From Cascade to Convergence: How Science Rethought the Way Blood Clots

For most of the 20th century, doctors and scientists described blood clotting using a single image: a waterfall. One protein activated the next, which activated the next, all the way down to a clot. It was elegant, teachable, and very useful in the lab. But it had a problem — it did not quite match what happened inside a living human body.

Over the following decades, researchers kept finding things the waterfall could not explain. Why did some patients bleed uncontrollably despite having parts of the system intact? Why did the same protein behave differently depending on where it was? The answers slowly pointed toward something the original model had largely ignored: the cells themselves.

The cascade model of coagulation was introduced in 1964 by two groups almost simultaneously — Davie and Ratnoff in one publication, MacFarlane in another. At the time, it was a genuine breakthrough. Scientists had just spent decades identifying the individual proteins involved in clotting, and now they had a framework showing how these proteins fit together in a logical chain.

The model divided clotting into two pathways. The "extrinsic" pathway was triggered from outside the bloodstream, primarily through a protein called tissue factor. The "intrinsic" pathway was activated by contact with charged surfaces inside the blood itself. Both pathways eventually merged and led to the generation of thrombin, which converted fibrinogen into fibrin — the mesh that holds a clot together.

This framework had real practical value. It directly explained two laboratory tests — the prothrombin time and the activated partial thromboplastin time — that are still used in hospitals today. When a patient bled unusually or clotted too easily, doctors could use these tests to trace the problem to a specific part of the cascade. The model also made sense of why warfarin and heparin worked as blood thinners, and it guided the development of clotting factor replacement therapies for patients with hemophilia.

But even in those early years, the model had awkward edges. Paul Morawitz had actually described coagulation decades earlier as a two-step process involving what he called thrombokinase — a tissue extract containing tissue factor — reacting with prothrombin in the presence of calcium. This older view captured something important that the newer cascade model would later struggle with: the idea that coagulation was inseparable from the physical environment in which it occurred.

As more patients were studied and more factor deficiencies described, the cascade model ran into serious trouble.

Take hemophilia A and B. Both conditions involve deficiencies in factors VIII and IX, which are part of the "intrinsic" pathway. According to the cascade model, the "extrinsic" pathway should have been able to compensate — after all, it also generates factor Xa, the same downstream product. But it cannot. Patients with hemophilia bleed severely, even though their extrinsic pathway is perfectly intact.

More puzzling still, deficiencies in factor XII, high molecular weight kininogen, and pre-kallikrein — all supposed initiators of the intrinsic pathway — cause a dramatic lengthening of the aPTT test in the lab, but cause no bleeding whatsoever in the patient. Some mammalian species, such as whales and dolphins, do not even have factor XII and appear to manage perfectly well.

These contradictions were not minor footnotes. They pointed to something fundamentally wrong with the idea of two independent, interchangeable pathways operating in parallel. The model that worked beautifully in a test tube was failing to describe what happened in a living body.

The rethinking came gradually through the 1990s, driven by a period of rapid advance in cell biology. Researchers began examining not just which proteins were present, but where reactions actually happened — and it turned out that location mattered enormously.

In 2001, Maureane Hoffman and Dougald Monroe published what became known as the cell-based model of hemostasis. Rather than two pathways merging into one, they described coagulation as occurring in three overlapping phases, each on a different cell surface.

The first phase is initiation, which begins when injury exposes tissue factor-bearing cells — such as fibroblasts in the vessel wall — to the flowing blood. Factor VIIa binds to tissue factor on these cells and generates small amounts of factors IXa and Xa, along with a tiny amount of thrombin. This thrombin is not enough to form a clot on its own, but it acts as a signal.

The second phase is amplification. That initial spark of thrombin drifts from the tissue factor-bearing cell to nearby platelets. Platelets, which have stuck to the site of injury, are activated by the thrombin. They release stored proteins, expose new surface receptors, and begin accumulating the cofactors needed for large-scale clotting.

The third phase is propagation. Now fully activated, the platelets become the main stage. The key clotting complexes assemble on their surfaces and generate a massive burst of thrombin — enough to convert fibrinogen into fibrin and seal the wound.

The model explained what the cascade had not. Factor IXa, generated on tissue factor-bearing cells, is relatively stable in solution and can travel to the platelet surface. Factor Xa, by contrast, is rapidly inactivated by plasma inhibitors the moment it leaves the cell surface. This is why factor VIII and factor IX are indispensable: without them, factor X cannot be activated directly on the platelet surface, and the burst of thrombin never happens. The extrinsic pathway generates factor Xa on the wrong surface — and there is simply no way for it to travel safely to where it is needed.

This insight also explained something that had puzzled clinicians for years: why very high doses of factor VIIa help patients with hemophilia. At high concentrations, factor VIIa can bind directly to activated platelet surfaces and generate factor Xa right there — bypassing the need for factors VIII and IX altogether.

The story did not stop with the cell-based model. By the 2020s, it was becoming clear that coagulation does not operate in isolation from the rest of the body's response to injury. The COVID-19 pandemic made this impossible to ignore. Patients developed unusual clots in unusual places. Standard anticoagulants had inconsistent effects. Something else was involved.

In 2023 and 2024, Jun Yong and Cheng-Hock Toh at the University of Liverpool proposed what they called the convergent model of coagulation. Their argument was that clotting, inflammation, and innate immune activation are not three separate systems that sometimes interact — they are one unified response to injury that evolved together over hundreds of millions of years.

Central to this model is a group of molecules called damage-associated molecular patterns, or DAMPs. These are substances — including cell-free histones, fragments of DNA, and proteins like HMGB1 — that are released when cells are injured or dying. In the cell-based model, DAMPs were barely considered. In the convergent model, they are everywhere: activating tissue factor, stimulating platelets, assembling clotting complexes, and resisting the dissolution of clots once they are formed.

One vivid example involves neutrophil extracellular traps — web-like structures that immune cells release to catch bacteria. These traps contain histones and DNA that also powerfully accelerate clot formation. In COVID-19, the body's immune response generated enormous quantities of these traps, and the result was a runaway clotting process that conventional anticoagulants could not adequately address.

Looking back across sixty years, the history of coagulation research tells a story of progressively widening vision. The cascade model gave us the proteins. The cell-based model showed us that location and cellular context determined everything. The convergent model reminds us that clotting is not a self-contained plumbing problem — it is a biological response woven tightly into immunity and inflammation, shaped by hundreds of millions of years of evolution.

Each model was right for its time and right about many things. None was wrong so much as incomplete. And each advance, from the identification of clotting factors to the discovery of cell surface receptors to the recognition of DAMPs, opened new doors in treatment — for hemophilia, for stroke, for the coagulopathies of sepsis and pandemic infection.

The question scientists are now asking is not simply "which proteins are present?" but "which cells are involved, what signals have they received, and what does the body think it is fighting?" That shift in thinking — from chemistry to biology, from pipeline to ecosystem — is what the convergent model represents.

Understanding how blood clots is, ultimately, understanding how the body keeps itself alive.