From the first research linking heparin to a reduction in fatal pulmonary embolism (PE), to the development and refinement of risk assessment tools to lower the risk of thrombosis, Joseph Caprini (University of Chicago Pritzker School of Medicine, Chicago, USA) traces the development of thrombosis research. Caprini—who has dedicated his career to the understanding of venous thromboembolism (VTE)—outlines important milestones, addresses key unanswered questions, and considers the latest findings regarding COVID-19 and VTE.
What have been some of the key milestones in thrombosis research to date?
One of the first important milestones can be dated to 1975, when The Lancet published a multicentre study by Vijay Kakkar and colleagues demonstrating that little doses of heparin given to surgical patients did not increase their bleeding, but caused a 66% reduction in fatal PE. A lot of people criticised that trial, but in 1988, Oxford scholar Rory Collins published 70 additional trials conducted around the world using the same protocol and produced exactly the same results. At this point, there were 15-year data on 20,000 surgical patients in 98 centres around the world showing that small doses of unfractionated heparin produced a 66% reduction in fatal PE, with no change in bleeding. In my opinion, to get a death from a prophylactic dose of anticoagulation, except in the case of an unusual type of allergic reaction, is a rarity. However, to have that prophylactic dose prevent a fatal PE is a reality, and that is the key.
In 2005, Kakkar’s son, Ajay Kakkar, produced an excellent article with Sylvia Haas, a brilliant investigator from Germany. They wanted to show that their new drug, low molecular weight heparin (LMWH), was better than unfractionated heparin. Therefore, they constructed a trial with 23,000 surgical patients and decided that the endpoint should be autopsy-adjudicated fatal PE—the hardest endpoint you could get. Kakkar and Haas reported that the incidence of death in 23,000 patients was one-tenth of a percent, whether they got LMWH or unfractionated heparin, showing that the appropriate anticoagulation can prevent nearly all fatal blood clots. In 2012, the Chest guidelines were published, showing that LMWH was better than unfractionated heparin based on the results of 51 trials.
The main point of all of this is that of the 4,000 Kakkar patients in 1975, the 13,000 more patients assessed by Collins in 1988, the 23,000 more patients in 2005, and the 40,000 patients collected from the 51 trials in 2012, the period of prophylaxis was seven to 10 days, which is as important now as it ever was.
How has the world of VTE prevention developed?
A big improvement in VTE prevention came with the move from unfractionated heparin to LMWH, which is a much more specific, much smaller molecule of which 90% of the drug goes to clotting proteins, compared to just 50% with unfractionated heparin.
Another key development was the introduction of direct oral anticoagulants (DOACs), which have revolutionised care. In fact, they are so good they do not need to be monitored under most circumstances. A lot of us who had been in the field for years were used to a culture of measuring, and so questioned how this could be safer. The answer was really very simple, but it took a while to come out, and was that DOACs have a short half-life. Despite the fact that you do not measure them, they go away quickly, and so bleeding deaths were actually lower with these new drugs.
In addition, it was finally recognised that you have to conduct a preoperative patient risk assessment. The community now understands that when you have surgery, it is not just the surgical procedure that is an issue, it is also the baggage you bring to the hospital with you, the risk assessment. In the beginning, the guidelines talked about protecting patients according to type of surgery. The trials were organised according to type of surgery, and so it was very logical that the guidelines said the answer to thrombosis is related to the type of surgery. By 2012, the guidelines acknowledged the presence of risk and they put out two risk assessments: the Roger score and the Caprini score. While the Caprini score has my name on it, the only reason the score is there is due to a dedicated group of individuals. Everybody put their piece into the puzzle and came out with 40 risk factors of varying weights. This changed everything in the surgical world.
What are the latest findings regarding COVID-19 and VTE, and what should be the focus of future research in this area?
There have been some interesting developments so far, one of the most significant being that full-dose anticoagulation, instead of prophylactic anticoagulation, has been shown in a large, multinational trial to improve a patient’s results regarding their need for organ transplants, incidence of deep venous thrombosis (DVT), and death, but that is only in moderate-risk patients. In severe-risk patients in the intensive care unit (ICU), full-dose anticoagulation does not improve results.
It has become clear that anticoagulation alone is not the answer in COVID-19 patients. I think the problem is that ‘clotters’ look at the clotting system, inflammatory doctors at their system, and immunologists at their system, but that is not how the body works. The body is a series of intricate pieces and it is only when they are put together do you get the whole puzzle, and that is where future research needs to go with COVID-19.
Another thing that is missing is risk assessment. Alex Spyropoulos (Northwell Health and Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, New York, USA) has taken the lead here, and has a brand-new publication showing that, by using a risk assessment tool, you can pick out the risk factors in those patients that are very high risk.
What are some of the key unanswered questions in the field of VTE prevention as a whole?
I think one of the key questions has to do with how we collect crucial patient data. What we have learned is that if you come into the hospital short of breath with a severe COVID-19 illness, if you are in a motor accident, or you have a stroke, nobody is going to find out if your aunt had a fatal PE, even though that may be a very important factor in determining your treatment. Therefore, you have got to collect the data ahead of time. One of the initiatives I am involved in is creating a patient-friendly validated form. We are finding that patients want to be involved in their own medical care, and if you can get them to sit down with their family, they can put together a good picture of their medical history. I had the good fortune to be involved with the Global Thrombosis Forum, which is dedicated to including young people from around the world in projects involving thrombosis. We assigned two high school students to go to their schools with the Caprini score to give to their friends at school to take home to their families. In one month, we got 1,500 responses and the number one finding was that there was a 20% incidence of family history of thrombosis. The critics would say this is biased because the families were trying to help the students with their school work, but who cares? We are finding out really important data, and this is crucial to developing patient care.
Research also needs to focus on adding risk factors to the current score, including sleep apnoea, diabetes requiring insulin, body mass index (BMI) of over 40, operations over three and four hours, AIDS, and COVID-19. I think that artificial intelligence and machine-learning tools will have a part to play here, and I envision eventually a universal forum where all of that data with all of the risk factors are there.
Finally, we need to address the fact that there are still a lot of PE deaths. PE may be down, but death rates have not gone down. There are only two scores that I know of that track family history of thrombosis, and I do not know why because family history of thrombosis itself is very important, but one of those, besides the Caprini score, is the British National Health Tool in the UK, which is mandatory, and over a two-year period actually lowered the death rate. These are very powerful data. I want to emphasise that there are many roads to Rome in determining VTE risk, and what we have learned is to use your favourite tool in order to lower the risk of thrombosis. That I think it important.