Impella Redux: JAMA Publication Post-AHA Renews Push for More Data
The second of two observational studies that first stirred debate at AHA 2019 is now in print, roiling the AMICS waters anew.
A controversial analysis of US registry data suggesting worse outcomes with the Impella percutaneous left ventricular assist device (Abiomed) versus an intra-aortic balloon pump (IABP) in patients with acute MI complicated by cardiogenic shock (AMICS) is now in print, prompting a fresh wave of discussion about safety and efficacy.
The publication, appearing online earlier this week in JAMA, comes 3 months after the study’s initial release stirred debate at the American Heart Association (AHA) 2019 Scientific Sessions. As reported by TCTMD, both this report and a second one that was simultaneously published online in Circulation at the time, were showcased during the same late-breaking clinical trial session at AHA and raised concerns as to how well newer forms of mechanical circulatory support are performing against the older IABPs in real-world practice in the United States.
Clinicians who spoke with TCTMD during AHA said the results must be interpreted with caution but highlight the need for better understanding of who should receive the device and when, with an eye toward reducing the variations in use and outcomes currently seen across US hospitals.
Speaking with TCTMD this week, senior author on the new JAMA paper, Nihar R. Desai, MD, MPH (Yale School of Medicine, New Haven, CT), said the response from the cardiology community since the AHA meeting has been positive. “For the most part, we’ve been quite heartened by the comments and support we’ve gotten from a number of individuals, colleagues. I think it’s reflective of the fact that there is relatively little evidence in this space,” he said.
Now, it falls to clinicians and researchers to advocate for their patients “and demand that we get additional high-quality evidence,” Desai said. He emphasized to TCTMD that their report is merely one piece of an as yet unfinished puzzle.
I think it’s on all of us to ensure that the trial is high quality, that it’s done well, and that we get answers to these questions. Nihar R. Desai
The authors of an accompanying editorial, however, point out that the recently published study is an important part of that puzzle. In the absence of sufficiently powered randomized controlled trials in this area, they say, “large-scale registry studies with propensity matching represent an important data source and method to provide additional evidence.”
These devices—which aim to “unload the left ventricle and improve forward blood flow”—are “theoretically appealing as mechanical circulatory support devices might improve perfusion of critical organs such as the heart, brain, and kidneys. However, any invasive measure is also associated with complications,” write Holger Thiele, MD, Steffen Desch, MD, and Anne Freund, MD (all from University of Leipzig, Germany).
But “until reliable evidence from RCTs is available,” they argue, this study, “together with other registry studies, provide evidence to support a more restrictive use of these devices and, as based on current guidelines, only in selected patients with refractory cardiogenic shock.”
New Indications Followed Approval
Impella has been on the market in the United States since 2008. In 2015, Impella 2.5 garnered premarket approval for use in high-risk PCI, and then in 2016, this approval was extended to acute MI and postcardiotomy cardiogenic shock for not only the Impella 2.5 but also its successors, the Impella CP, 5.0, and LD devices. Other indications have since followed.
Despite extensive observational data on Impella use, as well as RCT data specific to high-risk PCI, AMICS data has proven to be more elusive. Operators with firsthand experience with the device argue that the volatile scenario unfolds heartbeat to heartbeat; split-second decisions could be a matter of life or death, making this a difficult device to study. Clinicians and researches in the field are eagerly awaiting results from the first large randomized trial of Impella versus IABP, DanGer, as well as several others evaluating extracorporeal membrane oxygenation. Abiomed also recently announced another AMICS-related trial: RECOVER IV.
“It's a substantial move in the right direction,” Desai said of DanGer and RECOVER IV. “I will say that they’ve tried to do trials and they’ve been abandoned and they’ve never been brought to completion, so I think it’s on all of us to ensure that the trial is high quality, that it’s done well, and that we get answers to these questions.”
Numerous stakeholders, including payers, regulatory bodies, clinicians, and researchers, must come together to add clarity about the optimal management of AMICS patients and where not only mechanical circulatory support devices but specifically Impella fit in, he said.
It’s important “for us to get back to what matters most here, which is the patients: the patients that are in a very precarious clinical position where there is high morbidity, high mortality [and] everyone’s trying to do the best we can,” stressed Desai.
In-Hospital Mortality, Major Bleeding
For the retrospective cohort study, lead author Sanket S. Dhruva, MD (San Francisco VA Medical Center/University of California, San Francisco), and his co-investigators looked at 28,304 AMICS patients (mean age 65 years; 67% men) undergoing PCI from the National Cardiovascular Data Registry for the period between October 2015 and December 2017. In this group, 81.3% presented with STEMI and 43.3% with cardiac arrest.
Impella was used in 6.2% patients, while 29.9% received IABP alone, 6.5% received another form of mechanical circulatory support or multiple devices, and 57.3% were treated with medical therapy alone. Impella-treated patients tended to be younger and were less likely to have STEMI than those given IABPs.
Using propensity matching, Dhruva and colleagues compared 1,680 pairs that were balanced regarding age, sex, body mass index, race, insurance status, and clinical history. As the investigators previously showed at AHA 2019, patients treated with Impella had a nearly 10.9% higher rate of in-hospital death than those who received an IABP as well as a 15.4% higher rate of in-hospital major bleeding (P < 0.001 for both). These differences were more severe in patients who received MCS after PCI, but they were still apparent in cases where MCS use occurred before or during PCI.
Propensity-Matched AMICS Patients: In-Hospital Outcomes
MCS After PCI
MCS Before/During PCI
The authors conclude that “study interpretation is limited by the observational design. Further research may be needed to understand optimal device choice for these patients.”
In a lengthy rebuttal to the Circulation paper provided to TCTMD back in November, Abiomed chief medical officer Seth Bilazarian, MD, pointed out that prior peer-reviewed and real-world studies have showed more favorable outcomes and better cost-effectiveness with Impella.
Studies a Conversation Starter
Duane Pinto, MD (Beth Israel Deaconess Medical Center, Boston, MA), asked to comment on how the AHA presentations—and now the JAMA paper—have been received, told TCTMD he values the conversations inspired by the results. “In general, people are always worried about when there are safety signals with the devices that we use,” he said, also stressing the need for consumers of medical information to always keep in mind limitations specific to various designs.
“Observational trials are limited by confounding and difficulties in knowing the details of how patients were treated and why,” Pinto observed.
One clue that something is amiss, he suggested, is that the mortality rates in the study by Dhruva et al seem lower than what would be expected for such a high-risk population. “So are we really comparing apples to apples or not? . . . Remember, Impellas are still used extremely infrequently compared to balloon pumps. It still is a relatively smallish group of operators and hospitals who are utilizing this technology in a larger denominator of cardiogenic shock,” so the indication for device use is a relevant—but unknown—detail, Pinto said. “Was it prophylactic or was it for cause?”
Given nuances like these, the AHA reports have raised many questions, Pinto said. “Is it that the device is really unsafe? Or is it that the patients are not comparable? Is it that we need more training [on] how to safely use these devices or who the right patients [are]? And do we need higher-quality data? It’s probably [a mix of] all of the above. So that’s why I think the greatest contribution of these two papers was to make us think more about what we’re doing and to be really rigorous consumers of the data.”
Pinto urged that withholding treatment from patients who might benefit is not the right answer. Rather, he said, “it’s our job as both physicians and partners with industry to try to figure out who the right people are.”
https://t.co/obzUXQUn1D 34% mortality in IABP arm means they didn’t have cardiogenic shock as we know & study it. If results had have been the other way round, scepticism would rightly abound. Two RCTs in cardiogenic shock of ECMO & 1 of Impella active in Europe. Until then:🤷♂️— Andrew SP Sharp (@drandrewsharp) February 11, 2020
Another nail in the coffin for Impella?
Use of an Intravascular Microaxial Left Ventricular Assist Device vs Intra-aortic Balloon Pump — Conservative Cardiologist (@DavidLBrownMD) February 10, 2020
Major paper published by @JAMA_current today that uses real-world data to examine outcomes among patients undergoing PCI for AMI who received mechanical circulatory support, comparing the Impella to IABP (and to medical therapy) (1/9) https://t.co/EsoRfSj1eH— Joseph Ross (@jsross119) February 10, 2020
Though Pinto praised the idea of active discussion, he was less enthusiastic about how the Impella debate continues to play out on social platforms. “I think that there are different camps: there are people who are vociferous antagonists and vociferous proponents. We as a community have to not be so polarizing. I think we need to understand what we’re all on the side of is better outcomes for the patients. If somebody has a secondary gain that’s besides that, they need to be called out on that. If they . . . want to advance their career or sell more catheters or whatever, listen, I’m just taking care of patients.”
We as a community have to not be so polarizing. I think we need to understand what we’re all on the side of is better outcomes for the patients. Duane Pinto
“I have my own personal experience where I believe there’s a group of cardiogenic shock patients and high-risk PCI patients that I couldn’t have pulled through without the use of this device,” Pinto continued. “Now are there certain situations where I used the device and maybe it didn’t make a difference or led to a complication? Yes. But on balance, I feel like I have to make the decisions in the moment that I can and be informed by more-relevant data as it comes out. And I think we still need to generate that more-relevant data.”
To TCTMD, Desai said the editorial accompanying their paper nicely synthesizes the current situation and calls on the clinical community to look hard at how decisions are being made.
“[Although] everyone is trying to do the right thing and, in some ways, we believe mechanically in this device that it does all we want it to do, things are much more complicated. Things that may not have been foreseen can happen in real-world practice. We have many other examples of where that was the case. That just heightens the need for us to align our clinical practice with the best-available evidence,” he commented.
As for whether the evidence base does in fact, as the editorialists argue, indicate Impella should only be used in refractory shock, Desai replied: “If you really look at that, one would say it would be difficult to advocate for use of this device as a matter of routine for all patients with cardiogenic shock, particularly given the safety signal that has emerged not just in one analysis now but in two separate, independent analyses by independent investigators.”
“I don’t think we’re interested in kind of getting into a back and forth with particular . . . individuals or obviously the manufacturer,” he continued. “Everyone has their own sort of role to play in all of this, and I think for us our approach has always been that we think about the patients first, we try and do the best analysis possible, and we try to answer the question that we set forth to answer [no matter the results].”
Dhruva SS, Ross JS, Mortazavi BJ, et al. Association of use of an intravascular microaxial left ventricular assist device vs intra-aortic balloon pump with in-hospital mortality and major bleeding among patients with acute myocardial infarction complicated by cardiogenic shock. JAMA. 2020;Epub ahead of print.
Thiele H, Desch S, Freund A. Microaxial left ventricular assist devices: in search of an appropriate indication. JAMA. 2020;Epub ahead of print.
- Dhruva, Desai, Thiele, Desch, and Freund report no relevant conflicts of interest.
- Pinto reports serving as a consultant to Teleflex and Abiomed.