187
Psychiatry Research, 43:187-195
Elsevier
Cerebrospinal Fluid and Serum Levels of Neuron-Specific Enolase in Patients With Schizophrenia Michael F. Egan, Rif S. El-Mallakh, Richard H. Stefan Bracha, and Richard Jed Wyatt Received
October
7, 1991; revised
version received
L. Suddath,
April
James
14, 1992; accepted
B. Lohr,
June
20, 1992.
Abstract. Some patients with schizophrenia appear to have brain abnormalities, including enlarged third and lateral ventricles and reduced volumes of temporal lobe structures. These abnormalities could be attributed to a developmental abnormality or a neurodegenerative process. Neuron-specific enolase (NSE), a protein that is found primarily in neurons and neuroendocrine cells, has been used as an index of neuronal damage or degeneration. Levels of NSE in cerebrospinal fluid (CSF) and serum from 50 patients with acute and chronic schizophrenia were compared with those in normal and neurological control subjects. A double-antibody, solid phase iodinated radioimmunoassay was used to determine NSE levels. There was no evidence of elevated levels in patients with schizophrenia, whereas control subjects with neurological illnesses had increased levels of NSE in CSF. Because NSE is rapidly cleared from CSF, however, elevated levels could have been missed. Unmedicated patients tended to have lower levels than medicated patients. Key Words. Neurodegeneration,
neurodevelopment,
glycolytic
enzyme.
Schizophrenia is a chronic psychotic illness that is often associated with alterations in brain structure. Studies using pneumoencephalography, computed tomography (CT), and magnetic resonance imaging (MRI) to assess brain morphometry have repeatedly demonstrated enlarged ventricles and cortical thinning (Shelton et al., 1986; Suddath et al., 1989). Morphometric analyses of temporal lobe structures indicate that the volumes of the hippocampus, amygdala, and temporal lobes are also reduced (Brown et al., 1986; Suddath et al., 1989; Bogarts et al., 1990). A study of monozygotic twins discordant for schizophrenia found similar structural abnormalities in 14 of 15 affected twins, relative to their unaffected siblings. This finding suggests that most patients with schizophrenia may have structural brain abnormalities when compared with appropriate control subjects (Suddath et al., 1990).
Michael F. Egan, M.D., Rif S. El-Mallakh, M.D.. and Richard L. Suddath, M.D., were all Medical Staff Fellows in the Neuropsychiatry Branch of the National Institute of Mental Health at the time this study was done. James B. Lohr, M.D., is an Associate Professor in the Department of Psychiatry at the University of California at San Diego School of Medicine. H. Stefan Bracha, M.D., is an Associate Professor in the Departments of Psychiatry and Neurology at the University of Arkansas for Medical Sciences. Richard Jed Wyatt, M.D., is Chief of the Neuropsychiatry Branch of the National Institute of Mental Health, Washington, DC. (Reprint requests to Dr. M.F. Egan, Neuropsychiatry Branch, NIMH Neuropsychiatric Research Hospital at St. Elizabeths, 2700 Martin Luther King, Jr., Ave, S.E., Washington, DC 20032, USA.) 0165-1781/92/$05.00
@ 1992 Elsevier Scientific
Publishers
Ireland
Ltd
188 Neurodevelopmental models (Weinberger, 1987) suggest that structural brain changes occur early in development, possibly in utero, as a result of a viral infection (Torrey and Kaufmann, 1986; O’Callaghan et al., 1991) or aberrant neuronal migration (Jacob and Beckmann, 1986). A second, but not mutually exclusive, possibility is that a neurodegenerative or neurotoxic process leads to neuronal cell death or loss of neuronal terminals during acute phases of the illness (Wyatt et al., 1989; Wyatt, 1991). A recent study of first-break patients suggests that ventriculomegaly is present very early in the illness, whereas temporal lobe changes develop later (DeLisi et al., 1991). Although this finding needs confirmation, it offers support for the neurodegenerative hypotheses. Neuron-specific enolase (NSE) is a glycolytic enzyme that is found primarily in neurons and neuroendocrine cells. It has been used in neurological research (Royds et al., 1983; Vermuyten et al., 1990) as a marker for and measure of central nervous system (CNS) damage in patients suffering from strokes, Huntington’s disease, Alzheimer’s disease, and Creutzfeld-Jacob disease. NSE appears to be elevated during the acute phases of these illnesses when ongoing neuronal degeneration is more likely. One study of patients with schizophrenia (Vermuyten et al., 1990) found elevated NSE levels in 16 patients compared with 37 control subjects. While no clinical data were given regarding illness duration or onset (acute vs. gradual), this elevation is consistent with a neurodegenerative process. To investigate the possibility of ongoing neurodegeneration, we examined NSE concentrations in the cerebrospinal fluid (CSF) and serum of 50 patients with recentonset or chronic schizophrenia. This group included 14 patients who donated CSF while they were and were not receiving neuroleptics. We anticipated that a subgroup of patients would have a marked elevation of NSE and that this elevation would be correlated with the degree of psychosis or proximity of onset.
Methods Patients with schizophrenia were recruited from three hospitals. Informed consent was obtained for lumbar punctures from 3 I inpatients at the Neuropsychiatric Research Hospital of the National Institute of Mental Health (NIMH) at St. Elizabeths in Washington, DC (NIMH patients), from IO inpatients from the University of Oregon Health Sciences Center, and from 9 inpatients at the Veterans Administration Hospital in San Diego, California. This third group included five recent-onset, never-medicated schizophrenic patients. All subjects were diagnosed by board-certified psychiatrists according to DSM-III-R criteria (American Psychiatric Association, 1987). Subtype diagnoses were as follows: undifferentiated (n = 30), paranoid (n = 17), and disorganized (n = 3). Twenty-three patients had a documented history of previous drug or alcohol abuse. The average (& SD) duration of illness was 7.4 5 6.4 years, while the average age was 29.6 k 6.7 years. Of the patient group, 38 were men and 12 were women. For the control subjects, the average age was 36.4 * 10.4 years; 14 were men and 9 were women. Medical histories were obtained, and physical examinations were performed. All patients were screened with the following laboratory tests: electrolytes, glucose, creatinine, complete blood count, liver enzymes, thyroid studies, calcium, vitamin B12, folate, urinalysis, electrocardiogram. and chest x-ray. Patients at the NIMH Neuropsychiatric Research Hospital also were screened for the human immunodeficiency virus, and underwent electroencephalograms and brain CT or MRI scans. Patients were excluded if there was any clinical or laboratory evidence of medical or neurological illnesses other than schizophrenia or tardive dyskinesia (TD).
189
The NIMH patients were admitted to the research hospital for participation in a variety of protocols. Patients at the NIMH Neuropsychiatric Research Hospital also were screened for the human immunodeficiency virus, and underwent electroencephalograms and brain CT or MRI scans. Patients were excluded if there was any clinical or laboratory evidence of medical or neurological illnesses other than schizophrenia or tardive dyskinesia (TD). The NIMH patients were admitted to the research hospital for participation in a variety of protocols. Several weeks after admission, patients were placed on coded active neuroleptic medications. Both patients and nursing staff were unaware of whether medication was active or inactive. Coded medications were later switched to an identical placebo for 2 to 6 weeks, followed by a second trial with coded active medications. Lumbar punctures were usually performed 2 to 6 weeks after the discontinuation of active medication while patients were on placebo. Several patients had not been receiving neuroleptics for longer periods. Fourteen NIMH patients also had lumbar punctures while they were receiving coded active neuroleptics. Most patients also received coded active anticholinergic agents when they were given active neuroleptics. Two patients were taking birth control pills. Psychopathology was rated during active and placebo phases by nursing staff members who used the Psychiatric Symptom Assessment Scale (PSAS; Bigelow and Berthot, 1989). Only ratings from medication-free periods were used for correlational analysis. Patients from the University of Oregon Health Sciences Center were admitted for treatment of acute psychotic symptoms, Three patients were receiving medication and seven were medication-free at the time of the lumbar puncture. Patients from the San Diego VA Hospital were also admitted for treatment of acute psychotic symptoms. Five patients had recent-onset schizophrenia and had not previously been treated with neuroleptics. Four patients had chronic schizophrenia and had been medicated for variable lengths of time. Samples of both CSF and serum were not available for all patients. Control subjects were recruited from the community and staff at the NIM H Neuropsychiatric Research Hospital. After giving informed consent, subjects had a screening interview with board-certified psychiatrists and were excluded if medical or psychiatric diagnoses, including substance abuse. were present. A second control group of three patients with active neurological disease was also included. One of these patients had had a small stroke documented by CT scan. Two other patients had metastatic brain lesions. Samples were collected from 50 patients and 24 control subjects. CSF and serum were collected from the same subject in 26 cases. Fourteen NIMH patients donated CSF both while they were receiving and not receiving medication (referred to as the “paired samples”). CSF was collected from 34 patients with chronic schizophrenia who were not receiving medications (chronic unmedicated group), including I4 from the paired NIM H samples, I3 NIMH patients who donated CSF only while they were drug-free, and 7 unmedicated patients from Oregon. Twenty-two patients donated CSF while they were receiving medication (chronic medicated group), including the I4 paired samples, I NIMH patient who was only studied while receiving medication, 3 patients from Oregon, and 4 patients from San Diego, Five additional patients from San Diego had schizophrenia of recent onset and had never been medicated (new-onset group). From the 24 control subjects. one donated only serum while 23 donated CSF. including three control subjects with neurological illnesses (neurological control group). This included two patients with metastatic brain tumors and one patient who had recently (within 3 days) suffered a cerebral infarction. Serum was obtained from I4 patients with chronic schizophrenia while they were unmedicated. I I patients who were receiving medications, and 4 control subjects. Paired serum samples were available from three patients. No serum samples were available from the new-onset group. CSF was collected by lumbar puncture; assayed for glucose, protein, and cells; and cultured for evidence of infection. Samples with abnormalities. such as those containing red or white blood cells or protein elevations, were excluded from further study. Samples were collected over a IO-year period and stored at -70 “C. New-onset patient samples were collected over a 3-year period. Levels of NSE were measured in duplicate with a double-antibody solid phase iodinated
190 radioimmunoassay (RIA; Parma et al., 1981). The assay has been demonstrated to detect levels as low as 108 pg of NSE per assay. Standard curves are linear in the range of 100 pg to 20 ng. Levels of NSE as low as 1 to 2 ng/ml are easily detectable. There is no cross-reactivity with the alpha enolase subunit (Parma et al., 1981). Intra-assay variances are routinely less than 5%, while interassay variability ranges from 5% to 10%. For this sample, there was a significant correlation (Pearson’s r = 0.53; p = 0.0001) between split samples. Medicated and unmedicated chronic groups from different hospitals were compared by an unpaired t test. When no significant differences were found, all chronic medicated patients were grouped together. The same was done for all chronic, unmedicated patients. Unpaired t tests were used to compare the new-onset, chronic medicated, and chronic unmedicated patients with control subjects for CSF and serum levels of NSE. Both a paired t test and a Wilcoxon signed ranks test were used to compare the paired CSF samples from the 14 NIMH patients who had samples collected while they were and were not receiving medications. Additional 1 tests were used to compare subgroups of chronic patients on the basis of age, sex, and duration of illness. For these comparisons, p = 0.005 was accepted as the level of significance. Pearson’s correlation coefficient was used to examine correlations with age, duration of illness, and sample time in freezer storage.
Results There were no differences in serum NSE concentrations among any of the groups (see Table 1). For CSF, there were no differences between medicated or unmedicated chronic groups from different hospitals. These groups were therefore combined. Levels of NSE in CSF in chronic medicated patients (mean = 1.32, SD = 0.73) did not differ from those in control subjects (mean = 1.34, SD = 0.79). Compared with control subjects, chronic unmedicated patients had lower CSF NSE levels (t = 1.97, df= 55, p = 0.05). Chronic unmedicated patients also tended to have lower CSF
Table 1. Neuron-soecific enolase (NSEJ concentrations patients with sch’izophrenia ’ ’ Duration of illness n
Diagnosis
Gender
Age
WI
in controls and
NSE lev @g/ml) CSF
Serum
Controls Normal
20
13M/7F
NA
1.34i- 0.79 7.87 t 1.97
43.5 + 21.3
lMI2F
NA
3.9 + 0.44
29.6 +
18Ml4F
9.25 f 5.5
1.26 + 0.69
46
f
3.2
(n = 4) Neurologic
Patients Chronic
disorder
3
-
with schizophrenia (medicated]
22
6.1
6.25 i 2.21 (n = 11)
Chronic
(unmedicated)
Recent-onset
34
5
29.6 xk 6.7 24MilOF
35.2 k
2.4
5M
6.6
f 6.5
< 0.5
0.96 f 0.61
6.66 f 2.61 (n = 14)
0.58 + 0.38
-
Note. Serum levels were available for only a subset of each group. Then for each subset is listed below the respective serum NSE levels. Demographic data include all subjects for each group. The neurologic control group (neuro control) includes 2 patients with me&static brain tumors and 1 patient with a recent cerebral Infarct. Values are presented as mean f SD.
191 NSE levels when compared with those in chronic medicated patients (t = -1.90, df = 54, p = 0.06). In 14 paired samples, NSE concentrations for unmedicated patients (mean = 1.1, SD = 0.69) were lower in I1 cases (Wilcoxon signed rank, z = -2.04, p = 0.04) compared with those in the same patients when they were receiving medications (mean = 1.30, SD = 0.81). There did not appear to be a subgroup of patients with abnormally elevated levels (see Fig. 1).
Fig. 1. Scattergram for cerebrospinal fluid levels of neuron-specific enolase in patient and control groups 8 5
. F
.= CONTROL
NEURO CHRONIC CHRONIC CONTROLS UNMED MEDICATED
TEk-
New-onset patients had lower CSF concentrations of NSE when compared with chronic medicated patients (t = 2.17, df = 22, p = 0.04) and control subjects (t = 2.06, cif= 23, p = 0.05), but not when compared with chronic unmedicated patients (t = 1.47, df= 40, p = 0.15). Concentrations for the three neurological control subjects were higher than all other values except for that in one unmedicated patient with chronic schizophrenia. When compared with concentrations in control subjects, CSF concentrations in acute and chronic patients were significantly elevated (p < 0.002 for all groups). There were no significant correlations between NSE levels and age, duration of illness, PSAS total scores (average score = 28.9, SD = 16.5), or the length of time samples remained in freezer storage.
Discussion We did not find elevated NSE concentrations in the CSF of patients with chronic schizophrenia that would be consistent with ongoing neuronal degeneration. Furthermore, CSF NSE levels were not correlated with degree of psychosis and were not elevated in a small group of new-onset patients. This study fails to replicate a previous report (Vermuyten et al., 1990) of NSE elevations in patients with schizophrenia, relative to control subjects. A third of the patients in the previous study had levels higher than the highest control value. The absence of a significant elevation of NSE in the present study suggests that there may not be a progressive neurodegenerative process in these patient groups, or that such a process occurs at a time other than
192 when these samples were collected. Because NSE is cleared from CSF within several days of an acute insuh(Hay et al., 1984) increases could have been missed. Finally, our assay may not possess the required sensitivity, or neurodegenerative changes in this illness may not be associated with elevated NSE in the CSF. In general, NSE appears to be a specific and relatively sensitive measure of neurodegeneration. Studies of other diseases involving some degree of neurodegeneration, such as epilepsy, transient ischemic attacks, and even migraines, are associated with elevations of NSE in the majority of cases. Even higher levels are seen with acute processes such as herpes simplex encephalitis, strokes, and tumors (Royds et al., 1983). Studies of more chronic diseases, such as Huntington’s disease (Marangos and Paul, 1981), Alzheimer’s disease (Royds et al., 1983) and Creutzfeld-Jacob disease (Wakayama et al., 1987; Vermuyten et al., 1990; Van den Doe1 et al., 1988) have shown elevations in patients above the highest normal ranges in at least some cases. Results, however, are not always consistent. For example, in five studies that examined at CSF NSE in Alzheimer’s disease, two (Royds et al., 1983; Vermuyten et al., 1990) found at least 20% of cases elevated above the highest control levels. In three other studies, two found no differences between patients and control subjects (Sulkava et al., 1988; Van den Doe1 et al., 1988) while a third (Cutler et al., 1986) found a reduction of CSF NSE after several years of illness. The authors of the latter study concluded that lower CSF NSE levels were a reflection of CNS cell loss. A similar pattern may also be present in Creutzfeld-Jacob disease (for discussion, see El-Mallakh et al.. in press), where levels may also be elevated in the early stages (Wakayama et al., 1987; Van den Doe1 et al., 1988) and then fall below control levels at later stages (Vermuyten et al., 1990). It may be that NSE levels are only elevated during the acute phase of a neurodegenerative process. With progression, levels may drop to normal or below normal levels (El-Mallakh et al., in press). Control levels of CSF NSE in this study were lower than those of other studies. Royds et al. (1983) and Hay et al. (1984) reported an average (* SD) in normal control subjects of 3.0 f 2.4 ng/ml, while Sulkava et al. (1988) (8.6 f 0.67 ng/ml) and Cutler et al. (1986) (15.0 + 1.22 ng/ml) found even higher levels. Our normal control values were closer to those reported by Vermuyten et al. (1990) (2.0 ng/ml) and Parma et al. (1981) (2.2 f 0.5 ng/ml), the latter of whom used a procedure similar to ours. Differences between assays have been previously noted (Kaiser et al., 1989; Vermuyten et al., 1990) and could be due to different methods of antibody preparation or purification. A second methodological problem with studies of CSF NSE is the method of storage. It is unclear how the collection and storage process used in this and all other published CSF studies affects antigenicity. It is possible that there is a significant loss of NSE antigenicity that could be avoided by storage in a cryprotectant. Current studies are examining this issue. Nevertheless, other studies that used this method of collection and storage have reported marked elevations in CSF NSE in patients with acute neurodegenerative processes. While the neurological control group used in this study is very small, elevations in all three patients by at least 2 SD above the mean normal control level suggest that the method used here can detect abnormal CSF levels. It seems unlikely that this method would preferentially reduce antigenicity in
193 the CSF of patients with schizophrenia compared with normal control subjects. Further studies may help to clarify this issue. CSF NSE levels in medication-free schizophrenic patients tended to be slightly lower than in control subjects or patients who were receiving neuroleptic medication. Similarly, patients who donated CSF samples both while they were and were not receiving medication had lower NSE levels when they were drug-free. New-onset patients also had lower levels than control subjects, although the low number of patients in this group makes such comparisons difficult. Several factors could account for these differences. Medication-free and new-onset patients could have reduced levels due to a slight reduction in cerebral volume (Weinberger et al., 1979; Shelton and Weinberger, 1986). This could lead to a reduction in the normal amount of NSE produced by the brain, as has been suggested for other neurodegenerative diseases (Cutler et al., 1986; El-Mallakh et al., in press). The increase with neuroleptic medication could reflect a slight increase in neurodegeneration due to a neurotoxic effect (Cadet et al., 1986). On the other hand, NSE levels, in the absence of an acute neurological process, might reflect cerebral metabolic activity. Although this is highly speculative, it is noteworthy that NSE is an important enzyme in glycolysis. In this case, reduced NSE in medication-free patients would reflect reduced cerebral metabolism (Buchsbaum et al., 1982; Weinberger and Berman, 1988). Increases with neuroleptic treatment could reflect recovery of cerebral metabolism (Szechtman et al., 1988). It is also possible that reduced CSF NSE levels are due to trivial factors such as diet, activity, or changes in sleep patterns. While the finding of a slight reduction in CSF NSE levels in neuroleptic-free chronic patients is intriguing, several methodological problems necessitate a cautious interpretation. First, at these low levels, intra-assay variability is very high. A more sensitive assay is needed for an accurate investigation of such small differences. The use of a monoclonal antibody might provide increased sensitivity. A second methodological problem involves the permeability of the blood-brain barrier. Serum levels of NSE in this study were much higher than CSF levels. Alterations in bloodbrain barrier permeability due to the illness or medications could affect CSF levels. Third, little is known about the mechanisms involved in the efflux of NSE into CSF. This may be a passive process whereby protein is exuded as neurons lose membrane integrity. Alternatively, the efflux could be due to more active processes, such as synaptic loss or sprouting. Active pumps or transport mechanisms could also be involved. Thus, the finding of a difference between medicated and unmedicated patients may be unrelated to differences in neurodegeneration. In conclusion, we failed to find evidence for ongoing neurodegeneration when NSE levels were used as an index in patients with acute and chronic schizophrenia. We may have failed to detect changes due to the short half-life of NSE in the CSF. NSE is elevated in a number of other neurodegenerative illnesses during the acute neurodegenerative stages. It may be a relatively specific index of such disorders. As expected, elevations were seen in this study in three neurological control cases with tumors or stroke. A reduction in CSF NSE was observed in medication-free patients and in patients with schizophrenia of recent onset. Interpretation of this finding is confounded by the relatively low sensitivity of the assay at these low concentrations.
194
Further studies with more sensitive of this change.
assays are needed to understand
the significance
Acknowledgments. We thank Dianon Systems, Inc., of Stratford, CT, for donating NSE assays. We also thank Janice Stevens for providing CSF and Ahmed Elkashef for his thoughtful input.
References Psychiatric Association. DSM-III-R: Diagnostic and Statistical Manual of Mental 3rd ed., revised. Washington, DC: American Psychiatric Press, 1987. pp. 187-203. Bigelow, L.B., and Berthot, B.D. The Psychiatric Symptom Assessment Scale (PSAS).
American
Disorders.
Psychopharmacology
Bulletin,
25(2): 168-179,
1989.
Bogarts, B.; Ashtari,
M.; Degreef, G.; Alvir, J.M.J.; Bilder, R.M.; and Lieberman, J.A. limbic structure volumes on magnetic resonance images in first episode
Reduced temporal schizophrenia. Psychiatry
Research:
Neuroimaging,
35: I-1 3, 1990.
Brown, R.; Colter, N.; Corsellis,
J.A.N.; Crow, T.J.; Frith, C.D.; Jagoe, R.; Johnstone, EC.; and Marsh, L. Postmortem evidence of structural brain changes in schizophrenia. Archives of General Psychiatry, 43:36-42, 1986. Buchsbaum, M.S.; Ingvar, D.H.; Kessler, R.; Waters, R.N.; Cappelletti, J.; van Kammen, D.P.; King, A.C.; Johnson, J.L.; Manning, R.G.; Flynn, R.W.; Mann, L.S.; Bunney, W.E.; and Sokoloff, L. Cerebral glucography with positron tomography: Use in normal subjects and in patients with schizophrenia. Archives of General Psychiatry, 39:25 I-259, 1982. Cadet, J.L.; Lohr, J.B.; and Jeste, D.V. Free radicals and tardive dyskinesia. Trends in Neuroscience,
9: 107-108,
1986.
Cutler, N.R.; Kay, A.D.; Marangos, P.J.; and Burg, C. Cerebrospinal fluid neuron-specific enolase is reduced in Alzheimer’s disease. Archives of Neurology, 43:153-154, 1986. DeLisi L.E.; Hoff, A.L.; Schwartz, J.E.; Shields, G.W.; Halthore, S.N.; Gupta, S.M.; Henn, F.A.; and Anand, A.K. Brain morphology in first episode schizophrenia-like psychotic patients: A quantitative magnetic resonance imaging study. Biological Psychiatry, 29: 159-l 75, 1991.
El-Mallakh, R.S.; Egan, M.F.; Intracellular enzymes serving as neuropsychiatric disorders. Psychiatry, Hay, E.; Royds J.A.; Davies-Jones, Cerebrospinal fluid enolase in stroke 471724-729,
and Wyatt, R.J. Creatine phosphate and enolase: markers of central nervous system damage in in press. G.B.; Lewtas, N.A.; Timperley, W.R.; and Taylor, C.B. Journal
of Neurology,
Neurosurgery,
and Psychiatry,
1984.
Jacob, H., and Beckmann, H. Prenatal developmental disturbances in the limbic allocortex in schizophrenics. Journal of Neural Transmission, 65:303-326, 1986. Kaiser, E.; Kuzmits, R.; Pregnant, P.; Burghuber, 0.; and Worofka, W. Clinical biochemistry of neuron specific enolase. Clinica Chimica Acta, 183:13-32, 1989. Marangos, P.J., and Paul, S.M. Brain levels of neuron-specific enolase and nonneuronal enolase in Huntington’s disease. Journal of Neurochemistry, 37(5): 1338-1340, 198 1. O’Callaghan, E.; Sham, P.; Takei, N.; Glover, G.; and Murray, R.M. Schizophrenia after prenatal exposure to 1957 A2 influenza epidemic. Lancet, 337: 1248-1250, 1991. Parma, A.M.; Marangos, P.J.; and Goodwin F.K. A more sensitive radioimmunoassy for neuron-specific enolase suitable for cerebrospinal fluid determination. Journal of Neurochemistry, 36(3): 1093-1096, 198 1.
195 Royds, J.A.; Davies-Jones G.B.; Lewtas, N.A.; Timperlay, W.R.; and Taylor, C. Enolase isoenzymes in the cerebrospinal fluid of patients with diseases of the nervous system. Journal of Neurology, Neurosurgery. and Psychiatry, 46:1031-1036, 1983. Shelton, R.C., and Weinberger, D.R. Computerized tomography in schizophrenia: A review and synthesis. In: Nasarallah, H.A., and Weinberger, D.R., eds. Handbook of Schizophrenia, Vol 1: The Neurology of Schizophrenia. Amsterdam: Elsevier Science Publishers, 1986. pp. 207-250. Suddath, R.L.; Casanova, M.F.; Goldberg, T.E.; Daniel, D.G.; Kelsoe, J.R.; and Weinberger, D.R. Temporal lobe pathology in schizophrenia: A quantitative magnetic resonance imaging study. American Journal of Psychiatry, 146(4):464-472, 1989. Suddath, R.L.; Christison, G.W.; Torrey, E.F.; Casanova, M.F.; and Weinberger, D.R. Anatomical abnormalities in the brains of monozygotic twins discordant for schizophrenia. Nen, England Journal of Medicine, 3221789-794. 1990. Sulkava R.; Viinikka, L.; Erkinjuntti T.; and Roine, R. Cerebrospinal fluid neuron-specific enolase is decreased in multi-infarct dementia, but unchanged in Alzheimer’s disease. Journal of Neurology, Neurosurgery, and Psychiatry, 5 1549-55 1, 1988. Szechtman, H.; Nahmias, C.; Garnett, S.; Firnau, G.; Brown, G.M.; Kaplan, R.D.; and Cleghorn, J.M. Effect of neuroleptics on altered cerebral glucose metabolism in schizophrenia. Archives of General Psychiatry, 45523-532, 1988. Torrey, E.F., and Kaufman, C.A. Schizophrenia and neuroviruses. In: Nasrallah, H.A., and Weinberger, D.R., eds. Handbook of Schizophrenia, Vol I: The Neurology of Schizophrenia. Amsterdam: Elsevier Science Publishers, 1986. pp. 361-376. Van Den Doel, E.M.H.; Rijksen, G.; and Staal, G.E.J. Interet et limites de I’etude de la gamma-enolsae dans le liquide cephalo-rachidien comme temoin de la degradation neuronale. Revue
Neurologique.
144:452-455,
1988.
Vermuyten, K.; Lowenthal, A.; and Kracher, D. Detection of neuron specific enolase concentrations in cerebrospinal fluid from patients with neurological disorders by means of a sensitive enzyme immunoassay. Clinica Chimica Acta, 187:69-78, 1990. Wakayama, Y.; Shibuya, S.; Kawase, J.; Sagawa, F.; and Hashizume, Y. High neuron specific enolase level of cerebrospinal fluid in the early stage of Creutzfeld-Jacob disease. Khnische
Wochenschrtft,
65~798-801,
1987.
Weinberger, D.R. Implications of normal development for the pathogenesis of schizophrenia. Archives of General Psychiatry, 44:660-669, 1987. Weinberger, D.R., and Berman, K.F. Speculations on the meaning of cerebral metabolic hypofrontality in schizophrenia. Schizophrenia Bulletin, 14:157-168, 1988. Weinberger, D.R.; Torrey, E.F.; Neophytides, A.N.; and Wyatt, R.J. Structural abnormalities of the cerebral cortex in chronic schizophrenia. Archives of General Psvchiutry, 361735-739, 1979. Wyatt R.J. Neuroleptics and the natural course of schizophrenia. Schizophrenia Bulletin, 17:325-351,
1991.
Wyatt, R.J.; Fawcett, R.; and Kirch D.G. A possible animal model of defect state schizophrenia. In: Schulz, SC., and Tamminga, C.A., eds. Schizophrenia: Scientific Progress. New York: Oxford University Press, 1989. pp. 184-189.
Psychiatry Research, 43:187-195
Elsevier
Cerebrospinal Fluid and Serum Levels of Neuron-Specific Enolase in Patients With Schizophrenia Michael F. Egan, Rif S. El-Mallakh, Richard H. Stefan Bracha, and Richard Jed Wyatt Received
October
7, 1991; revised
version received
L. Suddath,
April
James
14, 1992; accepted
B. Lohr,
June
20, 1992.
Abstract. Some patients with schizophrenia appear to have brain abnormalities, including enlarged third and lateral ventricles and reduced volumes of temporal lobe structures. These abnormalities could be attributed to a developmental abnormality or a neurodegenerative process. Neuron-specific enolase (NSE), a protein that is found primarily in neurons and neuroendocrine cells, has been used as an index of neuronal damage or degeneration. Levels of NSE in cerebrospinal fluid (CSF) and serum from 50 patients with acute and chronic schizophrenia were compared with those in normal and neurological control subjects. A double-antibody, solid phase iodinated radioimmunoassay was used to determine NSE levels. There was no evidence of elevated levels in patients with schizophrenia, whereas control subjects with neurological illnesses had increased levels of NSE in CSF. Because NSE is rapidly cleared from CSF, however, elevated levels could have been missed. Unmedicated patients tended to have lower levels than medicated patients. Key Words. Neurodegeneration,
neurodevelopment,
glycolytic
enzyme.
Schizophrenia is a chronic psychotic illness that is often associated with alterations in brain structure. Studies using pneumoencephalography, computed tomography (CT), and magnetic resonance imaging (MRI) to assess brain morphometry have repeatedly demonstrated enlarged ventricles and cortical thinning (Shelton et al., 1986; Suddath et al., 1989). Morphometric analyses of temporal lobe structures indicate that the volumes of the hippocampus, amygdala, and temporal lobes are also reduced (Brown et al., 1986; Suddath et al., 1989; Bogarts et al., 1990). A study of monozygotic twins discordant for schizophrenia found similar structural abnormalities in 14 of 15 affected twins, relative to their unaffected siblings. This finding suggests that most patients with schizophrenia may have structural brain abnormalities when compared with appropriate control subjects (Suddath et al., 1990).
Michael F. Egan, M.D., Rif S. El-Mallakh, M.D.. and Richard L. Suddath, M.D., were all Medical Staff Fellows in the Neuropsychiatry Branch of the National Institute of Mental Health at the time this study was done. James B. Lohr, M.D., is an Associate Professor in the Department of Psychiatry at the University of California at San Diego School of Medicine. H. Stefan Bracha, M.D., is an Associate Professor in the Departments of Psychiatry and Neurology at the University of Arkansas for Medical Sciences. Richard Jed Wyatt, M.D., is Chief of the Neuropsychiatry Branch of the National Institute of Mental Health, Washington, DC. (Reprint requests to Dr. M.F. Egan, Neuropsychiatry Branch, NIMH Neuropsychiatric Research Hospital at St. Elizabeths, 2700 Martin Luther King, Jr., Ave, S.E., Washington, DC 20032, USA.) 0165-1781/92/$05.00
@ 1992 Elsevier Scientific
Publishers
Ireland
Ltd
188 Neurodevelopmental models (Weinberger, 1987) suggest that structural brain changes occur early in development, possibly in utero, as a result of a viral infection (Torrey and Kaufmann, 1986; O’Callaghan et al., 1991) or aberrant neuronal migration (Jacob and Beckmann, 1986). A second, but not mutually exclusive, possibility is that a neurodegenerative or neurotoxic process leads to neuronal cell death or loss of neuronal terminals during acute phases of the illness (Wyatt et al., 1989; Wyatt, 1991). A recent study of first-break patients suggests that ventriculomegaly is present very early in the illness, whereas temporal lobe changes develop later (DeLisi et al., 1991). Although this finding needs confirmation, it offers support for the neurodegenerative hypotheses. Neuron-specific enolase (NSE) is a glycolytic enzyme that is found primarily in neurons and neuroendocrine cells. It has been used in neurological research (Royds et al., 1983; Vermuyten et al., 1990) as a marker for and measure of central nervous system (CNS) damage in patients suffering from strokes, Huntington’s disease, Alzheimer’s disease, and Creutzfeld-Jacob disease. NSE appears to be elevated during the acute phases of these illnesses when ongoing neuronal degeneration is more likely. One study of patients with schizophrenia (Vermuyten et al., 1990) found elevated NSE levels in 16 patients compared with 37 control subjects. While no clinical data were given regarding illness duration or onset (acute vs. gradual), this elevation is consistent with a neurodegenerative process. To investigate the possibility of ongoing neurodegeneration, we examined NSE concentrations in the cerebrospinal fluid (CSF) and serum of 50 patients with recentonset or chronic schizophrenia. This group included 14 patients who donated CSF while they were and were not receiving neuroleptics. We anticipated that a subgroup of patients would have a marked elevation of NSE and that this elevation would be correlated with the degree of psychosis or proximity of onset.
Methods Patients with schizophrenia were recruited from three hospitals. Informed consent was obtained for lumbar punctures from 3 I inpatients at the Neuropsychiatric Research Hospital of the National Institute of Mental Health (NIMH) at St. Elizabeths in Washington, DC (NIMH patients), from IO inpatients from the University of Oregon Health Sciences Center, and from 9 inpatients at the Veterans Administration Hospital in San Diego, California. This third group included five recent-onset, never-medicated schizophrenic patients. All subjects were diagnosed by board-certified psychiatrists according to DSM-III-R criteria (American Psychiatric Association, 1987). Subtype diagnoses were as follows: undifferentiated (n = 30), paranoid (n = 17), and disorganized (n = 3). Twenty-three patients had a documented history of previous drug or alcohol abuse. The average (& SD) duration of illness was 7.4 5 6.4 years, while the average age was 29.6 k 6.7 years. Of the patient group, 38 were men and 12 were women. For the control subjects, the average age was 36.4 * 10.4 years; 14 were men and 9 were women. Medical histories were obtained, and physical examinations were performed. All patients were screened with the following laboratory tests: electrolytes, glucose, creatinine, complete blood count, liver enzymes, thyroid studies, calcium, vitamin B12, folate, urinalysis, electrocardiogram. and chest x-ray. Patients at the NIMH Neuropsychiatric Research Hospital also were screened for the human immunodeficiency virus, and underwent electroencephalograms and brain CT or MRI scans. Patients were excluded if there was any clinical or laboratory evidence of medical or neurological illnesses other than schizophrenia or tardive dyskinesia (TD).
189
The NIMH patients were admitted to the research hospital for participation in a variety of protocols. Patients at the NIMH Neuropsychiatric Research Hospital also were screened for the human immunodeficiency virus, and underwent electroencephalograms and brain CT or MRI scans. Patients were excluded if there was any clinical or laboratory evidence of medical or neurological illnesses other than schizophrenia or tardive dyskinesia (TD). The NIMH patients were admitted to the research hospital for participation in a variety of protocols. Several weeks after admission, patients were placed on coded active neuroleptic medications. Both patients and nursing staff were unaware of whether medication was active or inactive. Coded medications were later switched to an identical placebo for 2 to 6 weeks, followed by a second trial with coded active medications. Lumbar punctures were usually performed 2 to 6 weeks after the discontinuation of active medication while patients were on placebo. Several patients had not been receiving neuroleptics for longer periods. Fourteen NIMH patients also had lumbar punctures while they were receiving coded active neuroleptics. Most patients also received coded active anticholinergic agents when they were given active neuroleptics. Two patients were taking birth control pills. Psychopathology was rated during active and placebo phases by nursing staff members who used the Psychiatric Symptom Assessment Scale (PSAS; Bigelow and Berthot, 1989). Only ratings from medication-free periods were used for correlational analysis. Patients from the University of Oregon Health Sciences Center were admitted for treatment of acute psychotic symptoms, Three patients were receiving medication and seven were medication-free at the time of the lumbar puncture. Patients from the San Diego VA Hospital were also admitted for treatment of acute psychotic symptoms. Five patients had recent-onset schizophrenia and had not previously been treated with neuroleptics. Four patients had chronic schizophrenia and had been medicated for variable lengths of time. Samples of both CSF and serum were not available for all patients. Control subjects were recruited from the community and staff at the NIM H Neuropsychiatric Research Hospital. After giving informed consent, subjects had a screening interview with board-certified psychiatrists and were excluded if medical or psychiatric diagnoses, including substance abuse. were present. A second control group of three patients with active neurological disease was also included. One of these patients had had a small stroke documented by CT scan. Two other patients had metastatic brain lesions. Samples were collected from 50 patients and 24 control subjects. CSF and serum were collected from the same subject in 26 cases. Fourteen NIMH patients donated CSF both while they were receiving and not receiving medication (referred to as the “paired samples”). CSF was collected from 34 patients with chronic schizophrenia who were not receiving medications (chronic unmedicated group), including I4 from the paired NIM H samples, I3 NIMH patients who donated CSF only while they were drug-free, and 7 unmedicated patients from Oregon. Twenty-two patients donated CSF while they were receiving medication (chronic medicated group), including the I4 paired samples, I NIMH patient who was only studied while receiving medication, 3 patients from Oregon, and 4 patients from San Diego, Five additional patients from San Diego had schizophrenia of recent onset and had never been medicated (new-onset group). From the 24 control subjects. one donated only serum while 23 donated CSF. including three control subjects with neurological illnesses (neurological control group). This included two patients with metastatic brain tumors and one patient who had recently (within 3 days) suffered a cerebral infarction. Serum was obtained from I4 patients with chronic schizophrenia while they were unmedicated. I I patients who were receiving medications, and 4 control subjects. Paired serum samples were available from three patients. No serum samples were available from the new-onset group. CSF was collected by lumbar puncture; assayed for glucose, protein, and cells; and cultured for evidence of infection. Samples with abnormalities. such as those containing red or white blood cells or protein elevations, were excluded from further study. Samples were collected over a IO-year period and stored at -70 “C. New-onset patient samples were collected over a 3-year period. Levels of NSE were measured in duplicate with a double-antibody solid phase iodinated
190 radioimmunoassay (RIA; Parma et al., 1981). The assay has been demonstrated to detect levels as low as 108 pg of NSE per assay. Standard curves are linear in the range of 100 pg to 20 ng. Levels of NSE as low as 1 to 2 ng/ml are easily detectable. There is no cross-reactivity with the alpha enolase subunit (Parma et al., 1981). Intra-assay variances are routinely less than 5%, while interassay variability ranges from 5% to 10%. For this sample, there was a significant correlation (Pearson’s r = 0.53; p = 0.0001) between split samples. Medicated and unmedicated chronic groups from different hospitals were compared by an unpaired t test. When no significant differences were found, all chronic medicated patients were grouped together. The same was done for all chronic, unmedicated patients. Unpaired t tests were used to compare the new-onset, chronic medicated, and chronic unmedicated patients with control subjects for CSF and serum levels of NSE. Both a paired t test and a Wilcoxon signed ranks test were used to compare the paired CSF samples from the 14 NIMH patients who had samples collected while they were and were not receiving medications. Additional 1 tests were used to compare subgroups of chronic patients on the basis of age, sex, and duration of illness. For these comparisons, p = 0.005 was accepted as the level of significance. Pearson’s correlation coefficient was used to examine correlations with age, duration of illness, and sample time in freezer storage.
Results There were no differences in serum NSE concentrations among any of the groups (see Table 1). For CSF, there were no differences between medicated or unmedicated chronic groups from different hospitals. These groups were therefore combined. Levels of NSE in CSF in chronic medicated patients (mean = 1.32, SD = 0.73) did not differ from those in control subjects (mean = 1.34, SD = 0.79). Compared with control subjects, chronic unmedicated patients had lower CSF NSE levels (t = 1.97, df= 55, p = 0.05). Chronic unmedicated patients also tended to have lower CSF
Table 1. Neuron-soecific enolase (NSEJ concentrations patients with sch’izophrenia ’ ’ Duration of illness n
Diagnosis
Gender
Age
WI
in controls and
NSE lev @g/ml) CSF
Serum
Controls Normal
20
13M/7F
NA
1.34i- 0.79 7.87 t 1.97
43.5 + 21.3
lMI2F
NA
3.9 + 0.44
29.6 +
18Ml4F
9.25 f 5.5
1.26 + 0.69
46
f
3.2
(n = 4) Neurologic
Patients Chronic
disorder
3
-
with schizophrenia (medicated]
22
6.1
6.25 i 2.21 (n = 11)
Chronic
(unmedicated)
Recent-onset
34
5
29.6 xk 6.7 24MilOF
35.2 k
2.4
5M
6.6
f 6.5
< 0.5
0.96 f 0.61
6.66 f 2.61 (n = 14)
0.58 + 0.38
-
Note. Serum levels were available for only a subset of each group. Then for each subset is listed below the respective serum NSE levels. Demographic data include all subjects for each group. The neurologic control group (neuro control) includes 2 patients with me&static brain tumors and 1 patient with a recent cerebral Infarct. Values are presented as mean f SD.
191 NSE levels when compared with those in chronic medicated patients (t = -1.90, df = 54, p = 0.06). In 14 paired samples, NSE concentrations for unmedicated patients (mean = 1.1, SD = 0.69) were lower in I1 cases (Wilcoxon signed rank, z = -2.04, p = 0.04) compared with those in the same patients when they were receiving medications (mean = 1.30, SD = 0.81). There did not appear to be a subgroup of patients with abnormally elevated levels (see Fig. 1).
Fig. 1. Scattergram for cerebrospinal fluid levels of neuron-specific enolase in patient and control groups 8 5
. F
.= CONTROL
NEURO CHRONIC CHRONIC CONTROLS UNMED MEDICATED
TEk-
New-onset patients had lower CSF concentrations of NSE when compared with chronic medicated patients (t = 2.17, df = 22, p = 0.04) and control subjects (t = 2.06, cif= 23, p = 0.05), but not when compared with chronic unmedicated patients (t = 1.47, df= 40, p = 0.15). Concentrations for the three neurological control subjects were higher than all other values except for that in one unmedicated patient with chronic schizophrenia. When compared with concentrations in control subjects, CSF concentrations in acute and chronic patients were significantly elevated (p < 0.002 for all groups). There were no significant correlations between NSE levels and age, duration of illness, PSAS total scores (average score = 28.9, SD = 16.5), or the length of time samples remained in freezer storage.
Discussion We did not find elevated NSE concentrations in the CSF of patients with chronic schizophrenia that would be consistent with ongoing neuronal degeneration. Furthermore, CSF NSE levels were not correlated with degree of psychosis and were not elevated in a small group of new-onset patients. This study fails to replicate a previous report (Vermuyten et al., 1990) of NSE elevations in patients with schizophrenia, relative to control subjects. A third of the patients in the previous study had levels higher than the highest control value. The absence of a significant elevation of NSE in the present study suggests that there may not be a progressive neurodegenerative process in these patient groups, or that such a process occurs at a time other than
192 when these samples were collected. Because NSE is cleared from CSF within several days of an acute insuh(Hay et al., 1984) increases could have been missed. Finally, our assay may not possess the required sensitivity, or neurodegenerative changes in this illness may not be associated with elevated NSE in the CSF. In general, NSE appears to be a specific and relatively sensitive measure of neurodegeneration. Studies of other diseases involving some degree of neurodegeneration, such as epilepsy, transient ischemic attacks, and even migraines, are associated with elevations of NSE in the majority of cases. Even higher levels are seen with acute processes such as herpes simplex encephalitis, strokes, and tumors (Royds et al., 1983). Studies of more chronic diseases, such as Huntington’s disease (Marangos and Paul, 1981), Alzheimer’s disease (Royds et al., 1983) and Creutzfeld-Jacob disease (Wakayama et al., 1987; Vermuyten et al., 1990; Van den Doe1 et al., 1988) have shown elevations in patients above the highest normal ranges in at least some cases. Results, however, are not always consistent. For example, in five studies that examined at CSF NSE in Alzheimer’s disease, two (Royds et al., 1983; Vermuyten et al., 1990) found at least 20% of cases elevated above the highest control levels. In three other studies, two found no differences between patients and control subjects (Sulkava et al., 1988; Van den Doe1 et al., 1988) while a third (Cutler et al., 1986) found a reduction of CSF NSE after several years of illness. The authors of the latter study concluded that lower CSF NSE levels were a reflection of CNS cell loss. A similar pattern may also be present in Creutzfeld-Jacob disease (for discussion, see El-Mallakh et al.. in press), where levels may also be elevated in the early stages (Wakayama et al., 1987; Van den Doe1 et al., 1988) and then fall below control levels at later stages (Vermuyten et al., 1990). It may be that NSE levels are only elevated during the acute phase of a neurodegenerative process. With progression, levels may drop to normal or below normal levels (El-Mallakh et al., in press). Control levels of CSF NSE in this study were lower than those of other studies. Royds et al. (1983) and Hay et al. (1984) reported an average (* SD) in normal control subjects of 3.0 f 2.4 ng/ml, while Sulkava et al. (1988) (8.6 f 0.67 ng/ml) and Cutler et al. (1986) (15.0 + 1.22 ng/ml) found even higher levels. Our normal control values were closer to those reported by Vermuyten et al. (1990) (2.0 ng/ml) and Parma et al. (1981) (2.2 f 0.5 ng/ml), the latter of whom used a procedure similar to ours. Differences between assays have been previously noted (Kaiser et al., 1989; Vermuyten et al., 1990) and could be due to different methods of antibody preparation or purification. A second methodological problem with studies of CSF NSE is the method of storage. It is unclear how the collection and storage process used in this and all other published CSF studies affects antigenicity. It is possible that there is a significant loss of NSE antigenicity that could be avoided by storage in a cryprotectant. Current studies are examining this issue. Nevertheless, other studies that used this method of collection and storage have reported marked elevations in CSF NSE in patients with acute neurodegenerative processes. While the neurological control group used in this study is very small, elevations in all three patients by at least 2 SD above the mean normal control level suggest that the method used here can detect abnormal CSF levels. It seems unlikely that this method would preferentially reduce antigenicity in
193 the CSF of patients with schizophrenia compared with normal control subjects. Further studies may help to clarify this issue. CSF NSE levels in medication-free schizophrenic patients tended to be slightly lower than in control subjects or patients who were receiving neuroleptic medication. Similarly, patients who donated CSF samples both while they were and were not receiving medication had lower NSE levels when they were drug-free. New-onset patients also had lower levels than control subjects, although the low number of patients in this group makes such comparisons difficult. Several factors could account for these differences. Medication-free and new-onset patients could have reduced levels due to a slight reduction in cerebral volume (Weinberger et al., 1979; Shelton and Weinberger, 1986). This could lead to a reduction in the normal amount of NSE produced by the brain, as has been suggested for other neurodegenerative diseases (Cutler et al., 1986; El-Mallakh et al., in press). The increase with neuroleptic medication could reflect a slight increase in neurodegeneration due to a neurotoxic effect (Cadet et al., 1986). On the other hand, NSE levels, in the absence of an acute neurological process, might reflect cerebral metabolic activity. Although this is highly speculative, it is noteworthy that NSE is an important enzyme in glycolysis. In this case, reduced NSE in medication-free patients would reflect reduced cerebral metabolism (Buchsbaum et al., 1982; Weinberger and Berman, 1988). Increases with neuroleptic treatment could reflect recovery of cerebral metabolism (Szechtman et al., 1988). It is also possible that reduced CSF NSE levels are due to trivial factors such as diet, activity, or changes in sleep patterns. While the finding of a slight reduction in CSF NSE levels in neuroleptic-free chronic patients is intriguing, several methodological problems necessitate a cautious interpretation. First, at these low levels, intra-assay variability is very high. A more sensitive assay is needed for an accurate investigation of such small differences. The use of a monoclonal antibody might provide increased sensitivity. A second methodological problem involves the permeability of the blood-brain barrier. Serum levels of NSE in this study were much higher than CSF levels. Alterations in bloodbrain barrier permeability due to the illness or medications could affect CSF levels. Third, little is known about the mechanisms involved in the efflux of NSE into CSF. This may be a passive process whereby protein is exuded as neurons lose membrane integrity. Alternatively, the efflux could be due to more active processes, such as synaptic loss or sprouting. Active pumps or transport mechanisms could also be involved. Thus, the finding of a difference between medicated and unmedicated patients may be unrelated to differences in neurodegeneration. In conclusion, we failed to find evidence for ongoing neurodegeneration when NSE levels were used as an index in patients with acute and chronic schizophrenia. We may have failed to detect changes due to the short half-life of NSE in the CSF. NSE is elevated in a number of other neurodegenerative illnesses during the acute neurodegenerative stages. It may be a relatively specific index of such disorders. As expected, elevations were seen in this study in three neurological control cases with tumors or stroke. A reduction in CSF NSE was observed in medication-free patients and in patients with schizophrenia of recent onset. Interpretation of this finding is confounded by the relatively low sensitivity of the assay at these low concentrations.
194
Further studies with more sensitive of this change.
assays are needed to understand
the significance
Acknowledgments. We thank Dianon Systems, Inc., of Stratford, CT, for donating NSE assays. We also thank Janice Stevens for providing CSF and Ahmed Elkashef for his thoughtful input.
References Psychiatric Association. DSM-III-R: Diagnostic and Statistical Manual of Mental 3rd ed., revised. Washington, DC: American Psychiatric Press, 1987. pp. 187-203. Bigelow, L.B., and Berthot, B.D. The Psychiatric Symptom Assessment Scale (PSAS).
American
Disorders.
Psychopharmacology
Bulletin,
25(2): 168-179,
1989.
Bogarts, B.; Ashtari,
M.; Degreef, G.; Alvir, J.M.J.; Bilder, R.M.; and Lieberman, J.A. limbic structure volumes on magnetic resonance images in first episode
Reduced temporal schizophrenia. Psychiatry
Research:
Neuroimaging,
35: I-1 3, 1990.
Brown, R.; Colter, N.; Corsellis,
J.A.N.; Crow, T.J.; Frith, C.D.; Jagoe, R.; Johnstone, EC.; and Marsh, L. Postmortem evidence of structural brain changes in schizophrenia. Archives of General Psychiatry, 43:36-42, 1986. Buchsbaum, M.S.; Ingvar, D.H.; Kessler, R.; Waters, R.N.; Cappelletti, J.; van Kammen, D.P.; King, A.C.; Johnson, J.L.; Manning, R.G.; Flynn, R.W.; Mann, L.S.; Bunney, W.E.; and Sokoloff, L. Cerebral glucography with positron tomography: Use in normal subjects and in patients with schizophrenia. Archives of General Psychiatry, 39:25 I-259, 1982. Cadet, J.L.; Lohr, J.B.; and Jeste, D.V. Free radicals and tardive dyskinesia. Trends in Neuroscience,
9: 107-108,
1986.
Cutler, N.R.; Kay, A.D.; Marangos, P.J.; and Burg, C. Cerebrospinal fluid neuron-specific enolase is reduced in Alzheimer’s disease. Archives of Neurology, 43:153-154, 1986. DeLisi L.E.; Hoff, A.L.; Schwartz, J.E.; Shields, G.W.; Halthore, S.N.; Gupta, S.M.; Henn, F.A.; and Anand, A.K. Brain morphology in first episode schizophrenia-like psychotic patients: A quantitative magnetic resonance imaging study. Biological Psychiatry, 29: 159-l 75, 1991.
El-Mallakh, R.S.; Egan, M.F.; Intracellular enzymes serving as neuropsychiatric disorders. Psychiatry, Hay, E.; Royds J.A.; Davies-Jones, Cerebrospinal fluid enolase in stroke 471724-729,
and Wyatt, R.J. Creatine phosphate and enolase: markers of central nervous system damage in in press. G.B.; Lewtas, N.A.; Timperley, W.R.; and Taylor, C.B. Journal
of Neurology,
Neurosurgery,
and Psychiatry,
1984.
Jacob, H., and Beckmann, H. Prenatal developmental disturbances in the limbic allocortex in schizophrenics. Journal of Neural Transmission, 65:303-326, 1986. Kaiser, E.; Kuzmits, R.; Pregnant, P.; Burghuber, 0.; and Worofka, W. Clinical biochemistry of neuron specific enolase. Clinica Chimica Acta, 183:13-32, 1989. Marangos, P.J., and Paul, S.M. Brain levels of neuron-specific enolase and nonneuronal enolase in Huntington’s disease. Journal of Neurochemistry, 37(5): 1338-1340, 198 1. O’Callaghan, E.; Sham, P.; Takei, N.; Glover, G.; and Murray, R.M. Schizophrenia after prenatal exposure to 1957 A2 influenza epidemic. Lancet, 337: 1248-1250, 1991. Parma, A.M.; Marangos, P.J.; and Goodwin F.K. A more sensitive radioimmunoassy for neuron-specific enolase suitable for cerebrospinal fluid determination. Journal of Neurochemistry, 36(3): 1093-1096, 198 1.
195 Royds, J.A.; Davies-Jones G.B.; Lewtas, N.A.; Timperlay, W.R.; and Taylor, C. Enolase isoenzymes in the cerebrospinal fluid of patients with diseases of the nervous system. Journal of Neurology, Neurosurgery. and Psychiatry, 46:1031-1036, 1983. Shelton, R.C., and Weinberger, D.R. Computerized tomography in schizophrenia: A review and synthesis. In: Nasarallah, H.A., and Weinberger, D.R., eds. Handbook of Schizophrenia, Vol 1: The Neurology of Schizophrenia. Amsterdam: Elsevier Science Publishers, 1986. pp. 207-250. Suddath, R.L.; Casanova, M.F.; Goldberg, T.E.; Daniel, D.G.; Kelsoe, J.R.; and Weinberger, D.R. Temporal lobe pathology in schizophrenia: A quantitative magnetic resonance imaging study. American Journal of Psychiatry, 146(4):464-472, 1989. Suddath, R.L.; Christison, G.W.; Torrey, E.F.; Casanova, M.F.; and Weinberger, D.R. Anatomical abnormalities in the brains of monozygotic twins discordant for schizophrenia. Nen, England Journal of Medicine, 3221789-794. 1990. Sulkava R.; Viinikka, L.; Erkinjuntti T.; and Roine, R. Cerebrospinal fluid neuron-specific enolase is decreased in multi-infarct dementia, but unchanged in Alzheimer’s disease. Journal of Neurology, Neurosurgery, and Psychiatry, 5 1549-55 1, 1988. Szechtman, H.; Nahmias, C.; Garnett, S.; Firnau, G.; Brown, G.M.; Kaplan, R.D.; and Cleghorn, J.M. Effect of neuroleptics on altered cerebral glucose metabolism in schizophrenia. Archives of General Psychiatry, 45523-532, 1988. Torrey, E.F., and Kaufman, C.A. Schizophrenia and neuroviruses. In: Nasrallah, H.A., and Weinberger, D.R., eds. Handbook of Schizophrenia, Vol I: The Neurology of Schizophrenia. Amsterdam: Elsevier Science Publishers, 1986. pp. 361-376. Van Den Doel, E.M.H.; Rijksen, G.; and Staal, G.E.J. Interet et limites de I’etude de la gamma-enolsae dans le liquide cephalo-rachidien comme temoin de la degradation neuronale. Revue
Neurologique.
144:452-455,
1988.
Vermuyten, K.; Lowenthal, A.; and Kracher, D. Detection of neuron specific enolase concentrations in cerebrospinal fluid from patients with neurological disorders by means of a sensitive enzyme immunoassay. Clinica Chimica Acta, 187:69-78, 1990. Wakayama, Y.; Shibuya, S.; Kawase, J.; Sagawa, F.; and Hashizume, Y. High neuron specific enolase level of cerebrospinal fluid in the early stage of Creutzfeld-Jacob disease. Khnische
Wochenschrtft,
65~798-801,
1987.
Weinberger, D.R. Implications of normal development for the pathogenesis of schizophrenia. Archives of General Psychiatry, 44:660-669, 1987. Weinberger, D.R., and Berman, K.F. Speculations on the meaning of cerebral metabolic hypofrontality in schizophrenia. Schizophrenia Bulletin, 14:157-168, 1988. Weinberger, D.R.; Torrey, E.F.; Neophytides, A.N.; and Wyatt, R.J. Structural abnormalities of the cerebral cortex in chronic schizophrenia. Archives of General Psvchiutry, 361735-739, 1979. Wyatt R.J. Neuroleptics and the natural course of schizophrenia. Schizophrenia Bulletin, 17:325-351,
1991.
Wyatt, R.J.; Fawcett, R.; and Kirch D.G. A possible animal model of defect state schizophrenia. In: Schulz, SC., and Tamminga, C.A., eds. Schizophrenia: Scientific Progress. New York: Oxford University Press, 1989. pp. 184-189.
