Is a Plasmodium lactate dehydrogenase (pLDH) enzyme-linked immunosorbent (ELISA)-based assay a valid tool for detecting risky malaria blood donations in Africa?

Benin is characterized by tropical rainforest vegetation, Sudanese climate type with an annual rainfall of about 1,600 mm, and an atmospheric temperature of 32°C. There are four distinct seasons: two rainy and two dry. This descriptive, transversal study was conducted in Benin over ten consecutive months in six departments, which grouped into three pairs of Departmental Blood Transfusion Centres (DBTC): Atlantique-Littoral,Oueme-Plateau, and Mono-Couffo. They serve as referral centres for south Benin blood transfusion services. This study was approved by the Direction of Benin National Blood Transfusion Agency and the Research Ethics Committee of the Republic of Benin. Approval was granted under the following conditions: donor anonymity must be maintained, good laboratory practice quality control must be ensured, and every finding must be treated with utmost confidentiality and used only for this research purpose. Biological analyses of the samples were performed in the biochemistry and molecular biology laboratory in Cotonou (Benin) and the Institut de Parasitologie et de Pathologie Tropicale de Strasbourg (IPPTS) (France).

Individuals considered healthy who visited the three DBTCs for blood donation between May 2009 and March 2010 were considered eligible for the study. A total of 2,515 donors and regular voluntary blood donors were enrolled over a period of ten months, which was divided into a long rainy season (LRS) from May to July, a short dry season (SDS) from August to September, a short rainy season (SRS) from October to November, and a long dry season (LDS) from December to March. Donors were selected according to medical screening criteria: no fever, no clinical evidence of progressive disease, weight above 50 kg, age between 18 and 60 years old, or 65 years old for regular donors. Last blood donation had to have occurred within the previous four months for women and three months for men. A more systematic questionnaire was introduced providing the full name, age, sex, address, and telephone of each individual. Malaria-related questions were also included: date and confirmation via clinical manifestation of malaria episodes that occurred during the year, subsequent treatment and prevention measures taken against malaria. Each study participant signed an informed consent. Overall, 7 mL of blood were collected in EDTA tubes, with 1 mL transferred into a microtube and the remaining 6 mL centrifuged at 4°C, plasma being collected into a new microtube. The tubes were stored at −20°C and transported to the IPPTS laboratory in solid carbon dioxide for antibody and pLDH antigen detection.

A thick and a thin smear were performed on slides for each Benin blood donor. The thick film was dehaemoglobinized using distilled water at pH 7.2, and then thick and thin blood films were stained using a May-Grünwald-Giemsa (MGG) stain. The microscopic reading and parasite density calculation were performed by two laboratory technicians, and in the case of uncertainty, also by a third one. Parasite count was made on the basis of 500 parasites per 1,000 leukocytes. The reading was stopped after 500 parasites were counted, even though the 1,000 figures of leukocytes were not reached. Parasite density (PD) was expressed as the number of parasites per μL of blood.

pLDH is a glycolytic pathway enzyme secreted by the different Plasmodium species, but it possesses species-specific isomers [24]. The pLDH enzyme disappears within 24 hours of effective malaria treatment [25]. Therefore, the pLDH antigen is considered a specific marker for the presence of viable Plasmodium in blood, and is used for screening in malaria-endemic countries. The pLDH antigen detection was performed by a sandwich enzyme-linked immunosorbent assay (ELISA), notably an ELISA-malaria antigen test (apDianv, Belgium) that detects pLDH via immunocapture. The apDia Antigen ELISA is an in vitro diagnostic immunoassay (IVD) for the qualitative determination of Plasmodium spp. LDH in blood samples. The apDia Malaria antigen test can be used in order to detect the malaria pLDH antigen of any of the four species in blood samples. The test was performed according to manufacturer recommendations: pour 100 μL of ready-to-use lysing buffer into each well; add 50 μL of reconstituted positive control to one well and 50 μL of negative control to triplicate well; add 50 μL of homogenized fresh whole blood sample into corresponding well; incubate for 60 min at 37°C under continuous gentle shaking conditions; drain the wells via aspiration, and fill them completely with 350 μL of washing solution; allow the wells to soak for 1 min before washing again five times; pour 100 μL of conjugate 1 solution into each well, and incubate the plate for 30 min at 37°C; wash the wells five times and pour 100 μL of conjugate 2 into each wells. Incubate the plate for 15 min then wash the wells five times and pour 100 μL of chromogenic solution into each well; incubate the plate for 15 min at 37°C; add 50 μL of stopping solution to all wells and read absorbance of each well at 450 nm with reference wavelength of 620 nm within 15 min. Test results were interpreted as follows: the optical densities of positive control (ODpos) must be >0.500 and the average OD of negative control (ODneg) <0.100. The ODneg was used to calculate the cut-off by multiplying its average value by three. The antigen index (Ag Index) of each sample was calculated by dividing the OD value of the sample by the cut-off value. A sample was considered positive if the Ag Index was ≥1.0, indicating that the sample contained viable parasites. A sample was considered negative if the Ag Index was ≤0.8, indicating that there were no viable parasites in the blood, or that there was no Plasmodium multiplication because of anti-malarial drug intake. A sample was considered inconclusive if the Ag Index was between 0.8 and 1.

The 3D7 strain of P. falciparum was maintained in vitro using an adapted candle jar method for continuous culture [26]. Parasites were cultured in normal O-positive red blood cells (RBCs) from healthy donors (EFS Alsace) and malaria culture medium (MCM, pH 7.4) composed of Roswell Park Memorial Institute medium (RPMI) 1640 supplemented with 2 mM L-glutamine, 10 mM Hepes (Gibco, Invitrogen, Cergy Pontoise, France), 1 μg/ml hypoxanthine, 0.11 mg/ml Na pyruvate, 0.02 mg/ml gentamycin, and 10% (V/V) alpha calf serum (Perbio Science, Brebières, France). The culture was diluted at a parasitaemia and haematocrit of 1%, and the medium was changed every 48 hours.

Culture with 65% ring-stage parasites and 35% schizonts was used. Infected RBCs from the continuous culture at 1% parasitaemia determined on thin blood film by microscopy were serially diluted in uninfected RBCs by incubation with low agitation at 37°C. Dilution ranged from 20,000 infected RBCs (iRBCs)/μl to 0.1 iRBCs/μl. For microscopy examination, thin blood films were made from each dilution after careful suspension, while thick blood films were made for the lower parasitaemia. Three slides were prepared for each dilution and read by two microscopists in a blind manner. Parasitaemia was determined by counting fields with 100 RBCs, and the total level was expressed in μL of blood corresponding of 5.106 RBCs/μL. A linear regression curve based on mean parasitaemia levels was established in order to determine the parasitaemia, and a standard curve was generated using recombinant pLDH to assess the correlation between parasitaemia and pLDH concentration.

A synthetic gene (Genscript, NJ, USA) encoding for P. falciparum pLDH protein was inserted in the expression vector pMAL C2X (New England Biolabs, Ipswich, MA, USA) containing the maltose binding protein (MBP) fusion partner sequence. The protein was expressed in Escherichia coli and purified by affinity chromatography on amylose resin. The protein was used to generate the standard curve.

Antibody screening was performed using a homemade ELISA-antibody test. The serological ELISA test was derived from a commercial assay exhibiting high specificity and analytical sensitivity, namely 96.7 and 93.1%, respectively [27]. The test was based on binding of anti-Plasmodium antibodies present in serum or plasma samples to antigens immobilized on 96-well plates, with the antigen being an extract of in vitro P. falciparum culture (3D7 strain). Due to antigenic community, antibodies to other Plasmodium species may be detected. The diluent solution containing PBS and Tween 0.1%, (125 μL) was poured into each well, followed by 25 μL of test plasma. On the same plate, 25 μL positive control and negative control were poured in single and triplicate wells, respectively. The plate incubated for 60 min at 37°C before being washed five times. One hundred μL of horseradish peroxidase-conjugated rabbit anti-human IgG polyclonal antibodies (Sigma-Aldrich, St Quentin, France) were added to each well, and the plate incubated for a further 30 min at 37°C. The wells were washed five times, and 100 μL of TMB plus substrate solution (tetramethylbenzidine) (Kem-en-tec, Denmark) were added to each well. The plate was covered and incubated in the dark for 15 min at 37°C. Finally, 50 μL of 0.5 M sulphuric acid was added to each well, and absorbance was read within 15 min at 450 nm, with a reference wave-length of 620 nm. Test validation required that the positive OD be >0.500 and the negative OD≥0.200. The cut-off value was calculated by multiplying the negative control wells’ average OD by four. The antibody (Ab) index of each sample was calculated by dividing its OD value by the cut-off value. The sample was considered positive if the Ab index was >1.0, equivocal if the Ab index was between 0.8 and 1.0, and negative if the Ab index was ≤0.8. The test was not able to distinguish between antibodies directed to either P. falciparum, P. vivax, P. malariae, or P. ovale.

The data was analysed by the chi-squared, Bartlett's chi-squared, and Kruskal-Wallis tests using Sigma Plot 2001 software (Sigma Plot 2001 Statistical Analysis Software, San Jose, CA, USA) and Epi Info version 3.5.1. The results were expressed as mean ± standard deviation. Statistical significance was reached if P <0.05, with a 95% confidence interval (CI).

Among the 2,515 blood donors, the rate of asymptomatic Plasmodium carriage measured by microscopy was 295/2,515 (11.72%, 95% CI: 10.5-13.1%). Three Plasmodium species were detected: P. falciparum was identified in all infection cases, P. malariae in 14 mixed infection cases (5.1%), and P. ovale in one mixed infection case (0.34%) (Table 4). The study findings revealed a higher percentage of infection among males 252/295 (85.42%) compared to females 43/295 (14.6%). Microscopic prevalence varied according to seasons: P. falciparum and P. malariae detection was constant during all seasons (Table 4), whereas P. falciparum density was very low and highly variable. Among the 295 positive donors, 238 (80.67%) exhibited a parasite density between seven and 100 parasites per microliter (p/μL), with 0.3% exhibiting a density >5,000 p/μL (Table 5).

Plasmodium antibody detection by ELISA in 2,515 donors showed a high positivity rate with a prevalence of 73.9% (1,859/2,515) (95% CI: 72.1-75.6%). Equivocal results were found in 13.1% of samples (329/2,515), and there were no detectable antibodies in 327 donors (13.0%, 95% CI: 11.7-14.4%) (Table 6B). Antibody prevalence differed significantly according to the season (P <0.05) (Table 7 and Figure 2).

pLDH antigen was found in 38.4% of blood donors (966/2,515) (95% CI: 36.5-40.3%), with equivocal results (n=622/2,515, 24.7%) considered as positive for the calculation. The prevalence of pLDH antigen confirmed by microscopy was 230/966 (23.8%). The only sample that was positive by microscopy but negative by pLDH detection was a P. malariae infection case, notably a malaria-antibody positive sample (Table 6A).

pLDH antigen prevalence varied significantly according to the season (P < 0.05) (Table 7 and Figure 2). Interestingly enough, 25.1% of the blood donors with equivocal pLDH index were microscopically negative. However, a similar percentage was observed in microscopically positive blood donors, suggesting that the grey zone was correctly set up for pLDH (Table 8). The absence of antibodies in the blood of patients with detectable parasitaemia was noted in 30 cases (30/295; 10.2%), but 20 of them tested positive when using the pLDH ELISA. In 744/2,515 (29.6%) cases, both antibody and antigen were present in samples, and 199 samples were positive (n=112) or equivocal (n=87) for pLDH and negative for antibodies (Table 9).