Controlled human malaria infections by mosquito bites induce more severe clinical symptoms than asexual blood-stage challenge infections

Abstract

Background: Fever and inflammation are a hallmark of clinical Plasmodium falciparum (Pf) malaria induced by circulating asexual parasites. Although clinical manifestations of inflammation are associated with parasite density, this relationship is influenced by a complex network of immune-modulating factors of both human and parasite origin.

Methods: In the Controlled Human Malaria infection (CHMI) model, we compared clinical inflammation in healthy malaria-naïve volunteers infected by either Pf-infected mosquito bites (MB, n=12) or intravenous administration of Pf-infected red blood cells (BS, n=12).

Findings: All volunteers developed patent parasitaemia, but both the incidence and duration of severe adverse events were significantly higher after MB infection. Similarly, clinical laboratory markers of inflammation were significantly increased in the MB-group, as well as serum pro-inflammatory cytokine concentrations including IFN-γ, IL-6, MCP1 and IL-8. Parasite load, as reflected by maximum parasite density and area under the curve, was similar, but median duration of parasitaemia until treatment was longer in the BS-group compared to the MB-group (8 days [range 8 - 8 days] versus 5·5 days [range 3·5 - 12·5 days]). The in vitro response of subsets of peripheral blood mononuclear cells showed attenuated Pf-specific IFNγ production by γδ T-cells in the BS-arm.

Interpretation: In conclusion, irrespective the parasite load, Pf-infections by MB induce stronger signs and symptoms of inflammation compared to CHMI by BS infection. The pathophysiological basis remains speculative but may relate to induced immune tolerance.

Funding: The trial was supported by PATH's Malaria Vaccine Initiative; the current analyses were supported by the AMMODO Science Award 2019 (TB).

Keywords: Controlled Human Malaria Infection; Inflammation; Malaria; Plasmodium falciparum; Tolerance.

Figure 1

Schematic study design and course of parasitaemia. Malaria naïve volunteers were challenged by either bites of 5 infected mosquitoes or by injection with ∼2,800 infected red blood cells, both with the Pf 3D7 isolate. Following mosquito bite, Pf sporozoites first invade liver cells, within which they undergo schizogony over the course of ∼7 days before releasing a first wave of blood-stage parasites into circulation, which subsequently commence cyclical asexual blood-stage multiplication (blue curve). Following inoculation of infected red blood cells, in contrast, blood-stage multiplication commences immediately (red curve). Both curves represent schematic blood-stage parasite density in peripheral venous blood on a log-scale; the apparent wave-like pattern of parasitaemia observed in peripheral blood is due to sequential sequestration of mature blood-stage schizonts in the vasculature and subsequent release of next-generation parasites into peripheral circulation in the course of each multiplication cycle. Volunteers infected by mosquito bite (MB, n=12) received treatment (T, dashed lines) when parasitaemia reached 5,000 parasites/mL, or earlier if safety criteria were met. One subject received treatment at day 10•3 due to thrombocytopenia; this subject was included in all analyses. Volunteers who received a blood stage inoculum (BS, n=12) were all treated presumptively on day 8 post-challenge. PBMC sampling time points are indicated with arrows and days relative to challenge infection (C+#).

Figure 2

Asexual parasitaemia per study group. a) Peak asexual parasite density in mosquito bite group (MB, n=12, blue circles) and blood stage-infected subjects (BS, n=12, red squares). b) Area under the curve (AUC) asexual parasite density over time. Error bars present median with interquartile range. P-values by Mann-Whitney U test.

Figure 3

Adverse events and inflammaotry parameters per study group. a) Absolute number of severe adverse events per subject. b) Median duration in days of severe adverse events per subject in days. Bullets in a+b represent individual subjects infected by mosquito bite (MB, blue circles) or blood-stage inoculation (BS, red squares); bar and whiskers represent median and IQR; p-values for difference between groups by Mann-Whitney-U. c) Body temperature (tympanic). d) C-reactive protein concentrations. e) Serum IFN-γ concentrations. f) Serum IL-6 concentrations. Figures c-f show median values for MB (blue circles) and BS (red squares), with error bars for IQR. Dotted lines represent threshold for normal values. Time points are indicated as days relative to challenge infection (C+#) or treatment (T+#). Asterisks indicate a significant difference between inoculation groups at a given time point (p<0•05, Mann-Whitney U).

Figure 4

In vitro stimulation of γδ T-cells. The proportion of IFNγ+ γδ T-cells after in vitro stimulation with PfRBCs. Figures show values for individual mosquito bite infected- (MB, blue circles, n=12) or blood-stage infected subjects (BS, red squares, n=12) at baseline (one day before challenge C-1) and respectively day 9 or day 7 post-challenge (C+9, C+7); values corrected for proportions of IFNγ+ cells stimulated with uninfected RBCs (range 0·25 - 8·14 %). P-values by Wilcoxon signed-rank test.