The ‘low density lipoprotein (LDL)-cholesterol hypothesis’ is widely accepted as the best model explaining how atherosclerotic cardiovascular disease (ASCVD) develops. It states that a raised LDL level causes thickenings within arterial walls which gradually develop into larger plaques.
More vulnerable plaques may then ‘rupture’, triggering the final thrombotic event, fully occluding an artery, leading to events such as myocardial infarction (MI) and strokes.
However, the association between a raised LDL level and the development of ASCVD is weak. Indeed, many studies have found no association or even an inverse association, particularly in the older population, where most deaths from ASCVD occur [1▪].
The inability of LDL to accurately predict risk highlighted by the fact that the two most widely used calculators for ASCVD, in the United States and UK, do not use LDL levels to establish the risk of a future cardiovascular (CV) event [2,3].
Whilst it is true that statins reduce the risk of ASVCD, the absolute reduction in risk is small [4▪▪]. Additionally, several agents that significantly lower LDL demonstrated no benefit on ASCVD, including cholesterol ester transfer protein inhibitors such as torcetrapib and evacetrapib. Evacetrapib reduced LDL by 37% and increased high-density lipoprotein (HDL) by 120% but had no effect on CV risk .
Examining ASCVD from a process perspective, plaques have been found to contain lipoprotein remnants, assumed to originate from LDL. However, there remains the likelihood that the majority of these remnants actually originated from lipoprotein(a) (Lp(a)). This points to a different causal model, as Lp(a) plays a role in thrombus formation and lysis, not lipid transport.
The different causal model was first postulated in 1852 by Karl von Rokitansky, which he called the encrustation hypothesis. It is possibly better known as the ‘thrombogenic hypothesis’ . He proposed that ASCVD develops as a result of a dysfunction in a normal three-step process.
Step one damage to the endothelium/glycocalyx. This damage leads to step two, the formation of a thrombus to cover the damaged area. The thrombus is, in turn, covered by a new layer of endothelium which effectively draws the thrombus into the artery wall.
Step three is that the remnant thrombus is broken down or lysed by various repair mechanisms, such as macrophage action, which can break down and remove the remnants of damaged material.
However, if damage is accelerated, plaques are bigger and/or more difficult to break down, or repair is hampered, plaques can develop and grow.
Therefore, the thrombogenic hypothesis proposes that factors that can increase the risk of ASCVD are those that do one of three essential things:
- (1) Increase the rate of endothelial/glycocalyx damage beyond that which the repair mechanism can handle.
- (2) Drive the formation of larger and/or more difficult to lyse/remove thrombi.
- (3) Interfere with the repair processes (see Fig. 1).
To put this another way, if the process of damage occurs faster than repair, plaques will form, and grow larger. If the repair systems are working at a rate matching the damage, plaque growth will be slowed, or prevented.
This is a concept supported by the paper ‘Why atherothrombosis is in principle a hematologic disease’. The authors present evidence that conditions, and drugs, which affect thrombosis also affect the development of atherosclerotic plaques. They propose that plaques develop from the organisation of mural thrombi that have not been fully cleared away by the repair processes, thus creating a ‘vulnerable’ area for further thrombi to develop .
RECENT EVIDENCE IN SUPPORT OF THE THROMBOGENIC HYPOTHESIS
This review will look at the three interlinked processes involved in atherosclerotic plaque formation
- (1) Glycocalyx and endothelial damage
- (2) Formation of larger/more difficult to lyse thrombi
- (3) Interference with repair processes
GLYCOCALYX AND ENDOTHELIAL DAMAGE
All blood vessels are lined with endothelial cells which are, in turn, covered by a ‘gel’ layer, the glycocalyx, which is where nitric oxide (NO) is synthesised, along with many other anticoagulant factors (see Fig. 2).
The glycocalyx acts as the protective layer necessary to maintain vascular endothelial cell function, and homeostasis. It is also important to protect the underlying endothelial cells from direct physical damage .
It has been increasingly recognised that in many acute illnesses the glycocalyx is thinned and weakened. This, in turn, increases the risk of acute cardiovascular events [9▪▪].
For example, destruction of the glycocalyx/endothelium is critical in sepsis, where it is a major contributor to the underlying pathophysiology. The destruction of the glycocalyx is primarily due to exotoxins released by bacteria in the bloodstream. This, in turn, is followed by widespread thrombus formation, known as disseminated intravascular coagulation .
In Sars-Cov2 infection the glycocalyx/endothelium is also attacked and weakened, and this is the trigger for the blood clots that are commonly seen with coronavirus disease 2019 (COVID-19) infection [9▪▪].
Unlike sepsis, where the glycocalyx broken down by an external agent, with severe acute respiratory syndrome coronavirus 2 (SARS-CoV2), the damage occurs within endothelial cells. Endothelial cells in the lungs, and the vasculature, have a high concentration of Angiotensin Converting Enzyme 2 receptors, and SARS-CoV2 hijacks this receptor to gain entry before multiplying and ‘bursting out’. So, the cells are damaged from ‘within’, rather than in sepsis, from ‘without’. The end result is similar.
It is now also known that there is also considerable overlap between SARS-CoV2 infection and Kawasaki's disease. In both conditions there is a form of widespread vasculitis, with the glycocalyx/endothelium seriously compromised. With SARS-CoV2 the presence of virions within cells provides an additional target for ‘attack’ by the immune system, increasing the damage .
It seems that the process Kawasaki's is very similar, although an infectious agent has not been identified. In Kawasaki's, ASCVD can develop rapidly, and in the years following the acute episode the risk of a CV event can be increased by >50-fold .
The acute problems, seen shortly after an initial triggering event, highlight the role that the healthy glycocalyx plays in preventing thrombus formation. However, it is also hypothesised to be critical in the longer-term progression and development of ASCVD [13▪].
Looking at longer term conditions, one important factor known to cause chronic damage to the glycocalyx is a raised blood glucose level . Interestingly, patients with type II diabetes are at an increased risk of ASCVD, sepsis, and other infection diseases. Possibly due to a damaged glycocalyx allowing infectious agents to enter cells more easily .
Here, we have a possible ‘positive reinforcement loop’, increasing both infection risk and thrombus formation, which may also be why patients with type II diabetes are more susceptible to severe SARS-CoV2 infections.
Other chronic conditions known to damage the glycocalyx and increase the risk of CVD include:
- (1) Systemic lupus erythematosus (SLE) 
- (2) Smoking 
- (3) Hypertension 
- (4) Antiphospholipid syndrome 
With antiphospholipid syndrome (APS), the immune system attacks the endothelial cell membrane itself, creating widespread vascular damage. Many patients with SLE have APS, and it is this subset of patients at greatest risk of ASCVD .
It is not within the scope of this article to describe all conditions, or agents that can damage the endothelium/glycocalyx. However, an increasing body of evidence suggests that any factor, or disease, with this effect, increases the risk of ASCVD [21▪▪].
FORMATION OF LARGER/MORE DIFFICULT TO LYSE THROMBI
Once the endothelium has been damaged this triggers thrombus formation. One of the key mechanisms here is that an undamaged and healthy endothelium expresses tissue factor pathway inhibitor (TFPI). TPFI limits the actions of tissue factor (TF), which is perhaps the single most potent pro-coagulant factor.
Thus, when TFPI production falls, TF is released to drive thrombus formation [22▪]. In addition to this mechanism, the glycocalyx is where NO, a highly potent anticoagulant, is synthesised (indeed, there are a host of anticoagulant actions which require a healthy glycocalyx to function). This means that, when damage occurs, the endothelium shifts towards a prothrombotic state .
It would not be possible to review all the factors involved in thrombus formation and breakdown, as this is a highly complex process. So it may be most useful to look at three important players: von Willebrand factor (vWF), fibrinogen, and Lp(a) to review their impact on thrombi.
These three factors operate in very different stages of coagulation. vWF has a key role in formation of the initial platelet plug, fibrinogen is the raw material of the final ‘fibrin’ clot, and Lp(a) is critical in thrombus breakdown or lysis.
Von Willebrand factor
vWF is a glycoprotein required for the formation of haemostatic ‘plugs’ and arterial thrombi. It binds to platelets, factor VIII, and collagen in the underlying artery wall.
A meta-analysis of patients admitted to hospital with major adverse CV events found that they had significantly higher vWF levels. The difference in CV risk versus those with normal levels of vWF was very nearly 50% .
Fibrinogen has a key role in the final act of thrombus formation. After the platelet/erythrocyte plug has formed, strands of fibrinogen link to form fibrin, which wraps around the plug, binding it. This is the final step in the clotting ‘cascade’.
A Scottish study of risk factors for coronary heart disease showed that high levels of fibrinogen were associated with a greatly increased CV risk, for both men and women .
In a more recent meta-analysis, the increased risk was independent of other CVD risk factors including lipid levels, and antithrombotic and other medications, such as aspirin. It is of interest that the risk of ASCVD was much greater in patients with type II diabetes, who have higher levels of other pro-coagulant factors [26▪]”.
The issue of Lp(a) is complex, but important. Lp(a) is identical to LDL, apart from the attachment of an additional protein, known as apolipoprotein(a) (apo(a)).
This apparently small difference is highly significant, because apo(a) is virtually identical in structure to plasminogen. The key difference is that apo(a) has different folding structure (kringle) at one end. It is this ‘folding’ difference that turns apo(a) into, effectively, ‘antiplasminogen’.
Plasminogen is incorporated into all thrombi as they form. It is inactive until it is converted to plasmin by tissue plasminogen activator (TPa), which is an enzyme synthesised by endothelial cells. Once TPa has triggered this conversion plasmin lyses strands of fibrin, essential for the breakdown of thrombi.
However, if there is a higher concentration of Lp(a)/apo(a) within thrombi, this blocks the action of TPa on plasminogen, resulting in a thrombus that is highly resilient .
Lp(a) also binds to TFPI. Thus, Lp(a) can stimulate clot formation and make it more difficult to remove the resultant thrombi. A raised Lp(a) is associated with a tripling of CVD risk [28▪▪].
Why does Lp(a) have this function? Linus Pauling first hypothesised that, because humans cannot synthesise vitamin C, which plays a key role in the production of collagen, a lack of vitamin C will lead to a breakdown and ‘cracks’ in blood vessel walls.
Lp(a), which is almost exclusively found in animals that cannot synthesise vitamin C, binds very strongly to the endothelium. and the underlying arterial wall, creating thrombi that are particularly ‘tough’. This reduces blood loss, allowing the animal to survive until sufficient vitamin C can be consumed .
However, this protective role is a double-edged sword. As Lp(a) can also drive the formation of thrombi which are more difficult to lyse. This, in turn, results in accelerated plaque formation.
Whist this review has only looked in more detail at three of the pro-coagulant factors/conditions, all of which can significantly increase CV risk, there are many others including type II diabetes, Cushing disease, APS and raised factor VIII .
To briefly examine this area from a different perspective, it is worth looking at people with haemophilia. A recent study found that the relative risk of CV events in haemophilia was around one third of predicted risk. The authors concluded that ‘haemophilia protects against CVD’ .
INTERFERENCE WITH REPAIR PROCESSES
Endothelial damage, clot formation and repair, represents a continual process. This is highlighted in people who smoke – who have a significantly higher risk of CVD. Smoking leads directly to glycocalyx damage, reduced NO production and bioavailability which, in turn, creates a pro-coagulant and inflammatory environment [32▪].
This damage can, in turn, be measured by an increase in microparticles – the breakdown products of endothelial cells. However, smoking also stimulates the production and release of endothelial progenitor cells (EPCs), which cover over areas of endothelial damage, driving the healing process .
Therefore, with smoking, increased damage and repair occurs simultaneously. This means that damage to the endothelium/glycocalyx does not necessarily lead to (accelerated) atherosclerotic plaque formation.
However, if the repair systems are impaired, or the damage is simply too extensive, plaque formation/growth will occur. This is probably why, as people become older, and the repair systems become less robust, a risk factor such as smoking becomes more significant .
The conjecture that reduced repair is as important as damage is supported by the fact that low EPC levels are an initial sign of endothelial dysfunction and represent one of the early signs of ASCVD [35▪].
Of course, EPCs do not represent the only repair system(s). But it is clear that protection against ASCVD requires a positive balance between repair and damage. This can perhaps be most clearly seen when we look at agents, which have been designed to block endothelial cell growth and repair, simultaneously reducing NO synthesis.
These are vascular endothelial growth factor (VEGF)-inhibitors. VEGF is a hormone which drives endothelial cell proliferation, survival, and migration . It also stimulates the production of EPCs in the bone marrow.
VEGF-inhibitors are used in cancer treatments to prevent endothelial growth and thus angiogenesis, as new blood vessels to deliver the blood supply required for growth in many tumours.
It is no surprise, therefore, that the adverse effects of VEGF-inhibitors centre around vascular damage, including haemorrhage, retinal detachment, venous thrombosis, strokes, MI and heart failure .
At one time, the increased cardiovascular risks associated with bevacizumab, the first widely used VEGF-inhibitor, nearly led to its withdrawal from the market. This highlights the critical importance of repair in the triad of endothelial damage, clot formation, impaired repair.
For the last seventy years the LDL-cholesterol hypothesis has been the most widely accepted causal model for the development of ASCVD. However, it cannot explain how many different factors such as smoking, SLE, or chronic kidney disease can lead to ASCVD.
In addition, a raised LDL level is a relatively weak predictor of risk. Indeed, LDL is not used to calculate CV risk using Qrisk3, or the ACC/AHA risk calculator.
An alternative model is one that was first proposed over one hundred and fifty years ago by Karl von Rokitansky, which is that atherosclerotic plaques represent the build-up/metamorphosis of thrombi that have been deposited onto, then incorporated into the arterial wall.
Other researchers, for example, Elspeth Smith, promoted the thrombogenic hypothesis widely. Forty years ago, she wrote: ‘After many years of neglect, the role of thrombosis in myocardial infarction is being reassessed. It is increasingly clear that all aspects of the haemostatic system are involved. Not only in the acute occlusive event, but also in all stages of atherosclerotic plaque development, from the initiation of atherogenesis to the expansion and growth of large plaques’ .
It should be recognised that this was written before NO and EPCs had been identified, and also before the protective function of the glycocalyx and the structure of apo(a) had been established.
We now have a greater understanding of the endothelial/glycocalyx function and how endothelial damage links to thrombus formation and lysis, and these new findings strongly suggest that Rokitansky, Duguid, Ross and Elspeth Smith, and others, may well have been correct.
Thus, whilst this article is entitled “Assessing cardiovascular disease: looking beyond cholesterol”, it may have been more accurate to call it, ‘Assessing cardiovascular disease, looking before cholesterol’, because the thrombogenic hypothesis precedes the cholesterol hypothesis by many years. It is fascinating that recent research now provides an increasing evidence base.
Financial support and sponsorship
Conflicts of interest
The author has written three books critical of the LDL-cholesterol hypothesis. The Great Cholesterol Con, A Statin Nation and The Clot Thickens.
REFERENCES AND RECOMMENDED READING
Papers of particular interest, published within the annual period of review, have been highlighted as:
▪ of special interest
▪▪ of outstanding interest
1▪. Hee Kim S, Young Son K. Association between lipoprotein cholesterol and future cardiovascular disease and mortality in older adults: a Korean nationwide longitudinal study. Lipids Health Dis 2021; 20:3.
4▪▪. Byrne P, Demasi M, Jones M. Evaluating the association between low-density lipoprotein cholesterol reduction and relative and absolute effects of statin treatment: a systematic review and meta-analysis. JAMA Intern Med 2022; 182:474–481.
5. Michael Lincoff J, Nicholls S, Riesmeyer J. Evacetrapib and cardiovascular outcomes in high-risk vascular disease. N Engl J Med 2017; 376:1933–1942.
6. Minelli S, Minelli P, Montinari MR. Reflections on atherosclerosis: lesson from the past and future research directions. J Multidiscip Healthc 2020; 13:621–633.
7. Sloop G, Gheorghe Pop G, Weidman J. Why atherothrombosis is in principle a hematologic disease: the effect of disorders and drugs which affect thrombosis on the development of atherosclerotic plaques. Int Arch Cardiovasc Dis 2018; doi:10.23937/iacvd-2017/1710012.
8. Yamaoka-Tojo M. Vascular endothelial glycocalyx
as a mechanism of vascular endothelial dysfunction and atherosclerosis. World J Cardiovasc Dis 2020; 10:731–749.
9▪▪. Vollenberg R, Tepasse PR, Ochs K. Indications of persistent glycocalyx
damage in convalescent COVID-19 patients: a prospective multicenter study and hypothesis. Viruses 2021; 13:2324.
10. Sullivan RC, Matthew D. Endothelial glycocalyx
degradation during sepsis: causes and consequences. Matrix Biol Plus, 2021;12:100094.
11. Yamaoka-Tojo M. Endothelial glycocalyx
damage as a systemic inflammatory microvascular endotheliopathy in COVID-19. Biomed J 2020; 43:399–413.
12. Walsh N. CV risks persist long after Kawasaki disease — a quarter of kids experienced a CV event 20 years after the diagnosis. MedPage Today. November 8, 2020. Available at: https://www.medpagetoday.com/meetingcoverage/acr/89549
13▪. Kurihara O, Takano M, Miyauchi Y. Vulnerable atherosclerotic plaque features: findings from coronary imaging. J Geriatr Cardiol 2021; 18:577–584.
14. Sampei S, Hideshi Okada H, Tomita H. Endothelial glycocalyx
disorders may be associated with extended a diabetic mouse model. Front Cell Dev Biol. 2021; 9: doi: 10.3389/fcell.2021.623582.
15. Abe T, Ogura H, Shiraishi A. Characteristics, management, and in-hospital mortality among patients with severe sepsis in intensive care units in Japan: the FORECAST study. Crit Care 2018; 22:322.
16. Ki-Jo K, Ji-Young K, In-Woon B. Elevated serum levels of syndecan-1 are associated with renal involvement in patients with systemic lupus erythematosus. J Rheumatol 2015; 42:202–209.
17. Kahn K. Heart health: atherosclerosis and aging — insights into the role of the endothelial glycocalyx
in cardiovascular health. Today's Geriatric Med. 12;8. Available at: https://www.todaysgeriatricmedicine.com/archive/JA19p8.shtml
18. Triantafyllidi, Benas, Dimitrios. HDL cholesterol levels and endothelial glycocalyx
integrity in treated hypertensive patients over 40 years old. J Hypertens 2021; 39:e226.
19. Miranda S, Billoir P, Le Besnerais M. New insights into antiphospholipid-related endothelial dysfunction by assessment of vascular glycocalyx
layer: results from a preliminary cross-sectional study. Lupus 2020; 29 (2):157–164.
20. Ajeganova S, Hafström I, Frostegård J. Patients with SLE have higher risk of cardiovascular events and mortality in comparison with controls with the same levels of traditional risk factors and intima-media measures, which is related to accumulated disease damage and antiphospholipid syndrome: a case–control study over 10 years. Lupus Sci Med 2021; 8:e000454.
21▪▪. Qu J, Cheng Y, Wenchao Wu W. Glycocalyx
impairment in vascular disease: focus on inflammation. Front Cell Dev Biol 2021; 9:730621.
22▪. Neubauer K, Zieger B. Endothelial cells and coagulation. Cell Tissue Res 2022; 387:391–398.
23. Potje S, Tiano Dal-Cin P, Paulo M. The role of glycocalyx
and caveolae in vascular homeostasis and diseases. Front Physiol 2021; 11:620840.
24. Fan M, Wang X, Peng. Prognostic value of plasma von Willebrand factor levels in major adverse cardiovascular events: a systematic review and meta-analysis. BMC Cardiovasc Disord 2020; 20:72.
25. Tunstall-Pedoe H, Woodward M, Tavendale R. Comparison of the prediction by 27 different factors of coronary heart disease and death in men and women of the Scottish heart health study: cohort study. BMJ 1997; 315:722–729.
26▪. Surma S, Banach M. Fibrinogen and atherosclerotic cardiovascular diseases—review of the literature and clinical studies. Int J Mol Sci 2022; 23:193.
27. Riches K, Porter K. Lipoprotein(a): cellular effects and molecular mechanisms. Hindawi Rev 2012;Article ID 923289.
28▪▪. Finneran P, Pampana A, Khetarpal A. Lipoprotein(a) and coronary artery disease risk without a family history of heart disease. J Am Heart Assoc 2021; 10:e017470.
29. Shil, RM, Niedzwiecki. Dietary vitamin C and age-induced lipid and hormonal metabolic changes in a humanized mouse model not synthesizing vitamin c and producing lipoprotein(a) [Gulo (−/−); Lp(a)+]. J Nutr Metab 2021;Article ID 5591697.
30. Khandelwal D, Mathur V, Vyas A. Elevated factor VIII levels and arterial stroke: a review of literature with a case report. Egypt J Neurol Psychiatr Neurosurg 2021; 57:20.
31. Van Der Valk P, Makris M, Fischer K. Reduced cardiovascular morbidity in patients with hemophilia: results of a 5-year multinational prospective study. Blood Adv 2022; 6:902–908.
32▪. Hahad O, Arnold N, Jurgen H. Cigarette smoking is related to endothelial dysfunction of resistance, but not conduit arteries in the general population—results from the gutenberg health study. Front Cardiovasc Med 2021; 8:674622.
33. Mobarrez F, Antoniewicz L, Bosson J. The effects of smoking on levels of endothelial progenitor cells and microparticles in the blood of healthy volunteers. PLoS One 2014; 9:e90314.
34. Rethineswaran VK, Kim DY, Kim YJ. Augmented the function of late endothelial progenitor cells by preventing replicative senescence. Int J Mol Sci 2021; 22:4796.
35▪. Morrone D, Picoi M, Felice F. Endothelial progenitor cells: an appraisal of relevant data from bench to bedside. Int J Mol Sci 2021; 22:12874.
36. Wang A, Bove A, Simon G. Molecular bases of VEGFR-2-mediated physiological function and pathological role. Front Cell Dev Biol 2020; 8:
37. Tsiros D, Sheehy CE, Nugent MA. Heparin–Avastin complexes show enhanced VEGF binding and inhibition of VEGF-mediated cell migration. Int J Transl Med 2021; 10:101–115.
38. Smith E, Douglas Thompson W. Fbrin as a factor in atherogenesis. Thromb Res 1994; 73:1–19.