Scheduled Intermittent Screening with Rapid Diagnostic Tests and Treatment with Dihydroartemisinin-Piperaquine versus Intermittent Preventive Therapy with Sulfadoxine-Pyrimethamine for Malaria in Pregnancy in Malawi: An Open-Label Randomized Controlled Trial

Ethical approval was obtained from the Liverpool School of Tropical Medicine (LSTM) and the Malawian National Health Science Research Committee. Written informed consent was obtained from all participants prior to randomization.

This was a three-site, open-label, two-arm individually randomized superiority trial using a stratified design with one strata for primi- and secundigravidae (paucigravidae) and one for multigravidae (third pregnancy or higher). The study was conducted at the Mpemba and Madziabango Health Centers and the Chikwawa District Hospital in southern Malawi. The area has moderate to intense year-round malaria transmission and high levels of SP resistance, as evidenced by near fixation of parasites harboring mutations at codons 51, 59, and 108 of dhfr and 437 and 540 of dhps [16,17].

Women of all gravidae attending their first antenatal visit were eligible if they were HIV-seronegative, were resident in the study catchment area and willing to deliver at the study clinics/hospital, had a hemoglobin > 70 g/l, had a pregnancy between 16 and 28 wk gestation, and had not yet received IPTp-SP. Exclusion criteria included multiple gestation and other high-risk pregnancies according to national guidelines, previous enrollment in the same study, and history of allergy to any of the study drugs.

Randomization sequences were computer-generated by the study statistician at LSTM, one for each gravidity strata and study site, using variable block randomization and an allocation ratio of 1:1. In each clinic, eligible women were allocated to the IPTp-SP or ISTp-DP arm by the coordinating study staff in order of their study identification number by drawing sequentially numbered opaque envelopes containing the allocation arm from a box corresponding to each gravidity stratum. Following allocation, women and care providers were aware of the arm allocation. All laboratory staff were blinded to the treatment assignment. The study statistician remained blinded until after database lock and approval of the statistical analysis plan by the data and safety monitoring board.

At enrollment, demographic, socioeconomic, and educational information was collected, a medical and obstetric history taken, and the gestational age ascertained by ultrasound. A 5-ml venous blood sample was taken for malaria microscopy, PCR, immunology, and testing for syphilis, HIV serostatus, and hemoglobin concentration (Hemocue). All women received a long-lasting insecticide-treated net.

Participants were randomized to receive either IPTp-SP or ISTp-DP, at enrollment and all subsequent scheduled antenatal visits. The IPTp-SP arm received three tablets of SP (500 mg/25 mg sulfadoxine/pyrimethamine tablets). If they had fever or history of fever, they were tested for malaria by RDT. RDT-positive women were treated with AL and then received their first course of SP during the first scheduled follow-up visit. Women in the ISTp-DP arm were screened for malaria using the histidine-rich protein 2 (HRP2)/plasmodium lactate dehydrogenase (pLDH) combination RDT (First Response Malaria pLDH/HRP2 Combo Test, Premier Medical Corporation). All RDT-positive women in the ISTp-DP arm received a standard 3-d course of DP (Eurartesim, Sigma Tau; 40 mg/320 mg dihydroartemisinin/piperaquine tablets) at a dose of 2.5, 3, 3.5, and 4 tablets for women weighing <50, 50–59, 60–69, and ≥70 kg, respectively. All SP and DP doses were provided with a slice of dry bread as directly observed therapy. All doses in both arms were supervised. In case of vomiting within 30–60 min, the full dose was repeated. If the repeat dose was vomited, the women received AL. Sigma Tau provided the Eurartesim free of charge.

The follow-up schedule consisted of three or four scheduled antenatal visits every 4 to 6 wk: four if enrolled at 16–24 wk gestation or three if enrolled at ≥25 wk gestation. At each such visit, a clinical and obstetric examination was conducted, and a blood sample taken for RDT (ISTp-DP arm), malaria microscopy, and PCR. Hemoglobin was assessed during the last scheduled visit. Women were encouraged to make unscheduled visits if they felt ill or were concerned about their pregnancy. In the IPTp-SP arm, women with uncomplicated clinical malaria (fever/history of fever and RDT-positive) during or in between scheduled visits received AL. Women with uncomplicated malaria in the ISTp-DP arm received DP, or AL if they had received DP within the previous 4 wk.

At delivery, a maternal venous sample was taken for the same malaria metrics, and a placental and cord-blood sample for histology, RDT, microscopy, and PCR. Children were weighed and the gestational age assessed using the modified Ballard score [18]. The presence of congenital abnormalities and jaundice was assessed at delivery, at day 7, and at the final visit at 6–8 wk, coinciding with their childhood vaccination visit. In between scheduled visits, infants were followed passively.

RDT results were used to determine care. RDT positivity was defined as either pLDH or HRP2 antigen positivity. Seefor details of microscopy and real-time PCR used for detection and identification of parasites, as well as baseline parasite genotyping. S1 Text

The primary outcome among paucigravidae was “adverse live birth outcome,” defined as the composite of having a singleton baby born small for gestational age (SGA) [19] or with low birthweight (<2,500 g), or preterm (<37 wk) (S1 Text). The primary outcome among multigravidae was a composite of any evidence for plasmodium infection at delivery detected in peripheral maternal blood (microscopy, RDT, or PCR) or placenta (incision smear, impression smear, PCR, or active or past infection detected by histology) (S1 Text). The rationale for using a different primary outcome for multigravidae was based on systematic reviews showing that preventing plasmodium infection by IPTp-SP or long-lasting insecticide-treated nets is associated with improved birth outcomes primarily among women in their first and second pregnancies [20,21]. Plasmodium infection status at delivery was used as the primary outcome in multigravidae because plasmodium infection is associated with an increased risk of malaria [22–25] and anemia [26–28] in infancy, particularly in those born to multigravidae [24].

Key secondary efficacy outcomes included the individual components of the composite primary outcomes, fetal loss (spontaneous abortion at <28 wk gestation, stillbirth), any adverse birth outcome (adverse live birth outcome or fetal loss), maternal hemoglobin concentrations and anemia, clinical malaria (documented fever/history of fever plus positive malaria RDT), plasmodium infection, mean birthweight, mean gestational age at delivery, congenital plasmodium infection (cord blood positive at birth by microscopy, RDT, or PCR, or clinical malaria within 7 d of birth with parasitological confirmed diagnosis by microscopy or RDT), neonatal and infant (by 6–8 wk) clinical malaria, all-cause severe anemia and all-cause illness detected at scheduled or unscheduled postnatal visits, and perinatal and infant mortality by 6–8 wk.

The primary safety outcomes included maternal death, severe cutaneous skin reaction in the mothers within 30 d of drug intake, other serious adverse events (SAEs) in the mother or infant, congenital malformations, and neonatal jaundice.

See S1 Text for details about sample size calculations. Log binomial regression was used for binary endpoints to obtain relative risk (RR) values and corresponding 95% confidence intervals. The identity-link function was used to obtain risk differences. Linear regression was used for continuous variables, and results expressed as mean difference (95% CI). The unadjusted analysis, stratified by gravidity (pauci- and multigravidae), was considered the primary analysis. Secondary, covariate-adjusted analyses for the primary endpoints were conducted using seven prespecified covariates (in addition to gravidity and site) and simple imputation for missing covariates (<1%). These same covariates were included in subgroup analyses. Poisson regression with time of follow-up as an offset was used for count variables to obtain incidence rate ratios (95% CIs). A two-sided p-value < 0.05 was used to define statistical significance. The intention to treat (ITT) analytical population was defined as all eligible women who were randomized and contributed to the outcome. The per protocol population included women who attended every scheduled visit, who took all the daily study doses on each occasion, and who contributed to the endpoint. For the safety analysis, women in the ISTp-DP arm were considered overall and split by recipients and non-recipients of DP (i.e., those who were RDT-negative throughout). All analyses were prespecified, unless otherwise indicated, in a statistical analysis plan (see S2 Text) approved by the data and safety monitoring board. Analysis was done in SAS version 9.3 and Stata version 14.

Between 21 July 2011 and 18 March 2013, 3,214 women were screened for inclusion; 1,873 women were randomized (paucigravidae, n = 1,155; multigravidae, n = 718). Recruitment was stopped when the full sample size for paucigravidae had been reached (S1 Text). Of the randomized women, 1,743 (93.1%) were seen at delivery (Fig 1). Overall, 6,504 of 6,942 (93.7%) scheduled antenatal follow-up visits were attended (S1 Table), and 1,742 women (94.5%) attended all scheduled visits. Ultimately, 1,676 (89.5%) contributed to the primary endpoint, with proportions of participants equally distributed between the study arms (ISTp-DP arm, 89.3%; IPTp-SP arm, 89.6%) (S2 Table). The baseline characteristics were well balanced between the study arms, within each gravidity strata, and overall (Table 1). At baseline, about half of the women were infected with malaria parasites, and this proportion was slightly lower (not significant) in those not contributing to the primary analyses (S3 Table). Overall, 99.5% and 2.7% of the parasites harbored the dhps K540E and A581G mutation, respectively (S4 Table). In both arms, the median (interquartile range) follow-up time was 4.0 (3.2–4.7) mo, and the median (range) number of scheduled visits was 4 (1–4) (S1 Table). In the ISTp-DP arm, 48.8%, 38.0%, 12.4%, and 0.9% received 0, 1, 2, and 3 courses of DP, respectively (S1 Table). Overall, 3,048 and 604 courses of SP and DP were administered in the respective study arms, and in the IPTp-SP arm, 251 courses of AL were administered for clinical malaria (S1 Table).

Among paucigravidae, the prevalence of adverse live birth outcome was similar in the ISTp-DP (33.7%) and IPTp-SP (30.6%) arms (RR = 1.10 [95% CI 0.92–1.31], p = 0.282; Fig 2). The prevalence was also similar between arms among multigravidae.

Among multigravidae, the risk of malaria at delivery was higher in the ISTp-DP (34.9%) than in the IPTp-SP (27.2%) arm (RR = 1.29 [95% CI 1.02–1.63], p = 0.037). This increased risk was also evident among paucigravidae and all gravidae. In absolute terms, the risk of malaria was increased in multigravidae by 7.8% (95% CI 0.6%–14.9%) and amongst all gravidae by 7.9% (95% CI 3.1%–12.6%) (Fig 2).

Similar results for both primary outcomes were obtained from prespecified covariate-adjusted analyses, with and without prespecified imputation for missing covariates (), with per protocol population analysis (), and in a sensitivity analysis that restricted analysis to birthweight obtained within 24 h of delivery (). Results were also consistent across subgroups (andFigs), although the increased risk of malaria at delivery appeared lowest in primigravidae (). S5 Table S6 Table S7 Table S1 S2 S2 Fig

Following enrollment, 45.8% of women had ≥1 episode of plasmodium infection prior to delivery (PCR, microscopy, or RDT), and 11.4% had ≥1 episode of clinical malaria. These proportions were similar in both arms (Fig 3). At delivery, 22.2% of women had peripheral malaria detected by PCR, RDT, or microscopy. This value was higher in the ISTp-DP arm (RR = 1.34 [95% CI 1.12–1.61], p = 0.002; S3 Fig), particularly for subpatent infections (PCR-positive, RDT- or microscopy-negative; S3 Fig). The overall prevalence of placental malaria detected by histology, PCR, RDT, or microscopy was 38.0%, and this value was higher in the ISTp-DP arm (RR = 1.16 [95% CI 1.03–1.32], p = 0.018; S4 Fig), reflecting differences in acute rather than chronic or past histological infections (S4 Fig). Congenital malaria was common (12.0%) in both groups (Fig 4).

At delivery, relative to the IPTp-SP arm, paucigravidae in the ISTp-DP arm had higher mean hemoglobin concentrations (S8 Table) and a lower prevalence of anemia (hemoglobin < 110 g/l) (Fig 3). The individual components of the primary endpoint adverse live birth outcome are provided in Fig 4. Low birthweight was more common in the ISTp-DP arm (RR = 1.29 [95% CI 0.97–1.71], p = 0.079).

Overall, DP was well tolerated (S9 Table). There was no difference between arms in the number of maternal SAEs or deaths (S10 Table). There were no severe cutaneous reactions. Fetal loss was highest in the ISTp-DP arm (2.6% versus 1.3%; Fig 4). Further stratified analysis within the ISTp-DP arm showed fetal loss was highest among women who had never received DP (i.e., who remained RDT-negative throughout) (3.1% versus 2.2% in DP recipients; S10 Table). Perinatal and infant (by 6–8 wk) mortality were not statistically different between the arms (perinatal mortality: RR = 1.62 [95% CI 0.87–2.99], p = 0.127; infant mortality: RR = 1.42 [95% CI 0.63–3.17], p = 0.398) (Fig 4), but overall, the composite of fetal loss or infant death by the end of follow-up (6–8 wk) occurred more often in the ISTp-DP arm (4.0% versus 2.3%, RR = 1.76 [95% CI 1.04–2.98], p = 0.036; Fig 4; S10 Table). One case of neonatal jaundice was detected in the ISTp-DP arm (mother was a non DP-recipient), and none in the IPTp-SP arm. The frequency of congenital malformations was 1.2% in the ISTp-DP arm (0.9% in infants whose mother was a DP-recipient) and 1.0% in the IPTp-SP arm (RR = 1.11 [95% CI 0.45–2.71], p = 0.824).