Start · Catalogue · Profile · Table
Screening BODY HANDBOOK
Screening · §110
Coronary CT Angiography
Half of first heart attacks land on people who never noticed a thing — the trim colleague in the parking garage, the friend who didn't come back from the trail. The reason is usually soft plaque: lipid-rich, thin-capped, prone to rupturing before it ever turns to calcium, and a normal coronary calcium score can't see it. A coronary CT angiogram is a 20-minute contrast scan that maps the inside of every coronary artery — soft plaque, calcified plaque, lumen — and in the trial that put it on the standard chest-pain pathway, five years of acting on its findings cut fatal and non-fatal heart attacks by 41% SCOT-HEART, NEJM 2018. The biggest decision the result drives is not stenting — it is whether to get serious about the medicines and habits that quietly rewrite the next decade.
Test · As-needed Evidence Moderate Chapter Screening

Half an hour at the imaging centre, a beta-blocker pill, an IV of contrast, and a radiation dose roughly equivalent to a few months of background. Two large randomised trials and a 10-year follow-up put this first-line in the US, European, and UK chest-pain guidelines — the gain comes from acting on what it finds, not from stents. The catches: dense old calcium can blur the picture, a clean scan is a five-year answer not a forever one, and in adults without symptoms the mortality benefit is observed and consistent but not yet proven by a randomised screening trial.

The test is a CT scan with a difference: a bolus of contrast goes into a vein in your arm, the camera spins around your chest in step with your heartbeat (with the help of a beta-blocker that drops your pulse below 65), and the radiologist gets sub-millimetre slices of every coronary artery — inside and outside the wall. A regular calcium score, the test you may have heard of, is the same machine without contrast and only sees the white speckles of calcified plaque. Contrast is what makes the lumen — the actual hole the blood flows through — light up, and it is what makes soft plaque visible at all. Soft plaque is the lipid-rich, thin-capped kind that has not had time to mineralise: invisible to a calcium score, but it is the one that ruptures and starts the clot that becomes a heart attack.

The report you walk away with is standardised. The current scheme, CAD-RADS 2.0, grades stenosis from 0 (clean) to 5 (totally blocked), grades total plaque burden P1 through P4, and flags high-risk features — low-attenuation plaque, positive remodelling, spotty calcification, the napkin-ring sign Cury et al., 2022. The report ends with a recommended next step: nothing to do, intensify prevention, add a stress test, send to the cath lab. That recommendation is the actionable line.

Does scanning actually change the outcome?

Two large randomised trials and a 10-year follow-up are the spine of the answer.

The US comparison, PROMISE, came out closer to a tie: against a functional stress test in 10,003 stable-chest-pain patients, the CT pathway produced similar overall event rates with fewer normal-finding cath-lab visits and slightly more diagnostic clarity Douglas et al., NEJM 2015. Functional testing works; the CT just works differently. The settled position across the 2021 American chest-pain guideline, the 2019 European Society of Cardiology guideline, and the UK's NICE pathway is the same: for an adult with stable chest pain and no prior coronary disease, this is now the first-line test Gulati et al., 2021 Knuuti et al., 2020 NICE 2016.

The third piece is plaque-based prognosis. Follow-on analyses found that soft-plaque features predict future heart attacks independently of how narrow the artery is. A SCOT-HEART substudy showed that more than 4% low-attenuation plaque burden carried roughly five times the heart-attack risk of less — a better predictor than the calcium score or the worst stenosis on the scan Williams et al., JACC 2019. In a 3,158-patient cohort, two or more high-risk plaque features carried a 9% per-year heart-attack rate versus 0.5% per year without them Motoyama et al., JACC 2015. The composition of the plaque matters, not just the count.

"My calcium score was zero, so I'm fine"

The biggest miss in this whole space is treating the calcium score and the CT angiogram as the same test. They are not. The calcium score is non-contrast, cheaper, faster, and only sees plaque that has already calcified — a slow process, often decades old. The CT angiogram is contrast-enhanced, sees the actual lumen and the soft plaque too, and is what tells you whether you have early disease the calcium count missed.

In the SCOT-HEART plaque substudy, a meaningful share of heart attacks happened in patients whose calcium score had been zero — they had non-calcified plaque, visible only on the contrast scan Williams et al., JACC 2019. The "calcium zero, go home" reading is a five-year-and-low-baseline-risk reading, not a forever-young one. Under 50, with a strong family history or a high Lp(a), the calcium score is the wrong test to lean on alone.

The second misconception cuts the other way: an abnormal scan does not mean a stent. The ISCHEMIA trial randomised stable patients with confirmed obstructive disease to invasive versus medical management and found no mortality difference at 3.2 years Maron et al., NEJM 2020. SCOT-HEART's payoff was from prevention, not from stents SCOT-HEART, NEJM 2018. Stable coronary disease is treated with medicine first — and aggressively. The scan tells you who needs that medicine; it does not draft you to the cath lab.

What you can't feel

Roughly half of first heart attacks happen in people who had no symptoms the week before. Not mild ones — none. The colleague who runs marathons, the parent who passed every checkup, the body that has been showing up for fifty years without complaint — they don't get a warning siren because the disease that causes sudden cardiac events is silent by design. The artery is fine until the plaque ruptures, and then it isn't.

Risk calculators miss this in both directions. They over-treat the lean, fit reader whose arteries are actually clean, and they under-treat the average-looking reader whose grandfather died at 58 and whose blood markers sit in the higher percentiles. The CT angiogram is the test that resolves the disagreement: it shows you, today, whether you have the disease the family-history conversation has been hinting at, while there is still a decade of low-cost prevention available. Even non-obstructive plaque — the kind that does not pinch the artery yet — roughly doubles 10-year heart-attack risk versus clean arteries in the same age band Mortensen et al., Eur Heart J 2023. The thing the typical reader is buying is not drama. It is the ability to make the next ten years' worth of decisions with the actual map in front of them.

How to get one, what the day looks like

You don't walk in off the street: a coronary CT angiogram is ordered by a physician — typically your GP or a cardiologist — once the indication is clear. The standard indications are stable chest pain in an adult without known coronary disease, an intermediate ten-year cardiovascular risk where the statin conversation is on the fence, or a strong family history of premature heart attack. Preventive-cardiology clinics are increasingly ordering them at-request; insurance coverage varies by indication.

The result lands as a CAD-RADS 2.0 report a few days later: stenosis grade per artery, plaque burden, high-risk features flagged, and a recommended next step Cury et al., 2022. The conversation with the ordering physician is where the work actually happens — the scan only earns its keep if the result changes what gets prescribed, monitored, or repeated.

When this is the wrong test

A handful of situations push you toward a different workup.

The iodine and kidney concerns have softened over the last decade: well-matched cohort studies suggest the historical fear of "contrast-induced kidney injury" was overstated in stable outpatients with reasonable kidney function. But the call still goes through your physician, not around them, when any of the above applies.

Radiation, cost, where to get one

Radiation first, because it is the question most people walk in with. The PROTECTION VI registry across 4,502 patients at 61 international sites reported a median effective dose of 2.7 millisieverts on modern protocols — roughly the natural background radiation you absorb from being alive for nine months Stocker et al., Eur Heart J 2018. On a dual-source scanner with a slow, regular heart rate, the dose can drop below 1 mSv. For context, a routine chest CT is 5-7 mSv, a screening mammogram is 0.4 mSv, an invasive coronary angiogram is 2-10 mSv. The dose is meaningfully low. It is not zero.

Cost varies by country and facility. In the US, cash prices run $500-$2,000; insurance-negotiated rates often $250-$700; a 2025 Medicare reimbursement update is expanding availability. In the UK and most of Europe the test is publicly funded under the chest-pain pathway. The hidden cost is reader expertise: ask whether the report is being read by a radiologist or cardiologist with cardiac-CT certification (Society of Cardiovascular Computed Tomography Level II or III). At a non-specialist centre the report can drift toward over-reading stenosis severity, which sends people unnecessarily to the cath lab.

Where it goes sideways

Two common downstream errors. The first is over-reading: an inexperienced reader sees a moderate, non-obstructive plaque, calls it a tight stenosis, and the patient ends up in the cath lab where an invasive angiogram confirms the plaque is not actually flow-limiting. The big stable-disease trials say the right move there is usually medical anyway Maron et al., NEJM 2020 — but the cath itself carries a small complication risk that should not have been spent. Cardiac-CT-trained readers reduce this substantially.

The second is under-reading. The report says "no obstructive coronary disease" — meaning no artery is more than 50% narrowed — and the patient and the ordering physician both hear "all clear". But any plaque at all carries meaningful ten-year heart-attack risk that climbs with how much there is, even when no artery is narrowed enough to cause symptoms Mortensen et al., Eur Heart J 2023. A "non-obstructive" CT is not a green light to stop the statin conversation; it is an early-stage finding that earns aggressive prevention.

The third, smaller failure mode is incidental findings: a small lung nodule, a thyroid bump, a sliver of mediastinal lymph node show up on 10-30% of scans and pull the patient into a separate workup that often turns out to be nothing. It is the price of looking carefully at a chest.

The other tests and where each one wins

  • Coronary calcium score. Cheaper ($75-$200), lower dose (~1 mSv), no contrast, no heart-rate prep. Best for asymptomatic adults 40-75 with intermediate ten-year risk where the statin decision is genuinely uncertain. Misses pure soft plaque; not a substitute for the contrast scan when symptoms are present or family history is strong Nakanishi et al., EHJ-CI 2017.
  • Stress testing (treadmill ECG, stress echo, nuclear, stress cardiac MRI). Detects whether existing disease is starving the heart of blood, not whether the disease exists. The big head-to-head trial found similar event prediction to the CT scan but with lower diagnostic certainty Douglas et al., NEJM 2015. Useful when the question is whether blood flow is choked, not whether plaque is there.
  • Invasive coronary angiography. Catheter, contrast, can stent in the same session. Higher complication rate (around 1-2%); usually reserved for confirmed or strongly suspected obstructive disease after a non-invasive test points there. The advantage in the SCOT-HEART era is that fewer of these get done on people who turn out to have nothing.
  • Blood-based risk inputs — ApoB, Lp(a), high-sensitivity C-reactive protein. Complementary, not competing. They tell you about the cargo in the bloodstream and the inflammation; the scan tells you about the wall. A high Lp(a) with a clean CT still earns aggressive prevention; a low Lp(a) with extensive plaque also does.

What changes after the result

The week of the scan, almost nothing — you go home, you eat, you go to work. The result conversation a few days later is the actual intervention. If the scan is clean, your physician has a near-100% answer that obstructive disease is not driving your chest discomfort, and the cardiac thread of the workup closes for the next several years Knuuti et al., 2020. The version of you that has been carrying low-grade dread on stair climbs gets to drop it. That clarity is its own payoff.

If the scan finds plaque — calcified, soft, both — the months that follow look different. A high-intensity statin gets started or titrated up. Lp(a) gets checked if it hasn't been. Blood pressure gets a tighter target. The conversation with your family gets serious in a way it has not been. By six months your LDL is meaningfully lower; by five years, the SCOT-HEART trajectory says the chance of a fatal or non-fatal heart attack has dropped by close to half compared with the version of you who never had the scan SCOT-HEART, NEJM 2018. Ten years out, the gap is still there Newby et al., Lancet 2024. The colleague's story is not your story.

Worth saying plainly: the heart-attack-reduction benefit is from acting on the result. A scan that sits in a folder unread does nothing. The pay is in the prescription, the run, and the kitchen.

Three adjacent threads worth pulling: the coronary calcium score (the cheaper screen this entry's test partly replaces), ApoB and Lp(a) (the blood-side risk inputs that make the anatomical scan more informative), and high-intensity statin therapy (the medicine the scan most often triggers). For chest pain that turns out not to be coronary in origin, the workup branches into reflux, costochondritis, anxiety, and other shapes — a different kind of question.

·
110