J Antimicrob Chemother 2016; 71: 862 – 870
doi:10.1093/jac/dkv415 Advance Access publication 24 December 2015
Ceftaroline fosamil versus ceftriaxone for the treatment of
community-acquired pneumonia: individual patient data
meta-analysis of randomized controlled trials
Maria Taboada1*, David Melnick2†, Joseph P. Iaconis3, Fang Sun4, Nan Shan Zhong5, Thomas M. File6,7,
Lily Llorens8‡, H. David Friedland8§ and David Wilson1
2
1
AstraZeneca Research & Development, Biometrics & Information Sciences, Parklands, Alderley Park, Macclesfield, SK10 4TG, UK;
AstraZeneca, 1800 Concord Pike, Wilmington, DE 19803, USA; 3AstraZeneca, 35 Gatehouse Drive, Waltham, MA 02451, USA; 4AstraZeneca
China, 1168 Nan Jing Xi Road, Shanghai, 200041, P.R. China; 5State Key Laboratory of Respiratory Diseases, 1st Affiliate Hospital of
Respiratory Disease, Guangzhou Medical University, 151 Yan Jiang Road, Guangzhou, 501120, P.R. China; 6Northeast Ohio Medical
University, Rootstown, OH 44272, USA; 7Summa Health System, Akron, OH 44304, USA; 8Cerexa Inc., 2100 Franklin St,
Oakland, CA 94612, USA
*Corresponding author. Tel: +44 (0)1625 515770; E-mail: [email protected]
†Present address: Allergan, Inc., Harborside Financial Center, Plaza V, Suite 1900, Jersey City, NJ 07311, USA.
‡Present address: 302 Laguna Vista, Alameda, CA 94501, USA.
§Present address: Forest Research Institute, Inc., Harborside Financial Center, Plaza V, Suite 1900, Jersey City, NJ 07311, USA.
Received 3 June 2015; returned 19 July 2015; revised 5 November 2015; accepted 5 November 2015
Background: We conducted a meta-analysis of clinical trials of adults hospitalized with pneumonia outcomes
research team (PORT) risk class 3 –4 community-acquired pneumonia (CAP) receiving ceftaroline fosamil versus
ceftriaxone.
Methods: Three Phase III trials (clinicaltrials.gov registration numbers NCT00621504, NCT00509106 and
NCT01371838) including 1916 hospitalized patients with CAP randomized 1:1 to empirical ceftaroline fosamil
(600 mg every 12 h) or ceftriaxone (1 – 2 g every 24 h) for 5 – 7 days were included in the meta-analysis.
Primary outcome was clinical response at the test-of-cure visit (8 –15 days after end of treatment) in the PORT
risk class 3 –4 modified ITT (MITT) and clinically evaluable (CE) populations. Data were tested for heterogeneity
(x2 test) and, if not significant, results were pooled and OR and 95% CI constructed. A logistic regression analysis
assessed factors impacting cure rate and treatment interactions.
Results: Clinical cure rates in each trial consistently favoured ceftaroline fosamil versus ceftriaxone, with no
evidence of heterogeneity. In the meta-analysis, ceftaroline fosamil was superior to ceftriaxone in the MITT
(OR: 1.66; 95% CI 1.34, 2.06; P, 0.001) and CE (OR: 1.65; 95% CI 1.26, 2.16; P, 0.001) populations. Results
were consistent across various patient- and disease-related factors including patients’ age and PORT score.
Prior antimicrobial use within 96 h of starting study treatment was associated with diminished differences in
cure rates between treatments.
Conclusions: Ceftaroline fosamil was superior to ceftriaxone for empirical treatment of adults hospitalized with
CAP. Receipt of prior antimicrobial therapy appeared to diminish the observed treatment effect.
Introduction
Lower respiratory tract infections are a leading cause of global
mortality, responsible for an estimated 3.1 million deaths in
2012,1 and community-acquired pneumonia (CAP) results in substantial morbidity, mortality and healthcare costs.2 – 5 Sustained
and coordinated efforts have been made to improve the management of CAP, embodied in the development of international management guidelines and evidence-based quality indicators.6,7 For
CAP of sufficient severity to warrant hospitalization and intravenous (iv) antibiotics [usually defined as a pneumonia outcomes
research team (PORT) class ≥3], treatment recommendations
include cephalosporins such as ceftriaxone administered at
doses of 1 –2 g every 24 h.6,7 However, few clinical trials comparing empirical therapies in CAP have demonstrated clear differences in outcomes between treatments because most trials are
designed only to demonstrate non-inferiority, limiting the
strength of guideline recommendations. Moreover, there is
# The Author 2015. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: [email protected]
862
JAC
Systematic review
evidence that certain aspects of the design of CAP trials, specifically the receipt of prior antibiotic therapy, might bias results
toward non-inferiority.8
Ceftaroline is a cephalosporin with an in vitro spectrum of
activity that covers most of the common bacterial pathogens
associated with CAP. The prodrug ceftaroline fosamil has
been approved in the USA since 2010 for the treatment of
community-acquired bacterial pneumonia and acute bacterial
skin and skin structure infections, and is approved for similar indications in Europe and elsewhere. Approvals of ceftaroline fosamil
in the CAP indication were based on two Phase III multinational,
randomized controlled trials conducted in patients hospitalized
with CAP, FOCUS 1 (clinicaltrials.gov registration number
NCT00621504) and FOCUS 2 (clinicaltrials.gov registration number NCT00509106), in which ceftaroline fosamil 600 mg every
12 h demonstrated non-inferiority to ceftriaxone 1 g every 24 h,
with differences in clinical cure rates at the test-of-cure (TOC)
visit favouring ceftaroline fosamil in each trial.9,10 In a pooled analysis of FOCUS 1 and 2, the lower bound of the 95% CI for the treatment difference was .0% (i.e. the treatment difference met
standard criteria for superiority).11 In a subsequent Phase III randomized controlled trial in Asian patients with PORT class 3– 4 CAP
(clinicaltrials.gov registration number NCT01371838), ceftaroline
fosamil 600 mg every 12 h demonstrated superiority to ceftriaxone 2 g every 24 h.12
We conducted a meta-analysis of FOCUS 1 and 2 and the Asia
CAP trial to compare further the efficacy of ceftaroline fosamil versus ceftriaxone in the treatment of CAP. We also used logistic
regression to evaluate the impact of baseline patient variables
on cure rate and treatment interactions.
Methods
Trials included
The meta-analysis included all of the Phase III randomized controlled
trials of ceftaroline fosamil versus ceftriaxone for the treatment of CAP
conducted to date, FOCUS 1,9 FOCUS 210 and the Asia CAP trial. 12 In
most respects, the trials were similar: adult patients with acute infection
consistent with CAP requiring hospitalization and iv antibiotics were randomized (1:1) to receive ceftaroline fosamil 600 mg every 12 h or ceftriaxone
for 5 – 7 days; all three trials were conducted in accordance with the
Declaration of Helsinki and national and institutional standards; all
patients (or their legally authorized representatives) provided written
informed consent; and all study sites received approval for study conduct
from their independent Ethics Committee or Institutional Review Board. A
notable difference between the trials was the ceftriaxone dose (1 g every
24 h in FOCUS 1 and 2 and 2 g every 24 h in the Asia CAP trial). Pneumonia
severity was evaluated prior to enrolment by use of the PORT risk classification;13 patients with scores of 51– 130 (i.e. PORT risk class 2– 4) were eligible for inclusion in FOCUS 1 and 2, and those with scores of 71–130 (PORT
risk class 3–4) could be included in the Asia CAP trial. Patients with PORT risk
class 2 CAP were excluded from the primary efficacy analyses in FOCUS 1 and
2, and from this meta-analysis. Patients in each trial could have received a
single dose of an oral or iv short-acting antibiotic for CAP within 96 h prior to
study entry. In FOCUS 1, all patients also received oral clarithromycin
(2×500 mg doses administered on day 1).
Key exclusion criteria in each trial included CAP suitable for outpatient
therapy with oral antimicrobial agent(s); need for intensive care unit
admission; confirmed/suspected infection attributable to sources other
than community-acquired bacterial pathogens (e.g. ventilator-associated
pneumonia, hospital-acquired pneumonia, visible/gross aspiration
pneumonia, or suspected viral, fungal or mycobacterial lung infection);
non-infectious causes of pulmonary infiltrates (e.g. pulmonary embolism,
chemical pneumonitis from aspiration, hypersensitivity pneumonia or congestive heart failure); pleural empyema (not including non-purulent parapneumonic effusions); microbiologically documented infection with a
ceftriaxone-resistant pathogen or epidemiological/clinical context suggesting infection with a bacterial pathogen resistant to one or both
study treatments (e.g. Pseudomonas aeruginosa or MRSA).
Endpoints and analysis populations
The primary endpoint in each trial was clinical response at the TOC visit
(8 – 15 days after the last dose of study drug). In FOCUS 1 and 2, the
modified intent-to-treat population efficacy (MITTE) and the clinically
evaluable (CE) populations were co-primary. In the Asia CAP trial, the
CE population was the primary analysis population. The FOCUS 1 and 2
MITTE populations comprised all randomized patients who received
any amount of study drug and who had PORT risk class 3 – 4 CAP. The
Asia CAP modified ITT (MITT) population comprised all randomized
patients who received any amount of study drug. Thus, the FOCUS 1
and 2 MITTE populations are equivalent to the Asia CAP MITT population
and are referred to henceforth as the MITT population. For the
meta-analysis, clinical response at TOC was assessed in the pooled CE
and MITT populations. Patients in the MITT population who met minimal
disease criteria and had ≥1 bacterial pathogen commonly associated
with CAP identified at baseline were included in the microbiological modified MITT (mMITT) population, and those who met criteria for both the CE
and mMITT population were included in the microbiologically evaluable
(ME) population.
Microbiological assessments
Bacterial identification methods included culture of clinical specimens
(sputum, tracheal aspirates and blood), as well as serological testing for
Mycoplasma pneumoniae and Chlamydophila pneumoniae and urinary
antigen testing for Legionella pneumophila and Streptococcus pneumoniae. Bacterial isolates obtained by local laboratories were sent to a central
laboratory for confirmatory identification and susceptibility testing.
Statistical methods
For the meta-analysis, tests of heterogeneity of the treatment effects
across the three trials were conducted (based on the x2 test), and if no
significant heterogeneity was found, results from each trial were pooled
and OR and 95% CI for clinical cure at the TOC visit were constructed.
The meta-analysis was performed with study as a fixed effect, calculated
using an inverse variance approach. A sensitivity analysis in which
study was considered a random effect was also performed using the
DerSimonian and Laird method.14 Univariate logistic regression models
were fitted to assess various explanatory factors for the clinical response
and corresponding treatment by factor interaction terms. The following
baseline factors were assessed for their association with clinical response:
study (i.e. FOCUS 1, FOCUS 2 or Asia CAP), age (,65 or ≥65 years), gender,
region, BMI (≤25 or .25 kg/m2), receipt of prior systemic antibiotics, presence of bacteraemia, PORT risk class, presence of pleural effusion, creatinine
clearance, fever, presence of abnormal white blood cells or abnormal immature neutrophils, tachypnoea, hypoxia, smoking status, Gram stain result (no
visible pathogens; Gram-positive pathogens only; Gram-negative pathogens
only; or mixed Gram-positive/Gram-negative pathogens), presence of common bacterial pathogens (none identified; common bacterial pathogens
only; or both common bacterial and atypical pathogens). Patients with atypical pathogens only were excluded from the CE population and included in
the no pathogen group in the MITT population.
863
864
24 (2.5)
345 (36.1)
26 (2.7)
317 (33.0)
5 (1.3)
85 (22.3)
3 (0.8)
80 (21.0)
11 (4.0)
117 (42.9)
15 (5.2)
100 (34.6)
8 (2.7)
137 (47.1)
Bacteraemia, n (%)
Prior antibiotic use, n (%)
9 (3.0)
143 (47.7)
618 (64.7)
337 (35.3)
615 (64.0)
346 (36.0)
265 (69.4)
117 (30.6)
255 (66.9)
126 (33.1)
171 (62.6)
102 (37.4)
170 (58.8)
119 (41.2)
190 (65.3)
101 (34.7)
PORT risk class, n (%)
PORT 3
PORT 4
182 (60.7)
118 (39.3)
532 (55.7)
402 (42.1)
15 (1.6)
5 (0.5)
1 (0.1)
538 (56.0)
400 (41.6)
17 (1.8)
5 (0.5)
1 (0.1)
—
382 (100.0)
—
—
—
—
381 (100.0)
—
—
—
264 (96.7)
4 (1.5)
—
5 (1.8)
—
278 (96.2)
5 (1.7)
—
5 (1.7)
1 (0.3)
260 (89.3)
14 (4.8)
17 (5.8)
—
—
268 (89.3)
16 (5.3)
15 (5.0)
—
1 (0.3)
62.9
499 (51.9)
627 (65.2)
24.4
380 (39.5)
65.8
236 (61.8)
272 (71.2)
22.0
78 (20.4)
66.1
226 (59.3)
265 (69.6)
22.0
82 (21.5)
62.0
133 (48.7)
175 (64.1)
26.8
169 (61.9)
60.6
130 (45.0)
175 (60.6)
26.0
154 (53.3)
61.2
148 (49.3)
191 (63.7)
26.5
163 (54.3)
Ceftaroline
fosamil (n ¼289)
Ceftaroline
fosamil (n¼291)
Race, n (%)
White
Asian
Black or African American
American Indian/Alaskan
other
Rates of bacterial identification across the three trials were relatively low: the mMITT population comprised 23% – 29% of the
MITT population and the ME population 16% – 27%. In addition,
not all pathogens identified at baseline were available for microbiological analysis (e.g. some were identified by urinary antigen
test only) and some isolates were not viable for culture upon
receipt at the central laboratory. MICs of study treatments for
common CAP pathogens isolated at baseline and available for
61.0
143 (49.1)
187 (64.3)
26.0
144 (49.5)
Microbiological outcomes
FOCUS 1
Of the study-, patient- and disease-related factors included in the
logistic regression analysis, gender, prior systemic antibiotic use,
presence of pleural effusion, bacteraemia, Gram stain result and
smoking status had some evidence of impact on clinical cure rates
in both treatment groups (Table S1, available as Supplementary
data at JAC Online). Prior systemic antibiotic use within 96 h of
randomization, presence of pleural effusion, fever and smoking
status each had differential treatment effects on clinical cure
rates (Figure 3). Ceftaroline fosamil was superior to ceftriaxone
in both the MITT and CE populations in patients who did not
receive prior antimicrobial therapy, but not in those who received
prior antimicrobial therapy (Figure 3). There were no observed differences among other factors investigated not included in Figure 3
(including age or gender), with the data consistently indicating
superiority for ceftaroline fosamil across these subgroups (data
not shown).
Table 1. Patients’ baseline characteristics (MITT population)
Subgroup analyses
Ceftriaxone
(n¼300)
FOCUS 2
Ceftriaxone
(n ¼273)
Asia CAP
In each trial, a greater proportion of patients were clinically cured at
TOC in the ceftaroline fosamil group versus the ceftriaxone group in
both the MITT and CE populations (Figure 1). Of note, increasing the
dose of ceftriaxone in the Asia CAP trial (2 g every 24 h) relative
to FOCUS 1 and 2 (1 g every 24 h) was not associated with an
increased clinical cure rate in the ceftriaxone arm and did not
diminish the observed treatment difference between arms.
Meta-analysis using the fixed-effect method (Figure 2) demonstrated consistent results between trials with no evidence of heterogeneity (I 2 ¼0%). Based on the meta-analysis, ceftaroline fosamil
was concluded to be superior to ceftriaxone in both the MITT (OR
1.66, 95% CI 1.34, 2.06; P,0.001) and CE populations (OR 1.65,
95% CI 1.26, 2.16; P, 0.001). Results of the random effect
meta-analysis were similar (data not shown).
Mean age, years
Age ≥65 years, n (%)
Male, n (%)
Mean BMI, kg/m2
BMI ≥25 kg/m2, n (%)
Overall clinical response rates
Ceftaroline
fosamil (n¼381)
Ceftriaxone
(n¼382)
Pooled
The pooled MITT population comprised 1916 patients (ceftaroline
fosamil, n¼ 961; ceftriaxone, n¼ 955), and the pooled CE population comprised 1406 patients (ceftaroline fosamil, n ¼717; ceftriaxone, n¼689). In total, 19 patients, 7 in FOCUS 1 (5 assigned to
ceftaroline fosamil and 2 to ceftriaxone), 5 in FOCUS 2 (2 ceftaroline fosamil and 3 ceftriaxone) and 7 in the Asia CAP trial (3 ceftaroline fosamil and 4 ceftriaxone), were excluded from the MITT
population because of lack of receipt of study drug after randomization. Patients’ baseline characteristics were well balanced
across treatment groups within each trial, and were generally
comparable across the three trials (Table 1).
Ceftaroline
fosamil (n ¼961)
Ceftriaxone
(n ¼955)
Results
63.3
517 (54.1)
638 (66.8)
24.8
410 (42.9)
Systematic review
JAC
Systematic review
n/N (%) patients with clinical cure:
ceftaroline fosamil vs ceftriaxone
CE population
FOCUS 1
194/224 (86.6%) vs 183/234 (78.2%)
FOCUS 2
193/235 (82.1%) vs 166/215 (77.2%)
Asia CAP
217/258 (84.1%) vs 178/240 (74.2%)
MITT population
FOCUS 1
244/291 (83.8%) vs 233/300 (77.7%)
FOCUS 2
235/289 (81.3%) vs 206/273 (75.5%)
Asia CAP
305/381 (80.1%) vs 256/382 (67.0%)
–10
–5
0
5
10
Difference, %
Favours
ceftriaxone
15
25
Favours
ceftaroline fosamil
Figure 1. Difference in clinical cure and 95% CI (ceftaroline fosamil versus ceftriaxone) at the TOC visit in the individual CAP trials (CE and MITT
populations).
Study or subgroup
Ceftaroline fosamil Ceftriaxone
Cures Total Cures Total
CE population
224
183
FOCUS 1
194
235
166
FOCUS 2
193
258
178
Asia CAP
217
Fixed effect
Total events
527
604
Heterogeneity: c2 = 1.06, df = 2 (P = 0.59); I2 = 0.00%
MITT population
FOCUS 1
244
291
233
FOCUS 2
235
289
206
Asia CAP
305
381
256
Fixed effect
Total events
784
695
Heterogeneity: c2 = 1.92, df = 2 (P = 0.38); I2 = 0.00%
Weight
OR (95% CI)
OR (95% CI)
234
215
240
29.4%
33.7%
36.9%
100.0%
1.80 (1.11, 2.98)
1.36 (0.86, 2.16)
1.84 (1.19, 2.88)
1.65 (1.26, 2.16)
300
273
382
27.6%
28.9%
43.5%
100.0%
1.49 (0.99, 2.27)
1.42 (0.95, 2.13)
1.98 (1.42, 2.75)
1.66 (1.34, 2.06)
0.1
Favours ceftriaxone
1
10
Favours ceftaroline fosamil
Figure 2. ORs and 95% CI of clinical cure rates (ceftaroline fosamil versus ceftriaxone) at the TOC visit in the individual CAP trials and fixed effects
meta-analysis (CE and MITT populations).
microbiological analysis are shown in Table 2. Clinical responses by
pathogen and infection composition (Table 3) generally reflected
the overall results. Decreased pathogen susceptibility (defined as
≥4-fold increase in the study drug MIC relative to baseline) occurred
in 1 patient in the mMITT population in the ceftriaxone group, and no
patients in the ceftaroline fosamil group. Rates of superinfection,
colonization, new infection and reinfection/recurrence in the ME population are summarized in Table S2.
Adverse events
Safety outcomes were generally consistent across the trials. In
total, 460 of 994 (46.3%) ceftaroline fosamil and 444 of 998
(44.5%) ceftriaxone recipients in the safety population experienced ≥1 adverse event, and 34 of 994 (3.4%) and 32 of 998
(3.2%) patients, respectively, discontinued the trials due to
an adverse event. Mortality from all causes at 30 days after the
start of study treatment occurred in 15 patients (1.5%) in
each group.
Discussion
For various reasons, antimicrobial clinical trials are usually designed
to determine non-inferiority of an experimental intervention compared with an established standard of care. A meta-analysis provides
an opportunity to assess the consistency of treatment effects, as well
as to explore the potential for superiority that might not be possible
within single trial non-inferiority settings. This meta-analysis demonstrated that ceftaroline fosamil was superior to ceftriaxone, an
established standard of care recommended in international treatment guidelines,6,7 as empirical treatment for adult patients hospitalized with PORT risk class 3 – 4 CAP. This is significant, as the
expectation from most antibiotic trials and meta-analyses would
be that there would be no difference between treatment groups.
865
Systematic review
Study or subgroup
CE population
Prior antibiotics
Yes
No
Pleural effusion
Yes
No
Fever
Yes
No
Smoking
Previous
Current
Never
PORT risk class
3
4
Gram-stain resulta
No pathogens
Gram-positive pathogens only
Gram-negative pathogens only
Type of culture
No pathogen
Monomicrobial
Polymicrobial
Bacterial pathogens identified
None identified
Common bacterial pathogens only
Both common bacterial and
atypical pathogens
MITT population
Prior antibiotics
Yes
No
Pleural effusion
Yes
No
Fever
Yes
No
Smoking
Previous
Current
Never
PORT risk class
3
4
Gram-stain resulta
No pathogens
Gram-positive pathogens only
Gram-negative pathogens only
Type of culture
No pathogen
Monomicrobial
Polymicrobial
Bacterial pathogens identified
None identified
Common bacterial pathogens only
Both common bacterial and
atypical pathogens
Ceftaroline fosamil Ceftriaxone
Cures
Total
Cures Total
OR (95% CI)
OR (95% CI)
194
410
234
483
193
334
239
450
1.16 (0.72, 1.85)
1.95 (1.40, 2.72)
81
523
120
597
83
444
126
563
1.09 (0.64, 1.87)
1.89 (1.37, 2.60)
309
295
350
367
264
263
343
346
2.26 (1.50, 3.42)
1.31 (0.91, 1.87)
172
139
293
219
167
331
158
111
258
199
160
330
0.95 (0.59, 1.55)
1.99 (1.15, 3.43)
2.11 (1.37, 3.26)
397
207
460
257
343
184
443
246
1.83 (1.29, 2.59)
1.39 (0.91, 2.12)
369
56
89
480
85
106
423
81
85
506
96
98
1.54 (1.12, 2.12)
2.80 (1.36, 5.75)
1.17 (0.52, 2.65)
369
116
42
480
150
59
423
134
47
506
154
57
1.54 (1.12, 2.12)
1.94 (1.05, 3.59)
1.78 (0.70, 4.53)
370
137
20
477
184
28
428
154
22
509
181
27
1.54 (1.11, 2.12)
1.96 (1.15, 3.34)
1.14 (0.27, 4.82)
250
534
317
644
262
433
345
610
1.24 (0.85, 1.82)
1.84 (1.39, 2.43)
114
670
164
797
104
591
171
784
1.40 (0.89, 2.21)
1.65 (1.27, 2.15)
413
371
481
480
360
335
485
470
1.92 (1.37, 2.68)
1.36 (1.00, 1.85)
216
181
387
283
224
454
194
142
359
257
226
472
1.04 (0.68, 1.58)
2.17 (1.39, 3.39)
1.67 (1.17, 2.37)
520
264
615
346
460
235
618
337
1.24 (0.85, 1.82)
1.84 (1.39, 2.43)
497
71
113
684
107
143
576
86
106
711
104
127
1.61 (1.25, 2.08)
2.41 (1.25, 4.63)
1.28 (0.67, 2.46)
497
145
53
684
193
78
576
154
54
711
181
69
1.61 (1.25, 2.08)
1.85 (1.09, 3.14)
1.64 (0.75, 3.56)
499
165
31
682
229
44
583
173
28
714
207
40
1.64 (1.27, 2.12)
1.94 (1.21, 3.11)
0.94 (0.34, 2.61)
0.1
Favours ceftriaxone
1
10
Favours ceftaroline fosamil
Figure 3. Fixed effects meta-analysis of ORs and 95% CI of clinical cure (ceftaroline fosamil versus ceftriaxone) at the TOC visit in various patient
subgroups (CE and MITT populations). aORs and 95% CIs were not calculated for the mixed Gram-positive/Gram-negative subgroups because the
numbers were too small.
866
JAC
Systematic review
Table 2. MICs of study treatments for common CAP pathogens isolated at baseline pooled across trials (mMITT population)
Ceftaroline MIC (mg/L)
Ceftriaxone MIC (mg/L)
No. of isolates
testeda
range
MIC50b
MIC90b
range
MIC50b
MIC90b
Streptococcus pneumoniae
non-MDRSP
MDRSP
PSSP
PISP
PRSP
87c
67
20
81
2
3
≤0.004 to 0.25
0.008 to 0.03
≤0.004 to 0.25
≤0.004 to 0.12
0.06 to 0.12
0.06 to 0.25
≤0.015
≤0.015
≤0.015
≤0.015
—
—
0.03
≤0.015
0.12
≤0.015
—
—
≤0.015 to 2
≤0.015 to 0.5
≤0.015 to 2
≤0.015 to 2
0.5 to 1
0.5 to 2
≤0.015
≤0.015
0.5
≤0.015
—
—
0.5
0.06
1
0.12
—
—
Staphylococcus aureus d
MSSA
MRSA
63
57
6
0.12 to 0.5
0.12 to 0.5
0.25 to 0.5
0.25
0.25
—
0.5
0.25
—
2 to 64
2 to 4
4 to 64
4
4
—
4
4
—
Haemophilus influenzae
Haemophilus parainfluenzae
Klebsiella pneumoniae
Escherichia coli
52
38
61
38
≤0.008 to 0.5
≤0.004 to 1
≤0.03 to .16
≤0.03 to .16
0.015
0.008
0.12
0.06
0.03
0.12
.16
.16
≤0.008 to 0.03
≤0.008 to 0.25
≤0.03 to .64
≤0.03 to .64
≤0.008
≤0.015
0.06
0.06
0.03
0.06
.64
.64
Pathogen
MDRSP, MDR S. pneumoniae; PISP, penicillin-intermediate S. pneumoniae; PRSP, penicillin-resistant S. pneumoniae; PSSP, penicillin-susceptible S.
pneumoniae.
a
Not all pathogens identified at baseline were available for MIC testing.
b
Assessed for isolates where n≥ 10.
c
Eighty-six S. pneumoniae isolates were tested for ceftaroline MIC.
d
For Staphylococcus aureus, strains of MRSA and MSSA are considered distinct pathogens. Therefore, subjects with both MRSA and MSSA are counted
twice in these summaries. The same applies for S. pneumoniae, and strains of PISP, PSSP and PRSP.
Table 3. Clinical response rates at the TOC visit by pathogen in the individual trials and pooled across trials (ME population)
Patients, n/N (%)
FOCUS 1
FOCUS 2
Asia CAP
Pooled
ceftaroline
fosamil
(n ¼69)
ceftriaxone
(n¼71)
ceftaroline
fosamil
(n ¼85)
ceftriaxone
(n¼76)
ceftaroline
fosamil
(n ¼57)
ceftriaxone
(n¼62)
ceftaroline
fosamil (n¼ 211)
ceftriaxone
(n¼209)
All patients
monomicrobial
polymicrobial
62/69 (89.9)
51/57 (89.5)
11/12 (91.7)
54/71 (76.1)
41/51 (80.4)
13/20 (65.0)
69/85 (81.2)
44/54 (81.5)
25/31 (80.6)
57/76 (75.0)
39/51 (76.5)
18/25 (72.0)
50/57 (87.7)
39/43 (90.7)
11/14 (78.6)
47/62 (75.8)
36/48 (75.0)
11/14 (78.6)
184/211 (87.2)
137/154 (89.0)
47/57 (82.5)
166/209 (79.4)
120/150 (80.0)
46/59 (78.0)
Gram-positive
S. aureus
S. pneumoniae
8/10 (80.0)
21/24 (87.5)
7/12 (58.3)
18/27 (66.7)
10/15 (66.7)
33/39 (84.6)
8/15 (53.3)
23/32 (71.9)
4/4 (100.0)
19/22 (86.4)
2/4 (50.0)
13/15 (86.7)
22/29 (75.9)
73/85 (85.9)
17/31 (54.8)
54/74 (73.0)
5/6 (83.3)
6/8 (75.0)
9/10 (90.0)
3/4 (75.0)
2/4 (50.0)
13/15 (86.7)
9/9 (100.0)
6/6 (100.0)
4/6 (66.7)
11/12 (91.7)
6/7 (85.7)
7/8 (87.5)
3/3 (100.0)
11/12 (91.7)
0
11/14 (78.6)
5/6 (83.3)
6/7 (85.7)
4/6 (66.7)
12/16 (75.0)
13/15 (86.7)
26/30 (86.7)
16/16 (100.0)
24/27 (88.9)
14/18 (77.8)
23/27 (85.1)
19/23 (82.6)
22/28 (78.6)
Gram-negative
E. coli
H. influenzae
H. parainfluenzae
K. pneumoniae
8/8 (100.0)
2/3 (66.7)
7/7 (100.0)
7/7 (100.0)
In addition, the similar study designs and lack of heterogeneity
between the trials justified pooling the microbiological data to
increase the robustness of the dataset, as pathogen isolation
rates were relatively low within the individual trials. In this pooled
analysis, per-pathogen clinical response rates in the ME population
generally reflected the meta-analysis results for the CE and MITT
populations, numerically favouring ceftaroline fosamil.
A further problem with clinical trials for CAP in particular,
which was highlighted by a subgroup analysis of two trials of daptomycin versus ceftriaxone,8 is that because of the need for
867
Systematic review
emergent therapy as soon as patients are diagnosed, many
patients enrolled into clinical trials have already received some
antimicrobial treatment considered effective for CAP by the time
they are randomized. As early treatment is considered an important determinant of outcome in CAP trials,15 a consequence of
such prior therapy may be to dilute or even potentially negate
any observed treatment difference between comparators. This
problem has also been noted by the FDA in its recent guidance
on CAP antimicrobial clinical trial design. 16 The current
meta-analysis also provided an opportunity to explore the impact
of prior antimicrobial therapy. Across many patient-, disease- and
study-related factors, prior antimicrobial therapy was consistently
associated with a differential treatment effect across the three
trials (the other factors where some differences were observed
were pleural effusion, fever and smoking status). The overall treatment difference between ceftaroline fosamil and ceftriaxone in
patients who received prior antimicrobial therapy was diminished
such that the treatments appeared approximately similar, whereas
in patients who did not receive prior antimicrobial therapy, there
was a clear difference in favour of ceftaroline fosamil (P,0.001).
This suggests that future trials may need to balance the practical
considerations that many patients would have received prior antibiotics with the knowledge that this prior therapy may dilute the
treatment effect.
A limitation of the method chosen for the meta-analysis
is that it uses the underlying assumption that the results
of the three studies are ‘poolable’, based on a heterogeneity
test, which has low power; moreover, the DerSimonian and
Laird method 14 is impacted by the small number of studies.
However, a sensitivity analysis using the Hartung – Knapp
method17 produced similar results and demonstrated the superiority of ceftaroline fosamil versus ceftaroline: 95% CIs for the OR
were 1.04, 2.65 for the MITT population and 1.08, 2.65 for the CE
population; the superiority result can therefore be considered
conclusive. In this meta-analysis, an MITT analysis rather the
ITT populations from the individual trials was used for the following reasons: (i) to ensure consistency across the trials in analysing the PORT risk class 3 – 4 population and with the published
analyses of these trials9 – 11 (it was considered clinically justified
to exclude PORT risk class 2 CAP patients, since unless hospitalized for other reasons, such patients would normally be treated
as outpatients for CAP, and hence would be unlikely in practice to
be considered for iv antibiotic therapy with either ceftaroline
fosamil or ceftriaxone); and (ii) the only reason for exclusion
from the MITT population was lack of receipt of study drug
(and the numbers of such exclusions were balanced across
treatment groups, and thus are not expected to impact the
findings).
As the meta-analysis illustrated consistency of treatment
effect, the observed difference in efficacy between ceftaroline
fosamil and ceftriaxone appears unlikely to reflect differences
in the patient population (which could potentially influence
pharmacokinetics/exposure to one or both treatments) or
regional differences in bacterial aetiology or resistance patterns
(which could result in varying responsiveness to study treatments). It is notable that susceptibility testing demonstrated
that ceftaroline was 16-fold more potent than ceftriaxone
against S. pneumoniae and S. aureus (MSSA) isolates obtained
at baseline across the three trials; similar potency differences
have been documented in antimicrobial surveillance studies
868
contemporaneous to the FOCUS trials.18,19 Clinical response
rates by pathogen for patients infected with S. pneumoniae
and S. aureus in the ME population in the individual trials and
pooled analysis also generally favoured ceftaroline fosamil. In
the case of S. pneumoniae the treatment difference met the
standard criteria for superiority in the pooled analysis (the
lower limit of the 95% CI of the difference being .0%). In the
smaller S. aureus subgroup, CIs of the treatment difference
were wider and crossed 0 (data not shown). Since patients
infected with suspected/confirmed pathogens that were resistant to one (e.g. MRSA) or both study treatments (e.g.
Pseudomonas aeruginosa) were excluded from the ME population,
the observed differences in per-pathogen response rates for S.
pneumoniae and S. aureus appear to reflect differences in potency
within the range of MICs currently defined as susceptible for each
agent. Differences in potency may also underlie the observed rates
of superinfection and development of decreased susceptibility during treatment.
It is also possible that differences in protein binding and lung
tissue penetration between ceftaroline (active metabolite of the
pro-drug ceftaroline fosamil) and ceftriaxone, which could affect
pharmacodynamics at the site of infection, may have contributed
to the observed results. Ceftaroline plasma protein binding is typically about 20%,20 and that of ceftriaxone is 95%.21 However,
data on pulmonary tissue penetration for these antibiotics are
somewhat limited. In a study in patients with pleural effusions,
the ceftriaxone maximum concentration in pleural fluid was
10% of that in serum following a single iv dose of ceftriaxone
1 g, and the corresponding area under the curve was 18%.22
In a separate study in healthy subjects, the penetration of ceftaroline into epithelial lining fluid, defined as the ratio of free ceftaroline exposure in epithelial lining fluid to plasma on day 4
following a 600 mg every 12 h ceftaroline fosamil dosing regimen, was 22.5%.23 The pharmacodynamic index most closely
associated with efficacy for b-lactams, including ceftaroline
and ceftriaxone, is the proportion of the dosing interval that the
concentration of the free drug is above the MIC of the infecting
pathogen (i.e. %fT.MIC). Hence, although the relationship
between lung tissue penetration and clinical outcome is uncertain,24 it is plausible that even modest differences in tissue penetration and/or protein binding could have clinically relevant
pharmacodynamic effects, particularly when considering the
respective dosing schedules for ceftaroline fosamil (every 12 h)
and ceftriaxone (every 24 h) used in the trials.
The implications of these findings for the management of CAP
are significant. First, based on the meta-analysis, it is clear that
ceftaroline fosamil is superior to ceftriaxone for the treatment
of CAP in hospitalized patients and should be considered as a
replacement to ceftriaxone for the cephalosporin component of
empirical antibiotic regimens in adult patients hospitalized with
CAP. Second, the evidence from past antimicrobial trials should
be evaluated according to prior antimicrobial use, to determine
whether there are other treatment differences that have been
obscured by prior antimicrobial treatment (a systematic approach
using pooled data may be required), and future clinical trials
should be designed to facilitate such analyses. Third, the extent
of prior antimicrobial therapy in future CAP clinical trials should
be minimized as far as practically possible to ameliorate the
bias towards no treatment difference; the FDA has recommended
a limit of 25% of trial participants.16 Finally, given that the
Systematic review
increased ceftriaxone dose in the Asia CAP trial (2 g every 24 h) as
compared with the FOCUS trials (1 g every 24 h) did not result in
an increase in clinical cure rates in ceftriaxone recipients, there is
now further evidence to indicate that there is no clinical benefit in
increasing the ceftriaxone dose beyond 1 g every 24 h for the
treatment of patients with CAP.
In conclusion, the results of this meta-analysis demonstrate
that ceftaroline fosamil was superior to ceftriaxone as empirical
treatment for adult patients hospitalized with PORT risk class
3 – 4 CAP. Receipt of prior antimicrobial therapy appeared to
diminish the observed treatment effect.
Acknowledgements
We thank all of the patients, investigators and study personnel for their
contributions to the trials, and Helen Broadhurst, AstraZeneca, Macclesfield,
Cheshire, UK, for additional statistical analysis support.
A preliminary report of these results was presented at the 54th
Interscience Conference on Antimicrobial Agents and Chemotherapy,
5-8 September 2014, Washington DC, USA (Abstract L-1748a).
Funding
This study was funded by AstraZeneca. Medical writing support was
provided by Mark Waterlow of Prime Medica Ltd, Knutsford, Cheshire,
UK, funded by AstraZeneca. The design and conduct of the trials, as
well as analysis of the trial data and opinions, conclusions and interpretation of the data, are the responsibility of the authors. The
authors retained full control of the manuscript content and its
conclusions.
Transparency declarations
N. S. Z. received institutional research funding from AstraZeneca for
the conduct of the Asia CAP trial. N. S. Z. did not receive payment for
work on the manuscript. T. M. F. received recent research funding from
Actavis (formerly Forest Laboratories/Cerexa Inc.), Ortho-McNeil, Pfizer,
Boehringer-Ingelheim, Gilead and Tibotec. He is also a consultant for
Bayer, Cerexa, Inc., GlaxoSmithKline, Ortho-McNeil, Protez/Novartis,
Merck, Nabriva, Pfizer and Tetraphase. T. M. F. did not receive payment
for work on the manuscript. M. T., D. W., F. S. and J. P. I. are employees
of AstraZeneca. D. M. is a former employee of AstraZeneca. L. L. and
H. D. F. are former employees of Actavis. Cerexa, Inc. conducted the
FOCUS 1 and FOCUS 2 trials, prepared the statistical analysis plan and performed the analyses. AstraZeneca conducted the Asia CAP trial, prepared
the statistical analysis plan and performed the analyses and the
meta-analysis. Ceftaroline fosamil is being developed by AstraZeneca
and Actavis plc, a subsidiary of Allergan Inc.
Supplementary data
Tables S1 and S2 are available as Supplementary data at JAC Online (http://
jac.oxfordjournals.org/).
JAC
2 Wunderink RG, Waterer GW. Clinical practice. Community-acquired
pneumonia. N Engl J Med 2014; 370: 543– 51.
3 Cecere LM, Rubenfeld GD, Park DR et al. Long-term survival after hospitalization for community-acquired and healthcare-associated pneumonia. Respiration 2010; 79: 128–36.
4 Bruns AH, Oosterheert JJ, Cucciolillo MC et al. Cause-specific long-term
mortality rates in patients recovered from community-acquired pneumonia as compared with the general Dutch population. Clin Microbiol Infect
2011; 17: 763–8.
5 Broulette J, Yu H, Pyenson B et al. The incidence rate and economic burden of community-acquired pneumonia in a working-age population. Am
Health Drug Benefits 2013; 6: 494– 503.
6 Mandell LA, Wunderink RG, Anzueto A et al. Infectious Diseases Society
of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis
2007; 44 Suppl 2: S27–72.
7 Woodhead M, Blasi F, Ewig S et al. Guidelines for the management of
adult lower respiratory tract infections—full version. Clin Microbiol Infect
2011; 17 Suppl 6: E1– 59.
8 Pertel PE, Bernardo P, Fogarty C et al. Effects of prior effective
therapy on the efficacy of daptomycin and ceftriaxone for the
treatment of community-acquired pneumonia. Clin Infect Dis 2008; 46:
1142– 51.
9 File TM, Low DE, Eckburg PB et al. FOCUS 1: a randomized, doubleblinded, multicentre, Phase III trial of the efficacy and safety of ceftaroline
fosamil versus ceftriaxone in community-acquired pneumonia.
J Antimicrob Chemother 2011; 66 Suppl 3: iii19 –32.
10 Low DE, File TM, Eckburg PB et al. FOCUS 2: a randomized, doubleblinded, multicentre, Phase III trial of the efficacy and safety of ceftaroline
fosamil versus ceftriaxone in community-acquired pneumonia.
J Antimicrob Chemother 2011; 66 Suppl 3: iii33 –44.
11 File TM, Low DE, Eckburg PB et al. Integrated analysis of FOCUS 1
and FOCUS 2: randomized, doubled-blinded, multicenter phase 3 trials
of the efficacy and safety of ceftaroline fosamil versus ceftriaxone in
patients with community-acquired pneumonia. Clin Infect Dis 2010; 51:
1395– 405.
12 Zhong NS, Sun T, Zhuo C et al. Ceftaroline fosamil versus ceftriaxone for
the treatment of Asian patients with community-acquired pneumonia: a
randomised, controlled, double-blind, phase 3, non-inferiority with nested
superiority trial. Lancet Infect Dis 2015; 15: 161–71.
13 Fine MJ, Auble TE, Yealy DM et al. A prediction rule to identify low-risk
patients with community-acquired pneumonia. N Engl J Med 1997; 336:
243–50.
14 DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials
1986; 7: 177–88.
15 Powers JH. Reassessing the design, conduct, and analysis of clinical
trials of therapy for community-acquired pneumonia. Clin Infect Dis
2008; 46: 1152– 6.
16 Food and Drug Administration Center for Drug Evaluation and Research.
Guidance for Industry. Community-Acquired Bacterial Pneumonia: Developing
Drugs for Treatment. Revision 2, January 2014. http://www.fda.gov/
downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/
ucm123686.pdf.
17 Hartung J, Knapp G. A refined method for the meta-analysis of controlled clinical trials with binary outcome. Stat Med 2001; 20: 3875– 89.
References
1 World Health Organization. The 10 leading causes of death in the world,
2000 and 2012. Fact Sheet N8310. http://www.who.int/mediacentre/
factsheets/fs310/en/.
18 Jones RN, Farrell DJ, Mendes RE et al. Comparative ceftaroline activity
tested against pathogens associated with community-acquired pneumonia: results from an international surveillance study. J Antimicrob
Chemother 2011; 66 Suppl 3: iii69– 80.
869
Systematic review
19 Sader HS, Flamm RK, Jones RN. Antimicrobial activity of ceftaroline and
comparator agents tested against bacterial isolates causing skin and soft
tissue infections and community-acquired respiratory tract infections isolated from the Asia-Pacific region and South Africa (2010). Diagn Microbiol
Infect Dis 2013; 76: 61– 8.
22 Goonetilleke AK, Dev D, Aziz I et al. A comparative analysis of pharmacokinetics of ceftriaxone in serum and pleural fluid in humans: a study of
once daily administration by intramuscular and intravenous routes.
J Antimicrob Chemother 1996; 38: 969–76.
20 AstraZeneca AB. ZINFOROTM 600 mg powder for concentrate for solution for infusion: Summary of product characteristics. http://www.ema.
europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/
human/002252/WC500132586.pdf.
23 Riccobene T, Pushkin R, Jandourek A. Penetration of Ceftaroline into
Epithelial Lining Fluid in Healthy Adult Subjects. In: Abstracts of the Fiftythird Interscience Conference on Antimicrobial Agents and Chemotherapy,
Denver, CO, 2013. Abstract A-469. American Society for Microbiology,
Washington, DC, USA.
21 Roche. Rocephin powder for infusion: Summary of product characteristics. http://www.ema.europa.eu/docs/en_GB/document_library/Referrals_
document/Rocephin_30/WC500160113.pdf.
24 Drusano GL. What are the properties that make an antibiotic acceptable for therapy of community-acquired pneumonia? J Antimicrob
Chemother 2011; 66 Suppl 3: iii61– 67.
870