Epidemiology of otitis media with effusion and allergic rhinitis

Otitis media with effusion is a common childhood illness that often responds well to medical therapy. It is the most common cause of acquired hearing loss with a prevalence of 3 to 8 percent in children . A relationship between COME and allergic rhinitis was initially noted in a subset of OME patients who were resistant to conventional medical therapy. Many of these patients suffered from allergic rhinitis. Numerous epidemiologic studies have identified allergy as a risk factor for COME, citing a high frequency of allergic rhinitis in this patient population. One such study by Alles and colleagues reported an 89% prevalence of allergic rhinitis in patients with COME. This is significantly higher than the reported prevalence of allergic rhinitis in the general population, which ranges from 10 to 30 percent in adults and 20 to 40 percent in children . In this study, patients between 3 and 8 years of age from the “Glue Ear” clinic were surveyed for nasal symptoms. Allergic rhinitis was defined as the presence of 2 allergy nasal symptoms such as sneezing, nasal itch or nasal crease or one symptom with either a positive skin test or presence of nasal eosinophilia. Given the exceptionally high reported prevalence rate of 89% noted in this patient population, there was a concern than this high number may reflect a referral bias.

To address the issue of referral bias, the same group evaluated the presence of otologic, rhinologic and asthma symptoms via a community-based questionnaire of parents with children aged 5 to 6.5 years old. Consistent with reported prevalence of AR in children, they found 5.1% of children polled had symptoms strongly suggestive of AR and an additional 31.5% with at least one nasal symptom consistent with AR . Otologic symptoms were reported in 32.8% of the children. A cross-tabulation comparison of otologic and nasal symptoms revealed a significant association between allergic rhinitis and otitis media (p = 0.00000) . Other epidemiologic studies have reported significantly lower prevalence rates of AR for patients with COME than 89%. In a study from Brazil, 51 patients with ages ranging from 3 to 55 years old with COM undergoing an otologic surgical treatment were evaluated for the presence of allergic or non-allergic rhinitis with eosinophils syndrome (NARES). Allergic rhinitis and NARES was diagnosed in 33.3% and 15.7%, respectively, in these patients with chronic otitis media. They concluded an association between nasal atopy and COM . Similar to the previous study, referral bias may be a limitation. In addition, the lack of a local healthy control group for comparison is a significant shortcoming.

Caffarelli and colleagues had a similar methodology to the study published by Alles and colleagues with the exception of including a similar aged healthy control group. In this study, children aged 6 to 14 years of age with documented hearing loss from OME for at least 5 months were assessed for allergic rhinitis symptoms, were given a physical exam to document findings consistent with allergic rhinitis and underwent skin testing against multiple common regional allergens. A healthy control group for comparison was obtained from similar-aged children undergoing an annual school health exam. Only 16.3% of children with COME had allergic rhinitis as defined by the presence of nasal symptoms not related to an infection. When compared to the similar aged healthy control from the same region, there was a significant difference between patients with COME and healthy controls, 16.3% as compared to 5.5% .

Despite the discrepancy in the reported percentages of OME associated with AR, many studies support an association between OME and AR . Potential causes of the wide range of prevalence rates may stem from the commonality of the symptoms, differences in the diagnostic criteria for both OME and AR among studies, and the variation in the ages and location of the groups. More recent studies have addressed this issue by more specifically defining the symptoms, applying strict criteria for the diagnosis of OME and AR, limiting the age groups and including a regional healthy control group for comparison. Including only those studies with a matched control, the prevalence range of AR with OME tightens to 16 to 25% .

Different forms of chronic otitis media with effusion

Broadly defined, COME clearly represents a spectrum of inflammatory diseases sharing the commonality of fluid accumulation within the middle ear behind an intact tympanic membrane. In a recent study, Rezes and colleagues subtyped COME as either a transudative versus an exudative process based on the albumin-to-immunoglobulin ratio of the middle ear effusion (MEE). They then compared the profile of inflammatory mediators between these two forms of COME. Interestingly, there was a significant difference in the cytokine profile between the transudative and exudative forms. Specifically, interferon-gamma and tumor necrosis factor-alpha characterized the exudative effusion, whereas interleukin (IL)-4 and IL-10 were more predominate in the transudative effusion.

Subtyping the MEE as an exudate versus a transudate suggests different pathophysiology between these two forms. The extravasation of albumin from peripheral blood, where it is normally present, into the middle ear cleft, as characterized in a transudative process, is a consequence of the local microvasculature permeability. A number of vasoactive inflammatory mediators are capable of sustaining this type of transudate, including histamine, immune complexes, and prostaglandins. Histamine, found in mast cells, is one of several characteristic mediators of an IgE-mediated Type 1 hypersensitivity. Although not measured in this study, histamine is present in the effusion of patients with COME and at levels higher than in serum . This localization of histamine in the middle ear cleft and the unique cytokine profile associated with allergies suggest the role of an allergic response triggering and sustaining local inflammation that is responsible for the accumulation of middle ear transudate.

Different forms of chronic otitis media with effusion

Broadly defined, COME clearly represents a spectrum of inflammatory diseases sharing the commonality of fluid accumulation within the middle ear behind an intact tympanic membrane. In a recent study, Rezes and colleagues subtyped COME as either a transudative versus an exudative process based on the albumin-to-immunoglobulin ratio of the middle ear effusion (MEE). They then compared the profile of inflammatory mediators between these two forms of COME. Interestingly, there was a significant difference in the cytokine profile between the transudative and exudative forms. Specifically, interferon-gamma and tumor necrosis factor-alpha characterized the exudative effusion, whereas interleukin (IL)-4 and IL-10 were more predominate in the transudative effusion.

Subtyping the MEE as an exudate versus a transudate suggests different pathophysiology between these two forms. The extravasation of albumin from peripheral blood, where it is normally present, into the middle ear cleft, as characterized in a transudative process, is a consequence of the local microvasculature permeability. A number of vasoactive inflammatory mediators are capable of sustaining this type of transudate, including histamine, immune complexes, and prostaglandins. Histamine, found in mast cells, is one of several characteristic mediators of an IgE-mediated Type 1 hypersensitivity. Although not measured in this study, histamine is present in the effusion of patients with COME and at levels higher than in serum . This localization of histamine in the middle ear cleft and the unique cytokine profile associated with allergies suggest the role of an allergic response triggering and sustaining local inflammation that is responsible for the accumulation of middle ear transudate.

Eustachian tube function in allergic subjects

In a series of studies performed at the University of Pittsburgh in the 1980s, Eustachian tube obstruction was shown to develop in response to nasally administered cold, dry air, histamine, Dermatophagoides farinae , and house dust. Subjects with AR showed much higher rates of Eustachian tube obstruction compared with nonallergic patients on a nine-step test of Eustachian tube function in response to these challenges. Moreover, subjects pretreated with antihistamines/decongestants showed significantly less functional nasal airway obstruction when compared with placebo-treated subjects. Eustachian tube function was also improved by antihistamine/decongestant pretreatment as pretreated subjects required higher doses of antigen to produce similar levels of Eustachian tube obstruction. These authors note that the middle ear mucosa is just an extension of the upper airway mucosa identified in the nose and nasopharynx and that the mucosa of the Eustachian tube structurally resembles that of the bronchus .

Cellular milieu of middle ear effusion from chronic otitis media with effusion supports an allergic response

Antigen-activated T-helper lymphocytes, cells central to the immune response, differentiate into one of two major subtypes of effector cells, T helper type 1 (Th1) or T helper type 2 (Th2) cells. These subtypes release unique cytokine profiles which orchestrate different immune responses. The Th1 cells secrete interferon gamma (IFN-), tumor necrosis factor-beta (TNF-), and macrophage activating factor, whereas Th2 cells secrete a different profile of cytokines including IL-4, IL-5, IL-10 and IL-13. Th1 cells incite a strong cell-mediated immune response, targeting intracellular pathogens. Th2 cells, thru the production of Th2 interleukins, elicit a humoral immune response stimulating B-cell antibody production. IL-4 stimulates primarily immunoglobulin E (IgE). In addition, Th2 cells are essential in the defense against parasitic infections. IL-5 activates the differentiation and maturation of eosinophils, leukocytes central in this immune response.

Many foreign antigens incite both Th1 and a Th2 cells in order to activate both cell-mediated and humoral immunity. However, some antigens and disease processes skew the activation of T-helper lymphocytes to either Th1 or Th2. One such condition, atopy represents an abnormal activation of Th2 cells in response to allergens. To sort the etiology of the immune response resulting in chronic otitis media with effusion, analysis of the cytokine profile of this process may differentiate whether COME represents a cell-mediated versus a humoral immunity consistent with an allergic response.

By definition, an allergic response is a Type I hypersensitivity reaction that is IgE-mediated and driven by a Th2 cytokine response. IgE production is stimulated on initial exposure to an antigen in an atopic patient. IgE attaches to both mast cells and basophils. On subsequent exposure, the cross-linking of antigens to their respective IgE initiates the release of bioactive molecules, which results in a two-phase early and late response. The early-phase response occurs within minutes of antigen exposure and is characterized by histamine release. The late-phase response, which appears 4 to 6 hours after antigen exposure, is characterized by the production and stimulation of eosinophils, monocytes, and multiple inflammatory mediators, including IL-4, -5, -13, eosinophilic cationic protein (ECP), and myeloperoxidase.

In the course of years of clinical practice, Hurst and Venge noted the similarity of COME to rhinosinusitis and asthma. He reasoned that, therefore, similar investigative techniques to those used to study asthma and rhinitis should be applied to COME. He strongly suspected that the most resistant cases of COME were most likely to be of allergic etiology. Especially older children (5 years or older) whose Eustachian tubes had matured were believed most likely to have an atopic etiology.

Hurst noted that, by definition, allergic reactions involve Type I, IgE-mediated hypersensitivity in which activated mast cells and eosinophils precipitate a Th2 inflammatory reaction. Therefore, he investigated the participation of mast cells and eosinophils in the development of MEE in atopic children. Hurst and Venge was able to demonstrate that ECP, a marker for eosinophil activity, was statistically significantly higher in atopic children with COME than in nonatopic children and that the level of ECP was higher in the MEE than in serum. Similarly, myeloperoxidase from mast cells was shown to be statistically significantly higher in the effusion from the ears of 68 children with objectively documented allergy than in nonatopic children. Wright and colleagues demonstrated that IL-5 mRNA expression is also higher in atopic children with COME than in control children; this despite that serum IgE levels were not notably higher in atopic versus nonatopic children. From this series of studies demonstrating a pattern of Th2 inflammatory mediators in atopic children with COME, Hurst concluded that there was a very different response between atopic and nonatopic children with COME . It appears that OME can result from at least two different processes.

The above studies were then validated by Sobol and Nguyen at McGill University. They confirmed that neutrophils were more common in MEEs from nonatopic children as compared with atopic patients. Atopic children with COME (as verified on skin tests) had a higher T lymphocyte count and a higher percentage of eosinophils. Furthermore, the levels of IL-4 and IL-5 were higher in the atopic group . IL-4 induces local immunoglobulin switching to IgE production, and IL-5 is the main activator of eosinophils. Both cytokines play central roles in orchestrating an allergic response. Nguyen extended these findings by comparing the cellular and cytokine profiles of MEEs with the mucosa of the torus tubarius and adenoid tissues using immunohistochemical techniques and in situ hybridization. In atopic patients, a Th2 type of inflammatory response dominated both the MEE and the mucosa of the nasopharynx and adenoid .

Nagamine and colleagues at Teikyo University confirmed the subtype of COME associated with atopic children by describing an eosinophilic otitis media (EOM) found in adults with bronchial asthma. Characterized by the presence of a viscous MEE containing eosinophils and elevated systemic IgE levels, this entity was notably resistant to conventional treatment of COME. Similar to studies by Hurst, the MEE from patients with EOM also had elevated levels of ECP, confirming the local activation of the eosinophils .

The importance of IL-4 and IL-5 in the pathophysiology of this form of COME has been illustrated in presensitized rats. Rats, presensitized to ovalbumin (OVA), were exposed to 2 transtympanic injections of OVA with and without an antagonist to either IL-4 or IL-5. As expected, transtympanic exposure to OVA in presensitized rats resulted in an MEE. The production of MEE was prevented in seven presensitized rats pretreated with an IL-4 antagonist. In addition, histologic evaluation of the middle ear mucosa demonstrated no inflammatory changes. Similarly, four of the seven presensitized rats pretreated with an IL-5 antibody did not develop an effusion, although five of the seven had histologic inflammatory changes . Despite the inflammatory changes in this group, these rats pretreated with IL-5 antibody showed no eosinophilia within the middle ear mucosa. In this animal model for atopic otitis media, blocking the effects of IL-4 and IL-5 before antigen exposure in the middle ear aborted the production of MEE, supporting the role of an allergic Th2 inflammatory response in the pathophysiology of COME.