Procalcitonin, Pentraxine 3, and Molecules of the Vascular Endothelium are Hallmark of Pathogenicity and Predict Coma and Mortality Among Children with Severe and Cerebral Malaria

Background: Malaria-related deaths could be prevented if prompt diagnosis and prognostic biomarkers are available to allow rapid adequate treatment. Understanding of the mechanisms implicated in the progression from P. falciparum asymptomatic and uncomplicated malaria infections to severe life-threatening disease is necessary to identify such indicators. Methods: Using quantitative ELISA, we assessed the plasma concentrations of Procalcitonin, Pentraxine 3, Ang-2, sTie-2, suPAR, sEPCR, and sICAM-1 in a cohort of 337 Beninese children who presented with pediatric malaria to investigate the potential association with clinical manifestations and outcomes. Results: The molecules showed higher levels in children with severe or cerebral malaria compared to those with uncomplicated malaria. Plasmas concentrations of PTX3, PCT and the soluble receptors suPAR, sTie-2, sICAM-1 were signicantly higher in children with deep coma as dened by a Blantyre Coma Score < 3 (P <0.001 for PTX3, suPAR, sTie-2, p=0.004 for PCT and p=0.005 for sICAM-1). Moreover, the concentrations of PTX3, suPAR and sEPCR were higher among children who died from severe malaria compared to those who survived with (p=0.037, p=0.035, and p=0.002 respectively). Conclusion: Our ndings indicate the ability of these seven bioactive molecules to discriminate among the clinical manifestations of malaria and therefore, given their potential utility as prognostic biomarkers for severe and fatal malaria, they might be useful to improve severe cases management. therefore to help improve case management. In this study, we assessed in a cohort of 337 Beninese children presenting various clinical manifestations of malaria, the level of four soluble receptors (suPAR, sEPCR, sICAM-1, and sTie-2 with its ligand protein Ang-2) and other bioactive molecules including PTX3 and the calcitonin-precursor-hormone procalcitonine (PCT).


Background
Plasmodium falciparum malaria is associated with a global annual mortality of 380, 000 with 94% deaths occurring among African children [1]. The diverse clinical presentations of malaria ranging from mild uncomplicated malaria to severe life-threatening disease re ect distinct underlying pathogenic mechanisms involved in each disease state making often a challenge for the clinicians to provide reliable diagnosis and appropriate treatments [2]. Although, the factors associated with the transition from uncomplicated to severe forms of malaria remain not fully understood, several studies showed immune dysfunction and excessive endothelial activation in acute severe cases [3,4]. Speci cally, the types of cytokines and chemokines produced by the host play a critical role in the progression of uncomplicated malaria towards cerebral malaria and other severe complications. For instance, increased plasma levels of pro-in ammatory Tumor-Necrosis-Factor-alpha (TNF-α), Interferon-gamma (IFN-γ), and Interleukin-1-beta (IL-1β) [4][5][6], as well as a decreased levels of antiin ammatory cytokines such as Interleukin 10 (IL-10) and Transforming-growth-factor-beta-1 (TGF-β1) [4,7,8], are hallmarks of severe malaria. Furthermore, several cytokines are known to take part in the endothelial dysfunction associated with leucocytes and infected red blood cells sequestration via the induction of Intercellular-Adhesion-Molecule-1 (ICAM-1) and Endothelial-protein-C-receptor (EPCR) expression on the cell surface and also by modulation of their shedding in the blood circulation resulting in an increased level of soluble EPCR and ICAM-1 (sEPCR,sICAM-1) in children admitted to hospital for severe malaria [9][10][11][12]. Macrophage/Monocytes mediators including neopterine, the Monokine-induced by IFN-γ (MIG), and other molecules such as the patternrecognition-glycoprotein-pentraxine (PTX3), and the soluble urokinase-plasminogen-activator-receptor (suPAR), were also found to differentiate between uncomplicated and severe malaria. These results however were obtained on a limited number of malaria conditions missing fatal cases in some studies [13][14][15]. In addition, angiopoietin 2 (Ang-2) and its tyrosine-kinase-receptor 2 Tie-2) were found to be increased and associated with retinopathies in Malawian children with cerebral malaria. In combination with clinical parameters, Ang-2 improved mortality prediction among these children [16]. The results obtained so far highlights the interest to further investigate the potential of these molecules as biomarkers for accurate prognostic of patients at risk of dying from severe malaria and therefore to help improve case management. In this study, we assessed in a cohort of 337 Beninese children presenting various clinical manifestations of malaria, the level of four soluble receptors (suPAR, sEPCR, sICAM-1, and sTie-2 with its ligand protein Ang-2) and other bioactive molecules including PTX3 and the calcitonin-precursor-hormone procalcitonine (PCT).

Study design and participants
In this cross-sectional study carried out from December 2017 to July 2019, data were collected from children under 6 consulting at hospitals including Centre- Children were recruited in the study if they presented a positive rapid diagnostic test for P. falciparum (DiaQuick-Malaria-P. falciparum-Cassette, Dialab; Hondastrasse, Austria) and meet the World Health Organization clinical malaria de nition criteria (WHO; 2011). All malaria cases were de ned as a microscopically con rmed P. falciparum mono infection. Cerebral-malaria-group (CM) had Blantyre Coma Score (BCS) < 3, with the exclusion of any other causes of coma. The severe-non-cerebral-malaria-group (SM) were children presenting with one or more of the following symptoms; pulmonary edema, acute respiratory distress syndrome, acute kidney failure, abnormal liver function, hemoglobinuria, or severe anemia with BCS>2. The uncomplicated-malaria-group (UM) had P. falciparum parasitemia infection with fever, without signs of life-threatening malaria and evidence of vital organ dysfunction.
Ethics approval and consent to participate: Ethical clearance was obtained from the Comité National d'Ethique pour la Recherche en Santé (CNERS), Cotonou, Benin (No 50, 25 th October 2017, IRB00006860). For all the participants a signed informed consent was obtained from parents or legal guardians.
Blood and data collection Up to 4 ml peripheral blood was collected into citrate-phosphate-dextrose-adenine containing tubes from children at admission to the hospital. Clinical, biological and demographic data of patients were captured in a questionnaire and entered in an ad-hoc le for further analysis. All patients were treated according to the guidelines of the Beninese Ministry of Health.

Quanti cation of plasma bioactive molecules
Plasma samples were assayed in duplicate blindly. Standard ELISA experiments were performed according to the manufacturer's instructions (R&D Systems ® , Minneapolis, MN) for PTX3, Ang-2 and soluble receptors. Plasma samples were diluted to ½ for Ang-2, 1/4 for sTie-2, 1/10 for PTX3, suPAR, and sICAM-1, and 1/20 for sEPCR. For PCT quanti cation, the one-step enzyme-linked uorescent immunoassay (VIDAS BRAHMS bioMérieux, Lyon, France) was performed using 200µl plasma. In all experiments, analyte concentrations were calculated according to standard curves obtained by the assessment of speci c recombinant human proteins which systematically included negative control sera. The nal results were expressed as ng/ml.

Statistical analysis
All statistical analysis was conducted with Stata software v15 and GraphPad Prism v9. Continuous variables and categorical variables were compared between the three clinical groups (CM, SM, and UM) using the Kruskal-Wallis test in conjunction with Dunn's multiple comparison post-hoc tests and chisquare test respectively. The levels of analytes were compared between the groups, UM vs SM/CM, survived vs died and deep coma vs no deep coma among severe cases using Mann-Whitney U-test and binary logistic regression. Linear regression was used to examine the relationship between analyte levels and platelets, hemoglobin, leucocytes, glycemia, parasitemia, and age. Signi cance levels was set at P-value <0.05.
The validity of each molecule as biomarker for malaria related severity, coma and fatality was assessed by Receiver Operating Characteristic (ROC) curve analysis, a two-dimensional measure of classi cation performance where the area under the ROC curve (AUROCC) accurately measures discrimination, i.e., re ects the power of a quanti ed parameter to distinguish between two clinical groups. The greater the AUROCC, the better the test is [17].

Cohort description
Blood samples were collected from 337 children aged between 4 and 66 months who arrived at the hospitals for consultation or were admitted to intensive care. Among them, 228 had severe malaria, including 95 with cerebral malaria having deep coma with BCS< 3. Thirty children also had impaired consciousness with BCS=3 accompanied by other symptoms of severe malaria, while 103 children had severe malaria without deep coma and were all ranked in the group of SM. One hundred and nine children suffered from UM. Severe anemia was very common among children and was detected in 39% of children with SM and 41% of those with CM. Thirty-eight children (11.3%) died including 26 (68.42%) among the CM group and 12 (31.57%) among the SM group, 10 of them having consciousness alteration with BCS of 3, and 2 suffered from SM without neurological impairment with BCS of 5 (Table:1). Relationship between the analytes and other blood components parasitemia and age In the linear regression analysis, increased levels sTie-2, Ang-2, suPAR and sICAM-1 were associated with reduced level of hemoglobin with a signi cant pvalue of (p<0.001), (p=0.042) (p=0.031), (p=0.004) respectively. Elevated sTie-2 level was also associated with a reduction in platelets count (p=0.025) and glycemia (p=0.0001). However, elevated levels of Ang-2 and sTie-2 were associated with increased leucocytes count with a signi cant p <0.001 and p=0.036 respectively. There was no relationship between analytes levels and parasitemia or age (Table:2). Immune and endothelial activators are associated with malaria related severity, coma and mortality.
Geometric mean comparisons between the clinical groups including UM versus SM or CM and between SM versus CM showed that the levels of all tested analytes were higher in children with SM and CM compared to those with UM, all P< 0.0001 ( Figure:1 A-G), while only suPAR, sTie-2, sICAM-1, PTX3, and PCT were increased in children with CM as de ned by deep coma (BCS <3) compared to those with BCS>2 among children with SM. Besides, the levels of sEPCR, sICAM-1, sTie-2, Ang-2 and suPAR did not differ between children with deep coma BCS<3 and those with altered consciousness with BCS=3. However, the level of sEPCR and sICAM-1 was higher in children with BCS=3 who died than those who survived with p-value of 0.035 and 0.04 respectively (Figure:2 A-G). Furthermore, the levels of suPAR, sEPCR and PTX3 were higher in children who died from SM or CM than those who survived to these pathologies ( Figure:3 A-C).
The logistical regression analyses summarizing the Odds Ratios calculations are presented in Table 3. In these analyses we de ned disease severity (SM and CM versus UM), presence versus absence of deep coma and mortality versus survival among severe cases, as outcome variables while the effect described was independent of covariates such as ethnicity, age, parasitemia and sex. Table 3 Relationship between PTX3, PCT, Ang-2, sTie-2, suPAR, sEPCR, sICAM-1 level and malaria outcomes, Odds Ratios for various outcomes when testing 7 analytes as potential predictors of severity, coma and mortality. The odds ratios show an increase of 51% for Ang-2 and 21% for sTie-2 in children with CM or SM compared to UM indicating that SM is 1.51 and 1.21 times more likely to occur with a unit increase in Ang-2 and sTie-2. While a slight increase in the odds (between 1.03 and 1.06) for suPAR, PCT, and PTX3 was noticed with a statistically signi cant difference between the two groups. Except for sICAM-1 for which the increase in the odds was low close to zero but signi cant, the odds for sTie-2, PTX3, and suPAR levels were 1.09, 1.04, 1.04 respectively times higher in children with BCS<3 compared to those with BCS>2.
In addition, there was 1.03 1.02, 1.03 times increase in the odds of suPAR, PTX3 and sEPCR respectively in fatal cases compared to survivors from CM or SM pathologies which was statistically signi cant.
Predictive accuracy of analytes to discriminate between the various clinical groups of malaria

Discussion
Our ndings showed the importance of four soluble receptors including sEPCR, sICAM-1, suPAR, and sTie-2 with its ligand Ang-2, and two other molecules PCT, and PTX3, as informative biomarkers of malaria disease severity, coma and mortality.
The plasma levels of sEPCR, sICAM-1, sTie-2 and suPAR were signi cantly higher in children with CM or SM compared to children with UM using both logistical regression analysis (Table:3) and non-parametric tests ( Figures:1,2,3). These molecules have also been found to be increased in patients with other infectious or chronical diseases such as sepsis, HIV-1-AIDS, tuberculosis, rheumatoid-arthritis and various cancers with suPAR, sEPCR and PTX3, often predicting poor clinical outcome as in our present study [12,[18][19][20]. Furthermore, sEPCR, suPAR and PTX3 were associated with both deep coma and fatality.
sEPCR, sTie-2/Ang-2 and sICAM-1 are involved in cytoprotective activity to maintain micro-vessels endothelial barrier functions. The soluble as well as the membranous form of EPCR bind with similar a nity to activated C-reactive-protein (APC) to maintain endothelium integrity through the activation of proteaseactivator-1 [21,22] promoting anticoagulation and anti-in ammation. During malaria, EPCR has been found to bind the P. falciparum erythrocyte-membraneprotein-1 to the same region as APC does thus, decreasing anticoagulant activity of APC and promoting thrombosis and obstruction of blood circulation [11,23]. Our data supports the hypothesis that increased sEPCR in severe cases contributes for maintaining APC anticoagulant activity in these patients preventing acute thrombosis.
Similarly, the level of Ang-2 and its receptor sTie-2 was higher in SM or CM groups with sTie-2 highly increased in coma cases suggesting a pathophysiological role during CM. Interestingly, Ang-2 and Ang-2/Ang-1 ratio were shown as independent predictors of metabolic acidosis, coma and mortality [16, 24,25]. The Ang1/Ang-2-Tie-2 system is a paramount regulator of endothelial integrity with Ang-1 signals through Tie-2 to maintain vascular quiescence. This is antagonized by Ang-2 resulting in endothelial dysfunction [26]. The sTie-2 contains the ligand binding domain that binds angiopoietins and inhibit their interaction with cellular Tie-2. Given the importance of sTie-2 in regulating angiopoietins availability it contributes to vascular pathology when its level is impaired [27]. Therefore, the high production of sTie-2 may thwart the antagonizing effects of Ang-2 which was highly expressed in CM and SM.
Moreover, the enhanced level of Ang-2 and sTie-2 leads to sICAM-1 expression which was signi cantly higher in children with CM or SM as well as in children with coma ( Figure:1 C, D, E) ( Figure:2 B, C).
We also found that suPAR level was signi cantly higher in children with SM or CM compared to those with UM, in children with BCS<3 compared to those with BCS >2 and in children who died from SM or CM making this molecule a potential marker of severity, coma and fatality. Produced by activated monocytes and endothelial cells, suPAR plays several roles in innate immune defence and in ammation. It acts in the recruitment of effector cells (monocytes/macrophages) and clearance of pathogens at infection sites [28]. In addition to the uPA/suPAR binding which triggers the plasminogen activation system, suPAR also binds integrins and other receptors to activate different intracellular signaling pathways implicated in cell proliferation, invasion, angiogenesis, and metastasis.
Therefore, suPAR has been described to be increased in cancers, in coronary heart diseases (CHD) and in infectious diseases including malaria [15]. An immunohistochemistry analysis of CM patients brain showed an accumulation of suPAR in macrophages/microglial cells in Durks granuloma adjacent to petechial haemorrhages, as well as in astrocytes, and in endothelial cells while the expression of suPAR was low in normal brains suggesting the association of suPAR expression with tissue damage of the blood brain barrier during CM [14,29,30]. High suPAR level was associated with disease severity and poor prognosis in cancers, CHD and infectious diseases including SARS-CoV-2 [31][32][33][34].
We found that PCT came out as the best biomarker for diagnosing malaria severity, with an AUROCC of [0.83]. This result is consistent with previous ndings [35][36][37]. This molecule able to discriminate between viral and bacterial, fungi or parasitic infections is used in guiding antibiotic treatments in patients [38].
Interestingly, PCT was higher in children with BSC<3 than those with BCS>2 (P=0.004) and showed moderate predictive value for coma with an AUROCC of [0.65] which is to our knowledge a novel nding. Of note, PCT in combination with CRP, chitinase-3-like-protein, and S100beta was recently described to be a promising biomarker in determining the presence, location, and extent of traumatic intracranial lesions [39] The level of PTX3 was higher in children with SM or CM compared to children with UM, in children with BCS<3 compared to those with BCS >2 and in children who died compared to those who survived with signi cant diagnosis performance for severity, coma and mortality as shown by the signi cance of the AUROCC analyses ( Figure:4D,5A,6B).
PTX3 is involved in the humoral innate immunity by recognizing microbial moieties and damaged tissues and in regulating in ammation and autoimmunity. PTX3 is stored in speci c granules in neutrophils and is released in response to microbial recognition and in ammatory signals. Increased levels of plasma PTX3 was associated with the clearance of apoptotic cells by dendritic cells resulting from high immune response and endothelial dysfunction in cardiovascular disorders [40,41]. PTX3 was investigated as biomarker in infectious diseases and has emerged as strong independent predictor of mortality as recently described in SARS-CoV2 syndrome [42]. However, this study is only the second one that investigated PTX3 as potential biomarker during malaria and revealed the potential of this molecule at distinguishing SM and CM from UM cases, deep comas from mild and conscious cases and fatal cases from patients who survived. This result highlights the unique interest of PTX3 as biomarker to assess the progression of the disease and response to treatment in hospitalized patients.
Nine (75%) of the fatal cases among the SM group were children who had consciousness alteration with BCS=3. In these children the levels of sEPCR, sICAM, sTie-2, Ang-2 and suPAR were similar to those with CM BCS<3. This nding suggests that SM children with (BCS=3) who died are probably children with CM misclassi ed in the SM group for compliance with the WHO de nition of CM conditioned by the BCS <3 criterion. It might also re ect an underestimation of the BCS evaluation in these children by the clinicians.
In addition, sEPCR and sICAM -1 levels were higher in children with BCS = 3 who died than in those who survived emphasizing the interest of sEPCR as marker of mortality from malaria supporting earlier ndings (Figure: 2 F, G) [12].
The enhanced levels of sTie-2, Ang-2, suPAR and sICAM-1 were signi cantly associated with reduced hemoglobin with sTie-2 also associated with low platelets count and hypoglycemia and sTie-2, Ang-2, associated with high leucocytes count. These results emphasize a deleterious effect of these analytes during severe malaria as the fall of hemoglobin, platelets count and glycemia are hallmarks of SM and CM pathologies [43].
Beyond the observed association between the elevated levels of the analytes and SM and CM manifestations, these molecules might be implicated in the processes that control blood parameters and deserve to be fully investigated.

Conclusion
Our ndings show that in Beninese children with malaria suPAR, sICAM-1, sEPCR, sTie2, Ang-2, PCT, and PTX3 differentiate children with SM and CM from children with UM, with PCT and sTie-2 providing the highest diagnosis performance for severity. sICAM-1, sTie-2, suPAR and PTX3 were higher in children with BCS<3 with PTX3 and sICAM-1 presenting the best diagnosis performance for coma. In addition, sEPCR, PTX3 and suPAR levels discriminate between children who died from SM and CM and those who survived to these pathologies with the highest predictive value for sEPCR.
Overall, the expression of these molecules during malaria, either at the onset triggering the in ammatory response like suPAR and PTX3 or resulting from an acute in ammatory response like sICAM-1 and sEPCR, contributes to pathogenesis of SM in which the site of initiation of this response and its intensity play a major role. Therefore, further researches on the mechanisms of action of these molecules are needed to better understand their role in the pathogenesis of SM and CM and explore their potential in therapeutics.

Declarations
Authors' contributions RT, NN and CR designed the study and got the funding for its accomplishment, RA, AM, contributed to the sample collection and preparation, MJA, AA and CP were responsible of patient's enrollment in the study and patient's treatment, BT, LR, PT, LD, BA, HL and SOB carried out all the lab experiments, BAM performed the statistical analysis, SP facilitated PCT assessment. BT, RT, wrote the manuscript. All the authors contributed, read corrected and approved the submitted version of the manuscript.