Do PVADs and IABPs Offer Similar Survival in High-Risk PCI or Cardiogenic Shock?

Mechanical circulatory support (MCS) provided by a percutaneous ventricular assist device (PVAD) offers no reduction in mortality compared with the use of an intra-aortic balloon pump (IABP) and may lead to a higher risk of adverse events, according to a new meta-analysis that combined data on both high-risk PCI and cardiogenic shock.

The two therapies “are commonly used in different clinical scenarios,” the study investigators acknowledge. But as a whole, they say, “the use of more robust and sophisticated hemodynamic support devices during complex PCI or for patients with established cardiogenic shock does not seem to have decreased the morbidity and mortality of these conditions.”

But Joseph G. Rogers, MD (Duke University, Durham, NC), who commented on the research for TCTMD, had some reservations. “None of the studies that were cited really are large enough to draw the conclusions that you’d like. They’re very small sample sizes, and the population of patients is heterogeneous. . . . To try to come to some definite conclusion about whether one of these devices is superior to the other is very difficult.”

Various statistical approaches can help overcome these challenges, but not entirely, Rogers said. As such, the meta-analysis is interesting but not definitive. What it does do, he continued, is “set the stage for the fact that we really need to do some well-done clinical trials comparing [individual] devices head-to-head.”

With lead authors Saul A. Rios, MD (Jacobi Medical Center, New York, NY), and Claudio A. Bravo, MD (Montefiore Medical Center, New York, NY), the paper was published online July 24, 2018, ahead of print in the American Journal of Cardiology.

Bravo said that even though PVADs are very different devices than the older IABPs, and can be used in different ways, both classes of device were designed to provide hemodynamic support in high-risk interventions. Clinicians still frequently must choose between them in treating acute MI cardiogenic shock and performing high-risk PCI. Right now, the decision is mainly based on common sense, Bravo told TCTMD, but he and his colleagues had aimed to provide some clarity on the data through their meta-analysis. For now, there’s no final answer on which therapy is best, Bravo agreed. “We need a lot more studies and also a lot more patients to be able to answer this question. . . . I don’t think we have enough evidence to go either way.”

Indeed, earlier this spring there was much talk among cardiologists and other physicians about whether real-world evidence on MCS in acute MI complicated by cardiogenic shock is sufficient to support its widespread use. Some are calling for a large randomized trial comparing the newer percutaneous support devices against standard care, which varies from hospital to hospital and may or may not include IABPs. The major IABP trials, meanwhile, have been disappointing. In the 37-center, randomized IABP-SHOCK II trial of patients with acute myocardial infarction (AMI) complicated by cardiogenic shock, IABPs failed to show a benefit over standard care in terms of all-cause mortality. In the randomized BCIS-1 trial, planned IABP use failed to improve short-term survival or MACCE rates over no planned use of the device in patients with multivessel coronary disease and severe left ventricular dysfunction (LVEF ≤ 30%) undergoing PCI, although longer-term survival was inexplicably improved.

More Studies, More Patients Needed

For their meta-analysis, Rios et al looked at five randomized controlled trials (n = 596), all open-label, and one nonrandomized study (n = 78) that compared the TandemHeart (LivaNova), Impella (Abiomed), or both against an IABP. The therapies were being used during high-risk PCI in patients with multivessel or unprotected left main disease, or as a “bridge to recovery” in patients with cardiogenic shock.

All-cause mortality within 30 days did not differ between the PVAD and IABP arms in the five RCTs (RR 1.09; 95% CI 0.79-1.52), nor did composite MACE (RR 1.11; 95% CI 0.84-1.46). Results from one RCT also showed no difference in all-cause mortality or MACE by 1 year.

The risk of adverse events (acute kidney injury, limb ischemia, infection, major bleeding, and vascular injury), derived from four RCTs, was increased with PVAD use (fixed-effect RR 1.65; 95% CI 1.14-2.39).

However, when the researchers performed trial sequential analysis (TSA) “to control the risk of random errors due to sparse data and repetitive testing of the accumulating data,” this added risk disappeared. Further calculations estimated that a sample size of 13,170 individuals would be needed to tease out any difference in adverse events.

Studies in the meta-analysis also suggest PVADs and IABP use carry the same risks of nonfatal stroke over the short and long term, and of short-term nonfatal MI, though each of these comparisons was derived from only one randomized controlled trial. There were no data on the need for cardiac or vascular surgery, quality of life, or cardiovascular mortality.

“We simply need more RCTs with sufficient bias control and sufficient large groups of participants,” the researchers conclude, stressing, “The field of percutaneous mechanical circulatory support is rapidly evolving, and our results can likely not be extrapolated to higher-output devices currently in use [like Impella 5.0] as well as other emerging investigational low-profile/high-output devices.”

Do we actually know the answer to whether one device is better than another, [or] whether any device in that population is better than treating patients medically? We don’t know that. Joseph G. Rogers

The next step, said Bravo, is to test whether the Impella 5.0 is superior to IABP therapy in terms of reducing mortality or improving patient-centered outcomes. One other area of interest is the role of RV failure, he added. “I think the right ventricle is the forgotten chamber of the heart in the setting of cardiogenic shock. We usually try to support the left ventricle.” Supporting the right ventricle, both in cardiogenic shock and in high-risk PCI, might improve outcomes, Bravo suggested.

It is possible to conduct a randomized controlled trial in post-MI cardiogenic shock, as evidenced by the experience of IABP-SHOCK II trial, Rogers pointed out.

Even today, he asked, “do we actually know the answer to whether one device is better than another, [or] whether any device in that population is better than treating patients medically? We don’t know that.”

Lumping all cardiogenic shock together is also problematic, Rogers added. “I’m not sure that shock following a myocardial infarction is the same as the chronic shock that many advanced heart-failure teams are caring for. Even that is an interesting question. I think there are advanced heart-failure programs in the country who have done some work with balloon pumps and find them to work quite well. But balloon pumping in the setting of acute myocardial infarction really has been less beneficial.”

Here, what the results show is that using PVADs and IABPs in this diverse set of patients seems “equivalent,” with the possibility that there are more adverse events with Impella and TandemHeart, Rogers said. “It’s a statistical attempt to try to pull together the best literature, [but] even when you do that the sample sizes are so small and the populations are so heterogeneous that you really can’t come to a conclusion.”