Psychopharmacology (1992) 106: 268-274
Psychopharmacology 9 Springer-Verlag 1992
Determination of haloperidol and reduced haloperidol in the plasma and blood of patients on depot haloperidol Darryl W. Eyles 1'2, Harvey A. Whiteford 1, Terry J. Stedman I, and Susan M. Pond 2 1 Clinical Studies Unit, Wolston Park Hospital and 2 University of Queensland, Department of Medicine, Princess Alexandra Hospital, Brisbane 4102, Australia Received March 22, 1991 / Final version
Abstract. We developed a sensitive HPLC assay to measure haloperidol (HA) and its metabolite, reduced haloperidol (RH), in plasma and whole blood. The conditions under which HA might be converted to R H during collection and analysis of blood were examined. Provided the blood was kept at 0 ~ C, erythrocyte ketone reductase activity was insignificant. The solid phase extraction method did not generate RH. We studied ten patients taking 25-400 mg/month of HA decanoate and one patient for 4 weeks after the daily oral dose of 120 mg HA was ceased. In the patients on depot HA, the plasma and blood concentrations of HA were not significantly different ( P > 0.1). For the first time, R H was detected in plasma patients on depot drug, but only in three cases. In contrast, RH was present in the blood of eight of these patients. The accumulation of RH in red blood cells was also evident in the patient on oral HA, in whom the mean ratio of R H concentrations in whole blood to plasma was 3.6 • 1.1. Plasma concentrations of HA correlated highly with total neuroleptic activity measured by a radioreceptor assay. Compared to plasma, analysis of concentraOffprint requests to: D.W. Eyles, second address
F
__•0
KEIONE REDUCIASE
~_ F
C-(CH2)3--N~
tions of HA and RH in blood has the advantages of greater sensitivity, of using smaller volumes of blood and of avoiding the efflux of HA and RH during separation of plasma and red cells.
Key words: Haloperidol - Reduced haloperidot - Depot haloperidol - HPLC - Plasma - Blood - Analysis Ketone reductases - Pharmacokinetics - Metabolism Therapeutic drug monitoring - Schizophrenia
Haloperidol (HA) {4-[4-(p-chlorophenyl)-4-hydroxypiperidino]-4'-fluorobutyrophenone} (Fig. 1) is used widely as a neuroleptic agent, particularly in the treatment of schizophrenia. Recently a depot formulation, HA decanoate, has become available. HA has been a popular choice for seeking drug concentration-response relationships in patients with schizophrenia because it is thought to have only one major metabolite with dopaminergic activity, reduced haloperidol (RH) (Fig. 1). To date, however, no clear cut relationships between HA dose or plasma concentrations and response in schizo-
_~OH C- (CH2) 3--N i~t
el ~
HALOPERIDOL
Cl MICROSOMAL OXIDASE,P450?
(HA)
REDUCED HALOPERIDOL
(RH)
DOPANIINEANTAGONIST
LOWAFFINITYFORDOPAMINERECEPTOR
AGONISTAT HA-SENSITIVESIGMASITE
AGONISTAT HA-SENSITIVESIGMASITE
o
c1~-c- (cM2) 3--N C H L O R O H A L O P ERID O L ( C H )
Fig. I. Structures of haloperidol (HA), the only known metabolite with dopaminergic activityreduced haloperidol (RH), and the internal standard for this assay, chloro-haloperidol (CH). The enzymes responsible for the metabolic steps and the receptor activity of each drug in the brain are also indicated
269 phrenic patients have been demonstrated. Dosages and plasma concentrations of HA in clinically stable patients can range over as many as 3 orders of magnitude. The existence of a "therapeutic window" for plasma HA concentrations has been suggested but remains contentious (Froemming et al. 1989). The lack of a clear cut therapeutic range for plasma HA concentrations has been attributed to a number of factors. These include: errors or insensitivities in the analytical methods; failure to account for the pharmacological effects of RH; failure to measure concentrations of HA and RH in red blood cells; failure to account for the reversible metabolism of RH to HA; lack of consideration of compensatory responses in dopaminergic and other receptors in the brain to chronic drug exposure; incorrect patient selection criteria; patient non-compliance; multiple drug therapy; and variable dosing strategies (Midha et al. 1987a; Froemming et al. 1989). With respect to reversible metabolism, it has been demonstrated that RH is reconverted to HA by microsomal enzymes, presumably cytochrome P450, in both human (Tyndale and Inaba 1990), rat and guinea-pig liver (Korpi et al. 1985). Thus, RH can act as a reservoir of HA in vivo. Although RH is not a potent dopamine receptor antagonist, it has been reported to have the same affinity as HA for the " H A sensitive" sigma site, although the clinical relevance of this finding is unknown (Bowen et al. 1990). Several investigators have found a significant relationship between either the sum or the quotient of RH and HA concentrations and clinical response in schizophrenics. In particular, it has been suggested that patients who are classified as non-responders to treatment are more likely to have a high ratio of plasma or red cell RH: HA concentrations (Ereshefsky et al. 1984; Altamura et al. 1988; Bareggi et al. 1990). However, the results of others are not confirmatory (Chang et al. 1987; Ko et al. 1989). As for the HA concentration-response relationship, inconsistencies or errors in the analytical methods for RH could contribute to the conflicting results. Many high performance liquid chromatographic methods which measure HA and RH have been reported. We were concerned about two potential analytical shortcomings of these. The first relates to the use of strong alkalis such as 2 M N a O H to extract plasma and erythrocytes. This could lead to reduction in vitro of HA to RH, similar to the reduction ofchlorpromazine oxides reported by Hubbard et al. (1985). The proposed mechanism is the generation of reactive sulphydryl groups from disulphide bonds in plasma proteins such as albumin. These sulphydryl groups could then act as reducing agents for reactive oxygen species such as chlorpromazine oxide metabolites or the ketone group of HA. The second potential shortcoming relates to the presence of ketone reductases in erythrocytes (Inaba et al. 1989). These could also generate RH from HA after the blood had been collected. Concentrations of lipophilic drugs such as HA and RH can be higher in red blood cells than in plasma. Recently, Ko et al. (1989) reported that RH accumulates in erythrocytes. For pharmacokinetic studies, analysis of
such drugs in whole blood rather than in plasma has two major advantages: higher concentrations and, therefore, a reduction in the volume o f blood required. Bearing in mind the two potential mechanisms for conversion of HA to RH during the sample collection and assay procedure, and wishing to develop a reliable, sensitive assay for HA and RH in whole blood, we have developed an HPLC method which we report here. We make recommendations about blood collection and storage procedures, so that the in vitro production of RH is avoided. The HPLC assay, which involves solid phase extraction and electrochemical detection, is capable of high through-put, a necessity for our studies of the pharmacokinetics of HA. The solid-phase extraction procedure used was adapted from that for plasma reported by Eddington and Young (1988) but was extended to enable analysis of HA and RH in whole blood. The HPLC assay was used to measure steady state concentrations of H A and RH in plasma and whole blood from ten patients with schizophrenia who were receiving depot HA. In addition, the concentrations of H A and RH were measured in plasma and blood over 4 weeks after oral HA was ceased in one patient.
Materials and methods
Materials Haloperidol (HA), reduced haloperidol (RH) and chlorohaloperidol (CH) (Fig. 1) were donated by Janssen-Cilag (Beerse, Belgium). 3[H]-HA,purchased from the manufacturer, was purified by the HPLC method described below immediately before use. Solvents which interfered least with electrochemical detection were chosen. Methanol, acetonitrile, hexane, heptane and isopropyl alcohol were obtained from Mallinckrodt (Ky, USA). Isoamyl alcohol was purchased from BDH, Poole, UK, alkaline cetrimide from Ajax (Australia), menadione from Sigma (Mo, USA), and "EmulsifierSafe" liquid scintillant was supplied by Canberra-Packard (I11, USA). Drugs used to examine their potential interference with the assay were donated by manufacturers in Australia: diazepam, oxazepam, atenolol, pindolol, thioridazine (Alphapharm); pimozide (Janssen); thiothixene (Pfizer); fluphenazine (Squibb); benztropine (Merck Sharp and Dohme); dothiepin (Boots). All other chemicals were analytical reagent grade. Cyanopropyl "'Bond Elute" columns were obtained from Analytichem International (Ca, USA). To prevent the compounds of interest from adhering to glass, all glassware was pretreated by soaking it overnight in 2% alkaline cetrimide solution.
Collection of blood samples The study was approved by the Human Ethics Committees of the Wolston Park Hospital and the University of Queensland. Patients and drug free volunteers gave informed consent. Blood samples were collected from ten patients with chronic schizophrenia, diagnosed according to DSM-III criteria, who had been clinically stable on depot HA for at least 2 months. In all but one patient, depot HA was the only neuroleptic prescribed. Blood was collectedjust before an injection was due. Blood samples were also collected from one patient in whom oral HA dosage (120 mg/day for 2 months) was stopped abruptly for clinical reasons, and from five normal volunteers not taking any medications. In all cases, 10 ml blood was collected by venipuncture and withdrawal into a syringe. Blood was put into two polystyrene 10ml tubes containing ethylenediaminetetraacetic acid (EDTA). One tube was czippcd, mixed
270 and frozen immediately at - 70 ~ C for later analysis as whole blood. The other tube was placed on ice immediately until plasma could be obtained (usually within 10 min) by centrifugation at 4 ~ C. Plasma was stored at - 7 0 ~ C. In the case of the healthy volunteers, blood samples were obtained in a similar fashion except that menadione was added to some collection tubes and some samples were kept at room temperature or 0 ~ C for 30 min or prior to processing and storage.
produce RH from HA during sample collection and processing, whole blood and plasma from five healthy volunteers was collected in (I) the presence of the ketone reductase inhibitor, menadione (260 p.g/ml), (2) the absence of the inhibitor but placed immediately on ice, or (3) left at room temperature. 3[H]-HA, (95 ng/ml) was then added to all tubes. The mixtures were shaken intermittently for 30 min, then extracted and analysed as described. Fractions from each HPLC run were collected at 15-s intervals and aliquots assessed for radioactivity by liquid scintillation counting.
High performance liquid chromatographic assay for HA and R H
Generation of RH from HA during the extraction procedure. To assess the possibility that RH was generated from HA during the extraction process, 50 mM phosphate buffer pH 7.4 or plasma or whole blood samples (2 ml) from four normal volunteers were added to test tubes containing 3[H]-HA (final concentration 95 ng/ ml) and menadione (final concentration 260 lag/ml). The mixtures were shaken intermittently for 30 min then extracted by the above method or by the method of Larsson et al. (1983). Briefly, the latter method consisted of making samples alkaline with 2 M NaOH, extracting into heptane-isoamyl alcohol, back-extracting into 0.05 M sulphuric acid, again making samples basic by adding I M NaOH, and then extracting into hexane-isoamyl alcohol. The organic phase was evaporated, reconstituted in the mobile phase and analysed by the HPLC method described. Fractions from the HPLC runs were collected and the radioactivity counted.
Instrumentation. HPLC analyses were performed using a Waters system (Waters Associates, Milford, MA). This consisted of a 510 pump, 712 autosampler, column oven, 460 ampomeric electrochemical detector with a glassy carbon cell, an inline 0.22 p.m filter, a 3.9 • 150 mm 3 ktm particle size Nova-pak C18 reverse phase column and a frac-100 fraction collector, (Pharmacia LKB, Uppsala, Sweden). The mobile phase was delivered at a flow rate of 1.3 ml/min. The column temperature was maintained at 30 ~ C and the detector was set at a potential of 0.97 V between the reference and working electrodes.
Mobilephase. All solvents were degassed twice by vacuum filtration through 0.22p.m cellulose membranes. Deionised water was prepared by a Milli-Q water system (Millipore, Ma. USA). The mobile phase consisted of 80 mM NaHzPO4 adjusted to pH 7.25 by KOH and acetonitrile in a final ratio of 65: 25 and 0.02 or 0.01% w/v alkaline cetrimide. The column was pre-washed with mobile phase containing 0.02% cetrimide (40 ml), and finally equilibrated at a flow rate of 1.3 ml/min with mobile phase containing 0.01% cetrimide.
Standards. Standards were prepared by spiking plasma and blood with HA and RH to give a concentration range between 0.I142 ng/ml and 0.1-98 ng/ml, respectively. Assay procedure Plasma samples or standards were thawed at 4 ~ C and spun in a refrigerated centrifuge at 12000 g for 10 min. The supernatant (0.5-2 ml) was pipetted into a glass test tube containing CH and mixed briefly. The final concentration of CH was 20 ng/ml. Whole blood samples were thawed and centrifuged and the supernatant (0.5-2 ml) mixed with CH (20 ng/ml) and diluted 1 in 3 with deionised water. Supernatant prepared from plasma or whole blood was loaded onto a disposable 3 ml cyanopropyl "Bond Elute" column mounted in a 24 sample "Vac-Elute" manifold (Analytichem Int. Ca, USA). The column was pre-washed with two sequential 2 ml aliquots of 100% methanol and 3 % methanol/deionised water. After the sample was loaded, the column was washed with 2 ml aliquots of 3% then 20% methanol. The column was then eluted with 2 ml 100% methanol and the eluate collected into glass test tubes. This solution was evaporated under Nz at 60 ~ C. The tubes were cooled, and samples reconstituted in 200 lal mobile phase containing 0.01% cetrimide and centrifuged at 12 000 g. The supernatant was then injected onto the column. The concentrations of HA and RH in samples were determined by comparing the peak height ratios of H A : C H and R H : C H to those of the linear standard curves. Data were stored and processed by a computer utilising Maxima 820 software, (Waters Associates).
Examination of in vitro production of R H from HA during sample collection and extraction Generation of RH from HA by erythrocyte ketone reductases. To determine whether ketone reductase activity in erythrocytes could
Radioreceptor assay for HA All but one patient was receiving HA alone, without any other neuroleptic drug. The remaining plasma samples were also analysed for total neuroleptic activity by the radioreceptor method of Rao (1986). Whole blood was not used because haemoglobin interferes with the radioreceptor assay. Because RH has minimal activity at the dopamine receptor relative to HA, only plasma HA concentrations measured by the two assays were compared.
Statistics The statistical significance of difference between groups was assessed using Student's t test. Mean values are expressed + 1 standard deviation (SD).
Results A c h r o m a t o g r a m f o r w h o l e b l o o d o b t a i n e d f r o m a patient on d e p o t H A is s h o w n in Fig. 2. N o n e o f the o t h e r d r u g s tested (see m a t e r i a l s a n d m e t h o d s ) p r o d u c e d interfering peaks. T h e i n c l u s i o n o f c e t r i m i d e in the m o b i l e p h a s e r e d u c e d r e t e n t i o n t i m e s a n d p e a k tailing b u t did n o t p r o d u c e i n t e r f e r i n g p e a k s o r c o n t r i b u t e to b a s e l i n e levels o r noise. T h e effects o f c e t r i m i d e o n the c h r o m a t o g r a p h i c b e h a v i o u r o f H A a n d R H are p r o b a b l y d u e to a c o m b i n a t i o n o f t w o factors. First, the t e r t i a r y a m i n e s o f b o t h c e t r i m i d e a n d H A a n d R H are c h a r g e d . A t p H 7.25 t h e y will b o t h b i n d to a n y r e s i d u a l silica h y d r o x y l g r o u p s o n the c o l u m n . H o w e v e r , the c e t r i m i d e i n t e r a c t i o n w o u l d d o m i n a t e in this a s s a y b e c a u s e t h e c o n c e n t r a tion o f this c o m p o u n d was 104 times h i g h e r t h a n t h a t o f H A o r R H . T h u s p e a k tailing w o u l d be m i n i m i s e d . S e c o n d , the alkyl c h a i n s o f the c e t r i m i d e m o l e c u l e also i n t e r a c t w i t h the C18 c h a i n s o f the c o l u m n , in direct c o m p e t i t i o n w i t h H A a n d R H . T h u s , r e t e n t i o n times w o u l d be r e d u c e d .
271 Table 2. Comparison of the generation of RH from HA by the liquid phase method of Larsson et al. (1983) and by the solid phase method reported here Sample Extraction method
% RH produced from HA in plasma
% RH produced from HA in whole blood
1 2 3 4 t 2 3 4
1.6 < 0.1 < 0.1 1.1 < 0.1
Psychopharmacology 9 Springer-Verlag 1992
Determination of haloperidol and reduced haloperidol in the plasma and blood of patients on depot haloperidol Darryl W. Eyles 1'2, Harvey A. Whiteford 1, Terry J. Stedman I, and Susan M. Pond 2 1 Clinical Studies Unit, Wolston Park Hospital and 2 University of Queensland, Department of Medicine, Princess Alexandra Hospital, Brisbane 4102, Australia Received March 22, 1991 / Final version
Abstract. We developed a sensitive HPLC assay to measure haloperidol (HA) and its metabolite, reduced haloperidol (RH), in plasma and whole blood. The conditions under which HA might be converted to R H during collection and analysis of blood were examined. Provided the blood was kept at 0 ~ C, erythrocyte ketone reductase activity was insignificant. The solid phase extraction method did not generate RH. We studied ten patients taking 25-400 mg/month of HA decanoate and one patient for 4 weeks after the daily oral dose of 120 mg HA was ceased. In the patients on depot HA, the plasma and blood concentrations of HA were not significantly different ( P > 0.1). For the first time, R H was detected in plasma patients on depot drug, but only in three cases. In contrast, RH was present in the blood of eight of these patients. The accumulation of RH in red blood cells was also evident in the patient on oral HA, in whom the mean ratio of R H concentrations in whole blood to plasma was 3.6 • 1.1. Plasma concentrations of HA correlated highly with total neuroleptic activity measured by a radioreceptor assay. Compared to plasma, analysis of concentraOffprint requests to: D.W. Eyles, second address
F
__•0
KEIONE REDUCIASE
~_ F
C-(CH2)3--N~
tions of HA and RH in blood has the advantages of greater sensitivity, of using smaller volumes of blood and of avoiding the efflux of HA and RH during separation of plasma and red cells.
Key words: Haloperidol - Reduced haloperidot - Depot haloperidol - HPLC - Plasma - Blood - Analysis Ketone reductases - Pharmacokinetics - Metabolism Therapeutic drug monitoring - Schizophrenia
Haloperidol (HA) {4-[4-(p-chlorophenyl)-4-hydroxypiperidino]-4'-fluorobutyrophenone} (Fig. 1) is used widely as a neuroleptic agent, particularly in the treatment of schizophrenia. Recently a depot formulation, HA decanoate, has become available. HA has been a popular choice for seeking drug concentration-response relationships in patients with schizophrenia because it is thought to have only one major metabolite with dopaminergic activity, reduced haloperidol (RH) (Fig. 1). To date, however, no clear cut relationships between HA dose or plasma concentrations and response in schizo-
_~OH C- (CH2) 3--N i~t
el ~
HALOPERIDOL
Cl MICROSOMAL OXIDASE,P450?
(HA)
REDUCED HALOPERIDOL
(RH)
DOPANIINEANTAGONIST
LOWAFFINITYFORDOPAMINERECEPTOR
AGONISTAT HA-SENSITIVESIGMASITE
AGONISTAT HA-SENSITIVESIGMASITE
o
c1~-c- (cM2) 3--N C H L O R O H A L O P ERID O L ( C H )
Fig. I. Structures of haloperidol (HA), the only known metabolite with dopaminergic activityreduced haloperidol (RH), and the internal standard for this assay, chloro-haloperidol (CH). The enzymes responsible for the metabolic steps and the receptor activity of each drug in the brain are also indicated
269 phrenic patients have been demonstrated. Dosages and plasma concentrations of HA in clinically stable patients can range over as many as 3 orders of magnitude. The existence of a "therapeutic window" for plasma HA concentrations has been suggested but remains contentious (Froemming et al. 1989). The lack of a clear cut therapeutic range for plasma HA concentrations has been attributed to a number of factors. These include: errors or insensitivities in the analytical methods; failure to account for the pharmacological effects of RH; failure to measure concentrations of HA and RH in red blood cells; failure to account for the reversible metabolism of RH to HA; lack of consideration of compensatory responses in dopaminergic and other receptors in the brain to chronic drug exposure; incorrect patient selection criteria; patient non-compliance; multiple drug therapy; and variable dosing strategies (Midha et al. 1987a; Froemming et al. 1989). With respect to reversible metabolism, it has been demonstrated that RH is reconverted to HA by microsomal enzymes, presumably cytochrome P450, in both human (Tyndale and Inaba 1990), rat and guinea-pig liver (Korpi et al. 1985). Thus, RH can act as a reservoir of HA in vivo. Although RH is not a potent dopamine receptor antagonist, it has been reported to have the same affinity as HA for the " H A sensitive" sigma site, although the clinical relevance of this finding is unknown (Bowen et al. 1990). Several investigators have found a significant relationship between either the sum or the quotient of RH and HA concentrations and clinical response in schizophrenics. In particular, it has been suggested that patients who are classified as non-responders to treatment are more likely to have a high ratio of plasma or red cell RH: HA concentrations (Ereshefsky et al. 1984; Altamura et al. 1988; Bareggi et al. 1990). However, the results of others are not confirmatory (Chang et al. 1987; Ko et al. 1989). As for the HA concentration-response relationship, inconsistencies or errors in the analytical methods for RH could contribute to the conflicting results. Many high performance liquid chromatographic methods which measure HA and RH have been reported. We were concerned about two potential analytical shortcomings of these. The first relates to the use of strong alkalis such as 2 M N a O H to extract plasma and erythrocytes. This could lead to reduction in vitro of HA to RH, similar to the reduction ofchlorpromazine oxides reported by Hubbard et al. (1985). The proposed mechanism is the generation of reactive sulphydryl groups from disulphide bonds in plasma proteins such as albumin. These sulphydryl groups could then act as reducing agents for reactive oxygen species such as chlorpromazine oxide metabolites or the ketone group of HA. The second potential shortcoming relates to the presence of ketone reductases in erythrocytes (Inaba et al. 1989). These could also generate RH from HA after the blood had been collected. Concentrations of lipophilic drugs such as HA and RH can be higher in red blood cells than in plasma. Recently, Ko et al. (1989) reported that RH accumulates in erythrocytes. For pharmacokinetic studies, analysis of
such drugs in whole blood rather than in plasma has two major advantages: higher concentrations and, therefore, a reduction in the volume o f blood required. Bearing in mind the two potential mechanisms for conversion of HA to RH during the sample collection and assay procedure, and wishing to develop a reliable, sensitive assay for HA and RH in whole blood, we have developed an HPLC method which we report here. We make recommendations about blood collection and storage procedures, so that the in vitro production of RH is avoided. The HPLC assay, which involves solid phase extraction and electrochemical detection, is capable of high through-put, a necessity for our studies of the pharmacokinetics of HA. The solid-phase extraction procedure used was adapted from that for plasma reported by Eddington and Young (1988) but was extended to enable analysis of HA and RH in whole blood. The HPLC assay was used to measure steady state concentrations of H A and RH in plasma and whole blood from ten patients with schizophrenia who were receiving depot HA. In addition, the concentrations of H A and RH were measured in plasma and blood over 4 weeks after oral HA was ceased in one patient.
Materials and methods
Materials Haloperidol (HA), reduced haloperidol (RH) and chlorohaloperidol (CH) (Fig. 1) were donated by Janssen-Cilag (Beerse, Belgium). 3[H]-HA,purchased from the manufacturer, was purified by the HPLC method described below immediately before use. Solvents which interfered least with electrochemical detection were chosen. Methanol, acetonitrile, hexane, heptane and isopropyl alcohol were obtained from Mallinckrodt (Ky, USA). Isoamyl alcohol was purchased from BDH, Poole, UK, alkaline cetrimide from Ajax (Australia), menadione from Sigma (Mo, USA), and "EmulsifierSafe" liquid scintillant was supplied by Canberra-Packard (I11, USA). Drugs used to examine their potential interference with the assay were donated by manufacturers in Australia: diazepam, oxazepam, atenolol, pindolol, thioridazine (Alphapharm); pimozide (Janssen); thiothixene (Pfizer); fluphenazine (Squibb); benztropine (Merck Sharp and Dohme); dothiepin (Boots). All other chemicals were analytical reagent grade. Cyanopropyl "'Bond Elute" columns were obtained from Analytichem International (Ca, USA). To prevent the compounds of interest from adhering to glass, all glassware was pretreated by soaking it overnight in 2% alkaline cetrimide solution.
Collection of blood samples The study was approved by the Human Ethics Committees of the Wolston Park Hospital and the University of Queensland. Patients and drug free volunteers gave informed consent. Blood samples were collected from ten patients with chronic schizophrenia, diagnosed according to DSM-III criteria, who had been clinically stable on depot HA for at least 2 months. In all but one patient, depot HA was the only neuroleptic prescribed. Blood was collectedjust before an injection was due. Blood samples were also collected from one patient in whom oral HA dosage (120 mg/day for 2 months) was stopped abruptly for clinical reasons, and from five normal volunteers not taking any medications. In all cases, 10 ml blood was collected by venipuncture and withdrawal into a syringe. Blood was put into two polystyrene 10ml tubes containing ethylenediaminetetraacetic acid (EDTA). One tube was czippcd, mixed
270 and frozen immediately at - 70 ~ C for later analysis as whole blood. The other tube was placed on ice immediately until plasma could be obtained (usually within 10 min) by centrifugation at 4 ~ C. Plasma was stored at - 7 0 ~ C. In the case of the healthy volunteers, blood samples were obtained in a similar fashion except that menadione was added to some collection tubes and some samples were kept at room temperature or 0 ~ C for 30 min or prior to processing and storage.
produce RH from HA during sample collection and processing, whole blood and plasma from five healthy volunteers was collected in (I) the presence of the ketone reductase inhibitor, menadione (260 p.g/ml), (2) the absence of the inhibitor but placed immediately on ice, or (3) left at room temperature. 3[H]-HA, (95 ng/ml) was then added to all tubes. The mixtures were shaken intermittently for 30 min, then extracted and analysed as described. Fractions from each HPLC run were collected at 15-s intervals and aliquots assessed for radioactivity by liquid scintillation counting.
High performance liquid chromatographic assay for HA and R H
Generation of RH from HA during the extraction procedure. To assess the possibility that RH was generated from HA during the extraction process, 50 mM phosphate buffer pH 7.4 or plasma or whole blood samples (2 ml) from four normal volunteers were added to test tubes containing 3[H]-HA (final concentration 95 ng/ ml) and menadione (final concentration 260 lag/ml). The mixtures were shaken intermittently for 30 min then extracted by the above method or by the method of Larsson et al. (1983). Briefly, the latter method consisted of making samples alkaline with 2 M NaOH, extracting into heptane-isoamyl alcohol, back-extracting into 0.05 M sulphuric acid, again making samples basic by adding I M NaOH, and then extracting into hexane-isoamyl alcohol. The organic phase was evaporated, reconstituted in the mobile phase and analysed by the HPLC method described. Fractions from the HPLC runs were collected and the radioactivity counted.
Instrumentation. HPLC analyses were performed using a Waters system (Waters Associates, Milford, MA). This consisted of a 510 pump, 712 autosampler, column oven, 460 ampomeric electrochemical detector with a glassy carbon cell, an inline 0.22 p.m filter, a 3.9 • 150 mm 3 ktm particle size Nova-pak C18 reverse phase column and a frac-100 fraction collector, (Pharmacia LKB, Uppsala, Sweden). The mobile phase was delivered at a flow rate of 1.3 ml/min. The column temperature was maintained at 30 ~ C and the detector was set at a potential of 0.97 V between the reference and working electrodes.
Mobilephase. All solvents were degassed twice by vacuum filtration through 0.22p.m cellulose membranes. Deionised water was prepared by a Milli-Q water system (Millipore, Ma. USA). The mobile phase consisted of 80 mM NaHzPO4 adjusted to pH 7.25 by KOH and acetonitrile in a final ratio of 65: 25 and 0.02 or 0.01% w/v alkaline cetrimide. The column was pre-washed with mobile phase containing 0.02% cetrimide (40 ml), and finally equilibrated at a flow rate of 1.3 ml/min with mobile phase containing 0.01% cetrimide.
Standards. Standards were prepared by spiking plasma and blood with HA and RH to give a concentration range between 0.I142 ng/ml and 0.1-98 ng/ml, respectively. Assay procedure Plasma samples or standards were thawed at 4 ~ C and spun in a refrigerated centrifuge at 12000 g for 10 min. The supernatant (0.5-2 ml) was pipetted into a glass test tube containing CH and mixed briefly. The final concentration of CH was 20 ng/ml. Whole blood samples were thawed and centrifuged and the supernatant (0.5-2 ml) mixed with CH (20 ng/ml) and diluted 1 in 3 with deionised water. Supernatant prepared from plasma or whole blood was loaded onto a disposable 3 ml cyanopropyl "Bond Elute" column mounted in a 24 sample "Vac-Elute" manifold (Analytichem Int. Ca, USA). The column was pre-washed with two sequential 2 ml aliquots of 100% methanol and 3 % methanol/deionised water. After the sample was loaded, the column was washed with 2 ml aliquots of 3% then 20% methanol. The column was then eluted with 2 ml 100% methanol and the eluate collected into glass test tubes. This solution was evaporated under Nz at 60 ~ C. The tubes were cooled, and samples reconstituted in 200 lal mobile phase containing 0.01% cetrimide and centrifuged at 12 000 g. The supernatant was then injected onto the column. The concentrations of HA and RH in samples were determined by comparing the peak height ratios of H A : C H and R H : C H to those of the linear standard curves. Data were stored and processed by a computer utilising Maxima 820 software, (Waters Associates).
Examination of in vitro production of R H from HA during sample collection and extraction Generation of RH from HA by erythrocyte ketone reductases. To determine whether ketone reductase activity in erythrocytes could
Radioreceptor assay for HA All but one patient was receiving HA alone, without any other neuroleptic drug. The remaining plasma samples were also analysed for total neuroleptic activity by the radioreceptor method of Rao (1986). Whole blood was not used because haemoglobin interferes with the radioreceptor assay. Because RH has minimal activity at the dopamine receptor relative to HA, only plasma HA concentrations measured by the two assays were compared.
Statistics The statistical significance of difference between groups was assessed using Student's t test. Mean values are expressed + 1 standard deviation (SD).
Results A c h r o m a t o g r a m f o r w h o l e b l o o d o b t a i n e d f r o m a patient on d e p o t H A is s h o w n in Fig. 2. N o n e o f the o t h e r d r u g s tested (see m a t e r i a l s a n d m e t h o d s ) p r o d u c e d interfering peaks. T h e i n c l u s i o n o f c e t r i m i d e in the m o b i l e p h a s e r e d u c e d r e t e n t i o n t i m e s a n d p e a k tailing b u t did n o t p r o d u c e i n t e r f e r i n g p e a k s o r c o n t r i b u t e to b a s e l i n e levels o r noise. T h e effects o f c e t r i m i d e o n the c h r o m a t o g r a p h i c b e h a v i o u r o f H A a n d R H are p r o b a b l y d u e to a c o m b i n a t i o n o f t w o factors. First, the t e r t i a r y a m i n e s o f b o t h c e t r i m i d e a n d H A a n d R H are c h a r g e d . A t p H 7.25 t h e y will b o t h b i n d to a n y r e s i d u a l silica h y d r o x y l g r o u p s o n the c o l u m n . H o w e v e r , the c e t r i m i d e i n t e r a c t i o n w o u l d d o m i n a t e in this a s s a y b e c a u s e t h e c o n c e n t r a tion o f this c o m p o u n d was 104 times h i g h e r t h a n t h a t o f H A o r R H . T h u s p e a k tailing w o u l d be m i n i m i s e d . S e c o n d , the alkyl c h a i n s o f the c e t r i m i d e m o l e c u l e also i n t e r a c t w i t h the C18 c h a i n s o f the c o l u m n , in direct c o m p e t i t i o n w i t h H A a n d R H . T h u s , r e t e n t i o n times w o u l d be r e d u c e d .
271 Table 2. Comparison of the generation of RH from HA by the liquid phase method of Larsson et al. (1983) and by the solid phase method reported here Sample Extraction method
% RH produced from HA in plasma
% RH produced from HA in whole blood
1 2 3 4 t 2 3 4
1.6 < 0.1 < 0.1 1.1 < 0.1
