Clinical Consultation: Legionella pneumonia: The value of clinical and laboratory findings

What findings are most useful in the diagnosis of Legionella pneumonia?

In the October 2004 issue of The Journal of Respiratory Diseases, Morrison and Gupta¹ reviewed the clinical and laboratory approaches to the diagnosis ofcommunity- acquired pneumonia (CAP) caused by Legionella. They discussed the advantages and limitations of culture, direct fluorescent antibody (DFA) staining, serology, polymerase chain reaction, and Legionella urinary antigen assays. As the authors noted, DFA staining of respiratory secretions is an underused test that has a high specificity in patients with untreated Legionnaires disease.

The Legionella urinary antigen test is commercially available and represents an advance in diagnostic testing. The main problem with this test is that the results may not be positive when the patient presents--which is when the results would be most useful. The advantage of Legionella urinary antigen assays is that antigenuria persists for weeks, which can be helpful in making a retrospective diagnosis. The current urinary antigen assays for Legionella detect only Legion-ella pneumophila (serogroup 1) strains. Fortunately, most cases of Legionella CAP are caused by this serotype.

In the literature, clinical criteria have failed to distinguish typical from atypical pathogens and have been unsuccessful in identifying Legionella in patients with CAP.^2-4 Such studies have compared individual findings and have found them to be of equivalent diagnostic importance.

The Winthrop-University Hospital (WUH) diagnostic point system represents the first attempt to use a weighted point scale for clinicians to assess the probability of Legionella infectionin patients with CAP.⁵ This system assigns different diagnostic weight to each clinical finding and combines nonspecific findings to increase diagnostic specificity. If the results indicate high probability, then specific testing for Legionella is indicated.

The WUH diagnostic system was tested in a study of patients with Legionnaires disease (defined by a positive urinary antigen test for L pneumophila), who were compared with patients with pneumococcal bacteremia and CAP.⁶ The overall sensitivity of the WUH system was 78%, and the specificity was 65%, which was better than anticipated. The investigators concluded that the sensitivity of the WUH diagnostic criteria was not high enough to exclude Legionella and to withhold empiric anti-Legionella therapy.

The value of the WUH diagnostic system is its ability to differentiate Legionella from other causes of CAP, not to exclude Legionella.⁶The WUH criteria were not meant to be a basis for treatment decisions but were designed to help clinicians assess the likelihood of Legionella CAP, prompting them to proceed with definitive diagnostic testing when indicated.

Studies comparing clinical or laboratory features of Legionella pneumoniawith those of CAP caused by other organisms describe much overlap and nonspecificity, leading to the incorrect conclusion that these findings have no diagnostic value.^2-4 The strength of the WUH system is that it combines clinical and laboratory findings and assigns more diagnostic weight to some findings than to others.⁵

Clinicians should suspect atypical pneumonia in patients with CAP who have extrapulmonary findings. CAP caused by Legionella may be differentiated on a clinical basis from that caused by Chlamydiapneumoniae and Mycoplasma pneumoniae by the presence of relative bradycardia or CNS findings, which are not features of either M pneumoniae or Cpneumoniae CAP. Hepatic involvement (mild, transiently elevated serum transaminase levels) and renal involvement (microscopic hematuria) are frequently associated with Legionella CAP but are rarely, if ever, features of CAP caused by Cpneumoniae or M pneumoniae. Clinicians often overlook the early mild/transient elevations of serum transaminase levels in patients with Legionnaires disease.

Hyponatremia is more common in CAP caused by Legionella than in CAP caused by M pneumoniae or C pneumoniae, but it is less specif-ic than early mild/transient hypophosphatemia. In patients with CAP and extrapulmonary findings (atypical CAP), microscopic hematuria or elevated serum creatine phosphokinase (CPK) levels suggest the possibility of Legionnaires disease.⁷ These nonspecific laboratory findings are not found with other atypical forms of CAP (Q fever, psittacosis, tularemia, and pneumonia caused by M pneumoniae or Cpneumoniae)or with typical causes of CAP (Streptococcus pneumoniae, Haemophilus influenzae, Moraxella catarrhalis).^8,9

Legionella CAP can range from mild to severe. If the possibility of Legionella is considered only in patients with severe CAP admitted to the ICU, the diagnosis may be missed in less critically ill patients.The characteristic combination of nonspecific laboratory test results may suggest the possibility of Legionella infection that would otherwise have been missed or not considered in the differential diagnosis of CAP.

In my experience, relative bradycardia is a typical feature of Legionnaires disease but not of CAP caused by C pneumoniae or M pneumoniae. Relative bradycardia may occur with Q fever or psittacosis, but it does not generally occur with CAP caused by typical pathogens, such as S pneumoniae.¹⁰

Even if relative bradycardia is overlooked, and neurologic (headache, confusion) or GI (watery diarrhea) symptoms and signs are not present, the characteristic pattern of nonspecific laboratory test abnormalities should suggest the diagnosis of Legionnaires disease. The presence of highly elevated lactate dehydrogenase levels or cold agglutinin titers argues strongly against the diagnosis of Legionnaires disease. In patients with CAP and extrapulmonary manifestations, with otherwise unexplained minimal increases in serum transaminase levels, a low-normal or slightly decreased serum phosphorus level, an elevated CPK level, or microscopic hematuria, the diagnosis of Legionella CAP should be considered and the clinician should order specific Legionella testing.^5,7

In my opinion, recommending Legionella urinary antigen testing for all patients with CAP who require hospitalization is an approach that would eliminate clinical reasoning from the diagnostic process. Clinicians should appreciate the diagnostic clues provided by nonspecific laboratory abnormalities that suggest the possibility of Legionella and prompt appropriate testing. These nonspecific laboratory findings may be the only clue suggesting Legionella CAP, particularly in hospitalized patients with mild to moderate CAP.

In severely ill patients in the ICU, the WUH clinical criteria are useful for assessing the possibility of Legionella CAP. Most patients with severe pneumonia in the ICU do not have Legionella CAP, and the WUH system may be used to select patients for Legionella testing. The challenge is not what to treat the patient with once you know the diagnosis--or even how to make the diagnosis--but rather when to suspect the diagnosis.

References:

REFERENCES

1. Morrison AR, Gupta SK.

Legionella

pneumonia: assessing the latest diagnostic tests.

J Respir Dis.

2004;25:404-417.
2. Sopena N, Sabria-Leal M, Pedro-Botet ML, et al. Comparative study of the clinical presentation of

Legionella

pneumonia and other common pneumonias.

Chest.

1998;113:1195-2000.
3. Mulazimoglu L, Yu VL. Can Legionnaires' disease be diagnosed by clinical criteria? A critical review.

Chest.

2001;120:1049-1053.
4. Sopena N, Pedro-Botet ML, Garcia-Pares D, et al. Comparative study of community-acquired pneumonia caused by

Streptococcus pneumoniae, Legionella pneumophila,

or

Chlamydia pneumoniae. Scand J Infect Dis.

2004;36:330-334.
5. Cunha BA. Clinical features of Legionnaires' disease.

Semin Respir Infect.

1998;13:116-127.
6. Gupta SK, Imperiale TF, Sarosi GA. Evaluation of the Winthrop-University Hospital criteria to identify

Legionella

pneumonia.

Chest.

2001;120:1064-1071.
7. Cunha BA. Diagnostic significance of nonspecific laboratory abnormalities in infectious diseases. In: Gorbach SL, Bartlett JG, Blacklow NR, eds.

Infectious Diseases.

Philadelphia: Lippinocott Williams & Wilkins; 2004:158-165.
8. Bochud PY, Moser F, Erard P, et al. Community-acquired pneumonia. A prospective outpatient study.

Medicine (Baltimore).

2001;80:75-87.
9. Cunha BA. Hepatic involvement in

Mycoplasma pneumoniae

community-acquired pneumonia.

J Clin Microbiol.

2003;41:3456-3457.
10. Cunha BA. Diagnostic significance of relative bradycardia.

Clin Microbiol Infect Dis.

2000;6:633-634.