Overdosage of antidepressants: clinical and pharmacokinetic aspects

Automated Interlaboratory Comparison of Therapeutic Drug Monitoring Data and Its Use for Evaluation of Published Therapeutic Reference Ranges

Therapeutic drug monitoring is a tool for optimising the pharmacological treatment of diseases where the therapeutic effect is difficult to measure or monitor. Therapeutic reference ranges and dose-effect relation are the main requirements for this drug titration tool. Defining and updating therapeutic reference ranges are difficult, and there is no standardised method for the calculation and clinical qualification of these. The study presents a basic model for validating and selecting routine laboratory data. The programmed algorithm was applied on data sets of antidepressants and antipsychotics from three public hospitals in Denmark. Therapeutic analytical ranges were compared with the published therapeutic reference ranges by the Arbeitsgemeinschaft für Neuropsychopharmakologie und Pharmakopsychiatrie (AGNP) and in additional literature. For most of the drugs, the calculated therapeutic analytical ranges showed good concordance between the laboratories and to published therapeutic reference ranges. The exceptions were flupentixol, haloperidol, paroxetine, perphenazine, and venlafaxine + o-desmethyl-venlafaxine (total plasma concentration), where the range was considerably higher for the laboratory data, while the calculated range of desipramine, sertraline, ziprasidone, and zuclopenthixol was considerably lower. In most cases, we identified additional literature supporting our data, highlighting the need of a critical re-examination of current therapeutic reference ranges in Denmark. An automated approach can aid in the evaluation of current and future therapeutic reference ranges by providing additional information based on big data from multiple laboratories.

Nortriptyline serum concentration as a predictor for cardiac risk in amitriptyline-treated patients

PURPOSE

Tricyclic antidepressants have been shown to affect electrocardiogram (ECG) parameters, but there is limited evidence in relation to the serum concentrations. Therefore, we aimed to evaluate a prediction of cardiac risk in amitriptyline- and doxepin-treated patients by serum concentrations.

PATIENTS AND METHODS

The association between serum concentrations of amitriptyline (n = 100) and doxepin (n = 71) and ECG parameters was retrospectively examined using linear regression analysis. Mann-Whitney U tests were applied to evaluate differences in QTc intervals in patients with serum concentrations above and below the upper limit of the therapeutic reference range, as well as the alert level of each target drug.

RESULTS

The sum serum concentration of amitriptyline and the nortriptyline serum concentration were significantly associated with an increased PQ interval (p = 0.020, p = 0.007), as well as with increased QTcB (p = 0.012, p < 0.001) and QTcF intervals (p = 0.025, p < 0.001). The nortriptyline concentration was significantly associated with the QRS interval (p = 0.003). In patients with active moiety concentrations above the alert level (300 ng/ml) and nortriptyline concentrations above the reference range (170 ng/ml), the QTcB interval was significantly prolonged (p = 0.032, p = 0.007). No significant association with any ECG parameter was detected for doxepin serum concentrations.

CONCLUSION

The effect of amitriptyline on ECG parameters may be explained by nortriptyline alone. Accordingly, with increasing nortriptyline concentrations, the potential risk for an atrioventricular block, a bundle branch block, and prolongation of QTc interval may increase significantly.

Increase of Heart Rate and QTc by Amitriptyline, But Not by Venlafaxine, Is Correlated to Serum Concentration

Electrocardiographic pathologies are a common problem during antidepressant treatment. The authors investigated the association of serum concentrations of antidepressants and heart rate, QT, and QTc. Polymorphisms of NOS1AP (nitric oxide synthase 1 adaptor protein) rs10494366 and rs12143842 as potential influence factors also were considered. In the amitriptyline sample (n = 59), significant Spearman ρ correlations were found between serum concentration and QTc (r = 0.333, P = 0.010), as well as heart rate (r = 0.407, P = 0.001). Patients with a serum concentration greater than the therapeutic range (>200 ng/mL) exhibit significantly higher heart rates (87.0 ± 13.3 vs 80.0 ± 13.9, U test P = 0.011) and higher QTc values (443.8 ± 28.8 vs 427.9 ± 20.6, U test P = 0.022). Excluding the 26 patients with a serum concentration greater than the therapeutic range, patients with rs12143842 risk alleles exhibit higher heart rates and as a trend lower QT intervals with no difference in QTc. In the venlafaxine sample (n = 81), no significant association between serum concentration and heart rate, QT, or QTc was revealed. In summary, the risk for relevant electrocardiographic alterations induced by tricyclic antidepressants, such as amitriptyline, is dependent on serum concentrations. NOS1AP polymorphisms may be a genetic vulnerability factor.

The value of drug and metabolite concentration in blood as a biomarker of psychopharmacological therapy

Desirable and undesirable effects of a drug are related to its concentration at various sites of actions. For many psychotropic drugs, it has been shown that drug concentration in brain correlates with concentration in blood. The latter is also an available estimate of clearance and bioavailability. Its monitoring enables identification of multiple factors that have an impact on clinical outcomes, especially uncertain compliance and pharmacokinetic peculiarities. For this review we analysed for antidepressants if drug concentration in blood can be used as biomarker for psychopharmacological treatment. Systematic review of the literature revealed for new and old antidepressant drugs that drug and metabolite concentrations in blood are measures of the pharmacokinetic phenotype and related differentially to occupancy of primary target structures, therapeutic effects and unwanted anticholinergic, cardiac and other side effects. Drug concentration in blood can therefore be used as biomarker in clinical practice to guide psychopharmacological treatment with established antidepressant drugs. Monitoring of drug concentration is suitable to improve efficacy and safety of the pharmacotherapy, especially in elderly patients who require complex pharmacological therapies.

Prolonged toxicity after amitriptyline overdose in a patient deficient in CYP2D6 activity

INTRODUCTION

Amitriptyline and its metabolite, nortriptyline, are metabolized, in part, by CYP2D6, a polymorphic enzyme. About 8% of Caucasians are deficient in CYP2D6 activity.

CASE REPORT

We present the case of a comatose woman who intentionally overdosed on amitriptyline and displayed rising serum total tricyclic antidepressant concentrations for at least 6 days after admission. Serial immunoassay total tricyclic antidepressant concentrations in our patient showed gradual decline beginning day 7, although the patient did not regain normal mental status until day 12. Genotyping revealed the patient to be homozygous for the CYP2D6*4 allele, the most common explanation of CYP2D6 enzymatic deficiency among Caucasians. Patients taking tricyclic antidepressants who are homozygous for CYP2D6*4 demonstrate >3 times concentration-time curve (AUCs) and prolonged elimination half-lives, especially of secondary amines such as nortriptyline.

DISCUSSION

We believe this is the first report describing toxicokinetics after tricyclic antidepressant overdose in a CYP2D6-deficient patient.

Intralipid outperforms sodium bicarbonate in a rabbit model of clomipramine toxicity

STUDY OBJECTIVE

Previous investigators have demonstrated amelioration of lipid-soluble drug toxidromes with infusion of lipid emulsions. Clomipramine is a lipid-soluble tricyclic antidepressant with significant cardiovascular depressant activity in human overdose. We compare resuscitation with Intralipid versus sodium bicarbonate in a rabbit model of clomipramine toxicity.

METHODS

Thirty sedated and mechanically ventilated New Zealand White rabbits were infused with clomipramine at 320 mg/kg per hour. At target mean arterial pressure of 50% initial mean arterial pressure, animals were rescued with 0.9% NaCl 12 mL/kg, 8.4% sodium bicarbonate 3 mL/kg, or 20% Intralipid 12 mL/kg. Pulse rate, mean arterial pressure, and QRS duration were sampled at 2.5-minute intervals to 15 minutes. In the second phase of the experiment, 8 sedated and mechanically ventilated rabbits were infused with clomipramine at 240 mg/kg per hour to a mean arterial pressure of 25 mm Hg. Animals received either 2 mL/kg 8.4% sodium bicarbonate or 8 mL/kg 20% Intralipid as rescue therapy. External cardiac compression and intravenous adrenaline were administered in the event of cardiovascular collapse.

RESULTS

Mean difference in mean arterial pressure between Intralipid- and saline solution-treated groups was 21.1 mm Hg (95% confidence interval [CI] 13.5 to 28.7 mm Hg) and 19.5 mm Hg (95% CI 10.5 to 28.9 mm Hg) at 5 and 15 minutes, respectively. Mean difference in mean arterial pressure between Intralipid- and bicarbonate-treated groups was 19.4 mm Hg (95% CI 18.8 to 27.0 mm Hg) and 11.5 mm Hg (95% CI 2.5 to 20.5 mm Hg) at 5 and 15 minutes. The rate of change in mean arterial pressure was greatest in the Intralipid-treated group at 3 minutes (6.2 mm Hg/min [95% CI 3.8 to 8.6 mm Hg/min] Intralipid versus -0.25 mm Hg/min [95% CI -1.9 to 1.4 mm Hg/min] saline solution) and 5 minutes (4.4 mm Hg/min [95% CI 3.0 to 5.9 mm Hg/min] Intralipid versus 0.06 mm Hg/min [95% CI -0.9 to 1.1 mm Hg/min] saline solution). In the second phase of the experiment spontaneous circulation was maintained in all Intralipid-treated rabbits (n=4). All animals in the bicarbonate-treated group developed pulseless electrical activity and proved refractory to resuscitation at 10 minutes (n=4, P=.023).

CONCLUSION

In this rabbit model, Intralipid infusion resulted in more rapid and complete reversal of clomipramine-induced hypotension compared with sodium bicarbonate. Additionally, Intralipid infusion prevented cardiovascular collapse in a model of severe clomipramine toxicity.

The AGNP-TDM Expert Group Consensus Guidelines: focus on therapeutic monitoring of antidepressants

Therapeutic drug monitoring (TDM) of psychotropic drugs such as antidepressants has been widely introduced for optimization of pharmacotherapy in psychiatric patients. The interdisciplinary TDM group of the Arbeitsgemeinschaft für Neuropsychopharmakologie und Pharmakopsychiatrie (AGNP) has worked out consensus guidelines with the aim of providing psychiatrists and TDM laboratories with a tool to optimize the use of TDM. Five research-based levels of recommendation were defined with regard to routine monitoring of drug plasma concentrations: (i) strongly recommended; (ii) recommended; (iii) useful; (iv) probably useful; and (v) not recommended. In addition, a list of indications that justify the use of TDM is presented, eg, control of compliance, lack of clinical response or adverse effects at recommended doses, drug interactions, pharmacovigilance programs, presence of a genetic particularity concerning drug metabolism, and children, adolescents, and elderly patients. For some drugs, studies on therapeutic ranges are lacking, but target ranges for clinically relevant plasma concentrations are presented for most drugs, based on pharmacokinetic studies reported in the literature. For many antidepressants, a thorough analysis of the literature on studies dealing with the plasma concentration-clinical effectiveness relationship allowed inclusion of therapeutic ranges of plasma concentrations. In addition, recommendations are made with regard to the combination of pharmacogenetic (phenotyping or genotyping) tests with TDM. Finally, practical instructions are given for the laboratory practitioners and the treating physicians how to use TDM: preparation of TDM, drug analysis, reporting and interpretation of results, and adequate use of information for patient treatment TDM is a complex process that needs optimal interdisciplinary coordination of a procedure implicating patients, treating physicians, clinical pharmacologists, and clinical laboratory specialists. These consensus guidelines should be helpful for optimizing TDM of antidepressants.

TDM-Based Imipramine Treatment in Neuropathic Pain

Tricyclic antidepressants (TCA) are the best-documented treatment of neuropathic pain. TCAs have a pronounced interindividual pharmacokinetic variability and a narrow therapeutic index. The aim of this study was to characterize the plasma concentration-effect relationship of imipramine in neuropathic pain and to determine the usefulness of therapeutic drug monitoring (TDM) of TCA treatment in a population with noncancer chronic pain. To do this, 83 patients with chronic noncancer neuropathic pain were included. Information on previous use of TCA was collected, and patients were tested for the presence of hyperalgesia. Pain intensity and pain relief were recorded, and the Short Form McGill Pain Questionnaire and Major Depression Inventory were completed before and during a TDM-based imipramine treatment. Imipramine dose was increased in steps of 25 mg/d every second week, and blood samples were taken at every dose. Endpoints were best possible pain relief, unacceptable side effects, or insufficient pain relief despite plasma drug level > 500 nmol/L. Dose range used was 10-300 mg/d. The study showed that imipramine 75 mg/d caused a 36-fold interindividual variation in steady-state plasma drug concentrations. In 46 responders (global pain relief > 25%) the plasma drug concentration at which an individual maximal analgesic effect was obtained ranged from 50 to 1400 nmol/L, but for the majority it was below 400 nmol/L. The concentration-effect relationship was similar for patients with central versus peripheral neuropathic pain and independent of the presence of hyperalgesia. Previous treatment failure with non-TDM TCA treatment was not a predictor of poor response to TDM-based treatment. In conclusion, there is a pronounced interindividual variability in concentration-effect relationship for imipramine treatment in neuropathic pain, but the majority of patients obtain a maximal analgesic effect at drug levels below 400 nmol/L. The concentration-effect relationship is similar for patients with central and peripheral neuropathic pain. Further studies are needed to document if TDM improves pain relief; however, TDM reduces the risk for toxicity.

Treatment of severe tricyclic antidepressant overdose with extracorporeal sorbent detoxification

Tricyclic overdose can be a medical emergency, and therapy with intravenous bicarbonate is not always successful in preventing cardiac toxicity or coma. Mortality in patients developing these complications is from 1% to 15%. Extracorporeal detoxification with sorbents has been used in treatment of patients with very high drug levels and declining clinical condition. Ten patients with serious drug overdose caused by tricyclics failed to respond quickly to standard therapy and were in stage 3-4 encephalopathy. Nine of these patients were on respirator support, 5 had hypotension, and 6 had QRS widening. Average level was 1,423 microg/L at presentation. Enteral activated charcoal and intravenous (IV) bicarbonate were initiated in the emergency room. The patients were treated for 3 to 4 hours with the Liver Dialysis Unit, a hemodiabsorption device using a cellulosic plate dialyzer and sorbent suspension as dialysate. Inflow and outflow blood levels indicated that the hemodetoxifier removed modest amounts of the tricyclics, metabolites, and other consumed drugs. The clinical improvement of the patients was dramatic, with patients reaching stage 0 or 1 encephalopathy during the treatment. Ventilator support was removed at the end of treatment for 3 patients who had not already developed pneumonia, and for others was prolonged up to 48 hours because of pneumonia, rather than mental status. Average length of stay in the intensive care unit (ICU) was 4.8 days (range 1 to 7 days). None of the patients died despite their high risk for ventricular arrhythmias, seizures, and death. Clinical improvement may have been attributable to removal of free drug from the blood or to removal of drug metabolites.

Fatal poisoning with the antidepressive agent opipramol

An ingestion of an unknown quantity ( < or = 3000 mg) of opipramol dihydrochloride in a suicide case is described. Quantitation of opipramol and its deshydroxyethyl metabolite was performed simultaneously after liquid-liquid extraction from alkalinized samples prior to HPLC analysis. Postmortem concentrations of opipramol and of the metabolite in body fluids and organs are given. The distribution pattern of opipramol in tissue was comparable to findings from animal studies. The results showed the high uptake of opipramol in parenchymal tissues resulting in a low blood concentration. A possibly low contribution of deshydroxyethyl opipramol to the antidepressive effect may be concluded as the concentration was considerably lower in brain compared to other organs.

Pharmacotherapy of the suicidal patient

The incidence of self-poisoning and suicide accomplished with tricyclic antidepressants (TCA) is unacceptably high. TCA are undeniably highly toxic, and overdosage occurs frequently. This review presents data that more than suggest that the compounds are prescribed for patients for whom TCA treatment is not indicated. Therefore, the physicians must learn to face up to the benefit-risk elements with these drugs. As TCA for many years to come probably remain the drug of choice in treatment of many depressed patients, the only way out of this dilemma is to focus on prevention. In any event, more cautious prescribing of TCA outside hospital is desirable.

Hexapropymate self-poisoning causes severe and long-lasting clinical symptoms

Cases of hexapropymate poisoning requiring intensive care in an urban region of Sweden (420,000 inhabitants) were collected over 2.5 years (1985 to 1987). Only patients with serum hexapropymate concentrations above 5.5 mg/L (30 mumol/L) and with a negative history for intake of tricyclic antidepressants, phenothiazines, barbiturates, antihistaminic drugs and opiates were included. Clinical data about 8 intoxication events in 6 patients were evaluated retrospectively. Initial symptoms included coma, hypotension, hypothermia, and hypoventilation. Maximum coma depth (Glasgow coma score) was 3 to 5 in 5 out of 8 events. On 7 occasions assisted ventilation was required (for 12 hours or more in 5 events). There was no relationship between serum concentrations of hexapropymate and severity of clinical symptoms. All patients survived. Detailed analysis of the drug elimination in one patient showed a terminal elimination half-life of 21 hours, which is longer than previously reported (5 hours). The indications for use of this hypnotic drug may vary between doctors since an 8-fold variation was seen in drug prescription between Swedish counties in 1987. Poisoning with hexapropymate is a serious condition which may require symptomatic treatment in the intensive care unit. The clinical picture is similar to that seen in patients with burbiturate intoxication. There is no role for active forced elimination of the drug. It is questionable whether the clinical value of the drug is outweighted by its toxicity.

The cardiovascular effects of antidepressants

This monograph comprises a review of the cardiovascular effects of the various types of antidepressant drugs in clinical use. The frequency, severity and clinical importance of these effects are placed in perspective. Most antidepressants can cause changes in blood pressure. Both the tricyclic type (TCA) and the monoamine oxidase inhibitors (MAOIs) can produce postural hypotension which may be dose-limiting. In addition, the MAOIs may be associated with severe hypertension when amine-containing foods or medicines are ingested. It is unlikely that therapeutic doses of any available antidepressant drug could impair cardiac contractility. Typical TCAs can cause abnormalities of cardiac conduction and arrhythmias, but this affects less than 5% of patients, mostly to a clinically insignificant extent. Newer compounds such as lofepramine, mianserin, trazodone and viloxazine seem safer in this respect. Reports of an association between therapeutic use of TCAs and sudden death are far from convincing. Overdosage with the MAOIs, lithium and carbamazepine is dangerous but not common; overdose with a TCA is a major source of morbidity and mortality. Lofepramine, mianserin and trazodone are relatively safe in overdose. The use of various antidepressants in patients with hypertension, cardiac failure, angina pectoris, myocardial infarction, or cardiac arrhythmias is discussed and guidelines suggested for the selection and use of antidepressant medication.

Emergency psychopharmacology: a review and update

The emergency area is a difficult site for the practice of careful psychopharmacology. The detailed history taking and collaborative treatment relationship often found in the consultant's office is rarely present. Instead, rapid decisions are made under a variety of pressures, with patients who may be uncooperative and unwilling to be treated. Many emergency patients are concurrently in treatment with a psychotherapist or psychopharmacologist who is available for consultation to the emergency service. When this is possible it is of great potential value, both in arriving more quickly at an appropriate treatment and in preventing a harmful opposition of the patient's treatment resources. For patients who are not currently in another treatment, additional helpful information often may be obtained from friends or family members. This is especially useful in treating patients who are acutely psychotic, unable to communicate, or uncooperative with the emergency interview. Although psychopharmacological approaches properly chosen often yield rapid results, medications are only a part of most emergency treatment plans. Medications are ineffective in a variety of crises and cannot replace careful interviewing of patients and others aimed at understanding the exogenous stress, change in interpersonal relationships, intrapsychic conflict, or biological disorder that precipitated an emergency visit. Furthermore, the decision to treat with medications must include a consideration of the adverse effects that may occur. With these limitations in mind, the area of emergency psychopharmacology can provide powerful assistance to emergency clinicians.

Poisoning due to tricyclic antidepressant overdosage. Clinical presentation and treatment

Tricyclic antidepressants are among the commonest causes of both non-fatal and fatal drug poisoning in the world. Their toxicity is due to effects on the brain, the heart, the respiratory system and the parasympathetic nervous system. Symptoms usually appear within 4 hours of an overdose and all but the most seriously poisoned patients recover within 24 hours. The most common clinical features are dry mouth, blurred vision, dilated pupils, sinus tachycardia, pyramidal neurological signs, and drowsiness. In severe poisoning, there may be coma, convulsions, respiratory depression, hypotension and a wide range of electrocardiographic (ECG) abnormalities. The most frequent findings on the ECG are prolongation of the PR and QT intervals; the tracing may resemble bundle branch block or supraventricular or ventricular tachycardias. Treatment of poisoning due to the tricyclic antidepressants is essentially supportive, there being insufficient evidence at present to recommend the use of methods to increase elimination of the drug from the body. Gastric aspiration and lavage should be performed if more than 750 mg of drug have been taken. There must be regular monitoring for hypoxia, acidosis and hypokalaemia and these complications should be corrected enthusiastically. Convulsions should be treated with diazepam or chlormethiazole. Muscular paralysis and artificial ventilation should be employed if anticonvulsants are ineffective. Hypotension should be treated firstly by fluid replacement and then with sympathomimetic agents (dopamine or dobutamine). Antiarrhythmic drugs should only be employed if there is evidence of circulatory failure which fails to respond to correction of hypotension. Sodium bicarbonate infusions should be given to cardiotoxic patients who are acidotic and are worth trying even if the patient is not acidotic. Although physostigmine salicylate will reverse most of the features of tricyclic antidepressant poisoning, its effects are short-lived in serious toxicity and it can produce dangerous side effects; physostigmine should therefore be reserved for those patients who have complications of coma or who have resistant cardiotoxicity or convulsions. Drug screening and quantitative determination of tricyclic antidepressant serum concentrations are useful in a minority of patients who have severe, unusual or prolonged symptoms.

The role of the laboratory in acute poisoning: case histories

We describe case histories and management of four patients poisoned with different compounds. The role of the laboratory, particularly of toxicological analysis, in the management of poisoned patients is examined.

Clinical course in acute self-poisonings: a prospective study of 1125 consecutively hospitalised adults

The clinical course in an unselected group of 1125 consecutively hospitalised self-poisonings was studied during 1 year in Oslo. Mortality was 0.5%, but only 0.3% in those admitted without cardiac and respiratory arrest. Mortality among those in grade IV coma was 4.2%. The deepest comas (grade III or IV) occurred in 25.1% of the admissions with a mean duration of the coma of 5.8 h (range 1-80). Complications occurred in 21.7% of the admissions and 6.9% suffered more than one complication of which the most frequent were respiratory depression (13.5%), hypotension (5.3%), pneumonia (4.4%), and hypothermia (1.6%). The complication rate was highest in poisonings with opiates (60.7%), meprobamate (37.5%) and antihistamines (30.0%). Arrhythmias and respiratory depression were closely associated with poisonings with antidepressants and opiates, respectively. Owing to frequent polydrug overdoses it was difficult to associate other complications with other main toxic agents. Administration of antidotes (20.6%), cuffed intubation (4.4%) and forced alkaline diuresis (3.4%) were the most frequent special therapeutic measures taken. The change in pattern of self-poisonings in Oslo focuses on antidote therapy and intensive care, especially outside hospital, but limits the need for haemodialysis and haemoperfusion which were performed in only 1.0% of the admissions.

Suppression of plasma cortisol after oral administration of oxazepam in man

In seven healthy male subjects plasma cortisol was measured at 20 min intervals from 07.00 to 13.00 h, during a drug free control test and after oral administration of 30 and 60 mg oxazepam. A significant suppression of the plasma cortisol levels for about 2 h after 30 mg oxazepam, and for about 3 h after 60 mg oxazepam was observed. The suppression wore off before the oxazepam plasma concentration started to decline significantly.

Comparative animal studies on cardiovascular toxicity of tri- and tetracyclic antidepressants and citalopram; relation to drug plasma levels

The aim of the present study was to compare cardiovascular and/or cardiotoxic effects of eight anti-depressants (imipramine, chlorimipramine, amitriptyline, nortriptyline , doxepin, maprotiline, mianserin and citalopram) in anaesthetized cats after oral dosing and in conscious rabbits after intravenous infusion. In the cats drug plasma levels were determined as well. When estimated from ECG recordings, citalopram and chlorimipramine in particular, but also mianserin, appeared less cardiotoxic than the other drugs tested. The cardiovascular effects seen in the cats were with few exceptions identical for all the drugs tested but not seen at the same dose (concentration). Safety margins were defined as minimal doses or plasma levels when ECG changes (conduction or rhythm) or cardiovascular effects (+/- 10% change of initial value in a series of parameters) occurred in experimental animals divided by maximal therapeutic dose or mean plasma levels in patients. From comparisons of the safety margins it is concluded that except for citalopram and mianserin (safety margins 80 and 18 respectively in cats and greater than 15 in rabbits) all the other drugs tested (safety margins less than or equal to 9) have a cardiotoxic potential. The probability that cardiovascular side effects may occur is less pronounced for citalopram (safety margins 10-32) than for all the other drugs tested (safety margins ranging from 0.1 to less than 5).