Cytokine response during non-cerebral and cerebral malaria: evidence of a failure to control inflammation as a cause of death in African adults

Abstract

Background. With 214 million cases and 438,000 deaths in 2015, malaria remains one of the deadliest infectious diseases in tropical countries. Several species of the protozoan Plasmodium cause malaria. However, almost all the fatalities are due to Plasmodium falciparum, a species responsible for the severest cases including cerebral malaria. Immune response to Plasmodium falciparum infection is mediated by the production of pro-inflammatory cytokines, chemokines and growth factors whose actions are crucial for the control of the parasites. Following this response, the induction of anti-inflammatory immune mediators downregulates the inflammation thus preventing its adverse effects such as damages to various organs and death. Methods. We performed a retrospective, nonprobability sampling study using clinical data and sera samples from patients, mainly adults, suffering of non-cerebral or cerebral malaria in Dakar, Sénégal. Healthy individuals residing in the same area were included as controls. We measured the serum levels of 29 biomarkers including growth factors, chemokines, inflammatory and anti-inflammatory cytokines. Results. We found an induction of both pro- and anti-inflammatory immune mediators during malaria. The levels of pro-inflammatory biomarkers were higher in the cerebral malaria than in the non-cerebral malaria patients. In contrast, the concentrations of anti-inflammatory cytokines were comparable in these two groups or lower in CM patients. Additionally, four pro-inflammatory biomarkers were significantly increased in the deceased of cerebral malaria compared to the survivors. Regarding organ damage, kidney failure was significantly associated with death in adults suffering of cerebral malaria. Conclusions. Our results suggest that a poorly controlled inflammatory response determines a bad outcome in African adults suffering of cerebral malaria.

Keywords: Cerebral; Cytokine; Inflammation; Malaria; Plasmodium falciparum.

Figure 1. Serum levels of immune mediators during malaria.

The levels of 29 biomarkers were measured in control subjects (CT) and in non-cerebral (NCM) and cerebral (CM) malaria patients. The median value of each cytokine in the global population (CT + NCM + CM) was used as a cut-off value to determine the percentage of “high” (above median) biomarker producer individuals in each group. The ascendant biomarker profile of the CT (A) was determined and the resulting curve used as a reference to visualize the difference in the proportion of high biomarker producers with the NCM (B) and CM (C) groups. Hatched bars represent biomarkers for which there is a significant difference in Mann–Whitney U pairwise comparison with the CT reference group after Benjamini–Hochberg multiple test adjustment.

Figure 2. Serum biomarker levels in control individuals and in non-cerebral and cerebral malaria patients.

Biomarkers that significantly differed across the three groups in Kruskal–Wallis test after Benjamini–Hochberg adjustment are shown. Box plots represent medians with 25th and 75th percentiles, bars 10th and 90th percentiles, and dots outliers for biomarker concentrations. P, p[i] values in Kruskal–Wallis tests. C, critical values in Benjamini–Hochberg correction.

Figure 3. Levels of inflammatory immune mediators are higher in cerebral than in non-cerebral malaria patients.

The ascendant biomarker profile curve of the NCM (line) was plotted on the CM graph (bars) to visualize the difference in the proportion of high biomarker producers. Hatched bars represent biomarkers for which there is a significant difference in Mann–Whitney U pairwise comparison between the two groups after Benjamini–Hochberg multiple test adjustment.

Figure 4. Levels of inflammatory immune mediators are higher in deceased than in survivors of cerebral malaria.

The ascendant biomarker profile curve of the survivors (line) was plotted on the deceased graph (bars) to visualize the difference in the proportion of high biomarker producers. Hatched bars represent biomarkers for which there is a significant difference in Mann–Whitney U pairwise comparison between the two groups after Benjamini–Hochberg multiple test adjustment.