Malaria: a haematological disease

Abstract

Plasmodium falciparum malaria remains a major cause of mortality throughout the tropical world. Haematological abnormalities are considered a hallmark of malaria, bearing an impact on final outcome and representing indices of prognostic and follow-up value. These include severe anaemia, coagulation disturbances, leukocyte numerical or functional changes and spleen involvement. Anaemia involves red blood cell lysis due to parasite invasion, as well as mechanisms of intravascular haemolysis and decreased erythropoiesis. Exchange or blood transfusion is mainly recommended in the management of these patients. Haemorrhagic complications in severe malaria are relatively rare despite prominent thrombocytopenia and dysfunction in the coagulation pathway. Numerical, as well as functional changes in the white blood cell are less dramatic than other blood cell series, but still, remain a significant index of disease progression and ultimate prognosis. Finally, the role of the spleen in severe malaria is multifactorial. Care and vigilance should be taken against splenic rupture which is fatal and can occur despite appropriate antimalarial prophylaxis and treatment.

Keywords:

• Malaria
• Malarial anaemia
• Thrombocytopenia in malaria
• Malaria leukocytosis

Introduction

Malaria remains a major cause of mortality in the tropical world, with as many as 500 million cases annually.¹ Severe malaria is a multisystem disorder, presenting with multiple manifestations, requiring hospitalization, parenteral antimalarial therapy, as well as appropriate management of evolving complications. Setting an extreme example of parasitism, malaria plasmodia spend most of their complex life cycle intracellularly, primarily within their host–cell erythrocyte. Because of this association between the parasites and red cells, there are numerous consequences to the host's blood extending far beyond the direct effect of parasitized red blood cells, including severe anaemia, coagulation disturbances, leukocyte numerical or functional changes and spleen involvement.

Nevertheless, assessment of these disturbances in the vertebrate host can be complicated by several interacting factors. The species of the infecting parasite is one, for example, anaemia from Plasmodium falciparum tends to be much more severe than from other species. Furthermore, the endemicity of malaria is important in determining the relative incidence and patterns of haematological complications associated with malaria, as well as the age at which these manifestations occur.^2–4 Host genetic factors are relatively common in endemic areas resulting from selection pressure, including many red cell polymorphisms, such as the haemoglobin variants S and C, and thalassaemias can also complicate the course of disease.⁵ For example, much epidemiological and cellular evidence has accrued over the years to support the hypothesis that these variants provide relative protection against malaria. Finally, co-morbidities, particularly in patients living in holoendemic areas, including hookworm infestation, malnutrition or HIV, have been found to be other important contributing factors.⁶ As a result, assessment of haematological complications in such diverse conditions is far from an easy task, and should always be considered under the spectacle of individual settings. The purpose of this review is to set a general framework of the major haematological complications of severe malaria and elucidate the mechanisms underlying their pathogenesis.

Red Blood Cell (RBC) abnormalities

Anaemia

Severe malaria anaemia, defined as a haemoglobin level <5 g/dL in children or <7 g/dL in adults, represents a major public health burden in malaria-endemic areas, with prevalence varying from 30% to 90% in children and pregnant women.⁷ Severe malaria anaemia is quite pronounced in P. falciparum malaria, in contrast to other plasmodia that selectively invade only specific red cell populations. For example, P. vivax and P. ovale tend to invade reticulocytes whilst P. malariae invade more mature forms of erythrocytes. Together with cerebral malaria, severe anaemia is a leading cause of morbidity and mortality ranging between 5.6% and 16% in children^8,9 and 6% in pregnant women, particularly primigravidae.¹⁰

Pathogenesis of malaria anaemia remains both complex and multi-factorial. Parasite multiplication results in decreased haematocrit level due to rupture of RBCs during release of the different stages of maturing parasites. However, the severity of malarial anaemia, in which the haematocrit level may approach 15% or less, does not correlate with the degree of parasitaemia and is often far in excess of what can be accounted for by the loss of infected red blood cells (iRBCs) alone.^11,12 Therefore, this must, by inference, include increased destruction of both infected and uninfected cells, and implicate decreased erythropoiesis or potential miscalculation of actual parasitaemia due to sequestered parasite population in the deep vasculature of subcutaneous tissues.¹³

Indeed, non-iRBCs in malaria patients have a shortened life span compared to those of healthy individuals and their haemolysis seems to account for more than 90% of erythrocyte loss during acute malaria.^14,15 Mathematical modelling of haematological data from experimental human P. falciparum infections has shown that up to 12 non-iRBC are lost for every iRBC,^15,16 suggesting that the destruction of non-iRBCs could be the major cause of haemoglobin drop [reviewed by Ekvall¹⁷]. The mechanism underlying the shortened life span of RBCs, however, is unclear. It is known that, during malaria infection, the RBC undergoes oxidative stress and haemichromes (haemoglobin degradation products) are formed. Haemichromes possess a strong affinity for membrane protein band 3 cytoplasmic domain and, following their binding, lead to band 3 oxidation and clustering, similar to normal senescence.^18–20 Those band 3 clusters show increased affinity for auto-antibodies which activate complement and finally trigger the phagocytosis of altered RBC.^21,22 Immunoglobulin of the vertebrate host may also recognise pure parasite antigens expressed on the RBC surface, resulting in antibody deposition on the RBC membrane and subsequent phagocytosis by macrophages.^23–25 Loss of the complement regulatory proteins, including complement receptor type 1 (CR1) and CD55, from the red cell surface has also been shown to correlate with severe malarial anaemia.^23,26 As these proteins are important in protecting cells from complement-mediated damage by controlling the increased activation of complement cascade in malaria,^27,28 the loss of CR1 and CD55 from the RBC surface may result in increased lysis and/or phagocytosis of uninfected RBCs.

Malaria infection is generally associated with reduced total erythropoietic activity, since inappropriately low reticulocyte numbers have been observed in malaria patients.^29,30 Bone marrow cellularity is often abnormal, associated with erythroid hypoplasia and hypoproliferative erythropoiesis, most prominent during chronic infection.³¹ A number of mechanisms have been proposed to cause the defective erythropoietin stimulated proliferation of hematopoietic progenitors including cytotoxic parasite derived molecules, namely haemozoin resulting from digested haemoglobin.³² Levels of haemozoin have been shown to be associated with anaemia and reticulocyte suppression,³² while malarial pigment appears to directly promote apoptosis of erythroid precursors.³³ Moreover, ingestion of iRBCs or malarial pigment (haemozoin) induces the release of macrophage migration inhibitory factor (MIF) from macrophages, which in turn, inhibits erythroid, progenitor-derived colony formation.³⁴ Following defective proliferation, the differentiation and maturation of blood cell precursors is also inhibited, possibly due to a shift in transferrin receptor expression to non-erythroid cells and decreased iron uptake.^29,35 Finally, malaria toxins have been shown to stimulate macrophages to produce TNF-a³⁶ which, together with IFN-g and IL-12, are the major pro-inflammatory cytokines produced during the acute response to blood-stage malaria in both humans and mice.³⁷ TNF-a, IFN-g, and IL-12 as well as other cytokines have been shown in vitro and in vivo to mediate anaemia associated with chronic infections, including malaria.³¹ It has been demonstrated that TNF-a directly inhibits the growth of red cell precursors³⁸ and also indirectly inhibits erythropoiesis by inducing IFN-g production by accessory cells.³⁹ TNF-a has also been shown to be capable of reducing the life span of red cells, causing increased erythrophagocytosis,⁴⁰ and inducing apoptosis of erythroid precursors.⁴¹ Although several pro-inflammatory cytokines may inhibit kidney EPO production,⁴² multivariate regression analyses revealed that the overall contributions of cytokines to variations in EPO production are relatively minor compared with the significant impact exerted by the degree of anaemia, as reflected in a marked decrease in the haematocrit level during the acute phase of infection.³⁵ Whether a more pronounced impact of cytokines on EPO production would be evident during chronic malaria, when kidney damage may be severe, requires further investigation.

Blood transfusion

Blood transfusion has traditionally been the cornerstone of treatment of severe anaemia.⁴³ To date, there is no clear evidence supporting specific haemoglobin cut-off levels, and various figures have been quoted in the literature. In adults, the threshold for blood transfusion is traditionally set at a haematocrit <20%. In a recent study from Kenya, an absolute cut-off of haemoglobin levels of 4 g/dL was set for African children. This was extended to 5 g/dL in the presence of co-existing respiratory distress, impaired consciousness, or hyperparasitaemia.⁴⁴ However, there is a paucity of follow-up data in children with Hb 5.0–6.0 g/dL who are not transfused, leaving open questions as to the outcome in this subset of patients. On the other hand, a relatively recent Cochrane review reported that routine transfusion did not reduce mortality, but caused more adverse events including circulatory overload, transfusion reactions, and occasionally death.⁴⁵

Exchange transfusion

Exchange transfusion has been used in the treatment of severe malaria with apparent benefits.^46–48 Potential mechanisms include rapid reduction of parasitaemia by direct parasite removal, reduction of cytokines and/or toxic by-products, improved rheology by replacing unparasitized erythrocytes having reduced deformability with transfused cells, while correcting anaemia and acidosis.^49,50 A recent retrospective cohort study has shown significantly shorter parasite clearance times for patients with severe malaria treated with adjunct exchange transfusion than for patients treated with parenteral quinine only, recommending its beneficial effect in view of lack of fatalities.⁵¹

However, to date, no adequately powered randomized controlled clinical trial has been performed to establish its clear indications. Exchange transfusion is strongly recommended, particularly in well-resourced settings where safe blood is available for all patients in whom the parasitaemia exceeds an arbitrary 30%, even in the absence of other clinical manifestations of disease.⁵² Exchange transfusion should also be considered for patients with a parasitaemia higher than 10% in the presence of complications such as cerebral malaria, non-volume overload pulmonary oedema, renal compromise or if concomitant morbidity factors exist including pregnancy and old age.^52,53 Nonetheless, a meta-analysis of small studies and case series showed no clear benefit between groups of patients receiving exchange transfusion or not.^54,55

In the decision to use exchange transfusion, the potential risks including fluid overload, febrile and allergic reactions, metabolic disturbances, red blood cell alloantibody sensitization, transmissible infection, cerebral haemorrhage, and line sepsis, must be weighed against the potential benefits.⁵⁴ Erythrocytapheresis in which only the diseased red blood cell fraction is removed by apheresis, may be of advantage over classical exchange transfusion since leukocytes, platelet fractions or useful plasma components (i.e. clotting factors) are returned to the patient.^56–58 Moreover, an accurate fluid balance can be maintained while the reduction in parasitized cells is more than one should expect from the models used with exchange transfusion.^59,60 During treatment, useful parameters for monitoring progress should include parasite counts twice a day, regular pH and blood gas measurements and appropriate measurement of glucose, lactate concentrations, and renal function.

Bleeding disorders

Platelet derangement

Thrombocytopenia is one of the most persistent features of severe malaria, found in approximately 50–80% of malarial patients.^61,62 Normal platelet count in combination with a normal C-reactive protein (CRP) usually excludes the diagnosis.^63,64 Several mechanisms have been proposed for the low platelet counts observed in malaria. Demonstration of high molecular weight Von Willebrand factor in the plasma of patients suggested that thrombocytopenia in malaria is associated with endothelial damage and isolated platelet consumption.⁶⁵ The latter is in line with observations of normal or increased numbers of megakaryocytes in the marrow⁶⁶ and increased mean platelet volume in patients’ sera confirming the presence of giant platelets on stained blood films and reflecting early release from the bone marrow in response to peripheral platelet destruction.^63,67 Thrombopoietin levels (the main growth factor for megakaryocytes and thrombocytopoiesis) also appear be normal in patients with malaria⁶⁸ enhancing the hypothesis of isolated platelet consumption due to diffuse sequestration,⁶⁹ rather than dysmegakaryopoiesis.

Immune-mediated destruction of circulating platelets could also theoretically play a role in malaria thrombocytopenia, since elevated levels of platelet-bound IgG have occasionally been observed.⁷⁰ However, this would be expected late in the course of disease and cannot explain mild or moderate thrombocytopenia seen at the beginning of malarial infection. This comes in accordance with recent data from experimental infection with P. falciparum in healthy volunteers that failed to give rise to anticardiolipin antibodies.⁷¹ Recently, the role of elevated macrophage colony-stimulating factor (M-CSF) in malaria has also been studied. It appears that, by enhancing macrophage activity, M-CSF in malaria may result in increased macrophage-mediated platelet destruction. Furthermore, platelet/endothelial activation or damage, as measured by P-selectin could intensify thrombocytopenia in severe P. falciparum malaria.^72,73

Concerning platelet functional defects in malaria, abnormal platelet adhesion is reported in malaria.⁷⁴ Platelet microaggregates,⁷⁵ enhanced platelet activation⁷⁶ and increased levels of platelet factor 3,⁷⁷ have also been described in malarial patients. Raised concentrations of platelet specific proteins such as beta thromboglobulin (bTG), platelet factor 4 (PF4), and other changes such as enhanced production of thromboxane A2 and prostacyclin, have been observed. It has been postulated that these hypersensitive (hyperactive) platelets will enhance haemostatic responses, and may be the reason why bleeding episodes are rare in acute malaria infection, despite the thrombocytopenia.⁶⁶

Data relating to the clinical significance of deranged platelet profiles during active infection varies. Thrombocytopenia during infection is often associated with age, prostration and parasite density⁷⁸ but not with bleeding problems or mortality according to some authors.^79–81 It has actually been suggested that it only predicts the level of parasitaemia and not disease severity, potentially through an IL-10 mediated mechanism.^82,83 However, according to others, the extent of thrombocytopenia is a predictor of outcome as well as a predictor of both outcome and severity in children.^84–86

Coagulation derangement

Acute P. falciparum malaria also appears to activate or derange the coagulation pathways by diverse mechanisms.^87,88 Seventeen per cent of cases show evidence of intravascular coagulation,⁸⁹ whilst generalized bleeding or disseminated intravascular coagulopathy (DIC) are rare in these patients, despite accompanying thrombocytopenia.^90,91

In addition to coagulation pathway activation through cytokine dysregulation in sepsis, malaria also influences coagulation by its ability to activate coagulation directly and damage inactive endothelial cells leading to microparticle genesis and subsequent complement activation.⁹² It has also been hypothesized that infected erythrocytes possess pro-coagulant activity themselves, because of changes in the lipid distribution across the inner and outer surfaces of the infected erythrocyte membrane.⁹³ Activation of the coagulation cascade has been observed with a reduction in antithrombin concentration, an increase in thrombin–antithrombin complexes⁹⁴ and a reduction in factor XII and prekallikrein activities,⁹¹ which nonetheless does not appear to be clinically significant. Accordingly, prolonged prothrombin time (PT), and depletion of blood coagulation factors V, VII, VIII, IX have been reported, but liver involvement rather than a direct defect appears to cause those.⁹⁵ Fibrinolytic, antifibrinolytic and natural inhibitors of coagulation have been found to be significantly depressed in severe malaria.^96,97 Increased IL-6 levels in malaria result in an increase in fibrinogen and other inflammatory mediators, while TNF-α has been shown to cause tissue factor induction in monocytes, macrophages and in endothelial cells.^98,99

Whether this coagulopathy has any ‘functional’ importance in vivo in causing morbidity and/or fatal outcome since typical DIC (e.g., bleeding) occurs in 5% to 10% of severe malaria cases (1% of all cases) remains elusive.¹⁰⁰ Reports on the incidence of malarial bleeding are controversial varying from 4% to 25% depending on the setting.^101,102 Rare clinical manifestations include cases of hematochezia,^103,104 nose bleeding,¹⁰⁵ haematuria (even though mostly secondary to glomerulonephritis)¹⁰⁶ and hypermenorrhea^107,108 [reviewed by Wiwanitkit¹⁰⁹]. In cases of prolong clotting time, transfusion of fresh blood, clotting factors or platelets may be given. Administration of vitamin K (10 mg) by slow intravenous injection can also be useful under specific conditions.¹¹⁰ Aspirin and corticosteroids should be avoided. Patients in ITU should be monitored or receive prophylaxis against the risk of gastrointestinal bleeding.

White Blood Cell (WBC) abnormalities

Changes in the white blood cell (WBC) count are less dramatic than other blood cell series and have been controversial. White blood cell counts during malaria are generally characterized as being low to normal.^111–113 Observations show that WBC count decreases to its minimum at roughly the same time that fever begins and infection becomes detectable by microscopy.^114,115 Leukocytosis (reflected in the neutrophil count) is typically reported in a fraction of cases and may be associated with concurrent infections and/or poor prognosis,^80,116 but not always.¹¹⁷

The underlying mechanisms of apparent neutropenia (i.e. absolute neutrophil counts ≤1000 cells/μL) include a shift in neutrophils from the circulatory to the marginal pool, i.e. sites of inflammation, serum lymphotoxic factors, splenic localization and intercurrent bacterial infections. Observations from P. vivax infection support the view that neutropenia seems to be attributed to the rapid release of marrow granulocytes into the blood (left shift) coupled with a shift of neutrophils from the circulating to the enlarged marginal cell pools.¹¹⁸ Apparent neutropenia is attributed to changes in intravascular granulocyte distribution.

In addition, a markedly lower percentage and absolute numbers of T cells in the peripheral circulation and a transient inability of such cells to respond to antigenic stimulation in vitro have been observed. This temporary reallocation of T cells away from the periphery seems to correlate with disease severity, but the exact cause is as yet unknown.^119–122 Malaria induced apoptosis of the lymphocyte population has also been proposed.^123124125 It appears that IL-12 may be involved in the pathogenesis of malarial pancytopenia, thus leukopenia.³¹ Other authors have shown that malarial toxins, e.g. glycosylphosphatidylinositol, act directly on monocytes and macrophages and trigger the release of pro-inflammatory cytokines³⁶ which consequently may cause suppression of haemopoiesis and dyshaemopoiesis.¹²⁶

Macrophage function has also been found to be altered by the phagocytosis of debris – the so-called malaria pigment – released during schizogony. Malarial pigment and local release of reactive oxygen species inhibit macrophage function^127–129 and dendritic cell functions^130,131 and can suppress erythropoiesis.^34,83,132 Even though associated with disease severity,¹³³ the prognostic value of intraleukocytic pigment remains controversial.^133–135

Notably, WBC counts have been used for the microscopic estimation of parasitaemia¹³⁶ (even though ultimately found wanting), while automated detection of malaria pigments in WBC has been utilized for malaria diagnosis.¹³⁷ Either way, the clear normalizing trend of WBC counts with time and treatment can provide significant guidance for disease management.

Spleen involvement

Splenomegaly is frequently – though inconsistently – present in acute and chronic forms of malaria and represents an index marker of Plasmodium endemicity.^31,138 A multifactorial role for the spleen has been proposed in malarial infection, as reflected by its current pathology.¹³⁹ The spleen filters RBCs with reduced deformability, clears opsonized P. falciparum iRBCs and can extract drug-exposed Plasmodium parasites from their host RBCs.¹⁴⁰ Splenic macrophages also phagocytize both parasitized and normal red blood cells during malaria infection.¹⁴¹ Elimination of normal or immunologically damaged platelets by a hypertrophied reticuloendothelial system,^141,142 or excessive platelet pooling with a shortened life span has also been suggested.¹⁴³

A novel role for the spleen in malaria has recently been indicated by epidemiological¹⁴⁴ and experimental data.¹⁴⁵ Current paradigms, implicated the spleen in retention of mostly iRBCs of mature parasite forms, rather than young ring iRBCs, which importantly have a role in parasite multiplication capability. Apparently, a heterogeneous parasite ring stage population has been observed; the less deformable subset to be retained in the spleen, the more deformable circulating in the periphery.¹⁴⁵ As a result, the extent of RBC retention in the spleen seems to condition not only haemoglobin concentration and spleen size but also the rate of parasite load increase. These data indicate that cerebral malaria and severe malarial anaemia may result from distinct mechanisms with a variable involvement of the spleen.¹⁴⁴ Stringent splenic retention of ring stages and non-iRBCs could on one hand reduce the risk of cerebral malaria by preventing periphery sequestration, but also increase the risk of severe malarial anaemia, through increased RBC destruction.¹⁴⁵ Interestingly, severe malarial anaemia and cerebral malaria differ significantly with regard to prevalence of splenomegaly (57% versus 11%), and spleen size (Hackett grade 3.5 versus 0), while spleen size inversely correlates with haemoglobin level.¹⁴⁴

The spleen is palpable in 70–80% of patients, of which, in an estimated 2%, rupture can occur despite appropriate antimalarial prophylaxis and treatment,^146,147 especially in the context of P. vivax.¹³⁸ The exact mechanism of splenic rupture in malaria is still controversial. Potentially implicated mechanisms include: (i) an increase in intrasplenic tension due to cellular hyperplasia and engorgement; (ii) the spleen may be compressed by the abdominal musculature during physiological activities such as sneezing, coughing, defecation, etc.; and (iii) vascular occlusion due to reticulo-endothelial hyperplasia, resulting in thrombosis and infarction.¹⁴⁸ This leads to interstitial and subcapsular haemorrhage and stripping of the capsule, which leads to further subcapsular haemorrhage and capsule tear. In rare cases, subsequent organization can also lead to splenic cyst development. Splenic rupture should be considered in any patient with P. falciparum malaria who develops abdominal pain and/or signs or symptoms of cardiovascular collapse. Ultrasound and/or CT scanning are essential for diagnosis and management. Nonoperative management consists of inpatient observation, strict bed rest, and administration of fluid and blood transfusion as needed.¹⁴⁹ Abdominal ultrasonography should be performed regularly to assess healing. Splenectomy should be reserved for patients with severe rupture, continued or recurrent bleeding.¹⁵⁰

A rare form of malarial splenic activity is reflected in hyper reactive malarial syndrome characterized by gross splenomegaly, preferential stimulation of B lymphocytes as well as increased titres of serum IgM and malarial antibodies. Notably, the syndrome can rarely be found outside endemic areas and responds well (hence, its diagnosis is confirmed) to antimalarial treatment.^138,151,152

Conclusions

Haematological complications are prominent manifestations of severe malaria, commonly representing indices of prognostic and follow-up value. Elucidation of pathogenesis, which underlies blood-component derangements in those patients, even though extensively explored and observed, largely remains unknown. Understanding the Plasmodium life cycle and its interactions with its host erythrocytes, during its developmental journey along the vertebrate's circulatory system, can prove vital in future targeted management of haematological and other complications of severe malaria.