CT coronary angiography in ED chest pain patients

CT Coronary Angiography In ED Chest Pain Patients: An Emergency Physician’s Perspective.

Anne-Maree KellyDirector, Joseph Epstein Centre for Emergency Medicine Research@Western Health, Melbourne

@kellyam_jec

This presentation may be reproduced in part or whole for education purposes on the condition that each reproduced slide contains the following:

‘Reproduced with permission of Professor Anne-Maree Kelly, Joseph Epstein Centre for Emergency Medicine Research @Western Health, Melbourne, Australia’

Permissions

Support for this meeting and advisory boards from Astra Zeneca

Travel support to speak at a conference (on blood gases) by Radiometer

Advisory board membership MSD

No relationships with cardiac diagnostic or imaging companies

Co-author of NHF guidelines for the management of ACS and addenda

Editorial boards of:◦ Annals of Emergency Medicine◦ Emergency Medicine Australasia◦ Hong Kong Journal of Emergency Medicine

Conflicts of interest disclosure

To explore the role of CTCA in ED chest pain patients, with a focus on those that ‘rule out’ for ACS

To compare the cost-benefit of a CTCA compared to alternatives

To provoke debate about the rational place of CTCA in ED chest pain work-up!

Objectives

Why are we talking about this?

From Schussler JM. Cardiac computed tomography:Emergeing cardiac devices and technology. Asian Hospital and Healthcare Management. http://www.asianhhm.com/diagnostics/cardiac_computed_tomography.htm

• Non-invasive• Nice pictures• Can ‘see’ if there are lesions

or not

Why are we talking about this

Three recent studies have suggested that CTCA for ED chest pain patients:

• Reduces ED length of stay

• Reduces admissions

• Negative scans have good prognostic performance

• Maybe more ‘accurate’ in identification of CAD than alternatives

ROMICAT II

ACRIN-PA

CT-STAT

Magnitude of benefit

ACRIN-PA ROMICAT II

50% reduction in admissions (23% vs. 50%)

25% reduction in LOS (18 hours vs. 25 hours)

67% reduction in median LOS (9 hours vs. 27 hours)

19% reduction in ED costs

Litt HI et al. N Engl J Med 2012; 366:1393-403. Hoffmann U et al. NEJM 2012; 367:299-308

Magnitude of benefit

CT-STAT

54% reduction in time to diagnosis (3 hours vs 6 hours)

38% reduction in costs

Goldstein et al. J Am Coll Cardiol 2011;58:1414-22

In Victoria, estimated 37,500 patients undergo ACS rule out in ED annually

The ‘rule in’ rate for ACS is ~15-20%◦ Depends how you count

About 30-32,000 have ACS ruled out and (according to ACS guidelines) need a further assessment strategy to rule out clinically significant CAD

The shape (and size) of the problem

Based on Dept Health Victoria data and estimates of chest pain presentations by Goodacre (UK): Goodacre et al. Heart. 2005; 91: 229–230.

Highly variable

Options◦ Exercise test◦ Nuclear medicine studies◦ CTCA◦ GP or cardiologist can decide!◦ Nothing (active choice)

Current practice in Australia

What does this group ‘look’ like?

TIMI score Demographics

0 33% 1 18% 2 18% 3 11% 4 11% 5+ ~9%

Male =60%

Average age=62

Known CAD = 33%

Based on data from cohort study @ WH 2009

My Key Questions Is CTCA sensitive for the detection of CAD?

Is CTCA suitable for the patient cohort in question?

Does negative CTCA have good prognostic performance for future ACS events?

Does CTCA improve outcomes for patients?

How does CTCA perform in comparison to alternative investigation strategies?

Which patients should have this test rather than an alternative?

Depends on whether analysis is at patient level or segment level◦ Patient level is of prime importance in the cohort of primary interest

Simple answer is ‘YES’

In a recent systematic review/ meta-analysis, CTCA had 94% (61-99%) sensitivity and 87% (16-100%) specificity for CAD.

Another meta-analysis of 64-slice +, reports sensitivity of 99% (95% CI 97-99%)

But about 9% of tests are non-diagnostic/ inconclusive

Is CTCA sensitive for the detection of CAD?

• Goodacre et al. Health Technol Assess 2013;17:1-188 • Mowatt et al. Technol Assess. 2008; 12:iii-iv, ix-143.

The question being asked is “Is there CAD”?

Just over 50% of the patient cohort is suitable for CTCA

About 30-40% of patients already have known CAD◦ Other investigation pathways more suitable in most

Other ‘contra-indications’: 10-15%◦ Metformin ◦ Inability to control rate adequately◦ Renal failure◦ Thyroid disease◦ Irregular rhythms

Is CTCA suitable for the patient cohort in question?

Hamid S et al. Am J Emerg Med. 2010;28:494-8

Safety◦ Short term adverse events related to the scan are very rare◦ Contrast allergy at expected rate (1/2,500-1/25,000)◦ Adverse effects due to rate control-usually minor◦ Radiation risk

Feasibility◦ Limited by access to scanner and availability of experienced readers◦ ‘In hours’ only availability does not match ED 24/7 patient flow◦ ‘Competition’ with other patients needing CT scan

Is CTCA feasible and safe in an ED chest pain population?

Simple answer is ‘YES’

In meta-analysis: I death from 1334 patients No PCI, MI etc Rate = 0.07% (95% CI 0.01% to 0.4%)

Does negative CTCA have good prognostic performance for future ACS events?

Goodacre et al. Health Technol Assess 2013;17:1-188

In meta-analysis: 39 events in 332 cases 12 MI Two thirds of events were revascularisations Rate 12% (95% CI 9-16%) Only one study was blinded to CTCA results:

◦ Showed CTCA results (presence of stenosis) was independently associated with MACE (HR 17)

Does a positive or intermediate CTCA have good prognostic performance for future ACS events?

Goodacre et al. Health Technol Assess 2013;17:1-188. SSchlett CA et al. JACC Cardiovasc Imaging. 2011;4: 481–491.

Focus is the sub-population without known CAD◦ 65-70% of cohort◦ 19,500-22,500 patients annually in Victoria

Available data suggests rate of undiagnosed CAD ~8-10%.

It all depends on risk of adverse events (cardiac death, MI) vs. cost

NICE (UK) sets a willingness to pay threshold at $30,000 to $45,000/ QALY

How does CTCA perform in comparison to alternative investigation strategies?

CTCA asks “Is there plaque”?

I am not sure that is the right question

Are we asking the right question?

A. 5%

B. 2%

C. 1%

D. 0.5%

What is the risk of MACE in patients without known CAD, with non-diagnostic ECG and normal serial biomarkers in ED?

A. 5%

B. 2%

C. 1%

D. 0.5%

At what MACE risk level is ‘routine’ testing indicated?

What is the risk of MACE in patients without known CAD, with non-diagnostic ECG and normal serial biomarkers in ED?

A. 5%

B. 2%

C. 1%

D. 0.5%

So what is the risk of MACE?

Fitzgerald P et al. Acad Emerg Med 2011;18:488–95.

Test Sensitivity NPV (MACE)

CTCA 94-99% >99%

MPS 87% 97.2%

Exercise ECG (EST) 20-30% As low as 86%

Comparison alternative strategies

Conti et al. Nucl Med Commun 2011 32;1223

Varying study design, populations and outcomes studied

In meta-analysis Rate of MACE for negative EST 0.7% (95% CI 0.5-

1.2%) But sensitivity questionable

◦ Some studies around 30% sensitivity for occlusive CAD

Prognostic performance of negative EST

Goodacre et al. Health Technol Assess 2013;17:1-188. SSchlett CA et al. JACC Cardiovasc Imaging. 2011;4:481–491.

Not enough data in the specific population of interest to draw conclusions

What about other testing or no testing?

Positive predictive value for CAD at segment level is only moderate (78%)◦ False positives: over-estimation of lesion severity in presence

of calcified plaques

Scanning 15,000 patients in Victoria/year will pose access issues for CT scanners!

Some down-sides

An ‘elephant in the room’

Retrospectively gated protocols, risk estimated at:◦ 0.11 to 0.13% for men◦ 0.27-0.37% for women

Prospectively gated protocols, risk estimated at:◦ 0.014-0.017% for men◦ 0.035-0.06% for women

Risk is inversely related to age

Significant ethnic variation

Radiation-related cancer risk

Huang et al. Br J Radiol.  2010;83(986):152-8.

Revisiting magnitude of benefit

ACRIN-PA

50% reduction in admissions (23% vs. 50%)

25% reduction in LOS (18 hours vs. 25 hours)

No patient with negative CTCA had death, MI within 30 days

Only 2/1357 (0.15%) of patients not diagnosed with MI at index visit had MI within 30 days

Trial conditions re CT availability

TIMI 0-2◦ >85% TIMI 0 or 1

Litt HI et al. N Engl J Med 2012; 366:1393-403.

Revisiting magnitude of benefit

CT-STAT

54% reduction in time to diagnosis (3 hours vs. 6 hours)

38% reduction in costs

Only included ED costs

Trial conditions re CT availability

Highly selected cohort

In CTCA cohort, 6 times greater rate of additional non-invasive tests after ED discharge◦ Cost◦ Radiation, etc

Goldstein et al. J Am Coll Cardiol 2011;58:1414-22

Revisiting magnitude of benefit

ROMICAT II

67% reduction in median LOS (9 hours vs. 27 hours)

19% reduction in ED costs

Eventual hospital costs actually 50% higher in CTCA group

Higher rate of additional testing (27% vs.12%)

No difference in events

Trial conditions re CT availability

Selected population ◦ 40-74◦ No AF or renal disease or BMI<40Hoffmann U et al. NEJM 2012; 367:299-308

Data from administrative dataset◦ Age 66+◦ Non-emergent, non-invasive test for ?CAD◦ No known CAD

Compared CTCA vs. stress myocardial perfusion scan Results:

Down stream impact of CTCA

Outcome CTCA MPS

Cardiac catheter 23% 12%

PCI 7.8% 3.4%

CABG 3.7% 1.3%

All cause mortality 180 days

1.05% 1.28%

Hospitalization for MI 180 days

0.19% 0.43%

Schreibati et al. JAMA 2011; 306:2128-36

1. That a test to rule out CAD before discharge is needed in ED chest pain patients◦ This is unproven!◦ The rationale for any test (compared to no test) is that it improves

outcome◦ Event rates are so low (<1%) in all arms that it is impossible to tell if

CTCA provided benefit 2. All lesions found were cause of symptoms

◦ 5% rate of occlusive lesions found in screening of asymptomatic patients

Flawed assumptions?

With risk of dye, radiation, extra tests etc. harm is likely to seriously compete with any benefit!

In Australia:

◦ ~75% of patients are discharged from ED/SSU

◦ Most do not have additional testing before discharge

◦ Median LOS of the order of 10 hours, depending on centre and protocol

◦ LOS likely to reduce as accelerated diagnostic biomarker pathways are validated

Fit with the Australasian context?

SCCT/AHA/ACC:◦ Symptomatic patients without known CAD with ‘intermediate’

pre-test probability◦ Symptomatic patients without known CAD with ‘low’ pre-test

probability who cannot perform a functional test or with equivocal functional test results

◦ Not suitable for high pre-test probability patientsdue to: High likelihood of plaques Limited spatial and temporal resolution These should have CA or functional test

Current indications

Taylor AJ et al. J Am Coll Cardiol 2010:56:1864-94.

CTCA is not indicated as a ‘routine’ test in ED patients with chest pain without known CAD and with normal biomarkers and ECG

It may be useful in a subgroup based on risk, but how this risk might be defined in unclear

There is a reasonable case for no further testing in significant proportion of ED chest pain patients who have had ACS ruled out by ECG and biomarkers

My view

Comparison of DM, ‘metabolic syndrome’ and other (MPS study)

Metabolic syndrome defined as at least 3 of:◦ Fasting glucose >110mg/dl◦ High BP◦ Low HDL◦ High triglicerides◦ High waist circumference

Rate of MACE at 1 year ◦ DM 30%◦ Metabolic syndrome 26%◦ Others 15%

Patient selection: one approach?

Conti et al. Nucl Med Commun 2008; 29:1106-12.

Could similar parameters identify a subgroup of patients who might benefit from CTCA?

Questions