Transcription
[Applause] [Music] You know, this is just a brief overview of hemostasis. And the bit I want to, you know, basically, to me, the summary is that we are always clotting, we're always making blood clots, and we are always breaking those blood clots down with the help of plasmin involved in fibrinolysis. What this picture doesn't do, which I think is a vital part of blood clot formation, particularly when you have an occlusive thrombus, so an occlusive DVT, is actually looking at blood clot contraction as well. And if you want to read some interesting work on this, so John Weisel's work in Pennsylvania and Valerie Tutweiler wrote, did a lot of this. And it shows that very early on in the formation of a clot, the fibrin will actually contract the erythrocytes down. And they've done work which has now suggested that this is dependent on platelets and their activity, that they are the places which actually pull on the fibrin and contract. So the degree of contraction and degree of stability of the clot may be related to the platelet count or platelet activity. That will also determine, in the couple of studies they've done in patients, that perhaps the likelihood of a deep vein thrombosis breaking off and embolizing may depend on this as well. It's early days with that, but I think we've got to stop thinking that all clots are the same. They are not all the same. Different conditions: sepsis, stroke, DVTs in cancer, DVTs in infection, the actual structure of the clot changes, and we need to be understanding at that sort of more cellular level.
The other thing I'd say is just to be remember of the average timing. Now, this is average, and this is data from, say, orthopedic studies. But on average, if you look at the index moment of, um, damage or increased or risk factors, you know, DVT normally takes between 7 to 14 days to present clinically, and a PE will take up to 21 days. Now, we also know that about 80% of people who have PEs will have had no symptoms of DVT. But when we're looking at kind of cause and effect, did we think that it was that event that led this person to get to a DVT? Often the timing can be quite helpful in this. And then the other thing you need to be thinking of is how long does it take for that clot to adhere to the endothelium, to contract, to stabilize, and for optimal resolution to go on? And the initial stabilization is going to be the first five days, and that'll be why we often have, you know, in the first week, the higher dose of anticoagulation. Optimum anticoagulation, and we would expect a clot to be stable and stable sufficient for us to bridge, um, during the perioperative period after six weeks. And then we would hope after about 12 weeks, three months, we'd have had optimal resorption of that clot. Doesn't mean it's all gone, and you often find there is some residual thrombosis there. About three months is the time that you would feel that you would have, you would have had resolved maximum clot and stabilization.
[Music] So where does cancer differ? Well, there's lots of reasons why patients with cancer have a higher risk of thrombosis. First of all, the cancer itself may release, um, procoagulants. Now, tissue factor is one of them. It may also lead to inflammatory cytokines being released, and different cancers will have different thrombogenicities. So we're seeing data now coming through that patients with, um, glioblastomas, which appears to suggest that if the glioblastomas express the protein polar planing, there is an increased thrombotic risk, and it is lower in other glio patients without expressing that, um, that protein. The other thing is that that is one risk factor. We know the risks go higher if the patients have locally advanced disease, if they've got lymphatic involvement, and even higher if they have metastatic disease. But then we come in and we introduce extrinsic factors, predominantly cancer modifying interventions: so surgery, central line access, chemotherapy, hormones, anti-angiogenics, some of the new targeted therapies, plus immobility because of illness or because of local tumor or node compression. All of these things come together to contribute to this person getting a thrombus. And so rather than thinking about it as one single entity, I look at it as a sort of build-up of lots of different risk factors, and you need a certain thrombogenicity level before you will precipitate that clot. The one thing we don't yet have is a way of quantifying how much each one of those will contribute. We have a bit of an idea, but as, as things stand, we can't give an idea numerically.
This is an old slide now, but I use it for illustration. First of all, because it identifies the various thrombogenicities of different cancers. So we can see that lung cancer is a 22-fold increase in risk of clots, GI 20. It's slightly higher in patients with pancreatic cancer. It's one of the most thrombotic of cancers. And then hematological, particularly I'm thinking of myeloma patients here. And fortunately, as a group of colleagues, hematologists are very good at considering primary thromboprophylaxis in the myeloma patients receiving steroid and lenalidomide. Breast cancer, early stage breast cancer is probably twice as clotty as the average person, so two per thousand. You add an adjuvant chemotherapy to that, it can go up to about 2%. If you're looking at a patient who's got metastatic disease and breast cancer, and they receive palliative chemo, the risk goes up to about 8%. And you can see on this slide that distant metastases have a considerable impact on thrombotic risk. Think it's interesting as well where we look at time to cancer diagnosis, and the majority of clots occur in the first three months of cancer diagnosis. Now, I think that's because we are hoping that these people are being diagnosed early and then go straight on to cancer modifying treatments. And I think what we're looking at here are those patients in whom we are picking up incidental VTE on the staging scans. But I think this is predominantly due to provocation from anti-cancer therapies, chemotherapy in particular. My clinical practice, where we see about 400 new patients a year, I'd say that about 60% of those will have had their clot precipitated by systemic anti-cancer therapies. And the reason this is important is first of all, we need to appreciate the psychological impact of having cancer-associated thrombosis. This is within the first three months of them being told already they have a potentially life-threatening condition, that is the cancer, and now within three months of them having that, we're telling them that they've got what they will read up to be a potentially life-threatening complication of their treatments. And I think they are very rarely given that information in the same sensitive, timely kind way as we do when we break the news that they have cancer. It's usually a bit of an in-and-out job. You start them on an anticoagulant, say we'll get the district nurse round, and we'll see you in clinic soon, we'll refer you to the CAT clinic. I don't think these people get as much information or support as they would like.
Finally, to illustrate that not all clots are the same. This is an electron micrograph from colleagues in Swansea in the hemostasis and biomarker research group. And this is an electron micrograph of blood that has been allowed to coagulate normally. And this is the same magnification in a patient with stage four lung cancer. And they use a technique called fractal dimension. They published a paper in 2010 in Blood on this. And what this identifies is that at the point at which blood coagulates, when it goes from a liquid to a solid, there is a moment where you have this viscoelastic gel. And if that is exposed to various vibration frequencies on a rheometer, they will, you will find that there is a common point at which normal blood will always gel. And they can use this as a template for the incipient blood clot, which then works, works up into be the main clot. So you can tell at an early stage what sort of clot microstructure you're going to see. And so, um, we had been doing some work which was showing that lung cancer patients, the more advanced the cancer, the more dense the clot is, and therefore the more resistant that clot is going to be to lysis through normal lytic systems. So I think we need to look upon the cancer journey with respect to thrombosis as an ebbing, waxing, and waning of thrombotic risks. There'll be times you will have a lot of cancer releasing procoagulants, there may be chemotherapy which will increase that risk. There'll be times where there's very little tumor burden, or then metastases may come back, and they will then have metastases and palliative chemotherapy. We also know that as people near the end of life, we're seeing about 50% of patients with evidence of thrombus at postmortem, for example. So this is not something that just happens once. So your average, say, your orthopedic patient who has a broken hip, you know that you had your ground zero of thrombotic risk, you know that they are at greater risk for the next 90 days. Cover them for that, they'll be alright. With this, the risk is ongoing, and it will change. But not only that, whilst we're giving patients lots of different anti-cancer therapies, we're going to alter their risk of bleeding as well. As these people have chemotherapy nadirs, for example, as we are giving them various blood products, they, the risk factors may go up if we're giving them colony stimulators, etc. The other thing to know is this is not just a chemotherapy issue. We have data now which shows that the anti-angiogenics, the immunotherapies, the other targeted therapies, all have an increased risk of developing VTE. And it tends to be an accumulative risk. So the longer you are on something, the greater the risk is. We are facing an issue where our population is getting older. We are able to treat people with advanced cancer for longer. So it becomes very normal that patients with renal tumors, breast tumors, colorectal tumors, melanomas, for example, will often spend several years receiving systemic anti-cancer therapies with metastatic disease. And so therefore, the risk is ongoing. So this is a growth industry, and it becomes far more nuanced as well with the various chemotherapy and systemic anti-cancer therapy agents. And I would urge people to consider whether or not you need a more specialist approach to management of cancer-associated thrombosis rather than just having them all chucked in at your thrombosis clinic.
So let's just go to the data that has informed our practice for the last 20 odd years. So 2003, Agi Lee and colleagues published the landmark CLOT trial, which compared low molecular heparin with warfarin and identified that the rate of recurrent VTE was halved in patients receiving dalteparin, the low molecular weight heparin, compared to modern studies. The numbers here were very high for VTE, and when we look at these studies later, we'll see it's because we had a much more poorly, poor performance status, metastatic disease population of patients compared to the latest studies. And yet, all the studies looking at low molecular heparins, when put together on our forest plot, says quite conclusively that we should be looking at low molecular heparin rather than warfarin in the management of cancer-associated thrombosis. And let's not remember, let's not forget that these were patients in a clinical trial setting. Often we will find that warfarin, maintaining a stable INR, particularly with chemotherapies and supportive care drugs, the INR can go all over the shop. And actually, as the cancer progresses, the risk of bleeding, even when you maintain a stable INR, goes up.
Here is the late, great Prince, or he was for a while just known as AFcab, the artiste formerly known as Prince. And similarly with what we call the direct acting oral anticoagulants, they went through various iterations of names. When they first came out, they were the new oral anticoagulants, NOACs. And then when they were no longer new, but we liked calling them NOACs, we called them novel. So then they were the novel oral anticoagulants. And then there was a parting of the ways between cardiologists and hematologists. Cardiologists wanted to continue to call them NOACs, and so they were non-vitamin K oral antagonists. And the hematologists changed it over to DOACs, direct acting oral anticoagulants. Now, for the sake of this talk, I am not going to spend much time on dabigatran at all. Um, there's a lot of reasons for that. One in particular being that we don't have any cancer-associated thrombosis studies in that population. Its behavior, its safety profile, I think does differ to the others. So I'm not going to focus on that. With respect, the other three, rivaroxaban, apixaban, and edoxaban, they all have little nuanced differences. I think the most important thing to look at is the renal clearance, how that can vary between the different agents. And so theoretically, one would think that apixaban will be, um, safer in patients with, um, poor renal function, although their SPC would advise caution against using it in people with a creatinine clearance less than 30, which is not that different to rivaroxaban or edoxaban. Although edoxaban, they would just say half the dose if you've got creatinine clearance between 15 and 30. The other thing to be mindful of is CYP3A4, um, metabolism. Very minor with edoxaban. It is, it says minor on this, but actually I think, um, we're seeing now that apixaban does have a more significant metabolism with, um, CYP3A4. And then also P-glycoprotein, which is a substrate protein. So that's, these are the things where they vary somewhat.
So there's been a few studies. Originally, there were studies just for the treatment of thrombosis, and there was a smattering of very well, healthy, early stage cancer patients, and certainly not enough data to convince us that we could use them first line in CAT. The first study was AMPLIFY VTE Cancer, which gave edoxaban in its usual manner against dalteparin in its usual manner. And they used a primary outcome which was a composite of recurrent VTE and major bleeding. And so that was their, their primary outcome, SELECT D. They called it a pilot study. I mean, it was very large numbers for a pilot study, but this was looking at VTE recurrence at six months, and then looked at bleeding as secondary outcomes. So interestingly, AMPLIFY VTE Cancer, they, they showed this, this was non-inferior when you looked at the composite of VTE or major bleeding. There is no difference between the two drugs. But if you were to separate that and have a look at recurrent VTE rates and major bleeding, it looks like edoxaban is actually more successful at preventing VTE, but that is at the cost of a higher rate of major bleeding. Now, this was the first paper that came out on CAT, and it was absolutely scrutinized. And I am one of the sad individuals which has read all 32 pages of the appendix over and over. And what becomes very clear is that when you look at the bleeding, the major bleeding, it was highest in patients with GI cancers, and in particular upper GI cancers. It was also high, about 8%, in patients with urothelial cancers. People say it didn't reach statistical significance, but if you look at the numbers, you will see that it was actually, I'd say it was clinically significant. And our practice since then has shown that bleeding with urothelial cancers is a significant issue. They may not qualify as a major bleeding, but they are certainly clinically relevant non-major bleeding. Look at the SELECT D study. SELECT D was a similar methodology. The only major difference I'd say is that early on, they stopped recruiting esophageal and gastric cancer because they were seeing excessive bleeding in the DOAC arm. And so what you are seeing there again is the recurrent VTE rate is lower, but it is at the expense of major bleeding, and in particular, clinically relevant non-major bleeding. And I think this is where we've got to really think about what is important to us as researchers. And we're saying, well, fantastic, you know, we didn't have that many major bleeds, which are the ones that kill you, bleed into a critical area, or mean you have to go into hospital for a transfusion. But actually, they are clinically relevant, and they do have a massive impact on the patient's quality of life or how we manage the clot. So I think we do need to take more notice of those. And we can see here when we compare them together. So this was one of the first meta-analyses, if you can call a meta-analysis comparing two studies, and it showed DOACs are very good at preventing clots, but there is a bleeding risk. So one of the things we've had to do is actually identify who are those patients. So we should use them with caution with one thing. We do need to think about is renal insufficiency, and there are recommendations for each of those. The reason this is important in the cancer situation is first of all, we have a lot of, a lot of elderly patients in the cancer population, and they reckon about 50% of elderly patients will have a creatinine clearance less than 30. The other thing is we're very good at giving patients anti-cancer medicines which are good at knackering your kidneys, even if we just lack them for a short period of time. So it's not just the thrombotic risk that will wax and wane over time during a cancer patient's journey, also the bleeding risk may wax and wane as well. The potential for drug-drug interactions is quite significant as well. And I used to have a very nice leaflet with about, you know, 40 anti-cancer medicines and supportive medicines. And now you can go online and probably find about 40 that end in 'ib', let alone all the other anti-cancer medicines. But anything that is a strong inhibitor of CYP3A4 and P-glycoprotein, or strong inhibitor of CYP3A4, a moderate inhibitor of P-glycoprotein, is likely to lead to higher levels of the DOACs. And there is some data now which shows that that does clinically translate into an increased bleeding risk. Some colleagues may say, well, I will use edoxaban then because the CYP3A4 activity is so much less. My approach at the moment is, let's go for the big wins. Let's go for the majority of patients that we can manage who are not on these drugs. Make sure we've got them being used safely. And then if we have people who are unable to tolerate the low molecular heparin, yes, we can consider these in a few patients, but I would be very cautious in those you've got strong inhibitors of both. The other thing is to be mindful of some of the supportive care drugs. Dexamethasone can be quite a swine. I think it's okay in patients who just have a single dexamethasone boost at the beginning of their chemotherapy. But you whack someone on 8 milligrams of dex, and you will see a decrease in the efficacy of the DOACs.
Finally, the CARAVAGGIO study. And this came along. So you'd have thought that, you know, this is the one which was going to answer whether we are seeing a class effect or whether apixaban has any benefits itself. So patients received apixaban in standard way again, and dalteparin in dropping the dose on day 30, as it's licensed. And what you saw here was apixaban had a lower rate of VTE compared with dalteparin. And so once again, that did appear to be, um, more favorable from that point. But, you know, does that mean that the drug was better? It possibly, I mean, you know, it works well. But this, I think, is interesting when you look at this Kaplan-Meier curve, and this is in the appendix, actually. The curves don't start separating until 30 days in, which is actually at the same time that we drop the dalteparin dose by 25%. Now, we drop our dalteparin dose by 25% down to 150 units per kilogram because that's what was done in the CLOT study. Interestingly, the CATCH study, which was tinzaparin in cancer-associated thrombosis, they didn't drop the dose down, and there wasn't an increase in bleeding risk. And so, you know, many of my colleagues will not drop the dalteparin down, um, because, you know, they don't think there's an increased bleeding risk, but they do believe that it may help prevent more clots. And I certainly think this data suggests that perhaps. But interestingly, once again, when we look at the bleeding rates, although there wasn't a great difference in major bleeding, and certainly there wasn't a major difference in GI bleeding, although that may be because we had so few patients, clinically relevant bleeding that was higher in the apixaban group. So once again, when we have a meta-analysis of the RCTs, we are seeing here that from a recurrent VTE point perspective, it favors the DOACs. But from a major bleeding perspective, when you have not taken out those patients who we now know are at risk, it would favor low molecular heparin. So I think the message that I say is, if you have patients in whom there are no drug-drug interactions or no significant drug-drug interactions, in whom the renal and hepatic function is satisfactory, and they are not GI cancers with the two endoluminal tumors still present or urothelial cancers, I'd use DOACs. Now, if it's an endoluminal tumor has been removed, yes, I'm happy to use a DOAC. So someone who's had an esophageal tumor resected, someone has had the colorectal cancer or gastric cancer resected, yeah, go for it. But if you've got an intact urothelial tumor, especially the bladder tumors, or you've got a GI tumor, I wouldn't be using a DOAC.
Kind of, I was a bit surprised when I saw this, um, appear on my, um, appear on my website. And I actually had an email from a professor smugly who said, oh, should we be changing what you're doing with your CAT clinics at the moment? For safety, we suggest everyone gets started on a low molecular weight heparin, gets referred in, we'll see them as soon as possible, and those who we can convert over to a DOAC, we will. But that way, we feel it is saving people from making the error of giving, um, a DOAC inappropriately. So when I saw this, what on earth was going on? Was there some study that I'd missed out on and hadn't heard of? And then what I found was actually this was, um, a systematic review and meta-analysis. Will it be? It's a, all bells and whistles one, because this was a living interactive systematic review and network meta-analysis, but wasn't telling us anything we didn't already know. But it had analyzed things again without, um, you know, differentiation. And, you know, these headlines, I sometimes feel in the wrong hands can be, can be a concern. People need to understand the caveats. They need to understand the safety perspectives. And what's interesting is, you know, really, there should only be one systematic review and meta-analysis, because that's all you need. But since 2015, since the DOACs have been introduced, um, there's been over 40 guidelines for cancer-associated thrombosis, consensus documents, and also now people are making, um, you know, getting measure out of, um, writing commentaries on guidelines and comparing the guidelines as to which one had the greatest rigor. So which one should you believe? Because actually, the guidelines do, if they use different methodologies, will come out with different answers. Let's not forget that some will just look at efficacy, some will look at health economics, some will just be what we call GOBSAT, which stands for Good Old Boys Sat Around a Table. And people just sit there at a meeting which has been hosted by a pharma company with a, which has paid for a medical writer, and then everyone then puts their name to a paper. So, you know, there's lots of different guidelines out there. Whether we need them all, I don't know. Um, they're great when there is evidence, but they are not any use, I think, when there is no evidence. And sometimes I think the least helpful thing is to come out with a 60-page document and then after all of this work, it goes, we can neither recommend for or against the use of rivaroxaban in the treatment of whatever. That's no help. And yet you've spent time, you know, tens of thousands of pounds doing a review on that. So it's only good when you've got data. But the thing is, they don't apply for every patient. And actually, when you look at the studies, those who get into them, they've already been selected according to the inclusion and exclusion criteria to try and minimize the complications. You don't want a study where everyone is, is bleeding or thrombosing all the time. You don't want that. You want it to be a safe study. But likewise, you can make a subject study that is got so few risks at all that actually the study is only relevant to 10% of the population. So if we look at these studies, and if I'm wrong, please, you know, someone email me, I'm fine with that. But I've looked at apixaban cancer, and I know that there were 1050 patients randomized, but there's no documentation that I could find of a screening log, which would say how many patients were actually screened. So what proportion of the population with cancer-associated thrombosis that went through these services were actually approached for randomization? And I think that's important because otherwise, this is not going to tell us how representative of the population we're talking about. Likewise, CARAVAGGIO, 1770 patients underwent randomization, but we've got no idea how many were screened. The only kind of hint we can get is from SELECT D, where if you look here, 2060 patients were assessed for eligibility, of which just over half did not meet the inclusion criteria, and then another proportion declined to participate. So in the end, those that were assigned just represented 25% of the population. Although theoretically, assuming that those who declined to participate in the other reasons would have been eligible, let's for the sake of arguments say 50% of the patients screened did not meet the criteria. So that's telling me that these patients that we recruited are only representative of half our population of patients we see. And unfortunately, we don't get a choice. We still have these people with clots in front of us that we have to treat, even if the data in the studies isn't going to help us.
When we look at the exclusion criteria, any patient with a life expectancy estimated of less than three months wouldn't get in. Um, CARAVAGGIO and CATCH, they had to be even healthier, kind of less than six months. Now, one thing I do know is that, um, as best as we can, doctors are rubbish at prognostication. In fact, there are systematic reviews and meta-analyses which say we are rubbish at saying how many months someone's got to live. We're pretty good at suggesting when someone's in the last week of life, but otherwise, we're no good here. So this tends to be an estimate at best. Anyone who's not well enough to come into clinic to receive treatments, they're not in. So poor performance status, poor renal function, we said, and poor hepatic function. So that is a significant proportion of patients who we also know are far more thrombotic, the patients who poor performance status, those who've got more advanced cancer. And then if we look at the studies overall, and if you look at this six-month mortality, the CLOT study, 40% of patients were dead at six months compared to, you know, about 25% on average of the DOAC studies. Metastatic disease varies. The CARAVAGGIO says 68%, but that also included locally advanced cancers. But you can also see the number of patients with performance status 2, because they were having people with either 0, 1, or 2, the numbers with a low performance status, poor performance status was much lower in the DOAC studies. And then when you look at how many patients had systemic anti-cancer therapies, another risk factor for clots, you know, the CLOT study had 78%, which once again explains, we've got metastatic disease, poor performance status, and systemic anti-cancer therapies, that is why these people had much higher rates of VTE in the study. But overall, I'd suggest that we've got better prognosis patients and fitter patients in the, um, DOAC studies. Also, I think age. I mean, there's a paper I've just recently seen which is looking at cancer patients with clots with an average age of 77. I think the patients who went into these studies were younger than the general population. So they're great, these guidelines for a select group of patients, but they're not there for everyone. And we've got to remember that medicine is as much an art as a science, and we need to understand the pharmacokinetics, pharmacodynamics, patient preferences when we are looking at how to manage CAT. We can't just go, the guidelines say this, that is what we do. I think people forget that when David Sackett wrote his, um, his seminal editorial on evidence-based medicine, they talked about this Venn diagram of three things, which was the best relevant evidence, clinical experience, and patient views. But I think sometimes evidence-based medicine has got to the point where we're only looking at the data and not looking at these other things as well, which are as important.
DOAC-related GI bleeding. We have to be careful of that. There's various reasons. Dabigatran, we, we know has got a direct caustic effect. But I think what we also know is, say, apixaban is absorbed throughout the intestine, and so you do have a topical anticoagulant effect on the mucosa. Sometimes the DOACs will inhibit mucosal healing as well. And if you've got a highly vascular with friable blood vessels on a tumor, you know, this is going to have a topical effect which will lead to a greater risk of bleeding. Normally, we would look at thromboprophylaxis, and we would see the case for it. We've seen that the mandatory risk assessment and prescribing of appropriate stronger prophylaxis has led to a reduced death, uh, by 15.4% in patients admitted to acute hospitals. And we've seen some change in practice in hospices. I mean, in 2000, when I did a survey, only 2, 2% of hospices had policies, of which some of them, the policy was, we don't do thromboprophylaxis. This has increased in 2006. This kind of precedes the NICE guidelines, but then in 2010, NHS England mandated reporting risk of VTE with the target of 90%. Now, there's only a small proportion of hospices that are actually run by the NHS. The charitable sector runs the others, so they weren't subject to this mandate. And there was actually a real resistance to thromboprophylaxis, even though there were data to suggest there is a large proportion of people who have VTE in the advanced cancer setting. When we talked to patients with, we talked to people, directors of hospices, they didn't really feel that thrombus was a big problem. They didn't see many. And I think the argument to that is that actually there were a lot of patient patients who had symptoms attributable to VTE, but people would consider breathlessness more due to lung mets, plural effusions, fatigue, anemia, etc. Similarly for leg swelling. There were those who thought a large PE was a nice way to go. If you see me at ISH, ask me about that and watch me spin off the walls. I don't have time to go into that at the moment. They suggested patients wouldn't accept a daily injection of low molecular heparin. We did a qualitative study published in the BMJ a few years ago, which showed patients, some consider themselves like any other patient and are happy with it, um, if there is a reason for it. But also people felt that the data which informed our practice was not appropriate to the hospice population. So the landmark studies which showed us low molecular heparin had a use in preventing VTE, they're 20 years old now, and they were using outcomes looking at radiological recurrence. When we need to be thinking about symptoms perhaps, rather than whether or not someone had radiological evidence of a clot. So we undertook a study called HIDDEN, and we had five hospices in which we had specialist nurses trained up to undertake three-point compression ultrasonography to have a look at the true prevalence of VTE in cancer patients coming into hospital. Every scan was reviewed by a consultant radiologist and confirmed to confirm the results. And we screened 1,390 patients, of which 840, so 60% of the patients in the hospice were ineligible. Now, I think that's important to see that we did not feel that thromboprophylaxis was appropriate in patients who were coming in to die, or we thought were likely to die, because we're not there to prevent something, prevent early death. If this person's dying in the next few days, we're trying to prevent symptom burden. Some people, it was not possible to undertake ultrasound, some people were unable to consent, um, sometimes they were consented outside the time frame because of weekends. But for a hospice study to be able to recruit 343 patients, I thought that was pretty impressive.
So first of all, just to know the demographics were that this was a poorly bunch of patients. Mean survival 44 days, so you know, these are less than three months, these are less than two months to live, most patients. And they had an Australia Karnofsky performance status, which is a more detailed form of ECOG, really, of 49. Now, that means these people are predominantly dependent on nursing assistants to carry out their activities of daily living. Once again, poorly people. We found that when we looked at the initial valuable scans, there was a 34% prevalence of femoral DVT. So we're talking big, chunky clots. We didn't look at calves, we just looked at three-point femoral, um, and we found that those with the greatest number, it was associated with previous thrombosis or patients who had become bed-bound over the last 12 weeks, so a slow deterioration. Now, we acknowledged that there was a learning going on here, and certainly the first few months of scans, I would, I would guess or I presume that the staff weren't pushing hard enough. So we actually excluded the first few months of scans in case there was, um, you know, some error there. But even when we did that, um, and we looked at those where people had really, you know, developed their skill in scanning, we still saw, we would have a prevalence of 28%. So we're looking at about a third of patients admitted to a hospice already have radiological evidence of VTE but didn't have symptoms from it. And furthermore, there was no survival difference between those who had femoral DVT and those who did not. So these didn't go on to cause symptom problems, they didn't go on to shorten life. And so one could argue that patients who had an asymptomatic femoral DVT, if these are people who've had a progressive deterioration in their condition, that a Karnofsky less than 50, these people are likely to have clots, have no problems from them. So I would argue that there is no benefit in giving them thromboprophylaxis. We've already said that 29% of people going into a hospice are receiving end-of-life care. So we're looking over, you know, 50% of patients in a hospice are unlikely to benefit from thromboprophylaxis. What we don't know is, is this true for all palliative care patients? Because your average hospice will be able to look after a small proportion of patients with advanced cancer, and the majority are admitted to the acute hospital. And so we've just opened HIDDEN 2, where we're doing exactly the same study with, um, vascular scientists doing the dopplers this time, and we are looking to see whether or not the incidence is the same in the advanced cancer patient who may have actually a better performance status prior to coming in.
Finally, and we only have a few minutes, but I think it's worth doing this. There's another hospice study here in which our French colleagues. [Music] And I still haven't forgiven them for not giving us any points. Oh, no, they did give us a point, sorry. Who was it who didn't give us points in Eurovision? The French gave us 12. Sorry, guys. Apologies. Mondia. Um, 22 specialist palliative care units, just under 1200 patients. There was clinically relevant bleeding observed in just under 10%. So clinically relevant bleeding is major bleeding and clinically relevant non-major bleeding. When we look at the sort of patients we had in the hospice, their majority are those with metastatic cancer, um, 20% receiving chemotherapy, um, 44% receiving prophylactic low molecular weight heparin, just under 6% having treatment, and then 14% on anti-platelets. We found the greatest risk would be people who'd had previous bleeding, um, people who are on platelets, and people who are on anticoagulants. And so those are the things that increase your risk in a hospice of having a bleed. So once again, one needs to consider how beneficial it is to be on these drugs near the end of life. We looked, and this is finally, we looked at a patient group who'd been going through the cancer, cancer-associated thrombosis service, and we just followed them up until death and then looked at their notes leading up to their last, last few weeks of life. And our 450 patients, there's about 214 who died over this period of time. And what we saw first of all was that the majority continued anticoagulation up until the point of death. So even though we often will rationalize medicines as people are ending, entering near the end of life, people are scared of stopping these. And I think people have a fear they may have another clot. What do we do about it? When, of course, these people, even if they did have another clot, we could treat them with end-of-life meds. And also, when you stop the anticoagulation, how long is it going to be before that clot actually reforms and causes symptoms? The reason I'd suggest we should be stopping our anticoagulation as death approaches or before even is because 7% of these guys had clinically relevant bleeding. And I say that because this was retrospective notes, and so that was the amount that actually made it into the notes as being an issue.
So with just five minutes to go, one of my messages, first of all, clots are cool, guys. And palliative care, sorry, cancer patients, they're a great patient population. You know, I think one of my favorite parts of the week is my CAT clinics. But the trials we've got at the moment aren't in representative populations for all of our cancer patients. Some of them yes, but not all. I think in patients who are getting more poorly, have a low threshold for stopping the anticoagulants. And if these people have had, have got poor performance status, have been gradually deteriorating, and you think they have a short prognosis, I don't think there's any utility in giving thromboprophylaxis. So thank you very much for your time.
[Music] [Music]