| KQ | Evidence Gap | Potential Future Research |
|---|---|---|
| 1 | Evidence was insufficient to determine efficacy of some medications either because of inconsistency and imprecision or because we found 0 or just 1 small trial with low to medium risk of bias (e.g., amitryptiline, aripiprazole, atomoxetine, baclofen, buspirone, citalopram, desipramine, fluoxetine, fluvoxamine, imipramine, olanzapine, paroxetine, quetiapine). | Future studies could evaluate medications that have some evidence (often from 1 or 2 small trials) suggesting possible efficacy (e.g., baclofen) or medications that have not yet been studied with some theoretical basis to support their potential efficacy. |
| 1 | We found no head-to-head studies of oral naltrexone and injectable naltrexone. | Future studies could compare the benefits of harms of oral and injectable naltrexone. |
| 1 | Whether patients need to stop drinking before starting medications in order to benefit is somewhat unclear. Most studies required patients to abstain for at least a few days prior to initiating medication, but some studies enrolling patients who were not yet abstinent have reported reduction in heavy drinking with naltrexone97,234 or acamprosate.73 | Future studies could assess the efficacy of medications for patients who are not ready to abstain. |
| 2 | We found insufficienta direct evidence to conclude that treatment with acamprosate or naltrexone leads to improvement in health outcomes. | Future studies could focus on health outcomes, such as accidents, injuries, QoL, function, or mortality. These could include large prospective studies to evaluate harm and health consequences with various levels of drinking. |
| 3 | Relatively few studies reported information about suicide, suicidal ideation, or self-harmful behaviors. | Additional studies could be conducted to determine whether precautions about suicide, suicidal thoughts, or self-harmful behaviors are warranted. |
| 3 | Little evidence was available to determine whether naltrexone can be used for people with various liver conditions.b | Future studies could evaluate the use of naltrexone for people with various chronic liver conditions. |
| 4 | No eligible trials assessed the use of FDA-approved medications in primary care settings. | Future studies could evaluate the use of acamprosate and naltrexone in primary care settings. |
| 5 | Evidence on whether any medications are more or less effective than other medications for population subgroups was scant. | Future studies could compare the use of acamprosate and naltrexone for subgroups of patients (e.g., enrolling subjects who all have depression or other psychiatric conditions; comparing effectiveness for men or women or among older or younger patients) |
| 6 | Relatively few subjects contributed data to our analyses of variation in naltrexone response and OPRM1 polymorphisms. Patients with at least one G allele may be more likely to respond to naltrexone, but confidence intervals were wide and the effect was not statistically significant. | Additional studies are likely to change our confidence in the estimate of the effect and to change the estimate. |
| 6 | No studies assessed the clinical utility of genotype-guided dosing strategies or genotype-guided medication selection, and none randomized by genotype. | If variation in naltrexone response by OPRM1 polymorphisms becomes established, then future studies could assess the clinical utility of using genotype-guided dosing strategies. For example, studies might compare the use of genotype-guided dosing strategies (e.g., use naltrexone for patients with at least one G allele, but use acamprosate for A-allele homozygotes) with using naltrexone or acamprosate for all subjects. |
| 6 | Only 1 study was available for most polymorphism-medication response associations. | Future studies could explore other genotypic associations (i.e., not limiting future studies to OPRM1 polymorphisms). |
FDA = U.S. Food and Drug Administration; OPRM1 = mu-opioid receptor gene; QoL = quality of life.
Evidence was insufficient for health outcomes because we found no trials meeting inclusion/exclusion criteria rated as low or medium risk of bias (i.e., accidents and injuries) or mainly because of inconsistency and imprecision (i.e., quality of life and mortality). Very few trials reported any health outcomes, and the included trials were not designed or powered to assess impact on health outcomes—they typically focused on alcohol consumption outcomes.
The FDA removed the black box warning for hepatotoxicity for injectable naltrexone, but it is unclear whether naltrexone should be used in people with various chronic liver conditions.
From: Discussion
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.