Академический Документы
Профессиональный Документы
Культура Документы
Validity Assessment
A review of the quality of each RCT included in our metaanalysis was performed by using the Jadad score, which examines whether there is randomization, blinding, and information
on withdrawals from the study, and evaluates the appropriateness of randomization and blinding, if present.7,8 One point
was awarded for the presence of each of the former 3 criteria,
whereas the latter 2 criteria could be awarded the values of 1
(inappropriate), 0 (no data), and 1 (appropriate). Thus, the
maximum score for a study was 5; a score higher than 2 points
denotes a good quality RCT according to this methodology.
The reviewers calculated the Jadad score of each study independently.
Data Extraction
Two reviewers (D.K.M. and E.M.K.) independently extracted
and tabulated the data from the selected studies. The data extracted from each study included the author names and year of
publication, the population of interest, the number of patients,
the type of therapeutic regimens used, the adjunctive treatments used (if any), the location and extend of focus of osteomyelitis, the duration of therapy and of the follow-up, as well
as treatment success, bacteriological success, relapses, superinfections, and patients with adverse events. For studies not focusing exclusively on patients with osteomyelitis, we extracted
the data which were reported particularly for the subset of
patients with such infections.
Definition of Outcomes
Patients were considered to have osteomyelitis, when the diagnosis was confirmed by bone biopsy or alternatively was established by appropriate findings in at least 2 of the following
diagnostic methods: (a) clinical examination, (b) conventional
radiography, (c) bone scanning, and (d) computed tomography
or magnetic resonance imaging. The primary outcome of our
meta-analysis was defined as treatment success of each regimen. Cure was defined as resolution of all signs and symptoms
of active infection at the end of follow-up period. The secondary outcomes of our meta-analysis included bacteriological
success, relapses, superinfections, and number of patients experiencing adverse events due to the administered regimens.
Bacteriological success was defined as absence of all causative
organisms from the culture at the end of follow-up. Superinfection was defined as isolation of new pathogens from the site of
the infected bone during the course of therapy. Relapse was
defined as reappearance of the causative pathogen during follow-up, after prior clinical signs or microbiological data of
eradication.
Statistical Analysis
Statistical analyses were performed using the S-Plus 6.1 software (Insightful Corp., Seattle, WA). The heterogeneity between studies was assessed by using the 2-test; a P value less
than 0.10 was defined to denote the presence of statistically
significant heterogeneity between studies. Bias of small studies
was assessed by the funnel plot method using the Egger test.
Pooled odds ratios and 95% confidence intervals (CIs) for all
primary and secondary outcomes were calculated by using
both the Mantel-Haenszel fixed-effect model 9 and the
DerSimonian-Laird random-effects model.10 For all analyses,
results obtained with the fixed-effect model (FEM) are presented if there was no heterogeneity between studies; otherwise, results from the random effects model are presented. The
reported outcomes of the analyzed studies were weighted by
the inverse of their variance with the fixed-effect model.
Results
Study Selection Process
In Figure 1, we present a flow diagram depicting the
various steps of the study selection process. We identified
287 potentially relevant studies that included patients
with osteomyelitis, exclusively or as a subset of the whole
study population, who were treated with fluoroquinolones or other conventional agents. From those, 220 studies were excluded because they were case reports, reviews, animal or laboratory studies, imaging studies,
case series, or retrospective studies. Fifty-six studies were
further excluded because they constituted noncomparative studies. Also, 3 studies were excluded because they
were not randomized controlled trials and another study
was excluded because it included a small number of patients. The remaining 7 studies met the inclusion criteria
and were evaluated further.1117
Study Characteristics
In Table 1, we present the characteristics (patient population, administered drugs, duration of treatment, and
follow-up) of the RCTs included in our meta-analysis.
Patients in the first treatment arm of all RCTS received
fluoroquinolones. The fluoroquinolones used were ciprofloxacin in 3 studies, which were the older ones, performed,14,16,17 ofloxacin in 3 studies, which were the
newer ones,1113 and pefloxacin in 1 study.15 Fluoroquinolones were administered exclusively orally in 4 out of
7 RCTs,13,14,16,17 whereas in the other 3 RCTs, fluoroquinolones were administered parenterally and subsequently orally.11,12,15 Patients in the second treatment arm
received -lactams. In 5 out of 7 RCTs, -lactams were
administered exclusively via the parenteric route.11,1316 In
one RCT, -lactams were administered intravenously and
subsequently orally,12 whereas in one RCT, the route of
administration is not clear.17 With the exception of the
study by Gomis et al12 and by definition for the study by
Lipsky et al,12 no specific data were reported regarding the
location and extend of focus of osteomyelitis. Furthermore,
no data regarding surgical procedures undertaken after the
enrollment of patients in the trials were reported.
Treatment Success
In Table 2, we present data regarding the primary and
secondary outcomes. Data about treatment success were
provided in all 7 RCTs included in our meta-analysis.1117
There was no difference in treatment success between fluoroquinolones and -lactams [194 patients, FEM, odds ratio
(OR) 0.99, 95% CI 0.511.91, P 0.97]. The odds
ratios for treatment success in the individual RCTs, as well
as the pooled odds ratio, are presented graphically in Figure
2(A).
Superinfections
Data about superinfections were reported in 6 out of 7
RCTs.11,1317 Although superinfections were more frequent in patients receiving fluoroquinolones, there was
no statistically significant difference in the superinfections between fluoroquinolones and -lactams (173 patients, FEM, OR 1.75, 95% CI 0.63 4.90, P
0.28). The odds ratios for superinfections in the individual RCTs, as well as the pooled odds ratio, are presented
in Figure 3(A).
Bacteriologic Success
Data about bacteriological success of fluoroquinolones versus -lactams were provided in 6 out of 7 RCTs.11,1317
There was no difference in bacteriological success between
the 2 compared treatments (201 isolates, FEM, OR 0.88,
95% CI 0.451.70, P 0.71). The odds ratios for bacteriologic success in the individual RCTs, as well as the
pooled odds ratio, are presented in Figure 2(B).
Relapses
Data about relapses were reported in 5 out of 7 RCTs.1317
The follow-up period for the evaluation of relapses ranged
from 6 to 18 months. There was no difference in relapses
between fluoroquinolones and -lactams (153 patients,
FEM, OR 1.23, 95% CI 0.46 3.31, P 0.68). The
odds ratios for relapses in the individual RCTs, as well as
the pooled odds ratio, are presented in Figure 3(B).
Study Design
Population
Gomis et al,
199911
Open-label
RCT
Lipsky et al,
199712
Open-label
multicenter
RCT
Gentry and
RodriguezGomez,
199113
Fluoroquinolone
Regimen
Other Regimen
Adjunctive Treatments
Ofloxacin 400 mg
p.o. q12h (if
not possible it
was
administered
i.v.)
Imipenem/cilastatin 500 mg
i.v. q6h
Ofloxacin 400 mg
i.v. q12h
changed to
p.o. when
appropriate
Ampicillin 12 g/sulbactam
0.51 g i.v. q6h changed
to amoxicillin 500 mg/
clavulanic 125 mg p.o.
q8h when appropriate
Open-label
RCT
Ofloxacin 400 mg
p.o. q12h
Gentry and
Rodriguez,
199014
Open-label
RCT
Ciprofloxacin 750
mg p.o. q12h
Giamarellou
et al,
198915
Open-label
RCT
Broad spectrum
cephalosporin i.v.
(usually ceftazidime) or
nafcillin-aminoglycoside
combination (usually
amikacin)
Ceftazidime 2g i.v. q8h or
1g i.m. q8h
Snydman
et al,
198916
Open-label
RCT
Greenberg
et al,
198717
Open-label
RCT
Pefloxacin 400
mg i.v. or p.o.
q8h to q12h,
as well as i.v.
followed by
p.o.
Ciprofloxacin 750
mg p.o. q12h
Ciprofloxacin 750
mg p.o. q12h
NA
Alternative parenteral
antimicrobial therapy
with -lactams
NA
Another appropriate
antimicrobial therapy
NA
RCT indicates randomized controlled trial; OM, osteomyelitis; NA, not applicable; p.o, per os; i.v, intravenously; q12h, every 12 h; q8h, every 8 h; q6h, every 6 h.
*n 7 out of 8 cases pefloxacin was given for 180 d.
-Lactams, aminoglycosides, vancomycin, imipenem, clindamycin, trimethoprim/sulfamethoxazole.
Discussion
The main finding of our meta-analysis is that fluoroquinolones were as effective as -lactams regarding the primary
outcome of treatment success for patients with osteomyelitis. Also, no difference was found between the 2 treatments
in the analyses of secondary outcomes, namely, bacteriological success, superinfections, relapses, and patients with
adverse events. However, it should be noted that in 4 of the
7 included studies fluoroquinolones were administered
Table 1. Continued
Enrolled
Patients
32
21
108
21
NA
42
NA
67
NA
103
Follow-up
Period
Quality
Assessment
3045 d
6 mo
NA
54 d vs. 30 d
18 mo
56 d vs. 47 d
12 mo
15
46 mo* vs. 68
wk
12 mo
29
21
72 d vs. 44 d
NA
30
56 d vs. 44 d
Up to 13 mo
orally while the comparator -lactam agents were administered intravenously. Taking under consideration the fact
that fluoroquinolones exhibit high bioavailability when administered orally, fluoroquinolones appear as a valid alternative for initiation or continuation of the long-term treatment
required for osteomyelitis allowing better compliance.
To our knowledge, there has been only one relevant
meta-analysis, performed by Stengel et al,18 in which
several types of antibiotic therapies were examined for
the treatment of both bone and joint infections. In the
subanalysis of fluoroquinolones versus nonfluoroquinolone control regimens, the results were similar to the
ones obtained in our meta-analysis, despite the fact that
Mean Duration of
Therapy
Treatment Success
Bacteriological Success
Superinfections
Relapses
10/11 (90.1)
6/16 (37.5)
14/19 (73.7)
7/10 (70)
1/4 (25)
12/14 (85.7)
10/10 (100)
NA
12/20 (60)
8/10 (80)
NA
13/20 (65)
0/10 (0)
NA
1/19 (5.3)
1/10 (10)
NA
1/14 (7.1)
NA
NA
3/19 (15.8)
NA
NA
1/14 (7.1)
0/16 (0)
NA
7/19 (36.8)
3/16 (18.8)
NA
4/14 (28.6)
24/31 (77.4)
22/28 (78.6)
33/40 (82.5)
34/40 (85)
4/31 (12.9)
2/28 (7.1)
6/31 (19.4)
5/28 (17.9)
1/31 (3.2)
4/28 (14.3)
7/8 (87.5)
7/10 (70)
7/14 (50)
5/7 (71.4)
7/11 (63.6)
8/11 (72.7)
7/8 (87.5)
7/10 (70)
8/14 (57.1)
0/8 (0)
3/10 (30)
1/14 (7.1)
0/7 (0)
0/11 (0)
0/11 (0)
1/8 (12.5)
0/10 (0)
0/14 (0)
1/7 (14.3)
0/11 (0)
0/11 (0)
NA
3/10 (30)
2/14 (14.3)
NA
5/11 (45.5)
4/11 (36.4)
Figure 2. Odds ratios of treatment success, and bacteriological success, for studies comparing quinolones with -lactams
for the treatment of osteomyelitis; (A) treatment success, (B)
bacteriological success (Vertical
line no difference point between the 2 regimens. Square
odds ratio; the size of each
square denotes the proportion of
information given by each trial.
Diamond pooled odds ratio for
all RCTs. Horizontal lines 95%
CI).
6/7 (85.7)
7/11 (63.6)
9/11 (81.8)
Figure 3. Odds ratios of (A) superinfections, (B) relapses, and (C) patients with adverse events for studies comparing quinolones with
-lactams for the treatment of osteomyelitis. (Vertical line no difference point between the 2 regimens. Square odds ratio; the size
of each square denotes the proportion of information given by each trial. Diamond pooled odds ratio for all RCTs. Horizontal lines
95% CI).
even in environments with low pH, such as it is frequently encountered in sites of infection.19 However,
caution is advised in the elderly or in individuals with
renal failure or liver disease, in whom the peak serum
concentrations rise accordingly.6
The aforementioned properties of fluoroquinolones
could be considered as potential advantages over other
agents, which are expected to result in increased efficacy
in clinical practice. Nevertheless, in our meta-analysis,
there was no difference in efficacy between fluoroquinolones and -lactams for the treatment of osteomyelitis.
This finding may be appointed to the inherent characteristics of the osteomyelitis pathophysiology. Specifically,
active osteomyelitis lesions are surrounded by necrotic
tissues and debris and have poor vascular supply, thus
preventing adequate distribution of the antimicrobial
drugs in the site, as well as the adequacy of the host
inflammatory response. Also, since osteomyelitis frequently coexists with diabetes mellitus, diabetic microangiopathy could further hamper fluoroquinolone
penetration in the inflamed tissues.
It should be mentioned though, that fluoroquinolones
are one of the few oral agents available for satisfactory
coverage of the gram-negative pathogens that are frequently a cause of osteomyelitis. Moreover, fluoroquinolones initially exhibited substantial antistaphylococcal
activity, but resistance to these agents has greatly increased in parallel with the rise in the prevalence of methicillin-resistant Staphylococcus aureus. So, for empirical therapy, fluoroquinolones should be better used in
combination with an agent with sufficient activity
against gram-positive pathogens. Nevertheless, as the
findings of our study show, the value of fluoroquinolones
for the treatment of osteomyelitis lies in the option that
they offer for treatment in an outpatient setting, without
the cost of inferior treatment outcomes nor an increased
rate of adverse events.
There are certain limitations though that should be
taken under consideration in our meta-analysis. First,
the total number of patients included in our metaanalysis is rather small to allow for true differences between compared treatments to be shown. Second, the
included studies were conducted more than 15 years ago,
most of them in a time span of 5 years between 1987 and
1991. Since then, trends of susceptibility of the pathogens to the examined agents may have changed considerably, so it is arguable if the results of these studies
could be applied in current clinical practice. Third, the
quality assessment of the included studies yielded a score
of 2, which equals to the cut-off point for considering
them as good quality RCTs. It should be also emphasized
that factors such as acuity of osteomyelitis, host immu-
References
1. Waldvogel FA, Medoff G, Swartz MN. Osteomyelitis: a review of clinical
features, therapeutic considerations and unusual aspects. N Engl J Med
1970;282:198 206.