Azithromycin versus Doxycycline for Urogenital Chlamydia trachomatis Infection

Abstract

Background
Urogenital Chlamydia trachomatis infection remains prevalent and causes substantial reproductive morbidity. Recent studies have raised concern about the efficacy of azithromycin for the treatment of chlamydia infection.

Methods
We conducted a randomized trial comparing oral azithromycin with doxycycline for the treatment of urogenital chlamydia infection among adolescents in youth correctional facilities, to evaluate the noninferiority of azithromycin (1 g in one dose) to doxycycline (100 mg twice daily for 7 days). The treatment was directly observed. The primary end point was treatment failure at 28 days after treatment initiation, with treatment failure determined on the basis of nucleic acid amplification testing, sexual history, and outer membrane protein A (OmpA) genotyping of C. trachomatis strains.

Results
Among the 567 participants enrolled, 284 were randomly assigned to receive azithromycin, and 283 were randomly assigned to receive doxycycline. A total of 155 participants in each treatment group (65% male) made up the per-protocol population. There were no treatment failures in the doxycycline group. In the azithromycin group, treatment failure occurred in 5 participants (3.2%; 95% confidence interval, 0.4 to 7.4%). The observed difference in failure rates between the treatment groups was 3.2 percentage points, with an upper boundary of the 90% confidence interval of 5.9 percentage points, which exceeded the prespecified absolute 5-percentage-point cutoff for establishing the noninferiority of azithromycin.

Conclusions
In the context of a closed population receiving directly observed treatment for urogenital chlamydia infection, the efficacy of azithromycin was 97%, and the efficacy of doxycycline was 100%. The noninferiority of azithromycin was not established in this setting. (Funded by the National Institute of Allergy and Infectious Diseases; ClinicalTrials.gov number, NCT00980148.)

Full text

The
new england journal
of
medicine
n engl j med 373;26 nejm.org December 24, 2015
2512
From the Department of Medicine, Uni-
versity of Alabama at Birmingham, Bir-
mingham (W.M.G.); the Departments of
Preventive Medicine (A.U., P.R.K.) and
Internal Medicine (P.R.K), University of
Southern California, and Los Angeles
County Department of Health Services,
Juvenile Court Health Ser vices (R.C.W.P.,
C.M.K.) — both in Los Angeles; the De-
partment of Biostatistics, University of
Arkansas fo r Medical Sciences, L ittle Rock
(J.Y.L., S.Y.L.); and FHI 360, Durham, NC
(S.J.). Address reprint requests to Dr.
Geisler at the University of Alabama at
Birmingham, 703 19th St. S., 242 Zeigler
Research Bldg., Birmingham, AL 35294-
0007, or at wgeisler@ uab . edu.
N Engl J Me d 2015;373:2512-21.
DOI: 10.1056/NEJMoa1502599
Copyright © 2015 Massachusetts Medical Society.
BACKGROUND
Urogenital Chlamydia trachomat is infection remains prevalent and causes substantial
reproductive morbidity. Recent studies have raised concern about the efficacy of
azithromycin for the treatment of chlamydia infection.
METHODS
We conducted a randomized trial comparing oral azithromycin with doxycycline
for the treatment of urogenital chlamydia infection among adolescents in youth
correctional facilities, to evaluate the noninferiority of azithromycin (1 g in one
dose) to doxycycline (100 mg twice daily for 7 days). The treatment was directly
observed. The primary end point was treatment failure at 28 days after treatment
initiation, with treatment failure determined on the basis of nucleic acid amplifica-
tion testing, sexual history, and outer membrane protein A (OmpA) genotyping of
C. trachomatis strains.
RESULTS
Among the 567 participants enrolled, 284 were randomly assigned to receive
azithromycin, and 283 were randomly assigned to receive doxycycline. A total of
155 participants in each treatment group (65% male) made up the per-protocol
population. There were no treatment failures in the doxycycline group. In the
azithromycin group, treatment failure occurred in 5 participants (3.2%; 95% con-
fidence interval, 0.4 to 7.4%). The observed difference in failure rates between the
treatment groups was 3.2 percentage points, with an upper boundary of the 90%
confidence interval of 5.9 percentage points, which exceeded the prespecified ab-
solute 5-percentage-point cutoff for establishing the noninferiority of azithromycin.
CONCLUSIONS
In the context of a closed population receiving directly observed treatment for
urogenital chlamydia infection, the efficacy of azithromycin was 97%, and the
efficacy of doxycycline was 100%. The noninferiority of azithromycin was not es-
tablished in this setting. (Funded by the National Institute of Allergy and Infec-
tious Diseases; ClinicalTrials.gov number, NCT00980148.)
ABSTRACT
Azithromycin versus Doxycycline for
Urogenital Chlamydia trachomatis Infection
William M. Geisler, M.D., M.P.H., Apurva Uniyal, M.A., Jeannette Y. Lee, Ph.D.,
Shelly Y. Lensing, M.S., Shacondra Johnson, B.S.P.H.,
Raymond C.W. Perry, M.D., M.S.H.S., Carmel M. Kadrnka, D.O.,
and Peter R. Kerndt, M.D., M.P.H.
Original Article
The New England Journal of Medicine
Downloaded from nejm.org on June 16, 2021. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.

n engl j med 373;26 nejm.org December 24, 2015
2513
Azithromycin vs. Doxycycline for Chlamydia Infection
U
rogenital Chlamydia trachomatis in-
fection is the most prevalent bacterial
sexually transmitted infection in the
United States and worldwide.
1,2
Females are dis-
proportionately affected by this infection be-
cause of the risk of pelvic inflammatory disease,
which can lead to ectopic pregnancy and infer-
tility. Efforts to prevent and control chlamydia
infection, which have been aimed mainly toward
the reduction of sequelae, have not diminished
the high prevalence.
Along with screening, the provision of effec-
tive treatment is a cornerstone of chlamydia
control programs. For the treatment of chlamydia
infection, the Centers for Disease Control and
Prevention (CDC) recommends oral administra-
tion of either 1 g of azithromycin in a single dose
or 100 mg of doxycycline twice daily for 7 days.
3
These recommendations are supported by a meta-
analysis of 12 randomized clinical trials, which
showed that the efficacy of azithromycin against
chlamydia was 97%, and that of doxycycline was
98%
4
; however, these trials had limitations.
Most of the trials used tests that were less sensi-
tive than the currently recommended nucleic
acid amplification tests,
3
which may have led to
an underestimation of the rates of treatment
failure. Adherence to doxycycline treatment was
not ensured, which is important because non-
adherence may lead to treatment failure.
5,6
Re-
peat chlamydia exposure from partners could
not be controlled, which made it difficult to
determine whether repeat positive tests after
therapy indicated treatment failure or reinfec-
tion. Finally, the studies had a single test of cure
within 2 to 5 weeks after treatment but did not
have a repeat test at a later time to evaluate pa-
tients for relapse from incomplete eradication of
persistent noncultivable chlamydial forms, which
has been described in in vitro studies.
7,8
Some studies of chlamydia in which nucleic
acid amplification tests have been used have
raised concern about the efficacy of azithromy-
cin. Three studies of nongonococcal urethritis
showed azithromycin efficacy of less than 90%
in symptomatic chlamydia-infected males.
9-11
In
two chlamydia studies involving female partici-
pants — a randomized clinical trial of azithro-
mycin versus rifalazil and a longitudinal study
of repeat chlamydia infection — the efficacy of
azithromycin was 92%.
12,13
To address the limi-
tations of previous studies, we conducted a
phase 3, open-label, randomized trial of chla-
mydia treatment among youth in correctional
facilities to assess whether azithromycin is non-
inferior to doxycycline. Youth correctional facili-
ties were ideal sites for this study because the
prevalence of chlamydia infection in such facili-
ties is high,
14-1 6
residents of youth correctional
facilities are usually not reexposed to untreated
partners, treatment is directly observed, and
chlamydia exposure from new partners can be
minimized by screening and treating all persons
at intake and by constant staff supervision,
which limits the opportunities for sexual activi-
ties. We obtained a sexual history and per-
formed outer membrane protein A (OmpA) geno-
typing on C. trachomatis strains to more accurately
classify treatment outcomes.
Methods
Study Design and Participants
We enrolled males and females 12 to 21 years of
age who were residing in four long-term, sex-
segregated youth correctional facilities in Los
Angeles. The study began in December 2009 and
was initially limited to female participants. Be-
cause of the slow accrual of participants, higher-
than-expected rates of early discharge from the
facilities, and emerging data suggesting that
cure rates with azithromycin were lower among
chlamydia-infected males than previous studies
had indicated,
9
the protocol was amended to in-
clude male participants, beginning in August 2011.
Nucleic acid amplification testing to screen
for chlamydia (APTIMA Combo 2, Gen-Probe) is
routinely performed on first-catch urine speci-
mens obtained from residents of Los Angeles
County youth correctional facilities at intake;
routine genital examinations are also performed
within 96 hours after intake. Study staff re-
cruited residents who had a positive screening
nucleic acid amplification test result and, after
obtaining written informed consent, reviewed the
eligibility criteria and enrolled eligible residents.
The exclusion criteria were pregnancy, breast-
feeding, gonorrhea coinfection, allergy to tetra-
cyclines or macrolides, previous photosensitivity
from doxycycline, an inability to swallow pills,
receipt of an antibiotic with antichlamydial ac-
tivity within 21 days before screening or be-
tween screening and enrollment, concomitant
infection requiring treatment with an antibiotic
The New England Journal of Medicine
Downloaded from nejm.org on June 16, 2021. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.

n engl j med 373;26 nejm.org December 24, 2015
2514
The
new england journal
of
medicine
agent that had antichlamydial activity, and pelvic
inflammatory disease or epididymitis. Study par-
ticipation was voluntary. Refusal to participate
in the study by eligible residents was uncommon,
and no participants withdrew from the study.
After enrollment, follow-up study visits occurred
in up to 23 Los Angeles County youth correc-
tional facilities (some participants were trans-
ferred from the 4 facilities where they had been
enrolled to different facilities during the course
of the study). The full details of the study proce-
dures are provided in the protocol, available with
the full text of this article at NEJM.org.
Study Oversight
The study was approved by the institutional re-
view board at the University of Alabama at Bir-
mingham and at the County of Los Angeles
Public Health Department, as well as by the Su-
perior Court of California County of Los Angeles
Juvenile Division, the County of Los Angeles
Probation Department, and Office for Human
Research Protections, Department of Health and
Human Services. A data and safety monitoring
board convened annually. Data were collected by
Los Angeles County study staff, managed by FHI
360 (a nonprofit human development organiza-
tion), and analyzed by statisticians from the
University of Arkansas for Medical Sciences. The
study drugs were purchased from a pharmacy in
Los Angeles with study funds from the National
Institute of Allergy and Infectious Diseases. All
the authors vouch for the accuracy and com-
pleteness of the data and analyses presented and
for the fidelity of the study to the protocol.
Study Procedures
At enrollment, participants underwent an inter-
view (regarding demographic characteristics, pre-
vious sexually transmitted infections, sexual
behavior, contraception, and urogenital or gastro-
intestinal symptoms), provided first-catch urine
specimens for nucleic acid amplification testing
(to confirm chlamydia infection), and were ran-
domly assigned, in a 1:1 ratio, with the use of a
block randomization scheme (with randomiza-
tion performed separately within each facility) to
receive the CDC-recommended azithromycin
regimen or doxycycline regimen. The oral intake
of all doses of study drug was directly observed
by youth correctional facility staff. Two database
systems were reviewed for previous chlamydia
infections: STD Casewatch Millennium and the
Los Angeles County Public Health Laboratory
MISYS database system.
Participants whose chlamydia test results at
enrollment were negative were categorized as
not able to be evaluated, and their participation
in the study was discontinued. Participants who
had a positive chlamydia test result at enroll-
ment and who were still in a youth correctional
facility at day 28 attended a first follow-up on
that day; were interviewed regarding symptoms,
sexual behaviors, antibiotics taken, and fur-
loughs from the correctional facility; and pro-
vided a first-catch urine specimen for the test of
cure by nucleic acid amplification testing. The
timing of the test of cure was chosen to limit
possible false positive nucleic acid amplification
test results that can occur as a result of residual
nucleic acids from dead organisms that have not
yet been cleared; studies have reported C. tracho-
matis nucleic acid shedding 2 to 3 weeks after
therapy.
17-20
Youth correctional facility clinic rec-
ords were reviewed for medications and interim
pelvic inflammatory disease or epididymitis.
Participants who tested positive for chlamydia
at the first follow-up were initially classified as
having possible treatment failure, and their par-
ticipation was complete, with further treatment
administered in accordance with routine prac-
tice at the youth correctional facilities. Partici-
pants who tested negative for chlamydia at the
first follow-up and who were still in a youth
correctional facility on day 67 attended a second
follow-up on that day for repeat nucleic acid
amplification testing (to evaluate for persisting
chlamydia that was not identified at the first
follow-up) and for an interview in which the
same information was collected as at the day 28
follow-up. Participants who tested positive for
chlamydia at the first or second follow-up had
OmpA genotyping performed on their urine
specimens with the use of reported methods
21
to
assess for concordant strains (identical ompA
sequences); participants with suspected treat-
ment failure had genotyping performed on urine
specimens from both the enrollment visit and
the follow-up visit.
Outcomes and Populations Used for Analyses
The primary outcome was treatment failure at
the first follow-up, which was defined as a posi-
tive test for chlamydia and concordant C. tracho-
The New England Journal of Medicine
Downloaded from nejm.org on June 16, 2021. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.

n engl j med 373;26 nejm.org December 24, 2015
2515
Azithromycin vs. Doxycycline for Chlamydia Infection
matis strains at baseline and follow-up; if geno-
typing was unsuccessful (i.e., insufficient number
of ompA copies), participants could not have had
unsupervised furloughs and could not have
had sex (self-reported) between enrollment and
the follow-up. Participants with discordant strains
were presumed to have new infections and were
not considered to have treatment failure. The
secondary outcomes included treatment efficacy
based on the results of tests from both follow-up
visits, as well as safety.
The primary analysis population for efficacy
and safety was the per-protocol population, in
accordance with International Conference on
Harmonisation guidelines for a noninferiority
study.
22
An intention-to-treat approach is consid-
ered to be nonconservative in noninferiority
studies, because dropouts and withdrawals may
reduce the magnitude of the difference in effi-
cacy between treatments.
22
Thus, we used the
per-protocol population, which comprised partici-
pants who completed therapy, defined as patients
who received either a single dose of azithromy-
cin or at least 10 doses of doxycycline and whose
status with regard to treatment failure could be
established at the first follow-up; the use of at
least 10 doses of doxycycline for the evaluation
of outcome was based on a doxycycline adher-
ence study in which a 100% cure rate was shown
in association with this degree of adherence.
5
Participants were considered not able to be evalu-
ated for the primary outcome if they were not
tested at the first follow-up, had negative results
of the chlamydia test at enrollment, or vomited
the treatment within 1 hour after taking it (with-
out the medication being readministered).
Statistical Analysis
The goal of noninferiority trials is to determine
whether the efficacy of a treatment is no worse
than that of another treatment.
23
This noninferi-
ority study was designed to test the null hypoth-
esis that the absolute rate of azithromycin treat-
ment failure would be at least 5 percentage
points higher than the absolute rate of doxycy-
cline treatment failure against the alternative
hypothesis that there would be no difference
between regimens, with a failure rate of 3% for
both (a rate that was based on the results of the
meta-analysis).
4
The decision to use the differ-
ence cutoff of 5 percentage points was based on
the reported high cure rates for both treatments
4
;
this difference was considered by the investiga-
tive team to be an appropriate cutoff to establish
the clinical noninferiority of azithromycin to
doxycycline. It was estimated that for the study
to have 90% power to test the hypothesis at a
one-sided 0.10 significance level, the per-protocol
population would need to include 153 partici-
pants in each group. The failure rate was esti-
mated with binomial proportion and 95% confi-
dence intervals. One-sided 90% exact confidence
intervals were used to estimate the difference in
the failure rates between the two treatments,
which is appropriate for a noninferiority study
and which is consistent with the one-sided sig-
nificance level of 0.10 that was used for the de-
termination of the sample size. Analyses were
performed with SAS software, version 9.3 (SAS
Institute). Associations between participant char-
acteristics and study group or treatment failure
were evaluated with the use of Fisher’s exact test
or a Wilcoxon rank-sum test. Differences in self-
reported versus documented previous chlamydia
infection were evaluated with McNemar’s test.
Results
Study Participants
Of the 567 participants enrolled from December
2009 through April 2014, a total of 284 were
randomly assigned to receive azithromycin, and
283 were randomly assigned to receive doxycy-
cline (Fig. 1). After early discontinuation was
accounted for, 155 participants (55%) in each
group completed the first follow-up and made
up the per-protocol population. Discharge from
the youth correctional facility was the primary
reason for early discontinuation. The final study
visit was completed in May 2014; in the per-
protocol population, 90% of the patients in the
azithromycin group and 81% of the patients in
the doxycycline group completed both follow-up
visits.
The characteristics of the participants in the
per-protocol group at baseline are shown in Ta-
ble 1. Most male participants (150 of 201, 75%)
reported not having urogenital symptoms. The
majority of female participants (67 of 109, 61%)
reported urogenital symptoms, most commonly
abnormal vaginal discharge (54 of 109, 50%).
Pelvic inf lammatory disease was not diagnosed
in any female participant. Previous chlamydia
infection was self-reported by 22% of the male
The New England Journal of Medicine
Downloaded from nejm.org on June 16, 2021. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.

n engl j med 373;26 nejm.org December 24, 2015
2516
The
new england journal
of
medicine
participants (44 of 201) and was documented in
20% (40 of 201). Female participants self-report-
ed previous chlamydia infection more often than
it was documented (44 [40%] vs. 32 [29%],
P = 0.001). Demographic characteristics, frequen-
cies of symptoms, and the rate of reported previ-
ous chlamydia infection did not differ signifi-
cantly between the treatment groups. Sex with a
female in the previous 12 months was reported
by 198 male participants (99%), and sex with a
male was reported by 1 (0.5%). Sex with a male
during the previous 12 months was reported by
107 female participants (98%), and sex with a
female was reported by 17 (16%). Sex after en-
rollment was reported by 1 participant: a female
who reported having sex with another female in
the youth correctional facility. No participants
had a furlough from the youth correctional facil-
ity during the course of the study.
Chlamydia Treatment Failure and Predictors
No treatment failures occurred in the doxycy-
cline group (0%; 95% confidence interval [CI],
0.0 to 2.4). In the azithromycin group, seven
participants (six male and one female) tested
positive for chlamydia at the first follow-up;
however, two of the males were infected with
strains that were discordant with the infecting
strains at baseline and were not considered to
have treatment failure. Therefore, there were five
treatment failures among participants who re-
ceived azithromycin (3.2%; 95% CI, 0.4 to 7.4):
four in male participants (3.9%; 95% CI, 1.1 to
9.7) and one in a female participant (1.9%; 95%
CI, 0.0 to 10.1); all the participants with treat-
ment failure were asymptomatic. The difference
in failure rate between the treatments was 3.2
percentage points (one-sided 90% CI, 0 to 5.9).
Because the upper boundary of the 90% confi-
Figure 1. Screening, Randomization, and Follow-up.
A total of 567 participants in four youth correctional facilities were enrolled and underwent randomization. In each
study group, 155 participants completed the primary evaluation at the first follow-up visit 28 days after treatment
initiation and made up the per-protocol population. The most common reason for early discontinuation was discharge
from the youth correctional facility, which occurred more often among females than among males (60% vs. 31%).
Other reasons for discontinuing study participation in the azithromycin group were as follows: tested positive for
gonorrhea (1 participant), pelvic inflammatory disease diagnosis (2), vomited within 1 hour after treatment (2), preg-
nancy identified after treatment (1), and enrolled without a positive chlamydia screening test (2). Other reasons for
discontinuing study participation in the doxycycline group were as follows: pelvic inf lammatory disease diagnosis
(2), pregnancy identified after treatment (1), enrolled without a positive chlamydia screening test (1), pregnancy pos-
sibility (1), physician withdrawal (1), and enrolled while taking antibiotics with potential antichlamydial therapeutic
effects (3).
567 Were enrolled and underwent
randomization
567 Patients were screened
284 Were assigned to receive
azithromycin
283 Were assigned to receive
doxycycline
129 Discontinued study early
89 Were discharged from facility
20 Received other antichlamydia
treatment
12 Were negative for chlamydia
at enrollment
8 Had other reason
128 Discontinued study early
96 Were discharged from facility
14 Received other antichlamydia
treatment
9 Were negative for chlamydia
at enrollment
9 Had other reason
155 Completed primary evaluation 155 Completed primary evaluation
The New England Journal of Medicine
Downloaded from nejm.org on June 16, 2021. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.

n engl j med 373;26 nejm.org December 24, 2015
2517
Azithromycin vs. Doxycycline for Chlamydia Infection
dence interval exceeded 5 percentage points, the
null hypothesis was not rejected, and thus the
noninferiority of azithromycin to doxycycline
was not established. With only five treatment
failures, analyses of participant characteristics
associated with treatment failure were limited,
and no significant associations were identified.
No doxycycline treatment failures were iden-
tified at the second follow-up. A single female
participant who received azithromycin tested
negative for chlamydia at the first follow-up but
tested positive at the second follow-up. When
this treatment failure was included with the five
that were detected at the first follow-up, the
failure rate for azithromycin was calculated as
3.9% (95% CI, 1.4 to 8.2). Of note, OmpA geno-
typing was unsuccessful in two of the six partici-
pants with treatment failure (both were female
participants); however, these two participants did
not have unsupervised furloughs or report in-
terim sexual activity.
Treatment Adherence and Safety
In the azithromycin group, two participants (1%)
vomited azithromycin within 1 hour after taking
it, and a second dose was administered success-
fully. In the doxycycline group, 77% of partici-
pants received 14 doses; because of the logistic
challenges inherent in conducting the study in
youth correctional facilities, 2% of participants
received 11 doses, 3% received 12 doses, 12%
received 13 doses, 6% received 15 doses, and 1%
received 16 doses. No participants were excluded
from the per-protocol population because they
received an insufficient number of doxycycline
doses. Adverse events were reported by 23% of
the participants in the azithromycin group and
by 27% of the participants in the doxycycline
group; the most common adverse events reported
in both groups were gastrointestinal symptoms.
No severe or serious adverse events occurred,
and no participants discontinued participation
in the study because of an adverse event.
Discussion
In this trial, the noninferiority of azithromycin
to doxycycline for the treatment of chlamydia
infection was not established. This was not a
result of azithromycin having low efficacy; the
97% efficacy of the drug in our study was con-
sistent with that reported in a meta-analysis.
4
Rather, it reflects the 100% efficacy of doxycy-
cline. Because doxycycline is not given under
direct observation in practice outside institu-
tional settings, the generalizability of our find-
ings is unknown. The two studies that have
evaluated patient-reported adherence to doxycy-
cline treatment for chlamydia infection suggest
that 3 to 28% of patients miss at least 1 dose.
5,6
However, a study in which adherence to doxycy-
cline treatment was evaluated with micropro-
cessor-containing medication bottles suggested
that the degree of adherence based on reporting
by participants can be an overestimation
5
; the
study showed that although 90% of participants
reported taking all doses within 8 days, only
16% actually managed this degree of adherence.
Nonadherence to doxycycline therapy contributes
to treatment failure. Bachmann et al. reported no
treatment failures among 58 participants who
took 10 to 14 doses, as compared with treatment
failure in 4 of 20 participants (20%) who took
fewer than 10 doses.
5
Khosropour et al. evalu-
ated doxycycline adherence among males with
symptomatic chlamydia urethritis and found
treatment failure in 1 of 37 participants (3%)
who took 14 doses, versus 2 of 10 participants
(20%) who missed at least one dose.
6
In our
study, we determined adherence through the
staff recording directly observed treatment, and
our results suggest that doxycycline is up to
100% efficacious against chlamydia among pa-
tients who are mostly adherent, whereas azithro-
mycin may be slightly less efficacious, with an
occasional treatment failure. When chlamydia
treatment is provided in real-world clinical prac-
tice, the possibility that the efficacy of doxycy-
cline could be offset by limited adherence should
be taken into consideration.
It is unclear why all the treatment failures in
our study occurred in azithromycin-treated par-
ticipants. Resistance to the drug is a consider-
ation, although it is unlikely. High-level azithro-
mycin resistance in human C. trachomatis strains
has not been definitively shown, and only in rare
cases have “heterotypic” resistant strains been
associated with treatment failure.
24,25
More re-
cent studies of nongonococcal urethritis have
suggested that the efficacy of azithromycin may
be lower in males who have symptomatic chla-
mydia urethritis than previous studies had indi-
cated.
4,9 -11
Our study was not designed or pow-
ered to evaluate the efficacy of azithromycin in
The New England Journal of Medicine
Downloaded from nejm.org on June 16, 2021. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.

n engl j med 373;26 nejm.org December 24, 2015
2518
The
new england journal
of
medicine
Characteristic All Participants
(N = 310) Azithromycin
(N = 155) Doxycycline
(N = 155) P
Value†
All participants
Sex — no. (%) 0.74
Male 201 (65) 102 (66) 99 (64)
Female 109 (35) 53 (34) 56 (36)
Race — no./total no. (%) 0.49
White 109/307 (36) 59/154 (38) 50/153 (33)
Black 136/307 (44) 64/154 (42) 72/153 (47)
Other 62/307 (20) 31/154 (20) 31/153 (20)
Hispanic ethnic background — no. (%) 158 (51) 84 (54) 74 (48) 0.09
Median age (range) — yr 17.0 (13.5–20.4) 17.1 (14.6–18.9) 16.9 (13.5–20.4) 0.07
Previous chlamydia infection — no. (%) 0.71
No 222 (72) 113 (73) 109 (70)
Yes 88 (28) 42 (27) 46 (30)
HIV or AIDS — no. (%) NA
Not present 309/310 (100) 154/155 (99) 155/155 (100)
Information not ascertained 1/310 (<1) 1/155 (1) 0
Median no. of sexual partners in the previous 12 mo (IQR) 3 (2–6) 3 (2–6) 3 (2–6) 0.57
Median age at first sexual activity (IQR) — yr 14 (13–14) 14 (13–15) 13 (13–14) 0.04
Male participants
Painful urination — no./total no. (%) 0.86
No 160/201 (80) 82/102 (80) 78/99 (79)
Yes 41/201 (20) 20/102 (20) 21/99 (21)
Penile discharge — no./total no. (%) >0.99
No 191/201 (95) 97/102 (95) 94/99 (95)
Yes 10/201 (5) 5/102 (5) 5/99 (5)
Pain in the scrotum or testicle area — no./total no. (%)
No 199/201 (99) 100/102 (98) 99/99 (100) 0.50
Yes 2/201 (1) 2/102 (2) 0
No previous nongonococcal urethritis — no./total no. (%) 201/201 (100) 102/102 (100) 99/99 (100) NA
No previous epididymitis — no./total no. (%) 201/201 (100) 102/102 (100) 99/99 (100) NA
Female participants
Painful urination — no./total no. (%) >0.99
No 102/109 (94) 50/53 (94) 52/56 (93)
Yes 7/109 (6) 3/53 (6) 4/56 (7)
Abnormal vaginal discharge — no./total no. (%) 0.85
No 55/109 (50) 26/53 (49) 29/56 (52)
Yes 54/109 (50) 27/53 (51) 27/56 (48)
Irregular vaginal bleeding — no./total no. (%) >0.99
No 97/109 (89) 47/53 (89) 50/56 (89)
Yes 12/109 (11) 6/53 (11) 6/56 (11)
Table 1. Baseline Characteristics of the Participants in the Per-Protocol Population.*
The New England Journal of Medicine
Downloaded from nejm.org on June 16, 2021. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.

n engl j med 373;26 nejm.org December 24, 2015
2519
Azithromycin vs. Doxycycline for Chlamydia Infection
symptomatic chlamydia-infected males. Howev-
er, in a subanalysis of 102 males who received
azithromycin, the difference in treatment failure
rate between males who reported painful urina-
tion and those who did not report it (2 of 20
[10%] and 2 of 82 [2%], respectively) was not
significant (P = 0.17), but the results suggest that
the efficacy of azithromycin for the treatment of
symptomatic chlamydia urethritis deserves fur-
ther study. Azithromycin levels that are suffi-
cient to eradicate chlamydia may not be achieved
in some patients. Only limited studies have
shown the presence of therapeutic levels of
azithromycin in the female genital tract,
26,27
and
there are insufficient data on azithromycin levels
in the urethra. Even with sufficient levels, it is
possible that some organisms are not eradicated
in acute infection, as suggested by an in vitro
study that showed that doxycycline was more
effective than azithromycin in eradicating chla-
mydia from acutely infected human epithelial
cells.
28
An important aspect of our study was the use
of youth correctional facilities as the sites for the
trial. Conducting the trial at these sites mini-
mized the possibility of chlamydia re-exposure
from untreated partners, limited exposure from
new partners, and enhanced treatment adher-
ence. Another important aspect of this study was
the improved accuracy in the detection of treat-
ment failure with the use of nucleic acid ampli-
fication testing and OmpA genotyping; the use-
fulness of the latter method is illustrated by the
identification of two male participants who de-
nied having sex after therapy and had repeat
chlamydia infection with different strains, which
suggested that they were not forthcoming about
their sexual history. Because the residents of
youth correctional facilities do not undergo rou-
tine oropharyngeal or rectal chlamydia screen-
ing, it is possible that those two male partici-
pants had sex with male residents who had
extragenital chlamydia infection. Adding chla-
mydia testing at day 67 yielded only one treat-
ment failure, which suggests that relapsing or
persistent chlamydia infection is probably rare,
despite the findings of a study of same-serovar
chlamydial infections over a period of 2 to 5 years,
which suggested that it may occur.
29
Our study had some limitations. First, it was
performed in a single geographic location,
which may influence the generalizability of the
findings. Second, urine specimens were used for
chlamydia nucleic acid amplification testing in
female participants. During the course of the
study, vaginal swabs become the preferred spec-
imen for chlamydia screening by means of nu-
cleic acid amplification testing,
30
in part because
of the slightly higher sensitivity of this type of
testing when vaginal swab samples are used
31
;
however, we continued to test urine specimens
for comparison purposes, and it is possible that
Characteristic All Participants
(N = 310) Azithromycin
(N = 155) Doxycycline
(N = 155) P
Value†
Pelvic pain — no./total no. (%) 0.12
No 83/109 (76) 44/53 (83) 39/56 (70)
Yes 26/109 (24) 9/53 (17) 17/56 (30)
Previous pelvic inflammatory disease — no./total no. (%) 0.23
No 107/109 (98) 51/53 (96) 56/56 (100)
Yes 2/109 (2) 2/53 (4) 0
Hormone contraception use — no./total no. (%) >0.99
No 100/109 (92) 49/53 (92) 51/56 (91)
Yes 9/109 (8) 4/53 (8) 5/56 (9)
* All characteristics, with the exception of the sex of the participants, were self-reported. AIDS denotes acquired immunodeficiency syndrome,
HIV human immunodeficiency virus, IQR interquartile range, and NA not applicable.
† P values for the differences between the two treatment groups were determined with the use of a Wilcoxon rank-sum test (for age, sex part-
ners in the past 12 months, and age at first sex) or Fisher’s exact test (for all other characteristics).
Table 1. (Continued.)
The New England Journal of Medicine
Downloaded from nejm.org on June 16, 2021. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.

n engl j med 373;26 nejm.org December 24, 2015
2520
The
new england journal
of
medicine
a small number of infections were undetected.
Third, an unexpected challenge in our study was
the high rate of early discharge from youth cor-
rectional facilities; this may have been due in
part to changing trends in juvenile justice poli-
cies leading to fewer and shorter incarcerations.
The discharge rate led to a need to enroll more
participants to achieve the targeted per-protocol
population size and influenced sex distribution,
with the result that there were almost two times
as many males as females in the per-protocol
population. Finally, with our sample size and low
rate of treatment failure, a small change in the
number of failures in either treatment group
could alter our conclusions regarding noninferi-
ority. For example, 2 fewer failures in the azithro-
mycin group (3 of 155 in the azithromycin group
vs. 0 of 155 in the doxycycline group) would give
an upper boundary of the 90% confidence inter-
val of 4.3 percentage points. Therefore, 2 events
would influence the determination of noninferi-
ority. Furthermore, we determined that for the
observed rates of treatment failure, an additional
130 participants who could be evaluated in each
treatment group (almost double our sample size)
would be required in order to increase the preci-
sion and establish the noninferiority of azithro-
mycin with the use of the prespecified statistical
criteria. As we noted above, we used a one-sided
significance level of 0.10 for the determination
of sample size; decreasing the significance level
to 0.05 would have increased the sample-size
requirements to an unattainable level.
In conclusion, the noninferiority of azithro-
mycin to doxycycline was not established in our
study. However, the efficacy of both types of
treatment was high (97% and 100%) in the con-
text of a per-protocol analysis and directly ob-
served and monitored therapy.
Supported by the Division of Microbiology and Infectious
Diseases, National Institute of Allergy and Infectious Dis-
eases, National Institutes of Health (contract number
HHSN266200400073C).
Dr. Geisler reports receiving grant support from ActivBiotics
Pharma. No other potential conflict of interest relevant to this
article was reported.
Disclosure forms provided by the authors are available with
the full text of this article at NEJM.org.
We thank the staff at the County of Los Angeles Department
of Public Health, Los Angeles County youth correctional facili-
ties, and of the Count y of Los Angeles Juvenile Court Health
Services for their contributions to completing the study; the
staff of the County of Los Angeles Probation Department and
Los Angeles County Juvenile Court Health Services for permit-
ting the study to be conducted; Jill Stanton and Linda McNeil
from FHI 360 for assistance with study coordination and data
management; Melina Boudov, Staeci Morita, Lashawnda Royal,
Kirsten Wilson, Kimberly Coffee, Kimberly Givan, Marisol Mejia,
and Jennifer Vonghack for assistance with study coordination
throug h the County of L os Angeles Department of P ublic Health;
Carolyn Deal, Barbara Hahn, and Jill Long from the Division of
Microbiolog y and Infectious Diseases, National Institute of Al-
lergy and infectious Diseases, National Institutes of Health, for
their support and oversight throughout the development, imple-
mentation, analysis and reporting of the study; Dr. Edward
Hook from the Universit y of Alabama at Birmingham (UAB) for
his review of an earlier version of the manuscript; and Richa
Kapil and LaDraka Brown from UAB for their assistance with
C. trachomati s OmpA genotyping.
References
1. Centers for Disease Control and Pre-
vention. Sexually transmitted disease
sur veillance 2013. At lanta: Depart ment of
Health and Human Services, 2014.
2. World Health Organization. Global in-
cidence and prevalence of selected curable
sexually transmitted infections — 2008
(http://www .who .int/ reproductivehealth/
publications/ rtis/ stisestimates/ en).
3. Sexually transmitted diseases treat-
ment guidelines, 2010. MMWR Recomm
Re p 2010; 59(R R-1 2): 1-110.
4. Lau CY, Qureshi AK. Azithromycin
versus doxycycline for genital chlamydial
infections: a meta-analysis of randomized
clinical trials. Sex Transm Dis 2002; 29:
497-502.
5. Bachmann LH, Stephens J, Richey CM,
Hook EW III. Measured versus self-repor ted
compliance with doxycycline therapy for
chlamydia-associated syndromes: high
therapeutic success rates despite poor
compliance. Sex Transm Dis 1999; 26:
272-8.
6. Khosropour CM, Manhart LE, Colom-
bara DV, et al. Suboptimal adherence to
doxycycline and treatment outcomes
among men with non-gonococcal ure-
thritis: a prospective cohort study. Sex
Transm Infect 2014; 90: 3-7.
7. Beatty WL, Morrison RP, Byrne GI.
Persistent chlamydiae: from cell culture
to a paradigm for chlamydial pathogene-
sis. Microbiol Rev 1994; 58: 686-99.
8. Wyrick PB. Chlamydia trachomatis per-
sistence in vitro: an overview. J Infect Dis
2010; 201: Suppl 2: S88-S95.
9. Schwebke JR, Rompalo A, Taylor S, et
al. Re-evaluating the treatment of non-
gonococcal urethritis: emphasizing emerg-
ing pathogens — a randomized clinical
trial. Clin Infect Dis 2011; 52: 163-70.
10. Manhart LE, Gillespie CW, Lowens
MS, et al. Standard treatment regimens
for nongonococcal urethritis have similar
but declining cure rates: a randomized
controlled trial. Clin Infect Dis 2013; 56:
934-42.
11. Takahashi S, Matsukawa M, Kuri-
mura Y, et al. Clinical eff icacy of azithro-
mycin for male nongonococcal urethritis.
J Infect Chemother 2008; 14: 409-12.
12. Geisler WM, Pascual ML, Mathew J, et
al. Randomized, double-blind, multicenter
safet y and efficacy study of rifalazil com-
pared with azithromycin for treatment of
uncomplicated genital Chlamydia tracho-
matis infection in women. Antimicrob
Agents Chemother 2014; 58: 4014-9.
13. Batteiger BE, Tu W, Ofner S, et al. Re-
peated Chlamydia t rachomati s genital infec-
tions in adolescent women. J Infect Dis
2010; 201: 42-51.
14. Mertz KJ, Voigt RA, Hutchins K, et al.
Findings from STD screening of adoles-
cents and adults entering correctional fa-
cilities. Sex Transm Dis 2002; 29: 834-9.
15. Bauer HM, Chartier M, Kessell E, et al.
Chlamydia screening of youth and young
adults in non-clinical settings throughout
California. Sex Transm Dis 2004; 31: 409-
14.
16. Robertson AA, Thomas CB, St Law-
rence JS, Pack R. Predictors of infection
with chlamydia or gonorrhea in incarcer-
The New England Journal of Medicine
Downloaded from nejm.org on June 16, 2021. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.

n engl j med 373;26 nejm.org December 24, 2015
2521
Azithromycin vs. Doxycycline for Chlamydia Infection
ated adolescents. Sex Transm Dis 2005;
32: 11 5-22.
17. Renault CA, Israelski DM, Levy V, Fuji-
kawa BK, Kellogg TA, Klausner JD. Time
to clearance of Chlamydia trachom atis ribo-
somal R NA in women treated for chlamyd-
ial infection. Sex Health 2011; 8: 69-73.
18. Morré SA, Sillekens PT, Jacobs MV, et
al. Monitoring of Chlamydi a trachomati s in-
fections after antibiotic treatment using
RNA detection by nucleic acid sequence
based amplification. Mol Pathol 1998; 51:
149-54.
19. Workowski KA, Lampe MF, Wong KG,
Watts MB, Stamm WE. Long-term eradi-
cation of Chlamydia trachomatis genital in-
fection after antimicrobial therapy: evi-
dence against persistent infection. JAMA
1993; 270: 2071-5.
20. Gaydos CA, Crotchfelt KA, Howell
MR, Kralian S, Hauptman P, Quinn TC.
Molecular amplification assays to detect
chlamydial infections in urine specimens
from high school female students and to
monitor the persistence of chlamydial
DNA after therapy. J Infect Dis 1998; 177:
417-24.
21. Kapil R, Press CG, Hwang ML, Brown
L, Geisler WM. Investigating the epidemi-
ology of repeat Chlamydia trachomatis
detection after treatment by using C. tra-
chomatis OmpA genotyping. J Clin Micro-
biol 2015; 53: 546-9.
22. International Conference on Har-
monisation E9 Expert Working Group.
ICH Harmonised Tripartite Guideline:
statistical principles for clinical trials.
Stat Med 1999; 18: 1905-42.
23. Piaggio G, Elbourne DR, Pocock SJ,
Evans SJ, Altman DG. Reporting of non-
inferiority and equivalence randomized
trials: extension of the CONSORT 2010
statement. JAMA 2012; 308: 2594-604.
24. Somani J, Bhullar VB, Workowski KA,
Farshy CE, Black CM. Multiple drug-resis-
tant Chlamydia trachomatis associated with
clinical treatment failure. J Infect Dis
2000; 181: 1421-7.
25. Jones RB, Van der Pol B, Martin DH,
Shepard MK. Partial characterization of
Chlamydia trachomatis isolates resistant to
multiple a ntibiotics. J In fect Dis 1990; 162:
1309 -15.
26. Worm AM, Osterlind A. Azithromy-
cin levels in cervical mucus and plasma
after a single 1.0g oral dose for chlamyd-
ial cer vicitis. Genitourin Med 1995; 71:
244-6.
27. Krohn K. Gynaecological tissue levels
of azithromycin. Eur J Clin Microbiol In-
fect Dis 1991; 10: 864-8.
28. Reveneau N, Crane DD, Fischer E,
Caldwell HD. Bactericidal activity of first-
choice antibiotics against gamma inter-
feron-induced persistent infection of
human epithelial cells by Chlamydi a tracho-
matis. Antim icrob Agents Chemot her 2005;
49: 1787-93.
29. Dean D, Suchland R J, Stamm WE.
Evidence for long-term cer vical persis-
tence of Chlamydia trachomatis by omp1
genotyping. J Infect Dis 2000; 182: 909-16.
30. Recommendations for the laboratory-
based detection of Chlamydia trachomatis
and Neisseria gonorrhoeae — 2014. MMWR
Recomm Rep 2014; 63(RR-02): 1-19.
31. Chernesky M, Jang D, Gilchrist J, et al.
Head-to-head comparison of second-gen-
eration nucleic acid amplif ication tests
for detection of Chlamydia trachomatis and
Neisseria gonorrhoeae on urine samples from
female subjects and self-collected vaginal
swabs. J Clin Microbiol 2014; 52: 2305-10.
Copyright © 2015 Massachusetts Medical Society.
The New England Journal of Medicine
Downloaded from nejm.org on June 16, 2021. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.