Intermittent Preventive Treatment of Malaria in Pregnancy with Mefloquine in HIV-Negative Women: A Multicentre Randomized Controlled Trial

The study protocol and informed consent forms were reviewed and approved by the Ethics Committees from the Hospital Clínic of Barcelona (Spain), the Comité Consultatif de Déontologie et d'Éthique (CCDE) from the Institut de Recherche pour le Développement (IRD, in France), and all local regulatory authorities and National Ethics Review Committees from each malaria endemic country participating in the study (). The trial was conducted under the provisions of the Declaration of Helsinki and in accordance with Good Clinical Practices guidelines set up by the WHO and by the International Conference on Harmonization. An independent Data Safety Monitoring Board (DSMB) was created prior to the beginning of the trial and regularly reviewed and monitored the safety data collected. The trial was registered prior to the enrolment of the first participant in both the ClinicalTrials.gov (NCT0081121) and in the Pan African Clinical Trials (PACTR2010020001429343) registries. Table S1

The study was conducted between 2009 and 2013 in four sub-Saharan countries: Benin (Allada, Sékou, and Attogon), Gabon (Lambaréné and Fougamou), Tanzania (Makole and Chamwino), and Mozambique (Manhiça and Maragra). The characteristics of each site are shown in. Table S2

The study was designed as an open-label, randomized, three-arm trial to compare two-dose MQ with two-dose SP for IPTp, and to compare the tolerability of two different MQ administration regimens in the context of LLITN use. The three study arms were: (1) IPTp with SP, (2) IPTp with MQ (15 mg/kg) given once as a full dose, and (3) IPTp with MQ (15 mg/kg) split over two days. The primary endpoint of the study was the prevalence of LBW babies (<2,500 g). On the basis of previous estimations in the study sites [9],[25], of a LBW prevalence of 12% in the context of IPTp-SP and LLITN use and an estimated 25% reduction to 9% in the MQ group, 1,257 women in the SP arm and 2,514 women equally split between the two MQ arms were needed to show superiority of MQ compared to SP in reducing LBW rates, at the 5% two-sided level of significance with 80% statistical power (Texts S1 and S2).

Pregnant women of all gravidities attending an ANC clinic for the first time and who had not received IPTp during their current pregnancy were invited to participate in the study after provision of informed consent. Inclusion criteria were: permanent residence in the study area, gestational age ≤28 weeks, negative HIV-testing at recruitment, absence of history of allergy to sulfa drugs or MQ, absence of history of severe renal, hepatic, psychiatric, or neurological disease, and of MQ or halofantrine treatment in the preceding 4 weeks. Gestational age was determined from fundal height measurement by bimanual palpation. Women not meeting inclusion criteria received standard ANC following national guidelines. Hemoglobin (Hb), HIV test and the syphilis rapid plasma reagin test (RPR) were assessed at the first antenatal visit as per local standard procedures. In Mozambique and Tanzania, HIV-infected women were invited to participate in a placebo-controlled trial evaluating MQ IPTp in women on daily cotrimoxazole prophylaxis [26]. The allocation of the participants to the study arms was done centrally by randomization stratified by country according to a 1∶1∶1 scheme. The sponsor's institution biostatistician produced the computer-generated randomization list for each recruiting site. Treatment allocation for each participant was concealed in opaque sealed envelopes that were opened only after recruitment. Study participants were assigned a unique study number linked to the allocated treatment group. All participants received a LLITN (PermaNet, Vestergaard Fransen) at enrolment as part of the study intervention.

Following physical examination, recruited women with gestational age ≥13 weeks received their first dose of IPTp (either SP or MQ) under supervision. Women allocated to the SP group received standard IPTp (three tablets of the fixed combination therapy containing 500 mg of sulfadoxine and 25 mg of pyrimethamine, Malastop, Sterop), whereas participants allocated to the MQ groups received 15 mg/kg of the drug (Lariam, Roche, tablets of 250 mg of MQ base). The number of tablets was calculated according to body weight, thus a woman weighing 70 kg would receive four and a quarter tablets. The maximum dosage would not exceed 1,500 mg of MQ base corresponding to six tablets. For women allocated to the MQ split dose group, the 15 mg/kg dose was divided into two halves and administered over two consecutive days with the second half dose administered either at the ANC clinic or at home (by study personnel). All study participants were observed for 60 minutes following IPT administration. Women who vomited within the first 30 minutes were provided a second full IPT dose and those vomiting 30–60 minutes after drug intake were given a half replacement dose. Home visits by field workers were done two days after IPTp administration to assess drug tolerability and correct LLITN use. The second IPTp-SP/MQ administration was given at least one month later than the first one.

Women were encouraged to attend the ANC clinic whenever they had any health complaint. Health care was free of charge and in general there was little availability of antimalarial drugs over the counter at all sites. A health facility-based passive surveillance system was established at each site to capture unscheduled visits of the study participants during the study follow-up. At each unscheduled visit, a standardized questionnaire was completed documenting signs and symptoms. Blood smears were prepared for malaria parasite examination and hemoglobin was measured if there were current or reported symptoms and/or signs suggestive of malaria. Clinical malaria episodes were treated with oral quinine or artemether-lumefantrine in the first and subsequent trimesters, respectively, for uncomplicated malaria, and with parenteral quinine for severe malaria. Solicited and unsolicited adverse events (AEs) were assessed. The former was done by directed questioning of malaria related signs and symptoms during unscheduled visits, whereas the latter were assessed through open questioning during scheduled visits. Women who were withdrawn from the study received routine ANC treatment.

At delivery, women's peripheral blood, cord blood, and placental (biopsy and impression smears) samples were collected for hematological and parasitological evaluation. Newborns were weighed (weekly calibrated scales, either digital or three beam balances), and their gestational age at birth evaluated using the Ballard's score [27]. Newborn weights not captured at birth but within the first week of life were estimated using a linear regression model (Figure S1) [28]. One month after the end of pregnancy, a capillary blood sample from the mother was collected for malaria parasite determination. LLITN use was assessed at each study visit by questions about use the preceding night.

At enrolment, HIV and syphilis serostatus were assessed at each site according to local standard procedures using rapid diagnostic tests (Table S2). Hemoglobin was determined using mobile devices in capillary blood sample (HemoCue [www.eurotrol.com↗] and Hemocontrol [www.ekfdiagnostics.com↗]). Thick and thin blood films were stained and read for Plasmodium species detection according to standard, quality-controlled procedures [29],[30]. Tissue samples were collected from the maternal side of the placenta and placed into 10% neutral buffered formalin. Biopsies were processed, stained, and examined following standard procedures [31]. Impression smears from the placental blood were stained with Giemsa and read following a standardized protocol [32],[33].

The quality of the data recorded in the study source documents and case report forms (CRFs) were monitored regularly following Good Clinical Practices principles by the trials' clinical monitor before their shipment to the centralized database in Manhiça, Mozambique. Data were double-entered using the OpenClinica Enterprise software for clinical data management (www.openclinica.com↗). The analysis was done on the modified Intention to Treat (ITT) cohort that included all recruited women who met the inclusion criteria and had data on the specific outcomes, and the According to Protocol (ATP) cohort that included all women who had received the two doses of IPTp according to the pre-specified schedule, had delivered singletons, and whose babies' weight (including stillbirths) had been recorded. The safety cohort was defined as all recruited women who had received at least one dose of IPTp and whose data for analysis were available. The analysis of safety and tolerability was made on the safety cohort. The ITT analyses were adjusted by country. The analyses in the ATP cohort were adjusted by baseline covariates (seasonality, gestational age, gravidity, anemia, literacy, and middle upper arm circumference). To include seasonality in the adjusted analysis, the duration of recruitment was divided into eight periods, and the interaction terms between the periods of recruitment and country were included in the model, which allows modelling the effect of period in each country independently. Proportion of low birth weight babies (<2,500 g at birth) were compared between groups using a modified binomial regression [34]. Only birth weights captured during the first week of life were included in the analysis. The interpretation of the statistical analysis for efficacy followed a sequential approach [35]. First, based on the confidence intervals, non-inferiority between the proportion of LBW in the MQ (combined MQ full and split dose groups) and the SP groups was evaluated assuming a 25% reduction in LBW prevalence as non-inferiority margin. If non-inferiority between MQ and SP was achieved, then a superiority interpretation comparing the groups was planned. The statistical analysis plan is available (Text S3).

Malaria infection was defined as the presence of asexual P. falciparum parasites of any density in a blood smear. A clinical malaria episode was defined as the latter plus any sign and/or symptom suggestive of malaria including: fever (axillary temperature ≥37.5°C) in the last 24 hours, and/or pallor and/or arthromyalgias and/or headache and/or history of convulsions [36]. The incidence of all clinical malaria episodes was compared between groups using a negative binomial regression allowing for interdependence between episodes within the same subject, excluding from the time at risk the 28 days after a malaria episode. Failure curves were produced using the Kaplan-Meier methodology. Placental infection was defined as the presence of parasites with or without pigment in the histological examination, or in the impression smear [31]–[33]. Anemia was defined as a Hb level <11 g/dl and severe anemia as Hb<7 g/dl. Immediate tolerability to study drugs was assessed as observed vomiting within one hour of drug administration. An AE was defined as any untoward medical occurrence in a study participant, to whom the study drug was administered, including occurrences, which are not necessarily caused or related to that drug. Serious adverse events (SAEs) were defined as an AE that met any of the following criteria: (1) results in death, (2) is life-threatening, (3) requires hospitalization (or prolongation of existing hospitalization), (4) results in disability/incapacity, (5) is a congenital anomaly, or (6) any event of special interest (including miscarriage and stillbirths of women not admitted to hospital) [37]. The proportions of women with an AE or a SAE were presented by treatment group with 95% confidence intervals and p-values were calculated by Fisher-exact text. For safety and tolerability outcomes it was considered that there was no evidence of significant difference between treatment groups if the 95% confidence intervals overlapped. Data analysis was performed using Stata statistical software version 13 (Stata Corp.).

Figures 1 and S2 show the trial profile in the ITT and ATP cohorts, respectively. Overall, 4,749 pregnant women were randomized to receive IPTp (1,578 were allocated to SP, 1,580 to MQ full dose, and 1,591 to MQ split dose). Four women were not included in the ITT cohort: two did not finalize recruitment process, one was not pregnant, and one was recruited twice and only the first enrolment was included in the analysis. The main reasons for no enrolment into the trial were no permanent residence in the study area (41%) and gestational age>28 weeks (28%). The overall refusal rate for trial participation was 18%. Baseline characteristics were similar for women in the three treatment groups (Table 1). Syphilis prevalence was between 1% and 2% and anemia prevalence between 58% and 60%. Mean gestational age was 21 (standard deviation [SD] 7) weeks at the first IPTp dose and 26 (SD 6) weeks at the second IPTp dose. Median time between first and second dose was 35 (interquartile range [IQR] 12) days, and median time between the last dose and delivery was 94 days (IQR 48).

A total of 4,176 birth weights were collected and analyzed in the ITT cohort (81.4% were captured at birth; overall, 2.7% from stillbirths, 3.7% from twins), whereas 3,435 birth weights were analyzed in the ATP cohort (83.7% captured at birth; 2.5% from stillbirths, none from twins). There were no significant differences between the MQ and SP groups in either the prevalence of LBW infants (13.0% in the overall MQ group and 12.7% in the SP group, risk ratio [RR], 1.02 [95% CI 0.86–1.22; p = 0.80]) or in mean birth weight (Tables 2 and 3). On the basis of the confidence interval, non-inferiority could be accepted between the MQ and the SP groups at the pre-specified 25% margin for reduction in LBW prevalence in the ITT analysis, but not in the ATP analysis (Table 2). No difference in the prevalence of LBW was observed either between the full and split MQ dose (RR, 1.04 [95% CI 0.84–1.28]). The results were similar in the ATP adjusted (RR, 1.03 [95% CI 0.84–1.26]) and unadjusted (RR, 1.05 [95% CI 0.85–1.29]) analyses. Figure 2 shows the birth weight distribution by country and treatment group.

The risk of maternal peripheral malaria parasitemia at delivery was 30% lower in women who received MQ compared to those who received SP (RR, 0.70 [95% CI 0.51–0.96]; p = 0.03). The risk of overall maternal anemia (Hb<11 g/dl) at delivery was also lower (44.1% in the SP group versus 40.5% in the MQ group [RR, 0.92 (95% CI 0.85–0.99)]; p = 0.03) and the mean Hb was higher in women who received IPTp-MQ compared to those in the SP group, although these differences were only significant in the ITT analysis (Table 3). The frequency of severe anemia was also lower in women receiving MQ although the difference did not reach statistical significance (Table 3).

There were no differences between groups in the prevalence of placental infection, neonatal parasitemia, neonatal anemia, or maternal peripheral parasitemia one month after delivery (Table 3). Results stratified by country can be found in Tables S3, S4, S5 and placental histology results by treatment in Table S6. The incidence of clinical malaria episodes and of all-cause outpatient attendances during pregnancy were significantly lower in women receiving IPTp-MQ compared to those receiving IPTp-SP (RR, 0.67 [95% CI 0.52–0.88]; p = 0.004 and RR, 0.86 [0.78–0.95]; p = 0.003, respectively) (Figure 3; Table 4). The incidence of all-cause hospital admissions was lower in the MQ group although not statistically significantly different (RR, 0.88 [0.68–1.14]; p = 0.35) (Table 4). The overall reported use of LLITN at delivery was of 90% and of 96% one month after the end of pregnancy, with no difference between study groups.

There was no difference in the prevalence of adverse pregnancy outcomes (including miscarriages, stillbirths, and congenital malformations) between groups (Table 5). The number of SAEs, including maternal and neonatal deaths, was also similar among the three study arms. The number of women who had SAEs considered as drug-related by the site investigator was higher in the MQ groups: one in the SP group (0.1%; a miscarriage), 11 in the MQ full-dose group (0.7%; one urinary tract infection, one generalized urticaria, one stillbirth, one premature delivery, two miscarriages, and five vomiting episodes), and ten in the MQ split-dose group (0.6%; two miscarriages, two stillbirths, three preterm delivery, one malaria, and three vomiting episodes). Serious adverse pregnancy outcomes considered as drug-related by the study investigators were carefully reviewed by the trial's independent DSMB, which concluded that a causal relationship between the drug and these SAEs could not be established. No serious neurological AEs were reported among study participants. The frequency of non-serious reported sleeping disorders was higher in the MQ group (79/3,113, 2.5%; [95% CI 2.0–3.2]) than in the SP group (12/1,561, 0.8% [95% CI 0.4–1.3]). Two women with unknown psychiatric antecedents attempted suicide in the SP group.

The immediate tolerability of IPTp was poorer in the two MQ groups as compared to the SP group, with no difference between the full and split-dose groups (Table 6). The most frequently reported related AEs following the first MQ administration were dizziness (33.9% and 35.5% in the full and split-dose groups, respectively) and vomiting (31.7% and 30.1% in the full and split dose groups, respectively) (Table 7). The majority of these AEs started within 48 hours after drug administration. The mean duration of the aforementioned AEs was one day (IQR 2) and over 70% were classified as mild (Table 8). The prevalence of dizziness and vomiting related to MQ was reduced at the second IPTp administration compared to the first one (Figure 4; Table 7).

Table 9 shows the reported AEs by system organ class and treatment group and Table 10 the list of SAEs related to the drug.

The second IPTp administration was not given to 7% (111/1,559), 11% (178/1,550), and 12% (193/1,562) of women who had received the first administration of SP, MQ full, and split-dose, respectively. In addition, in the MQ split-dose group, 7% (101/1,562) and 8% (108/1,369) of the women did not receive the second half dose at the first and second IPTp administrations, respectively. The proportion of women who had a related AE within two weeks after receiving the first IPTp administration and who did not receive the second IPTp administration was 7% (23/293), 16% (116/721), and 14% (109/769) in the SP, MQ full, and split dose groups, respectively.