Naltrexone for Kleptomania Ahead of a Court Date
A quilting-circle regular faces a misdemeanor charge over an urge she's fought silently for over a decade. The one real trial behind the drug being offered is smaller, and the dosing further from anything labeled, than almost anywhere else in this project.
Thursday evenings have belonged to the same quilting circle for more than fifteen years for S.T., a 52-year-old woman, and it was on the way home from one of those evenings — stopping at a drugstore for thread she genuinely needed — that she was caught leaving with a small item she didn't need and could easily have afforded. It wasn't the first time, only the first time she was caught. She describes a pattern going back more than a decade: a rising, specific tension that builds until she takes something small and unremarkable, followed by a rush of relief that fades within hours into shame she has never told anyone about, including her husband of twenty-eight years, who knows about the arrest but not how long this has actually been going on. She has carried a low-grade, low mood for most of her adult life, never formally diagnosed or treated, the kind of baseline she describes as “just how I am” rather than something she thought to raise with a doctor. Her liver function tests, drawn today ahead of any medication decision, are normal, and she has no history of opioid use of any kind, past or present. Her attorney has asked for documentation that she is engaged in treatment before her court date in six weeks — a real, external deadline sitting alongside the clinical one.
What the team can actually offer her rests on a narrower evidentiary base than almost anything else in this kind of consultation. Grant, Kim, and Odlaug's 2009 trial (Biological Psychiatry) is the only randomized, placebo-controlled evidence naltrexone has for kleptomania at all — twenty-five patients, eight weeks, dosed from 50 up to 150 milligrams daily, well above naltrexone's labeled 50-milligram dose for opioid and alcohol use disorder. It found a real, statistically significant reduction in stealing urges and behavior, and it was well tolerated. But it is one small trial, and the dose that actually produced that result sits at up to three times the dose most clinicians reflexively think of when they hear the drug's name.
Initial consultation, six weeks before her court date
I'd start naltrexone, titrated toward the dose the actual evidence used. Grant, Kim, and Odlaug's 2009 trial is the only randomized, placebo-controlled evidence naltrexone — or really anything — has for kleptomania: twenty-five patients, eight weeks, and a real, statistically significant reduction in stealing urges and behavior, well tolerated, twenty-three of twenty-five completed it. She has a real, external deadline in six weeks, and being in a documented, evidence-based treatment by then matters both clinically and for her case.
I'm not arguing against naltrexone — I'm arguing we should be honest about what that trial actually asked us to do. The dose that produced the result, fifty to a hundred fifty milligrams a day, runs up to three times naltrexone's labeled dose for opioid and alcohol use disorder. That's not the same regimen most of us picture when we hear the drug's name, and the entire evidentiary case for kleptomania rests on this one small trial plus an earlier open-label chart review from the same investigator.
And I want to be precise about the liver question rather than alarming about it, because the history cuts both ways. Naltrexone carried a boxed warning for hepatotoxicity until 2013; the FDA removed it because the enzyme elevations behind it came from doses far above the labeled 50 mg, and the drug did not behave as a hepatotoxin at labeled doses. A warning and precaution stays in the label. But notice where that reassurance was actually earned — at 50 mg. We are proposing to titrate toward 150. That is precisely the dose range the original signal came from, and the range the 2013 reassurance does not cover. So: not a reason to withhold treatment, and not a reason to frighten her. A reason to draw the labs on a schedule rather than start this as if it were routine.
I want to raise something neither of you has named yet. She's carried a low-grade, untreated low mood for most of her adult life — something she's never brought to a doctor because she assumed it was just who she is. Kleptomania has real, documented comorbidity with mood and anxiety disorders, and an untreated depressive baseline could plausibly be amplifying the tension-relief cycle in ways naltrexone's opioid-receptor mechanism was never designed to touch.
I'm not proposing we delay naltrexone to sort this out first — I'm proposing we don't let a single medication targeted at one mechanism stand in for a real evaluation of something she's lived with, untreated, for thirty years.
Agreed: naltrexone starts today at 50 mg with a titration plan toward the trial's dosing range, liver function monitoring scheduled at four and eight weeks, and a separate mood and comorbidity evaluation scheduled this month rather than deferred.
Not agreed, and the reason the plan carries a real branch point rather than a single expectation:
Naltrexone continues at whatever dose achieved that response, with liver monitoring continuing on a quarterly schedule and the mood evaluation's findings folded in alongside it.
The team agreed in advance not to simply increase the dose further past the trial's own upper range without a real conversation about whether the mood evaluation's findings point toward a different primary target — the primary care physician's concern, accepted by both other voices as a genuine condition on continuing to escalate a single-mechanism drug.