Mechanical Circulatory Support in COVID-19: Mixed Results, Hints of Hope

ECMO, most often venovenous, may help in otherwise healthy, younger patients. What happens long term remains unclear.

Mechanical Circulatory Support in COVID-19: Mixed Results, Hints of Hope


(UPDATED) In the race to save lives, physicians have turned to extracorporeal membrane oxygenation (ECMO) and, more rarely, other forms of mechanical circulatory support (MCS) to overcome lung and sometimes cardiac complications of COVID-19. But with limited evidence to back their use and the amount of resources required, physicians on the front lines are urging a go-slow approach.

Normally the ECMO is given 5 to 6 days, “but now we are talking weeks,” Khaldoon Alaswad, MD (Henry Ford Hospital, Detroit, MI), pointed out. “You have to factor that into resource utilization.”

More of TCTMD's coverage on our COVID-19 hub.
More of TCTMD's coverage on our COVID-19 hub.

This lengthy course is one reason why it’s hard to know for sure whether the therapy is effective, observed Craig Thompson, MD, head of the cardiac cath lab at NYU Langone Health in New York City. COVID-19 patients on ECMO “languish for weeks, so we’ve kind of got an initial experience to say, ‘This seems good in selected patients and we’re feeling good about it.’ However, the data’s not totally in yet, and I think we’ll get a better read over the next few weeks,” he said. “Things are trending positively, but it’s impossible to know for sure now.”

Data from the 2009 influenza A (H1N1) epidemic hinted at improved survival with ECMO among patients with acute respiratory distress syndrome (ARDS), but the randomized EOLIA trial failed to confirm this finding. Other mixed data have come from studies of Middle East Respiratory Syndrome (MERS).

The isolated respiratory-failure patients can be supported and their lungs can recover. Deane Smith

But the concept of ECMO has remained appealing. By mid-February 2020, anecdotal reports from China on a handful of patients began suggesting improved survival. The World Health Organization (WHO) has since suggested that ECMO could be considered in patients who have refractory hypoxemia despite being on a ventilator, and in early April the US Food and Drug Administration issued guidance on how to use ECMO in COVID-19.

In this setting, where ARDS predominates over cardiac involvement, venovenous (VV) ECMO is far more common than venoarterial (VA) ECMO. Both use the same equipment to operate, though with different cannula placement.

Cardiologists who are expert at VA ECMO will also be able to perform VV, said Duane Pinto, MD (Beth Israel Deaconess Medical Center, Boston, MA). “In general, we’ve been repurposed as interventional cardiologists focusing upon respiratory support when it comes to ECMO, so lung failure from COVID and things like that. That’s what I’ve been called upon to do the most, rather than shock and arrest and rapidly declining patients where we may put them on VA ECMO,” he explained, adding that, in parallel, there were sharp reductions in the numbers of non-COVID patients presenting to their hospital with conditions that would require more-conventional MCS.

Despite differences in MCS modalities, though, Thompson said what’s happening in COVID-19 is an extension of growing knowledge about MCS among cardiologists as procedures have become ever more complex in recent years. “In the setting of COVID, I think things got a little bit more interesting,” he commented.

But Does It Work?

Cardiac surgeon Deane Smith, MD (NYU Langone Health), believes that ECMO does work, at least in certain cases. “The isolated respiratory-failure patients can be supported and their lungs can recover. I think that there was a lot of concern that might not be the case, that there was something different about this viral illness that made it so that patients wouldn’t recover or that the lungs for some reason could never recover—our experience would suggest otherwise,” he commented.

But, he cautioned, “if ECMO is only used as a bailout strategy and you don’t do the things that you need to do to facilitate lung recovery, then the patients won’t do better.”

Smith also isn’t convinced that MCS, whether VA ECMO, intra-aortic balloon pump (IABP), or percutaneous devices like Impella (Abiomed) or TandemHeart (LivaNova), would be helpful in patients with cardiac manifestations of COVID-19, which he emphasized is primarily a lung disease. “When you have these other things happen, whether its myocardial dysfunction or cardiogenic shock or cardiac arrest, by definition you no longer have a single-organ dysfunction. And so the likelihood of survival, meaningful survival, for those patients goes down incrementally,” Smith said. “Where there’s clearly a resource limitation, I think part of our responsibility is to put the resources to the patients that have the best chance to improve.”

When colleagues ask about outcomes of MCS in COVID-19, Pinto said that question is hard to answer. “We don’t really know because each geography that’s been a hot spot has different resources [and] different thresholds for who they put on ECMO,” he observed. “And then this is all pretty new. . . . A lot of these people are still on support and now just coming off of support.”

Since mid-March, Pinto said, four COVID-19 patients have undergone VV ECMO at his center and a similar number have been treated with Impella 5.0, Impella CP, or IABP.

On Monday, Abiomed announced that the FDA has issued an emergency use authorization for its Impella RP device in COVID-19-related right heart failure or decompensation, including pulmonary embolism. This isn’t the first interest the company has expressed in this arena—in late April, Abiomed bought ECMO manufacturer Breethe, in part “to more comprehensively serve the needs of patients whose lungs can no longer provide sufficient oxygenation, including patients suffering from cardiogenic shock or respiratory failure such as due to ARDS, H1N1, SARS, or COVID-19,” a press release notes.

What we realize is the playbooks that we have used for various cardiac manifestations in the past are likely going to need to be modified for those patients who are experiencing those issues because they have COVID. Craig Thompson

Thompson said that thus far at NYU, ECMO has been a “very lifesaving countermeasure” in patients selected to receive it. “And so I think this VV ECMO thing may have some teeth for COVID patients,” he agreed. “We’re finding that it can take some time, but we’re getting people decannulated from the ECMO, we’re getting them off the ventilators, we’re getting them discharged to home.”

Speaking with TCTMD late last week, Smith said a total of 30 COVID-19 patients at NYU have received VV ECMO—at one point they had 15 machines running simultaneously. Among these 30, 16 were discharged from the hospital and two died. Twelve remain in the hospital; within this group, five are off ECMO awaiting discharge and seven are on ECMO but weaning.

Sripal Bangalore, MD, based at Bellevue Hospital in New York, told TCTMD that theirs is the NYC Health + Hospitals public system’s only ECMO center. They have placed VV ECMO in six COVID-19 patients: two recovered, two died, one is still on MCS, and one was decannulated yesterday. Bellevue clinicians, too, have seen that many COVID-19 patients “require ECMO for prolonged periods of time” and are targeting patients with lung-only involvement. “There is a lot more we need to learn about this disease,” he stressed.

At Henry Ford, Alaswad said, they have given only one patient VV ECMO for COVID-related lung symptoms. Similar to some other hospitals facing a wave of COVID-19, “we decided to have very stringent criteria to put a patient on VV ECMO or ECMO in general because we were overtaken by the surge,” he commented. “Although we have a huge ICU, we had to increase our ICU from 125 beds to 175 beds. And at a certain point we had 150 patients intubated.”

As of late May, this patient, a 25-year-old man, was still alive. “The problem is that it was late in the course of his disease. He’s been intubated for more than 7 days. And he developed all the signs of multiorgan failure,” Alaswad said at the time. “But he’s on ECMO and surviving. You could say if it wasn’t for ECMO, he would’ve been dead. But the end of the story is not there yet. Is he going to be decannulated and be able to walk out of the hospital? That’s what matters.”

The possibility that COVID-19 damage may impart long-term effects is on the mind of Thompson, who gave the example of a 28-year-old NYU patient: “[He] has no health issues, comes in, gets sick as heck, and in a matter of hours goes on VV ECMO [then] hangs around with the worst chest X-ray I’ve ever seen in my life for a couple of weeks. Eventually that clears up, we get him off ECMO and send him home.” The question is whether that degree of lung damage sets people up for future pulmonary problems, he continued. “I’m worried about [whether] they’re going to be happy and healthy and we’re high-fiving them in a couple of years or if they’re going to be on a lung transplant list.”

A more immediate concern in the ECMO community has been whether the therapy might activate coagulation in COVID-19 patients. As such, some groups are being aggressive with anticoagulation, “with concerning consequences,” Thompson said. “There are some groups that are reporting a higher incidence of intracranial hemorrhage and other bleeding complications on ECMO than you would expect in a VV population. So the coagulation cascade questions as they relate to COVID are unanswered for sure, and the only thing I think we know so far is that people have started to do things differently without a lot of data.”

Their center, he said, “took the opposite approach in that we have historically run our anticoagulation for ECMO relatively low. And we’ve been doing that routinely even in COVID, and at least so far it seems to be working okay.”

Understanding when not to invest in circulatory support also is important. For patients who’ve decompensated and require hemodynamic support, VA ECMO may be futile; on the other hand, there may be benefit when it comes to myocardial injury that’s progressing to heart failure, Thompson said. Indeed, case reports have backed this idea.

Whatever benefits there are of MCS in COVID-19, these occur on the backdrop of pandemic-strained healthcare systems that may choose to direct resources elsewhere. And, as noted in a JAMA viewpoint published last month, “ECMO is clearly a finite resource.”

“In a large outbreak, additional limitations to providing ECMO may include a lack of ECMO consoles or disposable equipment, suitably trained staff, or isolation rooms with the requisite infrastructure. Many materials necessary to make ECMO circuitry are manufactured in China and it is conceivable that the outbreak may disrupt supply chains,” its authors write.

How to Do It Well

In general, successful MCS requires expertise, Alaswad stressed. “It’s a team sport. You cannot actually do it as one person. It has to be a team effort and everybody has to be involved, all the way from the social workers and palliative care to the pulmonologist, the cardiologist, the cardiothoracic surgeon, the perfusionist, and the nursing care.”

And in COVID-19, the decision-making is even more complex than usual, all the more so because of the need to don and doff personal protective equipment, he added. It may be that only one physician has actually examined the patient and others on the team must review the chart instead.

Alaswad said that, to be prepared for sudden deterioration, the Henry Ford team decided to talk during rounds about what options would make sense and develop selection criteria. “What we’ve found very, very important is to actually decide up front. Victims of COVID will code any time during the day and a lot of the time during the night,” he said.

At NYU, Smith said he worked closely with cardiothoracic surgeon Zachary Kon, MD; together they did the cannulations for all 30 VV ECMOs at bedside without ancillary imaging. “We did that number one because it seemed like the safest thing to do for the patient, but also to minimize the time and number of people who needed to be in the room at any point in time,” he said. Beyond them, there also was an assistant, perfusionist, and nurse.

What we’ve found very, very important is to actually decide up front. Victims of COVID will code any time during the day and a lot of the time during the night. Khaldoon Alaswad

In terms of operator skill, Thompson said the most important thing is being adept at large-bore access. “The mechanics of being able to put people on VA or VV ECMO are not very technically difficult other than that,” he noted, adding, “It’s more of a [question of] who you put it in, when you put it in, and when you can safely get it out. Those are the decision matrices. Running the unit on a technical level is not really rocket science.”

What shouldn’t happen, Smith cautioned, is inexperienced centers attempting to start ECMO just to deal with COVID-19. “Managing a patient on ECMO is about a lot more than whether you can put two cannulas in a patient. The care of severe ARDS is about a lot more than whether the pump is on or not. What a lot of people with experience with ECMO would advocate is that patients who are failing conventional therapies be referred to ECMO centers,” he stressed.

In fact, Henry Ford Hospital transferred a few patients to Michigan Medicine, Alaswad said, not due to lack of expertise but rather to being overwhelmed by the surge. “Detroit suffered tremendously. We had a very large number of patients,” he explained. “The minorities in our society suffered more and worse outcomes from COVID-19, and Detroit has an 85% minority population.”

Most crucial to success is patient selection, the physicians told TCTMD. Common criteria are younger age, better overall health, time on a ventilator of less than 1 week, and isolated lung involvement. There are pertinent risk scores, like RESP and the modified SOFA score, though these haven’t been validated in COVID-19. The severity of ARDS is based on the level of hypoxemia.

Intense discussions over the potential of MCS in COVID-19, said Alaswad, will hopefully “lead to national or multinational clinical trials where we can decide correctly the outcome of VV ECMO or VA ECMO in patients like this.”

Waiting until patients have failed every other option is not the right way to evaluate ECMO, he urged. What’s better is to design a study “not using it too early where we have a low-risk population that’s not going to show benefit and not using it too late where the horse is already out of the barn,” Alaswad suggested.

For now, the Extracorporeal Life Support Organization is tracking the evolving data in this area. Additionally, the National Cardiovascular Data Registry’s CathPCI and Chest Pain – MI Registries are gathering COVID-19 details, as is Blue Cross Blue Shield of Michigan, Alaswad said.

“What we realize,” Thompson concluded, “is the playbooks that we have used for various cardiac manifestations in the past are likely going to need to be modified for those patients who are experiencing those issues because they have COVID.”

  • Alaswad and Smith report no relevant conflicts of interest.
  • Pinto reports being a consultant to Abiomed and Teleflex.
  • Thompson reports being a noncompensated consultant to Abiomed, Boston Scientific, and SoundBite Medical Solutions.



Giuseppe Biondi-Zoccai

2 years ago
For those interested in early data informing on ECMO in COVID-19, the European data analyzed and reported by Marullo et al are worth considering: