TAG

PR

IZE

—O

RA

LP

RE

SE

NTA

TIO

NS

Heart, Lung and Circulation S92007;16:S7–S11 TAG PRIZE—ORAL PRESENTATIONS

2. ASA and SCA Task Force on Transesophageal Echocardiog-raphy. Practice guidelines for perioperative transesophagealechocardiography. Anesthesiology 1996;84:986–1006.

doi:10.1016/j.hlc.2007.02.004

ON VERSUS OFF PUMP CORONARY ARTERY BYPASSGRAFTING: A PROSPECTIVE RANDOMISED TRIAL

P. Davis, R. Bunton, G. McCrystal, S. Valentine, S. Gold

Department of Cardiothoracic Surgery, Dunedin Hospital, NewZealand

Introduction. Within the field of coronary artery bypassgrafting (CABG) there remains uncertainty about the trueplace of off pump coronary artery bypass (OPCAB). Likeany new surgical technique it has its supporters as well asthose who are unconvinced of its benefits and believe itmay in fact be inferior to traditional surgery using cardio-pulmonary bypass (CPB). CPB has long been blamed formuch of the morbidity related to CABG including neu-rovascular side effects, an area where OPCAB was felt tooffer an advantage.

Much of the early literature relied on highly selec-tive retrospective case series and early randomised trialsplaced tight inclusion criteria again leading to a selectpatient cohort. With this trial, we aimed to reduce inclu-sion criteria to allow a more widely applicable study.PdFOa

pwCsotnoc

pnotiifssdH

aedi

respectively. There was a decreased application of pacingwires in the OPCAB group (17 versus 88%), decreased rateof AF 28% versus 44%. There was a trend towards highertotal blood loss in OPCAB, but little difference in inotroperequirements, or Intensive Care Unit and Hospital lengthof stay. Only the difference in application of pacing wiresreached statistical significance.

Neuro-cognitive testing did not reveal any significantdifference at any point of testing.

Discussion. Whilst this is a small single centre study webelieve it demonstrates the results that can be expectedin similar sized units across a broad range of patients.It shows that for in-hospital endpoints OPCAB can beequivalent to traditional CPB. It suggests that the neu-ropsychological impact of CABG using either CPB orbeating heart technique is equivalent. However, as withmany of the randomised trials within cardiac surgery andspecifically OPCAB research, recruitment of large num-bers remains difficult as does the ability to infer or drawsignificant conclusions.

Indications from our study and data from elsewhereshow an equivalent in hospital course for the two groups.Adopting this as a starting point the importance oflong-term symptom free survival data becomes cru-cial. With this in mind further study of this cohort toexamine recurrence of symptoms, anti-anginal therapy,re-hospitalisation and re-intervention rates would be giveis

d

AGTPS

Pi

irnooeoaott

tggbt

atients with poor left ventricular function, recent myocar-ial infarction and triple vessel disease were not excluded.urther to prospective randomisation of patients intoPCAB and CPB groups, a blinded neuro-cognitive

ssessment of all patients was undertaken.Methods. The study undertaken at Dunedin Public Hos-

ital, New Zealand between July 2001 and October 2003as single operator. All patients on the waiting list forABG or in hospital requiring urgent surgery were con-ented and enrolled prior to being randomised to on orff pump surgery. In all, 70 patients were randomised, 37o receive OPCAB and 33 traditional CPB. OPCAB tech-ique used medtronic octopus stabilisation, a combinationf LIMA and Saphenous vein grafts with aortic side-bitinglamp for each top end anastamosis.

Data was collected prospectively with primary end-oints of hospital mortality, conversion rates, andeurovascular events as well as secondary end pointsf bleeding rates, length of ICU and hospital stay,ime to removal from ventilation, time to removal fromnotropic support, incidence of requirement for pac-ng, and incidence of arrhythmias. Patients receivedull neuro-cognitive testing pre-operatively, as well asix days, six weeks and 30 weeks post-operatively. Thetudy pre-operative, intra-operative and post-operativeata collection was undertaken at Dunedin Publicospital.Results. Two equivalent randomised groups, 37 OPCAB

nd 33 CPB with slightly higher rates of MI and decreasedjection fraction in the OPCAB group. There were noeaths in the study, a conversion rate of 5%, and an

ncidence of stroke of 5.4 and 3% for OPCAB and CPB,

nsight into the long-term results with regards to OPCABurgery.

oi:10.1016/j.hlc.2007.02.005

NGIOGRAPHIC PATENCY OF CORONARY ARTERYRAFTS: 5-YEAR OBSERVATIONAL DATA FROMHE RANDOMISED CONTROLLED ‘RADIAL ARTERYATENCY AND CLINICAL OUTCOMES’ (RAPCO)TUDY

hilip Hayward, Brian Buxton on behalf of the RAPCOnvestigators

Austin Hospital, Heidelberg, Vic., Australia

Introduction. The RAPCO study has enrolled patientsnto two trials to compare the radial artery with the freeight internal thoracic artery (if age < 70) or with saphe-ous vein (if age > 70) when grafted on to the largest targetther than the left anterior descending artery. Patencyutcomes in RAPCO will focus only on the trial graft inach case. However, post operative angiography from thisngoing study has reassessed all coronary grafts, both trialnd non-trial: together these offer a valuable observationalverview of likely graft patency in the current era, in con-rast to symptom directed studies which may overestimatehe incidence of graft failure.

Methods. In situ internal thoracic arteries were anas-omosed to the LAD, the second target received the trialraft and other grafts were at the discretion of the sur-eon. Protocol directed percutaneous angiography haseen performed at intervals weighted towards the end of

he study period, or earlier when directed by symptoms.

TAG

PR

IZE

—O

RA

LP

RE

SE

NTA

TIO

NS

S10 Heart, Lung and CirculationTAG PRIZE—ORAL PRESENTATIONS 2007;16:S7–S11

Figure 1. Study angiogram all graft patency.

184 patients have been studied to date as per protocol, attime points up to 10 years (mean 3.9 years), among whomthere are 430 non-trial grafts in addition to the 184 trialgrafts. Supplementary angiography (percutaneous or CT)was offered at a fixed five-year time point post operatively:193 trial grafts and 469 non-trial grafts have been imagedat this point. Angiography is reviewed by three investiga-tors blinded to each others’ reporting; grafts after definedas angiographically successful if patent from proximal todistal anastomosis inclusive. Failed grafts include >80%stenosis, string sign or occlusion.

Results. In protocol directed angiography (Fig. 1),overall patency of the in situ internal thoracic arterygrafts and free arterial grafts was 95.5% (210/220) and91.4%% (150/164) respectively which did not differ signif-icantly (p = 0.13). Saphenous vein graft patency was 83.0%(191/230) with a trend to higher failure rate compared withfree arterial grafts (p = 0.07). At the optional five-year timepoint, outcomes are similar to the protocol directed find-ings (Fig. 2), with patency of in situ left and right internalthoracic artery grafts of 95.8% (204/213) which was signif-icantly greater than that of the other two groups (p = 0.02and <0.001). Patency of free arterial conduits and saphe-nous vein grafts was 89.1% (139/156) and 82.4% (201/244)respectively, which did not differ significantly (p = 0.09) atfive years.

Discussion. Although trending to superior patency

tify for now our current practice of maximal use of in situconduits supplemented by free radial grafts.

doi:10.1016/j.hlc.2007.02.006

CEREBRAL MICROEMBOLI AND NEUROBEHAV-IORAL IMPAIRMENT AFTER PROSTHETIC AORTICVALVE REPLACEMENT: A TRANSCRANIAL DOPPLERULTRASOUND STUDY

Gregory D. Rice, Robert A. Baker, Rebecca Perry, John L.Knight

Flinders Medical Centre, Adelaide, SA, Australia

Introduction. Mechanical heart valves producegaseous microemboli via cavitation and degassing.These microemboli can be detected in the cerebralcirculation, using transcranial Doppler ultrasound, ashigh-intensity transient signals (HITS). The rate ofembolus production appears to relate to valve design.Previous research has shown that these microembolido not cause overt neurological injury, but recent evi-dence suggests that they may produce cognitive and/orneuropsychological deficits.

The present study aimed to: (1) demonstrate the differ-ent rates of embolus production in patients with bileafletmechanical valves, tilting disk mechanical valves, andbioprosthetic valves; and (2) correlate the rate of produc-

compared with vein grafts, free arterial conduits have notmatched the gold standard achieved by in situ internalthoracic arteries at trial midpoint. Until the results fromRAPCO are available, which will test types of free arte-rial graft against each other and will compare radial andvein grafts in the longer term, these observational data jus-

Figure 2. 5 year elective angiography graft patency.

tion of microemboli with the presence of neurobehavioralimpairment as measured by a computerised neuropsycho-logical test battery.

Methods. All patients aged over 18 years, living inmetropolitan Adelaide, who had undergone isolated aor-tic valve replacement at Flinders Medical Centre wereeligible for inclusion in the study. Exclusion criteria were:a history of coronary artery disease, stroke with resid-ual deficit, or major neurological disorder; significant(>70%) stenosis of either common or internal carotid arteryon duplex ultrasound examination; non-therapeutic INR;or primary language other than English. All subjectsunderwent transthoracic echocardiography, standardisedneurological examination, a 30-min transcranial Dopplerultrasound (TCD) examination, and neuropsychologicaltesting using the CNS Vital SignsTM computerised testbattery.

Results. This study is still in progress. Subjects havebeen divided into four groups by valve type, but valvetypes have not been identified at present and statisticalanalysis has not yet been performed. Fifty-two subjectshave been successfully enrolled (45 males, 7 females; 26in group 1, 5 in group 2, 3 in group 3, and 8 in group 4,plus 10 control patients with bioprosthetic valves). Com-plete testing has been performed in 40 subjects to date.Median HITS counts were 8.1/h (range 0–106) for group 1,448.0/h (range 276–702) for group 2, 526.0/h (range 112–940)for group 3, and 4.0/h (range 0–10) for group 4. None ofthe control patients demonstrated HITS during the TCDexamination. Subjects were categorised into high-HITS(>20/h) and low-HITS (0–20/h) for comparison of neurobe-havioral outcomes. Results of neuropsychological testing