Post on 16-Dec-2015
Cardiac Biomarkers
W. Frank Peacock, M.D., FACEPProfessor, Emergency Medicine
Cleveland Clinic
Biomarker?
• What is a biomarker?– An expensive lab test– Commonly Protein with levels that correspond to
• Diagnosis• Prognosis
– Most common method of measurement• ELISA
Lab-test-ology
Sensitivity
TP/(TP+FN)
Specificity
TN/(TN+FP)
Lab-test-ology
• LOB• LOD• CV• 99th %ile
CV vs LODAssay w LOD 5 pg/mL
LOD
99th %ile
CV and 99th%ile
10% CV
The box of undetectableness
Where is the 99th%ile?
CV and 99th%ile
10% CV20% CV
30% CV
01020
30405060
7080
0 2 4 6 8 12 18 24 32 48 72
Hours After Onset of MI
CKMB
Myo
TnI
Historical timing of cardiac necrosis markers
TIMI IIIB: Troponin I Levels Predict Mortality In UA/NSTEMI
1.0
1.7
3.4
3.7
6.0
7.5
0
2
4
6
8
0 to <0.4 0.4 to <1.0 1.0 to <2.0 2.0 to <5.0 5.0 to <9.0 >9.0
831 174 148 134 6750
Risk Cardiac Troponin I (ng/ml)Ratio 1.0 1.8 3.5 3.9 6.2 7.8
Mo
rtal
ity
at 4
2 D
ays
(% o
f P
atie
nts
)
Antman EA, et al. N Engl J Med. 1996;335:1342-1349
2000• Emergency docs– This crap is useless in almost all patients– Only helpful if positive
• Rarely positive, <5% of chest pain– The rest of chest pain requires other testing
• Cardiologists own troponin– Tactics-TIMI 18– IF it is detectable, it is an MI, otherwise forget it– If positive, don’t even bother thinking, just call the cath
lab
C StatisticC Statistic
AUC = C StatisticC Stat = 50%REALLY BAD TEST
C Stat = 1.0A PERFECT Test
C Stat = 75
C Stat = 0.69OK test
Area Under the Receiver Operator Characteristic (ROC) Curve
Reichlin T. N Engl J Med 2009;361:858-67.
718 consecutiveED suspect AMI
MI/USA 238 (33.1%)
ACEP Marker Recommendations
• Level A recommendations Don’t use markers to exclude non-AMI ACS (ie, unstable angina)
• Level B recommendations Use any of the following to exclude NSTEMI
• 8-12 hours after symptom onset– A single (-) CK-MB mass, TnI, or TnT
• Serial measures if < 8 hours after symptom onset– Baseline and 90 mins
» A (-) myoglobin with a (-) CKMB, or (-) Tn– (-) 2-hour delta
» CK-MB and Tn
Can you trust them?
01020
30405060
7080
0 2 4 6 8 12 18 24 32 48 72
Hours After Onset of MI
CKMB
Myo
TnI
Historical timing of cardiac necrosis markers
hsTnI
If It Moves, It Is Bad
Marker Comparator OR for 30 day MACE
95% CI
ing Tn
vs. stable Troponin
2.25 1.42-3.55
ing Tn 3.04 1.94-4.75
ing CKMB
vs. stable CKMB
0.67 0.48-0.95
ing CKMB 0.96 0.57-1.60
Logistic regression models showing the odds ratios for predicting ACS
MACE: MI, revascularization (PCI or CABG), or positive testing (>70% stenosis at
catheterization, [+] MPI or non-invasive stress testing) within 30 days of index visit. McMullin N. AJEM 2009.
Event Free vs cTnT ValuesEvent Free vs cTnT Values
Days after admission to the hospital
AJC 93:278, 2004
Normal<0.01 g/L
Marginal<0.01-0.09 g/L
Frank elevation>0.01 g/LP=0.004
STRIVE®
You can’t have it both ways…
Tn, its not just for AMI anymore
Tn Elevation w/o Overt Cardiac Ischemia• Trauma
• contusion, ablation, pacing, ICD firings, cardioversion, endomyocardial biopsy, cardiac surgery, interventional closure of ASDs
• CHF
• Aortic valve disease and HOCM with significant LVH
• HTN
• Hypotension, often with arrhythmias
• Postoperative noncardiac surgery patients who seem to do well
• Renal failure
• Critically ill patients, esp with diabetes, respiratory failure, gi bleeding, sepsis
• Drug toxicity, eg adriamycin, 5 FU, herceptin, snake venoms, carbon monoxide poisoning
• Hypothyroidism
• Abnormalities in coronary vasomotion, including coronary vasospasm
• Apical ballooning syndrome
• Inflammatory diseases
– myocarditis, eg. Parvovirus B19, Kawasaki disease, sarcoid, smallpox vaccination, or myocardial extension of BE
• Post PCI patients who appear to be uncomplicated
• Pulmonary embolism, severe pulmonary hypertension
• Sepsis
• Burns, esp if TBSA > 30%
• Infiltrative diseases including amyloidosis, hemachromatosis, sarcoidosis and scleroderma
• Acute neurological disease
– CVA, subarchnoid bleeds
• Rhabdomyolysis with cardiac injury
• Transplant vasculopathy
• Vital Exhaustion
What now?Cardiologists are in a tizzyCardiologists are in a tizzy
All these All these ““false positivesfalse positives””
Emergency docs think this is greatEmergency docs think this is great
There is no such thing as a false positive There is no such thing as a false positive when your talking about being when your talking about being
DEADDEAD
Reichlin T. N Engl J Med 2009;361:858-67.
Do we really gotta be doing serial troponin’sanymore???
2012• The decade of 2000-10 – Will be remembered as when the cardiologists owned troponin
• Used to be an MI marker– Those days are gone
• Emergency Medicine– Taking troponin back from the cardiologists!– IT IS NOT AN AMI MARKER ANYMORE
• Now it’s a 14 day death marker– I don’t care about 30 days or 180 days from now– I REALLY don’t care about a year from now
Myocardial InfarctionIt’s a changing world
• An MI used to be– >40 and sweating with chest pain– Positive markers in 8-12 hours
• Now– It aint >40– It aint sweating– It aint even chest pain
It would be really great if they had it written on their forehead!!
In 2011, you will miss 423,600
Acute Myocardial Infarction’s
1/3 have no chest pain
Canto JG et al. JAMA. 2000;283:3223-3229
If you think this is the way they look…
STRIVE®
% With Chest Pain During AMIStratified by Age
SOBW&DN/VSyncopeConfused
STRIVE®
28
When your laying naked around the ER, they all look the same……
This one is having an AMI
STRIVE®
Closing Time
You don’t have to go home, but you can’t stay here….
– Semisonic
STRIVE®
The ER docs challenge
Admit them all:
and let the insurance company sort them out…
Discharge them all and let God sort them out…
STRIVE®
Emergency Medicine Roulette
What % are discharged from the ED??
• 14 Asia-Pacific region EDs
• >18yo with >5 mins CP
• Risk stratification (blinded to care team)– TIMI<1, ECG non-dx, – Negative 0 & 2hr POC Tn, CKMB, myo
• Endpoint: 30 day MACE
Than M. Lancet, 2011. DOI:10.1016/S0140-6736(11)60310-3
STRIVE®
33
TIMI Risk Score
Risk factors:
– Age 65 years
3 risk factors for CAD
– Prior coronary stenosis 50%
– ST-segment deviation on ECG
2 anginal events in last 24 hours
– Use of ASA in last 7 days
– Elevated serum cardiac markers CK-MB or troponin
Each risk factor is assigned 1 point, and the total represents a given patient’s TIMI Risk Score1
Event rates (all-cause mortality, MI, or urgent revascularization) increase with each 1-point increase in score (P<0.001 by chi square test for trend)1
4.78.3
13.2
19.9
26.2
40.9
0
5
10
15
20
25
30
35
40
45
0/1 2 3 4 5 6/7
Number of Risk Factors1R
ate
of
Co
mp
os
ite
En
dp
oin
t
(Da
ys
1-1
4),
%
1. Antman EM et al. JAMA. 2000;284:835-842.
• N=3582– 30 day MACE in 421 (11·8%)– Most often NSTEMI
• ADP identified 9·8% (352/3582) as low risk– 3 (0·9%) had 30 day MACE
Than M. Lancet, 2011. DOI:10.1016/S0140-6736(11)60310-3
• Potential costs savings in low risk negative ADP patients
• Hospital LOS– Median 26·0 h (IQR 9·9–37·0) – Mean 43·2 h (95% CI 36·2–51·2)
Than M. Lancet, 2011. DOI:10.1016/S0140-6736(11)60310-3
STRIVE®
He is a 67 year old, hypertensive, obese man. He took an aspirin this morning, he still smokes and has high cholesterol. Many of his family have CAD. He has been a diabetic for 15 years, and 4 years
ago he had an MI.
Age > 65, 3 risk factors, H/O MI, took asa: TIMI Risk score = 419.9% chance of death, MI, or UTVR in the next 14 days
George is sitting in his barat his restaurant across the streetfrom the Emergency Department
STRIVE®
He is a 67 year old, hypertensive, obese man. He took an aspirin this morning, he still smokes and has high cholesterol. Many of his family have CAD. He has been a diabetic for 15 years, and 4 years ago he
had an MI.
TIMI Risk score = 419.9% chance of death, MI, or UTVR in the next 14 days
George is laying in the ED, diaphoretic, with crushing CP,
nauseated, BP = 100/70
Can we discharge you??Derivation by blinded sampling (N= 703)
• 130 (18.5%) AMI – None w initially undetectable hs-cTnT – Sn 100.0%, NPV 100.0%
• 27.7% would have ‘ruled out’ for AMI– 2 (1.0%) died or had AMI w/in 6 months
» (1 peri-procedural AMI, 1 non-cardiac death)
Validation by standard practice (N= 915)– 1 patient (0.6%) with initially undetectable hscTnT developed
subsequent elevation (to 17ng/L)• Sn 99.8% (99.1-100.0)• NPV 99.4% (96.6-100.0).
Body, et al. JACC, 2011
European Society of Cardiology
• A Tn @ presentation cannot R/O NSTEMI– Repeated Tn 3 hours after admit or more CP. – LOE 1B
• Tn is preferred over CKMB
• Myoglobin is not specific or sensitive enough– Is not recommended.
STRIVE®
ESC Guidelines
Due to improved analytical sensitivity, low troponin levels can be detected in stable angina and in healthy patients.
The mechanisms of this troponin release are not yet explained, but ANY measurable troponin is associated with an unfavourable prognosis.
Low Level TroponinsOne Cut-off or Two?
0
2
4
6
8
10
12
14
16
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 More
pg/ml cTNI
Fre
qu
ency
Myocardial necrosis AMI
Sp=99%Wait and see, do more tests
AMISp=85%
The Now and Then of High Sensitivity Troponins
Last decade
• Detectable Tn– 99th %ile cutpoint
• Great specificity• Better sensitivity
– No real clinical disposition impact for the ER
• Serial testing of less necessary
Next decade
• Good bye specificity• 2 cutpoints?
• Second marker – copeptin, ST-2, MPO,
IMA, etc
• Hello sensitivity• Exclude ischemia?
• Challenges– The role of cardiology
consults– EDUCATION…………
So next time you want to get a troponin….
• Risk stratify (after decide it might be ACS)
• You want to send that patient home?– Put on your thinking cap!– AMI? Something else?
• Can always repeat
Results 25 participating hospitals
N=1,360 patients
Overall
Mean DTBT 115.770.1 minutesMedian 100; IQR=73,138
Central lab
Mean DTBT 119.2 70.5 minutes Median 103; IQR=76,141
Point of Care
Mean DTBT 68.2 40.8 minutesMedian 62.5, IQR=43,83.5
Peacock WF et al. Acad Emerg Med. 2004;11:569–570.
Saves about1 hour
Delay = Death
N= 13,934,542• Adverse events increase with the mean LOS in similar patients in the same ED shift
• OR for Death if LOS ≥6 v <1 hr cohorts– Hi Acuity 1.79 Low Acuity 1.71
BMJ 2011; 342:d2983
Overcrowding = Long waits Long waits = Death
• N= 62,495• Risk ratio for DEATH
– Per hour of ED stay = 1.1 (p < 0.001)
– Per hour of ED wait = 1.2 (p=0.01)
MJA 2006; 184: 208–212
Delay = Bad Care
• N=42,780 • Long ED stays less often received guideline-recommended
NSTEMI therapiesAnn Emerg Med. 2007; 50; 489-96
Delay = Bad Care
• N=694 patients Delayed/No antibiotics– OR 1.05 for each additional WR patient– OR 1.14 for each additional WR hour
Ann Emerg Med. 2007;50:510-516
Delay = Bad Care
• N=13,758• Nontreatment of pain associated with waiting room number
OR = 1.03 for each additional waiting patient
Ann Emerg Med. 2008;51:1-5.]
Delay = Bad Care
• N=162 “boarded” patients (waiting for room)• Undesirable event
• Missed meds, lab results, arrhythmias, or other adverse events
• 27.8% had an undesirable event
Ann Emerg Med. 2009;54:381-385.]
14,054,431 patients:
waiting = bad outcomes or
death
What business intentionally kills its customers?
The era of POC needing to justify itself is over. We are now in the era where the central lab must prove it is not killing our patients.
If you had a way of getting data quickly, wouldn’t you?
POC vs Lab Singulex
• 295 MIDAS patients155 (52.5%) NCCP
67 (22.7%) USA
61 (20.7%) NSTEMI
12 (4.1%) STEMI
Sensitivit
y
Specificit
y
Negative
Predictive
Value
Positive
Predictive
Value
Area Underneath
ROC curve
(C-Statistic)
Point of Care
Alere
86 94 95 82 94
Central Lab
Singulex
90 86 96 68 94
Sn Sp NPV PPV AUC
POC 86 94 95 82 94
Lab 90 86 96 68 94