Given the novelty of COVID-19 and a lack of scientific evidence, many clinicians treating critically ill patients with the disease, particularly those with acute respiratory distress syndrome, have been using conventional supportive care strategies and mechanical ventilation. Questions have emerged, however, about whether other treatments may also prove beneficial in this patient population.
For instance, in a recent case series, which is currently under review for publication, Hooman D. Poor, MD, director of pulmonary vascular disease at the Icahn School of Medicine at Mount Sinai, and colleagues described how thrombolysis led to dramatic, immediate physiologic improvement in five critically ill patients with COVID-19.
In an interview with Healio Pulmonology, Poor discussed these cases as well as how the atypical presentation of ARDS may provide clues to finding optimal treatment for at least a subset of patients with COVID-19.
What prompted you and your colleagues to look more closely at these five cases?
Poor: Right now, our hospital is overrun by patients with COVID-19 and respiratory failure and shock, and what my colleagues and I have noted is that these patients have severe gas exchange abnormalities — particularly low oxygen and high dead space ventilation — but also have relatively preserved lung compliance. We have been calling this ARDS, but when you have this massive disconnect between gas exchange abnormalities and lung compliance, it calls into question whether this actually is ARDS in the traditional sense. In fact, what comes to mind most prominently is pulmonary vascular disease. Specifically, pulmonary embolism often presents in this way. Additionally, these patients often have elevated D-dimers, lots of clotting of dialysis catheters and higher rates of venous thromboembolism.
Therefore, it was for these five critically ill patients who were in shock with multiorgan failure and essentially at the end of life and who were also demonstrating that type of physiology — increased dead space ventilation and severe hypoxemia but relatively compliant lungs — where we questioned whether clot was playing a role in this. At that point, I decided to give tissue plasminogen activator (tPA) to the first patient and, to my surprise, her partial pressure of carbon dioxide dropped within 20 minutes. Her physiologic response, in essence, confirmed that clot is definitely playing a role in the decompensation of these types of patients. Over the course of that weekend, I found other patients who were also near the end of life with severe multiorgan failure and evidence of dead space ventilation with elevated D-dimers in whom we administered tPA and again saw a dramatic physiologic response.
Could you elaborate more on how the presentation of ARDS in these patients differed from a more typical presentation of ARDS?
Poor: ARDS is an acute injury to the lungs that results in bilateral infiltrates on chest imaging, gas exchange abnormalities defined as a ratio of partial pressure of arterial oxygen to fraction of inspired oxygen less than 300 mm Hg, and these features cannot completely be explained by heart failure or volume overload. These patients with COVID-19 pneumonia respiratory failure actually meet that criteria, which is the problem. Because they meet that criteria, they are being treated as if they have ARDS using low tidal volume ventilation, conservative fluid management and prone ventilation. However, the pathophysiology does not really make sense because severe gas exchange abnormalities in ARDS are due to increased permeability pulmonary edema. The lungs fill up with fluid and the more fluid in the lungs, the more abnormal the gas exchange. But, when that happens, the lungs also become stiffer and less compliant. This disconnect raises the question of whether we should be calling this atypical ARDS and suggests maybe we shouldn’t necessarily be applying the same principles of treatment with respect to prone ventilation, positive end-expiratory pressure and driving pressure because we may be dealing with a different type of disease.
I do want to stress, though, that this is a case series of only five patients. I worry that physicians will run around and start lysing people with COVID-19 pneumonia respiratory failure or that family members will think patients should be getting lysis as treatment. For me, what was more telling about these patients was that, clearly, in ARDS and other syndromes, coagulopathy can exist and, in those other syndromes at least, it does not seem to be the driver of disease. Instead, it is like a sideshow that may worsen or may be a marker of worse disease, so aggressively treating that coagulopathy does not necessarily lead to good outcomes. Here, it was about reframing and asking whether coagulopathy and thrombosis play a major role: Is it a driver of disease or a driver of respiratory failure in at least a subset of patients? If that is the case, then you would want to be more aggressive with treatment. You would want anticoagulation to prevent this from happening and perhaps you would want thrombolysis for patients who are demonstrating organ failure. Again, both anticoagulation and thrombolysis for these patients need to be studied in trials. But currently, we have been saying, “It looks like a duck but walks like a rabbit and let’s just call it a duck.” So, perhaps this isn’t the disease that we think it is, and it may truly be vascular in nature in at least a subset of patients. There may be other subsets of patients who have more of an ARDS phenotype than we think.
Do you have any thoughts about why this might be a driver of disease in certain patients with COVID-19?
Poor: There are many potential explanations, but it is possible that the disease causes vascular injury initially and the disease may be endothelial to begin with and other things happen later. It could also be similar to antiphospholipid antibody syndrome, in that it is an autoimmune condition that results in coagulopathy and end-organ dysfunction because of the coagulopathy and thrombosis. Some people are viewing this as ARDS complicated by endothelial dysfunction and thrombosis, but one could argue that this is endothelial dysfunction and thrombosis complicated by ARDS. This is why biopsies and autopsies can be difficult to interpret because there are perhaps features of both, and it is difficult to know which was the initial issue. In classic ARDS, we know that it is inflammation in the alveoli, and vascular disease may occur later or as a consequence. Here, the question is: Could it be possible that this is endothelial injury initially and then ARDS phenotype occurs later due to ventilator-induced lung injury or self-induced lung injury?
How did evaluation and subsequent treatment of these patients go up until tPA administration?
Poor: Most of these patients were receiving some form of steroids, conservative management with fluids, low tidal volume ventilation, etc. In general, though, we have been treating these patients as though they have ARDS. Because of the coagulopathy, some patients were receiving anticoagulation but did not improve. Two of these cases were actually on anticoagulation for at least 12 hours without substantial improvement. Again, this raises the question of whether this is a vascular disease that is contributing to a significant portion of their critical illness.
Have findings from other studies been consistent with those from your case series?
Poor: Yes. Another group of researchers in Colorado also lysed three patients and had a similar experience. After lysing, they observed physiologic benefit as well, but the patients ultimately died. And there is something to be said about that. These patients — at least the patients whom we have lysed — are clearly having end-organ dysfunction because of thrombosis and had a physiologic response. Now, that does not mean lysing these patients will result in outcomes that we actually care about, specifically survival and decreased morbidity.
There are several things to consider, though. First, this subset of patients includes some of the most thrombogenic patients I have ever seen, such that, with lysis, they re-thrombose almost immediately. If a paradigm like this is to be evaluated — and we are about to evaluate it — it needs to be done with concomitant anticoagulation to prevent re-thrombosis. Second, we should perhaps consider that it might just be too late in patients who are that critically ill. If — and this is a big if — thrombolysis is going to be beneficial, there is probably a window of time during which it will be effective, and once the patient is past that point, there is too much organ dysfunction, other processes will take over and it does not matter what you do. You also obviously would not want to use thrombolysis too early either because many patients do fine without it. Identifying that sweet spot would be necessary for clinical trials.
Has there been any guidance put in place regarding these types of patients?
Poor: Not yet, as the disease did not exist before December. There are patients who do develop more classic, run-of-the-mill massive PE, and in those cases, thrombolysis is indicated. However, perhaps in the absence of an obvious clot that you can see because it may be microvascular, further evaluation may be necessary. Additionally, we should bear in mind that there are likely different phenotypes in these patients. Some may be more thrombotic, whereas others may have endothelial dysfunction with other characteristics that would not benefit from thrombolysis and treatment would only expose them to risk.
Ultimately, the research right now is very sparse and we are desperate for any guidance. Unfortunately, that can lead us astray and we can do things that are dangerous, which is why we need to express caution, particularly with therapies like thrombolysis and anticoagulation. We should not run around saying that we have found a miracle cure because clinicians and family members are desperate and want to do what is best for these patients. In fact, it would be unwise for us to do things for patients who may get better on standard therapy because an unnecessary treatment could cause a potential complication.
In retrospect, the thrombolysis in these cases was almost a diagnostic test. Of course, I did not do it for diagnostic purposes, but when you look back on it and see how the patients’ situations changed, you see how we can gain information about this disease process based on what changed. That is what happened here, in that this terminal patient population had a dramatic response to treatment. This led us to say that, perhaps, this process is actually occurring earlier in the disease process. Currently, it is still a mystery, though, and people are trying to piece these observations together.
What more would you like to see studied at this time?
Poor: We would like to further study whether thrombolysis would benefit patients who are critically ill. I am also interested in determining whether early anticoagulation would benefit patients and reduce progression of this disease and subsequent respiratory failure, organ damage and survival. Additionally, are there ways that we can phenotype these patients? Are all patients who come into the hospital with respiratory failure the same or are there some who are more inflammatory? Are there some who are more thrombotic? What is the underlying driver of this? Is it a complement-driven process? If these are thrombotic issues, what is causing that thrombotic issue and instead of treating the sequelae with anticoagulation and thrombolytics, can we treat the underlying process with some other intervention or via some other pathway?
And then, clearly, there are risk factors predisposing certain people to worse outcomes with COVID-19, including hypertension, diabetes, obesity and male sex. Why is that? Is there some interaction with those risk factors? It could be that they already have endothelial dysfunction or are at risk for it.
Right now, there is a lot to be unpacked, but I do not think this is your run-of-the-mill ARDS that we see with influenza, and I believe most people agree on that.
Is there anything you would like clinicians to know at this time?
Poor: I would say to just be cognizant of what is going on. If observations seem odd, take note of that because it may be another piece to this puzzle. And again, I would emphasize the need for caution about using potentially dangerous therapies and advise clinicians not to jump to using them without significant guidance from clinical trials and professional societies.
For more information:
Hooman D. Poor, MD, director of pulmonary vascular disease at the Icahn School of Medicine at Mount Sinai in New York, can be reached at email@example.com.
Disclosure: Poor reports no relevant financial disclosures.
Editor's Note: This article was updated on May 11 with several clarifications to the text.