Survival and Quality of Life for Nonagenarians After Cardiac Surgery

Transcription

1 Survival and Quality of Life for Nonagenarians After Cardiac Surgery Manuel Caceres, MD, Wen Cheng, MD, Michele De Robertis, RN, James M. Mirocha, MS, Lawrence Czer, MD, Fardad Esmailian, MD, Ali Khoynezhad, MD, Danny Ramzy, MD, Robert Kass, MD, and Alfredo Trento, MD Divisions of Cardiothoracic Surgery, Cedars Sinai Heart Institute, Cedars Sinai Medical Center, Los Angeles, California; and Divisions of Cardiothoracic Surgery, Veterans Affairs Medical Center, Memphis, Tennessee Background. Reports of cardiac surgery in the elderly have focused primarily on septuagenarians and octogenarians. There are very limited data regarding riskadjusted models in nonagenarians. Methods. From 1983 to 211, patients with age 9 years or greater at the time of coronary artery bypass grafting (CABG) or valve surgery (aortic or mitral) were retrieved from a prospective institutional database. A Cox proportional hazard model was used to determine significant predictors of 5-year survival. In addition, a 12-month assessment of quality of life was conducted. Results. The CABG-only (n [ 46), valve-only (n [ 55), or CABG-valve (n [ 53) surgery was conducted in 154 patients. Demographic characteristics were similar in all groups except for congestive heart failure, which was more prominent in the valve-only or CABG-valve groups (p <.1). The 3-day mortality was 8.8%, 12.8%, and 18.9% in the CABG-only, valve-only, and CABG-valve groups, respectively, without significant difference among groups (p [.35). At 5-years followup, the Kaplan-Meier survival curves do not show a difference among groups (p [.62). Cox proportional hazard model for 5-year survival identified age (hazard ratio [HR] [ 1.25, confidence interval [CI] 1.9 to 1.43, p [.1, for 1-year increase), prior surgery (HR [ 2.23, CI 1.23 to 4.64, p [.7), and prior stroke (HR [ 2.39, CI 1.25 to 3.98, p [.1), as significant predictors of mortality. The 12-month quality of life questionnaire revealed an improvement in 83% of the patients, whereas only 4% reported a decline in cardiac status. Conclusions. Survival in nonagenarians is comparable after CABG or valve surgery. Redo surgery, stroke, and increasing age are significant hazards for mortality. Nonagenarians can undergo cardiac surgery with acceptable mortality and quality of life. (Ann Thorac Surg 213;95: ) Ó 213 by The Society of Thoracic Surgeons Life expectancy in the western world has continued to increase, creating a challenging demand in healthcare systems for future decades. As per the latest national projections in the United States, the population older than 85 years of age accounted for 5.7 million in 21, and by 25 it is projected to surpass 19 million [1]. Furthermore, there is evidence of a higher prevalence of cardiovascular disease with advanced age [2, 3]; thus, the cardiac surgeon is likely to be challenged by an increasing volume of elderly patients in need of heart surgery evaluation. Publications describing risk stratification and outcomes of cardiac surgery in the elderly have primarily focused on septuagenarians and octogenarians, and report favorable results in survival and quality of life [4 7]. Cardiac surgery in nonagenarians has been reported sparsely and has been largely limited to small patient samples not suitable to conduct meaningful risk stratification analyses; yet, it has rapidly increased in the last 2 decades. Furthermore, the majority of studies have Accepted for publication Feb 25, 213. Address correspondence to Dr Caceres, 13 Jefferson Ave, Memphis, TN 3814; caceres manuel@hotmail.com. focused on operative survival with limited information on long-term results and quality of life. With the progressive expansion of the elderly population and the current landscape of cost containment in healthcare, it is essential to analyze the role of cardiac surgery in this growing sector of the population. We present the largest single-center experience of cardiac surgery in nonagenarians, reporting outcomes of early- and long-term survival, and quality of life. Furthermore, a multivariable analysis is performed to identify predictors associated with long-term survival. Patients and Methods Study Population The Cardiothoracic Surgery Quality Assurance Database at the Cedars Sinai Medical Center, Los Angeles, California, was queried to identify all patients 9 years of age or older who underwent open-heart surgery from 1983 to 211. Three surgical categories were defined as the following: isolated coronary artery bypass graft (CABGonly); isolated aortic or mitral valve repair or replacement (valve-only), and combined CABG with aortic or mitral valve repair or replacement (CABG-valve). All patients Ó 213 by The Society of Thoracic Surgeons 3 49/$36. Published by Elsevier Inc

2 Ann Thorac Surg CACERES ET AL 213;95: SURVIVAL AND QOL FOR NONAGENARIANS AFTER CARDIAC SURGERY were prospectively followed and a quality of life (QOL) assessment was conducted at 12 months. Operative mortality was defined as death within 3 days, and survivors were followed until death or the end of the study through telephone interviews, clinic appointments, or contact with the primary care providers. In addition, the Social Security Death Index was queried to determine the survival status of patients not identified through the established institutional follow-up protocol. Approval to conduct this outcomes review was obtained by our Institutional Review Board. Statistical Analysis Patient characteristics were recorded for each of the 3 surgical categories. Continuous variables were presented as the mean standard deviation (SD), and compared using the nonparametric Kruskal-Wallis test or analysis of variance, as appropriate. Categoric variables were presented as percentages and compared using the c 2 test or the Fisher exact test, as appropriate. Kaplan Meier survival curves and median survivals are presented for each surgical category and compared with the log-rank test. Likewise, survival curves are presented for septuagenarians and octogenarians undergoing CABG-only, valve-only, or CABG-valve during the time frame of the study, and compared with nonagenarians. The following preoperative variables were initially selected to develop a multivariable risk model of 5-year survival based on Cox proportional hazards regression: Age, gender, diabetes, hypertension, renal insufficiency (defined as creatinine > 1.5 mg/dl), history of cerebrovascular accident (CVA), congestive heart failure (CHF), ejection fraction (EF), prior surgery, and type of surgery (CABG-only, valve-only, CABG-valve; valve-only served as the reference group). An analysis of the effects eligible for entry in the risk model was conducted with each of the selected preoperative variables and a forward stepwise selection method was used to select the variables independently associated with 5-year survival. Although an objective assessment of the frailty of each patient was not prospectively recorded, the CHF functional class (1 to 4) was used as a surrogate of preoperative functional assessment. To provide a more accurate estimate of renal function, the glomerular filtration rate (GFR) was calculated using the Modification of Diet in Renal Disease (MDRD) formula, and a GFR value lower than 6 was used to identify renal dysfunction. The multivariable risk model was recalculated using this definition. Variables with a p value less than.5 were retained in the final model. The proportional hazards assumption was checked using the supremum test. Statistical calculations were conducted by using SAS version 9.2 (SAS Institute, Cary, NC). Results 1599 From 1983 to 211, a total of 154 patients 9 years of age or greater underwent open-heart surgery. The mean age was 91.4 years (range, 9 to 97 years), 4.9% (63 of 154) were women, the mean EF was.518 (interquartile range [IQR] 4.% to 6.1%), 11.7% (18 of 154) had diabetes, 19.5% (3 of 154) had renal insufficiency (creatinine > 1.5 mg/dl), 62.3% (96 of 154) had hypertension, 62.3% (96 of 154) had CHF, 9.1% (14 of 154) had a history of CVA, and 14.9% (23 of 154) had a history of prior sternotomy. The mean MDRD-calculated GFR was The breakdown according to surgery type was as follows: 29.9% (46 of 154) CABG-only; 34.4% (53 of 154) CABG-valve surgery (mitral or aortic valve); and 35.7% (55 of 154) valve-only surgery (mitral or aortic valve). Demographic characteristics were similar among the 3 surgery categories except for CHF, which was more prominent in the valve-only and CABGvalve groups. Patient characteristics by surgical category are shown in Table 1. The 3-day operative mortality was 13.6% (21 of 154); 8.8% (4 of 46) for CABG-only, 12.7% (7 of 55) for valveonly, and 18.9% (1 of 53) for CABG-valve. The log-rank test did not show a significant difference among surgery ADULT CARDIAC Table 1. Patient Characteristics by Surgery Type Characteristic CABG Only (n 46) CABG Valve (n 53) Valve Only (n 55) p Value Age (years), mean SD Male gender, n (%) 29 (63.) 33 (62.3) 29 (52.7).51 EF, mean SD Diabetes, n (%) 2 (4.4) 11 (2.8) 5/52 (9.6).39 Creatinine > 1.5, n (%) 9/43 (2.9) 11/52 (21.2) 12/52 (23.1) >.99 MDRD, mean HTN, n (%) 28 (6.9) 32 (6.4) 36/52 (69.2).9 CHF, n (%) 13/43 (3.2) 38/53 (71.7) 45/53 (84.9) <.1 CVA, n (%) 3/41 (7.3) 5/48 (1.4) 6/5 (12.).78 Prior surgery, n (%) 4/46 (8.7) 8/53 (15.1) 11/55 (2.).27 Valve replacement, n (%) >.99 Aortic 43 (81.1) 42/54 (77.8) Mitral 6 (11.3) 7/54 (13.) Aortic and mitral 4 (7.6) 5/54 (9.3) CABG ¼ coronary artery bypass grafting; CHF ¼ congestive heart failure; CVA ¼ cerebrovascular accident; EF ¼ ejection fraction; HTN ¼ hypertension; MDRD ¼ modification of diet in renal disease.

3 16 CACERES ET AL Ann Thorac Surg SURVIVAL AND QOL FOR NONAGENARIANS AFTER CARDIAC SURGERY 213;95: groups at 3 days (p ¼.35). The median follow-up was 2.7 [IQR:.6 5.3] years, and the 1- and 5-year survival rates were 71.8% and 37.1%, respectively. The median survival was 3.5 [IQR: ] years; 4.1 [IQR: ] years for the CABG-only group, 2.8 [IQR: ] years for the CABG-valve group, and 2.7 [IQR: ] years for the valve-only group. Kaplan-Meier survival curves were plotted for the 3 surgical groups, and the CABG-only group appears to carry a better survival for the first 4 years; however, the log-rank test failed to identify any significant difference at 5 years (p ¼.62, Fig 1) or 1 years (p ¼.88) of follow-up. Survival curves were also plotted for patients 7 years or older during the time frame of the study, and the median survivals were 8.9 years ([95% confidence interval [CI]: 8.7 to 9.2] for septuagenarians, 5.7 years [95% CI: 5.4 to 6.] for octogenarians, and 3.5 years [95% CI: 2.6 to 4.3] for nonagenarians, showing a significant difference among groups (p <.1). Survival curves are shown in Figures 1 and 2. Living arrangement prior to surgery was; 64.5% (98 of 152) lived at home with a spouse or friend, 28.9% (44 of 152) lived alone, and 6.6% (1 of 152) lived in a board and care facility. At discharge, 35.3% (53 of 15) were released to home, 5.3% (8 of 15) to friends or relatives, 2% (3 of 15) to a board and care facility, and 39.3% (59 of 15) to extended care facilities. A risk model, based on Cox proportional hazards regression, was developed to identify the variables associated with 5-year survival. The univariate assessment through an analysis of effect of eligibility for each candidate variable revealed age (p ¼.7) to have the highest Wald c 2 value; thus, age was entered first in the stepwise selection process. After the testing of each variable in the model, age (p <.1), prior surgery (p ¼.6), and CVA (p ¼.1) were retained as significant predictors of a decreased 5-year survival. To better adjust for the significance of renal dysfunction, creatinine level was replaced by MDRD-calculated GFR; likewise, Percent Surv va Number at Risk CHF functional class was used as a surrogate of the frailty condition of the patients. Neither variable showed a significant association with survival. The results of the analysis of eligibility for each candidate variable and the multivariable analysis are shown in Tables 2 and 3. A QOL questionnaire was sent to all the survivors 1 year after surgery, with 71 responding of the 11 patients who were alive at the 1-year mark. The questionnaire revealed an improvement in 83% (59 of 71) of the respondents whereas only 4% (3 of 71) reported a decrease in their quality of life. Comment Septuagenarian Octogenarian Nonagenarian Log-Rank P <.1 Years Fig 2. Five year survival after open heart surgery; septuagenarians (blue), octogenarians (red), and nonagenarians (green) The present study reports the largest single-center experience of cardiac surgery in nonagenarians, presenting Percent Surv va Number at Risk CABG CABG + Valve Valve Only Log-Rank P =.62 Years Fig 1. Open heart surgery in nonagenarians: 5 year survival. (CABG coronary artery bypass grafting.) Table 2. Analysis of Eligibility of Candidate Variables Candidate Variable p Value Age.7 Gender: male.2 CABG valve vs valve only.66 CABG only vs valve only.61 Diabetes.42 Hypertension.137 Cerebrovascular accident.24 Creatinine > 1.5 mg/dl.39 MDRD < 6.1 Prior surgery.26 Ejection fraction.43 CHF NYHA 1 vs NYHA 4.56 CHF NYHA 2 vs NYHA 4.25 CHF NYHA 3 vs NYHA 4.41 CABG ¼ coronary artery bypass grafting; CHF ¼ congestive heart failure; MDRD ¼ modification of diet in renal disease; NYHA ¼ New York Heart Association class.

4 Ann Thorac Surg CACERES ET AL 213;95: SURVIVAL AND QOL FOR NONAGENARIANS AFTER CARDIAC SURGERY Table 3. Multivariable Analysis; Final Stepwise Selection Model Variable Hazard Ratio Confidence Interval p Value Age per year Prior surgery Cerebrovascular accident early and long-term outcomes, quality of life data, and a risk model of long-term survival. Lichtman and colleagues [8], using the Medicare Provider Analysis and Report (MedPAR) file, analyzed the trends and outcomes of 4,224 nonagenarian patients undergoing CABG. However, administrative databases commonly lack essential clinical information to develop reliable risk models, and although this is the largest registry-based report in the literature, it did not provide a short- or long-term risk-analysis. There are only 2 studies reporting a multivariable risk model of operative mortality [9, 1] and only 1 focusing on long-term survival in nonagenarians [11]. Speziale and colleagues [9] reported a case series of 127 patients from 8 institutions and concluded that nonelective surgery and preoperative myocardial infarction were associated with a higher operative risk [9]. Bridges and colleagues [1] reported an operative risk model based on the Society of Thoracic Surgeons (STS) cardiac surgical database and identified emergency or salvage clinical status, intraaortic balloon pump use, renal failure, peripheral vascular and cerebrovascular disease, and mitral regurgitation to be associated with an increased operative mortality. A classification tree was reported in this study and showed that 57% of nonagenarians did not present any of these risk factors, showing an operative mortality of 7.2%; an acceptable outcome considering the higher risk of this patient category [1]. Ullery and colleagues [11] reported a risk model of long-term survival on 49 nonagenarians, identifying chronic renal insufficiency, peripheral vascular disease, and a low EF as predictors of decreased survival. In our study, with a much larger study sample, we identified age, cerebrovascular disease, and prior surgery to be associated to a decreased survival. The operative mortality in our study was 13.6%, ranging from 8.8% for CABG-only to 18.9% for CABGvalve. The operative mortality of cardiac surgery in nonagenarians has ranged from 7.1% to 23.5% in previous publications, though most of the studies were centered in the 7% to 13% range [3, 8 17]. The STS database has the largest repository of clinical information on cardiac surgery and reports an operative mortality rate of 11.8% for CABG, 11.4% for valve-only, and 12.% for CABGvalve [1]. Although there has always been a trend for CABG-valve to have a higher operative mortality, similar to our findings none of the previous studies have reported a statistically significant difference among surgical groups; ie, CABG-only, valve-only, and CABGvalve [9, 11]. 161 Unlike younger age groups and the collective population undergoing cardiac surgery, in which women represent 27% of the cases [18], previous publications of cardiac surgery in nonagenarians and our study have consistently found a higher proportion of women, ranging from 33% to 65% [3, 12, 15, 16]. This finding is likely related to the longer life expectancy in women, which is over 5 years longer compared with men [19]. The 1- and 5-year survival rates in our series were 71.8% and 37.1%, respectively; rates consistent with previous publications that report 1- and 5-year survival rates ranging from 64% to 87% and 26% to 54%, respectively [3, 12, 15, 2]. The median survival in our series was 3.5 years (2.8 years for men and 4.3 years for women), which is in close proximity to the age-matched life expectancies for the US population (3.6 years for men and 4.3 years for women [19]). Easo and colleagues [12] compared the survival of nonagenarians undergoing cardiac surgery (17 patients) with data from life tables, without showing a significant difference; in contrast, Guilfoyle and colleagues [21] conducted a similar comparison (23 patients) and found a standardized mortality ratio of.52 in favor of nonagenarians undergoing cardiac surgery. These studies, although contradictory, included a very limited number of patients, and the results may not be as robust as the data presented in our study. Furthermore, it has to be noted that nonagenarians undergoing cardiac surgery are not representative of the unselected population described in life tables; in fact, the survival of the former is expected to be significantly limited due to their higher comorbidity index. The QOL assessment after cardiac surgery in nonagenarians has been addressed in few studies and has invariably reported favorable results, showing negative effects in very few instances, which is consistent with the 4% decline in QOL reported in our study [3, 11, 17, 21 23]. In 1997 our group presented the early experience of openheart surgery in nonagenarians and showed that 89% of the patients reported an improvement in QOL, while 78% of the survivors were in New York Heart Association class I or II within 2 years of the operation [17]. There is agreement from previous publications that cardiac surgery in the elderly carries an increased operative mortality. Adjusted and unadjusted analyses in this sector of the population have reported an increased operative risk in the setting of an urgent or emergency surgery, and several publications have suggested that a delay in referral to surgical treatment may be in part responsible [9, 2]. Nonetheless, the STS database reports comparable rates of nonelective surgery for CABG (56.1% vs 5.8%), valve-only (36.5% vs 38.7%), and CABG-valve (27.9% vs 23.4%) in nonagenarians and the collective of all age groups, respectively, suggesting that a delay in surgical treatment may not play a significant role [1, 18]. Despite a higher operative risk, there is consistent data from prior publications supporting that nonagenarians undergoing cardiac surgery are highly symptomatic and have significant limitations in survival and QOL, and that pursuing surgical treatment has beneficial long-term ADULT CARDIAC

5 162 CACERES ET AL Ann Thorac Surg SURVIVAL AND QOL FOR NONAGENARIANS AFTER CARDIAC SURGERY 213;95: effects at the expense of an increased but acceptable upfront operative mortality [1, 21, 24]. Limitations Our study shares the inherent weaknesses of any retrospective study. The patients undergoing cardiac surgery represent a highly selected group and the results may not be generalized to all nonagenarians considered for surgical treatment. Neither ours nor previous studies provide a control group for comparison, other than life table survival data from the unselected US population. Similar to previous publications, the QOL assessment was not 1% complete, and missing data may bias the results. Although we present the largest institutional study to date, the limited patient samples of our and previous studies limit the number of variables to be entered in a multivariable analysis. Despite the various limitations noted, our data present favorable results and support a strong consideration for the risks and benefits of cardiac surgery in this sector of the population. Conclusions The elderly population is rapidly expanding and cardiac surgery in nonagenarians has significantly increased in the last 2 decades, with a higher representation of women as compared with younger age groups. Although risk stratification modeling may be limited by small patient samples, our and previous reports do not support gender and type of surgery as significant risk factors, as opposed to younger age groups. Conversely, age, prior cerebrovascular accident, and prior cardiac surgery appear to be associated with a decreased long-term survival. Cardiac surgery in nonagenarians is a valid option with an increased but not prohibitive operative mortality; however, offset by favorable results in long-term survival and QOL. Proper patient selection with a heightened attention to the functional reserve and the expectation for a QOL improvement is paramount to ensure the most favorable outcomes. References 1. Available at projections/summarytables.html. Accessed on September 25th, Ohlow MA, Hassan A, Lotze U, Lauer B. Cardiac catheter isation in nonagenarians: single center experience. J Geriatr Cardiol 212;9: Jonsson A, Agnarsson BA, Hallgrímsson J. Coronary atherosclerosis and myocardial infarction in nonagenarians: a retrospective autopsy study. Age Ageing 1985;14: Vasques F, Messori A, Lucenteforte E, Biancari F. Immediate and late outcome of patients aged 8 years and older undergoing isolated aortic valve replacement: a systematic review and meta analysis of 48 studies. Am Heart J 212;163: Stoica SC, Cafferty F, Kitcat J, et al. Octogenarians under going cardiac surgery outlive their peers: a case for early referral. Heart 26;92: Edmunds LH Jr, Stephenson LW, Edie RN, Ratcliffe MB. Open heart surgery in octogenarians. N Engl J Med 1988;319: Rich MW, Sandza JG, Kleiger RE, Connors JP. Cardiac operations in patients over 8 years of age. J Thorac Car diovasc Surg 1985;9: Lichtman JH, Kapoor R, Wang Y, Radford MJ, Allen NB, Krumholz HM. Temporal trends of outcomes for nonage narians undergoing coronary artery bypass grafting, 1993 to Am J Cardiol 27;1: Speziale G, Nasso G, Barattoni MC, et al. Operative and middle term results of cardiac surgery in nonagenarians: a bridge toward routine practice. Circulation 21;121: Bridges CR, Edwards FH, Peterson ED, Coombs LP, Ferguson TB. Cardiac surgery in nonagenarians and cente narians. J Am Coll Surg 23;197: Ullery BW, Peterson JC, Milla F, et al. Cardiac surgery in select nonagenarians: should we or shouldn t we? Ann Thorac Surg 28;85: Easo J, Holzl PP, Horst M, Dikov V, Litmathe J, Dapunt O. Cardiac surgery in nonagenarians: pushing the boundary one further decade. Arch Gerontol Geriatr 211;53: Bacchetta MD, Ko W, Girardi LN, et al. Outcomes of cardiac surgery in nonagenarians: a 1 year experience. Ann Thorac Surg 23;: Samuels LE, Sharma S, Morris RJ, et al. Cardiac surgery in nonagenarians. J Card Surg 1996;11: Hovanesyan A, Moon MR, Rich MW. Cardiac surgery in nonagenarians. J Cardiovasc Surg (Torino) 27;48: Roberts WC, Ko JM, Matter GJ. Aortic valve replacement for aortic stenosis in nonagenarians. Am J Cardiol 26;98: Blanche C, Matloff JM, Denton TA, et al. Cardiac operations in patients 9 years of age and older. Ann Thorac Surg 1997;63: Shahian DM, O Brien SM, Filardo G, et al. The Society of Thoracic Surgeons 28 cardiac surgery risk models: part 1 coronary artery bypass grafting surgery. Ann Thorac Surg 29;88(1 Suppl):S Available at nvsr59 9.pdf. Accessed on September 25, Speziale G, Nasso G, Barattoni MC, et al. Short term and long term results of cardiac surgery in elderly and very elderly patients. J Thorac Cardiovasc Surg 211;141: Guilfoyle MR, Drain AJ, Khan A, Ferguson J, Large SR, Nashef SA. Cardiac surgery in nonagenarians: single centre series and review. Gerontology 21;56: Levy Praschker BG, Leprince P, Bonnet N, et al. Cardiac surgery in nonagenarians: hospital mortality and long term follow up. Interact Cardiovasc Thorac Surg 26;5: Miller DJ, Samuels LE, Kaufman MS, Morris RJ, Thomas MP, Brockman SK. Coronary artery bypass surgery in nonage narians. Angiology 1999;5: Edwards MB, Taylor KM. Outcomes in nonagenarians after heart valve replacement operation. Ann Thorac Surg 23;: 83 4.