those who have known coronary disease and severe left ventricular dysfunction (LVEF less than 0.35) with breathlessness but only mild angina. Viability testing is redundant in patients who have unstable angina, postinfarction angina, or chronic stable angina, because revasculariza- tion is indicated for relief of symptoms. For ische- mic patients who have heart failure symptoms but minimal angina, the combination of good target vessels with more than 25% myocardial viability suggests potential to benefit from CABG [46]. For those who have less than 25% viability or poor target vessels, or are reoperative candidates, surgery is unlikely to provide benefit. Observational studies have documented sub- stantial improvements in LV regional and global function following revascularization of hibernat- ing myocardium [47,48]. This procedure provides relief from heart failure symptoms, improved sur- vival, and important quality-of-life benefit. A comprehensive meta-analysis of largely retrospec- tive data has been performed by Allman and col- leagues [49]. In 24 studies reporting patient survival after viability testing, annual death rates were analyzed to produce a risk-adjusted relation- ship between the severity of LV dysfunction, pres- ence of viability, and survival benefit associated with revascularization. The study included 3088 patients who had LVEF 32% Æ 8% followed for 25 Æ 10 months. For patients who had stunned or hibernating myocardium revasculari- zation was associated with a 79.6% reduction in the annual mortality rate (16% versus 3.2%, P ! .0001) compared with medical treatment. In contrast, patients who did not have viability had intermediate mortality, tending to higher rates with CABG versus medical therapy (7.7% versus 6.2%). Patients who had viability showed a direct relationship between the severity of LV dysfunction and magnitude of benefit with revas- cularization. No benefit was associated with revascularization in patients who did not have vi- ability at any level of LVEF. The perioperative mortality rates were impressively high in patients who did not have viability (around 10%) but negligible in those who had viability. This differ- ence in mortality is curious given the time frame of improvement in wall motion following revascularization. Histopathologic studies have indicated a grad- ual increase in ultrastructural damage and degree of myocardial fibrosis as the ischemic process progresses through stunning, hibernation, non- transmural scar, and full-thickness scar [47]. Obviously scar never improves in function after revascularization even if covered with a veneer of healthy epicardial muscle following thromboly- sis. Bax and colleagues [50] prospectively studied patients who had ischemic cardiomyopathy and left ventricular dysfunction using preoperative as- sessment of regional perfusion, glucose use, and contractile function. Of the dysfunctional seg- ments, 22% were stunned, 23% were hibernating, and 55% were scar tissue. In stunned myocardium contractile function improved significantly at 3 months but without further improvement at 14 months. Some 30% of stunned segments did not improve. In hibernating segments 31% had improved by 3 months and 61% had recovered fully by 14 months. In a similar study using intra- operative myocardial biopsy from dysfunctional myocardium, Haas showed only 31% of stunned segments and 18% of hibernating segments to ob- tain complete functional recovery after 1 year [47]. Failure to improve was associated with more se- vere degenerative changes in the myocyte, includ- ing depletion of sarcomeres, accumulation of glycogen, loss of sarcoplasmic reticulum, and cel- lular sequestration. Using gadolinium-enhanced contrast MRI, Kim and colleagues [21] showed that 78% of dysfunctional segments identified as completely viable showed improvement in con- tractility after revascularization. In contrast, 90% of segments with 50% to 75% of wall thick- ness scar did not improve after revascularization. The realization that the globular ischemic cardiomyopathy ventricle can be surgically re- stored to an elliptic shape by exclusion of scar is largely attributable to Dor [51]. Dor’s endoven- tricular circular patch plasty followed pioneering attempts at physiologic reconstruction by Jatene and Cooley [52]. Dor’s major contribution was to remove endocardial scar, exclude akinetic sep- tum, restore the curvature of the anterolateral wall, and undertake complete myocardial revascu- larization and correction of mitral regurgitation. In the event of spontaneous or inducible ventricu- lar tachycardia, cryotherapy was also applied to the edges of the resection (50% of cases). Between development of the surgical principles in 1984 and 2002, the Dor group operated on 1050 patients who were predominantly NYHA III or IV with LVEF less than 35%, LVESVI greater than 50 mL/m 2 , LVEDVI greater than 100 mL/m 2 , and mean pulmonary arterial pressure greater than 25 mm Hg [53]. One third of the cases had mitral regurgitation requiring repair. A balloon inflated to the theoretic diastolic capacity of the patient 149SURGERY FOR HEART FAILURE Maximizing Survival Potential in Very High Risk Cardiac Surgery Stephen Westaby, MS, PhD, FRCS * , L. Balacumaraswami, MBBS, FRCS (C-Th), R. Sayeed, PhD, FRCS Oxford Heart Centre, John Radcliffe Hospital, Oxford, UK The mean age and risk profile of patients referred for cardiac surgery is constantly increas- ing. Surgeons are now inclined to accept high-risk patients because interventional cardiology pro- vides less invasive alternatives for an overlapping patient cohort. As risk profile increases so does hospital mortality. A survey of 8641 patients who underwent coronary artery bypass operations in New England showed an overall mortality of 4.48%, of which 65% could be directly attributed to postcardiotomy myocardial failure [1].In the PURSUIT trial, which randomized patients who had coronary bypass and unstable angina to a glycoprotein IIb/IIIa inhibitor or placebo, the 7-day mortality or myocardial infarction rate was 22.3% in almost 700 patients in the control arm [2]. A collective review of 279 patients who had dialysis-dependent coronary bypass reported a 12.2% hospital mortality [3]. Similarly the Mayo Clinic Group reported a 14% perioperative mortality for patients who had aortic valve replacement with a left ventricular ejection fraction (LVEF) less than 35% and a borderline transvalvular gradient [4]. Intraoperative myocardial injury remains prevalent in the in- creasingly elderly surgical population because tolerance to ischemia is reduced in aged myocar- dium [5]. Patients who are difficult to wean from car- diopulmonary bypass (CPB) and those who sub- sequently deteriorate into a low cardiac output state have mortality rates between 50% and 80% [6]. In established cardiogenic shock, conventional treatment with inotropes, the intra-aortic balloon pump (IABP), or temporary circulatory support devices has not substantially improved survival. In an analysis of risk factors and outcomes for postcardiotomy mechanical support in 19,985 Cleveland Clinic patients, 0.5% received circula- tory support with overall survival of 35% [7]. In- cluded were patients who were converted to the HeartMate I implantable system and bridged to transplantation with 72% survival. In the absence of the transplant option, more innovative circula- tory support strategies are required to improve survival in the postcardiotomy setting. Mechanisms of postcardiotomy myocardial dysfunction Efforts to improve surgical results in patients who have heart failure depend on myocardial protection and preservation of contractile func- tion in the postoperative period. The clinical scenario is well known. The patient who has myocardial ischemia or chronically impaired left ventricular function undergoes combined valve and coronary bypass surgery. The ischemic time exceeds 90 minutes and despite myocardial pro- tection with blood cardioplegia, inotropic support is required to separate from CPB. The vaso- constricted patient returns to the intensive care unit with borderline cardiac index and a blood pressure of 110/70 mm Hg. Over the next 4 hours the blood pressure remains acceptable on inotro- pic support but the urine output dwindles and the ankles are cold. An IABP is deployed, seems to function well, and optimism returns until the blood gases reveal lactic acidosis and a pH of * Corresponding author. Oxford Heart Centre, John Radcliffe Hospital, Headley Way, Headington, Oxford OX3 9DU, UK. E-mail address: swestaby@ahf.org.uk (S. Westaby). 1551-7136/07/$ - see front matter Ó 2007 Published by Elsevier Inc. doi:10.1016/j.hfc.2007.05.001 heartfailure.theclinics.com Heart Failure Clin 3 (2007) 159–180 . spontaneous or inducible ventricu- lar tachycardia, cryotherapy was also applied to the edges of the resection (50% of cases). Between development of the surgical principles in 19 84 and 2002, the Dor. third of the cases had mitral regurgitation requiring repair. A balloon in ated to the theoretic diastolic capacity of the patient 14 9SURGERY FOR HEART FAILURE Maximizing Survival Potential in Very High. myocardium [47 ,48 ]. This procedure provides relief from heart failure symptoms, improved sur- vival, and important quality -of- life benefit. A comprehensive meta-analysis of largely retrospec- tive