- Email: [email protected]
Synergistic effect of status epilepticus and ischemic brain injury on mortality
Epilepsy Research 29 (1998) 175 – 183
Synergistic effect of status epilepticus and ischemic brain injury on mortality E.J. Waterhouse a,*, J.K. Vaughan a, T.Y. Barnes b, J.G. Boggs a, A.R. Towne a, L. Kopec-Garnett a, R.J. DeLorenzo a a
Department of Neurology, Medical College of Virginia, Virginia Commonwealth Uni6ersity, Box 980599, Richmond, VA 23298 -0599, USA b Department of Biostatistics, Medical College of Virginia, Virginia Commonwealth Uni6ersity, Box 980032, Richmond, VA 23298 -0032, USA Received 18 July 1997; accepted 21 July 1997
Abstract Ischemic brain injury (stroke) is a major cause of status epilepticus (SE). In our database of 529 adult SE cases, acute or remote cerebrovascular accidents (CVA) were a primary cause of SE for 41% of the patients overall and for 61% of the elderly patients. SE in the setting of acute CVA has a very high mortality, approaching 35%. The degree to which mortality can be attributed to the severity of the underlying CVA etiology vs. the effect of SE has not been evaluated. To address this issue, we prospectively studied patients with SE and acute CVA and compared them to control populations with acute CVA alone or with SE and remote CVA. The groups did not significantly differ with regard to age, sex, or radiographic lesion size. Mortality was unrelated to lesion size in the CVA and SE group. Overall, acute CVA and SE patients had an almost three-fold increase in mortality compared to the CVA group and an eight-fold increase compared to the SE and the non acute (remote) CVA group. Logistic regression analysis demonstrated a statistically significant synergistic effect of SE and CVA on mortality. This is the first study to document that the high mortality of SE and acute CVA is not solely due to the severity of the underlying CVA etiology, but due to the synergistic effect of combined injuries from SE and cerebral vascular ischemia. © 1998 Elsevier Science B.V. All rights reserved. Keywords: Status epilepticus; Cerebrovascular accident; Mortality; Stroke; Ischemic brain injury
1. Introduction * Corresponding author. Tel.: + 1 804 8284323; fax: +1 804 8283667.
Status epilepticus (SE) is a major medical and neurologic emergency, requiring immediate treat-
0920-1211/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII S 0 9 2 0 - 1 2 1 1 ( 9 7 ) 0 0 0 7 1 - 5
176
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
ment (Aminoff and Simon, 1980; Leppik, 1985; Hankey et al., 1987; DeLorenzo, 1990a,b; DeLorenzo et al., 1992; Lowenstein and Alldredge, 1992; Dodson et al., 1993; Towne et al., 1994; DeLorenzo, 1997). Recent epidemiological studies have indicated that SE affects approximately 150 000 people in the United States annually, and may account for as many as 25 000 – 50 000 deaths per year (DeLorenzo et al., 1995, 1996). Despite improvements in the diagnosis and treatment of SE, SE in adults is still associated with a significant morbidity and mortality (Leppik, 1985; Hankey et al., 1987; Sung and Chu, 1990; DeLorenzo et al., 1992, 1995). Thus, evaluating prognostic indicators for outcome in SE would significantly improve our ability to clinically manage patients and provide new insights into the pathophysiology of SE. Several determinants of SE outcome have been described, including etiology, age, and seizure duration (Hankey et al., 1987; Towne et al., 1994). Leppik and others have suggested that etiology is a major determinant of SE outcome (Celesia et al., 1972; Leppik, 1985; Cascino et al., 1995), while others have suggested that the consequences of prolonged seizures may cause mortality, independently of etiology (Oxbury and Whitty, 1971; Aminoff and Simon, 1980; Terrence et al., 1981; Sechi et al., 1985; Simon, 1985). It is important to determine whether etiology alone determines poor prognosis or whether other factors related to prolonged seizures also contribute to morbidity and mortality. Multi-regression logistic analysis of a large patient population determined that etiology did contribute to mortality in SE, along with age and seizure duration (Towne et al., 1994). In adults, ischemic brain injury or cerebrovascular accidents (CVAs) are a major cause of SE (Celesia et al., 1972; Hankey et al., 1987; Sung and Chu, 1989; DeLorenzo et al., 1995, 1996; DeLorenzo, 1997). It has recently been shown in a prospective epidemiologic study that acute and remote CVAs account for 41% of SE in adults, and 61% of SE in the elderly population (DeLorenzo, 1997; DeLorenzo et al., 1996). SE in the setting of CVA has a significant mortality, approaching 35% in adults (DeLorenzo, 1997; DeLorenzo et al., 1996). We initiated this study to
evaluate the relative contributions of SE and ischemic CVA etiology to mortality in patients who have both. This is the first study to investigate whether the mortality of SE with CVA differs significantly from the mortality of CVA alone, and whether the high mortality in this setting results from severity of the underlying CVA or a synergistic effect of these two brain insults.
2. Methods We prospectively studied a group of Medical College of Virginia (MCV) patients with SE and acute or remote CVA from July 1989 through June 1994. These patients were prospectively identified, and entered into the Richmond, VA status epilepticus data base, which has been previously described and validated (DeLorenzo et al., 1996). In addition, we identified a ‘control’ cohort of 159 adults who had acute ischemic CVA without associated SE between 7/1/92 and 6/30/93, using the MCV computerized discharge data base and ICDM-9-CM codes. There were no differences in the methods used to evaluate, diagnose, or treat the patients with SE and CVA or CVA alone. Patients were treated using the MCV status epilepticus protocol (DeLorenzo, 1990b) and there were no significant differences in the treatment modalities offered to the patients. The patients were collected over a time period during which there were no changes in the referral or clinical care patterns for CVA or SE patients in the database.
2.1. Definitions SE was defined as a seizure lasting at least 30 min, or intermittent seizures lasting greater than or equal to 30 min, during which the patient did not regain consciousness (Dodson et al., 1993). Non-convulsive SE was included if it was confirmed by an EEG recording. Mortality was defined as death during hospital admission for all groups. Acute CVA was defined according to the World Health Organization criteria as rapidly developed clinical signs of focal disturbance of cerebral function, lasting longer than 24 h or leading to death,
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
with no apparent cause other than those associated with vascular origin (World Health Organization, 1989). The remote CVA +SE group was defined as described previously (DeLorenzo et al., 1996). This group provided a control population with non acute ischemic lesion and SE with which to compare the acute CVA and SE group. Brain CT scan or MRI was performed during hospital admission, often within the first 24 h following an event. If the initial study was negative, and a follow-up scan demonstrated an acute lesion not seen initially, the follow-up scan was assessed. The MRI or CT scans were evaluated by a board certified neuroradiologist and were used to obtain information regarding lesion parameters. In the cases where the reports did not contain the necessary detailed information, the scans themselves were reviewed. Patients with primary intraparenchymal hemorrhages, subarachnoid hemorrhages, epidural and subdural hemorrhages were excluded. Categorization of lesion size was carried out using established procedures (Brott et al., 1989a) with the following characterizations of lesion size: large (an entire territory of a major cerebral artery), moderate (]1/2 of the territory), small (B1/2 of the territory), and lacune. Those with multiple acute strokes were categorized according to the size of their largest lesion, and were also tallied separately.
2.2. Statistics We compared SE patients in whom acute stroke was identified as a primary etiology of SE, with the patients who had acute CVA alone, and those with remote CVA and SE. The groups were compared with regard to age, using the Kruskal – Wallis test (Hollander and Wolfe, 1973), and sex, using the chi-square test. Radiographic lesion sizes on head CT scan or MRI were compared using the chi-square test or Fisher’s exact test. Mortalities for the groups were compared using the chi-square test. Logistic regression analysis (Hosmer and Lemeshow, 1989) was used to fit a regression model incorporating a ‘synergy term’ for SE to the mortality data, to determine whether or not the effects of SE and acute stroke on mortality were purely additive. Significance was defined as p 50.05.
177
3. Results There were a total of 83 MCV patients with CVA and SE: 39 with remote stroke (old CVA+ SE) and 44 with acute stroke (acute CVA + SE). There were 159 patients with acute CVA only (CVA). Table 1 presents the age and sex distributions of these three groups. The groups did not statistically differ with regard to age or sex. Mortality for the three groups is shown in Fig. 1. Acute CVA+ SE had a mortality of 39%, which was statistically significantly higher than either the control group of CVA (14%) or old CVA+ SE (5%). This represents almost a threefold statistically significant increase in mortality for patients with acute CVA and SE compared to patients with acute CVA alone (pB 0.001). Remote CVAs clearly had the lowest mortality (5%), indicating that the acuteness of the CVA in association with SE was an important contributing factor to the increased mortality. The old CVA+ SE mortality was statistically significantly lower by almost eight times compared to acute CVA+ SE, but did not statistically significantly differ from the CVA group. The remote CVA group represents a control group reflecting the mortality of SE alone in the setting of a non-acute and healed brain injury. Thus, we could determine whether the increased mortality of the SE and CVA group resulted from the purely additive effect of the mortalities due to CVA and SE, or was synergistic in nature. Logistic regression analysis was used to determine whether or not the Table 1 Age and sex distribution (mean age, years) CVA only
Acute CVA+ SE
Old CVA+SE
Female
66.5 9 13.69 (n =82)
67.2 9 15.2 (n =18)
64.7 914.3 (n =25)
Male
62.2 9 13.7 (n =77)
63.2 9 11.1 (n =26)
62.1 911.3 (n =14)
No statistically significant differences were found in the age distributions of the three groups using Kruskal – Wallis test (Inoue et al., 1980). The male/female distributions for these three groups were not statistically significantly different (chisquare test).
178
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
Fig. 1. Mortality by study groups. The mortality rates for acute cerebrovascular accident (CVA), acute CVA +SE, and old CVA + SE are presented. The mortality of acute CVA + SE was statistically significantly different from CVA (p 5 0.001) and old CVA + SE (p 50.001). No statistically significant difference was found between the mortalities of CVA and old CVA +SE.
effect of SE and CVA on mortality was purely additive (methods). This analysis demonstrated that the synergistic term in the logistic regression model was highly significant (p 50.001), indicating that the effects of SE and CVA on mortality were not simply additive. The distributions of radiographic lesion sizes within the CVA and acute CVA+SE groups are shown in Fig. 2. As seen in Fig. 2, 63% of CVA patients, and 75% of acute CVA+SE patients had no acute lesions on MRI or CT scan. Eleven percent of CVA patients, and 3% of the acute CVA + SE group had lacunar or small strokes. Fifteen percent of the CVA group had moderatesized strokes, as did 7% of the acute CVA+ SE group. While similar percentages of patients in the CVA and acute CVA + SE groups had large strokes (11 and 15% respectively), a relatively larger proportion (29%) of the old CVA+ SE population demonstrated large lesions on imaging studies (data not shown). Multiple strokes were more common in the CVA and old CVA+ SE groups than in the acute CVA+SE group (data not shown). There was a similar distribution of lesion sizes for the CVA and acute CVA + SE groups, while the old CVA + SE group had more large strokes.
CVA+ SE patients did not have more severe CVAs based on lesion size, and the majority of cases in both acute CVA+ SE and CVA groups had no significant acute lesion on MRI or CT imaging. This finding is consistent with findings from other series of acute CVAs (Inoue et al., 1980; Kunitz et al., 1984; Brott et al., 1989a). There was no evidence of disproportionately larger strokes in the acute CVA+ SE group. The relationships between lesion size and mortality for CVA and acute CVA+ SE, are shown in Fig. 3. In the acute CVA + SE group, lesion size and mortality were not correlated. However, there was a significant association between maximum lesion size and mortality in the CVA group (p B0.001). Mortality was consistently higher for acute CVA+ SE than for CVA alone, even in patients who had small lesions or no acute lesion seen on MRI or CT scan.
4. Discussion In a prospective population-based epidemiologic study of SE, acute or remote non-hemorrhagic CVA was an etiologic factor in almost 50% of adult SE cases, 22% acute CVAs and 25%
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
179
Fig. 2. Distribution of lesion size for the CVA and acute CVA + SE Groups. Lesion sizes were designated as none, lacune/small, moderate, and large (see Methods).
remote CVAs with 33% mortality (DeLorenzo et al., 1995, 1996). In elderly patients with SE and stroke, the mortality approached 48% (DeLorenzo, 1997). Thus, it is important to evaluate the relationship between acute CVA and SE. This is the first study to systematically address the role of SE etiology in determining mortality by comparing SE patients with CVA to a demographically similar control population with comparable stroke lesions. Those with acute CVA and SE had a consistently higher mortality rate compared with the group of patients who had acute CVA alone. This highly significant difference in mortality between the two groups was not due to lesion size. Furthermore, the results indicate that SE and acute brain ischemia have a synergistic effect on mortality. Evaluations of etiology as a contributor to mortality in SE patients suggested that the underlying etiology alone was a determinant of outcome, with certain etiologies associated with a high and others with a low mortality (Celesia et al., 1972; Leppik, 1985; Hankey et al., 1987; Hauser, 1990; Towne et al., 1994; Cascino et al., 1995; DeLorenzo et al., 1996). An initial theory
was that CVA and SE patients had a higher mortality than those with CVA alone because CVA patients who developed SE had larger, more severe strokes. If this hypothesis were correct, the higher mortality of acute CVA+ SE patients would be accounted for by larger and more severe strokes in the group compared to those with CVA alone. Our data indicate that this is not the case, and in fact, the distributions of lesion sizes in the CVA and the acute CVA+ SE groups were similar, even though the mortality of acute CVA+ SE was much higher than CVA alone. Furthermore, acute CVA+ SE patients without significant radiographic lesions still had a three-fold higher mortality than CVA alone. Thus, lesion size clearly did not determine the increased mortality. Distribution of radiographic lesions sizes in stroke varies widely in the literature (Mohr et al., 1978; Gupta et al., 1988; Brott et al., 1989a; Lancman et al., 1993). The range and distribution of lesion sizes reported here fall within the range of previous studies, many of which included hemorrhagic events. The results of this study demonstrate that the increased mortality associated with SE and acute CVA compared with CVA alone is
180
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
Fig. 3. Mortality and lesion size. The data show the mortality for CVA only and acute CVA+ SE for each lesion size: none, lacune/small, moderate and large.
not due to lesion size. In fact, even in cases with small or no acute lesions on neuroimaging studies, acute CVA+ SE patients had a three-fold increase in mortality compared to those with stroke alone. Lesion size has been shown to reflect clinical severity and functional outcome in CVA (Mohr et al., 1978; Miller and Miyamoto, 1979; Hankey et al., 1987; Gupta et al., 1988; Brott et al., 1989a; DeLorenzo, 1990b). The similar distribution of lesion sizes for the CVA and acute CVA+ SE groups implies that the groups also did not differ in clinical severity of the cerebral vascular injury (Miller and Miyamoto, 1979; Hankey et al., 1987; Brott et al., 1989b; Lancman et al., 1993). Acute CVA + SE patients without focal CT lesions were as likely to have a high mortality as patients with large lesions, while in CVA patients, lesion size was significantly correlated with outcome, with large lesions having a high mortality. These results suggest that in the acute CVA+ SE group the combined insults to the brain produced a higher mortality. In this group the potential effect of lesion size on mortality was overshadowed by the synergistic effect of the seizure and ischemic injuries.
A significant proportion of CVA patients in this study had no acute lesion on their CT scan or MRI. Other series have documented that 42–64% of acute strokes do not show acute lesions by CT scanning (Hollander and Wolfe, 1973; Kertesz et al., 1987; Hosmer and Lemeshow, 1989; Bryan et al., 1991; Weingarten, 1992; Cinnamon et al., 1995). Magnetic resonance imaging is more sensitive, detecting 82% of acute infarcts (Bryan et al., 1991). However, most of these studies included primary hemorrhages, which tend to be easily visible radiographically in the acute phase, and likely increased stroke detection rates. We excluded primary hemorrhages from our patient populations, which may contribute to our slightly lower rate of radiographic acute stroke detection. In addition, so called fogging effects occur when infarcted brain tissue becomes radiographically isodense when compared to surrounding uninvolved tissue. These effects may obscure subacute stroke lesions on both CT and MR scans (Cinnamon et al., 1995). It is also well recognized that posterior fossa infarcts are difficult to visualize on CT due to streak artifacts (Cinnamon et al., 1995). Thus, our findings that 63–75% of acute stroke patients had no acute radiographic lesion is
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
in keeping with other studies from the radiographic and stroke literature. Our data on remote CVA and SE provide a relevant contrast. Old CVA+SE patients had a greater proportion of radiographic lesions classified as large (29%) compared with the acute CVA +SE (15%) and CVA only (11%) groups. Not surprisingly, old CVA+ SE had a smaller proportion of radiographically undetectable lesions (32%) compared with acute CVA + SE (75%) and CVA (63%). Although a greater proportion of the old CVA+ SE group had radiographically more severe lesions compared with the other two groups, the mortality rate of old CVA+ SE was extremely low, only 5%. These data indicate that SE in the absence of an acute ischemic lesion had a much lower mortality than SE in the setting of acute CVA. If the relationship between CVA and SE was additive with regard to mortality, then the sum of the mortality rates of the CVA and old CVA + SE groups would equal that of the acute CVA + SE group. This was not the case. In fact, logistic regression analysis demonstrated with a high level of statistical significance (p 5 0.001) that the relationship between acute CVA and SE was synergistic and not additive. Thus, the effect of these two brain insults together produced a greater mortality than would be expected. The head injury literature provides a relevant comparison of multiple brain injuries. The combination of two acute traumatic brain injuries has a worse outcome when compared with either injury alone (Saunders and Harbaugh, 1984; Kelly et al., 1991). These results indicate that the initial injury sets up a process in the brain that makes it more susceptible to a second injury (Phillips et al., 1994). Our results suggest that a similar process may be occurring clinically with the combination of ischemic brain injury and SE. There is considerable evidence that neuronal damage and cell death in both ischemia and SE are mediated by excitotoxicity and prolonged increases in intracellular calcium (Choi, 1988; Dichter and Choi, 1989; Perlin and DeLorenzo, 1992). This common mechanism of brain injury in SE and CVA may in part
181
provide an explanation for the synergistic effect seen with these conditions. It is interesting to consider why SE occurs in the setting of an acute lacunar stroke. Lacunar strokes tend to occur in the deep grey and white matter, while symptomatic seizures are thought to involve the cortical grey matter. Lacunar infarctions have been previously described in association with seizures (Avrahami et al., 1988; Sung and Chu, 1990; Lancman et al., 1993), and with late-onset epilepsy (De la Sayette et al., 1987). An acute lacunar stroke may be an indicator of more generalized cerebrovascular dysfunction (Miller, 1983; Shorvon et al., 1984), and this more generalized dysfunction may be associated with SE rather than the acute lacune per se. Further studies are needed to evaluate the role of lacunar infarcts in seizures or SE. This study indicates that there is almost a three-fold increase in mortality in patients with SE and acute stroke compared to those who had acute stroke alone, or remote stroke and SE. This highly statistically significant difference in mortality was not accounted for by age, sex, or radiographic lesion size. Thus, the mortality of SE was not determined solely by the underlying lesion. These results suggest that the effects of stroke on brain function make the brain vulnerable to the effects of prolonged seizure activity, as seen in SE. The increased mortality associated with CVA and SE cannot be accounted for by the mortality of the underlying etiologies alone. The pathophysiology of combined prolonged seizures and ischemic injury to the brain is an important area for further investigation, and may provide insights into new therapeutic strategies and management of this severe condition associated with a very high mortality.
Acknowledgements We would like to thank Carolyn Cole, Judy Little, Connie Pettaway and Desiree Slawski for their help in the preparation of this manuscript. Supported by: MCV Epilepsy Research Center Grant 1 P01 NS25630.
182
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
References Aminoff, M.J., Simon, R.P., 1980. Status epilepticus: causes, clinical features and consequences in 98 patients. Am. J. Med. 69, 657 – 666. Avrahami, E., Drory, V.E., Raby, M.J., Cohn, D.F., 1988. Generalized epileptic seizures as the presenting symptom of lacunar infarction in the brain. J. Neurol. 235, 472–474. Brott, T., Marler, J.R., Adams, H.P., Olinger, C.P., 1989a. Measurements of acute cerebral infarction: a clinical examination scale. Stroke 20, 864–870. Brott, T., Marler, J.R., Olinger, C.P., Adams, H.P., Tomsick, T., Barsan, W.G., Biller, J., Eberle, R., Hertzberg, V., Walker, M., 1989b. Measurements of acute cerebral infarction: lesion size by computed tomography. Stroke 20, 871 – 875. Bryan, R.N., Levy, L.M., Whitlow, W.D., Killian, J.M., Preziosi, T.J., Rosario, J.A., 1991. Diagnosis of acute cerebral infarction; comparison of CT and MR imaging. Amer. J. Neuroradiol. 12, 611–620. Cascino, C.D., Hauser, W.A., Hesdorffer, D., Rocca, W.A., 1995. Morbidity and mortality of status epilepticus. Epilepsia 36 (Suppl 4), S46. Celesia, G.G., Messert, B., Murphy, M.J., 1972. Status epilepticus of late adult onset. Neurology 22, 1047–1055. Choi, D.W., 1988. Calcium-mediated neurotoxicity: relationship to specific channel types and role in ischemic damage. Trends Neurosci. 11, 465–469. Cinnamon, J., Viroslav, A.B., Dorey, J.H., 1995. CT and MRI diagnosis of cerebrovascular disease: going beyond the pixels. Sem. Ultrasound CT MRI 16 (3), 212–236. De la Sayette, V., Cosgrove, R., Melanson, D., Ethier, R., 1987. CT findings in late-onset epilepsy. Can. J. Neurol. Sci. 14, 286 – 289. DeLorenzo, R.J., 1990a. Status epilepticus: Concepts in diagnosis and treatment. Semin. Neurol. 10, 396–405. DeLorenzo, R.J., 1990b. Status epilepticus: Current therapy in neurologic disease. In: Johnson, R.T. (Ed.). Current Therapy in Neurologic Disease. Decker, Philadelphia, PA, pp. 47– 53. DeLorenzo, R.J., Towne, A.R., Pellock, J.M., Ko, D., 1992. Status epilepticus in children, adults and the elderly. Epilepsia 33 (Suppl 4), S15–S25. DeLorenzo, R.J., Pellock, J.M., Towne, A.R., Boggs, J.G., 1995. Epidemiology of status epilepticus. J. Clin. Neurophysiol. 12 (4), 316 – 325. DeLorenzo, R.J., Towne, A.R., Ko, D., Hauser, W.A., Penberthy, L., Pellock, J.M., 1996. Prospective populationbased epidemiological study of status epilepticus in Richmond, VA. Neurology 46, 1029–1032. DeLorenzo, R.J., 1997. Clinical and epidemiological study of status epilepticus in the elderly. In: Rowan, A.J., Ramsey, R.E. (Eds.), Seizures and Epilepsy in the Elderly. Butterworth – Heinemann, Newton, MA, pp. 191–205. Dichter, M.A., Choi, D.W., 1989. Excitatory amino acid transmitters and neurotoxins. Curr. Neurol. 9, 1–26.
Dodson, W.E., DeLorenzo, R.J., Pedley, T.A., Shinnar, S., Treiman, D.M., Wannamaker, B.B., 1993. The treatment of convulsive status epilepticus: Recommendations of The Epilepsy Foundation of America’s working group on status epilepticus. J. Am. Med. Assoc. 270, 854 – 859. Gupta, S.R., Naheedy, M.H., Elias, D., Rubino, F.A., 1988. Postinfarction seizures: a clinical study. Stroke 19 (12), 1477 – 1481. Hankey, G.J., Davis, S.J., Stewart-Wynne, E.G., Chakera, T.M.H., 1987. Cranial CT scan appearances that correlate with patient outcome in acute stroke. Clin. Exp. Neurol. 23, 71 – 74. Hauser, W.A., 1990. Status epilepticus: epidemiology considerations. Neurology 40 (Suppl 2), 9 – 12. Hollander, M., Wolfe, D.A. (Eds.), 1973. Nonparametric Statistical Methods. John Wiley and Sons, New York. Hosmer, D.W. Jr., Lemeshow, S. (Eds.), 1989. Applied Logistic Regression. John Wiley and Sons, New York. Inoue, Y., Kazumasa, T., Miyamoto, T., Yoshikawa, N., Taniguch, S., Saiwai, S., Nishimura, Y., Komatsu, T., 1980. Sequential computed tomography scans in acute cerebral infarction. Radiology 135, 655 – 662. Kelly, J.P., Nichols, J.S., Filley, C.M., Lillehei, K.O., Rubinstein, D., Kleinschmidt-DeMasters, B.K., 1991. Concussion in sports: guidelines for the prevention of catastrophic outcome. J. Am. Med. Assoc. 266, 2867 – 2869. Kertesz, A., Black, S.E., Nickolson, L., Carr, T., 1987. The sensitivity and specificity of MRI in stroke. Neurology 37, 1580 – 1585. Kunitz, S.C., Gross, C.R., Heyman, A., Kase, C.S., Mohr, J.P., Price, T.R., Wolf, P.A., 1984. The pilot stroke data bank: definition, design, and data. Stroke 4, 740 – 746. Lancman, M.E., Golimstok, A., Norscini, J., Granillo, R., 1993. Risk factors for developing seizures after a stroke. Epilepsia 34 (1), 141 – 143. Leppik, I.E., 1985. Status epilepticus. Clin. Ther. 7 (2), 272 – 278. Lowenstein, D.H., Alldredge, B.K., 1992. Status epilepticus at an urban public hospital in the 1980s. Neurology 43, 483 – 488. Miller, L.S., Miyamoto, A.T., 1979. Computed tomography: its potential as a predictor of functional recovery following stroke. Arch. Phys. Med. Rehabil. 60, 108 – 109. Miller, V.T., 1983. Lacunar stroke: a reassessment. Arch. Neurol. 40, 129 – 134. Mohr, J.P., Caplan, L.R., Melsk, J.W., Goldstein, R.J., Duncan, G.W., Kistler, J.P., Pessin, M.S., Bleich, H.L., 1978. The Harvard cooperative stroke registry: a prospective registry. Neurology 28, 754 – 762. Oxbury, J.M., Whitty, C.W.M., 1971. Causes and consequences of status epilepticus in adults. Brain 94, 733 – 744. Perlin, J.B., DeLorenzo, R.J., 1992. Calcium and Epilepsy. In: Pedley, T.A., Meldrum, B.S. (Eds.), Recent Advances in Epilepsy, No. 5. Churchill Livingstone, New York, pp. 15 – 36.
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183 Phillips, L.L., Lyeth, B.G., Hamm, R.J., Povlishock, J.T., 1994. Combined fluid percussion brain injury and entorhinal cortical lesion: A model for assessing the interaction between neuroexcitation and deafferentation. J. Neurotrauma. 11, 641 – 656. Saunders, R.L., Harbaugh, R.E., 1984. The second impact in catastrophic contact sports head trauma. J. Am. Med. Assoc. 252, 538 – 539. Sechi, G.P., Dessx-Fulgher, P., Glorisos, N., Volta, G., Rosah, G., 1985. Myocardial infarction complicating status epilepticus. Epilepsia 26, 572–576. Shorvon, S.D., Gilliat, R.W., Cox, C.S., Yu, Y.L., 1984. Evidence of vascular disease from CT scanning in late onset epilepsy. J. Neurol. Neurosurg. Psychiatry 47, 225–230. Simon, R.P., 1985. Physiologic consequences of status epilepticus. Epilepsia 26 (Suppl 1), S58–66.
.
183
Sung, C.Y., Chu, N.S., 1989. Status epilepticus in the elderly: etiology, seizure type and outcome. Acta. Neurol. Scand. 30, 51 – 56. Sung, C.Y., Chu, N.S., 1990. Epileptic seizures in thrombotic stroke. J. Neurol. 237, 166 – 170. Terrence, C.F., Rao, G.R., Pepper, J.A., 1981. Neurogenic pulmonary edema in unexpected, unexplained death of epileptic patients. Ann. Neurol. 9, 458 – 463. Towne, A.R., Pellock, J.M., Ko, D., DeLorenzo, R.J., 1994. Determinants of mortality in status epilepticus. Epilepsia 35 (1), 27 – 34. Weingarten, K., 1992. Computed tomography of cerebral infarction. Neuroimaging Clin. N. Am. 2, 409 – 419. World Health Organization, 1989. Stroke 1989: recommendations on stroke prevention, diagnosis, and therapy. Stroke 20 (10), 1407 – 1431.
.
Synergistic effect of status epilepticus and ischemic brain injury on mortality E.J. Waterhouse a,*, J.K. Vaughan a, T.Y. Barnes b, J.G. Boggs a, A.R. Towne a, L. Kopec-Garnett a, R.J. DeLorenzo a a
Department of Neurology, Medical College of Virginia, Virginia Commonwealth Uni6ersity, Box 980599, Richmond, VA 23298 -0599, USA b Department of Biostatistics, Medical College of Virginia, Virginia Commonwealth Uni6ersity, Box 980032, Richmond, VA 23298 -0032, USA Received 18 July 1997; accepted 21 July 1997
Abstract Ischemic brain injury (stroke) is a major cause of status epilepticus (SE). In our database of 529 adult SE cases, acute or remote cerebrovascular accidents (CVA) were a primary cause of SE for 41% of the patients overall and for 61% of the elderly patients. SE in the setting of acute CVA has a very high mortality, approaching 35%. The degree to which mortality can be attributed to the severity of the underlying CVA etiology vs. the effect of SE has not been evaluated. To address this issue, we prospectively studied patients with SE and acute CVA and compared them to control populations with acute CVA alone or with SE and remote CVA. The groups did not significantly differ with regard to age, sex, or radiographic lesion size. Mortality was unrelated to lesion size in the CVA and SE group. Overall, acute CVA and SE patients had an almost three-fold increase in mortality compared to the CVA group and an eight-fold increase compared to the SE and the non acute (remote) CVA group. Logistic regression analysis demonstrated a statistically significant synergistic effect of SE and CVA on mortality. This is the first study to document that the high mortality of SE and acute CVA is not solely due to the severity of the underlying CVA etiology, but due to the synergistic effect of combined injuries from SE and cerebral vascular ischemia. © 1998 Elsevier Science B.V. All rights reserved. Keywords: Status epilepticus; Cerebrovascular accident; Mortality; Stroke; Ischemic brain injury
1. Introduction * Corresponding author. Tel.: + 1 804 8284323; fax: +1 804 8283667.
Status epilepticus (SE) is a major medical and neurologic emergency, requiring immediate treat-
0920-1211/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII S 0 9 2 0 - 1 2 1 1 ( 9 7 ) 0 0 0 7 1 - 5
176
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
ment (Aminoff and Simon, 1980; Leppik, 1985; Hankey et al., 1987; DeLorenzo, 1990a,b; DeLorenzo et al., 1992; Lowenstein and Alldredge, 1992; Dodson et al., 1993; Towne et al., 1994; DeLorenzo, 1997). Recent epidemiological studies have indicated that SE affects approximately 150 000 people in the United States annually, and may account for as many as 25 000 – 50 000 deaths per year (DeLorenzo et al., 1995, 1996). Despite improvements in the diagnosis and treatment of SE, SE in adults is still associated with a significant morbidity and mortality (Leppik, 1985; Hankey et al., 1987; Sung and Chu, 1990; DeLorenzo et al., 1992, 1995). Thus, evaluating prognostic indicators for outcome in SE would significantly improve our ability to clinically manage patients and provide new insights into the pathophysiology of SE. Several determinants of SE outcome have been described, including etiology, age, and seizure duration (Hankey et al., 1987; Towne et al., 1994). Leppik and others have suggested that etiology is a major determinant of SE outcome (Celesia et al., 1972; Leppik, 1985; Cascino et al., 1995), while others have suggested that the consequences of prolonged seizures may cause mortality, independently of etiology (Oxbury and Whitty, 1971; Aminoff and Simon, 1980; Terrence et al., 1981; Sechi et al., 1985; Simon, 1985). It is important to determine whether etiology alone determines poor prognosis or whether other factors related to prolonged seizures also contribute to morbidity and mortality. Multi-regression logistic analysis of a large patient population determined that etiology did contribute to mortality in SE, along with age and seizure duration (Towne et al., 1994). In adults, ischemic brain injury or cerebrovascular accidents (CVAs) are a major cause of SE (Celesia et al., 1972; Hankey et al., 1987; Sung and Chu, 1989; DeLorenzo et al., 1995, 1996; DeLorenzo, 1997). It has recently been shown in a prospective epidemiologic study that acute and remote CVAs account for 41% of SE in adults, and 61% of SE in the elderly population (DeLorenzo, 1997; DeLorenzo et al., 1996). SE in the setting of CVA has a significant mortality, approaching 35% in adults (DeLorenzo, 1997; DeLorenzo et al., 1996). We initiated this study to
evaluate the relative contributions of SE and ischemic CVA etiology to mortality in patients who have both. This is the first study to investigate whether the mortality of SE with CVA differs significantly from the mortality of CVA alone, and whether the high mortality in this setting results from severity of the underlying CVA or a synergistic effect of these two brain insults.
2. Methods We prospectively studied a group of Medical College of Virginia (MCV) patients with SE and acute or remote CVA from July 1989 through June 1994. These patients were prospectively identified, and entered into the Richmond, VA status epilepticus data base, which has been previously described and validated (DeLorenzo et al., 1996). In addition, we identified a ‘control’ cohort of 159 adults who had acute ischemic CVA without associated SE between 7/1/92 and 6/30/93, using the MCV computerized discharge data base and ICDM-9-CM codes. There were no differences in the methods used to evaluate, diagnose, or treat the patients with SE and CVA or CVA alone. Patients were treated using the MCV status epilepticus protocol (DeLorenzo, 1990b) and there were no significant differences in the treatment modalities offered to the patients. The patients were collected over a time period during which there were no changes in the referral or clinical care patterns for CVA or SE patients in the database.
2.1. Definitions SE was defined as a seizure lasting at least 30 min, or intermittent seizures lasting greater than or equal to 30 min, during which the patient did not regain consciousness (Dodson et al., 1993). Non-convulsive SE was included if it was confirmed by an EEG recording. Mortality was defined as death during hospital admission for all groups. Acute CVA was defined according to the World Health Organization criteria as rapidly developed clinical signs of focal disturbance of cerebral function, lasting longer than 24 h or leading to death,
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
with no apparent cause other than those associated with vascular origin (World Health Organization, 1989). The remote CVA +SE group was defined as described previously (DeLorenzo et al., 1996). This group provided a control population with non acute ischemic lesion and SE with which to compare the acute CVA and SE group. Brain CT scan or MRI was performed during hospital admission, often within the first 24 h following an event. If the initial study was negative, and a follow-up scan demonstrated an acute lesion not seen initially, the follow-up scan was assessed. The MRI or CT scans were evaluated by a board certified neuroradiologist and were used to obtain information regarding lesion parameters. In the cases where the reports did not contain the necessary detailed information, the scans themselves were reviewed. Patients with primary intraparenchymal hemorrhages, subarachnoid hemorrhages, epidural and subdural hemorrhages were excluded. Categorization of lesion size was carried out using established procedures (Brott et al., 1989a) with the following characterizations of lesion size: large (an entire territory of a major cerebral artery), moderate (]1/2 of the territory), small (B1/2 of the territory), and lacune. Those with multiple acute strokes were categorized according to the size of their largest lesion, and were also tallied separately.
2.2. Statistics We compared SE patients in whom acute stroke was identified as a primary etiology of SE, with the patients who had acute CVA alone, and those with remote CVA and SE. The groups were compared with regard to age, using the Kruskal – Wallis test (Hollander and Wolfe, 1973), and sex, using the chi-square test. Radiographic lesion sizes on head CT scan or MRI were compared using the chi-square test or Fisher’s exact test. Mortalities for the groups were compared using the chi-square test. Logistic regression analysis (Hosmer and Lemeshow, 1989) was used to fit a regression model incorporating a ‘synergy term’ for SE to the mortality data, to determine whether or not the effects of SE and acute stroke on mortality were purely additive. Significance was defined as p 50.05.
177
3. Results There were a total of 83 MCV patients with CVA and SE: 39 with remote stroke (old CVA+ SE) and 44 with acute stroke (acute CVA + SE). There were 159 patients with acute CVA only (CVA). Table 1 presents the age and sex distributions of these three groups. The groups did not statistically differ with regard to age or sex. Mortality for the three groups is shown in Fig. 1. Acute CVA+ SE had a mortality of 39%, which was statistically significantly higher than either the control group of CVA (14%) or old CVA+ SE (5%). This represents almost a threefold statistically significant increase in mortality for patients with acute CVA and SE compared to patients with acute CVA alone (pB 0.001). Remote CVAs clearly had the lowest mortality (5%), indicating that the acuteness of the CVA in association with SE was an important contributing factor to the increased mortality. The old CVA+ SE mortality was statistically significantly lower by almost eight times compared to acute CVA+ SE, but did not statistically significantly differ from the CVA group. The remote CVA group represents a control group reflecting the mortality of SE alone in the setting of a non-acute and healed brain injury. Thus, we could determine whether the increased mortality of the SE and CVA group resulted from the purely additive effect of the mortalities due to CVA and SE, or was synergistic in nature. Logistic regression analysis was used to determine whether or not the Table 1 Age and sex distribution (mean age, years) CVA only
Acute CVA+ SE
Old CVA+SE
Female
66.5 9 13.69 (n =82)
67.2 9 15.2 (n =18)
64.7 914.3 (n =25)
Male
62.2 9 13.7 (n =77)
63.2 9 11.1 (n =26)
62.1 911.3 (n =14)
No statistically significant differences were found in the age distributions of the three groups using Kruskal – Wallis test (Inoue et al., 1980). The male/female distributions for these three groups were not statistically significantly different (chisquare test).
178
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
Fig. 1. Mortality by study groups. The mortality rates for acute cerebrovascular accident (CVA), acute CVA +SE, and old CVA + SE are presented. The mortality of acute CVA + SE was statistically significantly different from CVA (p 5 0.001) and old CVA + SE (p 50.001). No statistically significant difference was found between the mortalities of CVA and old CVA +SE.
effect of SE and CVA on mortality was purely additive (methods). This analysis demonstrated that the synergistic term in the logistic regression model was highly significant (p 50.001), indicating that the effects of SE and CVA on mortality were not simply additive. The distributions of radiographic lesion sizes within the CVA and acute CVA+SE groups are shown in Fig. 2. As seen in Fig. 2, 63% of CVA patients, and 75% of acute CVA+SE patients had no acute lesions on MRI or CT scan. Eleven percent of CVA patients, and 3% of the acute CVA + SE group had lacunar or small strokes. Fifteen percent of the CVA group had moderatesized strokes, as did 7% of the acute CVA+ SE group. While similar percentages of patients in the CVA and acute CVA + SE groups had large strokes (11 and 15% respectively), a relatively larger proportion (29%) of the old CVA+ SE population demonstrated large lesions on imaging studies (data not shown). Multiple strokes were more common in the CVA and old CVA+ SE groups than in the acute CVA+SE group (data not shown). There was a similar distribution of lesion sizes for the CVA and acute CVA + SE groups, while the old CVA + SE group had more large strokes.
CVA+ SE patients did not have more severe CVAs based on lesion size, and the majority of cases in both acute CVA+ SE and CVA groups had no significant acute lesion on MRI or CT imaging. This finding is consistent with findings from other series of acute CVAs (Inoue et al., 1980; Kunitz et al., 1984; Brott et al., 1989a). There was no evidence of disproportionately larger strokes in the acute CVA+ SE group. The relationships between lesion size and mortality for CVA and acute CVA+ SE, are shown in Fig. 3. In the acute CVA + SE group, lesion size and mortality were not correlated. However, there was a significant association between maximum lesion size and mortality in the CVA group (p B0.001). Mortality was consistently higher for acute CVA+ SE than for CVA alone, even in patients who had small lesions or no acute lesion seen on MRI or CT scan.
4. Discussion In a prospective population-based epidemiologic study of SE, acute or remote non-hemorrhagic CVA was an etiologic factor in almost 50% of adult SE cases, 22% acute CVAs and 25%
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
179
Fig. 2. Distribution of lesion size for the CVA and acute CVA + SE Groups. Lesion sizes were designated as none, lacune/small, moderate, and large (see Methods).
remote CVAs with 33% mortality (DeLorenzo et al., 1995, 1996). In elderly patients with SE and stroke, the mortality approached 48% (DeLorenzo, 1997). Thus, it is important to evaluate the relationship between acute CVA and SE. This is the first study to systematically address the role of SE etiology in determining mortality by comparing SE patients with CVA to a demographically similar control population with comparable stroke lesions. Those with acute CVA and SE had a consistently higher mortality rate compared with the group of patients who had acute CVA alone. This highly significant difference in mortality between the two groups was not due to lesion size. Furthermore, the results indicate that SE and acute brain ischemia have a synergistic effect on mortality. Evaluations of etiology as a contributor to mortality in SE patients suggested that the underlying etiology alone was a determinant of outcome, with certain etiologies associated with a high and others with a low mortality (Celesia et al., 1972; Leppik, 1985; Hankey et al., 1987; Hauser, 1990; Towne et al., 1994; Cascino et al., 1995; DeLorenzo et al., 1996). An initial theory
was that CVA and SE patients had a higher mortality than those with CVA alone because CVA patients who developed SE had larger, more severe strokes. If this hypothesis were correct, the higher mortality of acute CVA+ SE patients would be accounted for by larger and more severe strokes in the group compared to those with CVA alone. Our data indicate that this is not the case, and in fact, the distributions of lesion sizes in the CVA and the acute CVA+ SE groups were similar, even though the mortality of acute CVA+ SE was much higher than CVA alone. Furthermore, acute CVA+ SE patients without significant radiographic lesions still had a three-fold higher mortality than CVA alone. Thus, lesion size clearly did not determine the increased mortality. Distribution of radiographic lesions sizes in stroke varies widely in the literature (Mohr et al., 1978; Gupta et al., 1988; Brott et al., 1989a; Lancman et al., 1993). The range and distribution of lesion sizes reported here fall within the range of previous studies, many of which included hemorrhagic events. The results of this study demonstrate that the increased mortality associated with SE and acute CVA compared with CVA alone is
180
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
Fig. 3. Mortality and lesion size. The data show the mortality for CVA only and acute CVA+ SE for each lesion size: none, lacune/small, moderate and large.
not due to lesion size. In fact, even in cases with small or no acute lesions on neuroimaging studies, acute CVA+ SE patients had a three-fold increase in mortality compared to those with stroke alone. Lesion size has been shown to reflect clinical severity and functional outcome in CVA (Mohr et al., 1978; Miller and Miyamoto, 1979; Hankey et al., 1987; Gupta et al., 1988; Brott et al., 1989a; DeLorenzo, 1990b). The similar distribution of lesion sizes for the CVA and acute CVA+ SE groups implies that the groups also did not differ in clinical severity of the cerebral vascular injury (Miller and Miyamoto, 1979; Hankey et al., 1987; Brott et al., 1989b; Lancman et al., 1993). Acute CVA + SE patients without focal CT lesions were as likely to have a high mortality as patients with large lesions, while in CVA patients, lesion size was significantly correlated with outcome, with large lesions having a high mortality. These results suggest that in the acute CVA+ SE group the combined insults to the brain produced a higher mortality. In this group the potential effect of lesion size on mortality was overshadowed by the synergistic effect of the seizure and ischemic injuries.
A significant proportion of CVA patients in this study had no acute lesion on their CT scan or MRI. Other series have documented that 42–64% of acute strokes do not show acute lesions by CT scanning (Hollander and Wolfe, 1973; Kertesz et al., 1987; Hosmer and Lemeshow, 1989; Bryan et al., 1991; Weingarten, 1992; Cinnamon et al., 1995). Magnetic resonance imaging is more sensitive, detecting 82% of acute infarcts (Bryan et al., 1991). However, most of these studies included primary hemorrhages, which tend to be easily visible radiographically in the acute phase, and likely increased stroke detection rates. We excluded primary hemorrhages from our patient populations, which may contribute to our slightly lower rate of radiographic acute stroke detection. In addition, so called fogging effects occur when infarcted brain tissue becomes radiographically isodense when compared to surrounding uninvolved tissue. These effects may obscure subacute stroke lesions on both CT and MR scans (Cinnamon et al., 1995). It is also well recognized that posterior fossa infarcts are difficult to visualize on CT due to streak artifacts (Cinnamon et al., 1995). Thus, our findings that 63–75% of acute stroke patients had no acute radiographic lesion is
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
in keeping with other studies from the radiographic and stroke literature. Our data on remote CVA and SE provide a relevant contrast. Old CVA+SE patients had a greater proportion of radiographic lesions classified as large (29%) compared with the acute CVA +SE (15%) and CVA only (11%) groups. Not surprisingly, old CVA+ SE had a smaller proportion of radiographically undetectable lesions (32%) compared with acute CVA + SE (75%) and CVA (63%). Although a greater proportion of the old CVA+ SE group had radiographically more severe lesions compared with the other two groups, the mortality rate of old CVA+ SE was extremely low, only 5%. These data indicate that SE in the absence of an acute ischemic lesion had a much lower mortality than SE in the setting of acute CVA. If the relationship between CVA and SE was additive with regard to mortality, then the sum of the mortality rates of the CVA and old CVA + SE groups would equal that of the acute CVA + SE group. This was not the case. In fact, logistic regression analysis demonstrated with a high level of statistical significance (p 5 0.001) that the relationship between acute CVA and SE was synergistic and not additive. Thus, the effect of these two brain insults together produced a greater mortality than would be expected. The head injury literature provides a relevant comparison of multiple brain injuries. The combination of two acute traumatic brain injuries has a worse outcome when compared with either injury alone (Saunders and Harbaugh, 1984; Kelly et al., 1991). These results indicate that the initial injury sets up a process in the brain that makes it more susceptible to a second injury (Phillips et al., 1994). Our results suggest that a similar process may be occurring clinically with the combination of ischemic brain injury and SE. There is considerable evidence that neuronal damage and cell death in both ischemia and SE are mediated by excitotoxicity and prolonged increases in intracellular calcium (Choi, 1988; Dichter and Choi, 1989; Perlin and DeLorenzo, 1992). This common mechanism of brain injury in SE and CVA may in part
181
provide an explanation for the synergistic effect seen with these conditions. It is interesting to consider why SE occurs in the setting of an acute lacunar stroke. Lacunar strokes tend to occur in the deep grey and white matter, while symptomatic seizures are thought to involve the cortical grey matter. Lacunar infarctions have been previously described in association with seizures (Avrahami et al., 1988; Sung and Chu, 1990; Lancman et al., 1993), and with late-onset epilepsy (De la Sayette et al., 1987). An acute lacunar stroke may be an indicator of more generalized cerebrovascular dysfunction (Miller, 1983; Shorvon et al., 1984), and this more generalized dysfunction may be associated with SE rather than the acute lacune per se. Further studies are needed to evaluate the role of lacunar infarcts in seizures or SE. This study indicates that there is almost a three-fold increase in mortality in patients with SE and acute stroke compared to those who had acute stroke alone, or remote stroke and SE. This highly statistically significant difference in mortality was not accounted for by age, sex, or radiographic lesion size. Thus, the mortality of SE was not determined solely by the underlying lesion. These results suggest that the effects of stroke on brain function make the brain vulnerable to the effects of prolonged seizure activity, as seen in SE. The increased mortality associated with CVA and SE cannot be accounted for by the mortality of the underlying etiologies alone. The pathophysiology of combined prolonged seizures and ischemic injury to the brain is an important area for further investigation, and may provide insights into new therapeutic strategies and management of this severe condition associated with a very high mortality.
Acknowledgements We would like to thank Carolyn Cole, Judy Little, Connie Pettaway and Desiree Slawski for their help in the preparation of this manuscript. Supported by: MCV Epilepsy Research Center Grant 1 P01 NS25630.
182
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183
References Aminoff, M.J., Simon, R.P., 1980. Status epilepticus: causes, clinical features and consequences in 98 patients. Am. J. Med. 69, 657 – 666. Avrahami, E., Drory, V.E., Raby, M.J., Cohn, D.F., 1988. Generalized epileptic seizures as the presenting symptom of lacunar infarction in the brain. J. Neurol. 235, 472–474. Brott, T., Marler, J.R., Adams, H.P., Olinger, C.P., 1989a. Measurements of acute cerebral infarction: a clinical examination scale. Stroke 20, 864–870. Brott, T., Marler, J.R., Olinger, C.P., Adams, H.P., Tomsick, T., Barsan, W.G., Biller, J., Eberle, R., Hertzberg, V., Walker, M., 1989b. Measurements of acute cerebral infarction: lesion size by computed tomography. Stroke 20, 871 – 875. Bryan, R.N., Levy, L.M., Whitlow, W.D., Killian, J.M., Preziosi, T.J., Rosario, J.A., 1991. Diagnosis of acute cerebral infarction; comparison of CT and MR imaging. Amer. J. Neuroradiol. 12, 611–620. Cascino, C.D., Hauser, W.A., Hesdorffer, D., Rocca, W.A., 1995. Morbidity and mortality of status epilepticus. Epilepsia 36 (Suppl 4), S46. Celesia, G.G., Messert, B., Murphy, M.J., 1972. Status epilepticus of late adult onset. Neurology 22, 1047–1055. Choi, D.W., 1988. Calcium-mediated neurotoxicity: relationship to specific channel types and role in ischemic damage. Trends Neurosci. 11, 465–469. Cinnamon, J., Viroslav, A.B., Dorey, J.H., 1995. CT and MRI diagnosis of cerebrovascular disease: going beyond the pixels. Sem. Ultrasound CT MRI 16 (3), 212–236. De la Sayette, V., Cosgrove, R., Melanson, D., Ethier, R., 1987. CT findings in late-onset epilepsy. Can. J. Neurol. Sci. 14, 286 – 289. DeLorenzo, R.J., 1990a. Status epilepticus: Concepts in diagnosis and treatment. Semin. Neurol. 10, 396–405. DeLorenzo, R.J., 1990b. Status epilepticus: Current therapy in neurologic disease. In: Johnson, R.T. (Ed.). Current Therapy in Neurologic Disease. Decker, Philadelphia, PA, pp. 47– 53. DeLorenzo, R.J., Towne, A.R., Pellock, J.M., Ko, D., 1992. Status epilepticus in children, adults and the elderly. Epilepsia 33 (Suppl 4), S15–S25. DeLorenzo, R.J., Pellock, J.M., Towne, A.R., Boggs, J.G., 1995. Epidemiology of status epilepticus. J. Clin. Neurophysiol. 12 (4), 316 – 325. DeLorenzo, R.J., Towne, A.R., Ko, D., Hauser, W.A., Penberthy, L., Pellock, J.M., 1996. Prospective populationbased epidemiological study of status epilepticus in Richmond, VA. Neurology 46, 1029–1032. DeLorenzo, R.J., 1997. Clinical and epidemiological study of status epilepticus in the elderly. In: Rowan, A.J., Ramsey, R.E. (Eds.), Seizures and Epilepsy in the Elderly. Butterworth – Heinemann, Newton, MA, pp. 191–205. Dichter, M.A., Choi, D.W., 1989. Excitatory amino acid transmitters and neurotoxins. Curr. Neurol. 9, 1–26.
Dodson, W.E., DeLorenzo, R.J., Pedley, T.A., Shinnar, S., Treiman, D.M., Wannamaker, B.B., 1993. The treatment of convulsive status epilepticus: Recommendations of The Epilepsy Foundation of America’s working group on status epilepticus. J. Am. Med. Assoc. 270, 854 – 859. Gupta, S.R., Naheedy, M.H., Elias, D., Rubino, F.A., 1988. Postinfarction seizures: a clinical study. Stroke 19 (12), 1477 – 1481. Hankey, G.J., Davis, S.J., Stewart-Wynne, E.G., Chakera, T.M.H., 1987. Cranial CT scan appearances that correlate with patient outcome in acute stroke. Clin. Exp. Neurol. 23, 71 – 74. Hauser, W.A., 1990. Status epilepticus: epidemiology considerations. Neurology 40 (Suppl 2), 9 – 12. Hollander, M., Wolfe, D.A. (Eds.), 1973. Nonparametric Statistical Methods. John Wiley and Sons, New York. Hosmer, D.W. Jr., Lemeshow, S. (Eds.), 1989. Applied Logistic Regression. John Wiley and Sons, New York. Inoue, Y., Kazumasa, T., Miyamoto, T., Yoshikawa, N., Taniguch, S., Saiwai, S., Nishimura, Y., Komatsu, T., 1980. Sequential computed tomography scans in acute cerebral infarction. Radiology 135, 655 – 662. Kelly, J.P., Nichols, J.S., Filley, C.M., Lillehei, K.O., Rubinstein, D., Kleinschmidt-DeMasters, B.K., 1991. Concussion in sports: guidelines for the prevention of catastrophic outcome. J. Am. Med. Assoc. 266, 2867 – 2869. Kertesz, A., Black, S.E., Nickolson, L., Carr, T., 1987. The sensitivity and specificity of MRI in stroke. Neurology 37, 1580 – 1585. Kunitz, S.C., Gross, C.R., Heyman, A., Kase, C.S., Mohr, J.P., Price, T.R., Wolf, P.A., 1984. The pilot stroke data bank: definition, design, and data. Stroke 4, 740 – 746. Lancman, M.E., Golimstok, A., Norscini, J., Granillo, R., 1993. Risk factors for developing seizures after a stroke. Epilepsia 34 (1), 141 – 143. Leppik, I.E., 1985. Status epilepticus. Clin. Ther. 7 (2), 272 – 278. Lowenstein, D.H., Alldredge, B.K., 1992. Status epilepticus at an urban public hospital in the 1980s. Neurology 43, 483 – 488. Miller, L.S., Miyamoto, A.T., 1979. Computed tomography: its potential as a predictor of functional recovery following stroke. Arch. Phys. Med. Rehabil. 60, 108 – 109. Miller, V.T., 1983. Lacunar stroke: a reassessment. Arch. Neurol. 40, 129 – 134. Mohr, J.P., Caplan, L.R., Melsk, J.W., Goldstein, R.J., Duncan, G.W., Kistler, J.P., Pessin, M.S., Bleich, H.L., 1978. The Harvard cooperative stroke registry: a prospective registry. Neurology 28, 754 – 762. Oxbury, J.M., Whitty, C.W.M., 1971. Causes and consequences of status epilepticus in adults. Brain 94, 733 – 744. Perlin, J.B., DeLorenzo, R.J., 1992. Calcium and Epilepsy. In: Pedley, T.A., Meldrum, B.S. (Eds.), Recent Advances in Epilepsy, No. 5. Churchill Livingstone, New York, pp. 15 – 36.
E.J. Waterhouse et al. / Epilepsy Research 29 (1998) 175–183 Phillips, L.L., Lyeth, B.G., Hamm, R.J., Povlishock, J.T., 1994. Combined fluid percussion brain injury and entorhinal cortical lesion: A model for assessing the interaction between neuroexcitation and deafferentation. J. Neurotrauma. 11, 641 – 656. Saunders, R.L., Harbaugh, R.E., 1984. The second impact in catastrophic contact sports head trauma. J. Am. Med. Assoc. 252, 538 – 539. Sechi, G.P., Dessx-Fulgher, P., Glorisos, N., Volta, G., Rosah, G., 1985. Myocardial infarction complicating status epilepticus. Epilepsia 26, 572–576. Shorvon, S.D., Gilliat, R.W., Cox, C.S., Yu, Y.L., 1984. Evidence of vascular disease from CT scanning in late onset epilepsy. J. Neurol. Neurosurg. Psychiatry 47, 225–230. Simon, R.P., 1985. Physiologic consequences of status epilepticus. Epilepsia 26 (Suppl 1), S58–66.
.
183
Sung, C.Y., Chu, N.S., 1989. Status epilepticus in the elderly: etiology, seizure type and outcome. Acta. Neurol. Scand. 30, 51 – 56. Sung, C.Y., Chu, N.S., 1990. Epileptic seizures in thrombotic stroke. J. Neurol. 237, 166 – 170. Terrence, C.F., Rao, G.R., Pepper, J.A., 1981. Neurogenic pulmonary edema in unexpected, unexplained death of epileptic patients. Ann. Neurol. 9, 458 – 463. Towne, A.R., Pellock, J.M., Ko, D., DeLorenzo, R.J., 1994. Determinants of mortality in status epilepticus. Epilepsia 35 (1), 27 – 34. Weingarten, K., 1992. Computed tomography of cerebral infarction. Neuroimaging Clin. N. Am. 2, 409 – 419. World Health Organization, 1989. Stroke 1989: recommendations on stroke prevention, diagnosis, and therapy. Stroke 20 (10), 1407 – 1431.
.