A randomized controlled trial of dihydroartemisinin-piperaquine, artesunate-mefloquine and extended artemether-lumefantrine treatments for malaria in pregnancy on the Thailand-Myanmar border

Background Artemisinin and artemisinin-based combination therapy (ACT) partner drug resistance in Plasmodium falciparum have spread across the Greater Mekong Subregion compromising antimalarial treatment. The current 3-day artemether-lumefantrine regimen has been associated with high treatment failure rates in pregnant women. Although ACTs are recommended for treating Plasmodium vivax malaria, no clinical trials in pregnancy have been reported. Methods Pregnant women with uncomplicated malaria on the Thailand-Myanmar border participated in an open-label randomized controlled trial comparing dihydroartemisinin-piperaquine (DP), artesunate-mefloquine (ASMQ) and a 4-day artemether-lumefantrine regimen (AL+). The primary endpoint for P. falciparum infections was the PCR-corrected cure rate and for P. vivax infections was recurrent parasitaemia, before delivery or day 63, whichever was longer, assessed by Kaplan-Meier estimate. Results Between February 2010 and August 2016, 511 pregnant women with malaria (353 P. vivax, 142 P. falciparum, 15 co-infections, 1 Plasmodium malariae) were randomized to either DP (n=170), ASMQ (n=169) or AL+ (n=172) treatments. Successful malaria elimination efforts in the region resulted in premature termination of the trial. The majority of women had recurrent malaria (mainly P. vivax relapses, which are not prevented by these treatments). Recurrence-free proportions (95% confidence interval [95% CI]) for vivax malaria were 20.6% (5.1–43.4) for DP (n=125), 46.0% (30.9–60.0) for ASMQ (n=117) and 28.7% (10.0–50.8) for AL+ (n=126). DP and ASMQ provided longer recurrence-free intervals. PCR-corrected cure rates (95% CI) for falciparum malaria were 93.7% (81.6–97.9) for DP (n=49), 79.6% (66.1–88.1) for AMSQ (n=55) and 87.5% (74.3–94.2) for AL+ (n=50). Overall 65% (85/130) of P. falciparum infections had Pfkelch13 propeller mutations which increased over time and recrudescence occurred almost exclusively in them; risk ratio 9.42 (95% CI 1.30–68.29). Among the women with falciparum malaria, 24.0% (95% CI 16.8–33.6) had P. vivax parasitaemia within 4 months. Nausea, vomiting, dizziness and sleep disturbance were more frequent with ASMQ. Miscarriage, small-for-gestational-age and preterm birth did not differ significantly among the treatment groups, including first trimester exposures (n=46). Conclusions DP was well tolerated and safe, and was the only drug providing satisfactory efficacy for P. falciparum-infected pregnant woman in this area of widespread artemisinin resistance. Vivax malaria recurrences are very common and warrant chloroquine prophylaxis after antimalarial treatment in this area. Trial registration ClinicalTrials.gov identifier NCT01054248, registered on 22 January 2010. Supplementary Information The online version contains supplementary material available at 10.1186/s12916-021-02002-8.


Congenital and placental malaria
Malaria smears of maternal peripheral blood, cord blood, placental blood, and/or the baby's peripheral blood were assessed in 76·7% (359/468) of women who had been followed until delivery. Of these 95·3% (342/359) were assessed for at least maternal peripheral blood and baby's peripheral blood, and 89·4% (321/359) for all four specimens. Rarely, the maternal peripheral smear was positive for P. vivax (4·0%, 14/353) or P. falciparum trophozoites (0·9%, 3/353). P. vivax trophozoites were found in the cord blood in one case (0·3%, 1/329) and the peripheral blood at delivery in another (0·3%, 1/348). In both cases, maternal peripheral blood was negative. Placental blood was positive for P. falciparum trophozoites in one case (0·3%, 1/333) and P. vivax trophozoites in three cases (0·9%, 3/333) and the maternal peripheral smear was positive for the same species in all of them. All three women with vivax placental malaria had three or four episodes of vivax in pregnancy. The woman with falciparum placental malaria had one vivax (27 days before delivery treated with ASMQ) and no falciparum malaria in her 85 days of follow-up during pregnancy.

Twins
There were three twin pregnancies (one in DP and two in AL + ) and among the twins, one of the six newborns were stillborn (in AL+). All twin births are excluded from description of other birth outcomes.

Neonatal death
There were nine neonatal deaths: three AL + , four DP and two ASMQ. Six were related to prematurity (including one due to placental abruption), one resulted from intrapartum asphyxia due to tight nuchal umbilical cord with severe intrauterine growth restriction (2·5 th centile), one had a congenital diaphragmatic hernia, and one had meconium aspiration syndrome (10 th centile) (S111Table).

Preterm birth (PTB)
The overall proportion of preterm birth was 8·3% (35/423) and was not different among different treatment groups (p=0·45). Treatment was not associated with estimated gestational age at birth by univariable and multivariable linear regression (S12 Table).
Small for gestational age (SGA) Birthweight of 397 (88%) normal live singletons was measured within 72 hours; 26·2% (104) were small for gestational age and proportions were not different among the treatment groups.