Op-Ed: A Practical Guide to Addiction Medication

Last weekend while I was rounding on our inpatient detox unit I worked with D. D is 45 and has struggled with alcoholism for more than half of her life. Experiencing relapse after relapse, she has had few periods of sustained sobriety despite many treatments. Unsurprisingly, she felt utterly defeated and too exhausted to try treatment again. However, I was able to change this resistance to treatment with one simple question: "Have you ever been prescribed naltrexone?"

She was dumb-struck. "There is such a medicine?" She asked. D had never heard of it, and learning about the medication was enough to move her from deep pre-contemplation to readiness to try treatment again.

D wondered why no one mentioned naltrexone as a treatment option before. Indeed, why would a patient who has repeatedly sought addiction treatment never have been informed of viable treatment options involving medications?

We know several answers already: failure to teach addiction medicine at the pre- and post-graduate levels, limited opportunities for practicing clinicians to learn about the pharmacotherapy of alcohol use disorders, insurance barriers and ideological barriers to accepting medication management. Still, other larger questions may linger.

Consider, for example, the question of how we judge whether or not a prescription works. A recent JAMA article got me thinking on this subject. The article presented a systematic review of the literature on the effectiveness of medications in treatment of alcohol use disorders by looking at all three FDA-approved medications (naltrexone, both oral and injectable, acamprosate and disulfiram) and two non-approved medications (topiramate and nalmefene).

I've always looked towards research to help me improve treatment. After positive animal studies were presented at ASAM's Med-Sci Conference in 1992, I rushed back to my practice to try it before the Volpicelli and O'Malley studies were published. When acamprosate was approved I hastened to add that to my tool kit - and then when Project Combine did not show a medication effect for it, I stopped prescribing it even though a meta-analysis of the European trials did demonstrate a strong effect.

More recently, I began prescribing topiramate, baclofen, and gabapentin. Off label, none of these will ever likely get an FDA indication. I still on occasion prescribe disulfiram when it's absolutely, "no-kidding," mandatory that a patient cannot drink (e.g. the patient risks losing a job, freedom, spouse, home, etc).

With all these experiences practicing addiction medicine, the JAMA piece brought me back to my central problem: "How do I know that my prescriptions work?" Part of the answer is provided in understanding what each medication has been shown to do: acamprosate is associated with increased total abstinence, but not decreased heavy drinking; oral naltrexone is associated with increased duration before a return to heavy drinking and fewer heavy drinking days; injectable naltrexone has little association with a return heavy drinking but does show a decrease in heavy drinking days; and topiramate results in fewer drinking days, less heavy drinking days, and less drinks per drinking day.

There is also the concept of 'numbers needed to treat' (NNT). This knocked me on my heels.  For acamprosate, the NNT for no return to drinking was 12, but for naltrexone 20. The NNT for naltrexone for a return to heavy drinking was 12. And so, develops my cognitive dissonance.

When I'm working with a patient, I try to use empirical reasoning in my selection of these medications. I prescribe naltrexone to patients with many relapses that are clearly related to conditioned stimuli resulting in powerful incentive salience. Acamprosate seems to be best in patients who are likely to have protracted withdrawal symptoms (European studies). Topiramate is for patients who have not responded well to naltrexone or who have counter indications such as opioid agonist treatment.  Baclofen for patients with advanced liver disease. And so on.

There are so many factors in the course of the disease. Even if naltrexone may be effective, psychosocial stressors may play a more immediate role in a patient's relapse. I rounded today on R. He is 57 years old and has been admitted to our inpatient unit for alcohol detoxification eight times this year and over 30 times since 2010. He is often homeless or evicted because of his drinking. We've discussed his experiences with naltrexone. He tells me, "Doc, it doesn't work." I explained what naltrexone does and does not do and how to assess its benefits. We both acknowledged that as soon as he would drink he would stop taking it.

While patients with higher socio-economic status, good support systems and with less disease severity may be better with medication adherence, for others adherence may be an insurmountable issue without external assistance. This is true for all of the anti-craving medications, including the depot formulation of naltrexone.

So, if a patient returns to heavy drinking how am I to decide if the naltrexone was effective and I should continue it? If my expectation is that the patient stops drinking entirely would I not regard it as a failure? How long should I advise that it be continued? How can a patient who has achieved sustained sobriety believe that it is playing an important role in her recovery?

We can run this exercise other medications. The studies show that these medications are effective, but I believe we need to be more thoughtful in their use and in how we promote them to non-addiction practitioners. We must not be reductionist: these medicines can change the course of the most destructive of diseases. However, let's remember that applying an action phase intervention (the prescription) to a patient in denial or not ready to change is doomed to failure. The physician will learn that the medicine doesn't work, that patients don't want to stop drinking and that there is no value in prescribing them.

Do you prescribe any of these medications? How do you select which one to use? How do you decide if the medication is effective? What barriers have you encountered? How have you dealt with them?