• 
• 

Three Part Question

In [patients following coronary arterial bypass surgery] can [Troponin T or I] be used as a diagnostic tool to [diagnose perioperative MI]

Clinical Scenario

You have just operated on a 72 year old man who was an urgent referral for revascularisation after 2 days of unstable angina despite maximal medical therapy and intravenous nitrates. You had to perform 5 grafts and the operation was long and 3 of the target vessels were small and difficult to graft. You used an on-pump technique with anterograde intermittent cold blood cardioplegia, but as you come off bypass the ECG demonstrates elevated ST segments anteriorly and laterally. You recheck the anastomoses which seem patent and the patient comes off bypass with a moderate dose of adrenaline. He is stable in the Intensive care postoperatively but you are concerned that he has had a perioperative MI. Your anaesthetist tells you that Troponin T is unreliable but Troponin I is really good for diagnosis of MI post cardiac surgery but is unable to tell you what level of TnI is sufficient to make the diagnosis.

Search Strategy

Medline 1990- March 2004 using the Ovid interface
[exp coronary artery bypass/ OR coronary arter$ bypass.mp OR CABG.mp OR exp cardiac surgical procedures/ OR cardiac surgical procedures.mp OR coronary bypass.mp OR cardiac surgery.mp] AND [troponin.mp. OR exp Troponin T/ OR exp Troponin/ OR exp Troponin I OR TnI.mp OR TnT.mp] AND [exp Myocardial Infarction/ OR Myocardial infarct$.mp OR MI .mp OR myocardial damage.mp OR myocardial injury.mp] AND [exp Intraoperative complications/ OR postoperat$.mp OR perioperat$.mp.]

Search Outcome

A total of 191 articles were found of which 17 were deemed to be relevant.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Katus et al 1991 UK	N=90 [(56pts undergoing cardiac surgery), (22 minor orthopaedic procedures and 12 lung surgery controls)] Diagnosis of Perioperative MI : New Q waves >0.04ms and 25% R wave reduction in 2 leads on day 7	Prospective Diagnostic Study Level 3b	Troponin T after surgery 2,8,12,24 hours postoperatively and then once daily up to six days)	Tn T levels were detectable [median 5 range1.3-11 µg/L] in all pts undergoing cardiac surgery. 5 pts had MI and Tn T release persisted and serum concentrations (mean 11 µg range 6-31 µg/L) reached peak on fourth post operative day. Troponin T was undetectable in both control groups.	Very small study with only 5 perioperative MIs. In addition no cut-of point for TnT identified for diagnosis of MI and thus performance of TnT as a diagnostic test not fully assessed. Control groups largely irrelevant. (TnT not raised in any control patients)
Hake et al 1993 Germany	N=90 Elective CABG patients with severe triple vessel disease TnI and CKMB Diagnosis of Perioperative MI New Q waves.	Prospective Diagnostic cohort study Level 3b	Troponin T after surgery at 6, 12, 24 and 48 hrs	72 patients with normal ECG TnT median of 0.37 µg/L at 24hrs (quartile 0.13-0.5 µg/l ). 7 patients with MI TnT median level 10.5 µg/L (quartile 6.3-12.5 µg/l)	Small study. No analysis for optimal cut-off point. Loose definition of MI and 2 patients with elevated TnT but no Q-waves had other evidence of MI, but were not included in the MI group
Mair et al 1994 Austria	N=28 26 Elective CABG 2 Salvage CABG Serial TnI. ECG monitoring, CKMB and echocardiography. Diagnosis of Perioperative MI ECG, Echo, and increased CK-MB activity	Prospective Diagnostic cohort Study. Level 3b	Troponin I	4 Pts with PMI, 1 with new Q-waves TnI 30µg/L, 3 with non Q-wave MI TnI ±5µg/L. 2 patients with peri-op myocardial damage by clinical, ECG or Echo findings – TnI 3.9 and 3.4 µg/L. (CKMB <20)	Suggested Cut-off values: Peak >3.7 µg/L 12 hr >3.1 µg/L 24 hr >2.5 µg/L High probability of PMI.
Sadony et al 1997 Germany	N=119 patients with severe diffuse three-vessel disease undergoing CABG. Diagnosis of Perioperative MI: Non transmural MI Gp II TnT>1.0ng/ml, CKMB >20U/l, ECG non specific, Echo unchanged or new hypokinesia Transmural MI Gp III : TnT >1.5ng/l, CKMB >20U/l, ECG new Q waves , Echo New akinesia (Also Gp IV defined as pre-op MI )	Prospective Diagnostic cohort study Level 3b	Troponin I at 2, 3, 6, 8, 12, 24, and 48 hrs	Group I no MI (n=87) : TnI peaked at 4.8 ±2.5ng/ml at 12 h and 5.2 ±3.2ng/ml at 24 h. Groups II and III evidence of MI (n=27) : TnI significantly higher but values not given	The gold standard of MI is dependant on TnT and CKMB in this paper which may be unreliable markers of MI. In addition 6 patients had problems in categorisation with TnT and CKMB results differing in these patients. Data not presented in full detail
			Cut off point to diagnose MI	TnI at 24 hrs of 11.6 ng/ml resulted in a sensitivity of 100% and specificity of 96.6% for diagnosis of MI
Alyanakian et al 1998 France	N=41 CABG (n=17) Valve Replacement (n=24) Diagnosis of MI : Group I : Q wave MI on ECG and new areas of dyskinesia on Echo Group II : non specific changes on ECG or inotropes post-op. Group III : no changes on ECG daily and echo on day 4.	Prospective Diagnostic Cohort study Level 3b	Troponin I at 3, 12, 20, 24, 48hrs post -op	Gp I, New MI (n=5) Peak TnI 59.0 ± 38.8 µg/L at 24 hrs. Gp II Non Q MI (n=12) Peak TnI 26.2 ± 14.8 µg/L at 24 hrs. Gp III No MI (n=24) Peak TnI 7.1 ± 4.1 µg/L at 12 hrs	Small study Data not presented in full detail
			Cut off point to diagnose MI	TnI level of 15 µg/L, was the optimal cut-off. All non MI patients had TnI less than 15, and most in gp I were higher
Bonnefoy 1998 France	N=82 Elective CABG Definition of MI : New Q waves of at least 0.03secs on ECG at 24 hours or on discharge	Prospective Diagnostic Cohort study Level 3b	Troponin T and Troponin I at 12, 24 and 48 hrs	Pts without MI (n=69) : TnI peak 2.1 ± 2.4 µg/L (median) at 12hrs. TnT peak 0.22±0.2 µg/L at 48hrs. Pts with MI (n=13) TnI peak 17 ± 16 µg/L at 12hrs TnT 1.4 ± 2.3 µg/L at 12hrs	Small study, unable to demonstrate that TnI was significantly better than TnT in diagnosing MI. Gold Standard only used ECG criteria to diagnose MI
			Cut off point to diagnose MI	Best cut-off value for TnI=5 µg/L. 91%sens, 82%spec,TnT =0.3 µg/L. 75% sens , 75% spec
Jacquet et al 1998 Belgium	N=110 Elective CABG patients Diagnosis of MI : Patients with new Q wave abnormalities, or Echo dyskinetic segments, or prolonged ischaemia with ST-T wave changes for >15 mins.	Prospective Diagnostic Cohort study Level 3b	Troponin I at 2, 6, 10 and 20 hours post-op	Normal patients n=99, at 10 hrs TnI mean 3.6, (2.7 - 5.5 µg/L). Pts with MI n=11, at 10 hrs TnI mean 27.6, (19-45 µg/L).	Diagnostic uncertainty at levels between 8.4 and 13.1 µg/L at 6 hours. Gold standard for MI included a group of patients with ischemia only, in direct contrast to most of the other studies listed here.
			Cut- off point to diagnose MI	At 6hrs post-op: TnI> 8.4 µg/L 100% sens and 89% spec. TnI >13.1 µg/L 90% sens and 100% spec. At 10 hrs TnI >14.9 µg/L 90% sens and 100% spec. At 20 hrs TnI >13.4 µg/L 63% sens and 100% spec.
Gensini et al 1998 Italy	N=42 Elective CABG Diagnosis of MI : CK-MB of >50iu/l in first 2 days, and persistent new Q waves > 0.04 secs ECG and new persistent wall abnormalities on Echo	Prospective Diagnostic Cohort study Level 3b	Troponin I at 4, 12, 24, and 48 hrs	8 pts had Q-wave MI. All had elevated TnI > 9.2ng/ml. 34 pts had No MI: TnI all <9.0ng/ml.	Small study, ranges for groups not given.
			Cut- off point to diagnose MI	100%sens, 100% spec for TnI 9ng/ml. Peaks occurred between 12-24h after surgery.
Fellahi et al 1999 France	N=102 Elective CABG, Pericardial and serum TnI compared Diagnosis of MI Allocated to 1 of 3 groups according to ECG in first 24hrs Gp1 : normal ECG, Gp2: Non specific abnormalities , Gp3: new Q waves >0.04s, or red R waves of >25% in two leads	Prospective Diagnostic Cohort study Level 3b	Serum Troponin I at 3, 6, 12, 24 hrs	Gp 1 no MI : Mean TnI 6.1±6.0 at 12 hrs. Gp 2: non specific changes: Mean TnI 7.1±7.7 at 12 hrsGp 3 Acute MI: Mean 40.2±46.7 at 12 hrs	Pericardial TnI was less useful than serum level in diagnosing PMI.
			Pericardial Troponin I	Gp 1 no MI : 367±339ng/ml. Gp 2 : 558±608ng/ml Gp 3 acute MI : 1,318±1,810ng/ml
			Cut-off point	No accurate cut-off point presented
Carrier et al 2000 Canada	N=590 Elective CABG, TnT levels in 493 pts TnI levels in 97 pts. Diagnosis of MI Two of : new Q wave or disappearance of R wave on ECG ;CK-MB >100iu/l;positive pyrophosphate myocardial scan	Prospective Diagnostic Cohort study Level 2b	TnT levels at 24, hrs	MI group(n=22) had TnT mean 7.1±4.7ng/ml No MI group(n=471) had TnT mean 0.6±0.6ng/ml.	Small number in Troponin I group, but good number in TnT group Good Gold standard Note patients with TnT >3.4ng/ml had higher mortality (4% vs 0.7%)
			Troponin I at 24 hrs	MI group (n=9) Had TnI mean 26±39ng/ml No MI group (n=70) Had TnI mean 1.8±0.8ng/ml
			Cut-off point for diagnosis of MI	TnT level above 3.4ng/ml at 48 hrs had sensitivity of 90% and specificity of 94%. TnI level above 3.9ng/ml at 24hrs had sensitivity of 80% and specificity of 85%
Greenson et al 2001 USA	N=100 71 CABG 29 AVR Diagnosis of MI either: Serial ECGS showing New Q waves, on 2 or more leads and a least 1/3rd height of QRS, or ST rise >0.2millivolts ; or Wall abnormality on Echo,	Prospective Diagnostic Cohort study Level 3b	TnI levels	No MI , : No patient with TnI <40ng/ml had a cardiac event. MI : 13 of 17 patients with a cardiac event had peak TnI over 40ng/ml	Confidence intervals for sensitivities and specificities not given
			Cut-off point for diagnosis of MI	Peak TnI > 40ng/ml 100% sensitivity 95%specificity Peak TnI > 60ng/ml. 76%sensitivity 98%specificity
Benoit et al 2001 France	N=260 Valvular surgery n=152 CABG n=76 CABG and valve n=32 Diagnosis of MI : New Q wave on ECG, alteration of ST segments > 12 hrs duration in 2 leads. TOE wall motion defect, increased CK-MB levels	Prospective Diagnostic Cohort study Level 3b	TnI at 12 and 24 hours	MI patients Mean TnI 86 ±27ng/ml , range 25 to 239 ng/ml. Non MI patients mean TnI 17.5 ± 1.4 ng.ml	Only 8 patients had perioperative MI. This paper has been criticised for using gelatin based colloid in these patients,which interferes with the assay used here to measure TnI
			Cut-off point for diagnosis of MI	TnI of > 19µg/L at 12h had 100%sens, 73% spec TnI of 36µg/L at 24h had 100%sens, 93% spec.
Peivandi et al 2001 Germany	Only 8 patients had perioperative MI. This paper has been criticised for using gelatin based colloid in these patients,which interferes with the assay used here to measure TnI	Prospective Diagnostic Cohort study Level 3b	TnT levels	Normal Patients TnT 1.1ng/ml [CI 0.4-1.6]. MI patients (N=4) median TnT 63.8ng/ml	Diagnosis of MI was based on Q waves on ECG AND TnI>8.35ng/ml. Therefore this 'gold standard' is not independent of the test in question, i.e. TnI
			Correlation between graft patency and MI	All grafts patent in normal patients. In 4 patients diagnosed as post-op MI no occluded grafts were found. One occluded graft was found in a patient diagnosed as minor myocardial injury
			Cut-off point for diagnosis of MI	TnI>8.35ng/ml . 75% sensitivity and 82% specificity.
Andres et al 2001 Poland	N=49 patients undergoing CABG Diagnosis of MI : On Basis of ECG criteria	Prospective Diagnostic Cohort study Level 3b	Cut-off point for diagnosis of MI	TnI cut-off of 0.8ng/ml gave sensitivity of 80% and specificity of 94%. TnT cut-off of 0.41ng/ml gave sensitivity of 86% and specificity of 88%	Article is in Polish, thus authors contacted for English translation
Castro et al 2002 Spain	N=64 patients undergoing CABG Diagnosis of MI New Q waves on ECG and Echo regional wall abnormalities, and CK >400iu/l and CK-MB>40iu/l.	Prospective Diagnostic Cohort study Level 3b	Cut-off point for diagnosis of MI	TnI higher than 12ng/ml gave sensitivity of 91% and specificity of 89%.	Article is in Spanish and thus authors contacted for English Translation.
Holmvang Denmark 2002	N=103 Elective CABG, with all patients undergoing repeat angiography. Diagnosis of MI : ECG showing : New Q waves >40ms in 2 or more leads; or LBBB ; or a 'QRS score' of 3 points or more Diagnosis of Graft occlusion 100% occlusion of graft on angiography at day 5-7 post-op.	Prospective Diagnostic Cohort study Level 3b	TnT level	No MI : TnT median 0.9 (IQ range 0.4 to 2.0ng/ml). Graft occlusion : TnT median 3.7 (range 1.4-8.8ng/ml)	8 % of vein grafts and 2 % of LIMA grafts occluded 5-7 days post-op Of 10 patients with new Q waves or new LBBB, only 3 patients had graft occlusion on post-operative angiography.
			TnI levels	No MI : TnI median 6.9 (IQ range 3.5 to 13.8 ng/ml). Graft occlusion : TnI 9.9 (range 0.7-56.4ng/ml)
			Cut-off point for diagnosis of MI or graft occlusion	For PMI on ECG: optimal discrimination level for TnT 1.75µg/L (sens 0.71 spec 0.68). For graft occlusion: TnT 3 µg/L (sens 0.67 spec 0.76).
Fransen et al 2002 Netherlands	N=181 Elective CABG Diagnosis of MI : ECG changes of new Q waves, or >1mm ST elevation, , or a typical rise and fall of CK, CK-MB and ASAT curves	Diagnostic Cohort study Level 3b	TnT levels	No MI : TnT peak level 3.3 ±0.7ng/ml. MI : TnT peak level 0.9 ± 0.05ng/ml	TnT >1.0ng/ml had a sensitivity of 77% and specificity of 78% at 8 hours for diagnosis of MI
			Cut-off point for diagnosis of MI	TnT >1.0ng/ml had a sensitivity of 77% and specificity of 78% at 8 hours for diagnosis of MI

Comment(s)

A very large number of studies of similar quality were found and thus we were unable to reduce the number of best evidence papers included here on the basis of the quality of studies. As may be expected, this large number of papers indicates that there are many difficult issues to address and that no papers have been able to fully answer the question 'what cut-off point diagnoses a perioperative MI'. There are several reasons for the problems encountered by all these papers. Firstly of paramount importance when assessing the utility of a biochemical test is the ability to compare it to an established 'gold standard'. Unfortunately there is no easily applicable gold standard for the diagnosis of post-operative MI in the immediate post-operative period. In these papers, most authors use ECG criteria of newly developed Q-waves, or LBBB or ST changes assessed by cardiologists blinded to the biochemical markers. In addition some papers also diagnosed an MI on Trans-thoracic or trans-oesophageal echocardiography if wall motion abnormalities were seen. Of concern some papers used CK-MB or other rises in biochemical markers to diagnose MI, when the validity of biochemical markers is the very subject in question. Of note ECG or Echo changes would not allow the diagnosis of smaller or non-transmural infarcts to be tested. Some attempts have been made to define a subgroup of patients with "non-specific" changes of the ST segment or conduction abnormalities on ECG and correlate this to marker activity in some papers. However the 'gold standard' varied considerably amongst the papers presented here. Secondly the assay for TnI was initially non-uniform and thus some variation may have been due to the measuring techniques used. In particular the large study by Benoit et al found a markedly higher TnI level in normal patients. They have been criticised because gelatine-based colloid fluid replacement interferes with the immunoassay for TnI that they used and they did not account for this. In addition it has more recently been found that the levels of TnI in pericardial blood is higher than serum. Thus if autotransfusion is used, this will artificially raise serum TnI in otherwise normal patients. Another important issue in these papers is the absence of a correlation between graft patency and myocardial infarction. Holmvang et al performed an angiogram on all of their patients between days 5 and 7 postoperatively. They found 10 patients with an MI on ECG criteria but only 3 of these patients had vein graft occlusions. In addition 12 patients had an occlusion with only 3 of these showing evidence of myocardial infarction. Peivandi et al performed an MRI to look at graft patency and found no occluded grafts in the 4 patients diagnosed with peri-operative MI and 1 graft occlusion was found in a patient with minor myocardial injury. However despite all these shortcomings the papers do show that both TnI and TnT levels do significantly increase in patients who have a final diagnosis of peri-operative MI. Troponin T tended to be less than 1 in normal patients, however in patients who had an MI, the level was invariably over 1, usually over 3 and occasionally over 10ng/ml. Troponin I tended to be around 6ng/ml in normal patients but in patients with a final diagnosis of MI the level was well above 10-20ng/ml and often was into the high 50s.

Clinical Bottom Line

Troponin I and T can both be used to indicate myocardial damage, with the level correlating well with the level of injury. However until issues such as a 'gold standard' for peri-operative MI are addressed, one single cut-off point cannot be recommended.

References

1. Katus HA, Schoeppenthau M, Tanzeem A, Bauer HG, Saggau W, Diederich KW, Hagl S, Kuebler W. Non-invasive assessment of perioperative myocardial cell damage by circulating cardiac troponin T. Br. Heart J 1991;65(5):259-264.
2. Hake U, Schmid FX, Iversen S, Dahm M, Mayer E, Hafner G, Oelert H. Troponin T--a reliable marker of perioperative myocardial infarction? Eur. J Cardiothorac. Surg 1993;7(12):628-633.
3. Mair J, Larue C, Mair P, Balogh D, Calzolari C, Puschendorf B. Use of cardiac troponin I to diagnose perioperative myocardial infarction in coronary artery bypass grafting. Clin. Chem. 1994;40(11 pt1):2066-2070.
4. Sadony V, Korber M, Albes G, Podtschaske V, Etgen T, Trosken T, Ravens U, Scheulen ME. Cardiac troponin I plasma levels for diagnosis and quantitation of perioperative myocardial damage in patients undergoing coronary artery bypass surgery. Eur. J Cardiothorac. Surg 1998;13(1):57-65.
5. Alyanakian MA, Dehoux M, Chatel D, Seguret C, Desmonts JM, Durand G, Philip I. Cardiac troponin I in diagnosis of perioperative myocardial infarction after cardiac surgery. J Cardiothorac. Vasc. Anesth. 1998;12(3):288-294.
6. Bonnefoy E, Filley S, Kirkorian G, Guidollet J, Roriz R, Robin J, Touboul P. Troponin I, troponin T, or creatine kinase-MB to detect perioperative myocardial damage after coronary artery bypass surgery. Chest 1998;114(2):482-486.
7. Jacquet L, Noirhomme P, El Khoury G, Goenen M, Philippe M, Col J, Dion R. Cardiac troponin I as an early marker of myocardial damage after coronary bypass surgery. Eur. J Cardiothorac. Surg 1998;13(4):378-384.
8. Gensini GF, Fusi C, Conti AA et al. Cardiac troponin I and Q-wave perioperative myocardial infarction after coronary artery bypass surgery. Crit Care Med 1998;26(12):1986-1990.
9. Fellahi JL, Leger P, Philippe E, Arthaud M, Riou B, Gandjbakhch I, Coriat P. Pericardial cardiac troponin I release after coronary artery bypass grafting. Anesth. Analg. 1999;89(4):829-834.
10. Carrier M, Pellerin M, Perrault LP, Solymoss BC, Pelletier LC. Troponin levels in patients with myocardial infarction after coronary artery bypass grafting. Ann Thorac. Surg 2000;69(2):435-440.
11. Greenson N, Macoviak J, Krishnaswamy P, Morrisey R, James C, Clopton P, Fitzgerald R, Maisel AS. Usefulness of cardiac troponin I in patients undergoing open heart surgery. Am. Heart J 2001;141(3):447-455.
12. Benoit MO, Paris M, Silleran J, Fiemeyer A, Moatti N. Cardiac troponin I: its contribution to the diagnosis of perioperative myocardial infarction and various complications of cardiac surgery. Crit Care Med 2001;29(10):1880-1886.
13. Peivandi AA, Dahm M, Hake U, Hafner G, Opfermann UT, Loos AH, Tzanova I, Oelert H. H Patterns and diagnostic value of cardiac troponin I vs. troponin T and CKMB after OPCAB surgery. Thorac. Cardiovasc. Surg 2001;49(3):137-143.
14. Andres J, Stepien E, Szajna-Zych M et al. Levels of troponin I, tropoinin T, isoenzyme MB creatine kinase and myoglobins in blood serum for perioperative diagnosis of myocardial infarction in patients after coronary artery bypass graft... [Polish]. Folia Med Cracov. 2001;42(4):263-271.
15. Castro MJ, Vazquez RS, Velayos AC, Herranz VJ, Almeria VC, Iloro Mora MI. MI [Cardiac troponin I in perioperative myocardial infarction after coronary artery bypass surgery]. [Spanish]. Rev. Esp. Cardiol. 2002;55:245-250.
16. Holmvang L, Jurlander B, Rasmussen C, Thiis JJ, Grande P, Clemmensen P. Use of biochemical markers of infarction for diagnosing perioperative myocardial infarction and early graft occlusion after coronary artery bypass surgery. Chest 2002;121(1):103-111.
17. Fransen EJ, Diris JH, Maessen JG, Hermens WT, Dieijen-Visser MP. Evaluation of "new" cardiac markers for ruling out myocardial infarction after coronary artery bypass grafting. Chest 2002;122(4):1316-1321.