Are you ready for the child with status epilepticus (SE)? There is a good chance you are not fully prepared. The things that will be needed are:
* an understanding of the definitions and causes of SE and how SE can cause death or central nervous system (CNS) residua;
* a knowledge base about the different categories of SE and which are most dangerous;
* an understanding of how and why the diagnosis of SE can be missed; and
* ready availability of equipment and medications to manage SE.
This editorial discusses these issues.
Approximately one-third of patients with SE have idiopathic epilepsy and SE may be the first manifestation of epilepsy or occur after other seizures.1"3 Slightly more than half of SE events are symptomatic of another condition, such as head trauma, an electrolyte disturbance, or a CNS infection ("acute symptomatic" SE). The approximate percentages of SE children with various precipitating conditions are as follows3: febrile seizures - 36%; change in anticonvulsants - 20%; unknown - 9%; metabolic - 8%; congenital - 7%; anoxia - 5%; CNS infection - 5%; head trauma- 4%; cerebrovascular accident - 3%; poisoning, drug related- 2%; and CNS tumor1%.
Atypical febrile seizures can meet criteria for SE, and because febrile seizures are so common, they represent the largest group. Fortunately, SE from febrile seizures is much less damaging than that from most other acute symptomatic types, so it is generally classified separately.3,4 In addition, although focal seizures without loss of consciousness or recurrent petit mal seizures can last longer than 30 minutes and thus qualify as SE, they are also less dangerous and not the same medical emergency as an equal duration of continuous or repeated generalized seizures with loss of consciousness or persistent postictal depression.3
The mortality of SE in general remains significant, although it has dropped during the past 50 years to current death rates somewhat below 8% for children and 30% for adults. Most deaths are due to whatever produced an acute symptomatic SE, but as many as 3% of children and 10% of adults with SE seem to die from the SE itself. SE can thus be connected with morbidity and mortality in three ways. (1) Sequelae and seizures both may be caused by another condition (as above). (2) Seizures may produce systemic abnormalities that cause damage. (3) SE electrical activity and its disturbances inside the brain may cause death or disability. Dissecting which of these produces various components of injury is difficult in patients, but important for optimal management of SE.
MEANWHILE. IN THE LAB
Animal experiments should be reduced or avoided whenever the same information can be obtained in some less obtrusive way. But, animal experiments have helped sort this out and are a good example of how such studies can directly enhance clinical practice.5 By giving specific seizure-producing drugs, damage due to (2) and (3) can be studied without worrying about (1). Then, by reducing or eliminating systemic effects of SE (2), the injury that remains must be from (3).
For example, SE can be induced in baboons by giving bicuculline. During SE, brain metabolism increased as much as ninefold. Initial hypertension to systolic pressures above 200 mm Hg eventually gave way to hypotension. Blood pH transiently dropped from 7.4 to an average low of 6.6, whereas arterial PCO increased dramatically. There was venous dilatation and increased circulation in the brain. Serum bicarbonate levels fell, whereas lactate concentrations increased and eventually normalized. The animals became hyperpyrexic, some to as high as 43°C. The most significant electrolyte change was hyperkalemia. Arterial oxygen temporarily dropped to 70 to 80 mm Hg. Serum glucose initially rose, but later dropped to normal and eventually to severe hypoglycemia. Most of these systemic changes would be expected to produce cerebral edema or to cause or augment brain damage in other ways.
Confirming this, 10 baboons were subjected to prolonged SE and 5 had cellular injury in the cerebral cortex, hippocampus, thalamus, and cerebellum. This resembled ischemia and had similarities to autopsy findings of children who die of SE. Four of the 5 experienced herniation (which is rare in children with SE). The remaining 5 had no histologic brain abnormalities. To separate systemic effects (2) from direct seizure effects (3), the scientists compared the two groups for systemic effects and found that the 5 with CNS damage had been more hyperpyrexic during SE.
Next, they determined whether SE produced damage in paralyzed baboons on ventilators with oxygen. This intensive treatment blocks tonic-clonic muscle activity and eliminates or mutes most systemic effects. Such animals do not develop severe acidosis or hyperpyrexia above 400C and do not have hypoxia or hypotension, but still have the brain electrical disturbances of SE. This treatment spared the cerebellum completely (suggesting its damage was caused by systemic abnormalities), but histologic changes in the cortex, hippocampus, and thalamus were still found (suggesting these were at least partly from the seizures themselves). How seizures cause damage separate from systemic changes is not clear, but the abnormalities look ischemic, suggesting an intrinsic failure of cortical oxygen delivery. In summary, studies such as these provide evidence that SE can damage the cortex, hippocampus, and thalamus directly, that systemic effects could increase this damage, and that cerebellar damage is predominantly from systemic causes, especially hyperpyrexia.
Experimental animal models have also demonstrated that CNS damage can begin as early as 30 to 60 minutes after seizure onset. These findings led to a tightening of the duration that defines SE from ". . . seizures that persist for a sufficient length of time . . ." to those lasting more than 1 hour to the current 30-minute limit.1"3 Animal experiments also suggested that repeated seizures without recovering consciousness are as dangerous as or more dangerous than continuous seizures for the same duration. And, current theories about the pathogenesis of seizures are, in part, based on animal experiments demonstrating that seizures may be induced by drugs that either increase neuronal stimulation or block neuronal inhibition at neuronal cell membrane, postsynaptic receptor, or neurotransmitter levels.
MEANWHILE. IN THE OFFICE
The principles behind management of SE can be derived from these experiments as well as from clinical experience. Time is of the essence and any generalized convulsive seizure lasting longer than 10 minutes should be treated with anticonvulsants. Besides anticonvulsants, patients should have an IV when possible and should be monitored for (and treated for) abnormalities of blood pressure, serum electrolytes and glucose, hyperpyrexia, and arterial oxygen, CO2, bicarbonate, and pH as determined by blood gas.
What other pitfalls are there for the pediatrician? First, it is important to recognize SE. This sounds easy but after prolonged seizures, motor activity may be subdued so that the patient appears to have stopped seizing but remains unconscious. Such patients can still be in SE. The same goes for those paralyzed for ventilator management, although here it may also be impossible to judge level of consciousness. In these situations, an emergency EEG is the best way to tell whether the patient is in SE.
Do you have the drugs and equipment to manage SE, as defined in this issue? There is evidence that appropriate prehospital management makes control easier and faster after hospitalization.2,3 Besides O, suction, and at least an appropriate mask and bag, you will need a way to deliver anticonvulsants for rapid action. Optimally, the drugs and equipment to give these intravenously are available. But surveys in 1990 and 1991 demonstrated that only 55% to 66% of offices had the supplies to start an IV.6 If so, your choices are rectal diazepam or IM fosphenytoin, as described in the article by Bebin in this issue. Many offices may have midazolam for sedation, and limited experience indicates that seizures can often be brought under rapid control by giving 0.2 mg /kg to a maximum 7 mg intramuscularly.6 And, do not forget that most anticonvulsive agents, especially the benzodiazepines and phénobarbital, occasionally produce apnea or respiratory depression, so be ready to resuscitate.
1. Lowenstein DH, Alldredge BK. Status epilepticus. Massachusetts Medical Society. 1998;338:970-977.
2. Delgado-Esqueta AV, Waasterlain C, Treiman DM, Porter RJ. Management of status epilepticus: current concepts in neurology. N Engl J Med. 1982;306:1337-1340.
3. Dodson WE, DeLorenzo RJ, Padley TA, Shinnar S, Treiman DM, Wannamaker BB. Treatment of convulsive status epilepticus: recommendations of the Epilepsy Foundation of America's Working Group on Status Epilepticus. ]AMA. 1993;270:854-859.
4. Verity CM, Ross EM, Golding J. Outcome of childhood status epilepticus and lengthy febrile convulsions: findings of national cohort study. BMJ. 1993;307:225-228.
5. Dean JM, Senger HS. Status epilepticus. In: Rogers MC, ed. Textbook of Pediatric Intensive Care. Baltimore: Williams & Wilkins; 1987:615-627.
6. Chamberlain JM, Alteri MA, Futterman C, Young GM, Ochsenschlager DW, Waisman Y. A prospective, randomized study comparing intramuscular midazolam with intravenous diazepam for the treatment of seizures in children. Pediatr Emerg Care. 1997;13: 92-94.