Table of Contents


Case of the month:

Your treatment is as good as your diagnosis. 



A 80 year-old male seeks consultation for severe treatment-refractory depression. He had been depressed for 7 months, without improvement despite treatment with venlafaxine 150 mg/d for 6 weeks, which was then changed to citalopram 50 mg/d for 9 weeks, which made him more depressed and suicidal. Then Zyprexa (Olanzapine) was added to citalopram, up to 15 mg/d, without benefit. Then, fluoxetine was increased up to 80 mg/d plus zyprexa, and valproate was added at the same time at 500 mg/d; these last changes made him feel even worse, and valproate was switched to risperidone. At consultation, he was taking Prozac (fluoxetine) 80 mg/d, Zyprexa 15 mg/d, and risperidone 3 mg/d, along with multiple cardiological medications. 

He complained greatly about anxiety and indecision. “I have a lot of regrets; they bother me a lot. My worry has a basis, but it is overblown.” He had a great deal of nihilistic thinking; family reported: “He worries about the least likely thing that could happen.” He could not function at work. 

He had clear discrete depressive episodes in 1968 (his first episode, received ECT), 1975 (recovered in 3-4 months), 1996 (recovered in 3-6 months), 2001 (lasted longer, about 8-9 months, saw a consultant, who diagnosed bipolar illness but didn't stop antidepressants), and the current episode (2013). Unlike prior episodes, there was not a major psychosocial trigger for his last episode.  In between episodes, His family reported that he was "perfectly normal: very social, very gregarious" and functional as a prominent lawyer.   

No prior psychiatrist except for one consultant had observed that each of the depressive episodes described above had been preceded by a phase of  5-6 months of elevation.  According to family:  “He was a bit high, even more outgoing, would go and visit so many relatives when he went out for an errand that he’d get distracted, would sleep little, and would read furiously about all different kinds of topics that have nothing to do with his field.  He had a lot of energy, get up very early and would be very active.” 

He had no past suicide attempts or psychiatric hospitalizations, a medical history of hypertension and coronary artery disease, and a normal head MRI.  He had no drug allergies and no substance abuse history. The patient had no history of trauma or abuse, had been married for 55 years, had four adult children, and worked as a lawyer.  

Medication trials were multiple antidepressants (most of the SRIs), multiple neuroleptics, but only one mood stabilizer (recent treatment with low-dose valproate).  

In family history, his brother was diagnosed with “schizophrenia” 40 years ago – “He was quite lively, became obsessed with reading all medical books”, and was treated with antipsychotics.  This brother also had severe depressive episodes with marked loss of appetite and a catatonic state. Bipolar disorder was diagnosed in two second cousins, treated with lithium.  A niece committed suicide. 

The PL diagnosis and clinical impression

The diagnosis was bipolar disorder, type II, currently depressed.  The diagnosis of bipolar illness, is clear, as previously also noted by another consultant.  Unfortunately, this diagnosis had not been taken seriously, and thus had not been treated with the appropriate treatment, namely, an adequate dose of a mood stabilizer without antidepressants, which destabilize bipolar illness and counteract the benefits of mood stabilizers. (For further explanation of this idea, see the PL website).

Prior evaluations missed the fact that each depressive episode is preceded by a clear hypomanic episode that lasts months.  These hypomanic episodes are not questionable and are not brief. They exist, yet he has been treated as if they didn’t exist.  Prior psychiatrists either didn't make the bipolar diagnosis, or saw it as therapeutically unimportant. The consultant who had diagnosed bipolar illness made no effort to stop antidepressants or recommend mood stabilizers. He even focused instead on psychosocial explanations, such as Erikson's stage of "generativity versus despair" in old age.  (How this explanation would explain the patient's depressive episodes in his 20s, 30s, and 40s was left unexplained). 

Prior treatment had focused on treating each acute depressive episode with antidepressant and antipsychotic combinations, rather than trying to prevent those depressive episodes with mood stabilizers.  Divalproex had been used recently for his acute bipolar depression, but at a subtherapeutic dose of 500 mg/d, added to very high dose Prozac.  If antidepressants counteract the benefits of mood stabilizers, that trial would not have been effective.  This case is a good example of how antidepressants simply can be ineffective for acute bipolar depression, as found in the largest randomized clinical trials. Hence the mistaken label of "treatment-refractory depression":  the depression is not refractory when antidepressants have been proven ineffective for it.  

Prior treatment had focused on treating each acute depressive episode with antidepressant and antipsychotic combinations, rather than trying to prevent those depressive episodes with mood stabilizers.

When the patient improved in the past, it likely was not because of antidepressant treatment, but rather as part of the natural course of bipolar illness, namely that each depressive episode tends to last a certain amount of time (usually about 3-6 months) and then resolves on its own. 

When ineffective treatments are multiplied (such as combining olanzapine and risperidone) and increased to maximal doses (such as 80 mg/d of fluoxetine), in an 80-year-old person, then the likelihood of harm is magnified. These doses are meant for non-elderly adults. In elderly persons, it is a maxim that doses should be halved, because decreased renal excretion over time leads to lengthening of drug half-lives. Thus 80 mg/d of Prozac in a nearly 80 year old gentleman is quite excessive, and likely to be causing important physical harm. It is especially notable that Prozac is a very potent inhibitor of drug metabolism in the liver. Thus, this very high dose is also markedly increasing doses of almost all other medications, including olanzapine and risperidone, which would notably increase their side effects, including parkinsonian tremor and rigidity, akathisia, and cardiovascular risks.

For all these reasons, the patient's medications were proven ineffective, and at the very high doses given, posed major medical risks to him, and were likely causing, or would soon cause, serious side effects.  

The PL treatment

Prozac has the advantage of less serotonin withdrawal than other SRIs.  It was tapered off over one 2 weeks (40 mg/d for 1 week, then 20 mg/d for 1 week, then stop). The patient had no SRI withdrawal.  Risperidone was discontinued immediately and olanzapine was tapered off (10 mg/d for 1 week, then 5 mg/d for 1 week, then stop). 

One week after the initial visit, a titration of lamotrigine was initiated, reaching a target dose of 100 mg/d over another month (25 mg/d for 1 week, then 50 mg/d for 1 week, then 75 mg/d for 1 week, then 100 mg/d.) Lorazepam 1 mg  BID was given for anxiety and insomnia and to aid with SRI withdrawal.  

Two months later, the patient had improved markedly on lamotrigine 100 mg/d, and lorazepam 1 mg BID.

The depression isn't refractory when antidepressants have been proven ineffective for it.

The PL Bottom Line

  • This apparent treatment-resistant depression is really mistreated bipolar depression - a common scenario.
  • Very high dose fluoxetine with high dose olanzapine plus risperidone proved ineffective, and was no better than lamotrigine alone.
  • At least we were able to get the patient off medically harmful drugs, like olanzapine, and drugs with major drug interactions (fluoxetine).
  • Lamotrigine is now in place to help prevent future depressive episodes, with minimal long- term medical risks in this older gentleman with coronary artery disease.
  • Erikson's stages are irrelevant if you get the diagnosis and/or treatment wrong.
  • This type of case illustrates the adage: Your treatment is only as good as your diagnosis.

Meet our expert EDITORIAL BOARD, composed of clinicians and researchers from around the world. 

Subscribe to the RSS feed below to follow our "What's new" blog posts