An update on therapeutic drug monitoring and pharmacogenetic testing for the optimization of therapy with psychiatric medication

Introduction

The Therapeutic Drug Monitoring (TDM) of psychopharmaceutical drugs is carried out in numerous laboratories to an ever more frequent extent, as are pharmacogenetic studies. Based on numerous studies that have shown that pharmacogenetic polymorphisms, among other factors, cause large inter-individual differences in the pharmacokinetics of psychopharmaceutical drugs, doctors who prescribe such drugs increasingly order TDM tests. As part of the diagnostics updates 2013 [1] a compilation of the literature on the above topics published in 2011 and 2012 was prepared, and the individual papers and studies were evaluated. This review article is based on this literature compilation.

Basis for the therapeutic drug monitoring of psychopharmaceutical drugs

The Association for Neuropsychopharmacology and Pharmacopsychiatry (AGNP) published an update of its consensus guidelines on TDM in psychiatry at the end of 2011 [2]. These guidelines contained recommendations for all steps of TDM with respect to psychoactive substances.

The guidelines defined the term “therapeutic reference range”. When falling below the lower limit, a response to treatment is improbable, and when the upper limit is exceeded, any further therapeutic improvement becomes impossible. In addition, warning limits were specified for laboratories, above which the doctor should be informed immediately that an overdose has occurred. It was further specified for all drugs described whether therapeutic drug monitoring was highly recommended, recommended, useful or maybe useful. The “highly recommended” category was set aside for all drugs for which controlled clinical studies had demonstrated beneficial effects of TDM. In case of a sub- or supertherapeutic level, the dose should be adjusted until it is within the therapeutic reference range again. The concentrations for drugs in the “recommended” category were arrived at by way of plasma concentration measurements in connection with therapeutic doses of the drugs that were assigned to clinical effects. The “useful” category contained therapeutic reference ranges where drug concentrations were derived from pharmacokinetic studies. The “maybe useful” category, finally, listed the drugs for which plasma concentrations did not correlate with the clinical effects due to the unique pharmacological effect of the drug or for which the dose could be clearly derived from clinical symptoms (e.g., sedatives inducing sleep) [2].

In the case of psychopharmaceutical drugs, the drug concentration should be determined at the start of therapy or at the time of dose modifications if the therapeutic reference range is well defined or if the therapeutic index is narrow. In the event of suspected non-compliance, a determination of all psychopharmaceutical drugs may be useful, regardless of the recommendations for TDM. This is also true if the therapeutic effect is absent. If the drug concentration is within the therapeutic reference range in such a situation, a different treatment should be considered. In the presence of adverse drug reactions, a determination of the drug concentration can show whether the dose is too high and should be reduced. If a patient requires various drugs that are metabolized by the same cytochrome enzymes, interactions in terms of induction or inhibition of the metabolism can occur relative to the individual drugs prescribed. If one of the drugs involved affects the metabolism of another drug, the concentration of the affected drug should be analyzed in order to detect levels of the active ingredient that are either too high or too low. The working group of AGNP recommends that the drug concentration should be determined every 3–6 months in connection with long-term psychopharmaceutical treatment to prevent a recurrence. Apart from that, TDM should be performed for special patient groups, such as pregnant women, children and adolescents or older patients [2].

AGNP recommended TDM as routine monitoring according to the above indications and also when specific problems occur. Blood samples should always be taken as steady-state trough concentrations (i.e., immediately before the next dose). Any order should always contain information on the patient, the diagnosis, medication and therapeutic outcome. Laboratories should use a validated method for analysis, which is used to conduct both internal and external quality checks. The findings report should list the concentration of the drug (and any active metabolites), the unit, a reference range, as well as an interpretation of the results. The ordering physician should then incorporate the results into the further treatment of the patient. If necessary, the dose may be adjusted or the medication may be changed. Finally, recommendations on and help with interpretation were provided, as well as recommendations regarding the combination of TDM with pharmacokinetic tests [2].

These guidelines represent an extensive and widely-covered compilation of the literature published on this broad topic. They should be read by every laboratory that analyzes psychopharmaceutical drugs and by every doctor ordering such analyses, and become part of clinical daily routine.

Pharmacokinetic parameters of psychopharmaceutical drugs

The article by Patteet et al. [3] contains a compilation of the most important pharmacokinetic parameters of common antipsychotics, which are necessary for the determination of drug concentrations and also the interpretation. The effect of the first-generation antipsychotics was traced back mainly to the strong antagonism of the dopamine D2 receptor. An antipsychotic effect was achieved with an occupancy of the D2 receptor in the striatum of 65%–70%; with an occupancy of >80%, the risk of extrapyramidal side effects (EPS) increased. Thus, the therapeutic window with a minimal risk of EPS was at a D2 occupancy of between 65% and 80%. The second-generation antipsychotics were different mainly due to the blocking of various dopamine or serotonergic receptors (with the exception of amisulpride and sulpiride). Here, too, an occupancy of the dopamine receptors of between 65% and 80% resulted in therapeutic success without adverse side effects of the drug. Clozapine was given as an exception, where a receptor occupancy of as little as 30%–40% was associated with a therapeutic effect (Table 1).

According to the authors [3], the psychotic symptoms change after the initiation of treatment with antipsychotics only slowly. If there has been no improvement after 1–4 weeks from the start of therapy, an increased dose or a change of drug will have to be considered. Antipsychotics lead to a number of adverse drug reactions, the most important being EPS. These include Parkinson-like symptoms, dyskinesia, dystonia and akathisia. Sometimes it is very difficult to make the differential diagnosis between EPS, negative symptoms and depression. Tardive dyskinesia is an adverse drug reaction that is very severe and usually irreversible. It can result from long-term treatment with antipsychotics. A common endocrine effect of the dopamine receptor antagonism is an increase in prolactin levels, associated with amenorrhea, erectile dysfunction, infertility and reduced libido [3].

Interactions and monitoring of antipsychotic drugs

In the context of an individualized antipsychotic pharmacotherapy, many patients require more than one drug, because of their inadequate response to monotherapy. Other patients may require, as a result of comorbidity, drugs from other pharmacological classes. But combinations of drugs can lead to pharmacokinetic or pharmacodynamic drug interactions. Pharmacokinetic interactions lead to an increase or a decrease in the drug concentration. The article by Hiemke et al. [4] summarized these issues very nicely. The influence of inhibitors and inducers on the drug concentration is illustrated in Figure 1. When looking at pharmacokinetic interactions, antipsychotics need to be analyzed with respect to their status as a substrate, inhibitor, or inducer of cytochrome P450 (CYP) isoenzymes (Table 2). Not only drugs affect the activity of CYP isoenzymes; for example grapefruit juice, too, is a potent inhibitor of CYP3A4. Psychopharmaceutical drugs themselves have not been described as inducers of CYP isoenzymes. A very potent inducer of CYP1A2 and CYP3A4 is rifampicin, and St. John’s wort is a potent inducer of CYP3A4.

Figure 1

Effect of inhibitors and inducers on the drug concentration when administered either in monotherapy or when omitted in combination therapy (adapted from [4]).

For pharmacodynamic interactions, it is particularly important to look at the effects on the dopamine D2, histamine H2 and acetylcholine M1 receptors. If additive pharmacological effects on these target structures are generated, adverse drug reactions, such as EPS, dizziness, metabolic disorders with weight gain and heart problems, cognitive impairment, delirium or ventricular arrhythmias, may occur [4].

The determination of drug concentrations in all cases of drug interactions is a valuable tool for dose adjustment.

Monitoring of antidepressant therapy with biomarkers or imaging techniques

The determination of drug concentrations represents only a surrogate marker for the assessment of an adequate drug dosage. Biomarkers that change their concentration as part of a successful therapy are generally better markers for therapy optimization, since they reflect the pharmacokinetics and pharmacodynamics, that is, the effect of the drug.

Nichkova et al. [5] have developed a test for determining the concentration of serotonin in urine as a potential biomarker for depression. Most antidepressants, monoamine oxidase inhibitors and 5-hydroxytryptophan (5-HTP) have one or more biogenic amines as a biological target, such as serotonin. These are critical molecules in metabolic pathways that affect mood and depressive expression. Normally, serotonin secretion in urine reflects the release from the enterochromaffin cells in the intestine. But also the synthesis in the kidney and the release from platelets may also include a release from the brain. In the case of neurological diseases, the concentration of serotonin is partially increased in urine, plasma and/or platelets, which happens, for instance, with autism and when taking selective serotonin reuptake inhibitors (SSRI). In depressed patients, the serotonin concentration is reduced in the body fluids mentioned.

The authors of this study developed an ELISA test, which enabled them to determine the concentration of serotonin in urine samples. The validation data showed good levels of precision and accuracy and a very good correlation with the LC-MS/MS method. For a small cohort of 13 patients, they showed that depressed patients’ serotonin levels in urine increased significantly following the start of antidepressant treatment with SSRIs from 39.2±2.4 μg/g creatinine to 183.4±53.2 μg/g creatinine. A similar picture emerged after treatment with 5-HTP [5].

Since there is as yet no easy way in the laboratory to monitor the success of an anti-depressive therapy, determining serotonin in urine using an immunoassay would be a technically simple way to obtain at least an indication of the pharmacodynamic effects. The results of the small patient study are promising, to the effect that serotonin may be a good biomarker for the diagnosis and/or therapy monitoring with SSRI and 5-HTP.

Positron emission tomography (PET) technology is an interesting way to represent the composition of the target receptors for psychoactive drugs in the brain. Grunder et al. [6] have summarized numerous PET studies in a review article conclusively and clearly.

PET technology is already being used in drug development to determine the doses for new drugs. In addition, the pharmacokinetics of psychotropic drugs, in combination with PET, can represent important information about the relationship between blood concentrations of the drug and the occupied target molecules (e.g., receptors, transporters) over the course of time. This allowed for the possibility of calculating the pharmacodynamic drug action in the brain, which constitutes the basis for developing dosing strategies [6].

Using aripiprazole, it has been shown that the dopamine D2 receptors have to be almost fully occupied for an antipsychotic response. This is achieved through plasma concentrations >100 ng/mL. There was no upper limit at which the risk of EPS increased; the probability of side effects was based on individual sensitivities. But it was unclear whether concentrations >300 ng/mL could yield any additional benefit at all. In the case of clozapine, D2 receptor occupancy in the striatum was relatively low at the recommended plasma concentrations between 350 and 600 ng/mL. The blockage of D2 receptors in other brain areas, however, seemed to be significantly higher [6].

For SSRIs, the serotonin transporter (SERT) had to be about 80% occupied by the drug, so that an anti-depressive effect occurred. For selective serotonin/norepinephrine reuptake inhibitors, only the binding of the drugs to the SERT was examined in the PET studies. Typically, at least 80% of the SERT were occupied at the therapeutically effective plasma concentrations. Of the tricyclic antidepressants, only clomipramine was tested for binding to the SERT. In this context, the PET studies found a binding of 80% of the SERT at much lower doses and plasma concentrations than are normally used today. However, the therapeutic active principle of tricyclic antidepressants is not an inhibition of the serotonin reuptake. Therefore, the binding of clomipramine to SERT might represent only an epiphenomenon [6].

The selective norepinephrine reuptake inhibitors and dopamine reuptake inhibitors bind only to the norepinephrine transporter (NET), and only in a small proportion to the dopamine transporter (DAT) [6]. A summary of the PET studies is shown in Table 3.

Studies on factors influencing the concentration of psychotropic drugs

The following two studies retrospectively examined the influence of age and gender on the serum concentrations of antidepressants. The patient groups varied in size, and the results matched only partially. Both studies showed that higher dose-corrected serum concentrations were measured for some medications in elderly patients and that women had higher dose-corrected serum concentrations for a number of drugs. However, the same drugs in the two studies were not always assigned to the same categories. These contradictory results may surely be attributed to the collectives of various sizes and mainly to the retrospective approach of the two studies, because confounding factors can never be ruled out in retrospective studies.

LC-MS/MS analytical methods for the determination of several psychotropic drugs

Many laboratories today analyze psychopharmaceuticals by means of LC-MS/MS multi-analyte methods. In general, these methods allow for a quick, highly sensitive and specific determination of several psychopharmaceuticals. Table 4 provides a summary of four of these analytical methods published in 2011 and 2012 [10–13]. The use of non-deuterated standards in LC-MS/MS methods is no longer state-of-the-art and should be avoided. The validation results were very good for all methods with the exception of a few substances per method that were accepted just barely in terms of accuracy and precision under the applicable guidelines. Unfortunately, when it comes to multi-analyte methods, this is something that must be accepted, because no allowance can be made for specific requirements of an analyte when using these methods.

The role of pharmacogenetics in clinical psychiatry

Lombard and Doraiswamy [14] have prepared a very nice review to illustrate the key points of pharmacogenetics in psychiatry. Today medication for patients in psychiatry is primarily selected on the basis of a process of trial and error: drugs are selected deliberately because of their adverse drug reactions (e.g., administering a sedative antidepressant to a depressed patient suffering from insomnia). Although the serum concentrations often vary greatly, TDM has shown to be rather useless in connection with most of the new psychotropic drugs. Various studies have shown that >75% of patients with schizophrenia discontinue drug therapy within 18 months. For depressed patients it has been shown that >50% of patients did not achieve remission, although they were treated with two or more antidepressants.

The authors defined as the goal of pharmacogenetics the ability to predict the benefit of a drug for a particular patient in order to design an individually tailored therapy with improved efficiency and minimal adverse effects. The pharmacogenetic biomarkers can be used both for the prediction of adverse drug reactions and for the selection of the drug [14].

In psychopharmacogenetics, it is especially the cytochrome P450 enzymes CYP1A, CYP2C19, CYP2D6 and CYP3A that play a role. Polymorphisms in these genes result in an altered metabolism and changes in the drug concentrations of many psychotropic drugs. Depending on the constellation of the aforementioned genes, individuals can be classed into “poor metabolizers” (PMs), “extensive metabolizers” (EMs) and “ultrarapid metabolizers“ (UMs). The PM phenotype leads, for drugs metabolized by this enzyme, to a reduced degradation, while the UM phenotype results in low serum concentrations due to accelerated elimination.

One study showed that CYP2D6 PMs had a risk of significant adverse drug reactions about three times greater after taking risperidone, and about six times the risk of discontinuing treatment due to side effects [14].

The identification of genes is another important area of daily psychiatric routine for pharmacodynamics. Thus, genetic differences in the serotonin transporter (5-HTT) affected both the serotonin concentration and the availability of the transporter as a target of antidepressant therapy. In addition, the efficiency of the therapy and the likelihood of the occurrence of adverse drug reactions were affected [14].

The use of pharmacogenetic tests in psychiatry is limited by different factors, according to the authors, such as the lack of knowledge about this topic and the lack of evidence-based data from randomized studies to demonstrate an improved outcome. The results of several studies and meta-analyses were often inconclusive, and many analytical tests failed to detect all the genetic variability. In addition, there is uncertainty regarding the accountability of these tests, because the costs are generally not covered by health insurance [14].

Pharmacogenetics in psychiatry – the path from research to clinical practice

The publication of Malhotra et al. [15] shows very clearly why pharmacogenetic tests have not played a major role in the clinical practice of psychiatry so far. Due to the poor data from clinical studies, there is now too little solid data to propagate the routine use of tests in psychiatry as well as to achieve a settlement with health insurers.

The DRD2 gene encodes the dopamine D2 receptor, which is the common target structure of all approved antipsychotics. A meta-analysis has demonstrated that functional polymorphisms in the DRD2 promoter region, which modulate gene expression, significantly affect the efficiency of antipsychotic treatment. The same has also been found for the serotonin transporter, which represents the common target structure of serotonin reuptake inhibitors. Carriers of certain genetic constellations with reduced protein expression accounted only for about half to two-thirds of the therapeutic success in comparison to non-carriers of this mutation. Even though these effects were proved in a statistically significant manner, they had insufficient sensitivity and specificity, thus preventing them from entering clinical practice [15].

The absence of convincing data from pharmacogenetic studies hampered the development of clinical pharmacogenetic tests. In order for such tests to be approved by the FDA, they must have a higher sensitivity and specificity than is the case today. Thus, only the Roche AmpliChip R CYP450 test is approved by the FDA at the moment, which examines 27 alleles in CYP2D6 and 3 alleles in CYP2C19. The approval was granted due to the fact that the CYP450 genotype can affect drug action and safety [15].

A problem of pharmacogenetic studies in psychiatry has to do with their clinical endpoints, which were defined by assessing clinical symptoms. The detection of clinical symptoms depends on the high level of subjectivity of patients. Another problem is the non-compliance of patients. Especially the detection of adverse drug reactions is negatively affected by patients not taking their drugs, so that the statistical power is greatly reduced, making it necessary to study a much larger collective. So far, there have been no prospective pharmacogenetic studies in psychiatry [15].

In a second review article by the same authors, numerous studies were summarized that examined the genetic influences on life-threatening adverse drug reactions or those crucial to health policy [16]. In an individualized therapy, a drug would have to be not only individually dosed, but would also have to be identified as such because of the genetic basis for effects and side effects. Numerous studies have examined possible genetic markers for weight gain induced by antipsychotics. This represents a significant public health problem and can be easily determined as a phenotype. In the second-generation antipsychotics (especially clozapine, olanzapine, quetiapine), weight gain is the most common adverse drug reaction, but is also observed in many first-generation antipsychotics. In a recent study, weight gain was examined in 139 pediatric patients receiving for the first time risperidone, aripiprazole or quetiapine. A mutation in the MC4R (melanocortin 4 receptor) gene was identified, which occurred more frequently in patients with weight gain. The results were confirmed in two independent cohorts [15].

The study of genetic influences on psychiatric drugs therapy is often performed retrospectively and on patients already having undergone intensive treatment. Thus, various studies looked at the HTR2C C-759T polymorphism (serotonin receptor 2C) and the role it plays in the weight gain caused by antipsychotic drugs. A meta-analysis showed that the odds ratio was much higher for the few studies with patients undergoing their initial treatment than in studies with patients who had been treated with antipsychotics for longer periods of time [15].

Clozapine-induced agranulocytosis is a dreaded adverse effect of this drug, which prevents its wide application. In the human leukocyte antigen, a mutation was detected (HLA-DQB1, 667G>C) that was encountered with high specificity in patients who developed agranulocytosis as a result of clozapine, but only 21.5% of the affected patients carried this mutation, resulting in a small sensitivity for the biomarker. Given the various studies on this subject, one must assume that a risk profile should be created, instead of determining a single polymorphism [15].

Examples of current pharmacogenetic studies