The last major development in the pharmacotherapy of major depressive disorder (MDD) was indeed revolutionary. The approval in the 1980s of selective serotonin reuptake inhibitors (SSRIs) for depression, which appeared to match the effectiveness of older antidepressants without as much of a side effect burden, may have been one of the main reasons that the use of antidepressants tripled in the following years, with SSRIs accounting for more than half of all antidepressant medication prescriptions by 2006 (1).

In the two decades since, much of the action in antidepressant development has revolved around expanding upon the currently known antidepressant medication classes by identifying other selective reuptake inhibitors, developing other medications that mimic the action of already-known antidepressants, or testing medications approved for other indications as depression augmentation or monotherapy agents. Still, these new medications have not yielded a substantial advance in the rate of treatment response for depression, with only half of patients in the STAR*D trial achieving symptomatic remission within the first two treatment stages (2, 3).

More recently, however, advances in knowledge about the pathophysiology of depression have opened up additional potential directions for drug development. For example, discoveries about the role of other neurotransmitters in MDD have led to the development of potential antidepressants tapping the glutamatergic and nicotinic neurotransmitter systems (Table 1). Other avenues for drug development have targeted hormonal systems such as the hypothalamic-pituitary-adrenal axis, known to be involved in the body’s response to stress, and the circadian system, which might affect depression via its effects on sleep. At the same time, the increasing ease of gene sequencing and the growing body of research on genetic polymorphisms associated with antidepressant response means that it may soon be possible not only to profile an individual patient’s likelihood of experiencing side effects with a given medication but also to choose the most promising pharmacological treatment. This article will outline some of the mechanisms for antidepressant action that are attracting the most attention from drug developers and that could represent the next wave of antidepressants.

All of the antidepressant medications prescribed today capitalize on some aspect of monoamine neurotransmission, whether serotonin, norepinephrine, or dopamine. The most common are the serotonergic agents, which includes selective serotonin reuptake inhibitors (SSRIs) and serotonin-norepinephrine reuptake inhibitors (SNRIs). These agents all inhibit the serotonin transporter, with varying degrees of agonism or antagonism at specific receptor subtypes; some also inhibit the norepinephrine transporter.

One of the newer variations on this class is vilazodone, which was approved by the FDA for depression in 2011. This medication both inhibits the serotonin transporter and acts as a partial agonist at 5HT1A receptors; thus, it is sometimes also called a SPARI (serotonin partial agonist reuptake inhibitor) (4, 5). Whether this additional action of 5HT1A partial agonism boosts the efficacy of serotonin inhibition or mitigates the side effects of 5HT1A partial agonism remains unknown, but this combination of mechanisms may result in lower rates of sexual dysfunction and weight gain. (4)

Another multimodal serotonergic medication in development is vortioxetine, a 5HT3a receptor antagonist, 5HT7 receptor antagonist, 5HT1B partial agonist, 5HT1A agonist, and 5HT transporter inhibitor (6). Vortioxetine has been found to reduce symptoms of depression in several Phase III trials (7–9) and has been associated with a more favorable cognitive effect profile (9); it has also been found to be effective in preventing relapse of depression in an open-label trial (10). In a “failed study,” while vortioxetine did not separate from placebo, neither did the other active treatment, duloxetine (11).

Among medications that combine noradrenergic and serotonergic activity is the serotonin-norepinephrine reuptake inhibitor levomilnacipran. This compound, which is an enantiomer of the fibromyalgia medication milnacipran, has received approval as an antidepressant in Europe. The manufacturer recently presented evidence of greater improvement in depressive symptoms with levomilnacipran compared with placebo from two Phase III trials (12). Another possibility is the selective norepinephrine reuptake inhibitor edivoxetine. While it primarily has been studied for attention deficit hyperactivity disorder, it is also being considered as a potential treatment for major depression; a recent study showed greater symptomatic improvement and higher response and remission rates with edivoxetine compared with placebo in patients with major depression (13).

In addition, there is interest in developing antidepressants with dopaminergic activity, perhaps in combination with serotonergic and noradrenergic actions. The idea is that targeting dopaminergic neurotransmission will enhance overall efficacy; reduce anhedonia, apathy, and cognitive impairment; and minimize residual fatigue and sleepiness (as suggested by the dopamine reuptake inhibitor modafinil augmentation studies of SSRIs (14)). In addition, given that dopaminergic medications have been used to treat SSRI-induced sexual dysfunction (15, 16), there is hope that dopaminergic antidepressants may cause less sexual dysfunction than SSRIs and SNRIs.

One dopaminergic compound currently under study is the dopamine (D2) and 5HT1a partial agonist/5HT2 antagonist brexpiprazole, which is structurally related to the atypical antipsychotic (and depression augmentation agent) aripiprazole. Brexpiprazole showed promise in a Phase II trial as an augmentation treatment for MDD (17). Another potential antidepressant with dopaminergic activity is lisdexamfetamine. This ADHD medication is a prodrug of the stimulant dextroamphetamine, which blocks presynaptic reuptake of norepinephrine and dopamine and increases their release. Shire has reported positive Phase II results from a trial using lisdexamfetamine as an augmentation treatment for major depression in full or partial remission but with continued executive dysfunction (18, 19).

Because of the role of dopamine in the brain’s reward circuit and addictive behaviors, there has been some concern about the risk of abuse with dopaminergic medications, perhaps limiting enthusiasm for studying dopaminergic antidepressants. But there have been animal studies that have found no evidence of abuse liability with certain dopaminergic medications. For example, in a study of rats given a medication with dopaminergic as well as serotonergic and noradrenergic activity, the rats did not exhibit self-administration of the medication (a marker of abuse liability) (20); another study found that administration of a similar medication was associated with decreased ethanol consumption in alcohol-preferring rats (21).

Triple uptake inhibitors (TUIs) are attracting particular interest because of the promise of modulating three monoamine neurotransmitter systems at once. The hope is that these “all-in-one” compounds would have the synergistic effects of triple inhibition and lead to more robust antidepressant effects without requiring high occupancy of the serotonin transporter, thus minimizing many of the side effects seen with SSRIs. The TUI amitifadine has demonstrated positive results on several depression outcome measures, including an anhedonia measure, without significant weight gain or sexual dysfunction in a small phase 2 trial (22).

Finally, in a parallel line of research, pharmaceutical companies are pursuing variations on monoamine oxidase inhibitors, which decrease the breakdown of serotonin, norepinephrine, and dopamine and thus increase their levels. One of the major limitations of the early monoamine oxidase inhibitors was that they affected both the MAO-A and MAO-B forms of the enzyme and are generally irreversible, making it particularly dangerous to ingest dietary tyramine. Without either form of the enzyme available to break down tyramine for 2 weeks (the length of time it takes to regenerate the enzyme), patients were at high risk for hypertensive crisis after consuming foods containing tyramine. This was the impetus behind the development of reversible MAOIs, such as moclobemide and the newer CX 157 (TriRima). TriRima is thought to be more potent than moclobemide and also to be a powerful inhibitor of MAO-A (23), which is primarily responsible for breaking down serotonin, norepinephrine, melatonin and epinephrine (dopamine is equally metabolized by MAO-A and MAO-B).

Breaking away from the traditional focus on serotonin, norepinephrine, and dopamine, a number of researchers are investigating drugs that work on other neurotransmitters that have been implicated in depression, including glutamatergic medications and compounds that target nicotinic receptors.

Glutamatergic drugs

Glutamate is one of the main excitatory neurotransmitters in the CNS, exerting its effects primarily via one of several receptor subtypes: N-methyl-d-aspartate (NMDA) receptors, alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA) receptors, kainate receptors, and type I, II, III metabotropic glutamate receptors. Because of evidence of glutamate dysfunction in major depression and of neuroprotective properties of NMDA receptor antagonists (24), there is speculation that they may be effective antidepressants.

Of the NMDA antagonists currently under study for depression, the anesthetic ketamine may have the most buzz. A number of small studies have found rapid and in some cases sustained antidepressant effects after ketamine injections, for a total sample size of 163 patients (25). A number of trials are attempting to confirm these effects in larger, more rigorous trials and to examine how the effects can be sustained. The main drawback of ketamine at this point is that it must be administered intravenously in a hospital setting.

Other NMDA antagonists receiving attention for depression include riluzole, amantadine, traxoprodil, and dextromethorphan. Riluzole, a medication that noncompetitively inhibits NMDA receptors as well as the release of glutamic acid and that was originally approved for amyotrophic lateral sclerosis, has demonstrated antidepressant-like effects in several small studies (26, 27). Amantadine, which possesses dopaminergic activity in addition to antagonizing NMDA receptors, showed positive effects in a double-blind augmentation study for depressed imipramine nonresponders (28). Another NMDA receptor antagonist, traxoprodil, has demonstrated antidepressant effects in patients who have not responded to SSRI treatment (29). Finally, the cold remedy dextromethorphan is attracting attention as well because it has some actions similar to ketamine, such as NMDA antagonism, serotonin transporter inhibition, and mu (opiate) receptor potentiation. As Nuedexta (dextromethorphan/quinidine), it has already received FDA approval for the treatment of pseudobulbar affect, and some researchers speculate that it could find utility as a conventional antidepressant, rapid-acting antidepressant, or treatment for treatment-refractory depression (30).

However, results with memantine have been somewhat disappointing. Despite promising results from a small open-label study with depressed patients (31) and a double-blind randomized controlled trial of memantine plus escitalopram in major depression comorbid with alcohol dependence (32), Zarate et al. found no effect in a double-blind, randomized, controlled study of memantine for depression monotherapy (33).

Additionally, several concerns about glutamatergic medications have reduced enthusiasm for them, most notably questions about psychedelic effects. Some of these agents possess hallucinogenic properties and may induce psychosis-like symptoms in subjects who have no previous history of psychosis (34, 35).

The potential role of glutamatergic agents that act on AMPA, kainate, or metabotropic glutamate receptors in treating psychiatric disorders is not as well studied, although there is considerable interest in these compounds as well. Given the beneficial effects of glutamatergic agents such as AMPA receptor modulators on cognition, there was hope that agents could be effective in the treatment of cognitive dysfunction in depression, or in the treatment of MDD presenting with prominent cognitive dysfunction (36–37). Unfortunately, one of the first trials of an AMPA receptor modulator, farampator (CX-691/Org 24448), was terminated early due to concerns about adverse effects observed in other studies (38).

Nicotinic Medications

Several lines of evidence have led to interest in nicotinic medications for depression. Several antidepressants, such as the tricyclic antidepressant imipramine, have antagonist activity at the nicotinic receptor, and there is evidence of nicotinic receptor dysfunction in depression (39). The link between nicotine dependence and depression is well established, with higher rates of smoking in patients with major depressive disorder (40) and vice versa (41). In addition, the nicotinic receptor may be involved in memory, cognition and behavioral reinforcement/addiction. For example, the alpha4beta2 subtype of nicotinic receptors has been reported to play a role in acetylcholine-mediated dopamine release in areas of the brain that are involved in behavioral reinforcement and addiction—specifically, the striatum, ventral tegmental area, and nucleus accumbens (42–44). The alpha7 receptors have been linked to learning and memory in preclinical studies (45).

An early open-label study of the alpha4beta2 partial agonist and alpha7 full agonist varenicline, already approved for smoking cessation, found an improvement in mood when depressed smokers started taking the medication to augment their existing psychotropic treatment (46). However, despite promising Phase II studies, the noncompetitive nicotinic receptor antagonist mecamylamine did not meet primary endpoints in Phase III trials and has been pulled from development (47).

With evidence accumulating for the role of the endogenous opioid system in regulating mood, attention is turning to opioid-based medications for their potential role in depression treatment. There are three types of opioid receptors—delta, kappa, and mu—and all three have been linked variously to monoaminergic activity (48–50), behaviors in stressful situations such as the forced swim test (51), and antidepressant response (52, 53). In humans, depressed individuals and healthy comparison subjects exhibited different patterns of mu opioid receptor availability in emotionally neutral and sad states (54), supporting a role for opioids in mood. In addition, polymorphisms in the mu opioid receptor gene were associated with citalopram response in STAR*D (55).

While much remains to be understood about the relationship between the opioid system and mood disorders—it may play a role in schizophrenia and impulse control disorders as well as mood disorders and addiction (56, 57)—researchers are already proceeding with studies of opioid-based medications for depression. As early as 1982, some studies were finding antidepressant effects of buprenorphine, a mu opioid receptor agonist and kappa opioid receptor antagonist (58), and it is currently being tested as monotherapy or augmentation for depression. Another drug in Phase II testing is ALKS 5461, which combines buprenorphine and the mu antagonist samidorphan; it showed improvements in depression symptoms in treatment-resistant patients after one week in Phase II trials (59).

Unsurprisingly given their powerful and far-ranging effects on the body, a number of hormones can influence psychiatric symptoms. Hypothyroidism, for example, often causes depressive symptoms, while corticosteroids can induce mood lability, mood disturbance, or even psychosis. Thus a number of antidepressant development approaches are focusing on hormonal pathways.

Treatments That Act on the Hypothalamic-Pituitary-Adrenal Axis

Disturbances in the hypothalamic-pituitary-adrenal (HPA) axis are a well-established feature of depression. In particular, basic and clinical studies have found evidence of increased secretion of the hypothalamic neuropeptides vasopressin and corticotrophin-releasing factor (CRF) in depression and anxiety, leading to interest in treatments addressing these mechanisms. Vasopressin is released from the pituitary during stress and potentiates the body’s release of adrenocorticotropin in response to CRF. Animal studies in rats and birds first suggested that vasopressin may be essential to the body’s ability to adapt to stress (60–62). Because the V1b receptor is responsible for the pituitary response to vasopressin, it has become a target for drug development in depression and anxiety. The nonpeptide V1b receptor antagonist SSR149415 showed mixed results on efficacy for depression in several phase IIb studies for efficacy and tolerability, though the studies did not yield significant results for efficacy as treatment for generalized anxiety disorder. The authors concluded that “the antidepressant potential of SSR149415 needs to be further evaluated” (62).

Melatonergic Medications

The circadian system plays an important role in mood disorders. Sleep disturbances are a core feature of both major depression and bipolar disorder; seasonal affective disorder is tied to the length of daylight in different seasons; and there is even some evidence that sleep deprivation can cause mood symptoms. Thus, the hormone melatonin, which is a serotonin precursor that maintains the body’s internal clock and may also play a neuroprotective role, has drawn significant attention for its potential in depression treatment.

One melatonergic drug under study for depression as well as anxiety and obsessive-compulsive disorder is agomelatine, which is a melatonergic receptor agonist (MT1/MT2) as well as a 5HT2c antagonist. While early research concentrated on its utility for sleep disorders, the first dose-finding trials for its use in major depression were published in 2002 (63), and it gained approval for treating depression in Europe in 2009. Because it has minimal serotonergic action other than 5HT2c antagonism, agomelatine is thought to be less likely to cause gastrointestinal, sexual, and metabolic side effects; its main side effect appears to be dizziness (64). Discontinuation syndrome may also be less frequent with this medication (65). However, in some studies, approximately 1%–4.5% of patients taking the medication developed a transient and reversible elevation in hepatic transaminases, sometimes only at higher doses (66–68). Due to these concerns as well as poor Phase III trial results in the U.S (69), Novartis discontinued development of agomelatine in the U.S. in October 2011, although clinical trials by other groups are continuing.

The hot word in medicine is pharmacogenetics, and the field of antidepressant research is no exception. The idea is that by knowing a person’s genetic profile with regard to variants that affect response to medication treatment, it would be possible to personalize his or her treatment to be more effective while reducing side effects.

The targets attracting the most interest in depression research are genes coding for neurotransmitter or drug transporters or receptors and genes coding for enzymes that break down those neurotransmitters or drugs. Some of the earliest studies focused on polymorphisms in the promoter region of the serotonin transporter gene, such as 5-HTTLPR, which was linked to affective disorders more than 15 years ago (70) and has been implicated in response to antidepressants and antidepressant-induced mania (71). More recently, it has become possible to screen the genome with hundreds or thousands of single-nucleotide polymorphisms (SNPs) in genome-wide association studies (GWAS) to identify other potential targets. For example, STAR*D collected DNA from approximately 2,000 study participants with nonpsychotic major depressive disorder; this data were used to identify new genes associated with citalopram response and resistance, evaluate the role of genes previously thought to be associated with some aspect of depression treatment, and examine markers associated with side effects such as suicidal ideation (72). Another GWAS, the Genome-based Therapeutic Drugs for Depression (GENDEP) study of 760 adults with moderate-to-severe depression, focused on serotonergic, noradrenergic, neurotrophic, and glutamatergic mechanisms (73).

Perhaps not unexpectedly given the limited sample sizes and the large number of polymorphisms being evaluated, these large studies have sometimes produced contradictory results. For example, the STAR*D data did not confirm earlier suggestions of a significant relationship between treatment response and polymorphisms in the BDNF gene (74) and provided a complicated picture about the role of the serotonin transporter gene and its promoter region (reported variously to have no relationship to citalopram response, a relationship to tolerability but not treatment response, and a relationship to citalopram response in non-Hispanic whites in STAR*D) (72). The Munich Antidepressant Response Signature (MARS) study of 387 adults with depressive disorders (including about 10% with bipolar depression) failed to replicate the STAR*D findings, although a predictive model including effects of and interactions between three genes involved in serotonergic, glutamatergic, and HPA activity accounted for 13% of the variance in remission after 5 weeks, consistent with the STAR*D findings (75). Part of the difficulty is that the magnitude of the relationship between individual variants and treatment response has generally been modest (76). Another problem is that multiple genetic variations may influence response to an individual drug, making it even more difficult to predict response with the knowledge currently available.

Another major question with pharmacogenetics and personalized medicine is cost effectiveness. Perlis et al. calculated that testing 40-year-old men with major depression for response to SSRIs and using bupropion in likely nonresponders would cost $93,520 per additional quality-adjusted life-year (QALY) compared with treating all patients with an SSRI first and switching in the case of nonresponse (77); by contrast, the average cost per QALY for dialysis compared with the next cheaper options is $129,090 (78). An analysis of the economics of using 5-HTTLPR genotyping in Europe concluded that testing would be cost-effective in middle-income countries if it cost $100 or less (79).

Ethical and legal questions pose another area of concern. Among the potential risks cited in a survey of 75 psychiatrists at “early adopter” institutions offering pharmacogenetic testing clinically, respondents expressed the most concern about the risks of getting secondary information about disease risks and endorsed the importance of confidentiality and informed consent; however, there was a lack of consensus overall about risks and the safeguards needed to protect patients (80).

Nevertheless, pharmacogenetic testing is already available for several hundred dollars a test (81) at a number of institutions. In the survey by Hoop et al. cited above, 64% of respondents had ordered at least one pharmacogenetic test in the last year. Genetic testing appears to be ordered most often for cases of treatment-resistant depression or medication intolerance (80, 82). Thus researchers have recommended developing guidelines for testing that take into account clinical severity and indicators of treatment resistance (82).

Despite mixed or disappointing results from some of these drug trials and difficult times for psychiatric drug development in general, there are nevertheless promising signs in antidepressant medication research. Our increasing ability to target individual neurotransmitter receptors with specific agonist or antagonist activity could lead to more effective and more tolerable medications. After a long drought in the development of truly novel pathways for antidepressant action, the fact that mechanisms other than monoamine neurotransmission are being targeted is encouraging. Furthermore, in addition to allowing for the personalization of medication treatment, the study of psychiatric genetics and biomarkers of antidepressant response could expand knowledge about the pathophysiology of depression and unearth new treatment targets for the future.