Patients with Opioid Use Disorder Choosing Treatment with Extended-Release Naltrexone: A 6-Month Naturalistic Study

The harmful consequences of opioid use disorders (OUDs) require a continuous search for effective and available treatment options [1]. Opioid agonist treatment (OAT), either with the full-agonist methadone or the partial-agonist buprenorphine, is recommended by the World Health Organization as the treatment approach most likely to reduce mortality, morbidity, and substance use [2]. In Norway, a national OAT program is accessible free of charge for all citizens diagnosed with moderate or severe OUD [3]. Medication options in OAT include buprenorphine, buprenorphine-naloxone, methadone, and, more recently, long-acting buprenorphine and morphine. The majority of the OAT population in Norway is treated with buprenorphine or buprenorphine-naloxone (58.3%). The minority are receiving methadone (37%), and this proportion is gradually decreasing. Long-acting buprenorphine was introduced into OAT during the last couple of years. Daily, supervised medication intake at OAT clinics or pharmacies is a common practice. The take-home dosing schedule is based on individual assessments and negative urine test results [4]. Despite free access to OAT and a variety of medications, it is estimated that as many as 30% of people with moderate or severe OUD do not participate in the national program [5]. There may be many reasons why people do not choose OAT, such as negative treatment perceptions and a desire to abstain from all opioids. Patients experienced a lack of flexibility due to weekly or daily medication pick-up at pharmacies or the OMT Clinics [6]. The expansion of medication options to optimize personalized treatment has been suggested as a way to increase the number of patients in treatment [7]. Since 2010, antagonist treatment with extended-release naltrexone (XR-NTX) has gained recognition as a viable treatment alternative for OUD in the USA but has not yet been approved for clinical use in Western Europe and is only available through clinical studies. A naturalistic study is required to enhance the applicability of these results and guide clinical practice.

XR-NTX is an intramuscular injection that blocks the effects of opioids, prevents overdose, and reduces cravings for 4 weeks [8‒12]. Extensive research has proven that XR-NTX is a safe and feasible treatment option for OUD [13, 14], and studies have suggested that XR-NTX may be appealing to individuals who want to avoid OAT [15, 16]. In contrast to OAT, XR-NTX is not considered a life-long treatment [17]. However, there are few recommendations regarding the duration of XR-NTX treatment in clinical trials. Additionally, studies regarding preferred treatment time, reasons for treatment discontinuation, possible correlations between the chosen time in XR-NTX treatment, and other treatment outcomes are lacking. The effectiveness of XR-NTX treatment in naturalistic settings, where OAT is commonly used and free of charge for the patient, has rarely been explored [12, 18, 19].

This naturalistic, clinical study includes individuals with OUD who voluntarily chose treatment with XR-NTX instead of OAT, and aimed to investigate treatment retention; reason for treatment discontinuation; changes in use of opioids, illicit substances, and alcohol; heroin craving scores; treatment satisfaction; and adverse events (AE). We also explored the differences in treatment outcomes between participants who completed the 24-week treatment study and those who chose to discontinue the study before 24 weeks.

This 24-week open-label clinical naturalistic prospective cohort study included patients who received monthly intramuscular XR-NTX injections. The design of this study and the optional 28-week treatment extension were described in detail by Weimand et al. [20]. The patients were recruited between September 2018 and September 2020. The final patient visit at week 24 was in March 2021.

Participants were recruited from outpatient clinics and inpatient units in five urban hospitals in Norway. The study information was disseminated through health personnel and social workers, the Internet, and by word-of-mouth. Potential participants could contact study personnel directly or through local clinicians and social workers. Eligible participants were men and women aged 18 to 65 years with moderate or severe OUD according to the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM) criteria [21]. Exclusion criteria were severe alcohol use disorder (AUD), serious somatic (e.g., liver failure) or severe mental disorders (e.g., psychosis) that were regarded as contraindications for study participation. Women of childbearing potential could not be pregnant or lactating and agreed to use effective birth control. The use of opioids or any other substances during the study period was not a reason for exclusion.

Eligible participants were referred to an inpatient unit for medically managed withdrawal following screening and were included on the day of admission. An injection with naloxone was given prior to administrating the first XR-NTX injection to confirm opioid detoxification.

Participants were discharged from the inpatient units 1–3 days after the initial XR-NTX dose to further outpatient follow-up. Participants who chose to discontinue treatment before completing the study were not reenrolled in the study. When discontinuing or completing XR-NTX treatment, the participants were given thorough information about the overdose risk and the possibility of continued follow-up in the OAT program (with or without agonist medication). Simultaneously, OAT counselors and other healthcare professionals in contact with the participants were informed of their treatment termination in accordance with participant consent.

Study follow-up after XR-NTX induction was conducted in a naturalistic outpatient setting. Participants attended visits every 4 weeks, where they received an XR-NTX injection, completed study interviews, and provided urine samples.

All participants were required to be enrolled in the Norwegian OAT program on their study inclusion. This safety measure ensured the participants’ immediate access to OAT if needed after XR-NTX discontinuation and, additionally, gave the participants access to individual counseling if needed, both during the study period and after discontinuation or study completion. Participants who completed the 24-week study period had the choice of continuing the study and receiving XR-NTX treatment for additional 28 weeks.

Participants were interviewed using the European version of Addiction Severity Index (EuropASI) [22]. The outcome variables were retention in treatment; days of overall opioid use including heroin, buprenorphine, methadone, and other illicit opioids; days of use of benzodiazepines, amphetamines, cannabis, and alcohol; craving for opioids; treatment satisfaction; and AEs. Data regarding substance use were collected every 4 weeks through an interview using the timeline follow-back technique, whereby participants reported the number of days of use within the last 4 weeks [23].

Participant-reported data regarding the use of opioids, benzodiazepines, amphetamines, and cannabis were corroborated by a urine sample provided every 4 weeks before a new injection was administered. AEs are expected to occur when participants discontinue opioid use before being inducted on XR-NTX. AEs were continuously reported during the study. The intensity of AE was classified using the Medical Dictionary for Regulatory Activities (MedDRA) [24].

Craving for heroin, treatment satisfaction, and willingness to recommend XR-NTX to others during the 4 weeks preceding each study visit were measured on an ordinal scale from 0 (low) to 10 (high). AEs were reported in any participant who received a minimum of one dose of XR-NTX during the study period and for 3 months after study discontinuation.

Treatment retention was defined as the number of weeks that the participants remained in treatment. Data on the reasons for treatment discontinuation were collected for participants, who chose to discontinue the study before the end of 24 weeks.

Descriptive statistics was used to describe participant characteristics, using means and standard deviations, 95% confidence intervals (CIs), or frequencies and percentages, for all the included participants (intention-to-treat [ITT] population, n = 162). Differences between the groups were examined using t tests for continuous variables and Pearson’s χ² tests for categorical variables. Retention was illustrated by a Kaplan-Meier curve for both the for ITT population (n = 162) and the participants receiving a minimum of one XR-NTX injection (modified intention-to-treat [MITT] population, n = 138). Opioid use was explored by growth mixture modeling, which identified groups of participants based on individual profiles using several statistical criteria: Bayes information criterion, in which a smaller value indicates a better model, reasonable group sizes, and average within-group probability of at least 0.80. Further, linear mixed models with random intercepts for participants and random slopes were estimated to assess the between-group (those who completed the 24-week treatment study [completers] vs. those who chose to discontinue the study before 24 weeks [non-completers]) differences in trends regarding substance use, cravings, and treatment satisfaction.

The models contained fixed effects for continuous time variables (which were included as linear variables or higher order polynomials when necessary), groups, and interactions between the time variables and groups. Log-rank test was used to reveal difference in retention between the participants who were included and those who were not included in OAT prior to entering the study. Results with p values below 0.05 were considered statistically significant. All tests were two-sided. All statistical analyses were performed using STATA v.17.

Among the 309 participants assessed for eligibility, 180 were screened and 162 were included in the study (shown in Fig. 1). One participant did not provide essential data for this study and was excluded from further analysis, resulting in an ITT population of n = 161.

Thirty-nine participants (24.2%) were women, which is somewhat lower than the proportion in the general OAT population (30.5%) [4]. Men and women displayed similar age distributions, with means of 37.6 and 38.4 years, respectively, which is younger than in the general OAT population (mean; 46.1 years) [4]. The majority of participants (n = 101; 63.3% of the ITT population) were already included in OAT prior to study inclusion. The 84 completers and 77 non-completers displayed overall similar characteristics. However, the non-completers exhibited a higher number of women (p = 0.019), significantly earlier age at onset of injecting use (p = 0.005) and higher lifetime prevalence of hepatitis B or C (p = 0.038) than the completers (shown in Table 1). Baseline participant characteristics of the ITT population have been previously reported by Weimand et al. [20].

Most participants (n = 138, 85.7%) completed induction and received at least one XR-NTX injection (MITT population). Mean time in the study for the ITT and MITT populations was 15.5 weeks (95% CI: 14.0–17.0) and 18.1 weeks (95% CI: 16.8–19.5), respectively (shown in Fig. 2). However, the median time in treatment could not be estimated.

Eighty-four of the participants (52.2% of the ITT population and 60.9% of the MITT population) completed the 24-week study and received a total of six XR-NTX injections. Of the 54 participants in the MITT population who discontinued treatment, 46 (85.2%) received 1–3 injections, while eight (14.8%) received 4–5 injections. Log-rank test revealed no significant difference in retention between the participants who were included and those who were not included in OAT prior to entering the study.

The 54 participants who chose to discontinue the study before the end of the 24 weeks stated different reasons for leaving: 18 participants experienced worsening physical or mental problems (these are also registered as AEs [16] or serious adverse events [SAEs] [2], shown in Table 2). Seven felt able to remain abstinent from opioids without further support from XR-NTX, four were discontinued due to protocol violations, two wanted to start OAT, one wanted to use heroin, and five stated other reasons for discontinuation. Data on reasons for treatment discontinuation on 17 participants were missing because they did not attend follow-up appointments (shown in Fig. 1).

Analyses of substance use were performed in the ITT population. The distribution of all substance use variables was skewed. While most participants reported 0 days of use, a minority reported 1–28 days of use within a 4-week period (shown in Fig. 3a–h; online suppl. Table 3; for all online suppl. material, see https://doi.org/10.1159/000541431).

There was a significant (p < 0.001) nonlinear reduction in opioid use among the total study population during the study period (shown in Fig. 3 and online suppl. Table 3). Additionally, there was an overall significant difference (p < 0.001) in the use of opioids during the study period between the 84 (52.2%) completers of the 24-week study and the 77 (47.8%) non-completers (p < 0.001) (shown in Fig. 3a). Applying growth mixture modeling did not identify any groups with similar trajectories of opioid use.

The use of amphetamines throughout the study period was all over low and displayed an overall nonlinear slightly increasing trend among non-completers. However, overall, the differences between the two groups were not significant (shown in Fig. 3b).

The overall use of benzodiazepines was low, and no significant changes were observed during the study period. There was no difference between the non-completers and completers regarding the use of benzodiazepines (shown in Fig. 3c).

The overall use of cannabis remained stable in both groups; the non-completer group showed only negligible changes during the study. No significant differences were observed between the groups during the study period (shown in Fig. 3d).

Overall use of alcohol was low, but there was an overall significant difference (p = 0.01) between the non-completers and the completers throughout the study period. The non-completers reported none or a few days of use throughout the study, while the completers reported moderate and stable use, on average, with 1 day of use per week (shown in Fig. 3e).

Urine samples were provided at 89.3% of the visits. Overall, the urine samples corresponded well with the participant-reported previous week’s data on use of opioids (98.3%), amphetamines (93.6%), cannabis (98.1%), and benzodiazepines (96.6%).

Both non-completers and completers reported low and decreasing scores for heroin cravings during the study period, with no significant difference between the groups. Few participants in the non-completer group at the end of the study period may have accounted for the observed changes from week 12 to week 16 (shown in Fig. 3f).

Regarding treatment satisfaction, completers reported a significantly higher overall score for treatment satisfaction (p < 0.001) compared to non-completers. Both groups reported a high willingness to recommend XR-NTX to others throughout the study period, with no overall difference between the two groups (shown in Fig. 3g, h).

Among the 138 participants who received at least one XR-NTX injection, 116 (84%) reported at least one non-SAE and 24 (17.4%) reported at least one SAE (shown in Table 2). The most frequently reported AE was withdrawal-like symptoms (e.g., nausea, chills, diarrhea, and muscle cramps) associated with the termination of opioid use and the XR-NTX induction (65.2%). All these AEs were reported within 4 weeks after the first XR-NTX injection, and one-quarter of them was reported to have started prior to the induction on XR-NTX. Other reported AEs included infections, minor injuries, and various pain conditions. Two participants chose to discontinue treatment because of SAEs, and 16 participants discontinued treatment because of AEs. Only two of the reported SAEs were likely to be associated with XR-NTX (serious withdrawal reactions after XR-NTX induction), but none were considered to be a primary side effect of XR-NTX treatment. No nonfatal or fatal opioid overdoses were reported during the 24-week treatment period. Two nonfatal and two fatal opioid overdoses among four participants were reported between 8 and 12 weeks after last XR-NTX injection. No new or previously unreported AEs or suspected, but unexpected SAEs were reported.

This study marked the first time that XR-NTX was offered in a naturalistic setting in Western Europe and the second time that XR-NTX was offered to opioid-dependent individuals in a clinical study in Western Europe, where OAT is widely available in most countries. Most participants who were inducted on XR-NTX completed the 24-week study, with a mean time of 18.1 weeks in treatment. During the study, there was an overall significant reduction in opioid use and use of other illicit substances, and alcohol consumption was low; participants reported low overall heroin cravings and high treatment satisfaction. No SAEs were directly related to XR-NTX.

The main findings correspond well with a previous Norwegian study of XR-NTX treatment for OUD [14]. Both the induction rate (86%) and treatment retention (61%) were favorable compared with findings from recent studies in the USA [13, 25].

Longer time in treatment is usually associated with favorable treatment outcomes, such as reduced use of illicit substances and a lower risk of mortality [26, 27]. The majority of participants (63.3%) in this study transitioned from OAT to XR-NTX. Some of these participants had been enrolled in OAT program for a longer period and made significant progress in their individual recovery processes. We suggest that some of these participants wanted XR-NTX in a transitional support phase to prevent relapse, with a main goal toward opioid abstinence [19]. Previous studies have found that some patients may choose to leave treatment after a shorter period because they feel no further need for XR-NTX treatment [28]. Seven participants stated that they had no further need for XR-NTX to remain abstinent from opioids and discontinued the treatment before completing the 24-week study. We posit that these participants may have felt that they had reached their opioid abstinence goal before the end of the 24-week study. Nonetheless, posttreatment outcomes of patients who choose to discontinue XR-NTX treatment before week 24 need to be further examined. One-third of participants stated increased physical or mental problems as their reason for treatment discontinuation. Qualitative findings have demonstrated that the transition from opioid use to XR-NTX may lead to overwhelming emotional reactions, which some may seek to manage through increased substance use [18, 19]. A previous Norwegian study also found that anxiety and depression levels are positively correlated with substance use during XR-NTX treatment [29]. We observed no overall increase in the use of other illicit substances following induction on XR-NTX; however, these findings, taken together, emphasize the challenges during the early phases of XR treatment and underscore the importance of tailored psychosocial support for patients to reach the full potential of XR-NTX treatment. The associations between mental health and treatment outcomes should be examined further.

A significant reduction in the overall opioid use was observed during the study period. Participants who completed the 24-week treatment period reported fewer days of opioid use than those who discontinued the study before the end of the 24 weeks. There was also a decrease in the overall use of illicit substances, benzodiazepines, and amphetamines, and many participants used less or no substances at all. The high proportion of participants who reported near abstinence supports the participant’s high motivation for and the effectiveness of XR-NTX as a treatment for OUD even in naturalistic settings.

There was significantly higher alcohol use among completers during the 24-week treatment period; they reported an average of 1 day of alcohol use per week. Comparatively, 34% of the Norwegian population reported drinking alcohol every week [30]. This may illustrate the role of alcohol as a socially sanctioned substance and that participants were engaging more frequently in social settings where alcohol is consumed. Studies on the use of XR-NTX to treat AUD have found that XR-NTX reduced days of heavy drinking for people with AUD by reducing craving and the effects of alcohol [31, 32]. Our study only investigated changes in days of alcohol use, but not the amount of alcohol or the social setting of alcohol use. This aspect should be explored in future studies.

The participants who completed the 24-week study period reported significantly higher treatment satisfaction during the study period than those who discontinued the study. This corresponds with findings from a previous XR-NTX study [33] and suggests a correlation between experienced treatment satisfaction and time spent in XR-NTX treatment. To improve clinical practice, it is important to further investigate both psychological and social factors that may impact treatment satisfaction and combine XR-NTX treatment with psychosocial support to increase treatment satisfaction and retention rate.

Although having access to OAT, participants chose treatment with XR-NTX, and we suggest that motivation to try XR-NTX was due to a genuine interest in antagonist effects. Many participants recruited from the OAT program stated that their motivation to enter the study was a desire to discontinue agonist medication and avoid the procedures associated with the OAT program, such as frequent urine testing and monitored intake [18, 19]. Autonomy in treatment decisions may have inherent value, as emphasized by user organizations [7], and may be an important factor for recovery. We suggest that the advantages of low opioid use and overall reduction in the use of other illicit substances can partially be explained by the participants’ opportunity to choose a treatment that corresponded with their personal long-term treatment goals. This inference is supported by studies that demonstrate that treatment preference is associated with better treatment outcomes in mental health and methadone treatments [16, 34, 35]. These findings also emphasize the heterogeneity of participants’ recovery processes and indicate that XR-NTX may have impacted these processes in diverse ways. Previous studies have found significant associations between the strength of patients’ preference for antagonist treatment and favorable XR-NTX treatment outcomes such as reduced opioid use [16]. Patient involvement in treatment decisions regarding content (e.g., choice of medication) and time in treatment have been emphasized in treatment guidelines [35‒37].

The reported incidence of AEs was consistent with the results of previous studies [9, 10, 13, 14, 33]. As previously observed, opioid withdrawal-like AEs were reported during the first 4-week period and could significantly overlap with the side effects of the first XR-NTX injection, thus complicating the differentiation between opioid withdrawal and XR-NTX-mediated AEs. One-third of the participants reported worsening of physical or mental problems as a reason for early treatment discontinuation and reported withdrawal-like AEs after the first XR-NTX injection. Two nonfatal and two fatal opioid overdoses were reported after treatment discontinuation, all between 8 and 12 weeks after the last XR-NTX injection when the opioid-blocking effect had ceased. These findings highlight the need for enhanced focus and close monitoring during induction on and discontinuation of XR-NTX treatment. Further investigation is required to understand how detoxification from various opioids and dosages may impact induction and treatment retention, as well as to explore AEs, and mortality rates in longer term and post-XR-NTX treatment.

Due to the lack of a matched control group, we could not infer any relative treatment effectiveness of XR-NTX in this study. Generalizing our findings across all OUD populations requires caution; however, our results provide valuable insights into the outcomes and choices of individuals who voluntarily select XR-NTX treatment over OAT.

Urine samples were collected at most visits and corresponded with participant-reported data on substance use within the previous week. However, one limitation is that we could only examine the correlation between urine samples and self-reported data for 1 of the 4 weeks. Nevertheless, previous studies have found that self-reported substance use corresponds well with urine toxicology, particularly in research contexts with no negative consequences [15, 33, 38, 39].

This study corroborates the efficacy of XR-NTX in managing OUD in a naturalistic setting in Western Europe, where OAT is widely available. The key outcomes were as follows: high retention rates, reduced use of opioids and low use of other illicit substances, high treatment satisfaction, and minimal cravings, all of which support XR-NTX as a feasible treatment option. Participants’ ability to choose treatment that aligns with their personal recovery goals may improve treatment satisfaction and potentially their overall treatment outcomes. This study highlights the need for tailored psychosocial support during XR-NTX treatment to enhance both retention and effectiveness of XR-NTX treatment for OUD. Further research should focus on the long-term effects of XR-NTX, explore the dynamics of patient choice in terms of treatment duration, and study the impact of patients’ choice on recovery.

The authors are grateful to the participants. We also thank the manufacturer Alkermes Inc., who, at no cost and in accordance with an investigator trial agreement, provided extended-release naltrexone (Vivitrol^®) for use in the study.

The study was approved by the Regional Committee for Medical and Health Research Ethics Southeast Norway, Approval No. 2018/132, Norwegian Medicines Agency, and boards of research ethics at the participating hospitals. Study monitoring was conducted by publicly funded Regional Monitoring Authorities and followed Good Clinical Practice standards. Before inclusion, participants provided written informed consent. The study treatment incurred no costs for the participants. Participants did not receive any payment or economic compensation for study participation, except for the reimbursement of travel expenses for using public transportation. This study was registered at Clinicaltrials.gov # NCT03647774 on August 28, 2018, before patient enrolment.

The authors have no conflicts of interest to declare.

This work was supported by unrestricted grants from the Research Council of Norway (269864 and 302977) and the South-Eastern Norway Regional Health Authority (2019105). Extended-release naltrexone was provided by Alkermes at no cost in accordance with an Investigator Initiated Trial agreement. The funder had no role in the design, data collection, data analysis, and reporting of this study.

I.H.B. and K.M.W. contributed equally to this paper and they share first authorship. K.K.S., B.W., and L.T. designed the study. I.H.B., K.M.W., K.K.S., and J.M. collected the data. I.H.B., K.M.W., and J.Š.B. analyzed the data. I.H.B. and K.M.W. prepared the manuscript. K.K.S., E.-M.L., B.W., and L.T. conducted supervision. I.H.B., K.M.W., J.Š.B., K.K.S., J.M., E.-M.L., B.W., and L.T. participated in revising, editing, and final approval of the manuscript, and agreed to be accountable for all the aspects of the work.

Ida Halvorsen Brenna and Karin Merethe Waleur shared first authorship.

The ethical approval for this study does not open up for sharing data with any third party not mentioned in the protocol. However, anonymous data can be made available based on an agreement with the National Coordinating Investigator (Lars Tanum) and after having notified the Regional Board of Research Ethics for South-East Norway.