Background and Objectives: People with opioid use disorder (OUD) who are in safety-sensitive occupations are often not allowed to work if taking methadone or buprenorphine due to concerns about cognitive impairment from opioid agonist effects. Naltrexone, an antagonist which lacks opioid agonist effects, has fewer restrictions. However, comparisons of neurocognitive effects across treatments are limited. Methods: This post hoc analysis examined observed Trailmaking Test (TMT) difference scores and Stroop Color-Word accuracy and interference scores at 4, 8, 16, and 24 weeks after treatment initiation with either extended-release injection naltrexone (XR-NTX) (N=283) or sublingual buprenorphine-naloxone (BUP-NX) (N=287) in a randomized, nonblinded, trial of patients with OUD. Scores were compared between medication groups across time points with linear mixed-effects regression. Results: There were no significant effects of medication group or time for either the TMT or the Stroop tests. TMT difference scores at baseline were 48.40±24.12 (XR-NTX) and 48.26±25.45 seconds (BUP-NX), with model-estimated means (SE) across post-initiation time points: XR-NTX: 37.53 (1.62); BUP-NX: 37.80 (1.51); difference: −0.26 (1.59), P =0.87; 95% CI −3.38, 2.86. Stroop Color-Word accuracy and interference scores at baseline were 49.65 (XR-NTX) versus 50.88 (BUP-NX) and 51.11 (XR-NTX) versus 51.68 (BUP-NX), respectively. Model estimated means (SE) across post-initiation timepoints were: accuracy: XR-NTX: 56.54 (0.90); BUP-NX: 56.22 (0.75); difference: 0.33 (0.73), P =0.65; 95% CI −1.1, 1.76; interference: XR-NTX: 54.15 (0.76); BUP-NX: 54.67 (0.73); difference: −0.52 (0.62), P =0.41; 95% CI −1.74, 0.70. Conclusions: Neurocognitive performance did not significantly differ between patients with OUD treated with XR-NTX or BUP-NX. These results suggest that absolute restrictions on buprenorphine use in safety-sensitive occupations may be unwarranted. Individual cognitive assessments may still be appropriate based on specific job demands.
Castillo et al. (Mon,) studied this question.