GI Bleeding With LVADs

Pinterest LinkedIn Tumblr

Left ventricular assist devices (LVADs) are gaining more traction in the world of heart failure primarily as a bridge-to-transplantation or destination therapy (ie, ineligible for a heart transplant). These patients often times develop gastrointestinal (GI) bleeds requiring transfusion and endoscopic interventions. But why?

By unloading the LV with an LVAD, many patients can no longer generate the conditions required to open their aortic valve. Therefore, all of their LV cardiac output is through the LVAD in a “continuous flow” model of circulation. These patients often don’t have palpable pulses and have narrow pulse pressures (ie, BP of 80/78 mmHg).

Patients develop an acquired von Willebrand disease (vWD) caused by LVAD motor shear forces destroying high molecular weight vWF multimers normally involved in stabilizing primary hemostasis. Furthermore, a low pulse pressure creates relative hypoperfusion to the GI tract. In response, submucosal vascular proliferation occurs. So now you have extra blood vessels in a patient with defective platelet aggregation and on pharmacologic anticoagulation (heparin, Coumadin)… yeah, there’s a good reason to bleed. GI bleeding from angiodysplasia in the context of aortic stenosis (Heyde’s syndrome) has a similar mechanism. Interestingly in prior studies, vWF multimeter levels increased once pulsatility was regained in LVADs or after an aortic valve replacement in Heyde’s.

Addressing this recurring issue can be extremely challenging! If one decreases the LVAD pump speed to increase native pulsatility, they risk overt heart failure (the patient has an LVAD for a reason!) If anticoagulation is decreased, they risk clot formation. This remains an interesting area of research.

Drop me a comment below with questions or your experience regarding this topic!


  1. I’ve had good results with DDAVP in selected patients, both those with VADs and in one case a gentleman with a GI bleed on DAPT who had a mechanical valve. Risks and benefits to be weighed for sure, but have you ever used it for this indication?

    • Hey Andrew! I can’t say I’ve used it for those situations outside of overt, superimposed uremia; however I’ve used DDAVP in the OR during cardiac surgery (also after a fresh DAPT load) knowing the evidence really isn’t convincing… but the benefits outweigh the risks.

  2. We are starting to get more VAD patients in the main OR and GI suites at our hospital now. What type of monitoring do you do for these cases? If its just a colonoscopy for a GIB are you comfortable doing it without and aline? Often times can be difficult to obtain an O2 sat as well. If it is a more invasive case then is an aline definitely indicated? If you experience hypotension is it better to adjust pump speed or bolus a pressor?

    • Arterial lines are up to the discretion of the provider (ie, probably a good thing if large volume fluid shifts are anticipated due to bleeding/ongoing transfusions or they have a relatively high oxygen requirement). That being said, I can’t remember the last time I placed an arterial line for a patient with a VAD coming to the GI suite or a minor procedure in the OR like a MAC-based driveline washout. I document the patient’s baseline VAD flow and pulsatility index (except in a HeartMate 3 device) and try to maintain those values during the anesthetic. All of this has to be clearly documented in the record as most EMRs/paper charts are not created with VAD patients in mind.

      As far as hypotension, you’ve gotta know how VADs work. I can count on one hand how many times I’ve adjusted pump speed outside of the OR/ICU setting (usually in the context of a bad suction event while I’m simultaneously bolusing fluids). If the native heart has some underlying function, I’ll try to augment that with inotropes if my pulsatility begins to drop. If the flow suddenly increases with no change in power (ie, acute vasodilation), then I’ll consider a pressor.

Write A Comment

Pin It