~

ORIGINAL ARTICLE

Comparative study of levofloxacin and amoxycillin/clavulanic acid

in adults with mild-to-moderate community-acquired pneumonia
Clin Microbiol Infect 1999;5: 724-732
Claude Carbonl , Horacio Ariza2, Willem J Rabie3, Carlos R. Salvarezza4,
David Elkharrat 5, Manikam Rangaraj and Paola Decosta6
*HospitalBichat-Claude Bernard, Paris, France; 2Hospital Mi Pueblo, Florencio Varela,Buenos Aires,
Argentina, 3Department of Family Practice, National Hospital, Bloemfontein, South m c a ;
4Hospital Centenario, Rosario, Argentina; 5HospitalLariboisiire, Paris, France; 6HoechstMarion
Roussel, Romainville, France

Objective: To compare the efficacy and safety of two different doses of levofloxacin with amoxycillintclavulanic acid
in the treatment of community-acquired pneumonia.
Mathod.: A double-blind, randomized (l:l:l), double-dummy, three-arm parallel design, multicenter study was
conducted in adult patients with mild-to-moderate community-acquired pneumonia. In total, 518 patients were
randomized to receive levofloxacin 500 mg once daily, levofloxacin 500 mg twice daily or amoxycillin/clavulanic acid
625 mg three times daily for 7-10 days.
M u b The clinical cure rates post-therapy (2-5days after the end of treatment) in the intent-to-treatpopulation were
84.2% (1441171)in the levofloxacin oncedaily group, 80.2% (142/177)in the levofloxacin twicedaily group and 85.7%
(1W168)in the amoxycillin/clavulanic acid group. In the per-protocol population, the clinical cure rates post-therapy
were 95.2% (138/145),93.8% (137/146)and 95.3% (141/148),respectively. The total pathogen eradication rates were
97.896,10W0 and 97.546, respectively. Both drugs were equally well tolerated and no major adverse events were
observed.
Conclusions: Levofloxacin was effective, safe and well tolerated in the treatment of mild-to-moderate community-
acquired pneumonia. A complementary analysis indicatedthat there was no difference in therapeutic outcome between
levofloxacin 500 mg once daily and twice daily. Levofloxacin 500 mg once daily, for 7-10 days, is an effective and safe
treatment for mild-to-moderate community-acquired pneumonia in adults.
Kev w o n k Levofloxacin, amoxycillin/clavulanic acid, community-acquired pneumonia, clinical trial

INlRODUCTlON
Community-acquired pneumonia (CAP) is an acute
lower respiratory tract illness which is usually associated
with fmer, focal signs and symptoms in the chest and
recent pulmonary infiltrate in the chest X-ray [1,2].
The occurrence of CAP is influenced by geographic
and seasonal factors as well as population variables such
as age and the presence of underlying disease(s). Despite
the availability of modern antibacterial agents and a
Correspondingauthor and reprint requests:
Claude Carbon, CHU Bichat Claude Bernard, Service
Medecine Interne, 46 rue Henri Huchard, 95877,
Paris Cedex 18,France
Tel: +33 1 4025 7000 Fax: +33 1 4025 8845
E-mail: claude.carbon@bch.aphop-paris.fr
Accepted 30 July 1999
slow decrease in the mortality rate over the last five
decades, it remains a common and serious disease and
is the fifth or sixth leading cause of death in counties
such as the USA, the UK and Sweden.
Although the etiology of CAP continues to be
investigated, it is generally agreed that Streptococnrs
pneumoniue is the most common cause, accounting
for 30-70% of cases [3-7]. Other common pathogens
include Huemophilus infuenzue (2-18% of cases),
Mytoplusmu spp. (2-24%), Chtamydiu spp. (1-3%),
Legionella spp. (1-15%), S t i z p h y l o m u u m (1-lo%),
Moraxella caturrhulis (1-5%), Streptocouxcs spp. (l-3%),
Gram-negative bacilli such as Escherkhiu coli, Mebsiellu
spp. and Pseudomonus spp. (1-1 0%) and viruses (1-21%).
Up to 100?of cases may be due to two or more aerobic
bacteria, with Streptococcus pneumonia and H.i n f m a e
being the most common combination [5,8,9].
The choice of initial antibacterial agent is usually
empirical, and it should be filly effective against

724
Carbon e t al: Levofloxacin versus amoxycillin/clavulanic acid i n CAP
Streptococcus pneumoniae and other common pathogens,
including the atypical, intracellular pathogens Myco-
plasma and Chlamydia. The worldwide emergence of
Streptococcus pneumoniae with reduced susceptibility or
resistance to penicillin and macrolides, combined with
the resistance problems associated with p-lactamase-
producing H . injuenzae and Moraxella catarrhalis, has
provided a therapeutic challenge [lo-121. Penicillin-
resistant pneumococci are ofien also resistant to
macrolides and azalides. In addition, atypical pathogens
are not susceptible to p-lactam antibiotics, and their
susceptibility to macrolides is variable. The intro-
duction of new antibiotics may simplify the initial
treatment of CAP through broader coverage of
potential pathogens and possibly by shortening the
course of therapy.
Levofloxacin is the L-isomer of the racemate
ofloxacin and is approximately twice as active as the
equivalent amount of ofloxacin in vitro [13]. It has
a broad spectrum of activity which includes Gram-
positive aerobic bacteria such as Streptococcus pneumoniae
and Staphylococcus aureus, Gram-negative bacteria such
as Escherichia coli, H . inzuenzae, Klebsiella spp. and
Pseudornonas aeruginosa, and atypical bacteria such as
Chlamydia pneumoniae, Mycoplasma pneumoniae and
Legionella pneurnophila [13-151. Oral levofloxacin is
rapidly absorbed with 100% bioavailability, and peak
plasma concentrations are reached within 1 h. It is
eliminated relatively slowly, primarily by the kidneys,
with a half-life of 6-8 h, allowing once-daily dosing
[16,171.
The spectrum of activity of levofloxacin, combined
with its good penetration into lung tissue and bronchial
secretions [18-201 and previous clinical studies [21-241,
indicates that it may be effective in CAP. The aim of
this study was therefore to compare the efficacy and
safety of levofloxacin, at two different doses, with that
of amoxycillin/clavulanic acid in the treatment of CAP
in adults. Amoxycillin/clavulanic acid was chosen as
the comparator because it is widely used for this
indication and its efficacy is well documented [25-271.
PATIENTS AND METHODS
Study design
This was a double-blind, randomized (l:l:l),double-
dummy, three-arm parallel design, multinational, multi-
center study conducted in 50 centers in nine countries
in patients with mild-to-moderate pneumonia. Patients
were randomized to receive levofloxacin 500 mg once
My,levofloxacin 500 mg twice daily or amoxycillin/
clavulanic acid 625 mg three times daily for 7-10
days. Randomization was made in blocks of six, and
the study drug was randomly assigned to the patient
125
numbers in advance within each block. The primary
analysis was the cltnical cure rate, determined 2-5 days
after the end of treatment (post-therapy), in patients
evaluable for clinical efficacy in the per-protocol
population. The clinical and bacteriologic efficacy
assessments, performed by the investigator and a
computerized evaluation program (per-protocol assess-
ment), were included in the efficacy analysis. The
primary analysis was the per-protocol analysis deter-
mined by computer, and an intent-to-treat analysis was
also performed. The study was conducted in accor-
dance with the Good Clinical Practice Guidelines of
the European Community and The Declaration of
Helsinki.
Patients
In- or outpatients of either sex, aged 18-65 years, with
clinical signs and symptoms of mild-to-moderate
pneumonia and physical examination findings consistent
with the clinical diagnosis were included in the study.
Chest X-ray results confirming the clinical diagnosis of
pneumonia had to be present. Patients were excluded
if they were pregnant or of childbearing potential and
not taking adequate contraceptive measures or if they
had pneumonia occurring more than 72 h after
hospitalization; pneumonia requiring parenteral anti-
biotic treatment; one or more indicators of severe
pneumonia; pneumonia expected to be a terminal
event; glucose-6-phosphate deficiency; hypersensitivity
to ofloxacin or other fluoroquinolones or penicillin/
p-lactams; or any concomitant clinical condition likely
to interfere with the conduct of the study. Patients were
also excluded if they: required parenteral antibiotic
treatment for pneumonia; had received ofloxacin or
amoxycillin/cladanic acid for this infectious episode;
required probenecid or maintenance systemic corti-
costeroid therapy or a systemic antibiotic for another
infection; or had received antibiotic pretreatment for
more than 24 h in the 5 days before study entry or
azithromycin in the 7 days before study entry. All
patients provided written informed consent and the
study protocol was approved by each local ethics
committee.
Clinical
Assessments were performed after 3 and 6 days of
treatment, and 2-5 (post-therapy) and 14-21 days after
the end of treatment. Clinical signs and symptoms of
pneumonia were documented at each visit. The clinical
response was defined as follows: cure (all infection signs
and symptoms disappeared or returned to preinfection
state and chest X-ray improved, or at least one
infection-related sign and symptom, including chest X-
ray, improved and no subsequent antibiotic treatment
726
started); failure (all infection-related symptoms un-
changed or worsened; new clinical findings developed
consistent with active infection; death due to pneu-
monia; study drug discontinued because of clinical
and/or bacteriologic treatment failure; one or more
antibiotics added to the study drug because of
treatment failure or at least one infection-related sign
and symptom, including chest X-ray, improved but
subsequent antibiotic treatment started) or indeter-
minate (circumstances precluded classification as cure
or failure, e.g. missing follow-up information, pre-
mature discontinuation because of non-efficacy-related
reasons, major protocol violations). Patients were
withdrawn and switched to another antibiotic regimen
if: the clinical response was classed as failure after at least
48 h of treatment; the bacteriologic response was
unsatisfactory; or serious adverse events occurred
which were possibly related to study drug.
Bacteriologic
Bacteriologic cultures were obtained firom respiratory
tract secretions and blood samples within 48 h before
the start of treatment in order to isolate and identify the
causative pathogen. The majority of respiratory tract
cultures were obtained fiom non-invasive samples,
e.g. sputum, although some were obtained by inwive
methods such as bronchoscopic brushings, transtracheal
aspiration and bronchoalveolar lavage. Follow-up
cultures were obtained on day 3 in patients who had
not responded clinically, on day 6 in the case of drug
failure, 2-5 days after the end of treatment and 14-21
days &er the end of treatment in the case of failure.
Blood cultures were repeated in patients with persistent
fever. Serum samples were also obtained at inclusion
and 14-21 days after treatment for the determination
of antibody titer to Mycoplasma pneumoniae, Chlamydia,
Legionella and influenza viruses A and B.
The bacteriologic response was defined as: satis-
factory (eradication of baseline pathogen; presumed
eradication; colonization), unsatisfactory (persistence
of baseline causative pathogen; relapse; superinfection;
eradication and reinfection; new or additional anti-
biotic because of presumed persistence or resistance),
or indeterminate (lack of opportunity to perform
further cultures). If more than one causative pathogen
was isolated h m the pretreatment culture and the
microbiological response was not the same for all the
pathogens, the patient was classed as unsatisfactory if
the response of at least one pathogen was unsatisfactory.
safay
Safety was assessed at each visit accordmg to adverse
events and laboratory variables. Adverse events were
assessed by the investigator for intensity (mild,moderate
Clinical Micr obiology and Infection, Volume 5 Number 12
or severe), nature (serious or non-serious) and possible
relationship to the study drug. Patients were withdrawn
h m the study immediately in the case of a serious
adverse event possibly related to the study drug.
Statistical analyses
Analyses were performed on the responses in the
intent-to-treat population, the per-protocol population
and a l l patients with clinical signs and symptoms of
infection excluding major protocol violators. The
primary efficacy variable was the clinical cure rate,
post-therapy (determined 2-5 days after the end of
treatment), in the per-protocol population. Assuming a
success rate of 80% in all the treatment groups and a 6
of 15% (the maximum Merence between the two
levofloxacin groups and the amoxycillin/clavulanic acid
group to be accepted as equivalent), 125 evaluable
patients per group were required to provide an 80%
chance @ower=80%) of having a successfid trial,
defined as that producing a two-sided 95% confidence
interval (CI) for the difFerence between the clinical cure
rates of the two treatments that excluded the pre-
specified 6 value of 15% [28-301. Each lmfloxacin
group was considered to be equivalent to the amoxy-
cillin/clavulanic acid group if the upper and lower
bounds of the 95% CI for the efficacy rate difference
were >Ct% and > - 15%, respectively. A comparison
between the two levofloxacin groups was also per-
formed as a complementary analysis, using the same
rule as above.
In total, 1411 patients were screened, of whom 518
were included in the study, h m 50 centers in nine
countries. Two patients were withdrawn at their own
request before the study drug was administered, so 516
patients were included in the intent-to-treat popu-
lation. The demographic and baseline characteristics of
the intent-to-treat population are shown in Table 1.
The only statistically significant Merence between the
groups was a higher proportion of men in the
amoxycillin/cladanic acid group than in the lam-
floxacin twice-daily group (67.9% versus 57.6%;
p=0.05). This Merence was accounted for in the
efficacy analysis of robustness.
At inclusion, 297 (57.6%) patients had con-
comitant illnesses in the form ofunderlying diseases, of
which the most tiequent were respiratory, other than
pneumonia. Surgical history was positive in 171
patients (33.1%),and a history of drug/alcohol abuse
and smoking was observed in 14 (2.7%) and 295
(57.2%) patients, respectively. One hundred and four
(20.2%)patients were receiving concomitant non-anti-
Carbon e t al: L e v o f l o x a c i n versus amoxycillinlclavulanic a c i d i n CAP 7 21

Table 1 Demographic characteristics (intent-to-treat population)

Amoxycillin/
Levofloxacin Levofloxacin clavulanic acid
1X5OO mg 2x500 mg 3x625 mg
(n=171) (n=177) (tt=168)
Mean (+SD) age (years) 41.19k15.78 40.96214.20 40.93k14.23
Male 101 (59.1%) 102 (57.6%) 114 (67.9%)
Female 70 (40.9%) 75 (42.4%) 54 (32.1%)
Mean (kSD) weight (kg) 68.36214.99 66.76213.46 66.85k12.51
Race
whlte 130 (76.0%) 131 (74.0%) 124 (73.8%)
Black 25 (14.6%) 30 (16.9%) 29 (17.3%) .
Asian 1 (0.6%) 1 (0.6%) 0
Other 15 (8.8%) 15 (8.5%) 15 (8.9%)
Mean BMI in kg/rnZkSD 24.4624.82 23.68k3.97 23.4023.83
Diagnosis
Lobar pneumonia 140 (81.9%) 141 (79.7%) 131 (78.0%)
Bronchopneumonia 27 (15.8%) 29 (16.4%) 29 (17.3%)
Other 4 (2.3%) 7 (4.0%) 8 (4.8%)
Severity
Mild 82 (48.0%) 86 (48.6%) 87 (51.8%)
Moderate 89 (52.WA) 91 (51.4%) 81 (48.2%)
General condition
Good 131 (76.6%) 137 (77.4%) 134 (79.8%)
Poor 39 (22.8%) 39 (22.0%) 34 (20.2%)
Critical 1 (0.6%) 1 (0.6%) 0

infective medication. Prior antibiotic treatment occur-
red in 54 (10.5%) patients. During treatment, 218
(42.2%) patients received concomitant non-anti-
infective medications.
At inclusion, 370 patients were assigned to 7 days'
treatment and 148 to 10 days' treatment. The treatment
duration was prolonged to 10 days on day 6 in 72
patients originally assigned to 7 days' treatment. The
mean treatment duration was 8.1 days in all three
groups. Treatment was discontinued prematurely in 37
(7.2%) patients (11 in the levofloxacin once-daily group,
16 in the levofloxacin twice-daily group and 10 in the
amoxycillin/clavulanic acid group). One hundred and
five major protocol violations occurred in 75 patients
in the intent-to-treat population, giving a per-protocol
population of 441. No causative pathogen was isolated
in 321 patients, so bacteriologic efficacy was assessed in
120 patients. The number of patients included in each
analysis is shown in Table 2.
Clinical efficacy
The clinical cure rates are shown in Table 3. In
the intent-to-treat population post-therapy, the two-
sided 95% CI for the differences in cure rates revealed
therapeutic equivalence between amoxy&in/clavu-
lanic acid and the levofloxacin once-My (-9.7%;
+6.7%) and twice-daily groups (-14.0%; +3.0%).
The clinical response in the per-protocol population
post-therapy as assessed by the computerized evaluation
program was the primary efficacy outcome measure.
The two-sided 95% CI for the differences in cure rates
also revealed therapeutic equivalence between amoxy-
cillin/clavulanic acid and the levofloxacin once-daily
group (-5.7%; +5.5%) and the levofloxacin twice-
dady group (-7.3%; +4.4%). A complementary
analysis comparing the two levo%oxacingroups did not
show any difference. Of the 23 patients classed as
failures, 14 improved with subsequent antibiotic.
In the per-protocol population with bacterio-
logically proven infection, the two-sided 95% CI for
the S e r e n c e s in cure rates also revealed therapeutic
equivalence between amoxycillin/clavulanic acid and
the levofloxacin once-daily group (- 12.2%; 13.6%) +
and the levofloxacin twice-daily group (-14.1%;
+ 12.2%).
The clinical response post-therapy in all three
populations, as assessed by the investigator, supported
the findings of the computerized evaluation program.
In the intent-to-treat population, the clinical cure rates
were 90.6%, 87.6% and 91.1% in the levofloxacin
once-daily, levofloxacin twice-daily and amoxycillin/
clavulanic acid groups, respectively. In the per-protocol
population, the clinical cure rates were 95.9%, 95.2%
and 96.0%, respectively. The corresponding 95% CI
values for levofloxacin once daily versus amoxy-
cillin/clavulanic acid and levofloxacin twice daily
728 Clinical M i c r o b i o l o g y a n d Infection, Volume 5 N u m b e r 12

Table 2 Number of Datients included in each anal*

~~~ ~~

Amaxycillid
Levoflcaacin LCvOflaXaCin clavulanic acid
1x500 mg 2x500 mg 3x625 mg Total

Patients enrolled 172 177 169 518

Withdrawn fium the study
before start of study drug 1 0 1 2
Intent-*treat 171 177 168 516
Excluded fium per-protocol
d Y = 26 30 19 75
Per-protocol 145 147 149 441
No pathogen isolated 105 102 114 321
Per-protocol+bacteriologically
p m n infection 40 45 35 120

Table 3 Clinical response post-therapy

Number of patients (%)
Ammcydllin/
LWOflaXaCin LwOflOXaCin davuLnic acid
Population Assessnent 1X500mg 2x500 rng 3 x 6 s mg

Intention-to-treat No. of patients 171 177 168

CurC 144 (84.2%) 142 (80.290) 144 (85.7%)
Fail- 27 (15.8%) 35 (19.8%) 24 (14.3%)
Per-protocol No. of patients 145 147 149
Indeterminate 0 1 1
Total analyzed 145 (100%) 146 ( l W o ) 148 (100%)
Cure 138 (95.2%) 137 (93.8%) 141 (95.3%)
FdurC 7 (4.8%) 9 (6.Wo) 7 (4.7%)
BacteriologicaUy proven infection No. of patients 40 45 35
CurC 38 (95.0%) 42 (93.3%) 33 (94.3%)
Fail- 2 (5.0%) 3 (6.7%) 2 (5.7%)

95% CISarc given in the results.

versus amoxycillin/clavulanic acid were -5.3% and
+5.1%, and -6.1% and +4.6%, respectively, indicating
therapeutic equivalence. In the per-protocol popu-
lation with bacteriologically proven infection, the cure
rates were 97.5%, 93.3% and 97.1%, respectively.
Evaluation of the clinical response performed at the end
of the study (14-42 days after the end of treatment) also
revealed no differences between the three treatment
groups. Efficacy evaluations in subgroups of patients
(presumed pneumococcd pneumonia; fever at baseline;
lobar pneumonia; and smokers) revealed similar
findings to the main analysis.
Bacteriologic efficacy
The pathogens isolated at inclusion are shown in Table
4. The majority of respiratory tract cultures were
obtained fiom non-invasive samples; only 4.4% of
patients (61136) had invasive samples (two in the
levofloxacin once-daily group, three in the levofloxacin
twice-daily group and one in the amoxycillin/
clavulanic acid group). Bacteremia was found in 2.3%
(1/43) of levofloxacin once-daily patients, 19.6%
(10/51) of levofloxacin twice-daily patients and 16.7%
(7/42) of amoxycillin/clavulanic acid patients. A single
pathogen was isolated in 113 patients, two pathogens
in 21 patients and three pathogens in two patients. A
four-fold increase in serum antibody titer between the
acute and convalescent serum for atypical infections
was observed in 5.8% (16/274) of patients (Mycoplarma
pneumoniae eight patients, Chlamydia two, Legionella
one, influenza A one and influenza B four).
More pathogens, both Gram-positive and Gram-
negative, were resistant to amoxycillin/clavulanic acid
than to levofloxacin. At inclusion, using the disk
diffusion method, only one (Streptococcus pneumoniae) of
the 156 (0.6%) isolated pathogens tested was mistant
to levofloxacin, while 14 of 138 (10.1%) pathogens
tested (three Gram-positive and 11 Gram-negative (five
enterobacteria and six non-enterobacteria)) were
resistant to amoxycillin/clavulanic acid. None of the
C a r b o n e t al: L e v o f l o x a c i n v e r s u s a m o x y c i l l i n l c l a v u l a n i c a c i d i n C A P 729

Table 4 Causative pathogens isolated at inclusion in the intent-to-treat population

AmOXy&/
Levofloxacin Levofloxacin davulanic acid
Pathogen 1X500mg 2x500 mg 3x625 mg Total
StreptouKncs pneumoniae 16 28 19 63 (39.1%)
Haemophilus inauenzae 21 17 17 55 (34.2%)
StqhylouKncs aureus 2 3 6 11 (6.8%)
H.parainzuenzae 3 6 1 10 (6.2%)
Moraxella catanhalis 2 3 3 8 (5.0%)
Klebsiella pncumoniae 3 1 0 4 (2.5%)
Pseudomonas amginosa 1 1 1 3 (1.9%)
Enterobacter cloacae 2 1 0 3 (1.9%)
Pmteus mirabilis 0 1 1 2 (1.2%)
Klebsiella orytoca 1 0 0 1 (0.6%)
Streptocorn viridans 0 0 1 1 (0.6%)
Total 51 61 49 161

Table 5 Bacteriologic response in the per-protocol + bacteriologically proven infection population post-therapy
AmoXycillin/
Levofioxacin Levofloxacin clavulanic acid
1X500me 2x500 me 3x625 me Total
Total treated 40 45 35 120
Indeterminate 1 1 0 2
Satisfactory-ut of time window 0 1 1 2
Total analyzed 39 (100%) 43 (100%) 34 (100%) 116
Satisfactory 37 (94.9%) 43 (100%) 32 (94.1%) 112
Eradication 3 7 4 14
Presumed eradication 34 . 36 28 98
Unsatisfactory 2 (5.1%) 0 2 (5.9%) 4
Persistence 1 1 2
Presumed persistence 1 0 1
Eradication+superinfection 0 1 1

Table 6 Bacteriologic eradication rate by pathogen in the per-protocol+bacteriologically proven infection population post-
theravv
Amoxycillin/
Lmfloxacin Levofloxacin clavulanic acid
1 X500 mg 2x500 m g 3x625 mg
Total pathogens 44/45 (97.8%) 53/53 (100%) 39/40 (97.5%)
Gram-negative 28/29 (97%) 28/28 (100%) 16/17 (94.1%)
Pseudomom amginosa 1/1 1/1 1/1
Moraxella catarrhalis 2/2 2/2 1 /2
Haemophifus injuenzae 19/19 16/16 12/12
H . parainauenzae 2/3 6/6 1/1
Other 4/4 3/3 111
Gram-positive 16/16 ( l W ? ) 25/25 (100%) 23/23 (100%)
Sraphylocoms aureus 1/1 313 6/6
Sfreptormuspneumoniae 15/15 22/22 16/16
0th- - 1/1

isolates from the levofloxacin once-daily group (0/51)
and one of the 59 isolates (1.7%) h m the twice-daily
group was resistant to levofloxacin, while five of 40
isolates (12.5%) from the amoxycillidclavulanic acid
group were resistant to amoxycillin/clavulanic acid.
The bacteriologic response post-therapy in the
per-protocol population with bacteriologically proven
infection as assessed by the computerized evaluation
program is shown in Table 5. All three groups showed
a similar bacteriologic response rate. Similar findings
730 Clinical M i c r o b i o l o g y a n d I n f e c t i o n , V o l u m e 5 N u m b e r 12

were observed in the investigator’s assessment of
bacteriologic response (levofloxacin once daily 95.0%,
levofloxacin twice daily 95.6%, amoxycillin/clavulanic
acid 94.3%). The bacteriologic eradication rate by
pathogen at post-&erapy is shown in Table 6. All Gram-
positive isolates (Streptococcus pneumoniae, Stapkylocorcus
a u m ) and all H. injuenzae isolates were eradicated by
all three treatments. Post-therapy, one H. parainzuenzae
isolate &om a multipathogen infection in a patient in
the levofloxacin once-daily group and one Moraxella
catawhalis isolate in a patient in the amoxycillin/
clavulanic acid group persisted. The only pathogen
which was resistant to levofioxacin in the levofloxacin
twicedaily group was presumed to be eradicated.
Evaluation of the bacteriologic response performed at
the end of the study (14-42 days after the end of
therapy) also revealed no differences between the three
treatment groups.
safety
All patients included in the intent-to-treat population
were assessed for safety. The incidence of adverse events
was similar in all three treatment groups. The incidence
of adverse effects by body system is shown in Table 7.
The mjority of adverse events in a l l three groups were
mild to moderate in intensity. Five events reported by
four patients in the levofloxacin once-daily group, four
events by four patients in the twice-daily group and six
events reported by five patients in the amoxycillid
clavulanic acid group were considered by the investi-
gator to be severe in intensity.
There was a low or no incidence of adverse events
associated with the fluoroquinolone class of antibiotics,
i.e. musculoskeletal, central nervous system, cardio-
vascular, digestive system, skin rash/phototoxicity and
eye disorders. The majority of events in all three p u p s
were mild to moderate in intensity.
Serious adverse events irrespective of their
relationship to the study drug were reported by seven
patients in the levofloxacin once-dady p u p (4.1%), 11
in the levofloxacin twice-daily group (6.2%) and six in
the amoxycillin/clavulanic acid group (3.6%). Serious
adverse events in 10 of 24 patients were considered to
be possibly related to the study drug by the investigator
(four in the levofloxacin once-daily group, four in the
levofloxacin twice-daily group and two in the
amoxycillin/clavulanicacid group).
Two deaths occurred, both in the amoxycillin/
clavulanic acid group. One patient died 13 days after
the end of treatment (due to tuberculosis) and the other
died 2 h afker the first dose of study drug (due to status
asthmaticus). Treatment was discontinued in 18patients
(3.5%) due to adverse events (five in the levofloxacin
once-daily group, eight in the levofloxacin twice-daily
group and five in the amoxycillin/clavulanic acid
group). There were no clinically relevant abnormalities
in laboratory parameters.
DISCUSSION
The worldwide emergence of resistance to B-laaams
and macrolides among ,respiratory tract pathogens has

Table 7 Number of patients with adverse events (AEs) by body system (patients could have more than one AE)
Number of patien6 (%)
AUAES Possibly related AEs
Amqcillin/ AmcaYcillin/
h5oxacin Lcvofloxacin clavulanicacid Lcvo5oxacin Lcvofloxaciu clawl?nicacid
BodYsyJtun 1x500 mg 2x500 mg 3x625 mg 1X500mg 2X500mg 3x625 mg
Total no. 171 ( l W ? ) 177 (100%) 168 (lW!) 171 ( l W ? ) 177 (1Wo) 168 (100%)
Total with AEs 67 (39.24 74 (41.8%) 71 (42.3%) 46 (26.9%) 51 (28.8%) 50 (29.8%)
Body as whole 11 (6.4%) 7 (4.P?) 6 (3.6%) 6 (3.5%) 1 (0.6%) -
Cudiovasl-.ukr 4 (2.3%) 7 (4.Ph) 3 (1.8%) - - -
DigCStiVC 14 (8.2%) 18 (10.2%) 26 (15.5%) 11 (6.4%) 14 (7.9%) 20 (11.9%)
Hematologic/
lpph?tiC 26 (15.2%) 35 (19.8%) 25 (14.9s’) 23 (13.5%) 30 (16.9%) 21 (12.5%)
Injection site reaction - 1 (0.6%) - - - -
Metabolic and
nutritional 10 (5.8%) 14 (7.9%) 12 (7,1%) 6 (3.5%) 8 (4.5%) 10 (6.0%)
Musculcwkeletal - 4 (2.3%) - - 1 (0.6%) -
NCIVOLYS 3 (1.8%) 4 (2.3%) 1 (0.6%) 2 (1.2%) - 1 (0.6%)
Respiratory 15 (8.8%) 7 (4.0%) 17 (10.1%) 4 (2.3%) 1 (0-6%) 2 (1.2%)
Skin and appendages 5 (2.9%) 8 (4.5%) 6 (3.6%) 2 (1.299) 4 (2.3%) 3 (1.8%)
special JensTs 1 (0.6%) 1 (0.6%) - - - -
Urogmital 5 (299) 4 (2.3%) 4 (2.4%) - 3 (1.7%) 2 (1.2%)
Carbon e t al: Levofloxacin versus amoxycillinlclavulanic acid i n CAP
led to the reassessment of treatment for CAP [8,31].
The newer fluoroquinolones, such as levofloxacin and
grepafloxacin, have good activity against respiratory
tract pathogens and are well tolerated [13,32]. They
should, therefore, provide a usefd alternative treatment
in the face of the increasing spread of resistance to
traditional antibiotics [33].
Sparfloxacin was the first of the newer fluoro-
quinolones to be introduced but it has been associated
with a high incidence of photosensitivity reactions and
cardiovascular adverse events [34]. In studies of CAP,
sparfloxacin was shown to be as effective as the
comparators,with overall efficacy rates of 92% and 87%
compared with 82% for the combination of amoxy-
cillin and ofloxacin [35], 80% for amoxycillin/
clavulanic acid and 85% for erythromycin [36].
Previous studies have demonstrated the efficacy of
levofloxacin in the treatment of CAP [22,23], and the
results of our double-blind study confirm these
findings. Our results show that levofloxacin, at a dose
of 500 mg once or twice daily for 7-10 days, was
equivalent to amoxycillin/clavulanic acid three times
daily, in terms of efficacy, tolerability and safety for the
treatment of mild-to-moderate CAP. The clinical cure
and bacteriologic response rates with levofloxacin once
daily (95.2% and 94.9%, respectively) and twice daily
(93.8% and loo%, respectively) were. equivalent to
those with amoxycillin/clavulanic acid (95.3% and
94.1%, respectively) and this was confirmed by the
two-sided 95% CIS for the differences in cure rates.
There was no difference in therapeutic outcome between
levofloxacin 500 mg once daily and twice daily
Levofloxacin demonstrated good activity against
Gram-positive pathogens, including Streptococcus pneu-
moniae, with a 100% eradication rate. This is an
important consideration, since Streptococcus pneumoniae
is the commonest cause of CAP and treatment is often
empirical. Older fluoroquinolones, such as cipro-
floxacin, have not been beneficial in the treatment of
respiratory tract infections because of their limited
activity against Streptococcus pneumoniue [37]. The
improved Gram-positive activity of levofloxacin is
therefore an important advance.
Both levofloxacin and amoxycillin/clavulanic acid
were well tolerated. The absence of severe cases of
pneumonia in this study is reflected in the low
mortality rate (0.4%), with only two deaths. Higher
mortality rates of 4.5% for sparfloxacin and 2.0% for
amoxycillin/clavulanic acid were reported by Lode et
al in patients with non-severe, mild-to-moderate
CAP [36]. However, the same mortality rate of 0.4%
was reported by Carbon et al in a study comparing
temafloxacin with amoxycillin in patients with CAP of
average severity [38].
731
In conclusion, the results show that levofloxacin
was as effective, safe and well tolerated as amoxycillin/
clavulanic acid three times daily in the treatment of
mild-to-moderate CAP. In addition, levofloxacin
provides the advantage of once-daily dosing. Thus,
levofloxacin 500 mg once daily, for 7-10 days, is an
effective and safe treatment for mild-to-moderate CAP
in adults.
Acknowledgments
The authors would like to thank the following
members of the international study group for their
assistance in conducting this study: Argentina-
Ambasch and Bergallo, Farias, Harris, Jasovich, Luna
and Jolly, Mingrone, Ruvinsky; Finland-Kuosmanen,
h e , Sanna Kesa; France-Balmes, Baron, Canton,
Muir, Zuck; Germany-CloB, Koch, Petri, Menke,
Schnorr; Ireland O’Doherty, Dwyer, McGarry,
Quinn, Moran, Lee, McVey, McDonnell; Italy-
Giuntini, Rizzato, Todisco; The Netherlands-Van
Herzik; South Africa-Aboo and Minnie, Nel and
Smit, Van Vuuren, Foden; UK-Ah, Cantor, Khong,
Leak, Scott, Sheldon, Sheridan, Warriner and Spira,
Rees-Jones, Northfield, Hosie. This study was sup-
ported by a grant from Hoechst Marion Roussel.
References
1. Marnie TJ. Community-acquired pneumonia. State-of-the-art
article. Clin Infect Dis 1994; 1 8 501-15.
2. Bartlett JG, Mundy LM. Community-acquired pneumonia.
N Engl J Med 1995; 333: 1618-24.
3. Ashby BL. Treatment of pneumonia. new strategies for changing
pathogens. Clin Ther 1991; 13: 637-50.
4. Venkatesan P, Macfarlane JT. Epidemiology, pathogenesis and
prevention of pneumonia. Curr @in Infect Dis 1991; 4: 145-
9.
5. Macfklane JT, Colville A, Guion A, Macfarlane RM, Rose DH.
Prospective study of aetiology and outcome of adult lower
respiratory tract infections in the community. Lancet 1993; 341:
5 11-30.
6. Bartlen JC. Community-acquired pneumonia: current status.
Infect Dis Clin Pract 1995; 4(suppl4): S223-31.
7. Johnson JA. Pathogenesis of bacterialinfectionsof the respiratory
tract. Br J Biomed Sci 1995; 52: 157-61.
8. Fass RJ. Aetiology and treatment of community-acquired
pneumonia in adults: an historical perspective. J Antimicrob
Chemother 1993; 32(suppl A): 17-27.
9. Carbon C, Leophonte I? Management of community-acquired
pneumonia. J Antimicrob Chemother 1993; 32: 1-3.
10. Schuae GE, Kaplan SL, Jacobs RE Resistant pneumococcus: a
worldwide problem. Infection 1994; 22: 233-7.
11. Yee YC, Thornsberry C. Penicillin-resistant Sheptococnrs pneu-
m o n k on the rise in the United States: its effect on oral cephalo-
sporins. Antimicrob Infect Dis Newslett 1994; 13: 49-60.
12. Felmtngham D, Griineberg RN, The Alexander Project Group.
A multicentre collaborative study of the antimicrobial suscepti-
bility of community-acquired, lower respiratory tract pathogens
1992-1993 The Alexander project. J Antimicrob Chemother
1996; 3 8 ( ~ ~ pA):
p l 1-57.
732 Clinical Microbiology and Infection, Volume 6 N u m b e r 12
13. Langtry HD,Lvnb HM. Levofloxacin. Its uses in infiections of
the respiratory tract, skin, soft tissues and urinary tract. Drugs
1998; 5 6 487-515.
14. Fu Kp, &do SC, Foleno’B, et al. In’vitm and in vivo
antibacterial activities of levofloxacin (l-ofloxacin), an optically
active of&. AntimiaobAgents Chunother 1992; 3 6 860-6.
15. Yvnvle N, Jones RN,Frci R, Hoban DJ, plgnatvi AC, Marc0
E Levofimudn in vitro activity: results fhm an international
comparativestudy with ofloxacin and ciprofloxacin.J Chemothcr
1996,6 83-91.
16. Chien S-C, Rogge MC, Gisclon LG, et 11. Pharmacokinetic
profile of le~floxacinfollowing once-daily 500-milligr+m oral
or intnvcnous doses. Antimiaob Agents Chcmother 1997; 41;
2256-60.
17. Fish DN, Chow AT. The clinical phannacokinetics of
1 ~ ~ 0 f l ~ cClin
i n .Phvmacokinet 1997; 3 2 101-19.
18. Lee LJ, Sha X, Gotfiicd MH, H d JR,Dix RK,Fish DN.
Penetration of lmfloxacin into lung tissue after oral admini-
stration to subjects undergoing lung biopsy or lobectomy.
Pharmacothcnpy 1998; 18 35-41.
19. Nagai H, Yvnwki T, Matsuda M. Goto Y, Tviro T, Nuu M.
Penetration of levofloxacin into bronchoalveolar lavage fluid
Drugs 1993; 45(suppl3): 259.
20. N h o n Y, Ikbyaduta Y, Nakatani I, Nakata K. Levofloxacin:
penetration into sputum and once-daily trratmcnt of respiratory
tract infiections. Drugs 1995; 49(suppl2): 418-19.
21. DeMate CA, Russell M, McElvaine P,et al. Safety and efficacy
of oral levofloxacin versus c.e&uximeaxed in acute bacterial
exacerbation of chronic bronchitis. Respir Cue 1997; 42:
20613.
22. File TMJr. Segreti J, Dunbar L, et al. A multicenter, randomized
s t u d y comparing the &cacy and safety of inmvenous and/or
oral lcvdoxacin versus ceftriaxone and/or cefurmdme axed in
treatment of adults with community-acqukd pneumonia.
Antimicmb Agents Chemother 1997; 41: 1965-72.
23. Norrby Rs. h f l o x a c i n 2x500 mg iv (sequential therapy
iv/oral) vs ceftriaxonc 1X4.0 g iv in hospitalid patients with
pneumonia. Sand J Infect Dis 1998; 3 0 397-404.
24. Shah €? Le~floxaciuw. cefurmdme axctil in the treatment of
acute uucerbation of chronic bronchitis (AECB). J Antimicrob
Chcmother 1999: 43: 529-39.
25. Cunha BA.Amqcillin/cla- therapy of respiratory &act
infections: a microbiologic perspective. Qin Tha 1992; 1 4
418-25.
26. Lcgnan~D. Role of antiiiotics in treatment of community-
acquired lower respiratory tnct infections. D i a p Miaobiol
I&ct Dis 1997; 27 41-7.
27. Ncu HC, Whon APFL, Griinebcrg RN.AmqciUin/cladanic
xi& a review ofits efficacy in 38 500 patients from 1979to 1992.
J Chemother 1993; 5: 67-93.
28. Makuch K, Simon R.Sample size requirements for 4uating a
conservative therapy. Cancer Tmt Rcv 1978; 62: 1037-40.
29. Harkius RD,Albrecht R Design and YlaEjSis ofdinicaltrials for
anti-infidv~drug prod~ctr.Drug IofectJ 1990; 4 213-24.
30. Machin D, Campbell MJ. Statistid tables for the design of
clinical trials. M o d . Blackwcll Scientific Publications, 1987.
31. Bvry AL. Antimicrobialagents for commdty-acquired q i r a -
tory tnct infections. Infection 1995; 23(suppl2): S59-S64.
32. W A J , B&UI JA. Grepdomdn. Drugs 1997; 53: 817-24.
33. Finch RG. T h e role of new quinolones in the aeatment of
respintory tract infections. 1995; 49(suppl2): 144-51.
34. Goa KL, Bryson HM, Markham A. Spuflaxldn: a &cw ofits
antibacterialactivity, pharmacokincticproperties, clinical eftiacv
and tolerability in lowcr respiratory tract infections. L h q s 1997;
5 3 700-25.
35. Pomer H, May TH, the French S ~ d yGroup. Comparative
&acy of sparfloxacin in comparison with ainoxycillin plus
ofloxacin in the treatment of community acquid pneumonia.
J Antimicrob Chcmother 1996; 37(suppl A): 83-91.
36. Lode H, Garau J, Gnssi C, et al. Tnatment of community-
acquired pneumonia: a randomized compviron of sparfloxacin,
amoxycillin/&vulanic acid and erythromyciq. Eur Respir J
1995; 8 1999-2007.
37. Lee BL, Padula AM, Kimbrough RC, et al. Infectious
complications with respiratory pathogens despite aprofloxacin
therapy. N Engl J Med 1991; 325: 520-1.
38. Carbon C, Uophonte P, Petitpretz P, ChauvinJp, Hazebmucq
J. Efficacy and safety of tunaflcnacin versus those ofamoxidin
in hospitalized adults with community-acquid pneumonia.
Antimiaob Agents Chemothcr 1992; 36:833-9.