In 2011 in the pages of TCPR, we asked, “What’s new in antidepressant treatment?” The answer was “not much” (TCPR, April 2011). In 2012, unfortunately, the answer isn’t very different.
Given the high cost ($1.8 billion) and long time (almost nine years, on average) to bring a new antidepressant to market, drug companies are putting fewer eggs into more selective baskets.
As you know, fluoxetine (Prozac) entered the U.S. market in 1987, setting off the introduction of a cascade of SSRIs—and, eventually, SNRIs—which have, collectively, become some of the most prescribed drugs in the United States (Olfson M and Marcus SC, Arch Gen Psychiatry 2009;66:848–856). We have seen six different SSRIs and three different SNRIs approved for major depression and a smattering of other indications. The final SSRI to enter the market, escitalopram (Lexapro), lost its patent and went generic in March of 2012. Duloxetine (Cymbalta), one of the SNRIs, will also lose its patent in 2013.
That leaves only a few new name-brand antidepressants left—and a more limited selection of free samples, too, for those of you who still use them. For the most part, these remaining name-brand medications are “me-too” drugs and patent extenders, like desvenlafaxine (Pristiq, the active metabolite of venlafaxine) and extended-release trazodone (Oleptro). Vilazodone (Viibryd), from Forest Labs, the people who brought you citalopram (Celexa) and escitalopram, could best be described as a combination of buspirone (BuSpar) and an SSRI. Drug companies are now turning their attention to augmenting agents, like the atypical antipsychotics aripiprazole (Abilify) and quetiapine (Seroquel XR) (and others in clinical trials), but concerns about metabolic safety—not to mention cost— make these less attractive as firstline treatments.
Despite the panoply of options, treating depression remains more of an “art” than a science, as there are still no 100% reliable ways to predict a patient’s response to a given antidepressant, and no way to determine the molecular basis of an individual patient’s symptoms. Some psychiatrists have their own “favorite” antidepressants and use them regularly, switching or augmenting as needed. Some prefer to mix drugs from different classes to create combinations with novel mechanisms. This latter approach is not supported by the evidence. For example, a recent randomized trial showed that neither the combination of bupropion (Wellbutrin) plus escitalopram, nor mirtazapine (Remeron) plus venlafaxine (Effexor) provided any benefit over escitalopram alone (Rush AJ et al, Am J Psychiatry 2011;168(7):689–701).
Since most of the talk about new medication options focuses on the neurotransmitters involved, we will attempt to summarize the state of current antidepressants under development by separating them into groups based on the specific pathways they affect. Hopefully this gives a better sense of what may lie ahead in the antidepressant pipeline.
The melatonin system seems like a reasonable target for antidepressants, since disturbances of circadian rhythm may cause depressed mood, fatigue, low energy, and poor concentration. Agomelatine (Valdoxan), a melatonin analogue approved in Europe, is similar to ramelteon (Rozerem), which was approved by the FDA in 2005 for insomnia characterized by difficulty with sleep onset. Both are melatonin (MT1 and MT2) receptor agonists, helping to promote sleep. Agomelatine is also a 5-HT2B and 5-HT2C receptor antagonist; its 5-HT2C antagonism hypothetically increases dopamine and norepinephrine release in the frontal cortex.
Does it work? Not very well. In trials comparing agomelatine to placebo or SSRIs, agomelatine was just about as effective as venlafaxine, fluoxetine, and sertraline—but not in all trials—and may actually be worse than paroxetine (Paxil). Furthermore, its efficacy relative to placebo was only, on average, a few HAM-D points (Hickie IB and Rogers NL, Lancet 2011;378:621–631). These observations, together with the finding of potential hepatotoxicity at higher doses, might explain why Novartis halted the development of agomelatine for the U.S. market in October of 2011. It’s not clear whether it will ever become available here.
One well-known disadvantage of current antidepressants is their slow onset of action. A series of trials at the NIMH (which have received much popular attention in the last several months) has shown that ketamine, an antagonist of the N-methyl-D-aspartic acid (NMDA) subtype of the glutamate receptor, can have a powerful and rapid antidepressant effect. And yes, it’s that ketamine, also used as an anesthetic/ sedative by veterinarians and pediatric anesthesiologists and abused on the streets as “Special K.” In one NIH-funded study of 18 patients with treatment-resistant depression who received an infusion (0.5 mg/kg) of ketamine, a 50% reduction in HAM-D score (“response”) was seen in half the patients within two hours. Within 24 hours, 71% met criteria for response and 29% for remission, and 35% of subjects maintained their response one week later (Zarate C et al, Arch Gen Psychiatry 2006;63:856–864).
Basic research shows that ketamine raises intracellular levels of a protein called mTOR and increases dendritic spines in pyramidal cells of the frontal cortex, which might account for its antidepressant effect. Ketamine must be given intravenously, and may cause dissociative side effects. Some facilities are now studying repeated ketamine infusions, for example at weekly intervals, to observe for prolonged responses. This treatment remains off-label but is an intriguing possibility for the acute treatment of severe depression, including suicidality (DiazGranados N et al, J Clin Psychiatry 2010;71(12):1605–1611).
Acetylcholine may be involved in the control of mood, cognition, and attention. We know, for example “pro-cholinergic” drugs such as physostigmine, an inhibitor of cholinesterase, have the potential to exacerbate depression. You may recall that there are two subtypes of acetylcholine receptors. The anticholinergic drug scopolamine, which binds to the muscarinic subtype, may have a rapid antidepressant effect similar to ketamine. In an NIH-funded study of 18 depressed patients receiving intravenous scopolamine (4 mg/kg), antidepressant effects were seen within four days, and 56% achieved remission after three infusions three to four days apart (Furey ML and Drevets WC, Arch Gen Psychiatry 2006;63(10):1121–1129).
Another investigational antidepressant, S-mecamylamine (TC-5214), binds to the nicotinic subtype of the acetylcholine receptor. This is the same receptor responsible for the effect of varenicline (Chantix) and, of course, nicotine. An initial trial of TC-5214, a “nicotinic receptor modulator,” in 265 subjects found an average six-point decrease in HAM-D scores, but two phase III trials failed to replicate these findings. In March 2012, AstraZeneca announced that it would no longer be pursuing this agent due to the negative clinical trial data.
While it doesn’t look like there are any more SSRIs on the horizon, drugs that affect the serotonin system remain targets for antidepressant development. Vortioxetine (Lu-AA-21004), from Lundbeck and Takeda Pharmaceuticals, is currently in phase III trials. It is a partial agonist of 5-HT1A receptors (like buspirone), but it also blocks 5-HT3 and 5-HT7 receptors. The function of these receptors in depression is not well understood (although the 5-HT7 antagonist properties might remind you of the novel antipsychotic lurasidone [Latuda]). A six-week, placebo-controlled trial of 429 subjects showed response (68%) and remission (49%) rates that were comparable to venlafaxine and superior to placebo (Connolly KR and Thase ME, Expert Opin Emerg Drugs 2012;online ahead of print).
There may still be a market for multifunctional reuptake inhibitors, too. Forest Labs is conducting phase III trials of levomilnacipran, the active enantiomer of milnacipran, an SNRI like duloxetine and venlafaxine. Milnacipran is already marketed in the U.S. as Savella. However, most psychiatrists are probably unfamiliar with it, because its sole FDA indication (received in 2009) is for fibromyalgia. It’s not clear why Forest never sought approval of milnacipran for depression in the U.S. (it’s approved overseas for this indication); perhaps because earlier research showed no benefit over SSRIs or TCAs (Nakagawa A et al, Milnacipran versus other antidepressive agents for depression. Cochrane Database of Systematic Reviews, 2009).
Finally, a study of a novel norepinephrine reuptake inhibitor called edivoxetine showed favorable response (49%) and remission (30%) rates in a 10-week, placebo-controlled trial (Pangallo B et al, J Psychiatr Res 2011;45(6):748–755). Eli Lilly is currently conducting four separate phase III trials of this agent. The company is seeking an indication for adjunctive treatment of depression (basically converting an SSRI alone into an “SNRI-style” regimen) and for ADHD, where it may replace its sister drug Strattera (atomoxetine), also from Lilly, which is projected to lose its patent protection in 2016.
The future treatment options for depression appear to be limited. But rather than give up hope, we might see this as a golden opportunity to refine our use of existing treatments, or characterize exactly which medications are most appropriate for specific patients.
In addition, a “back-to-basics” emphasis on evidence-based psychotherapies (practiced by us or our colleagues) might help us treat our patients more effectively. Finally, we might see more use of somatic interventions such as transcranial magnetic stimulation (TMS) or deep brain stimulation (DBS). The FDA’s regulatory hurdles are somewhat lower for these interventions, but that means we need to be even more diligent in looking for the evidence that these treatments work.
TCPR’S VERDICT: Even though glutamate- and acetylcholine-based drugs might hold some promise in the future, other new antidepressant pathways seem to have led to dead ends, at least so far. For now, it’s probably best to understand the currently available drugs but not hold out hope that the next miracle cure is right around the corner.