arznei-telegramm 2009; 40: 49



The neuroleptic sertindole (SERDOLECT) was taken off the market in 1998 shortly after being launched due to severe cardiac effects including fatalities. Even so, in 2005, it was re-licensed by the European Medicines Agency (EMEA) as a second line therapy (a-t 2006; 37: 72-3). There has never been given a scientific justification for this. The large safety study required by the EMEA (SCoP*) has been over since February 2008, but the results have not yet been published. Now specific data from this study and other studies on the cardiac risk and mortality associated with sertindole are available through the American drug regulatory agency, FDA (1). Based on these results FDA's advisory committee has spoken out against the widespread use of this drug, which is not approved for use in schizophrenia in the United States.

According to studies with different doses of sertindole available to the American agency for licensing, a total of 1.3% of patients have a QTc value over 500 milliseconds (msec). With daily doses of 20 mg sertindole-the maximum dose allowed in Germany-1.9% of users had this QTc prolongation. This is three times higher than with ziprasidone, which is similarly known as markedly cardiotoxic (ZELDOX; a-t 2002; 33: 48-9). On average, the QTc time with 20 mg sertindole was almost 27 msec longer than with a placebo (1.2).

The open-label SCoP study also suggests a serious arrhythmogenic potential for sertindole. It included nearly 10,000 patients with schizophrenia from 38 countries mostly in Eastern Europe and Asia and from nearly 600 facilities, including private practices and peripheral hospitals. The patients were selected "naturalistically," that is, not according to strict diagnostic criteria using standard classification systems (1-4). According to the European licensing requirements, a first-line treatment must have failed before the patients were randomly assigned to either the sertindole or risperidone group (RISPERDAL, generics). Primary endpoints were all-cause mortality and hospitalization for cardiac events including arrhythmia (1-4). Secondarily, fatalities were documented according to cause of death and-due to a protocol change suggested by FDA after just under 2,000 participants had been enrolled-suicide and attempted suicide, as well. In agreement with the EMEA, but not FDA, which emphasizes its disassociation from this, it was defined in the SCoP study itself that sertindole is regarded as not inferior to risperidone, if it comes off 50% worse in terms of mortality (1.3).

The observation period is shorter for sertindole users as a whole (6,600 patient years) than for risperidone users (7,600 patient years). The difference is apparently due to more patients dropping out due to problematic side effects, particularly due to asymptomatic, but suspicious ECG readings. So patients are excluded even with QTc values under 500 msec. But since regular ECG controls are only performed in the sertindole group and not in the risperidone group (1), it is assumed there is bias there.

Despite the shorter observation period by one thousand patient years and the presumed risk minimization for sertindole because of the ECG controls, the all-cause mortality with sertindole (64 deaths) was numerically higher than for risperidone (n = 61). With an upper limit for the confidence interval (CI) of 1.6, even the high non-inferiority limit of 50% was exceeded according to the FDA's analysis (1.2).

There are three times as many fatalities with sertindole (31) than with risperidone (12). Even the number of sudden fatalities, which are considered the most relevant safety endpoint by FDA, is five times higher with sertindole (13 vs. 3; hazard ratio 5.1; 95% CI 1.5-17.9) (1.2).

Although the life-threatening arrhythmia Torsade de Pointes is not usually discovered in studies - the patients usually die in acute situations as sudden cardiac death - this condition was diagnosed in a total of three patients taking sertindole in the clinical study program and in the SCoP study (1). Since the last ECG for each of the patients who died of sudden cardiac death was unremarkable according the FDA's assessment (1), regular ECG controls do not appear to be able to prevent fatalities. In addition to long-term risk factors such as female sex, slow metabolisers and QT syndrome, acute risks such as liver damage, cardiac insufficiency, electrolyte imbalance and interactions with new concomitant medication also need to be taken into account over the course of treatment.

With regard to suicides and suicide attempts, sertindole and risperidone are not significantly different (47 vs. 63 events, p = 0.258) (1.2).

  The neuroleptic sertindole (SERDOLECT), which was taken off the market in 1998 because of early serious risk signals of cardiotoxicity, was re-licensed in Europe by the European Medicines Agency (EMEA) in 2005, without giving any scientific explanation.

  Publication of the SCoP safety study required by the EMEA is long overdue.

  Sertindole has a high risk of cardiotoxicity according to data available to the United States drug regulatory agency, FDA: With daily doses of 20 mg QTc prolongation over 500 msec occurs in 1.9% of patients and, according to the results of the SCoP study, compared to risperidone (RISPERDAL, generics) sudden death five times more frequently.

  We strongly advise against the use of this drug.



FDA: Meeting of the Psychopharmacologic Drugs Advisory Committee, 7 April 2009



LAUGHREN, T.P. (FDA Memorandum) from 10 March 2009



PEUSKENS, J. et al.: Pharmacoepidemiol. Drug Saf. 2008; 17: 425-33


4, SCoP-Studie, registriert 5. März 2009;



SCoP = Sertindole Cohort Prospective Study

© arznei-telegramm 5/09