Bleeding common in LVAD recipients

Recipients of left ventricular assist devices (LVADs) usually develop haemorrhagic events, which may lead to excess mortality, according to a recent study. While gastrointestinal (GI) bleeding is the most common, intracranial haemorrhages are most strongly associated with death.

“Haemorrhagic events in LVAD recipients remain a frequent complication that occurs in nearly half of patients. In addition, it is associated with a significant increase in the risk of death,” the researchers said. “Management remains complex and empirical.”

The current retrospective, single-centre, cohort study included 88 LVAD recipients (median age 58 years, 88 percent men). Electronic medical files were accessed for data on bleeding events, heart transplantation, and death. The primary endpoint was the occurrence of any haemorrhagic event or bleeding event leading to medical consultation or hospitalization.

Forty-three patients eventually developed at least one bleeding event, yielding an incidence rate of 49 percent. Eighty-seven events were recorded during follow-up. Each patient saw a median of one event, occurring a median of 9 months after implantation. [ESC Heart Fail 2022;doi:10.1002/ehf2.13899]

Most bleeding events occurred in the GI tract (n=56; 64 percent), followed distantly by epistaxis (n=13; 15 percent) and intracranial haemorrhages (n=10; 12 percent). The remaining eight (9 percent) bleeding episodes occurred in other sites. Haemorrhage incidence was not affected by age, sex, body mass index, comorbidities, and type of device used.

Thirty-two patients died over a median follow-up of 2.4 years, corresponding to an all-cause mortality rate of 36 percent. Death was higher among those with vs without bleeding episodes (42 percent vs 31 percent). Time-dependent Cox survival curves showed that bleeding events amplified all-cause death risk by 3.8 times (95 percent confidence interval [CI], 1.5–9.3; p<0.01).

GI bleeding occurred a median of 2.4 years after LVAD implantation, commonly manifested as melaena and haematochezia. Patients with GI bleeding saw an increasing trend of all-cause mortality, but the effect failed to reach statistical significance (hazard ratio [HR], 3.0, 95 percent CI, 0.9–9.3; p=0.05).

Most GI haemorrhage patients required transfusions (82 percent), each receiving a median of 4 units of packed red blood cells. Other management options included therapeutic endoscopic intervention (20 percent) and definitive discontinuation of antiplatelet therapy (55 percent).

On the other hand, intracranial haemorrhages, while less common, led to a significant excess in mortality risk (HR, 14.6, 95 percent CI, 4.2–51.1; p<0.0001). Such bleeding events occurred a median of 1.7 years after LVAD implantation. Management included discontinuing antiplatelet and antithrombotic therapies, and a switch to heparin. One patient underwent invasive management.

“Management is complex because of thrombotic risk, empirical, and almost exclusively medical based on a multidisciplinary approach,” the researchers said. “It was mainly based on temporary interruption of curative anticoagulation and antiplatelet aggregation with possible switch for heparin or even reversion of anticoagulation.”

Important study limitations included single-centre and retrospective design, and small sample size. The observational protocols also limited the study by only generating hypotheses and lines of research.