Aortic Valve Replacement With or Without Coronary Artery Bypass Graft Surgery: The Risk of Surgery in Patients ≥80 Years Old

ORIGINAL ARTICLES

Aortic Valve Replacement With or Without Coronary Artery Bypass Graft Surgery: The Risk of Surgery in Patients >80 Years Old Andrew Maslow, MD,* Paula Casey, RPN,† Athena Poppas, MD,‡ Carl Schwartz, MD,* and Arun Singh, MD† Objective: The purpose of this study was to evaluate the outcomes for elderly (>80 years) patients undergoing aortic valve replacement (AVR) with or without coronary artery bypass graft surgery (AVR/CABG). The authors hypothesized that the mortalities of AVR and AVR/CABG are lower than that predicted by published risk scores. Design: A retrospective analysis of data from a singlehospital database. Setting: Single tertiary care, private practice. Participants: Consecutive patients undergoing AVR or AVR/CABG. Measurements: Two hundred sixty-one elderly (>80 years) patients undergoing isolated AVR (145) or AVR/CABG (116) were evaluated. The majority (94.6%) underwent AVR for aortic valve stenosis. Outcomes were recorded and compared between the 2 surgical procedures with predicted mortalities based on published risk assessment scoring systems.

Results: The overall short-term mortality for the elderly group was 6.1% (AVR 5.5% and AVR/CABG 6.9%). The median long-term survival was 6.8 years. There were no significant differences in either morbidity or mortality between the AVR and AVR/CABG groups. Although predicted mortalities were similar for each surgical procedure, they overestimated observed outcome by up to 4-fold. Conclusions: Short- and long-term mortality was low for this group of elderly patients undergoing AVR or AVR/CABG and not significantly different between the 2 surgical groups. Predicted outcomes were worse than that observed, consistent with the hypothesis, and supportive of a more aggressive surgical treatment for aortic valve disease in the elderly patient. © 2010 Elsevier Inc. All rights reserved.

L

paring observed outcomes to predicted outcomes based on risk assessment from a large database would improve comparisons among different investigations as well as among institutions.18-22 The present study aimed to evaluate the short- and long-term surgical outcomes in elderly patients ⱖ80 years old undergoing either isolated AVR or AVR with CABG surgery (AVR/ CABG). The authors hypothesized that surgical outcome for elderly patients with severe aortic valve (AoV) disease is similar for the 2 procedures and better than that predicted by known risk assessment scores.

IFE-SAVING SURGERIES such as aortic valve replacement (AVR), with or without coronary artery bypass graft (CABG) surgery, are underutilized in the elderly (ⱖ80 years) patient.1-3 For more than 50% of elderly patients with severe aortic valve disease, the risk of surgery is overestimated, and often these patients are not offered aortic valve replacement.3 Seventy percent of these patients have documented symptoms, with a 2-year survival without surgery less than 50%.3,4 A major reason for the underutilization of AVR (⫾CABG) is the lack of definitive and consistent outcome data, yielding misconceptions, among referring clinicians, of the morbidity and mortality surrounding surgery. Previously reported hospital or 30-day mortality for patients ⱖ80 years old varied from 5% for isolated AVR to greater than 20% for combined AVR with CABG surgery.5-17 Although the varied outcome data may reflect the successes and failures of different cardiac centers over the past 2 decades, it could also reflect different patient populations characterized by differing demographic data. Com-

From the Departments of *Anesthesiology, †Surgery, Division of Cardiothoracic Surgery, and ‡Medicine, Division of Cardiology, The Warren Alpert Medical School of Brown University, Providence, RI. Presented in part at the American College of Cardiology Meeting, Chicago, IL, March 2008. Address reprint requests to Andrew Maslow, MD, Department of Anesthesiology, Rhode Island Hospital, Davol 129, 593 Eddy Street, Providence, RI 02903. E-mail: [email protected] © 2010 Elsevier Inc. All rights reserved. 1053-0770/10/2401-0005$36.00/0 doi:10.1053/j.jvca.2009.07.010 18

KEY WORDS: aortic valve replacement, coronary artery bypass graft, elderly, mortality, morbidity, cardiopulmonary bypass, short-term

METHODS A retrospective review of the cardiac surgical database from a single institution between 1996 and 2006 was performed to evaluate the early and late outcomes for elderly patients (ⱖ80 years) after either isolated AVR or AVR/CABG. All patients presented with significant aortic valve dysfunction (aortic valve stenosis [AS] or regurgitation [AI]). Patients with other valvular heart disease requiring additional valve surgery were excluded. Two hundred sixty-one patients were included. The indication for surgery was severe and symptomatic aortic valve disease. This was defined as having presented with one or more symptoms attributable to AS (angina, congestive heart failure, syncope, and reduced exercise tolerance) or AI (congestive heart failure, dyspnea on exertion, and reduced exercise tolerance) in conjunction with severe valve dysfunction/pathology. Severe AS was defined as an indexed aortic valve area, measured by echocardiogram, of ⱕ0.65 cm2/m2. Severe AI was defined by a color Doppler proximal jet width ⱖ40% of the width of the left ventricular outflow tract. Left ventricular ejection fraction (LVEF) was obtained from the preoperative TTE or cardiac

Journal of Cardiothoracic and Vascular Anesthesia, Vol 24, No 1 (February), 2010: pp 18-24

AORTIC VALVE REPLACEMENT IN ELDERLY PATIENTS

catheterization. Significant coronary artery stenosis was defined as a narrowing of ⱖ70% of the native luminal diameter by coronary artery angiography. Demographic and operative data were collected as defined by the Society of Thoracic Surgeons (STS) database18 and the EuroSCORE.19-22 Patients requiring surgery other than isolated AVR or AVR/CABG were excluded. Prespecified perioperative morbidity data collected included perioperative atrial fibrillation (prolonged [ⱖ1 hour] and/or requiring treatment), stroke (new gross neurologic deficit ⱖ24 hours after surgery), acute renal insufficiency (ⱖ50% increase in serum creatinine and a peak serum creatinine ⱖ2.0 mg/dL), renal failure (requiring dialysis), reoperation for bleeding, prolonged (greater than 24 hours) ventilation, pneumonia (new radiologic lesion associated with evidence of infection), and acute respiratory distress syndrome. Perioperative lung injury was defined as prolonged mechanical ventilation, pneumonia, and/or acute respiratory distress syndrome. Short-term mortality was defined as death within 30 days of surgery or during the same hospital stay. Anesthetic, surgical, and perioperative managements were standardized per divisional protocols. All patients were monitored with standard noninvasive monitors, an intra-arterial catheter, a pulmonary artery catheter, and intraoperative transesophageal echocardiography. All surgical procedures were performed though a median or hemisternotomy with cardiopulmonary bypass (CPB) under systemic normothermic perfusion (37°) and cold cardioplegic arrest. Hematocrits were maintained above 21% while on CPB and then ⬎24% after separation from CPB. During valvular surgery, continuous cardioplegia was delivered directly through a self-inflating balloon catheter (Vitalcor, Westport, IL) placed into the coronary ostia and/or via the newly constructed vein grafts. Management during and after CPB was standardized as previously described.23 For all patients, manipulation of the ascending aorta was minimized by using a single cross-clamp technique. For patients undergoing CABG surgery, distal anastomoses were performed before aortotomy. Proximal vein graft anastomoses were performed after AVR and closure of aortotomy. Perioperative management goals were standardized. Hemodynamic goals included a systemic blood pressure ⱖ100/50 mmHg, a heart rate ⱕ100 beats/min, central filling pressures within 20% of pre-CPB data, and a cardiac index of ⱖ2.5 L/min/m2. Other goals included weaning from mechanical ventilation within 24 hours and mobilization of patients. Categoric data, such as sex, were compared between surgical procedure groups (AVR v AVR/CABG) using the chi-square test or Fisher exact test. Continuous data, such as age, were compared between surgical procedure groups using the Student t test or Kruskal-Wallis test. Survival was estimated by using the Kaplan-Meier method, and comparisons of these distributions were made using the log-rank test. Observed mortality data were compared with the expected (observed: expected [O:E] ratio), the latter value determined by the STS risk calculation assessment (STS-PROM)18 and by the logistic EuroSCORE.19-22 Independent predictors of short- and long-term mortality were determined with a multivariate Cox model adjusted for significant (p ⬍ 0.05) univariate predictors of mortality. All p values ⬍0.05 were considered statistically significant. For aortic cross-clamp time and the logistic EursoSCORE, variables included categoric analysis based on previous data (logistic EuroSCORE low risk [ⱕ10%], medium risk [10% ⬍ ES ⬍ 20%], and high risk [ES ⱖ 20%] groups24; aortic cross-clamp [AoXCl] ⱖ75 minutes25). A similar categoric analysis was performed for the STS-PROM data; however, this was not based on any previous reference (low ⱕ 7, 7 ⬍ medium ⬍ 13, and high ⱖ 13).

19

Table 1. Demographic Data ⱖ80 Years Old

Age (y [SD]) Sex (%) Male (n) Female (n) Hypertension (n) (%) Tobacco (n) (%) COPD (n) (%) NYHA (mean) II (n) (%) III (n) (%) IV (n) (%) Atrial fibrillation (n) (%) LVEF (% [SD]) LVEF ⬍45% (n) (%) MI (n) (%) DM (n) (%) Renal dysfunction (n) (%) AS (n) (%) LVH (n) (%) Reoperation (n) (%)

Overall (n ⫽ 261)

AVR (n ⫽ 145)

83.1 (2.96)

83.1 (3.18)

83.2 (2.69)

142 (54.4) 119 (45.6) 204 (78.2) 109 (41.8) 36 (13.8)

70 (48.3) 75 (51.7) 114 (78.6) 57 (39.3) 16 (11.0) 3.27 (0.51) 10 (6.9) 96 (66.2) 39 (26.9) 3 (2.1) 53.7 (8.9) 21 (14.4) 18 (12.4) 26 (17.9) 12 (8.3) 138 (95.2) 143 (98.6) 9 (6.2)

72 (62.1)* 44 (37.9) 90 (77.6) 52 (44.8)* 30 (25.9)* 3.47 (0.52)† 2 (1.7) 59 (50.9) 55 (47.4) 5 (4.3) 47.7 (20.8)* 26 (22.4) 42 (36.2)† 33 (28.4)* 12 (10.3) 109 (94.0) 113 (97.4) 5 (4.3)

12 (4.6) 155 (59.4) 94 (36.0) 8 (3.1) 51 (13.4) 47 (18) 60 (23) 59 (22.6) 24 (9.2) 247 (94.6) 256 (98.0) 14 (5.4)

AVR/CABG (n ⫽ 116)

NOTE. Data are presented as number and percentage (n [%]) or mean and standard deviation (SD). Abbreviations: COPD, chronic obstructive pulmonary disease; AS, aortic stenosis; NYHA, New York Heart Association; LV, left ventricle; LVEF, left ventricular ejection fraction; LVH, left ventricular hypertrophy; MI, myocardial infarction; DM, diabetes mellitus. *p ⬍ 0.05. †p ⬍ 0.01.

RESULTS

Over a 10-year period, a total of 1,111 patients were identified as having undergone AVR or AVR/CABG for symptomatic AoV disease; 850 were ⬍80 years old. and 261 (23.5%) were ⱖ80 years old. Of these 261 patients, 145 patients had AVR, and 116 had AVR/CABG. These 261 patients constituted the study group (Table 1). Preoperative valvular hemodynamics revealed an indexed aortic valve area of ⱕ0.6 cm2/m2 for all patients operated on for aortic valve stenosis (94.6% of the study population). For 5.4%, AVR was performed based on the finding of ⱖ3⫹ (ⱖmoderate/ severe) aortic valve insufficiency. There were no differences in the number of cases performed each year during the study period. The mean and median age for surgery was 83.0 and 82.0 years (80-98), respectively, (range, 80-98). Ninety-five percent of patients were New York Heart Association class III/IV (249/261). The median follow-up for all patients was 6.3 years. Differences between those undergoing AVR versus those undergoing AVR/CABG included a greater percentage of male sex (p ⫽ 0.03), chronic obstructive pulmonary disease/tobacco use (p ⬍ 0.05), diabetes (p ⫽ 0.05), myocardial infarction (p ⬍ 0.01), a higher New York Heart Association class (p ⫽ 0.001), and a lower LVEF (p ⬍ 0.05) for those receiving AVR/CABG. Bioprosthetic valves were placed in almost all elderly pa-

20

MASLOW ET AL

tients, the size ranging between 19 and 23 mm in more than 90% (Table 2). Approximately 31% of cases were urgent/ emergent in nature and did not differ between the 2 surgical groups. Durations of aortic cross-clamp and cardiopulmonary bypass were significantly longer in the AVR/CABG group. In this group, 89% of the patients received 1 or 2 distal coronary artery bypasses, whereas 11% received 3 or 4. For all patients undergoing AVR and AVR/CABG (n ⫽ 1,111) short-term mortality was 3.9%. The younger age group (⬍80 years old) had an overall mortality of 3.3%. There were no significant mortality differences between the surgical groups in the younger patient populations (AVR 2.8% v AVR/CABG 4.0%). For the elderly group, the overall short-term mortality was 6.1% (16/261) and was significantly greater than the younger population (p ⬍ 0.01). Short-term mortality for isolated AVR (5.5% [8/145]) was not significantly different from that of AVR/CABG (6.9% [8/116], p ⫽ 0.39) (Table 3 and Fig 1). There were no intraoperative deaths. No hospital death occurred as a result of heart failure. Life support was withdrawn at the request of the family as a result of debilitating neurologic injury (n ⫽ 5), the need for hemodialysis (n ⫽ 4), or prolonged mechanical ventilation (n ⫽ 7). The short-term mortality did not change significantly over the 10 years of the study. Survival at 2 years was 80% (83% for isolated AVR and 77% for AVR/CABG). The long-term median survival was 6.8 years (7.3 years for isolated AVR and 6.2 years for AVR/ CABG).

Table 2. Operative Details ⱖ80 Years Old All (n ⫽ 261)

Operative priority (%) Elective 176 (67.4) Urgent 80 (30.7) Emergent 5 (1.9) Prosthesis (%) Biological 257 (98.5) Mechanical 4 (1.5) Prosthetic valve size Mean (mm) 21.4 (1.9) ⱕ19 mm (%) 66 (25.3) 21 mm (%) 102 (39.1) 23 mm (%) 73 (28.0) ⱖ25 mm (%) 20 (7.7) Coronary artery bypasses (%) No bypasses Single Double Triple Quadruple AoXCl (min ⫾ SD) 92.1 (45.5) CPB (min ⫾ SD) 115.4 (55.0)

Isolated AVR (n ⫽ 145)

AVR/CABG (n ⫽ 116)

99 (68.3) 43 (29.6) 3 (2.1)

77 (66.4) 37 (31.9) 2 (1.7)

142 (97.9) 3 (2.1)

115 (99.1) 1 (0.9)

21.3 (2.0) 42 (29.0) 53 (36.6) 41 (28.2) 9 (6.2)

21.5 (1.8) 24 (20.7) 49 (42.2) 32 (27.6) 11 (9.5)

145 (100) 0 0 0 0 72.0 (27.7) 117.3 (49.9)

0 65 (56.0) 38 (32.8) 12 (10.3) 1 (0.9) 92.42 (37.4)* 143.9 (60.4)*

NOTE. Data are presented as the number and percentage (n [%]) with the exception of mean prosthetic valve size, aortic cross-clamp, and CPB times, which are presented as minutes and standard deviation (SD). *p ⬍ 0.001.

Table 3. Morbidity and Mortality of Cardiac Surgical Patients >80 Years of Age Undergoing Either AVR or AVR/CABG From 1996 to 2006

Hospital mortality Stroke Atrial fibrillation Reoperation for bleeding Acute renal failure Renal failure (dialysis) Prolonged ventilation Pneumonia Acute respiratory distress syndrome Any lung injury Sepsis

All (n ⫽ 261) (%)

AVR (n ⫽ 145) (%)

AVR/CABG (n ⫽ 116) (%)

16 (6.1) 8 (3.1) 81 (31.0) 12 (4.6) 22 (8.4) 7 (2.7) 35 (13.4) 5 (1.9)

8 (5.5) 5 (3.4) 50 (34.4) 6 (4.1) 12 (8.3) 2 (1.4) 16 (11.0) 3 (2.1)

8 (6.9) 3 (2.6) 31 (26.7) 6 (5.2) 10 (8.6) 5 (4.3) 19 (16.4) 2 (1.7)

4 (1.5) 36 (13.8) 15 (5.7)

2 (1.4) 17 (11.7) 7 (4.8)

2 (1.7) 19 (16.4) 8 (6.9)

Pre- and perioperative variables associated with short- and long-term mortalities are listed in Tables 4 and 5. Perioperative morbidity was associated with short-term mortality. Specifically, this included acute perioperative renal failure (p ⬍ 0.001), reoperation for bleeding (p ⬍ 0.001), sepsis (p ⬍ 0.0001), and perioperative lung injury (p ⬍ 0.0001). Variables associated with long-term mortality included preoperative myocardial infarction (p ⬍ 0.05), COPD (p ⬍ 0.01), and postoperative sepsis (p ⬍ 0.001), stroke (p ⫽ 0.002), and lung injury (p ⬍ 0.01). Short-term morbidity was similar for patients undergoing AVR versus AVR/CABG. Hospital morbidities are listed in Table 3. The most common complication was atrial fibrillation followed by any kind of lung injury. There were no predictors of preoperative bleeding. Multivariate predictors of perioperative sepsis included preoperative

Fig 1. Overall survival by surgical procedure. The probability (percentage) of survival up to 10 years after surgery for all elderly (>80 years) patients undergoing AVR or AVR/CABG.

AORTIC VALVE REPLACEMENT IN ELDERLY PATIENTS

21

Table 4. Short Term: Univariate and Multivariate Predictors of Short-term Outcome of Elderly Patients Undergoing AVR or AVR/CABG Variable Univariate p Value

STS-PROM actual Logistic EuroSCORE actual Chronic obstructive pulmonary disease Preoperative renal dysfunction Urgent/emergent surgery Reoperation for bleeding Postoperative sepsis Acute perioperative renal dysfunction Postoperative pneumonia Prolonged mechanical ventilation Postoperative lung dysfunction

Multivariate Odds Ratio

p Value

⬍0.05 ⬍0.05

1.54 1.20

0.12 0.35

0.03 0.04 0.0004 0.0004 ⬍0.0001

0.67 0.88 0.55 3.55 2.98

0.12 0.70 0.87 0.0005 ⬍0.0001

⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001

1.02 2.56 0.44 2.32

0.96 0.04 0.28 ⬍0.05

renal dysfunction (p ⬍ 0.005), acute worsening of renal function (p ⬍ 0.001), and a prolonged aortic cross-clamp (⬎75 minutes) (p ⬍ 0.05). Predictors of perioperative lung dysfunction included preoperative tobacco abuse (p ⬍ 0.05), prolonged aortic cross-clamp time (p ⬍ 0.01), prolonged CPB (p ⬍ 0.01), and acute worsening of renal function (p ⬍ 0.05). The authors found no predictors of stroke in this study. Acute worsening of renal function was associated with preoperative renal dysfunction (p ⬍ 0.05), sepsis (p ⬍ 0.05), pneumonia (p ⬍ 0.05), and lung dysfunction (p ⬍ 0.05). The mean STS-PROM mortality prediction for the entire study population was 10.6% ⫾ 5.7% (Table 6). The STSPROM was similar between the 2 surgical groups (isolated AVR 10.7% ⫾ 5.4%; AVR CABG 10.4% ⫾ 6.1%). This yields an overall O:E ratio of 0.51 and 0.66, respectively. The mean STS-PROM was not significantly different from year to year or between the first and latter halves of the study period. The STS-PROM was not significantly different between short- or long-term survivors and nonsurvivors. The STSPROM score was not a predictor of short- and long-term mortality. The mean logistic EuroSCORE for the study population was 18.9% ⫾ 12.9%. The logistic EuroSCORE was similar between the 2 surgical groups (AVR 17.96 ⫾ 2.77 v AVR/CABG 20.19 ⫾ 13.08, p ⫽ 0.17). These yield an O:E ratio of 0.31 and 0.34 for isolated AVR and AVR/CABG, respectively. Although for the entire study group, a higher logistic EuroSCORE was significantly associated with shortterm mortality (18.53 ⫾ 12.8 v 25.39 ⫾ 13.3, p ⬍ 0.05), it was not a significant predictor of short- or long-term mortality. Similar to the STS-PROM, data did not differ between years or over the duration of the study. When the predictive scores were categorized based on STSPROM into low-, medium-, and high-risk groups, there appeared to be little, if any, overestimation of observed mortality for the low- and medium-risk groups but not for the high-risk group (Table 3). However, at all levels of risk, the logistic EuroSCORE overestimated observed mortality (Table 6).

DISCUSSION

This retrospective analysis of a consecutive cohort of patients ⱖ80 years undergoing aortic valve surgery, with or without CABG surgery, found a short-term mortality of 6.1% and a long-term median survival of 6.6 years. Short-term and long-term morbidities and mortalities for isolated AVR were not significantly different from AVR/CABG, whereas observed short-term mortalities were lower than that predicted by known risk scoring analyses.18-22 The data presented here, and in the literature, prompt discussion regarding the wide range results reported for the 2 surgical procedures. Furthermore, they support a more aggressive referral strategy for elderly patients, especially those who are thought to be at a higher risk. Previously reported short-term mortalities for patients ⱖ80 years old undergoing AVR or AVR/CABG varied significantly between 5% and 26.5%.7-9,11,15,26-31 Although short-term mortality for isolated AVR varied around 8%,8-11,15-17,28,29,31,32 the majority of data report a significant increase when AVR was combined with CABG surgery.6,8,10,11,13,16,17,30,32,33 Although short-term mortality varied widely, the 5-year survival for the elderly population was consistently reported to be around 60%,6,7,9,13,14,15,27,29-31 with long-term mortalities more likely the result of malignancy, pneumonia, and stroke.29 Only a handful of studies reported similar mortalities for the 2 surgical procedures (AVR and AVR/CABG) in the elderly population,5,7,9 with 1 reporting short- and long-term benefits for the AVR/CABG group.7 In contrast to the present study, these investigations either reported higher short-term mortalities for both surgical procedures (⬎10%)5,7 or an increase in stroke for the AVR/CABG group (10.8% v 1.0% for isolated AVR).9 None of these studies included any risk assessment scores to help describe the risk level of the patient population. Referral for valve replacement depends on the assessment of preoperative function and the predicted risks and benefits of surgery. However, the wide range of reported outcomes and outcome predictors makes it difficult to predict who is

Table 5. Long-term: Univariate and Multivariate Predictors of Long-term Outcome of Elderly Patients Undergoing AVR or AVR/CABG Variable Univariate p Value

STS-PROM actual Logistic EuroSCORE actual Myocardial infarction Chronic obstructive pulmonary disease Preoperative renal dysfunction Cross-clamp time Cardiopulmonary bypass time Postoperative sepsis Perioperative new stroke Acute perioperative renal dysfunction Postoperative ARDS Postoperative pneumonia Prolonged mechanical ventilation Postoperative lung dysfunction

Multivariate Odds Ratio

p Value

⬍0.01 0.01 0.02

1.42 1.11 1.77

0.08 0.20 0.02

0.002 0.001 0.01 0.003 ⬍0.001 0.02

4.81 1.45 0.99 1.01 7.29 4.16

0.002 0.49 0.27 0.08 ⬍0.001 0.002

0.0002 0.02 0.0003 ⬍0.05 0.002

2.56 0.75 0.63 1.22 0.30

⬍0.01 0.67 0.40 0.29 0.03

22

MASLOW ET AL Table 6. Short-term Mortality as Categorized by the Logistic EuroSCORE According to Leontyvev et al24 EuroLog

Total (n ⫽ 261) Low (n ⫽ 71) Medium (n ⫽ 104) High (n ⫽ 86)

STS-PROM

Predicted

Observed

18.9% (⫾12.9) 7.8% (⫾1.3) 14.5% (⫾2.9) 33.6% (⫾12.5)

6.1% (n ⫽ 16) 2.8% (n ⫽ 2) 4.8% (n ⫽ 5) 10.4% (n ⫽ 9)

Total (n ⫽ 261) Low (n ⫽ 82) Medium (n ⫽ 71) High (n ⫽ 108)

Predicted

Observed

10.6% (⫾5.7) 5.6% (⫾0.9) 9.6% (⫾1.8) 17.8% (⫾5.5)

6.1% (n ⫽ 16) 6.1% (n ⫽ 5) 9.9% (n ⫽ 7) 3.7% (n ⫽ 4)

NOTE. A similar categorization was written for the STS-PROM data; however, this was arbitrarily determined (ie, no previous reference). Predicted data are presented as the number (n) in each group and the percent mortality ⫾ standard deviation. The observed data are presented as the percentage with the actual number of nonsurvivors in parentheses.

and who is not an appropriate surgical candidate.3,5,31 A large database of 4,743 elderly cardiac surgical patients reported that any preoperative comorbidity was associated with a 2-fold increase in major perioperative morbidity and mortality.31 By contrast, a smaller study found no preoperative predictors of short-term mortality.5 The reported risk factors for adverse outcomes among the elderly included increasing age,6,19-22,34 preoperative functional classification (NYHA III/IV),6,9,11,12,16,26 urgent or emergent surgery,6,8,10,11,16,19-22,34 a reduced preoperative LVEF,6,8,10,15,19-22,34,35 and baseline pulmonary6,9,34 or renal9,14,19-22,34,35 dysfunction. In the present study, no preoperative variable was associated with short-term mortality; however, myocardial infarction and chronic obstructive lung disease were associated with long-term mortality. Given the varying significance of preoperative variables to predict outcome, investigators have developed risk assessment scores, such as the STS-PROM and logistic EuroSCORE, to improve the preoperative risk assessment and add perspective (O:E ratio) to the recorded data.18-22,26,36-39 Of studies reporting low mortalities for the elderly patient undergoing AVR ⫾ CABG,9,17,25,36,40 only 2 described outcomes in light of predicted data.25,36 In a multicenter study including 928 hospitals and 4,291 elderly patients undergoing isolated AVR, the authors reported a short-term mortality ranging from 4% to 8% over a 10-year period.36 Alone, these mortality data were similar to that reported in the present study; however, when further described using the STS-PROM, it could appear as though a lower-risk population had been referred to surgery because the O:E ratio ranged between 0.8 and 1.2.36 In the second study, Urso et al25 recorded a mortality of 8% and a mean logistic EuroSCORE of 13.3% (O:E ratio of 0.6). Neither of these studies categorized the patients into low-, medium-, or highrisk groups. Urso et al did find that a logistic EuroSCORE ⱖ13.5% was associated with a significant increase in mortality. Geissler et al21 found the logistic EuroSCORE to be the best predictor of outcome among 6 scoring systems; however, it overestimated mortality by nearly 2-fold. Leontyvev et al24 categorized patients, using the logistic EuroSCORE, into low-, medium-, and high-risk groups. Although the logistic EuroSCORE appeared to be similar to the hospital mortalities of low- (7.3%/7.5%) and medium- (13.7%/12.6%) risk groups, the observed mortality for the higher-risk group (Logistic EuroSCORE ⱖ20%) was significantly lower than the predicted value (12.5% v 33%). In the present study, predicted mortalities using the logistic EuroSCORE overestimated observed mortalities overall and for each risk category as defined by Leontyvev

et al. However, when patients were categorized using the data, predicted and observed outcomes appeared similar in the lowand medium-risk groups, whereas the predicted value overestimated the observed outcome of the high-risk group by 3- to 4-fold. It is possible that these scoring systems are more accurate in lower-risk groups; however, they fail to accurately predict outcome for the higher-risk patient. In contrast to much of the literature, the present study did not find a significant outcome difference between the 2 surgical procedures. Furthermore, an outcome difference between the 2 surgical procedures was not predicted by either the STS-PROM or the logistic EuroSCORE. Outcome appeared to be more strongly related to differences in the patient risk profile or perioperative morbidities than whether or not they underwent AVR or AVR/CABG. In the present study, the CPB and AoXCl times were significantly longer in the combined procedure; however, longer procedural times were not associated with adverse outcomes. This is in contrast to only a few studies reporting associations between CPB and AoXCl times with outcome.16,25,40 The impact of the severity of coronary artery disease and/or number of bypasses performed on outcome after AVR/CABG is unclear.6,9 With the exception of 2 previous investigations,6,9 the extent of coronary artery disease for patients undergoing AVR (with or without CABG surgery) was not reported. In 1 study, involving 233 patients of all ages (mean age, 73 years) undergoing AVR/CABG, no significant correlation was found between the number of coronary vessels bypassed and shortterm outcome (short-term mortalities were 9.3%, 11.1%, and 7.6% for 1, 2, and 3-vessel CABG surgery, respectively).6 Akins et al9 reported a 27% incidence of coronary artery disease in patients undergoing isolated AVR (ie, no bypasses performed). The reported short-term mortalities for both AVR and AVR/ CABG were similar and ⬍10%. Based on these data and the data from the present study, the number of bypasses performed may not be as important as the sequelae of clinically significant coronary artery disease reflected by previous myocardial infarction and/or a reduced ejection fraction.6,8-10,15,19-22,34,35 Investigations reporting an outcome benefit with surgical revascularization (AVR/CABG) may include a relatively larger percentage of patients with clinically significant coronary artery disease presenting primarily for CABG surgery.5,7 In the present study, the indication for surgery was significant aortic valve disease. Therefore, a lack of contribution of CABG surgery may reflect a greater secondary clinical role of coro-

AORTIC VALVE REPLACEMENT IN ELDERLY PATIENTS

23

nary artery disease toward the patient’s cardiac and overall function. A relatively low mortality can be achieved for a high-risk group of elderly patients scheduled for AVR or AVR/CABG. It should, however, be anticipated that the elderly patient will have a greater complication rate and resource allocation compared with a younger cohort.4,33,41 Regardless, excellent outcomes should be anticipated despite high preoperative risk scores. The data presented here should be considered in light of newer “less invasive” treatments of aortic valve disease such as the percutaneous or transcatheter aortic valve implantation (TAVI).37 Caregivers have used both the STS-PROM and the logistic EuroSCORE to identify high-risk patients for whom these newer treatments may be of benefit.38,39 In these reports, high-risk patients were identified by an STS-PROM of 10.9% ⫾ 3.5% and/or a EuroSCORE of 33.1% ⫾ 13.7%, both of which were similar to that reported in the higher-risk groups in the present study.38,39 Mortalities reported with TAVI were 9.5% as compared with 6.3% in a cohort of similar risk patients undergoing conventional AVR.38 For a similar population, the mortality after TAVI has been reported to be as high as 23%.39 Although TAVI may offer benefits for specific populations, it is not yet apparent that it offers any mortality benefits over conventional AVR.30,38,39 In addition, long-term outcome is not yet known. The present study was a retrospective review of data obtained from the hospital’s STS database. Therefore, possible selection bias cannot be excluded. Nevertheless, preoperative risk assessment showed that a high-risk population was studied. However, other data were not available such as the number of patients who were eligible for AVR and did not undergo surgery. Also, the number of patients with significant coronary

artery disease who were treated with percutaneous revascularization was not reported. The analysis of short-term mortality was based on only 16 patients. Although multivariate analysis showed that short-term mortality was not related to preoperative variables, it is possible that a larger study may have found significant associations. Therefore, the authors did not minimize other data. However, it is possible that the immediate improvement in cardiac function following AVR results in a low short-term mortality, despite the presence of significant preoperative dysfunctions. For those dysfunctions not related to aortic valve disease, their contribution to long-term function may become significant. Although a significant outcome difference between isolated AVR and AVR/CABG surgery was not reported in the present study, it is possible that enough patients were not studied. For this study to detect a difference in mortality between 5.5% and 6.9% with 80% power, the authors would have had to study approximately 9,500 patients. In a significantly larger study of 49,000 patients of all ages (mean age 61-65 years), a similar difference (4.0% v 6.0%) in 30-day mortality was reported to be minimally significant. Regardless, the mortalities for both AVR and AVR/CABG were relatively low compared with previous data. In conclusion, elderly patients undergoing AVR and AVR/ CABG surgery have good short- and long-term outcome and should not be denied surgical therapy based on age alone. Short- and long-term morbidities and mortalities were related to preoperative cardiopulmonary function and/or perioperative complications and not whether or not the patient underwent isolated AVR or AVR/CABG surgery. Predictive analyses overestimated mortality, with a greater difference seen for the higher-risk group. Based on these data, a more aggressive referral surgical strategy for elderly patients with significant aortic valve disease is supported.

REFERENCES 1. Bach DS, Radeva JI, Birnbaum HG, et al: Prevalence, referral patterns, testing, and surgery in aortic valve disease: Leaving women and elderly patients behind? J Ht Valve Dis 16:362-369, 2007 2. Lindroos M, Kupari M, Heikkila J, et al: Prevalence of aortic valve abnormalities in the elderly: An echocardiographic study of a random population sample. J Am Coll Cardiol 21:1220-1225, 1993 3. Bach DS, Cimino N, Deeb GM: Unoperated patients with severe aortic stenosis. J Am Coll Cardiol 50:2018-2019, 2007 4. Brown JL, Pellikka A, Schaff HV, et al: The benefits of early valve replacement in asymptomatic patients with severe aortic stenosis. J Thorac Cardiovasc Surg 135:308-315, 2008 5. Roberts WC, Ko JM, Garner WL, et al: Valve structure and survival in octogenarians having aortic valve replacement for aortic stenosis (⫹/⫺ aortic regurgitation) with versus without coronary artery bypass grafting at a single US medical center (1993 to 2005). Am J Cardiol 100:489-495, 2007 6. Kobayashi KJ, Williams JA, Nwakanma L, et al: Aortic valve replacement and concomitant coronary artery bypass: Assessing the impact of multiple grafts. Ann Thorac Surg 83:969-978, 2007 7. Melby SJ, Zierer A, Kaiser SP, et al: Aortic valve replacement in octogenarians: Risk factors for early and late mortality. Ann thorac Surg 83:1651-1657, 2007 8. Freeman WK, Schaff HV, O’Brien PC, et al: Cardiac surgery in the octogenarian: Perioperative outcome and clinical follow-up. J Am Coll Cardiol 18:29-35, 1991

9. Akins CW, Daggett WM, Vlahakes GJ, et al: Cardiac operations in patients 80 years and older. Ann Thorac Surg 64:614-615, 1997 10. Ngaage DL, Cowen ME, Griffin S, et al: Are initial valve operations still high-risk in the current era? J Heart Valve Dis 17:227-232, 2008 11. Kolh P, Kerzmann A, Lahaye L, et al: Cardiac surgery in octogenarians: Perioperative outcome and long-term results. Eur Heart J 22:1235-1243, 2001 12. Gatti G, Cardu G, Lusa AM, et al: Predictors of perioperative complications in high-risk octogenarians undergoing cardiac operations. Ann Thorac Surg 74:673-677, 2002 13. Vicchio M, Della Corte A, De Santo LS, et al: Tissue versus mechanical prostheses: Quality of life in octogenarians. Ann Thorac Surg 85:1290-1295, 2008 14. De Vincentiis C, Kunkl AB, Trimarchi S, et al: Aortic valve replacement in octogenarians: Is biologic valve the unique solution? Ann Thorac Surg 85:1296-1302, 2008 15. Elayda MA, Hall RJ, Reul RM, et al: Aortic valve replacement in patients 80 years and older. Operative risks and long-term results. Circulation 88:II11-II16, 1993 16. Kolh P, Kerzmann A, Honore C, et al: Aortic valve surgery in octogenarians: Predictive factors for operative and long-term results. Eur J Cardiothorac Surg 31:600-606, 2007 17. Ennker J, Mortasawi A, Gehle S, et al: Aortic valve replacement with or without concomitant coronary artery bypass grafting in the ninth decade of life. Langenbecks Arch Surg 386:272-277, 2001

24

18. Society of Thoracic Surgeons: National Adult Cardiac Surgical Databases. Data Analysis of the Society of Thoracic Surgery. Data version 2.61. Fall 2005, pp 27-32. Available at: http:// 209.220.160.181/STSWebRiskCalc261/de.aspx 19. Nashel SAM, Roques F, Hammill BG, et al: Validation of European System for Cardiac Operative Risk Evaluation (EuroSCORE) in North American Cardiac Surgery. Eur J Cardiothorac Surg 22:101-105, 2002 20. Roques F, Nashef SAM, Gauducheau ME, et al: Risk factors and outcome in European cardiac surgery: Analysis of the EuroSCORE multinational database of 109030 patients. Eur J Cardiothorac Surg 15:816-823, 1999 21. Geissler HJ, Holzl P, Marohl S, et al: Risk stratification in heart surgery: Comparison of six score systems. Eur J Cardiothorac Surg 17:400-406, 2000 22. Roques F, Nashef SA, Michel P; The EuroSCORE Study Group: Risk factors for early mortality after valve surgery in Europe in the 1990s: Lessons from the EuroSCORE pilot program. J Heart Valve Dis 10:572-578, 2001 23. Bert AA, Sterns G, Feng WC, et al: Normothermic cardiopulmonary bypass—A review. J Cardiothorac Vasc Anesth 11:91-99, 1997 24. Leontyvev S, Walther T, Borger MA, et al: Aortic valve replacement in octogenarians: Utility of risk stratification with EuroSCOR. Ann Thoac Surg 87:1440-1445, 2009 25. Urso S, Sadaba R, Greco E, et al: One-hundred aortic valve replacements in octogenarians: Outcomes and risk factors for early mortality. J Heart Valve Disease 16:139-144, 2007 26. Olsson M, Granstrom L, Lindblom D, et al: Aortic valve replacement in octogenarians with aortic stenosis: A case-control study. J Am Coll Cardiol 20:1512-1516, 1992 27. Bessou JP, Bouchart F, Angha S, et al: Aortic valvular replacement in octogenarians. Short-term and mid-term results in 140 patients. Cardiovasc Surg 7:355-362, 1999 28. Gilbert T, Orr W, Banning AP: Surgery for aortic stenosis in severely symptomatic patients older than 80 years: Experience in a single UK centre. Heart 82:138-142, 1999 29. Asimakopoulos G, Edwards MB, Taylor KM: Aortic valve replacement in patients 80 years of age and older. Survival and cause

MASLOW ET AL

of death based on 11100 cases: Collective results from the UK Heart Valve Registry. Circulation 96:3403-3408, 1997 30. Thourani VH, Myung R, Kilgo P, et al: Long-term outcomes after isolated aortic valve replacement in octogenarians: A modern perspective. Ann Thorac Surg 86:1458-1465, 2008 31. Zingone B, Gatti G, Rauber E, et al: Early and late outcomes of cardiac surgery in octogenarians. Ann Thorac Surg 87:71-78, 2009 32. Medallion B, Lytle BW, McCarthy PM, et al: Aortic valve replacement for octogenarians: Are small valves bad. Ann Thorac Surg 66:699-706, 1998 33. Alexander KP, Anstrom KJ, Muhlbaier LH, et al: Outcomes of cardiac surgery in patients ⬎ or ⫽ 80 years: Results from the national cardiovascular network. J Am Coll Cardiol 35:731-738, 2000 34. Likosky DS, Dacey LJ, Baribeau YR, et al: Long-term survival of the very elderly undergoing coronary artery bypass grafting. Ann Thorac Surg 85:1233-1238, 2008 35. Ullery BW, Peterson JC, Milla F, et al: Cardiac surgery in select nonagenarians: Should we or shouldn’t we? Ann Thorac Surg 85:854861, 2008 36. Brown JM, O’Brien SM, Wu C, et al: Isolated aortic valve replacement in North America comprising 108,687 patients in 10 years: Change in risks, valve types, and outcomes in the Society of Thoracic Surgeons National Database. J Thorac Cardiovasc Surg 137:82-90, 2009 37. Billings T, Kodali SK, Shanewise JS: Transcatheter aortic valve implantation: Anesthetic considerations. Anesth Analg 108:1453-1462, 2009 38. Dewey TM, Brown DL, Das TS, et al: High-risk patients referred for transcatheter aortic valve implantation: Management and outcomes. Ann Thorac Surg 86:1450-1457, 2008 39. Ye J, Cheung A, Lichtenstein SV, et al: Transapical transcatheter aortic valve implantation: 1-year outcome in 26 patients. J Thorac Cardiovasc Surg 137:167-173, 2009 40. Zangrillo A, Sparicio D, Crivellari M, et al: Low perioperative mortality for cardiac surgery in octogenarians. Minerva Anestesiol 70:717-726, 2004 41. Scott BH, Seifert FC, Grimson R, et al: Octogenarians undergoing coronary artery bypass graft surgery: Resource utilization, postoperative mortality, and morbidity. J Cardiothorac Vasc Anesth 19: 583-588, 2005