More on “alternative endpoints”

I’ve posted before on “alternative endpoints” for treating opioid use disorders, which is the idea that research on treatments should not just focus on abstinence as an outcome.

There is now a push for alternative endpoints for alcohol use disorders:

Reductions in alcohol use bring about significant improvement in adverse consequences, mental health status, and quality of life, even if the reductions do not reach the level of abstinence, according to recent research presented at the annual meeting of the American Academy of Addiction Psychiatry.

The findings, experts say, demonstrate how a fixation on abstinence or elimination of all heavy drinking – the endpoints the Food and Drug Administration now require in pivotal clinical trials on alcohol use disorder (AUD) treatments – is shortsighted and can unnecessarily discourage people with alcohol use disorder from pursuing treatment.

A fixation on abstinence?

“A fixation on abstinence” sounds like a pretty loaded phrase and I have thoughts. But first, Stuart Gitlow, a past president of ASAM, wrote a response:

I am confused as to whether the entire discussion within the article revolves around alcohol use or whether instead it revolves around alcohol use disorder. And I worry that those advocating for a new endpoint may also have some confusion. I especially am concerned that the referenced article is based on COMBINE data, which itself was quite flawed due to the requirement of a certain volume of alcohol intake for inclusion, thereby studying only a subset of those with alcohol use disorder rather than a representative disease population.

Alcohol use disorder does not have a severity level based upon volume or frequency of alcohol intake. Because of that, the severity of alcohol use disorder does not improve if volume or frequency of intake declines. On the other hand, morbidity secondary to alcohol use disorder may indeed be related to volume and frequency of alcohol intake. And therefore, if volume/frequency declines, morbidity would likely improve, at least for a short period of time.

Similarly, patients with tuberculosis have a cough. The frequency and intensity of coughing is not utilized to determine the severity of TB. However, the coughing can be impairing and can be considered an aspect of morbidity from TB. So here’s the question: If we give something (methadone, perhaps) to improve the cough in TB, resulting in improved morbidity, are we treating the TB? Are we altering the long term TB disease course?

The answer to both is no, of course we’re not.

My thoughts

• This is a predictable outcome of the DSM 5’s shift from a categorical model of alcohol problems to a continuum model.
  • All alcohol problems were put on a single continuum. This put low severity problems under the same diagnosis as high severity, high chronicity alcohol problems.
  • Abstinence has long been considered the most appropriate treatment goal for high severity, high chronicity problems (addiction), while moderation has been considered an appropriate goal for lower severity problems. (See here, here, and here for examples.)
  • This is exacerbating the pre-existing problems that the research too often fails to distinguish between addiction and lower severity substance use problems.
• There are important ways in which this is a nonevent, at least for how medications are used.
  • The evidence-base for the current medications focuses on reductions in drinking, not abstinence.
  • The current medications are not very effective at facilitating abstinence, making the current standard seem more theoretical than practical.
• Do people with low severity alcohol problems (for whom reduced use is a good target) need medications to help them moderate?
• Who will benefit from this?
  • The best case is that patients will benefit. However, it’s hard to imagine that lowering the bar will improve patients’ options when the current meds only meet that lower standard.
  • It seems very likely that drug manufacturers will benefit from lowering the bar for FDA approval.

Finally, if someone can achieve sustainable quality of life improvements, that they find satisfactory, through reduced use that is achieved with or without medication, that is an unqualified success.

The question is not whether or not that’s a good thing. The question is about which endpoints deliver the best quality of life for which kinds of problems. (And, who decides, for what reasons, and who benefits.)