Portal de Periódicos da CAPES

Role of viruses in the pathogenesis of acute otitis media

2000; Lippincott Williams & Wilkins; Volume: 19; Issue: Supplement Linguagem: Inglês

10.1097/00006454-200005001-00004

1532-0987

Terho Heikkinen,

Respiratory and Cough-Related Research

Although acute otitis media (AOM) is generally considered a bacterial infection and therefore usually treated with antibiotics, pathogenic bacteria can be isolated from the middle ear fluid (MEF) in ∼70% of cases of AOM. 1 The fact that about one-third of AOM cases remain without a proven bacterial etiology, together with vast clinical experience connecting AOM with viral upper respiratory tract infections (URIs), has prompted investigators to search for the role of viruses in the pathogenesis of this condition. At present there is convincing evidence to support the crucial role of respiratory viruses in the development of AOM. EPIDEMIOLOGIC STUDIES The occurrence of AOM shows an extensive seasonal variation. The incidence rates of AOM are highest during the winter months and lowest in the summertime, which parallels the incidence of viral URI. 2, 3 In addition the incidence of AOM is highest among children 4500 cases of AOM. The monthly occurrence of AOM correlated significantly with the recovery of respiratory viruses in the nasopharyngeal specimens of the children. The rates of detection of respiratory viruses in nasopharyngeal specimens from children with AOM have usually ranged between 30 and 50%. 12–18 Until recently detection of respiratory viruses has been based on viral culture and/or antigen detection methods, and it is obvious that the limitations of these techniques have resulted in underdetection of viruses in nasopharyngeal specimens. In a study of 363 children with newly diagnosed AOM, respiratory viruses could be found by culture or rapid antigen detection in nasopharyngeal aspirates of 42% of the children, although clinical symptoms indicative of a viral infection were present in 94% of the patients. 17 The development of PCR-based assays for a wide range of respiratory viruses may substantially increase the rates of viral detection in nasopharyngeal specimens of children. 19–21 By using PCR for detection of rhinovirus, RSV and coronavirus, Pitkäranta et al. 21 found at least one of these viruses in nasopharyngeal aspirates from 62% of 92 children with AOM. Recently we searched for viruses in nasopharyngeal aspirates with rapid antigen detection for RSV, influenza viruses, parainfluenza viruses and adenovirus and PCR for rhinovirus, enterovirus and coronavirus. Viral infection was documented in the nasopharyngeal specimens of 90% of 67 children with AOM (A Ruohola, unpublished data, September, 1999). Different types of respiratory viruses may vary in their ability to predispose to AOM. Several studies have indicated that RSV may be the principal virus leading to the development of AOM as a complication. 3, 11, 22 Other viruses that frequently cause AOM include influenza A virus, adenovirus and parainfluenza type 3 virus. 3, 11 However, the relative ability of different viruses to predispose to AOM is difficult to determine because the development of AOM is dependent on multiple, potentially confounding factors, such as the age of the child. EUSTACHIAN TUBE DYSFUNCTION Dysfunction of the Eustachian tube (ET) is considered the most important factor in the pathogenesis of AOM. 23 The results of experimental animal studies as well as human studies provide strong evidence supporting a causal role for respiratory viruses in the disruption of normal ET function. Animal studies Most experimental studies on the role of viruses in the pathogenesis of OM have been performed in chinchillas. In one of the first of these studies by Giebink et al., 24 OM developed in 67% of chinchillas inoculated intranasally with both influenza A virus and S. pneumoniae, whereas only 4% of chinchillas inoculated with influenza virus alone and 21% of animals inoculated with S. pneumoniae alone developed OM. A further histopathologic study clearly demonstrated epithelial damage in the ET, together with accumulation of cellular debris in the tubal lumen, in association with the development of negative middle ear pressure during influenza A virus infection in chinchillas. 25 Suzuki and Bakaletz 26 inoculated chinchillas intranasally with adenovirus type 1 and nontypable H. influenzae. All animals inoculated with both pathogens developed OM of greater severity than animals receiving either pathogen alone. The viral-bacterial interaction may, however, vary among different types of viruses and bacteria, because a subsequent experiment in which H. influenzae was replaced with Moraxella catarrhalis did not demonstrate a similar viral-bacterial synergy. 27 Impairment of the ET has been demonstrated in several studies in various animal models. 28–32 In these studies impairment of the transport function and the ciliary activity of the ET was observed after inoculation of influenza A virus or adenovirus, either intranasally or directly into the middle ear. Direct inoculation of influenza A virus into the middle ear of chinchillas resulted in a substantial damage of ciliated epithelial cells in the ET, and the restoration of the mucosal lining of the ET was not seen until 4 weeks after viral inoculation. 32 Human studies Several studies have demonstrated the development of ET dysfunction in children during URI. 33–35 Sanyal et al. 34 did a prospective tympanometric study of preschool children to determine the effect of URI on ET function. Although significant negative middle ear pressure was present in 13% of tympanograms in asymptomatic children, the rate of abnormal tympanograms increased to 75% during URIs. Eighty-one percent of the abnormalities were already detectable by the second day of illness, and 97% were detectable by the fourth day of illness. Although AOM is an infrequent complication of URI in adults, experimental viral challenge studies in adult volunteers have provided important data on the development of ET dysfunction during URI. After intranasal challenge with rhinoviruses, deterioration of normal ET function was observed in 50 to 80%, and significant negative middle ear pressure developed in 30 to 50% of the volunteers. 36–38 The effect of influenza A virus on induction of ET dysfunction may even exceed that of rhinoviruses. In adult volunteers challenged with influenza A virus, >80% of the infected subjects developed ET dysfunction, and high middle ear underpressures were detected in 60 to 80% of the persons by Day 4 to 5 after viral challenge. 39–40 Further one of the subjects developed AOM, and both influenza A virus and S. pneumoniae were detected in the MEF by PCR analysis. 40 In addition to experimental studies, the development of significant middle ear underpressures has also been demonstrated during natural rhinovirus infections in adults. 41 Recent studies have disclosed potential mechanisms by which respiratory viruses may cause ET dysfunction. There is increasing evidence that respiratory viruses are capable of inducing a release of various inflammatory mediators from target cells in the nasopharynx. Numerous in vitro studies have demonstrated the increased production of cytokines by human epithelial cells infected by respiratory viruses. 42–45 Several clinical studies have documented the presence of increased levels of inflammatory mediators [e.g. histamine, bradykinin, interleukin (IL)-1, IL-6, IL-8, leukotriene C4 and tumor necrosis factor] in human nasopharyngeal secretions in response to a viral infection. 45–50 After intranasal challenge many of these substances have been shown to provoke ET dysfunction in adult volunteers. 51 ALTERATION OF THE HOST'S IMMUNE DEFENSE Several studies indicate that some respiratory viruses have an immunosuppressive effect on polymorphonuclear leukocyte function; by this mechanism they may increase the host's susceptibility to secondary bacterial infections. 52 In a study in chinchillas intranasal inoculation with influenza A virus induced polymorphonuclear leukocyte dysfunction that was not observed after inoculation with S. pneumoniae alone. 53 The greatest incidence of pneumococcal OM occurred when the bacteria were inoculated 4 days after the virus, just before the time of influenza-induced polymorphonuclear leukocyte dysfunction. Suppression of neutrophil function after influenza A virus infection has also been shown in humans. 54–56 Some studies have also demonstrated an alteration of cell-mediated immune function during RSV and rhinovirus infections. 57–59 EFFECT ON BACTERIAL COLONIZATION AND ADHERENCE Viral infection of the upper respiratory tract may have a substantial impact on the nasopharyngeal bacterial flora. A study in cotton rats by Patel et al. 60 showed that colonization with H. influenzae increased significantly within 4 days of RSV infection. In an experimental study in adult volunteers, influenza A virus infection promoted colonization of the oropharynx with S. pneumoniae. 61 Indirect evidence of the effect of viral URI on the nasopharyngeal bacterial flora was provided by Faden et al., 62 who reported that both the occurrence and the quantity of the most common middle ear pathogens were increased in children during AOM. Some studies have suggested that respiratory viruses may also increase the adherence of bacteria to human epithelial cells. Håkansson et al. 63 demonstrated that certain types of adenovirus significantly increased the binding of S. pneumoniae to human A549 cells. Similarly, increased adherence of S. pneumoniae and H. influenzae to pharyngeal cells during experimental influenza A virus infection in adult volunteers was reported by Fainstein et al. 64 Studies in different species of animals have provided controversial results, but at least one study in ferrets inoculated with influenza A virus showed a significant increase in bacterial adherence to the respiratory mucosa. 65 PRESENCE OF VIRUSES IN MEF Further evidence for the role of respiratory viruses in the pathogenesis of AOM has been obtained by searching for them in MEF directly. Because in the early studies during the 1950s and 1960s the detection of viruses was based on viral culture alone, the overall detection rate of viruses was ∼5%. 66, 67 Since the 1980s improved viral culture techniques and the development of viral antigen detection methods have allowed demonstration of viruses or viral antigens in ∼20% of children with AOM (Table 1). 12–16, 18, 68–70 The most common viruses identified in MEFs are RSV, influenza viruses, parainfluenza viruses, rhinovirus and adenovirus; these same viruses have been found to be closely associated with AOM in epidemiologic studies. In about two-thirds of cases, when virus has been found in MEF, bacteria have also been isolated, indicating a mixed infection. With the use of viral culture and antigen detection methods, virus as the only pathogen in MEF has been detected in 6% of AOM cases. Table 1: Detection of viruses in middle ear fluid by culture and/or antigen detection in patients with acute otitis media, 1982 to 1996The development of PCR technique has further increased the rates of viral detection in MEF. Okamoto et al. 71 studied 44 children with AOM and reported that RSV genomic sequences were detected in 53% of the MEF by PCR. Further the RSV sequences were detectable at an even higher rate (82%) in the MEF of children in whom RSV could be cultured in nasopharyngeal secretions. Recently Pitkäranta et al. 21 used PCR to determine the frequencies of rhinovirus, RSV and coronavirus in the MEF of 92 children with AOM. Viral RNA was detected in the MEF of 48% of the children. The considerably higher rates of viral detection in MEF by PCR, compared with those by viral culture or antigen detection methods, raise the question of the real significance of the PCR findings. Whether viral nucleic acids detected by PCR represent viruses with a real pathogenetic role in the middle ear remains to be determined. Recent data indicate that there are differences between various respiratory viruses in their ability to invade the middle ear. In our study of 456 children with AOM, RSV was detected in the MEF of 74% of the children infected by this virus. 72 The relative rate of middle ear invasion by RSV was significantly higher than the rates for any of the other viruses searched for. Also parainfluenza viruses and influenza viruses were found to invade the middle ear significantly more often than enteroviruses or adenoviruses. 72 In another study using the PCR technique for viral detection, rhinovirus was found in the MEF of 69% of the children infected with this virus, whereas the corresponding figures for RSV and coronavirus were 65 and 44%, respectively. 21 The different relative prevalences of viruses in the MEF clearly suggest that although some viruses may enter the middle ear passively along with nasal secretions, other viruses may actively invade the middle ear and contribute significantly to the inflammatory process in the middle ear. VIRAL VACCINES IN THE PREVENTION OF AOM Perhaps the strongest evidence for the crucial role of respiratory viruses in the pathogenesis of AOM is provided by direct clinical intervention trials investigating the efficacy of viral vaccination in the prevention of AOM. In a prospective study of 374 children in day care, we administered trivalent, inactivated influenza vaccine to one-half of the children before the expected influenza epidemic. 73 During the influenza A outbreak the incidence of AOM associated with influenza A was decreased by 83% in the vaccinees. Even more importantly there was a 36% reduction in the overall incidence of AOM in the vaccine group compared with the control group. Essentially similar results were subsequently reported by Clements et al., 74 who studied the efficacy of the inactivated influenza vaccine in 186 children in day care. During the influenza epidemic the vaccinated children had 32% fewer episodes of AOM. Recently Belshe et al. 75 investigated the efficacy of a live attenuated, trivalent, intranasal influenza vaccine in 1602 children ages 15 to 71 months. The vaccine efficacy was 93% against culture-confirmed influenza, and the vaccinated children had also 30% fewer episodes of febrile AOM. The results of the influenza vaccine trials indicate that prevention of the preceding viral URI is an effective way to prevent the development of AOM. At present influenza vaccine is the only commercially available vaccine for the control of respiratory virus infections. It is obvious, however, that effective vaccines also against other major viruses predisposing to AOM, particularly against RSV, might have a great effect on the incidence of AOM in children. SUMMARY To date there is ample evidence suggesting a crucial role for respiratory viruses in the pathogenesis of AOM. Respiratory viral infection appears to initiate the cascade of events that finally leads to development of AOM (Fig. 1). The pathogenesis of AOM is complicated, involving a network of factors, some probably not yet identified, which affect each other in a time-dependent manner. Increased knowledge of the detailed mechanisms of viral infection, the host inflammatory response during URI and the interaction between viruses and bacteria could lead to major advances in the prevention of AOM. Fig. 1: The cascade of events in the pathogenesis of acute otitis media during viral upper respiratory tract infections.DISCUSSION Dr. Bluestone: Studies done in volunteers have demonstrated that patients who had preexisting negative ET pressure and a history of OM (vs. those who did not) did worse when challenged with influenza A with regard to their chance of developing MEF. Those who had no history of OM had normal tympanograms and otoscopic findings and less chance of getting middle ear effusion. There is a predisposing factor that was thought to be ET dysfunction. One woman in this study had no pneumococcus in her nasopharynx prior to the virus challenge, but when she developed AOM, the PCR identified the challenged influenza A virus and S. pneumoniae. So this may be an example of acquisition of pneumococcal OM following influenza A infection. Dr. Giebink: In chinchillas influenza studies show a very clear predisposing effect to pneumococcal OM. Baby chinchillas, but not adult animals, are susceptible to adenovirus infection, which predisposes them to nontypable H. influenzae OM. RSV produces a nonreplicating nasopharyngeal infection in chinchillas but does nothing to enhance the incidence of pneumococcal OM when the organism is introduced in the nose. Therefore the same animal species demonstrates age-specific differences in susceptibility to RSV challenge and corresponding differences in susceptibility to either pneumococcal or H. influenzae OM. The entire area of RSV and OM is complex, but there is an opportunity to understand this interaction using animal models. In a mouse model for gene therapy to the inner ear, an adenoviral vector targeted to inner ear tissue in young animals targeted only sensory cells. In the adult they lost that capability. The inference is that the receptor expression on the cell changes with age rather than an immune response to the vector. In primates ciliated epithelial cells are exquisitely sensitive to influenza virus replication. They are not the cells that RSV targets. I am not sure if anyone has data from the inner or middle ear that shows that it is cells other than ciliated epithelium that are infected with RSV. It is easy to understand predisposition to otitis if ciliated epithelium is being destroyed, such as by influenza virus infection. It is another question as to the mechanism of RSV and OM. RSV is probably replicating in other epithelial cells, but not in the ciliated cells. In the chinchilla model RSV was replicating in the nasopharynx, but when the temporal bones were examined there was no evidence of OM and no effect on ET epithelium. In the influenza model there was a profound effect on ET epithelium. This work was supported by the Academy of Finland.