Treatment-Resistant Depression: Clinical Conundrum

Last updated: November 15, 2018
Host: Wegdan Rashad, MD
Today’s question is: How to manage treatment-resistant depression?

Here is a summary of this episode:

  • The first step in the management of treatment-resistant depression (TRD) is adequate history taking and assessment for comorbidities and bipolar depression.
  • Augmentation is more effective than switching. Adding lithium is effective particularly in suicidal patients. Modafinil and T3 have been shown to be good add-ons in patients with fatigue.
  • Psychosocial interventions such as mindfulness CBT and behavioral activation have been shown to improve outcomes and prevent relapse of depression.
  • Depression may have an underlying inflammatory component. It has been found in some studies that patients with raised CRP above 3 mg/dL did better with nortriptyline than with SSRIs.


Hello everyone and welcome to the Psychopharmacology Institute podcast. I’m your host, Dr. Wegdan Rashad. In this series we will be discussing topics that matter to you in your clinical practice, sharing expert opinions and latest research.

When I started psychiatry practice I had the notion that depression was one of those conditions that always have a good outcome when treated adequately. I realized, however, that in some patients multiple trials of medications don’t seem to help them at all. The concept of treatment-resistant depression (or TRD for short) can be a real conundrum for psychiatrists and obviously a frustration for patients suffering from it.

Today, you will first join me on a journey in exploring the options available for TRD and then we’ll wrap up with a treatment algorithm you can apply in your practice. I will do this with the help of snippets from Professor Philip Cowen’s lecture on treatment-resistant depression available for our premium members on Dr. Cowen is a Professor of psychopharmacology at Oxford University in the UK.

So before we jump in, how do we define treatment-resistant depression? In simple terms, a patient who has failed to respond to two separate, adequate trials of antidepressant medication could be thought of as treatment-resistant. Now that’s out of the way, how to treat it?

Not so fast! Before we rush to label a patient treatment-resistant we need to fully assess him or her properly. Here’s Dr. Cowen on assessment of TRD.

Dr. Cowen:
In practice, every patient needs to be assessed as an individual. So you want to know exactly what treatment they’ve had, how many kinds of medication, what psychological treatment they’ve had. One also wants to get an understanding of how long the condition has been going on for and what is the level of symptomatology. Are they completely disabled? Can they work still? What’s their social life like?
Another key factor in the assessment is to check the diagnosis. Is it major depression? Are there other comorbidities that we know that tend to make response to treatment less satisfactory? For example, comorbid anxiety, substance misuse.
The other issue to think about is might they have a depression in the bipolar spectrum? Because we know now for patients with bipolar disorder, standard antidepressant treatments don’t seem as effective as they do in major depression. And other kinds of approaches might be necessary.

So, in short, we take a sufficient history of past therapy, how disabling their illness is and of course, to revise the diagnosis. I could add that some patients may need further workup to exclude a possible secondary medical cause to their depression.

Now, how about depression with psychotic features?

Dr. Cowen:
It’s often quite hard to know when the negative thinking of depressed patient crosses the line into a psychotic symptom. But one just wants to look out for patients who feel rather persistently they’ve made a failure of everything, that they’re never going to recover and really hang on to this with a great intensity. And here, it’s important to remember that treatment with an antidepressant monotherapy is rarely successful. And patients will normally need an antipsychotic drug fully dosed as if one was treating a psychosis and an antidepressant. And undiagnosed depressive psychosis counts for five or six cases I see each year. And you normally find when you add an antipsychotic drug at a good dosage, patients will make a good response.

It can be easy to omit asking about psychosis in a patient with depression, so this is an excellent point. Now, you might be thinking, after how long do I wait on an antidepressant trial to judge its response, or lack thereof? On average if the patient is not responding within 4 weeks, then it’s unlikely that waiting will make much of a difference.

Just like combatting any tough dilemma, we employ multiple strategies to tackle it from all sides. We will touch upon four modalities namely; pharmacologic strategies, psychosocial interventions, brain stimulation and amusing, experimental strategies, not yet approved, but showing promise.

So for pharmacologic strategies, we can switch, augment and combine. Over to Dr. Cowen.

Dr. Cowen:
So most guidelines suggest that where a patient doesn’t respond to a first line medication which generally nowadays is going to be an SSRI you should switch to a different treatment.
Intuitively, you’d think if a patient hasn’t responded to a drug such as an SSRI that potentiates serotonin function there wouldn’t be much point switching to another drug that has the same mechanism, so in this case, a second SSRI. In fact, randomized trials suggest it doesn’t really matter too much if you switch to a drug with the same mechanism of action or a different one.

So, switching to an SSRI or an SNRI or something like bupropion for instance, doesn’t actually matter too much, they are just as effective according to the famous STAR*D study.

My Psychopharmacology Institute colleague, Dr. Dana Wang hosted Dr. Cowen in one of our “Expert Consultations” interviews and she posed a very interesting question.

Dr. Wang:
In your lecture on switching after a failed trial of SSRI, you briefly mentioned the “Prozac poop out”. Would it still be advised for the patient to switch to another SSRI or SNRI? Would waiting one week before switching SSRI enough time to reverse this phenomenon? In your clinical experience, what are some of the strategies you use to deal with this problem?

Dr. Cowen:
So I don’t think this is a problem that’s being studied systematically. So what I say will be based on my own experience. My sense is that when someone has stopped responding to an SSRI, switching to a further one isn’t very helpful and the same applies to lowering or stopping the SSRI for a week or two and then trying again. So my usual practice is to try something rather different either as a treatment by itself or added to the SSRI. And I’m thinking of something here pharmacologically rather different so adding another antidepressant such as the noradrenaline and dopamine reuptake blocker, bupropion or perhaps bupropion by itself. The other approach would be to use an augmentation approach with, for example, a drug such as aripiprazole.
I think it’s an interesting idea. So the notion would be I suppose is that if you took away the 5-HT reuptake blocker, then perhaps the desensitized receptors will be able to restore themselves to normal again. I’m just not sure whether one or two weeks would be long enough for that to occur. Often, one is looking at some neuroadaptive changes in the brain that might have occurred over several months and might take several weeks to go back to normal again. During that time, the patient might be quite depressed and despairing. So it will be hard just to stop the drug without doing anything else.

And this leads us, ever-so-swiftly to our next line of defense; combination and augmentation. “What’s the difference between combination and augmentation? I hear you ask. As for those who didn’t ask… here’s the answer anyway.

Dr. Cowen:
Conventionally, the phrase combination treatment refers to adding a second drug that’s also an antidepressant in its own right, so a drug such as mirtazapine. Augmentation consists of adding a drug that’s not normally thought to be an antidepressant by itself but can produce an antidepressant effect when added to an ineffective antidepressant treatment. An example of this will be lithium. In fact, people tend to swap between the two kinds of expressions. Combination or augmentation are used fairly interchangeably. Generally, augmentation treatment seems a bit more effective than switching particularly if patients have made a partial response to the first treatment.

Interesting that you should mention mirtazapine, Dr. Cowen, because I recently came across a multicenter, randomized placebo-controlled trial published in October 2018 revealing that they found no evidence of benefit with mirtazapine added to SSRIs or SNRIs compared to placebo. In fact, the side effects were responsible for relatively more participant dropout. But in general, when combining we start low and go slow and I’d like to add, keep your eyes open for adverse effects like serotonin syndrome for example.

The third and most effective pharmacologic strategy is augmentation. On the augmentation menu, we have many delicious options to choose from including; second-generation antipsychotics, lithium, stimulants and thyroid hormones.

Dr. Cowen:
So the augmentation strategy with the best evidence base is the addition of atypical antipsychotic drugs. So it’s been found in a series of placebo-controlled randomized trials in over 3000 patients that adding drugs such as olanzapine, quetiapine, aripiprazole, brexpiprazole can be useful in securing an antidepressant response in patients who failed to respond to an SSRI. The same kind of effect seems to occur in patients taking SNRIs though this is not being studied so systematically. The doses of atypicals used for this purpose is distinctly lower than that used to treat psychosis. So one is thinking of a dose of about 2.5 to 10 mg of aripiprazole, 50 to 300 mg of quetiapine. Pharmacologically, 5-HT2 receptor blockade has been implicated.

The cool thing about augmentation is that you can choose your add-on tailored to your patient’s symptoms. Dr. Cowen explains more.

Dr. Cowen:
So for a patient who is sleeping poorly with anxiety, adding quetiapine at night might be helpful. Patients who seem to have more problems with motivation and anhedonia can be helped by the addition of aripiprazole.

But for how long do we keep second-generation antipsychotics for, given their unsavory side effects like movement disorders and metabolic syndrome?” The answer, dear listener, is not going to satisfy you, because simply we don’t have conclusive data on that yet. But, Dr. Cowen does mention in his lecture that it may be used for months and if you want to withdraw it, do it slowly and with a close eye on signs of relapse as you go.

Now for augmentation with mood stabilizers, let’s start with our historical friend, lithium.

Dr. Cowen:
I think also lithium is a less popular treatment. It’s harder to use because of the need for blood tests and continual monitoring with blood tests. Nevertheless, the meta-analyses suggests that it is an effective treatment and the number needed to treat is about four or five which if that’s generally the case, it would be rather better than augmentation with atypicals.
People sometimes ask what’s the right dose of lithium in patients with resistant depression when you’re adding it to antidepressant treatment? And generally, you might be able to use rather more modest doses than in bipolar disorder. I would tend to go between 0.4 to 0.6 mmol/L.

So, lithium is an effective option and I’m sure you remember that lithium is the drug to use in suicidal patients with a mood disorder, so keep this in mind.

Have you ever used psychostimulants like methylphenidate to augment antidepressants in TRD? It is not very common practice but here’s the evidence in a nutshell.

Dr. Cowen:
Even though psychostimulants don’t seem to be very helpful in treatment-resistant depression, there is some benefit from modafinil augmentation in SSRI-resistant patients particularly those who are sleepy or feeling fatigued.
And there is evidence from a case series that high dose pramipexole might be useful in patients with treatment-resistance but it would need to be used with some care.

So the verdict says…hmm, maybe not the best.
The strategy of adding thyroid hormone, triiodothyronine, is an old strategy when it was typically added to tricyclic antidepressants. Who would make a good candidate for T3 augmentation, though?

Dr. Cowen:
So T3 was tried in STAR*D and actually appeared numerically better to the addition of lithium in antidepressant-resistant patients. And it was slightly better tolerated. On the other hand, remission rates of both treatments in STAR*D were fairly low. The dose given in STAR*D was between 25 to 50 mcg daily. I would generally start at a low dose something like 20 mcg daily and see how people feel with that.
The things to watch out for will be typical signs of excess thyroid function such as increased pulse rate, sweating and feeling hot. It’s also wise to precede treatments with an assessment of thyroid function and an ECG. Clearly, if there’s any evidence of cardiovascular disease, it’s prudent to avoid T3 treatment. There are no clearly established predictors for T3 augmentation in terms of which patients are going to do best. But intuitively, patients with tiredness, low energy, apathy would seem to be the right kind of candidates. Anecdotally, women patients are said to respond rather better than men to T3 augmentation.

So in simple terms, patients with TRD who happen to be female, tired all the time and do not have cardiac problems, could be great candidates for thyroid augmentation.

Of course, there are more augmenting agents available discussed by Dr. Cowen and I would recommend you watching the full lecture to find out more.

Now, as psychiatrists we know very well that medications are not everything. Ideally, psychotropic medications should be prescribed hand-in-hand with psychotherapy. Imagine going through months and maybe even years of failed treatments and disabling depression, the least you could feel is hopeless. Dr. Cowen sheds some light on this matter.

Dr. Cowen:
Taking time to see the patient with family or partners can be very helpful for both. Just to hear how things are going at home and to give advice about the best way to try and help people and to avoid criticism where this is possible. Patients with depression often don’t have a structured day particularly if they aren’t working. So working out some kind of activity scheduling which hopefully will include some rewarding activities is always worth doing.
There is a rather simpler treatment called behavioral activation which focuses again on people’s activity and trying to incorporate rewarding experiences. This might be easier for the more depressed patients to deal with. And in fact, a recent controlled trial showed that behavioral activation was possibly more effective than CBT in severely depressed patients.

It has been found that not only are psychosocial interventions effective, but they prevent relapses of depression as well. So, definitely, it’s something worth trying with your patients.

Now we have two tactics left. One, brain stimulation modalities like electroconvulsive therapy, transcranial magnetic stimulation, and deep brain stimulation and two, experimental strategies like using hallucinogens, ketamine and even anti-inflammatory drugs. So, stay tuned.

Medications and psychotherapy may not be doing it for your patient. This may call for more invasive procedures. I will briefly touch upon each one. Electroconvulsive therapy or ECT is effective for more serious symptoms like psychotic depression, suicidality, and catatonic features. Transcranial Magnetic Stimulation or TMS is less effective but does not have known cognitive side effects. As for Deep Brain Stimulation or DBS, well, the evidence base is still inconclusive for our patients with TRD.

And now for some fun and games! Well, it’s not exactly fun and games, but certainly, there are exciting, experimental strategies on the horizon that you should know about! It is likely that you’ve heard of the rapid-acting antidepressant, ketamine, that within hours can rid a patient of depressed mood and suicidal thinking. It is used off-label in some places, but its safety profile for long-term use is still inconclusive.

Now, do you imagine in the near future prescribing an anti-inflammatory agent like celecoxib along with your everyday sertraline or escitalopram? Over to Dr. Wang and Dr. Cowen.

Dr. Wang:
You talked about some fascinating evidence and clinical applications of using inflammatory markers. Have you had experiences of using this data in your clinical practice? What would you like to see more research in this area?

Dr. Cowen:
So there are a number of studies going on in patients with raised inflammatory markers and that normally means raised levels of C-reactive protein, CRP in the blood and various biological agents. So things like drugs that block TNF-alpha, drugs that block IL-6, so these are quite powerful biological agents. And there are a number of registered clinical trials going on at the moment. I think at the end of those in a year or two we’ll have quite a good idea about whether these drugs work in patients with raised inflammatory markers and what the range of side effects might be.
Before then, what can we do as clinicians? I have started measuring CRP in my patients because I think it’s an interesting marker. And above a level of 3 mg/mL, then people do think that patients have signs of raised inflammation. There are one or two approaches one might try there particularly if patients aren’t doing well with conventional treatment. So one, I may have mentioned was a study that found that patients with CRP greater than 2 or 3 did better with nortriptyline, the noradrenaline related tricyclic antidepressants than with the SSRI escitalopram. So when patients haven’t responded well and they’ve also raised CRP, then I sometimes think of a trial of nortriptyline. Equally, there are lifestyle things patients can do if they have raised inflammatory markers. And these would include things like exercise regularly and perhaps the use of omega-3 fish oils.
It’s still a bit early to say that these patients should be given specific anti-inflammatory drugs. I think we’ll know the answer to that in a couple of years. But prior to that, as I said, there are some things one can do on an everyday level ranging from things like the use of nortriptyline to various other activities which could be designed to lower inflammation. And that would certainly include things like diet and exercise.

Very interesting! I am sure we will hear more about the use of antiinflammatory agents in treatment-resistant depression, so watch this space.

Another intriguing development is the increasing interest and studies on the use of hallucinogens like shrooms or ecstasy also known as psilocybin and MDMA respectively. Actually, I hosted an interview with Dr. Robin Carhart-Harris at the 2018 ASCP Annual Meeting and he spoke more in detail about this controversial modality. Check out the podcast entitled “Psilocybin for Treatment-Resistant Depression” in the podcast section.

Phew! Now let’s wrap it up in an algorithm you can take home with you, or rather, take to your clinic. But you must keep in mind, there is no one-size-fits-all algorithm, each patient has their own unique history and circumstances which need to be taken into account.

Dr. Cowen:
In general what we have to start with are the two newer antidepressants that are thought to be among the most effective from the various network meta-analysis. So here, we’re starting with either the SSRI escitalopram or the SNRI venlafaxine. Both are given at reasonable doses. So if patients are taking either those drugs and they haven’t responded, we could think about adding an atypical and antipsychotic. Mirtazapine is an alternative. Another possibility might be to think of lithium. Moving down, I think there’s a place for ECT in patients who don’t respond well to these treatments, who are more treatment refractory. But ECT can be considered earlier in patients with very severe symptoms. On the final line, just to point out that psychological treatment is a key part of the management of resistant depression with CBT and behavioral activation always worthwhile trying.

And that’s all folks, a short and sweet algorithm right there.
Now for those who have dozed off, listen up because here come the long-awaited key points!

Key Points

  • The first step in the management of treatment-resistant depression (TRD) is adequate history taking and assessment for comorbidities and bipolar depression.
  • Augmentation is more effective than switching. The evidence base supports augmenting with second-generation antipsychotics at lower doses than for psychosis. Adding lithium is also effective, particularly if the patient is suicidal. Modafinil and T3 have been shown to be good add-ons in patients with fatigue.
  • Psychosocial interventions such as mindfulness CBT and behavioral activation have been shown to improve outcomes and prevent relapse of depression.
  • Electroconvulsive therapy has shown efficacy for severe and psychotic depression, however, TMS is better in terms of cognitive adverse effects.
  • Depression may have an underlying inflammatory component and it has been found in some studies that patients with raised CRP above 3 mg/dL did better with nortriptyline than with SSRIs.

Liked our podcast? Visit us at where you can find more podcasts and some lectures absolutely free. You can choose to become a premium member to access all lectures and interviews. Also, we would be thrilled to hear from you, drop us an email at

The following people participated in this episode: Dr. Flavio Guzman as the general editor, Rosario Anon Suarez as the audio engineer, Pamela Gonzalez as the project manager and myself Dr. Wegdan Rashad as the host.

Thank you for joining us in today’s podcast until the next episode, goodbye!


  1. Rush, A. J., Trivedi, M. H., Wisniewski, S. R., Nierenberg, A. A., Stewart, J. W., Warden, D., … & McGrath, P. J. (2006). Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR* D report. American Journal of Psychiatry, 163(11), 1905-1917.
  2. Kessler, D. S., MacNeill, S. J., Tallon, D., Lewis, G., Peters, T. J., Hollingworth, W., … & Shepherd, T. (2018). Mirtazapine added to SSRIs or SNRIs for treatment resistant depression in primary care: phase III randomised placebo controlled trial (MIR). BMJ, 363, k4218.

Earn CME/SA credits

  • Up to 36 CME credits/year
  • Up to 16 SA credits/year
  • Download more than 70 videos
  • Exclusive updates

Free Course: “SSRIs: The Essentials”

Learn the essentials of SSRIs with our free online course.

One module per week:

  • The Mechanism of Action of SSRIs
  • Fluoxetine
  • Fluvoxamine
  • Paroxetine
  • Sertraline
  • Citalopram and Escitalopram

2018 Psychopharmacology Online Update

Earn Up to 36 CME Credits / Year

  • Access exclusive updates: new videos every month
  • Download our entire library: more than 70 videos
  • Earn up to 36 CME credits /year