Prioritizing Impella Support May Boost Cardiogenic Shock Survival, Pilot Study Suggests

WASHINGTON, DC—A protocol focusing on early use of the Impella ventricular support device (Abiomed) in patients with acute MI complicated by cardiogenic shock may help lower the stubbornly high mortality rates seen in that population, a small study suggests.

Concerns about study design, however, particularly given the high cost of the device, should temper enthusiasm for applying results in practice before a randomized comparison is done, one expert says.

The Detroit Cardiogenic Shock Initiative—which stresses early identification of shock, unloading with Impella support prior to PCI, a reduction in vasopressors and inotropes, and hemodynamic monitoring to guide next steps—was applied to 37 patients in this analysis. As William O’Neill, MD (Henry Ford Hospital, Detroit, MI), reported at the American College of Cardiology 2017 Scientific Session here, the rate of survival to explant in this small series was 89%. Survival to discharge was 84%, and all patients who were able to leave the hospital had recovery of native heart function without the need for LV assist devices or cardiac transplants, he said.

Implementing such a protocol in a systematic way “can dramatically improve outcomes, and I believe that this will set the new standard of care for the management of acute myocardial infarction cardiogenic shock,” O’Neill said. He noted that the only patients excluded from the approach are those with anoxic brain injury. “But apart from that, I think we can do this systematically [in] all comers and have a great improvement in outcome.”

David Cohen, MD (Saint Luke’s Mid America Heart Institute, Kansas City, MO), however, was more circumspect in comments to TCTMD, citing concerns with the methods.

The main issues limiting the strength of the findings, he said, were the lack of a control group, the use of survival to device explant as an endpoint, and the lack of risk adjustment for certain comparisons. He said that without a control group, the impact of Impella is uncertain, adding that survival to explant is an endpoint difficult to compare against historical data. Also, Cohen said, patients given Impella earlier versus later may differ in ways that are not adequately addressed in observational studies.

Cohen said that, conceptually, the protocol described by O’Neill “makes pretty good sense.” For patients in very severe cardiogenic shock, “putting in the Impella first will give you the opportunity to do a more complete job of revascularizing,” he said. “That’s what the benefit has always potentially been.”

But the improvements suggested by the pilot study should be confirmed in a future trial, Cohen added, noting that there is sufficient equipoise to support assigning patients to early or late use of Impella because the device is not used in the vast majority of cardiogenic shock cases.

Wide Variation in Cardiogenic Shock Outcomes Across Centers

Over the past few decades, mortality rates among patients with acute MI complicated by cardiogenic shock have lingered around 50% despite advances in cardiac care. And this is an increasing problem, O’Neill noted, with studies showing that the numbers of patients with cardiogenic shock has continued to rise.

To explore the potential impact of Impella in this setting, O’Neill and colleagues examined the Impella Quality Database, which includes information gathered by Abiomed clinical personnel on nearly all US patients treated with the device since it was first approved in 2008. Roughly one-third of the 46,949 patients included in the database were treated for cardiogenic shock; the device was most commonly used for high-risk PCI.

The researchers identified several factors associated with survival—which averaged 58% in 2016, but ranged from 30% in the worst-performing centers to 76% in the best performers—by analyzing that database and a smaller registry that incorporated more detailed information from high-volume centers.

Use of Impella before PCI was associated with higher survival compared with use of an intra-aortic balloon pump and inotropes before the procedure, with Impella only used when other measures failed (59% vs 52%; P < 0.001).

In addition, patients were more likely to survive to device explant when operators used invasive hemodynamic monitoring (63% vs 49%; P < 0.0001). “So use of Swan-Ganz monitoring to look at cardiac power, determine volume status, discuss inotropic [therapy], and then decrease support is very important in improving outcomes,” O’Neill said.

Moreover, mortality increased along with the number of inotropes/vasopressors used. About one-third of patients who didn’t receive any of the medications died, whereas the mortality rate was 74% in those taking four or more. “This is really a warning to us that using inotropes in patients with cardiogenic shock actually can have lethal consequences,” O’Neill said.

The researchers used those findings to inform best practices and the development of the protocol that forms the basis of the Detroit Cardiogenic Shock Initiative. O’Neill said there are currently nine institutions using the approach to treat patients with shock.

In their pilot study, O’Neill and colleagues achieved an average door-to-unloading time of 82 minutes, with 62% of patients supported with Impella before PCI. Invasive hemodynamic monitoring was used in 84%, and in a similar proportion of patients (86%), operators were able to establish TIMI 3 flow. Use of inotropes/vasopressors was reduced in 80% of cases, due to a relative 58% increase in cardiac power output on Impella support, O’Neill said.

O’Neill pointed out that the 89% rate of survival to device explant observed in these patients was much higher than the 51% rate seen in metropolitan Detroit before the initiative was implemented.

Validation Needed

Commenting for TCTMD, Srihari Naidu, MD (Westchester Medical Center, Valhalla, NY), said the study “is food for thought and hypothesis-generating, [suggesting] that maybe these types of algorithms may move the needle on cardiogenic shock if we generalize them to a larger population.”

The obvious next step would be to study the impact of implementing these protocols over a wider network of hospitals treating larger numbers of patients to evaluate whether they are, in fact, effective when performed by teams other than the highly experienced investigators involved in this study, he added.

“Where they have to go now is to really prove that it was the algorithm and the concept of prioritizing unloading with mechanical support and invasive hemodynamic assessment as opposed to just their innate skill level and the team-based approach to care,” Naidu said. “My suspicion is that it will work, and—regardless of what type of device is used—I do believe that a balloon pump is likely not sufficient in this situation and more advanced support such as Impella probably will be.”

Asked to comment on the 6% overall rate of Impella use in US patients with acute MI complicated by cardiogenic shock that was reported by O’Neill, Cohen said it’s hard to justify greater use without strong evidence of benefit, pointing out that the device “costs many tens of thousands of dollars when we put it in.”

“If they do a good trial and they show us that this really improves short- and long-term survival for these patients, then 6% would be far too few,” he said. “If, on the other hand, they do a trial and the result turns out to show that the survival is no better than it is without the device—as has been shown with the balloon pump—then it’s 6% too many.”