Acute Variceal Hemorrhage: Vasoactive Agent Choice and Antibiotic Timing
Active variceal bleeding in a cirrhotic patient turns on two decisions that have to be made before endoscopy, not after it: which vasoactive agent to start, and whether antibiotic prophylaxis can wait for the scope or has to start now.
Ray T., 54, has been sober for six years after a long stretch of heavy drinking that left him with cirrhosis, and spends most of that sobriety now running a small engine-repair shop with his brother-in-law — work he says gives his days a structure he never had before. He arrived by ambulance an hour ago vomiting bright red blood, twice, at home; his blood pressure on arrival was 88/54 with a heart rate of 118, and he was pale and diaphoretic. He carries a known diagnosis of cirrhosis with esophageal varices, last scoped fourteen months ago and graded medium-sized without prior bleeding, and had been on a non-selective beta-blocker for primary prophylaxis until he ran out three months ago and never refilled it.
Two decisions have to be made before the endoscopy team can even get to him, not after. Octreotide and terlipressin both reduce splanchnic blood flow and portal pressure, but through different mechanisms and different real-world evidence bases — octreotide is a somatostatin analog with a short half-life requiring continuous infusion, the more commonly used agent in the U.S.; terlipressin, a vasopressin analog now FDA-approved and increasingly available, has trial data specifically supporting a mortality benefit against no vasoactive drug (Ioannou and colleagues' Cochrane review) and comparable-or-favorable variceal-bleeding control, though it carries a real ischemic risk terlipressin's own labeling explicitly flags in patients with known cardiovascular disease. Separately, antibiotic prophylaxis in variceal bleeding isn't a discretionary add-on to consider once he's stable — Chavez-Tapia and colleagues' Cochrane review found it reduces both rebleeding and mortality, largely by preventing the bacterial translocation and SBP that a cirrhotic, actively bleeding patient is already primed for, and the actual argument is whether that timing can wait for endoscopy confirmation or has to start on arrival.
Network meta-analyses (Wang and colleagues among them) have found broadly comparable control of active variceal bleeding between octreotide and terlipressin, without a clearly established mortality advantage for either agent specifically in that comparison — terlipressin's more clearly demonstrated advantage is in hepatorenal syndrome, a related but genuinely distinct indication, which is part of why the team's disagreement here is really about logistics and risk-screening rather than a clear efficacy gap between the two drugs.
Two decisions before the scope even starts
I'd start octreotide now — it's the standard, immediately available vasoactive agent, and there's no reason to add a cardiovascular screening step to a hemodynamically unstable patient before starting the drug we already know works and can start this minute.
I'd at least name terlipressin as a real option, not dismiss it for logistics alone. Its own trial data shows comparable, in some populations favorable, control of variceal bleeding, and it doesn't require continuous infusion the way octreotide does. What I do want taken seriously is terlipressin's own labeling — it specifically flags ischemic risk and calls for cardiovascular risk assessment before use.
Ray's history is negative for known cardiovascular disease right now, but 'negative history' in an emergent setting isn't the same as a completed risk assessment, and that distinction is exactly why the label calls it out separately.
I want us not to lose the antibiotic-timing question while we're debating the vasoactive agent, because that's the piece with the clearest trial evidence behind it. Whichever agent you two land on, IV ceftriaxone should start now, on arrival, not held for endoscopic confirmation — his presentation already meets the threshold the antibiotic-prophylaxis trials were built around, and delaying it risks losing a real, well-established mortality benefit while we're still deciding between two vasoactive agents with a genuinely closer efficacy comparison.
Given octreotide's immediate availability and Ray's currently negative cardiovascular history, I'd support starting it now rather than adding a delay to screen for terlipressin's specific risk in an actively bleeding patient.
Agreed: octreotide started immediately alongside ceftriaxone on arrival, with endoscopy performed within the next several hours per standard timing rather than delayed for antibiotic confirmation, since the antibiotics were never contingent on the scope in the first place.
Not agreed: the pharmacologist's view that terlipressin should have at least been screened for as a real option before defaulting to octreotide on availability alone. That question wasn't resolved — it was tabled for the group's next unstable variceal bleed, with a note to build cardiovascular screening into the initial workup rather than treat it as a delay to avoid.