Legionellosis

Background: Erythromycin has been the treatment of choice for Legionnaires disease (LD). However, treatment failure and experimental evidence of its bacterlostatic effect have led to evaluation of new drugs such as fluoroquinolones. This study compared the evolution of patients with LD treated with macrolides and fluoroquinolones.

Methods: A prospective observational study was performed, and 130 patients from three centers were included. Diagnoses were made using Legionella urinary antigen assay in all patients. Patients receiving any antibiotic > 36 h before starting the study therapy were excluded. Group 1 included 76 patients who received macrolides (33 patients with erythromycin and 43 patients with clarithromycln), and group 2 included 54 patients treated with fluoroqninolones (50 patients with levofloxacin and 4 patients with ofloxacin).

Results: No significant differences were seen between the two groups regarding age, sex, smoking, alcohol intake, underlying diseases, or community/hospital acquisition. The time from onset of LD symptoms until the initiation of antibiotic treatment was 78.5 h and 92.7 h in groups 1 and 2, respectively (p = 0.1). Time to apyrexia was significantly longer in the macrolide group (77.1 h vs 48 h for groups 1 and 2, respectively; p = 0.000). There were no differences according to radiology, clinical complications, or mortality. Nevertheless, a trend to a longer hospital stay was observed in the macrolide group (9.9 days vs 7.6 days in groups 1 and 2, respectively; p = 0.09).

Conclusions: Fluoroqninolones were as effective as erythromycin in the treatment of LD. It is of note that time to apyrexia was significantly shorter and hospital stay tended to be shorter in patients receiving fluoroquinolones.

Key words: fluoroquinolones; Legionella; Legionnaires disease; macrolides; pneumonia

Abbreviations: AUC = area under concentration 24-h time curve; LD = Legionnaires disease; MIC = minimum inhibitory concentration

**********

Erythromycin has been the treatment of choice for Legionnaires disease (LD) since a retrospective study (1) of the epidemic outbreak in Philadelphia in 1976 showed a significantly lower death rate in patients treated with this antibiotic. Nevertheless, reports of treatment failure with erythromyein, the interference of this antibiotic with the metabolism of numerous drugs, especially in imnmnosuppressed patients, as well as the appearance of secondary effects with the use of high doses have led to the evaluation of new antibiotics in the treatment of this infection. (2-4) Despite good extracellular and intracellular activity, (5,6) the effect of clarithromycin is still bacteriostatic against Legionella. (7,8)

Legionella is an intracellular pathogen, and antibiotics with adequate intracellular penetration are more likely to be efficacious. Fluoroquinolones achieve high intracellular levels and have a lower minimum inhibitory concentration (MIC) against Legionella than erythromyein. (9,10) Moreover, experimental studies (11-14) have demonstrated that these antibiotics irreversibly inhibit the multiplication of this microorganism at an intracellular level. Some clinical observations (15,16) have shown great efficacy with fluoroquinolones in the treatment of this infection, particularly in severe illness or immunosuppressed patients.

Fluoroquinolones and macrolides have been indistinctly used in many hospitals as the standard treatment of LD. This study evaluated different variables related to the clinical evolution of Legionella infection in patients with LD treated with fluoroquinolones and macrolides.

MATERIALS AND METHODS

The study was observational. Cases of LD from Hospital Universitario Germans Trias i Pujol, in Badalona, Spain (n = 53); Virgen de la Arriaxaca Hospital, in Murcia, Spain (n = 68); and Nostra Senvora de Meritxell Hospital, in Escaldes, Principality of Andorra (n = 9) were included. Data from Virgen de Arriaxaca Hospital were prospectively col letted during the outbreak of LD in Mureia in 2002, (17) and data from Hospital Universitario Germans Trias i Pujol and Nostra Senyora de Meritxell Hospitals were obtained from a Legionella database prospectively collected from 1995 until the present (from 2001 for Nostra Senyora de Meritxell Hospital). All the patients from this last hospital were exclusively treated with clarithromycin.

In order to exclude additional pathogens, blood culture and urinary Streptococcus pneumoniae antigen test results were negative in all the patients. Sputum cultures were performed in only a few patients because of the difficulty to obtain a good specimen and the retrospective nature of this study. Results of serologic tests in order to rule out Chlamydia psittaci, Chlamydia pneumoniae, Coxiella burnetti, and Mycoplasma pneumoniae infection were recorded from 50 patients (38.4%) and were negative in all.

Inclusion and Exclusion Criteria

Patients receiving either macrolides (erythromycin or clarithromycin) or fluoroquinolones (ofloxacin or levofloxacin) for definitive LD were included in the study. Doses considered appropriate for inclusion were as follows: 500 to 1,000 mg of erythromycin q6h, or 500 mg of clarithromycin q12h by the IV or oral route; ofloxacin at 400 mg or levofloxacin at 500 mg q12h IV until apyrexia, and thereafter 400 mg or 500 nag po qd. The treatment was > 14 days in all the cases. Patients who had received another antibiotic > 36 h before starting the study therapy and those who were treated at doses considered to be inadequate or over a treatment period of < 14 days were excluded from the study.

Selection Groups

One hundred thirty patients fulfilled the inclusion criteria and were classified into two groups. Diagnoses were made in all patients using Legionella pneumophila serogroup 1 urinary antigen assay. Group 1 included 76 patients who had received treatment with erythromycin (n = 33) or with clarithromycin (n = 43), and group 2 included 54 patients treated with fluoro-quinolones (50 with levofloxacin and 4 with ofloxacin).

Group Characteristics

Demographic data and predisposing factors are shown in Table 1. No significant differences were observed between the two groups with respect to these variables.

Data Collected

A protocol for data collection including time to apyrexia, clinical evolution, mortality, and mean hospital stay was applied.

Statistical Methods

The Kolmogorov-Smirnov test and the Levene test were performed to assess population normality and the homogeneity of variances, respectively. Means were compared by the Student t test or the Mann-Whitney U test when appropriate. Proportions were compared using the [chi square] test with Yates correction or Fisher Exact Test when necessary. All p values and confidence intervals were two sided. All confidence interval estimates were 95%.

RESULTS

Comparative data on evolution are shown in Table 2. The mean time to apyrexia in group 1 was 77.1 h (range, 24 to 408 h), and 48 h (range, 24 to 192 h) in group 2 (p = 0.000). The breakdown on duration of IV vs oral therapy was 5.3 days for patients receiving macrolides and 9.9 days for those receiving fluoroquinolones (p = 0.000). The following complications were observed during the evolution of Legionella pneumonia: pleural effusion, empyema, mechanical ventilation, and septic shock. The global incidence of complications was greater in group 1 (23.6%) compared to group 2 (16.6%) [p = 0.4]. Thirteen of 76 patients (17.1%) vs 5 of 54 patients (9.2%) [p = 0.2] presented pleural effusion in groups 1 and 2, respectively. Only one patient in group 1 and no patients in group 2 had pleural empyema. Mechanical ventilation was required in 5 of 76 patients (6.5%), compared to 3 of 54 patients (5.5%) [p = 1] in groups 1 and 2, respectively. In group 1, 6 of 76 patients (7.8%) presented with septic shock, compared to 3 of 54 patients (5.5%) in group 2 (p = 0.7). Cavitation was not observed in either group. The mean length of hospitalization for Legionella pneumonia was 9.9 days in group 1 (range, 2 to 59 days) and 7.6 days in group 2 (range, 1 to 19 days) [p = 0.09]. Finally, the mortality rates were 7.8% in group 1 and 5.5% in group 2 (p = 0.7).

DISCUSSION

Erythromycin has been the treatment of choice in LD for many years. The poor outcome reported in some clinical reports (18-20) in patients with severe LD, especially in immunosuppressed patients, and the numerous side effects observed when using high doses of erythromycin have led to the consideration of new antimicrobial agents. To date, azithromycin, fluoroquinolones, and ketolides are among the most promising alternative antimicrobials to be considered according to experimental data and anecdotal clinical observations. Nevertheless, no comparative studies have been published on the treatment of LD. Furthermore, the relative rarity of LD until the availability of Legionella urinary antigen and its widely variable outcome make an adequate comparative clinical trial of the disease unlikely to be completed. Fluoroquinolones have shown the greatest activity against L pneumophila in experimental models. This group of drugs is very active against intracellular L pneumophila and is more effective than erythromycin in inhibiting the growth of L pneumophila in different intracellular models. (21)

Moreover, many fluoroquinolones kill, rather than just inhibit, intracellular bacterium replication. (12,14,22) Ciprofloxacin, levofloxacin, sparfloxacin, and trovafloxacin were more active than erythromycin in guinea pig models of Legionella pneumonia and Legionella peritonitis. (21) Ofloxacin and levofloxacin have been successfully used in the treatment of LD despite heavy immunosuppression in some of the patients treated. (15,23,24)

To our knowledge, this is the first study comparing macrolides vs fluoroquinolones in the treatment of LD. Assuming the low mortality rate of LD in the community or even in a hospital scenario, a comparative study to detect differences in this variable would probably need hundreds of patients in each study arm. (25) However, to demonstrate differences in some aspects of clinical evolution, the size of the study arms does not necessarily need to be so large. Thus, time to apyrexia was found to be significantly shorter in the group of patients with LD treated with fluoroquinolones. Consequently, the switch from IV to oral therapy was achieved sooner in patients receiving fluoroquinolones. In addition, a trend to a shorter hospital stay, as a consequence of the faster clinical improvement, was observed in patients treated with ofloxacin and levofloxacin.

It is important to point out that the starting doses of fluoroquinolones were double (500 mg twice) those of standard doses or doses recommended by the manufacturer (Sanofi-Aventis; Paris, France) in ease of pneumonia (500 [micro]g/24 h). In the experience of some of the authors, doses of 500 mg q24h do not seem to be as effective as the schedule recommended in this study. However, reports of treatment failure have been described using low doses of ofloxacin (26) or ciprofloxacin (16,27) in LD. Dosing every 12 h should lead the levofloxacin concentration to exceed the MIC for a longer period, improving the ratio of the area under concentration 24-h time curve (AUG) to the MIC (AUC/MIC ratio). For concentration-dependent drugs such as fluoroquinolones, the AUC/MIC ratio is the dynamic variable most clearly linked to the killing rate and outcome. (28) The impact of these pharmacokinetics is reinforced by the experimental data showing a statistically significant decrease in colony forming units per milliliter with increasing concentrations of levofloxacin using a human monocyte model against L pneumophila. (29)

The reason for early improvement in the group treated with fluoroquinolones is unknown, although it is probably related to a more marked decrease in pulmonary bacterial load. Nonetheless, recovery of these patients is faster, and they may be discharged sooner. Even a reduction in hospital stay of only 24 h can greatly reduce health-care costs, thereby justifying a relatively more expensive treatment. Although not evaluated in this study, the rate of adverse effects is probably lower in patients receiving fluoroquinolones than in those receiving macrolides, as suggested by other authors. (21)

However, complications and mortality rates were similar in both groups, which may be expected due to the reasons pointed out previously. However, complications and mortality rate in LD depend on the timing of the administration of appropriate antimicrobial therapy as well as other very important variables, such as the presence of immunosuppression, the severity of the underlying diseases, and the severity of the pneumonia. (24,30-32)

The advantages of choosing macrolides or fluoroquinolones for LD in healthy patients or community-acquired pneumonia may only be those reported in this study: shorter time to apyrexia, and consequently a more rapid achievement of good health status. Nevertheless, in these patients, complications and mortality are probably not influenced by the choice of either type of drug. Thus, each physician should determine whether these advantages are sufficient to indicate treatment with one antibiotic or another.

To the contrary, in immunosuppressed patients or hospital-acquired LD, the experimental data, the anecdotal clinical observations, and the results of our study should be taken into account as a whole. (16,18,33) A comparative study in this subgroup of patients with a worse clinical evolution and greater mortality would probably show significant differences in terms of mortality and complications that would favor the use of fluoroquinolones. The main weakness of this study is that it is not a randomized interventional trial. Nevertheless, we consider that the data available in the literature are, at present, sufficiently important to advise the use of fluoroquinolones when bad evolution of LD is suspected (immunosuppression, severe or hospital-acquired LD). (21,34,35) Nonetheless, the present study is ongoing in the three hospitals participating in order to obtain a larger population sample for further information on the use of these two groups of antibiotics.

APPENDIX

Participating investigators are as follows: Miquel Sabria, MD, PhD (Badalona, Spain); Marfa Luisa Pedro-Botet, MD, PhD (Badalona, Spain); Joaquin Gomez, MD, PhD (Murcia, Spain); Jorge Roig, MD, PhD (Escaldes, Principality of Andorra); Blanca Vilaseca, MD (Badalona, Spain); Nieves Sopena, MD, PhD (Badalona, Spain); Victor Banos, MD, PhD (Murcia, Spain); Esteban Reynaga, MD (Badalona, Spain); Marian Gracia Nunez, Bsc (Badalona, Spain); Manuel Lorenzo, MD (Murcia, Spain); Javier Casal, MD (Escaldes, Principality of Andorra); Nicolas Ortega, MD (Murcia, Spain); Joaquin Ruiz, MD (Murcia, Spain) and Pere Tudela, MD (Badalona, Spain).

REFERENCES

(1) Meyer RD, Finegold SM. Legionnaires' disease. Annu Rev Med 1980; 31:219-232

(2) Hays JH, Hinthorn DR, Chonko A, et al. Failure of oral erythromycin therapy for Legionnaires' disease in a renal transplant recipient. South Med J 1981; 74:1422-1423

(3) Mercatello A, Frappaz D, Robert D, et al. Failure of erythromycin/rifampicin treatment of Legionella pneumonia. J Infect 1985; 10:282-283

(4) Rudin JC, Evans TL, Wing EJ. Failure of erythromycin in treatment of Legionella micdadei pneumonia. Am J Med 1984; 76:318-320

(5) Stout J, Arnold B, Yu VL. Activity of azithromycin, clarithromycin, roxithromycin, dirithromycin, quinupristine/dalfopristin and erythromycin against Legionella species by intracellular susceptibility testing in HL-60 cells. J Antimicrob Chemother 1998; 41:289-291

(6) Stout J, Arnold B, Yu VL. Comparative activity of ciprofloxacin, ofloxacin, levofloxacin, and erythromycin against Legionella species by broth microdilution and intracellular susceptibility testing in HL-60 cells. Diagn Microbiol Infect Dis 1998; 30:37-43

(7) Dubois J, St-Pierre C. In vitro activity, postantibiotic effect and human monocyte activity of grepafloxacin against Legionella species. Clin Microbiol Infect 1999; 5:205-212

(8) Edelstein PH, Edelstein MAC. In vitro activity of the ketolide HMR 3647 (TU6647) for Legionella spp., its pharmacokinetics in guinea pigs, and use of the drug to treat guinea pigs with Legionella pneumophila pneumonia. Antimicrob Agents Chemother 1999; 43:90-95

(9) Schulin T, Wennersten CB, Ferraro MJ, et al. Susceptibilities of Legionella spp. to newer antimicrobials in vitro. Antimicrob Agents Chemother 1998; 42:1520-1523

(10) Walz A, Nichterlein T, Hof H. Excellent activity of newer quinolones on Legionella pneumophila in J 774 macrophages. Zentralbl Bakteriol 1997; 285:431-439

(11) Edelstein PH, Edelstein MAC. WIN 57273 is bactericidal for Legionella pneumophila grown in alveolar macrophages. Antimicrob Agents Chemother 1989; 33:2132-2136

(12) Edelstein PH, Edelstein MAC, Lehr KH, et al. In vitro activity of levofloxacin against clinical isolates of Legionella spp., its pharmacokinetics in guinea pigs, and use in experimental Legionella pneumophila pneumonia. J Antimicrob Chemother 1996; 37:117-126

(13) Saito A, Sawatari K, Fukuda Y, et al. Susceptibility of Legionella pneumophila to ofloxacin in vitro and in experimental Legionella pneumonia in guinea pigs. Antimicrob Agents Chemother 1985; 28:15-20

(14) Stout JE, Arnold B, Yu VL. Comparative activity of ciprofloxacin, ofloxacin, levofloxacin, and erythromycin against Legionella species by broth microdilution and intracellular susceptibility testing in HL-60 cells. Diagn Microbiol Infect Dis 1998; 30:37-43

(15) Wynkel A, Toupance O, Melin JP, et al. Traitement des legionelloses par ofloxacine chez le transplant renal: abscence d'interference avec la ciclosporine A. Presse Med 1991; 20:291-293

(16) Unertl KE, Lenhart FP, Forst H, et al. Ciprofloxacin in the treatment of legionellosis in critically ill patients including those cases unresponsive to erythromycin. Am J Med 1989; 87(Suppl):128S-131S

(17) Barrio M, Ramos M, Page MA, et al. Brote epidemico de neumonia por Legionella pneumophila en Murcia 2001. Rev Clin Esp 2002; 202(Suppl):126-127

(18) Fernandez-Aviles F, Batlle M, Ribera JM, et al. Legionella spp. pneumonia in patients with hematologic diseases: a study of 10 episodes from a series of 67 cases of pneumonia. Hematologica 1999; 84:474-475

(19) Stout JE, Yu VL. Current concepts: legionellosis. N Engl J Med 1997; 337:682-687

(20) Parker MM, Macher AM, Shelhamer JH, et al. Unresponsiveness of Legionella bozemanii pneumonia to erythromycin administration despite in vitro susceptibility. Am Rev Respir Dis 1983; 128:955-956

(21) Sabria M, Yu VL. Legionella species (Legionnaires' disease). In: Yn VL, Weber R, Raoult D, eds. Antimicrobial therapy and vaccines. 2nd ed. New York, NY: Apple Trees Productions, 2002; 395-417

(22) Manzor O, Majithia D, Saravolatz L. In vitro activities of seven quinolones and two macrolides against Legionella isolates [abstract no E2284]. In: Program and Abstracts of the 1999 International Conference on Antimicrobial Agents and Chemotherapy. San Francisco, CA: American Society for Microbiology, 1999

(23) Sanchez-Conde M, Perez-Molina JA, Moreno L, et al. Treatment with levofloxacin of community-acquired pneumonia caused by Legionella spp [abstract no P846]. In: Program and Abstracts of the 2001 European Congress of Clinical Microbiology and Infectious Diseases. Istanbul, Turkey: European Society of Clinical microbiology and Infectious Diseases, 2001

(24) Pedro-Botet ML, Sabria M, Sopena N, et al. Legionnaires' disease and HIV infection. Chest 2003; 124:543-547

(25) Edelstein PH. Antimicrobial chemotherapy for Legionnaires' disease: a review. Clin Infect Dis 1995; 21:S265-S276

(26) Salord JM, Matsiota-Bernard P, Staikowsky F, et al. Unsuccessful treatment of Legionella pneumophila infection with a fluoroquinolone. Clin Infect Dis 1993; 17:518-519

(27) Kurz RW, Graninger W, Egger TP, et al. Failure of treatment of Legionella pneumophila with ciprofloxacin. J Antimicrob Chemother 1988; 22:389-391

(28) Drusano G, Lebro MT, Cars O, et al. Pharmacokinetics and pharmacodynamics of fluoroquinolones. Clin Microbiol Infect 1998; 4:2S27-2S41

(29) Baltch AL, Smith RP, Franke MA, et al. Antibacterial effects of levofloxacin, erythromycin, and rifampin in a human monocyte system against Legionella pneumophila. Antimicrob Agents Chemother 1998; 42:3153-3156

(30) Pedro-Botet ML, Sabria-Leal M, Sopena N, et al. Role of immunosuppression in the evolution of Legionnaires' disease. Clin Infect Dis 1998; 26:14-19

(31) Heath CH, Grove DI, Looke DF. Delay in appropriate therapy of Legionella pneumonia associated with increased mortality. Eur J Clin Microbiol Infect Dis 1996; 15:286-290

(32) Gacouin A, Le Toulzo Y, Lavoue S, et al. Severe pneumonia due to Legionella pneumophila: prognostic factors, impact of delayed appropriate antimicrobial therapy. Intensive Care Med 2002; 28:686-691

(33) Dournon E, Mayaud C, Wolff M, et al. Comparison of the activity of three antibiotic regimens in severe Legionnaires' disease. J Antimicrob Chemother 1990; 26(Suppl):129-139

(34) Edelstein PH. Antimicrobial chemotherapy for Legionnaires' disease: time for a change. Ann Intern Med 1998; 129:328-330

(35) Roig J, Rello J. Legionnaires' disease: a rational approach to therapy. J Antimicrob Chemother 2003; 51:1119-1129

* From the Infectious Diseases Unit (Drs. Sabria, Pedro-Botet, Vilaseea, and Sopena), Hospital Universitari Germans Trias i Pujol, Universitat Autonoma de Barcelona, Badalona, Spain; Infectious Diseases and Pneumology Units (Drs. Gomez and Banos), Hospital Virgen de la Arriaxaca, Universidad de Murcia, Murcia, Spain; and Pulmonary Division (Dr. Roig), Hospital Nostra Senyora de Meritxell, Escaldes Principality of Andorra.

Manuscript received June 2, 2004; revision accepted February 10, 2005.

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml).

Correspondence to: Miquel Sabria, MD, PhD, Unitat de Malalties Infeccioses, Hospital Universitari Germans Trias i Pujol, Carretera del Canyet s/n, Badalona 08916, Barcelona, Spain; e-mail: msabria@ns.hugtip.scs.es

Miquel Sabria, MD, PhD; Maria Luisa Pedro-Botet, MD, PhD; Joaquln Gomez, MD, PhD; Jorge Roig, MD, PhD; Blanca Vilaseca, MD; Nieves Sopena, MD, PhD; and Victor Banos, MD, PhD; for the Legionnaires Disease Therapy Group ([dagger])

([dagger]) A list of participants is given in the Appendix.

COPYRIGHT 2005 American College of Chest Physicians
COPYRIGHT 2005 Gale Group