. II. Variability in results of skin testing done in octuplicate. Ann Allergy 1973;31:195-204. 16. Sampson HA. Comparative study of commercial food antigen extracts for the diagnosis of food hypersensitivity. J Allergy Clin Immnol 1988;82:718-26. 17. Leinhas .IL, McCaskill C, Sampson HA. Food allergy challenges: guidelinesand implications.J Am Diet Assoc 1987;87: 604-8. 18. Fasano MB Wood RA, Cooke SK, Sampson HA. Egg hypersensitivity and adverse reactions to measles, mumps, and rubella vaccine. J PEDIATR1992;120:878-81.
Volume 121 Number 3
Development o f tolerance to cow milk
19: Neuberger MS, Williams GT, Mitchell EB, Joubal SS. Flanogan JG, Rabbitts TH. A hapten-specific chimeric IgE antibody with human physiologic effeetor function. Nature 1985;314:268-70. 20. Bruggemann M; Williams GT, Bindon CL, et al. Comparison of the effector functions of human immunoglobulins using a matched set of chimeric antibodies. J Exp Med 1987;166:135161. 21. Hamilton RG. Engineered human antibodies as immunologic quality control reagents. Ann Biol Clin (Paris) 1990;48:473-7. 22. Hamilton RG. Application of engineered chimeric antibodies to the calibration of human antibody standards. Ann Biol Clin (Paris) 1991;49:242-8. 23. Hill D J, Firer MA, Ball G, Hosking CS. Recovery from milk
377
allergy in early childhood: antibody studies. J PEDIATR 1989; 114:761-6. 24. Daul CB, Morgan JE. Lehrer SB. The natural history of shrimp-specific immunity. J Allergy Clin Immunol 1990;86:8893. 25. Morgan JE, Daul CB, Lehrer SB. The relations among shrimp-specific IgG subclass antibodies and immediate adverse reactions to shrimp challenge. J Allergy Clin Immunol 1990;86:387-92. 26. Isolauri E, Suomalainen H, Kaila M, et al. Local immune response in patients with cow milk allergy: follow-up of patients retaining allergy or becoming tolerant. J PEDIATRi992; 120:915.
BOUND VOLUMES AVAILABLE TO SUBSCRIBERS Bound volumes of the 1992 issues of THE JOURNAL OF PEDIATRICS are available to subscribers (only) from the Publisher, at a cost of $58.00 for domestic, $78.06 for Canadian, and $74.00 for international subscribers, for Vol. 120 (January-June) and Vol. 12l (July-December), shipping charges included. Each bound volume contains subject and author indexes, and all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durable buckram, with the Journal name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact M o s b y - Y e a r Book, Inc., Subscription Services, 11830 Westline Industrial Dr., St. Louis, M O 63146-3318, USA/800-325-4177, ext. 4351, or 314-453-4351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular Journal subscriptio n .