Narrowing Down the Choices for Alcohol Use Disorder Treatment

This post is Part 2 in a five-part Series entitled "Narrowing Down the Choices." Part 1 can be found here.

J. was an educated, talented, well-respected professional and a loving and well-loved wife. She had always been a drinker, but it had never caused any problems until after she hit age 30, when her consumption spun out of control. She lost her job and separated from her husband and was devastated.

In efforts to get her life back, she went to multiple rehabs and detoxes, tried numerous medications for depression, and saw several different therapists. Despite all these treatments, her heavy drinking persisted for 10 years, during which time she had several DUIs, endured one stint in jail, went through a divorce, and underwent three week-long stays in a medical hospital for severe, life-threatening alcohol withdrawal.

One day, a doctor gave her a shot of injectable naltrexone, an established treatment for alcohol use disorder, known to block the craving for alcohol and reduce drinking. Suddenly, she was struck sober. Ten years later, she still hasn’t had a drink, and she's back at work, and her life is on track.

However, M., T., and L., also patients with alcohol use disorder, with entirely different personalities and histories, were all prescribed naltrexone for their drinking and had different experiences. M. had significant side effects (depression and nausea), which just made him want to drink more, and T. and L. felt like the medication was like a placebo. The urge to drink overcame them just as often, and the amount they drank didn’t abate. But over time, M. T. and L. tried other approaches and each did, finally, find their way to recovery.

Wouldn’t it be better if we knew ahead of time which treatment will work best for who?

There are now numerous available medications to help people quit or cut back on their alcohol drinking. Often referred to as “medication-assisted treatment," pharmacotherapeutic relapse prevention agents are an evidence-based, usually non-habit-forming option for people with alcohol and other substance use disorders. They are often used in combination with behavioral treatment to help support people in their recovery goals. They help reduce craving and the risk of return to use, by rewiring the brain so that people have more control over their decisions around the substance.

According to guidelines, naltrexone, topiramate, disulfiram, and gabapentin are some of the first and second-line choices for alcohol. But preliminary research shows signs that prazosin, varenicline, ketamine, ondansetron, baclofen, and others may end up having practical utility, too.

In the real world, providers and patients often have to march through two or several treatments before they find the best. Therefore, a major research focus of the field has been to look for better ways to make medication decisions.

Alcohol use disorder, like other behavioral health disorders, is heterogeneous. Each individual is different, and, in fact, there may be subgroups, or subtypes, within the alcohol use disorder population, such that a group of people is more likely to respond to a treatment than another. Subtyping has already shown promise as a way to streamline clinical decision-making.

Sub-typing in Alcohol Use Disorder to Improve Medication Prescribing: The Early Days

Within the addiction field, there have been numerous attempts over decades of research to define alcohol use disorder subtypes. It’s easy to pick any old way to categorize people within a group (say, males versus females), but it’s less straightforward to identify subgroups that are valid (e.g., when you look at them statistically, they separate out from one another, forming distinctly characteristic groups) and have clinical relevance (e.g., they can be used to predict whether or not someone is going to respond to a medication or not).

Much of this research on addiction has initially been done in the alcohol use disorder field. (We will talk about illicit substances and nicotine in Part 3 of this series.) According to one earlier method of subtyping, developed by Babor, there are two types of people with alcohol use disorder: Type A and Type B. Type A individuals have a later onset of their problem drinking, and tend to be less severe, with less comorbid psychopathology. Type B individuals have an earlier onset, more premorbid risk and vulnerability (family history, childhood behavior problems, certain personality traits), more negative consequences, and more are likely to have problems with other drugs.

This method of subtyping has shown promise. Type A’s, with later-onset alcohol use, appear to experience some reduction in drinking when they take antidepressants, like selective serotonin reuptake inhibitors (SSRIs), over the course of several months. By contrast, Type B’s, who have an earlier onset, appear to get worse.

Alcoholism Essential Reads

Does Alcohol Really Make Us Happier?

The Truth About the Health Benefits of Alcohol

In similar studies, naltrexone, the most commonly prescribed anti-craving medicine for people with alcohol use disorder, works best in people who fall in the Type A group, whereas ondansetron (a medication that is rarely used for treatment of alcohol use disorder) works best in those with Type B, showing further clinical applicability.

A limitation of the Babor method of subgrouping is that to identify a person's type requires a complex analysis: a practical tool, like a short questionnaire, has not yet been developed to allow clinicians to quickly categorize individuals into one of the two subgroups. However, age of onset of the alcohol problem (with a cutoff at 25) is a proposed, albeit less effective, approximation that could be done in a clinical setting.

Other subgrouping approaches are also being trialed, using measures such as genetics, presence or absence of a particular mental health diagnosis, blood pressure, alcohol withdrawal severity, Lesch Typology, Cloninger Typology, degree of brain activation during certain tasks (the subject of Part 5 of this series), and motivations for drinking (e.g., reward versus relief drinkers).

A couple of these additional sub-typing methods also seem promising and could even be deployed by clinicians, today. First, individuals with high blood pressure and more severe withdrawal severity benefit more from prazosin than those without. Second, one study found that people who had post-traumatic stress disorder (PTSD) prior to the onset of their alcohol use problems were more likely to get better on sertraline than those whose PTSD came later. Although less clinically relevant, a third set of studies have shown that those with high brain activation in response to an alcohol-related stimulus did better with naltrexone than those without, which is especially interesting as naltrexone is also known to reduce brain activation to alcohol stimuli (more on this in Part 5 of this series).

Project MATCH

A classic study in alcohol use disorder, project MATCH, was a psychotherapy rather than a medication treatment-matching study. In this large clinical trial, individuals with alcohol use disorder were randomized to one of three types of therapy: 12-step facilitation, motivational enhancement, or cognitive behavior therapy. The study results were largely inconclusive, however, because none of the baseline characteristics predicted robustly who was most likely to respond to which therapy, although researchers did find that clients high in anger did better with motivational enhancement than the other two therapies.

Conclusion

Precision medicine involves matching treatments to particular subtypes within a diagnostic category. Most of the treatment-matching research in addiction has been done in the alcohol use disorder field, and to some degree there has been progress in that area. It appears that blood pressure and withdrawal severity, a history of PTSD prior to the onset of the alcohol problem, and Babor typology (age of onset) may give us some clues about which treatment will work best.

Although many of these findings are preliminary, meaning whether or not using them to choose medications improves clinical outcomes has not yet been prospectively studied, we can still keep this information in mind when we are developing treatment plans, and some of the subtyping approaches are easy enough to perform.