Effect of sulfadoxine-pyrimethamine doses for prevention of malaria during pregnancy in hypoendemic area in Tanzania

This was a cross-sectional study conducted between April and November, 2018 and involved 1161 pregnant women with at least 18 years of age, admitted in the delivery unit at Mwananyamala regional referral hospital in Kinondoni Municipality, in Tanzania. Pregnant women in delivery wards were recruited consecutively. Those who had complicated pregnancy with high risk of haemorrhage, preeclampsia, eclampsia, HIV positive, with incomplete ANC cards (not included either IPTp-SP use, mebendazole use, FEFO use or gravidity, delivered twins, delivered by caesarian section, resided in Dar es Salaam less than six months and used co-trimoxazole were excluded from the study. Dar es Salaam region is considered a low malaria transmission area with malaria prevalence of 1.1% among children below 5 years of age [3]. As per the WHO, low malaria transmission is when the prevalence of malaria in below 10% among children aged 2–9 years [5]. The region generally experiences tropical climatic conditions, typified by hot and humid weather throughout much of the year with an average temperature of 29 °C. Annual rainfall is approximately 1100 mm (lowest 800 mm and highest 1,300 mm), and in a normal year there are two rainy seasons: the long rains from March/April to May and the short rains from October to November/December [20]. Therefore, the study was conducted to cover both rain and dry season which could have influenced malaria transmission.

Validated case report form (CRF) was used to collect information on socio-demographic characteristics such as place of residence, age, gender, marital status and education level; obstetric characteristics; the use of anaemia preventive measures example FEFO (Ferrous, 60 mg/folic acid, 400 µg) and mebendazole, 500 mg; malaria preventive measures including IPTp-SP, mosquito repellents or spray and ITN; the number of IPTp-SP doses used as documented on ANC card; history on the previous malaria infection and anti-malarial used; and, birth outcomes. Gestation age was determined by last normal menstrual period and documented.

Three EDTA tubes were used to collect maternal blood, cord blood (fetal blood) and placental blood. The EDTA tubes were inverted ten times (to enable mixing of whole blood with EDTA) then transferred to the laboratory at Mwananyamala Regional Referral Hospital for malaria and haemoglobin examination. Maternal and fetal Hb levels were determined using HemoCue^® Hb 201 + HemoCue AB, Angelholm, Sweden. The WHO cut off points on haemoglobin were used to characterize maternal and fetal haemoglobin status as normal or anaemia [21]. Pregnant women and fetus were considered anaemic at Hb < 11.0 g/dL and Hb < 12.5 g/dL, respectively [21, 22].

Laboratory analysis of blood samples Maternal and fetal Hb levels were determined using HemoCue^® Hb 201 + HemoCue AB, Angelholm, Sweden. About 5 μL of blood was used for testing malaria infection using malaria Rapid diagnostic test, (RDT), SD BIOLINE Malaria Ag P.f/pan, Standard Diagnostics, INC. Microscopy was used to confirm the results of RDT. Thick and thin blood smears were stained with 2% Giemsa. Microscopic examination was performed by two competent laboratory technologists, any discordant in results was resolved by the third reader. A blood smear was considered negative when the examination of 100 high power fields did not reveal asexual parasites.

After delivery, placenta was sliced using a disposable lancet halfway between the edge of placenta and insertion of the cord. The placental biopsy with approximately 2 cm³ was placed in a container with 10% neutral buffered formalin and stored at room temperature for transfer and processing at Muhimbili University of Health and Allied Sciences (MUHAS) pathology laboratory. Paraffin-embedded placental specimens were sectioned, stained with haematoxylin and eosin (H&E) and Giemsa. Malaria parasites and pigments were examined under light microscopy. Malaria parasites were identified by their presence in the erythrocytes in intervillous space. On the other hand, malaria pigments were identified by their presence in the erythrocytes and monocytes in intervillous space and pigment in fibrin [23]. Two experienced laboratory scientists assessed placental malaria. The third laboratory scientist was involved to read the slides that had discordant readings between the two readers. Placental malaria was recorded as infected RBCs, haemozoin and infected RBCs with haemozoin pigment.

Two drops of placental blood were spotted on a filter paper (Whatman^®3MM, Maidstone, UK), air dried overnight and preserved in plastic bags for PCR analysis. After histological analysis of placental biopsy for placental malaria, 286 DBS samples with negative placental malaria were randomly selected to include participants who used ≤ 2 and ≥ 3 IPTp-SP doses. The 286 DBS samples were used to determine submicroscopic malaria infection by PCR. The DBS were transported to the laboratory at the National Institute for Medical Research, Tanga Centre for detection of submicroscopic malaria infection. The DNA was extracted from DBS using the QIAamp DNA Minikit (Qiagen, Hilden, Germany) following the manufacturer’s protocol. DNA was eluted in 150 μl of buffer and samples were stored at − 20 °C freezer until time of use. The Plasmodium species were identified using nested PCR according to the Snounou et al. protocol [24].

Aliquots of maternal venous blood were collected and centrifuged at 2000 g for 10 min [25]. Plasma was stored in 2ml cryotubes and stored at − 20 ℃ at MUHAS, Sida-bioanalytical laboratory for analysis. High Performance Liquid Chromatography with Ultraviolet detection (HPLC-UV) technique was used as described by Virendra et al. [25] with minor modifications. The extraction process used a mixture of diethyl ether and ethyl acetate at the ratio of 2:1 and a reversed-phase column (ReproSil-Pur Basic C18, 5 μ, 250 × 4.6 mm) was used. Sulfamethoxazole was used as internal standard, and the low limit of quantification (LLoQ) and low limit of detection (LLoD) was 3 μg/ml and 0.5 μg/ml, respectively. Plasma samples were analysed together with quality controls and calibration standards. Detection of SDX was used as a proxy for the use of SP for IPTp. SDX can be detected in plasma up to 63 days from the last dose. On the other hand, pyrimethamine has short half-life of 3 to 5 days, and can be detected up to 42 days after oral intake [6, 26]. Studies have reported the median concentration of SDX among pregnant women 7 days and 42 days after the last dose of SP to be approximately 75 μg/ml and 2 μg/ml, respectively [25, 26].

The primary outcome of the study was placental malaria and the secondary outcomes were adverse birth outcomes including maternal anaemia, fetal anaemia and premature delivery. Plasma SDX concentrations were log-transformed and the geometric means with standard deviation were presented. Chi square and Fischer exact tests were used to compare grouped data (such as gestation age, anaemia status, age groups, gravidity, marital status and IPTp-SP dose groups). Spearman’s correlation was used to establish the relationship between continuous variables such as maternal hemoglobin concentration and SDX concentration. Variables that had p-value ≤ 0.2 were subjected to multivariate analysis to establish their influence on the relationship between IPTp-SP doses and outcomes of interest. Multivariable logistic regression was performed to determine the effect of IPTp-SP doses on placental malaria, maternal anaemia, fetal anaemia and preterm delivery. Significance level was set at 0.05 and the confidence level at 95%. Exposure variables that had p-value < 0.05 were considered significant predictor of the outcome variables. The IPTp-SP doses were grouped as sub-optimal and optimal doses when participants used ≤ 2 doses and ≥ 3 doses, respectively. Analysis was conducted using a Statistical Package for Social Sciences (SPSS) program version 23.0.

The total of 1161 pregnant women participated in the study with median (IQR) age of 25 (18–44) years old. Most of them were married (73.7%) with primary education level (60.9%) and attended ANC visits at least 4 times (66.9%) during their recent pregnancies. They reported to have used various malaria preventive measures, such as mosquito sprays/repellents (57.9%) and ITN (98.1%), and anaemia preventive strategies including mebendazole (96.6%) and FEFO (97.9%) (Table 1). Twenty-eight (2.4%) women reported to have contracted malaria at least once during their recent pregnancies; 19 used artemether-lumefantrine, 3 artesunate injection, 1 quinine tablets, 1 dihydroartemisinin-piperaquine (DP) and 4 could not recall the names of the anti-malarial drugs they used. Fifteen (1.3%) pregnant women had fever with no confirmed no malaria.

At delivery, half of the women gave birth to male or female babies at mean (± SD) gestation age of 38.8 (± 1.6) weeks with birth weight of 3.1 (± 0.46) Kg. The proportions of primigravida, secundigravida and multigravida among the study participants were 38.4%, 28.3% and 33.3%, respectively (Table 1).

The total of 397 (34.2%) women used ≤ 2 doses while 764 (65.8%) used ≥ 3 doses of IPTp-SP (Fig. 1). Most of the participants’ characteristics such as participants age, gravidity and education level were not associated with the uptake of IPTp-SP (p > 0.05). On univariate analysis the uptake of IPTp-SP was associated with marital status (p = 0.02) and ANC attendance (p < 0.01). After adjusting for covariates, the use of ≥ 3 doses of IPTp-SP was significantly associated with increased attendance to ANCs (p = < 0.01). Those who attended ≥ 4 ANC visits had six times higher odds of taking ≥ 3 doses of IPTp-SP than those who attended ≤ 3 ANC visits (aOR, 5.91; 95% CI 4.49–7.76; p = < 0.01) (Table 2).

Placental malaria by histology was detected in 42 (3.62%) out of the 1161 study participants. Three (7.1%) out of the 42 pregnant women with peripheral malaria had placental malaria (p = 0.002), accounting for 21-fold risk of placental malaria than those without peripheral parasitaemia (aOR, 21.37; 95% CI 4.47–102.08; p < 0.001). The prevalence of placenta malaria with active infection, active-chronic infection and past infection were 2.5%, 0.4% and 0.5% respectively. Out of 397 pregnant women who used ≤ 2 doses of IPTp-SP, 16 (4.0%) of them had placental malaria. For 764 pregnant women who used ≥ 3 doses of IPTp-SP, 26 (3.4%) of them had placental malaria. Sub-optimal doses (≤ 2 doses) of IPTp-SP did not increase the risk of placental malaria among pregnant women (p = 0.97) (Table 3).

Out of 1119 placental blood samples that were microscopically confirmed to be negative for malaria parasites, 4/286(1.4%) had submicroscopic placental malaria. Of these participants with submicroscopic placental malaria, 3 (1.58%) of them used ≥ 3 IPTp-SP doses while 1 (1.04%) used ≤ 2 IPTp-SP doses. PCR positive samples revealed that, all the submicroscopic infections were due to P. falciparum (Table 2).

Out of 1161 pregnant women, the prevalence of maternal anaemia and fetal anaemia was 43.8% and 10.1%, respectively. Peripheral malaria was significantly associated with maternal anaemia and not fetal anaemia. Women with peripheral malaria had six times risk of maternal anaemia than those who had no malaria infection (aOR, 5.83; 95% CI 1.10–30.92; p = 0.04). In multivariable analysis, the use of ≤ 2 IPTp-SP doses increased the risk of maternal anaemia by 1.36 times higher compared to the use of ≥ 3doses (aOR, 1.36; 95% CI 1.04–1.79; p = 0.02), (Table 4).

The use of sub-optimal IPTp-SP doses did not increase the risk of fetal anaemia (cOR, 0.84; 95% CI, 0.56–1.27; p = 0.41). Further analysis revealed that, pregnant women who had anaemia were 2 times at increased risk of delivering anaemic babies (aOR, 1.9; 95% CI 1.31–2.87; p = < 0.01). Other characteristics of pregnant women such as age, marital status, education level, number of ANC visits, use of FEFO, mebendazole, ITN, mosquito spray/repellents, gestation age, gravidity and placental malaria were not associated with maternal and fetal anaemia.

Also, the risk of preterm delivery was not increased by the use of sub-optimal IPTp-SP doses, however, factors such as primigravidity and mosquito spray/repellents were associated with preterm delivery as reported previously [27]. Women who did not use ITN had increased risk of preterm delivery three-times higher than those who used ITN (aOR,3.39; 95% CI 1.078–10.67; p = 0.04). In addition, women who had < 4 ANC visits had two-fold risk of preterm delivery compared to those who attended ≥ 4 ANC visits (aOR, 2.05; 95% CI 1.26–3.33; p = 0.004).

SDX was detected in 218 (92.0%) out of 237 participants. A majority (60.1%) of those who had detectable SDX used ≥ 3 doses of IPTp-SP. In the analysed samples (n = 237), 79 (33.3%) samples had detectable SDX levels but could not be quantified (plasma concentration was < 3 μg/ml). Three (1.3%) participants had detectable and quantified SDX concentration despite of reporting to have not taken any dose of IPTp-SP throughout their recent pregnancies. Out of 19 (8.0%) women who had undetectable SDX in plasma, 13 (68.4%) used ≤ 2 doses of IPTp-SP. Six (2.5%) participants who reported to have used ≥ 3 doses of IPTp-SP had no detectable SDX in plasma. The overall geometric mean plasma SDX concentration was 10.76 ± 2.51 μg/mL. The geometric mean concentration for women who used ≥ 3 IPTp-SP doses was 10.46 ± 2.50 μg/mL, while those who used ≤ 2 doses was 11.24 ± 2.54 μg/mL. There was no difference in the geometric mean of plasma concentration between women who used sub-optimal versus optimal IPTp-SP doses (p = 0.65) (Fig. 1a).

There was no statistical association between prevalence of peripheral malaria and low SDX concentration (p = 0.37). In addition, the differences in SDX plasma concentration at birth had no effect on placental malaria among the study participants (p = 0.24) (Table 5).

There was a slight increase in maternal Hb concentration with plasma concentration of SDX at delivery, but this relationship was weak (spearman’s correlation = 0.02) (Fig. 2b). The geometric mean of plasma SDX between anaemic and non-anaemic women was not statistically different (p = 0.49). Similarly, the differences of geometric SDX concentration at birth was not associated with low birth weight, fetal anaemia and preterm delivery (p > 0.05), (Table 5).