Screening · Heart
The coronary plaque ladder
Rung 1: the calcium score (CAC). A non-contrast CT that counts calcified plaque only. Cheap, fast, genuinely predictive, and structurally blind to soft, non-calcified plaque.
Rung 2: CCTA with AI plaque quantification. A contrast coronary CT angiogram read by FDA-cleared AI software that also sees the soft plaque rung 1 cannot. The step up is real detection, bought at a real price.
Read this before the ladder: rung 2's outcome evidence (SCOT-HEART, PROMISE) sits in symptomatic and stable-chest-pain populations. For asymptomatic screening, its outcome-change axis is unproven, so the no-outcome cap logic applies to rung 2 itself. The ladder is a step-up path for a narrow group, never a claim that rung 2 is simply better.
The ladder in one sentence: rung 2 sees more, and seeing more has not yet been shown to change outcomes in people without symptoms.
Already have a result that worries you?
That includes a calcium score or a CCTA result you are trying to interpret on your own. Your next step is the clinician who ordered the test, or your own doctor, before anything else on this page. Bring the result itself, not a grade from a website.
Rung 1: Coronary artery calcium scan (CAC, ladder rung 1)
Non-contrast, ECG-gated CT of the heart producing an Agatston score for calcified coronary plaque. Structurally blind to soft, non-calcified plaque (the ladder's rung 2 exists because of this limit).
CAC is the only test in the first wave whose result flips a decision a guideline has already made.
Outcome-change evidence
No randomized trial shows a CAC-guided screening strategy reduces events; USPSTF gave an I statement in 2018 (insufficient evidence). Indirect support: DANCAVAS (NEJM 2022, n=46,611, screening bundle including CAC) all-cause mortality HR 0.95, p=0.06; EISNER (JACC 2011) found better risk-factor control in the scanned group with no increase in downstream testing or costs. The benefit chain is one link removed: CAC changes a statin decision, and statins carry outcome-trial evidence.
Detection performance
The strongest reclassifier in this class. MESA (Polonsky et al., JAMA 2010): net reclassification improvement 0.25 overall and 0.55 in the intermediate-risk group (16% up, 39% down); C-statistic 0.76 to 0.81. Ten-year event rates run roughly 1.3 to 5.6% at CAC 0 and 13.1 to 25.6% above 300.
Actionability
HIGHEST IN THIS CLASS. ACC/AHA cholesterol guidance makes CAC Class IIa for adults 40 to 75 at borderline or intermediate risk when the statin decision is uncertain: CAC 0 supports deferring a statin (with carve-outs), CAC 100 or above (or 75th percentile or above) supports starting one. About 40% of intermediate-risk adults score zero, so the downward reclassification changes real management.
Cost and cadence
Per screen: $50 to $400 self-pay (market range, commonly $75 to $250; most insurers do not cover)
Downstream: Small: EISNER found scanning did not raise downstream testing rates or costs in a randomized comparison. Serial scanning once positive and treated is not supported.
Repeat cadence: One decision scan; rescan a zero at 3 to 7 years depending on risk (basis: guideline plus cohort (MESA warranty analysis, Dzaye et al., 2021; updated ACC/AHA dyslipidemia guidance))
Why this grade
The USPSTF I statement and the missing screening-strategy outcome RCT bar the A range. Top of class because reclassification is large and measured, the attached action is a proven therapy under a Class IIa recommendation, and harms are small and priced.
Harm ledger
| Harm | Rate or scale | Source |
|---|---|---|
| Radiation | About 1 mSv, comparable to a mammogram | MESA protocol data |
| Extracoronary incidental findings (lung and thyroid nodules predominate) | About 1 in 10 scans; follow-up pursued in 4 to 8%, almost all benign | MESA-based analyses |
| False reassurance from a zero score in younger adults, smokers, and people with diabetes (groups where non-calcified plaque matters most) | Qualitative, structural | Longevity Lab Report screening evidence review, 2026-10-18 |
Rung 2: CCTA with AI plaque quantification (ladder rung 2)
Contrast coronary CT angiography read by FDA-cleared AI software that quantifies total plaque volume, plaque composition including the soft non-calcified fraction, and stenosis, artery by artery. Cleerly is the exemplar implementation readers will meet; the graded item is the modality class, not the vendor.
What the AI actually changes, and its limits
Where AI genuinely changes what a test can see, we say how, with the primary evidence named. Here it is real: AI quantification reads total plaque volume and the soft non-calcified fraction that a calcium score structurally cannot, validated against invasive reference standards (CREDENCE substudy: sensitivity 94%, specificity 68% at the 50% stenosis threshold). Cleerly is the exemplar implementation readers will meet; the graded item is the modality class, not the vendor.
The scope limits, stated plainly: the FDA clearance for Cleerly ISCHEMIA (January 2024) is as a diagnostic aid in patients undergoing CCTA for known or suspected coronary artery disease, not for screening. Its validation datasets (CREDENCE, PACIFIC, CONFIRM) are symptomatic or referred populations; none is screening evidence. Medicare coverage and the CPT code are payment facts, never evidence: CMS's own coverage review flags retrospective designs and absent long-term outcomes.
The price of the step up: contrast exposure and clinician-gated access, radiation at roughly 2 to 4 times rung 1 repeated in full at every rescan, significant extracardiac findings in 16.0% of screening-series scans, and an AI analysis that reimburses about $1,021 under CPT 75577 (CMS 2026 Physician Fee Schedule), on top of the scan charge.
Who rung 2 actually answers a question for
The defensible niche is narrow: a calcium score of zero or low where a statin decision genuinely hinges on plaque presence, with strong family history of premature coronary disease, current smoking, or diabetes. In SCAPIS, CCTA found visible atherosclerosis in 5.5% of CAC-zero adults overall, 6.0% with strong family history, 6.8% in smokers, and 8.1% in diabetes: exactly the groups where rung 1's blind spot matters most. Detection real, outcomes unproven, niche narrow. Outside that niche, in low-risk adults whose treatment would not change, and as serial plaque tracking, rung 2 is not defensible on current evidence.
Outcome-change evidence
For the graded use (asymptomatic screening): NONE, and the one screening RCT was null. FACTOR-64 (JAMA 2014, n=900 asymptomatic diabetics): CCTA screening changed treatment recommendations in 70% of patients and did not reduce events (HR 0.80, p=0.38). The positive outcome evidence (SCOT-HEART 2018) lives in symptomatic chest-pain patients; PROMISE 2015 was null even there. Cleerly's validation datasets (CREDENCE, PACIFIC, CONFIRM) are symptomatic or referred populations; none is screening evidence. The FDA clearance for Cleerly ISCHEMIA (January 2024, verified) is as a diagnostic aid in patients undergoing CCTA for known or suspected CAD, not for screening. CMS's own coverage review (LCD L39851) flags retrospective designs and absent long-term outcomes; payment is not evidence.
Detection performance
AI quantification against invasive reference: CREDENCE substudy sensitivity 94%, specificity 68% at the 50% stenosis threshold. Low-attenuation plaque burden above 4% predicted myocardial infarction (HR 4.65). In SCAPIS, CCTA saw atherosclerosis in 5.5% of CAC-0 healthy adults (8.1% in diabetes): real detection in the niche, which is narrow.
Actionability
REAL BUT NARROW. Defensible where a statin decision genuinely hinges on plaque presence: CAC 0 or low with strong family history of premature CAD, current smoking, or diabetes (SCAPIS found CCTA-visible atherosclerosis in 5.5% of CAC-0 adults, 6.0% with strong family history, 6.8% in smokers, 8.1% in diabetes). Not defensible in low-risk asymptomatic adults, where anyone whose treatment would not change, or as serial plaque tracking, a use no guideline supports.
Cost and cadence
Per screen: AI analysis reimburses about $1,021 under CPT 75577 (CMS 2026 Physician Fee Schedule), on top of the CCTA scan charge; billed totals commonly in the low thousands
Downstream: Repeat imaging and, for a minority, invasive angiography. Compare rung 1 at $50 to $400 all-in.
Repeat cadence: None supported. Rescans every 2 to 5 years are a concierge practice pattern (basis: guideline absence; concierge practice pattern labeled as such)
Why this grade
No outcome-change evidence in the graded use caps at C+; the null screening RCT, plus harm and cost ledgers at multiples of rung 1, discount to C-. It grades above whole-body MRI because its detection science is validated against invasive standards, its key prognostic claim is primary-literature real, and a narrow population exists where it answers a question rung 1 cannot.
Harm ledger
| Harm | Rate or scale | Source |
|---|---|---|
| Radiation at roughly 2 to 4 times CAC, repeated in full at every rescan | Multiples of rung 1's ~1 mSv | Longevity Lab Report screening evidence review, 2026-10-18 |
| Contrast exposure (allergy, kidney considerations) and clinician-gated access | Per-scan, structural | Longevity Lab Report screening evidence review, 2026-10-18 |
| Significant extracardiac findings | 16.0% in screening series | Longevity Lab Report screening evidence review, 2026-10-18 |
| Population labeling: atherosclerosis identified in a large share of healthy middle-aged adults, most of whom will never have an event | 42.1% of healthy 50 to 64 year olds in SCAPIS | Bergstrom G et al., SCAPIS, 2021 |
| Downstream invasive angiography for a minority, an event with a five-figure price tag | Minority of screened positives | Longevity Lab Report screening evidence review, 2026-10-18 |
The two rungs, side by side
| Rung 1: CAC | Rung 2: CCTA + AI | |
|---|---|---|
| Grade | B | C- |
| Sees | Calcified plaque only (Agatston score) | Total plaque volume, soft fraction, stenosis, artery by artery |
| Outcome evidence | Reclassification that flips a guideline statin decision; no screening-strategy outcome trial | None for asymptomatic screening; the one screening trial was null (FACTOR-64) |
| Cost | $50 to $400 self-pay | AI analysis about $1,021 (CPT 75577, CMS 2026) plus the scan charge; totals in the low thousands |
| Repeat cadence | Guideline basis: one decision scan; rescan a zero at 3 to 7 years | None supported; periodic rescans are a concierge practice pattern |
Grades are research summaries, not medical advice. Decisions about your own stack belong with your clinician.
These pages grade a test in people without symptoms, as a population question. They never grade a reader, and they cannot read your result.
Longevity Lab Report
Screening news, graded when it happens.
The grades on this page stay free. The email is the week's longevity and biotech news in plain English: biotech news, expert moves, what is working, and new supplements and peptides, each story linked to the grades on this site.
Frequently asked questions
What does a calcium score miss?
Soft, non-calcified plaque. A CAC scan counts calcified plaque only, so a score of zero in a younger adult, a smoker, or a person with diabetes can coexist with soft plaque that carries risk. That blind spot is exactly why rung 2 exists, and also why a zero is not a lifetime pass in those groups.
Is CCTA with AI plaque analysis better than a calcium score?
It detects more: total plaque volume, the soft fraction, and stenosis, artery by artery, validated against invasive standards. Better detection is not the same as better outcomes. In people without symptoms, no evidence shows acting on the extra detection changes events, and the one screening trial (FACTOR-64) was null. That is why rung 1 grades B and rung 2 grades C- on the same rubric.
Who is rung 2 actually for?
A narrow group: a calcium score of zero or low where the statin decision genuinely hinges on plaque presence, with strong family history of premature coronary disease, current smoking, or diabetes. In the SCAPIS cohort, CCTA found visible atherosclerosis in 5.5% of CAC-zero adults, rising to 6.0% with strong family history, 6.8% in smokers, and 8.1% in diabetes. Outside that niche, and as serial plaque tracking, no guideline supports it.
How often should a calcium score be repeated?
Guideline-anchored: one decision scan; rescan a zero at 3 to 7 years depending on risk. For CCTA with AI analysis there is no supported screening cadence; rescans every 2 to 5 years are a concierge practice pattern, labeled as such.
What does the AI analysis cost?
The AI plaque analysis reimburses about $1,021 under CPT 75577 (CMS 2026 Physician Fee Schedule), on top of the CCTA scan charge; billed totals commonly run in the low thousands. Rung 1, by comparison, runs $50 to $400 self-pay all-in.
Sources
- Polonsky TS et al. Coronary artery calcium and risk prediction (MESA). JAMA, 2010. (2010)
- USPSTF. Risk assessment for cardiovascular disease with nontraditional risk factors: I statement on CAC. JAMA, 2018. (2018)
- Lindholt JS et al. DANCAVAS trial. NEJM, 2022. (2022)
- Dzaye O et al. MESA CAC warranty-period analysis. JACC: Cardiovascular Imaging, 2021. (2021)
Last reviewed October 18, 2026
- 2026-10-18|Initial grading from the 2026-10-18 screening evidence review.
Sources, rung 2
- Muhlestein JB et al. FACTOR-64 randomized clinical trial. JAMA, 2014. (2014)
- Choi AD et al. CREDENCE substudy, AI vs invasive reference. JACC: Cardiovascular Imaging, 2022. (2022)
- Bergstrom G et al. SCAPIS. Circulation, 2021. (2021)
- FDA 510(k) records for Cleerly LABS and ISCHEMIA (K202280; K242338, cleared March 7, 2025), verified 2026-10-18. (2024)
- 2026-10-18|Initial grading from the 2026-10-18 screening evidence review, including the full Cleerly verification record.
- 2026-10-18|Update: the AI analysis payment is now stated from the CMS 2026 physician fee schedule (about $1,021). The vendor's own release cites $1,026 and is labeled as the company's figure wherever it is mentioned.