SEARCH

SEARCH BY CITATION

Keywords:

  • CNS infections;
  • Epilepsy;
  • Seizures;
  • Epidemiology;
  • Antiepileptic drugs

Summary

  1. Top of page
  2. Burden of CNS Infections
  3. Burden of Epilepsy
  4. Burden of Epilepsy Due to CNS Infections
  5. Epidemiological Approaches to CNS Infections and Epilepsy
  6. Microbiological Considerations
  7. Epileptological Considerations
  8. Problems in Studies of the Association between CNS Infections and Epilepsy
  9. Conclusions
  10. Acknowledgment
  11. References

The causal association between central nervous system (CNS) infections and epilepsy is predictable but poorly documented due to constraints in epidemiological, epileptological, and microbiologic methods. The large number of CNS infections with varied geographic distributions means that epidemiological studies in many different regions are required in order to establish their association with epilepsy. Whenever infectious diseases occur in the community, the majority of the infected cases are either asymptomatic or develop only mild symptoms. Those with neurological involvement and hence at risk of developing epilepsy constitute the tip of the iceberg. Furthermore, there is no one-to-one relationship between surrogate markers of infection used in epidemiological studies (e.g., seropositivity or brain imaging abnormalities) and neurological involvement (and hence epilepsy). As a result, there are individuals in the community who have no neurological symptoms but may either be seropositive or demonstrate imaging abnormalities compatible with the neurological infectious disorder and conversely, those who have seizures (or epilepsy) but may be seronegative. Relevant to the epidemiological study of seizures and epilepsy in relation to CNS infections is the classification of seizures as provoked and unprovoked. Accordingly, seizures that occur during the active stage of infection are considered provoked and those that occur later are unprovoked. Finally, with the burden of infections being concentrated in the less-developed countries, epidemiological studies are required to be carried out in these locations, which present logistic, financial and technical barriers for them to be accomplished.

Central Nervous System (CNS) infections continue to be recognized worldwide and despite variations in their incidence across the world, these constitute public health problems of considerable magnitude. This is largely due to the serious nature of morbidity produced not only by the active infection episodes but also their immediate and long-term sequelae. Focal neurological deficits, cognitive impairments, and neurobehavioral disorders are among the well-known long-term effects of CNS infections. Hence, efforts (in the form of drug intervention trials) have been undertaken in order to prevent and reduce the impact of these long-term complications. Epilepsy is an under-recognized long-term complication of CNS infections. Assesment of the size and impact of the association between CNS infections and epilepsy requires a population-based approach. In this review, we draw attention to the link between CNS infections and epilepsy and recapitulate some of the epidemiological, microbiological, and epileptological principles and drawbacks thereof in the study of this association.

Burden of CNS Infections

  1. Top of page
  2. Burden of CNS Infections
  3. Burden of Epilepsy
  4. Burden of Epilepsy Due to CNS Infections
  5. Epidemiological Approaches to CNS Infections and Epilepsy
  6. Microbiological Considerations
  7. Epileptological Considerations
  8. Problems in Studies of the Association between CNS Infections and Epilepsy
  9. Conclusions
  10. Acknowledgment
  11. References

There is a large variety of neurological infections and the exact incidence of each of these has not been determined. Regional and global estimates of the burden of some of the more common infections are now available. In the United States, about 25,000 individuals are hospitalized with a diagnosis community-acquired bacterial meningitis and viral encephalitis every year, giving an upper estimate of incidence of 14/100,000/population/year for the two disorders combined (Beghi et al., 1984; Schuchat et al., 1997). To compare these figures with two other common causes of epilepsy, the estimated number of cerebrovascular accidents and traumatic brain injuries that occur every year in the United States are 700,000 and 1.4 million, respectively (Bruns & Hauser, 2003; http://www.cdc.gov/stroke/online_resources.htm). The incidence of CNS infections is perhaps much higher in the rest of the world. The World Health Organization (WHO) estimates that 448,000 cases of bacterial meningitis (due to Streptococcus pneumoniae, Haemophilus influenzae, and Neisseria meningitides) and 408 million cases of malaria occur every year (http://www.who.int/healthinfo). Not all cases of malaria involve the CNS though; cerebral malaria one of the severe manifestations, which is associated with a significant burden of neurological sequelae is estimated to affect 600,000 children (of age <5 years) in sub-Saharan Africa each year (Murphy & Breman, 2001).

Burden of Epilepsy

  1. Top of page
  2. Burden of CNS Infections
  3. Burden of Epilepsy
  4. Burden of Epilepsy Due to CNS Infections
  5. Epidemiological Approaches to CNS Infections and Epilepsy
  6. Microbiological Considerations
  7. Epileptological Considerations
  8. Problems in Studies of the Association between CNS Infections and Epilepsy
  9. Conclusions
  10. Acknowledgment
  11. References

The WHO estimates that there are at present 50 million people with epilepsy worldwide (http://www.who.int/healthinfo/statistics). Three-fourths of these people reside in less-developed countries (LDCs). The enormous burden of epilepsy in LDCs is largely due to the majority of the world's population living in these countries. This burden may also be attributable in part to a higher incidence of epilepsy in the LDCs. Only limited data are available regarding incidence of epilepsy in LDCs. Even so, some of these data suggest incidence rates in excess over the incidence estimate of 50/100,000 population/year for developed countries (Sander & Shorvon, 1996). For instance, in two separate incidence studies (in Ecuador and Chile), the estimated incidence of a first unprovoked seizure was 190/100,000 population/year and of recurrent unprovoked seizures was 113/100,000 population/year (Lavados et al., 1992; Placencia et al., 1992).

Burden of Epilepsy Due to CNS Infections

  1. Top of page
  2. Burden of CNS Infections
  3. Burden of Epilepsy
  4. Burden of Epilepsy Due to CNS Infections
  5. Epidemiological Approaches to CNS Infections and Epilepsy
  6. Microbiological Considerations
  7. Epileptological Considerations
  8. Problems in Studies of the Association between CNS Infections and Epilepsy
  9. Conclusions
  10. Acknowledgment
  11. References

The etiological fraction of epilepsies due to CNS infectious disorders can be determined in studies of the etiology of newly diagnosed epilepsies in community-based samples. These studies are few and not really comparable due to differences in design and methodology. In the Rochester Epidemiology Project, 3% of newly diagnosed epilepsies were attributed to prior CNS infection. In comparison, cerebrovascular disease and traumatic brain injury were responsible for 11% and 4% of all newly diagnosed cases in this community (Annegers et al., 1996). There are virtually no studies of etiology in incident populations of epilepsy from LDCs. A study of newly diagnosed epilepsy drawn from five main hospitals in three cities of Ecuador revealed that epilepsy was etiologically related to cysticercosis in 9.6%, other CNS infections in 5.2%, perinatal insults in 9.6%, traumatic brain injury in 3.9%, and cerebrovascular disease in 5.7% cases (Carpio et al., 2001). This study suggested that CNS infections may account in part for the higher incidence of epilepsy in LDCs. Additional studies in other geographic regions are, however, required to substantiate this view.

Epidemiological Approaches to CNS Infections and Epilepsy

  1. Top of page
  2. Burden of CNS Infections
  3. Burden of Epilepsy
  4. Burden of Epilepsy Due to CNS Infections
  5. Epidemiological Approaches to CNS Infections and Epilepsy
  6. Microbiological Considerations
  7. Epileptological Considerations
  8. Problems in Studies of the Association between CNS Infections and Epilepsy
  9. Conclusions
  10. Acknowledgment
  11. References

Broadly speaking, the epidemiological study of the association between CNS infections and epilepsy may be approached using either cohort or case-control designs. Cohort studies are difficult to assemble. Since the infective episodes often occur several years prior to onset of epilepsy, follow-ups may have to be back-dated. The back-dated follow-up design was effectively used in several epidemiological studies in the Rochester community in the United States because of excellent access to medical records of individuals in the community from as early as the 1920s (Hauser et al., 1996). Back-dated cohort designs are, however, not suited to most LDCs since medical record-keeping is inadequate or sometimes nonexistent in these countries. Alternatively, case-control studies are immensely useful in testing etiological hypotheses as well as convenient because full enumeration of the cohort is not required. Regardless of the design, the sample collected thereof is divided in exposed and unexposed subgroups based on ascertained exposure to the infection (the explanatory variable) (Gerstman, 2003). Measures of association (e.g., risk difference, risk ratio, or attributable risk) may then be calculated after determining the frequency of the outcome variable (i.e., epilepsy or single unprovoked seizure) in exposed and unexposed fractions of the population.

An initial step in developing population-based association studies between a neurological infectious disorder and epilepsy comprises of the formulation of case definitions for the infection. The case definitions should be highly sensitive and specific and are traditionally based either on fairly characteristic symptoms and signs of the infectious disease or laboratory criteria for diagnosis of the infection. In population studies, laboratory evidence is often provided by serological studies performed upon biological samples (e.g., serum or saliva). The latter are objective, convenient, and inexpensive means of measuring exposure. Positive serological reactions are implicit with recent (IgM-based) and past (IgG-based) exposure to the infecting organism.

Microbiological Considerations

  1. Top of page
  2. Burden of CNS Infections
  3. Burden of Epilepsy
  4. Burden of Epilepsy Due to CNS Infections
  5. Epidemiological Approaches to CNS Infections and Epilepsy
  6. Microbiological Considerations
  7. Epileptological Considerations
  8. Problems in Studies of the Association between CNS Infections and Epilepsy
  9. Conclusions
  10. Acknowledgment
  11. References

The risk factors for epilepsy are many but for the purpose of this review may be divided into infectious and noninfectious conditions. The noninfectious causes include traumatic brain injury, cerebrovascular disease, tumors and genetic factors. The infectious risk factors differ inherently from noninfectious conditions because of their biological nature as opposed to chemical, environmental, or genetic nature of the latter. The biological nature of agents that cause infections confer certain special epidemiological characteristics to epilepsy associated with CNS infections as opposed to posttraumatic and poststroke epilepsy. These characteristics are outlined below.

A complex interaction exists among the infectious agent, the host, and the environment (Fig. 1). An understanding of this epidemiological triad is vital for appreciation of the mechanisms by which infective agents produce clinical symptoms and signs (e.g., seizures or epilepsy). Infective agents include prions, viruses, bacteria, fungi, and helminths. The biological life cycle of the infective agent, which may in certain cases, be rather complex, the mode and dynamics of transmission, and the natural history of the infectious disorder it produces in the human host determine the epidemiological proportions of the infectious disease.

image

Figure 1. The epidemiological triad operates in the causation of infectious diseases and includes the agent, host and environment. Examples of factors responsible for transmission of infectious agents are shown.

Download figure to PowerPoint

In population studies, exposure to the infectious agent is not necessarily synonymous with clinical disease and may not lead to infection at all. On the other hand, it may lead to subclinical disease with no overt clinical manifestations. The iceberg analogy (Fig. 2) is often used to describe the clinical spectrum of a variety of infectious disorders, including both epidemic encephalitides and more chronic endemic disorders such as Taenia solium-cysticercosis. For instance, during an epidemic of Japanese-B-encephalitis (JE) infection, symptomatic illness occurs in only 1 in 250 cases (Grossman et al., 1973). Furthermore, the majority of the symptomatic cases have only mild manifestations. Those that present with neurological manifestations and sequelae (including epilepsy) represent merely the tip of the iceberg. Likewise, in the case of cysticercosis, caused by the larval stage of the tapeworm, T. solium, community-based serological surveys reveal high levels of seropositivity in endemic populations. The majority of the seropositive individuals in the populations surveyed do not have epilepsy or other neurological symptoms due to cysticercosis or evidence of neurocystcercosis (NCC) on imaging studies (e.g., computed tomography [CT]). In the community, therefore, there is a large pool of individuals who are seropositive, suggesting exposure to the parasite, T. solium but no manifestations of NCC (Montano et al., 2005). Conversely, there are individuals in the community, who demonstrate evidence of neurological infestation by the parasite (i.e., NCC) on imaging studies but have no symptoms (or seizures). A study undertaken in an endemic community in Guatemala found evidence of NCC on imaging studies in 47% of those with seizures as well as 24% of asymptomatic individuals (Garcia-Noval et al., 1996). Indeed, CT undertaken in asymtomatic individuals in T. solium–endemic communities may frequently reveal single- or multiple-calcified specks (Montano et al., 2005). These calcified specs signify dead cysticerci, which presumably were live and active in the past but nonetheless did not manifest with symptoms (i.e., seizures) and hence remained subclinical. Thus, both, in the case of endemic infections and during epidemics, there is a proportion of individuals in the community, who are exposed to the infective organism but do not develop the infection (asymptomatic seroposivity) as well as those who get infected but do not develop symptoms and signs (subclinical cases) in addition to those who manifest with symptoms (seizures) in addition to imaging evidence of neurological involvement by the infecting organism as well as demonstrate immunological evidence of exposure (i.e., seropositivity correlating with symptoms). Finally, there may be individuals who present with seizures due to neurological infection but are seronegative (Montano et al., 2005). Presumably, the immunological response to the infective agent is insufficient to be detected by the serological assays in these individuals.

image

Figure 2. The iceberg analogy to the spectrum of certain CNS infectious disorders (e.g., arboviral encephalitis and T. solium cysticercosis).

Download figure to PowerPoint

Epileptological Considerations

  1. Top of page
  2. Burden of CNS Infections
  3. Burden of Epilepsy
  4. Burden of Epilepsy Due to CNS Infections
  5. Epidemiological Approaches to CNS Infections and Epilepsy
  6. Microbiological Considerations
  7. Epileptological Considerations
  8. Problems in Studies of the Association between CNS Infections and Epilepsy
  9. Conclusions
  10. Acknowledgment
  11. References

In relation to brain insults such as traumatic brain injury and cerebrovascular accidents, clear mechanistic and prognostic differences exist between early seizures (those that occur within 7 days and are also referred to as provoked or acute symptomatic seizures) and late (after 7 days) seizures (also referred to as, unprovoked or remote symptomatic seizures or epilepsy) in order to justify their separate analysis in terms of outcome and treatment. Early seizures as opposed to late unprovoked seizures do not require long-term antiepileptic drug treatment. The time-line drawn (of 7 days) for the distinction between early (provoked) and late (unprovoked) seizures is based on recommendations of the International League Against Epilepsy (ILAE)-Commission for Epidemiology and Prognosis (Commission on Epidemiology and Prognosis of the ILAE, 1993). Likewise, in the case of acute CNS infections such as bacterial meningitis, viral encephalitis, and cerebral malaria, the distinction between “early” and “late” seizures is both of semantic and pragmatic importance and fairly unproblematic, being simply based on the temporal relationship to onset and resolution of the infection. However, in the case of chronic or recurrent infectious disorders such as chronic meningitis (e.g., tubercular or fungal) or chronic granulomatous conditions (e.g., NCC or intracranial tuberculoma), for which the active phase of infection persists for long periods of time (in month-years), the differentiation between provoked and unprovoked seizures using a 7-days time-line is not possible. In chronic infectious disorders, seizures that occur as long as the infection is active are designated as provoked and those that occur following resolution of the active phase of infection are referred to as unprovoked.

Since CNS infectious disorders cause both acute and remote symptomatic seizures, it is pertinent to review the epidemiology and natural history of acute symptomatic seizures. In the Rochester community, roughly 15% of all incident acute symptomatic seizures were attributed to CNS infections, giving an age-adjusted incidence of acute symptomatic seizures due to CNS infection of 5.2/100,000 persons/year (Annegers et al., 1995). The mechanisms for seizures in CNS infection are many and are summarized in Table 1 (Vaughan & Delanty, 2001). The occurrence of acute symptomatic seizures due to CNS infection increases the risk for late unprovoked seizures or epilepsy but this risk is less than the average risk of a second seizure following a single unprovoked seizure (Berg & Shinnar, 1991; Hesdorffer et al., 1998). The risk may depend upon the underlying mechanism responsible for the acute seizure during the infection episode. If the acute infection episode leaves a structural residue involving the brain parenchyma, the 10-year risk may be 24% (Hesdorffer et al., 1998). Also, the risk increases substantially following status epilepticus in comparison to an isolated seizure during the CNS infection episode.

Table 1.  Mechanistic considerations in acute symptomatic seizures in various CNS infectious disorders (adapted from Vaughan & Delanty, 2001)
S. No.Mechanistic basisExample (Infectious disorder)
 1Focal, inflammatory and hemorrhagic cortical necrosisHerpes simplex encephalitis
 2Vasculitis and infarctionTubercular meningitis, aspergillosis, subarachnoid cysticercosis
 3Cortical granuloma/sCNS tuberculosis, parenchymal cysticercosis, aspergillosis
 4Cerebral abscessPyogenic, tubercular, and fungal abscess
 5Mass lesion/sHuman immunodeficiency virus infection
 6Neurotoxin mediated (cytokines, prostaglandins, free radicals)Various infectious disorders
 7Drug-inducedHuman immunodeficiency virus infection, CNS tuberculosis
 8Electrolyte imbalance (e.g., hyponatermia due to syndrome of inappropriate antidiuretic hormone secretion)Various (acute and chronic meningitis), HIV infection
 9Hyperthermia-inducedVarious
10Disruption of blood–brain barrierVarious
11Intracranial hypertensionVarious

Problems in Studies of the Association between CNS Infections and Epilepsy

  1. Top of page
  2. Burden of CNS Infections
  3. Burden of Epilepsy
  4. Burden of Epilepsy Due to CNS Infections
  5. Epidemiological Approaches to CNS Infections and Epilepsy
  6. Microbiological Considerations
  7. Epileptological Considerations
  8. Problems in Studies of the Association between CNS Infections and Epilepsy
  9. Conclusions
  10. Acknowledgment
  11. References

In comparison to epilepsy following CNS infectious disorders, the frequency, course, and outcome of posttraumatic seizures and epilepsy have been fairly well characterized (Frey, 2003). For instance, studies of the natural history of posttraumatic seizures reveal that the incidence of late posttraumatic seizures is 2–30%, remission rates for posttraumatic epilepsy are in the order of 25–40% and the estimated fraction that is likely to become intractable is 13% (Frey, 2003). Furthermore, several studies have shown that the incidence of posttraumatic seizures correlates with the severity of the traumatic brain injury as measured by several parameters including the Glasgow coma scale, duration of loss of consciousness, and the presence of posttraumatic amnesia and focal neurological deficits.

Tempting though it may be to extrapolate data regarding posttraumatic epilepsy to seizures and epilepsy associated with CNS infections, the likewise analysis of the latter, however, presents certain difficulties. There are no validated scales for assessing the severity of brain injury during CNS infections; therefore, correlation between the severity of infection and incidence of late seizures remains impractical. Another problem is that it may be difficult to differentiate between active and inactive phases of certain chronic CNS infectious disorders. For instance, it is not uncommon to find both active (involuting or degenerating) and inactive (fibro-calcified) parenchymal cysticerci at the same time in patients with NCC (Fig. 3). Because anatomic substrates for both provoked and unprovoked seizures coexist, the designation of seizures as provoked or unprovoked can be problematic. In such situations, careful correlation between the semiology of seizures and imaging findings with reference to the stage of the parasitic infestation is helpful in categorizing the seizures as provoked or unprovoked. A problematic issue in remote CNS infections is that the diagnosis of CNS infection in many cases is at best presumptive, especially so in cases of suspected viral encephalitis for which the etiological agent often remains unidentified. Finally, as in the case of posttraumatic seizures and epilepsy, the practice of administration of antiepileptic drugs at the time of active infection confounds the assessment of natural history of seizures due to CNS infection. Many incidence studies of late unprovoked seizures and epilepsy are hospital-based even more so in impoverished, developing countries. The patient sample in such studies is highly selected with overrepresentation of subjects with either severe illnesses or easy access to health-care on account of geographic and economic reasons.

image

Figure 3. Computed tomographic (CT) scan revealing the coexistence of live vesicular (arrow) and calcified cysticerci (arrowhead).

Download figure to PowerPoint

Studies of the association between CNS infections and epilepsy often have to be performed in resource-poor countries as the latter are host to many infectious disorders not found elsewhere. The conduct of these studies present complex logistic and ethical challenges. Living conditions, social services, and health infrastructure are often suboptimal in these countries. Acquisition of information suffers from the lack of an organized system for medical record-keeping and often inappropriate and incorrect reporting. In addition, there are limitations of funding sources, manpower to conduct research, and quality assurance programs. Linguistic and other intangible barriers may hold back effective research partnership between agencies and institutions in developed and developing countries. These include inadequate protection of human rights, understanding of scientific research, and the limited ability of people to provide informed consent.

Conclusions

  1. Top of page
  2. Burden of CNS Infections
  3. Burden of Epilepsy
  4. Burden of Epilepsy Due to CNS Infections
  5. Epidemiological Approaches to CNS Infections and Epilepsy
  6. Microbiological Considerations
  7. Epileptological Considerations
  8. Problems in Studies of the Association between CNS Infections and Epilepsy
  9. Conclusions
  10. Acknowledgment
  11. References

Further studies of association between CNS infections and epilepsy and of the natural history of seizures following CNS infection are desirable. These should address the incidence of late unprovoked seizures and epilepsy following each of the several neurological infectious disorders and the remission and intractability rates of epilepsy following CNS infections. The studies should ideally document the anatomical substrates of epilepsy associated with CNS infections with the help of contemporary imaging tools. The estimation of the burden of epilepsy due to CNS infections logically precedes the initiation of measures to effectively control and treat it. The ultimate aim is to reduce the burden of epilepsy by preventing infectious disorders through improvements in preventive health-care, the overall socio-economic climate, and the will to change in LDCs.

Acknowledgment

  1. Top of page
  2. Burden of CNS Infections
  3. Burden of Epilepsy
  4. Burden of Epilepsy Due to CNS Infections
  5. Epidemiological Approaches to CNS Infections and Epilepsy
  6. Microbiological Considerations
  7. Epileptological Considerations
  8. Problems in Studies of the Association between CNS Infections and Epilepsy
  9. Conclusions
  10. Acknowledgment
  11. References

We acknowledge Deepika Tikoo for help in preparation of the manuscript.

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. Gagandeep Singh received travel grants from Sanofi Aventis, Esai and Janssen Cilag. Sudesh Prabhakar declares no conflicts of interest.

References

  1. Top of page
  2. Burden of CNS Infections
  3. Burden of Epilepsy
  4. Burden of Epilepsy Due to CNS Infections
  5. Epidemiological Approaches to CNS Infections and Epilepsy
  6. Microbiological Considerations
  7. Epileptological Considerations
  8. Problems in Studies of the Association between CNS Infections and Epilepsy
  9. Conclusions
  10. Acknowledgment
  11. References