Cerebrospinal Fluid Amine in Acute Schizophrenia Robert M. Post, MD; Ed
Metabolites
Fink, MD; William T. Carpenter, Jr, MD; Frederick K. Goodwin, MD
The metabolites of serotonin, dopamine, and norepinephrine, 5-hydroxyindoleacetic acid (5HIAA), homovanillic acid (HVA), and 3-methoxy-4-hydroxy-phenylethylene glycol (MHPG), respectively, were studied in cerebrospinal fluid of patients with
acute schizophrenia. Base line levels of these metabolites were not significantly different from those in normal, neurological, and affectively ill controls. Accumulations of 5HIAA and HVA following probenecid administration, which provide a measure of serotonin and dopamine turnover, were also not significantly different in patients with acute schizophrenia and affective illness. After patients had recovered from their acute schizophrenic illness, HVA accumulations were significantly reduced. We discuss results in relation to amine hypotheses of schizophrenia and the suggestion that altered dopamine metabolism may reflect a biological change predisposing to acute schizophrenia.
in central catecholamine and indoleamine metabolism have been hypothesized in the schizo¬ but there have been relatively few at¬ tempts to verify these hypothesized alterations in pa¬ tients. In an attempt to directly assess central amine metabolism in patients with acute schizophrenia, 5-hydroxyindoleacetic acid (5HIAA), homovanillic acid (HVA), and 3-methoxy-4-hydroxy-phenylethylene glycol (MHPG), major metabolites of serotonin, dopamine, and norepinephrine, respectively, were measured in cerebrospinal fluid (CSF). Early work with LSD and the more recent study of methylated serotonin metabolites have indirectly impli¬ cated abnormal serotonin function in acute psychosis.'"6 Deficits in norepinephrine and in dopamine ß-hydroxylase, an enzyme in its biosynthetic pathway, have like-
Alphrenic teratisyndromes, ons -
Accepted for publication March 25, 1975. From the Adult Psychiatry Branch (Drs. Post, Fink, and Carpenter) and the Laboratory of Clinical Science (Dr. Goodwin), National Institute of Mental Health, Bethesda, Md. Dr. Fink is now with Butler Hospital, Providence, RI. Reprint requests to the Section on Psychobiology, Adult Psychiatry Branch, National Institute of Mental Health, Bldg 10, Rm 3S 239, 9000 Rockville Pike, Bethesda, MD 20014 (Dr. Post).
wise been postulated in schizophrenia.7·8 Finally, on the basis of a wide variety of indirect data, an important role for dopamine in schizophrenia has been postulated.912 This evidence implicating dopamine includes the observa¬ tions that the neuroleptics (in apparent relationship to their antipsychotic acitivity) overlap with the preferred conformation of dopamine,13 block dopamine receptors,1416 increase dopamine turnover,1420 alter firing rates in single dopamine neurons,21·22 and block dopamine mediated in¬ creases in cyclic adenosine monophosphate.23·24 The precip¬ itation of schizophreniform psychoses by amphetamine or cocaine likewise gives credence to the association of dopa¬ mine (and norepinephrine) with schizophrenia.2529 Although two early studies reported low CSF 5HIAA in acute schizophrenic subjects,30·31 other base line CSF stud¬ ies have failed to demonstrate statistically significant dif¬ ferences in amine metabolites in schizophrenic patients compared to controls3136 (Table 1). Probenecid can inhibit the transport system responsible for the removal of the acid metabolites HVA and 5HIAA from brain CSF to blood3740; when administered in sufficiently high doses to patients, probenecid can uncover differences in amine me¬ tabolite accumulations that were not apparent from base line data.41 Recently, Bowers, administering probenecid in this way,42 reported that accumulations of HVA are sub¬ stantially reduced in schizophrenic patients compared with patients with affective illness, poor prognosis schizo¬ phrenics having lower HVA accumulations than good
prognosis schizophrenics.
METHODS The CSF samples were obtained from 20 acute schizophrenic pa¬ tients hospitalized on two clinical research units at the National Institute of Mental Health. Patients and their families gave in¬ formed consent for the research procedures. Patients (age range, 16 to 54 years; mean, 25) were selected for admission and study on the basis of acute onset of psychotic symptomatology without evi¬ dence of organic illness or brain syndrome, and adequate pre-
Downloaded From: http://archpsyc.jamanetwork.com/ by a University of California - San Diego User on 12/13/2015
Table 1.—CSF Amine Metabolites in
Fotherby et al32
5HIAA NS (11)
HVA
MHPG
phrenic^_
Acute & chronic
Low (7)
schizophrenic
1966
Persson & Roos33
NS (40)
NS (4)
Chronic schizo¬
Low (7) NS (22)
NS (6) NS (30)
Unavailable Acute schizo¬
phrenic_
1969
Bowers3' 1969 Rimon et al34
Diagnosis Acute schizo¬
1963
Ashcroft et al30
Schizophrenic Patients Compared to Controls*
phrenic^^
1971
Shopsin et al36
NS (26)
Acute schizo¬
phrenic
Controls
Comment
Psychiatric
Psychiatric
&
neurological
Psychiatric
& normal
schizophrenic low
Patients taking
neuroleptics
Psychiatric & normal Psychiatric Psychiatric
Only acute
Higher HVA in paranoid
& normal
1973 Post et al 1975
NS (18)
NS (17)
Bowers 197335
NS (18)
NS
Bowers 1974«
NS (17)
Low (17)
Acute & chronic
Psychiatric
Lowest HVA in poor
Cohen et al83
NS (9)
NS (9)
Autism
Psychiatric & neurological
Autistic children
Psychiatric, normal, & neurological phrenic CSF Studies Employing Probenecid Accumulations Acute & chronic (18) Psychiatric, normal, & Inmates schizophrenic NS (17)
Acute schizo¬
schizophrenic
1974
Post et al 1975
NS (18)
NS (20)
NS (17)
Acute schizo¬
phrenic
Low HVA in
Schneiderian
prognosis
Psychiatric
intermediate between
atypical
lower on recovery
Numbers in parentheses indicate number of subjects. 5-Hydroxyindoleacetic acid indicated by 5HIAA; homovanillic acid, oxy-4-hydroxy-phenylethylene glycol, MHPG. *
morbid social history. They had hallucinations, delusions, thought disorder, and bizarre behavior and were diagnosed independently by two psychiatrists as schizophrenic, employing DSM II cate¬ gories and criteria.43 These diagnostic judgments were buttressed by criteria defined by Carpenter et al," and (except for acuteness) by Feighner et al.45 Four patients with schizoaffective illness were included in the sample; their amine metabolites did not differ sig¬ nificantly from those of the schizophrenic patients. Patients were interviewed and rated by physicians on a shorts ened version of the Present State Examination46·'" or Brief Psy¬ chiatric Rating Scale.48 Prognostic scales of Strauss and Car¬ penter,49 Stephens and Astrup,50 and assessment of Schneiderian First Rank symptoms51 were completed. Daily ratings of a variety of cognitive, affective, and behavioral measurements52 were per¬ formed by trained research nurses who had no knowledge of the biological data. All patients were treated with intensive individual and group psychotherapy and family therapy where indicated. On one of the clinical research units, patients were systematically treated without psychopharmacological intervention.53 Five of the 12 patients studied in a recovered or compensated phase of their illness had not been treated with neuroleptics during the interval between the first and second set of lumbar punctures. After patients had been off all medication regimens a minimum of two weeks, a base line lumbar puncture was performed in the lateral decubitus position at 9 am and after nine hours of bed rest. A second lumbar puncture was performed at 3 pm the following day, after administration of probenecid (100 mg/kg) during an 18hour period. Probenecid was administered in four divided doses with 30% of the total dose being administered at 9 pm and the re¬ mainder at 2 am, 7 am, and 12 noon.41 Again, bed rest was main¬ tained as closely as clinically possible from midnight until the lumbar puncture. The nature of the research procedures was dis¬ cussed on an ongoing basis on the wards and, with few exceptions, even the most psychotic patients were able to cooperate for the lumbar punctures. The lumbar puncture was carried out in the pa¬ tient's own bed and physical restraint was not employed other than a familiar nurse or aide helping the patient maintain the
& seizure
patients Good prognosis patients only; HVA HVA: and 3-meth-
position. Postlumbar puncture headaches were in¬ frequent, although nausea and, rarely, vomiting were associated with probenecid administration in some patients. After 12 pa¬ tients had recovered from their psychotic episodes, base line and probenecid lumbar punctures were repeated in a similar fashion, again after at least two weeks without all psychotropic medica¬ decubitus
tion. Similar
procedures were followed for ten normal and 19 neuro¬ logical patients (studied in collaboration with T.N. Chase, MD, and E. Gordon) and for patients with depressive and manic illness.54
The first 8 ml of CSF were collected in 20 mg of ascorbic acid and immediately frozen at -60 C. The 5HIAA and HVA were assayed fluorometrically, as previously described,*1·54 and MHPG was as¬ sayed by the gas Chromatographie technique of Gordon and Oli¬ ver.55
Since probenecid inhibits the process by which 5HIAA and HVA are removed from the central nervous system, their accumu¬ lation in CSF after probenecid administration can be used as a re¬ flection of central serotonin and dopamine turnover.3741 Although CSF probenecid levels are not available for these patients, it is un¬ likely that they would account for the differences in HVA accumu¬ lations observed. As discussed by Bowers,56 a decreased probene¬ cid effect would be reflected in decreased accumulations of both 5HIAA and HVA, while our data on recovered schizophrenic pa¬ tients demonstrates a decrease in HVA coincident with no change or an increase in 5HIAA. Moreover, Perei et al57 have demon¬ strated, utilizing the high-dose probenecid technique employed here, that probenecid levels in CSF are not significantly correlated with accumulations of 5HIAA and HVA. The MHPG increases af¬ ter probenecid are small and do not appear to adequately reflect norepinephrine turnover in man, but have been shown to provide a reliable independent replication of base line differences.58 In our patients, the increase in MHPG, following probenecid adminis¬ tration is largely accounted for by free MHPG59 and not the sul¬ fate congregate as reported in studies on rats60 and rabbits.61 A substantial body of evidence suggests HVA in lumbar CSF is de¬ rived largely from rostal sources of dopamine in the striatum,6266
Downloaded From: http://archpsyc.jamanetwork.com/ by a University of California - San Diego User on 12/13/2015
300
Probenecid 160
250
140
f
200
Probenecid
120 100
80 150
60 40
100
20 E c
50
N=16 0
i I
N=15
< >
N=17:
N=48
40 r Base Line
I
30
Base Line 25
Í
20
i-
Í
N=10
N=21
Normal
Manic Controls
20
15
10
10
5 0
N=10
N=22
Normal
Manic Controls
N=69
Depressed Schizophrenics Fig 1 —Cerebrospinal fluid 5-hydroxyindoleacetic acid (5HIAA) in acute schizophrenia. Table 2.—Correlations of 5HIAA Accumulation After Probenecid and Initial Psychopathology*
Acutely III (n=18) Delusions Visual hallucinations Auditory hallucinations
Psychosis Schneiderlan First Rank Symptoms
Recovered (n=11)
.14 -.46 —.13 -.28
NS
Metabolites
Fink, MD; William T. Carpenter, Jr, MD; Frederick K. Goodwin, MD
The metabolites of serotonin, dopamine, and norepinephrine, 5-hydroxyindoleacetic acid (5HIAA), homovanillic acid (HVA), and 3-methoxy-4-hydroxy-phenylethylene glycol (MHPG), respectively, were studied in cerebrospinal fluid of patients with
acute schizophrenia. Base line levels of these metabolites were not significantly different from those in normal, neurological, and affectively ill controls. Accumulations of 5HIAA and HVA following probenecid administration, which provide a measure of serotonin and dopamine turnover, were also not significantly different in patients with acute schizophrenia and affective illness. After patients had recovered from their acute schizophrenic illness, HVA accumulations were significantly reduced. We discuss results in relation to amine hypotheses of schizophrenia and the suggestion that altered dopamine metabolism may reflect a biological change predisposing to acute schizophrenia.
in central catecholamine and indoleamine metabolism have been hypothesized in the schizo¬ but there have been relatively few at¬ tempts to verify these hypothesized alterations in pa¬ tients. In an attempt to directly assess central amine metabolism in patients with acute schizophrenia, 5-hydroxyindoleacetic acid (5HIAA), homovanillic acid (HVA), and 3-methoxy-4-hydroxy-phenylethylene glycol (MHPG), major metabolites of serotonin, dopamine, and norepinephrine, respectively, were measured in cerebrospinal fluid (CSF). Early work with LSD and the more recent study of methylated serotonin metabolites have indirectly impli¬ cated abnormal serotonin function in acute psychosis.'"6 Deficits in norepinephrine and in dopamine ß-hydroxylase, an enzyme in its biosynthetic pathway, have like-
Alphrenic teratisyndromes, ons -
Accepted for publication March 25, 1975. From the Adult Psychiatry Branch (Drs. Post, Fink, and Carpenter) and the Laboratory of Clinical Science (Dr. Goodwin), National Institute of Mental Health, Bethesda, Md. Dr. Fink is now with Butler Hospital, Providence, RI. Reprint requests to the Section on Psychobiology, Adult Psychiatry Branch, National Institute of Mental Health, Bldg 10, Rm 3S 239, 9000 Rockville Pike, Bethesda, MD 20014 (Dr. Post).
wise been postulated in schizophrenia.7·8 Finally, on the basis of a wide variety of indirect data, an important role for dopamine in schizophrenia has been postulated.912 This evidence implicating dopamine includes the observa¬ tions that the neuroleptics (in apparent relationship to their antipsychotic acitivity) overlap with the preferred conformation of dopamine,13 block dopamine receptors,1416 increase dopamine turnover,1420 alter firing rates in single dopamine neurons,21·22 and block dopamine mediated in¬ creases in cyclic adenosine monophosphate.23·24 The precip¬ itation of schizophreniform psychoses by amphetamine or cocaine likewise gives credence to the association of dopa¬ mine (and norepinephrine) with schizophrenia.2529 Although two early studies reported low CSF 5HIAA in acute schizophrenic subjects,30·31 other base line CSF stud¬ ies have failed to demonstrate statistically significant dif¬ ferences in amine metabolites in schizophrenic patients compared to controls3136 (Table 1). Probenecid can inhibit the transport system responsible for the removal of the acid metabolites HVA and 5HIAA from brain CSF to blood3740; when administered in sufficiently high doses to patients, probenecid can uncover differences in amine me¬ tabolite accumulations that were not apparent from base line data.41 Recently, Bowers, administering probenecid in this way,42 reported that accumulations of HVA are sub¬ stantially reduced in schizophrenic patients compared with patients with affective illness, poor prognosis schizo¬ phrenics having lower HVA accumulations than good
prognosis schizophrenics.
METHODS The CSF samples were obtained from 20 acute schizophrenic pa¬ tients hospitalized on two clinical research units at the National Institute of Mental Health. Patients and their families gave in¬ formed consent for the research procedures. Patients (age range, 16 to 54 years; mean, 25) were selected for admission and study on the basis of acute onset of psychotic symptomatology without evi¬ dence of organic illness or brain syndrome, and adequate pre-
Downloaded From: http://archpsyc.jamanetwork.com/ by a University of California - San Diego User on 12/13/2015
Table 1.—CSF Amine Metabolites in
Fotherby et al32
5HIAA NS (11)
HVA
MHPG
phrenic^_
Acute & chronic
Low (7)
schizophrenic
1966
Persson & Roos33
NS (40)
NS (4)
Chronic schizo¬
Low (7) NS (22)
NS (6) NS (30)
Unavailable Acute schizo¬
phrenic_
1969
Bowers3' 1969 Rimon et al34
Diagnosis Acute schizo¬
1963
Ashcroft et al30
Schizophrenic Patients Compared to Controls*
phrenic^^
1971
Shopsin et al36
NS (26)
Acute schizo¬
phrenic
Controls
Comment
Psychiatric
Psychiatric
&
neurological
Psychiatric
& normal
schizophrenic low
Patients taking
neuroleptics
Psychiatric & normal Psychiatric Psychiatric
Only acute
Higher HVA in paranoid
& normal
1973 Post et al 1975
NS (18)
NS (17)
Bowers 197335
NS (18)
NS
Bowers 1974«
NS (17)
Low (17)
Acute & chronic
Psychiatric
Lowest HVA in poor
Cohen et al83
NS (9)
NS (9)
Autism
Psychiatric & neurological
Autistic children
Psychiatric, normal, & neurological phrenic CSF Studies Employing Probenecid Accumulations Acute & chronic (18) Psychiatric, normal, & Inmates schizophrenic NS (17)
Acute schizo¬
schizophrenic
1974
Post et al 1975
NS (18)
NS (20)
NS (17)
Acute schizo¬
phrenic
Low HVA in
Schneiderian
prognosis
Psychiatric
intermediate between
atypical
lower on recovery
Numbers in parentheses indicate number of subjects. 5-Hydroxyindoleacetic acid indicated by 5HIAA; homovanillic acid, oxy-4-hydroxy-phenylethylene glycol, MHPG. *
morbid social history. They had hallucinations, delusions, thought disorder, and bizarre behavior and were diagnosed independently by two psychiatrists as schizophrenic, employing DSM II cate¬ gories and criteria.43 These diagnostic judgments were buttressed by criteria defined by Carpenter et al," and (except for acuteness) by Feighner et al.45 Four patients with schizoaffective illness were included in the sample; their amine metabolites did not differ sig¬ nificantly from those of the schizophrenic patients. Patients were interviewed and rated by physicians on a shorts ened version of the Present State Examination46·'" or Brief Psy¬ chiatric Rating Scale.48 Prognostic scales of Strauss and Car¬ penter,49 Stephens and Astrup,50 and assessment of Schneiderian First Rank symptoms51 were completed. Daily ratings of a variety of cognitive, affective, and behavioral measurements52 were per¬ formed by trained research nurses who had no knowledge of the biological data. All patients were treated with intensive individual and group psychotherapy and family therapy where indicated. On one of the clinical research units, patients were systematically treated without psychopharmacological intervention.53 Five of the 12 patients studied in a recovered or compensated phase of their illness had not been treated with neuroleptics during the interval between the first and second set of lumbar punctures. After patients had been off all medication regimens a minimum of two weeks, a base line lumbar puncture was performed in the lateral decubitus position at 9 am and after nine hours of bed rest. A second lumbar puncture was performed at 3 pm the following day, after administration of probenecid (100 mg/kg) during an 18hour period. Probenecid was administered in four divided doses with 30% of the total dose being administered at 9 pm and the re¬ mainder at 2 am, 7 am, and 12 noon.41 Again, bed rest was main¬ tained as closely as clinically possible from midnight until the lumbar puncture. The nature of the research procedures was dis¬ cussed on an ongoing basis on the wards and, with few exceptions, even the most psychotic patients were able to cooperate for the lumbar punctures. The lumbar puncture was carried out in the pa¬ tient's own bed and physical restraint was not employed other than a familiar nurse or aide helping the patient maintain the
& seizure
patients Good prognosis patients only; HVA HVA: and 3-meth-
position. Postlumbar puncture headaches were in¬ frequent, although nausea and, rarely, vomiting were associated with probenecid administration in some patients. After 12 pa¬ tients had recovered from their psychotic episodes, base line and probenecid lumbar punctures were repeated in a similar fashion, again after at least two weeks without all psychotropic medica¬ decubitus
tion. Similar
procedures were followed for ten normal and 19 neuro¬ logical patients (studied in collaboration with T.N. Chase, MD, and E. Gordon) and for patients with depressive and manic illness.54
The first 8 ml of CSF were collected in 20 mg of ascorbic acid and immediately frozen at -60 C. The 5HIAA and HVA were assayed fluorometrically, as previously described,*1·54 and MHPG was as¬ sayed by the gas Chromatographie technique of Gordon and Oli¬ ver.55
Since probenecid inhibits the process by which 5HIAA and HVA are removed from the central nervous system, their accumu¬ lation in CSF after probenecid administration can be used as a re¬ flection of central serotonin and dopamine turnover.3741 Although CSF probenecid levels are not available for these patients, it is un¬ likely that they would account for the differences in HVA accumu¬ lations observed. As discussed by Bowers,56 a decreased probene¬ cid effect would be reflected in decreased accumulations of both 5HIAA and HVA, while our data on recovered schizophrenic pa¬ tients demonstrates a decrease in HVA coincident with no change or an increase in 5HIAA. Moreover, Perei et al57 have demon¬ strated, utilizing the high-dose probenecid technique employed here, that probenecid levels in CSF are not significantly correlated with accumulations of 5HIAA and HVA. The MHPG increases af¬ ter probenecid are small and do not appear to adequately reflect norepinephrine turnover in man, but have been shown to provide a reliable independent replication of base line differences.58 In our patients, the increase in MHPG, following probenecid adminis¬ tration is largely accounted for by free MHPG59 and not the sul¬ fate congregate as reported in studies on rats60 and rabbits.61 A substantial body of evidence suggests HVA in lumbar CSF is de¬ rived largely from rostal sources of dopamine in the striatum,6266
Downloaded From: http://archpsyc.jamanetwork.com/ by a University of California - San Diego User on 12/13/2015
300
Probenecid 160
250
140
f
200
Probenecid
120 100
80 150
60 40
100
20 E c
50
N=16 0
i I
N=15
< >
N=17:
N=48
40 r Base Line
I
30
Base Line 25
Í
20
i-
Í
N=10
N=21
Normal
Manic Controls
20
15
10
10
5 0
N=10
N=22
Normal
Manic Controls
N=69
Depressed Schizophrenics Fig 1 —Cerebrospinal fluid 5-hydroxyindoleacetic acid (5HIAA) in acute schizophrenia. Table 2.—Correlations of 5HIAA Accumulation After Probenecid and Initial Psychopathology*
Acutely III (n=18) Delusions Visual hallucinations Auditory hallucinations
Psychosis Schneiderlan First Rank Symptoms
Recovered (n=11)
.14 -.46 —.13 -.28
NS
