Schizophrenia is one of the most debilitating and costly illnesses worldwide. First-generation antipsychotics such as chlorpromazine and haloperidol succeeded in controlling the positive symptoms of schizophrenia, but had significant extrapyramidal effects that led to the search for new agents and the release of second-generation (or atypical) antipsychotics. These drugs had a lower risk of adverse motor symptoms. Therapeutic drug monitoring has become a useful tool to optimize schizophrenia treatment and HPLC-MS/MS has been considered the primary technique to monitor antipsychotics. This review comprises three sections: schizophrenia pathophysiology and treatment; recent advances in LC-MS/MS methods designed to measure levels of antipsychotics and their metabolites in plasma samples (selectivity, matrix effect and sensitivity); and the importance of therapeutic drug monitoring.

The 1975 publication of Seeman et al. (Proc Nat Acad Sci, USA), reporting the discovery of the antipsychotic receptor in the brain, is a classic example of translational medicine research. In searching for a pathophysiological mechanism of psychosis, the team sought to identify sites that bound the antipsychotic drug haloperidol. Their criterion was that haloperidol bound to the site at one to two nanomoles per liter, corresponding to haloperidol concentrations found in spinal fluid or plasma water in treated patients. They requested de novo synthesis of tritiated haloperidol, and it readily detected specific haloperidol binding sites in brain striatum. With dopamine binding the haloperidol-labeled sites with higher potency than other neurotransmitters, the sites were named antipsychotic/dopamine receptors (now designated dopamine D2 receptors). Most significantly, they found that all antipsychotics bound these sites at concentrations and with a rank order of potencies that were directly related to the mean daily antipsychotic dose taken by patients with schizophrenia. Their findings enabled screening for new antipsychotics, initiated D2 receptor measurements in brain of living patients, and determination of minimum occupancy (65%) of D2 receptors for antipsychotic benefit. The collective work is generally viewed as providing a fundamental basis for the dopamine hypothesis of schizophrenia.

The aim of this review is to provide information for interpreting outcome results from monitoring of antipsychotics in biological samples. A brief overview of the working mechanisms, pharmacological effects, drug interactions, and analytical methods of classical and atypical antipsychotics is given. Nineteen antipsychotics were selected based on their importance in the worldwide market as follows: amisulpride, aripiprazole, asenapine, bromperidol, clozapine, flupenthixol, haloperidol, iloperidone, lurasidone, olanzapine, paliperidone, perphenazine, pimozide, pipamperone, quetiapine, risperidone, sertindole, sulpiride, and zuclopenthixol. A straightforward relationship between administered dose, plasma or serum concentration, clinical outcome, or adverse effects is often lacking. Nowadays, focus lies on therapeutic drug monitoring and individualized therapy to find adequate treatment, to explain treatment failure or nonresponse, and to check patient compliance. However, extensive research in this field is still mandatory.

After a 12-year search for the antipsychotic receptor, the binding site was discovered and labelled by [3H]haloperidol in 1975. Of the various neurotransmitters, dopamine was the most potent in inhibiting the binding of [3H]haloperidol, indicating that the antipsychotic receptor was a dopamine receptor, now named the dopamine D2 receptor, a major targeting site in schizophrenia. All antipsychotic drugs, including traditional and newer antipsychotics, either bind to D2 in direct relation to their clinical potencies or hinder normal dopamine neurotransmission, as in the case of partial dopamine agonists. In fact, the antipsychotic concentrations found in the plasma water of treated patients closely match the predicted therapeutic absolute concentrations, adjusted for the 60-75% D2 occupancy needed for clinical efficacy. Antipsychotics that elicit low or no Parkinsonism or prolactinaemia are loosely attached to D2 and rapidly dissociate from D2, whereas those eliciting Parkinsonism stay tightly attached to D2 for many hours. Because animal models of psychosis (amfetamine sensitisation, brain lesions) all show a marked elevation in the number of high-affinity states of D2, the antipsychotics are thought to specifically target these D2High states in psychosis in general and schizophrenia in particular.

The cyclic voltammetric behavior of haloperidol at a hanging mercury drop electrode was studied in Britton-Robinson buffer series of pH 2.5-11 containing 40% (v/v) ethanol. A single two-electron irreversible cathodic peak was obtained which attributed to reduction of the CO double bond. In addition, a small enhanced adsorptive pre-wave was observed at less negative potentials over the pH range 3.5-11. Controlled adsorptive accumulation of haloperidol onto the hanging mercury drop electrode provided the basis for its direct trace assay in bulk form, pharmaceutical formulation and human biological fluids using square-wave adsorptive cathodic stripping voltammetry. Following preconcentration of bulk haloperidol onto the HMDE a well-developed square-wave cathodic peak was generated in Britton-Robinson buffer especially at pH values 9-10; its peak current showed a linear dependence on the concentration of haloperidol over the range 1 x 10(-9)M to 1.5 x 10(-6)M depending on the preconcentration duration. The procedural parameters for assay of haloperidol were studied. The achieved limits of detection (LOD) and quantitation (LOQ) were 3.83 x 10(-10)M and 1.28 x 10(-9)M bulk haloperidol, respectively. The procedure was successfully applied to assay haloperidol in tablets (Safinace) and in spiked human serum and urine. LOD of 3.3 x 10(-9)M and 5.46 x 10(-9)M, and LOQ of 1.10 x 10(-8) and 1.82 x 10(-8)M haloperidol were achieved in spiked human serum and urine samples, respectively.

A quantitative method for the analysis of haloperidol in human plasma is described. Sample clean-up was performed by means of solid-phase extraction using 3M Empore extraction disk plates in the 96-well format, automated with a Canberra Packard pipetting robot. Separation was performed by reversed phase high performance liquid chromatography with turbo ionspray tandem mass spectrometric detection by monitoring the decay of protonated haloperidol of m/z 376 to its fragment at m/z 165, versus the decay of protonated haloperidol-D4 at m/z 380 to its fragment at m/z 169. The validated concentration range was from 0.100 to 50.0 ng ml(-1), with an inaccuracy and overall imprecision below 10% at all concentration levels. Validation results on linearity, specificity, precision, accuracy and stability are shown and are found to be adequate. The average sample preparation time for a batch of 96 samples is approximately 50 min. The chromatographic run time is 3 min. A sample throughput of at least 240 samples per day can be achieved.

1. p-Fluorobenzoyl-propionic acid, 4-(4-chlorophenyl)-4-hydroxy-piperidine, and reduced haloperidol were confirmed as metabolites of haloperidol. Their formation was catalysed by hepatic microsomes and was NADPH dependent. 2. The pyridinium metabolite of haloperidol (HP+) was identified. It is proposed that haloperidol first undergoes dehydration to form its 1,2,3,6-tetrahydropyridine analogue (HTP). HTP is then further metabolized to HP+, HTP N-oxide and its N-dealkylated product, 4-chlorophenyl-1,2,3,6-tetrahydropyridine (CPTP). HTPN-oxide was metabolized to CPTP and HTP. All these metabolites were confirmed by comparison with synthesized compounds using h.p.l.c. and h.p.l.c.-mass spectrometry. 3. Three unknown metabolites were present in microsomal metabolic incubations of haloperidol. One of them was tentatively characterized by h.p.l.c.-mass spectrometry as an oxygenated product of haloperidol, another appears to be the 2-pyridine analogue of haloperidol. The third metabolite was shown to be a neutral compound of unknown structure, which was not haloperidol N-oxide nor 4-hydroxy-4'-fluorobutyrophenone. In addition, HTP was metabolized to a further unknown product with a similar u.v. spectrum to that of HTP. 4. The identification of these metabolites led to the hypothesis that the metabolism of haloperidol is similar to that of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) and may therefore produce motor neurone toxicity by a similar mechanism.

Following cannulation of the right common carotid artery of female Sprague-Dawley rats, 3 microCi (10 micrograms) of either [3H]haloperidol or [3H]amphetamine were infused. At various time intervals, drug concentrations were determined in the right and left striata, anterior forebrains, posterior forebrains and cerebella. One minute following unilateral intracarotid infusion of haloperidol, approximately a 90-100-fold right/left (ipsilateral/contralateral) difference in drug concentrations was attained in the striatum and the posterior forebrain, while more than a 75-fold difference was evident in the anterior forebrain. One minute following amphetamine infusion, a difference greater than 40-fold was seen in all forebrain structures. The right-left differences steadily declined with time as a result of the declining drug concentrations of the infused hemisphere. The pharmacokinetic parameters of both the distribution and elimination phases were similar in each forebrain region for both haloperidol and amphetamine. The kinetic parameters did, however, show specific drug differences. Bilateral drug concentrations in the striatum following intraperitoneal administration of amphetamine to unilaterally cannulated rats were nearly identical. Therefore, the cannulation procedure did not significantly alter the blood supply to either hemisphere. This is the first study to quantify drug concentrations and to analyze pharmacokinetic parameters following unilateral cerebral drug administration in conscious animals. This technique should be useful in studying functional and biochemical interhemispheric relationships as well as lateralized behaviors.

The discovery of neuroleptic drugs in 1952 provided a new strategy for seeking a biological basis of schizophrenia. This entailed a search for a primary site of neuroleptic action. The Parkinsonian effects caused by neuroleptics suggested that dopamine transmission may be disrupted by these drugs. In 1963 it was proposed that neuroleptics blocked "monoamine receptors" or impeded the release of monoamine metabolites. The neuroleptic concentration in plasma water or cerebrospinal fluid was of the order of 2 nM for haloperidol in clinical therapy. A systematic research was made between 1963 and 1974 for a primary site of neuroleptic action which would be sensitive to 2 nM haloperidol and stereoselective for (+)-butaclamol. Direct evidence that neuroleptics selectively blocked dopamine receptors occurred in 1974 with the finding that nanomolar concentrations of these drugs stereoselectively inhibited the binding of [3H]-dopamine or [3H]-haloperidol. These binding sites, now termed D2 dopamine receptors (which inhibit adenylate cyclase), are blocked by neuroleptics in direct relation to the antipsychotic potencies of the neuroleptics. No such correlation exists for D1 receptors (which stimulate adenylate cyclase). Based on the fact that dopamine-mimetic drugs elicited hallucinations, and that neuroleptics caused rigidity, Van Rossum in 1966 had suggested a hypothesis that dopamine pathways may be overactive in schizophrenia. The D2-selective blockade by all neuroleptics (except the monoamine-depleting reserpine) provided strong support for the dopamine hypothesis. Further support now comes from postmortem data and in vivo positron tomographic data, both of which indicate that the density of D2 receptors are elevated in the schizophrenic brain. The postmortem data indicate a bimodal pattern with half the schizophrenics having striatal D2 densities of 14 pmol/g (control is 13 pmol/g) and the other half having 26 pmol/g. Current positron tomographic data indicate D2 densities of 14 pmol/g in control subjects, but values of 34 pmol/g in drug-naive schizophrenics. Future tests of the dopamine hypothesis of schizophrenia may entail an examination of the amino acid composition and genes for D2 receptors in schizophrenic tissue, an examination of the ability of the D2 receptor to become phosphorylated and to desensitize into the low-affinity state, and an examination of the interaction of D2 receptors with D1 receptors or other neurotransmitters.

An improved method for the analysis of haloperidol in human serum, utilizing gas chromatography-ammonia chemical ionization mass spectrometry is described. A tetradeutero analogue of haloperidol is utilized as the internal standard, while a second drug, thioridazine, is added as a priming compound. The characteristic high sensitivity and selectivity of selected-ion monitoring combined with the added accuracy provided by incorporation of a labeled internal standard provide a reliable method for the quantitation of low levels of haloperidol.

A method for the identification of butyrophenone and bisfluorophenyl neuroleptics and their predominant basic metabolites in urine after acid hydrolysis is described. The acetylated extract is analysed by computerized gas chromatography-mass spectrometry. An on-line computer allows rapid detection using mass fragmentography with the masses m/e 112, 123, 134, 148, 169, 257, 321 and 189, 191, 223, 233, 235, 245, 287, 297. The identity of positive signals in the reconstructed mass fragmentograms is established by a comparison of the entire mass spectra with those of standards. The mass fragmentograms and the underlying mass spectra are documented.

Haloperidol serum concentrations were determined after IM or oral treatment in 15 schizophrenic patients. No correlation was found between drug levels and therapeutic effect. However, a good relationship was found between the half-life calculated after the first IM injection and the BPRS decrease after 3 weeks. Therefore a serum level study to the first day may forecast the therapeutic response.

Tissue distribution studies of [18F]haloperidol and [82Br]bromperidol, two potent neuroleptic drugs, were performed in rats by serial sacrifice. The usefulness of external scintigraphy in obtaining tissue distribution data in large animals is demonstrated by the tissue distribution of [18F]haloperidol in rhesus monkeys. Both serial sacrifice and external scintigraphic studies demonstrated that uptake of the two drugs after intravenous administration into their target organ, the brain, was very fast and that the ratio of brain to blood levels was high throughout the 2-hr observation. Bromperidol appeared to reach peak brain levels faster than its chloro analog, haloperidol. Both bromperidol and haloperidol concentrated overwhelmingly in the rat lung. Haloperidol also showed a high affinity for the monkey lung. The disposition pattern in rats of [18F]-beta-(4-fluorobenzoyl)propionic acid, an apparent intermediate in butyrophenone metabolism, was entirely different from that of the parent drugs. This metabolite did not concentrate in the rat brain.

A new gas chromatographic mass spectrometric chemical ionization assay for haloperidol with selectedion monitoring is presented which provides for better combined selectivity and sensitivity than previous assays. Levels of haloperidol in 2 ml of human plasma were reproducibly measured down to subnanogram levels. Both methane and methane--ammonia chemical ionization spectra are presented for haloperidol and the internal standard trifluperidol.

A sensitive gas-chromatographic method for quantitative analysis of haloperidol in human plasma is described. The use of nitrogen-phosphorus selective detection reduces the time required for analysis. Azaperone is used as the internal reference standard. The method is suitable for the determination of haloperidol plasma levels in patients treated with doses ranging from 1.2 to 200 mg/day. No interference from drugs needed in the associated antipsychotic therapy has been found. The simplicity, specificity and sensitivity of the method make it suitable for routine analysis of haloperidol plasma levels in psychotic patients undergoing chronic treatment.