The role of histamine in allergic rhinitis

The role of histamine in allergic rhinitis Robert M. Naclerio, MD Baltimore, Maryland Studies using nasal provocation followed by nasal lavage have demonstrated that histamine plays a n important role in the mediation of allergic rhinitis but not of rhinovirus infection. The response to antigen challenge is often biphasic. During the early response, increases in histamine levels appear to be associated with activation of mast cells. In a subset of persons who exhibit an additional late response, however, the rise in histamine is concomitant with an increase in the number of basophils. Farther evidence of the role of histamine in allergic rhinitis has emerged from nasal provocation experiments involving pretreatment with a variety of drugs known to antagonize or affect release of histamine. Nasal provocation with histamine causes sneezing, itching, rhinorrhea, and nasal congestion. Most H, antihistamines have been found to inhibit sneezing and to lessen the increase in vascular permeability, but they do not affect histamine release. The H1 antihistamine terfenadine, however, inhibits histamine release during the early response; effects on late response remain unknown. Prednisone decreases histamine levels during the late, but not the early, response. One-week pretreatment with topical steroids, on the other hand, affects both the early and the late response. (J ALLERGY CLIN 1MMUNOL 1990;86:628-32.) Nasal provocation with antigen, followed by nasal lavage, is a valuable method for furthering our understanding of the role of inflammatory mediators and ceils in the upper airway. In both experimental and clinical settings, we have used nasal provocation to investigate the role of histamine in allergic rhinitis. We have also assessed the role of histamine by preventing its release or antagonizing its action with various drugs. The data gathered thus far have helped to clarify patterns of mediator activity during different phases of the allergic response. MEDIATION OF ALLERGIC RHINITIS Our early work involved a comparison of the responses of allergic (as confirmed by history and skin test) and nonallergic (control) subjects to nasal challenge with antigen. The findings showed that subjects with allergic rhinitis sneeze, in contrast to nonallergic subjects, and had related increases in levels of histamine and of prostaglandin D2. ~ These results strongly suggested that mast cells are activated to release histamine when asymptomatic, allergic persons are challenged with antigen. Castells and Schwartz 2 demonstrated increases in From the Johns Hopkins University School of Medicine. Supported in part by grant DC00320 from the National Institutes of Health. Reprint requests: Robert M. Naclerio, MD, Johns Hopkins Asthma and Allergy Center, Unit Office 7, 301 Bayview Blvd., Baltimore, MD 21224. 1/0/23310 628 mast cell tryptase among allergic persons stimulated with antigen; these data support a role for the mast cell during the early response to nasal provocation with antigen. In another study, biopsy specimens obtained by Gomez et al. 3 demonstrated that mast cells are degranulated after acute antigen exposure. MEDIATION OF RHINOVIRUS A commonly raised question is whether the observed increase in histamine is simply related to increased nasal secretions. One way to control for this possibility is to examine another disease process in which the amount of nasal secretion is known to increase and then to assess whether histamine levels rise as well. For 5 days we challenged a group of persons with rhinovirus, then compared their responses with those in a control group that underwent sham inoculation: First, we found that when we placed rhinovirus in the noses of persons with negative neutralizing titers, some subjects became symptomatic and infected (defined as an increase in the serum neutralizing titer or the ability to culture virus in nasal secretion). We observed that symptoms began to increase 24 to 48 hours after inoculation, peaked on days 2 and 3, then slowly declined during the subsequent several days. Nasal lavage showed that the levels of kinins rose dramatically during the period that these persons became symptomatic, then decreased during the next several days. In contrast, no significant increases in kinins were evident in either the persons who were infected but did not become ill, or in those who underwent sham inoculation. This evidence suggests that

VOLUME 86 Histamine in allergic rhinitis 629 NUMBER 4, PART 2 the rhinovirus response is associated with elevations in kinin levels. In all subjects--those who were inoculated with rhinovirus and either remained asymptomatic or became symptomatic, and those in the control group-- levels of histamine did not rise at any point during the 5 days of exposure, despite an increase in nasal secretions. Work in the clinical setting has also furthered our understanding of the role of mediators in rhinovirus infection. One study focused on 100 persons who underwent nasal lavage when they were well and again after naturally acquired rhinovirus infection developed. The presence of infection was associated with a significant increase in the level of kinins, in parallel to the findings in the experimental model. Histamine levels did not change significantly in the presence of infection, again confirming the findings in the earlier experiments. 5 These observations suggest that histamine is not involved in rhinovirus-induced upper respiratory tract illness. We may also conclude, therefore, that mast cells and basophils are not involved in these processes. If provocation with histamine induces symptoms and if histamine is present when symptoms are present, blocking the mediator (histamine) should have some effect on symptoms. If histamine is not produced in rhinovirus colds, the administration of an antihistamine should have no effect. To explore this possibility, Gaffey et al. 6 evaluated with naturally acquired symptoms of upper respiratory tract viral infections in 250 patients. The subjects were seen at the peak of their symptoms. Treatment was randomized such that patients received either terfenadine (60 mg twice daily) or placebo and were followed for 2 to 3 days. No difference was apparent between the symptom responses of the two groups, further supporting the concept that histamine does not play a role in upper respiratory tract rhinovirus or nonrhinovirus infection. PATTERN OF MEDIATOR ACTIVITY Further experiments have been undertaken to expand our allergic rhinitis model beyond acute mast cell degranulation to include the later inflammatory events. Nine persons were selected for study, because they had shown a biphasic response to antigen provocation. All the subjects were examined at baseline, after antigen challenge, during the early response, during the next 1 to 3 hours (quiescent phase) and 3 to 11 hours after challenge (late response). 7 Histamine levels rose dramatically during the acute response, then tended to return toward baseline during the quiescent phase, only to rise again during the late phase. This pattern was not evident in control subjects (who were either the same subjects undergoing sham 8" 6" E 4- Z 2. 30-20" 10" 0 Cetirizine Placebo SNEEZES n=10 p<0.01 ~ - N = 9 _ HISTAMINE /~ Cetirizgne /? 300. ALBUMIN - 200' E ::L 10o ---D-- Cetirizine Placebo p = 0.02 FIG. 1. Effect of cetirizine on the early response to nasal challenge. (From Naclerio RM, Proud D, Kagey- Sobotka A, Freidhoff L, Norman PS, Lichtenstein LM. Laryngoscope 1989;99:596-9.) challenge or nonallergic subjects undergoing antigen.challenge). Having previously determined that the role of histamine in the acute or early response is associated with activation of mast cells, we sought to identify the source of histamine in the late response. Assessment of the influx of alcian blue-positive cells into nasal secretions showed that the number of these ceils increased significantly during the hours after antigen provocation. 8 They represent approximately 1% of cells entering the nasal mucosa (the others being neutrophils, eosinophils, and mononuclear cells). On light microscopy the alcian blue-positive cells appeared to be composed of approximately 70% basophils and 7% mast cells; the remainder were unidentifiable. The number of alcian blue-positive cells showed a strong correlation to the amount of histamine recovered during the late response (r = 0.78, p < 0.01). This observation suggests that the alcian bluepositive cells not only were entering the nasal mucosa

630 Naclerio J. ALLERGY CLIN. IMMUNOL. OCTOBER 1990 6- SNEEZES 4. -~ 60 mg ---- ~.~3. -0-300 mg 9.~ -I~ Placebo, Z 2. n=12 B~4' e- 8" 6" 100' 80, O o 60' E o_ 40' O 20, = = J 0\\3 o\~'2 to e~'a too e~'a ~ooo e~'3 ~D\k.. t HISTAMINE "m- 60 mg 300 mg NS '~ -o- Placebo n=12 to too eeo ~obo e~'j TAME-ESTERASE ACTIVITY -11-60 mg -~ -O- 300 mg N S, -0- Placebo P<' n=12 FIG. 2. Effect of terfenadine on the early response to nasal challenge. (From Naclerio RM, Kagey-Sobotka A, Lichtenstein LM, Freidhoff L, Proud D. Reprinted from American Review of Respiratory Disease July 1990;142:167-71.) but were also degranulating and releasing histamine to participate in the allergic response. Iliopoulos et al. 9 obtained leukocytes from the nasal cavity during the late response and stimulated them to see whether they would behave as mast cells or basophils ex vivo. Cells were first challenged with anti-ige and then with f-met peptide. A dosedependent increase in the amount of histamine released from the cells occurred. The optimal anti-ige concentration proved to be 0.1 I~g/ml. In parallel studies of the blood of these persons, the dosedependent pattern showed the same optimal concentration of anti-ige for maximal histamine release. This result suggests that these cells were basophils. (For mast cells the optimal concentration would be higher, ->4 ixg/ml.) Additionally, f-met peptide has been shown to cause (peripheral blood) basophils to release histamine, whereas (human lung) mast cells do not respond to this stimulus. On the basis of these two characteristics, therefore, the cells within the nasal mucosa during the late response appear to be more like basophils than mast cells. We also investigated the effects of dimaprit on anti- IgE-induced nasal leukocyte histamine release. In control subjects, nasal leukocytes had a high spontaneous release of histamine, a phenomenon that remains somewhat puzzling at this time. Premedication with increasing concentrations of dimaprit resulted in suppression of the amount of anti-ige-induced histamine release; at the highest levels of dimaprit the level of suppression was greater than that of the spontaneous release. If basophils in the nasal secretion act as targets, reexposure to antigen should cause histamine release. The peripheral blood basophils of subjects allergic to ragweed were therefore compared with leukocytes obtained from the nose approximately 24 hours after antigen provocation. The optimal responses to antigen E (the major antigenic component of ragweed) and anti-ige occurred at approximately the same point in both nasal and peripheral blood basophils. With both stimuli, however, the response of the peripheral blood basophils was greatly attenuated. It is possible that these cells had already been partially activated to release histamine during the late response, which is when they were obtained. Because allergic rhinitis results from continuous exposure to low doses of antigen, rather than from one exposure to a large dose, the effects of challenging subjects with the same stimuli (e.g., ragweed) at different time intervals were examined. When subjects were given placebo and then challenged with increasing antigen doses, we observed a dose dependent increase in the amount of histamine released into nasal secretion. Ten hours ]ater, reexposure of subjects to a low dose (10 protein nitrogen units) of antigen greatly augmented the response as compared with the same dose given hours before, suggesting that less antigen caused a greater amount of histamine released. ~o Prednisone pretreatment had no effect on the acute response but dramatically reduced the augmented response to rechallenge that was seen after pretreatment with placebo. The finding that prednisone (which is effective in the treatment of the clinical disease) did not affect acute mast cell degranulation highlights the importance of the other inflammatory events. As mentioned earlier, a pivotal concern for researchers studying the role of histamine is whether the mediator induces symptoms when introduced to the nasal mucosa. When we conducted such a challenge in asymptomatic persons, we observed that sneezing responses increased with increasing doses of

VOLUME 86 Histamine in allergic rhinitis 631 NUMBER 4, PART 2 TABLE I. Effects of drugs on histamine release in the early, late, and rechallenge response to antigen provocation Effect on histamine release Treatment Dosage Duration ER LR RC Cetirizine 20 mg/day 2 days - ND ND Terfenadine 60 mg twice daily 7 days + ND ND Theophylline 400 mg twice daily 7 days + ND ND Flunisolide 100/xg twice daily 7 days + + + Prednisone 60 mg twice daily 2 days - + + Immunotherapy 2 p,g/injection 10 mo + +??, Data equivocal; ER, early response; LR, late response; ND, not determined; RC, rechallenge. histamine. Furthermore, we noted increased levels of tosyl-l-arginine methyl ester esterase activity and albumin (markers of vascular permeability). When we exposed the persons to antigen and then rechallenged them with histamine 24 hours later, we saw that the number of eosinophils in nasal secretions before challenge increased and that the response to histamine changed; the subjects became more sensitive to histamine, sneezing more and exhibiting more evidence of increases in vascular permeability. During the next 10 days the responsiveness to histamine decreased toward baseline from the levels present after a single antigen exposure. The number of eosinophils in nasal secretions was still elevated 10 days after a single antigen challenge. 11 In summary, then, we have described an immediate response and four subsequent events to antigen provocation: an increase in cells, an increased spontaneous release of histamine (the late response), change in reactivity to histamine, and augmented responsiveness to antigen. These late events do not occur in all persons with allergic rhinitis. EFFECTS OF DRUGS Studies have examined the effects of various drugs on the acute allergic response. One such investigation evaluated the effect of cetirizine, an H1 antihistamine, on the early response to nasal challenge with antigen. In this double-blind, crossover study, 10 subjects allergic to either ragweed or grass were premedicated with 20 mg/day of the drug for 2 days before challenge. In the presence of placebo, increasing doses of antigen produced an increase in the amount of histamine and in the amount of sneezing and vascular permeability; premedication with the drug had no effect on histamine release and thus mast cell activation (Fig. 1). Cetirizine did, however, significantly suppress sneezing and vascular permeability--two of the major actions of histamine on the nasal mucosa. 12 Another investigation examined the effect of terfenadine (also an H~ antihistamine) on the early re- sponse to nasal challenge with antigen. In this threeway, double-blind, crossover study, 12 subjects allergic to either ragweed or grass were premedicated with terfenadine (in oral dosages of 60 or 300 mg twice daily) for 7 days before challenge. Like cetirizine, terfenadine clearly reduced the amount of sneezing and vascular permeability; however, terfenadine produced a different effect on the levels of histamine recovered after challenge (Fig. 2). In the presence of placebo, histamine level increased with increasing doses of antigen. Premedication with either 60 mg (the standard clinical dose) or 300 mg of terfenadine significantly reduced this rise in histamine level. ~3 Steroids also provide relief of symptoms in allergic rhinitis. Prednisone affects the late release of histamine, as discussed previously, whereas 1 week of pretreatment with topical flunisolide significantly decreases the late release of histamine and reduces the increase that follows a rechallenge. In addition, flunisolide exerts a significant effect on the acute response, or mast cell degranulation, suggesting that the route of administration (and perhaps the duration and concentration) is important. 14 Table I presents a summary of our studies examining the effects of various drugs on the allergic response. Early histamine release was affected by terfenadine but not by cetirizine. The role of Hi antihistamines in late and rechallenge responses has not yet been studied. Theophylline had a significant effect on early histamine release as well as on the symptoms of acute allergic rhinitis. 15 Flunisolide affected all three phases of histamine release. Prednisone, in contrast, affected late histamine re/ease and the response to rechallenge but had no effect on acute response. Studies have suggested that immunotherapy, even in modest doses, will affect the release of histamine during all phases of allergic inflammation. CONCLUSION Nasal provocation studies have demonstrated that histamine levels rise in allergic and nonallergic rhi-

632 Naclerio J. ALLERGY CLIN. IMMUNOL. OCTOBER 1990 nitis, suggesting a role for mast cells and basophils. In some persons provocation results in both early and late responses, which are associated with different patterns of mediator activity. Mast cells appear to be activated during the early reaction and basophils during the late reaction. Experiments have also shown that most H1 antihistamines inhibit sneezing and the increase in vascular permeability without affecting histamine release. Terfenadine, however, inhibits histamine release during the early response. The role of Hi antihistamines in late responses remains to be assessed. At therapeutic doses theophylline suppresses the early reaction to antigen challenge, including histamine release. Prednisone reduces histamine levels during the late, but not the early, response, whereas 1- week pretreatment with topical steroids affects both phases. On the basis of the data obtained in the various nasal provocation studies, we conclude that histamine plays an important role in allergic rhinitis. REFERENCES 1. Naclerio RM, Meier HL, Kagey-Sobotka A, et al. Mediator release after nasal airway challenge with allergen. Am Rev Respir Dis 1983; 128:597-602. 2. Castells M, Schwartz LB. Tryptase levels in nasal lavage fluids as an indicator of the early allergic reaction. J ALLERGY CL1N IMMUNOL 1988;82:348-55. 3. Gomez E, Corrado OJ, Baldwin DL, Swanston AR, Davies RJ. Direct in vivo evidence for mast cell degranulation during allergen-induced reactions in man. J ALLERGY CLIN IMMUNOL 1986;78:637-45. 4. Naclerio RM, Proud D, Lichtenstein LM, et al. Kinins are generated during experimental rhinovims colds. J Infect Dis 1988;157:133-42. 5. Proud D, Naclerio RM, Hendley JO, Gwalthey J. Kinins are generated in nasal secretions during natural rhinovirus cold. J Infect Dis 1990;161:120-3. 6. Gaffey MJ, Kaiser DL, Hayden FG. Ineffectiveness of oral terfenadine in natural colds: evidence against histamine as a mediator of common cold symptoms. Pediatr Infec Dis J 1988;7:223-8. 7. Naclerio RM, Proud D, Togias AG, et al. Inflammatory mediators in late antigen-induced rhinitis. N Engl J Med 1985;313:65-70. 8. Bascom R, Wachs M, Naclerio RM, Pipkorn U, Galli SJ, Lichtenstein LM. Basophil influx occurs after nasal antigen challenge: effects of topical corticosteroid pretreatment. J AL- LERGY CLIN IMMUNOL 1988;81:580-9. 9. Iliopoulos O, Naclerio RM, Baroody F, Kagey-Sobotka A, Lichtenstein LM. Histamine release by nasal leukocytes [Abstract]. J ALLERGY CLIN IMMUNOL 1989;83:236. 10. Pipkorn U, Proud D, Lichtenstein LM, et al. Effect of shortterm systemic glucocorticoid treatment on human nasal mediator release after antigen challenge. J Clin Invest 1987;80: 957-61. 11. Walden SM, Proud D, Lichtenstein LM, Kagey-Sobotka A, Norman PS, Naclerio RM. Nasal challenge with antigen (AG) leads to increased sensitivity to histamine [Abstract]. J AL- LERGY CLIN IMMUNOL 1988;81:282. 12. Naclerio RM, Proud D, Kagey-Sobotka A, Freidhoff L, Norman PS, Lichtenstein LM. The effect of cetirizine on the early allergic response. Laryngoscope 1989;99:596-9. 13. Naclerio RM, Kagey-Sobotka A, Lichtenstein LM, Freidhoff L, Proud D. Terfenadine, an H1 antihistamine, inhibits histamine release in vivo in man. Am Rev Respir Dis 1990; 142:167-171. 14. Pipkorn U, Proud D, Lichtenstein LM, Kagey-Sobotka A, Norman PS, Naclerio RM. Inhibition of mediator release in allergic rhinitis by pretreatment with topical glucocorticoids. N Engl J Med 1987;316:1506-10. 15. Naclerio RM, Bartenfelder D, Proud D, et al. Theophylline reduces histamine release during pollen-induced rhinitis. J AL- LERGY CLIN IMMUNOL 1986;78:874-6. DISCUSSION Question. One of the supposed advantages of the newer antihistamines is a lower incidence of anticholinergic side effects. Would the addition of an anticholinergic agent to therapy with a newer antihistamine be of benefit in the treatment of rhinitis? Dr, Naclerio. In Europe, where anticholinergics have been available for some time, they have been most effective in stemming watery nasal secretion but are not potent in treating the other symptoms of allergic rhinitis. Question. Your data on nasal secretions in persons with viral upper respiratory tract infections appear to show that kinin levels fluctuate with almost a circadian rhythm, peaking at approximately 4 AM each day. Do you have an explanation for that observation? Dr. Naclerio. I do not have an explanation, although circadian rhythms in the nose have been demonstrated. Levels of albumin, for instance, have been found to fluctuate within the nose. It is possible, therefore, that kinin fluctuates with a circadian rhythm. Another possible explanation might relate to the fact that at 4 AM most people are supine. In our experiments we actually woke up our subjects to tavage their noses. We cannot say, therefore, whether the findings were related to the difference in position or to a circadian rhythm. Question. The baseline level of histamine in nasal lavage fluid is 10 to 100 times higher than that in the circulation, even in nonallergic, asymptomatic persons. Are you satisfied with the concept that the source of the histamine is primarily mast cells and basophils, or do you believe we should be looking for another source? Dr. Naclerio. Basal levels of histamine are indeed present in the nose; this is why we must lavage the nose repeatedly to reduce the level to a stable baseline before conducting experiments to assess changes in histamine. Some of our subjects with rhinovirus who had elevated histamine levels early were repeatedly washed, and the level did not rise again. One potential source might be bacteria, which can synthesize histamine. Some evidence suggests that Haemophilus influenzae can generate histamine. The point is, however, that we were not looking at absolute levels as a basis for interpreting our findings; we were examining changes from baseline as an indication of mast cell or basophil activation,