 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* University of Virginia Health System, Charlottesville, Virginia
University of Sao Paulo School of Medicine, Ribeirao Preto, Sao Paulo, Brazil
Hoffman-LaRoche, Nutley, New Jersey
|| Roche Products Ltd, Welwyn, United Kingdom
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods. Children (ages 2–12) with normal tympanograms before onset of illness had bilateral tympanometry daily except weekends for 2 weeks after the onset of a cold. Nasopharyngeal secretion obtained at onset of illness was cultured for bacterial pathogens of otitis media using selective agars and tested for rhinovirus, coronavirus, respiratory syncytial virus, influenza A and B, and parainfluenza 1–3 by reverse transcriptase polymerase chain reaction technology. Tympanometry was designated as abnormal with peak pressure of 
-100 daPa or 
50 daPa and/or a compliance peak of <0.2 cm3.
Results. Eighty-six colds were studied, 82 in schoolchildren (5–12 years old) and 4 in 2- to 3-year-olds. Abnormal negative middle ear pressure occurred at least once during the 2 weeks after onset in 57 (66%) of the 86 colds. Tympanometry was abnormal in the first week after onset in 50 (88%) of the 57 colds and was abnormal on a single day in 17 (30%) of the 57. The middle ear pressure abnormalities were intermittent and shifted from one ear to the other ear from day to day. Reverse transcriptase polymerase chain reaction was positive for a respiratory virus in 56 (65%) of the 86 illnesses. Rhinovirus was found in 48% and respiratory syncytial virus in 14%. Pathogenic bacteria (Streptococcus pneumoniae, Haemophilus influenzae, or Moraxella catarrhalis) were detected in nasopharyngeal secretion in 29 (34%) of the 86 colds; the bacteria were in high titer (103 cfu/mL) in 26 of the 29 positive specimens. None developed illness that required a visit to a physician.
Age, detection of a respiratory virus, and presence of bacterial pathogen in the nasopharyngeal secretion had a negligible effect on the occurrence of abnormal tympanometry. Occurrence of negative middle ear pressure in winter-spring colds was significantly greater than in fall colds for unexplained reasons.
Conclusions. Transient negative middle ear pressure occurred in two thirds of uncomplicated colds in healthy children. This negative pressure, which may facilitate secondary viral or bacterial otitis media, seems to result from viral infection of the nasopharynx and distal tube causing bilateral eustachian tube dysfunction. Tympanometry provides an objective measure of the potential beneficial effects of investigational treatments on the risk of eustachian tube dysfunction/otitis media.
Key Words: otitis media • eustachian tube • viral respiratory infections • tympanometry
Abbreviations: RSV, respiratory syncytial virus • HRV, rhinovirus • HCV, PCR, polymerase chain reaction • RT-PCR, reverse transcriptase polymerase chain reaction • HCV, coronavirus • AOM, acute otitis media
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The pathophysiology of otitis media varies with cause. Bacterial otitis media is usually attributable to Streptococcus pneumoniae, Haemophilus influenzae, or Moraxella catarrhalis, which are common inhabitants of the nasopharynx of healthy children. A viral respiratory infection is thought to be the primary event that precedes secondary bacterial infection of the middle ear by causing impairment of eustachian tube ventilatory function and resulting development of negative middle ear pressure and aspiration of nasopharyngeal contents into the middle ear.2 Although all of the recognized respiratory viruses contribute to the problem, viral otitis media is most commonly caused by respiratory syncytial virus (RSV) or rhinovirus (HRV).3,4 Viruses have been detected in middle ear fluid both alone and in coinfection with bacterial pathogens.4–6 Detection of viruses in middle ear fluid has increased recently with the use of polymerase chain reaction (PCR) for amplification of viral genomes.4,7 Finally, retraction of the tympanic membrane attributable to negative middle ear pressure may produce redness of the eardrum, in the absence of microorganisms. Negative pressure may result from eustachian tube dysfunction attributable to nasopharyngeal viral infection or, in barotitis, from ambient pressure changes unaccompanied by equalization of pressure in the middle ear cavity.
In this study, we examined the effect of viral respiratory tract infections (colds) on middle ear pressure in healthy children. To estimate middle ear pressure, we used tympanometry, which provides an objective and reproducible measure of pressure in the middle ear cavity behind an intact tympanic membrane8 and is easy to use in children. Pneumatic otoscopy can provide information on inflammation of the tympanic membrane and presence of fluid in the middle ear cavity, but it is operator-dependent. Transient negative middle ear pressure has been shown to occur commonly during viral respiratory infections in adults.9–13 Development of negative middle ear pressure because of viral colds in children has been more difficult to study, because young children have a high point prevalence of abnormal tympanograms between cold episodes.14 Sanyal et al15 demonstrated that negative middle ear pressure developed early (first 2 days) during symptomatic colds in 28 children in a research child care center. Recently, Moody et al16 demonstrated a decrease in average middle ear pressure during colds in 20 children followed over a winter season with daily tympanometry performed by parents.
To evaluate the effect of an uncomplicated cold on middle ear pressure in a child, tympanometry should be normal before onset of illness. Consequently, we enrolled predominantly school-aged children with colds, because determination of day of onset of the cold is more precise and abnormal middle ear before onset of illness is less common among this age group than in younger children.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Symptom Diary
The parent (or child with the help of the parent) recorded daily symptoms (feverish, congestion/stopped up nose, runny nose, cough) during illness. The use of any medications (analgesics, cough/cold medication, and/or antibiotics) was also recorded.
Tympanometry
Bilateral tympanometric measurements were done each weekday for 2 weeks after onset of illness by a trained study nurse or otolaryngologist using a handheld microtympanometer (GSI, Grason-Stadler, Inc, Milford, NH). Recordings of abnormal pressure were repeated twice before being accepted. For comparison with other studies, abnormal middle ear pressure recordings were divided into 3 types: C1 (-100 to -199 caPa); C2 (-200 to -399 daPa); and B (no measurable pressure peak or compliance <0.2 cm3).17,18
Nasopharyngeal Aspirate for Microbiology
The aspirate of nasopharyngeal secretions was collected in a Lukens trap (Sherwood Medical, St Louis, MO) by suctioning with a Vacu-Aide portable aspirator (Sunrise Medical, Carlsbad, CA) (-50 to -375 mm Hg). A Yankauer suction device (Allegiance Healthcare Corp, McGaw Park, IL) was placed in each nostril for 5 to 10 seconds. If no secretions were obtained, a few drops of saline were instilled into the contralateral nostril with the head tilted back and aspiration then performed. The aspirate was maintained at 4°C until inoculation onto plated media for bacterial detection within 12 hours. Aliquots for virology were frozen and stored at -70°C. Nasopharyngeal aspirate samples from each child were obtained either at the first visit after onset of illness or at the first 2 visits and pooled for virologic identification.
Bacteriology
Isolation and quantitation of the common nasopharyngeal pathogens (S pneumoniae, H influenzae, and M catarrhalis) was done by a selective agar method.19 A 0.1 mL-nasopharyngeal sample was inoculated onto each selective agar plate and incubated for 48 hours at 35°C. S pneumoniae was isolated on gentamicin agar (tryptone blood agar base with 5% sheep blood and 5 µg/mL gentamicin).20 H influenzae was isolated on GCYSB agar (GC agar base containing chocolated sheep blood, yeast autolysate, vancomycin, bacitracin, and clindamycin.21,22 M catarrhalis was isolated on B-AVTA plates (brucella agar base with 5% sheep blood, amphotericin B, vancomycin, trimethoprim, and acetazolamide.23 The plates for M catarrhalis were incubated in air and the other 2 in a candle extinction jar. The number of colonies of each species on the selective plates was used to determine titer per mL of nasopharyngeal sample.
Isolates with appropriate colonial and Gram stain morphology were identified according to routine methods including X and V factor requirement for growth of H influenzae, optochin sensitivity for S pneumoniae, and oxidase positivity and hydrolysis of indoxyl acetate24 for M catarrhalis.
Urea Nitrogen Assay
The quantitative bacterial culture method provided titer/mL of sample. The bacterial titer/mL of nasal secretion was ascertained by estimating the amount of dilution of nasal secretion in the aspirate by measurement of the urea concentration in the sample using coupled enzyme reaction involving urease and glutamate dehydrogenase (Sigma Diagnostics Kit No 66-UV, Sigma Chemical Co, St Louis, MO). Twenty microliters of aspirate was added to 300 µL blood urea nitrogen (endpoint) reagent at room temperature, and the absorbance after 5 minutes measured at 340 nm on a spectrophotometer. Validity of each run was assessed by inclusion of a diluent blank and a urea standard (Sigma). Because the urea concentration in nasal lining fluid is the same as that in blood,25 the dilution factor in an aspirate was calculated by dividing the urea concentration in mg/dL into the assumed normal blood urea concentration in children of 10 mg/dL. Bacterial titer in nasopharyngeal secretion was determined by multiplying titer in aspirate times the dilution factor. Isolates that had 103 cfu/mL were considered to be high titer.
Virology
RNA for reverse transcriptase-polymerase chain reaction (RT-PCR) testing was extracted from thawed aspirate samples by commercial matrix affinity chromatography (see below). Bovine serum albumin to a final concentration of 0.025% was added to the RNA extracts of all RT-PCR negative samples to block PCR inhibitor,26 and RT-PCR was repeated. Consequently, all samples were either positive for RNA of one of the viruses or were negative in the presence of 0.025% bovine serum albumin in the RT and PCR steps.
HRV and Coronavirus (HCV)
RT-PCR was used for identification of the RNA of HRV and HCV according to previously described methods.4 Briefly, RNA was extracted from 100 µL of aspirate/wash diluted in an equal volume of PBS by matrix affinity chromatography (QIAmp blood kit; Qiagen, Chatsworth, CA). In reverse transcription, the virus-specific oligonucleotide HRV primer 1 was 5'GCACTTCTGTTTCCCC-3'; for HCV 229E, 5'-GGTACTCCTAAGCCTTCTCG-3' and HCV OC43, 5'-AGGAAGGCTGCTCCTAATTC-3'. The 5' biotinylated primers were 5'-CGGACACCCAAAGTAG-3' for HRV, 5'-GACTATCAAACAGCATAGCAGC-3' for HCV 229E, and 5'-GCAAAGATGGGGAACTGTGG-3' for HCV OC43. Thirty-five cycles of PCR were performed according to previously published methods from our laboratory.4 HRV type 39 and human HCV OC43 and 229E (ATCC, Rockville, MD) were used as positive controls, and sterile PBS was used as a negative control in each reaction series. Unincorporated primers and deoxynucleoside triphosphates were removed from the PCR products by Select-B spin columns (5Prime-3Prime, Boulder, CO).
Oligonucleotide probes for HRV, 5-GCATTCAGGGGCCGGAG-3; HCV 229E, 5'-ACAACACCTGCACTTCCAAA-3'; and HCV OC43, 5'-TATTGGGGCTCCTCTTCTG-3' were labeled at the 3' end with digoxigenen d-UTP. Amplified product was detected by using microplate hybridization.27
RSV, Parainfluenza 1–3, Influenza A and B
The assay for these viruses was conducted using Hexaplex according to the manufacturer’s directions (Prodesse, Inc, Waukesha, WI). Viral genomic RNA was extracted from 280 µL of the nasal specimen. cDNA was synthesized with the use of random hexamers and reverse transcription.28 PCR amplification was conducted combining a mix containing primers for the 6 viruses (Supermix, Prodesse Incorporated, Waukesha, WI) with AmpliTaq Gold II (Applied Biosystems, Foster City, CA) and the synthesized cDNA. Both positive (RNA transcripts) and negative controls (water, negative respiratory samples) were added to the assays. PCR products were then purified with the QIA Quick PCR purification kit (Qiagen) and probe hybridization conducted according to the manufacturer’s instructions. Purified product was added to 96-well avidin-coated microtiter plates (Prodesse) and peroxidase-labeled probe solutions added, each to a single well. After hybridization the plates were washed, a substrate solution was added to each well, the reaction was stopped after 10 minutes, and the optical density of each well was measured at 450 nm on a spectrophotometer. The positive cutoff value was 0.400 or about at least 3 times higher than the negative control.
RSV Antigen Detection
Enzyme immunoassay for rapid detection of RSV was performed on nasopharyngeal aspirate samples. Abbott Testpack (Abbott Laboratories, North Chicago, IL) was used according to the manufacturer’s instructions.
Adenovirus
Frozen (-70°C) specimens of nasopharyngeal aspirate were thawed, clarified by centrifugation at 800 x g, and 0.2 mL of supernatant was inoculated into monolayers of A549 cells. Tubes were incubated at 33°C on roller apparatus. The tubes were examined every other day for cytopathic effect for 2 weeks.
Statistics
A Fisher exact test (2-tailed) was used to compare proportions. Multiple logistic regression analysis was used to test the effect of season of the year and presence of bacterial pathogens on occurrence of abnormal tympanometry during a cold. Model covariates included an indicator of season (fall, winter/spring) when the cold occurred and an indicator for presence of pathogenic bacteria at the time of the cold.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Microbiology of Nasopharyngeal Aspirate Samples
A respiratory virus was detected by RT-PCR in nasopharyngeal samples from 56 (65%) of the 86 illnesses. One virus was found in 48 (56%) of the colds and 2 viruses were detected in 8 (9%). HRV was found in 41 (48%) of the 86 colds, and RSV was found in 12 (14%). A small number of illnesses were attributable to influenza, parainfluenzae, or HCV, but adenovirus was not isolated in cell culture from any of the 53 illnesses studied. The virus positivity rate varied little by season of year (67%, 53%, and 75% for fall, winter, and spring, respectively). HRV was the dominant virus detected in all 3 seasons, and it is notable that a third of the 19 colds in the winter (December, January, February) were HRV-positive (Table 1). RSV, which is thought to be a wintertime virus, was detected in 9 (16%) of 55 colds in the fall (September, October, November). The virus-positive rate (65%) in the 23 colds in children 7 years old was identical to the rate in the 63 colds in children 8 years old. The viruses in the 8 colds in which there were 2 viruses were HRV and RSV (3 colds), HRV and HCV 229E (2), HRV and influenza A (1), RSV and parainfluenza virus (1), and RSV and influenza A (1). The RSV rapid antigen test was not reliable for detection of RSV infection in these children with colds as compared with results obtained by RT-PCR. Only 1 of 14 samples which were rapid test-positive were positive by RT-PCR, and 1 of 12 samples which were RSV-positive by RT-PCR were positive by rapid antigen test.
|
103 cfu/ml of secretion) in 26 (90%) of the 29 positive specimens. Bacterial detection in younger children (11/23, 48%) was similar to the rate in children 8 to 12 years old (18/63, 29%; P = .16). The detection rate was low in fall (25%) and tended to increase in winter (42%) and spring (58%, P = not significant).
Middle Ear Findings
In 84 colds, there were 5 to 1 tympanometric assessments; in 1 cold there were 4 measurements, and in 1 there were 3. Abnormal middle ear pressure by tympanometry was demonstrated at least once during the 2 weeks after onset in 57 (66%) of the 86 colds. The initial abnormal tympanogram occurred during the first week of illness in 50 of the 57 colds. In 17 (30%) of the 57 colds, tympanometry was demonstrated to be abnormal on a single day; 10 of these had C1 and 7 had B curves (Table 2). Sixty percent of abnormal tympanograms were type C1; type C2 and type B constituted 19% and 24% of abnormal tympanograms, respectively. Abnormal positive pressure (+45 daPa) was detected in 1 ear on day 8 of illness.
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
3 years) who are at higher risk for development of overt otitis media. There was a trend toward higher frequency of developing negative pressure in colds in winter and spring compared with fall colds, and the duration of abnormal pressure was clearly prolonged in winter-spring colds. The reason(s) for this difference in winter-spring versus fall colds is not obvious. HRV and RSV were prominent in all 3 seasons, and the frequency of abnormal tympanometry did not vary appreciably with the specific causative virus. The occurrence of abnormal middle ear pressure was similar in frequency in the one third of colds in which an etiologic virus was not detected to that in the two-thirds in which virus was detected. Some of the virus-negative colds may have been attributable to adenovirus for which PCR was not done. It is more likely that the causative virus was not detected because of technical factors, such as sampling time during illness, adequacy of sampling, or presence of PCR inhibitors in the tested sample. The PCR assays used are known to be very sensitive for viral detection compared with cell culture isolation.28,29 The pathophysiologic mechanisms by which viral respiratory infections cause negative middle ear pressure are not understood. There are at least 3 mechanisms by which the viral infection could interfere with eustachian tube function: 1) the virus could cause destructive infection of the mucociliary epithelium of the eustachian tube to block ventilation and prevent clearance; 2) a nondestructive infection of the tubal epithelium by virus could stimulate mucus hypersecretion and/or decrease mucociliary clearance interfering with middle ear ventilation; and 3) nondestructive viral infection of the nasopharynx and the distal tubal epithelium could initiate an inflammatory response with release of proinflammatory mediators and cytokines with effect on eustachian tube ventilation function.
The middle ear pressure abnormalities occurring during colds in this study were intermittent and shifted from one ear to the other from day to day. This pattern of tympanometric abnormalities is more consistent with eustachian tube dysfunction attributable to inflammatory changes in the nasopharynx affecting tubal physiology, rather than to cytopathic injury from virus localized in the tube or middle ear. If there was viral injury localized in the tube or middle ear, progressive decline in tubal ventilation of the ear on the affected side might be expected to remain for some days before recovery. The more common pattern was for a C or B curve to be present in 1 or both ears on 1 day followed by return to normal on the next day, or for 1 ear to be abnormal on 1 day and the other ear to be abnormal at the next examination. This suggests that the driving force for the tympanometric abnormalities was in the nasopharynx, at the terminus of both tubes, with variable effects on both tubes.
Intermittent, brief appearance of B and C2 curves during these uncomplicated colds was of particular interest, because these tympanometric types of abnormality have high correlation with the presence of fluid in the middle ear cavity. The predictive accuracy of tympanometry for the presence of chronic middle ear effusion at surgery has been reported to be 79% to 89%,30,31 but this may not apply to middle ear effusion during colds. Negative middle ear pressure may also produce inflammatory changes in the eardrum. In the current study, we did not have information on the mobility or appearance of the tympanic membrane, because pneumatic otoscopy was not done.
Prevention of complications of bacterial and/or viral otitis media (abbreviated as acute otitis media or AOM) after viral respiratory tract infection is of great interest. Bluestone32 outlined a likely sequence of events in the pathogenesis of AOM: 1) an antecedent viral upper respiratory tract infection produces congestion of the mucosa of the nasopharynx and eustachian tube; 2) congestion of the eustachian tube mucosa results in functional obstruction of the tube; 3) impairment of ventilatory function attributable to the tubal obstruction leads to development of negative middle ear pressure; and 4) prolonged negative middle ear pressure results in "aspiration" of virus and bacteria from the nasopharynx into the middle ear. In the presence of continued tubal obstruction organisms in the middle ear may replicate and produce viral and/or bacterial otitis media. In this pathogenic sequence the microbial infection of the middle ear is secondary.
The relative importance of viruses and bacteria in the pathogenesis of AOM is important in formulating prevention strategies. There is some information on the role of influenza virus. Administration of influenza vaccine to preschool children reduced occurrence of AOM during influenza season by 32% to 36% in vaccinated children compared with controls.33,34 In very recent studies, antiviral treatment of children with influenza infections with a neuraminidase inhibitor (oral oseltamivir) reduced the risk of development of acute otitis media by 44% in 1- to 12-year-olds who did not have otitis media at the time oseltamivir therapy was initiated.35,36 As to the role of bacteria, the effect of pneumococcal vaccine on AOM has been disappointing. In a Finnish study of the efficacy of a 7-valent conjugated pneumococcal vaccine (PncOMPC), AOM caused by the 7 vaccine serotypes was reduced by 56%, but the overall frequency of AOM episodes was reduced only 6%.37 Thus, the available information is consistent with the primary importance of viral infections in the pathogenesis of AOM.
Transient eustachian tube dysfunction resulting in negative middle ear pressure is common in children with uncomplicated colds, particularly during the first week, and resolves without treatment. The findings in this study are most consistent with bilateral eustachian tube dysfunction being caused by a nondestructive viral infection of the nasopharynx and distal tubes producing a local inflammatory response, in the absence of direct viral infection of the middle ears. Therapeutic interventions that might reduce viral replication and the host response to infection deserve study as means of ameliorating middle ear pressure changes and the risk of otitis media. However, interventions directed at modifying host inflammatory responses need to be carefully tested. One recent study of intranasal fluticasone found higher rates of AOM in HRV-positive colds compared with placebo.38
|
ACKNOWLEDGMENTS |
|---|
We thank Kathleen Ashe and Paula Joyner for their microbiologic and virologic work; study nurses Bonnie Stevens, Vanessa John, Nancy Burton, Florence Williams, and Lori Elder for their competent work; and James T. Patrie for statistical analysis.
|
FOOTNOTES |
|---|
Reprint requests to (B.W.) University of Virginia Health System, Department of Otolaryngology, Head and Neck Surgery, Box 800713, Charlottesville, VA 22906. E-mail: bw8b{at}virginia.edu
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. B. Turner, R. Bauer, K. Woelkart, T. C. Hulsey, and J. D. Gangemi An Evaluation of Echinacea angustifolia in Experimental Rhinovirus Infections N. Engl. J. Med., July 28, 2005; 353(4): 341 - 348. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Blomgren and A. Pitkaranta Is it possible to diagnose acute otitis media accurately in primary health care? Fam. Pract., October 1, 2003; 20(5): 524 - 527. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Heikkinen and T. Chonmaitree Importance of Respiratory Viruses in Acute Otitis Media Clin. Microbiol. Rev., April 1, 2003; 16(2): 230 - 241. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. O. Hendley Otitis Media N. Engl. J. Med., October 10, 2002; 347(15): 1169 - 1174. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2002 American Academy of Pediatrics. All rights reserved. |
||
|
||
