Collateral circulation

Why can one person survive a blocked artery that kills another with the same blockage?

Two people, two coronary angiograms, and on each of them the same artery blocked in the same place. One is dead within the hour. The other has been walking around with it for years and turns up complaining of mild breathlessness on stairs.

The instinct is to reach for a difference in the patients — age, fitness, luck. But hold that instinct and ask a plainer question first. If the pipe is shut, the muscle beyond it must be getting blood from somewhere, or it would not be alive. So the interesting difference may not be in the blockage at all. It may be in what else already existed around it. And that raises a stranger question: where would a spare route have come from, and why would one person have it and another not?

Think about what a backup route has to be, physically. It cannot be conjured at the moment of the emergency — growing a vessel takes weeks, and heart muscle starved of blood begins dying in minutes. So any route that helps must have been there beforehand, at least in outline.

And in outline it usually is. The coronary arteries are not entirely separate trees; there are small connecting channels between neighbouring territories, present in most hearts but ordinarily tiny — wide enough to be irrelevant to flow. So the raw material for redundancy is nearly universal. What varies enormously between people is whether that material was ever developed.

Now ask what would develop it, and this is the part worth slowing down for. Suppose one artery narrows gradually over years. Pressure downstream of the narrowing falls, while pressure in the neighbouring territory does not. What does that difference do to those tiny connecting channels sitting between the two? It drives flow through them — and with flow comes shear stress against their walls. That mechanical signal is what triggers the vessels to remodel and enlarge, a process called arteriogenesis. Notice the shape of that: the stimulus that builds the backup is the very stenosis the backup will later protect against.

So the redundancy is not standing spare capacity. It is conditional capacity, grown in response to a threat that arrived slowly enough to be noticed.

That immediately explains our two patients, and it also explains something that surprises people about heart attacks. The person with the long-standing tight narrowing had years of the signal, and built a network; when the artery finally closed, the muscle was already partly supplied and the event was small or silent. The person who dropped dead may have had a plaque that was never severely narrowing at all — and this is well established, that many infarcts arise from plaques that were not the tightest ones. Because that plaque never restricted flow, it never generated the pressure gradient, so no collaterals grew. Then it ruptured, a clot formed on it in minutes, and the muscle beyond had nothing to fall back on.

Put crudely: the artery that warned you built you a defence. The one that said nothing left you undefended.

How much does this matter? Observational studies and their meta-analyses consistently report lower mortality in patients with well-developed collaterals, and smaller infarcts when occlusion occurs. But keep the qualification, because it is a real one: those are associations, and the same things that let a person grow good collaterals — a slower disease, better vascular health, more habitual exertion — may independently predict survival.

And the protection is partial, which is worth being precise about. Collateral flow is usually enough to keep muscle alive at rest but not enough to meet the demand of exertion. That is why the well-collateralised patient is often not symptom-free but merely alive and limited — angina on the stairs rather than a cardiac arrest.

One honest disappointment belongs here too. If growth factors drive vessel development, the obvious therapeutic idea is to inject them and grow collaterals on demand. That was tried extensively — trials of growth factor proteins and gene therapy for coronary and peripheral disease — and they have largely failed to show durable clinical benefit. Whatever the natural stimulus is doing, it has not been reproduced in a syringe.

The misconception worth correcting is that collaterals are new vessels grown because the tissue is short of oxygen. Two different processes are being run together there. Oxygen shortage does drive capillary sprouting — angiogenesis — but capillaries are far too small to carry the flow of a blocked artery. The vessels that rescue a territory are enlarged pre-existing connections, remodelled by mechanical shear rather than summoned by hypoxia, and they lie upstream of the starved tissue rather than within it.

A second refinement. It is tempting to conclude that a severe narrowing is therefore good for you. It is not; it is simply the lesser of the two exposures, and only in the specific sense that it buys time to develop an alternative. The patient with the tight stenosis still has the underlying disease, still has other plaques, and still has a limited reserve.

Finally, the general lesson is about redundancy itself. A backup that is never exercised is not really a backup — it is a claim about capacity that has never been tested under load. Biology's version of this solves the problem elegantly by making the load itself the trigger for growth, at the cost that the redundancy is only ever ready for the threats that approached slowly.

• Why the circle of Willis in the brain provides standing redundancy while the heart's must be grown, and what that costs each organ.
• Whether exercise training measurably improves collateral function, and how well that has been shown.
• Why therapeutic angiogenesis trials failed, and what that suggests about the missing part of the natural signal.

Every term the collection defines is gathered in the glossary.