Transcript Multicenter evaluation of a 1h-algorithm in the diagnosis of

Multicenter evaluation of a 1h-algorithm in the
diagnosis of myocardial infarction using
high-sensitivity cardiac Troponin T
Christian Mueller, Evangelos Giannitsis, Michael Christ, Jorge OrdóñezLlanos, Christopher R. deFilippi, James K. McCord, Richard Body, Mauro
Panteghini, Tomas Jernberg, Mario Plebani, Franck Verschuren, John K.
French, Robert H. Christenson, Silvia Weiser, Garnet Bendig, Peter Dilba,
Bertil Lindahl,
for the TRAPID-AMI Investigators
High sensitivity cardiac Troponin T assay for RAPID rule-out of Acute Myocardial Infarction
Sponsored by
Disclosures
• Swiss National Science Foundation
• .
• .
..
•
Background: Early diagnosis of AMI
• Critical for timely initiation of evidence based
therapies, but still an unmet need in many patients
• Delayed “rule-in”  morbidity + mortality 
• Delayed “rule-out”  time in ED + costs 
patient anxiety 
• 3 key tools:
History incl. chest pain characteristics
ECG
cTn
Background: Early diagnosis of AMI
• Critical for timely initiation of evidence based
therapies, but still an unmet need in many patients
• Delayed “rule-in”  morbidity + mortality 
• Delayed “rule-out”  time in ED + costs 
patient anxiety 
• 3 key tools:
History incl. chest pain characteristics
ECG
hs-cTn
1.Rule-in
2.Rule-out
0h 1h 2h 3h 4h 5h 6h 7h
ECG
cTn
ESC 2011:
hs-cTn
TRAPID-AMI: hs-cTnT
hs-cTnT
cTn
hs-cTn
APACE: hs-cTnT 1h-algorithm
Mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
600 pts
NPV 100%
95% CI 99-100%
150 pts
PPV 76%
95% CI 69-82%
Reichlin T, et al. Arch Intern Med 2012
APACE: hs-cTnT 1h-algorithm
491 Patients (56%)
NPV 100%
95% CI 99-100%
212 Patients (24%)
Prevalence of AMI 9%
169 Patients (19%)
PPV 76%
95% CI 69-82%
Aim of TRAPID-AMI:
to prospectively validate the hs-cTnT 1h-algorithm
Reichlin T, et al. Arch Intern Med 2012
12 sites on 3 continents
Methods II
• TRAPID-AMI ≈ APACE
• Prospective diagnostic study in the ED
• Inclusion: acute chest pain suggestive of AMI
onset < 6h (<12h in APACE)
written informed consent
First blood draw <45min since presentation
• Exclusion: terminal kidney failure on dialysis
Trauma
DC or defibrillation prior to inclusion
CABG <1month or AMI <3 weeks
• Blinded measurement of hs-cTnT at 0h+1h
Methods III
• Final diagnosis: central adjudication by two
independent cardiologists according to the universial
definition of MI using all information including serial
cTnI-ultra (hs-cTnT in APACE)
• Primary endpoint:
NPV for AMI
• Secondary endpoints:
PPV for AMI, ….
• Follow-up: 30d mortality
• Sample size calculation: based on APACE
761 patients in rule-out zone, 1233 patients overall
Results
Table 1
Baseline Characteristics of the Patients
All patients
Acute MI
Others
P Value
(n=1282)
(n=213)
(n=1069)
Age – years
62 (50-74)
68 (57-78)
60 (49-73)
< 0×001
Male gender – no. (%)
805 (62×8)
150 (70×4)
655 (61×3)
0×013
Hypertension
805 (62×8)
156 (73×2)
649 (60×7)
< 0×001
Diabetes
270 (21×1)
53 (24×9)
217 (20×3)
0×141
Current smoking
288 (22×8)
46 (21×9)
242 (23×0)
0×788
History of smoking
468 (37×1)
93 (44×3)
375 (35×6)
0×019
Previous myocardial infarction
319 (24×9)
71 (33×3)
248 (23×2)
0×002
Coronary intervention
388 (30×3)
78 (36×6)
310 (29×0)
0×033
Stable angina pectoris
148 (11×5)
29 (13×6)
119 (11×1)
0×292
Unstable angina pectoris
164 (12×8)
40 (18×8)
124 (11×6)
0×007
Congestive heart failure
107 (8×3)
30 (14×1)
77 (7×2)
0×002
Time since chest pain onset (h)
3×4 (2×1-6×0)
3×6 (2×4-5×9)
3×4 (2×1-6×0)
0×492
81 (65-97)
73 (57-90)
82 (67-99)
< 0×001
Systolic blood pressure (mmHg)
141 (127-157)
143 (130-162)
141 (126-155)
0×037
Diastolic blood pressure (mmHg)
81 (72-90)
82 (71-94)
81 (72-90)
0×376
Heart rate (beats/min)
76 (66-88)
79 (67-91)
76 (65-87)
0×013
Risk factors – no. (%)
History – no. (%)
Creatinine clearance - (ml/min/1.73m2)
Table 1
Baseline Characteristics of the Patients
All patients
Acute MI
Others
P Value
(n=1282)
(n=213)
(n=1069)
92 (7·2 %)
23 (10·8 %)
69 (6·5 %)
0·029
1148 (89·5 %)
182 (85·4 %)
966 (90·4 %)
0·037
42 (3·3 %)
8 (3·8 %)
34 (3·2 %)
0·673
Left ventricular hypertrophy (%)
65 (5·1 %)
14 (6·6 %)
51 (4·8 %)
0·303
Complete LBBB (%)
36 (2·8 %)
12 (5·6 %)
24 (2·2 %)
0·011
Complete RBBB (%)
55 (4·3 %)
13 (6·1 %)
42 (3·9 %)
0·192
Paced ventricular complex (%)
23 (1·8 %)
8 (3·8 %)
15 (1·4 %)
0·041
Pathological Q-waves (%)
119 (9·3 %)
31 (14·6 %)
88 (8·2 %)
0·006
Significant ST-segment elevation (%)
60 (4·7 %)
20 (9·4 %)
40 (3·7 %)
0·001
St-segment depression (%)
164 (12·8 %)
63 (29·6 %)
101 (9·4 %)
< 0·001
T-inversion (%)
184 (14·4 %)
54 (25·4 %)
130 (12·2 %)
< 0·001
ECG
Rythm (%)
atrial fibrillation
sinus
other rhythm
Results
95% CI 98.2-99.7%
95% CI 70.4-83.0%
Subgroups:
Results
0.1%
Limitations
• ED patients vs general practitioner (pre-test probability)
• Observational diagnostic study. Physicians were
blinded to the investigational hs-cTnT results and
patients were not managed based on these results.
• May underestimate the true NPV of the hs-cTnT 1halgo due to enrichment with early presenters
• Performance in patients on dialysis is unknown
Conclusions
• The hs-cTnT 1h-algorithm performs well for early
rule-out as well as rule-in of AMI.
• Effectiveness (% rule-out/in) and safety (NPV & 30dmortality in rule-out, PPV in rule-in ≈ APACE.
• Clinical implication: Used in conjunction with all
other clinical information including ECG, the hs-cTnT
1h-algorithm will be a safe and effective approach to
substantially reduce time to diagnosis in ≈75% of
acute chest pain patients.