Trial of Three Rounds of Mass Azithromycin Administration for Yaws Eradication

Abstract

Background: Treponema pallidum subspecies pertenue causes yaws. Strategies to better control, eliminate, and eradicate yaws are needed.

Methods: In an open-label, cluster-randomized, community-based trial conducted in a yaws-endemic area of Papua New Guinea, we randomly assigned 38 wards (i.e., clusters) to receive one round of mass administration of azithromycin followed by two rounds of target treatment of active cases (control group) or three rounds of mass administration of azithromycin (experimental group); round 1 was administered at baseline, round 2 at 6 months, and round 3 at 12 months. The coprimary end points were the prevalence of active cases of yaws, confirmed by polymerase-chain-reaction assay, in the entire trial population and the prevalence of latent yaws, confirmed by serologic testing, in a subgroup of asymptomatic children 1 to 15 years of age; prevalences were measured at 18 months, and the between-group differences were calculated.

Results: Of the 38 wards, 19 were randomly assigned to the control group (30,438 persons) and 19 to the experimental group (26,238 persons). A total of 24,848 doses of azithromycin were administered in the control group (22,033 were given to the participants at round 1 and 207 and 2608 were given to the participants with yaws-like lesions and their contacts, respectively, at rounds 2 and 3 [combined]), and 59,852 doses were administered in the experimental group. At 18 months, the prevalence of active yaws had decreased from 0.46% (102 of 22,033 persons) at baseline to 0.16% (47 of 29,954 persons) in the control group and from 0.43% (87 of 20,331 persons) at baseline to 0.04% (10 of 25,987 persons) in the experimental group (relative risk adjusted for clustering, 4.08; 95% confidence interval [CI], 1.90 to 8.76). The prevalence of other infectious ulcers decreased to a similar extent in the two treatment groups. The prevalence of latent yaws at 18 months was 6.54% (95% CI, 5.00 to 8.08) among 994 children in the control group and 3.28% (95% CI, 2.14 to 4.42) among 945 children in the experimental group (relative risk adjusted for clustering and age, 2.03; 95% CI, 1.12 to 3.70). Three cases of yaws with resistance to macrolides were found in the experimental group.

Conclusions: The reduction in the community prevalence of yaws was greater with three rounds of mass administration of azithromycin at 6-month intervals than with one round of mass administration of azithromycin followed by two rounds of targeted treatment. Monitoring for the emergence and spread of antimicrobial resistance is needed. (Funded by Fundació "la Caixa" and others; ClinicalTrials.gov number, NCT03490123.).

Figure 1. Trial profile

MDA: Mass drug administration. TTT: total targeted treatment. The estimated population was based on a survey conducted in 2016 for bed net distribution. The study area consisted of three Local Level Government (LLG) areas in Namatanai District. Each LLG in the study area has between 7-22 wards that are identified with consecutive numbers. Wards are the lowest administrative unit encompassing a group of 3 to 5 villages that share the same school and/or church. We used the ward as the randomization unit to reduce the risk of spill over between treatment arms; ward residents often attend the same school or church, but this is less common among residents of different wards. The mean population of the wards in Namatanai is 1,177 individuals (range 679-1,902), except for the district capital ward n# 10 in the Namatanai Rural LLG with 3,578. We selected 7/7 wards in Matalai Rural LLG, 14/14 wards in Namatanai Rural LLG, 17/22 wards in Sentral Niu Ailan LLG. Five island wards of Sentral Niu Ailand (1-4) were excluded due to remoteness. The coverage at the 6-month survey (Oct 2018) was lower due to a tropical cyclone that caused roadblocks and reduced access to many villages.

Figure 2. Allelic profile and macrolide resistance in PCR-confirmed yaws ulcer episodes.

We applied a variation of the three-amplicon multi-locus sequence typing (MLST) scheme previously described for T. pallidum subs. pertenue (see Supplementary Methods for details) which allowed us to grouping bacterial strains according to genetic relatedness, leading to three allelic profiles: J11, S22, and T13. TPE evolutionary distance (K80/K2P model) for the different rounds (not separated by arm) was 0.00207 at baseline, <0.0001 at 6 month, 0.00151, at 12 month, and 0.00055 at 18 months (P-value for the AMOVA when comparing baseline to 18 month 0. 2248)