General discussion I

General discussion I

Galli: There are two topics that I would like us to address in this general discussion. One is the clinical di¡erences in anaphylaxis between patients who develop anaphylaxis in the context of severe atopy, as opposed to those who have no evidence of atopy. The other topic is the controversy regarding the potential reclassi¢cation of anaphylaxis and the utility of the concept that everything that clinically looks like anaphylaxis shall be called anaphylaxis, as opposed to attempting to ferret out the different mechanistic causes of this phenomenon. Perhaps ¢rst we could consider anaphylaxis in atopic and non-atopic individuals.

Golden: I’ll start by commenting on some ¢ndings we made some years ago in our epidemiological survey of insect sting anaphylaxis. We found a dissociation between symptoms and atopy (Golden et al 1989). The presence of IgE to inhalant allergens correlated clearly with the presence of IgE to venom allergens. But atopic history of rhinitis or asthma had no correlation to anaphylaxis to insect stings. It complicates the discussion about ‘what is atopy?’ and ‘how do IgE level and atopy relate to the expression clinically of anaphylaxis?’.

Galli: I’d like to be more speci¢c about my question. There seems to be no doubt that patients who do not ¢t the classical de¢nition of atopy can develop anaphylaxis to, for example, venom and peanuts. The question is, however, when atopic individuals develop an anaphylactic reaction, is it in any way more severe or di¡erent than the anaphylaxis that develops in patients without atopy?

Ring: We did a study involving large numbers of patients which showed that insect venom anaphylaxis is not more frequent in atopics than in non-atopics. I think everyone agrees with this. But the concentration of IgE antibodies against bee venom was higher in atopics, independent of the severity. For practical purposes, in the atopic patients high IgE means less for their clinical symptomatology compared to non-atopic patients. An atopic patient may have class 5 or 6 without very severe symptoms, whereas if a non-atopic class 5 is probably more signi¢cant.

Golden: Is it clinically important? Do the non-atopics have more severe reactions with high IgE?

Ring: This is what we guess, but there is no good study.

Mˇller: There has been an interesting ¢nding in beekeepers who are allergic to bee venom. In this group atopy is signi¢cantly more frequent than in non- beekeeper patients who are allergic to bee venom. There was a suggestion that this is caused by the fact that atopic beekeepers inhale tiny amounts of the venom during their work in the beehive and thus get sensitized through the mucosal route (Miyachi et al 1979). If you look at symptoms, atopic patients with bee venom allergy more commonly have respiratory symptoms than non-atopics.

Vercelli: I have a general question. We have here a remarkable group of people who know everything about mast cells and basophils. From my point of view, if we think about potential genetic mechanisms, the question is, are there individuals with a genetically determined deregulation in their pathways that leads to degranulation and mediator release because of a lower threshold? In such individuals, stimuli that would normally be harmless may become pathogenic.

Galli: This is an interesting topic. Individual variation in the responsiveness of basophils and mast cells derived from di¡erent patients has been studied by a number of investigators, including Don MacGlashan and Gianni Marone. Perhaps we could discuss this after Larry Schwartz’ presentation on the e¡ector cells. However, getting back to one of the general discussion topics, I am still not certain as to whether we have a consensus on the issue of anaphylaxis being more or less severe in patients with atopy. Or does it depend on the circumstances?

Sampson: At least when we look at patients with food allergy, the patients who die virtually all have asthma and are atopic. Among the food allergic group atopy seems to be extremely harmful.

Galli: I would like to attempt to link this observation to one of the points that I made in my introduction. Mast cells and basophils in mice or humans with high levels of IgE have high levels of surface expression of the FceRI, and these cells thereby have heightened ability to release mediators when activated by IgE- dependent mechanisms (Hsu & MacGlashan 1996, Yamaguchi et al 1997, 1999, Lantz et al 1997, Williams & Galli 2000, MacGlashan et al 1997). Speci¢cally, in vitro studies indicate that cells with high levels of surface expression of FceRI both undergo activation at lower concentrations of antigen and also, upon activation, can release larger amounts of pre-formed and lipid mediators, and cytokines (Yamaguchi et al 1997, 1999, Williams & Galli 2000, MacGlashan et al 1997). This could be one explanation for why people with very high levels of IgE may experience more severe anaphylaxis reactions to the same antigen than those with low levels of IgE.

Sampson: The one other observation that we have made is that we don’t see a good correlation between the level of speci¢c antibody and the severity of the allergic response. We see people with very low levels of food-speci¢c IgE who die, and people with extremely high levels that get just some eczematous symptoms. We haven’t seen a good correlation there, but there does seem to be this thing with asthma.

Pumphrey: Looking at the fatal cases and the cases we see in clinic, the severity of the responses does not seem to depend on the severity of allergy, but on concomitant pathology. We recently had a striking illustration of this. A child at the age of six weeks was given cow’s milk for the ¢rst time. He had a generalized allergic reaction with urticaria, local symptoms in the mouth and so on. That child then had bronchiolitis, went on to have viral-associated wheeze, and at the age of 5.5 months was given a similar dose of cow’s milk, resulting in a fatal asthmatic anaphylactic reaction. I suggest the severity of the allergy hadn’t changed; there was simply concomitant pathology that made the reaction more dangerous.

Galli: This certainly seems to be a well documented clinical observation. I am having a little trouble with the conclusion that perhaps the allergy hadn’t changed in the interval, though. Isn’t that an assumption? The child was allergic in the ¢rst instance and the second, but couldn’t there have been some signi¢cant change in an aspect of the allergy during the intervening time?

Pumphrey: There are several possibilities. The level of IgE at the time of death was not outstandingly high (milk-speci¢c 6 kUA/L, total 22 kIU/L). Looking at that, there was no reason to assume that the child had extreme anaphylactic sensitivity, but it is possible that he died from an asthmatic reaction because of concomitant pathology in the lungs.

Golden: We have reported a number of times that there is not a consistent correlation between the level of venom-speci¢c IgE and the outcome (Golden et al 2001). There is a paradox, if you will. However, there is a statistical correlation but it is the outliers that are the puzzle. It is the patient with barely detectable IgE who dies from sting anaphylaxis, or the patients who have late phase large local reaction to insect stings but rarely get systemic reactions despite markedly elevated venom IgE. I’d also like to put a slightly di¡erent spin on something that Hugh Sampson said. I’ll suggest that there are at least two di¡erent phenotypes in anaphylaxis: the respiratory pattern and the vascular pattern. They can coexist, but perhaps what is occurring in the paediatric population is that they rarely get vascular symptoms - they are predominantly respiratory or cutaneous - and this may correlate more with atopy. The atopic children may have more of that respiratory pattern and fatal asthmatic deaths, whereas the vascular pattern which is more common in adults may not behave with that atopic or respiratory correlation.

Schwartz: In any de¢nition one comes up with for anaphylaxis, it is just as important to consider the cells involved as it is to distinguish between whether it is IgE or non-IgE. It is important to decide whether anaphylaxis is mast cell- dependent or whether it depends upon other cell types. We participated in a study with John Yunginger looking at fatal anaphylaxis (Yunginger et al 1991). We divided the groups into those that had been exposed to the precipitating factor parenterally versus orally. The parenterally exposed subjects had low levels of antigen-speci¢c IgE in their postmortem serum, but very high levels of mature tryptase. In those that succumbed to an oral challenge, they had high levels of antigen-speci¢c IgE, but low levels of mature tryptase. This suggests that there is a di¡erent magnitude of mast cell responsiveness depending on the route of administration. In a study that Peter van der Linden (van der Linden et al 1992), sensitive individuals were systemically challenged with insect stings; anaphylaxis occurred in a number of them. If you look at the baseline levels of total tryptase in their blood prior to anaphylaxis, there were higher baseline total tryptase levels in those who were destined to have a more severe anaphylactic reaction, perhaps re£ecting higher mast cell numbers. There have been some reports in the literature suggesting that having a high baseline level of tryptase may be a risk factor for anaphylaxis. Although this remains to be con¢rmed, if true, does it re£ect simply an increase in mast cell number, or instead some kind of priming or hyper-releasability of the mast cells?

MacGlashan: We have discussed several times the correlation (or lack of) for an antigen-speci¢c IgE response and anaphylaxis. The other part of this is the ratio of the antigen-speci¢c IgE to the total IgE. It is generally not the case that you can cross-calibrate between antigen-speci¢c IgE, measured by radioallergosorbent test (RAST), and total IgE, measured by radioimmunosorbent test (RIST). In fact, the absolute ratio is never known (unless a laboratory explicitly cross-calibrates the two assays). In terms of the studies mentioned, we discuss this relationship between speci¢c IgE and the anaphylactic response, but I take it it is not actually known what the speci¢c to total IgE ratios are.

Golden: In the insect sting cases we have not systematically looked at the ratio of speci¢c versus total serum IgE. We are in the process of doing that in our latest series of sting challenges, but I have no data yet.

Ring: Coming back to the relationship between atopy and anaphylaxis, I think the route of contact is crucial. Atopy is a hypersensitivity of the surface: the mucous membranes and the skin. Therefore, when the antigen comes via the surface, there is a connection. When the antigen comes parenterally, like a drug, there is no relation.

Lasser: It may be a mistake to look at death as an endpoint for the kinds of things that we are discussing. In contrast material reactions, for example, it is known that the incidence of total reactions is highest from the ages of 10^45. The incidence of total IgE is likewise highest in these ages, but death rates peak about age 70 (Lasser et al 1997). In other words, I think there are concomitant circumstances that go into the production of death at that time, versus earlier times. The patients are more fragile.

Galli: Most people would probably agree that if a patient who undergoes anaphylaxis also has signi¢cant pre-existing cardio-pulmonary problems, then it will be more likely that the anaphylactic reaction will be fatal. However, severe anaphylactic reactions resulting in death also occur in the younger age group. That outcome can occur both in the atopic and non-atopic populations. It is probably fair to say that the most compelling data indicating that an atopic history is a predictor of a severe outcome are those from the food allergic patients. So, let’s go back to the publication entitled ‘A revised nomenclature for allergy’ (Johansson et al 2001). I think this is the report that Johannes Ring was referring to in his paper when he mentioned the proposal that ‘anaphylaxis’ be used as a general clinical term. A number of you have commented that it is critical to understand the actual mechanisms that are responsible for producing this clinical picture in a number of di¡erent settings, but that these mechanisms may not always be known at the time when it may be necessary to explain the cause of death. Therefore, as I understand the recent proposal, it is to use the term anaphylaxis for all of these cases and then leave to the basic and clinical investigators the problem of ferreting out the mechanism(s) responsible. Would anyone like to comment on this proposal?

Metzger: As a non-clinician, I would opt for Frank Austen’s position. It seems to me that at the autopsy or death report one shouldn’t go beyond one’s knowledge. If the patient died of respiratory death or shock, this should be put down. As soon as you use a term such as anaphylaxis you are implying something for which you may lack ¢rm evidence. At one stage sarcoidosis was not distinguished from tuberculosis: you could make the same argument there. To the extent that we know at least one cause of a particular syndrome, it is useful to have a speci¢c term for that syndrome caused by that pathology. That similar symptoms may occur due to other forms of pathology shouldn’t justify the use of the same term.

Pumphrey: My understanding is that this new de¢nition does allow for that. Anaphylaxis is divided into allergic and non-allergic, and the allergic is divided into IgE-mediated and non-IgE mediated. What you are talking about is IgE- mediated anaphylaxis, so you could possibly defend the European Academy of Allergology and Clinical Immunology de¢nition.

Schwartz: How would you then classify the hypotensive shock that occurs from complement activation, due to dialysis membranes, for example? This doesn’t involve mast cells or IgE. Would you call that anaphylaxis?

Pumphrey: That’s a non-IgE-mediated anaphylaxis.

Marone: Frankly, I am a little confused by that position paper on the classi¢cation of allergic diseases. As previously mentioned by Frank Austen I fully understand the meaning of IgE-mediated anaphylaxis. It is more di⁄cult for me to understand non-IgE-mediated anaphylaxis. Also in that paper it has been suggested to eliminate the use of the term ‘anaphylactoid’ reactions. We have to discuss this.

Galli: I have always liked that term, because it covers a whole variety of situations where we lack the information to say that something is anaphylaxis.

Mosbech: From a clinical point of view we have to keep things simple. We have

heard that a lot of these patients are not treated as rapidly as they should be. If the physicians have to think about whether this is one or the other, this might hinder prompt treatment. So it might be a good idea to call all these responses anaphylaxis.

Sampson: I can think of a scenario where that would be harmful to the patient. We have seen a syndrome in young children called milk-induced enterocolitis. These children present with severe repetitive vomiting, marked hypotension and look extremely sick. If you then diagnose that as anaphylaxis and give epinephrine, this will do nothing for them. What they need is large volume expansion. This is part of the therapy, but the ¢rst response would be to give them epinephrine, and this is not what these children need.

References

Golden DBK, Marsh DG, Kagey-Sobotka A et al 1989 Epidemiology of insect venom sensitivity. J Am Med Assoc 262:240^244 Golden DBK, Kagey-Sobotka A, Norman PS, Hamilton RG, Lichtenstein LM 2001 Insect

sting allergy with negative venom skin test responses. J Allergy Clin Immunol 107:897^901 Hsu C, MacGlashan D Jr 1996 IgE antibody up-regulates high a⁄nity IgE binding on murine bone marrow-derived mast cells. Immunol Lett 52:129^134

Johansson SG, Hourihane JO, Bousquet J et al 2001 A revised nomenclature for allergy. An EAACI position statement from the EAACI nomenclature task force. Allergy 56:813^824 [Erratum in Allergy 56:1229]

Lantz CS, Yamaguchi M, Oettgen HC et al 1997 IgE regulates mouse basophil Fc epsilon RI expression in vivo. J Immunol 158:2517^2521

Lasser EC, Lyon SG, Berry CC 1997 Reports on contrast media reactions: analysis of data from reports to the U.S. Food and Drug Administration. Radiology 203:605^610 [Erratum in Radiology 204:876]

MacGlashan DW Jr, Bochner BS, Adelman DC et al 1997 Down-regulation of Fc(epsilon)RI expression on human basophils during in vivo treatment of atopic patients with anti-IgE antibody. J Immunol 158:1438^1445

Miyachi S, Lessof MH, Kemeny DM, Green LA 1979 Comparison of the atopic background between allergic and non-allergic beekeepers. Int Arch Allergy Appl Immunol 58:160^166 van der Linden P-WG, Hack CE, Poortman J, Vivie¤ -Kipp YC, Struyvenberg A, van der Zwan JK 1992 Insect-sting challenge in 138 patients: relation between clinical severity of anaphylaxis and mast cell activation. J Allergy Clin Immunol 90:110^118

Williams CM, Galli SJ 2000 The diverse potential e¡ector and immunoregulatory roles of mast cells in allergic disease. J Allergy Clin Immunol 105:847^859

Yamaguchi M, Lantz CS, Oettgen HC et al 1997 IgE enhances mouse mast cell Fc(epsilon)RI expression in vitro and in vivo: evidence for a novel ampli¢cation mechanism in IgE-dependent reactions. J Exp Med 185:663^672

Yamaguchi M, Sayama K, Yano K et al 1999 IgE enhances Fc epsilon receptor I expression and IgE-dependent release of histamine and lipid mediators from human umbilical cord blood derived mast cells: synergistic e¡ect of IL-4 and IgE on human mast cell Fc epsilon receptor I expression and mediator release. J Immunol 162:5455^5465

Yunginger JW, Nelson DR, Squillace DL et al 1991 Laboratory investigation of deaths due to anaphylaxis. J Forensic Sci 36:857^865