Atypical Pathogen Infection in Adults with Acute Exacerbation of Bronchial Asthma

Transcription

1 Atypical Pathogen Infection in Adults with Acute Exacerbation of Bronchial Asthma David Lieberman, Devora Lieberman, Shmuel Printz, Miriam Ben-Yaakov, Zilia Lazarovich, Bella Ohana, Maureen G. Friedman, Bella Dvoskin, Maija Leinonen, and Ida Boldur Pulmonary Unit and Division of Internal Medicine, Soroka Medical Center, and the Department of Virology, Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer-Sheva; Savyon Diagnostics Ltd., Ashdod; Microbiology Laboratory, Asaf Harofeh Medical Center, Zerifin; and Bar-Ilan University, Ramat-Gan, Israel; National Public Health Institute, Oulu, Finland An in-depth analysis leads to the conclusion that the ab- sence of consistent findings in these studies of atypical patho- gens stems from the difference in the age composition of the study populations (adults versus children), the presence or absence of appropriate control groups, and, in particular, the substantial differences in the techniques used to identify specific atypical etiologies for AEBA. There are no reports, to our knowledge, of studies in which an association between Coxiella burnetii and Legionella spp., established members of the atypical pathogen group, and AEBA was investigated. In light of the above, the primary objective of this study was to assess, by advanced serologic methods, the rate of acute infections with these four atypical pathogens in a group of adult patients hospitalized for AEBA compared with a matched control group. Because infection with one of these atypical pathogens can be associated with infection with an additional respiratory pathogen, a second objective of the study was to test and compare the study groups for infections caused by a group of respiratory viruses and the bacterium Streptococcus pneumoniae. In a serologically based prospective study, acute infections with four atypical pathogens were determined in 100 adults hospitalized for acute exacerbation of bronchial asthma, and compared with the corresponding rate in a matched control group. Paired sera were tested using immunofluorescence or enzyme immunoassay methods to establish the serologic diagnosis. In 18 patients (18%), there was evidence of acute infection with Mycoplasma pneumoniae, compared with 3% in the control group (p ). In 10 of these patients there was evidence of infection with at least one additional pathogen, a respiratory virus in 7. There was no significant difference between the study groups in the rates of acute infection by Chlamydia pneumoniae (8% in the hospitalized patients versus 6% in the control subjects), Legionella spp. (5 versus 3%, respectively), or Coxiella burnettii (no patients in either group). We conclude that of these four atypical pathogens, only infection with M. pneumoniae is associated with hospitalization for acute exacerba- tion of bronchial asthma. In most of these M. pneumoniae patients there is evidence of infection with a respiratory virus as well. The pathophysiologic and therapeutic significance of these findings should be tested in further studies specifically designed to address these questions. Keywords: serology; Mycoplasma pneumoniae; Chlamydia pneumoniae; Legionella spp.; Coxiella burnettii (Received in original form September 5, 2002; accepted in final form October 28, 2002) Correspondence and requests for reprints should be addressed to David Lieberman, M.D., Pulmonary Unit, Soroka Medical Center, Beer-Sheva, Israel Lieberma@bgumail.bgu.ac.il This article has an online supplement, which is accessible from this issue s table of contents online at Am J Respir Crit Care Med Vol 167. pp , 2003 Originally Published in Press as DOI: /rccm OC on November 8, 2002 Internet address: METHODS Patients (Study Group) The pivotal role of respiratory tract infections (RTI) in induc- ing acute exacerbations of bronchial asthma (AEBA) in chil- dren and adults is well established (1 5). In most of these episodes the specific infectious etiology is a respiratory virus, usually rhinovirus, coronavirus, or influenza (1, 2, 6). The atypical pathogens are also recognized as causes of RTI, but their role in AEBA is not clear. Some studies found an association between acute infection with Mycoplasma pneumoniae (Mp) and AEBA (7 9), but others did not (10 12). A similar association between Chlamydophila (Chla- mydia) pneumoniae (Cpn) and AEBA was found in two studies (13, 14), but was not confirmed in a later one (15). Recent studies have shown that serologic evidence of chronic infection with Cpn was more frequent in patients with asthma (16), but a trial of roxithromycin for 6 weeks in subjects with asthma and serologic evidence of infection with Cpn leaves the question of the importance of Cpn in asthma unanswered (17, 18). The study included patients hospitalized for AEBA over a 12-month period who provided informed consent to participate and met all of the following inclusion criteria: (1) age over 18 years; (2) a past history of typical bronchial asthma; (3) reversibility of at least 15% of the FEV 1 value with the patient in stable condition; (4) a smoking history of no more than 10 pack-years; (5) the present hospitalization for a new-onset or exacerbation of dyspnea over the week before admission; (6) negation during the hospitalization of causes of dyspnea other than bronchial asthma; and (7) ruling out of pneumonia by admission chest X-ray. The control group included patients hospitalized over the same time period in the orthopedic surgery ward who provided informed consent to participate and met all of the following conditions: (1) age over 18 years; (2) hospitalization for trauma or elective surgery; (3) no evidence of febrile illness or RTI over the month before hospitalization and during the period between the two serum samples; and (4) negation of any lung disease. Recruitment of control patients was conducted in parallel with the study group to ensure a similar distribution by seasons of the year. An acute blood sample was drawn for serologic testing within 48 hours of hospitalization and a second (convalescence phase) sample 3 to 5 weeks later. The etiologic workup in this study was based, exclusively, on serologic tests. Serologic tests were conducted for 12 pathogens known to cause upper or lower RTI that can be identified by serologic testing. The antibody levels for Mp were determined using three commercial enzyme immunoassay (EIA) kits: SeroMp-IgA, SeroMp-IgG, and Sero- Mp-IgM (Savyon Diagnostics, Ashdod, Israel). These three kits are an improved version of a previous commercial kit, SeroMp (Savyon Diagnostics), whose sensitivity and specificity were assessed and were found to be very good (19). The sensitivity and specificity of the novel SeroMp-IgA kit (with which 85% of the patients were identified as positive for Mp in the present study) was recently validated and found to have high sensitivity and specificity for the serologic diagnosis of

2 Lieberman, Lieberman, Printz, et al.: Atypicals in Asthma Exacerbation 407 TABLE 1. DEMOGRAPHIC DATA, BASELINE FEV 1, FEV 1 TABLE 2. CLINICAL MANIFESTATIONS OF EXACERBATION, REVERSIBILITY, AND O 2 SATURATION IN STABLE ARTERIAL PO 2, PCO 2, AND PH AT ADMISSION, RATES CONDITION, HOME OXYGEN, CHRONIC ORAL STEROID OF ADMISSION TO INTENSIVE CARE AND INVASIVE THERAPY, RATE OF INFLUENZA AND PNEUMOCOCCAL MECHANICAL VENTILATION, DURATION OF VACCINATION, AND CHRONIC COMORBIDITY IN THE HOSPITALIZATION, READMISSION WITHIN 30 d, AND 100 PATIENTS WITH ASTHMA RECOVERY TO BASELINE FUNCTIONAL CONDITION WITHIN 30 d AFTER DISCHARGE FROM THE HOSPITAL Variable Result Range IN THE 100 PATIENTS WITH ASTHMA Sex, % males 36 Mean age, yr SD Variable Result Range Baseline FEV 1, % of expected* Abrupt onset of exacerbation, %* 22 % baseline FEV 1 reversibility, mean SD* Feeling of shaking chills, % 28 Baseline O 2 saturation, mean SD* Fever during exacerbation, % 54 Home oxygen, % 3 PO 2 at admission, mean SD at room air Chronic oral steroid therapy, % 24 PCO 2 at admission, mean SD at room air Influenza vaccination, % 28 ph at admission, mean SD Pneumococcal vaccination, % 7 Admission to intensive care, % 26 Diabetes mellitus, % 13 Invasive ventilation, % 7 Ischemic heart disease, % 3 Mean length of hospital stay, d SD Left-sided heart failure, % 4 Readmission, % 5 Recovery within 30 d, % 94 * Baseline FEV 1, FEV 1 reversibility, and O 2 saturation values were determined 6 months before hospitalization, or within 1 2 mo after hospitalization, at room * Development of all symptoms of exacerbation within less than 12 h. air with the patient in stable condition. as closely as possible. This matching process was performed acute Mp infection (20). Acute Mp infection was diagnosed if a significant before the serologic tests were conducted; these 100 patients change (according to the formula in the manufacturer s instruc- comprised the final control group. The mean age ( SD) for tions) in the antibody level was found between the acute and convales- the control group was years (range, 19 74), and 34 cence serum samples. The methods, kits, and criteria used to reach patients were males. The mean time between the serum samples serologic diagnoses of infection with the other 11 pathogens were identiwas days (range, 19 37). There were no significant cal to those described in our previous publication (21), with the lone exception that acute Cpn infection was also diagnosed in the presence differences in these parameters between the study and control of unchanged titers of IgG 512 or IgM 8. groups. The results were analyzed using the statistical software Epi Info Table 1 shows the demographic characteristics of the study (Epidemiology Program Office, Center for Disease Control and Prevention, group and clinical data such as baseline FEV 1,FEV 1 reversibility, Atlanta, GA). The 2 test or its equivalent served to compare O 2 saturation in stable condition, home oxygen use, chronic oral proportions between groups, and ANOVA was done to compare contin- steroid therapy, rates of influenza, and pneumococcal vaccinauous variables among two or more groups. Statistical significance was tion and chronic comorbidity. set at p Table 2 details clinical and laboratory data relating to the An expanded Patients and Methods section with a full study protocol exacerbation that led to hospitalization, the clinical course, and and a detailed description of the laboratory etiologic workup appears in the online supplement of this article. the convalescence phase, including arterial Po 2,Pco 2 and ph at admission, rates of admission to intensive care, rates of mechanical ventilation, duration of hospitalization, rates of readmission RESULTS within 30 days of discharge, and recovery to baseline function One hundred thirteen patients with a working diagnosis of AEBA who were hospitalized during the study period were recruited into the study. All the patients survived the hospitaliza- TABLE 3. COMPARISON OF THE FREQUENCY DISTRIBUTION tion and the period between the hospitalization and the follow- OF ACUTE INFECTIONS WITH THE VARIOUS PATHOGENS up appointment, and all came to the follow-up. Eleven patients BETWEEN THE PATIENTS WITH ASTHMA (N 100) AND for whom there was no FEV 1 reversibility test before hospitaliza- THE CONTROL GROUP (N 100) tion and who did not meet the reversibility criteria at the followup visit after discharge were not included in the data analysis. Pathogen Asthma Control p Value Two patients with a polyclonal antibody reaction to all tested Viral agents, % pathogens were also excluded from analyses. Thus, the final data Influenza virus type A analysis was conducted on the study population of 100 patients Influenza virus type B 5 1 NS Parainfluenza virus type NS hospitalized for AEBA. For all of these patients, without excep- Parainfluenza virus type NS tion, a convalescence serum sample was obtained at a mean of Parainfluenza virus type NS days after the initial serum sample was taken (range, Adenovirus 6 1 NS days). Respiratory syncytial virus 2 0 NS One hundred sixty-eight patients were recruited during the One or more of the above study period into the control group. Sixteen reported a febrile Bacterial agent, % Streptococcus pneumoniae 3 3 NS illness and/or symptoms of respiratory tract infection in the pe- Atypical bacterial agents, % riod between the first serum sample and the scheduled time for Legionella spp. 5 3 NS the second sample, so they were excluded from the study. Eight Mycoplasma pneumoniae patients did not provide convalescence serum sample and were Coxiella burnetii 0 0 NS also excluded. The final control group of 100 patients was consti- Chlamydia pneumoniae 8 6 NS One or more of the above tuted from the remaining 144 patients so that the distribution No infectious etiologies found by sex, age, and season of the year would match the study group

3 408 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL TABLE 4. A COMPARISON OF THE NUMBER OF ACUTE INFECTIONS PER PATIENT BETWEEN THE PATIENTS WITH ASTHMA (N 100) AND THE CONTROL GROUP (N 100) Positive serologic results Asthma Control p Value DISCUSSION period, and the seasonal peaks of influenza were not different from other years The medical community has become increasingly aware of the importance of the atypical bacteria over the last five decades Today this group is considered an important etiologic factor in NS NS RTI together with respiratory viruses and classical bacteria. In addition to community-acquired pneumonia, in which these pathogens play an important role (22), they are also important in upper and lower RTI in the community (23) and in acute exacerbation of COPD (21). In light of the association between within 30 days of discharge. All blood cultures from febrile RTI and AEBA described above, it made sense to evaluate the patients in the study group on admission were sterile. involvement and contribution of these pathogens to AEBA. There was serologic evidence of acute infection in 49 of the Several methodologic issues relating to our study have to be patients in the study group with at least one of the 12 pathogens addressed. The etiologic diagnoses for atypical bacteria in the tested. This rate was significantly higher than the rate in the present study were based entirely on serologic tests. At the control group. Table 3 shows the rates of acute infection with planning stage of the study we took into consideration the theothe various pathogens in the study and control groups and a retical possibility that the antibody response upon which the comparison between them. There were significant differences diagnoses were based could be nonspecific. The solution to this between the study and control groups in the rates of acute infec- problem in this study was the absolute requirement that paired tion with respiratory viruses and atypical bacteria, but not S. sera be obtained for all patients in both study group together pneumoniae. Among the atypical bacteria there was a statistically with the very strict diagnostic criteria that were based, except significant difference between the study groups for acute infec- for Cpn, on a significant change in the antibody titer between tion with M. pneumoniae, but not the other three atypical patho- the paired sera. We preferred serologic testing to isolation of gens. pathogens by culture or identification by PCR from respiratory In 13 of the 49 patients with acute infection (27%) there secretions in light of the difficulty in ascribing a cause effect was evidence of infection with more than one pathogen. The relationship between a positive finding by these methods and distribution of the number of acute infections per patient in the acute infection. This problem is particularly critical in patients study groups is shown in Table 4. with chronic asthma, because colonization of the lower respira- In 8 of the 18 patients with acute M. pneumoniae infection tory tract by Mp and respiratory viruses has been found under this was the only pathogen, whereas in 10 others (56%) at least stable disease conditions at the same time that serology for acute one other pathogen was identified. In eight of these patients one infection with Mp was negative (24, 25). The only one exception other pathogen was found, in one patient two were found, and to our absolute requirement for a significant antibody titer in one patient three other cause of acute infection were identi- change between the paired sera as a diagnostic criterion was for fied. In 7 of the 10 patients the additional pathogen was a respira- Cpn. In this study, as in previous studies (13, 15), acute infection tory virus, in five an atypical bacterium (C. pneumoniae or Legio- with Cpn was also diagnosed in the presence of high but unnella spp.), and in one patient S. pneumoniae. changed titers. These criteria are somewhat problematic in that To identify unique characteristics of the group of patients an unchanged titer may reflect chronic infection with this pathowith asthma with evidence of acute Mp infection, we compared gen and not necessarily acute infection (26). In any event, we all parameters appearing in Tables 1 and 2 between these 18 found no difference in the rates of acute Cpn infection between patients and the 82 patients with asthma without evidence of the study and control groups with the criteria that were used, acute infection with Mp (non-mp). No statistically significant nor would a difference have been found with various criteria differences were found for any of these parameters between based on changing titers. these two groups. However, interesting trends were found be- It would have been appropriate to present sensitivity and tween the groups in some parameters, even though these differ- specificity data for each of the serologic tests used for the various ences did not reach statistical significance. This was the case in pathogens in this study. This issue of the sensitivity and specificity relation to sex (28% males in the Mp group versus 38% in the of serologic tests for the diagnosis of acute respiratory infections non-mp group), chronic oral steroid therapy (33% versus 22%, is very problematic. An accepted gold standard is required for respectively), diabetes mellitus (6% versus 15%, respectively), each pathogen to determine these characteristics. To date there and return to baseline function within 30 days (100% versus is no single laboratory test for the pathogens tested in the present 93%). In the Mp group there were three patients on home oxygen study that could be considered as a gold standard for acute therapy, seven patients who underwent invasive ventilation, and infection with that pathogen. Even the most advanced methods, five patients who were readmitted to the hospital, compared such as PCR, can identify with high sensitivity the presence of with none in any of these categories in the non-mp group. For a specific pathogen in the respiratory tract, but they cannot all other parameters appearing in Tables 1 and 2 the results were negate the possibility that the pathogen is present as a chronic very similar for the Mp and non-mp groups of patients with agent rather than the cause of acute infection (24, 25). The asthma. absence of precise and reliable sensitivity and specificity data There was not a single case of a fourfold or greater increase for each of the laboratory tests used in this study posed a serious in antibody titer for Cpn with any of the three antibody types methodologic challenge for us, because the significance of the tested in any of the 200 patients in both study groups. The eight study findings is dependent upon these unavailable data. We positive results in the study group and the six positive results in dealt with this challenge by adopting two strategies. First, we used standard, accepted serologic criteria for the difference be- tween two serum samples as the basis for diagnosing acute infection. The application of these criteria necessitated a large effort the control group were determined by the index of an unchanged titer of IgG 512 or IgM 8. There was no epidemic of Mp in Israel during the study

4 Lieberman, Lieberman, Printz, et al.: Atypicals in Asthma Exacerbation 409 to obtain two serum samples from each of the participants in believe that this lack of significant difference between the Mp the study. Second, we compared the results of the study group and non-mp groups of patients with asthma stems from the with a control group of patients, matched to the study group by relatively small size of the group of patients with Mp among the age, sex, and season of the year, who did not have evidence of patients with asthma. In addition, the fact that most of these 18 asthma or RTI. This comparison neutralized, in effect, the issue patients also had evidence of infection with an another respiratory of the specificity of the results for acute exacerbation of bronchial pathogen around the time of the acute episode raises the asthma. An example of this can be seen in the absence of a possibility that the clinical expression of the exacerbation was significant difference between the study and control groups for caused solely or predominantly by the other pathogen, making acute infection with Cpn, Legionella spp., or S. pneumoniae. it impossible to demonstrate unique clinical characteristics Taken together, these two strategies significantly reduced the among the 18 patients with Mp infection. importance of the absence of precise sensitivity and specificity The rate of acute infection with Cpn in the present study data for the various serologic methods in relation to the significance among patients with AEBA was 8%. This rate was not signifi- of the study results. cantly different from the corresponding rate in the control group. The serologic tests for Legionella spp. in this study were This finding is consistent with the result of a previous study conducted using an in-house kit. It is important to note that conducted in a similar population and with identical methods. there is no commercial kit on the market with which it is possible That study also reported no significant difference between the to test for all 41 serogroups of Legionella spp. that were tested. AEBA group and a matched control group (15). The positivity Although the use of an in-house kit is problematic, this is a rate for acute Cpn infection in the control group in that study universal problem and not a specific one for the present study. was 5.7%, a rate that is identical to our finding. This rate can The problem stems, primarily, from the absence of a gold standard be explained as true asymptomatic acute infection in the control against which results of serologic tests can be validated. group or as a false positive rate because of the diagnostic criterion Cultures of pathogens as well as PCR detection have low sensitivity of a high, but unchanged, titer of specific antibodies that may and cannot be used as a gold standard. The accepted substitute reflect a chronic, rather than an acute, infection. for validations of test results is strict QC/QA procedures together Approximately one-quarter of the patients in the study group with rigid diagnostic criteria, using paired sera. Both of these with evidence of an acute infection had at least one additional conditions were strictly adhered to in this study. pathogen. This phenomenon is well recognized in a broad spec- We checked the sera from the study and control groups for trum of RTI (21, 23, 27, 29, 30), so it is not surprising to find it seven respiratory viruses and S. pneumoniae with the aim of in the context of AEBA. In 10 of the 13 patients with acute identifying cases in which AEBA was caused by simultaneous infection with more than one pathogen, one of the pathogens infection with an atypical pathogen and an additional respiratory was Mp. These 10 patients represented the majority of the pa- pathogen. Rhinovirus and coronaviruses are conspicuously ab- tients who were positive for Mp, and in most cases the additional sent from the list of viruses tested in this study, despite their pathogen was a respiratory virus. This high rate of coinfection dominant position among the viruses that cause AEBA (1, 3). with a respiratory virus in patients positive for Mp has been The reason they were not included in this study was the known reported previously (8, 9). It is important to note that the rate methodologic difficulty in diagnosing acute infection with these of patients with evidence for more than one pathogen in general, pathogens by serologic methods. This difficulty is also reflected and the combination of Mp and respiratory viruses in particular in the absence of a commercial kit that can be used to diagnose in the present study, was affected apparently to a great extent infections caused by these viruses. The choice of the seven respiratory by the absence of rhinovirus and coronavirus from the list of viruses that were tested in this study was based on the tested viruses. In light of the data showing the dominant role practical consideration of the availability of a reliable kit for played by these viruses in acute infection among patients with their serologic identification, Our decision to include S. pneu- AEBA, it is reasonable to assume that the true number of patients moniae was based on our ability to reliably identify infection with more than one infection, and in particular those with this bacterium with the battery of serologic tests used in patients with Mp and a respiratory virus, is much higher than the study, as proven in our two previous studies on communityacquired the rate identified in this study. pneumonia (27, 28). Any attempt to reach therapeutic conclusions for AEBA The main and most important finding in this study is that, in from the results of this study is problematic. The association comparison with the control group, there was a significantly between Mp RTI and AEBA is apparently mediated by a TH2 higher rate of acute infection with Mp in the study group, but type immunologic reaction. This type of immunologic response not with the other three atypical pathogens. The specific rate of can be induced by Mp, but is also a hallmark of the immunologic 18% acute Mp infections in this population of adults hospitalized response in AEBA (31). If this is indeed the basis for the inducwith AEBA that was found in this study is very close to the rate tion of AEBA by Mp RTI, there is no assurance that specific of 21% that was found in a previous study in a similar population antibiotic therapy for Mp, after the infection has already induced (7). In that study, 4 of 20 AEBA patients who were positive for the immunologic response, could stop or reverse that response. acute Mp infection also had community-acquired pneumonia, Another difficult problem concerning the therapeutic applica- an exclusion criterion in our study. Exclusion of the patients tion of the results of this study is that in most patients with with community-acquired pneumonia from the population of AEBA with Mp infection another pathogen was also identified, the previous study would reduce the positivity rate to 18%, a rate usually a respiratory virus. If this is the case, one cannot negate that is surprisingly identical to the present result. The authors of the possibility that the Mp infection followed the viral infection that study state that their use of reliable, sensitive methods is so that antibiotic therapy would not change the course of AEBA. the reason for the high rate of Mp infections found in their study However, and despite the two qualifications just discussed, we compared with previously published studies. The results of the believe that the findings of the study justify the implementation present study strengthen that contention. of prospective, well-controlled trials designed to address the There were no statistically significant differences in any pa- question of the effectiveness of specific antibiotic therapy for rameter between the 18 patients with asthma with Mp infection Mp during AEBA. and the 82 patients with asthma without Mp infection, although We conclude that of these four atypical pathogens, only infection in a few parameters a trend toward a difference was found. We with M. pneumoniae is associated with hospitalization for

5 410 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL acute exacerbation of bronchial asthma. In most of these M. Increased frequency of Chlamydia pneumoniae antibodies in patients pneumoniae infected patients there is also evidence of infection with asthma. Am J Respir Crit Care Med 2001;163: Johnston SL. Is Chlamydia pneumoniae important in asthma? The first with a respiratory virus. The pathophysiologic and therapeutic controlled trial of therapy leaves the question unanswered. Am J significance of these findings need to be evaluated in future Respir Crit Care Med 2001;164: studies designed specifically to address these issues. 18. Black PN, Blasi F, Jenkins CR, Scicchitano R, Mills GD, Rubinfeld AR, Ruffin RE, Mullins PR, Dangain J, Cooper BC, et al. Trial of References roxithromycin in subjects with asthma and serological evidence of infection with Chlamydia pneumoniae. Am J Respir Crit Care Med 1. Atmar RL, Guy E, Guntupalli KK, Zimmerman JL, Bandi VD, Baxter 2001;164: BD, Greenberg SB. Respiratory tract viral infections in inner-city 19. Lieberman D, Lieberman D, Horowitz S, Horovitz O, Schlaeffer F, Porasthmatic adults. Arch Intern Med 1998;158: ath A. Microparticle agglutination versus antibody-capture enzyme 2. Teichtahl H, Buckmaster N, Pertnikovs E. The incidence of respiratory immunoassay for diagnosis of community-acquired Mycoplasma pneutract infection in adults requiring hospitalization for asthma. Chest moniae pneumonia. Eur J Clin Microbiol Infect Dis 1995;14: ;112: Watkins-Riedel T, Stanek G, Daxboeck F. Comparison of SeroMP IgA 3. Johnston SL, Pattemore PK, Sanderson G, Smith S, Lampe F, Josephs L, with four other commercial assays for serodiagnosis of Mycoplasma Symington P, O Toole S, Myint SH, Tyrrell DA. Community study pneumoniae pneumonia. Diagn Microbiol Infect Dis 2001;40: of role of viral infections in exacerbations of asthma in 9 11 year old 21. Lieberman D, Lieberman D, Ben-Yaakov M, Lazarovich Z, Hoffman S, children. BMJ 1995;310: Ohana B, Friedman MG, Dvoskin B, Leinonen M, Boldur I. Infectious 4. Nicholson KG, Kent J, Ireland DC. Respiratory viruses and exacerbations etiologies in acute exacerbation of COPD. Diagn Microbiol Infect Dis 2001;40: of asthma in adults. BMJ 1994;307: Lieberman D. Atypical pathogens in community-acquired pneumonia. 5. Johnston SL, Pattemore PK, Sanderson G, Smith S, Campbell MJ, Clin Chest Med 1999;20: Josephs L, Cummingham A, Robinson BS, Myint SH, Ward ME, et 23. Lieberman D, Lieberman D, Korsonsky I, Ben-Yaakov M, Lazarovich al. The relationship between upper respiratory tract infections and Z, Friedman MG, Dvoskin B, Leinonen M, Ohana B, Boldur I. A hospital admissions for asthma: a time-trend analysis. Am J Respir comparative study of the etiology of adult upper and lower respiratory Crit Care Med 1996;154: tract infections in the community. Diagn Microbiol Infect Dis 2002; 6. Papadopoulos NG, Johnston SL. Viruses and asthma exacerbations. Tho- 42: rax 1998;53: Kraft M, Cassell GH, Henson JE, Watson H, Williamson J, Marmion 7. Seggev JS, Lis I, Siman-Tov R, Gutman R, Abu-Samara H, Schey G, BP, Gaydos CA, Martin RJ. Detection of Mycoplasma pneumoniae Naot Y. Mycoplasma pneumoniae is a frequent cause of exacerbation in the airways of adults with chronic asthma. Am J Respir Crit Care of bronchial asthma in adults. Ann Allergy 1986;57: Med 1998;158: Huhti E, Mokka T, Nikoskelainen J, Halonen P. Association of viral and 25. Martin RJ, Kraft M, Chu HW, Berns EA, Cassell GH. A link between mycoplasma infections with exacerbations of asthma. Ann Allergy chronic asthma and chronic infection. J Allergy Clin Immunol 2001;107: 1974;33: Berkovich S, Millian SJ, Snyder RD. The association of viral and myconen 26. von Hertzen L, Alakarppa H, Koskinen R, Liippo K, Surcel H-M, Leinoplasma infections with recurrence of wheezing in the asthmatic child. M, Saikku P. Chlamydia pneumoniae infection in patients with Ann Allergy 1970;28: chronic obstructive pulmonary disease. Epidemiol Infect 1997;118: Hudgel DW, Langston L Jr, Seiner JC, McIntosh K. Viral and bacterial 164. infections in adults with chronic asthma. Am Rev Respir Dis 1979;120: 27. Lieberman D, Schlaeffer F, Boldur I, Lieberman D, Horowitz S, Fried man MG, Leinonen M, Horovitz O, Manor E, Porath A. Multiple 11. Minor TE, Dick EC, Baker JW, Ouellette JJ, Cohen M, Reed CE. pathogens in adult patients admitted with community-acquired pneu- Rhinovirus and influenza type A as precipitants of asthma. Am Rev monia: a one year prospective study of 346 consecutive patients. Thorax 1996;51: Respir Dis 1976;113: Lieberman D, Lieberman D, Gelfer Y, Varshavsky R, Dvoskin D, Leino- 12. Tarlo S, Broder I, Spence L. A prospective study of respiratory infection nen M, Friedman MG. Pneumonic vs. non-pneumonic acute exacerbain adult asthmatics and their normal spouses. Clin Allergy 1979;9:293 tion of COPD. Chest 2002;122: Neill AM, Martin JR, Weir R, Anderson R, Chereshsky A, Epton MJ, 13. Allegra L, Blasi F, Centanni S, Consentini R, Denti F, Raccanelli R, Jackson R, Schousboe M, Frampton C, Hutton S, et al. Community Tarsia P, Valenti V. Acute exacerbations of asthma in adults: role of acquired pneumonia: aetiology and usefulness of severity criteria on Chlamydia pneumoniae infection. Eur Respir J 1994;7: admission. Thorax 1996;51: Hahn DL, Dodge RW, Galubjatnikov R. Association of Chlamydia pneu- 30. Ragnar NS. Atypical pneumonia in the Nordic countries: aetiology and moniae (Strain TWAR) infection with wheezing, asthmatic bronchitis, clinical results of a trial comparing fleroxacin and doxycycline. Nordic and adult-onset asthma. JAMA 1991;266: Atypical Pneumonia Study Group. J Antimicrob Chemother 1997;39: 15. Cook PJ, Davies P, Tunnicliffe W, Ayres JG, Honeybourne D, Wise R Chlamydia pneumoniae and asthma. Thorax 1998;53: Daian CM, Wolff AH, Bielory L. The role of atypical organisms in 16. Gencay M, Rudiger JJ, Tamm M, Soler M, Perruchoud AP, Roth M. asthma. Allergy Asthma Proc 2000;21: