MEDICAL PROOF

The proof has been published in various medical references and is presented here:

The concept that otitis media with effusion is in some way related to allergy is not new. As outlined in the Introduction Section, this has been a consideration since Proetz noted a relationship between patients with allergic rhinitis and chronic OM 65 years ago[126] and Koch observed eosinophilia in the otorrhea from 222 children, “supporting the contention that the middle ear takes part in allergic reactions similar to those seen in the nose and sinuses”.[89] In 1965 Fernandez and McGovern[126] suggested that an allergic mechanism, while not the major cause of chronic otitis media, is a predisposing factor in as many as 85% of children with acute otitis. Shambaugh suspected allergy as an etiology of chronic draining mastoid cavities or middle ears of patients with OME, citing empirical data. He cautioned that “surgical mastoidectomy, simple or radical is not indicated. With competent allergic diagnosis and management, preferably by the otologist trained in allergic methods, the otorrhea is finally brought under control.”[143] Sprinkle’s review of the literature on the pathophysiology of OME found “solid evidence....to suggest that Type I immune injury can be considered a major contributing factor to persistent middle ear effusion.” He also stated that Type III hypersensitivities that require the presence of microorganisms were “very important,” and Type IV may also “play a role in causing and potentiating serous otitis media in man.”[150]

Most negative opinion regarding the “allergy connection” is based on the results published by Suehs.[154] Using the technology of 1952, he was unable to observe any eosinophils in 50 patients, nor did Senturia[141] who in 1960 also did not find eosinophils in MEE of patients suspected of having allergic otitis media. Senturia[141] and Sade[137] attributed the basic cause of OME to infection in the nasopharynx with retrograde contamination of the middle ear and edema of the eustachian tube. Their thinking has dominated otology to this day.

Epidemologic studies in Japan[164] and Sweden[74] have shown a significant relation of allergy to OME. Although only 6 to 20% of the general population are atopic and among atopics only 21% have OME, over 87% of OME patients were found to be atopic and/or have allergy symptoms.[164] Irander[74] found that among 54 Swedish infants 38% with OME had respiratory tract allergy. Infants with allergy symptoms were 5 times more likely to develop OME than non-atopics! Karma found similar results in Finland where allergy posed a risk factor of 4.4 for children failing to clear an acute otitis.[77] Among our own patients in Study I, 62% had documentation of additional atopic signs and symptoms including asthma 22%, allergic rhinitis 48%, eczema 4%, and chronic nasal congestion 8%. It is noticeable though that several studies have refuted a relationship between allergy and OME. In contrast to the above studies these were epidemiological studies in which the diagnosis of allergy was determined from questioners of allergy of among siblings and parents. [100],[48] Demonstration of any relationship of otitis to allergy depends on determining if the patient is atopic. Thus, there seems to be a discrepancy between those studies and other studies that attempted to objectively establish the relationship between OME and allergy either by investigating the prevalence of other allergic symptoms in the actual patients or by the demonstration of specific IgE in the patients.

Eustachian tube dysfunction, either extrinsic or intrinsic, is regarded as the underlying pathophysiologic event which leads to chronic middle ear disease. Obstruction of the ET in humans has been clearly demonstrated to result from antigen challenge.[50],[14],[2] The most frequently cited objection to the hypothesis that allergy might in any way attribute to middle ear disease is that an allergen is unlikely to get into the middle ear itself due to the structural gatekeeper function of the E.T.[43],[73]

Several possible mechanisms of immune response in the ear have been proposed. These do not rely on the premise of direct allergen transport to the middle ear, but rather are dependent on an understanding of both humoral and cell mediated immunology. The middle ear is not a privileged site, devoid of immune response mechanisms as was taught in the 1960’s. Middle ear mucosa, which evolves from the same ectoderm as the rest of the upper respiratory epithelium, has in recent years been found in animal studies to have the same active intrinsic immunologic responsiveness to antigenic stimuli as do the nasal tract, sinuses and bronchi.[18],[159] Chronic inflammation in the middle ear generates mucosal hypertrophy, fibrosis, osteolysis,[116],.[11] cartilage damage[102] and hearing loss.[56] If an allergen does not have direct access to the middle ear then how is it that such a hypersensitivity reaction may occur? Physicians who understand immune-mediated diseases acknowledge that in asthmatics an immune mediated response by respiratory mucosa to an allergen may produce general edema, bronchospasm, increased mucous production and effusion within pulmonary alveoli[75] even though some allergens may not actually penetrate the airway beyond the nose or bronchi. The current hypothesis of a common mucosal immune system suggests that activated T lymphocytes are able to migrate from one mucosal site to another. Exposure of one epithelial surface by an infectious or antigenic agent leads to secretion of locally produced antibodies.[171] This is demonstrated classically in eczema or asthma secondary to peanut allergy. Current studies support the hypothesis that an abnormally increased absorption of allergen is pathogenic in airway diseases of rhinitis and asthma.[57] It is now recognized that human nasal airway mucosa is the focus of absorption of allergens as well as microorganisms which activate the mucosal defense systems. Secretory immunity is the best defined effector mechanism of antigen presentation and stimulation of the immune system in humans and is elegantly described by the work of Jahnsen.[75]

The great difficulty in attempting to demonstrate that the middle ear might be involved in allergic disease is the ethical impossibility of performing direct challenge studies. Therefore we are forced to rely on circumstantial evidence in the form of animal experiments, histologic studies and clinical outcomes. The mechanisms that regulate the activity and accumulation of eosinophils and mast cells and their mediators which lead to tissue injury in the atopic patient are governed by the Th2 type of lymphocyte with activation of cytokines such as IL-4 and IL-5.[33],[75] The middle ear has clearly been demonstrated to be an immunoreactive site in many animal models.[18],[106],[92] ,[105] ,[84],[42] Takeuchi has concluded that the T-cell antigen receptor is involved in the regulation of all antigen specific T-cell responses in human middle ears. [159] He also raised the possibility that antigens retained in MEE following an acute infection are responsible for activating T-cells and subsequent cytokine production. It has been proposed that the late-phase reaction in the ear is under the control of cytokines and is produced by cells in the tubo-tympanic cavity[18] instead of selective recruitment of lymphocytes from the circulation, as previously thought.

Evidence based medicine proves that various well established past explanations for chronic OME are myths. It is still taught that “Otitis will stop once the eustachian tube grows to a normal size” even though Sade showed that the ET of OM patients does not have an “immature morphology”. The pharyngeal and isthmus portions of the ET from OM patients is no different in size than that of normal children.[138] Parents are told that their otitis prone child cannot equalize the pressure in their middle ears, yet the fact is that there is no organic obstruction or stenosis of the ET in OME patients. Actually only 11% of active OME patients have abnormally high opening pressures.[157] Finally, there is the myth of antibiotics. Rosenfeld and Bluestone[134] have shown that meta analysis of all treatment studies demonstrate antibiotics to be no better than placebo. In fact 70% of effusions resolve whether treated or not. They noted that their review included no studies of allergy management because there are no Double Blind Placebo Controlled studies involving allergy management of otitis to review. Our concern is those 30% of children who’s effusions do not resolve despite conventional therapies and time. However, several studies have suggested that conventional allergy management may reduce the incidence of relapses or complications of OME.[66],[101],[110],[127]

There are few studies in which humans or animals have been challenged with an allergen which have resulted in the development of OME, in part because of the lack of an natural model, and in part by design. Experiments with Rhesus monkeys produced ET obstruction after a four day challenge with ragweed. These animals failed to develop OME, perhaps because MEE does not form in monkeys without transection of the tensor veli palatini muscle.[28] Experiments with chinchillas and mice are more successful in producing MEE.[106],[73],[156] Surgical obstruction of the ET will only produce OME in 25% of chinchillas.[106] Fireman and Bluestone demonstrated that exposure of the human nose to antigen leads to ET obstruction and is dose dependent. The authors were careful to point out that: “The fact that OME did not occur in our preselected adult patient cohort was anticipated and an expectation of our experimental design.” They chose allergic adults who had no history of OME “so as to avoid creating complications”.[2]

Some animal experiments do support the concept that an immune-mediated response is localized within the middle ear itself, involving eosinophils, mast cells and neutrophils. Ryan has shown that in middle ears of mice local immunoregulation mediated by cytokines may be responsible for the amplification and differentiation of IgA-producing cells located in the middle ear mucosa. Ears with a chronic immune response showed extensive mucosal hyperplasia consisting of lymphocytes and neutrophils in the submucosa with neutrophils, macrophages and diffuse eosinophilic material in the effusions. This is accompanied by numerous IL-5+ cells in the subepithelial tissue and effusion in the chronic stages.[18] In asthma IL-5 has been shown to be a critical cytokine associated with eosinophil and lymphocyte recruitment following exposure to allergen.[155] Nakata also demonstrated high levels of histamine and prostaglandins in MEE that he suggested are local products rather than the result of a transudate from plasma.[106] Study IV demonstrated IL-5 in human MEE. In an elegant study using a rat model, Labadie[92] was able to demonstrate that allergen presentation to the middle ear of allergic rats caused ET dysfunction which resulted in middle ear effusion. Pollock[124]showed that when allergen at a dose 1/50 of that needed to cause local inflammation is presented to the nose it results in failure of those allergic rats to be able to demonstrate normal ET function.

An immune response involving IgG, complement activation, neutrophils, eosinophils and mast cells has been implicated in the generation of immune-mediated OME by Suzuki.[156] The perpetuation of chronic inflammation by cell mediators is further supported by the experiments of Nakata who was 100% successful in creating effusion when he either exposed the middle ear of pre-immunized animals to antigen or when he injected effusion fluid itself into a sterile ear. He concluded that middle ear “effusion itself induces inflammatory reactions in the middle ear cavity.”[106] Prostaglandins[106]and leukotrienes (LTC4)[81] have also been demonstrated in human ears. Study I similarly demonstrated neutrophils, eosinophils and mast cells and their mediators in human MEE in atopics.

Miglets[103] and Yamashita[178] performed histologic studies which showed that allergen presentation to the middle ear did result in inflammation, but were reluctant to attribute this to allergy because of the lack of eosinophils. Mogi reviewed animal studies and concluded that although the Eustachian tube “is involved, both functionally and morphologically, in Type I allergic reactions of the nose and that the tubal dysfunction evoked by nasal allergic reactions is transient.” He postulates that since this inflammation does not culminate in the development of MEE, then “it is unlikely that Type I allergy is a casual factor” in the development of OME “even if the middle ear is an allergic shock organ.” One might argue that what in fact Mogi describes as subjects being in a “chronic state of disease” is exactly what constitutes an atopic patient: namely an individual whose eosinophils, neutrophils and mast cells are primed so that when exposed to an antigen they react and release their various mediators, in contrast to the same cells in a non-atopic individual which do not react on similar exposure.[45],[153] Fireman concluded that the interaction between viral infection and nasal allergy might enhance certain responses and that such an enhancement might explain why allergic rhinitis is a risk factor for otitis media..[51]

The etiological relationship of allergy to OME is very controversial especially since from 1958 to 1984 Suehs,[154] Senturia,[142] Boor,[22] Siirala,[145] Lim,[96] and Reisman[131] could not find any clinical basis of allergy in the formation of OME. Typical was the guarded negative conclusion by Boedts whose study “failed to support the concept of allergy as a major causative factor in OME....although it may occur as a complication of an allergic (nasal) disease.”[21] Several reviews include critical analyses of decades of studies, both clinical and histological, that were designed to evaluate the relation between allergy and OM.[8],[44],[112],[178],[21],[164],[136] Unfortunately many reports including those published since the discovery of IgE are often less than critical, quote old data,[19],[93] are not based on present day standardized methods of allergy testing, apply allergy treatments to all patients regardless of diagnosis, or narrowly define their clinical population of allergic to include only those with a history of rhinitis.[13],[178]],[149] Tomonaga criticized many of the aforementioned works. He found 21% of 605 allergy patients had OME, but among 259 OME patients, 87% were atopic by skin testing, although only 50% of these had nasal allergy.[164] He determined the incidence of allergies among patients with OME to be four to five times that expected in a similar age-adjusted population of normal individuals. This suggests that the relationship may be more than coincidental and is supported by the high prevalence among our patients.

Since OME often appears to be a disease of atopics, it seemed logical to ask: “Is OME an allergic disease?” As stated, an allergic response involves activation of Th-2 cells by environmental antigens with resultant up regulation and production of IL-4 and 5 and an expression of IgE on mast cells and eosinophils with a resultant release of tryptase, ECP and other mediators of inflammation.[151 This combination of events has heretofore not been described in OME. Clinically, the diagnosis of Type I hypersensitivity is based on the detection of allergen specific IgE by means of skin testing and/or in vitro testing. This thesis includes studies I,II,IV which are the first to our knowledge to correlate the expression of markers of eosinophil, neutrophil, mast cell, CD+3, and IL-5 mRNA activity found in middle ear effusion and/or mucosal biopsies from patients with OME to their atopic status.

Prior studies which led otolaryngologists to believe that less than 30% of OME was related to allergy were based on unusually narrow definitions of atopy requiring both rhinitis and a total IgE > 100µg/L or results of prick testing,[16] Our data shows that the mean total IgE among atopics was 93.8µg/L with two-thirds of atopics with OME having a serum IgE < 100 µg/l.(Fig. 5) Otitis is thus similar to rhinitis in having no relation to total IgE, unlike asthma which does show correlation.[30] Our reported prevalence of atopy of 81% among the group of OME patients in our studies(I,II) may reflect the increased sensitivity of modern in vitro testing as compared to results obtained from prick tests or questionnaires. This percentage of allergy among OME patients is consistent with prior reports based on in vitro testing by McMahan (88%),[101] Tomonaga (72%).[164], Nsouli (86%),[110] Corey (60%),[32] and Psifidis (78%).[127] and as discussed in the Methods Section, does not reflect a selection bias.

Histologic Studies: The fact that conventional histology does not readily detect degranulated or activated neutrophils, mast cells or eosinophils leads to various conclusions[102],[116],[112],[156] and is the major reason this controversy has been perpetuated. There is also disagreement as to whether mediators in middle ear effusion come from the plasma or local tissue. Using animal studies Nakata found few eosinophils in the effusion of immunized chinchillas in the acute phase of inflammation but also concluded that “middle ear effusion is a local product of the middle ear mucosa rather than a transudate from plasma”.[106] Histopathologic examination of effusion demonstrates that eosinophils and neutrophils are integral components in these secretions,[82],[116] but not necessarily present in the mucosal lining. Despite Koch’s initial description of cytologic changes in chronic OM, in which he found eosinophils in the mucosa from 52 of 62 cases of allergic otitis,[89] eosinophil involvement is not usually observed and mast cells are not described in the middle ears of children with OME.

Previous biopsies of middle ear mucosa from atopic patients have shown that the active degranulation of eosinophils correlated with elevated levels of effusion mediators.).[69] The flat epithelial cells found in middle ear mucosa are known to be able to change into either secretory or ciliated cells when challenged by pathological stimuli; this was also seen in our earlier biopsy study[69] and was confirmed in Study III. These studies suggest that the inflammation observed in the promontory mucosa of 25 patients(Table III) with OME, proven by in vitro testing to be atopic, appear to be accompanied by the continual recruitment and activation of eosinophils and mast cells. The histologic findings in the ears from normal subjects(II,III,IV) correlated with the descriptions by previous investigators.[47],[86]

The high numbers of eosinophils and the high density of EG2/MBP staining deposits within the tissue and mucous (III,IV) as well as the appearance of mast cells in middle ear epithelium (II) is an indication of disease and not a normal feature.[76] Anti-tryptase antibody(AA1) staining (II) showed that mast cells are present in the mucosa and submucosa in atopics (Fig. 7) as compared to the controls. Study II found 61% of 33 chronic OME patients had extensive activation of mast cells in their middle ears.(Fig. 6) The presence of mast cells and eosinophils or their mediators in the target organ is essential, although not pathognomonic, for a disease process to be considered allergic.

Inflammation is exclusively an in vivo phenomenon that only occurs in living tissues with an active microcirculation.[120] Perpetuation of that inflammation, regardless of origin, is the crucial difference between recurrent acute otitis media and OME. A basic question is whether the middle ear inflammation was the result of infection or allergy or both. In Study I atopy seemed to have a significant relationship to whatever was producing a response from eosinophils and mast cells in the middle ear(Fig. 2, 4). This is a significant observation not only in regards to the association to allergy but also because human eosinophils are much more toxic than neutrophils, making them particularly harmful to host tissues.[132] Demoly notes that “chronic sinusitis, which occurs in some patients with allergic rhinitis, can be distinguished on the basis of T cells and intraepithelial mast cells.”[35] He found 8 times more T cells in patients with allergic rhinitis than in patients with nonallergic rhinitis. In Study IV we have demonstrated that 8 of 8 atopic patients with chronic OME had increased CD3+ T cells, vs none of the controls. This, coupled with the presence of degranulating mast cells(I,II) and EG2 positive eosinophils(III) in the mucosa, is further evidence that the inflammatory response described in the middle ears of our patients(I-IV) may represent a Th-2 mediated disease.

In addition to the cytotoxic effects of ECP itself, these mediators are always accompanied by the other proteases, lysozomasol enzymes and oxidizers that are released simultaneously from their respective cells. This may be why the atopic patient continues to produce additional fluid, as compared to the nonatopic patient. This was seen in allergic mice which produced twice the amount of MEE as non-allergic mice on antigen challenge.[92] ECP, known to decrease ciliary function, also impairs ET clearance. The very high concentrations of mediators released by eosinophils and mast cells may also account for the great destruction, osteitis and granulation tissue described on histologic examination of temporal bones from patients with chronic OM.[102] The destructive potential of these mediators in effusion is often overlooked, as is hearing loss, yet serves as a further justification for the removal of this fluid at the time of myringotomy and for prevention of its reoccurrence by appropriate medical management e.g. allergy management.

Purulence in the middle ear has previously been shown to elevate both eosinophil and neutrophil mediators.[72],[68] In Studies I and II gross pus had a similar elevating effect on ECP and MPO(Fig. 2,4) but did not appear to have an effect on mast cell activity(Fig. 3,6). Our findings are in concordance with studies of the histopathology of both serous OM and OME which have demonstrated that eosinophils and neutrophils are integral components in these infiltrates.[82],[116] One of the unique features of the middle ear response demonstrated in Study I was the involvement of neutrophils. Most atopic patients had increased levels of MPO in addition to ECP in their ear effusion, even though there was no evidence of acute inflammation. Paired samples(I) confirmed that in an acutely infected ear purulent OM was associated with a significant elevation of both mediators, but with a disproportionately greater elevation of MPO(Fig. 4), in the infected side as compared to the nonpurulent ear.[72] Yet even in the non-purulent ear the levels of MPO were much higher in atopics than non-atopics. The late asthmatic response in animal models is dependent on neutrophil availability and neutrophil chemotactic factor. This factor has been demonstrated in both atopic and nonatopic subjects.[31] Neutrophils from atopics produce more superoxide in the absence of stimuli than neutrophils from nonatopic patients. The total capacity of cells from atopics for oxidative metabolic activity apparently is not abnormal, but perhaps reflects a greater responsiveness to low levels of stimulation. Styrt found that “the neutrophils from atopics may be both easier to stimulate and more difficult to suppress than cells from normals.”[153] It was therefore expected that neutrophils would be present in the purulent OME,[85] but we were surprised to find that neutrophils were so active in non-acute OME. Neutrophils however, are reported in cutaneous IgE-late-phase reactions in the nose and skin.[39],[114]. In Study I, purulence in the middle ear was shown to elevate both eosinophil and neutrophil mediators, although MPO rose twice as high in purulent ears as in OME ears. One explanation for the increased MPO we report in non-purulent MEE is that an increase in neutrophils may occur as a result of weak stimulation of these cells as bacteria are being cleared from the site of inflammation.[61] Recent studies have indicated that bacterial mRNA present in otherwise “sterile” MEE may serve as a stimulus to T cell activation.[159] Endotoxins have been demonstrated in 52 to 87% of effusions.[107],[37] Atopy may thus contribute to the elevated levels of MPO (Study I) by causing the atopic child to respond differently to the products of acute inflammation due to its primed inflammatory cells.

It is our contention that the middle ear thus behaves like the rest of the respiratory tract and that what has been learned about the atopic response in the sinuses and lungs may be applied to the study of the middle ear to help in understanding OME. For the middle ear to emulate the respiratory tract and as such be capable of an allergic response it is essential that the cells and mediators necessary for the recognition of antigen, production of specific IgE and eventual release of inflammatory mediators and cytokines associated with a Th-2 response are indeed present in the middle ear effusion and mucosa of patients with OME.

Determining whether or not inflammatory mediators in the middle ear are produced locally is important in supporting or rejecting the hypothesis that the middle ear might be involved in an allergic Th-2 response. Just as our earlier studies showed no relation between levels of serum and effusion ECP[72] or specific IgE,[70] the concept that active immunologic processes may be a localized phenomena in the middle ear is supported by these studies which demonstrate local release of ECP(I,III) and tryptase(II) in middle ear mucosa, as well as CD3+ cells and IL-5 mRNA expression.(IV) A specific immune response in the middle ear has been established in animal models[94],[103],[92] as well as in humans. Lymphocytes necessary for antibody production have been shown to be recruited nonspecifically into the mucosa in otitis media. IL-5 is increased during the late-phase reaction of chronic middle ear disease.[18] T lymphocytes are common in serous or mucoid effusions.[115] The finding of CD3+ T cells has been thought to be a marker which can be used to distinguish between allergic and non-allergic sinusitis.[62] IL-5 is produced predominately by stimulated Th-2 and not Th-1 cells.[46] Having CD3+ cells and more importantly, cells actually manufacturing IL-5 mRNA in our biopsies(IV), provide strong evidence that the middle ear is actively participating in a TH-2 response. Similarily, in Studies I,II tryptase was not present in the effusion of any non-atopic patient, nor was it present in all atopics, suggesting that elevated effusion tryptase reflects a local activation of mast cells, not a general systematic response to being atopic.(II) Eight patients with paired samples had different values of tryptase from their opposite ears ( i.e. patient AL: 16µg/l left, 1.5 µg/l right) which also suggested a local response. Among the 13 patients (16 ears) with apparent undetectable levels of tryptase, 56% had evidence of atopy as signified by increased effusion ECP and positive serum ELISA. Thus not all atopics always show mast cell mediator release, again evidence for local, Th-2 allergic-like response in the middle ear.

Future Studies:
Patients may be atopic and by definition capable of making IgE antibodies to common allergens in their environment, although not necessarily symptomatic. Only if those antibodies are associated with disease of a specific target organ is the patient allergic.[119] Just as it is difficult to know with certainty which positive antigens are responsible for the actual disease in allergic rhinitis or asthma without direct challenge tests, the same is true of otitis. Clinical proof of this mechanism will require either provocative challenge studies or a large randomized therapeutic trial. Determination of whether local elevated IgE titers and the presence of mast cells and their products are directly due to an atopic reaction will also require advanced research involving direct challenge studies or immunohistostaining of middle ear mucosal biopsies. The development and use of newer technologies to improve on the sensitivity of in vitro testing so as to more accurately identify which patients are atopic will improve our ability to know in which OME patients allergy may be contributing to the disease.

The hypothesis that viral and/or bacterial infections may induce OME is not contradictory to an allergy hypothesis. Perhaps only in an atopic individual does an initial viral or bacterial infection lead to the chronic recruitment of inflammatory cells with the resultant release of their potent inflammatory mediators, attractants, cytokines, and other factors that results in the production and maintenance of effusion in a sterile ear. Given the economy of nature, it is difficult to explain the presence of plasma cells,[121] IL-5 +mRNA cells,(IV) CD3+ T cells,(IV) eosinophils,(I,II,III) mast cells,(II) specific IgE,[70] tryptase(I,II) and ECP(I,II) in any middle ear, let alone 80% of those tested, if a Th-2 immune-mediated “allergy-like” response were not involved.

The fact that 93.5% of our 77 consecutive children with persistent effusion were atopic(I) suggests that whatever the initiating event responsible for otitis might be, it results in more chronic disease in atopics and constitutes an important justification for the need to develop a new treatment approach for this disease. As an atopic child is known to be more likely to develop asthmatic bronchitis following a viral infection, so too an atopic with otitis may be more prone to develop chronic effusion. Certainly many children in day care get upper respiratory infections and acute otitis media. Irander suggests that it is predominantly those who are atopic who progress to chronic OME.[74] The observation that the inflammatory pattern in the middle ears of atopics differs markedly from that seen in non-atopics(Figs. 3,4,6,) deserves further investigation. The paucity of traditional randomized clinical trials dealing with OME, especially as it might relate to allergy, clearly indicates the need to devise alternate methods to study this disease. It may be that current methods used to control chronic OME, otorrhea, and draining mastoid cavities will continue to fail as long as the effects of allergy on middle ear mucosa are ignored. There is a need to develop outcome research instruments and methods to validate the efficacy of various treatments, including allergy management.

Conclusions
Eighty-five percent (85%) of 131 OME patients randomly selected for these four studies were atopic by in vitro testing. The 20 clinical studies cited in Table I demonstrate that there is a relationship of allergy to OME. Twelve of these studies present evidence that allergy management of chronic OME, when applied to patients proven by RAST, ELISA and/or IDT to be atopic, has a superior outcome to M&T or antibiotic/decongestant therapy alone; with effectiveness ranging from 82% to 100%. These empirical studies challenge scientists to investigate whether there is more direct evidence of allergic Th-2 like activity in the middle ear which might explain such results. Elevated effusion ECP(Fig. 2) and tryptase (Fig. 3,6) values in patients with non-purulent OME, coupled with the presence of ECP(I,II), degranulated eosinophils(I,III), tryptase(I,II), mast cells(II), CD3+ T cells(IV), and mRNA + cells for IL-5(IV) in the middle ears of atopic individuals but not nonatopic subjects supports our stated hypothesis that 1) there is a relation of atopy to OME, and that 2) the middle ear epithelium of atopics has all the components required to behave in a manner similar to that of the rest of the upper respiratory system and is itself capable of an allergic immune response.

These studies suggest on a cellular level that the mediators measured in effusion arise from cells identified in the local tissue lining the middle ear cleft. They refute the null hypothesis as they confirm that many of the participants essential to a Th-2 driven immune response are indeed present in the majority of ears with chronic effusion. They suggest that the inflammation within the middle ear associated with the maintenance of many cases of OME is “allergy like”. When viewed in the context of the Th-2 paradigm for allergic inflammation, the results are compelling. The Th-2 model purports that an increased expression of Th-2 cytokines (e.g. IL-4 and IL-5) is associated with T cell and eosinophil infiltration and IgE production (i.e. Type I hypersensitivity). The demonstrated expression of IL-5 mRNA in middle ear mucosa is support of an allergic etiology in refractory OME.

Two findings in this thesis may be keys to the understanding of the development of OME. First is the fact that atopics have elevated levels of inflammatory mediators in contrast to nonatopics and the other is the fact that neutrophil mediators in addition to the “Th-2-mediators” are raised in atopics only.(Fig. 4) The first finding tells us that there is some unique quality associated with being atopic. This quality is most likely related to a patient’s allergen sensitivity and its being exposed to that allergen. Indeed, Labadie[92] showed in animal studies that allergen-sensitized and challenged animals responded by producing more middle ear fluids to a secondary stimulus than non-sensitized and challenged animals. Labadie also showed that allergen challenge per se did not induce any fluid production, which is in agreement with others. The unique reactivity of atopic ears has also been documented in man with the demonstration of higher fluid levels of the adhesion molecule VCAM-1 than found in nonatopics.[111]

Our second finding regarding elevated MPO in effusion tells us that the response in the atopic ears is not restricted to those cells and mediators that are involved in the classical allergic reaction in which neutrophils are usually non-participants. Thus, the findings of MPO as a sign of neutrophil involvement suggest that the reaction accountable for middle ear effusion may be a more general inflammatory response to some inciting agent, but that this response is so much more prominent in the primed atopic ear. The relationship between allergy and OME, therefore, may be indirect and the allergic reaction one of the mechanisms that may prime the middle ear to a vigorous and prolonged response to microbial products.

Identification of factors responsible for the chronicity of otitis media is an essential step in understanding the disease. We present effusion and mucosal biopsy studies demonstrating levels of ECP, tryptase, and/or IL-5 mRNA cells, CD3+ T cells, eosinophils, and mast cells which indicate that many of the mediators and cells essential to the production of a Th-2 immune mediated response are present in ears with chronic effusion. The increased levels of MPO in atopic patients further suggest that the general inflammatory response to putative inciting agents such as bacterial and viral products may be altered in atopy. These studies support the hypothesis that the exaggerated inflammation within the middle ear associated with most cases of OME is possibly the result of an atopic response within the middle ear itself.

Remembering that children with hearing loss due to OME constitute potentially the largest group in the world with a reversible learning disorder, the significance of these studies as regards the treatment and ultimate prevention of this chronic disease cannot be understated. These findings should be relevant to a physician’s diagnostic and therapeutic approach to patients with chronic MEE. Physicians must become disengaged from the discussions of “What to do?” regarding antibiotics, adenoids, tubes, etc. long enough to pursue the “Why?”. Having evidence that otitis media with effusion may be a sign of allergy (although not pathognomonic) and having evidence that the middle ear mucosa, like a “5th sinus,” appears to be capable of responding in a manner similar to the rest of the respiratory tract and participate in a Th-2 like allergic response, the standard evaluation for a patient demonstrating persistent effusion with or without purulence warrants including a full evaluation for food and inhalant allergies. To be efficacious, antibiotics should be used for infection; antihistamines, steroids, and immunotherapy should be reserved for allergy.

Summary
I. Most striking in this random prospective study of 97 individuals is the very different response seen among groups of atopic vs non atopic patients(p<0.001) by all 3 inflammatory cells in what clinically appears to be identical disease. Histologically, OME in the 2 groups involves 2 very distinct disease processes. These findings provide evidence that eosinophils and mast cells, both essential to a Th-2 driven immune response, are active in the majority of ears from atopics with chronic OME.

II. This study was the first to our knowledge to provide confirmation on a cellular level that the mast cell mediator tryptase measured in effusion of atopic patients arises from actively degranulating mast cells identified in the local tissue lining the middle ear cleft. Mast cells and their mediator tryptase, both essential to the production of a Th-2 immune response, are present in ears with chronic effusion. This study provides further evidence in support of the hypothesis that middle ear mucosa itself is similar to that of the rest of the upper respiratory tract and is capable of an allergic response.

III. This study demonstrated that plastic embedding techniques provide superior structural preservation compared to the cryosections used in previous studies. Because thinner sections can be cut from plastic embedded material, this would make it preferable to paraffin embedding. The different cell types could be identified without problems and cell counts could be easily made. This study also shows that by the use of specific antibodies it is possible to quantitatively determine the frequency with which eosinophils and neutrophils occur in tissue and mucus in the middle ear. Such a quantitative determination is necessary to allow conclusions about the possible role of these inflammatory cells in pathologic conditions such as OME. The results suggest that eosinophils and neutrophils occur with a higher frequency in the middle ear of atopic children with OME as compared to non-OME controls. The present study shows a covariation between the occurrence of eosinophils and neutrophils in OME and supports our earlier findings.[72],[69]

IV. This study demonstrated for the first time the expression of IL-5 mRNA in the middle ear of patients with refractory OME. It is of note that increased expression of IL-5 mRNA cells, CD3+ cells and eosinophils (MBP) were present in the majority of OME ears and furthermore, that all children with these cell markers were atopic. This data suggests that IL-5 and other Th-2 type mediators may perpetuate the inflammatory response and contribute to the inability of these patients to clear their effusions. The presence in OME patients’ mucosa of activated eosinophils and IL-5 lends further support to the hypothesis that a TH-2 inflammation takes place in OME.



Our results indicate that atopy may be a major determinant for the development of OME.