Antipsychotics and VMAT2 inhibitors linked to faster decline in Huntington disease

Design

Post hoc analysis of the placebo arm of a randomised controlled trial

Population

247 adults with Huntington disease, TFC 7–13

Primary outcome

Change in composite UHDRS and Total Functional Capacity

Effect

At week 52, cUHDRS difference 0.66 (95% CI 0.31 to 1.01); TFC 0.85 (0.47 to 1.22), favouring off-drug

This post hoc analysis used the placebo arm of PROOF-HD, a trial of pridopidine in adults with early to moderate Huntington disease. Of 247 placebo participants, 135 were taking antidopaminergic drugs (antipsychotics, VMAT2 inhibitors or both) and 112 were not.

Those on these drugs declined faster across function, cognition and motor scores. At week 52, the difference in change favoured those off the drugs by 0.66 points on the composite UHDRS (95% CI 0.31 to 1.01) and 0.85 on Total Functional Capacity (0.47 to 1.22). Antipsychotic-only, VMAT2-only and higher-dose use were each associated with greater decline. Propensity-weighted analyses pointed the same way.

Psychiatrists are often asked to manage irritability, psychosis and agitation in Huntington disease, and antipsychotics are a mainstay. This is an association: people with more advanced or more symptomatic disease are more likely to be prescribed these drugs, and that confounding cannot be fully removed. It is a reason to use the lowest effective dose and review the indication, not to withdraw needed treatment.

• Review the indication and dose of any antipsychotic or VMAT2 inhibitor in a patient with Huntington disease at each visit.
• Use the lowest dose that controls psychosis, agitation or chorea.
• Consider non-dopaminergic options for irritability or depression where they are adequate.
• Coordinate with the neurologist so motor and psychiatric prescribing are reviewed together.

Why it matters

Drugs widely used for psychiatric symptoms in Huntington disease may carry a functional cost that has been underappreciated.

Don't overread it

A post hoc, non-randomised comparison cannot show that these drugs cause faster decline.

The statistics, in plain English

Both confidence intervals sit clear of zero, so the faster decline among those on the drugs is unlikely to be chance. But because the groups were not randomised to drug use, a sicker group may simply have been the one receiving treatment.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.