Cannabis at the End of Life: What the Evidence Actually Shows
A man dying of end-stage COPD wants to try cannabis for his appetite and anxiety before hospice regulations and his own decline make that impossible, and the team has to separate what patients and families genuinely believe about cannabinoids at the end of life from what the actual comparative trial evidence supports.
V.M. still sits on his porch most afternoons, restoring model ships in miniature at a workbench his son built for him once the COPD made the basement stairs too much, though his hands have grown less steady and his appetite has largely disappeared over the past month. He has heard from two other hospice families in his support group that cannabis helped their appetite and anxiety near the end, and he has asked his team directly, with real conviction, whether he should try it himself before his own decline or his hospice's own restrictions make the option harder to arrange.
What patients hear anecdotally about cannabinoids at the end of life and what the comparative trial evidence actually shows diverge in a way worth stating honestly rather than softening. Jatoi and colleagues' randomized trial directly compared dronabinol against megestrol acetate for cancer-related anorexia and found megestrol produced significantly greater appetite improvement and weight gain than dronabinol — a head-to-head trial where the cannabinoid lost to an older, less novel drug most patients have never heard mentioned in the same breath as cannabis. V.M.'s indication is anorexia in non-cancer terminal illness rather than cancer-related cachexia specifically, so that trial's population is a directional guide rather than an exact match, but it's the strongest comparative data available and it points away from dronabinol as the stronger appetite agent. Federal Schedule I status also creates real, practical friction — inconsistent state-level access, potential interactions with his other CYP450-metabolized medications, and hospice billing and formulary restrictions that vary considerably by program.
The two families he mentioned from his support group aren't imaginary comparisons to him — he can describe each of their situations in real detail, down to which specific product one of them used and how soon the appetite change showed up. He is not asking to be talked out of anything; he is asking, with a matter-of-factness that makes it harder rather than easier, whether the thing his friends found useful is available to him while he can still get to the porch — and the team has to answer that today, in a visit where the better-evidenced drug, the weaker-evidenced one he came in asking about, and a formulary that covers neither in the way he expects are all on the table at once.
On the porch, where he still sits most afternoons
I want to be honest with you about something that runs against what you've heard from your support group. The best comparative trial we have — Jatoi and colleagues, testing dronabinol directly against megestrol for appetite in a similar population — found megestrol worked better, not the cannabinoid. That doesn't mean cannabis can't help you, but it does mean the evidence doesn't back it as the strongest option for appetite specifically, even though it's the one most people have heard about.
I'd add one more layer to that honestly — that trial was in cancer-related anorexia, and V.M.'s appetite loss is from end-stage COPD, a different mechanism. So it's a directional guide, not a precise match, and I don't want to overstate how settled the answer is for his specific situation. What I'd flag independently is a real interaction concern, though I want to name it precisely rather than reach for the word everyone expects: dronabinol has no serotonergic activity, so this isn't a serotonin-syndrome question. It's that dronabinol is cleared by CYP2C9 and CYP3A4, sertraline inhibits CYP2C9 among others, and the predictable result is higher dronabinol exposure — which in a 69-year-old means additive sedation and confusion, in a man whose stated priority is staying alert at his workbench.
Given both of those points, I'd offer megestrol as the better-evidenced first try for his appetite specifically, and keep the cannabis conversation open separately for his anxiety, where the evidence is genuinely more mixed and his own reported experience matters more relative to a thin trial base. I also want to be upfront that our program's formulary doesn't cover cannabis directly given its federal status, so if he wants to pursue it, that's a real logistics conversation, not just a clinical one, and I'd rather he hear that from me now than run into it later.
Agreed: megestrol acetate started for appetite, with dronabinol tabled rather than ruled out entirely — held in reserve specifically for anxiety if megestrol doesn't move his appetite and his anxiety remains prominent, with the sertraline interaction named to him directly as something the team would watch for.
Not fully agreed: the Hospice Medical Director's formulary point left V.M. genuinely disappointed, and the team didn't reach full alignment on how hard to advocate for an exception given his explicit interest — the Palliative Care Physician favored pursuing a formulary appeal now, given his declining trajectory, while the Hospice Medical Director preferred waiting to see the megestrol trial's result first.