Last updated on
Hi, everyone. My name is Hanny Al-Samkari, and I’m a physician and clinical investigator in hematology at the Massachusetts General Hospital in Boston. My focus is in disorders of thrombosis and hemostasis. In this episode of Clot Chronicles, I’ll be discussing bleeding and thrombosis in COVID-19, which has been really a hot topic since the beginning of the pandemic. I’m going to talk briefly about a couple of studies, the first of which evaluated the role of various coagulation markers, as well as inflammatory markers, as potentially predictive of bleeding and thrombosis in COVID-19.
We evaluated 400 patients with COVID-19 admitted to the hospital. So, these are patients that 144 of them were critically ill in the ICU, and the remainder were out on the hospital ward. And we found rates of thrombosis that were actually a bit lower than what had been reported in a number of the European and Asian cohorts – radiographically confirmed thrombosis in the 7% range, and rates of major bleeding in the critically ill patients in the study in the 5% range, which was, I think, one of the first studies to report any bleeding rate in COVID-19.
We since followed that up with a study of over 3,000 patients across 67 sites throughout the United States and found very similar rates of thrombosis – and these were all critically ill patients in this study. The time period that we’re looking at was 14-day rates of thrombosis. In the initial study of 400 patients, the median follow-up was 10 days, so a shorter follow-up than some studies that reported higher rates of thrombosis in smaller patient populations. But in this case, a rather large study of over 3,000 patients did find these rates.
The larger study, the >3,000-patient study of critically ill patients, found a rate of approximately 3% of major bleeds. And that used a much more stringent definition of major bleeding—a very stringent definition actually—including bleeding in a critical site, for example, a brain bleed, or bleeding requiring a procedural intervention to address. So, hemoglobin drop alone, for example, was not adequate to be counted as a major bleed, or blood transfusion alone was not adequate. So using relatively stringent criteria, we found relatively high rates of major bleeding.
This is relevant because of the empiric anticoagulation that we give these patients. A lot of centers now are giving escalated doses empirically, either intermediate dose or some centers are even doing therapeutic-dose anticoagulation. That question was also evaluated in our larger study where we, essentially, emulated a clinical trial of these patients using rather complex statistical methods but did not find a benefit in survival for therapeutic-dose anticoagulation started early in the ICU stay (so during days 1 or 2). And this used inverse probability weighting to make sure that each of the groups were evenly balanced, and I think was a well-done analysis and did not have any improvements in survival in those patients receiving therapeutic-dose anticoagulation.
So, a lot of questions remain. We have to rely on randomized trial data, I think, to understand what the proper dose of anticoagulation is in critically ill patients with COVID-19 given the bleeding risk that we observed in these studies.
The other thing to talk about is D-dimer in this disease, which we know is clearly elevated in the vast majority of hospitalized patients. And in our initial 400-patient study, we looked at this, as well as a number of other inflammatory and other coagulation markers, and found, in particular, at the time of initial presentation to the hospital, elevations in D-dimer were significantly predictive of both clotting problems during hospitalization as well as bleeding.
And so, D-dimer seems to have significant prognostic importance in this disease, at least from what we can tell – not just for critical illness and mortality, which we also evaluated in that study and also found similar findings to prior studies that had looked at that, but also for these coagulation-associated complications of bleeding and clotting.
So, what do we take from this? We take from this that coagulation-associated complications—both bleeding and clotting—are a concern in patients with COVID-19. That we need to be very deliberate about the anticoagulation doses we give these patients recognizing that they have a hypercoagulable state, as we have seen from a number of studies, but also that they can bleed just like any other critically ill patient can. And recognizing that, hopefully, randomized studies will answer the question as to the appropriate dose of anticoagulation. And finally, considering the D-dimer that patients have drawn on admission could potentially be predictive of things to come during hospitalization, in particular bleeding and thrombotic manifestations.
It’s been a pleasure to chat with you today, and good luck taking care of these patients.