Metabolism of haloperidol to pyridinium species in patients receiving high doses intravenously: is HPTP an intermediate?

Classics in Chemical Neuroscience: Haloperidol

The discovery of haloperidol catalyzed a breakthrough in our understanding of the biochemical basis of schizophrenia, improved the treatment of psychosis, and facilitated deinstitutionalization. In doing so, it solidified the role for chemical neuroscience as a means to elucidate the molecular underpinnings of complex neuropsychiatric disorders. In this Review, we will cover aspects of haloperidol's synthesis, manufacturing, metabolism, pharmacology, approved and off-label indications, and adverse effects. We will also convey the fascinating history of this classic molecule and the influence that it has had on the evolution of neuropsychopharmacology and neuroscience.

Brain levels of the neurotoxic pyridinium metabolite HPP+ and extrapyramidal symptoms in haloperidol-treated mice

The typical antipsychotic haloperidol is a highly effective treatment for schizophrenia but its use is limited by a number of serious, and often irreversible, motor side effects. These adverse drug reactions, termed extrapyramidal syndromes (EPS), result from an unknown pathophysiological mechanism. One theory relates to the observation that the haloperidol metabolite HPP+ (4-(4-chlorophenyl)-1-[4-(4-fluorophenyl)-4-oxobutyl]-pyridinium) is structurally similar to MPP+ (1-methyl-4-phenylpyridinium), a neurotoxin responsible for an irreversible neurodegenerative condition similar to Parkinson's disease. To determine whether HPP+ contributes to haloperidol-induced EPS, we measured brain HPP+ and haloperidol levels in strains of mice at high (C57BL/6J and NZO/HILtJ) and low (BALB/cByJ and PWK/PhJ) liability to haloperidol-induced EPS following chronic treatment (7-10 adult male mice per strain). Brain levels of HPP+ and the ratio of HPP+ to haloperidol were not significantly different between the haloperidol-sensitive and haloperidol-resistant strain groups (P=0.50). Within each group, however, strain differences were seen (P<0.01), indicating that genetic variation regulating steady-state HPP+ levels exists. Since the HPP+ levels that we observed in mouse brain overlap the range of those detected in post-mortem human brains following chronic haloperidol treatment, the findings from this study are physiologically relevant to humans. The results suggest that strain differences in steady-state HPP+ levels do not explain sensitivity to haloperidol-induced EPS in the mice we studied.

P450-catalyzed vs. electrochemical oxidation of haloperidol studied by ultra-performance liquid chromatography/electrospray ionization mass spectrometry

The metabolites formed via the major metabolic pathways of haloperidol in liver microsomes, N-dealkylation and ring oxidation to the pyridinium species, were produced by electrochemical oxidation and characterized by ultra-performance liquid chromatography/electrospray ionization mass spectrometry (UPLC/ESI-MS). Liver microsomal incubations and electrochemical oxidation in the presence of potassium cyanide (KCN) resulted in two diastereomeric cyano adducts, proposed to be generated from trapping of the endocyclic iminium species of haloperidol. Electrochemical oxidation of haloperidol in the presence of KCN gave a third isomeric cyano adduct, resulting from trapping of the exocyclic iminium species of haloperidol. In the electrochemical experiments, addition of KCN almost completely blocked the formation of the major oxidation products, namely the N-dealkylated products, the pyridinium species and a putative lactam. This major shift in product formation by electrochemical oxidation was not observed for the liver microsomal incubations where the N-dealkylation and the pyridinium species were the major metabolites also in the presence of KCN. The previously not observed dihydropyridinium species of haloperidol was detected in the samples, both from electrochemical oxidation and the liver microsomal incubations, in the presence of KCN. The presence of the dihydropyridinium species and the absence of the corresponding cyano adduct lead to the speculation that an unstable cyano adduct was formed, but that cyanide was eliminated to regenerate the stable conjugated system. The formation of the exocyclic cyano adduct in the electrochemical experiments but not in the liver microsomal incubations suggests that the exocyclic iminium intermediate, obligatory in the electrochemically mediated N-dealkylation, may not be formed in the P450-catalyzed reaction.

Effects of haloperidol and its pyridinium metabolite on plasma membrane permeability and fluidity in the rat brain

The use of antipsychotic drugs is limited by their tendency to produce extrapyramidal movement disorders such as tardive dyskinesia and parkinsonism. In previous reports it was speculated that extrapyramidal side effects associated with the butyrophenone neuroleptic agent haloperidol (HP) could be caused in part by the neurotoxic effect of its pyridinium metabolite (HPP(+)). Although both HPP(+) and HP have been shown to induce neurotoxic effects such as loss of cell membrane integrity, no information exists about the difference in the neurotoxic potency, especially in the potency to induce plasma membrane damage, between these two agents. In the present study, we compared the potency of the interaction of HPP(+) and HP with the plasma membrane integrity in the rat brain. Membrane permeabilization (assessed as [(18)F]2-fluoro-2-deoxy-d-glucose-6-phosphate release from brain slices) and fluidization (assessed as the reduction in the plasma membrane anisotropy of 1,6-diphenyl 1,3,5-hexatriene) were induced by HPP(+) loading (at >or=100 microM and >or=10 microM, respectively), while comparable changes were induced only at a higher concentration of HP (=1 mM). These results suggest that HPP(+) has a higher potency to induce plasma membrane damage than HP, and these actions of HPP(+) may partly underlie the pathogenesis of HP-induced extrapyramidal side effects.

Delayed onset of oculogyric crisis and torticollis with intramuscular haloperidol

OBJECTIVE

To report a case of delayed-onset dystonic reactions, oculogyric crisis (OGC), and torticollis after treatment with intramuscular haloperidol lactate injection.

CASE SUMMARY

A 22-year-old Mexican American woman received intramuscular haloperidol lactate 7.5 mg followed 4 hours later by 10 mg. Twenty-six hours after the first injection, the patient reported that she was unable to lower her gaze and that her neck was stiff. She was immediately given intramuscular benztropine 2 mg; there was a nearly complete remission of symptoms within 15 minutes of treatment. An objective causality assessment revealed a probable relationship between the OGC/torticollis and haloperidol therapy.

DISCUSSION

Dystonic reactions have been reported in 10-60% of patients treated with neuroleptic medication, most commonly when patients just start or increase the dose of the drug. The highest frequency of dystonic reactions has occurred in patients receiving high-potency neuroleptics. It has also been suggested that haloperidol-induced dystonic reactions are a result of the toxic metabolites of that agent.

CONCLUSIONS

OGC and torticollis reactions may occur 12-24 hours after treatment with a high-potency neuroleptic, even in the absence of symptoms of extrapyramidal side effects (EPSEs). The delayed dystonic reaction may begin suddenly (no early EPSE symptomatology).

Binding of 4-(4-chlorophenyl)-1-[4-(4-fluorophenyl)-4-oxobutyl]pyridinium ion (HPP+), a metabolite of haloperidol, to synthetic melanin: Implications for the dopaminergic neurotoxicity of HPP+

The toxicity of 4-(4-chlorophenyl)-1-[4-(4-fluorophenyl)-4-oxobutyl]pyridinium ion (HPP+), a metabolite of haloperidol, toward dopaminergic neurons was investigated. When HPP+ (approximately 100 microM) was added to primary cultures prepared from rat embryonic mesencephalon for 1 h, the survivability of dopaminergic neurons decreased significantly, and this effect was not inhibited by the dopamine transporter (DAT) inhibitor GBR 12909. In addition, HPP+ bound to neuromelanin, which is abundant in dopaminergic neurons. A binding analysis using the Scatchard method showed that there are two classes of binding sites: high affinity sites with a dissociation constant K(d1) of 20.2 nM, and low affinity sites with a K(d2) of 4.0 microM. HPP+ was released easily from synthetic melanin using phosphate buffer (pH 7.0), suggesting that this binding is reversible. The results suggest that the toxicity of HPP+ in dopaminergic neurons is due not to DAT-mediated uptake, but to the binding to neuromelanin.

Assessment of the contributions of CYP3A4 and CYP3A5 in the metabolism of the antipsychotic agent haloperidol to its potentially neurotoxic pyridinium metabolite and effect of antidepressants on the bioactivation pathway

As a plausible explanation for the large interindividual variability in the pharmacokinetics of the neuroleptic agent haloperidol, the contributions of CYP3A isozymes (CYP3A4 and the polymorphic CYP3A5) predominantly involved in haloperidol bioactivation to the neurotoxic pyridinium species 4-(4-Chlorophenyl)-1-[4-(4-fluorophenyl)-4-oxobutyl]-pyridinium (HPP(+)) were assessed in human liver microsomes and heterologously expressed enzymes. Based on recent reports on drug-drug interactions between haloperidol and antidepressants including selective serotonin reuptake inhibitors, the inhibitory effects of antidepressants on the CYP3A4/5-mediated haloperidol bioactivation were also evaluated. HPP(+) formation followed Michaelis-Menten kinetics in microsomes, recombinant CYP3A4, and CYP3A5 with K(m) values of 24.4 +/- 8.9 microM, 18.3 +/- 4.9 microM, and 200.2 +/- 47.6 microM, respectively, and V(max) values of 157.6 +/- 13.2 pmol/min/mg of protein, 10.4 +/- 0.6 pmol/min/pmol P450, and 5.16 +/- 0.6 pmol/min/pmol P450, respectively. The similarity in K(m) values between human liver microsomal and recombinant CYP3A4 incubations suggests that polymorphic CYP3A5 may not be an important genetic contributor to the interindividual variability in CYP3A-mediated haloperidol clearance pathways. Besides HPP(+), a novel 4-fluorophenyl-ring-hydroxylated metabolite of haloperidol in microsomes/CYP3A enzymes was also detected. Its formation was consistent with previous reports on the detection of O-sulfate and -glucuronide conjugates of a fluorophenyl ring-hydroxylated metabolite of haloperidol in human urine. Finally, all antidepressants except buspirone inhibited the CYP3A4/5-catalyzed oxidation of haloperidol to HPP(+) in a concentration-dependent manner. Based on the estimated IC(50) values for inhibition of HPP(+) formation in microsomes, the antidepressants were ranked in the following order: fluoxetine, nefazodone, norfluoxetine, trazodone, and fluvoxamine. These inhibition results suggest that clinically observed drug-drug interactions between haloperidol and antidepressants may arise via the attenuation of CYP3A4/5-mediated 4-(4-chlorophenyl)-1-[4-(4-fluorophenyl)-4-oxobutyl]-4-piperidinol biotransformation pathways.

Liquid chromatographic-mass spectrometric determination of haloperidol and its metabolites in human plasma and urine

Haloperidol and its two metabolites, reduced haloperidol and 4-(4-chlorophenyl)-4-hydroxypiperidine (CPHP) in human plasma and urine were analyzed by HPLC-MS using a new polymer column (MSpak GF-310), which enabled direct injection of crude biological samples without pretreatment. Recoveries of haloperidol and reduced haloperidol spiked into plasma were 64.4-76.1% and 46.8-50.2%, respectively; those for urine were 87.3-99.4% and 94.2-98.5%, respectively; those of CPHP for both samples were not less than 92.7%. The regression equations for haloperidol, reduced haloperidol and CPHP showed good linearity in the ranges of 10-800, 15-800 and 400-800 ng/ml, respectively, for both plasma and urine. Their detection limits were 5, 10 and 300 ng/ml, respectively, for both samples. Thus, the present method was sensitive enough for detection and determination of high therapeutic and toxic levels for haloperidol and its metabolites present in biological samples.

Metabolic studies on haloperidol and its tetrahydropyridinyl dehydration product (HPTP) in C57BL/6 mouse brain preparations

The neuroleptic agent haloperidol (HP) and its tetrahydropyridinyl dehydration product HPTP are biotransformed by humans, baboons and rodents to the HP pyridinium (HPP(+)) and reduced HP pyridinium (RHPP(+)) species, potential neurotoxic metabolites that have been detected in the brain. HPP(+), however, does not pass the mouse blood-brain barrier since it is not detected in the brain following systemic administration. We report here that C57BL/6 mouse brain preparations catalyze the oxidation of HP and HPTP to HPP(+). The initial rate of HPP(+) formation from HPTP by whole brain homogenates was estimated to be approximately 20 times faster than that observed with HP as substrate. HPTP also was converted to HPP(+) by mouse brain microsomal preparations and brain slices. These results suggest that the presence of HPP(+) in the C57BL/6 mouse brain following systemic administration of HPTP may be due primarily to its in situ metabolism to HPP(+). Attempts to identify the catalyst responsible for these biotransformations, however, have not been successful.

Potent inhibition of CYP2D6 by haloperidol metabolites: stereoselective inhibition by reduced haloperidol

AIMS

We evaluated the inhibitory effect of haloperidol and its metabolites on CYP2D6 activity in order to better understand the potential role of these metabolites in drug interactions involving haloperidol.

METHODS

The inhibitory effects of haloperidol and five of its metabolites on dextrorphan formation from dextromethorphan, a marker probe of CYP2D6 activity, were measured in human liver microsomal preparations. Apparent kinetic parameters for enzyme inhibition were determined by nonlinear regression analysis of the data.

RESULTS

Racemic reduced haloperidol and its metabolite, RHPTP competitively inhibited dextromethorphan O-demethylation with estimated Ki values (0.24 microM and 0.09 microM, respectively) that were substantially lower than that of haloperidol (0.89 microM). The inhibitory effect of S(-)-reduced haloperidol was more potent than the R(+)-enantiomer, with estimated Ki values of 0.11 microM and 1.1 microM, respectively. The pyridinium metabolite of haloperidol, HPP+ inhibited the enzyme activity noncompetitively with a Ki value of 0.79 microM. The N-dealkylated metabolites of haloperidol (FBPA and CPHP) had a diminished inhibitory potency. While FBPA showed no notable inhibitory effect on dextrorphan formation, CPHP showed moderate competitive inhibition with a Ki value of 20.9 microM.

CONCLUSIONS

The principal metabolites of haloperidol inhibit CYP2D6, suggesting that they might contribute to the inhibitory effects of the drug. Reduced haloperidol seems to inhibit CYP2D6 activity in an enantioselective manner with the physiologically occurring S(-) enantiomer being more potent.

Disposition of haloperidol pyridinium and reduced haloperidol pyridinium in schizophrenic patients: no relationship with clinical variables during short-term treatment

In an open clinical trial, serum concentrations of haloperidol pyridinium (C(HP+)) and reduced haloperidol pyridinium (C(RHP+)), as well as haloperidol (CH) and reduced haloperidol (C(RH)), were measured in 57 schizophrenic and schizoaffective inpatients during 6 weeks of short-term treatment. Psychopathology was monitored with the Brief Psychiatric Rating Scale (BPRS), and extrapyramidal adverse effects were assessed with the Extrapyramidal Symptom Rating Scale (EPS). Significantly linear relationships were found between haloperidol dose (D) and pyridinium metabolite serum concentrations, as well as between C(H) and the pyridinium metabolite serum concentrations. C(HP+) (range, 0.2-4.9 ng/mL) and C(RHP+) (range, 0.03-6.23 ng/mL) were low compared with C(H) and C(RH), being as mean values approximately 7% and 14% of C(H) and C(RH), respectively. Additionally, the values of C(RHP+) and the slope of the correlation of C(H) with the C(RHP+)/C(HP+) ratio were considerably lower than in a previous report of long-term treatment with haloperidol. This is explained by the shorter time of treatment of the present study. Carbamazepine comedication was found to not influence relative pyridinium metabolite serum concentrations C(HP+)/D and C(RHP+)/D. However, the aromatization ratios of haloperidol (C(HP+)/C(H)) and reduced haloperidol (C(RHP+)/C(RH)) were increased by concomitant carbamazepine. As the main result, no relationships between the pyridinium metabolite serum concentrations and clinical variables (BPRS change, EPS, dose of biperiden) were detected. For instance, the aromatization ratios C(HP+)/C(H) and C(RHP+)/C(RH) did not predict clinical improvement or extrapyramidal adverse effects. Therefore, no confirmation of the "pyridinium hypothesis," which suggests haloperidol pyridinium metabolites to be the origin of adverse effects and decreased therapeutic effect, can be derived from this study. However, the authors emphasize that pyridinium metabolites cannot be excluded as the origin of decreased therapeutic effect in long-term treatment and of adverse effects not investigated in the present study, such as tardive dyskinesia. Finally, it is concluded that the serum concentration of the parent drug remains the main variable of interest in the therapeutic drug monitoring of haloperidol during short-term treatment.

Pharmacokinetics of haloperidol: an update

Haloperidol is commonly used in the therapy of patients with acute and chronic schizophrenia. The enzymes involved in the biotransformation of haloperidol include cytochrome P450 (CYP), carbonyl reductase and uridine diphosphoglucose glucuronosyltransferase. The greatest proportion of the intrinsic hepatic clearance of haloperidol is by glucuronidation, followed by the reduction of haloperidol to reduced haloperidol and by CYP-mediated oxidation. In studies of CYP-mediated disposition in vitro, CYP3A4 appears to be the major isoform responsible for the metabolism of haloperidol in humans. The intrinsic clearances of the back-oxidation of reduced haloperidol to the parent compound, oxidative N-dealkylation and pyridinium formation are of the same order of magnitude, suggesting that the same enzyme system is responsible for the 3 reactions. Large variation in the catalytic activity was observed in the CYP-mediated reactions, whereas there appeared to be only small variations in the glucuronidation and carbonyl reduction pathways. Haloperidol is a substrate of CYP3A4 and an inhibitor, as well as a stimulator, of CYP2D6. Reduced haloperidol is also a substrate of CYP3A4 and inhibitor of CYP2D6. Pharmacokinetic interactions occur between haloperidol and various drugs given concomitantly, for example, carbamazepine, phenytoin, phenobarbital, fluoxetine, fluvoxamine, nefazodone, venlafaxine, buspirone, alprazolam, rifampicin (rifampin), quinidine and carteolol. Overall, drug interaction studies have suggested that CYP3A4 is involved in the biotransformation of haloperidol in humans. Interactions of haloperidol with most drugs lead to only small changes in plasma haloperidol concentrations, suggesting that the interactions have little clinical significance. On the other hand, the coadministration of carbamazepine, phenytoin, phenobarbital, rifampicin or quinidine affects the pharmacokinetics of haloperidol to an extent that alterations in clinical consequences would be expected. In vivo pharmacogenetic studies have indicated that the metabolism and disposition of haloperidol may be regulated by genetically determined polymorphic CYP2D6 activity. However, these findings appear to contradict those from studies in vitro with human liver microsomes and from studies of drug interactions in vivo. Interethnic and pharmacogenetic differences in haloperidol metabolism may explain these observations.

Clinical and neuropathological abnormalities in baboons treated with HPTP, the tetrahydropyridine analog of haloperidol

Tardive dyskinesia (TD) is relatively common among psychiatric patients on maintenance therapy with typical neuroleptics and persists in more than 20% even after withdrawal of the medication. Such persistence suggests an underlying pathology due to neurotoxicity. We present evidence for such a neurotoxic mechanism in a baboon model of TD. Four baboons were treated chronically with the dehydration product of haloperidol, 4-(4-chlorophenyl)-1-[4-(4-fluorophenyl)-4-oxobutyl]-1,2,3,6- tetrahydropyridine (HPTP), which is metabolized, similarly to haloperidol, to two neurotoxic pyridinium species. The animals developed orofacial dyskinesia which persisted after HPTP was ceased. Serial sections of the entire brain from the four treated animals and four vehicle-treated controls revealed volume loss in the basal forebrain and hypothalamus. Histological evaluation demonstrated a reduction in the density of magnocellular neurons in the anterior region of the nucleus basalis of Meynert (NbM). We speculate that the loss of these NbM neurons may be associated with the persistent orofacial dyskinesia observed in the HPTP-treated animals. These findings may contribute to a better understanding of neuroleptic-induced TD.

Effects of haloperidol metabolites on neurotransmitter uptake and release: possible role in neurotoxicity and tardive dyskinesia

This research explored the effects of haloperidol (HP) metabolites on biogenic amine uptake and release, and compared them to those of MPTP and its toxic metabolite, MPP+. In synaptosome preparations from mouse striatum and cortex, the HP metabolites haloperidol pyridinium (HPP+), reduced haloperidol pyridinium (RHPP+), and haloperidol tetrahydropyridine (HPTP) inhibited the presynaptic uptake of dopamine and serotonin, with greater affinity for the serotonin transporter. HPP+ was the most potent inhibitor of dopamine uptake, and HPTP of serotonin uptake, both with IC50 values in the low micromolar range. RHPP+ was less active than the other metabolites, but was more active than the parent compound, HP. Inhibition of uptake was reversed when free drug was removed by centrifugation and then resuspension of the synaptosomes in fresh buffer, suggesting that inhibition of uptake was due to interaction with the transporters and was not due to irreversible cytotoxicity. HPP+ showed noncompetitive inhibition of both serotonin and dopamine uptake, suggesting that it has a relatively slow dissociation rate for its interaction with the transporter proteins. In experiments on amine release, HPP+ and HPTP were four-fold less potent than MPP+ for releasing preloaded dopamine from striatal synaptosomes, and only MPP+-dependent release was antagonized by the uptake blocker, mazindol. In contrast, RHPP+ displayed little ability to release either amine neurotransmitter. HPTP was about two-fold more potent than MPP+ for releasing serotonin from cortical synaptosomes, whereas HPP+ was less active than MPP+. The specific serotonin transport blocker fluoxetine was only able to antagonize release induced by MPP+. These results suggest that HP metabolites bind to the transporters for dopamine and serotonin, but are not transporter substrates. In contrast to their potent effects on amine release, HPP+ and HPTP were unable to release preloaded GABA from cortical synaptosomes. The implications of these results concerning a possible role of HP metabolites in the development of tardive dyskinesia are discussed.