Etiology of community-acquired pneumonia treated in an ambulatory setting

Respiratory Medicine (2005) 99, 60 65 Etiology of community-acquired pneumonia treated in an ambulatory setting Thomas J. Marrie a, *, Melanie Poulin-Costello b, Michael D. Beecroft b, Zeljka Herman-Gnjidic b a Department of Medicine, University of Alberta, Edmonton, AB, 2F1.30 WMC, 8440 112 Street, Edmonton, Alberta, Canada T6G 2B7 b Bayer Inc, 77 Belfield Rd, Toronto, Ont., Canada M9W 1G6 Received 5 November 2003; accepted 17 May 2004 KEYWORDS Pneumonia; Etiology; Streptococcus pneumoniae; Mycoplasma pneumoniae; Chlamydia pneumoniae Summary Background: Very few studies have addressed the etiology of communityacquired pneumonia (CAP) treated in an ambulatory setting. Methods: Patients were recruited from physicians offices and from Emergency Rooms in Canada. Pneumonia was defined as two or more respiratory symptoms and signs and a new opacity on chest radiograph interpreted by a radiologist as pneumonia. Blood and sputum for culture as well as acute and convalescent serum samples for serology were obtained. Antibodies to Mycoplasma pneumoniae and Chlamydia pneumoniae were determined using enzyme-linked immunosorbent assays. Results: Five hundred and seven patients were enrolled in the study; 419 (82%) had blood cultures done, seven (1.4%) of which were positive for Streptococcus pneumoniae; 241 (47.5%) had a sputum processed for culture, 31% of which were positive for a potential respiratory pathogen. 437 (86.2%) had both acute and convalescent serum samples obtained, 148 (33.8%) of which gave a positive result. Overall an etiological diagnosis was made in 48.4% of the patients. M. pneumoniae accounted for 15% of the cases, C. pneumoniae 12%, S. pneumoniae 5.9% and Haemophilus influenzae 4.9%. Conclusions: Despite considerable effort an etiological diagnosis of CAP treated on an ambulatory basis was made in only half the patients. The most commonly identified pathogens were M. pneumoniae, C. pneumoniae, S. pneumoniae, H. influenzae. & 2004 Elsevier Ltd. All rights reserved. Introduction Community-acquired pneumonia (CAP) is a common disease in North America with significant morbidity and mortality. 1 3 About 80% of all patients with CAP *Corresponding author. Tel.: þ 1-780-407-6234; fax: þ 1-780- 407-3132. E-mail address: tom.marrie@ualberta.ca (T.J. Marrie). are treated in an ambulatory setting, 4 however, we know very little about the etiology of pneumonia affecting these patients. Most ambulatory patients with CAP are seen in their physician s office and do not have any diagnostic testing. Indeed in a study of 944 patients with pneumonia treated either in their physicians office or in an Emergency Room only 8% had blood cultures, 11% had sputum cultured and 1.3% had acute and convalescent blood samples 0954-6111/$ - see front matter & 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.rmed.2004.05.010

Etiology of CAP 61 collected for serological testing. 5 A randomized trial of the treatment of pneumonia in ambulatory setting gave us the opportunity to determine the etiology of this important subset of patients with CAP. Methods Study design This was a prospective, randomized, double-blind, multicentre study to compare the efficacy and safety of moxifloxacin 400 mg PO OD for 10 days, to clarithromycin 500 mg PO BID for 10 days in the treatment of non-hospitalized patients with CAP. Patients 18 years of age or older, with two or more of the following signs and symptoms (fever, cough, sputum production, pleuritic chest pain, dyspnea, crackles, bronchial breathing) consistent with bacterial pneumonia of mild to moderate severity, not requiring hospitalization (Fine risk class I III were enrolled). 6 In order to be classified as having community-acquired pneumonia, patients had to have radiographic evidence of a new or progressive opacity. Patients were excluded from the trial if they met any of the following criteria: a history of allergy to carboxyquinolone or macrolide derivatives; pregnant or nursing women; severe hepatic impairment (baseline SGOT or SGPT, and/or total bilirubin X3 times the upper limit of normal); co-existent disease considered likely to affect the outcome of the study (e.g., lung cancer, lung abscess, collagen vascular disease affecting the lungs, empyema, cystic fibrosis, known or suspected active tuberculosis, alcoholism); use of systemic corticosteroids 415 mg per day oral prednisone (or its equivalent). Diagnostic workup Two sets of blood cultures, sputum for culture and a sample of clotted blood for serology were collected at enrollment. Additional specimens for serological testing were collected at the 2- and 6-week followup visits. Laboratory methods Blood and sputum were processed for culture at the hospital microbiology laboratory of the site investigator. Only sputum samples with p10 squamous epithelial cells per low power field and X25 leucocytes per low power field were processed for culture. Standard methods were used for isolation and identification of microorganisms. Serology All serum samples were tested at a central laboratory. Serum samples (acute phase, 14 and 42 day) were received frozen on dry ice. Prior to testing, samples were thawed and then centrifuged at 3000g for 10 min. Antimicrobial susceptibility testing This testing was performed at each site using standard methods. Chlamydia pneumoniae IgG and IgM Samples were tested using a qualitative enzymelinked immunosorbent assay (ELISA) (Savyon Diagnostics Ltd., Ashdod, Israel) according to the manufacturer s instructions. Results were expressed as the cut-off index which is equal to the absorbance of the patient s serum sample at 450 nm divided by twice the absorbance of the negative control at 450 nm. A cut-off index of less than 1 was considered negative and a value of greater than 1.1 was considered as positive. Values between 1 and 1.1 were considered borderline. Results of enrollment, day 14 and 6-week follow-up serum were interpreted based on both the IgG and IgM results as follows: IgG IgM Interpretation Negative Negative No evidence of infection Negative Positive Recent/current infection Positive Positive Recent/current infection Positive Negative Past infection Mycoplasma pneumoniae IgG and IgM Serum samples were tested using a semi-quantitative enzyme-linked immunosorbent assay (ELISA) (Savyon Diagnostics Ltd., Ashdod, Israel) according to the manufacturer s instructions. A standard curve was generated for each assay using the calibrators provided by the manufacturer. The absorbance values of the patient samples were then plotted on the standard curve and the concentration of antibodies interpolated and expressed as binding units per ml (BU/ml). Patients with BU/ml values of 415 were considered

62 T.J. Marrie et al. positive. Results of pre-study, day 14 and 6-week follow-up serum were interpreted based on both the IgG and IgM results as follows: IgG IgM Interpretation Negative Negative No evidence of infection Negative Positive Recent/current infection Positive Positive Recent/current infection Positive Negative Past infection Patients with negative serology at Pre-Study but whose IgG and/or IgM became positive at day 14 or 6-week follow-up were considered as having recent/current infection. Sputum culture interpretation All sputum culture results were interpreted in conjunction with Gram stain results. In particular any aerobic Gram negative bacillus that was considered a pathogen had to be present on Gram stain. Results The study population as described in Table 1 was young with a mean age of 47.8 years and most had pneumonia of mild severity. One patient who had small cell carcinoma of the lung died. Almost 8% of the study population required subsequent admission to hospital, with half of these patients requiring admission because of the pneumonia and the remainder because of worsening of comorbidities. The results of blood cultures are given in Table 2. Note that only 1.4% were positive and that all isolates were Streptococcus pneumoniae. One patient had yeast isolated from the blood which was felt to be a contaminant. This person had serological evidence of acute infection with Mycoplasma pneumoniae. Three of the seven patients with pneumococcal bacteremia were admitted to hospitalftwo for worsening pneumonia and one for sepsis. All of the S. pneumoniae isolates were susceptible to clarithromycin and moxifloxacin. Three were susceptible to penicillin and one was intermediately resistant and another was resistant. Penicillin susceptibility results were not available for two isolates. Table 3 gives the sputum culture results. Only 47.5% had a sputum specimen obtained and 37.7% of these were positive for a respiratory pathogen. Overall 75 (15%) of patients had a positive sputum culture. One of the patients who had S. pneumoniae Table 1 Characteristics of the study population. Number of patients 507 N(%) male 263 (51.9%) Age (years) mean (SD) 47.8 (16.2) FINE class I 286(56.4%) II 150 (29.6%) III 64 (12.6%) IV 7(1.4%) Subsequent hospitalization Respiratory related N(%) 20(3.9%) Non-respiratory related N(%) 20(3.9%) N (%) deaths 1(0.2%) Systolic blood pressurefmean (SD) 125.6 (18.0) Diastolic blood pressurefmean (SD) 75.1 (10.9) Heart rate (BPM)Fmean (SD) 88.8 (17.2) Respiration ratefmean (SD) 21.1 (4.2) Temperature(1C)Fmean (SD) 37.3 (1.0) Hemoglobin (g/l)fmean (SD) 134.7 (15.5) White blood cell count (giga/l)f 10.5 (5.4) mean (SD) Platelet count (giga/l)fmean (SD) 270.9 (91.2) Table 2 Blood culture results. No. of patients 507 No. (%) with blood cultures done 419 82% No. (%) with blood cultures positive 7 1.4% for a pathogen Organism isolated from blood S. pneumoniae 7 isolated from the sputum also had this microorganism isolated from the blood. Table 4 gives the results of the serological testing. Fifteen percent had M. pneumoniae pneumonia and 12% had Chlamydia pneumoniae pneumonia. A small number, 11 (2.5%), met the criteria for dual infection with these pathogens. It is noteworthy than 9 of the 11 were positive for C. pneumoniae and M. pneumoniae IgM antibodies at the time of enrollment. For the remaining two, one was positive for C. pneumoniae at the time of enrollment and M. pneumoniae was positive at the 2-week follow-up visit. The reverse was true for the other. Overall an etiological diagnosis was made in 48.4% (Table 5). M. pneumoniae, C. pneumoniae and S. pneumoniae were the three most common

Etiology of CAP 63 Table 3 Sputum culture results. No. of patients 507 Sputum sample obtained 241 47.5% Not processed 6 1.1% No. of patients with a positive 75 31% sputum culture Organisms isolated from sputum Normal flora 92 38% No growth 68 28% H. influenzae 25 10% S. pneumoniae 24 10% H. parainfluenzae 10 4% S. aureus 6 2.5% Moraxella catarrhalis 6 2.5% Streptococcus species 5 2% Klebsiella pneumoniae 2 0.8% Acinetobacter spp 2 0.8% Streptococcus pyogenes 1 Streptococcus agalactiae 1 Neisseria meningitidis 1 Escherichia coli Enterobacter cloacae 1 Enterbacter aerogenes 1 Pseudomonas aeruginosa 1 Agrobacterium radiobacter 1 Haemophilus parahaemolyticus 1 Pasteurella multocida 1 Some patients had more than 1 microorganism isolated. Table 4 Results of testing serum samples for antibodies to M. pneumoniae and C. pneumoniae. No. of patients 507 No. with acute and convalescent 437 86.2% serology No. with acute phase serology 67 13.2% only No. with M. pneumoniae 76 15% No. with C. pneumoniae 61 12% No. with both 11 2.5% pathogens. Haemophilus influenzae was surprisingly common at 4.9%. Discussion This study has several important findings. The first is that an etiological diagnosis was made in only half the patients. The rank order of the pathogens, M. pneumoniae, C. pneumoniae and S. pneumoniae is similar to that found in other studies. 7 10 Indeed Table 5 Etiology of pneumonia in 507 ambulatory patients. Unknown 262 51.6% M. pneumoniae 76 15% C. pneumoniae 61 12% Both C. pneumoniae and M. 11 2.1% pneumoniae S. pneumoniae 30 5.9% H. influenzae 25 4.9% H. parainfluenzae 10 1.9% S. aureus 6 1.1% Moraxella catarrhalis 6 1.1% Streptococcus species 5 0.9% Other 14 2.8% in all studies including the present one M. pneumoniae was the most commonly identified pathogen, accounting for 17 34% of all cases. M. pneumoniae undergoes cyclical variation with peaks every 4 7 years, 11 thus it is not surprising that the rate of M. pneumoniae as a cause of ambulatory CAP will vary from study to study. In the two studies in which C. pneumoniae was sought 7,10 and the current study, 5.3%, 10.7% and 12% were felt to have pneumonia due to this pathogen. In our study 11 patients had dual infection with M. pneumoniae and C. pneumoniae. Whether this represents dual infection or antibody cross reactivity is a question that needs to be addressed. In one study the sensitivity and specificity of the EIA test that we used for the detection of M. pneumoniae IgM and IgG antibodies was 100% 12 while the sensitivity and specificity of the C. pneumoniae test was 92% and 95%, respectively. 13 Bochud et al. 14 carried out a comprehensive study of the etiology of pneumonia in 184 patients with ambulatory pneumonia. An etiological diagnosis was made in 45.9%. S. pneumoniae was most common, followed by M. pneumoniae, Chlamydia species and Influenza A virus. The second important finding in our study is the low rate of bacteremia at 1.4% and the observation that all of the isolates were S. pneumoniae. The population in this study reflects that seen in clinical practice in that most were recruited from physicians offices and some from Emergency Rooms. In a study of 236 patients with lower respiratory tract infections in Nottingham, UK, 37% of whom had radiological evidence of pneumonia only 1 of 219 (0.5%) had a positive blood culture. 15 In Bochud s study of 184 patients only six had S. pneumoniae isolated from the blood. 14 The number of patients with positive blood cultures is not stated in the paper but if these were the only ones with positive blood cultures the positivity rate would be 3%.

64 T.J. Marrie et al. Campbell et al. 16 studied 289 patients with CAP who were seen in an Emergency Room setting and treated on an ambulatory basis. Six (2.1%) had positive blood culturesf4 were S. pneumoniae and 2 Escherichia coli. Three of the six were subsequently admitted to hospital. H. influenzae accounted for 4.9% of the pneumonias in our study. It was uncommon in Bochud s study 14 at 1.8%. In contrast in a study of 610 patients with CAP attending primary care clinics in the United States 12.6% had H. influenzae isolated from sputum. 17 This latter population was characterized by a history of chronic obstructive pulmonary disease in 18% or current tobacco smokers at 34%. Patients with chronic obstructive pulmonary disease frequently carry H. influenzae in their respiratory secretions. 18 Unfortunately, there have been no studies of H. influenzae pneumonia that define the risk factors associated with certain patient characteristics. 19 Haemophilus parainfluenzae accounted for 4% of the cases of pneumonia. This microorganism has not been reported with this frequency in many other studies of CAP. 1 3 Haemophilus bacteria constitute approximately 10% of the normal bacterial flora of the healthy upper respiratory tract and H. parainfluenzae accounts for 75% of the Haemophilus flora both in the oral cavity and in the pharynx but is absent from the nasal cavity. 20 Thus it is possible that the isolation of H. parainfluenzae represent colonization rather than infection. However, there are well documented cases of H. parainfluenzae pneumonia. 21,22 In a recent study carried out in 15 countries H. parainfluenzae accounted for 2.2% of the pneumonias and 10% of the patients who had sputum submitted for culture. 23 Staphylococcus aureus was the fifth most common cause of pneumonia in our study accounting for 2.5% of the cases. In Hoeffken s study 23 it accounted for about 1% of the cases and ranked in the 9th position. In three other studies S. aureus was not mentioned as a cause of ambulatory CAP. 8,14,15 Nine patients had aerobic gram negative bacteria isolated from sputum and using our criteria these agents were felt to be the cause of the pneumonia. Pneumonia due to these bacteria is usually hospital acquired 24 or is present in the setting of aspiration in elderly patients. 25,26 Any severe illness predisposes to colonization of the oropharynx with aerobic gram negative bacilli 27 thus the recovery of these organisms may represent colonization of the upper airway in patients who were ill for several days prior to presentation. In a review of the changes in etiology of pneumonia that have occurred over the past two decades Kayser noted that Enterobacteriaceae and S. aureus are now seen more frequently as causes of CAP. 28 We performed serological studies for M. pneumoniae and C. pneumoniae only. Studies that performed a battery of serological tests have found that Influenza viruses A and B, respiratory syncytial virus, adenovirus, Coxiella burnetii, C. psittaci, and Legionella species account for a small but definite number of cases of CAP treated on an ambulatory basis. 7,8,14 Thus more comprehensive testing may have allowed us to diagnose 10 15% more of the cases of pneumonia, leaving 35 40% of the cases as of unknown etiology. The issue of antimicrobial resistance is of concern when choosing therapy on an empiric basis. Zhanel et al. 29 collected a total of 6991 unique patient isolates of S. pneumoniae from October 1997 to June 2002 from 25 medical centers in 9 of the 10 Canadian provinces. They found that 20.2% were penicillin non-susceptible, with 14.6% being penicillin intermediate (MIC, 0.12 1 mg/ml) and 5.6% being penicillin resistant (MIC, 4/ ¼ 2 mg/ ml). Almost 10% of the isolates were resistant to macrolides. Overall resistant to the quinolones was low with 0.3% of the strains being resistant to moxifloxacin and 0.6% resistant to levofloxacin and gatifloxacin. 29 These rates increased to 1.2% for the penicillin non-susceptible strains. The same authors also tested 7566 unique patient isolates of H. influenzae and 2314 unique patient isolates of M. catarrhalis and found that 1.6% of the H. influenzae were resistant to clarithromycin and 0.1% to moxifloxacin. 30 All of the M. catarrhalis isolates were susceptible to both moxifloxacin and clarithromycin. In our study all of the bacteremic S. pneumoniae isolates were susceptible to clarithromycin and moxifloxacin. Our study has several strengths and limitations. The strengths are the large number of patients studied and enrollment from multiple sites in Canada. The limitations are that we did not test urine samples for Legionella and S. pneumoniae antigens. We also did not do a comprehensive serological work-up. In conclusion, we have shown that despite considerable efforts an etiological diagnosis of community-acquired pneumonia in an ambulatory setting is made in only half the patients. References 1. Bates JH, Campbell GD, Barron AL, et al. Microbial etiology of acute pneumonia in hospitalized patients. Chest 1992; 101:1005 12.

Etiology of CAP 65 2. Fang GD, Fine M, Orloff J, et al. New and emerging etiologies for community-acquired pneumonia with implications for therapy: a prospective multi-center study of 359 cases. Medicine 1990;69:307 16. 3. Marrie TJ, Durant H, Yates L. Community-acquired pneumonia requiring hospitalization: a 5 year prospective study. Rev Infect Dis 1989;11:586 99. 4. Niederman MS, Mandell LA, Anzueto A, et al. Guidelines for the management of adults with community-acquired pneumonia (American Thoracic Society). Am J Respir Crit Care Med 2001;163:1730 54. 5. Fine MJ, Stone RA, Singer DE, et al. Processes and outcomes of care for patients with community-acquired pneumonia. Results from the pneumonia patient outcomes research team (PORT) cohort study. Arch Intern Med 1999;159: 970 80. 6. Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to identify low-risk patients with community-acquired pneumonia. NEJM 1997;336:243 50. 7. Marrie TJ, Peeling RW, Fine MJ, et al. Ambulatory patients with community-acquired pneumonia: the frequency of atypical agents and clinical course. Am J Med 1996; 101:508 15. 8. Berntsson E, Lagergard T, Strannegard O, et al. Etiology of community-acquired pneumonia in outpatients. Eur J Clin Microbiol 1986;5:446 7. 9. Erard PH, Moser F, Wenger A, et al. Prospective study on community-acquired pneumonia diagnosed and followed up by private practitioners. CHUV, Lusanne, Switzerland. 1991; Abstract no. 56. ICAAC, American Society of Microbiology, Washington, DC. 10. Langille DB, Yates L, Marrie TJ. Serological investigation of pneumonia as it presents to the physician s office. Can J Infect Dis 1993;4:328 32. 11. Foy HM, Kenny GE, Cooney MK, Allan ID. Long-term epidemiology of infections with Mycoplasma pneumoniae. J Infect Dis 1979;139:681 7. 12. Tuuminen T, Suni J, Kleemola M, Jacobs E. Improved sensitivity and specificity of enzyme immunoassays with P1-adhesin enriched antigen to detect acute Mycoplasma pneumoniae infection. J Microbiol Methods 2001;44:27 37. 13. Persson K, Boman J. Comparison of five serologic tests for diagnosis of acute infections by Chlamydia pneumoniae. Clin Diag Lab Immunol 2000;7:739 44. 14. Bochud PY, Moser F, Erad P, et al. Community-acquired pneumonia: a prospective outpatient study. Medicine 2001;80:75 87. 15. Woodhead MA, Macfarlane JT, McCracken JS, et al. Prospective study of the aetiology of pneumonia in the community. Lancet 1987;1:671 4. 16. Campbell SG, Anstey R, Ackroyd-Stolarz S, et al. Utility of blood cultures in the outpatient management of communityacquired pneumonia. J Emerg Med 2003;20:521 3. 17. Gotfried MH. Epidemiology of clinically diagnosed community-acquired pneumonia in the primary care setting: results from the 1999 2000 respiratory surveillance program. Am J Med 2001;111:25S 9S. 18. Murphy TF, Sethi S. Bacterial infection in chronic obstructive pulmonary disease. Am Rev Respir Dis 1992;146:1067 83. 19. Strausbaugh LJ. Haemophilus influenzae. In: Marrie TJ, editor. Community-acquired pneumonia. New York: Kluwer Academic/Plenum Publishers; 2001. 20. Kilian M. Haemophilus. In: Murray PR, Barron EJ, Jorgensen JH, Pfaller MA, Yolken RH, editors. Manual of clinical microbiology, 8th ed. Washington, DC: ASM Press; 2003. 21. Pillai A, Mitchell JL, Hill SL, et al. A case of Haemophilus parainfluenzae pneumonia. Thorax 2000;55:623 4. 22. Feldman C, Smith C, Kaka S, et al. The clinical significance of Haemophilus influenzae and H. parainfluenzae isolated from the sputum of adult patients at an urban general hospital. SAMJ 1992;81:495 511. 23. Hoeffken G, Meyer HP, Winter J, et al. The efficacy and safety of two oral moxifloxacin regimens compared to oral clarithromycin in the treatment of community-acquired pneumonia. Respir Med 2001;95:553 64. 24. Parodi S, Bedwell Goetz M. Aerobic gram negative bacillary pneumonia. Curr Infect Dis Rep 2002;4:249 56. 25. El-Solh AA, Pietrantoni C, Bhat A, et al. Microbiology of sever aspiration pneumonia in institutionalized elderly. Am J Respir Crit Care Med 2003;167:1650 4. 26. Millns B, Gosney M, Jack CI, et al. Acute stroke predisposes to oral gram-negative bacillifa cause of aspiration pneumonia? Gerontology 2002;49:173 6. 27. Johanson WG, Woods DE, Chaudri TR. Association of respiratory tract colonization with adherence of gramnegative bacilli to epithelial cells. J Infect Dis 1979;139: 667 73. 28. Kayser FH. Changes in the spectrum of organisms causing respiratory tract infections: a review. Postgrad Med J 1992;68(Suppl 3):S17 23. 29. Zhanel GG, Palatnick L, Nichol KA, et al. Antimicrobial resistance in respiratory tract Streptococcus pneumoniae isolates: results of the Canadian Respiratory Organism Susceptibility Study, 1997 to 2002. Antimicrob Agents Chemother 2003;47:1867 74. 30. Zhanel GG, Palatnick L, Nichol KA, et al. Antimicrobial resistance in Haemophilus influenzae and Moraxella catarrhalis respiratory tract isolates: results of the Canadian Respiratory Organism Susceptibility Study, 1997 to 2002. Antimicrob Agents Chemother 2003;47:1875 81.