On August 24th, 2011, the American FDA had issued a safety warning, in which they pointed out that there is a risk of prolongation of the QT interval associated with the use of citalopram. Since the risk would increase dose-dependently, it was recommended not to administer citalopram daily doses above 40 mg. In patients aged > 65 years, the dose should not exceed 20 mg daily. On March 28th, 2012, the FDA published an update to this safety warning.In Germany the company Lundbeck together with the German Federal Institute for Drugs and Medical Devices (BfArM) had published a so-called “Red Hand”-letter on October 31st, 2011, with an almost identical warning. In a second Red Hand letter from December 5th, 2011, Lundbeck also warned of a possible QTc prolongation associated with use of escitalopram, although the extent of the possible QTc prolongation with escitalopram was only about half the size of citalopram. The German Psychiatric Association (Deutsche Gesellschaft für Psychiatrie und Psychotherapie, Psychosomatik und Nervenheilkunde, DGPPN) published a comment on February 6th, 2012, written under my auspices, which warned against an overreaction to the Red Hand letters: “The DGPPN advises a rational approach to the two Red Hand letters on citalopram and escitalopram. A waiver of the substances would be an overreaction. Citalopram and escitalopram are safe and well-tolerated drugs. This also applies to intoxications. Other SSRIs such as fluoxetine, paroxetine, and sertraline pose similar risks. A prudent and cautious handling of citalopram and escitalopram in certain patient groups (impaired renal function, advanced age, combined with CYP inhibitors) had already been advised before the appearance of the Red Hand letters. The fact that Red Hand letters do not exist for other substances does not make these compounds any more secure (e.g., relatively high interaction potential of fluoxetine or paroxetine) and does not release prescribing physicians of particular care in handling these substances, especially in the above groups of patients.” (DGPPN Comment of Feb 6th, 2012).
The discussion is now being fueled again by a publication titled “Evaluation of the FDA Warning Against Prescribing Citalopram at Doses Exceeding 40 mg” in the May issue of the American Journal of Psychiatry (Zivin et al., Am J Psychiatry, online first). The authors analyzed data from the Veterans Health Administration (VHA) of all patients with depression in the years between 2004 (the year citalopram became available generically) and 2009 who had received at least one prescription of citalopram. In addition, the data for patients treated with sertraline were considered as reference data, because sertraline was not affected by the warnings. Patient numbers were enormous: 618,450 patients had received citalopram, 365,898 sertraline. According to the authors, these are the largest groups of patients that were ever investigated. They examined associations between the two doses of antidepressants on the one hand and ventricular arrhythmias, cardiac, non- cardiac, and all-cause mortality on the other hand. This was made possible by connection of the data from the VHA National Registry for Depression (NARDEP) and data from the National Death Index. Approximately 90% of patients were male due to the nature of the cohort. The median age was 57 years. 36% of patients in both groups had a risk or a potential risk of torsade de pointes.
6,754 patients, or 1.1%, who had received citalopram in the specified 5-year period, had ventricular arrhythmias. In the patients treated with sertraline, the rate was the the same: 4,198 (= 1.1%) had ventricular arrhythmias. 64,970 (= 10.5%) of patients treated with citalopram died during the observation period, with 20,509 (= 3.3%) patients with a cardiac cause identified. Again, the numbers were the same inpatients treated with sertraline: 45,768 (= 12.5%) of the patients died, 14,808 (= 4.0%) due to a cardiac cause.
Particularly interesting are the results when cardiac events and mortality rates were related to the administered dosages. The incidence of ventricular arrhythmias and all observed mortality rates decreased with increasing doses of both antidepressants! 18.6% of all patients received citalopram doses > 40 mg/day. 44.6% of all ventricular arrhythmias were noted at doses of 1-40 mg/day, 6.1% at doses > 40 mg/day. 29.6% of deaths from all causes were noted at doses of 1-40 mg/day, but only 3.9% at > 40 mg/day. 31.6% of the deaths with cardiac causes were observed at doses of 1-40 mg/day, but only 4.1% at doses > 40 mg/day. 28.7% of deaths with non-cardiac causes were noted at doses of 1-40 mg/day, 3.8% at > 40 mg/day. Very similar results were found for sertraline when doses of 1-100 mg/day were compared with those of > 100 mg/day.
When the authors controlled for demographic and clinical covariates, they found that citalopram doses of > 40 mg/day were associated with a statistically significantly lower risk of ventricular arrhythmias, all-cause mortality and non-cardiac mortality compared to dosages of 1-20 mg/day. Cardiac mortality was not increased. Doses of 21-40 mg/day were associated with a significantly lower risk of ventricular arrhythmias than doses 1-20 mg / day. The analyzes for sertraline led to similar results.
Zivin et al. point out that their study is no proof for the harmlessness of citalopram (or sertraline), because they did not include a group that received no medication as a control group. However, higher doses – of which the FDA and the BfArM had warned – are associated with lower rather than higher risks.On the basis of their results, the authors question the merit of the FDA warning and ask the question whether “the warning itself will cause more harm than good”. Nevertheless, clinicians and patients must continue to make the decision whether they adhere to the warnings of the authorities or accept potentially worse treatment with lower doses of citalopram.
The authors do not comment on escitalopram, since there was no warning in this respect in the USA. As mentioned above, the observed prolongation of the QTc interval was only half as pronounced with escitalopram as with citalopram. The risks are therefore likely to be even lesser.
A PDF of the article can be ordered from me.
This post is also available in: German