BIOL PSYCHIATRY 199o;27:1179-1182
1179
Correlations Between Cistemal CSF and Plasma Concentrations of HVA, MHPG, 5-HIAA, DA, and NA Istvfin Degrell and Erzsdbet Nagy
Introduction The concenWations of homow~fiUic acid (HVA), a metabolite of dopamine (DA), 3-metboxy-4hydroxyphenyl glycole (MHPG), a noradrenaline (NA) metabofite, and 5-hydroxyindole acetic acid (5-HIAA), a metabolite of serotonin (5-HT) in the cerebrospinal fluid (CSF), are commonly used as possible indicators of the neurotransmitter metabolism of the brain. There are marked concentration gradients for both HVA and 5HIAA in CSF. Comparing the cisternal and lumbar HVA and 5-HIAA levels, ne significant correlation was found for HVA, but the 5-HIAA levels correlated well (p < 0.05) (Degreil and Nagy 1990). Consequently, lumbar HVA data can hardly be used as an indicator of the more rostral (in this case cisternal) levels. In the present work we compared the levels of HVA, MHPG, 5-HIAA, DA, and NA in cisternal CSF (cCSF) with their plasma concentrations. Our question was whether or not the levels of these substances are in good accordance with the levels in cisternal CSF. Methods Cistema! punctures for routine neurological examination were performed in 18 neurological and psychiatric patients with the consent of the
From the Department of Nem'ology and Psychiatry, University of Debmcen Medical School, i~:brecen, Hungary. Address reprint requests to Dr. I. Degrell, Department of Neurology and Psychiatry, l.Tniv~,:sity of Debrecen Medical School, Nagyerdei kit. 9~, H-4012 Debrecen, Hungary. © 1990 Society of Biological Psychiatry
patients or of their relatives. The diagnoses, age, and gender of the patients are shown in Table 1. In the 2 weeks prior to the punctures, the patients had taken no drugs that would influence the DA, NA, or 5-HT metabolism. At 9 AM, after bedrest and before bre~ast, 5 ml of CSF was drawn with the patient in a sitting position. After the cisternal puncture, 20 ml of blood was collected from the cubital vein into a heparinized tube. The samples were divided and frozen in liquid nitrogen and stored at - 1 8 0 ° C until analysis. HVA, MHPG, 5-HIAA, DA, NA, and adrenaline (A) were measured by high performance liquid chromatography with electrochemk:al detection in CSF (Krstnlovic et al. 1982) and in plasma (Minegishi and Ishizaki 1984), as modified by us (Nagy and Degrell 1988). Results and Discussion Mean values of HVA, DA, MHPG, NA, and 5-HIAA in cis~mal CSF and in plasma are shown in Table 2. There is a strict correlation between the CSF and plasma levels of HVA (p < O.001) and 5-HIAA (p < 0.01). No significant correlation was found for DA and NA (p < 0. l for both) and for MHPG (r = O. 174). In human CSF the ratio of HVA concentration in cisten~al:lumbar CSF is appro~:i.mately 6:1, whereas for 5-HIAA, it is 3:1 (Deg~ll and Nagy 1990). The HVA and 5-HIAA concentration of the lumbar CSF is greatly influenced by the position, movement of the patients, anatomical variations of the lumbar sac, differences in 0006-3223/90/$03.50
1180
BIOLPSYCHIATRY 1990;27:1179-1182
diffusion, active transport processes, the mixing effect of pneumoencephalography, and the volume of the collected CSF. The nonhomogenous and uncertain distribution of HVA and 5-HIAA in CSF make it difficult to draw any conclusion regarding brain metabolism. Comparing cistemal CSF and plasma substance levels, we found high correlations for HVA and 5-HIAA. Both are of acidic nature, produced in the CNS and the periphery. MI-IPG is pH neutral, and its concentration in plasma and CSF does not correlate. This result is in contrast to the finding reported by Kopin et al. (1984) who suggest that there is a slow but free e~:h~ge of plasma and CSF MHPG based on the fate of intravenous administration of deuterium-labeled MHPG. The blood-brain barrier is not penetrable for either DA or NA, and no significant correlation was found between their cCSF and plasma levels. However, the value ofp < 0. !. may indicate a tentative correlation: they are produced in isolation but probably not totally independently of each other. There is some evidence that the levels of HVA in plasma correlate significantly with the psychopathological changes. Bacopoulos et al. (1979) observed parallel enhancements in rat brain and plasma HVA concentrations after stimulation of the nigrostriatal pathway or after the administration of haloperidol. A decrease of HVA level was found after lesion of the nigrostriatal pathway or after the administration of pargyline, a monoamine oxidase inhibitor. The decrease in plasma HVA level observed after 6OHDA treatment suggests that at least 40% of circulating HVA originates from the brain, as 6-OHDA does not cross the blood-brain barrier. The correlation between the brain and plasraa HVA levels was stronger after haloperidol + debrisoquin than after haloperidol treatment alone (Sternberg et al. 1983). Similar results were observed after apomorphine and bromocriptine treatment, supporting the idea that plasma HVA level indicates the changes in the brain's DA metabolism (Kendler et al. 1982). In schizophrenic patients, Davis et al. (1985) found a high correlation between the plasma
Brief Reports
Table 1. Age, Gender and DJagnosis of the Patients Age Patients (years) Gender 1
30
F
2
85
M
3
40
F
4 5
58 58
F F
6
67
F
7
64
M
8
60
M
9 10 11
66 78 38
F F F
12
45
F
13 14
71 41
M F
15
55
M
16
50
M
17
66
F
18
80
F
Diagnosis Orthostatichypotension. Vestibularlesion? Subac:atesubdural hematoma Chr,,nicalcoholism. Or~hostatic hypotension Multi-infarctdementia Preseniledementiaof Alzheimer'stype Transientischemieattack (right middlecerebral artery) Cerebralinfarct(left middlecerebralartery) Cerebralinfarct(left middlecerebralartery) Multi-infarctdementia Multi-infarctdementia Chronicalcoholism. Deliriumtremens Chronicalcoholism. Narcomania Multi-infarctdementia Transientischemicattack (rightanteriorcerebral artery) Hypertension. Eneephalomalacia (right posterior cerebral artery) Malignantmelanoma Multi-infarctdementia Multi-infarctdementia
HVA concentrations and the severity of the illness both before and after treatment. Pickar et al. (1986) described a high positive correlation between psychosis rating and plasma HVA. Bowers et al. (1984) measured plasma HVA concentration in respect to treatment with halop eridol and found that the good responders showed higher initial HVA concentrations than poor responders. Bondy et al. (1984) observed elevated plasma DA and NA levels in schizophrenic patients. Though no correlation is found between psychiatric symptoms and DA and 5-HT metabolite
Brief Reports
BIOLPSYCH~T~Y 1990;27:1179-1182
1181
Table 2. Mean Values (ng/ml) and Coefficients of Correlation of Neurotransmitters and their Metabolites in Cisternal CSF and in Plasma of the Same Patients N HVA DA MH[N3 HA 5-HIAA
18 15 11 15 18
cCSF 168.7 0.23 14.2 0.127 55.3
-+ -±
43.3 0.04 4.4 0.03 22.9
Plasma 13.5 -0.27 3.66.40.217 7.4 ±
5.3 0.11 1.1 0.05 3.4
Regression equations y = O.091x y = -0.792x -y = 0.647x + y = 0.155x +
2.11 + :/44.9 134.9 0.909
r 0.709 ¢ 0.422 • 0.174 0.429 • 0.627 b
'p < 0.1; bp < 0.01, Cp< 0.001.
levels in CSF, on the other hand, significant correlations exist between plasma catecholamine metabofite levels and the ,-,aeat~ symptoms. Would plasma level of HVA be a better indicator for the psychic disturbances than the HVA lev,~l in lumbar CSF? The HVA level in CSF varies depending on the circumstances (see above) of the lumbar puncture. The constancy of these conditions can be maintained only with great difficulty. Plasma HVA level might be a better indicator for the dopamine metabolism of the brain because its concentration is more constant in the blood than in lumbar CSF. The dopamine metabolism in the periphery is probably not so high as to mask the variability of HVA level from central origin. We also found a strict correlation (p < 0.01) between the cistemal CSF ::nd plasma levels of 5-HIAA. This relationship is closer than that found between cistemal and lumbar CSF 5-HIAA levels.
N (1984): Plasma catecholamine metabolites and
early response to haloperidol. J Clin Psychiatry 45:248-251. Davis Kid, Davidson M, Mohs RC, et al. (1985): Plasma homovaniUic scid concentration and the severityof schizophrenicillness. Science 227:16011602. Degrell I, Nagy E (1990): Concentration gradientq for HVA, 5-HIAA, ascorbic acid and uric acid in cerebrospinal fluid. Biol Psychiatry (in press). Kendler KS, Heninger GR, Roth RH (1982): Influence of dopamine agonists on plasma and brain levels of homovan;llic acid. Life Sci 30:20632069. Kopin IJ, Blombery P, Ebert MH, et al. (1984): Disposition and metabolism of MHPG-CD3 in humans: Plasma MHI~ as the principal pathway of norepinephrine metabolism and as an important determinant of CSF levels of MHPG. In Usdin E, Asberg M, Bertilsson L, SjOqvist F (eds), Frontiers in Biochemical and Pharmacological Research in Depression. Advances in Biochemical Psychopharmacology, vol. 39. New York: Raven Press, pp 57-68. The authors thank Drs. IAszl6Csiba and Daniel Bereczki Krstulovic AM, Bertani-Dziedzic L, Bautista-Cerqueira S, Gitlow SE (1982): Simultaneous deter(NeurologicalDepartment)forthe CSFandplasmasamples. mination of 4-hydmxy-3-methoxyphenylacetic (homovanillic) acid and other monoamine metabolites in human lumbar cerebrospinal fluid. An imp~'oved high-performance liquid chromatoReferences graphic study with electro-chemical detection. 3 Bacopoulos NG, Hattox SE, Roth RH (1979): 3,4Chromatography 227:379-389. dihydroxy-phenyl-aceticacid and homovanillicacid in rat plasma: Possible indicators of central do- Minegishi A, Ishizaki T (1984): Determination of free 3-methoxy-4-hydroxyphenylglycol with several paminergic activity. Eur Y Pharmacol 56:225other monoamine metabolites in plasma by high236. performance liquid chromatography with amperBondy B, Ackenheil M, Birzle W, Elbers R, Fr6hier ometric detection. J Chromatography 311:51-57. M (1984): Catecholamines and their receptors in blood: Evidence for alterations in schizophrenia. Nagy E, Degrell I (1988): Determinations of catecholamines in human lumbar cerebrospinal fluid Biol Psychiatry 19:1377-1393. by high-performance liquid chromatography with Bowers MB Jr, Swigar ME, Jatlow PI, Goicoechea
1182
BIOLPSYCHIATRY !990;27:1179--1182
elecu'ochemical detection, Acta Pharm Hung 58:159-165. Pickar D, Labarca R, Doran AR, et al. (1986): Longitudinal measurement of plasma homovanillic acid levels in schizoplh~enicpatients. Correlation with psychosis and response to r~euro-
Brief Reports
leptic treatment. Arch Gen Psychiatry 43:669676. Stemberg DE, Heninger GR, Roth RH (1983): Plasma homovanillic acid as an index of brain dopamine ~¢t~,bolism: Enhancement with d~brisoquin. Life Sci 32:2447-2452.
1179
Correlations Between Cistemal CSF and Plasma Concentrations of HVA, MHPG, 5-HIAA, DA, and NA Istvfin Degrell and Erzsdbet Nagy
Introduction The concenWations of homow~fiUic acid (HVA), a metabolite of dopamine (DA), 3-metboxy-4hydroxyphenyl glycole (MHPG), a noradrenaline (NA) metabofite, and 5-hydroxyindole acetic acid (5-HIAA), a metabolite of serotonin (5-HT) in the cerebrospinal fluid (CSF), are commonly used as possible indicators of the neurotransmitter metabolism of the brain. There are marked concentration gradients for both HVA and 5HIAA in CSF. Comparing the cisternal and lumbar HVA and 5-HIAA levels, ne significant correlation was found for HVA, but the 5-HIAA levels correlated well (p < 0.05) (Degreil and Nagy 1990). Consequently, lumbar HVA data can hardly be used as an indicator of the more rostral (in this case cisternal) levels. In the present work we compared the levels of HVA, MHPG, 5-HIAA, DA, and NA in cisternal CSF (cCSF) with their plasma concentrations. Our question was whether or not the levels of these substances are in good accordance with the levels in cisternal CSF. Methods Cistema! punctures for routine neurological examination were performed in 18 neurological and psychiatric patients with the consent of the
From the Department of Nem'ology and Psychiatry, University of Debmcen Medical School, i~:brecen, Hungary. Address reprint requests to Dr. I. Degrell, Department of Neurology and Psychiatry, l.Tniv~,:sity of Debrecen Medical School, Nagyerdei kit. 9~, H-4012 Debrecen, Hungary. © 1990 Society of Biological Psychiatry
patients or of their relatives. The diagnoses, age, and gender of the patients are shown in Table 1. In the 2 weeks prior to the punctures, the patients had taken no drugs that would influence the DA, NA, or 5-HT metabolism. At 9 AM, after bedrest and before bre~ast, 5 ml of CSF was drawn with the patient in a sitting position. After the cisternal puncture, 20 ml of blood was collected from the cubital vein into a heparinized tube. The samples were divided and frozen in liquid nitrogen and stored at - 1 8 0 ° C until analysis. HVA, MHPG, 5-HIAA, DA, NA, and adrenaline (A) were measured by high performance liquid chromatography with electrochemk:al detection in CSF (Krstnlovic et al. 1982) and in plasma (Minegishi and Ishizaki 1984), as modified by us (Nagy and Degrell 1988). Results and Discussion Mean values of HVA, DA, MHPG, NA, and 5-HIAA in cis~mal CSF and in plasma are shown in Table 2. There is a strict correlation between the CSF and plasma levels of HVA (p < O.001) and 5-HIAA (p < 0.01). No significant correlation was found for DA and NA (p < 0. l for both) and for MHPG (r = O. 174). In human CSF the ratio of HVA concentration in cisten~al:lumbar CSF is appro~:i.mately 6:1, whereas for 5-HIAA, it is 3:1 (Deg~ll and Nagy 1990). The HVA and 5-HIAA concentration of the lumbar CSF is greatly influenced by the position, movement of the patients, anatomical variations of the lumbar sac, differences in 0006-3223/90/$03.50
1180
BIOLPSYCHIATRY 1990;27:1179-1182
diffusion, active transport processes, the mixing effect of pneumoencephalography, and the volume of the collected CSF. The nonhomogenous and uncertain distribution of HVA and 5-HIAA in CSF make it difficult to draw any conclusion regarding brain metabolism. Comparing cistemal CSF and plasma substance levels, we found high correlations for HVA and 5-HIAA. Both are of acidic nature, produced in the CNS and the periphery. MI-IPG is pH neutral, and its concentration in plasma and CSF does not correlate. This result is in contrast to the finding reported by Kopin et al. (1984) who suggest that there is a slow but free e~:h~ge of plasma and CSF MHPG based on the fate of intravenous administration of deuterium-labeled MHPG. The blood-brain barrier is not penetrable for either DA or NA, and no significant correlation was found between their cCSF and plasma levels. However, the value ofp < 0. !. may indicate a tentative correlation: they are produced in isolation but probably not totally independently of each other. There is some evidence that the levels of HVA in plasma correlate significantly with the psychopathological changes. Bacopoulos et al. (1979) observed parallel enhancements in rat brain and plasma HVA concentrations after stimulation of the nigrostriatal pathway or after the administration of haloperidol. A decrease of HVA level was found after lesion of the nigrostriatal pathway or after the administration of pargyline, a monoamine oxidase inhibitor. The decrease in plasma HVA level observed after 6OHDA treatment suggests that at least 40% of circulating HVA originates from the brain, as 6-OHDA does not cross the blood-brain barrier. The correlation between the brain and plasraa HVA levels was stronger after haloperidol + debrisoquin than after haloperidol treatment alone (Sternberg et al. 1983). Similar results were observed after apomorphine and bromocriptine treatment, supporting the idea that plasma HVA level indicates the changes in the brain's DA metabolism (Kendler et al. 1982). In schizophrenic patients, Davis et al. (1985) found a high correlation between the plasma
Brief Reports
Table 1. Age, Gender and DJagnosis of the Patients Age Patients (years) Gender 1
30
F
2
85
M
3
40
F
4 5
58 58
F F
6
67
F
7
64
M
8
60
M
9 10 11
66 78 38
F F F
12
45
F
13 14
71 41
M F
15
55
M
16
50
M
17
66
F
18
80
F
Diagnosis Orthostatichypotension. Vestibularlesion? Subac:atesubdural hematoma Chr,,nicalcoholism. Or~hostatic hypotension Multi-infarctdementia Preseniledementiaof Alzheimer'stype Transientischemieattack (right middlecerebral artery) Cerebralinfarct(left middlecerebralartery) Cerebralinfarct(left middlecerebralartery) Multi-infarctdementia Multi-infarctdementia Chronicalcoholism. Deliriumtremens Chronicalcoholism. Narcomania Multi-infarctdementia Transientischemicattack (rightanteriorcerebral artery) Hypertension. Eneephalomalacia (right posterior cerebral artery) Malignantmelanoma Multi-infarctdementia Multi-infarctdementia
HVA concentrations and the severity of the illness both before and after treatment. Pickar et al. (1986) described a high positive correlation between psychosis rating and plasma HVA. Bowers et al. (1984) measured plasma HVA concentration in respect to treatment with halop eridol and found that the good responders showed higher initial HVA concentrations than poor responders. Bondy et al. (1984) observed elevated plasma DA and NA levels in schizophrenic patients. Though no correlation is found between psychiatric symptoms and DA and 5-HT metabolite
Brief Reports
BIOLPSYCH~T~Y 1990;27:1179-1182
1181
Table 2. Mean Values (ng/ml) and Coefficients of Correlation of Neurotransmitters and their Metabolites in Cisternal CSF and in Plasma of the Same Patients N HVA DA MH[N3 HA 5-HIAA
18 15 11 15 18
cCSF 168.7 0.23 14.2 0.127 55.3
-+ -±
43.3 0.04 4.4 0.03 22.9
Plasma 13.5 -0.27 3.66.40.217 7.4 ±
5.3 0.11 1.1 0.05 3.4
Regression equations y = O.091x y = -0.792x -y = 0.647x + y = 0.155x +
2.11 + :/44.9 134.9 0.909
r 0.709 ¢ 0.422 • 0.174 0.429 • 0.627 b
'p < 0.1; bp < 0.01, Cp< 0.001.
levels in CSF, on the other hand, significant correlations exist between plasma catecholamine metabofite levels and the ,-,aeat~ symptoms. Would plasma level of HVA be a better indicator for the psychic disturbances than the HVA lev,~l in lumbar CSF? The HVA level in CSF varies depending on the circumstances (see above) of the lumbar puncture. The constancy of these conditions can be maintained only with great difficulty. Plasma HVA level might be a better indicator for the dopamine metabolism of the brain because its concentration is more constant in the blood than in lumbar CSF. The dopamine metabolism in the periphery is probably not so high as to mask the variability of HVA level from central origin. We also found a strict correlation (p < 0.01) between the cistemal CSF ::nd plasma levels of 5-HIAA. This relationship is closer than that found between cistemal and lumbar CSF 5-HIAA levels.
N (1984): Plasma catecholamine metabolites and
early response to haloperidol. J Clin Psychiatry 45:248-251. Davis Kid, Davidson M, Mohs RC, et al. (1985): Plasma homovaniUic scid concentration and the severityof schizophrenicillness. Science 227:16011602. Degrell I, Nagy E (1990): Concentration gradientq for HVA, 5-HIAA, ascorbic acid and uric acid in cerebrospinal fluid. Biol Psychiatry (in press). Kendler KS, Heninger GR, Roth RH (1982): Influence of dopamine agonists on plasma and brain levels of homovan;llic acid. Life Sci 30:20632069. Kopin IJ, Blombery P, Ebert MH, et al. (1984): Disposition and metabolism of MHPG-CD3 in humans: Plasma MHI~ as the principal pathway of norepinephrine metabolism and as an important determinant of CSF levels of MHPG. In Usdin E, Asberg M, Bertilsson L, SjOqvist F (eds), Frontiers in Biochemical and Pharmacological Research in Depression. Advances in Biochemical Psychopharmacology, vol. 39. New York: Raven Press, pp 57-68. The authors thank Drs. IAszl6Csiba and Daniel Bereczki Krstulovic AM, Bertani-Dziedzic L, Bautista-Cerqueira S, Gitlow SE (1982): Simultaneous deter(NeurologicalDepartment)forthe CSFandplasmasamples. mination of 4-hydmxy-3-methoxyphenylacetic (homovanillic) acid and other monoamine metabolites in human lumbar cerebrospinal fluid. An imp~'oved high-performance liquid chromatoReferences graphic study with electro-chemical detection. 3 Bacopoulos NG, Hattox SE, Roth RH (1979): 3,4Chromatography 227:379-389. dihydroxy-phenyl-aceticacid and homovanillicacid in rat plasma: Possible indicators of central do- Minegishi A, Ishizaki T (1984): Determination of free 3-methoxy-4-hydroxyphenylglycol with several paminergic activity. Eur Y Pharmacol 56:225other monoamine metabolites in plasma by high236. performance liquid chromatography with amperBondy B, Ackenheil M, Birzle W, Elbers R, Fr6hier ometric detection. J Chromatography 311:51-57. M (1984): Catecholamines and their receptors in blood: Evidence for alterations in schizophrenia. Nagy E, Degrell I (1988): Determinations of catecholamines in human lumbar cerebrospinal fluid Biol Psychiatry 19:1377-1393. by high-performance liquid chromatography with Bowers MB Jr, Swigar ME, Jatlow PI, Goicoechea
1182
BIOLPSYCHIATRY !990;27:1179--1182
elecu'ochemical detection, Acta Pharm Hung 58:159-165. Pickar D, Labarca R, Doran AR, et al. (1986): Longitudinal measurement of plasma homovanillic acid levels in schizoplh~enicpatients. Correlation with psychosis and response to r~euro-
Brief Reports
leptic treatment. Arch Gen Psychiatry 43:669676. Stemberg DE, Heninger GR, Roth RH (1983): Plasma homovanillic acid as an index of brain dopamine ~¢t~,bolism: Enhancement with d~brisoquin. Life Sci 32:2447-2452.
