| Clinical Microbiology | Research Article

Daily fosfomycin versus levofloxacin for complicated urinary

tract infections
Nadine Rouphael,1 Patricia Winokur,2 Michael C. Keefer,3 Jessica Traenkner,1 Ana Drobeniuc,1 Yohei Doi,4,5 Sonal Munsiff,3 Vance G.
Fowler,5,6 Scott Evans,5,7 Randolph E. Oler,8 Bonifride Tuyishimire,8 Marina Lee,9 Varduhi Ghazaryan,9 Henry F. Chambers,5,10 DMID
15-0045 study group

AUTHOR AFFILIATIONS See affiliation list on p. 11.

ABSTRACT Fosfomycin, approved in the United States only for cystitis, is an attractive
alternative for oral treatment of outpatient complicated urinary tract infections (cUTIs)
as it has antimicrobial activity against most common uropathogens. The study was a
multicenter, randomized, open-label pragmatic superiority clinical trial evaluating the
efficacy of oral fosfomycin versus oral levofloxacin strategies in cUTIs (FOCUS study).
The trial compared two strategies for initial or step-down oral therapy of cUTI without
bacteremia after 0–48 hours of parenteral antibiotic therapy. Subjects were assigned to
3 g of fosfomycin or 750 mg (or dose adjusted for kidney function) of levofloxacin daily
for 5–7 days. Clinical and microbiological cures were assessed at the end of therapy
(EOT) and test of cure (TOC) (approximately 21 days from the start of antibiotics). The
trial did not meet accrual goals; thus, the results were descriptive. Only 51 subjects were
included in the microbiological intention-to-treat population. The subjects were mainly
females (76%), with a mean age of 46.7 years (standard deviation [SD] = 20.8) and acute
pyelonephritis (88%). At the end of therapy, clinical cure remained similar (69% and
68% for fosfomycin and levofloxacin strategies, respectively), and microbiological success
was 100% for both strategies. At the test of cure, clinical cure was similar (84% and
86% in the fosfomycin and levofloxacin strategies, respectively); however, a numerically
lower microbiological success was observed for fosfomycin (69% compared to 84% for
levofloxacin). These limited data suggest that fosfomycin could be an oral alternative as a
step-down therapy for the treatment of cUTIs (registry number NCT 03697993).

IMPORTANCE Concerns over resistance and safety have been identified in the current
treatment regimen for complicated urinary tract infections. Fosfomycin is a drug that
is routinely used for the treatment of uncomplicated cystitis. This study shows that
fosfomycin could be an oral alternative as step-down therapy for the treatment of
complicated urinary tract infections, with a clinical cure rate comparable to levofloxacin
but a lower microbiological success rate 3 weeks from start of antibiotics.
Editor Robert A. Bonomo, Louis Stokes Veterans
Affairs Medical Center, Cleveland, Ohio, USA
KEYWORDS complicated UTI, fosfomycin, oral antibiotic
Address correspondence to Nadine Rouphael,
nroupha@emory.edu, or Jessica Traenkner,

C omplicated urinary tract infections (cUTIs) are a widespread clinical problem

associated with substantial burden to the healthcare system (1, 2) and high
morbidity (3) and mortality (4), all exacerbated when cUTIs are caused by multidrug-
jessica.jones.traenkner@emory.edu.

The authors declare no conflict of interest.

See the funding table on p. 11.

resistant (MDR) organisms.
Quinolones tend to be the most commonly used antibiotics in the treatment of cUTI Received 5 July 2023
(5); however, their use is limited by antimicrobial resistance and potential side effects. Accepted 17 July 2023
Published 12 September 2023
Escherichia coli, the most common uropathogen, currently shows resistance rates of
5%–32% to quinolones in developed countries and 55%–85% in developing countries This is a work of the U.S. Government and is not
subject to copyright protection in the United States.
(6). More than half of quinolone-resistant strains also express extended-spectrum
Foreign copyrights may apply.

September/October 2023 Volume 14 Issue 5 10.1128/mbio.01677-23 1

Research Article mBio

beta-lactamases (ESBLs) (7), making them resistant to most beta-lactam antibiotics

except carbapenems. Furthermore, in European countries, 15%–60% of uropathogenic
E. coli isolates are resistant to trimethoprim-sulfamethoxazole (6), eliminating most
available oral antibiotics for cUTI. In addition, the U.S. Food and Drug Administra­
tion (FDA) has warned against many potential severe side effects of quinolones
such as tendinitis/tendon rupture, worsening myasthenia gravis, potentially irreversible
peripheral neuropathy, psychiatric effects, hypoglycemic risks, and increased risks of
ruptures or tears of the aorta (8–10). Thus, based on current resistance patterns and
safety concerns, there is a critical need for alternative oral therapeutic strategies to treat
cUTIs.
In the United States, a tromethamine salt of fosfomycin has been FDA-approved
since 1996 as a single oral sachet for the treatment of uncomplicated urinary tract
infections (UTIs) (e.g., cystitis [11]). Fosfomycin has broad in vitro antibacterial activity
against many clinically significant multidrug-resistant uropathogens (12) and excellent
penetration into the urinary tract (13). For more serious infections, oral administration
may provide inadequate concentrations due to its limited systemic bioavailability (37%)
and dose-limiting gastrointestinal tolerability (e.g., predominately diarrhea) (11). While
prior studies on the use of more than one dose of fosfomycin for the treatment of cUTIs
were promising, the findings were limited by their retrospective design (14–16).
The current trial assessed the safety and efficacy of a strategy of fosfomycin com­
pared to a strategy of levofloxacin as initial or step-down oral therapy in patients with
cUTI including pyelonephritis. The study utilized a unique pragmatic design, comparing
personalized antibiotic strategies (COMPASS) (17). Recent pharmacokinetic data support
the use of daily 3-g oral dose of fosfomycin (18). Furthermore, randomized clinical trials
have shown that 750-mg levofloxacin once daily for 5 days is non-inferior to 10 days of
ciprofloxacin in adults (19) with 7 days of quinolones non-inferior to 14 days of treatment
for acute pyelonephritis (20). We therefore elected to compare two strategies in terms of
safety and efficacy: 3 g of fosfomycin (Strategy 1) or 750 mg of levofloxacin (Strategy 2)
for 5–7 days to ensure clinically relevant and potentially highly effective oral comparator
arms.

MATERIALS AND METHODS

Trial design, oversight, and procedures
The FOCUS study was a multicenter, randomized, open-label pragmatic superiority
clinical trial evaluating the efficacy of oral fosfomycin versus oral levofloxacin strategies
in complicated urinary tract infections. The trial compared two strategies for initial
or step-down oral therapy of cUTI without bacteremia after 0–48 hours of parenteral
antibiotic therapy. Strategy 1 was 3 g of fosfomycin daily, and Strategy 2 was 750 mg of
levofloxacin daily (or the equivalent dosing based on kidney function [21]). Assessment
of compliance was performed at each of the three follow-up visits and by observing the
first dose. The trial was designed in accordance with FDA guidelines (22) using a dynamic
treatment strategy (17) allowing for investigator-directed adjustment. The adjustment
to another adequate oral therapy was allowed (i) if the causative pathogen was not
susceptible in vitro to quinolone initial or step-down therapy in a subject randomized to
the levofloxacin strategy; (ii) if the subject developed intolerance or allergy to the initial
or step-down oral therapy, at the investigator’s discretion; or (iii) if the subject had an
underlying condition posing increased risk of adverse events from quinolone therapy
(Fig. 1). The duration of oral study therapy (initial + investigator-directed adjustment if
indicated) in each strategy was 5–7 days of any per protocol antibiotic to which the
pathogen was susceptible, such that the total duration of effective antibacterial therapy
(including pre-study administration of oral therapy escalated to parenteral therapy or
parenteral therapy alone) was 7 days. All microbiology laboratories used matrix-assisted
laser desorption/ionization-time of flight (MALDI-TOF) to identify uropathogens and the
automated broth microdilution (VITEK 2) for quinolone susceptibility testing. Fosfomycin

September/October 2023 Volume 14 Issue 5 10.1128/mbio.01677-23 2

Research Article mBio

FIG 1 Schematic of the study design. QD, once daily; PO, orally; BID, twice a day; DS, double strength.

antimicrobial susceptibility testing results were generally not available in real time.
Centralized testing for fosfomycin was adapted to assist in the interpretation of the trial
data as the agar dilution method recommended for minimum inhibitory concentration
(MIC) testing is not routinely performed in clinical microbiology laboratories.
The study was conducted both in the outpatient and inpatient settings from
November 2018 to October 2019 when it was halted due to slow enrollment. Subjects

September/October 2023 Volume 14 Issue 5 10.1128/mbio.01677-23 3

Research Article mBio

were enrolled at five sites in different states in the United States (Georgia, Illinois, New
York, and Iowa).

Eligibility criteria
Eligible subjects were adults (aged ≥18 years) who (i) had documented or suspected
microbial pathogen isolated on urine culture, (ii) had not received fosfomycin in the past
year or any oral antibiotic prior to presentation, (iii) were anticipated to be able to be
stepped down or initially started on study oral antibiotic therapy within 48 hours of
enrollment, (iv) had a creatinine clearance of more than 20 mL/min, (v) had pyuria, (vi)
had absence of bacteremia/sepsis, and (vii) presented with clinical symptoms of cUTI.
cUTI was indicated by at least two signs or symptoms (chills, rigors, fever, or hypothermia;
dysuria or urinary frequency or urgency; lower abdominal or pelvic pain or tenderness;
nausea or vomiting; new onset of foul smell of urine or increased cloudiness of urine
per subject or their caregiver); and at least one complicating factor (indwelling urinary
catheter, current obstructive uropathy, or any functional or anatomical abnormality with
voiding disturbance). Acute pyelonephritis was indicated by at least two of the following
signs or symptoms: chills, rigors, or fever; flank pain; tenderness in the costovertebral
angle; dysuria or urinary frequency or urgency; and nausea or vomiting. Subjects were
excluded if they had a complete, permanent obstruction of the urinary tract, perinephric
or intrarenal abscess, suspected prostatitis, or an ileal loop or known vesicoureteral
reflux.

Randomization
Subjects were randomly assigned in a 1:1 ratio to receive oral therapy from Strategy
1 (initial or step-down to oral fosfomycin) or Strategy 2 (initial or step-down to oral
levofloxacin). Randomization was performed by statisticians at the Data Coordinating
Center (the Emmes Corporation) using an interactive web system and was stratified
according to site location and baseline diagnosis (acute pyelonephritis or other cUTIs).
There were no masking procedures as the trial was open label.

Analysis populations, assessments, and end points

The intention-to-treat (ITT) or safety population included all randomly assigned subjects
who received at least one dose of study drug. The analyses on the safety population were
performed per treatment actually received. The microbiological intention-to-treat (mITT)
population included all subjects in the ITT population who had a positive baseline urine
culture.
Study procedures included assessment of study outcomes, collection of urine for
urinalysis and cultures, and blood for creatinine (renal function), complete blood count
and bacterial culture, as well as pregnancy testing (if not already collected through
standard of care) and optional pharmacokinetic analyses. Assessments of adverse
events, vital signs (including body temperature), and physical examinations were also
conducted. Investigator assessment was also supported by an iterative structured
subject questionnaire and subjects’ memory aid. Therapeutic drug monitoring was
not implemented. Identification and susceptibility testing were conducted as part of
standard of care at local laboratories (using the Clinical and Laboratory Standards
Institute [CLSI] methods [23]). Levofloxacin susceptibility was tested at the participating
site as per routine clinical care. Fosfomycin susceptibility was centrally tested in duplicate
by the agar dilution method in Mueller-Hinton agar supplemented with 25 mg/L of
glucose-6-phosphate as stipulated by the CLSI. If there was more than a twofold
difference between the two results, a third round was performed, and those runs in
agreement were accepted. Thus, the result of fosfomycin susceptibility testing was not
available at the time of treatment. Positive baseline culture was defined as a culture
grown from a urine sample collected prior to treatment that had ≥105 colony-forming
units (CFU)/mL (50,000 and above was an allowed cut-off, provided it was a causative

September/October 2023 Volume 14 Issue 5 10.1128/mbio.01677-23 4

Research Article mBio

uropathogen) for non-catheter specimens or ≥104 CFU/mL for catheter specimens of a

single species of bacteria that causes cUTI. Up to two isolated pathogens were allowed
per urine culture. Urine cultures with three or more bacterial organisms were considered
contaminated unless there was a causative uropathogen that was growing at ≥50,000
CFU/mL and the contaminant was growing at ≤40,000 CFU/mL.
Clinical cure was defined as resolution of core symptoms from presentation with
no new UTI symptoms with avoidance of parenteral antibiotic therapy, in or out of
hospital, or oral antibiotic therapy different from per protocol at any time after random­
ization. Microbiological success was defined as reduction of the baseline uropathogen
to <104 CFU/mL for non-catheter specimens or <103 CFU/mL for catheter specimens
on urine culture (22). Clinical cure and microbiological success were assessed at end of
therapy (EOT) and test of cure (TOC) (approximately 21 days from start of antibiotics).

Statistical analysis
While originally designed and powered to evaluate superiority of Strategy 1 over
Strategy 2 (sample size of 634), study enrollment was terminated prematurely due to
slow recruitment, yielding a sample size with low power for hypothesis testing. All
resulting analyses on the mITT population are thus considered descriptive in nature.

RESULTS
A total of 79 subjects were screened, and 62 were randomized. Fifty-one subjects were
included in the mITT population (with a positive baseline bacterial urine culture) and 58
in the safety population (Fig. 2), and 3 subjects were lost to follow-up. Demographics and
baseline characteristics are described in Table 1. The subjects were mostly females (39 of

FIG 2 Consolidated standards of reporting trials (CONSORT) diagram. CC, complete cases.

September/October 2023 Volume 14 Issue 5 10.1128/mbio.01677-23 5

Research Article mBio

TABLE 1 Summary of demographic and baseline characteristics by treatment group-mITT populatione

Variable, statistic Characteristics Strategy 1 (n = 24) Strategy 2 (n = 27) All subjects (n = 51)
Age (years), mean (SD) – 46.7 (20.4) 46.7 (21.6) 46.7 (20.8)
Sex, n (%) Male 4 (17) 8 (30) 12 (24)
Female 20 (83) 19 (70) 39 (76)
Race, n (%) American Indian or Alaska Native – – –
Asian – – –
Native Hawaiian or other Pacific Islander – – –
Black or African American 11 (46) 13 (48) 24 (47)
White 10 (42) 11 (41) 21 (41)
Multiracial – – –
Unknown 3 (13) 3 (11) 6 (12)
BMI (kg/m2), mean (SD) – 26.20 (7.19) 28.94 (8.49) 27.65 (7.95)
cUTI type, n (%) Acute pyelonephritis 23 (96) 22 (81) 45 (88)
Other cUTI 1 (4) 5 (19) 6 (12)
Calculated creatinine clearance, n (%) ≥90 mL/min 11 (46) 17 (63) 28 (55)
60–89 mL/min 5 (21) 5 (19) 10 (20)
30–59 mL/min 8 (33) 3 (11) 11 (22)
≤29 mL/min – 2 (7) 2 (4)
Medical history, n (%) With recurrent UTIs 9 (38) 8 (30) 17 (33)
With obstructive uropathy 1 (4) – 1 (2)
With indwelling urinary catheter 1 (4) 3 (11) 4 (8)
Diabetes mellitus 8 (33) 6 (22) 14 (27)
All uropathogens, n (%) – 24 (100) 27 (100) 51 (100)
Enterobacteriaceae uropathogens,a n (%) Not susceptible to quinolonesb 1 (4) 2 (7) 3 (6)
Not susceptible to fosfomycin – – –
Not susceptible to carbapenems – – –
ESBLc 1 (4) 1 (4) 2 (4)
Multidrug resistantd 5 (21) 3 (11) 8 (16)
Any prior use of antibiotics, n (%) Yes 18 (75) 17 (63) 35 (69)
No 6 (25) 10 (37) 16 (31)
Prior use of antibiotics, n (%) Sulfamethoxazole and trimethoprim – 1 (4) 1 (2)
Ceftriaxone 18 (75) 13 (48) 31 (61)
Levofloxacin – 1 (4) 1 (2)
Meropenem – 2 (7) 2 (4)
a
Escherichia coli, Klebsiella pneumoniae, Proteus mirabilis, and Enterobacter cloacae.
b
Levofloxacin and ciprofloxacin.
c
Resistance to either ceftazidime, aztreonam, or ceftriaxone.
d
Resistance to at least one antibiotic from at least three different classes.
e
mITT, microbiologic intention-to-treat population, denotes all randomized patients who have a positive baseline bacterial culture of urine; n, number of subjects in the mITT
population; BMI, body mass index, weight (kg)/height (m)2.

52 or 76%); the mean age was 46.7 years (SD = 20.8); almost half (24 of 51 or 47%) were
African American; and the most common presentation was acute pyelonephritis (45 of 51
or 88%). The mean body mass index was 27.7 kg/m2 (SD = 8). The majority (38 of 51 or
75%) had preserved kidney function (calculated creatinine clearance above 60 mL/min)
with some having a history of recurrent UTIs (17 of 51 or 33%) or diabetes (14 of 51 or
27%). The majority of the uropathogens were E. coli (63%), Klebsiella pneumoniae (9.3%),
and Staphylococcus saprophyticus (5.6%) (Fig. 3). All E. coli isolates tested were susceptible
to fosfomycin. Six subjects had a uropathogen that was resistant to at least one antibiotic
from at least three different antimicrobial categories (multidrug-resistant) or resistant to
quinolones (Enterobacter cloacae, E. coli, and Proteus mirabilis).
The maximum duration of antimicrobial therapy was 7 days (median 5–6 days for
Strategies 1 and 2, respectively). Six subjects had adjustment in therapy in Strategy 2
based on having a medical condition for which quinolones are contraindicated (n = 3) or
having urine culture result showing resistance (n = 3) (Fig. 2). One subject, after being
switched from Strategy 2 to Strategy 1, could not tolerate fosfomycin because of diarrhea

September/October 2023 Volume 14 Issue 5 10.1128/mbio.01677-23 6

Research Article mBio

FIG 3 Distribution of uropathogens at baseline in the microbiological intention-to-treat population.

and required a different antibiotic (sulfamethoxazole-trimethoprim). All other subjects

had some moderate to severe side effects while on antibiotics; however, these side
effects did not result in adjustment in therapy. In Strategy 1 (fosfomycin group), 67% (20
of 30) had moderate to severe adverse events with more than half experiencing diarrhea
(53% or 16 of 30). The only severe side effects in Strategy 1 were gastrointestinal
symptoms. Similar rates of moderate to severe adverse events were seen in Strategy 2
(levofloxacin group and other adjustment therapies); however, symptoms were mainly
insomnia and back pain (28% or 8 of 28) (though may not necessarily be differentiated
from costovertebral angle pain). The majority of side effects seen in Strategy 2 were
related to central nervous system symptoms (insomnia, dizziness, and headache) (Fig. 4).
No allergic reactions were observed in either strategy. At the end of therapy, the clinical
cure rates were similar in Strategy 1 (69%) compared to Strategy 2 (68%), and the
microbiological success rate was 100% for both strategies. At the test of cure, the clinical
cure rates were similar in both strategies (84% in Strategy 1 and 86% in Strategy 2);
however, lower microbiological success rates were observed for Strategy 1 at 69%
compared to 84% in Strategy 2, with a composite cure in Strategy 1 of 55% compared to
73% in Strategy 2 (Table 2). Among subjects with microbiological failure, one subject in
Strategy 2 had both recurrent UTI and an indwelling urinary catheter, while two subjects
in Strategy 1 had a recurrent UTI. No subject required readmission for intravenous
antibiotics.

DISCUSSION
Each year in the United States alone, UTIs result in more than 10 million office vis­
its, 2 million emergency department visits, and half a million hospitalizations costing
approximately $3.5 billion (24, 25). The rise in antimicrobial resistance frequently limits
oral treatment options and requires the use of parental therapy, further increasing
healthcare costs and leading to risk of complications and patient discomfort (26).

September/October 2023 Volume 14 Issue 5 10.1128/mbio.01677-23 7

Research Article mBio

FIG 4 Solicited grade 2 and 3 adverse events in subjects who received at least one dose of study drug by maximum severity and by strategy (Strategy 1: initial or
step-down to oral fosfomycin; Strategy 2: initial or step-down to oral levofloxacin).

TABLE 2 Comparison of composite cure, clinical cure, and microbiological cure rates at TOC and EOT
between treatment group and mITT populationc

Strategy 1 Strategy 2 Difference (95% CI)b

a b a b
End point n No. missing % n No. missing % %
Test of cure
Composite cure 24 5 55 27 5 73 −18 (−43.4 to 8.7)
Clinical cure 24 5 84 27 4 86 −2 (−21.1 to 17.9)
Microbiological cure 24 5 69 27 5 84 −15 (−37.5 to 8.7)
End of therapy
Composite cure 24 3 67 27 4 67 0 (−26.3 to 25.3)
Clinical cure 24 3 69 27 4 68 1 (−24.4 to 26.8)
Microbiological cure 24 3 100 27 4 100 0 (NE)
a
n, number of subjects in the mITT population at that end point. Multiple imputations are used to impute missing
values.
b
Composite, clinical, or microbiological cure rates will be estimated from the multiple imputation model and 95%
Wald CI without continuity correction will be provided. Clinical cure is defined as (i) resolution of UTI symptoms
from presentation, (ii) no new UTI symptoms, and (iii) avoidance of parenteral antibiotic therapy, in or out of
hospital, at any time after randomization or oral antibiotic therapy different from per protocol. Microbiological
cure is defined as a reduction of the pathogen found at presentation to <104 CFU/mL for non-catheter specimens
or <103 for catheter specimens on urine culture. Treatment success (composite cure) is defined as a combination of
clinical cure and microbiological success. Strategy 1 is the initial or step-down to oral fosfomycin; strategy 2 is the
initial or step-down to oral levofloxacin.
c
Notes: NE, not estimated. Confidence interval (CI) and P values were not estimated since the microbiological cure
rates were 100% for both treatment strategies. Missing values were ignored for this result.

September/October 2023 Volume 14 Issue 5 10.1128/mbio.01677-23 8

Research Article mBio

Fosfomycin is considered an acceptable alternative to quinolones for the treatment of

cUTIs as it is given orally, is well tolerated, and has low resistance rates against uropatho­
gens (0%–7%) (27). However, its utility as a multiple dose regimen for treatment of cUTIs
is not well described.
Though our study did not reach its accrual goal, the strategies appeared to be equally
well tolerated, and only one subject could not tolerate fosfomycin due to diarrhea,
resulting in change in antibiotic regimen. The only severe side effects seen in Strat­
egy 1 were related to gastrointestinal symptoms, while insomnia and back pain were
mostly noted in Strategy 2. The clinical cure rates 3 weeks after starting oral step-down
therapy were similar in the fosfomycin strategy (84%) compared to the levofloxacin
strategy (86%). However, the microbiological success rates were numerically lower in
the fosfomycin strategy (69%) compared with the levofloxacin strategy (84%) at test of
cure (although not at end of therapy), consistent with data on cUTI from retrospective
studies on fosfomycin (31%–84%) (27). The microbiological success rates were within
range of those observed in contemporary prospective cUTI clinical trials with other
agents (28, 29). Interestingly, discordant outcomes with respect to microbiological failure
and clinical cure were also reported for other agents tested in cUTI clinical trials at TOC
(doripenem at 14.8% versus levofloxacin 6.3% [30]). In our study, microbiological failure
was observed mainly in subjects with non-identifiable risk factors for microbiological
relapse. No resistance to fosfomycin nor quinolones in the microbiological failures was
reported in the study. Though the follow-up was limited to 3–4 weeks, we were aware of
only four subjects who required additional oral antibiotics outside the study, suggesting
that microbiological failure does not necessarily translate into clinical failure. Test of cure
cultures are discouraged in clinical settings when clinical cure is achieved, and therefore
the long-term clinical significance of microbiological failure is unclear in the setting of
clinical cure.
Eighty-seven percent of uropathogens were gram-negative bacteria, mostly E. coli
as expected for cUTIs, but a surprisingly low number (n = 6) of uropathogens were
MDR, given the study was conducted at tertiary academic centers. The susceptibility
characteristics of the causative uropathogens could reflect a selection bias where
some sicker subjects with MDR organisms were not eligible for de-escalation within
48 hours. Additionally, many of these subjects had acute pyelonephritis rather than
known structural abnormalities.
Single-dose oral fosfomycin is approved for uncomplicated UTIs. Multidose regimens
(e.g., 3 g once every 2–3 days for three doses) have been described, particularly for
MDR UTIs (31) and prostatitis (32). In our cUTI study, fosfomycin was given daily for 5–7
days and only resulted in one therapy adjustment due to gastrointestinal side effects,
suggesting that repeated dosing of fosfomycin could be a safe alternative to quinolones
for UTIs.
The study utilized a unique pragmatic design, COMPASS (17), valuable in the setting
of antibiotic resistance. Management of patients with cUTI is not based on a single
decision point. It is dynamic and based on a sequence of decisions, with personalized
therapeutic adjustments, tailored to individual patients as new information becomes
available. The most important question for informing patient management is how
strategies, decision rules that guide empirical and definitive therapy, compare with
respect to ultimate outcome. Traditional trial designs comparing drugs rather than
strategies are unable to address this most important question. COMPASS was used
to compare two strategies mirroring clinical practice: (i) empirical fosfomycin with
therapeutic adjustments for intolerance or allergy and (ii) empirical levofloxacin with
therapeutic adjustments for non-susceptibility, increased risks for adverse events, or
intolerance/allergy.
The study has many limitations. Our trial did not meet the originally planned
target enrollment, making our results strictly descriptive. Of note, more than 15% of
interventional adult cUTI trials registered in clinicaltrials.gov do not meet enrollment
goals. Though this was a multicenter study, 71% of patients were enrolled from one

September/October 2023 Volume 14 Issue 5 10.1128/mbio.01677-23 9

Research Article mBio

site. The duration of follow up was short, though typical of cUTI studies, and not
long enough to detect resistance development nor relapse. In addition, imbalances
between the treatment groups could have affected our results, though none were
statistically significant. Though no documented fosfomycin resistance was noted in our
population, established CLSI breakpoints are applicable only for E. coli isolates, and there
were several non-E. coli uropathogens that demonstrated MICs greater than 64 mg/L,
thereby exceeding the current susceptibility breakpoint for E. coli. In addition, while the
majority of pathogens were E. coli as typically seen in UTIs, we cannot conclude whether
fosfomycin is effective against other pathogens such as Klebsiella spp. with concerns
for suboptimal response to fosfomyin (33). Certain patient groups were excluded from
the study, including patients with end-stage renal disease as well as pregnant women
and the study mainly enrolled females, making the results less generalizable. The study
allowed the inclusion of immunocompromised patients, including those with renal
transplantation, to enrich the population at risk of MDR infection, though our MDR rates
were low. The study mostly included patients with pyelonephritis who typically have
better cure rates than patients with cUTIs other than pyelonephritis. “Foul-smelling” or
“cloudy urine” was part of the inclusion criteria as many patients and caregivers seek
testing and treatment solely based on these reasons. Of the 29 patients with these
symptoms, 28 already had two or more other qualifying symptoms; in addition, all
patients included in the analysis had a positive urine culture, making the inclusion of the
subjective findings less relevant.

Conclusion
These limited data suggest that fosfomycin could be an oral alternative as step-down
therapy for treatment of cUTIs, though the study included less than 10% of the planned
sample and was underpowered to draw formal conclusions. Further clinical studies
are warranted to evaluate the efficacy and safety of repeated dosing regimens of oral
fosfomycin in patients with urogenital infections in order to establish an appropriate
benefit/risk ratio.

ACKNOWLEDGMENTS

This project was funded in whole or in part with federal funds from the National Institute
of Allergy and Infectious Diseases (NIAID) to the Vaccine and Treatment Evaluation Units
at Emory (HHSN272201300018I) and Iowa (HHSN2722013000201). Additional support
was provided by award UM1 AI104681 to the Antibacterial Resistance Leadership Group
(ARLG) and National Center for Advancing Translational Sciences award number CTSA
UL1TR002378 (Emory). Statistics and data management support for the trial were
provided by the EMMES Corporation under NIAID award HHSN272201500002C.
The content is solely the responsibility of the authors and does not necessarily
represent the official views of the National Institutes of Health.
The DMID 15-0045 study group includes the following: Emory University School of
Medicine, Annette Esper, Paulina A. Rebolledo, Zanthia Wiley, Jesse T. Jacob, Aneesh
Mehta, Colleen S. Kraft, Yun F. Wang, Rody G. Bou Chaaya, Danielle Fayad, Amer Bechnak,
Hollie Macenczak, Alexandra Dretler, Michele Paine McCullough, Sara Jo Johnson, Nour
Beydoun, Youssef Saklawi, Mark Mulligan, and Ghina Alaaeddine; University of Iowa
College of Medicine, Karl Kreder and Elizabeth B. Takacs; University of Rochester School
of Medicine and Dentistry, David Adler, Catherine Bunce, Dwight Hardy, Susan Ante­
nozzi, and Andrew Moran; Northwestern Medicine, Margaret Mueller; Duke University,
Antibacterial Resistance Leadership Group; The EMMES Company, Malcolm Almuntazar-
Harris, Alison Wall, and John Sumerel.
We would like to thank the following: at ARLG, Heather Cross, Lauren Komarow,
Bob Gazak, Holly Geres, Carolyn Rugloski, Norman Mustafa, Smitha Zaharoff, Rena
Hodges, Nyssa Schwager, Michael Woodworth, Nancie Deckard, Christi McElheny, and
Ryan Shields; at DMID, Jane Knisely, Venus Shahamatdar, Janie Russell, Gail Tauscher,
Liz Formentini, Jae Arega, Claudia Baxter, Michelle Wildman, Eliza Sindall, Blaire Osborn,

September/October 2023 Volume 14 Issue 5 10.1128/mbio.01677-23 10

Research Article mBio

Rick Fairhurst, Tammy Yokum, Mohamed Elsafy, Ranjodh Gill, Megan Gordon, Chidi Obasi,
Baoying Liu, and Ruth Ebiasah; Emory, Laura Oh, Varun Phadke, Mari Hart, Srilatha
Edupuganti, Colleen Kelley, Vanessa Raabe, Amy Sherman, Daniel Reichman, Alexis
Ahonen, Tigisty Girmay, Brandi Johnson; Lilin Lai, Juliet Morales, Rijalda Deovic, Ann
Lasseter, Lisa Harewood, Dilshad Rafi Ahmed, Delaney Morris, Juton Winston, Francine
Dyer, Terra Winter, Laurel Bristow, Chieutate Stallworth, Andrew Cheng, Mary Bower,
Wendy Nesheim, Chad Robichaux, Candace Miller, Jessica Ingersoll, Jean Winter, Philip
Powers, Jianguo Xu, Caitrin Carroll, Andrew Favre, Giselle Melville, Anna Morison, Lovie
Negrin, Matthew Romine, Maryam Roosta, Nicholas Stanley, Alaina Williams, Negrin
Lovie, David Weiss, Cecilia Losada, Ron Trible, and Julia Paine; at Iowa, Nancy Wagner,
Mary Eno, David Bush, Michelle Rodenburg, Geraldine Dull, Cathy Flanders, Brad Franzwa,
Alfred Carr, Stacy McMichael, Pam Nauerth, Debra Pfab, and Theresa Hegmann; and at
Northwestern, Veronica Munoz, Margaret Mueller, Meera Tavathia, and Sylwia Borowska

AUTHOR AFFILIATIONS
1
Emory University School of Medicine, Atlanta, Georgia, USA
2
Division of Infectious Diseases, Department of Internal Medicine, University of Iowa
College of Medicine, Iowa City, Iowa, USA
3
Division of Infectious Diseases, Department of Medicine, University of Rochester School
of Medicine and Dentistry, Rochester, New York, USA
4
University of Pittsburgh, Pittsburgh, Pennsylvania, USA
5
Antibacterial Resistance Leadership Group, Duke University Medical Center, Durham,
North Carolina, USA
6
Division of Infectious Diseases, Department of Medicine, Duke University Medical
Center, Durham, North Carolina, USA
7
George Washington University, Rockville, Maryland, USA
8
The Emmes Company, LLC, Rockville, Maryland, USA
9
Division of Microbiology and Infectious Diseases, NIAID, NIH, Rockville, Maryland, USA
10
University of California at San Francisco, San Francisco, California, USA

AUTHOR ORCIDs

Nadine Rouphael http://orcid.org/0000-0002-2512-7919

Yohei Doi http://orcid.org/0000-0002-9620-2525

FUNDING

Funder Grant(s) Author(s)

HHS | NIH | National Institute of Allergy and HHSN272201300018I Nadine Rouphael
Infectious Diseases (NIAID)
HHS | NIH | National Institute of Allergy and HHSN2722013000201 Patricia Winokur
Infectious Diseases (NIAID)
HHS | NIH | National Institute of Allergy and HHSN272201500002C Nadine Rouphael
Infectious Diseases (NIAID)
HHS | NIH | National Institute of Allergy and UM1 AI104681 Yohei Doi
Infectious Diseases (NIAID) Vance G. Fowler
Scott Evans
Henry F. Chambers

REFERENCES
1. Flores-Mireles AL, Walker JN, Caparon M, Hultgren SJ. 2015. Urinary tract 2. Foxman B. 2010. The epidemiology of urinary tract infection. Nat Rev
infections: epidemiology, mechanisms of infection and treatment Urol 7:653–660. https://doi.org/10.1038/nrurol.2010.190
options. Nat Rev Microbiol 13:269–284. https://doi.org/10.1038/ 3. Lee YC, Hsiao CY, Hung MC, Hung SC, Wang HP, Huang YJ, Wang JT.
nrmicro3432 2016. Bacteremic urinary tract infection caused by multidrug-resistant
Enterobacteriaceae are associated with severe sepsis at admission:

September/October 2023 Volume 14 Issue 5 10.1128/mbio.01677-23 11

Research Article mBio

implication for empirical therapy. Medicine 95:e3694. https://doi.org/10. 2018. Phase I study to evaluate the pharmacokinetics, safety, and
1097/MD.0000000000003694 tolerability of two dosing regimens of oral fosfomycin tromethamine in
4. Zilberberg MD, Nathanson BH, Sulham K, Fan W, Shorr AF. 2017. healthy adult participants. Antimicrob Agents Chemother 62:e00464-18.
Carbapenem resistance, inappropriate empiric treatment and outcomes https://doi.org/10.1128/AAC.00464-18
among patients hospitalized with Enterobacteriaceae urinary tract 19. Peterson J, Kaul S, Khashab M, Fisher AC, Kahn JB. 2008. A double-blind,
infection, pneumonia and sepsis. BMC Infect Dis 17:279. https://doi.org/ randomized comparison of levofloxacin 750 mg once-daily for five days
10.1186/s12879-017-2383-z with ciprofloxacin 400/500 mg twice-daily for 10 days for the treatment
5. Cek M, Tandoğdu Z, Wagenlehner F, Tenke P, Naber K, Bjerklund- of complicated urinary tract infections and acute pyelonephritis.
Johansen TE. 2014. Healthcare-associated urinary tract infections in Urology 71:17–22. https://doi.org/10.1016/j.urology.2007.09.002
hospitalized urological patients--a global perspective: results from the 20. Sandberg T, Skoog G, Hermansson AB, Kahlmeter G, Kuylenstierna N,
GPIU studies 2003-2010. World J Urol 32:1587–1594. https://doi.org/10. Lannergård A, Otto G, Settergren B, Ekman GS. 2012. Ciprofloxacin for 7
1007/s00345-013-1218-9 days versus 14 days in women with acute pyelonephritis: a randomised,
6. Kot B. 2019. Antibiotic resistance among uropathogenic Escherichia coli. open-label and double-blind, placebo-controlled, non-inferiority trial.
Pol J Microbiol 68:403–415. https://doi.org/10.33073/pjm-2019-048 Lancet 380:484–490. https://doi.org/10.1016/S0140-6736(12)60608-4
7. Salah FD, Soubeiga ST, Ouattara AK, Sadji AY, Metuor-Dabire A, Obiri- 21. Cockcroft DW, Gault MH. 1976. Prediction of creatinine clearance from
Yeboah D, Banla-Kere A, Karou S, Simpore J. 2019. Distribution of serum creatinine. Nephron 16:31–41. https://doi.org/10.1159/
quinolone resistance gene (qnr) in ESBL-producing Escherichia coli and 000180580
Klebsiella spp. Antimicrob Resist Infect Control 8. https://doi.org/10. 22. Anonymous. 2018. Complicated urinary tract infections: Developing
1186/s13756-019-0552-0 drugs for treatment. U.S Food & Drug Administration, June. https://www.
8. Anonymous.Dec 21, 2018. FDA warns about increased risk of ruptures or fda.gov/media/71313/download.
tears in the aorta blood vessel with fluoroquinolone antibiotics in 23. Anonymous. Clinical and laboratory standards Institute IMEA. Available
certain patients. U.S Food & Drug Administration. https://www.fda.gov/ from: https://clsimena.org
drugs/drug-safety-and-availability/fda-warns-about-increased-risk- 24. Stamm WE, Norrby SR. 2001. Urinary tract infections: disease panorama
ruptures-or-tears-aorta-blood-vessel-fluoroquinolone-antibiotics. and challenges. J Infect Dis 183 Suppl 1:S1–4. https://doi.org/10.1086/
Accessed 318850
9. Anonymous.July 10, 2018. FDA reinforces safety information about 25. Simmering JE, Tang F, Cavanaugh JE, Polgreen LA, Polgreen PM. 2017.
serious low blood sugar levels and mental health side effects with The increase in hospitalizations for urinary tract infections and the
fluoroquinolone antibiotics; requires label changes. U.S Food & Drug associated costs in the United States, 1998-2011. Open Forum Infect Dis
Administration. Accessed 4:fw281. https://doi.org/10.1093/ofid/ofw281
10. Anonymous.March 8, 2018. FDA Drug Safety Communication: FDA 26. Laupland KB, Valiquette L. 2013. Outpatient parenteral antimicrobial
updates warnings for oral and injectable fluoroquinolone antibiotics therapy. Can J Infect Dis Med Microbiol 24:9–11. https://doi.org/10.1155/
due to disabling side effects. U.S Food & Drug Administration. https:// 2013/205910
www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety- 27. Derington CG, Benavides N, Delate T, Fish DN. 2020. Multiple-dose oral
communication-fda-updates-warnings-oral-and-injectable-fluoroqui- fosfomycin for treatment of complicated urinary tract infections in the
nolone-antibiotics. Accessed outpatient setting. Open Forum Infect Dis 7:ofaa034. https://doi.org/10.
11. Anonymous. 2007. Monurol Sachet. U.S Food & Drug Administration, 1093/ofid/ofaa034
July. https://www.accessdata.fda.gov/drugsatfda_docs/label/2008/ 28. Wagenlehner FM, Sobel JD, Newell P, Armstrong J, Huang X, Stone GG,
050717s005lbl.pdf. Yates K, Gasink LB. 2016. Ceftazidime-avibactam versus doripenem for
12. Falagas ME, Kastoris AC, Kapaskelis AM, Karageorgopoulos DE. 2010. the treatment of complicated urinary tract infections, including acute
Fosfomycin for the treatment of multidrug-resistant, including pyelonephritis. Clin Infect Dis 63:754–762. https://doi.org/10.1093/cid/
extended-spectrum beta-lactamase producing, Enterobacteriaceae ciw378
infections: a systematic review. Lancet Infect Dis 10:43–50. https://doi. 29. Kaye KS, Bhowmick T, Metallidis S, Bleasdale SC, Sagan OS, Stus V,
org/10.1016/S1473-3099(09)70325-1 Vazquez J, Zaitsev V, Bidair M, Chorvat E, Dragoescu PO, Fedosiuk E,
13. Dijkmans AC, Zacarías NVO, Burggraaf J, Mouton JW, Wilms EB, van Horcajada JP, Murta C, Sarychev Y, Stoev V, Morgan E, Fusaro K, Griffith
Nieuwkoop C, Touw DJ, Stevens J, Kamerling IMC. 2017. Fosfomycin: D, Lomovskaya O, Alexander EL, Loutit J, Dudley MN, Giamarellos-
pharmacological, clinical and future perspectives. Antibiotics (Basel) Bourboulis EJ. 2018. Effect of meropenem-vaborbactam vs piperacillin-
6:24. https://doi.org/10.3390/antibiotics6040024 tazobactam on clinical cure or improvement and microbial eradication
14. Giancola SE, Mahoney MV, Hogan MD, Raux BR, McCoy C, Hirsch EB. in complicated urinary tract infection. JAMA 319:788–799. https://doi.
2017. Assessment of fosfomycin for complicated or multidrug-resistant org/10.1001/jama.2018.0438
urinary tract infections: patient characteristics and outcomes. 30. Naber KG, Llorens L, Kaniga K, Kotey P, Hedrich D, Redman R. 2009.
Chemotherapy 62:100–104. https://doi.org/10.1159/000449422 Intravenous doripenem at 500 milligrams versus levofloxacin at 250
15. Bielen L, Likic R. 2019. Experience with fosfomycin in the treatment of milligrams, with an option to switch to oral therapy, for treatment of
complicated urinary tract infections caused by extended-spectrum beta- complicated lower urinary tract infection and pyelonephritis. Antimicrob
lactamase-producing Enterobacteriaceae. Ther Adv Infect Dis Agents Chemother 53:3782–3792. https://doi.org/10.1128/AAC.00837-
6:2049936119858883. https://doi.org/10.1177/2049936119858883 08
16. Cai T, Cocci A, Verze P, Rizzo M, Palmieri A, Liguori G, Trombetta C, 31. Babiker A, Clarke L, Doi Y, Shields RK. 2019. Fosfomycin for treatment of
Adembri C, Carini M, Bartoletti R, Wagenlehner FM, Bonkat G, Mirone V, multidrug-resistant pathogens causing urinary tract infection: a real-
Bjerklund Johansen TE, Novelli A. 2018. The use of oral fosfomycin- world perspective and review of the literature. Diagn Microbiol Infect Dis
trometamol in patients with catheter-associated urinary tract infections. 95:114856. https://doi.org/10.1016/j.diagmicrobio.2019.06.008
J Chemother 30:290–295. https://doi.org/10.1080/1120009X.2018. 32. Kwan ACF, Beahm NP. 2020. Fosfomycin for bacterial prostatitis: a
1500110 review. Int J Antimicrob Agents 56:106106. https://doi.org/10.1016/j.
17. Evans SR, Follmann D, Liu Y, Holland T, Doernberg SB, Rouphael N, ijantimicag.2020.106106
Hamasaki T, Jiang Y, Lok JJ, Tran TTT, Harris AD, Fowler VG, Boucher H, 33. Abbott IJ, van Gorp E, Wyres KL, Wallis SC, Roberts JA, Meletiadis J, Peleg
Kreiswirth BN, Bonomo RA, Van Duin D, Paterson DL, Chambers H. 2019. AY. 2022. Oral fosfomycin activity against Klebsiella pneumoniae in a
Sequential, multiple-assignment, randomized trials for comparing dynamic bladder infection in vitro model. J Antimicrob Chemother
personalized antibiotic strategies (SMART-COMPASS). Clin Infect Dis 77:1324–1333. https://doi.org/10.1093/jac/dkac045
68:1961–1967. https://doi.org/10.1093/cid/ciy912
18. Wenzler E, Bleasdale SC, Sikka M, Bunnell KL, Finnemeyer M, Rosenkranz
SL, Danziger LH, Rodvold KA, Antibacterial Resistance Leadership Group.

September/October 2023 Volume 14 Issue 5 10.1128/mbio.01677-23 12