Extended-Release Naltrexone vs. Agonist Maintenance for Opioid Use Disorder
Two patients have both just completed detoxification. One’s history and support system make her a strong antagonist candidate; the other’s own account of four failed attempts points firmly the other way.
C.W. completed a 21-day residential detoxification program eleven days ago and has been fully opioid-free since, confirmed by observed urine toxicology on three separate occasions. She is thirty-one, staying at a structured sober-living house, attending a daily recovery meeting she has genuinely come to look forward to, and started back up a part-time bookkeeping job this week — small pieces of a routine she is rebuilding on purpose, one at a time.
She has one prior treatment episode, four years ago, that ended in relapse after roughly eight months — which she attributes clearly, in her own account, to stopping her medication early because she “felt fine” and wanted to be opioid-free entirely, not maintained on anything. She is explicit and consistent across two separate conversations that she wants an antagonist approach specifically this time. She does not want to take a daily opioid, even a partial one, and names her prior maintenance experience as something that made her feel like she “never actually left” the disorder behind — a distinction she has discussed at length with her sponsor and that clearly matters to how she understands her own recovery. She has no other significant medical history, no cardiac or hepatic disease.
Her sober-living house supports monthly injectable appointments and has case managers who track medical follow-through closely, which was not true of her housing during her prior relapse, when she lived alone with no one checking whether she'd kept a single appointment. She has read about the injection in advance and understands it requires her to be fully opioid-free before each dose — which she describes as the part she actually wants: a monthly appointment she either keeps or doesn't, rather than a decision she has to make correctly every morning on her own, alone with it.
Case A — a well-supported antagonist candidate
Extended-release naltrexone fits her directly. She's already completed the detoxification the antagonist approach requires, she has a clear, consistent, well-reasoned preference against agonist maintenance grounded in her own relapse history, and her sober-living house is structured specifically to support monthly-injection follow-through — the exact piece that determines whether this medication actually works in practice.
I want to name the X:BOT trial data plainly before we proceed: induction failure was substantially higher for extended-release naltrexone than buprenorphine — 28% versus 5.9% in that trial — almost entirely because full detoxification is required first, and any lapse before that first injection reopens the gap. Her eleven days are real, but they're recent.
And her eleven days aren't self-reported — they're three observed toxicology screens inside a case-managed house. That's a materially different situation from the X:BOT participants who never made it to a first injection, most of whom were still working through detox without that structure around them.
That induction failure rate is a real risk during induction specifically, not a case against the drug for someone already eleven days clean with confirmed toxicology and structured support — once successfully inducted, X:BOT itself found relapse outcomes converged, and concluded both medications were comparably safe and effective from that point on.
Agreed on the rest — proceed with her first injection today given her confirmed status, and build in a same-day contingency plan with her case manager for exactly what happens if she uses before then, rather than treating today's plan as the only path forward.
Agreed: first injection given today, with a documented contingency plan for any lapse before the next dose and her case manager looped in on the medical follow-through required.
R.G. has been through inpatient detoxification four times in six years. He is thirty-eight. Each time he relapsed within weeks of discharge — twice before ever starting any medication, once after a short trial of oral naltrexone he stopped within ten days, and once after leaving a sober-living placement that didn't work out because of a conflict with another resident that he still describes, unprompted, in detail. Fourteen months ago he survived a witnessed overdose, reversed with naloxone by a friend who happened to be in the room; he has been hospitalized twice since for infections related to injection use, including one episode of endocarditis he still talks about as the closest he's come to dying, closer than the overdose itself, in his own reckoning. He has hepatitis C, diagnosed three years ago, not yet treated.
He arrives today five days out from his fourth detox, housed temporarily with an aunt who has told him — gently but firmly, in a conversation he repeated to the team almost word for word, as though he'd rehearsed it — that this is the last time she can take him in. He has no other stable housing option lined up if the placement ends, and says, flatly, that he understands what that would likely mean for his chances this time.
He says directly that he does not believe he can sustain full abstinence long enough to reach a naltrexone injection safely again, given his own documented pattern across four attempts. What actually kept him engaged longest — roughly five months, his longest stretch by a wide margin — was the one period he was on buprenorphine maintenance rather than pursuing full abstinence from the start.
He has never had that treatment through a structured program with counseling attached; the five-month period came through a primary care prescriber he lost access to when that practice stopped accepting his insurance — not because the medication itself stopped working for him, a distinction he is careful, almost insistent, about making, as though the record needed correcting on this one point specifically.
He brings up the endocarditis again near the end of the visit, unprompted a second time, and says he thinks about the cardiologist's warning that his valve might not tolerate a repeat more than he thinks about the overdose itself — it felt less like an accident and more like a number he doesn't get many more chances to gamble with. His aunt is not in the room for this part; he asked that she not be, so he could speak plainly without watching her face while he said it.
Case B — when his own history is the strongest evidence in the room
Start buprenorphine maintenance, not an antagonist approach. Sordo's pooled meta-analysis found substantial mortality reduction associated with retention in either agonist medication, and unlike C.W., R.G. has already told us directly, from his own four-episode history, which approach actually worked for him — his longest stable period, roughly five months, was specifically on buprenorphine maintenance.
I agree with buprenorphine as today's starting point, but I want to name the underlying principle more broadly: given how destabilized his history is, and that he survived a witnessed overdose, the actual goal right now is engagement today in whatever medication he can sustain, not optimizing for a particular treatment philosophy. If buprenorphine stops working for him again, the door to naltrexone or any other option should stay genuinely open, not closed by an earlier commitment to one approach.
Where I'd push back is on leaning as hard as you are on his own account of what worked. Those five months were also the one stretch he had stable housing and a prescriber he could reach — "buprenorphine worked" and "his life was briefly stable" are badly confounded in that memory, and I'd rather we start it because it's what he'll actually take today than because his retrospective read is strong evidence.
Fair — we're not actually disagreeing about today's medication. Buprenorphine started now, hepatitis C treatment referral made in parallel rather than deferred until his housing stabilizes, and an explicit, standing agreement that this isn't his last option if it doesn't hold.
Agreed: buprenorphine started today, hepatitis C referral made in parallel, and an explicit standing commitment that a future switch to naltrexone remains open if he stabilizes and later prefers it.
Not agreed: how much weight a single patient's own retrospective account of "what worked" should carry against population-level evidence when the two happen to point the same direction here but might not always. Both voices agreed on today's plan; the harm reduction physician wanted this tension named explicitly for future cases rather than treated as resolved by today's convenient alignment.