Clinically relevant pathogens

Following is a taxonomic classification of neurologically relevant bacterial pathogens:

  • Aerobic gram-positive cocci
  • Streptococcus pneumoniae
  • Streptococcus agalactiae (group B)
  • Enterococcus
  • Aerobic gram-positive rods
  • Listeria monocytogenes
  • Aerobic gram-positive cocci
  • Neisseria meningitidis
  • Aerobic gram-negative rods
  • Escherichia coli
  • Haemophilus influenzae (type b)
  • Legionella pneumophila
  • Salmonella
  • Pseudomonas aeruginosa
  • Anaerobic gram-positive cocci
  • Peptostreptococcus
  • Anaerobic-gram positive rods
  • Actinomyces
  • Propionibacterium
  • Clostridium botulinum, Clostridium tetani, Clostridium perfringens
  • Anaerobic gram-negative rods
  • Bacteroides fragilis
  • Spirochetes
  • Borrelia burgdorferi (Lyme disease)
  • Treponema pallidum (syphilis)
  • Intracellular
  • Rickettsia (Rocky Mountain spotted fever)

The relationship of bacterial infections and epilepsy

A number of studies have investigated epileptic complications associated with bacterial infection of any type or severity. Iavanainen, et al. found that increased serum levels of various bacterial antibodies were more common in patients with recent seizures than in healthy control subjects (17 of 29 vs. 2 of 31; p <.001).6 In most of these cases, no infections were recognized clinically or bacteriologically. It is unclear whether seizure production was related to the bacterial infection or to a nonspecific immune response. Besides bolstering long-standing theories regarding the epileptogenic effect of infection in general, these results also suggest that epileptic seizures can be triggered by bacterial infections even when no clinically apparent infection is recognized.

During an investigation of an animal model for the study of the neurotoxicity of bacterial products, pentylenetetrazol (PTZ), a convulsant drug, was injected into mice.7 Increased seizure induction sensitivity to PTZ was used as an indicator of neurotoxicity. If the mice were given a preinjection of sonicates of Shigella dysenteriae or Escherichia coli, the seizure response to PTZ was enhanced: the mean convulsion score was higher, as was the number of mice responding to PTZ. The induction of seizures in animals pretreated with a subepileptic dose of PTZ was also increased.

In an attempt to formulate a quantitative means of predicting seizure risk in patients with serious acute infection, Guess and colleagues developed a predictive score of clinical risk for seizure occurrence based on records of intensive care unit patients with gram-negative infections.8 They devised an inventory of clinical risk factors with individual seizure risk weights (derived from regression coefficients determined by the frequency with which each clinical condition was associated with seizure occurrence in retrospective chart review):


Factor Risk index
Before admission  
 Prior seizure history 2
 CNS tumor history 4
Concurrent factors (within 1 month)  
 Acute stroke (ischemic or hemorrhagic) 3
 CNS surgery 3
 CNS infection 3
While in intensive care unit  
 Coma or anoxic encephalopathy 1
 Renal impairment (serum creatinine >1.6 mg/dL) 2
 Acute hypotensive episode 1


They considered an index of 3 or more to be suggestive of relatively high seizure risk, whereas an index of 2 or less was considered low seizure risk.

To evaluate the predictive value of this scoring system, the authors applied it to a separate population of seriously ill inpatients with similar gram-negative infections. Seizure risk index scores and actual seizure occurrences were then compared. Positive predictive values were low (26%) but negative predictive values were high (³97%). Thus, this index better identifies patients at low risk for seizures than it predicts seizure occurrence in individual patients.

In another study,9 about 40% of patients with encephalopathy related to bacterial sepsis manifested seizures. Systemic sepsis may cause seizures by disturbing cortical function (e.g., synaptic transmission, cerebral energy production) via diffuse toxic or metabolic mechanisms. Autopsies revealed other potential causes of seizures, however:

  • bacterial intraparenchymal invasion with microabscess formation
  • cortical proliferation of astrocytes and microglia
  • microinfarcts
  • multiple small white-matter hemorrhages
  • central pontine myelinolysis

Seizure Complications of Bacterial Endocarditis

Neurologic complications occur in about 30% of all patients with bacterial endocarditis, and seizures are not uncommon. In one series of 218 patients with endocarditis, 39% had a neurologic complication. Of these, 58% died, versus 20% mortality among endocarditis patients without neurologic complications.

Cerebral embolism is the most frequent and most clinically significant complication. It is particularly common in patients with mitral valve infection and with endocarditis due to virulent organisms (e.g., Staphylococcus aureus and enteric gram-negative bacilli). Emboli are important not only for their direct morbidity via cerebral infarction, but also because of their role as epileptogenic foci in the form of mycotic aneurysms and brain abscesses. Focal seizures are more commonly associated with acute emboli.29 Generalized seizures have diverse etiologies, with systemic (e.g., metabolic) and iatrogenic (e.g., drug-induced) seizures being most frequent.30

Table adapted from HA Guess, LJ Resseguie, LJ Melton, et al. Factors predictive of seizures among intensive care unit patients with gram negative infections. Epilepsia 1990;31:567–573.
Adapted from: Goldstein MA and Harden CL. Infectious states. In: Ettinger AB and Devinsky O, eds. Managing epilepsy and co-existing disorders. Boston: Butterworth-Heinemann; 2002;83-133. 
With permission from Elsevier (

Reviewed By: 
Steven C. Schachter, MD
Sunday, February 29, 2004