PHILADELPHIA – Use of the Impella ventricular-assist device in patients with cardiogenic shock having percutaneous coronary interventions (PCI) has increased rapidly since its approval in 2008, but two studies comparing it with intra-aortic balloon pumps in PCI patients have raised questions about the safety, effectiveness, and cost of the ventricular-assist device, according to results of two studies presented at the American Heart Association scientific sessions.
The results of an observational analysis of 48,306 patients and a national real-world study of 28,304 patients may not be telling the complete story of the utility of ventricular assist in patients requiring mechanical circulatory support (MCS), one interventional cardiologist said in an interview. “It’s concerning; it’s sobering,” said Ranya N. Sweis, MD, of Northwestern University, Chicago. However, the data didn’t parse out patients who would have been routed to palliative care and otherwise wouldn’t have been candidates for PCI without MCS.
“What I take from it is that we need to get more randomized data,” she said. “Who are the patients that were doing worse? Who are the patients who really needed the Impella support for the PCI after cardiogenic shock?”
Dr. Amit P. Amin
In the observational study, Amit P. Amin, MD, of Washington University, St. Louis, said that the use of MCS devices increased steadily to 32% of all PCI patients receiving MCS from 2008 to 2016 while use of intra-aortic balloon pump (IABP) declined, but that Impella was less likely to be used in critically ill patients. The study analyzed patients in the Premier Healthcare Database who had PCI with MCS at 432 hospitals from 2004 to 2016.
Outcomes in what Dr. Amin called “the Impella era,” showed significantly higher risks for death, acute kidney injury, and stroke, with odds ratios of 1.17, 1.91 and 3.34, respectively (P less than .001 for all). In the patient-level comparison of Impella versus IABP, Impella had a 24% higher risk of death (P less than .0001), 10% for bleeding (P = .0445), 8% for acute kidney injury (P = .0521) and 34% for stroke (P less than .0001). The findings were published simultaneously with the presentation (Circulation. 2019 Nov 17. doi: 10.1161/CIRCULATIONAHA.119.044007)
“The total length of stay, as well as the ICU length of stay, were actually lower with Impella use, by approximately a half day to 1 day,” Dr. Amin said. “Despite that, the total costs were approximately $15,000.”
Yet, the study found wide variation in the use of Impella among hospitals, some doing no cases with the device and others all of them, Dr. Amin said. The risk analysis also found wide variations in outcomes across hospitals using Impella. “We saw a 2.5-fold variation in bleeding across hospitals and a 1.5-fold variation in acute kidney injury, stroke and death,” he noted. The study found less variation in hospital stays and total cost of Impella, “perhaps related to the uniformly high device acquisition costs.”
“These data underscore the need for defining the appropriate use of mechanical circulatory support in patients undergoing PCI,” Dr. Amin said.
Dr. Sweis wasn’t surprised by the cost findings. “New technology is going to cost more,” she said in an interview. “I’m actually surprised that the cost wasn’t more significantly different just knowing the cost of some of these devices.
Patients who require MCS represent a small portion of PCI cases: 2%, according to Dr. Sweis. “It’s not like all PCI has increased because of MCS, and there’s a potential improvement in the length of stay so there are going to be cost savings that way.”
The national real-world study that Sanket S. Dhruva, MD, MHS, of the University of California San Francisco, reported on focused on Impella and IABP in PCI patients with acute MI complicated by cardiogenic shock (CS). The study used outcomes of patients with AMI-CS who had PCI from October 2015 to December 2017 in the National Cardiovascular Data Registry’s CathPCI and Chest Pain–MI registries. An estimated 4%-12% of AMIs present with CS.
Most patients in the study population had medical therapy only, but this study focused on the 1,768 who had Impella only and the 8,471 who had IABP only. The rates of in-hospital death and bleeding were 34.1% 16% in the IABP group, and 45% and 31.3% in the Impella group, Dr. Dhruva said. In this study population, the rate of Impella use increased from 3.5% in 2015 to 8.7% by the end of 2017 (P less than .001).
Dr. Dhruva acknowledged a number of limitations to the study findings, including residual confounding. However, the “robust propensity match” of 95% of the Impella-only patients and the results were consistent across multiple sensitivity analyses. “There may have been questions about the clinical severity of AMI-CS patients in the NCDR Registry,” he said. “However, the registry definition is similar to that used in the trials.”
The trial also failed to distinguish between the different types of Impella devices, but the results mostly pertain to the Impella 2.5 and CP because the 5.0 device requires a surgical cutdown, and the study excluded patients who received multiple devices.
“Better evidence and guidance are needed regarding the optimal management of patients with AMI-CS as well as the role of mechanical circulatory support devices in general and Impella in particular,” he said, adding that Impella has been on the U.S. market since 2008, but with limited randomized clinical trial evidence in cardiogenic shock.
The study population of patient’s with CS is “only a piece of the puzzle,” Dr. Sweis said. “We know that there are sick hearts that aren’t in shock right now, but you’re going to do triple-vessel intervention and use atherectomy. Those patients would not do very well during the procedure itself and it may not even be offered to them if there weren’t support.”
Impella is not going away, Dr. Sweis said. “It provides an option that a patient wouldn’t otherwise have. This is really stressing to me that we need to get rid of that variability in the safety related to these devices.”
Dr. Amin disclosed financial relationships with Terumo and GE Healthcare. Dr. Dhruva had no financial relationships to disclose. The study was supported in part by a Center of Excellence in Regulatory Science and Innovation grant from the Food and Drug Administration and the American College of Cardiology’s National Cardiovascular Data Registry.