Decerebrate posturing following traumatic brain injury: MRI findings and their diagnostic value.

AIM: To determine the pathomorphological and clinical background to decerebrate posturing in humans following serious traumatic brain injury. MATERIALS AND METHODS: One hundred and twenty patients who had been unconscious for more than 24 h underwent diagnostic MRI within 8 days after trauma. The presence of decerebrate rigidity as the clinical parameter was correlated to MRI findings, such as traumatic lesions in defined brain areas. Significance was presumed as p < 0.05. RESULTS: On the day of MRI 43 (36%) patients exhibited decerebrate posturing: 19 (23%) cases were unilateral and 24 (77%) bilateral. There was a significant correlation between midbrain lesions and the presence of rigidity. If a midbrain lesion was found in the absence of pontine lesions, decerebrate rigidity could be concluded (p < 0.05). There was no significant correlation to the rigidity in the case of midbrain lesions accompanied by pontine lesions, and no correlation to the rigidity could be detected for other regions of the brain. Both the occurrence of decerebrate posturing and the detection of brainstem lesions at MRI correlated with the Glasgow Outcome Scale. The combination of both parameters improved the probability of predicting the outcome. CONCLUSION: The rate of decerebrate posturing increases significantly in the presence of midbrain lesions. The presence of pontine lesions appears to be of secondary importance. The chances of predicting the Glasgow Outcome Scale are improved by the combination of clinical information (decerebrate posturing) and radiological parameters (type of brainstem lesion).

A filmed hanging without decerebrate and decorticate rigidity: a case report and pathophysiological considerations.

The study of filmed hangings in the recent has reveal a striking similarity in the agonal sequences, with a sequence of rapid loss of consciousness, convulsions, and then a complex pattern of decerebrate rigidity and decorticate rigidity. We report a case of filmed hanging not presenting with decerebrate and decorticate rigidity. A 52-year-old man stepped off a stool, hanging himself in a complete suspension in an upright position. The movement of the body stepping off the stool created a rotary movement around the ceiling's ring and the body of the hanging man immediately started to revolve around the ring. Apart for the rolling around the ceiling's ring, the body stayed motionless for the duration of the movie, without any evidence of decerebrate or decorticate rigidity. A review of the pathophysiology of these reflex posturing gives some insight as to the possible elucidation for their absence in this specific case.

Midbrain and medullary control of postinspiratory activity of the crural and costal diaphragm in vivo.

Studies on brain stem respiratory neurons suggest that eupnea consists of three phases: inspiration, postinspiration, and expiration. However, it is not well understood how postinspiration is organized in the diaphragm, i.e., whether postinspiration differs in the crural and costal segments of the diaphragm and what the influence is of postinspiratory neurons on diaphragm function during eupnea. In this in vivo study we investigated the postinspiratory activity of the two diaphragm segments during eupnea and the changes in diaphragm function following modulation of eupnea. Postinspiratory neurons in the medulla were stereotaxically localized extracellularly and neurochemically stimulated. We used three types of preparations: precollicularly decerebrated unanesthetized cats and rats and anesthetized rats. In all preparations, during eupnea, postinspiratory activity was found in the crural but not in the costal diaphragm. When eupnea was discontinued in decerebrate cats in which stimulation in the nucleus retroambiguus induced activation of laryngeal or abdominal muscles, all postinspiratory activity in the crural diaphragm was abolished. In decerebrate rats, stimulation of the midbrain periaqueductal gray abolished postinspiration in the crural diaphragm but induced activation in the costal diaphragm. In anesthetized rats, stimulation of medullary postinspiratory neurons abolished the postinspiratory activity of the crural diaphragm. Vagal nerve stimulation in these rats increased the intensity of postinspiratory neuronal discharge in the solitary nucleus, leading to decreased activity of the crural diaphragm. These data demonstrate that three-phase breathing in the crural diaphragm during eupnea exists in vivo and that postinspiratory neurons have an inhibitory effect on crural diaphragm function.

Propagation of sinusoidal electrical waves along the spinal cord during a fictive motor task.

We present for the first time direct electrophysiological evidence of the phenomenon of traveling electrical waves produced by populations of interneurons within the spinal cord. We show that, during a fictive rhythmic motor task, scratching, an electrical field potential of spinal interneurons takes the shape of a sinuous wave, "sweeping" the lumbosacral spinal cord rostrocaudally with a mean speed of approximately 0.3 m/s. We observed that traveling waves and scratching have the same cycle duration and that duration of the flexor phase, but not of the extensor phase, is highly correlated with the cycle duration of the traveling waves. Furthermore, we found that the interneurons from the deep dorsal horn and the intermediate nucleus can generate the spinal traveling waves, even in the absence of motoneuronal activity. These findings show that the sinusoidal field potentials generated during fictive scratching could be a powerful tool to disclose the organization of central pattern generator networks.

Disruption of cutaneous feedback alters magnitude but not direction of muscle responses to postural perturbations in the decerebrate cat.

Quadrupeds and bipeds respond to horizontal perturbations of the support surface with muscular responses that are broadly tuned and directionally sensitive. Since the discovery of this directional sensitivity, interest has turned toward the critical sensory systems necessary to generate these responses. We hypothesize that cutaneous feedback affects the magnitude of muscle responses to postural perturbation but has little effect on the directionality of the muscle response. We developed a modified premammillary decerebrate cat preparation to evaluate the sensory mechanisms driving this directionally sensitive muscle activation in response to support surface perturbation. This preparation allows us the flexibility to isolate the proprioceptive (cutaneous and muscle receptors) system from other sensory influences. We found that loss of cutaneous feedback leads to a significant loss in background activity causing a smaller muscular response to horizontal perturbations. However, the directional properties of the muscular responses remained intact.

Distribution of dendrites from longissimus lumborum motoneurons stained intracellularly with biocytin in adult cats.

The three-dimensional distribution of dendrites from motoneurons innervating longissimus lumborum (Long Motoneurons) in the L4 spinal segment was examined in the adult cat using intracellular staining techniques. Long Motoneurons were electrophysiologically identified, stained with injection of biocytin and reconstructed from serial histological sections. Somas of Long Motoneurons were mainly located in the lateral-ventral area of the ventral horn. The dendritic distribution followed an orderly pattern in all motoneurons examined. Long Motoneurons showed a multi-directional distribution of dendrites, and the dendritic distribution pattern varied depending on the motoneuron. All studied motoneurons distributed dendrites from the spine into the white matter. The most significant morphological characteristic of the Long Motoneurons was the variation in their dendritic distribution. No relationship was observed between the effects of peripheral afferent inputs from the hindlimb and morphological characteristics of motoneurons.

Caudal brainstem processing is sufficient for behavioral, sympathetic, and parasympathetic responses driven by peripheral and hindbrain glucagon-like-peptide-1 receptor stimulation.

The effects of peripheral glucagon like peptide-1 receptor (GLP-1R) stimulation on feeding, gastric emptying, and energetic responses involve vagal transmission and central nervous system processing. Despite a lack of studies aimed at determining which central nervous system regions are critical for the GLP-1R response production, hypothalamic/forebrain processing is regarded as essential for these effects. Here the contribution of the caudal brainstem to the control of food intake, core temperature, heart rate, and gastric emptying responses generated by peripheral delivery of the GLP-1R agonist, exendin-4 (Ex-4), was assessed by comparing responses of chronic supracollicular decerebrate (CD) rats to those of pair-fed intact control rats. Responses driven by hindbrain intracerebroventricular (fourth i.c.v) delivery of Ex-4 were also evaluated. Intraperitoneal Ex-4 (1.2 and 3.0 microg/kg) suppressed glucose intake in both CD rats (5.0+/-1.2 and 4.4+/-1.1 ml ingested) and controls (9.4+/-1.5 and 7.7+/-0.8 ml ingested), compared with intakes after vehicle injections (13.1+/-2.5 and 13.2+/-1.7 ml ingested, respectively). Hindbrain ventricular Ex-4 (0.3 microg) also suppressed food intake in CD rats (4.7+/-0.6 ml ingested) and controls (11.0+/-2.9 ml ingested), compared with vehicle intakes (9.3+/-2.1 and 19.3+/-4.3 ml ingested, respectively). Intraperitoneal Ex-4 (0.12, 1.2, 2.4 microg/kg) reduced gastric emptying rates in a dose-related manner similarly for both CD and control rats. Hypothermia followed ip and fourth i.c.v Ex-4 in awake, behaving controls (0.6 and 1.0 C average suppression) and CD rats (1.5 and 2.5 C average suppression). Intraperitoneal Ex-4 triggered tachycardia in both control and CD rats. Results demonstrate that caudal brainstem processing is sufficient for mediating the suppression of intake, core temperature, and gastric emptying rates as well as tachycardia triggered by peripheral GLP-1R activation and also hindbrain-delivered ligand. Contrary to the literature, hypothalamic/forebrain processing and forebrain-caudal brainstem communication is not required for the observed responses.

Postural performance in decerebrated rabbit.

It is known that animals decerebrated at the premammillary level are capable of standing and walking without losing balance, in contrast to postmammillary ones which do not exhibit such behavior. The main goals of the present study were, first, to characterize the postural performance in premammillary rabbits, and, second, to activate the postural system in postmammillary ones by brainstem stimulation. For evaluation of postural capacity of decerebrated rabbits, motor and EMG responses to lateral tilts of the supporting platform and to lateral pushes were recorded before and after decerebration. In addition, the righting behavior (i.e., standing up from the lying position) was video recorded. We found that, in premammillary rabbits, responses to lateral tilts and pushes were similar to those observed in intact ones, but the magnitude of responses was reduced. During righting, premammillary rabbits assumed the normal position slower than intact ones. To activate the postural system in postmammillary rabbits, we stimulated electrically two brainstem structures, the mesencephalic locomotor region (MLR) and the ventral tegmental field (VTF). The MLR stimulation (prior to elicitation of locomotion) and the VTF stimulation caused an increase of the tone of hindlimb extensors, and enhanced their responses to lateral tilts and to pushes. These results indicate that the basic mechanisms for maintenance of body posture and equilibrium during standing are present in decerebrated animals. They are active in the premammillary rabbits but need to be activated in the postmammillary ones.

Leptin responsiveness in chronically decerebrate rats.

Peripheral infusions of physiological doses of leptin decrease body fat mass, but it is not known whether this results from direct effects on peripheral tissue or activation of central leptin receptors. In this study, we infused chronically decerebrate (CD) rats, in which the forebrain was surgically isolated from the caudal brainstem, with 60 microg leptin/d or PBS for 14 d from ip mini-osmotic pumps. The CD rats were tube fed an amount of food equivalent to the intake of ad libitum-fed intact controls or 75% of this amount to account for their reduced energy expenditure. Control rats fed ad libitum or tube fed 75, 100, or 125% of their ad libitum intake also were peripherally infused with leptin or PBS. CD rats had a lower serum testosterone, energy expenditure, and lean body mass compared with controls but had increased levels of adiponectin and leptin and were obese. Leptin increased body fat and decreased energy expenditure during the light period in 100%-fed CD rats, but not 75%-fed CD rats. Leptin decreased body fat of ad libitum- and 100%-fed but not 75%-fed or 125%-fed intact controls. Energy expenditure did not change in any control group. These results show that leptin can change body fat independent of a change in food intake or energy expenditure, that the forebrain normally prevents leptin from inhibiting energy expenditure through mechanisms initiated in the caudal brainstem or peripheral tissues, and that the leptin response in both intact and CD rats is determined by the energy status of the animal.

Topography of spinal neurons active during hindlimb withdrawal reflexes in the decerebrate cat.

There exists a spatial organization of receptive fields and a modular organization of the flexion withdrawal reflex system. However, the three dimensional location and organization of interneurons interposed in flexion reflex pathways has not been systematically examined. We determined the anatomical locations of spinal neurons involved in the hindlimb flexion withdrawal reflex using expression of the immediate early gene c-fos and the corresponding FOS protein. The flexion withdrawal reflex was evoked in decerebrate cats via stimulation of the tibial or superficial peroneal nerve. Animals that received stimulation had significantly larger numbers of cells expressing FOS-like immunoreactivity (42.7+/-2.3 cells/section, mean+/-standard error of the mean) than operated unstimulated controls (18.6+/-1.4 cells/section). Compared with controls, cells expressing FOS-like immunoreactivity were located predominantly on the ipsilateral side, in laminae IV-VI, at L6 and rostral L7 segments, and between 20% and 60% of the distance from the midline to the lateral border of the ventral gray matter. Labeled neurons resulting from tibial nerve stimulation were medial to neurons labeled following superficial peroneal nerve stimulation in laminae I-VI, but not VII. The mean mediolateral positions of labeled neurons from both nerves shifted medially as the transverse plane in which they were viewed was moved from rostral to caudal and as the coronal plane in which they were viewed was moved from dorsal to ventral. The mediolateral separation between populations of labeled cells was consistent with primary afferent projections and the location of reflex encoders. This topographical segregation corresponding to different afferent inputs is a possible anatomical substrate for a modular organization of the flexion withdrawal reflex system.

Purkinje cell activity during classical eyeblink conditioning in decerebrate guinea pigs.

Purkinje cells are the sole output from the cerebellar cortex and play a critical role during classical eyeblink conditioning. The present study revealed for the first time a learning-related change in individual Purkinje cell activity during successive eyeblink conditioning in decerebrate guinea pigs which permitted continuous single unit recording from the simplex lobe of the cerebellar cortex. The pair-conditioned group received paired presentation of the conditioned stimulus (CS) and unconditioned stimulus (US) until the frequency of the conditioned response (CR) exceeded 80%. The control group received a comparable number of the CS and US in a pseudorandom fashion. Responses of Purkinje cells to the CS were classified into four types: excitatory, inhibitory, a combination of the two, or no response. Approximately half of the recorded cells from both groups changed their response type at various conditioning stages. The firing frequency of a Purkinje cell to the CS showed a tendency to decrease in the pair-conditioned group, while it had a tendency to increase in the pseudoconditioned group. This learning-related difference in change of response type was attributable to a difference in the change between the no response and the inhibitory response types. Correlation analysis of the temporal pattern between the neural activity and the CR revealed that most of the cells that developed an inhibitory response by paired conditioning acquired the CR-related temporal pattern. These results suggest that the learning-related Purkinje cells gain an inhibitory response with a temporal pattern correlated with the CR topography.

Frequency organization of the dorsal cochlear nucleus in cats.

Sensory epithelia are often spatially reiterated throughout their representation in the central nervous system. Differential expression of this representation can reveal specializations of the organism's behavioral repertoire. For example, the nature of the central representation of sound frequency in the auditory system has provided important clues in understanding ecological pressures for acoustic processing. In this context, we used electrophysiological techniques to map the frequency organization of the dorsal cochlear nucleus in nine cats. Frequency responses were sampled in increments of 100-200 microns along electrode tracks that entered the dorsomedial border of the nucleus and exited at the ventrolateral border. Electrode tracks were oriented parallel to the long (or strial) axis of the nucleus so that each penetration sampled neural responses for most of the cat's audible frequencies and remained in or near the pyramidal cell layer for several millimeters. Nearly identical distance versus frequency relationships were obtained for different rostral-caudal locations within the same cat as well as for different cats. Frequency responses systematically decreased from above 50 kHz at the most dorsomedial locations in the nucleus to below 1 kHz in the most ventrolateral regions. The rate of frequency change was roughly three times greater in high frequency regions than in low frequency regions. In addition, the highest pyramidal cell density and longest rostral-caudal axis was observed for the middle third of the dorsal-ventral axis of the nucleus. As a result, roughly half of all pyramidal cells responded to frequencies between 8-30 kHz. The representation of neural tissue for these frequencies may be related to the importance of spectral cues in sound locations.

Spatiotemporal response properties of cerebellar Purkinje cells to neck displacement.

The activity of 184 Purkinje cells and 58 unidentified neurons located within the cerebellar anterior vermis was recorded in decerebrate cats during wobble of the body under a fixed head. This stimulus induced a neck displacement of constant amplitude (2.5 degrees) whose direction rotated at the constant velocity of 56.2 degrees/s on the horizontal plane, both in the clockwise and counterclockwise directions. It was then possible to evaluate the spatiotemporal characteristics of unit responses to neck displacement in the vertical planes; 131 of 184 Purkinje cells (71%) and 35 of 58 unidentified cells (60%) responded to clockwise and/or counterclockwise rotations. In particular, among the responsive units, 44% of the Purkinje cells and 37% of the unidentified cells showed an equal amplitude modulation during clockwise and counterclockwise rotations. These units are expected to show a maximal response sensitivity for neck displacement in a preferred direction, a null response for perpendicularly oriented stimuli and a constant temporal phase (narrowly tuned neurons). In 28% of the Purkinje cells and 40% of the unidentified cells, responses of different amplitudes were observed during clockwise and counterclockwise rotations. These neurons should display a preferred direction of response to neck displacement, lack of null response directions and a temporal phase changing with the stimulus direction (broadly tuned neurons). Finally, 27% of the Purkinje cells and 23% of the unidentified cells responded only to wobble in the clockwise or counterclockwise direction (unidirectional units). This behavior predicts equal sensitivities for all the directions of neck displacement and a response phase changing linearly with the direction of neck displacement. A maximal sensitivity vector (Smax), aligned with the preferred direction of the neuron, was evaluated for the bidirectional narrowly tuned and broadly tuned units. Its amplitude and temporal phase corresponded to the response characteristics expected for stimuli in the preferred direction of the cell. Smax directions were distributed over the horizontal plane. Most of them, however, were closer to the pitch than to the roll axis and pointed towards the animal's tail. Among pitch-related Purkinje cells, the temporal phase of Smax was small with a predominance of phase lags; phase leads of rather large amplitude were usually observed for roll-related Purkinje cells. The possibility that the recorded population of units coded the direction of neck displacement was tested by assuming that each cell gave a vectorial contribution related to its response properties and that the vectorial sum of such contributions represented the outcome of the population code. Dynamic body-to-head displacements in four different directions were simulated and for each direction 12 population vectors were evaluated at regular intervals of the stimulus cycle. The direction of the population vector was related to that of the stimulus, but the correspondence was close only for the pitch direction. Moreover, the amplitude of the population vector depended upon the direction of the stimulus, being larger for pitch than for roll displacements. Due to the efferent connections of the explored cerebellar region, the neuronal signals generated by the Purkinje cell population are probably transferred to the spinal cord, where they may differentially affect the amplitude and the spatial properties of the neck reflexes according to the direction of neck displacement.

Changes in fusimotor activity during repetitive lengthening muscle contractions in decerebrate cats.

Responses of fusimotor neurons to lengthening vs isometric contractions have been studied in decerebrate cats. Spike discharges of fusimotor neurons to the medial gastrocnemius muscle were recorded from this muscle nerve filament during sequences of contractions and/or stretches of the lateral gastrocnemius and soleus muscles. The sequences lasted for 250-450s (duty cycle 4:2 s). Isometric contractions were elicited by electrical stimulation (40 Hz, 1.3 times motor threshold) of the muscle nerves. Lengthening contractions were elicited in the same way while the muscles were stretched by 4 mm at a velocity of 1 mm/s. Of 25 fusimotor neurons studied, 23 responded to muscle contractions with an increase in firing rate, subsiding towards the end of the sequence. The increase was either modulated with each subsequent contraction or smooth throughout the sequence. Approximately 64% of fusimotor neurons, responding to muscle contractions, responded in a similar way to the sequences of muscle stretches, applied alone. Responses to sequences of the lengthening contractions were significantly larger, on average, than those to the isometric ones, but smaller than the sum of the responses to the contractions and stretches applied separately. On the other hand, they were also larger in fusimotor units, showing no overt responses to muscle stretches alone.

Cardiovascular responses during spontaneous overground locomotion in freely moving decerebrate cats.

To examine whether the cerebrum is essential for producing the rapid cardiovascular adjustment at the beginning of overground locomotion, we examined heart rate (HR), mean arterial blood pressure (MAP), and integrated electromyogram (iEMG) of the forelimb triceps brachialis muscle in freely moving decerebrate cats during locomotion. Two to four days after decerebration surgery performed at the level of the precollicular-premammillary body, the animals spontaneously produced coordinated overground locomotion, supporting body weight. HR began to increase immediately before the onset of iEMG, and MAP began to rise almost simultaneously with the iEMG onset. Their increases in HR and MAP (24 +/- 3 beats/min and 22 +/- 4 mmHg) were sustained during locomotion. Sinoaortic denervation (SAD) did not affect the abrupt changes in HR and MAP at the beginning of locomotion (0-4 s from the onset of iEMG), whereas SAD had a contrasting effect during the subsequent period, a decrease in the HR response (9 +/- 1 beats/min) and an increase in the MAP response (30 +/- 3 mmHg). These results suggest that the cerebrum and the rostral part of the diencephalon are not essential for producing the rapid cardiovascular adjustment at the beginning of spontaneous overground locomotion. The arterial baroreflex does not contribute to this rapid adjustment but plays an important role in regulating the cardiovascular responses during the later period of spontaneous locomotion.

Decerebration by global ischemic stroke in rats.

In this study we present a fast and simple technique to decerebrate rats. By injecting polyvinylsiloxane (PVS) into both common carotid arteries we occluded the circle of Willis and all cerebral arteries, thereby completely interrupting the blood supply to the cerebrum. High viscosity PVS was used so that it only entered the large arteries, and did not pass into the capillary beds. This procedure reliably resulted in an anemic decerebration, without interfering with the blood supply to the brainstem. Long-term survival was achieved by reducing intracranial pressure by the application of steroids and/or opening the skull.

A comparison of cathepsin B processing and distribution during neuronal death in rats following global ischemia or decapitation necrosis.

The objective of this study was to examine the possible role of the cysteine protease cathepsin B (E.C. 3.4.22.1) in the delayed neuronal death in rats subjected to the two-vessel occlusion model of global ischemia. Immunohistochemistry of the hippocampus showed an alteration in the distribution of cathepsin B in CA1 neurons from a lysosomal pattern to a more intense label redistributed into the cytoplasm. This change was not detected until the neurons had become morphologically altered with obvious shrinkage of the cytoplasmic region. Western blotting and enzyme activity measurements of subcellular fractions, including lysosomes and a cell soluble fraction, demonstrated that there was an overall decrease in cathepsin B activity at this time but an increase in the proenzyme form, particularly in the soluble fraction. This was found to be completely different from the marked loss of all forms of cathepsin B in necrotic neurons following decapitation.

The effects of strychnine, bicuculline, and ketamine on 'immersion-inhibited' dorsal horn convergent neurons in intact and spinalized rats.

In both intact and spinalized rats, this study examined the effects of strychnine (a glycine antagonist), bicuculline (a GABAA antagonist), and ketamine (a non-competitive NMDA receptor antagonist) on one particular class of lumbar dorsal horn convergent neurons. This group of convergent neurons are inhibited when a rat's entire ipsilateral hindpaw is immersed in 50 degrees C water and has a strong afterdischarge as soon as the paw is removed from the water. Strychnine (2 mg/kg, iv) increased ongoing activity and blocked the 'inhibition phase' in both intact and spinalized rats demonstrating that a spinal-related glycine mechanism was involved in the inhibition. However, only in intact rats did the firing rate of the 'afterdischarge phase' increase significantly from pre-drug levels, suggesting that supraspinal sites may be involved in modulating this phase. Ketamine (15 mg/kg, iv) depressed ongoing activity and the firing rate in the afterdischarge phase of these neurons. Additionally, ketamine reversed the strychnine-induced increase in ongoing activity. Bicuculline (2 mg/kg, iv) had no effect on the activity of this cell class. As shown previously, and replicated here, these 'immersion-inhibited' neurons invariably have both inhibitory and excitatory mechano-receptive fields on the ipsilateral hindpaw. Thus, the response of this class of convergent neurons to noxious stimulation may be a function of relative inputs of glycine and EAA's, each possibly triggered by the stimulation of different receptive fields/regions on the same paw. Furthermore, when both fields are co-stimulated during noxious immersion of the entire paw, glycine has a stronger influence on activity than does the EAA's.

Pressor responses to stimulation of non-NMDA receptors in the superficial laminae of the cat spinal cord.

Microinjection of L-glutamate (109 mM; 12-18 nl) or AMPA (150-300 microM; 12 nl) into the superficial laminae of the L7 dorsal horn of decerebrate or chloralose anesthetized cats significantly increased mean arterial pressure. In contrast, microinjection of NMDA (300 microM; 12 nl) had no effect on mean arterial pressure. The pressor response to L-glutamate microinjection was blocked by prior microinjection of CNQX, an antagonist to non-NMDA receptors, but not by AP-5, an antagonist to NMDA receptors. We conclude that stimulation of non-NMDA receptors in the superficial laminae of the lumbar dorsal horn increases arterial blood pressure.

Seizures in Japanese encephalitis.

OBJECTIVE: Although Japanese encephalitis (JE) has been reported to be associated with seizures, there is no report on its frequency, pattern and severity and their correlation with electroencephalography (EEG) and radiological findings. In this communication, the analysis of seizures in 65 JE patients seen during 1991-1999 is presented. METHODS: The diagnosis of JE was based on clinical, radiological and serological criteria. The patients were subjected to a detailed clinical evaluation. Frequency and pattern of seizures and other neurological findings were recorded. Cranial CT scan or MRI was carried out in all the patients. Electroencephalograph was recorded using the 10-20 system of electrode placement. Patients' outcome was defined, on the basis of clinical status for 3 months, into poor, partial and complete recovery. The relationship between seizures and other clinical, radiological and EEG parameters was evaluated by the Z test of proportion using the two-tail approach. RESULTS: The patient's age ranged between 2 and 57 years, 24 were females and 19 were below 13 years of age. Thirty patients had seizures within the first week of encephalitis. The seizures were generalised tonic-clonic in 17 patients and partial motor with secondary generalization in 13 patients. Eleven patients had a single seizure, 8 had two seizures and 11 had multiple seizures. Two patients were presented with status epilepticus. Interictal EEG in the patients with seizures revealed theta to delta slowing in all patients and epileptiform discharges in four patients only. MRI revealed thalamic lesions bilaterally in 24, cortical in 7, basal ganglia in 8 and brainstem in 3 patients out of 26 patients with seizure. In the seizure group, three patients died, nine had poor, eight had partial and nine had complete recovery. Upon comparison of the JE patients with and without seizures, Glasgow coma scale (GCS) score, focal weakness, EEG slowing and cortical and thalamic lesion on CT or MRI were significantly related to the occurrence of the seizures; however, it was not associated with poor outcome. CONCLUSIONS: JE is associated with seizures in 46% of the patients in the acute stage of encephalitis which is easily controlled by monotherapy. Patients with severe encephalitis were associated with higher frequency of seizures.