Venlafaxine and cardiovascular toxicity

People with depression have a higher incidence of cardiovascular disease and mortality from cardiovascular disease than people without depression.1 Ideally, the use of antidepressants should not add to this risk. Depression may also provoke suicide, so toxicity in overdose is a crucial consideration when choosing an antidepressant.

Older tricyclic antidepressants may increase mortality from cardiovascular disease when used therapeutically and are often fatally toxic in overdose, whereas selective serotonin reuptake inhibitors (SSRIs) show little if any cardiotoxicity at any dose, although, their antiplatelet effect has potential for harm.2 The serotonin-noradrenaline reuptake inhibitor venlafaxine has been thought to show cardiovascular toxicity somewhere between that of tricyclics and SSRIs, so its use has been restricted in some countries. In the linked case-control study (doi:10.1136/bmj.c249), however, Martinez and colleagues challenge the notion that venlafaxine is more cardiotoxic than SSRIs.3

In their study using data from the general practice research database in the United Kingdom, which looked at 207 384 first time users of antidepressants, venlafaxine was not associated with an increased risk of sudden cardiac death or near-death compared with fluoxetine, citalopram, or (generally low dose) dosulepin. Risk was lower for venlafaxine than for other drugs examined, and confidence intervals effectively excluded any clinically important increased risk for venlafaxine. If considered in isolation, this study strongly suggests that the therapeutic use of venlafaxine is as safe as use of the two SSRIs and low dose dosulepin. Because SSRIs are considered to have minimal cardiovascular toxicity, the same might now be assumed for venlafaxine.

Evidence suggesting that venlafaxine is cardiotoxic is limited and is derived from diverse sources. Two studies in animals have shown that venlafaxine can inhibit cardiac ion channels,4 5 but the concentrations of venlafaxine associated with inhibition were much greater than those seen in humans taking therapeutic doses. Studies of overdoses in humans suggest that venlafaxine has a much higher fatal toxicity index than SSRIs,6 and that prolongation of QRS and QTc intervals (electrocardiographic indicators of inhibition of potassium and sodium channels) are often seen in overdose.7 8 Therapeutic doses of venlafaxine also increase blood pressure to a small extent.2

Against this is a substantial body of evidence suggesting that venlafaxine has minimal arrhythmogenicity even in overdose. In an analysis of 5510 overdoses where venlafaxine alone was ingested, only 12 (0.22%) were fatal, and cardiac conduction defects were seen in fewer than 2% of all cases.9 Even in the two studies noting changes on electrocardiography in venlafaxine overdoses, serious arrhythmias did not occur in one (51 overdoses observed) and were rare (3/235 overdoses), benign, and transient in the other.7 8 In addition, the drug’s relatively high fatal toxicity index may be related to non-cardiac toxicity (seizures, for example7) or its prescription to people who are more likely to attempt suicide.10 Moreover, the fatal toxicity index is probably a poor measure of inherent toxicity—the index has been reported to be as low as 5.5 and as high as 53.6 for the antidepressant nortriptyline, effectively making it both one of the least toxic and most toxic antidepressants available.11 Lastly, as discussed by Martinez and colleagues,3 formal premarketing and postmarketing observations of venlafaxine did not suggest a significant potential for cardiotoxicity.

Regulatory authorities, particularly in the UK, have been inconsistent with their advice on the safe prescribing of venlafaxine. Its use was at first severely restricted in the UK because of “concerns about cardiotoxicity and toxicity in overdose,” although many (but not all) of the original restrictions were later lifted after the Medicines and Healthcare Products Regulatory Agency (MHRA) conceded that “available data do not point towards an increased risk of cardiotoxicity for venlafaxine versus SSRIs.”12 A key aspect to regulatory deliberations seems to have been the high fatality rate in venlafaxine overdoses, with the MHRA citing the case fatality rate of single overdoses in the UK adverse drug reactions online information tracking (ADROIT) database to be around 27%.12 This fatality rate is very different from that described in more robust observational studies, and it probably represents extensive reporting bias, as the MHRA seemed ultimately to conclude.

So, therapeutic use of venlafaxine seems not to increase risk of sudden death or near death compared with SSRIs, and venlafaxine causes arrhythmia only rarely in overdose. Its toxicity in overdose may be higher than that for SSRIs, but this is not clearly substantiated and toxicity in overdose is not linked to any cardiovascular effects of the drug. Evidence suggesting any arrhythmogenic potential for venlafaxine is weak, especially where normal therapeutic use is concerned. Future research should centre on meaningful outcomes (arrhythmia, cardiac mortality) in humans, rather than surrogates of toxicity in animals. For now, continued restriction on the use of venlafaxine on the grounds of cardiovascular toxicity seems inappropriate.