David L. Ginsberg, MD

Expert Perspective

Insights Into Treatment-Resistant Depression: Current and Emerging Therapies

Practice Community
New York, NY
Practice Niche
Hospital and Institute Affiliations
NYU Langone Health


Why are high rates of partial or inadequate responses to first-line antidepressant therapy seen in patients with major depressive disorder (MDD)?


When speaking about treatment-resistant major depressive disorder (MDD), it is best to first define treatment resistance: it refers to the failure to achieve response or remission to at least 2 different antidepressants of different classes of adequate dosage and duration.

There are different contributors to treatment resistance; one is misdiagnosis. Sometimes patients may meet some or many criteria for MDD but actually have a variant of depression that is a little different that makes them less responsive to a trial of a certain antidepressant. That misdiagnosis may include bipolar disorder I or bipolar disorder II.

There are also other subtypes within depression that need to be diagnosed for appropriate treatment, such as psychotic depression, atypical depression, and others. There may also be comorbid psychiatric conditions — including panic disorder, anxiety disorder, personality disorders, and substance use disorders — that can complicate treatment of MDD.

We also know that depression that is very severe or long-lasting is associated with treatment resistance. There may also be comorbid general medical conditions that, when undiagnosed, may make it more difficult to treat MDD.

Other factors may also include the level of adherence to the treatment regimen, as patients may not be taking medications as prescribed. Contributing factors may include the stigma of mental illness, adverse effect issues, and cost. In addition, there may be underlying pharmacokinetic factors and patient genetic factors that make them less responsive to traditional therapy. They may be rapid metabolizers of antidepressant medication, which lowers the effective antidepressant serum levels and renders them less responsive to treatment.

Broadly speaking, I like to think of different subtypes of treatment-resistant depression. There are those patients whom I would consider truly treatment resistant: those who have failed multiple classes of antidepressants at adequate doses and duration of treatment (generally at least 6 to 8 weeks). Then there are those are treatment intolerant; these are patients who have been exposed to a lot of medication but had never stayed on a single agent for an adequate amount of time. This may be because of adverse effects that are not tolerable, true drug insensitivity, or ambivalence to taking medication. Personality traits or comorbid personality disorders also may make it more challenging to treat MDD. Patients may also have ongoing significant life stressors such as the loss of a job or relationship, physical abuse, sexual abuse, or other ongoing issues.


What are the risks associated with partial or inadequate MDD treatment?


In general, we define response to antidepressant therapy as a 50% reduction in symptoms and remission as achieving below an absolute score on the Hamilton Depression Rating Scale or the Montgomery-Asberg Depression Rating Scale. We should remember that 20 to 30 years ago, antidepressant clinical trials only looked at percentage of patients who exhibited a response to treatment (not necessarily achieving remission). Although it is true that a patient with 50% reduction of symptoms will feel better and look better, they are still not completely “well.” The field of psychiatry later discovered that we really needed to go beyond response and move toward complete disease remission. The most obvious reason for this is because we want to improve our patients’ quality of life as much as we can, and this can be difficult if a patient still has residual symptoms.

In addition, there are data demonstrating that the presence of residual depression symptoms predicts a higher rate of relapse, so that patients may not even remain partially well when you have these residual symptoms. So the goal should not only be to get patients well but also to keep them well for as long as possible, which is another reason why we are now seeking to get patients to complete remission.

An analogy I like to give my patients is, if you are near-sighted and I’m your ophthalmologist, my goal is not to get your vision to 20/80, my goal is to get your vision to 20/20.


What are the current challenges associated with treating patients with selective serotonin reuptake inhibitors (SSRIs) and serotonin-norepinephrine reuptake inhibitors (SNRIs)?


There are several challenges. The first is that we currently do not have an antidepressant approved by the US Food and Drug Administration that works very fast. In contrast to fast-acting benzodiazepines that work almost immediately, antidepressants may not have noticeable effects on mood until after about 2 to 3 weeks. We typically explain to patients that it takes about 4 to 6 weeks to feel any appreciable benefit from the medication, with a subset of patients taking up to 12 weeks. We also have to remember that these periods only begin after we have achieved therapeutic drug levels, as we may have to sometimes start at much lower doses and titrate up because of adverse effect issues. So the process of getting patients up to an adequate dosage requires time, and patients must have patience. This may be difficult if you have a patient who is already reluctant to start a medication, as we are not sure whether this class of antidepressant is going to be effective at a time when we would really like a treatment response. Remember, this whole time, a patient may be really suffering: some may be unable to go to work or attend school. So this lag from starting medication to getting an appreciable response is an issue.

In addition, although SSRIs and SNRIs are a significant improvement compared with the earlier-generation tricyclic antidepressants and monoamine oxidase inhibitors, they still have adverse effects. These adverse effects can be very challenging for patients. In my experience, one of the more problematic adverse effects involves sexual functioning. All phases of sexual functioning (libido, arousal, and orgasm) can be affected both in men and women. The incidence rate is about 35% to 45%. These adverse effects tend to be dosage related, and unfortunately these adverse effects tend not to abate but rather persist. This is in contrast to other adverse effects such as gastrointestinal upset and nausea, which tend to be temporary. So oftentimes psychiatrists have to manage these long-term adverse effects as another complicating factor of SSRI/SNRI treatment that may be a limiting factor.

Another SSRI and SNRI adverse effect that patients may have involves weight gain. This weight gain may be a result of a slowing of the basal metabolic rate or, in some cases, a result of overeating. The end result is that it can be very difficult for patients to maintain or lose weight despite maintaining healthy eating habits, especially if their metabolic rate has slowed down.


Can you discuss the potential therapeutic role of modulating pathways outside of the monoamine oxidase pathway in MDD? This might include opioidergic, ionotropic, and metabotropic glutamatergic receptor modulation.


There have been studies showing that subanesthetic doses of intravenous ketamine have been effective in patients with treatment-resistant depression. There are actually a number of specialty clinics across America that are giving ketamine intravenously to patients with MDD in an off-label manner. One of the biggest challenges with ketamine is that its antidepressant effects are not durable. Typically, the duration of effect after injection is about 1 week on average, with some patients having beneficial effects for several weeks. So patients would need to come back in for repeated intravenous ketamine administration, and there’s still a lot of questions regarding the relative safety of this technique over a longer period of time. What needs to be further elucidated is the long-term effect of ketamine on the brain, and in particular on working memory and other cognitive functions. There are also concerns regarding the addictive potential of ketamine.

Another category of interest involves drugs with anti-inflammatory properties, as there has been research on the link between inflammation and neuropsychiatric disorders. We know that certain patients with particular subtypes of depression have increased inflammation, and we also know that patients with inflammatory disorders also have higher rates of depression. Therefore, there seems to be some commonality between the pathways of inflammation and depression. As a result, there has been interest in studying anti-inflammatory drugs for MDD, either as monotherapy or combination therapy with traditional antidepressants. One example is celecoxib, a nonsteroidal anti-inflammatory drug that has shown some efficacy as an augmenting agent. Another example is infliximab, a monoclonal antibody that targets tumor necrosis factor-alpha, which has been found to reduce depressive symptoms in patients with systemic inflammations, as manifested by increased inflammatory biomarkers including C-reactive protein.

A separate category of drugs includes the use of buprenorphine for MDD. A study published in the Journal of Clinical Psychiatry several years ago looked at using low-dose buprenorphine in middle-aged and older patients with treatment-resistant depression and found sustained improvements in Montgomery-Asberg Depression Rating Scale scores.1 The investigators noted appreciable improvements in mood within 3 weeks, representing a relatively fast action. Since then, there has been renewed interest in this category of drugs, with one pharmaceutical company pursuing approval of a buprenorphine-samidorphan combination drug. There has been some concern about the potential for misuse or diversion of buprenorphine, despite the fact that it has been used safely to treat opiate addiction.

Another compound that has been studied for MDD is brexanolone (formerly SAGE-547 injection), which is an allosteric neurosteroid modulator of both synaptic and extrasynaptic GABA receptors that is being evaluated for postpartum depression. There has been a lot of excitement over this compound, given how quickly patients responded to this medication.

The final modality I want to mention is sleep deprivation, which is interesting. There was a meta-analysis published in the Journal of Clinical Psychiatry looking at sleep deprivation as a treatment for depression.2 This is not new. We have known about this for a long time, but it has never really made it into clinical practice because it is not durable. So overall, the response rate to sleep deprivation for MDD is around 45%. The interesting thing about sleep deprivation is that it also works very fast. The most obvious problem, however, is that this treatment effect is short lived: after the patient sleeps the next night, the effects stop. But there are a lot of research groups looking at how to make this more practical and durable, including evaluating strategies such as partial sleep deprivation. Other researchers are looking at including bright light therapy as antidepressant therapy.


Can you briefly review your general treatment algorithms for patients with treatment-resistant depression?


I begin with a thorough patient history to first confirm the diagnosis of MDD. I also screen very carefully for bipolar disorder and other comorbid psychiatric disorders such as generalized anxiety disorder, panic disorder, posttraumatic stress disorder, alcohol and drug use disorders, and personality disorders. These comorbid psychiatric disorders may affect what medications I give them and at what doses. Another thing I specifically look for is the presence of medical disorders; I carefully review all laboratory tests and results from recent physical examinations from general practitioners. If we do not address any potential underlying medical conditions, for example, Hashimoto’s thyroiditis or anemia, we will not be able to effectively treat these patients. So we want to start with making sure these conditions are adequately treated before starting treatment with antidepressants.

I carefully screen for alcohol and drug abuse, so that I can address proper treatment for these conditions. If patients are smokers, I counsel them and note that smoking may increase the metabolism of certain drugs, rendering them less effective. I also make sure to screen for stressful life events to ensure that patients receive appropriate talk therapy that is evidence based. Basic wellness, including adequate sleep, needs to be addressed with patients, as well.

Then, as I mentioned earlier, I ensure that adequate drug trials of correct dose and duration have been performed. All these issues must first be addressed before affixing a label of treatment resistance on a patient with MDD.


1 Karp JF, Butters MA, Begley AE, et al. Safety, tolerability, and clinical effect of low-dose buprenorphine for treatment-resistant depression in midlife and older adults. J Clin Psychiatry. 2014;75(8):e785-e793.

2 Boland EM, Rao H, Dinges DF, et al. Meta-analysis of the antidepressant effects of acute sleep deprivation. J Clin Psychiatry. 2017;78(8):e1020-e1034.