Monday, November 9, 2015

Duloxetine (Cymbalta)

Duloxetine (Cymbalta) is another newer antidepressant, approved in the US in 2004, and in Canada in 2007.  It is a reuptake inhibitor of both serotonin and norepinephrine, and is most similar in this regard to venlafaxine (Effexor).  

In a study of medication treatment options for severe depression, a switch to duloxetine was compared with a dose increase of escitalopram.  The escitalopram group had better results, including a remission rate of 54% for escitalopram vs. 42% for duloxetine.  ( )

Another similar comparative study also favoured escitalopram 10-20 mg daily over duloxetine 60 mg, both in terms of effectiveness and side effect profile.  In this study 2% of the escitalopram group dropped out due to side effects, compared to 13% of the duloxetine group.
( )

In this well-done 2011 review by Schueler et al. comparing venlafaxine and duloxetine with SSRIs.  They concluded the following:
1) Venlafaxine had superior efficacy in response rates but inferior tolerability to SSRIs
2)  Duloxetine did not show any advantages over other antidepressants and was less well tolerated than SSRIs and venlafaxine.
( )

Another study, done in 2006, also showing evidence that venlafaxine is superior to duloxetine:

In one of the few well-designed comparative studies of venlafaxine vs. duloxetine, done by Perahia et al (2008), the two medications are found to have similar effectiveness, but with a higher dropout rate due to side effects in the duloxetine group. ( )   A look at graphs of symptom change show that the two medications appear identically effective.  But duloxetine caused more side effects, especially nausea.  It is true that discontinuing venlafaxine causes more side effects than discontinuing duloxetine, but this could be framed as a technical matter that just needs to be managed by very slow tapering.

This 2012 study (sponsored by the manufacturer!) by Martinez et al ( ) compares duloxetine with SSRI treatments for major depression, in a 12 week prospective trial.  Duloxetine performed well, but on the primary outcome measure there was no significant difference in response or remission rates.  On secondary measures there appeared to be some advantages for duloxetine, particularly for pain symptoms.  But the study was not intended to be for treating pain syndromes!  SSRIs are known to be ineffective for pain!
 Duloxetine is often touted as a good treatment for neuropathic pain.  And numerous studies do show that it can help.  But how does it actually compare to other options?  Specifically, how does it compare with a much cheaper and similar antidepressant, venlafaxine?   Rudroju et al (2013) looked at comparative effectiveness of various medications for treating neuropathic pain.  Many medications helped, including duloxetine.  But in this study, gabapentin and venlafaxine had the best odds ratio of helping, followed by pregabalin. Duloxetine was farther down the list.  With a benefit-risk analysis, which takes into account side effects and tolerability, gabapentin, pregabalin, and venlafaxine were once again at the top of the list of best agents, with duloxetine farther down.

Duloxetine (Cymbalta) costs about $4.23 for a 60 mg dose, compared to $0.38 for an equivalent 150 mg dose of Effexor XR.  So it is about 10 times more expensive than an alternative which is shown to work as well if not better. 

In conclusion, Cymbalta is yet another newer antidepressant which is not necessarily better than alternatives; in fact the alternatives such as Effexor XR are probably equally effective or more effective.   It is marketed intensely as a treatment for neuropathic and other pain syndromes, but alternatives such as Effexor XR work better, with fewer side effects, at a lower cost.   Therefore, just as with the other antidepressants mentioned in the previous posts, Cymbalta could be considered a third-line option, which might suit some people well if they have tried other things unsuccessfully.


Vortioxetine is one of the newest antidepressants on the market, released in the U.S. in 2013.  It has serotonin and norepinephrine reuptake inhibition effects, plus a variety of direct effects on serotonin receptors. 

This is a negative study of vortioxetine, showing that it did not lead to any difference in rating scores compared to placebo, when used at doses of 10 mg or 15 mg daily, to treat depression for 8 weeks:

In another study, by Jacobson et al (2015), looking at doses of 10 mg or 20 mg daily, they found slight improvements in the vortioxetine groups compared to placebo, with "significant" differences in the MADRS score only for the 20 mg dose ( ).  If you look at the symptom changes vs. placebo on a graph, the clinical relevance of the vortioxetine effect appears questionable.  Yet, typically with papers of this type, despite the results being very unimpressive, the authors try to frame it in a very positive way, as though they had discovered a fantastically effective new treatment.  Vortioxetine is supposed to be helpful for managing sexual side effects as well, but the measures of this done in the study once again do not show a spectacular benefit.  For those who did not have sexual side effects previously, about half in the vortioxetine group developed sexual side effects, at a rate 10-20% greater than placebo.   Here are the authors' final assertions at the end of their paper:   "In conclusion, vortioxetine 20 mg significantly reduced MADRS total score at 8 weeks in adults with MDD. Overall, vortioxetine was well tolerated in this study."   Perhaps a more fair conclusion could be "vortioxetine produced small differences compared to placebo in the MADRS score, but only at a dose of 20 mg daily.  The degree of improvement does not compare favourably with similar studies using other antidepressants.  Rates of side effects, including sexual side effects, were higher in the vortioxetine groups compared to the placebo groups."   

A 2015 meta-analytic review paper by Rosenblat et al ( ) showed in general that antidepressants appear to help with cognitive function when used to treat depression.  But they conclude that "no statistically significant difference in cognitive effects was found when pooling results from head-to-head trials of SSRIs, SNRIs, TCAs, and NDRIs."

In this article by Llorca et al (2014), which is a "meta-regression analysis", it appears to favour vortioxetine as being better than other antidepressants.  ( article is then quoted elsewhere, such as on Wikipedia, as supporting the claim that vortioxetine is a superior antidepressant.  But the article shows indirect information only, there is no actual comparative study referred to at all.  And the findings, even from this study, really only show that vortioxetine is in the "same ballpark" in terms of effects, compared to other agents-- it certainly doesn't show superiority.

It was hoped that vortioxetine might help with generalized anxiety, but after several negative studies (, ), the latter of which showing that it was significantly inferior to another antidepressant (duloxetine), it is no longer claimed by anyone that it is an appropriate treatment for GAD.

Vortioxetine costs about $3.25 for a 20 mg dose.  This is about 10 times more than a 20 mg dose of citalopram.  

In conclusion, vortioxetine is another new option for treating depression.  It could be something to think about for treating anxious depression.  But there is no evidence that it is superior to other options, and is probably inferior in many cases.  There is no evidence of any specific benefit for treating anxiety disorders such as GAD.    I would consider it to be a third-line alternative at this point. 

Sunday, November 8, 2015

Desvenlafaxine (Pristiq)

Desvenlafaxine (Pristiq) is an antidepressant that has been available since 2008-2009.  It is another example, similar to escitalopram, of a new drug being marketed which is simply a chemical "tweak" of another very similar drug.  Pristiq is an active metabolite of another common antidepressant, venlafaxine (Effexor).  Effexor had been on the market since 1993.

Being new, many studies were done, usually comparing it with placebo, showing that it works.  Yet, very few studies were done comparing it with other antidepressants.

Laoutidis and Kioulos (2015) have recently published a review and meta-analysis of desvenlafaxine.
They found that while it clearly works better than placebo in short-term trials, it is significantly inferior to other agents in comparative studies (i.e. those studies in which desvenlafaxine is compared with a different antidepressant prospectively).

In a 2014 study by Maity et al., desvenlafaxine and escitalopram were found to be equally effective (actually with a non-statistically-significant" edge favouring escitalopram) for anxious depression.  But in this study, it caused more side-effects than escitalopram.

Soares et al (2010) similarly showed no advantage to using Pristiq instead of Cipralex for treating depression in post-menopausal women.  Once again, Cipralex had a non-statistically-significant advantage in effectiveness over Pristiq.

(to be clear about what I mean by "non-significant," it is important to know that all statistical findings are probability statements.  "Significant" usually refers to a finding which has a less than 5% chance of being due to random variation alone.    For many findings, one measure might exceed another, but with a higher than 5%  likelihood of the difference being due to chance.  It should be considered, though, that from a Bayesian point of view, if you have results which differ, even at a so-called "non-significant" level of confidence, this finding still increases the likelihood somewhat that there is a significant difference.  For example, if we toss a coin 10 times, and find that we get 7 heads (instead of the expected 5), we know that there is 17% chance of getting 7 or more out of 10 heads from a fairly balanced coin.  Thus this would be "non-significant" with respect to showing that the coin was not fairly balanced.  But even so, if one indeed did see 7 heads in 10 tosses, it should increase one's suspicion (in a quantifiable way) that the coin is actually imbalanced.  Thus, one should not entirely dismiss "non-significant" results, they should optimally be considered in a large fund of data about an issue, each part of which should reasonably sway our judgment slightly)

In this interesting study by Liebowitz et al (2013)
( ), Pristiq was offered at two different doses (10 mg and 50 mg), compared with placebo, for treating depression.  Both doses were superior to placebo, but were equally effective to each other!   Yet, the "recommended dose" is 50 mg.  Pristiq is only available in 50 mg and 100 mg tablets!  

Pristiq costs about $3.00 per 50 mg pill.  A similar drug, Effexor XR, costs $0.75 for a similar dose.  Celexa at an equivalent dose costs $0.27, according to Pharmacy Compass (

So I do not see any reason to recommend Pristiq, except as one of a list of alternatives to try after other options have been tried.  There is no reason to expect that it would work better than any other antidepressant, unless a particular person just happens to prefer it (as is sometimes the case).  There is evidence to suggest that it has more side effects than alternatives.  I do not necessarily think it is a bad drug though:  I'm sure that there are some who might try it, and find it very helpful after exploring other options.  But based on current evidence it should not be included as a first-line agent. 

Escitalopam vs. Citalopram (Cipralex or Lexapro vs. Celexa)

It is interesting how professional opinion can be swayed by trends in practice.   Escitalopram (Cipralex, or Lexapro) is a newer antidepressant than citalopram (Celexa).   In fact, citalopram itself is a mixture of "enantiomers," which are molecules that are identical to each other except for being mirror-images of each other geometrically.  In many chemical processes, different enantiomers are formed in fairly equal amounts, as a mixture.   But escitalopram, unlike citalopram, consists of just one of these entantiomers, rather than being a mixture.   Citalopram is literally a mixture of escitalopram with an inactive enantiomer.  Therefore, you literally are taking escitalopram when you are taking citalopram.  You are also taking the inactive enantiomer of escitalopram. 

Here we have it again, that escitalopram has more recently been on patent, while citalopram has been available in a generic form for a longer time.  Of course, there would be many more industry-sponsored research studies done recently on escitalopram. 

There's no doubt about it, that escitalopram can be a good antidepressant.  But many professionals (including in one formal instructive report I recently read), assert that escitalopram is clearly "better" than citalopram.

I think this belief is mainly due to cognitive biases.   There has been much more marketing favouring escitalopram in the past decade.  The trends in practice among psychiatrists tend to favour the personal belief that "escitalopram is better."  Because it is used more often these days than citalopram, any positive report about escitalopram is likely to be more salient.  Also, with recurrent trials of antidepressants, any switch to almost any new agent has a reasonable probability of leading to some improvement, irrespective of the properties of the new agent.  For many people, a given antidepressant does not work well enough.  In this cohort, it is much more likely that a given person would have tried citalopram at some point in the past, and would now be looking at trying escitalopram.  There might be about a 30% chance of the escitalopram helping in this scenario.  For the thousands of people in this group, there would then be hundreds who would have the experience of escitalopram appearing to work better than citalopram.   This feeds the notion that escitalopram is in fact a better antidepressant.

The bias here is that very few people in this cohort would have tried escitalopram first, then tried citalopram later on.   This is because escitalopram is newer, more highly marketed, and is more likely to be used when other antidepressants have not worked.   But the prevailing evidence is that most any new antidepressant (or other therapy) trial has a similar chance of helping, when a previous trial has not helped.  Therefore, I predict that there would be an equal likelihood of citalopram working when escitalopram failed, compared to escitalopram working when citalopram failed.  It is possible that the only reason escitalopram appears to work more commonly is that it is simply used more often!

Some of my patients, over the years, have tried both of these medications.  Some have ended up preferring escitalopram.   Others have ended up preferring citalopram.  For most, there has been no difference, either in side effects or effectiveness. 

Are there any recent research studies which compare the two?  One recent study, by Li et al (2014), reviews and pools results from 3 previous clinical studies.  They conclude that there is no difference in response or remission rates between escitalopram and citalopram:

It is interesting to look at the data from previous studies, including a Cochrane review done in 2012, which conclude that escitalopram is better than citalopram:   The authors slip in the caution that "As with most systematic reviews in psychopharmacology, the potential for overestimation of treatment effect due to sponsorship bias and publication bias should be borne in mind when interpreting review findings."  Yet the reader of this article is left with the impression that escitalopram is much better than citalopram.  

I note that escitalopram is about 30% more expensive than an equivalent dose of citalopram, according to PharmacyCompass, a Canadian service which helps people find the best local prices for medications at local pharmacies.   

In conclusion, I think that with respect to antidepressant choice, there is no doubt that escitalopram is appropriate and works at least as well as other available medications.  But it is not necessarily true that escitalopram is "better."  The problem with this biased view of "betterness" is that it could cause a person (a psychiatrist or patient) to overlook other options, and favour escitalopram as a first choice automatically, and unnecessarily.  It could also cause many to overlook citalopram as a possibility for someone who has unsuccessfully tried escitalopram in the past.

Monday, April 27, 2015


Here's another update of this post, to account for studies between 2009 and 2015.

Marijuana use is quite common in the university population I see in my clinic.

It is my opinion that sporadic recreational marijuana use is less dangerous than alcohol use, for many people.  For others, it is more problematic, and the risks may be underestimated. 

Cannabis is an acute intoxicant, which could make activities such as driving much more dangerous. Also, smoking marijuana undoubtedly causes harm to the lungs, though probably not quite to the same degree as smoking tobacco cigarettes (see references below).

There is strong evidence that marijuana use increases the risk of developing a psychotic disorder, probably by about 40%.

People who have a psychotic illness, or who have a family history of psychotic illnesses, are at higher risk for having new or continuing psychotic symptoms if they use marijuana.

Also, based on some of the evidence cited below,  children and adolescents are probably much more vulnerable to negative, long-term emotional and cognitive effects from marijuana use. 

Many regular consumers of cannabis have problems with motivation. This may be reflected in poor grades in school, lack of success in building a career, etc.  This is possibly a non-causal association, but if someone has low motivation to begin with, the addition of cannabis is not likely to help.

There may be some selected exceptions.  For example, some have claimed that a culture of cannabis use has had a catalytic role in helping reclusive technical geniuses relax their social and creative inhibitions, to permit some examples of very successful scientific and business innovation, such as in Silicon Valley.

There is strong evidence that marijuana use is associated with more severe psychiatric symptoms, of almost every type; but much of this association could be due to the fact that those with more severe symptoms are more likely to use marijuana, not the other way around. In any case, those who choose to use marijuana more regularly as a cultural pursuit may be surrounding themselves with others who have more severe symptoms.   This is similar to the case  of alcohol:  part of the harmful effect of drinking heavily is due to proximity to places (such as rough bars) where there are a lot of other heavy drinkers -- in this environment, there is likely to be more physical danger, and much less breadth of social or cultural opportunity.  Ironically, decriminalization should probably reduce this effect, and therefore reduce some of the potential social harms.

There is some evidence that marijuana or other cannabinoids could be helpful to treat a variety of medical ailments. This evidence needs to be taken seriously.

Here is a brief survey of the very large literature on this subject:

Evidence of Risk and Harm

     Psychiatric Risks

This 2007 review from Lancet shows convincing evidence that marijuana use increases the risk of developing a psychotic disorder, and that the risk is dose-dependent (i.e. the more marijuana one uses, the higher the risk is of developing a psychotic disorder):

It concluded that the evidence is less clear linking marijuana to other problems, such as depression and anxiety: many of the studies looking at this did not sufficiently address non-causal reasons for the association between marijuana and other problems. For example, people who are more depressed or anxious may have a higher likelihood of using marijuana to treat their symptoms. Or, people whose cultural style may lead them away from conventional treatments for depression, may be more likely to use marijuana regularly.    Use of psychotherapy and antidepressants are also more common among those with depression, but this does not prove that psychotherapy and antidepressants cause depression! 

In this 2008 review from the British Journal of Psychiatry, the authors conclude that marijuana use is associated with worse outcome in psychotic disorders--but they say that the existing studies show only an association, not causality. Once again, confounding variables may cause this association to exist:

A significant cannabis withdrawal syndrome is described in the literature, particularly for heavy, long-term users. The syndrome involves about 2 weeks of irritability, restlessness, and insomnia, which could be quite destabilizing for someone struggling with mood symptoms, therefore leading to continued marijuana/cannabis use. Here is a 2006 review of the subject:

A few recent prospective studies have demonstrated increased dysphoria, anxiety, tiredness, ideas of reference, and schizotypal symptoms as a result of marijuana intoxication. In particular, individuals with pre-existing schizotypal personality traits had a more substantial increase in schizotypal symptoms following THC exposure. This adds to an evidence base suggesting that marijuana use carries a significant risk of exacerbating a variety of psychiatric symptoms, particularly psychosis-spectrum symptoms, and particularly in those with risk factors for psychotic illness.
Here are the references, which are both from Psychological Medicine in 2009:

This interesting study involved administration of THC to healthy volunteers who did not use THC.  Some members of the cohort experienced transient psychotic phenomena, while others did not.  These differences were associated with differences in cognitive impairment and functional MRI results. This supports the common-sensical observation that some individuals may be more vulnerable than others, to having adverse neuropsychiatric effects from THC use.

Many other studies looked at populations who used different amounts of marijuana over time, and compared them in terms of various symptoms and intellectual functions, etc. Unfortunately, I find this type of retrospective analysis to be weak, and highly prone to confounding variables. In order to understand marijuana's long-term effects for sure, we would need to do a long-term, prospective, randomized, controlled study.

     Physical Risks

Here are some studies looking at risk to the lungs associated with marijuana smoking:

These studies show an increased risk of lung cancer in marijuana smokers:

These studies show a likely causal association between long-term marijuana smoking and obstructive lung disease:

     Prospective Animal Studies

Animal studies could add a little bit more information into the picture, since these have been done in a prospective, controlled fashion. Here is what I've found from the animal research literature:

This study showed that chronic marijuana exposure impairs spatial memory & learning in rats:

This study showed that chronic marijuana exposure impairs social and cognitive functions in rats, but especially when the period of exposure is during the pubertal ("adolescent") phase of development:

Another study showing that marijuana exposure may be particularly harmful to the "adolescent" brain in rats:

This study from UBC suggests that high-dose cannabinoids increase emotionality and "sensitize the stress axis" in rats:

Evidence of Benefits or Therapeutic Uses

This study shows that a synthetic cannabinoid promotes neurogenesis in the hippocampus, and may have antidepressant and anxiolytic effects:

Here is a reference to a good 2008 review of the pharmacology and potential therapeutic applications of cannabinoids such as marijuana:

     Neurological Diseases

Here's a 2012 study showing relief in muscle stiffness in multiple sclerosis patients, due to cannabis administration:

Another 2012 study from CMAJ showing relief of spasticity and pain in MS patients, following cannabis administration:

This study shows immediate relief of the symptoms of Parkinson's Disease following cannabis treatment:

     Bowel Disease

This study, from a major journal of gastroenterology, shows that cannabis dramatically improved symptoms  of Crohn's disease (a type of inflammatory bowel disease), in a prospective, placebo-controlled trial.

Another prospective study, showing that cannabis improves quality of life in inflammatory bowel disease:

     Pain Disorders

Here's a good 2013 study showing that cannabis compares favorably with other standard pharmacological treatments for neuropathic pain:

     Heart Disease

This 2005 study from the prestigious journal Nature suggests that cannabinoids could reduce the progression of atherosclerosis (the main cause of heart disease):

In conclusion, I think that marijuana use is dangerous, and harmful to your health in a variety of ways, due to acute intoxication, increased risk of psychosis, possible cognitive side-effects, and lung damage. It may be particularly harmful to adolescents. As a cultural pursuit, it may distract people from other life activities, or meaningful life roles, just as any habit or addictive behaviour can. But it may have beneficial effects for a variety of medical problems.

I have to admit, to be fair, that some people have psychological benefits from marijuana use -- certainly there are many testimonial accounts of this, but evidence beyond this is not clear on this point.  The few studies touting this application tend to be of short-duration, which leads to a similar criticism as that pertaining to mainstream pharmaceuticals:  short-term benefits for symptom relief do not always translate into long-term benefits, if the use continues for years.   More research is needed to gain a better understanding of the potential risks or benefits of cannabinoids, especially over longer-term use.

I have certainly seen people for whom cannabis appears to have a better benefit:risk profile than alternative treatments, for example to treat chronic pain symptoms and associated insomnia.    It may be preferable to use cannabis instead of a benzodiazepine, opiate, pregabalin, etc., particularly if these latter agents are causing a much higher load of side effects in a given person.

For some people, cannabis could be a relatively harmless entertainment, or even a catalyst for enjoying life more richly in various settings.  In this way, it could be analogous to having a glass of wine with meals, etc.

Another angle to the analysis is to consider relative risks of cannabis compared to other accepted intoxicants, such as alcohol.  With this type of risk analysis, one could often see greater risks with alcohol compared to cannabis, on a case-by-case basis, but we don't have good group data on this.  Suppose we had two adjacent similar countries, and prospectively allowed free access to alcohol in one country, and free access to cannabis in the other.  Then, suppose we were to assess health outcomes in these countries 20 years later.  I suspect we would have more examples of ruined families, criminal assaults or manslaughter, chronic diseases, and traffic fatalities, in the "alcohol" country compared to the "cannabis" country. 

The issue is complicated by the fact that those who are more apt to use cannabis are statistically also more apt to use alcohol and other street drugs.  It is possible that cannabis use could have "gateway" effects, leading people into a higher-probability zone of trying or using more dangerous drugs. But this is an open question. 

A proliferation of cannabis dispensaries have appeared in Vancouver in the past year.  While I do think that legalization is a positive step, in terms of the various pros and cons for public health,  I am not happy with the idea being touted by some, that cannabis is some kind of health food, or panacea.   There is an issue of cultural freedom as well, which I support, though I think that many in this "4-20" movement have an exaggerated view of the benefits of cannabis, with an underestimation of risks.