Alterations of blood coagulation in controlled human malaria infection

Malaria is the most important parasitic disease of mankind, leading to approximately 600,000 deaths per year [1]. The majority of fatal cases are due to infections with the protozoan parasite Plasmodium falciparum. Leading causes of death in patients with malaria infection are severe anaemia and cerebral malaria. Moreover, in a review of imported malaria cases requiring intensive care unit admission it was reported that about 10 % of severe malaria cases develop disseminated intravascular coagulation (DIC), a serious coagulopathy associated with poor survival [2]. Also in malaria patients without clinically evident bleeding or thrombotic complications, alterations of the blood coagulation system, such as decreased levels of plasma antithrombin or elevated levels of plasminogen activator inhibitor (PAI)-1, and thrombocytopaenia are frequently found [3, 4]. Thrombocytopaenia is observed in 60–80 % of malaria cases and presents more frequently and severe in complicated P. falciparum malaria [4]. In severe malaria patients, also reduced levels of von Willebrand factor (vWF) cleaving protease, ADAMTS13, and increased vWF levels have been found [5, 6]. Interestingly, it has been discussed that the pathogenesis of severe malaria, especially of cerebral malaria, might be closely linked to alterations of the blood coagulation system [7, 8] and alterations of coagulation markers have been used as prognostic markers [8, 9]. Platelets and the coagulation system have important roles in sequestration of infected red blood cells, causing microvessel thrombosis, especially within the brain vasculature. Recently, the link between the innate immune response and coagulation as part of the host defence mechanism has been termed ‘immunothrombosis’ [10], and this phenomenon might be relevant in malaria.

Interestingly, platelets, the cellular component of the coagulation system, were found to be able to directly kill P. falciparum parasites [11]. Further research on the interplay between malaria infection and the coagulation system is, therefore, of great interest. Novel and promising coagulation tests that are currently intensively studied include the thrombin generation assay, which is a global coagulation test that reflects the ability of an individual to generate thrombin, the central molecule of the coagulation cascade. Clinical studies showed that this assay correlates with thrombotic and also with bleeding complications in different patient populations [12, 13]. Another currently intensively studied biomarker is soluble P-selectin (sP-selectin), which is a marker of platelet activation. sP-selectin has been extensively studied in patients with vascular thrombotic diseases and might be a promising biomarker of thrombotic events [14, 15].

Standardized investigations of the effects of malaria on humans are difficult to conduct in malaria patients due to frequent co-infections and widely varying clinical manifestations, time of infection and levels of parasitaemia. One way to standardize such investigations is to deliberately infect healthy malaria volunteers by controlled human malaria infection (CHMI). Trials of CHMI are primarily conducted to assess the efficacy of investigational malaria vaccines and drugs. Until recently, these studies were primarily conducted by exposure of volunteers to the bites of P. falciparum-infected mosquitoes. In 2013, the first report of CHMI by intradermal injection of aseptic, purified, cryopreserved P. falciparum sporozoites (PfSPZ Challenge) was published [16]. The incubation period was prolonged after intradermal injection, and therefore development of a better model using intravenous (IV) injection of PfSPZ was needed. This was accomplished at the University of Tübingen, Germany [17].

The thrombin generation assay is a global coagulation test that gives information on the potential of an individual to generate thrombin, which reflects the overall haemostatic capacity [12]. Thrombin is the central molecule in the coagulation cascade and measurement of the amount of thrombin that can be generated gives additional information compared to conventional clotting times, such as, for example, the activated partial thromboplastin time (aPTT), which mirrors only the time until the start of clot formation. The thrombin generation assay uses tissue factor (TF) and phospholipids to activate the coagulation cascade and by use of a fluorogenic thrombin substrate a thrombin generation curve can be obtained, which describes the concentration of thrombin generated over time and represents all three phases of coagulation (initiation, propagation and termination) [13]. Several parameters can be read out from the thrombin generation curve, which are the lag phase (time until thrombin burst), peak thrombin (peak amount of thrombin), time to peak, velocity index of thrombin generation curve, and the total amount of thrombin generated [area under the curve (AUC)].

High thrombin generation is associated with hypercoagulable states, such as obesity [18], and increased thrombin generation was shown to predict risk of first and recurrent venous thromboembolism (VTE) [19, 20] and risk of VTE in patients with cancer [21]. Moreover, thrombin generation is diminished in patients with bleeding tendencies and it therefore not only reflects thrombotic, but also bleeding tendency, providing a laboratory test that covers a broader spectrum of haemostasis in comparison to conventional clotting times [13].

The coagulation system is increasingly recognized to play an important role in malaria [7]. Obstruction of small vessels and binding of parasitized red blood cells to endothelial cells are crucial events in the pathogenesis of severe malaria, and endothelial cell activation and activation of the coagulation cascade are proposed to be involved in this process [4]. Several studies have shown altered levels of coagulation parameters in malaria patients [22]. Recent data revealed underlying pathophysiological mechanisms leading to hypercoagulability of malaria patients: Ndonwi et al. showed that the parasite-derived protein, histidine rich protein II (HRPII), has an inhibitory effect on the anti-coagulatory protein antithrombin, which acts physiologically as an inhibitor of thrombin [23]. Furthermore, Moxon et al. found that P. falciparum-infected erythrocytes can cause loss of endothelial protein C receptor (EPCR). EPCR physiologically enhances conversion of protein C to activated protein C (APC); APC is an anticoagulatory protein and decreased APC due to loss of EPCR consequently leads to the generation of high levels of thrombin [24]. According to some case reports, recombinant APC (Drotrecogin alfa) was also successfully used for treatment of patients with severe P. falciparum malaria who did not respond to standard treatment [25]; however, a clinical trial failed to show benefit for this drug in patients with severe sepsis [26]. Our in vivo data showed that the thrombin generation potential began to change when parasites were at a low density, approximately 5–6 days after release from the liver, and before most subjects were symptomatic. This supports the hypothesis that P. falciparum specifically increases thrombin generation potential at the same time as parasites can be first detected by highly sensitive thick blood smear microscopy. However, it also has to be mentioned that in one-third of study participants no change in thrombin generation was observed.

No significant changes in levels of the fibrin degradation products D-dimer and prothrombin fragment 1 + 2 were found in early stage malaria in our study, although these parameters indirectly reflect in vivo thrombin generation. The in vitro thrombin generation potential might be a more sensitive parameter for subclinical coagulation alterations, whereas other parameters might indicate coagulopathy in more advanced or severe malaria cases. In the current study, patients were treated immediately with anti-malarial drugs as soon as first microscopically detectable parasitaemia occurred.

In the current study, also no changes in vWF, ADAMTS13-Activity, sP-selectin or platelet count during early stages of malaria infection were observed. These results are in part contrary to previous studies investigating alterations of coagulatory parameters in P. falciparum infection: a study of CHMI induced by bites of P. falciparum infected mosquitos found decreased blood platelet counts and elevated levels of vWF at very early stages of malaria infection [27], which was not seen in our study cohort. Underlying explanations for the different results remain obscure. It is unlikely that the differences were related to route of infection (IV injection of PfSPZ versus ID injection of PfSPZ versus bites of infected mosquitoes), time point of the assessment during the course of the disease or the severity of the infection. In concordance with previous studies ADAMTS13-activity and sP-selectin remained unchanged in early malaria in this study [27, 28]. However, in symptomatic Indonesian malaria patients ADAMTS13-activity was found to be reduced, indicating that a decrease in this enzyme probably occurs at later stages of malaria [29].

In the current study, alterations of the thrombin generation potential were not observed before microscopic detection of parasites in peripheral blood. Measurement of coagulation parameters in malaria, such as thrombin generation, might therefore not be relevant prior to a microscopically detectable parasitaemia, although the routinely achieved limit of detection of microscopy is usually at least tenfold higher than in this study [17]. Whether the thrombin generation potential might be helpful in predicting severity and outcome of P. falciparum infection remains to be studied in appropriate populations, since in the present study parasitaemia was several orders of magnitude lower than in severe or even uncomplicated malaria.