Wyeth Pharmaceuticals is heavily promoting Effexor XR for the treatment of depression and Generalized Anxiety Disorder (GAD), and the gist of the promotion is that Effexor XR is more powerful at producing remission than the SSRIs. How believable is this new data? Even if it is believable, how should this influence our prescribing decisions?

As practitioners, we have all been bombarded by the Effexor promotional literature, with slogans such as: THE GOAL IS REMISSION OF SYMPTOMS and LETS GET IT RIGHT THE FIRST TIME. These are hardly news flashes to psychiatrists, and the implication that we dont already try our darndest to get our patients all the way well is mildly insulting.

Of course, we all know where Wyeth is coming from here. They are trying to convince us to think in terms of remission (Ham-D depression scale 7 or less) rather than response (Ham-D 50% improved) because Effexor XR claims to be better than the SSRIs at bringing about remission, but not at producing response.

Lets look at the basis for Wyeths claim, which is the now famous study published in the British Journal of Psychiatry in 2001 by Thase et al (1) showing that Effexors remission rates were 45% (vs. 35% for SSRIs and 25% for placebo). The study had excellent methodology, comprising a pooled analysis of all the data the company had collected comparing Effexor with Prozac, Paxil, and Luvox. By psychiatric research standards, the numbers were huge: 851 patients in the Effexor XR group, 748 in the SSRI group (Prozac, Paxil, and Luvox), and 446 in the placebo group. The doses of comparator SSRIs were robust enough to mirror what we actually use in clinical practice, and treatment duration was reasonable, between 6 and 8 weeks. There are a few minor quibbles, including the fact that there was no comparison with either Celexa or Zoloft, but overall the study was solid, and the higher remission rate with Effexor seems pretty convincing, if not absolutely compelling.

Nonetheless, one might ask how meaningfulin the real world of patient careis this 45% to 35% difference in remission rates? The figures mean that if you were to put 10 patients on Effexor XR instead of putting those 10 on an SSRI, Effexor would bring one extra patient to remission vs. the SSRI. One out of ten certainly isnt anything to sniff at. But wouldnt that one non-remitted SSRI-treated patient eventually make it to remission if continued on the SSRI for a few weeks beyond the 8 weeks reported in the article? Without continuation data, it is impossible to say.

Ultimately, Wyeth is asking us to prescribe Effexor as a first line agent, as readily as we now prescribe the SSRIs. Recall that Effexor is a Serotonin Norepinephrine Reuptake Inhibitor (SNRI), and presumably any advantage in efficacy is related to the fact that at higher doses (beginning at around 150 mg.) the medication increases levels of norepinephrine as well as serotonin. This is similar to the action of the old tricyclic clomipramine, and there is in fact evidence that clomipramine has an advantage over SSRIs in efficacy.

So why not prescribe Effexor over the SSRIs, given this fairly convincing data of greater efficacy? Alas for Wyeth, Effexor has three things going against it: 1) A perceived risk of hypertension; 2) A nasty discontinuation reaction; 3) No FDA indication for panic disorder, social anxiety disorder, or OCD. A word on each of these issues in turn.

1) Blood pressure. This is probably less of a problem than most prescribers think. The 1998 paper entitled Effects of Venlafaxine on Blood Pressure: A meta-analysis of original data from 3744 depressed patients (2) shows that as long as you dont go above 300 mg per day, the rate of sustained diastolic hypertension was no higher than the rate seem with placebo (2.9% on Effexor vs. 2.3% on placebo). The PDR insert reports additional data, showing a 0.5% rate of hypertension for dosages up to 225 mg. Thus at doses below 300 mg., hypertension does not appear to be a significant problem with Effexor. My strategy is to inform my patients that there is a very small risk of blood pressure changes and I ask them to get their BP checked by their PCP at some point after we achieve our final dose. Some of my colleagues check their own patients BPs, which is probably not necessary but is a nice touch.

2) Discontinuation reactions. According to a brief report from Mass General (3), 7 of 9 patients discontinued from Effexor XR had discontinuation reactions, vs. only 2 of 9 placebo treated patients. Of course, the severity of discontinuation reactions varies, from barely perceptible light-headedness to severe vertigo, nausea, insomnia, and tearfulness. A psychopharmacologic pearl is to try to minimize this reaction by using the long-acting SSRI Prozac to taper patients off the medication, but there are no studies endorsing this approach. Of course, Effexor is not the only antidepressant to cause withdrawal effects: Paxil is infamous for this drawback as well.

3) Indicationo-penia. The nice thing about SSRIs is that one member of the club can often claim the advantages of another member. Thus, Paxil, which is the indication-hog with no less than 6 FDA-approvals and growing, has provided an indication halo over the other SSRIs, but not over Effexor, an SNRI. Effexor is indicated for depression and GAD, and rumor has it that social anxiety may be awarded soon. One double blind study of Effexor for panic disorder had pretty positive results (4), and probably suffered from too small a sample size. One double blind study of OCD (5) showed no differences between Effexor and placebo, but a recent single blind study showed it to be as effective as clomipramine (6). Bottom-line: Effexor has a robust spectrum of anti-anxiety activity, though probably not as broad as the SSRIs.

So, should we be using Effexor XR first-line? For patients who suffer from the disorders for which it is approved (depression and GAD), it would be hard to argue against its first-line use. TCR would be even more enthusiastic about Effexor if its higher remission rate can be shown to hold up over the long term, like 6 months to a year.

TCR VERDICT:We like it, and we want more data!