Mass drug administration for the acceleration of malaria elimination in a region of Myanmar with artemisinin-resistant falciparum malaria: a cluster-randomised trial

Abstract

Background: To contain multidrug-resistant Plasmodium falciparum, malaria elimination in the Greater Mekong subregion needs to be accelerated while current antimalarials remain effective. We evaluated the safety, effectiveness, and potential resistance selection of dihydroartemisinin-piperaquine mass drug administration (MDA) in a region with artemisinin resistance in Myanmar.

Methods: We did a cluster-randomised controlled trial in rural community clusters in Kayin (Karen) state in southeast Myanmar. Malaria prevalence was assessed using ultrasensitive quantitative PCR (uPCR) in villages that were operationally suitable for MDA (villages with community willingness, no other malaria control campaigns, and a population of 50-1200). Villages were eligible to participate if the prevalence of malaria (all species) in adults was greater than 30% or P falciparum prevalence was greater than 10% (or both). Contiguous villages were combined into clusters. Eligible clusters were paired based on P falciparum prevalence (estimates within 10%) and proximity. Community health workers provided routine malaria case management and distributed long-lasting insecticidal bed-nets (LLINs) in all clusters. Randomisation of clusters (1:1) to the MDA intervention group or control group was by public coin-flip. Group allocations were not concealed. Three MDA rounds (3 days of supervised dihydroartemisinin-piperaquine [target total dose 7 mg/kg dihydroartemisinin and 55 mg/kg piperaquine] and single low-dose primaquine [target dose 0·25 mg base per kg]) were delivered to intervention clusters. Parasitaemia prevalence was assessed at 3, 5, 10, 15, 21, 27, and 33 months. The primary outcomes were P falciparum prevalence at months 3 and 10. All clusters were included in the primary analysis. Adverse events were monitored from the first MDA dose until 1 month after the final dose, or until resolution of any adverse event occurring during follow-up. This trial is registered with ClinicalTrials.gov, NCT01872702.

Findings: Baseline uPCR malaria surveys were done in January, 2015, in 43 villages that were operationally suitable for MDA (2671 individuals). 18 villages met the eligibility criteria. Three villages in close proximity were combined into one cluster because a border between them could not be defined. This gave a total of 16 clusters in eight pairs. In the intervention clusters, MDA was delivered from March 4 to March 17, from March 30 to April 10, and from April 27 to May 10, 2015. The weighted mean absolute difference in P falciparum prevalence in the MDA group relative to the control group was -10·6% (95% CI -15·1 to -6·1; p=0·0008) at month 3 and -4·5% (-10·9 to 1·9; p=0·14) at month 10. At month 3, the weighted P falciparum prevalence was 1·4% (0·6 to 3·6; 12 of 747) in the MDA group and 10·6% (7·0 to 15·6; 56 of 485) in the control group. Corresponding prevalences at month 10 were 3·2% (1·5 to 6·8; 34 of 1013) and 5·8% (2·5 to 12·9; 33 of 515). Adverse events were reported for 151 (3·6%) of 4173 treated individuals. The most common adverse events were dizziness (n=109) and rash or itching (n=20). No treatment-related deaths occurred.

Interpretation: In this low-transmission setting, the substantial reduction in P falciparum prevalence resulting from support of community case management was accelerated by MDA. In addition to supporting community health worker case management and LLIN distribution, malaria elimination programmes should consider using MDA to reduce P falciparum prevalence rapidly in foci of higher transmission.

Funding: The Global Fund to Fight AIDS, Tuberculosis and Malaria.

Figure 1. Map of study region in Eastern Kayin state with P. falciparum prevalences found at screening survey.

Figure 2

Study profile. MDA = Mass Drug Administration; RDT = Malaria Rapid Diagnostic Test; M=month i.e. M3, M33 = Month 3, Month 33; * Individuals present during census but away from the cluster for all rounds of medication delivery.

Figure 3

Study Timeline. Small circles denote ongoing activities; vertical arrows denote activities at specific times; M and C denote activities at specific times in MDA (M) and Control (C) clusters and black triangles represent a round of MDA.

Figure 4. Malaria species detected by uPCR at baseline survey.

Figure 5. Weighted P. falciparum prevalence (assessed by uPCR) by arm.

Weighted P. falciparum prevalence by survey time-points in MDA (A) and control (B) arms with 95% Confidence Intervals (capped bars). S = screening; M = month.

Figure 6. Individual P. falciparum prevalences (assessed by uPCR) by cluster pairs.

A; Red circles (MDA clusters) and blue squares (control clusters) denote point estimates of P. falciparum prevalences, capped bars denote 95% confidence intervals.

Figure 7. Weighted mean difference between the P. falciparum prevalence in the MDA and control arms.

Circles denote point estimates and capped bars denote the 95% confidence intervals. Red dashed line denotes start of MDA.

Figure 8. Weighted P. vivax prevalences (assessed by uPCR) by arms.

Weighted P. vivax prevalences by survey time-points in MDA (A) and control (B) arms with 95% Confidence Intervals (capped bars). S = screening; M = month.

Figure 9. Individual P. vivax prevalences (assessed by uPCR) by cluster pairs.

A; Red circles (MDA clusters) and blue squares (control clusters) denote point estimates of P. vivax prevalence, capped bars denote 95% confidence intervals.

Figure 10. Weighted mean difference in the P. vivax prevalence between the MDA relative and control arms.

Black circles denote point estimates and capped bars denote the 95% confidence intervals.

Figure 11. Monthly (A) P. falciparum and (B) P. vivax incidences in the intervention (red) and control arms (blue).

Data were excluded if the matched cluster in the other arm did not yet have an operational CHW. Lines connect time points with the same number of cluster pairs included. 10 clusters provide data from study month -10; 12 from study month -7; 14 from study month-3 and 16 from study month 7. Vertical dashed grey line indicates MDA start (Month 0). For incidence of balanced cohorts of clusters see Supplementary Figures 2-4. For incidence of each individual cluster see Supplementary Figures 5-6.