Jeri Sever

Trends in coronary artery bypass surgery results: a recent, 9-year study

BACKGROUND The demographics of patients undergoing coronary artery bypass grafting (CABG) have changed over time and may contribute to differing operative mortality and the combination of mortality and morbidity (M + M). In this study, the trends in results are analyzed and causes are suggested. METHODS Prospectively collected data concerning 4,839 CABG operations was divided into three time cohorts (1990 to 1992, 1993 to 1995, 1996 to 1998) and analyzed by univariate and multivariate techniques. RESULTS Mean age and female gender frequency increased in the later time cohorts (60.7 +/- 9.0 to 63.4 +/- 9.9 years and 16.5% to 21.4%, respectively). The following comorbidities were more prevalent in the later time cohorts: diabetes (26.7% versus 18.6%), renal failure (8.5% versus 2.2%), peripheral vascular disease (20.7% versus 11.0%), previous cerebrovascular accident (6.7% versus 5.0%), urgent procedures (41.5% versus 26.9%), unstable angina (47.8% versus 31.7%), urgent CABG following myocardial infarction (17.1% versus 7.3%), previous percutaneous transluminal coronary angioplasty (8.0% versus 4.5%), ejection fraction less than 35% (20.5% versus 10.4%), (all p < 0.05). Procedurally, increased utilization of the left internal mammary artery, multiple arterial conduits, and warm blood cardioplegia occurred in the later cohorts (91.2%, 22.2%, and 80.4% versus 78.7%, 3.4%, and 38.0%, respectively). The mortality rate was 2.0% and the M + M rate was 15.6% in all 4,839 patients. The mortality and M + M for the three cohorts were 1.6%, 2.0%, and 2.3% and 18.4%, 17.2% and 12.5%, respectively. The risk-adjusted mortality and M + M decreased from 2.4% and 15.9%, respectively, in 1990 to 1992 to 1.8% and 8.4% in 1996 to 1998 (p < 0.001). The difference in adjusted event rates was minimized when the surgical factors were entered into the model. CONCLUSIONS Over time, there has been a trend toward operating on older patients with more comorbidities. Though hospital mortality has been stable, risk-adjusted M + M has been in a constant decline. This decline was associated with an increased use of left internal mammary artery grafts, multiple arterial conduits, and warm blood cardioplegia during the later years of the study.

Left Ventricular Mass Regression Early After Aortic Valve Replacement

BACKGROUND Regression of left ventricular hypertrophy is an important and well-recognized salutary effect of aortic valve replacement. The earliest evidence of left ventricular mass regression after aortic valve replacement and the influence of prosthesis type are not well known, and were the focus of this study. METHODS Transthoracic echocardiography was used to measure left ventricular mass index preoperatively and before discharge in 57 consecutive patients undergoing isolated aortic valve replacement (with or without coronary artery bypass grafting). RESULTS Three patients were excluded from the study because of inability to obtain accurate M-mode echocardiographic images for left ventricular mass measurement preoperatively (1) or postoperatively (2). Of the remaining 54 patients, mechanical bileaflet valves were used in 19, stented tissue bioprostheses were implanted in 15, and a stentless porcine bioprosthesis was chosen for 20. Postoperative echocardiograms were obtained 4.9 +/- 2.3 days after aortic valve replacement (range, 2 to 9 days). A two-way repeated-measures analysis of variance demonstrated a significant reduction of left ventricular mass index before discharge (preoperative 141.4 +/- 45.2 g/m2, postoperative 127.5 +/- 32.8 g/m2; p = 0.0005) but no differences between prostheses. CONCLUSIONS Left ventricular mass regression begins early after aortic valve replacement, probably because of reduction of transvalvular gradients and left ventricular wall stress. At least in the very early postoperative period, the type of prosthesis does not influence the extent of mass regression.

Troponin after cardiac surgery: a predictor or a phenomenon?

BACKGROUND Increased cardiac troponin is observed after virtually every cardiac operation, indicating perioperative myocardial injury. The clinical significance of this elevation is controversial. This study aimed to correlate postoperative troponin levels with major adverse cardiac events (MACE). METHODS The study included 1918 consecutive patients undergoing adult cardiac operations, including 1515 isolated coronary procedures, 229 valvular operations, and 174 combined coronary/valve procedures. Peak troponin T (normal value < 0.1 microg/L) was measured at less than 24 hours postoperatively. Excluded were 506 patients with a recent myocardial infarction (< 30-days of operation). The primary outcome was a composite of death, electrocardiogram-defined infarction, and low output syndrome (MACE). RESULTS Mortality rates were 1.4%, 6.1%, and 7% in the coronary bypass, valve, and combined groups, respectively (p < 0.001). The rates of MACE were 17%, 35%, and 44% (p < 0.0001), and mean troponin T levels were 0.9 +/- 1.5, 1.2 +/- 2.9, and 1.3 +/- 1.2 microg/L (p < 0.001), in the coronary bypass, valve, and combined groups, respectively. All patients were divided into quintiles based on their peak postoperative troponin level (Q1, 0.0 to 0.39; Q2, 0.4 to 0.59; Q3, 0.6 to 0.79; Q4, 0.8 to 1.29; and Q5, > 1.3 microg/L). Adverse outcomes were similar and stable in the lower quintiles. A stepwise increase in adverse outcomes was observed in the higher quintiles. Receiver operating characteristic curve analysis revealed a troponin cutoff of 0.8 microg/L was the most discriminatory for MACE (area under the curve, 0.7). Multivariable analyses showed a troponin value of more than 0.8 microg/L was independently associated with MACE. CONCLUSIONS Moderate elevations in troponin are common after cardiac operations; troponin is a well-described predictor of outcomes. Troponin levels exceeding 0.8 microg/L are associated with increased MACE in patients without a history of preoperative myocardial infarction within 30 days of operation.

Are stentless valves hemodynamically superior to stented valves? Long-term follow-up of a randomized trial comparing Carpentier–Edwards pericardial valve with the Toronto Stentless Porcine Valve

OBJECTIVE The benefit of stentless valves remains in question. In 1999, a randomized trial comparing stentless and stented valves was unable to demonstrate any hemodynamic or clinical benefits at 1 year after implantation. This study reviews long-term outcomes of patients randomized in the aforementioned trial. METHODS Between 1996 and 1999, 99 patients undergoing aortic valve replacement were randomized to receive either a stented Carpentier-Edwards pericardial valve (CE) (Edwards Lifesciences, Irvine, Calif) or a Toronto Stentless Porcine Valve (SPV) (St Jude Medical, Minneapolis, Minn). Among these, 38 patients were available for late echocardiographic follow-up (CE, n = 17; SPV, n = 21). Echocardiographic analysis was undertaken both at rest and with dobutamine stress, and functional status (Duke Activity Status Index) was compared at a mean of 9.3 years postoperatively (range, 7.5-11.1 years). Clinical follow-up was 82% complete at a mean of 10.3 years postoperatively (range, 7.5-12.2 years). RESULTS Preoperative characteristics were similar between groups. Effective orifice areas increased in both groups over time. Although there were no differences in effective orifice areas at 1 year, at 9 years, effective orifice areas were significantly greater in the SPV group (CE, 1.49 +/- 0.59 cm(2); SPV, 2.00 +/- 0.53 cm(2); P = .011). Similarly, mean and peak gradients decreased in both groups over time; however, at 9 years, gradients were lower in the SPV group (mean: CE, 10.8 +/- 3.8 mm Hg; SPV, 7.8 +/- 4.8 mm Hg; P = .011; peak: CE, 20.4 +/- 6.5 mm Hg; SPV, 14.6 +/- 7.1 mm Hg; P = .022). Such differences were magnified with dobutamine stress (mean: CE, 22.7 +/- 6.1 mm Hg; SPV, 15.3 +/- 8.4 mm Hg; P = .008; peak: CE, 48.1 +/- 11.8 mm Hg; SPV, 30.8 +/- 17.7 mm Hg; P = .001). Ventricular mass regression occurred in both groups; however, no differences were demonstrated between groups either on echocardiographic, magnetic resonance imaging, or biochemical (plasma B-type [brain] natriuretic peptide) assessment (P = .74). Similarly, Duke Activity Status Index scores of functional status improved in both groups over time; however, no differences were noted between groups (CE, 27.5 +/- 19.1; SPV, 19.9 +/- 12.0; P = .69). Freedom from reoperation at 12 years was 92% +/- 5% in patients with CEs and 75% +/- 5% in patients with SPVs (P = .65). Freedom from valve-related morbidity at 12 years was 82% +/- 7% in patients with CEs and 55% +/- 7% in patients with SPVs (P = .05). Finally, 12-year actuarial survival was 35% +/- 7% in patients with CEs and 52% +/- 7% in patients with SPVs (P = .37). CONCLUSION Although offering improved hemodynamic outcomes, the SPV did not afford superior mass regression or improved clinical outcomes up to 12 years after implantation.

Cardiac Pacing Following Surgery for Acquired Heart Disease

Abstract Background: This study is comprised of 3493 consecutive patients who underwent open heart surgery at our institution. Data on all patients were collected prospectively. Methods: In 45 patients (Group P) (1.3%), a permanent pacemaker (PP) was inserted postoperatively. For the purpose of the study, these patients were compared to 3448 patients (Group NP) who did not require insertion of a PP after surgery. Mean follow‐up was 33 months (range 1.5 to 66). Results: We found Group P patients were older (64.8 ± 11.0 vs 61.0 ± 11.0 years, p < 0.05), had a higher proportion of elderly (> 70 years) (36% vs 19%, p = 0.01), and of female patients (48.8% vs 22.7%, p < 0.001) compared to Group NP. Group P also had a higher incidence of preoperative rhythm abnormalities (26.6% vs 5.7%, p < 0.0001), redo surgery (13.3% vs 4.6%, p = 0.02), aortic valve surgery (48.8% vs 10.8%, p < 0.001), and tricuspid valve surgery (repair 3, replacement 1) (8.8% vs 0.5%, p < 0.001), in addition to a higher proportion of patients in whom cold (vs warm) blood cardioplegia was used (68.8% vs 52.3%, p = 0.03). Indication for postoperative PP was sick sinus syndrome (SSS) in nine patients; atrial fibrillation in eight patients; atrioventricular block (AVB) in 27 patients; and combined AVB/SSS in 1 patient. There were no operative deaths in Group P. Necessity for PP after heart surgery had a significant impact on resource utilization resulting in prolonged ventilation (3.1 ± 7.5 vs 1.4 ± 3.3 days, p < 0.01), intensive care unit (5.1 ± 10.2 vs 2.5 ± 4.0 days, p < 0.01), and postoperative hospital stay (18.0 ± 13.4 vs 8.1 ± 9.4 days, p < 0.01). Conclusions: By multivariate logistic regression (odds ratio and p value in parentheses), aortic valve surgery (8.23, p = 0.001), the absence of preoperative sinus rhythm (5.60, p = 0.001), postoperative myocardial infarction (3.46, p = 0.024), and female gender (2.52, p = 0.003), were found to be independent predictors for PP requirement post surgery.

Can patients with left main stenosis wait for coronary artery bypass grafting

BACKGROUND The economic impact of health care reforms may result in waiting lists for coronary artery bypass grafting. This study was designed to examine the clinical results of patients with left main stenosis who were placed on a triaged wait list for operation. METHODS Data were collected prospectively on 2,145 patients undergoing isolated coronary artery bypass grafting between 1989 and 1994. Critical left main stenosis (LMS, 50% or more stenosis) was present in 281 patients, and 1,864 patients had no left main disease, or a left main stenosis of less than 50% (no LMS). RESULTS The average time from angiography to operation was shorter in patients with LMS (LMS 38 +/- 46 days versus no LMS 84 +/- 71 days; p = 0.0001). Two patients in the LMS group died; they had declined operation. Four patients suffered non-Q wave myocardial infarctions, all of whom subsequently underwent operation with no perioperative complications. The presence of LMS did not influence operative mortality (LMS 2.8% versus no LMS 1.3%), the incidence of low output syndrome (LMS 8.3% versus no LMS 5.4%), or the incidence of perioperative myocardial infarction (LMS 3.8% versus no LMS 4.2%). To examine the effect of waiting time on outcomes, patients with LMS were divided into early (operation 10 days or less after angiography) and late revascularization groups (more than 10 days). Operative mortality, low output syndrome, and myocardial infarction were similar in the early and late groups. Patients in the early group were more likely to have New York Heart Association functional class IV symptoms (64% versus late 22%; p < 0.0001), unstable angina (87% versus late 65%; p < 0.0001), or a recent preoperative myocardial infarction (17% versus late 2%; p < 0.0001). CONCLUSIONS Carefully selected patients with significant left main stenosis can safely wait for operation with a low risk of complications. Early surgical intervention is allocated to patients with severe symptoms or recent preoperative myocardial infarction.

Cut‐Off Values for Transit Time Flowmetry: Are the Revision Criteria Appropriate?

Graft Imaging to Improve Patency (GRIIP), a single‐center, randomized blinded clinical trial, reported that intraoperative graft assessment with graft revision according to a priori criteria of transit time flowmetry (TTF) and intraoperative fluorescent angiography did not improve graft patency at one year after coronary artery bypass grafting (CABG) when compared with standard intraoperative management. The objective of this study is to investigate whether other TTF values are more predictive of the saphenous vein graft (SVG) failure and/or clinical outcomes.

The Current Status of Myocardial Revascularization: Changing Trends and Risk Factor Analysis

Abstract From November 1989 to December 1994, we performed 2264 bypass procedures. Data were collected prospectively. The population was divided into three subgroups: group 1 = single internal mammary artery (IMA) ± veins (n = 1584); group 2 = veins only (n = 503); and group 3 = two or more arterial conduits ± veins (n = 177). Patients who received only saphenous vein conduits (group 2) were significantly older (66.7 ± 8.9 years) than either group 1 (60.3 ± 8.3 years) or group 3 (51.6 ± 9.2 years). Furthermore, this cohort group had the highest percentage of females (28.6%), urgent cases (43.5%), preoperative myocardial infarction (MI) (18.5%), and redo surgery (5.4%). In contrast, patients who received two or more arterial conduits were 94.9% male, and had the lowest incidence of urgent cases (18.1%) and redo surgery (0.5%). Mortality was 1.4% in group 1 and 3.2% in group 2; there were no deaths in group 3. Furthermore, group 2 patients had the highest incidence of perioperative MI (6.6%), low output syndrome (22.1%), intra‐aortic balloon pump (IABP) assist (6.2%), and stroke (2.7%). By multivariate logistic regression analysis (odds ratio in parentheses), redo surgery (7.92), preoperative IABP (5.53), poor LV function (4.01), renal impairment (3.94), and advanced age (2.12) were all predictors of operative mortality. When mortality and morbidity (death, infarction, low output syndrome, IABP assist) were combined, regression analysis revealed that in addition to the above variables, female gender and cold cardioplegia were also independent predictors of combined mortality and morbidity. Resource utilization was determined for the three patient groups. There was concern that the increased technical demands of multiple arterial grafting along with longer periods of aortic occlusion and pump times may lead to increased complications and prolonged hospital stay. However, we found that group 3 had the lowest ventilation time, intensive care unit stay, and hospital stay. The results no doubt were influenced by case selection. Whether or not this approach to revascularization will increase long‐term survival and freedom from reoperation will require further study.

Impact of clopidogrel use on mortality and major bleeding in patients undergoing coronary artery bypass surgery.

Patients who received clopidogrel prior to coronary bypass surgery are at increased risk for bleeding that must be balanced with risk of ongoing ischemia if coronary artery bypass grafting is delayed. This study aimed to evaluate the impact of clopidogrel on mortality and major bleeding in patients undergoing urgent coronary bypass surgery. We reviewed 451 consecutive patients who underwent urgent isolated coronary bypass surgery; 262 had not received clopidogrel, whereas 189 received clopidogrel < or = 5 days preoperative. The primary endpoint was in-hospital death, massive transfusion or massive blood loss. Patient characteristics were almost similar between groups. There was no difference in in-hospital death or massive bleeding indices between groups (clopidogrel: 7% vs. no clopidogrel: 6%, P = 0.9). No difference was observed even after adjusting for the date of stopping clopidogrel preoperatively. Multivariate regression analysis showed that clopidogrel or the duration it was stopped preoperatively, did not predict adverse outcomes. Significant independent predictors included preoperative renal dysfunction, hemoglobin level and peripheral vascular disease. clopidogrel, or the time it was stopped prior to surgery, was not a risk factor for in-hospital death, massive bleeding, or other poor early outcomes in patients undergoing urgent coronary artery bypass surgery.

Coronary bypass with arterial conduits

To assess the efficacy of arterial revascularization 368 patients were studied who underwent myocardial revascularization with two or more arterial conduits (group M) and compared with 2092 patients in whom a single internal mammary artery +/- veins (group S), and to a third group in which only saphenous vein conduits (group V, n = 602) were used. Group M patients were younger (aged 54.0(9.5) years; 4.1% aged > 70 years) than either group V (67.6(8.9) years, 40.7% > 70 years, P < 0.0001) or group S patients (62.0(8.5) years, 15.7% > 70 years, P < 0.0001). Furthermore, this cohort group had the lowest percentage of females (8.4%), of urgent cases (21.7%), of preoperative myocardial infarction (6.0%), and of redo surgery (0.8%). In contrast, patients who received only saphenous vein conduits had the highest proportion of female patients (29.2%), of urgent cases (47.4%), of preoperative myocardial infarction (16.5%), and of redo surgery (5.5%). By multivariate logistic regression analysis (odds ratio in parentheses), redo surgery (6.06, P = 0.0001), preoperative intra-aortic balloon pump assist (6.11, P = 0.0001), diabetes (1.97, P = 0.03), urgent surgery (1.80, P = 0.05), and advanced age (2.14, P = 0.01) were all predictors of operative mortality. In contrast, while choice of conduit appeared to influence outcome by univariate analysis (4.2% mortality in group V, P < 0.001), it was not found to be a predictor of either mortality or morbidity by regression analysis. The present results indicate that, in carefully selected patients, despite increased technical demands with longer periods of aortic occlusion and longer pump times, multi-arterial grafting (compared with conventional revascularization) is a safe and efficacious procedure. Whether or not this approach to revascularization will increase long-term survival and freedom from reoperation will require further study.