Extended-Release Injection vs Sublingual Buprenorphine for Illicit Opioid Use in Pregnant Patients with Depression and/or Anxiety

Abstract

Background:

Active depression and/or anxiety (D/A) symptoms among pregnant individuals with opioid use disorder (OUD) may impact the efficacy of new OUD medication treatment formulations.

Methods:

We conducted an exploratory secondary analysis of an intent-to-treat, parallel group, randomized trial that compared the effectiveness of injectable extended-release buprenorphine (XR-BUP) to sublingual (SL-BUP). Participants included pregnant adults (N=140) seeking or receiving buprenorphine for OUD treatment. A total of 52.1% (n=73) of participants screened positive for active D/A symptoms at baseline, assessed using the Hospital Anxiety and Depression Scale. The primary outcome of illicit opioid abstinence, measured by urine drug screens, was assessed from baseline through one-year postpartum. We also assessed buprenorphine adherence, opioid craving, and withdrawal during the same period. We employed descriptive statistics to characterize baseline population characteristics and longitudinal multiple regression to test treatment effects across time.

Results:

Opioid abstinence during pregnancy was higher in the XR-BUP group than the SL-BUP group who were D/A positive (group difference=12.5%, p=0.03) compared to D/A negative (group difference=6.7%, p=0.27). Withdrawal was lower during the postpartum period in the XR-BUP group than the SL-BUP group for those D/A positive (group difference −2.9, p<0.01) compared to those D/A negative (group difference= −0.7, p=0.47). We observed a time effect for decreases in the proportion of participants who were D/A positive during pregnancy (Odds Ratio=0.27, p<0.01).

Conclusion:

Among pregnant individuals with OUD and co-occurring D/A symptoms, XR-BUP may confer greater benefits related to opioid abstinence and reduced withdrawal symptoms compared to SL-BUP.

1.1. INTRODUCTION

Pregnant individuals are disproportionately impacted by opioid use disorder (OUD) and overdose, which has been particularly prominent in the US opioid epidemic (Hirai et al., 2021). Depression and/or anxiety (D/A) are highly prevalent co-occurring psychiatric conditions among pregnant individuals who use substances (Pentecost et al., 2021), particularly those with OUD (Pentecost et al., 2021). D/A are associated with decreased medication adherence and worse substance use treatment outcomes (Pentecost et al., 2021; Worley et al., 2012). Buprenorphine use, however, may confer therapeutic benefits for D/A symptom reduction (Ahmadi and Jahromi, 2017; Serafini et al., 2018). Understanding whether D/A symptoms among pregnant individuals with OUD influence the effectiveness of OUD treatments and if OUD treatment influences D/A trajectories is clinically important.

CTN-0080: Medication Treatment for Opioid Use Disorder in Expectant Mothers (MOMs): A Pragmatic Randomized Trial Comparing Two Buprenorphine Formulations was a large pragmatic trial comparing extended-release injectable buprenorphine (XR-BUP) to sublingual buprenorphine (SL-BUP) for reducing illicit opioid use among pregnant people. This trial demonstrated that XR-BUP was well-tolerated, produced superior efficacy for illicit opioid use during pregnancy, and produced non-inferior efficacy postpartum for illicit opioid use compared to SL-BUP (Winhusen et al., 2026). Within the study sample, 52.1% (n=73) of participants screened positive for D/A using the Hospital Anxiety and Depression Scale (HADS) (Bjelland et al., 2002; Winhusen et al., 2026).

First, this short communication presents the results of an exploratory secondary data analysis evaluating the efficacy XR-BUP vs. SL-BUP among participants with and without D/A symptoms at baseline and compared outcomes related to illicit opioid use, buprenorphine adherence, and opioid craving and withdrawal. Second, among those who screened positive for D/A at baseline, we also examined whether XR-BUP was associated with becoming negative for D/A during pregnancy and into the postpartum period. These results may assist clinicians in determining if XR-BUP or SL-BUP differentially assist pregnant persons who screen positive for D/A to improve opioid use, related symptoms, and psychiatric status.

2.1. METHODS

2.1.1. Design, setting, and participants

We conducted an exploratory secondary analysis of data from CTN-0080. Methods of this parent study have been reported in detail previously (Winhusen et al., 2020) and are summarized herein. CTN-0080 was an intent-to-treat, open label, parallel group randomized clinical trial conducted between July 2020 and October 2024 (Winhusen et al., 2026).

Participants in CTN-0080 were recruited from 13 addiction specialty settings, prenatal health clinics, or academic health systems across the US. Eligible participants were 18–41 years of age, with a singleton pregnancy and an estimated gestational age of 6–30 weeks at randomization, eligible to receive/already receiving buprenorphine, and planning to deliver at a hospital with a protocol to treat neonatal opioid withdrawal syndrome. Participants were excluded if they had alcohol or sedative dependence requiring withdrawal management, had a psychiatric condition that would impede participation, or reported criminal legal involvement that would increase the risk of being lost to follow up. Participants were also excluded if they were being actively treated with methadone or naltrexone or were enrolled or had pending enrollment in a treatment program above the American Society of Addiction Medicine Criteria level 3.3 (American Society of Addiction Medicine, 2015). All participants provided informed consent and were randomized (1:1 ratio) to XR-BUP (Brixadi^®) or SL-BUP.

2.1.2. Intervention conditions

Participants assigned to XR-BUP received the weekly formulation during the pregnancy phase and while breastfeeding. In the postpartum phase, participants who were not breastfeeding were eligible for the monthly formulation. Participants receiving SL-BUP received buprenorphine tablets and/or buprenorphine/naloxone films, based on site or participant preference. Target dose for all participants was 16mg SL-BUP per day or the extended-release equivalent (24mg). Dose adjustments were based on patient needs and clinician judgement.

2.1.3. Assessments

Assessments occurred at baseline (prior to randomization), study-week 3, monthly until delivery (i.e. pregnancy phase), and at months 1, 3, 6, 9, and 12 postpartum (i.e. postpartum phase). Urine was collected weekly throughout the study. The primary outcome for this exploratory study was illicit (non-prescribed) opioid abstinence defined as the proportion of urine drug tests negative for fentanyl, opiates/morphine, methadone, or oxycodone. Missing results were imputed as illicit opioid positive. We also assessed additional exploratory opioid-related outcomes, which included buprenorphine adherence, opioid craving, and opioid withdrawal. Adherence was assessed weekly and included both study-provided and non-study buprenorphine (i.e., clinically provided outside of the study). Buprenorphine adherence was calculated as the number of adherent days divided by the number of expected treatment days. XR-BUP adherence was defined as 7 days of adherence per weekly and 28 days per monthly injection. SL-BUP adherence was defined as receipt of study dispensation/prescription and self-reported use, reduced by 7 days for each negative buprenorphine/norbuprenorphine urine result. Craving was assessed using the Opioid Craving Scale (McHugh et al., 2021), and opioid withdrawal severity was assessed with the Short Opiate Withdrawal Scale (Vernon et al., 2016).

Current baseline depression and anxiety were assessed using the HADS, which contains a 7-item subscale that asks about active symptoms of depression and a 7-item subscale that asks about active symptoms of anxiety, with individual items scored from 0–3 and total scores ranging from 0–21 (Bjelland et al., 2002). In line with previous literature, we considered a score of ≥8 on these subscales as a positive screen for a depressive or anxiety disorder (Snaith, 2003). It is important to note that participants were screened with the HADS regardless of D/A history or current D/A treatment. We captured sociodemographic characteristics of participants at the baseline assessment using the PhenX toolkit (Hamilton et al., 2011).

2.1.4. Analyses

We compared participant baseline demographic characteristics, substance use disorders (e.g. amphetamine, cannabis), and pre-study buprenorphine and anti-depressant/anxiolytic use by D/A status using Wilcoxon and Fisher exact tests.

First, we used linear regression to estimate outcome means for participants with active D/A symptoms at baseline (the D/A positive group), for those without D/A symptoms at baseline (the D/A negative group), and most important, for treatment cohorts (XR-BUP vs. sublingual) within each D/A group. We used these mean estimates to test for treatment differences within each D/A group and for differences between the D/A groups, themselves. For each outcome, mean estimates, mean differences, and 95% confidence intervals for the mean differences were all derived from a single linear regression, with treatment, D/A, and treatment × D/A as the covariates of interest. Mean differences were each tested for significance using a Wald Chi Square test with an alpha of 0.05. Except for adjusting the covariates and estimating procedures to accommodate the D/A groups, these analyses mirrored their respective counterparts in the main outcome paper (Winhusen et al., 2026).

Second, for participants D/A positive at baseline, we also compared the impact of XR-BUP vs. SL-BUP on D/A positive vs. D/A negative status during pregnancy and postpartum using longitudinal logistic regression with treatment, time, and a treatment × time interactions as covariates. Site and baseline estimated week of gestation were tested as covariates and retained, if significant.

3.1. RESULTS

3.1.1. Participant characteristics

A total of 140 individuals were randomized to XR-BUP (n=69) or SL-BUP (n=71). Our comparisons of baseline characteristics by D/A status showed no differences between groups (p>0.05). Mean age for both groups was 31 years (standard deviation [SD]: D/A positive SD= 4.7, D/A negative SD= 4.6); most identified as White (D/A positive = 87.7% n=64, D/A negative = 77.6%, n=52) followed by Black race (D/A positive = 4.1% n=3, D/A negative = 10.4%, n=7); with 6.8% of those D/A positive (n=5) and 7.5% of those D/A negative (n=5) reporting being from Hispanic heritage. Slightly more than half had completed ≤high school (D/A positive = 59.2% n=42, D/A negative = 55.4%, n=36), and approximately one-third were employed (D/A positive = 32.9% n=24, D/A negative = 34.3%, n=23). Related to co-occurring substance use disorders, amphetamine (D/A positive = 43.8% n=32, D/A negative = 40.3%, n=27) and cannabis (D/A positive = 30.1% n=22, D/A negative = 37.3%, n=25) use disorders were most common. Similar proportions of participants D/A positive and D/A negative had pre-study buprenorphine treatment (D/A positive = 97.3%, n=71, D/A negative =100%, n=67) and anti-depressant and/or anxiolytic use (D/A positive = 52.1% n=38, D/A negative = 49.3%, n=33).

3.1.2. Medication efficacy among patients with baseline depression and anxiety

The proportion of illicit opioid-negative urine samples during the pregnancy phase in the XR-BUP relative to the SL-BUP group was higher among those who were D/A positive at baseline (group difference=12.5%, p=0.03) compared to those D/A negative (group difference=6.7%, p=0.27) (Table 1). During the postpartum phase, opioid withdrawal symptoms were lower in the XR-BUP compared to the SL-BUP group in those D/A positive at baseline (group difference −2.9, p<0.01) relative to those D/A negative (group difference= −0.7, p=0.47). We did not detect differences by medication condition for buprenorphine adherence or craving by D/A status.

Table 1.

Maternal Outcomes as a Function of Study Treatment and Baseline Anxiety/Depression Status

	Extended-Release Buprenorphine	Sublingual Buprenorphine	Difference (95% CI)	P	Total	Difference (95% CI)	P
Pregnancy	(N=69)	(N=71)			(N=140)
Illicit opioid-negative urine samples, % of UDS’s
Depression/Anxiety+	84.3% ± 5.0%	71.7% ± 5.2%	12.5% (1.0%, 24.0%)	0.03	78.0% ± 4.2%	1.1% (−7.2%, 9.5%)	0.79
Depression/Anxiety-	80.2% ± 5.4%	73.5% ± 5.0%	6.7% (−5.2%, 18.6%)	0.27	76.9% ± 4.2%
Buprenorphine Adherence, % days
Depression/Anxiety+	85.8% ± 3.8%	86.1% ± 4.1%	−0.3% (−11.4%, 10.9%)	0.96	86.0% ± 2.8%	−0.7% (−8.8%, 7.4%)	0.87
Depression/Anxiety-	87.1% ± 4.5%	86.3% ± 3.9%	0.8% (−11.0%, 12.6%)	0.89	86.7% ± 3.0%
Opioid craving scale
Depression/Anxiety+	2.1 ± 0.3	2.7 ± 0.3	−0.6 (−1.4, 0.3)	0.19	2.4 ± 0.2	0.5 (−0.1, 1.1)	0.10
Depression/Anxiety-	1.8 ± 0.3	2.0 ± 0.3	−0.2 (−1.0, 0.7)	0.72	1.9 ± 0.2
Short Opiate Withdrawal Scale
Depression/Anxiety+	6.5 ± 0.7	7.5 ± 0.7	−1.0 (−3.0, 1.0)	0.34	7.0 ± 0.5	2.3 (0.7, 3.8)	0.01
Depression/Anxiety-	4.0 ± 0.8	5.4 ± 0.7	−1.4 (−3.6, 0.7)	0.18	4.7 ± 0.6
Postpartum	(N=67)	(N=69)
Illicit opioid-negative urine samples, % of UDS’s
Depression/Anxiety+	58.2% ± 5.5%	58.7% ± 5.9%	−0.4% (−15.9%, 15.0%)	0.96	58.5% ± 4.2%	−3.0% (−14.1%, 8.1%)	0.60
Depression/Anxiety-	62.6% ± 6.2%	60.2% ± 5.5%	2.4% (−13.5%, 18.4%)	0.76	61.4% ± 4.3%
Buprenorphine Adherence, % days
Depression/Anxiety+	70.6% ± 5.5%	76.7% ± 5.8%	−6.1% (−21.9%, 9.8%)	0.45	73.7% ± 4.0%	−3.0% (−14.4%, 8.5%)	0.61
Depression/Anxiety-	74.9% ± 6.3%	78.4% ± 5.5%	−3.5% (−20.1%, 13.1%)	0.68	76.6% ± 4.2%
Opioid craving scale
Depression/Anxiety+	1.4 ± 0.3	1.7 ± 0.3	−0.3 (−1.0, 0.5)	0.50	1.6 ± 0.2	0.0 (−0.6, 0.5)	0.87
Depression/Anxiety-	1.8 ± 0.3	1.4 ± 0.3	0.4 (−0.4, 1.3)	0.29	1.6 ± 0.2
Short Opiate Withdrawal Scale
Depression/Anxiety+	3.7 ± 0.7	6.7 ± 0.7	−2.9 (−4.8, −1.1)	<0.01	5.2 ± 0.5	0.9 (−0.5, 2.4)	0.21
Depression/Anxiety-	3.9 ± 0.8	4.6 ± 0.7	−0.7 (−2.7, 1.2)	0.47	4.3 ± 0.5

3.1.3. Outcomes as a function of baseline depression and anxiety

During the pregnancy phase, opioid withdrawal symptoms were significantly higher in the participants who were D/A positive at baseline relative to those who were D/A negative at baseline (group difference 2.3, p=0.01). We did not detect differences for any of the other outcomes (Table 1).

3.1.4. Association of buprenorphine type on screening for depression and anxiety

Results evaluating the association of XR-BUP and SL-BUP with reductions in D/A scores below the cut-off score of 8 over time in those who were D/A positive at baseline are provided in Table 2. While there were no significant treatment or treatment × time interaction effects, the significant time effect observed during pregnancy suggests there may be a significant decrease over time in the proportion of participants scoring at or above the HADS cutoff during pregnancy.

Table 2.

Effect of Extended Release vs. Sublingual Buprenorphine on Depression/Anxiety ± During Pregnancy and Postpartum

Pregnancy Phase	Odds Ratio	95% Confidence Interval	p
Extended-release BUP	1.50	(0.21–11.01)	0.69
Time	0.27	(0.11–0.70)	<0.01
Extended-release BUP × time	0.57	(0.17–1.93)	0.36
Gestational age at study enrollment	2.29	(0.86–6.16)	0.10
Postpartum Phase
Extended-release BUP	0.28	(0.06–1.40)	0.12
Time	0.91	(0.70–1.18)	0.46
Extended-release BUP × time	1.02	(0.73–1.44)	0.89

4.1. DISCUSSION

This study sought to elucidate associations of XR-BUP and SL-BUP with reductions in illicit opioid use and related symptomology among pregnant people as well as the impact of these medications on D/A status during the pregnancy and postpartum phases. This study suggests that for those who are D/A positive, XR-BUP may confer greater benefits with regard to illicit opioid abstinence during pregnancy and withdrawal symptoms in the postpartum phase. This study found that patients with co-occurring D/A at baseline had significantly higher opioid withdrawal scores during the pregnancy phase but not during the postpartum phase which might reflect the decrease in the proportion of participants scoring positive for D/A over the course of the trial and/or the significant decrease in postpartum opioid withdrawal in the D/A positive group receiving BUP/XR. As withdrawal management is particularly important in preventing return-to-use episodes following delivery (Park et al., 2012), this latter finding is notable. If confirmed through additional clinical research and/or leveraging administrative data sources, these findings may support increased efforts to offer XR-BUP to pregnant patients with co-occurring D/A.

Identifying potential causal pathways underlying these findings is an important next step, particularly those related to formulation differences. XR-BUP formulations, including Brixadi^®, demonstrate more favorable pharmacokinetic and pharmacodynamic profiles compared with sublingual buprenorphine as they produce reduced peak-trough variability and support a more stable and consistent plasma concentration, resulting in steadier bioavailability over the dosing interval (Blair, 2024). For individuals with D/A symptoms, XR-BUP may be better tolerated, in part because its steadier pharmacologic profile could reduce physiological stress and contribute to a more comfortable treatment experience, as suggested in prior patient satisfaction studies (Blair, 2024). Moreover, addressing external social and individual contextual factors may also benefit this population. Pregnant individuals with OUD, including those in our study, often have histories of trauma and maltreatment, along with related mental health challenges (Cash et al., 2025). These stressors can be overwhelming and may hinder long-term adherence and abstinence. For individuals experiencing active OUD-related symptoms, these medications may help reduce treatment-related burdens by simplifying care. It could also be the case that such simplified care has greater alignment with patient preference for MOUD management; future research should seek to better understand the importance of such differences on participant outcomes.

4.1.1. Limitations

While this study includes several strengths, including its multi-site, randomized design and use of well-validated objective measures, several limitations should be noted. First, this study was not specifically designed to answer the question of efficacy of XR-BUP for reducing illicit opioid use and related symptoms among pregnant people with OUD and co-occurring D/A. However, given the longitudinal randomized design employed along with validated measures, this secondary analysis has a high degree of internal validity and a design that may support causal inference. Additionally, the sample of this study was also limited in terms of size and demographic generalizability. Future studies should seek to include a more diverse population. Fortunately, this study included sites from across the US, thus providing a greater degree of geographic generalizability. In addition, this study used a D/A screening tool rather than clinical diagnoses; future studies should clarify what diagnoses are represented (e.g., substance-induced D/A, generalized anxiety disorder, major depressive disorder). This study only examined one type of XR-BUP product. Consequently, the findings may not generalize among other long-acting buprenorphine products. Future studies should seek to evaluate additional products to verify findings. Lastly, while we reported similar baseline levels of medication treatment for D/A in our results section above, we did not assess long term D/A medication impact or capture or track psychotherapy for these conditions overtime. While randomization of participants likely equalized impact of any such treatments across groups over time, future research may benefit from more closely examining the interaction between other treatments, MOUD type, and patient outcomes.

5.1. CONCLUSION

This study provides initial evidence that XR-BUP products may be superior to SL-BUP for reducing illicit opioid use and withdrawal among pregnant individuals with co-occurring D/A. These findings may be helpful for clinicians considering which buprenorphine products to recommend to individuals with these unique care needs.