Cerebral malaria and epilepsy

Authors

  • Edgard Brice Ngoungou,

    1. Institute of Neuroepidemiology and Tropical Neurology (EA 3174), Faculty of Medicine, Limoges, France
    2. Department of Parasitology-Mycology and Tropical Medicine, Faculty of Medicine, University of Health Sciences, Libreville, Gabon
    Search for more papers by this author
  • Pierre-Marie Preux

    1. Institute of Neuroepidemiology and Tropical Neurology (EA 3174), Faculty of Medicine, Limoges, France
    Search for more papers by this author

Address correspondence to Edgard Brice Ngoungou, Department of Parasitology-Mycology and Tropical Medicine, Faculty of Medicine, University of Health Sciences, BP 4009 Libreville, Gabon. E-mail: ngoungou2001@yahoo.fr

Summary

Malaria, one of the most common parasitic diseases worldwide, is responsible for more than one million deaths among African children every year. Its neurological form, known as cerebral malaria (CM) is a potential cause of epilepsy in malaria-endemic regions of the world, primarily made up for the most part by the sub-Saharan Africa. Herein, we review recent African studies that examine the association between CM and epilepsy. Three studies suggest a modestly strong association between CM and epilepsy. Furthermore, there appears little doubt that this association is causal. Speculative considerations that may explain this causal association are discussed in this review. Additional research is however required in order to determine the clinical and electrographic behavior, the underlying structural and molecular basis, and course and outcome of this condition.

Malaria is one of the most common parasitic diseases and is a major public health problem in many tropical, impoverished, or developing countries. Worldwide, between 350 and 500 million clinical episodes of malaria occur, resulting in over one million estimated deaths every year (World Health Organization [WHO], 2003, 2004). Children, less than 5 years of age are particularly susceptible because of low levels of immunity (Snow et al., 1999). Over 80% of cases and 90–95% of deaths occur in the sub-Saharan Africa, where an estimated 600,000 children of age less than 5 years are affected by cerebral malaria (CM) (representing 10% of all malaria episodes) every year (Fig. 1) (Murphy & Breman, 2001). Of these, about 100,000 children die every year, amounting to a mortality rate of 20% and nearly 20,000 experience neurological sequelae lasting beyond 6 months. In pediatric wards of many hospitals in the sub-Saharan Africa, malaria and CM may account for 40% and 10%, respectively of all admissions (Marsh et al., 1995).

Figure 1.


Global distribution of malaria transmission risk, 2003 (from Hay et al., 2004).

Malaria is caused by the microparasite, Plasmodium. Of the four species of the parasite (P. vivax, P. ovale, P. malariae, and P. falciparum) that are responsible for disease, P. falciparum and P. vivax are most common with widespread distribution throughout the world. Although globally distributed, the highest levels of endemicity have been documented in the sub-Saharan Africa and possibly parts of South and Southeast Asia and South America. The other two species rarely produce human disease in selected geographic distributions. P. falciparum is the foremost amongst all species in the propensity to cause severe disease, while the other species of Plasmodium rarely cause death or serious immediate and long-term morbidity. Indeed, acute infection due to P. falciparum frequently affects the central nervous system (CNS). Coma, headaches, seizures, and impaired consciousness are frequent manifestations of this infection.

The malarial parasite, P. falciparum is transmitted by the female mosquito of the genus, Anopheles. When the mosquito bites human beings in order to ingest a bloody meal, it incidentally injects sporozoites into the human host (Fig. 2). The sporozoites are then transported to the liver through the blood stream, where they undergo replication and proliferation (known as exoerythrocytic shizogony). The next stage of the parasite, that is, merozoites are released from the liver into the blood stream, where they invade the erythrocytes, and over a period of 48 h proceed through morphologically distinct stages, before they (the schizonts) eventually rupture the erythrocytes. Ring stages are seen in peripheral blood smears, but trophozoites and meronts are usually absent, as they are sequestered within the deep vascular beds. Thus, although the parasite undergoes development in the liver, the erythrocytic phase is responsible for clinical manifestations of falciparum malaria.

Figure 2.


Life cycle of Plasmodium falciparum.

Falciparum Malaria: Clinical Manifestations and Pathophysiology

Malaria commonly presents as an acute febrile illness, though at times it may remain asymptomatic. Two main categories of illness are recognized: simple (uncomplicated) malaria and complicated (severe) malaria. The clinical spectrum of severe malaria, caused by P. falciparum includes CM, acute hemolytic anemia, acute respiratory distress syndrome, acute renal failure, metabolic acidosis, and hypoglycemia. CM is one of the most dreaded manifestations of infection. Mortality approaches 20% in children and most of those who die do so within the first 24 h (Molyneux et al., 1989). Although, mortality rates due to CM are high, recovery promptly follows institution of antimalarial treatment as most cases improve dramatically within 24–48 h of initiation of treatment. Several authors have proposed case definitions for CM and these have been widely used in clinical practice and epidemiological studies in the sub-Saharan Africa (Table 1) (Warrell et al., 1982; Newton et al., 1990). Seizures are common manifestations of the acute stage of CM but may also occur in the absence of manifestations of CM, that is, uncomplicated malaria with seizures (Carter et al., 2004). There are some geographical variations in the incidence of seizures as a feature of acute disease (Molyneux et al., 1989; Warrell, 1997; Kochar et al., 2002). In series from South and South-east Asia, seizures constitute 20% of presenting manifestations. In African children, however, seizures have been described in up to 80% of acute CM episodes treated in hospital facilities.

Table 1.  Case definition for cerebral malaria for use in clinical and epidemiological work-up
  1. Source: Adapted from Warrell et al., 1982; Newton et al., 1990.

1. Profound loss of consciousness (Blantyre coma scale <3) lasting for >1 h (so as to exclude postictal state)
2. Demonstration of asexual parasitemia on thick and thin blood films (at least three films made from EDTA-treated blood samples collected 12 h apart)
3. Exclusion of alternative infectious causes of encephalopathy (e.g., bacterial meningitis and viral encephalitis ruled out by lumbar puncture examination)

The clinical manifestations of acute CM are thought to be related to the sequestration of parasitized erythrocytes in the cerebral microvasculature (White & Ho, 1992; Kaul et al., 1998; Newton et al., 2000). The latter appears to be a preferred site for the parasitized erythrocytes to settle because of the favorable hypoxic conditions therein. Sequestration is mediated through proteins expressed on the parasitized erythrocyte surface and specific receptors on the vascular endothelium including possibly intracellular adhesion molecule-1, leading thereby to adhesion between the erythrocytes and endothelium. In addition, there occurs loss of deformability of parasitized erythrocytes and aggregration of nonparasitized erythrocytes to the parasitized erythrocytes, thereby forming rosettes, all of which lead to profound impairment in cerebral microcirculation (Cranston et al., 1984; Rowe et al., 1997). Finally, evidence for a role of tumor necrosis factor-α and other interleukins, which possibly mediate the release of nitric oxide at a tissue level has been presented (reviewed in Newton et al., 2000). The demonstrated upregulation of interleukins may merely be an epiphenomenon. However, as several interleukins are involved in the generation of febrile seizures and epileptogenesis, it is tempting to speculate their role in seizures and epilepsy associated with CM. The net result of the above processes is microvascular congestion, breakdown of the blood–brain barrier, and eventually raised intracranial tension leading to herniation and both micro- and large vessel infarcts. Raised intracranial tension largely determines acute mortality but the micro- and large vessel occlusion are perhaps responsible for long-term neurological sequelae.

CM is a medical emergency demanding immediate clinical assessment and treatment. Impairment of consciousness, convulsions, and other neurological features should raise the possibility of CM in any person who might possibly have been exposed to this infection during the previous year. Such cases deserve transfer to the highest available level of care, where an appropriate antimalarial drug should be administered as soon as possible, ideally by the parenteral route. Currently, the treatment of choice for severe or complicated malaria (including CM) is intravenous quinine. Complications of CM, such as convulsions, hypoglycemia, and hyperpyrexia should be prevented or detected and treated early. Fluid, electrolyte, and acid-base balance need skilful management. Nursing care of the unconscious patient crucially determines outcome. Ancillary treatments should be avoided inasmuch as they have not proved safe and effective thus far (Newton et al., 2000).

Neurological sequelae occur in up to 10–17% children affected by CM in clinical series; their incidence in the general population is estimated to be 0.1–0.2/1000 child-years (Murphy & Breman, 2001). Persistent neurological sequelae may also contribute to the long-term mortality associated with malaria but there is lack of data of mortality consequence due to them. The persistent neurological sequelae include ataxia, deafness, hemiparesis, cortical blindness and cognitive, neurodevelopmental, and behavioral disorders, which may range from mild to severe. Some deficits are transient, whereas others improve over months but may not resolve completely. Other sequelae, for example, epilepsy or behavioral disorders, may develop after discharge and persist (Newton & Krishna, 1998). Neurological sequelae are likely to be underestimated inasmuch as very few studies report follow-up of patients after discharge from hospital and fewer still on the outcome beyond 6 months. Seizures are recognized among neurological sequelae, but the incidence of unprovoked seizures and epilepsy has been systematically explored only recently. In the subsequent section of this article, we review studies that suggest an association between malaria and epilepsy (defined as recurrent unprovoked seizures).

Methods (Search Strategy)

Literature was searched via PubMed search engine. Research articles and reviews were identified by keyword searches. Key words included malaria, CM, neurological sequelae and epilepsy, and seizures. Bibliography of identified papers were browsed to identify additional articles. Retrieved studies that examined an association between CM and epilepsy, employing standard definitions for CM and epilepsy were included for detailed appraisal (Warrell et al., 1982; Newton et al., 1990; Commission on Epidemiology and Prognosis, International League Against Epilepsy, 1993). The aim of this review was to explore the strength of available epidemiological evidence for an association between CM and epilepsy, to determine if this association were of causal significance and to speculate on the biological basis of this association if any.

Association between CM and Epilepsy: Epidemiological Evidence

CM has been identified as a potential cause of epilepsy in tropical regions. Many experts have frequently conjured upon the link between CM and epilepsy and proposed a causal association between the two (Senanayake & Roman, 1991; Commission International League Against Epilepsy (ILAE), 1994; de Bittencourt et al., 1996). However, very few studies have systematically explored and quantified the relationship. Three recent studies with complementary epidemiological approaches performed in the sub-Saharan Africa (Kenya, Mali, and Gabon) deserve mention in this connection (Carter et al., 2004; Ngoungou et al., 2006a, 2006b).

A cross-sectional study among children of 6–9 years of age identified CM cases admitted to a district hospital in Kenya over an 8-year period. In all, 487 children were recruited to the study, including 152 with CM, 156 with uncomplicated malaria and acute symptomatic seizures (M/S), and 179 who were presumably unexposed to CM. The study revealed that the neurological sequelae following CM were much more frequent than previously appreciated. The prevalence of epilepsy was significantly increased among children with prior CM (odds ratio [OR]: 4.4; 95% confidence interval [CI]: 1.4–13.7; p = 0.01) and M/S (OR: 6.1; 95%CI: 2.0–18.3; p = 0.001) in comparison to unexposed children (Carter et al., 2004). The second study was an incidence-based cohort study with backdated follow-up obtained from a database maintained for malaria among children, aged 6 months to 15 years in Mali (Ngoungou et al., 2006a). The cohort was sub-classified in to those previously exposed to CM (n = 101) and those without prior exposure (n = 222). The incidence rate of epilepsy in the CM group was 17.0/1000 person-years and 1.8/1000 person-years in the unexposed group. The age-adjusted relative risk for developing epilepsy after exposure to CM was 14.3 (95%CI: 1.6–132.0; p = 0.01) suggesting a strong association between CM and epilepsy. Finally, a hospital-based, matched, case-control study of 296 people with epilepsy was undertaken in Gabon (Ngoungou et al., 2006b). A prior diagnosis of CM was confirmed among cases and controls using hospital records. The risk of developing epilepsy was 3.9 (95%CI: 1.7–8.9; p < 0.001) times greater among those with prior exposure to CM than unexposed people in this source population.

CM and Epilepsy: From Association to Causation

The studies hitherto referred to vary in design and in the point estimates of the risk of epilepsy associated with CM. The studies may have also been subject to some degree of bias as the source populations were hospital-based and not truly community-based, and enrollment of all eligible subjects was not possible due to logistic reasons. Furthermore, the CI for the OR or relative risk for epilepsy in some of the studies were rather wide. Nevertheless, the studies are remarkable in having been undertaken in an environment in which logistic, economic, and ethnic considerations present considerable operational difficulties for their conduct. Notably, the measures of association used in the studies suggest a modestly strong association between CM and epilepsy. The results are consistent across the studies despite the diverse methodology used, source populations, and geographic locations within Africa. Finally, by excluding cases of epilepsy with onset prior to the episode of CM, the studies establish temporal plausibility of the association between CM and epilepsy implying thereby that epilepsy followed the occurrence of CM. All of these provide reasonable ground for inferring a causal association between CM and epilepsy.

Association between CM and Epilepsy: Pathological and Pathophysiological Considerations

The mechanisms that engender the process of epileptogenesis following the acute CM episode are unclear. Magnetic resonance imaging studies are likely to be critical in clarifying these mechanisms. There is evidence for occurrence of structural damage to the brain mostly in the form of vascular-ischemic lesions that develop due to the sequestration of parasitized erythrocytes during the acute attack of CM. Another histological feature of CM, the Durck's malarial granuloma, which essentially comprises of reactive astrocytes might also represent a potentially epileptogenic lesion (Aleem, 2005). In addition, generalized hemorrhage and deep ischemic lesions have been observed in postmortem studies of CM (Toro & Roman, 1978).

The role of acute malarial seizures in the development of subsequent epilepsy requires consideration. These occur frequently during the acute CM episode and occur in a variety of forms including overt generalized tonic–clonic, focal seizures with or without secondary generalization, and subtle or purely electrographic seizures. Not infrequently, seizures are recurrent and prolonged. In one series of children with CM from Kenya, 28% developed status epilepticus during the acute episode (Crawley et al., 1996). While the possibility that the seizures in acute CM may merely be febrile seizures cannot be discounted, there are certain differences between acute malarial seizures and febrile seizures. A good proportion of acute malarial seizures are complex, recurrent, and prolonged in contrast to simple febrile seizures. In addition, seizures may occur in the later part of the acute febrile illness or sometimes even after the fever has settled in malaria. Recorded rectal temperatures are often less than 38.0°C in malarial seizures (Waruiru et al., 1996). It is plausible that the occurrence seizures in uncomplicated malaria is a sign of CM without other neurological manifestations. The role of acute malarial seizures and status epilepticus in development of subsequent epilepsy needs to be better clarified even though parallel may be drawn from the increased risk of epilepsy following febrile seizures and early seizures in other CNS infections (Annegers et al., 1988; Maher & McLachlan, 1995; Berg, 2003).

Factors other than cerebral damage caused by the acute CM episode may be involved in the development of subsequent epilepsy. One factor that has been studied to some extent is the role of genetic propensity to develop epilepsy. Interesting in this regard is the observation that relatives of children admitted with severe falciparum malaria are more likely to have seizure disorders than controls (Versteeg et al., 2003). In a follow-up study that however did not address epilepsy, the risk of neurological sequelae was increased with younger age at occurrence of the CM episode, deep coma, multiple acute malarial seizures, and associated malnutrition (Idro et al., 2006). Finally, the relevance of neurotoxins such as quinolinic acid (Dobbie et al., 2000) and autoantibodies to voltage-gated calcium channels, which have been demonstrated in high titers in children with severe malaria and seizures, requires further consideration (Lang et al., 2005).

Seizures and Epilepsy in CM: Outcome and Treatment

Specific antimalarial treatment, if promptly instituted unequivocally reduces mortality due to acute severe malaria but its impact on long-term neurological sequelae including epilepsy has not been studied (van der Torn et al., 1998). As early seizures complicating CNS infections are predictive of the later development of epilepsy, it is relevant to review the impact of control and prevention of early seizures on the development of later epilepsy. In a Kenyan study, the prophylactic administration of phenobarbital (20 mg/kg, intramuscular) reduced the incidence of acute symptomatic seizures by half and also of neurological sequelae (Crawley et al., 2000). However, mortality in the treated group was twice that of the untreated group, primarily due to a high frequency of respiratory arrest in the treated group. Thus, though the option of preventing seizure-induced brain damage with the prophylactic administration of phenobarbital may sound appealing, this intervention has practical limitations in the African environment with limited resources and manpower to carry out resuscitation even in the hospital setting.

The semiological and electrographic characteristics (i.e., neocortical vs. temporal lobe epilepsy) and the course and outcome of epilepsy due to prior CM have not been systematically reported. Limited information suggests that both focal and secondary-generalized seizures with a range of frequencies can occur. The response to antiepileptic drugs and hence propensity to intractability is also largely undetermined, given the magnitude of the treatment gap for epilepsy in the African environment.

Conclusions

Causality between CM and epilepsy remained unheard of till very recently when epidemiological evidence for an association between the two conditions was presented. The elucidation of pathogenic basis of this association requires a number of sophisticated imaging, pathological, molecular, and experimental studies. These studies would provide insight not only into the mechanisms of development of epilepsy following CM, but also in the development of rational interventions directed to reducing the impact of CM on neurological sequelae including epilepsy. The burden of epilepsy in sub-Saharan Africa due to malaria needs to be determined. Even if the epilepsy is not the most frequent among neurological sequelae of CM, the infinitely large numbers of cases of CM episodes in the sub-Saharan Africa means that CM would contribute substantially to the burden of epilepsy in this continent.

Conflict of interest: We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. The authors have declared no conflicts of interest.

Ancillary