Prazosin didn't beat placebo for PTSD nightmares in its largest trial

Prazosin, the medication most associated with treating the nightmares of post-traumatic stress disorder, was no better than placebo in the largest and most rigorous trial of it — 304 military veterans followed for 26 weeks [s1]. The treatment that sleep medicine formally recommends for nightmare disorder is not a drug at all but a brief behavioural therapy, image rehearsal therapy [s2].

That gap — between the drug people reach for and the therapy the evidence points to — is the useful thing to know about a common and distressing problem.

The trial that changed the picture

Earlier, smaller randomised trials had found prazosin, an alpha-1-adrenoreceptor blocker, effective against PTSD-associated nightmares, which is how it became a mainstay [s1]. The PACT trial was built to test that finding at scale. It recruited veterans with chronic PTSD and frequent nightmares from 13 Department of Veterans Affairs medical centres and randomly assigned them to prazosin or placebo: 152 to each [s1]. The drug was escalated over five weeks to a daily maximum of 20 mg in men and 12 mg in women, and continued double-blind to 26 weeks [s1].

At 10 weeks, across all three primary outcomes, prazosin and placebo were indistinguishable. On the Clinician-Administered PTSD Scale item for recurrent distressing dreams, the between-group difference was 0.2 (95% confidence interval −0.3 to 0.8; P=0.38) [s1]. On the Pittsburgh Sleep Quality Index it was 0.1 (95% CI −0.9 to 1.1; P=0.80), and on the Clinical Global Impression of Change it was 0 (95% CI −0.3 to 0.3; P=0.96) [s1]. There were no significant differences at 26 weeks either [s1].

The drug was doing something — just not to the nightmares. Prazosin lowered supine systolic blood pressure by 6.7 mm Hg more than placebo, confirming it was pharmacologically active at the doses used [s1]. On safety, the adverse event of new or worsening suicidal ideation was recorded in 8% of the prazosin group versus 15% of the placebo group [s1].

What the guideline recommends instead

The American Academy of Sleep Medicine's position paper on nightmare disorder — which affects about 4% of adults, either on its own or as part of conditions such as PTSD, and can significantly impair quality of life — reviewed the behavioural, psychological and drug options [s2]. Its categories are deliberately blunt: "recommended" and "not recommended" mark treatments judged clearly useful or clearly ineffective or harmful for most patients, while "may be used" covers everything in between, where the evidence is less settled [s2]. Only one therapy earned the "recommended" position: image rehearsal therapy [s2]. A range of other psychological approaches and medications landed in the "may be used" tier rather than alongside it [s2].

Image rehearsal therapy is a short course of cognitive-behavioural work in which a person, while awake and with a clinician, changes the storyline of a recurring nightmare and rehearses the new version. It asks nothing of a pharmacy, and it is the intervention the field's own body places at the top of the list [s2].

What this settles, and what it does not

It settles that prazosin, tested properly and at scale, did not beat placebo for PTSD nightmares or sleep quality in veterans [s1] — a result that should temper how confidently the drug is offered for that purpose. It does not mean prazosin never helps any individual, and the trial's own authors were testing a population that had already been living with chronic PTSD. Nor does it settle nightmare disorder outside the PTSD context, where the recommended first move is behavioural [s2]. This piece carries a held rating because it concerns a drug, its dosing and a trauma condition; it describes evidence and is not medical advice.

Sources

• Trial of Prazosin for Post-Traumatic Stress Disorder in Military Veterans — New England Journal of Medicine, 2018-02-07
• Position Paper for the Treatment of Nightmare Disorder in Adults: An American Academy of Sleep Medicine Position Paper — Journal of Clinical Sleep Medicine, 2018-06-14