Prazosin for Alcohol Dependence

Thanks to Dr. Saxon for sharing this article at the 2013 Fundamentals of Addiction Medicine Conference

A Pilot Trial of the Alpha-1 Adrenergic Antagonist, Prazosin, for Alcohol Dependence.
Simpson, T. L., Saxon, A. J., Meredith, C. W., Malte, C. A., McBride, B., Ferguson, L. C., Gross, C. A., Hart, K. L. and Raskind, M. (2009), Alcoholism: Clinical and Experimental Research, 33: 255–26

The authors had noted reports of decreased alcohol use when prazosin as used for the treatment of PTSD, prompting this pilot study.

24 patients were enrolled (out of 120 screened) and randomized to placebo vs prazosin.
A lot of patients were excluded.  The reasons for not being enrolled were:

not interested once learned what was involved, including not willing to be abstinent (n=37)
current PTSD or other psychiatric condition requiring medication other than an anti-depressant (n=28)
in other treatment or in controlled living environment (n=21)
no drinking past month or no AD (n=12)
scheduling difficulties or no reliable way to contact (n=11)
medical contraindications or already taking prazosin (n=6)
currently using opiates (n=>4)
outside of age range (n=>1)

The prazosin was given at a fairly high dose, 4 mg AM, 4 mg PM and 8 mg HS. They note that all the patients but one was able to titrate up to the full dose. They started at 1 mg HS for the first two days and titrated up to the full dose over two weeks. Patients were seen twice weekly the first two weeks then weekly thereafter.

The medication was well tolerated. Systolic BP decreased 6.6 mm Hg on average and diastolic decreased 3.3 mm Hg. No patient had orthostatic hypotension. 4 patients dropped out, none related to medical effects. The prazosin group reporting taking 92% of their study medication doses, the placebo group 87%.

Patients had brief medical management counseling visits, similar to the COMBINE trial.

Baseline alcohol use was not the most severe. In the 42 days prior to the study there were about 28 drinking days, 20 heavy drinking days and 4-5 drinks per day.

The data analysis is somewhat complex. There was no differences over the 6 weeks in mean drinks per week or mean drinking days per week, although both groups decreased use 2.78 drinks per week.

 If the first 3 weeks are excluded, as the medication was still being titrated up, there were fewer drinking days per week in the prazosin patients than the placebo group (β = −1.22; 95% CI = −2.29 to −0.14; p = 0.027). The average total number of drinking days was 5.9 for the placebo group, and only 3.2 for the prazosin group. There was no difference between the groups in mean drinks per week during those final 3 weeks.

Data was reanalyzed looking just at the male subjects as only three women completed the trial and only one woman was in the control group.  Looking at the just the 17 male study completers (7 prazosin and 10 placebo) efficacy appeared quite impressive.  6 of the 7 prazosin men were abstinent at study completion vs 4 of 10 placebo men.   The placebo men drank on average 20.8 total drinks over the last 3 trial weeks vs 2.6 for the prazosin men.  The placebo men had on average 5.7 and the prazosin men 0.9 drinking days over those 3 weeks.

The authors noted they did not see a decrease in craving for alcohol in the prazosin (vs placebo) patients even as used decreased.

This small pilot study demonstrates a new possible medication option for treatment of alcohol dependence. The authors explain a plausible mechanism or prazosin to decrease alcohol relapse.  The data for the men only subgroup shows impressive efficacy although the larger analysis did not.

With this degree of evidence, I think it would be reasonable to offer prazosin to help male patients reduce drinking if there is also some evidence of PTSD.  I look forward to larger study of prazosin.  The major change in my practice will be to move doses to the higher range demonstrated in this study.