Author + information
- Received January 11, 2007
- Revision received January 19, 2007
- Accepted January 22, 2007
- Published online May 8, 2007.
- ↵⁎Reprint requests and correspondence:
Dr. Robert H. Jones, Duke Clinical Research Institute, P.O. Box 17969, Durham, North Carolina 27715.
In past years, a January submission date of this yearly review series required a Medline search covering the first 10 months supplemented by a late December manual review of selected print journals to include all important cardiovascular surgery articles published during the entire calendar year. Important cardiac surgical literature seems to be concentrated in relatively few journals, which are all now available online (Table 1).Therefore, the Medline search has been replaced this year by a more efficient process of scanning the tables of contents of all journals used as information sources for the past 4 years. Only articles judged to be of broad general interest to most cardiovascular specialists were selected. Therefore, a large body of information about how to best do cardiac surgery has been purposefully omitted. High-interest articles with major methodologic flaws were included in prior years along with a critique cautioning readers about weak methodology. High-profile but fatally flawed reports are simply omitted from the calendar year 2006 summary, thereby decreasing the methodologic detail necessary to include for all articles. In general, the order of topics and articles described under subtopics reflect the strength and scope of impact of individual articles reported. Therefore, the first article, “Left Ventricular Assist Device and Drug Therapy for the Reversal of Heart Failure,” is honored as the best cardiovascular surgical article of the year.
Surgery for Heart Failure
Surgery for Nonischemic Cardiomyopathy
Recovery of left ventricular function on left ventricular assist device (LVAD)
The Best Cardiac Surgical Paper of 2006 Award goes to the investigators of the Royal Brompton and Harefield National Health Service Trust for tenacious application of scientific thought and energy to a clinical problem often considered futile without access to a donor heart (1).
Between December 1999 and July 2001, 27 patients underwent LVAD insertion at Harefield Hospital for severe heart failure and low cardiac output evidenced by organ dysfunction. Exclusions to form a subgroup for this study were 3 patients with ischemic cardiomyopathy, 5 patients who never could be stabilized on LVAD support, and 4 patients who did not complete the first stage of pharmacologic therapy. Included patients were those who first tolerated maximum titrated doses of lisinopril (40 mg daily), carvedilol (50 mg twice daily), spironolactone (25 mg daily), and losartan (100 mg daily). These 15 remaining patients with severe heart failure caused by nonischemic cardiomyopathy also completed a second stage of pharmacologic therapy instituted after the left ventricular diastolic diameter had remained constant for 2 weeks. Carvedilol was replaced by bisoprolol, and clenbuterol was administered in an initial dose of 40 μg twice daily, then at a dose of 40 μg 3 times daily, and finally at a dose of 700 μg 3 times daily. After the LVAD could be safely stopped temporarily, 6-min walk tests were performed. Criteria to be met while the LVAD was off for 15 min used to consider explantation included: 1) left ventricular end-diastolic diameter <60 mm; 2) left ventricular end-systolic diameter <50 mm; 3) left ventricular ejection fraction >45%; 4) left ventricular end-diastolic pressure <12 mm Hg; 5) resting cardiac index >2.8 l/min/m2; and 6) maximal oxygen consumption with exercise of more than 16 ml/kg body weight/min and an increase in minute ventilation consistent with production of carbon dioxide <34.
Four patients did not meet explantation criteria and received heart transplantation, and 3 of these survived the perioperative period. After a mean LVAD support duration of 320 ± 186 days (range 63 to 603 days), 11 (73%) of the 15 patients completed the full course of combination therapy and met the explantation criteria. Their mean ejection fraction was 64 ± 8% before explantation as compared with 12 ± 6% before implantation (p = 0.001). The mean left ventricular end-diastolic diameter was 55.9 ± 8.3 mm as compared with 75.1 ± 16.3 mm (p = 0.002). The mean left ventricular end-systolic diameter was 39.6 ± 6.5 mm as compared with 66.9 ± 16.3 mm (p = 0.002). The 6-min walk, maximal oxygen consumption, minute ventilation, and carbon dioxide production slope, right atrial pressure, pulmonary-capillary wedge pressure, cardiac output, and pulmonary artery oxygen saturation all improved. At a minimum of 4 years follow-up, 8 (73%) of the 11 explantation patients were surviving free of heart transplant and heart failure symptoms.
Surgery for Ischemic Cardiomyopathy
Surgical ventricular reconstruction (SVR) outcomes
Before joining the STICH (Surgical Treatment for Ischemic Heart Failure) trial, a single center conducted a retrospective review and 85 of the 122 patients undergoing SVR between 1991 and 2003 who would have the STICH hypothesis 2 entry criteria (2). These criteria include an ejection fraction ≤0.35, coronary artery disease (CAD) amenable to coronary artery bypass grafting (CABG) and dominant akinesia or dyskinesia of the anterior wall. Eighty-two (96.5%) of these 85 patients underwent CABG, and 16 (18.8%) of the 85 patients had mitral valve surgery. Hospital mortality was 7.1%. The 79 patients discharged were followed up for a mean of 43.7 months, during which 14 patients died after a mean interval of 24.3 months. Survival rates after 1, 3, 5, and 10 years were 89%, 79%, 75%, and 75%, respectively. The left ventricular equatorial diastolic diameter of >70 mm was the dominant predictor of long-term mortality in a multivariable model. Moreover, patients whose anterior dysfunction was akinesia and not dyskinesia also had significantly poorer long-term survival. Postoperatively, all parameters of left ventricular volume and function improved compared with baseline.
At a second center, 190 (86%) of 220 consecutive patients who underwent SVR were alive 2.3 years after surgery and most had fewer heart failure symptoms (3). Over 10 years at a third institution, 89 patients with a left ventricular ejection fraction ranging from 0.08 to 0.45 underwent SVR with 3.4% operative mortality and a 5-year survival rate of 82% (4).
The SVR operation was shown to have a much higher perioperative mortality and morbidity on review of clinical data entered into the Society of Thoracic Surgeons’ National Database from January 2002 to June 2004 (5). Operative mortality was 9.3% in 731 patients undergoing SVR at 141 centers. Re-operation was required in 14.1%. Death or major complications occurred in 33.5%. Only 20 (14%) of these 141 hospitals performed 10 or more SVR procedures. Operative mortality and morbidity was greatest in patients operated on in centers with a <5 procedure total volume. Multivariable modeling showed nonelective status to be the strongest predictor of operative mortality among the 9 significant clinical variables (Table 2).The volume of procedures performed at the site was not an independent predictor of clinical outcomes. This extremely high operative mortality and morbidity may simply reflect cardiac surgeons’ commitment to try all possible remedies for patients who cannot be weaned from cardiopulmonary bypass. An SVR would be reasonable to perform after CABG in any patient with anterior myocardial scar who could not be weaned from mechanical support after CABG because of left ventricular dysfunction.
No prospective observational studies exist that assess the potential benefit of adding SVR to CABG in patients with regional dysfunction. Until definitive information from the 1,000 patients randomized in the STICH trial to undergo CABG with or without SVR is available in 2008, there is no compelling reason to perform elective SVR.
Surgery for Ischemic Heart Disease
Comparison of CABG With Alternate Treatment Strategies
Confidential recording of treatment preference enhances value of randomized trial results
The MASS II (Medicine, Angioplasty, or Surgery Study II) randomized 611 patients with stable multivessel CAD with ≥70% stenotic lesions amenable to percutaneous coronary intervention (PCI) and preserved left ventricular function among 3 treatment options of medical therapy only (n = 203), revascularization by PCI (n = 205), or CABG (n = 203) (6). The randomization date was regarded as the time treatment began, and analysis was by intention to treat. However, before randomization, a confidential preferred treatment decision representing the consensus of 2 experienced noninterventional cardiologists was recorded. Figure 1compares the primary combined 1-year end point of survival free of death, myocardial infarction (MI), or recurrent ischemia requiring revascularization for each of the 3 randomized cohorts stratified by concordance or discordance between the treatment preferred by the clinicians and the treatment assigned by randomization. Percutaneous coronary intervention was the only treatment for which the clinically defined preferred treatment was sufficiently discordant from the randomized treatment assignment to make a significant difference in the primary end point (Fig. 2).A multivariable logistic regression model developed to identify baseline characteristics that influenced the clinical treatment preference decision showed the number of diseased coronary arteries to be the only independent predictor of discordant status. The presence of 3-vessel CAD decreased the odds for a PCI treatment decision by more than 2-fold. Therefore, in a care setting outside the randomized trial, these experienced noninterventional cardiologists would have chosen PCI for fewer patients, and those they referred to PCI would be expected to have a more event-free outcome than the PCI cohort actually experienced by random treatment assignment to PCI.
Percutaneous coronary intervention had more events that were dominated by need for subsequent revascularization when analyzed by randomization treatment assigned. Without subsequent secondary testing of their preferred treatment allocation, the published trial results might have convinced the 2 clinicians that their hesitancy about PCI use in patients with 3-vessel CAD was unwarranted bias and not reflective of superior clinical judgment.
Prior randomized trials have attempted to document generalizability of results observed in a randomized cohort to a broader population by comparing baseline characteristics of registry with randomized patients. However, this approach is costly, and even if baseline demographic and clinical characteristics assessed are similar, the diverse reasons patients or their physicians make a choice against randomization introduce bias and confounding. Experienced clinicians at equipoise to randomize a patient for whom definitive evidence does not exist for benefit between alternate treatment strategies can readily record a treatment preference they do not consider to be evidence-based. Comparison of outcomes between the randomized and clinically preferred treatment assignments would confirm the generalizability of the randomized result in 2 settings. No difference in treatment effectiveness between concordant or discordant groupings or discordant differences that show the randomized assignment to be more effective than the preferred clinical treatment would confirm broad generalizability of the randomized study outcome. This comparison in MASS II suggests that a patient cohort selected with fewer patients with severe 3-vessel CAD would have had fewer events. Therefore, PCI would have had outcomes more similar to the other 2 treatments if the criteria the clinicians used to make their preferred treatment selection also had been entry criteria for randomization.
Prospective 17-year observational study of medical therapy, PCI, and CABG
An institutional database was the resource used to compare long-term outcomes of medical therapy with revascularization and CABG to PCI as the mode of revascularization (7). A total of 18,481 patients were followed up annually for survival during a mean 7.4 years and maximal 17-year follow-up. Percutaneous coronary intervention or CABG within 30 days of cardiac catheterization defined the 2 revascularization subgroups for analysis, and the remaining patients formed the medical treatment group. To avoid bias associated with early deaths before patients could undergo revascularization, all patients were excluded from the study that died without a revascularization procedure within 5 days (defined by the median CABG waiting time). Cox multivariable modeling identified only 8 major baseline variables contained 94% of the information predictive of long-term death (Table 3).Models examining interactions with treatment choice identified the Duke CAD index to be the only variable to interact independently with treatment. The low-severity group included patients with 1- or 2-vessel CAD ≥75%, but none with stenoses ≥95%. This group had a significant survival benefit with PCI. The high-severity group was characterized by 3-vessel CAD ≥75% and 2- or 3-vessel CAD ≥95%. This coronary anatomy identified patients with a survival benefit from CABG. The intermediate-severity group had equivalent survival with PCI and CABG.
Separate Kaplan-Meier curves adjusted for baseline clinical characteristics and propensity for a specific treatment and stratified by CAD severity were created for patients receiving medical therapy and revascularization by either PCI or CABG. A plot of survival differences between patients treated with the 2 treatment strategies showed revascularization to increase survival for all levels of CAD severity over the 17 years of follow-up (Fig. 3).However, this medical cohort included only patients whose clinical symptomatology or noninvasive test evaluation suggested a need for coronary angiography. Moreover, the medical therapy that many patients received during the early years of the enrollment interval spanning 1986 to 2000 would now be considered suboptimal.
Subtraction of the PCI from the CABG risk-adjusted Kaplan-Meier survival curves for the high-severity CAD group showed a CABG survival advantage of about 6 patients per 100 treated during years 4 to 12 of follow-up (Fig. 4).The intermediate-severity and low-severity groups did not show significant differences over time. Figure 5depicts additional months of survival over 15 years added by revascularization compared with medical therapy. Revascularizations by either method added significant survival time at all CAD levels. Comparisons of choice of revascularization strategy showed a clear survival difference favoring CABG over PCI in the high-severity CAD group. However, both CABG and PCI add similar survival benefit over medical therapy to patients with low-severity and intermediate-severity CAD. This information used in the context of the clinical status and life situation adds quantitation of anticipated benefit of alternate treatment strategies for individual patients.
Bedside estimation of PCI and CABG procedure risk
Separate multivariable logistic regression models developed from 46,060 PCI procedures (8) and 16,120 CABG procedures (9) performed in New York State during 2002 were simplified into a risk index that, when applied to all 2003 data, performed with a C-index of 0.905 for PCI and 0.782 for CABG. These indices can be applied to estimate the revascularization procedure mortality for an individual patient or defined patient cohort. Moreover, these scores also were shown to relate to procedure morbidity and length of stay in New York. Clear criteria for scoring baseline clinical information needed to calculate total PCI and CABG risk scores (Table 4)are converted to actual predicted mortality that would have been expected should the patient have been treated in 2003 in New York (Tables 5 and 6).⇓⇓These predicted mortality data may simply be described to patients during the informed consent process. Alternatively, the ratio of actual procedure mortality in any patient cohort to the observed 2003 New York State CABG mortality (1.61%) or PCI mortality (0.58%) can be used to recalibrate the score to reflect mortality rates expected for a patient or patient cohort in local care environments. Cardiac surgeons and cardiologists who now do not regularly monitor their revascularization procedure outcomes now have a no-technology, easy-to-use tool to monitor success of their quality improvement programs and thereby bring ongoing quantitation of outcomes to their practice.
Outcomes of CABG within 6 h of PCI
The Society of Thoracic Surgeons’ National Cardiac Surgery Database was used to examine the 1994 to 2003 incidence and outcomes of patients undergoing CABG within 6 h of PCI (10). Patients were stratified into groups based on whether an MI had occurred within 24 h of PCI. The database did not permit further timing of the occurrence of the MI relative to intervals before, during, or after PCI. The CABG procedures done emergently after PCI decreased from 2.9% in 1994 to 0.8% in 1999 of the total CABG operations. The subset sustaining acute MI within 24 h of CABG decreased less dramatically over this interval and stabilized in 1999 at 0.3% of the total CABG population (Fig. 6).Although severity of baseline predictors of operative risk increased as the numbers of total patients undergoing CABG decreased over the 10-year interval, risk-adjusted CABG operative mortality and combined morbidity remained stable. Operative death occurred in 6% of patients presenting for operation with no MI and in 15% of patients with a preoperative MI. Combined mortality and morbidity rates were above 20% for patients with no MI and averaged 40% in patients with MI (Fig. 7).Therefore, the need for urgent CABG after PCI is decreasing, but this event continues to impose substantial mortality and morbidity to the CABG population.
Strategies for Optimizing CABG Outcomes
Positron emission tomography (PET) myocardial perfusion and viability assessment does not discriminate patients who need revascularization
The value of quantitative PET fluorodeoxyglucose (FDG) imaging of myocardial perfusion and viability for identifying patients that receive a survival benefit from revascularization was assessed in 765 consecutive patients with ejection fraction ≤0.35 and without significant valvular heart disease (11). Within the first 6 months of study entry between 1997 and 2002, 230 (30%) of the 765 patients underwent early surgical (25%) or PCI (5%) intervention. Seventy percent of patients without early intervention were considered to be medically treated, and death occurred in 184 (34.7%) of these 535 patients. Baseline, demographic, clinical, and PET/FDG variables were used to propensity match 153 of the 230 early intervention patients with 153 patients who did not undergo early intervention. In this propensity-matched cohort of 306 patients, there were 84 deaths (27.5% mortality) during a median of 3 years of follow-up. In the early intervention cohort of 153 patients, CABG with or without valve surgery was performed in 113, PCI was used in 30, and another open heart surgery was done in 10. Early intervention was associated with an adjusted 3-year mortality hazard ratio of 0.52, 95% confidence interval 0.33 to 0.81, p = 0.0004. The amount of total combined ischemia and hibernation tended to predict mortality, but early intervention was associated with a lower risk of death across the full range of ischemia and hibernation. Therefore, careful analysis of this largest single study showed no discriminatory power of PET viability testing to predicting patients who benefit from revascularization.
Off-pump CABG does not offset hemorrhage from clopidogrel pretreatment
Clopidogrel had been given to 281 (17.9%) of a total of 1,572 patients who underwent isolated off-pump CABG over a 30-month interval that ended in June 2002 (12). These patients had either received a 300-mg oral loading dose or a daily oral regimen of 75 mg of clopidogrel within 7 days of CABG. Multivariable logistic regression analysis examined the effect of clopidogrel on bleeding necessitating re-operation, blood and transfusion rates, and operative mortality. A propensity score matched pair analysis was performed for variables determined to be related to perioperative hemorrhage. This analysis showed the cohort that received clopidogrel had a 3.9-fold increase (p < 0.01) for re-exploration related to hemorrhage compared with the matched cohort not on the drug. Need for any blood transfusion was 2 times greater in patients with preoperative clopidogrel administration (p < 0.01), and the odds of receiving multiple blood units was increased by 60% (p = 0.02). Clopidogrel administration was found to be associated with an increase in need for packed red blood cells (odds ratio 2.7, 95% confidence interval 1.86 to 3.92, p < 0.01) and platelet transfusions (odds ratio 2.3, 95% confidence interval 1.48 to 3.71, p < 0.01). The clopidogrel group did not have increased operative mortality.
Surgery for Congenital Heart Disease
Benchmarks for operative mortality for congenital heart disease repair
The diversity of congenital heart presentations has made it difficult to risk-adjust outcomes of operations for congenital heart disease to document expected operative mortality. The RACHS-1 (Risk Adjustment for Congenital Heart Surgery-1) trial is increasingly being accepted as a system for consistent operative outcome reporting after a multi-institutional report was made from 1994 to 1996 data. All pediatric cardiac surgical procedures performed between 2001 and 2004 were categorized at 29 Congenital Heart Surgeon’s Society member institutions by RACHS-1 category for reporting of in-hospital mortality rates (13). The overall in-hospital mortality for the 12,672 operations that could be placed into RACHS-1 categories was 2.9%. The RACHS-1 classification functioned as a good discriminator of mortality with a C-statistic of 0.77. Individual center institutional total operative mortality ranged from 1.0% to 6.0%, and no center was consistently better or worse than another. Minimal variation was apparent at the lowest risk categories, and categories 5 and 6 had the least numbers of patients and the greatest variability in mortality rate (Fig. 8).These data from high-quality centers provide a performance benchmark for the entire pediatric cardiac surgical community.
1-year mortality after heart operations in children
California hospital discharge data from 1989 to 1999 was used to assess death within 1 year after undergoing cardiac surgery for all children <18 years old (14). In-hospital death occurred in 1,505 (0.59%) of 25,402 pediatric cardiac surgery cases. Death within the first year after discharge occurred in 148 (0.62%) of the 23,897 discharged children. The rate of death in children discharged alive decreased with time after discharge, so there were 37 deaths within 30 days (1.23/day), 44 deaths between 31 and 90 days (0.73/day), and 67 deaths between 91 and 365 days (0.24/day). Logistic regression showed that young age was an important risk factor for postdischarge death, with an odds ratio of 4.8 for neonates and 3.5 for infants when compared with children older than 1 year. Operation performed was the major risk factor. Gender, race/ethnicity, home income, and hospital case volume did not relate significantly to postdischarge deaths.
Ischemic preconditioning in legs provides myocardial protection during cardiac surgery
Remote ischemic preconditioning has been evaluated in children undergoing repair of congenital heart defects (15). Four 5-min cycles of lower limb ischemia and reperfusion were induced with a blood pressure cuff within a 10-min interval of initiating cardiopulmonary bypass for the cardiac operation. By random assignment, 17 children received the remote preconditioning protocol and 20 children served as control patients with a blood pressure cuff positioned on a leg without inflation. Preoperative and postoperative measurements were made of lung mechanics, cytokines, and troponin I. Despite similar baseline clinical characteristics and operative conduct, control patients had higher troponin I (p = 0.04), more need for postoperative inotropic support (p = 0.03), and higher airway resistance at 6 h postoperatively (p = 0.009). This simple noninvasive technique shows sufficient promise to warrant further study.
Serum markers of mortality after cardiac surgery in children
A retrospective cohort study of perioperative clinical and laboratory data from 1,001 consecutive children undergoing cardiac surgery during a 5-year period were evaluated by multivariable analysis to document the independent information that troponin T contributed to prediction of death after pediatric cardiac surgery on cardiopulmonary bypass (16). Troponin T and lactate level were the strongest predictors of 30-day all-cause death with lactate level contributing the most information. A troponin T-1 level >5.9 μg/l on the first postoperative day predicted death with 95% specificity and 30% sensitivity.
ABO-incompatible heart transplants perform well in infants
All 91 patients <18 months of age listed to undergo cardiac transplantation at a single institution from 1992 to 2002 were evaluated for the influence of use of ABO blood group-incompatible donor hearts on patient outcome (17). A strategy to accept ABO-incompatible organs was adopted in 1995. Higher waitlist status resulted in greater mortality regardless of care strategy. However, for any waitlist status, more patients underwent transplantation and fewer died using a strategy to accept ABO-incompatible organs. Parametric modeling of time-related freedom from death or re-transplantation showed no significant difference at 4 years for ABO-compatible or ABO-incompatible transplants. Using ABO-incompatible donor hearts for infant transplantation significantly improves the likelihood of transplantation and reduces waiting list mortality while not adversely altering outcomes after transplantation.
Percutaneous compares favorably with operative atrial septal defect closure
Short-term outcomes were compared on 533 patients who underwent surgical and 751 patients who underwent percutaneous atrial septal defect closure at a single institution without procedure deaths (18). Multivariable logistic regression analysis showed that surgery was an independent predictor of total and major complications. However, the surgical population had more complex defects, and 16 patients had major complications from the percutaneous procedure that required operation.
LVAD bridge to transplant improves in children
A multi-institutional database was used to analyze outcomes of 99 patients listed for ventricular assist devices by waiting for heart transplant between January 1993 and December 2003 (19). Diagnoses were cardiomyopathy in 78% and congenital heart disease in 22%. Successful bridge to transplant was more common in the modern era (Fig. 9).The 5-year survival was similar to that of patients who underwent heart transplantation without requiring ventricular assist device placement.
Surgery for Cardiac Valve Disease
Noninvasive screening effectively identifies CAD in patients scheduled for heart valve operations
From a consecutive group of 145 patients scheduled for cardiac valve surgery, 97 patients were eligible and 70 consented to undergo 64-slice computed tomography coronary angiography for assessment of CAD and comparison with an existing quantitative coronary angiogram (20). The question addressed was whether computed tomography coronary angiography could provide adequate characterization of CAD to obviate the need for coronary angiography before valve surgery. Twenty-six (37%) of 70 patients had previous myocardial infarction or angina. The prevalence of significant CAD defined as having at least one ≥50% stenosis per patient was 26%, and the test was 100% sensitive for ≥50% stenosis on the quantitative coronary angiogram. This noninvasive imaging modality would seem to be a useful screening test for significant CAD in patients with planned cardiac valve surgery.
Structured care without operation is acceptable in asymptomatic patients with severe mitral regurgitation
A total of 132 consecutive asymptomatic patients with severe degenerative mitral regurgitation were prospectively followed up for a median of 69.2 months (21). Serial clinical and echocardiographic examinations were conducted at least every 12 months, and patients were referred to surgery at the onset of mild symptoms. Asymptomatic patients also were referred on development of 1 or more of the following criteria: left ventricular end-systolic diameter ≥45 ml or end-systolic diameter index ≥25 ml/m2, fractional shortening <0.32 or ejection fraction <0.60, systolic pulmonary pressure >50 mm Hg, or recurrent atrial fibrillation. Patients were continuously followed up after surgery to monitor the late results of the preoperative evaluation strategy. Follow-up information was complete for 129 (98%) of the 132 consecutive asymptomatic patients entered into the study. During follow-up, 8 deaths occurred, and 38 patients developed criteria for operation. There was no operative mortality. This approach of systematic patient monitoring with watchful waiting provides a useful population to compare with similar patients who also would be at low risk for mitral valve repair.
Age and not valve type is the strongest predictor of mortality after cardiac valve replacement
A meta-analysis was performed on overall mortality after aortic valve replacement on information from 32 articles published from 1989 to 2004 that reported operations performed between 1975 and 2002 with a maximum follow-up of at least 10 years (22). The 17,439 patients were followed up for 101,819 patient-years, and 5,170 deaths (936 early and 4,234 late) occurred. The mean patient age was 58 years in the 15 articles describing mechanical valve use and 69 years in the 23 articles describing biological valve use. For mechanical valves, the overall death rate was 3.99%/patient-year, and for biological valves the overall death rate was higher at 6.33%/patient-year. Median age was strongly related to death rate (p < 0.0001) (Fig. 10).When corrected for age in the regression model, variables that retained a significant effect on death rate included New York Heart Association functional classes III and IV, which increased, and aortic regurgitation, which decreased mortality. After correction for the impact of these factors, valve type did not have an independent influence on mortality. This analysis with patient summary data and not individual patient data as the unit of analysis did not permit analysis of an interaction term between age and valve type. These data suggest that number of years lived may not be as important a determinant in valve choice as combinations of other specific risk factors that interface with age. For example, risk of bioprosthetic valve degeneration would be especially troublesome in young otherwise healthy patients, but also in older patients with a long life expectancy.
Tricuspid valve repair provides better outcomes than replacement
Outcomes of tricuspid valve repair were compared with valve replacement in 250 patients with organic tricuspid valve disease at a single institution from 1979 to 2003 (23). Tricuspid repair was performed in 178 (71%). Tricuspid replacement in 72 (29%) was done using a bioprosthetic valve in 54 and a mechanical valve in 18 patients. Cox regression analysis showed tricuspid valve replacement to be an independent predictor of lower event-free survival (hazard ratio 2.0, 95% confidence interval 1.1 to 3.6, p < 0.02). Echocardiography showed more moderate-to-severe tricuspid regurgitation in patients undergoing repair, but no difference in functional class or re-operation rates over a mean follow-up time of 5.2 ± 4.1 years.
Atrial Fibrillation in Cardiac Surgical Patients
Left atrial volume and age predict atrial fibrillation after heart operations
The power of clinical and transthoracic echocardiogram variables to predict postoperative atrial fibrillation was assessed in 205 patients undergoing cardiac surgery (24). Multivariable modeling of prospectively acquired data showed left atrial volume and age to be the 2 significant predictors of postoperative atrial fibrillation that occurred in 84 (41.4%) of the 205 patients a median of 1.8 days after cardiac surgery. Stratification of patients by age and left atrial volume suggests special attention should be directed to prevention and treatment of postoperative atrial fibrillation in patients who are older than age 65 years and have a left atrial diameter >32 mm Hg/m2(Fig. 11).
Atorvastatin reduces atrial fibrillation after cardiac surgery
Seven days before a planned elective cardiac operation requiring cardiopulmonary bypass, 200 patients were randomized to receive atorvastatin (40 mg/day, n = 101) or placebo (n = 99) (25). Atorvastatin reduced the incidence of postoperative atrial fibrillation compared with placebo (95% vs. 57%, p = 0.003). Therefore, the length of stay was longer for patients in the placebo compared with those in the atorvastatin arm (6.9 ± 1.4 days vs. 6.3 ± 1.2 days, p = 0.001). Multivariable analysis showed that atorvastatin reduced the risk of atrial fibrillation (odds ratio 0.39, 95% confidence interval 0.18 to 0.85, p = 0.017), whereas a high postoperative C-reactive protein increased risk (odds ratio 2.0, 95% confidence interval 1.2 to 7.0, p = 0.01). Major adverse cerebrovascular events at 30 days did not differ between the 2 groups.
Status Report of Major Ongoing Randomized Clinical Trials in Cardiac Surgical Patients
Outcomes After Myocardial Revascularization: On and Off Cardiopulmonary Bypass (ROOBY) trial
Sponsor: Veterans Affairs Cooperative Study.
Principal investigator: Frederick L. Grover, MD ().
Hypothesis: In properly selected patients, 30-day and 1-year mortality and morbidity are equivalent when CABG is conducted with or without using cardiopulmonary bypass.
Enrollment: 2,100 of 2,200 patients enrolled by 17 U.S. Veterans Affairs hospitals.
Sponsor: National Institutes of Health.
Principal investigator: Robert H. Jones, MD ().
Hypothesis 1: CABG combined with intensive medical therapy improves long-term survival compared with intensive medical therapy alone in 1,000 patients with LVEF ≤0.35 and CAD amenable to CABG.
Hypothesis 1 enrollment: 1,039 of 1,200 patients enrolled by 99 sites worldwide. Completion expected June 2007.
Hypothesis 2: Adding surgical ventricular restoration to CABG and intensive medical therapy improves survival free of cardiac hospitalization in 1,000 patients with LVEF ≤0.35, CAD amenable to CABG, and dominant anterior akinesia or dyskinesia.
Hypothesis 2 enrollment: 1,000 patients completed January 24, 2006, by 96 sites worldwide. Primary outcome report expected in 2008.
Future Revascularization Evaluation in Patients with Diabetes Mellitus: Optimal Management of Multivessel Disease (FREEDOM) trial
Sponsor: National Heart, Lung, and Blood Institute.
Principal investigator: Valentin Fuster, MD, PhD ().
Hypothesis: In 2,400 diabetic patients with multivessel disease amenable to CABG and PCI and indication for revascularization, either CABG or PCI will be superior in preventing death, nonfatal myocardial infarction, and stroke.
Enrollment: 537 of 2,400 patients enrolled. Activation completed at 117 clinical sites.
The author acknowledges Vanessa Moore with gratitude for her contributions in the preparation of this manuscript.
↵1 Dr. Jones is the principal investigator of the National Institutes of Health-sponsored Surgical Treatment for Ischemic Heart Failure (STICH) trial and received no support for production of this article.
- Received January 11, 2007.
- Revision received January 19, 2007.
- Accepted January 22, 2007.
- American College of Cardiology Foundation
- O’Neill J.O.,
- Starling R.C.,
- McCarthy P.M.,
- et al.
- Pereira A.C.,
- Lopes N.H.M.,
- Soares P.R.,
- et al.
- Wu C.,
- Hannan E.L.,
- Walford G.,
- et al.
- Hannan E.L.,
- Wu C.,
- Bennett E.V.,
- et al.
- Tarakji K.,
- Brunken R.,
- McCarthy P.M.,
- et al.
- Kapetanakis E.I.,
- Medlam D.A.,
- Petro K.R.,
- et al.
- Cheung M.M.H.,
- Kharbanda R.K.,
- Konstantinov I.E.,
- et al.
- Blume E.D.,
- Naftel D.C.,
- Bastardi H.J.,
- Duncan B.W.,
- Kirklin J.K.,
- Webber S.A.,
- Pediatric Heart Transplant Study Investigators
- Meijboom W.B.,
- Mollet N.R.,
- Van Mieghem C.A.G.,
- et al.
- Rosenhek R.,
- Rader F.,
- Klaar U.,
- et al.
- Patti G.,
- Chello M.,
- Candura D.,
- et al.