Legionella.org2

Levofloxacin Efficacy in the Treatment
of Community-Acquired Legionellosis*
Victor L. Yu, MD; Richard N. Greenberg, MD; Neringa Zadeikis, MD, MBA;Janet E. Stout, PhD; Mohammed M. Khashab, BS; William H. Olson, PhD; andAlan M. Tennenberg, MD, MPH Background: Although fluoroquinolones possess excellent in vitro activity against Legionella, few
large-scale clinical trials have examined their efficacy in the treatment of Legionnaires disease.
Even fewer studies have applied rigorous criteria for diagnosis of community-acquired Legion-
naires disease, including culture of respiratory secretions on selective media.
Methods: Data from six clinical trials encompassing 1,997 total patients have been analyzed to
determine the efficacy of levofloxacin (500 mg qd or 750 mg qd) in treating patients with
community-acquired pneumonia (CAP) due to Legionella.
Results: Of the 1,997 total patients with CAP from the clinical trials, 75 patients had infection with
a Legionella species. Demographics showed a large portion of these patients were < 55 years of
age and nonsmokers. More than 90% of mild-to-moderate and severe cases of Legionella
infection resolved clinically at the posttherapy visit, 2 to 14 days after treatment termination. No
deaths were reported for any patient with Legionnaires disease treated with levofloxacin during
the studies.
Conclusions: Levofloxacin was efficacious at both 500 mg for 7 to 14 days and 750 mg for 5 days.
Legionnaires disease is not associated only with smokers, the elderly, and the immunosuppressed,
but also has the potential to affect a broader demographic range of the general population than
previously thought.
(CHEST 2004; 125:2135–2139)
Key words: clinical efficacy; community-acquired pneumonia; Legionella; levofloxacin
Abbreviations: BCYE ϭ buffered charcoal yeast extract; CAP ϭ community-acquired pneumonia; ELISA ϭ enzyme-
linked immunosorbent assay; MIC ϭ minimum inhibitory concentration; PSI ϭ pneumonia severity index Legionella are facultative, intracellular bacteria among immunocompetent hosts, and even higher
that penetrate and proliferate within the phago- somes of alveolar macrophages and blood mono- In the past, erythromycin was typically used for cytes.1–3 Thus, antimicrobials that cannot penetrate the treatment of Legionnaires disease. However, the host’s cellular membrane, such as penicillins and newer antimicrobials have been shown to be safer cephalosporins, are relatively ineffective againstthese bacteria.4 Legionella is a major cause of lethal For editorial comment see page 1979
pneumonia, and mortality rates range from 5 to 25% and more effective.6 The Infectious Diseases Societyof America and the American Thoracic Society cur- *From the VAMC and University of Pittsburgh (Drs. Yu andStout), Pittsburgh, PA; Department of Veterans Affairs Medical rently recommend doxycycline, azithromycin, and Center, University of Kentucky Medical Center (Dr. Greenberg), various fluoroquinolones for treating Legionella re- Lexington, KY; and Ortho-McNeil Pharmaceutical, Inc. (Drs.
Zadeikis, Olson, Tennenberg, and Mr. Khashab), Raritan, NJ.
Drs. Yu and Stout have received research funding from Ortho- Levofloxacin, one of the recommended fluoro- McNeil Pharmaceutical, Inc., but not for this study.
quinolones, has excellent in vitro activity against Drs. Tennenberg, Olson, Zadeikis, Kahn, and Mr. Khashab areemployed by Ortho-McNeil Pharmaceutical, Inc.
Legionella, with a minimum inhibitory concentration Reproduction of this article is prohibited without written permis- (MIC) equal to 0.03 ␮g/mL.9 In vitro studies3,10,11 sion from the American College of Chest Physicians (e-mail: using intracellular systems have shown that levo- [email protected]).
Manuscript received July 10, 2003; revision accepted December floxacin is effective in intracellular killing of various Legionella strains. In addition, levofloxacin concen- Correspondence to: Neringa Zadeikis, MD, MBA, Ortho-McNeil trates in the intrapulmonary compartments at levels Pharmaceutical, Inc., 1000 Route 202, Room 3121, Raritan, NJ08869-0602; e-mail: [email protected] well above the MIC.12 Steady-state concentrations of levofloxacin in the epithelial lining fluid and alveolar resistant to a study drug prior to the patient’s entry into the study, macrophages are significantly higher than the plasma (2) a diagnosis of cystic fibrosis or fungal infection, (3) empyema, concentrations.12 The ratio of the area under the (4) HIV infection and CD4 counts Ͻ 200 cells/␮L, (5) neutro-penia (Ͻ 500 cells/␮L), (6) hospital-acquired infections, concentration-time curve or peak concentration to (7) requirement of an additional nonstudy systemic antimicrobial MIC is assumed to be a key pharmacodynamic agent, (8) a history of seizures or a major psychiatric disorder, indication for clinical and microbiologic success of (9) a history of allergy to a study drug or drugs, (10) pregnancy or the fluoroquinolones.13,14 In addition, unlike the breast feeding, (11) severe renal impairment (creatinine clear- macrolides, which are bacteriostatic for Legionella, ance Ͻ 20 mL/min [creatinine clearance Ͻ 50 mL/min for study5]), or (12) any investigational agent within 30 days of entry into Legionnaires disease cannot be readily diagnosed Patients in studies 1, 2, and 3 were classified as having severe because of its nonspecific clinical and radiologic pneumonia on the basis of one or more of the following: manifestations, and the need for specialized labora- bacteremia, hypotension (diastolic BP Ͻ 60 mm Hg) in the tory tests.16 Culture and urinary antigen are consid- absence of volume depletion, altered mental status, baselinerespiratory rate exceeding 30 breaths/min, or a requirement for ered the most specific tests for diagnosis. In the intubation or mechanical ventilation. Patients in study 4 were prospective analysis of levofloxacin presented herein, considered to have a high risk of mortality and had to have at least efficacy and toxicity data are presented for what we three American Thoracic Society criteria for hospital admission,8 believe to be the largest population of cases of and either mechanical ventilation or two of the following: ele- sporadic community-acquired Legionnaires disease vated temperature (oral Ն 39°C) or hypothermia (oral Յ 35°C),respiratory rate Ͼ 30 breaths/min, systolic hypotension (systolic BP Ͻ 90 mm Hg), pulse rate of at least 130 beats/min, or alteredmental status. The inclusion criteria of study 4 limited enrollmentto those patients with severe CAP. In studies 5 and 6, the patient’s PSI score20 determined the severity of the disease.
Mild-to-moderate cases belonged to PSI class I or II (score Յ 70)and were treated as inpatients or outpatients, while severe cases belonged to PSI class III or IV (PSI score Ͼ 70 but Յ 130) andwere hospitalized for at least 24 h.
This analysis integrated data from five prospective phase III clinical trials (studies 1, 2, 3, 5, and 6) and one phase IV clinicaltrial (study 4). Study 1 (data on file; R.W. Johnson Pharmaceu- tical Research Institute; Raritan, NJ) and study 317 were open-label, noncomparative studies that evaluated the efficacy and Patients in studies 1, 2, and 3 received levofloxacin, 500 mg/d safety profiles of levofloxacin in the treatment of CAP. Study 218 for 7 to 14 days, while patients in study 4 received levofloxacin, was an open-label, randomized, active-controlled study that 500 mg/d for 10 to 14 days. (Two patients in study 2 with compared levofloxacin with parenteral ceftriaxone and/or oral Legionnaires disease underwent extended therapies of 15 days cefuroxime axetil in the treatment of CAP. Study 419 was an and 18 days, respectively.) In study 5, patients received levofloxa- open-label, randomized, active-controlled study that compared cin, 500 mg/d, for 10 days or 750 mg for 5 days. Patients in study levofloxacin with ceftriaxone plus erythromycin, followed by 6 received levofloxacin, 750 mg for 5 days. Patients in all trials amoxicillin-clavulanate plus clarithromycin. Additionally, patients were started on either IV or oral therapy, and those receiving IV in study 4 had diagnoses of “serious” CAP; they fulfilled criteria therapy could be switched to oral medication at the investigator’s that predicted a higher probability of death, roughly comparable discretion. Clinical and microbiologic responses were determined to the pneumonia severity index (PSI) class IV population (score at a posttherapy visit occurring 2 to 14 days after cessation of drug of 91 to 130).20 Study 5 was a multicenter, randomized, double- therapy, and at a poststudy visit occurring 3 to 5 weeks after the blind study that compared the efficacy of 5-day, high-dose (750 completion of drug therapy. At posttherapy, a cured or improved mg) levofloxacin therapy with a 10-day course of levofloxacin 500 patient had resolution or partial resolution of clinical signs and mg for CAP.21 Study 6 was a noncomparative, open-label study symptoms associated with active infection, along with improve- that evaluated the efficacy of the 5-day, high-dose levofloxacin ment or stabilization on chest radiograph. Failure indicated an regimen for CAP (data on file; Ortho-McNeil Pharmaceutical, inadequate response to therapy with additional antibiotic treat- ment required for the original infection. Failure was also desig-nated for patients who had a clinical failure but in whom theadmission pathogen appeared to have been eradicated (negative The investigational review board of each center approved the studies, and all patients provided written informed consent. All enrolled patients required a primary diagnosis of CAP confirmedby radiographic evidence. For studies 1, 2, and 3, the clinical The Special Pathogens Laboratory at the VA Pittsburgh signs and symptoms of CAP also included at least two of the Healthcare System, Pittsburgh, PA, performed Legionella cul- following: elevated temperature (oral Ն 38°C, rectal Ն 39°C), ture, urinary antigen, and serologic testing. Cases were diagnosed new or increased cough, production of purulent sputum, rales or as Legionnaires disease based on any of the following: pleuritic chest pain, and shortness of breath. Studies 5 and 6 (1) fourfold increase in the level of IgM or IgG determined by required at least one of the following: fever (oral Ն 38°C) or enzyme-linked immunosorbent assay (ELISA), (2) seroconver- hypothermia (oral Յ 35°C), leukocytosis (WBC count Ͼ 10,000/ sion by the Carter Wallace IgG/IgM ELISA (Carter-Wallace- ␮L), or bands Ͼ 10%. Excluded from these studies were patients Wampole; Cranbury, NJ), (3) positive urinary antigen test result with the following: (1) infections due to organisms known to be using EIA for Legionela pneumophila serogroup 1 (Binax; S.
Portland, ME), or (3) isolation of Legionella from pretreatment Table 1—Clinical Response After Levofloxacin at
sputum cultures. Culture was performed using three media: Posttherapy Visit in the Clinically Evaluable
buffered charcoal yeast extract (BCYE) agar; BCYE with poly- Population of Patients With CAP and Legionella
mixin B, anisomycin, and vancomycin (PAV); and BCYE withpolymixin B, anisomycin, and cefamandole (PAC). Specimens were pretreated with an acid buffer (HCl-KCl, pH 2.2) for 4 min The intent-to-treat population consisted of all patients in the trials who were treated with levofloxacin and who fulfilled the previously mentioned criteria for CAP caused by a Legionella species. The clinically evaluable population was a combination of all Legionella-infected patients in the clinically evaluable popu- lations of the six individual clinical trials.
Patient Demographic and Baseline Characteristics From a total of 1,997 patients with a diagnosis of CAP, 75 patients (3.8%) had Legionella infection *Thirteen patients received 750 mg of levofloxacin for 5 days.
†All five patients survived, but additional antibiotics were prescribed based on at least one of the following criteria: seroconversion (either fourfold increase in IgM orIgG levels determined by ELISA [55 patients] or byWampole EIA [14 patients]), positive urinary anti- 71 patients) overall; 74.6% were cured and 18.3% gen (5 patients), or positive sputum culture (6 pa- were improved. Clinical failure occurred in five tients). The numbers total Ͼ 75 because some pa- patients (7.0%) based on the criteria that additional tients had multiple positive test results. In the antibiotics were prescribed by the physician. None of clinically evaluable population, 71 of 1,551 patients the five clinical failures resulted in patient death.
(4.6%) were determined to have Legionella infection.
Among the 13 patients receiving 750 mg of levofloxa- Patients were stratified according to the severity of cin for 5 days, 12 patients (92.3%) achieved clinical the disease, with 47 patients classified with mild-to- success at posttherapy; 5 of the patients had severe moderate CAP and 24 patients with severe CAP.
Patients were categorized according to whether they During the posttherapy visit, 93.6% of mild-to- had traditional risk factors for contracting Legio- moderate cases were assessed as clinical successes, nella-induced CAP. Forty-six percent (33 of 71 compared with 91.6% of the severe cases. Also, no patients) of the clinically evaluable population had a patient was found to have a documented microbio- chronic respiratory disease, including bronchitis, em- logical relapse during the poststudy visit (3 to 5 physema, sinusitis, and COPD, and 41% (29 of 71 patients) had a history of smoking. Thirty-four per-cent (24 of 71 patients) had none of these risk factors, and over half of these (14 of 24 patients)were also Ͻ 55 years of age.
Adverse effects in 7% (5 of 75 patients) were Seroconversion to Mycoplasma pneumoniae and judged by the investigator to be probably or defi- Chlamydia pneumoniae was observed in 25% (19 of nitely drug related. The most common complaints 75 patients) and 17% (13 of 75 patients), respec- included anxiety, insomnia, headache, nausea, and tively. Fifteen percent (11 of 75 patients) and 9% (7 diarrhea. Most of the reported adverse events were of 75 patients) had Streptococcus pneumoniae and judged to be mild in character by the investigators, Haemophilus influenzae isolated from sputum cul- and no patients discontinued therapy because of Seventy-one patients with community-acquired Legionnaires disease were clinically evaluable, and Levofloxacin was found to be highly effective clinical success (cured plus improved) was observed against Legionella infections, leading to clinical suc- during the posttherapy visit (Table 1) in 92.9% (66 of cess in Ͼ 90% of patients. It should be noted that, while not all patients with Legionnaires disease in However, the demographics of the patients with these six studies fulfilled the criteria for cure, not a Legionella in this large study deviated from the single patient died during the course of hospitaliza- typical patient profiles, especially with Ͼ 70% of the tion and/or treatment. Levofloxacin treatment was as subjects being Ͻ 65 years of age. Although approxi- successful in patients with severe CAP as in those mately half of the patients had a history of respira- with mild-to-moderate disease. During the post- tory illness and more than a fourth had a history of therapy visit, clinical success was seen in 93.6% of smoking, we found that approximately 34% (24 of 71 patients with mild-to-moderate pneumonia, com- patients) did not have either of these risk factors. In pared with 91.6% with severe pneumonia, with no addition, 14 of these 24 patients were also Ͻ 55 years documented microbiologic relapse in either patient of age, providing evidence that Legionnaires disease is not limited to elderly patients with chronic respi- For 13 patients with Legionella, a high-dose (750 ratory illnesses or a history of smoking. None of the mg), short-course (5 day) levofloxacin treatment was patients were known to be receiving immunosup- administered. This regimen was based on the ratio- pressive therapy, which is another key risk factor for nale that higher concentration peaks lead to in- Legionnaires disease. Interestingly, in a retrospec- creased killing of the pathogen, decreased resistance tive review of Legionnaires disease in Allegheny development, and higher patient compliance with County, PA, Squier et al25 also found that 28% of the shorter course. The 750-mg dose increases peak reported cases had none of the classic risk factors for plasma concentration twofold over the 500-mg dose Legionnaires disease. These findings suggest that at steady state, while maintaining a high drug con- testing for Legionnaires disease is warranted in centration in the alveolar macrophages (105.1 patients with CAP with broader demographics than ␮g/mL 4 h after dosing).12 The clinical cure rate was previously appreciated. These data also suggest the 92.3% (12 of 13 patients) at posttherapy; 5 of the need for an agent with intracellular activity when patients had severe pneumonia. Although definitive treating CAP regardless of the presence or absence conclusions cannot be drawn from the limited num- of the typical Legionella risk factors.
ber of patients in this study, a high-dose, short- This is the largest antibiotic study of patients with course therapy warrants scrutiny as a treatment Legionella-induced CAP ever published. Levofloxa- cin proved to be highly efficacious at both the In a meta-analysis of 13 studies of CAP in which an 500-mg and 750-mg doses, and mortality was negli- oral antibiotic could be administered, the respiratory gible. With greater routine use of Legionella culture quinolones showed a modest therapeutic advantage on selective media, it is likely that undiagnosed cases compared with other alternative antibiotics (such as may be uncovered. Moreover, other Legionella se- macrolides, ␤-lactams, and doxycycline)22; 100% (10 rogroups and species not diagnosable by serology of 10 patients) of fluoroquinolone-treated patients and urinary antigen may be identified as the actual with Legionella infection were cured as compared to 33% (4 of 12 patients) treated with a ␤-lactam agent.
In another retrospective observational study15 of 33patients with Legionnaires disease, patients treated with a fluoroquinolone had fewer complications, 1 Roig J, Domingo C, Morera J. Legionnaires disease. Chest more rapid defervescence, and lower mortality than patients treated with erythromycin; the differences, 2 Nguyen MLT, Yu VL. Legionella infection. Clin Chest Med however, were not statistically significant.
Legionella infections account for up to 16% of 3 Baltch AL, Smith RP, Franke MA, et al. Antibacterial effects of levofloxacin, erythromycin, and rifampin in a human cases of CAP, and in numerous observational studies monocyte system against Legionella pneumophila. Antimi- Legionella is among the top four microbial causes of crob Agents Chemother 1998; 42:3153–3156 hospitalization due to CAP.6,23 The classic risk fac- 4 Edelstein PH. Antimicrobial chemotherapy for Legionnaires tors for Legionnaires disease include cigarette smok- disease: a review. Clin Infect Dis 1995; 21(suppl 3):S265– ing, chronic lung disease, and immunosuppression; 5 Marston BJ, Lipman HB, Breiman RF. Surveillance for the disease most frequently occurs in the elderly.6 Legionnaires disease: risk factors for morbidity and mortality.
Based on a large-scale study of CAP in Ohio, the Centers for Disease Control and Prevention (CDC) 6 Stout JE, Yu VL. Legionellosis. N Engl J Med 1997; 337:682– estimated that only 3% of sporadic cases of Legion- naires disease are correctly diagnosed.24 It is proba- 7 Bartlett JG, Dowell SF, Mandell LA, et al. Practice guidelines for the management of community-acquired pneumonia in ble that detection bias occurs such that those pa- adults. Clin Infect Dis 2000; 31:347–382 tients with the classic risk factors are more likely to 8 Niederman MS, Mandell LA, Anzueto A, et al. Guidelines for undergo Legionella laboratory testing.
the management of adults with community-acquired pneu- monia: diagnosis, assessment of severity, antimicrobial ther- community-acquired pneumonia in adults. Infect Dis Clin apy, and prevention. Am J Respir Crit Care Med 2001; 18 File TM Jr, Segreti J, Dunbar L, et al. A multicenter, 9 Critchley IA, Jones ME, Heinze PD, et al. In vitro activity of randomized study comparing the efficacy and safety of intra- levofloxacin against contemporary clinical isolates of Legio- venous and/or oral levofloxacin versus ceftriaxone and/or nella pneumophila, Mycoplasma pneumoniae and Chlamydia cefuroxime axetil in treatment of adults with community- pneumoniae from North America and Europe. Clin Microbiol acquired pneumonia. Antimicrob Agents Chemother 1997; 10 Stout JE, Arnold B, Yu VL. Comparative activity of cipro- 19 Fogarty C, Siami G, Kohler R, et al. Multicenter, open-label, floxacin, ofloxacin, levofloxacin, and erythromycin against randomized study to compare the safety and efficacy of Legionella species by broth microdilution and intracellular levofloxacin vs ceftriaxone sodium and erythromycin followed susceptibility testing in HL-60 cells. Diagn Microbiol Infect by clarithromycin and amoxicillin/clavulanate in the treat- ment of serious community-acquired pneumonia in adults.
11 Smith RP, Baltch AL, Franke M, et al. Effect of levofloxacin, Clin Infect Dis 2004; 38(Suppl):S16 –S23 erythromycin or rifampicin pretreatment of growth of Legio- 20 Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to nella pneumophila in human monocytes. J Antimicrob Che- identify low-risk patients with community-acquired pneumo- 12 Gotfried MH, Danziger LH, Rodvold KA. Steady-state plasma 21 Dunbar LM, Wunderink RG, Habib MP, et al. High-dose, and intrapulmonary concentrations of levofloxacin and cipro- short-course levofloxacin for community-acquired pneumo- floxacin in healthy adult subjects. Chest 2001; 119:1114–1122 nia: a new treatment paradigm. Clin Infect Dis 2003; 37:752– 13 Wright DH, Brown GH, Peterson ML, et al. Application of fluoroquinolone pharmacodynamics. J Antimicrob Che- 22 Salkind AR, Cuddy PG, Foxworth JW. Fluoroquinolone treatment of community-acquired pneumonia: a meta-analy- 14 Craig WA. Pharmacokinetic/pharmacodynamic parameters: sis. Ann Pharmacother 2002; 36:1938 –1943 rationale for antibacterial dosing of mice and men. Clin Infect 23 Muder RR, Yu VL, Fang G-D. Community-acquired Legion- naires disease. Semin Respir Infect 1989; 4:32–39 15 Pedro-Botet M, Vilaseca Z, Sopena N, et al. Erythromycin 24 Marston BJ, Plouffe JF, Breiman RF, et al. Preliminary versus fluoroquinolones in the treatment of Legionnaires findings of a community-based pneumonia incidence study.
disease. Presented at the 41st Interscience Conference for In: Barberee JM, Breiman RF, Dufour AP, eds. Legionella: Antimicrobial Agents and Chemotherapy, Chicago, IL, Sep- current status and emerging perspectives. Washington, DC: American Society for Microbiology, 1993; 36 –37 16 Mulazimoglu L, Yu VL. Can Legionnaires disease be diag- 25 Squier C, Lin YE, Stout JE. Waterborne pathogens: Legio- nosed by clinical criteria? A critical review. Chest 2001; nella and other opportunistic pathogens. Presented at the American Society of Healthcare Engineers of the American 17 Fogarty CM, Sullivan JG, Chattman MS, et al. Once a day Hospital Association, July 16 to 18, 2001; Tampa, FL; abstract levofloxacin in the treatment of mild to moderate and severe

Source: http://www.legionella.org/LD%20Levofloxacin%20Efficacy%20in%20the%20Treatment%20of%20Community-Acquired%20Legionellosis.pdf