Most patients arrive after an angiogram, holding a number. Seventy per cent. Eighty. The number feels like a sentence, and the next sentence they expect to hear is that a stent will be placed today.
That number describes how a narrowing looks. It does not describe whether the narrowing is restricting blood flow, whether it is causing the symptoms that brought you in, or whether opening it will make you live longer. Those are three separate questions, and the evidence answers them differently.
What the randomised evidence actually shows
Over nearly two decades, cardiology has repeatedly tested the same question in stable coronary disease: does adding a stent to good medical therapy prevent death and heart attack? The answer has been remarkably consistent.
In COURAGE (2007), 2,287 patients with stable disease were randomised to medical therapy with or without PCI. Over a median 4.6 years, adding the stent did not reduce death or myocardial infarction. BARI 2D (2009) asked the same question specifically in patients with type 2 diabetes and found the same answer.
ISCHEMIA (2020) was designed to close the remaining objection — that earlier trials had enrolled patients whose disease was not severe enough to benefit. It recruited 5,179 patients with moderate or severe ischaemia on stress testing, the group most expected to gain from intervention, across 320 sites in 37 countries. Over a median 3.2 years, an initial invasive strategy did not reduce ischaemic cardiovascular events or death from any cause.
REVIVED-BCIS2 (2022) extended the question to severely weakened hearts — ejection fraction 35% or below, with viable muscle. Across 700 patients followed a median 41 months, PCI added to medical therapy did not reduce death or heart-failure hospitalisation, and the improvement in pumping function did not materialise.
The uncomfortable trial, included deliberately
It would be easy to stop there. Honesty requires going further.
ORBITA (2018) did something unusual: it gave half of 200 patients a real stent and half a placebo procedure, and nobody — patient or assessor — knew which. The stent did not significantly improve exercise time. That finding suggested a substantial share of the benefit patients report after angioplasty comes from having undergone a procedure at all.
But ORBITA-2 (2023) refined the design. It withdrew antianginal medication first, so the stent was compared against placebo rather than against placebo-plus-drugs, and measured angina daily through a smartphone app. In 301 patients, PCI clearly reduced angina — patients were around three times more likely to be free of symptoms at twelve weeks.
So stents work. They work on symptoms. ORBITA-2 did not show a reduction in death or heart attack, and its authors did not claim one. The distinction between a treatment that makes you feel better and a treatment that makes you live longer is the single most important thing a patient facing this decision can understand — and it is the distinction most often blurred in the consultation room.
How that translates into a consultation
The protocol is not a refusal to stent. It is a sequence of questions asked before stenting, in order:
- Is this an emergency? If you are having a heart attack, the reasoning below stops and the catheter lab starts. See the notice on this page.
- Are you symptomatic, and how limiting is it? Angina that stops you climbing stairs is a different problem from an incidental finding on a screening scan.
- Is there objective ischaemia? Stress testing, stress echocardiography or perfusion imaging — not the angiogram alone.
- Does the lesion actually restrict flow? FAME 2 established that physiological assessment, such as fractional flow reserve, selects lesions differently from visual assessment.
- Has medical therapy been given a fair trial? In most of the trials above, the comparator was not “nothing”. It was aggressive, properly titrated medical therapy — and that is what the stent failed to beat.
- What does the patient want, knowing all of this? A patient who understands they are choosing symptom relief rather than survival benefit makes a different decision from one who believes the stent is saving their life.
What “medical therapy” actually means
Choosing not to stent is not choosing to do nothing, and it is not the easier path. Done properly it means high-intensity statin therapy titrated to a lipid target, blood pressure controlled to goal, antiplatelet therapy where indicated, glucose management in diabetes, structured exercise, complete tobacco cessation, and scheduled follow-up that actually happens. It asks more of the patient than a procedure does, and more of the physician than a referral does.
That regimen is the comparator in COURAGE, ISCHEMIA and REVIVED-BCIS2. The reason those trials read as they do is not that stents are ineffective. It is that well-delivered medical therapy is very effective, and the trials were fair to it.
Where a stent is clearly right
An honest protocol names its own limits. Intervention is indicated in ST-elevation myocardial infarction and high-risk acute coronary syndromes, in significant left main stem disease, in angina that persists despite genuinely optimised medical therapy, and where objective testing shows a large ischaemic burden with symptoms that are limiting daily life. Those patients are stented, without hesitation.
Written and medically reviewed by Dr. Gurbir Singh Gill — Interventional Cardiologist & Diabetologist, Oxford Hospital.
Last reviewed: 1 September 2026