Figure 1. Change in Scores of Primary Outcome Measures from Baseline.The three primary outcome measures were the change in score from baseline to 10 weeks on the Clinician- Administered PTSD Scale (CAPS) item B2 (scores range from 0 to 8, with higher scores indicating more frequent and more distressing dreams) (Panel A), the change in Pittsburgh Sleep Quality Index (PSQI) score from baseline to 10 weeks (scores range from 0 to 21, with higher scores indicating worse sleep quality) (Panel B), and the Clinical Global Impression of Change (CGIC) score at 10 weeks (range, 1 to 7, with lower scores indicating greater improvement and a score of 4 indicating no change from baseline; the CGIC assessed the participant’s ability to function in daily activities and the participant’s sense of well-being) (Panel C). Over the entire 26 weeks, the mean change from baseline on the CAPS item B2 was −2.0 (95% confidence interval [CI], −2.1 to −1.8) with prazosin and −2.1 (95% CI, −2.2 to −1.9) with placebo (P = 0.56); the mean change from baseline on the PSQI was −2.6 (95% CI, −2.9 to −2.3) with prazosin and −2.6 (95% CI, −2.9 to −2.3) with placebo (P = 0.99); and the mean CGIC score was 3.1 (95% CI, 3.0 to 3.2) with prazosin and 3.1 (95% CI, 3.0 to 3.2) with placebo (P = 0.86). I bars indicate 95% confidence intervals.
During my internal medicine residency, I was privileged to be the primary care physician for many veterans. Practicing in a Veterans Affairs (VA) medical center is unique in many ways. Besides the predominantly male population, many patients suffered from mental illnesses with post-traumatic stress disorder (PTSD) as a common culprit. One memorable young patient struggled with persistent sleep disturbance. When a consulting psychiatrist recommended treatment with prazosin, I thought it was a curious choice because alpha blockers sat firmly in my mental buckets of “not great anti-hypertensives” and “helps with urination.” I subsequently learned that alpha-1 antagonists can lower the high adrenergic activity that is thought to drive PTSD symptoms. Further, prazosin crosses the blood-brain barrier quite effectively, and evidence from small randomized clinical trials (RCT) gave us hope that it would work.
Consequently, I read with great interest the results from the PACT trial published in this week’s NEJM. In this largest RCT to date, investigators randomized 304 veterans with PTSD and frequent nightmares from 12 VA centers to receive prazosin or placebo for 26 weeks. During that time, existing therapies were continued but no new pharmacotherapy and psychotherapy could be added.
Unfortunately, the results of this trial did not show any benefit from prazosin for the three primary outcomes: Change from baseline at 10 weeks in the recurrent distressing dream component of the Clinician-Administered PTSD Scale (CAPS; score range, 0–8), the Pittsburgh Sleep Quality Index (PSQI; score range, 0–21), and the absolute score of the Clinical Global Impression of Change (CGIC; score range, 1–7). In general, higher scores indicate worse symptoms for all three scoring systems. The mean between-group differences in the change from baseline for the CAPS recurrent distressing dreams component and PSQI were 0.2 (95% CI, -0.3 to 0.8; P=0.38) and 0.1 (95% CI, -0.9 to 1.1; P=0.80), respectively. The mean difference in the CGIC was 0 (95% CI, -0.3 to 0.3; P=0.96). Of the adverse events, dizziness (34% vs. 21%), lightheadedness (34% vs. 20%), and urinary incontinence (12% vs. 4%) were more common in the prazosin group, and new or worsening suicidal ideation was less common in the prazosin group (8% vs. 15%).
With any clinical trial, the devil is in the detail. The authors offer some possible explanations for the lack of benefit of prazosin, primarily selection bias. At the time of the study, prazosin was already available for off-label use in the VA medical system and was often prescribed as an adjunct to first-line selective serotonin reuptake inhibitors (SSRIs), as with the patient I described above. Therefore, clinicians may have been hesitant to enroll patients with more severe symptoms and risk for clinical deterioration in a randomized trial with the possibility of receiving placebo when prazosin was available outside of the trial. Furthermore, patients treated with trazodone were excluded from the trial because of its alpha-1 antagonist activity, thus removing a population of potential responders.
In an accompanying editorial, Dr. Kerry Ressler from Harvard Medical School reminds us that PTSD is a heterogenous disorder with different manifestations of symptoms. Perhaps not every patient shares the same adrenergic hyperarousal response that is thought to be the driver of nightmares. He encourages the development of better biomarkers to identify the phenotype of patients who may respond to prazosin.
I left the VA before I had the chance to see whether my patient benefited from prazosin, but other patients have told me that prazosin is the only intervention that lets them sleep at night. Although the PACT trial did not show clear efficacy, prazosin may still have therapeutic benefit in select patients.
Browse more From Pages to Practice
Mike is a 2017-2018 NEJM Editorial Fellow and a hospitalist at Beth Israel Deaconess Medical Center. He graduated from Harvard Medical School and completed his internal medicine training at BIDMC.