Bleeding After MCS in DanGer Shock: When, Where, and What to Do?

WASHINGTON, DC—In the DanGer Shock trial of Impella CP (Abiomed) in patients with STEMI complicated by cardiogenic shock (STEMI-CS), bleeding occurs not just in the cath lab but also while on mechanical support in the intensive care unit and after patients have been weaned off the device.

These and other details on the bleeding events, which emerged from a fresh look at the data, present a window of opportunity for clinicians to address the complication, researchers say.

In a surprise, however, there was no apparent link between the bleeding events and mortality risk.

Presentation TCT 2024

TCT 265: Bleedings Related to Use of Microaxial Flow Pump in Cardiogenic Shock: A Secondary Analysis of Danger Shock

Rikke Sørensen, MD, PhD (Rigshospitalet, Copenhagen, Denmark), presenting the results in a moderated abstract session here at TCT 2024, reminded the audience that the 36% relative reduction in 180-day mortality with the microaxial flow pump compared with standard care “came at a cost of increased bleeding and also increased risk of limb ischemia” in DanGer Shock. This analysis is one of several released this week at TCT seeking to better understand the mechanisms and pitfalls of Impella support.

In the main DanGer Shock trial, the rate of moderate or severe bleeding was significantly higher with Impella than with standard care (21.8% vs 11.9%).

For this analysis, Sørensen and colleagues delved into BARC type 3-5 bleeds. A comparison of the 281 patients without bleeding and the 74 with BARC 3-5 bleeding showed many similarities, with nearly equivalent systolic BP, lactate, and LVEF levels at baseline. Around three-quarters had multivessel disease, and nearly all patients in both groups underwent PCI of the culprit lesion. Given the setting—with most undergoing primary PCI—the majority were on dual antiplatelet therapy, with 80% receiving a potent P2Y12 inhibitor, she reported.

Just 2.7% of the bleeds occurred in the cath lab, while 47% developed when patients were on mechanical circulatory support and 50.3% after weaning off support. Most bleeding events occurred at the access site (23%) or were other/not specified (23%), while 21.6% were diffuse, 14.9% gastrointestinal, 9.5% related to cardiac surgery, 5.4% intracranial, and 2.7% pericardial.

While on Impella support, 30.9% of patients had a BARC 3-5 bleed, whereas the corresponding rate was 74.1% during venoarterial extracorporeal membrane oxygenation (VA-ECMO) support.

Over time, the highest rate of BARC 3-5 bleeding was seen in patients who’d received both VA-ECMO and Impella (64.3%) and lowest among those patients who did not receive mechanical support (6.2%). Patients treated with Impella only had a BARC 3-5 bleeding rate of 24.5%, while those treated with VA-ECMO had a rate of 34.5%.

“So the more aggressive or more complex the mechanical support is, the higher the risk of bleeding,” Sørensen noted.

Complications including mechanical ventilation, renal replacement therapy, sepsis, pneumonia during ventilation, and limb ischemia all were significantly more common in the patients who bled versus those who did not.

When examining predictors of bleeding, the researchers adjusted for age, sex, SCAI Shock severity at admission, history of MI, history of diabetes, prehospital treatment by emergency medical services, and whether the patient had undergone emergency CABG surgery. Independent predictors in the overall cohort were use of an Impella device (OR 3.48; 95% CI 1.82-6.64), renal replacement therapy (OR 2.66; 95% CI 1.49-4.79) and escalation to VA-ECMO (OR 5.01; 95% CI 2.51-10.00). Among the patients assigned to Impella, renal replacement therapy (OR 2.55; 95% CI 1.21-5.38) and escalation to VA-ECMO (OR 3.93; 95% CI 1.48-10.47) again predicted bleeding.

Yet BARC 3-5 bleeding was not associated with an increase in mortality risk, even when accounting for admission lactate, blood pressure, and LVEF as well as randomization group (HR 0.96; 95% CI 0.67-1.39).

So, said Sørensen, “what can we draw from the description of these bleedings from the DanGer Shock trial?” A key lesson is that the bleeds “occur throughout the whole hospital stay, not just during the time in the cath lab and not just during the mechanical support,” she said.

Care must be taken in the cath lab, with safe puncture and management of the access site, to avoid hemolysis and suction events, which can provoke hemolysis, while the patient is on mechanical support in the ICU. Furthermore, when considering escalation of mechanical circulatory support, the impact on bleeding risk should enter into decision-making. Then, even after weaning off support, “we need to remember what cohort we are dealing with—we still have critically ill patients on DAPT, many of them treated with renal replacement, intubated, instrumented, and often with an infection,” Sørensen advised.

“Very important data, from a very important trial,” commented co-moderator Joost Daemen, MD, PhD (Erasmus University Medical Center, Rotterdam, the Netherlands). What’s fascinating, he said, is that only one out of four bleeds (out of a bleeding rate of 25%) was actually related to the access site, which works out to a rate of 5%. “Maybe that also puts a little bit into perspective the fact that bleeding in this [study] was not associated with outcome, which is basically in contrast to everything we’ve seen in the past,” he said.

Srihari Naidu, MD (Westchester Medical Center, Valhalla, NY), the session’s other co-moderator, pointed out that earlier studies on bleeding have suggested its impact on survival typically is seen after longer follow-up. “So it’s possible that you’re not seeing much early because of the population, but that later term it may track with mortality in shock patients as well,” he suggested.

Mohammad Reza Movahed, MD, PhD (University of Arizona Sarver Heart Center, Tucson), said that one potential reason for the lack of relationship in DanGer Shock is that the study wasn’t double blinded. The Impella group tended to be in the hospital longer, and during that hospitalization “you knew you put in Impella, you know this is a study, [so] you’re going to do everything you can,” he said, adding that ill effects might be delayed until after discharge. Registry data, said Movahed, have indicated that “in real life, that’s not the case, unfortunately.”