GWAS-identified genetic variants associated with medication-assisted treatment outcomes in patients with opioid use disorder: a systematic review and meta-analysis protocol

Opioid use disorder (OUD) is characterized by the Diagnostic and Statistical Manual of Mental Disorders, 5^th edition (DSM-5) as a series of physical and psychological symptoms that promote compulsive opioid-seeking behaviors and hinder the constraint of opioid consumption [1]. The World Health Organization (WHO) reports that roughly 27 million people suffered from OUD in 2016, and about 118 thousand died due to OUD-related drug use in 2015 [2]. The continual increase of opioid-related deaths in North America has called the U.S. Department of Health and Human Services (HHS) and the Ministry of Health in Canada to declare an opioid crisis and take appropriate federal action, in 2017 and 2016, respectively [3, 4].

The most prevalent OUD treatments are a combination of pharmacological and behavioral therapies, commonly known as medication-assisted treatments (MATs) [5]. The medications act as either agonists or antagonists to endogenous opioid receptors, regulating the inhibition or stimulation of the opioid reward system [6, 7]. FDA-approved MATs include methadone, buprenorphine, buprenorphine in combination with naloxone, and naltrexone [5]. In addition to those listed, Health Canada has also recently approved the use of injectable heroin-assisted treatment for severe OUD cases [8].

The regulated administration of these MATs at an individual-based dose is essential in ensuring the effectiveness of the treatment and safety of the patients, as well as averting overdose or mortality cases [9]. Methadone dosing, for example, has been shown to be a key factor in predicting treatment outcomes. Very low doses of this agonist put patients at a higher risk of relapse [10, 11], while too high doses and the induction of methadone have been associated with a higher risk of cardiac arrhythmia and mortality, respectively [9, 12].

MAT efficacy in keeping patients from illicitly using opioids has been variable [10, 11, 13], calling into question whether a genetic basis for how patients respond to treatment exists. Several genetic studies have identified variants associated with a higher risk of developing OUD and MAT metabolism or clearance [14, 15]. However, no clear consensus has been formed regarding genes that contribute to treatment outcomes, including negative ones, in OUD patients seeking treatment. Furthermore, literature has not been systematically reviewed for genetic variants of genome-wide significance in this area, to date.

This systematic review aims to assess all the identified genetic variants from genome-wide association studies (GWASs) significantly associated with treatment outcomes for OUD patients receiving MAT.

The specific objectives of this study include:

This protocol has been reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA-P) reporting guidelines [16]. An accompanying checklist could be found in Additional file 1.

Information collected on this form will include: author(s), year of publication, country, cohort population, number of participants (separated by MAT), ethnicity of participants, mean age, sex ratio, type and dose of MAT, MAT outcomes (as outlined under “Outcome Measures”), any genetic variants found to be significantly associated with the outcomes, method of statistical measures, and p values. The traditional genome-wide significance threshold reported in the literature is p ≤ 5 × 10⁻⁸. However, since a considerable number of studies with a borderline genome-wide significance have been shown to be replicable and showcase genuine associations, p ≤ 1 × 10⁻⁷ will be used as the significance threshold for this review [20].

The main focus of this systematic review will be to assess GWAS-identified genetic variants significantly associated with MAT outcomes.

The primary MAT outcome of interest is illicit (unprescribed) opioid use throughout the duration of the MAT and at follow-up periods, the duration of which are to be determined based on the different studies reviewed. Continued illicit opioid use and abstinence from opioids will be assessed from urine toxicological screens and/or self-reported data.

Secondary outcomes of MAT to be considered in this review are:

Quality assessment and risk of bias scores of included studies will be provided independently by each reviewer. The Quality of Genetic Association Studies (Q-Genie) tool [Version 1.1] developed by McMaster University will be used to assess both the qualitative and quantitative aspects of each study [21]. It is tailored to assess the validity and reliability of genetic association studies. Through Q-Genie, a quality score that corresponds to “low”, “moderate”, or “high” quality would be calculated for each study. The Grading of Recommendations Assessment, Development, and Evaluation (GRADE) tool will be used to assess the risk of bias, strength of evidence, and consistency of included studies [22]. Disagreements occurring between two reviewers regarding the risk of bias score will be resolved through discussion. If a unanimous decision is not reached, then, a third senior author will be consulted.

If appropriate, quantitative methods of synthesis will be applied. Heterogeneity between the studies will be assessed through the I² statistic and 95% confidence interval. If low heterogeneity levels are observed, quantitative methods of synthesis applied will include a random effects metaanalysis with pooled odds ratios for main and secondary outcomes previously mentioned. If a large number of studies are identified in this systematic review, subgroup analyses will be used, where the studies will be separated based on the ethnicities of their respective populations and analyzed accordingly, as genetic associations might be more predominant in certain ethnic groups than others. Other subgroup analyses to be considered are based on variables observed to influence MAT outcomes. These include sex, type of MAT, type of illicit opioid used (for example, heroin versus prescription opioids), and alcohol use comorbidity, if discussed in the original studies. All statistical analysis will be conducted via the RStudio [1.1.456] interface of R statistical software [23].

To address the potential publication bias that might be encountered, PROSPERO and ClinicalTrials.​gov databases will be searched for relevant clinical trial protocols that might not have been followed by a publication of results [24, 25].

To assess the risk of bias within and across studies in the systematic review proposed, GRADE will be used [22]. It will be implemented to evaluate the study limitations and biases that contribute to each outcome of interest reported. The GRADE approach will assess the effect of the limitations on the results, effects being “not serious”, “serious”, or “very serious”. Downgrading of the quality of the study will take place depending on the assessed effect level.

Results will be reported according to PRISMA guidelines, with special considerations to Human Genome Epidemiology Network (HuGENet) guidelines when applicable to GWAS data presentation [19, 26]. Though HuGENet guidelines are more pertinent to systematic reviews and meta-analyses of candidate gene studies with foci on single or multiple related genes, they will be used to uphold a standard when presenting genetic association data, when feasible. Tables will be used to present information on each genetic variant-phenotype association reported, including the study details, population, findings, and source of data. Forest plots will be used to display meta-analysis results, should a meta-analysis be appropriate to conduct. The overall quality of each published result will be discussed, taking into account the risk of bias scores.

This systematic review will be able to identify GWASs that have been conducted regarding MATs for OUD. Having a clear list of relevant studies will enable easier access to published results by the public and researchers alike. Results of the meta-analysis will be informative in determining if any genetic markers have been identified to have an impact on MAT outcomes in patients. This will help direct which genes are of interest for future candidate gene studies or GWASs. It will also allow for a consensus to be made regarding whether genetics affect treatment outcomes in the OUD population. Furthermore, if performed, stratified meta-analyses based on population ethnicities will contribute to the breadth of knowledge of genetic differences between ethnic groups. In addition, this review will allow for more informed treatment plans for individuals with differing ethnicities and genetic makeup. A potential limitation that could arise would be the inability to conduct subgroup meta-analyses due to high calculated heterogeneity between studies or small study numbers. In that case, the studies will be qualitatively reviewed and critically assessed according to their risk of bias scores. Another limitation of the proposed review is the exclusion of results obtained from candidate gene studies. Although some relevant SNP-outcome associations will not be reported on, the level of those reported will be of genome-wide significance, highlighting associations that can be expected and replicated in GWASs.