Can citalopram cause QT-prolongation? If so, how should the patient´s drug treatment be adjusted? W

Fråga: Can citalopram cause QT-prolongation? If so, how should the patient´s drug treatment be adjusted? What other antidepressants could be recommended? A 59-year-old man with hypertension and unipolar depression is being treated with citalopram 60 mg, lithium 129 mg, propiomazin 25 mg, risperidone 1.0 mg, mianserin 60 mg, metoprolol 100 mg, bendroflumethiazide 2.5 mg and zopiclone 7.5 mg. Before electroconvulsive therapy (ECT), an ECG reveals a prolonged cQT-time of 632 ms (normal value < 450 ms for men). S-creatinine and other laboratory tests, including liver tests, are normal. S-lithium concentration is 0.69 mmol/L. Blood pressure is 150/95. Upon dose reduction of citalopram to 40 mg daily, the QTc-time decreases to 582 ms. The dose of citalopram is further reduced to 20 mg daily and ECT is completed successfully.

Sammanfattning: Citalopram has been reported to cause QT-prolongation and torsades de pointes. Risperidone may in this case contribute to QT-prolongation. We recommend analysis of this patient´s plasma concentrations of citalopram and of risperidone. The patient may tolerate a lower dose of citalopram without effects on QT-time. The dose of mianserin may be increased under observation.

Svar: The mechanism for prolongation of QT-interval is not clear. However, it is known that drugs, which prolong the QT-interval, and cause torsades de pointes, also block a major repolarizing potassium current, termed Ikr, regulated by a potassium channel protein (1). Reduced Ikr prolong action potentials in individual cells, similarly to what is found in the congenital long-QT-syndrome. There is no linear relationship between QT-interval and the risk of torsades de pointes, but a QT-interval of more than 500 ms is commonly regarded as involving an increased risk. It is believed that multiple factors contribute to the development of torsades de pointes, including high drug concentrations, two or more drugs affecting the same pathway, hypokalemia, hypomagnesemia, bradycardia, heart failure, high age, female sex and probably genetic disposition.

Prolongation of the QT-interval during treatment with the selective serotonin reuptake inhibitor (SSRI) citalopram has been reported (2). The Swedish adverse drug reaction register has four cases of QT-prolongation reported and 6 cases of torsade de pointes (out of a total of 893 reports), some of which are cited below.

An 88-year-old lady treated with citalopram 10 mg daily developed VT attacks type torsades de pointes and was defibrillated 8 times. There was a prolonged QT-time, 630 ms. After citalopram was discontinued there were no more VT attacks and the QT-time was 400 ms when the patient was released from hospital 17 days later (2).

There are two cases of suicidal overdose. One 45-year-old patient took 840 mg of citalopram and vomited. She had QT-intervals of 490-504-458 ms. The other patient (27-year-old) took between 200 and 400 mg. Her QT-time was 509 ms on admission and 492 ms after two days and no persistent arrhythmias were recorded. A 60-year-old man had probably taken higher doses than prescribed (20 mg daily) and developed prolonged QT-time. Plasma concentration of citalopram was 735 nmol/L (range 100-150 nmol/L). Citalopram was stopped and plasma concentrations declined to 110 and later <20 nmol/L. ECG was normal after 4 weeks (2).

An earlier Drugline document cites a case report concerning a 70-year-old woman, who was admitted to hospital with a documented episode of torsades de pointes (3). Upon arrival this was resolved, but a prolonged QT-interval was present. She had been treated with citalopram and mianserin for two years. One year earlier, she had had an attack of seizures and prolonged QT-interval. Upon discontinuation of citalopram and mianserin, the prolongation of the QT-interval regressed.

The present patient is treated with several drugs, and is thus subject to an increased risk of both pharmacokinetic and pharmacodynamic interactions. E.g. citalopram, risperidone, mianserin and metoprolol are all metabolised by the same liver enzyme, CYP2D6, and even though none of these drugs is known to be a strong enzyme inhibitor, an interaction leading to increased drug concentrations cannot be excluded. Plasma concentrations of citalopram and risperidone should be monitored in this patient. A dynamic interaction could occur between citalopram and risperidone, which has been reported in international literature to cause QT-prolongation of 10 ms (4). For risperidon, there are no reports of torsades de pointes (4). There is one report each of tachycardia, cardiac asystole, bradycardia and cardiac failure, but none of QT-prolongation in the Swedish adverse drug reaction register (2).

For mianserin, there are no reports of QT-prolongation (2). There are four cases of palpitations reported to the Swedish adverse drug reaction register, one of tachycardia, one of bradycardia and one of sick sinus syndrome. Mianserin did not affect the QT-interval in a comparative study with placebo and tricyclic antidepressants (5). Some authors suggest that mianserin is the antidepressant drug of choice in patients with underlying cardiac disease. The effective dose is reported to be 30-90 mg daily, usually 60 mg daily. In this case, the dose may be tentatively increased.

Lithium may also affect ECG. However, there are no reports of QT-prolongation (2). One fourth of patients treated with lithium have low or inverted T wave (6). The following cardiac effects are listed as less common (<1/100-1/1000): altered sinus activity, AV-block, premature ventricle contraction, paroxysmal left ventricular bundle branch block and "arrhythmias". In this patient, plasma concentration of lithium is within the therapeutic interval and a dose increase is not advisable.

As for alternative antidepressant treatment, tricyclic antidepressants have, in general, a greater potential to cause cardiac adverse effects than do SSRIs, and would not offer any advantage (7). The following drugs have been reported to increase QT-time: amitriptyline, nortriptyline, imipramine and klomipramine (4). Fluoxetine, an SSRI, was studied in comparison with doxepin (a tricyclic antidepressant) in 39 patients (8). After 6 weeks of treatment with a mean dose of 24 mg fluoxetine there were no effects on the QT-time or QRS duration, while doxepin treated patients had a small increase in QT-time.

For venlafaxine, there have been no reports of QT-prolongation to the Swedish adverse drug reaction register (2). (There are 13 cases of hypertension, four of hypotension, one of tachycardia and two of palpitations out of a total of 29 reports). However, venlafaxine has been in use for a shorter time than citalopram. Venlafaxine tends to increase blood pressure, and because of these two facts, it may not be a very good alternative.

We recommend that this case is reported to the regional adverse drug reaction monitoring centre.