Chasing an Unapproved Drug for the Highest-Risk Patient on the Schedule
A 70-year-old man with an ejection fraction of 22% is scheduled for high-risk reoperative valve surgery. The drug with the most promising early signal for exactly his risk profile isn't approved in this country — and the trial that was supposed to settle the question came back negative for almost everyone.
Bernard K., a 70-year-old man, spent forty years as a high school shop teacher and has been managing heart failure since a large anterior infarct fifteen years ago left him with an ejection fraction that never recovered above the low 20s despite years on maximally titrated guideline-directed therapy — sacubitril/valsartan, carvedilol, spironolactone and dapagliflozin, all at target dose. A degenerated bioprosthetic mitral valve placed a decade ago now requires reoperative replacement, and his surgeon has been direct with him about what a second sternotomy in a heart this weak actually means: a real, substantial risk of needing inotropic support simply to separate from bypass, on a ventricle with almost no reserve left to draw on.
His EF of 22% and reoperative field place him in a category where low cardiac output syndrome after bypass is not a remote possibility but close to an expected complication, and the team is discussing whether to try to prevent it rather than treat it once it happens. Levosimendan, a calcium-sensitizing inotrope with a mechanistically distinct profile from the catecholamines and PDE3 inhibitors the team already knows well, has real appeal for a ventricle this reduced — it augments contractility without raising myocardial oxygen demand the way beta-agonism does. Two things complicate reaching for it. It is not approved in the United States, so obtaining it means a special-access import pathway carrying its own delay. And LEVO-CTS, the trial built specifically to test prophylactic levosimendan in cardiac surgery, enrolled patients with an ejection fraction of 35% or less — which means Bernard at 22% is not outside the studied population looking in. He is squarely inside the population that was studied and came back negative on both co-primary endpoints, and any subgroup argument the room is about to make has to be made against a trial that already described him.
Deciding whether to chase an unapproved drug before this valve is even opened
I want us to seriously consider levosimendan for him, even knowing what pursuing it involves. His EF is 22% going into a reoperative field — this is close to the highest-risk population there is for post-bypass low output, and levosimendan's calcium-sensitizing mechanism doesn't raise myocardial oxygen demand the way our usual options do.
Earlier trials and meta-analyses suggested a real mortality benefit specifically in high-risk cardiac surgery populations, before the larger confirmatory trials came back negative overall. I think there's a reasonable argument those larger trials enrolled a broader population than his — diluting any true benefit in the sickest patients below what the trial could statistically detect.
I understand the appeal, but I want to be precise about what we actually know. LEVO-CTS was the definitive trial built specifically to answer this question in high-risk cardiac surgery, and it found no significant difference in its primary composite outcome. It wasn't a trial in a general population that happened to miss a high-risk subgroup — high-risk surgery was the whole population it enrolled.
The dilution argument you're making is a real statistical possibility, but it's a post-hoc explanation for a negative result, not something the trial itself demonstrated. I don't think we should pursue an unapproved drug — with real cost, delay, and logistical complexity attached — on the strength of a subgroup argument the definitive trial wasn't designed or powered to confirm.
That's a fair distinction, and you're right that I was treating a plausible explanation for the negative result as though it were confirmed — it isn't. I don't have a trial that actually shows benefit in a population as severe as his; I have a story about why the trial that looked might have missed it, and that's a meaningfully weaker basis than I was presenting it as.
I'll drop the request for special access. What I'd still want is milrinone staged and ready the moment we're off bypass, given how likely low output is for him regardless of which inotrope philosophy we go in with — prevention through an unproven drug isn't available to us, but fast reactive treatment should be.
Agreed: no prophylactic levosimendan pursued; milrinone staged and ready for immediate reactive use if low cardiac output develops after bypass separation, which the team continued to treat as a likely rather than a remote possibility given his EF.
Not agreed as a closed question: whether the very-low-EF subgroup levosimendan might still benefit remains genuinely unanswered by LEVO-CTS's overall result, and both voices acknowledged that a targeted trial in exactly that subgroup doesn't yet exist. The decision not to pursue special access was framed as the right call given current evidence, not as a claim that the underlying scientific question is settled.