Duration:  18 minutes, 35 seconds

KELLIE NEWSOME: Should we follow the FDA and give SSRIs for PTSD, or venture off-label with the fabled prazosin? Experts are divided, and today we’ll look at both sides of the argument.

CHRIS AIKEN: Welcome to the Carlat Psychiatry Podcast, keeping psychiatry honest since 2003. I’m Chris Aiken, the editor-in-chief of The Carlat Report. 

KELLIE NEWSOME: And I’m Kellie Newsome, a psychiatric NP and a dedicated reader of every issue.

CHRIS AIKEN: We wrap up our PTSD series today with a focus on medications. We saved it for last – not because it’s the best – but because meds are generally not as effective as therapy here. I’d start with that message for patients, as we already live in a culture that prioritizes meds over all else in mental health. But if they still prefer medication, or, haven’t had success with therapy, I’ll start it. People tend to respond best to the treatment they prefer.

KELLIE NEWSOME: For first line treatments, we usually go with the FDA, but we have some reasons to second guess to FDA’s picks in PTSD. There are two FDA approved options: sertraline and paroxetine. But these SSRIs have problems. Sertraline only worked in 2 out of 7 trials, and some analyses find it only effective in women with civilian trauma. For paroxetine, the data is more robust, but the side effects are worse. This one has some of the highest rates of withdrawal problems, fatigue, sexual side effects, and anticholinergic effects, and it has the worst pregnancy safety of any SSRI.

CHRIS AIKEN: If you’re going with an SSRI, I prefer fluoxetine. It may be off-label, but it has more studies in PTSD than paroxetine and a larger effect size than sertraline, based on two meta-analyses. Fluoxetine also has a few tolerability advantages. Compared to other SSRIs, it has a lower risk of withdrawal problems and weight gain. The dose in PTSD is 20-40 mg/day. For paroxetine, the dose is similar, 20-50 mg/day, while sertraline has a wider range in PTSD of 50-200 mg/day.

KELLIE NEWSOME: Another serotonergic option for PTSD is venlafaxine 75–225 mg/day. This SNRI is about as effective as fluoxetine and paroxetine, but we’re a bit less likely to start with it because like paroxetine it has a high risk of withdrawal symptoms. Some patients respond really well to SSRIs, particularly women and those with civilian trauma, but if the response isn’t meaningful or the side effects are too meaningful, we’ve got a lot of options up ahead in this episode. But, before we lose our CME eligibility going off-label in this podcast, let’s pause for a preview of the CME test:

1. Which medications are FDA approved in PTSD?

A. Sertraline and fluoxetine

B. Venlafaxine and fluoxetine

C. Paroxetine and venlafaxine

D. Sertraline and paroxetine

CHRIS AIKEN: PTSD is an area where I deviate from the FDA. Not only do I prefer fluoxetine over sertraline and paroxetine, but I often use prazosin first-line for PTSD. Why? Because it doesn’t just treat nightmares; it helps daytime symptoms of PTSD as well. It works in military as well as civilian trauma. Its studies may not be as large as those for SSRIs, which means we’re not as certain in what we say about prazosin as we are with the SSRIs. But while our knowledge of SSRIs may be certain, it’s not very encouraging. Their effect size is small – 0.3 – some like sertraline didn’t consistently work – and they cause problems that I don’t like to see in PTSD. Problems like sleep disruption, apathy, sexual dysfunction – think about that for people who are trying to rebuild their romantic life after a sexual trauma. These are subtle, long-term problems that don’t tend to raise alarms in acute trials, giving number crunchers the impression that these meds are better tolerated than they are.

KELLIE NEWSOME: Prazosin has been around since 1988, so we are pretty certain about its safety and tolerability. The main risk is low blood pressure. This is serious enough that it earned the drug a black box warning about syncope and falls even when treating high blood pressure. In that population, the FDA estimates about 1% of people have syncope on prazosin, and the rate is particularly high after the first dose so you’d want to start low and use fall precautions while starting it. In the PTSD trials, falls were not more common than placebo, but there was a higher risk of orthostatic blood pressure changes with prazosin – we’d estimate this is about 4% risk based on the largest trial.

CHRIS AIKEN: Prazosin’s psychiatric benefits were discovered in the late 1990s by Murray Raskind, who was working at the University of Washington VA when he noticed a striking pattern. A few of his group therapy patients had dramatic improvement in PTSD nightmares after starting prazosin for benign prostatic hypertrophy. Prazosin treats that by relaxing smooth muscle and improving urine flow. From those early observations, Raskin went on to confirm prazosin’s benefits – not just for nightmares but for daytime symptoms of PTSD as well – in multiple randomized controlled trials that were replicated by independent groups. 

This may seem like a serendipitous discover, the kind that any of us could have made by paying close attention to our patients. But as Louis Pasteur put it, Chance favors the prepared mind, and Raskin’s mind was prepared. He had studied prazosin’s effects on the hypothalamus in basic science research and probably had a keener eye than most of us when his patients started taking the adrenergic antagonist.

KELLIE NEWSOME: As the positive trials rolled in, nearly 2 dozen in all, there was a problem. Most were small, involving fewer than 100 patients each. So, Dr. Raskin launched a more definitive trial. This one would enroll 304 patients across 13 VA hospitals. Prazosin was titrated over 5 weeks to an average daily dose of around 15 mg per day, so a respectable dose, and they kept it going for a respectable time as well – 6 months in all. But at the end of the trial, there was no significant difference on the medication – not for sleep or nightmares, not for daytime symptoms, and not even for secondary measures. There was less suicidality in the prazosin group, an encouraging finding given that another adrenergic antagonist – trazodone – was associated with an increased risk of suicidality in the veteran population.

CHRIS AIKEN: That study came out in 2018 in the New England Journal of Medicine, and for many, it signaled the death knell for prazosin. Others rushed in to explain the results. Military trauma doesn’t respond as well to treatment, but prazosin has worked in lots of studies of military trauma – something we can’t say for the SSRIs which largely failed there. There was a high placebo response, and the study criteria biased the sample toward milder cases who may be as likely to get better with placebo as with prazosin. They also required that patients come off trazodone to start the trial. That medication is widely used in the VA, so it biased the sample against trazodone responders – which is a problem – because trazodone and prazosin share a common mechanism: blocking adrenergic receptors. 

But it’s not enough to make excuses. To prove that prazosin works, we need a large trial showing that, and we don’t have that for you. In place of that, we have meta-analyses published after the negative trial, and even though that large trial casts a heavy shadow over the results, they still turn up positive, with a large effect size for nightmares, a medium-to-large benefit in sleep, and a small benefit – similar to that for SSRIs – for daytime symptoms of PTSD. Narrowing in on those daytime benefits, prazosin helps – in order from greatest benefit to least - anhedonia, difficulty concentrating, and hypervigilance. It’s possible those benefits are all a result a sleeping better, but if they are, then prazosin is addressing the underlying problem. 

You’ll hear some experts who discourage prazosin, and others who recommend it first line. The difference hinges on that large trial, and on how much weight the expert puts more weight on scientific certainty. For me, we may not be as certain about what prazosin does as we are with the SSRIs, but what we do know points to a more meaningful benefit for patients with a lot less problems, and most patients would rather have, say, a highly effective and tolerable treatment that we are 70% certain of than a minimally effective treatment that we’re 90% certain of, particularly when it comes with risks of weight gain, apathy, sexual side effects, withdrawal problems, bone thinning, withdrawal problems and – like prazosin – SSRIs are associated with falls. 

KELLIE NEWSOME: That’s our bias, and we’re not alone in it. The psychopharmacology algorithm project, lead by David Osser, arrived at a similar conclusion. They recommend starting with prazosin if the patient nightmares or hyperarousal at night, which the vast majority of patients.

After that, Osser recommends SSRIs, and here his concerns about the sertraline’s restricted efficacy and paroxetine’s tolerability also lead him off-label. However, he cautions about jumping into mirtazapine. It’s tempting to think mirtazapine must work in PTSD, as it is one of the best antidepressants for anxious depression and insomnia, but the studies for anxiety or even insomnia in PTSD are limited, mixed, and don’t come close to the level of certainty that even we with these lax biases would demand. 

CHRIS AIKEN: But wait – some of you may have heard that mirtazapine is good for sleep apnea – a common problem in PTSD. If you were thinking that, major bonus points for knowing that mirtazapine relaxes pharyngeal muscles. But, as with many ideas that pass the intuitive test, it was debunked by controlled trials 10-15 years ago. Instead, current studies are looking to clonidine, trazodone, and – a surprising one – atomoxetine as leading pharmacologic candidates to open the airways at night. Of those, clonidine and trazodone may benefit sleep and nightmares in PTSD, but this little aside reminds us of an important warning. A recent study from 2019 found that patients with hypertension who use alpha1 antagonists like our favorite prazosin have a higher risk of sleep apnea. That study was not controlled, so we’ll hold on any serious warnings, but await more on this complex and counterintuitive area.

KELLIE NEWSOME: Prazosin may prevent PTSD – it did so in a small trial of acute stress disorder, but that study was too small to put much weight on. It may also help migraines associated with PTSD. That’s also from a small trial, published last year by Raskind’s group. When nightmares don’t improve with prazosin, other options with smaller studies include clonidine and prazosin’s cousin – doxazosin.

Here’s how to use prazosin. Start at 1 mg QHS and raise by 1–2 mg every four to seven days based on response and tolerability. Once you reach a daily dose of 3-5 mg, you can start giving some of the dose – around 25% of it – in the morning. The target dose is much higher than most clinicians reach. Average doses were 12-16mg in most trials, and the max dose is different for women and men –likely because of different cardiovascular responses. Those are a max of 25 mg/day for men and 12 mg/day for women. Check blood pressure and pulse and monitor for falls during titration. If a patient is taking other antihypertensives, consult with the provider managing those and consider tapering them after titrating prazosin.

CHRIS AIKEN: If your patient has alcohol problems, prazosin is less likely to work, but here’s a tip. Prazosin does reduce alcohol use when combined with cyproheptadine 8-12 mg at night. The combo significantly reduced alcohol use in a recent controlled trial, and it’s likely they have a synergistic effect because neither of them reduce alcohol when used on their own. One caveat: the trial did not involve PTSD. Another option for PTSD with alcohol use disorder is topiramate, and we’ll get more into that and other off-label options in next week’s episode.

But first, a few points of clarification

1. We mentioned a 1% syncope risk with prazosin. That is when starting at 2 mg, and the PDR recommends starting lower to reduce that risk – that’s why Kellie listed a starting dose of 1 mg.

2. Dr. Osser’s guidelines for PTSD don’t exactly recommend antidepressants 2^nd line. They recommend starting with prazosin when the patient is having nightmares or nocturnal hyperarousal, which is most patients, and starting with an SSRI if those aren’t pressing problems.

Finally, we’d like to acknowledge an error in our August 5 podcast on Research Updates. We mentioned that iloperidone – which was recently approved for bipolar mania – may have a big effect in that disorder because of its small p-value. P-values are a marker for certainty, not for efficacy – they tell us how certain we are that a drug works. You can get a small p-value by having a big effect – as big effects are easy to detect – but you can also have it by getting a small effect, such as when the study has a very large population or the results have a very narrow statistical spread or standard deviation. To correct our error I’ve calculated the effect size for iloperidone in mania from their phase III trial of 414 patients, and indeed it is small, 0.22. So, while we’re certain it works, we are also certain it doesn’t work very much.

KELLIE NEWSOME: Want to keep up with the latest in psychiatric research? We post new studies in the Daily Psych feed – search for Chris Aiken MD on LinkedIn, Twitter, Facebook, and that new one – Threads. It’s a first glimpse of the trials that inform this podcast. Thanks for tuning in and helping us stay free of industry support.

The Carlat CME Institute is accredited by the ACCME to provide continuing medical education for physicians. Carlat CME Institute maintains responsibility for this program and its content. Carlat CME Institute designates this enduring material educational activity for a maximum of one quarter (.25) AMA PRA Category 1 Credits^TM. Physicians or psychologists should claim credit commensurate only with the extent of their participation in the activity.