Heart 1997;77:314-318

Ischaemic preconditioning reduces troponin Trelease in patients undergoing coronary arterybypass surgery

David P Jenkins, Wilf B Pugsley, Abdul M Alkhulaifi, Michael Kemp, James Hooper,DerekM Yellon

AbstractObjective-To investigate whether ischae-mic preconditioning could reduce myo-cardial injury, as manifest by troponin Trelease, in patients undergoing electivecoronary artery bypass surgery.Design-Randomised controlled trial.Setting-Cardiothoracic unit of a tertiarycare centre.Patients-Patients with three vessel coro-nary artery disease and stable anginaadmitted for first time elective coronaryartery bypass surgery were invited to takepart in the study; 33 patients were ran-domised into control or preconditioninggroups.Intervention-Patients in the precondi-tioning group were exposed to two addi-tional three minute periods of myocardialischaemia at the beginning of the revas-cularisation operation, before the isch-aemic period used for the first coronaryartery bypass graft distal anastomosis.Main outcome measure-Serum troponinT concentration at 72 hours after cardio-pulmonary bypass.Results-The troponin T assays were per-formed by blinded observers at a differenthospital. All patients had undetectableserum troponin T (< 0a1 ugIl) beforecardiopulmonary bypass, and troponin Twas raised postoperatively in all patients.At 72 hours, serum troponin T was lower(P = 0.05) in the preconditioned group(median 0 3 ugIl) than in the controlgroup (median 144ugl).Conclusions-The direct application of apreconditioning stimulus in clinical prac-tice has been shown, for the first time, toprotect patients against irreversiblemyocyte injury.

(Heart 1997;77:314-318)

Keywords: myocardial ischaemia; coronary arterybypass grafts; ischaemic preconditioning; troponin T

It is established that exposing the myocardiumto brief periods of ischaemia and reperfusioninduces greater tolerance to a subsequent moreprolonged ischaemic insult. This endogenousadaptation to ischaemia, termed "ischaemicpreconditioning", was initially shown to delaymyocardial necrosis in an in vivo canine modelof myocardial infarction.' Since then an exten-sive body of reports on preconditioning hasaccumulated, the phenomenon has been char-

acterised, and its mechanisms explored, butdespite increasing evidence for the existence ofpreconditioning in human myocardium,2 fewclinical studies have been performed.

It is our hypothesis that the protectiveeffects of preconditioning can be induced inpatients requiring cardiac surgery. The majorityof cardiac surgical procedures involve theintentional interruption of coronary arteryblood flow and myocardial ischaemia is there-fore inevitable. Although techniques ofmyocardial protection during cardiac surgeryhave improved in the past 30 years, patientswith more severe disease are now being offeredsurgery, and any additional treatment whichattenuates myocardial injury should be investi-gated. It has been shown that ischaemic pre-conditioning may be induced in patientsundergoing coronary artery bypass surgery.3 Inthe latter study an ischaemic preconditioningprotocol instituted before the longer ischaemicperiod (of 10 minutes) for fashioning the firstsaphenous vein to coronary artery anastomosisresulted in relative preservation of myocardialATP levels at the end of the 10 minutes ofischaemia. The study was limited to the meta-bolic changes occurring at a single time point,but the similarity between the results inhumans and the metabolic results from animalmodels4 indicated for the first time thatischaemic preconditioning could be induced inhuman myocardium. However, the initial clin-ical study was not designed to evaluatemyocardial protection throughout the wholecourse of the operation.One of the difficulties in assessing interven-

tions designed to protect the myocardium inpatients undergoing cardiac surgery is theavailability of end points with which to quan-tify ischaemic injury. Troponin T is a contrac-tile apparatus regulatory protein that is onlydetectable in serum following severe ischaemicinjury.5 Measurement of cardiac troponin Tcan detect minor degrees of myocyte necrosiswhich had previously not been recognisedusing conventional cardiac enzymes and ECGtests.6 Troponin T has been shown to increasesignificantly in all patients recovering normallyafter uncomplicated cardiac surgery,78 and theamount of release correlates with theischaemic time.9 10 Therefore measurement oftroponin T release into serum may be anappropriate marker of ischaemic injury follow-ing cardiac surgery and an effective means ofassessing and comparing cardioprotectiveinterventions.

In this study we investigated whetherischaemic preconditioning reduced myocardial

University CollegeLondon Hospitals andMedical School,University CollegeHospital, LondonWC1, UnitedKingdom: The HatterInstitute forCardiovascularStudiesD P JenkinsD M YellonDepartment ofCardiothoracicSurgeryW B PugsleyA M AlkhulaifiRoyal BromptonHospital, London,United Kingdom:Department of ClinicalBiochemistryM KempJ HooperCorrespondence to:Professor D M Yellon, TheHatter Institute, Departmentof Academic and ClinicalCardiology, UniversityCollege Hospital, GraftonWay, London WC1E 6DB,United Kingdom.Accepted for publication10 January 1997

314

on March 24, 2022 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.77.4.314 on 1 A

pril 1997. Dow

nloaded from

Ischaemic preconditioning reduces troponin T release in patients undergoing coronary artery bypass surgery

PreconditioningF

3 2 3 2-4 10 10 8-12 6-8 8-12 10min-*l_ _ _Preconditioned

Paced at90 First graft Second graft Third graft

beats/minControl periodF 10 10 8-12 6-8 8-12 10min

l~

monary bypass techniques were standardised.The coronary artery bypass grafts were per-formed using the technique of intermittentischaemic arrest with fibrillation for the distalvein to coronary artery anastomosis, and theheart reperfused and beating for the proximalvein to aorta anastomosis. Ventricular ventswere not used. Whole body temperature wasmaintained at 36 ± 1°C for the period of thepreconditioning protocol and first distal anas-tomosis in both groups; following this periodall patients were cooled to 32°C.

A B C DVentricular biopsies

4 Cardiopulmonary bypass

Protocolfor operative management in control and preconditioned groups. Filled bo.indicate periods of ischaemia and empty boxes indicate periods of reperfusion. Temkept constant at 36°C during the preconditioning protocol and the first graft in bot,

injury in patients undergoing routine c

artery bypass surgery by measuring theof troponin T into serum during the peative recovery. We also measured AT]in ventricular muscle samples takenoccasions during both the ischaemreperfusion periods of the operation.

MethodsThe investigation was approved by tiethics committee and all patients enterthe study gave informed written cPatients with three vessel coronary artiease and stable angina admitted for filelective coronary artery bypass surge]invited to take part in the study. Thirnpatients were randomised into contrc16) or preconditioning groups (nbetween March and October 1995. ]with unstable angina, left venaneurysm, or very poor left ventriculation (ejection fraction < 30%), valve (

and those taking sulphonylurea anti(drugs were not eligible for inclusion.

SURGICAL TECHNIQUEAll operations were performed by a sinssultant surgeon (WBP) at the MiHospital and the anaesthetic and car

Table 1 Patient demographics and operation data

IschaemicControl preconditioning

Number of patients 16 17Age, years 62 (2) 57 (2)Sex 15 male: 1 female 15 male: 2 femPrevious MI 6 9Number of grafts 3-2 (0-1) 3-1 (0-2)LIMA usage 16 16CPB time, min 96 (4) 96 (4)Ischaemia time, min 34-1 (1 8) 33-3 (1-6)Defibrillation energy, joules 58 (11) 58 (11)

Values are means (SEM), no significant differences between groups. CPB, cardiopbypass; LIMA, left internal mammary artery; MI, myocardial infarction.

PRECONDITIONING PROTOCOLPatients randomised to ischaemic precondi-

Control tioning were pretreated with the same precon-ditioning protocol as used in our previous

t study3 (figure). After instituting cardiopul-E monary bypass, two three-minute periods of

ischaemia were applied by cross clamping the> aorta, each separated by two minutes of reper-

fusion. During this 10 minute period, heartsxes were paced at 90 beats/min and myocardialrperature and whole body temperature was maintained,h groups.

at 36 ± 1°C. Patients in the control group alsoreceived 10 minutes of normothermic car-

oronary diopulmonary bypass (without the precondi-release tioning protocol) before the first anastomosis.)stoper-P levels TROPONIN T ASSAYon five Blood samples for troponin T assay were takenLia and immediately before cardiopulmonary bypass,

one hour after bypass, and at six, 24, and 72hours. Blood was collected into plain tubesand the serum separated by centrifugationwithin one hour of collection. The serum was

le local frozen and stored at -20°C until analysis.red into Cardiac troponin T was measured in aonsent. blinded fashion using an enzyme linked.ery dis- immunosorbent assay (ELISA) at the Royalrst time Brompton Hospital. A commercially availablery were standard assay kit (ELISA troponin-T,ty three Boehringer Mannheim) and batch ELISA)I (n = analyser (Enzymun test system ES 300,i = 17) Boehringer Mannheim) were used. The assayPatients detected cardiac troponin T in the rangeitricular 0-1-18,ug/1l.Lr func-disease, VENTRICULAR BIOPSIES AND ATP ASSAYdiabetic Samples of left ventricular muscle were

obtained with a "Trucut" biopsy needle(Baxter) from the territory of the left anteriordescending coronary artery. Biopsies were

gle con- taken at the following time points: (A) base-iddlesex line, before the 10 minute preconditioningrdiopul- protocol; (B) following the preconditioning

protocol; (C) at 10 minutes of ischaemia, atthe end of the first distal anastomosis; (D)after 10 minutes of reperfusion; (E) following10 minutes of reperfusion after completion ofthe final anastomosis and before discontinuingcardiopulmonary bypass at the end of theoperation (figure).

isle Samples were immediately frozen in liquidnitrogen and then freeze dried for at least 12hours. Samples were assayed in a blinded fash-ion in random order. After accurately weigh-ing each freeze dried sample, protein was

pulmonary extracted by homogenisation with 6% perchlo-ric acid. ATP content was determined using

315

on March 24, 2022 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.77.4.314 on 1 A

pril 1997. Dow

nloaded from

_6enkins, Pugsley, Alkhulaifi, Kemp, Hooper, Yellon

Table 2 Troponin T in serum

TnT,ugll Pre CPB I h post CPB 6 h post CPB 24 h post CPB 72 h post CPB

Control < 0-1 1-0 (0 4 to 1-5) 1-8 (0-8 to 3-5) 1-4 (0 5 to 2-2) 1-4 (0 7 to 3-0)IPC < 0.1 1-0 (0-5 to 1-4) 1. 1 (0 5 to 3 3) 0 4 (0-3 to 1-7) 0 3 (0-2 to 2 0)*

Median and (interquartile range) of serum troponin T.CPB, cardiopulmonary bypass; IPC, ischaemic preconditioning; TnT, troponin T.*P = 0 05, Mann-Whitney U test.

an enzymatic assay and observing changes inoptical density of the extracts with a spec-trophotometer. ATP content of the biopsieswas expressed in ,umol/g dry weight.

ELECTROCARDIOGRAPHIC CHANGESTwelve-lead electrocardiograms were re-corded preoperatively, following surgery onreturn to the intensive care unit, and on thefirst and fourth postoperative days. Peri-operative transmural infarction was defined onelectrocardiographic criteria as the appearanceof new persistent Q waves (one third QRSheight and > 0 04 s duration). Other changeswere noted (ST segment, T wave and reduc-tion in R wave height of > 25%) if they per-sisted in two or more adjacent leads.Postoperative arrhythmias were also recorded.

STATISTICSDemographic, operative, and ATP data arepresented as mean (SEM). Differences withinand between groups were analysed with apaired or unpaired t test as appropriate. Thetroponin T levels at 72 hours after bypass waspreselected as the major end point. TroponinT results are presented as medians (withinterquartile range) and between group differ-ences were analysed by a non-parametric test(Mann-Whitney U) because of the non-Gaussian distribution of the data. Comparisonof proportions was performed using the X2with Yates correction for small sample size(X2y). Statistical significance was defined as aP value of 0-05.

ResultsPATIENTSThree patients who were eligible for inclusionin the study refused to consent. The demo-graphic and operative data of the 33 patientscompleting the study are presented in table 1.There were no significant differences betweenthe groups. No patients required inotropic orintra-aortic balloon support postoperatively.One patient (from the control group) devel-oped a perforated peptic ulcer on the fifthpostoperative day. This patient developed sep-ticaemia and eventually died of multiple organ

Table 3 ATP content (,umol/g) of myocardial biopsies

Biopsy A Biopsy B Biopsy C Biopsy D Biopsy E

Control 19 (1-2) 18 (1-0) 13 (1-0) 17 (1-0) 15 (1-1)IPC 21 (1-0) 17 (1-4)* 14 (0 9) 17 (1-6) 17 (1-3)Values are means (SEM).IPC, ischaemic preconditioned. Biopsy A, baseline; B, after preconditioning protocol; C, at 10minutes of ischaemia following first anastomosis; D, after 10 minutes of reperfusion; E, follow-ing 10 minutes of reperfusion after completion of final anastomosis.*P = 0 016 compared with IPC biopsy A. No significant differences between groups.

failure. There were no other major complica-tions.

TROPONIN T

Troponin T concentrations were below thedetectable range of the assay (< 0 1 ,ug/l) in allpatients before cardiopulmonary bypass.There was a rise in serum troponin T in allpatients postoperatively, indicating somemyocyte injury during the operation (table 2).Peak troponin T release occurred at six hoursafter completion of cardiopulmonary bypass inboth groups: preconditioned group, 1-1 ug/l;control group, 1.8 jug/l. At 72 hours troponinT was lower (P = 0-05) in preconditionedpatients (0-3,ug/1) than in the control group(1 4 jg/l). Ten patients in the preconditionedgroup had troponin T values of < 0 5 jg/l atthis time compared with only three patients inthe control group (X2y = 5*54, P = 0 04). Inthe control group 10 patients had troponin Tvalues of > 1 0 ug/l at 72 hours, but in the pre-conditioned group only five patients had levelsabove 1.0 ug/l (x2y = 3-64, P = 0 12).

ATPATP content in ventricular biopsies is shownin table 3. Baseline biopsies (A) were not sig-nificantly different between the groups. In thepreconditioned group there was a decline (P= 0-016) in ATP after the preconditioningprotocol (biopsy A to biopsy B) before theischaemic period of the first anastomosis. Thedecline in ATP during the ischaemia of thefirst anastomosis (biopsy B to biopsy C) was28% in control hearts and 18% in precondi-tioned hearts. However, there was no signifi-cant difference in ATP content betweengroups at any stage. The recovery in ATP onreperfusion following the first ischaemicperiod (biopsy D) was similar to that followingthe third ischaemic period (biopsy E) in bothgroups.

ELECTROCARDIOGRAPHYNo patient developed new Q waves on theECG postoperatively. A reduction in R waveheight occurred in five patients postoperatively(two from the control and three from the pre-conditioned group). ST segment changes werepresent in three patients postoperatively (onecontrol and two preconditioned). Six patientsdeveloped atrial fibrillation during the first fivepostoperative days (four control and two pre-conditioned). These differences did not reachstatistical significance.

DiscussionIn patients with ischaemic heart disease under-going myocardial revascularisation by coro-

316

on March 24, 2022 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.77.4.314 on 1 A

pril 1997. Dow

nloaded from

Ischaemic preconditioning reduces troponin T release in patients undergoing coronary artery bypass surgery

nary artery bypass grafting, ischaemic precon-ditioning reduced perioperative myocardialnecrosis, as manifest by cardiac troponin Trelease into serum. This is the first confirma-tion that ischaemic preconditioning maydirectly delay myocardial necrosis in humans.

In this study patients in the control and pre-conditioned groups had very similar operativeischaemia times (34 and 33 minutes, respec-tively) and troponin T concentration increasedin all patients. However, 72 hours after theoperation patients in the preconditioned groupwere releasing less cardiac troponin T intoserum.The subcellular compartmentation of tro-

ponin T is reflected by its release kinetics fol-lowing ischaemia/reperfusion injury. A smallunbound cytoplasmic pool accounts for theearly peak in troponin T; the later sustainedrelease of troponin T, over the 24 hours fol-lowing ischaemia, reflects washout of struc-turally bound protein from continuingdegradation of myofibrils in irreversiblyinjured cells.1' 12 The threshold for a positivetroponin T release was set at 0-5 pg/l in arecent report investigating troponin T in braindead organ donors"3 and all patients sustaininga Q wave or non-Q wave myocardial infarctionhad values > 1 pg/l in the original report ofdiagnostic efficiency.'3 In the present study nopatients showed measurable troponin T inserum (< 0 1 pg/l, the detection limit of theassay) before cardiopulmonary bypass, and 72hours postoperatively serum troponin T con-centrations were < 0-5 yg/l in 59% of precon-ditioned patients but > 1 yg/l in 63% of thecontrol group.

Troponin T is not present in the serum ofnormal individuals and unlike creatine kinasewas not detectable in serum followingorthopaedic and pulmonary surgery.9 All stud-ies measuring troponin T release after cardiacsurgery, including those employing cardiople-gia for myocardial protection, have reportedsignificant increases, even in patients who havean apparently uncomplicated postoperativecourse.7-10 1416 Serum troponin T values at24-72 hours after operation in the latter studieswere similar to the results obtained in the con-trol group reported here. Patients sustainingmajor perioperative myocardial infarction withdevelopment of new Q waves on the ECGform a clearly distinct group with very hightroponin T levels and a worse prognosis. 9The prognostic implications of moderate tro-ponin T elevation are not known, but mea-surement of troponin T has shown thatpreviously unrecognised myocardial damageoccurs in all patients during cardiac surgery.This probably reflects diffuse tissue necrosisscattered throughout the myocardium, whichwould not be manifest as specific ECGchanges. This concept is corroborated by thefinding that troponin T is also detectable(median 05 ,ug/1) in a subgroup of patientswith unstable angina in whom it indicated aworse prognosis-and this was thought to bethe result of localised myocyte necrosis causedby thrombotic microembolisation.'7 Takentogether, these results suggest that the human

myocardium is perhaps more vulnerable toirreversible ischaemic injury than previouslyassumed when less precise markers of myocar-dial injury were available.The above findings emphasise the scope for

improved myocardial protection during car-diac surgery. Although our study was per-formed in patients with stable angina andmoderately good left ventricular function, wewould expect that myocyte injury would begreater in patients with more unstable diseaseand longer ischaemic times. The present studywas too small for morbidity or mortality to beused as end points and most patients hada completely uncomplicated postoperativecourse. Indeed, the clinical significance of therise in troponin T after cardiac surgery is notyet understood, but since continuing release oftroponin T at 72 hours indicates irreversiblemyocyte injury, and myocytes cannot bereplaced, it must have biological significance.

Analysis of the ATP content in samples ofventricular muscle showed the expected deple-tion of ATP during the 10 minute ischaemicperiod of the first distal anastomosis (biopsyC) in both groups. There was equal recoveryin ATP content following the 10 minute reper-fusion for the proximal anastomosis (biopsyD). The ATP content following reperfusion atthe end of the third period of ischaemia(biopsy E) was not different from that followingreperfusion after the first ischaemic period(biopsy D). This indicates that there is nocumulative depletion in ATP with successive10 minute ischaemic challenges in the humanheart and corroborates the observations in thecanine model.'8 Although there was a greaterdecline in ATP content during the ischaemicperiod of the first graft (biopsy B to biopsy C)in control hearts, there was no significant dif-ference in ATP content between control andpreconditioned groups at the end of thisperiod (biopsy C). This finding is in contrastto our previous smaller study, which hadshown a relative preservation in ATP in pre-conditioned hearts at the end of the firstischaemic period (biopsy C).3 Indeed, in theabsence of a measure of myocyte necrosis, thesimilarity between the ATP results in our pre-vious study and those reported by Jennings inpreconditioned dogs4 had given us the confi-dence to believe that it was possible to precon-dition the human heart during cardiac surgery.

There are two potential reasons for the fail-ure to observe differences in ATP contentbetween control and preconditioned groups inthe current study. (1) It is possible that pre-conditioning does not necessarily result inpreservation of ATP, and that in this largerseries of less selected patients no real differ-ence in ATP content exists. Several recentexperiments with isolated rat hearts, in whichthe time course of ATP changes has been fol-lowed in control and preconditioned hearts bynuclear magnetic resonance, have reportedthat protection from preconditioning occurs inthe absence of significant ATP preserva-tion.'9 20 (2) In spite of the latter findings thereare many reports suggesting that precondition-ing is at least associated with a relative preser-

317

on March 24, 2022 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.77.4.314 on 1 A

pril 1997. Dow

nloaded from

38enkins, Pugsley, Alkhulaifi, Kemp, Hooper, Yellon

vation of myocardial ATP content in largeranimal models during the first 20 minutes of

42122sustained ischaemia, even if the ATPchanges have not been proven to be the causeof the delay in infarction. The differencebetween these studies and ours in humans isthat in the animal models it was possible tobiopsy the myocardium for ATP analysis moreoften and plot the ATP changes in bothgroups over time. These curves do show aslower depletion in ATP during ischaemia inpreconditioned hearts, so that the ATP con-tent is transiently higher than in controls atabout 10 minutes of ischaemia. However, withthe limited number of samples possible in ahuman study, a single biopsy at 10 minutes ofischaemia (biopsy C) may have missed thistransient difference. In this study it was notpossible to measure other high energy phos-phates and metabolites accurately or calculatethe energetic charge on the small tissue sam-ples available. It is unlikely that the availabilityof the latter data would alter our conclusionsbecause in laboratory experiments with largeanimal models, the most notable differencebetween control and preconditioned groups isusually apparent in the ATP content data.The fact that there was no difference in

ATP content between the groups at the end ofthe operation does not contradict the troponinT results. Although ATP is the cellular energysource and it is rational to assume that more isbetter than less, it is not possible to correlateATP or other metabolite content with cell via-bility, and the concept of a "critical" level ofATP below which cell death occurs is nowknown to be incorrect.23

In summary, we have shown that in patientstreated by coronary artery bypass grafting,ischaemic preconditioning results in less peri-operative myocardial necrosis as determinedby serum concentrations of cardiac troponin Tpostoperatively. We believe this is the firsttime that preconditioning has been shown tooffer patients some protection against irre-versible myocyte injury associated with a thera-peutic procedure in clinical practice. Theseresults indicate that the direct application of apreconditioning stimulus at the beginning ofan operation could result in better myocardialprotection. It had been argued that the tech-nique of intermittent ischaemic arrest in theperformance of the bypass grafts or in theinstitution of cardiopulmonary bypass itself4may induce preconditioning; the improvedmyocardial protection in the preconditionedhearts compared with control hearts reportedhere suggests that these assumptions wereincorrect. Although we are not recommendingthat patients should be subjected to additionalischaemia in the course of routine cardiacsurgery, this observation highlights the poten-tial of exploiting endogenous myocardialadaptation. If the adenosine A, receptor ago-nists and ATP dependent potassium channelopeners, which initiate preconditioning in lab-oratory models,2526 are shown to be as effectivein forthcoming clinical trials, then precondi-tioning might become a practical adjunct tomyocardial protection during cardiac surgery.

We are indebted to the assistance and patience of the followingpeople without whom the study would not have been possible:Dr Hulf, Dr O'Brien, and their anaesthetic colleagues; the the-atre sisters and the perfusionists at the Middlesex Hospital.DPJ was supported by a grant from the Middlesex HospitalSpecial Trustees through the North East Thames locally organ-ised research scheme. We thank the British Heart Foundationand the Hatter Foundation for the continuing support of theInstitute.

1 Murry CE, Jennings RB, Reimer KA. Preconditioning withischaemia: a delay of lethal cell injury in ischaemicmyocardium. Circulation 1986;74: 1124-36.

2 Kloner RA, Yellon DM. Does ischaemic preconditioningoccur in patients? JAm Coll Cardiol 1994;24: 1133-42.

3 Yellon DM, Alkhulaifi AM, Pugsley WB. Preconditioningthe human myocardium. Lancet 1993;342:276-7.

4 Murry CE, Richard VJ, Reimer KA, Jennings RB.Ischaemic preconditioning slows energy metabolism anddelays ultrastructural damage during a sustainedischaemic episode. Circ Res 1990;66:913-31.

5 Katus HA, Remppis A, Neumann FJ, Scheffold T,Diederich KW, Vinar G, et al. Diagnostic efficiency oftroponin T measurements in acute myocardial infarction.Circulation 199 1;83:902-12.

6 Mair J, Dienstl F, Puschendorf B. Cardiac troponin T inthe diagnosis of myocardial injury. Crit Rev Clin Lab Sci1992;29:31-57.

7 Mair J, Wieser C, Seibt I, et al. Troponin T to diagnosemyocardial infarction in bypass surgery [letter]. Lancet1991;337:434-5.

8 Hake U, Schmid FX, Iversen S, Dahm M, Mayer E,Hafner G, et al. Troponin T-a reliable marker of periop-erative myocardial infarction? Eur _J Cardiothorac Surg1993;7:628-33.

9 Katus HA, Schoeppenthau M, Tanzeem A, Bauer HG,Saggau W, Diederich KW, et al. Non-invasive assessmentof perioperative myocardial cell damage by circulatingcardiac troponin T. Br HeartrJ 1991;65:259-64.

10 Kallner G, Lindblom D, Forssell G, Kallner A. Myocardialrelease of troponin T after coronary bypass surgery.Scand7 Thor Cardiovasc Surg 1994;28:67-72.

11 Katus HA, Remppis A, Scheffold T, Diederich KW,Kubler W. Intracellular compartment of cardiac troponinT and its release kinetics in patients with reperfused andnonreperfused myocardial infarction. Am JI Cardiol 1991;67:1360-7.

12 Remppis A, Scheffold T, Greten J, Haass M, Greten T,Kubler W, et al. Intracellular compartmentation of tro-ponin T: release kinetics after global ischaemia and cal-cium paradox in the isolated perfused rat heart. Jf MolCell Cardiol 1995;27:793-803.

13 Riou B, Dreux S, Roche S, Arthaud M, Goarin J-P, Leger P,et al. Circulating cardiac troponin T in potential hearttransplant donors. Circulation 1995;92:409-14.

14 Taggart DP, Bhusari S, Hooper J, Kemp M, Magee P,Wright JE, et al. Intermittent ischaemic arrest and cardio-plegia in coronary artery surgery: coming full circle? BrHeart _J 1994;72:136-9.

15 Uchino T, Belboul A, Roberts D, Jagenburg R.Measurement of myosin light chain I and troponin T asmarkers of myocardial damage after cardiac surgery. JCardiovasc Surg 1994;35:201-6.

16 Anderson JR, Hossein-Nia M, Kallis P, Pye M, Holt DW,Murday AJ, et al. Comparison of two strategies formyocardial management during coronary artery opera-tions. Ann Thorac Surg 1994;58:768-72.

17 Hamm CW, Ravkilde J, Gerhardt W, Jorgensen P, PeheimE, Ljungdahl L, et al. The prognostic value of serum tro-ponin T in unstable angina. N Engl J Med 1992;327:146-50.

18 Reimer KA, Murry CE, Yamasawa I, Hill ML, JenningsRB. Four brief periods of ischaemia cause no cumulativeATP loss or necrosis. Am_JPhysiol 1986;251:H1306-15.

19 Steenbergen C, Perlman ME, London RE, Murphy E.Mechanism of preconditioning. Ionic alterations. Circ Res1993;72:112-25.

20 de Albuquerque CP, Gerstenblith G, Weiss RG.Importance of metabolic inhibition and cellular pH inmediating preconditioning contractile and metaboliceffects in rat hearts. Circ Res 1994;74:139-50.

21 Kida M, Fujiwara H, Ishida M, Kawai C, Ohura M, MiuraI, et al. Ischaemic preconditioning preserves creatinephosphate and intracellular pH. Circulation 1991;84:2495-503.

22 Jennings RB, Murry CE, Reimer KA. Energy metabolismin preconditioned and control myocardium: effect oftotal ischaemia.3 Mol Cel Cardiol 199 1;23: 1449-58.

23 Opie LH. Cardiac metabolism-emergence, decline, andresurgence. Part 2. Cardiovasc Res 1992;26:817-30.

24 Burns PG, Krukenkamp IB, Caldarone CA, Gaudette GR,Bukhari EA, Levitsky S. Does cardiopulmonary bypassalone elicit myoprotective preconditioning? Circulation1995;92(suppl II):JI447-5 1.

25 Walker DM, Walker JM, Pattison CW, Pugsley WB,Yellon DM. Preconditioning in isolated superfusedhuman muscle. J Mol Cell Cardiol 1995;27:1349-57.

26 Speechly-Dick ME, Grover GJ, Yellon DM. Doesischaemic preconditioning in the human involve proteinkinase C and the ATP-dependent K channel? Circ Res1995;77: 1030-5.

318

on March 24, 2022 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.77.4.314 on 1 A

pril 1997. Dow

nloaded from