new england

The
journal of medicine
established in 1812 february 19, 2015 vol. 372  no. 8

Mass Treatment with Single-Dose Azithromycin for Yaws

Oriol Mitjà, M.D., Ph.D., Wendy Houinei, H.E.O., Penias Moses, H.E.O., August Kapa, B.Sc., Raymond Paru, B.Sc.,
Russell Hays, M.D., Sheila Lukehart, Ph.D., Charmie Godornes, B.Sc., Sibauk Vivaldo Bieb, M.D., Tim Grice, Ph.D.,
Peter Siba, Ph.D., David Mabey, M.D., Ph.D., Sergi Sanz, M.Sc., Pedro L. Alonso, M.D., Ph.D.,
Kingsley Asiedu, M.D., M.P.H., and Quique Bassat, M.D., Ph.D.

A BS T R AC T

BACKGROUND
Mass treatment with azithromycin is a central component of the new World Health From Lihir Medical Center, International
Organization (WHO) strategy to eradicate yaws. Empirical data on the effective- SOS, Newcrest Mining, Lihir Island (O.M.,
P.M., A.K., R.P.), the Disease Control
ness of the strategy are required as a prerequisite for worldwide implementation Branch, National Department of Health,
of the plan. Port Moresby (W.H., S.V.B.), and the Papua
New Guinea Institute of Medical Re-
search, Goroka, Eastern Highland Prov-
METHODS ince (P.S.) — all in Papua New Guinea;
We performed repeated clinical surveys for active yaws, serologic surveys for latent Barcelona Institute for Global Health,
yaws, and molecular analyses to determine the cause of skin ulcers and identify Barcelona Center for International Health
Research, Hospital Clinic, University of
macrolide-resistant mutations before and 6 and 12 months after mass treatment Barcelona, Barcelona (O.M., S.S., P.L.A.,
with azithromycin on a Papua New Guinean island on which yaws was endemic. Q.B.); the College of Public Health, Medi-
Primary-outcome indicators were the prevalence of serologically confirmed active cal and Veterinary Sciences, James Cook
University, Cairns, QLD (R.H.), and the
infectious yaws in the entire population and the prevalence of latent yaws with Centre for Social Responsibility in Mining,
high-titer seroreactivity in a subgroup of children 1 to 15 years of age. Sustainable Minerals Institute, University
of Queensland, Brisbane (T.G.) — both
in Australia; the Departments of Medi-
RESULTS cine (S.L., C.G.) and Global Health (S.L.),
At baseline, 13,302 of 16,092 residents (82.7%) received one oral dose of azithromy- University of Washington, Seattle; the
cin. The prevalence of active infectious yaws was reduced from 2.4% before mass Department of Clinical Research, London
School of Hygiene and Tropical Medicine,
treatment to 0.3% at 12 months (difference, 2.1 percentage points; P<0.001). The London (D.M.); and the Department of
prevalence of high-titer latent yaws among children was reduced from 18.3% to Control of Neglected Tropical Diseases,
6.5% (difference, 11.8 percentage points; P<0.001) with a near-absence of high-titer World Health Organization, Geneva (K.A.).
Address reprint requests to Dr. Mitjà at
seroreactivity in children 1 to 5 years of age. Adverse events identified within 1 week the Department of Community Health,
after administration of the medication occurred in approximately 17% of the par- Lihir Medical Center, P.O. Box 34, Lihir
ticipants, included nausea, diarrhea, and vomiting, and were mild in severity. No Island, New Ireland Province, Papua New
Guinea, or at [email protected].
evidence of emergence of resistance to macrolides against Treponema pallidum sub-
species pertenue was seen. N Engl J Med 2015;372:703-10.
DOI: 10.1056/NEJMoa1408586
Copyright © 2015 Massachusetts Medical Society.
CONCLUSIONS
The prevalence of active and latent yaws infection fell rapidly and substantially 12
months after high-coverage mass treatment with azithromycin, with the reduction
perhaps aided by subsequent activities to identify and treat new cases of yaws. Our
results support the WHO strategy for the eradication of yaws. (Funded by Newcrest
Mining and International SOS; YESA-13 ClinicalTrials.gov number, NCT01955252.)

n engl j med 372;8 nejm.org february 19, 2015 703

The New England Journal of Medicine
Downloaded from nejm.org on April 21, 2016. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Y
aws, an infectious disease caused by hidden infection in the community.12 We used
Treponema pallidum subspecies pertenue, af- clinical surveys, serologic surveys, and etiologic
fects mainly children in poor rural com- studies of skin ulcers to measure the effect of
munities in tropical countries. This bacterium is mass azithromycin treatment on the community
transmitted by direct skin-to-skin, nonsexual burden of yaws.
contact and causes a chronic, relapsing disease
that is characterized by highly contagious pri- ME THODS
mary and secondary cutaneous lesions and by
noncontagious tertiary destructive lesions of the STUDY SETTING AND PARTICIPANTS
bones.1 Cases are reported currently in 12 coun- From April 2013 through May 2014, we conduct-
tries in Africa, Asia, and the western Pacific re- ed a longitudinal study in all the villages of Lihir
gion, and 42 million people are estimated to be Island, New Ireland Province, Papua New Guinea.
at risk for yaws.2 Lihir Island consists of 28 villages with a range of
A major campaign in the 1950s and 1960s to 400 to 600 inhabitants each, and the estimated
eradicate yaws by means of community-wide population in 2013 was 16,092 according to a
treatment with long-acting, injectable penicillin regularly updated census. All the villages have
reduced the number of cases of the disease by been reported to have a relatively high prevalence
A Quick Take
animation is
95% worldwide; however, yaws was not eradi- of active yaws (range, 0.5 to 3.8%),13 and mass
available at cated.3 The discovery that a single dose of oral treatment with azithromycin had never been at-
NEJM.org azithromycin is at least as effective as injectable tempted before our baseline surveys. An increased
penicillin G benzathine4 prompted the World number of active yaws cases is observed during
Health Organization (WHO) to develop a new the wet season (January through June).
azithromycin-based treatment policy in 2012 (the The implementation of the Morges elimina-
Morges strategy).5 The recommendation is initial tion strategy in the villages started after baseline
mass treatment of the entire population (often assessment surveys, in accordance with standards
called total community treatment) with a single advocated by the WHO.5 During the initial mass
dose of oral azithromycin, followed by resurveys treatment program, everyone older than 2 months
every 6 months in a targeted treatment program of age in the study villages was offered azithro-
to detect and treat newly identified persons with mycin (Medopharm) at a single oral dose of
active yaws and their contacts. 30 mg per kilogram of body weight, up to a
The Morges strategy has several advantages maximum dose of 2 g. Pregnant women and
over previous campaigns, including oral versus people with a known allergy to macrolides were
parenteral administration of the drug 6 and mass offered penicillin G benzathine at a dose of
treatment regardless of prevalence of yaws in the 50,000 U per kilogram, administered intramus-
community versus selective treatment of active cularly. The WHO provided generic azithromy-
cases, without treatment of latent cases, which cin, which was purchased at full cost from Medo-
can develop into infectious yaws lesions.7,8 In pharm.
earlier campaigns, it was neglect of latent infec- The initial mass treatment program was fol-
tion that resulted in the rapid return of the lowed every 6 months by a targeted treatment
disease.9 This risk is greatly reduced by the new program, in which surveys were performed that
mass treatment recommendation.10,11 consisted of clinical examination of the resident
Empirical data on the effectiveness of the population, with treatment of all persons with
Morges strategy to stop the transmission of yaws active clinical cases and their contacts (household
in a geographically defined area are required as members, classmates, and playmates). Between
a prerequisite for worldwide implementation of those surveys, patients with active yaws who
the plan. Data to determine the burden of yaws presented to a health care facility (passive case
in a community include the prevalence of active finding) and their close contacts were treated.
infectious yaws (i.e., skin ulcers or papilloma) Treatment was provided without cost to the
and the prevalence of seroreactivity among pa- study participants. Passive surveillance for ad-
tients without active yaws; these data indicate, verse events was undertaken throughout the
respectively, the extent to which yaws transmis- study at the Lihir Medical Center and all periph-
sion is occurring and the extent of the latent or eral health posts. Specific training for health

704 n engl j med 372;8 nejm.org february 19, 2015

The New England Journal of Medicine

Downloaded from nejm.org on April 21, 2016. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.
 MASS TREATMENT WITH SINGLE-DOSE AZITHROMYCIN FOR YAWS

care staff was conducted before the interven- guide.14 A clinical diagnosis of active infectious
tions began, with a special emphasis on report- lesions was based on chronic (symptomatic for
ing any allergic event or any other adverse event >2 weeks) solitary or multiple ulcers or papillo-
that was deemed to be possibly related to the mas (Fig. S1 and S2 in the Supplementary Appen-
intervention, with the use of a standard case- dix, available at NEJM.org).15 Serologic confir-
report form for adverse events. We also per- mation of treponemal infection with the use of
formed household surveys 1 week after the ini- a T. pallidum hemagglutination assay (TPHA) and
tial distribution of the antibiotic agents to RPR testing was performed in persons with ac-
monitor for potential adverse events in 60 ran- tive lesions. In resurveys at 6 and 12 months,
domly selected households from 28 villages. persons with new cases of active yaws were
All the participants, or their parent or guard- classified as absentees from initial treatment
ian, provided oral informed consent for screening surveys; previously untreated visitors, returning
and treatment. In addition, we obtained written laborers, or migrants with clinically active le-
informed consent from persons with suspected sions; or previously treated local residents. For
active yaws before their enrollment in clinical previously treated persons, we performed a me-
surveys and etiologic studies of ulcers and from ticulous inquiry to find the source of reinfection.
parents or guardians of children recruited for Serologic screening to detect latent yaws was
serologic surveys; we obtained verbal agreement conducted in a subgroup of asymptomatic chil-
from the children. The study protocol was ap- dren 1 to 15 years of age in six randomly selected
proved by the Medical Research Advisory Com- villages (chosen with the use of computer-gener-
mittee of the Papua New Guinea National De- ated random numbers). Because the random
partment of Health (approval number, 12.36). sample was regenerated at each survey, villages
Full details of the study conduct are provided in might have been chosen repeatedly. The age cri-
the protocol, available with the full text of this teria for inclusion in the serologic surveys were
article at NEJM.org. The sponsors of the study intended to reduce the likelihood of reactive sero-
had no role in the design of the study, the col- logic findings related to venereal syphilis. Venous
lection, analysis, or interpretation of the data, or blood samples were obtained from assenting
the writing of the manuscript. The first author children for TPHA and qualitative and quantita-
had full access to all the study data and had the tive RPR testing.
final responsibility for the decision to submit All asymptomatic children with a reactive TPHA
the manuscript for publication. and an RPR titer of at least 1:2 were classified as
having latent yaws; those with a high RPR titer
PROCEDURES (≥1:16) were classified as having high-titer latent
Primary-outcome indicators were the prevalence yaws. Children with latent yaws with high titers
of serologically confirmed active infectious yaws would be more likely than those with low titers
in the entire population and the prevalence of to have a clinical reactivation with active lesions.9
latent yaws with high-titer seroreactivity (rapid We classified asymptomatic children according
plasma reagin [RPR], ≥1:16) in a subgroup of to age (1 to 5 years vs. 6 to 15 years). Seroreac-
children 1 to 15 years of age, at baseline and at tivity in young children (1 to 5 years of age) can
6 and 12 months. Secondary-outcome indicators indicate recent infection, because these children
included the proportion of ulcers caused by are new entrants in the potential pool of suscep-
T. pallidum subspecies pertenue, as assessed with tible persons. The WHO criteria to certify the
the use of a polymerase-chain-reaction (PCR) interruption of transmission include no young
assay, and the proportion of yaws samples with children with RPR seroreactivity and no new
genetic mutations associated with macrolide re- cases of active yaws for 3 consecutive years.5
sistance at each time point. All serologic tests for syphilis were per-
Clinical surveys for active infectious yaws le- formed at the Lihir Medical Center laboratory,
sions were undertaken in the entire resident with external quality-control testing (i.e., 5% of
population. Active yaws is usually a visible dis- positive and negative samples) performed at Sul-
ease; hence clinical surveys provide useful knowl- livan Nicolaides Pathology in Queensland, Aus-
edge. We undertook screening examination of tralia. The laboratory scientists were unaware of
all the villagers, using the WHO yaws-pictorial the antibiotic coverage and clinical outcomes.

n engl j med 372;8 nejm.org february 19, 2015 705

The New England Journal of Medicine
Downloaded from nejm.org on April 21, 2016. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

PCR surveillance was performed by analysis multinomial logistic-regression model. A P value

of swabs obtained during clinical surveys from of less than 0.05 was considered to indicate sta-
each participant with papillomatous or ulcerative tistical significance. All P values are two-sided.
lesions. The PCR methods have been described
previously.16 After every survey, we forwarded R E SULT S
90 randomly selected specimens to the labora-
tory at the University of Washington, Seattle, for STUDY POPULATION
molecular testing with the use of PCR to detect At baseline, we examined 13,490 of the 16,092
T. pallidum DNA,17-19 a molecular signature spe- residents (83.8%). Of the 13,490 participants
cific to subspecies pertenue (confirming yaws examined, 13,302 received azithromycin, 177 re-
infection),19 evidence of mutations conferring ceived penicillin G benzathine, and 11 declined
resistance to azithromycin,20 and Haemophilus treatment. Treatment with oral drugs was ob-
ducreyi DNA, which may coexist with yaws as a served directly. The overall rate of treatment cov-
cause of skin ulcers.16 erage during the mass treatment program was
83.8%, and all the study villages had a coverage
STATISTICAL ANALYSIS rate of more than 70.0%.
We double-entered data in Access software, ver- No severe adverse events attributable to the
sion 14.0 (Microsoft), with discrepancies checked study drug were reported by means of passive
against original forms. Using Stata software, ver- surveillance during the study. Active surveillance
sion 13.1 (StataCorp), we calculated the prevalence of 316 participants from 60 households yielded
of active yaws, assessing everyone in the study 54 participants (17.1%) who reported adverse
population at three time points. We estimated events (all mild), including 30 (9.5%) with nausea
the prevalence of latent yaws in subgroups of or abdominal pain, 25 (7.9%) with diarrhea, and
children in randomly selected villages. We calcu- 15 (4.7%) with vomiting (Table S1 in the Supple-
lated that a sample of 875 children would provide mentary Appendix).
the study with 80% power to estimate the preva- Coverage rates of screening for active cases
lence of high-titer latent yaws with a precision of during the targeted treatment program in 28
1.5%, at a two-sided significance level of 5%. We villages were 81.8% at 6 months and 82.2% at
assumed that the prevalence of high-titer latent 12 months. The most common reasons for ab-
yaws at 12 months would be 5%.10 We estimated sence during treatment and follow-up were travel
that this sample size would provide the study and work.
with 100% power to detect a prevalence differ-
ence of 11.8 percentage points between baseline CHANGES IN THE PREVALENCE OF ACTIVE DISEASE
and the 12-month follow-up. The prevalence of infectious active yaws fell from
We estimated the prevalence ratio for the 2.4% at baseline to 0.3% at 6 months and re-
comparison of active and latent yaws at three mained at 0.3% at 12 months (difference from
time points using a log-binomial regression baseline, 2.1 percentage points; 95% confidence
model. We evaluated the decline in the preva- interval [CI], 1.9 to 2.4; P<0.001) (Table 1). In all
lence of PCR-detected infection over time using a the surveys, the community burden of yaws-­

Table 1. Prevalence of Clinically Active Yaws.

Time No. of Persons Serologically Confirmed Clinically Suspected Lesions

Point Examined Active Yaws with Negative Serologic Findings
No. of Persons Prevalence Ratio No. of Persons Prevalence Ratio
(%) (95% CI)* (%) (95% CI)*
Baseline 13,490 323 (2.4) 1.00 367 (2.7) 1.00
6 mo 13,166 44 (0.3) 0.14 (0.10–0.19) 77 (0.6) 0.22 (0.17–0.27)
12 mo 13,204 34 (0.3) 0.11 (0.08–0.15) 82 (0.6) 0.23 (0.18–0.29)

* The prevalence ratio was calculated by means of the log-binomial regression model. The baseline prevalence is the ref-
erence value. P<0.001 for the comparison with baseline.

706 n engl j med 372;8 nejm.org february 19, 2015

The New England Journal of Medicine

Downloaded from nejm.org on April 21, 2016. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.
 MASS TREATMENT WITH SINGLE-DOSE AZITHROMYCIN FOR YAWS

related ulcers was disproportionately borne by

children 15 years of age or younger (Fig. 1). Sim- Baseline 6 Mo 12 Mo
ilarly, there was a decline in the proportion of 140

participants with clinically suspicious lesions and 120

negative serologic findings, with significant reduc-

Infectious Active Yaws

(no. of participants)
100
tions from 2.7% at baseline to 0.6% at 12 months
(difference, 2.1 percentage points; 95% CI, 1.8 to 80

2.4; P<0.001). Because H. ducreyi is also a major 60

pathogen in skin ulcers on Lihir Island,17 mass
40
treatment probably also reduced the burden of
lesions caused by H. ducreyi. 20
Most active cases of yaws that were identified 0
at resurveys occurred in local residents who had 1– 5 6– 10 11– 15 16– 20 21– 25 26– 30 31– 35
been absent from initial treatment surveys: 33 of Age Group (yr)
44 cases (75.0%) at 6 months, and 21 of 34
Figure 1. Prevalence of Active Disease over Time in Various Age Groups.
(61.8%) at 12 months. Some yaws lesions oc-
curred in nontreated visitors or migrants: 2 cases
(4.5%) at 6 months, and 4 (11.8%) at 12 months.
A total of 9 cases of ulcers (20.5%) at 6 months at the baseline, 6-month, and 12-month surveys,
and 9 (26.5%) at 12 months occurred in previ- respectively. No significant differences were seen
ously treated long-term residents of Lihir Island. in the ratio of boys to girls at the three time
Of these 18 cases, 11 could be traced directly to points.
contact with infectious visitors or untreated lo- The prevalence of high-titer latent yaws de-
cal residents (i.e., probable reinfection), but no creased from 18.3% at baseline to 6.5% at 12
source of reinfection was found in 7 cases that months (difference, 11.8 percentage points; 95%
could have been due to a relapse of an inade- CI, 8.9 to 14.7; P<0.001) (Table 2), with major
quately treated latent infection. declines in titer from 6 to 12 months. This result
suggests that decreases in titer require a pro-
CHANGES IN THE PREVALENCE OF LATENT YAWS longed time after treatment. We also noted a
Table 2 shows cross-sectional surveys for latent significant reduction in the prevalence of all
yaws in asymptomatic children from randomly seropositive results, from 32.8% at baseline to
selected villages. At baseline, we obtained blood 16.4% at 12 months (difference, 16.4 percentage
samples from 991 children; 874 children were test- points; 95% CI, 12.6 to 20.2; P<0.001).
ed at 6 months, and 910 were tested at 12 months. In subgroup analyses, the prevalence of high-
The mean (±SD) ages of the children were titer seropositivity in the group of children 1 to
10.4±3.6 years, 10.4±3.2 years, and 10.0±3.7 years 5 years of age fell significantly, from 13.7% at

Table 2. Prevalence of Latent Yaws.

Time No. of Children

Point Tested All Cases of Latent Yaws* High-Titer Latent Yaws†
No. of Prevalence Ratio No. of Prevalence Ratio
Children (%) (95% CI)‡ Children (%) (95% CI)‡
Baseline 991 325 (32.8) 1.00 181 (18.3) 1.00
6 mo 874 261 (29.9) 0.91 (0.80–1.04) 123 (14.1) 0.77 (0.62–0.95)
12 mo 910 149 (16.4) 0.50 (0.42–0.59) 59 (6.5) 0.36 (0.27–0.47)

* The analysis included all seropositive children with a reactive result on the Treponema pallidum hemagglutination assay
(TPHA) and with a rapid plasma reagin (RPR) titer of at least 1:2.
† The analysis included children with a reactive result on the TPHA and with an RPR titer of at least 1:16.
‡ The prevalence ratio was calculated with the use of a log-binomial regression model. The baseline prevalence is the ref-
erence value. P<0.001 for the comparison with baseline.

n engl j med 372;8 nejm.org february 19, 2015 707

The New England Journal of Medicine
Downloaded from nejm.org on April 21, 2016. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 3. Subgroup Analyses of the Prevalence of Latent Yaws.

Time Point Children 1–5 Yr of Age Children 6–15 Yr of Age

No. of No. of
Children All Cases High-Titer Children All Cases High-Titer
Tested of Latent Yaws* Latent Yaws† Tested of Latent Yaws* Latent Yaws†
prevalence prevalence prevalence prevalence
no. of ratio no. of ratio no. of ratio no. of ratio
children (%) (95% CI)‡§ children (%) (95% CI)‡¶ children (%) (95% CI)‡‖ children (%) (95% CI)‡‖
Baseline 117 26 1.00 16 1.00 874 299 (34.2) 1.00 165 1.00
(22.2) (13.7) (18.9)
6 mo 77 10 0.58 6 0.57 797 251 (31.5) 0.92 117 0.78
(13.0) (0.30–1.14) (7.8) (0.23–1.39) (0.80–1.06) (14.7) (0.63–0.97)
12 mo 114 6 0.24 1 0.06 796 143 (18.0) 0.53 58 0.39
(5.3) (0.10–0.55) (0.9) (0.01–0.48) (0.44–0.63) (7.3) (0.29–0.51)

* The analysis included all seropositive children with a reactive TPHA and RPR titer of at least 1:2.
† The analysis included children with a reactive TPHA and RPR titer of at least 1:16.
‡ The prevalence ratio was calculated with the use of a log-binomial regression model. The baseline prevalence is the reference value.
§ P = 0.003 for the comparison with baseline.
¶ P = 0.02 for the comparison with baseline.
‖ P<0.001 for the comparison with baseline.

baseline to 0.9% at 12 months (difference, At baseline, T. pallidum subspecies pertenue alone

12.8 percentage points; 95% CI, 6.3 to 19.3; was identified in 19 of 90 participants (21.1%),
P = 0.02). The change in prevalence in the group H. ducreyi alone in 42 (46.7%), and co­infection
of children 6 to 15 years of age was also signifi- with both organisms in 12 (13.3%), as reported
cant (difference, 11.6 percentage points; 95% CI, previously.16 Although the prevalence of active
8.4 to 14.8; P<0.001) (Table 3). lesions due to any cause was greatly reduced
6 months after treatment, the proportion of pa-
PROPORTION OF PCR-Confirmed ULCERS DUE tients with lesions due to yaws did not decrease
TO YAWS significantly, as compared with baseline, and
PCR analyses of lesion swabs to detect T. pallidum the proportion of patients with dual infection
subspecies pertenue and H. ducreyi DNA were increased (Table 4). However, at 12 months,
available for 90 participants at baseline, for 84 at there was a significant reduction in the propor-
6 months, and for 114 at 12 months. A bacterial tion of ulcers containing either T. pallidum sub-
cause was identified in 73 participants (81.1%) species pertenue alone (risk difference, 10.6 per-
at baseline, in 73 (86.9%) at 6 months, and in 72 centage points; 95% CI, 0.4 to 20.7; P = 0.04) or
(63.2%) at 12 months. Overall, 70 of 288 people coinfection (risk difference, 7.2 percentage points;
(24.3%) who were tested in the three rounds had 95% CI, −1.1 to 15.5; P = 0.08), whereas the pro-
a negative result on all molecular tests. portion of ulcers containing H. ducreyi alone

Table 4. Results of Polymerase-Chain-Reaction Assay of Lesion Swabs.*

Treponema pallidum
Time Participants Subspecies pertenue Dual Infection Hae­mophilus ducreyi Negative
Point Tested Only Detected Detected Only Detected in All Tests
no. no. of participants (%)
Baseline 90 19 (21.1) 12 (13.3) 42 (46.7) 17 (18.9)
6 mo 84 14 (16.7) 27 (32.1) 32 (38.1) 11 (13.1)
12 mo 114 12 (10.5) 7 (6.1) 53 (46.5) 42 (36.8)

* P<0.001 by the chi-square test for the between-group comparison within each type of infection.

708 n engl j med 372;8 nejm.org february 19, 2015

The New England Journal of Medicine

Downloaded from nejm.org on April 21, 2016. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.
 MASS TREATMENT WITH SINGLE-DOSE AZITHROMYCIN FOR YAWS

remained at a level similar to the baseline level was no untreated control group. Although a
(risk difference, 0.1 percentage points; 95% CI, cluster-randomized trial design would have pro-
−13.6 to 14.0; P = 0.98). At all the time points, vided stronger evidence of effect, we can reason-
clusters of T. pallidum subspecies pertenue were ably assume that the decrease in transmission and
identified among family members. endemicity of yaws is attributable to mass treat-
All 91 confirmed specimens for yaws in the ment with an antibiotic agent. Second, the effect
three surveys were positive for several different of rainfall on the prevalence of active yaws has
T. pallidum gene targets and had a tprL molecular been well documented, and this effect may have
signature that identified them as subspecies biased our results.21 However, both the baseline
pertenue. All T. pallidum samples obtained before survey and the 12-month survey were conducted
and after mass treatment had wild-type 23S ribo- during the wet season, and there was a decline
somal DNA findings at positions 2058 and 2059, in active yaws cases between these two surveys.
a finding that is consistent with susceptibility to Finally, elimination of disease is generally easier
azithromycin. to accomplish on islands than in contiguous com-
munities, which raises questions regarding the
DISCUSSION generalizability of our findings. Records show
more than 1000 visitors per month between the
Our study showed that one round of mass treat- mainland and Lihir Island, but the risk of local
ment with azithromycin greatly reduced the transmission from imported cases is low be-
transmission and endemicity of yaws in Papua cause staff at rural health facilities have been
New Guinean villages that had high baseline carefully trained to recognize yaws lesions. Peri-
rates of infection. Implementation of the WHO odic resurveys must be maintained, however, to
strategy reduced clinical manifestations (ulcers consolidate the achievements of mass treatment
and papillomas) by 90%, thus reducing the likeli- and to ensure the elimination of yaws.
hood of transmission of infection to susceptible In summary, we observed a sustained quanti-
(uninfected) persons. This hypothesis is support- tative reduction in the prevalence of yaws that
ed by the near-absence of high-titer seroreactivity was most likely due to mass treatment with
among children 1 to 5 years of age (with seroreac- azithromycin. In addition, macrolide resistance
tivity in these young children indicating relatively in T. pallidum subspecies pertenue was monitored,
recent infection) and by reductions in the propor- and no emergence of common resistance muta-
tion of skin ulcers attributable to yaws (confirmed tions was seen; however, close monitoring is
by means of species-specific PCR assay). The in- required both for the emergence of resistance
tervention also reduced the proportion of sero- mutations in yaws and for the effect of macro-
positive persons, who may be at risk for clinical lide resistance on other colonizing flora. Although
reactivation, and there was no evidence of resis- the reintroduction of infection is a potential
tance to azithromycin after mass treatment. These risk, the magnitude of this risk depends on the
results support the use of the Morges strategy for adequacy of surveillance. Our findings provide
the elimination of yaws. evidence of the effectiveness of the WHO strategy
However, our findings suggest that 80% popu- regarding yaws. If this strategy is similarly effec-
lation coverage in mass treatment is not suffi- tive in other communities, if a high level of aware-
cient to extinguish local transmission, findings ness among health workers and the populations
that highlight the importance of WHO recom- is maintained, and if political and financial
mendations of nearly 100% coverage during ini- commitments are forthcoming, the control and
tial mass treatment and implementation of sub- potential eradication of yaws may be attainable.
sequent resurveys to detect and treat residual
cases. The number and frequency of resurveys Supported by Newcrest Mining and International SOS.
Disclosure forms provided by the authors are available with
that are necessary to achieve elimination is un- the full text of this article at NEJM.org.
known. Monitoring will be needed until no more We thank the village chairmen, elders, and villagers of Lihir
cases of active yaws are found and serologic tests Island for their willingness to be involved in the study; our field
teams for efforts with study implementation; and the National
among children 1 to 5 years of age prove negative. Department of Health of Papua New Guinea for guidance and
Our study has several limitations. First, there oversight of the trial and continued cooperation.

n engl j med 372;8 nejm.org february 19, 2015 709

The New England Journal of Medicine
Downloaded from nejm.org on April 21, 2016. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.
 MASS TREATMENT WITH SINGLE-DOSE AZITHROMYCIN FOR YAWS

REFERENCES
1. Mitjà O, Asiedu K, Mabey D. Yaws. 10. Zahra A. Yaws eradication campaign cohort study. Lancet Glob Health 2014;
Lancet 2013;381:763-73. in Nsukka Division, Eastern Nigeria. Bull 2(4):e235-e241.
2. Global Health Observatory Data Re- World Health Organ 1956;15:911-35. 17. Centurion-Lara A, Castro C, Shaffer
pository: yaws. Geneva: World Health Or- 11. Samame GE. Treponematosis eradica- JM, et al. Detection of Treponema palli-
ganization, 2013 (http://apps.who.int/gho/ tion, with special reference to yaws eradi- dum by a sensitive reverse transcriptase
data/node.main.NTDYAWS). cation in Haiti. Bull World Health Organ PCR. J Clin Microbiol 1997;35:1348-52.
3. Meheus A, Antal GM. The endemic 1956;15:897-910. 18. Marra CM, Sahi SK, Tantalo LC, et al.
treponematoses: not yet eradicated. World 12. Hackett CJ, Guthe T. Some important Enhanced molecular typing of Treponema
Health Stat Q 1992;45:228-37. aspects of yaws eradication. Bull World pallidum: geographical distribution of
4. Mitjà O, Hays R, Ipai A, et al. Single- Health Organ 1956;15:869-96. strain types and association with neuro-
dose azithromycin versus benzathine ben- 13. Mitjà O, Hays R, Ipai A, et al. Out- syphilis. J Infect Dis 2010;202:1380-8.
zylpenicillin for treatment of yaws in chil- come predictors in treatment of yaws. 19. Centurion-Lara A, Giacani L, ­Godornes
dren in Papua New Guinea: an open-label, Emerg Infect Dis 2011;17:1083-5. C, Molini BJ, Brinck Reid T, Lukehart SA.
non-inferiority, randomised trial. Lancet 14. Yaws: recognition booklet for commu- Fine analysis of genetic diversity of the tpr
2012;379:342-7. nities. Geneva: World Health Organization, gene family among treponemal species,
5. Eradication of yaws — the Morges strat- 2013 (http://apps.who.int/iris/bitstream/ subspecies and strains. PLoS Negl Trop
egy. Wkly Epidemiol Rec 2012;87:189-94. 10665/75360/1/9789241504096_eng.pdf ? Dis 2013;7(5):e2222.
6. Mabey D. Oral azithromycin for treat- ua=1). 20. Lukehart SA, Godornes C, Molini BJ,
ment of yaws. Lancet 2012;379:295-7. 15. Summary report of a consultation on et al. Macrolide resistance in Treponema
7. Report of Second International Con- the eradication of yaws: 5–7 March 2012, pallidum in the United States and Ireland.
ference on Control of Yaws: Nigeria, 1955. Morges, Switzerland. Geneva: World N Engl J Med 2004;351:154-8.
J Trop Med Hyg 1957;60:62-73. Health Organization, 2012 (http://apps 21. Hill KR. Non-specific factors in the
8. Li HY, Soebekti R. Serological study .who.int/iris/bitstream/10665/75528/1/ epidemiology of yaws. Bull World Health
of yaws in Java. Bull World Health Organ WHO_HTM_NTD_IDM_2012.2_eng.pdf). Organ 1953;8:17-51.
1955;12:905-43. 16. Mitjà O, Lukehart SA, Pokowas G, et al. Copyright © 2015 Massachusetts Medical Society.
9. Hackett CJ. Some epidemiological as- Haemophilus ducreyi as a cause of skin
pects of yaws eradication. Bull World ulcers in children from a yaws-endemic
Health Organ 1960;23:739-61. area of Papua New Guinea: a prospective

an nejm app for iphone

The NEJM Image Challenge app brings a popular online feature to the smartphone.
Optimized for viewing on the iPhone and iPod Touch, the Image Challenge app lets
you test your diagnostic skills anytime, anywhere. The Image Challenge app
randomly selects from 300 challenging clinical photos published in NEJM,
with a new image added each week. View an image, choose your answer,
get immediate feedback, and see how others answered.
The Image Challenge app is available at the iTunes App Store.

710 n engl j med 372;8 nejm.org february 19, 2015

The New England Journal of Medicine

Downloaded from nejm.org on April 21, 2016. For personal use only. No other uses without permission.
Copyright © 2015 Massachusetts Medical Society. All rights reserved.