Prevalence of Plasmodium falciparum infection among pregnant women at first antenatal visit in post-Ebola Monrovia, Liberia

Aim and design

This was an observational cross-sectional study that aimed to assess the burden of malaria among pregnant women attending ANC in Monrovia. Specific objectives of the study included determining the prevalence of P. falciparum infection in this particular vulnerable group through molecular assays, and assessing molecular markers of drug resistance in P. falciparum isolates.

Study site and population

The study was conducted at the Outpatients Department (OPD) of the not-for-profit Saint Joseph’s Catholic Hospital (SJCH) in Congo Town neighbourhood, Monrovia. The SJCH was founded in 1963 by the Hospitaller Order of the Brothers of St John of God. In 2014, the SJCH closed for 4 months after nine of its staff members died to Ebola. Since its reopening in 2015, the SJCH provides general services to the population of Monrovia. Although the SJCH applies a cost recovery system for the general public, the institution has a charity arm to subsidize healthcare-related costs for the most deprived ones. Its official fee for a first ANC is of 5.00 USD. However, as the SJCH is supported by the National HIV and Malaria Programmes, diagnostics and therapeutics for HIV/AIDS and malaria are provided free-of-charge.

Inclusion criteria set for sampling eligible participants at the OPD included: pregnant women of any gestational age attending ANC at the SJCH for the first time in their current pregnancy. Women unwilling to give consent to participate or who reported having received any IPTp dose during their current pregnancy were excluded. Calculations indicated that a sample of 198 women was needed to detect a prevalence of PCR-confirmed parasitaemia of 50% (conservative estimate) with a ± 5% precision and a confidence level of 95%.

Recruitment and specimen collection

After an initial period of training to the SJCH staff, recruitment and specimen collection was done between October 2016 and June 2017. All women arriving at the OPD who met the inclusion criteria were invited to participate and informed of the study objectives and specimen collection procedures.

After providing written informed consent, they were queried regarding basic socio-demographic and malaria prevention-related data (i.e. use of ITN the preceding night, or last time their house was sprayed indoor-residual insecticides). Participants were also asked to volunteer their HIV and tuberculosis status, if known. An infrared thermometer was used to measure forehead temperature. Gestational age was assessed by date of last menstrual period and by measurement of fundal height. All data were manually captured by the recruiting midwife using individual standardized case report forms.

All participants were referred to the SJCH Laboratory for specimen collection. Results of routine tests done at the SJCH laboratory as part of standard ANC care for all pregnant women (i.e. HIV, haemoglobin and syphilis) were also captured in their case report forms. Haemoglobin level was obtained using point-of-care Hemocue Hb 301 (^®Radiometer Group, Ängelholm, Sweden). HIV serostatus and syphilis screening were assessed using rapid tests SD Bioline HIV-1/2 3.0 (^®Alere Inc., Waltham, MA) and SD Bioline HIV/Syphilis Duo (^®Alere Inc., Waltham, MA), respectively.

Laboratory methods

Capillary blood from the women’s fingers was collected on two Whatman 903 filter papers that were shipped to ISGlobal Laboratory (Barcelona, Spain) for molecular testing. The presence of P. falciparum was determined by real-time Quantitative PCR (qPCR). Briefly, DNA extracted by Chelex method from the filter paper was used for detection of 18S rRNA in duplicate [40, 41]. Parasite density was quantified by extrapolation of cycle thresholds (Ct) from a standard curve of P. falciparum ring-infected erythrocytes. A negative control with no template DNA was run in all reactions.

Samples that tested positive for P. falciparum were subjected for the analysis of molecular markers of anti-malarial drug resistance using previously described methods [42–44]. DNA from samples with less than 5 parasites/μL was subjected to pre-amplification using PicoPLEX WGA Kit (^®Rubicon Genomics: Cat. No. —R30050) as per manufacturer’s instructions. DNA templates were then subjected to PCR amplification of molecular markers of k13, pfcrt, pfmdr1 and pfdhps genes using a 2720 thermal cycler (^®Applied Biosystems) followed by direct DNA Sanger sequencing bi-directionally [44]. A set of primer pair (Fw: caaattctatagtgtagttc; Rv: aattgtgtgatttgtccacaa) was used to sequence pfdhps 522–646 amino acid positions. For pfpm2 and pfmdr1 genes, copy number (CN) was also determined by SYBR-based qPCR (2× Power SYBR Green PCT Master Mix; ^®Thermo Fisher Scientific, Washington, WA, UK) [44, 45]. All samples with estimated CN > 1.5 were repeated for confirmation. Water and human genomic DNA were used as negative controls. A total of six positive controls with known k13 alleles; four parasite lines (3D7, 7G8, Dd2 and V1/S) with known pfcrt and pfmdr1 alleles; four plasmids (DHPS-V1S, DHPS-PERU, DHPS-MALI and DHPS-DD2) with known pfdhps alleles; one positive control (C4) with known copies (3–4) of pfpm2 gene; another positive control (Dd2) with known copies (3–4) of pfmdr1 were also processed, amplified and sequenced. Isolates with mixed alleles were considered as mutated for the purposes of polymorphism frequency estimation.

Data management

A unique identification number was used to link signed consent forms to case report forms. Case report forms did not include personal identifiers and were used to collect socio-demographic, malaria care-related data, and ANC routine laboratory tests data. Data from the case report forms were digitized by a SJCH trained clerk into an online case record form manager built in open source software OpenClinica (^®OpenClinica LLL., Waltham, MA) and kept safely in a locked cabinet at the SJCH. ISGlobal Laboratory data were directly captured by the laboratory technician into the OpenClinica-generated dataset. Access to OpenClinica was password protected. OpenClinica server is kept at the Hospital Clínic (Barcelona, Spain) server.

Definitions and statistical analysis

All blood specimens collected were included in the qPCR analysis. A positive P. falciparum infection in peripheral blood by qPCR, irrespective of parasite density, was the primary outcome in the analysis presented in this paper. Fever was defined as temperature ≥ 38 °C, and maternal anaemia as haemoglobin levels < 11 mg/dL (being mild if between 10 and 10.9 mg/dL, moderate if between 7 and 9.9 mg/dL, and severe if < 7 mg/dL). Women were classified as primigravidae (first pregnancy) or multigravidae (at least one previous pregnancy).

Cross tabulations and t-test, Pearson Chi squared test (χ²) or Fisher’s exact test were used, where appropriate, to explore whether socio-demographics and other health-related variables of interest were associated with P. falciparum infection. Data analysis was conducted using STATA v.12.0 (^®StataCorp LLD., College Station, Texas, USA). Significance level was set at p < 0.05.

Ethics

The study was conducted in accordance with the ICH E6 Good Clinical Practice [46], and under the provisions of the Declaration of Helsinki [47], and local rules and regulations. Written informed consent was sought for all women if 18 years of age or older. Parental consent in addition to minor assent was sought for all women aged younger than 18 years. Participants did not receive any retribution for their engagement as study subjects. Refusal to participate in this study did not affect service provision as per standard ANC practice. All women invited to enrol in this study, both consenting and non-consenting, received their first dosage of IPTp and treatment for anaemia, if necessary.

This research protocol was approved by the local University of Liberia-Pacific Institute Research and Evaluation Institutional Review Board (UL-PIRE, Monrovia, Liberia) and by the Hospital Clínic Health Research Ethics Committee (CEIC, Barcelona, Spain).

Clinical and socio-demographics characteristics of study women

Over a fourth (27.3%) of the participants were primigravidae (Table 1). Multigravidae women (n = 141) reported an average of 3.5 (range 2–10) pregnancies. Mean gestational age at first ANC visit was 17 weeks (SD 7.14).

Of the 193 women with haemoglobin test results, 29 (14.9%) and 51 (26.1%) had moderate and mild anaemia, respectively. No severe anaemia cases were detected. Six women (3.2%) had fever, and 55 (28.2%) self-reported having had taken anti-malarials in their current pregnancy.

All women but two (98.9%) reported that their home was never targeted in an indoor residual spraying program. Two-thirds of the total (71.3%) reported that they had not slept under an ITN the preceding night.

Prevalence of Plasmodium falciparum infection and association with demographic and health-related variables

Twenty-four women (12.3%) were confirmed as P. falciparum positive by qPCR (Table 2). Parasite densities tended to be lower in multigravidae than in primigravidae (Fig. 1). Low density P. falciparum infections (parasite densities below 10 parasites/µL) were more frequent among multigravidae (62% [8/13]) than among primigravidae (18% [2/11]; p = 0.047).

Compared to the uninfected women (Table 2), P. falciparum-infected women were significantly younger (24.9 vs. 27.6 years; p = 0.044) and in their first pregnancy (45.83 vs. 24.7%; p = 0.030). No associations between infection and education or occupation were found.

No statistically significant differences in terms of haemoglobin levels were found between the P. falciparum-infected (mean 10.89; range 9.2–13.7) and uninfected women (mean 11.26; range 7–16; Table 2). In total, 50.0% and 40.2% of the P. falciparum-infected and the uninfected women had anaemia, respectively (p = 0.364). Occurrence of fever was significantly more frequent among the infected women (14.3% vs. 1.8%; p = 0.019). No associations were found between P. falciparum infection and self-reported tuberculosis status, positive syphilis test, positive HIV test, and the use of ITNs (Table 2).

Molecular markers of anti-malarial resistance

The proportion of the pfdhps polymorphisms S436F, A437G, A581G and A613S described as associated with sulfadoxine resistance was 29.4% (5/17), 17.6% (3/17), 5.9% (1/17) and 11.8% (2/17), respectively. All the isolates had wild-type allele at K540E polymorphism of pfdhps gene. In the pfcrt gene, mutant alleles were found at codons M74I, N75E and K76T [88.2% (15/17)]. Two novel synonymous polymorphisms (I77I and I103I [5.9% (1/17)]) were observed in the pfcrt gene. A total of four polymorphisms were identified across the pfmdr1 gene. The prevalence of N86Y, L155L, Y184F and D1246Y mutant alleles was 47.1% (8/17), 5.9% (1/17), 58.8% (10/17) and 11.8% (2/17), respectively. The other most frequent S1034C and N1042D polymorphisms of pfmdr1 gene were not observed. The most frequent polymorphisms in k13 observed previously in Cambodian isolates [25] were absent in the P. falciparum isolates analysed. However, a novel synonymous polymorphism V510V [5.9% (1/17)] was observed. All the isolates carried a single copy of pfpm2 and pfmdr1 genes.

Limitations

An important limitation to consider is that the recruitment period took place from October to June. Hence, no participant was recruiting during the peak months of the rainy season, which may justify the relatively low prevalence observed. Another limitation is that this study was conducted in an institution that charges pregnant women for the first ANC, a factor that could have limited recruitment of certain women who may be more prone to exposure to malaria infection. The findings of this study might be extrapolated only to the group of women of low-middle class with access to employment and education.

Additionally, recall bias could have compromised participants’ self-reporting of previous use of malaria prevention and treatment. Potential effect of recall bias was minimized by intensive training on study specific procedures for data collection to the recruiting staff.