Moderate Aortic Stenosis in Coronary Artery Bypass Grafting Patients More Than 70 Years of Age: To Replace or Not to Replace?

Moderate Aortic Stenosis in Coronary Artery Bypass Grafting Patients More Than 70 Years of Age: To Replace or Not to Replace? François Dagenais, MD, Patrick Mathieu, MD, Daniel Doyle, MD, Éric Dumont, MD, and Pierre Voisine, MD Department of Cardiac Surgery, Institut Universitaire de Cardiologie et de Pneumologie de Québec, Québec, Ontario, Canada ADULT CARDIAC Background. Moderate aortic stenosis in coronary artery bypass graft surgery (CABG) patients more than 70 years old is not unusual. The risk-benefit of performing a concomitant aortic valve replacement (AVR) is often difficult to assess. To stratify the risk-benefit ratio, we reviewed outcomes of CABG patients more than 70 years old with preoperative moderate aortic stenosis (valve area 1.0 to 1.6 cm 2 or indexed valve area 0.6 to 1.0 cm 2 /m 2 ). Methods. Among 263 CABG patients more than 70 years old with moderate aortic stenosis, 167 patients underwent only CABG and 96 had CABG AVR. Results. Cross-clamp time (p < 0.0001) and perioperative transient ischemic attack cerebrovascular accident (p < 0.04) were significantly higher in the CABG AVR group. In-hospital mortality was comparable among groups (CABG 6.0% versus CABG AVR 4.2%; p 0.8). At a mean follow-up of 4.5 3.0 years, 5-year survival (CABG 64.2% 4.3% versus CABG AVR 62.3% 5.5%) and freedom from AVR (CABG 97.8% 1.2% versus CABG AVR 98.9% 1.1%; p 0.13) were comparable among both groups. Among patients treated with CABG alone, receiver operating characteristic curve analysis identified 26 mm Hg and 15 mm Hg as maximum and mean aortic valve gradients, respectively, for increased risk of reoperation for late AVR. Multivariate analyses for predictors of operative mortality were preoperative renal failure (odds ratio [OR] 7.64, p < 0.001) and intubation more than 48 hours (OR 11.10, p < 0.0002); for late death, ejection fraction less than 40% (OR 3.35, p < 0.02), New York Heart Association functional class III or IV (OR 2.37, p < 0.002), chronic obstructive pulmonary disease (OR 2.26, p < 0.02), and renal failure (OR 3.03, p < 0.003); for perioperative transient ischemic attack cerebrovascular accident, cross-clamp time (OR 1.02, p < 0.02) and Parsonnet score (OR 1.09, p < 0.05). Conclusions. For CABG patients more than 70 years old with minimal comorbidities especially in the presence of aortic gradients of 26/15 mm Hg or greater, concomitant AVR for moderate aortic stenosis should be performed during CABG and may be performed with minimal additional operative risk. Patients with significant comorbidities should be managed with CABG alone, owing to an increased perioperative risk, poor midterm survival, and minimal risk of AVR at 5 years. (Ann Thorac Surg 2010;90:1495 1500) 2010 by The Society of Thoracic Surgeons With the aging population, incidental aortic valve disease among patients undergoing coronary artery bypass graft surgery (CABG) is common. While it is widely accepted that severe aortic stenosis should be managed with concomitant aortic valve replacement (AVR), treatment of moderate aortic stenosis remains controversial [1, 2]. The controversy to perform or not perform a concomitant AVR for moderate aortic stenosis in a CABG setting is raised by the difficulty in assessing the rate of progression of the aortic stenosis and whether the patient will outlive the natural progression of the Accepted for publication June 7, 2010. Presented at the Forty-sixth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 25 27, 2010. Address correspondence to Dr Dagenais, Department of Cardiac Surgery, Institut Universitaire de Cardiologie et de Pneumologie de Québec, 2725 Chemin Sainte-Foy, Québec, ON G1V 4G5, Canada; e-mail: francois. dagenais@chg.ulaval.ca. aortic stenosis. This decision-making process is especially difficult for elderly patients. Concomitant AVR avoids the risk of long-term reoperation for progressive native aortic valve stenosis. However, the patient may be exposed to additional perioperative risks and long-term prosthesis-related complications. Conversely, performing only CABG may expose the patient to a higher risk of late reoperation for AVR. Reoperative AVR after CABG has been reported with acceptable results but still remains a challenging procedure, especially in the elderly [3, 4]. Establishment of risk factors of perioperative mortality and midterm survival may allow to better stratify the decision whether to replace the aortic valve in elderly patients who present for CABG with moderate aortic stenosis. To investigate this issue, we herein reviewed our institutional experience with elderly patients who have moderate aortic stenosis and are undergoing CABG. 2010 by The Society of Thoracic Surgeons 0003-4975/$36.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2010.06.036

ADULT CARDIAC 1496 DAGENAIS ET AL Ann Thorac Surg MODERATE AORTIC STENOSIS IN OLDER CABG PATIENTS 2010;90:1495 1500 Patients and Methods Patient Cohort Selection Study approval was obtained from the review boards of the Institut Universitaire de Cardiologie et de Pneumologie de Québec, Québec, Canada. Patients more than 70 years old undergoing CABG and presenting with a moderate aortic stenosis (defined as aortic valve area 1.0 to 1.6 cm 2 or an indexed valve area 0.6 to 1.0 cm 2 /m 2 ) were identified between 1995 and 2009 within the databank of the Division of Cardiac Surgery, Institut Universitaire de Cardiologie et de Pneumologie de Québec. Data were collected prospectively. No concomitant procedure was included. Decision to replace the aortic valve was based on the surgeon s evaluation at the time of the surgery. Operative techniques were conducted according to the current surgical practices and were standard among surgeons. Variable Definitions and Follow-Up Preoperative renal failure was defined as baseline creatinine greater than 150 mol/l; chronic obstructive pulmonary disease (COPD) by forced expiratory volume of air in 1 second (FEV 1 ) of less than 75% of the predicted value, or by daily use of a bronchodilator. Postoperative transient ischemic attack cerebrovascular accident (TIA-CVA) was diagnosed by the presence of a new focal deficit on neurologic examination as either transient or permanent with or without a new lesion on brain computed tomography; postoperative renal failure was diagnosed by a creatinine increase of 50 mol/l from baseline value. All patients were followed yearly in a dedicated valve clinic either through a clinic visit or by phone contact. Survival was validated by interrogating the Quebec Provincial Death Registry. Statistical Analysis Qualitative data are presented as percentages and quantitative data as mean SD. Differences between CABG and AVR CABG in proportions were tested with the 2 test or Fisher s exact test, and continuous variables were compared using Student s t test. The logistic regression analysis was performed to predict death and CVA. Variables with a probability value of less than 0.20 were candidates for the multivariate regression model building. The selection variables were performed using the forward and backward statistical approaches. The selection variable with interaction terms was performed using a forward approach. The survival function was obtained from the Nelson-Aalen estimator of the cumulative hazard rate. The Cox regression model was used to predict the distribution of the time of death from the following variables: Parsonnet score, ejection fraction less than 40%, New York Heart Association (NYHA) functional class, preoperative and postoperative CVA, COPD, preoperative and postoperative renal failure, peripheral vascular disease, reoperation for bleeding, preoperative atrial fibrillation, intubation longer than 48 hours, and septicemia. Variables with a probability value less than 0.20 were candidates for the multivariate Cox regression model building. Martingale s residuals were used to examine the functional form of continuous variables and to conclude that all variables needed no transformation. The graphic representation of the logarithm cumulative hazard rates versus time was used first to assess the proportionality, whereas continuous variables were stratified into four to eight disjoint strata. Second, an artificially time-dependent covariate was added to the model to test the proportionality assumption. The results were considered significant with p values less than 0.05. The data were analyzed using the statistical package program SAS version 9.2 (SAS Institute, Cary, NC). Results Patient Characteristics Among 263 patients more than 70 years old operated on for CABG with moderate aortic stenosis since 1995, 167 patients underwent only CABG whereas 96 had a concomitant AVR performed. Groups were comparable for major preoperative variables (Table 1), although higher maximal aortic valve gradients and lower indexed valve area were documented in the CABG AVR group. Number of distal grafts were significantly higher in the CABG group (CABG 3.4 0.9, CABG AVR 3.0 1.2; p 0.01). Perioperative outcomes are outlined in Table 2. Cardiopulmonary bypass and cross-clamp times were significantly longer in the CABG AVR group. Transient ischemic attack-cva (p 0.04) and postoperative renal failure (p 0.03) were more frequently encountered in the CABG AVR group as was mediastinitis (p 0.03) and reoperation for bleeding (p 0.04). Table 1. Patient Characteristics Variables CABG n 167 CABG AVR n 96 p Value Male 102 (61.1%) 69 (71.9%) 0.082 Redo 5 (3.0%) 4 (4.2%) 0.73 Peak gradient, mm Hg 20.6 7.0 42.2 14.7 0.0001 Aortic valve area, cm 2 1.44 0.24 1.15 0.16 0.0001 Indexed aortic valve area, 0.81 0.13 0.65 0.12 0.0001 cm 2 /m 2 Ejection fraction 40% 150 (90.9%) 86 (92.5%) 0.82 Angina III or IV 98 (59.0%) 47 (49.0%) 0.12 NYHA class III or IV 71 (43.6%) 36 (37.9%) 0.43 Elective/urgent 165 (99.4%) 96 (100%) 1.0000 Previous CVA 24 (14.4%) 14 (14.6%) 1.0000 Diabetes mellitus 47 (28.1%) 28 (29.2%) 0.89 COPD 36 (21.6%) 17 (17.7%) 0.52 Hypertension 135 (80.8%) 74 (77.9%) 0.63 Renal failure 32 (19.2%) 14 (14.6%) 0.40 Peripheral vascular 47 (28.1%) 19 (19.8%) 0.14 disease Previous atrial fibrillation 21 (12.6%) 9 (9.4%) 0.55 AVR aortic valve replacement; CABG coronary artery bypass graft surgery; COPD chronic obstructive pulmonary disease; CVA cerebrovascular accident; NYHA New York Heart Association.

Ann Thorac Surg DAGENAIS ET AL 2010;90:1495 1500 MODERATE AORTIC STENOSIS IN OLDER CABG PATIENTS Table 2. Perioperative Results Outcomes CABG n 167 CABG AVR n 96 p Value TIA-CVA 7 (4.2%) 11 (11.5%) 0.040 Renal failure 17 (10.2%) 20 (20.8%) 0.026 Mediastinitis 1 (0.6%) 5 (5.2%) 0.026 Atrial fibrillation 78 (46.7%) 50 (52.1%) 0.44 Reoperation for bleeding 5 (3.0%) 9 (9.4%) 0.042 Pacemaker 0 (0%) 2 (1.1%) 0.13 Intubation 48 hours 11 (6.6%) 11 (11.5%) 0.17 Transfusion 123 (73.7%) 80 (83.3%) 0.093 Hospital mortality 10 (6.0%) 4 (4.2%) 0.78 Hospital and late 38 (22.8%) 23 (24.0%) 0.88 mortality Cardiopulmonary 77.6 20.2 132.5 32.3 0.0001 bypass, minutes Cross-clamp time, minutes 52.9 16.2 101.7 25.3 0.0001 AVR aortic valve replacement; CABG coronary artery bypass graft surgery; TIA-CVA transient ischemic attack cerebrovascular accident. Risk Factors for Postoperative TIA-CVA By univariate analysis, necessity for AVR (p 0.04), ejection fraction less than 40% (p 0.17), lower indexed valve area (p 0.14), higher transaortic valve gradient (p 0.07), CPB time (p 0.02), cross-clamp time (p 0.06), and Parsonnet score (p 0.01) showed p values less than 0.20 and were included in the multivariate model. By multivariate analysis, only cross-clamp time (odds ratio [OR] 1.02; 95% confidence interval [CI]: 1.00 to 1.03, p 0.02) and Parsonnet score (OR 1.09; 95% CI: 1.00 to 1.19, p 0.05) were significantly associated with a postoperative TIA-CVA. Risk Factors for Operative Mortality By univariate analysis, presence of angina Canadian Cardiovascular Society class III or IV (p 0.1), emergent surgery (p 0.05), body mass index greater than 30 (p 0.12), COPD (p 0.17), preoperative renal failure (p 0.0001), peripheral vascular disease (p 0.05), preoperative myocardial infarct (p 0.10), perioperative reoperation for bleeding (p 0.03), postoperative septicemia (p 0.008), and intubation longer than 48 hours (p 0.0001) were associated with perioperative mortality (p 0.20) and were included in a multivariate analysis. By multivariate analysis, preoperative renal failure (OR 7.64, 95% CI: 2.25 to 25.97, p 0.001) and intubation longer than 48 hours (OR 11.10, 95% CI: 3.14 to 39.19, p 0.0002) were significantly associated with perioperative mortality. 1497 late mortality were left ventricular ejection fraction less than 40% (p 0.01), preoperative NYHA III or IV (p 0.0001), preoperative Canadian Cardiovascular Society angina class III or IV (p 0.03), COPD (p 0.003), preoperative renal failure (p 0.0001), peripheral vascular disease (p 0.06), and intubation longer than 48 hours (p 0.01). These factors were then studied by multivariate analysis, which showed left ventricular ejection fraction less than 40% (OR 3.35, 95% CI: 1.17 to 9.65, p 0.02), NYHA III or IV (OR 2.37, 95% CI: 1.37 to 4.09, p 0.002), COPD (OR 2.26, 95% CI: 1.17 to 4.38, p 0.02), and preoperative renal failure (OR 3.03, 95% CI: 1.45 to 6.31, p 0.003) to be significantly associated with late mortality. Figure 2 shows the negative additive impact on survival of preoperative renal failure and COPD. Whereas patients with renal failure or COPD alone have a 58.7% and 44.8% 5-year survival, respectively, patients with combined COPD and preoperative renal failure have a 5-year survival of 25.0% (p 0.003 versus COPD alone). Figure 3 illustrates the freedom from AVR for both groups. At 5 years, 97.8% 1.2% of CABG patients and 98.9% 1.1% of CABG AVR patients (p 0.13) were free from reoperation for AVR. Seven of the 167 patients in the CABG group required reoperation for AVR at a mean of 3.8 2.7 years. Among these, 6 patients were reoperated on for symptoms due to severe aortic stenosis, and 1 patient was reoperated on 1 year postoperatively for coronary graft occlusion, at which time a decision was made to replace the moderate aortic stenosis. Within the CABG AVR group, 1 patient required reoperation for prosthetic endocarditis 3 months postoperatively. Among the CABG alone group, risk of reoperation was significantly increased in the presence of vascular disease (OR 6.98,; 95% CI: 0.97 to 49.99, p 0.053) and a higher mean gradient at operation (OR 1.40, 95%CI: 1.12 to 1.76, p 0.0037. Receiver operating characteristic curve analysis identified 26 mm Hg as the peak and 15 mm Hg as ADULT CARDIAC Follow-Up and Long-Term Outcomes Mean follow-up was similar for both groups: CABG 4.4 2.9 years, and CABG AVR 4.5 3.1 years (p 0.91). Figure 1 depicts comparable survival (including perioperative deaths) for both groups. At 5 years, 64.2% 4.3% of CABG patients and 62.3% 5.5% of CABG AVR patients were alive. By univariate analysis, risk factors for Fig 1. Survival of patients undergoing coronary artery bypass graft surgery (CABG) only (solid line) and CABG plus aortic valve replacement (AVR [dashed line]).

ADULT CARDIAC 1498 DAGENAIS ET AL Ann Thorac Surg MODERATE AORTIC STENOSIS IN OLDER CABG PATIENTS 2010;90:1495 1500 Fig 2. Survival function of preoperative renal failure (broken line), chronic obstructive pulmonary disease (COPD [dotted line]), renal failure plus COPD (dashed line), and no renal failure or COPD (solid line). the mean aortic valve gradients cutoff at which a higher rate of reoperation was observed. Comment The management of moderate aortic stenosis at the time of CABG is highly debated [5, 6]. The problem is even more complex for elderly patients owing to an increased operative risk with aging and the difficulty in assessing whether the patient will outlive the natural history of the aortic valve disease. The current study was designed to evaluate perioperative risk factors of combining AVR and CABG, establish prognostic factors of long-term survival, and assess reoperation rates for elderly patients with moderate aortic stenosis and CABG. Although perioperative mortality did not differ by performing concomitant AVR, morbidity was significantly enhanced. The CABG AVR was linked to a significant increase in postoperative TIA-CVA and renal failure. Interestingly, although concomitant AVR was associated with postoperative TIA-CVA by univariate analysis, multivariate analysis established length of cross-clamp time and higher Parsonnet score as the sole significant predictors of postoperative TIA-CVA. The prolonged CPB time may possibly account for the higher rate of mediastinitis and reoperation for bleeding observed in the CABG AVR group. The incremental risk of morbidity of performing a concomitant AVR and CABG has been reported by others [7 9]. Furthermore, presence of preoperative renal failure and intubation longer than 48 hours were significantly associated with perioperative mortality, thus delineating the importance of considering patient s comorbidities in the decision process to perform a concomitant AVR in elderly patients with moderate aortic stenosis and CABG, an opinion supported by other authors [10]. No survival benefit was observed by performing a concomitant AVR in this series. This finding is corroborated by other reports [7, 11]. Long-term survival was significantly decreased by the presence of a preoperative ejection fraction less than 40%, preoperative NYHA III or IV/IV, COPD, and preoperative renal failure. Furthermore, combining COPD and renal failure had an additive negative impact on long-term survival. At 5 years, 97.8% of patients undergoing CABG alone were free from AVR. Other studies have reported higher rates of reoperation at 5 years for patients with moderate aortic stenosis treated by CABG alone [12, 13]. Differences may be explained by the fact that elderly patients may not outlive the natural history of their aortic valve disease and may become symptomatic at higher gradients owing to their more sedentary lifestyle. Peak and mean aortic gradients greater than 26 mm Hg and 15 mm Hg, respectively, were identified as an increased risk of late reoperation for AVR in the CABG alone group, thus suggesting that concomitant AVR should be performed in healthy patients more than 70 years old who have such gradients. Reoperation for AVR after previous CABG poses a technical challenge. Initial studies have showed increased perioperative risks in CABG patients, necessitating remote AVR [14, 15]. More recently, excellent outcomes have been reported [15]. However, patient age and comorbidities remain significant risk factors increasing perioperative morbidity and mortality [16]. The recent advent of transcatheter aortic valve implantation has been shown feasible in a setting of previous CABG [17]. Results are encouraging, and with further technical refinements and longer follow-up, transcatheter aortic valve implantation may become the preferred treatment Fig 3. Freedom from aortic valve reoperation: coronary artery bypass graft surgery (CABG [solid line]) only, or CABG plus aortic valve replacement (AVR [dotted line]).

Ann Thorac Surg DAGENAIS ET AL 2010;90:1495 1500 MODERATE AORTIC STENOSIS IN OLDER CABG PATIENTS modality for elderly patients with comorbidities requiring AVR after a previous initial CABG strategy [18, 19]. Study Limitations Although both study groups showed similar characteristics, severity of aortic valve disease differed, and thus may have biased the interpretation of results. Furthermore, variables such as valve mobility and valvular calcification have been suggested as predictors of valve progression and were not analyzed in the present study. Identification of vascular disease as a risk factor of late AVR supports that other risk factors are implicated in the progression of aortic stenosis. Longer follow-up may yield a higher rate of annual reoperation rate for progression of aortic valve stenosis in patients with excellent survival. In addition, the long-term morbidity of implanting a bioprosthetic valve has not been considered in the current study. Conclusions The performance of concomitant AVR in elderly patients with moderate aortic stenosis undergoing CABG surgery does not increase perioperative mortality but is associated with a higher incidence of postoperative complications while adding no significant benefit for long-term survival. Outcome is especially poor among patients with significant comorbidities. Moreover, late reoperation for remote AVR remains low after CABG alone in this subset of patients. Thus, elderly patients presenting for CABG with moderate aortic stenosis with significant comorbidities such as renal failure or COPD should undergo CABG alone, owing to an increased perioperative risk and decreased postoperative survival. However, patients more than 70 years old with minimal comorbidities and with good life expectancy should be considered for concomitant AVR especially in the presence of aortic valve gradients above 26/15 mm Hg. A better understanding of the variables influencing the progression of the aortic stenosis may allow us to better identify the patients who would benefit from a concomitant AVR approach. Transcatheter aortic valve implantation may become a late bailout strategy in elderly patients with comorbidities presenting with a late symptomatic aortic stenosis after an initial CABG alone approach. References 1. Phillips BJ, Karavas AN, Aranki SF, et al. Management of mild aortic stenosis during coronary artery bypass surgery. J Card Surg 2003;18:507 11. 2. Smith W, Ferguson B, Ryan T, Landolfo CK, Peterson ED. Should coronary artery bypass graft surgery patients with 1499 mild or moderate aortic stenosis undergo concomitant aortic valve replacement? J Am Coll Cardiol 2004;44:1241 7. 3. Hochrein J, Lucke JC, Harrison JK, et al. Mortality and need for reoperation in patients with mild-to-moderate asymptomatic aortic valve disease undergoing coronary artery bypass graft alone. Am Heart J 1999;138:791 7. 4. Sundt TM, Murphy SF, Barzilai B, et al. Previous coronary artery bypass grafting is not a risk factor for aortic valve replacement. Ann Thorac Surg 1997;64:651 8. 5. Wong PS, Davies SW, Youhana A, Wright JE, Magee PG. Coronary artery bypass surgery and minor aortic stenosis to replace or not to replace? J Heart Valve Dis 1993;2: 649 52. 6. Tam JW, Masters RG, Burwash IG, Mayhew AD, Chan KL. Management of patients with mild aortic stenosis undergoing coronary artery bypass grafting. Ann Thorac Surg 1998; 65:1215 9. 7. Sareyyupoglu B, Sundt TM, Schaff HV, et al. Management of mild aortic stenosis at the time of coronary artery bypass surgery: should the valve be replaced? Ann Thorac Surg 2009;88:1224 31. 8. Collins JJ, Aranki SF. Management of mild aortic stenosis during coronary artery bypass graft surgery. J Card Surg 1994;9:145 7. 9. Smith WT, Ferguson TB, Ryan T, Landolfo CK, Peterson ED. Should coronary artery bypass graft surgery patients with mild or moderate aortic stenosis undergo concomitant aortic valve replacement? J Am Coll Cardiol 2004;44:1241 7. 10. Brunvand H, Offstad J, Nitter-Hauge S, Svennevig JL. Coronary artery bypass grafting combined with aortic valve replacement in healthy octogenarians does not increase postoperative risk. Scand Cardiovasc J 2002;36:297 301. 11. Fiore AC, Swartz MT, Naunheim KS, et al. Management of asymptomatic mild aortic stenosis during coronary artery operations. Ann Thorac Surg 1996;61:1693 7. 12. Ahmed Alsir AM, Graham ANJ, Lovell D, O Kane HO. Management of mild to moderate aortic valve disease during coronary artery bypass grafting. Eur J Cardiothorac Surg 2003;24:535 40. 13. Hochrein J, Lucke JC, Harrison JK, et al. Mortality and need for reoperation in patients with mild-to-moderate asymptomatic aortic valve disease undergoing coronary artery bypass graft alone. Am Heart J 1999;138:791 7. 14. Odell JA, Mullany CJ, Schaff HV, Orszulak TA, Daly RC, Morris JJ. Aortic valve replacement after previous coronary artery bypass grafting. Ann Thorac Surg 1996;62:1424 30. 15. Verhoye JP, Merlicco F, Sami IM, et al. Aortic valve replacement for aortic stenosis after previous coronary artery bypass grafting: could early reoperation be prevented? J Heart Valve Dis 2006;15:474 8. 16. Khaladj N, Shrestha M, Peters S, et al. Isolated surgical aortic valve replacement after previous coronary artery bypass grafting with patent grafts: is this old-fashioned technique obsolete? Eur J Cardiothorac Surg 2009;35:260 4. 17. Walther T, Falk V, Borger MA, et al. Transapical aortic valve implantation in patients requiring redo surgery. Eur J Cardiothorac Surg 2009;36:231 5. 18. Ye J, Cheung A, Lichtenstein SV, et al. Transapical transcatheter aortic valve implantation: 1-year outcome in 26 patients. J Thorac Cardiovasc Surg 2009;137:167 73. 19. Mack MJ. Transapical aortic valve implantation: is it ready for prime time? Eur J Cardiothorac Surg 2009;36:229 30. ADULT CARDIAC DISCUSSION DR J. SCOTT RANKIN (Nashville, TN): Very nice study. I have a couple of comments to make. First, when we looked at long-term outcomes after coronary bypass surgery in the Duke data bank over 20 years, the number one predictor of late mortality was mild to moderate valve disease not treated during coronary bypass, and that is really a consistent finding (Rankin JS, et al. Techniques and benefits of multiple internal mammary artery bypass at 20 years of follow-up. Ann Thorac Surg 2007;83: 1549 55; Table 2). Thus, with more follow-up information here, this analysis might turn out differently.

ADULT CARDIAC 1500 DAGENAIS ET AL Ann Thorac Surg MODERATE AORTIC STENOSIS IN OLDER CABG PATIENTS 2010;90:1495 1500 Second, I hate to see studies that subgroup patients into the elderly or whatever subgroup. It is better in our models to have all ages included and then specifically look with formal treatment interactions at the age of interest. We lose statistical power when we just look at a subgroup like this. But it is clear, every time we look at long-term survival in the Duke data bank that the presence of mild aortic stenosis is a significant late risk factor for mortality. Perhaps you have just not followed your patients long enough. So there is a third surgical option, and that is valve repair. Over the years, I have been prone to repair the aortic valves with mild to moderate calcific stenosis during other procedures. It is clear that ultrasonic decalcification for heavily calcified aortic valves doesn t work. About a third of patients developed scarred retracted leaflets and aortic insufficiency (Leithe ME et al. Surgical aortic valvuloplasty using the Cavitron ultrasonic surgical aspirator: an invasive hemodynamic follow-up study. Cath Cardiovasc Diag 1991;24:16 21). But for this group with mild to moderate AS, the patient often will have just a couple of strategically placed spicules of calcium that can be debrided out with the ultrasonic device, and the valve will open a lot better on the postoperative transesophageal echocardiogram. It is a quick, easy operation. And I can tell you, they don t come back with progressive aortic stenosis. So I just want to suggest valve repair as another option: a quick, simple, and effective operation, with a low complication rate and good long-term results. DR DAGENAIS: Thank you for these comments. For sure, I don t think I will ever have 25-year follow-up in these kind of patients, but one thing that was a bit surprising was the midterm survival in these patients, which is 63% at 5 years. So this is still a subgroup of patients that still has a pretty good life expectancy. I think this has to be taken into account. But the other take-home message is if you have significant comorbidities, patients more than 70 years old don t live very long at the midterm, and these are the patients for whom we should avoid replacing their valve, especially that they have a higher perioperative complication rate risk. On the other hand, I do not have any experience with decalcification. DR MARTIN CZERNY (Berne, Switzerland): Francois, thank you very much. I think besides hemodynamics there is another factor that should be considered in these patients, and this is the degree of calcification of the valve. In our experience, we have clearly seen that the very patients who do have severe calcification despite moderate hemodynamic compromise are the ones who will come back. DR DAGENAIS: I think that is a very good point in the sense that there are different markers now we are starting to identify that may have an impact on the rate of progression of an aortic stenosis. Better characterization of these markers will certainly allow us to better decide which patient will require aortic valve replacement. DR WILLIAM T. CAINE (Salt Lake City, UT): I am curious, your operative mortality results are certainly different from the Society of Thoracic Surgeons (STS) database, and I am wondering how you account for that. Do you think it is a selection bias when the surgeon made the decision in the operating room, or how do you account for that difference? DR DAGENAIS: Well, I think if you look at the observed and expected mortality, we are not much different than what we should have expected from the STS database. DR CAINE: My point is in the STS database operative mortality for an AVR CABG is significantly higher than the operative mortality for CABG alone, but you didn t see that difference. So that is what I am trying to figure out. Why do you think you didn t see that difference? DR DAGENAIS: Okay. I better understand your point. For sure, one can not eliminate a bias introduced by the surgeon s decision to replace or not the valve. On the other hand, it is difficult to compare this cohort to a CABG AVR with severe aortic stenosis. Issues such as myocardial protection in a hypertrophied heart may impact outcome. One can, however, consider with our study that performing an AVR for mild to moderate aortic stenosis in a more than 70-year-old CABG patient may be done with minimal added mortality for a patient with significant comorbidities. DR WALTER MERRILL (Jackson, MS): Let me ask another question. I did note in your presentation that several patients received a late aortic valve replacement, but I don t think you mentioned what were the results of those operations and whether there was a high operative mortality or not. DR DAGENAIS: Like I said, there were 7 patients, 6 in the CABG alone group, 1 in the CABG AVR, and all these patients survived their reoperation. DR MERRILL: I have one follow-up question. Was there any difference in long-term symptomatic status or functional capacity of the patients in either group? DR DAGENAIS: Well, those are things we haven t yet assessed, and I think that is one point, how the patients evolved on a clinical status and also on an echocardiographic status. We are now presently taking all these echocardiograms and trying to find what kind of progression we can also get in these patients.