Transcatheter Aortic Valve Replacement Versus Aortic Valve Bypass

Transcatheter Aortic Valve Replacement Versus Aortic Valve Bypass

Transcatheter Aortic Valve Replacement Versus Aortic Valve Bypass: A Comparison of Outcomes and Economics John W. Brown, MD, Jack H. Boyd, MD, Parth M...

231KB Sizes 1 Downloads 82 Views

Transcatheter Aortic Valve Replacement Versus Aortic Valve Bypass: A Comparison of Outcomes and Economics John W. Brown, MD, Jack H. Boyd, MD, Parth M. Patel, MS, Mary L. Baker, BSN, Amjad Syed, MD, Joe Ladowski, MS, and Joel Corvera, MD Sections of Cardiothoracic Surgery, Indiana University Health and Indiana University School of Medicine, Indianapolis, IN

Background. Transcatheter aortic valve replacement (TAVR) is currently offered to patients who are high-risk candidates for conventional surgical aortic valve replacement. For the past 37 years, off-pump aortic valve bypass (AVB) has been used in elderly patients at our center for this similarly high-risk group. Although TAVR and AVB were offered to similar patients at our center, comparisons of clinical outcomes and hospital economics for each strategy were not reported. Methods. We reviewed the clinical and financial records of 53 consecutive AVB procedures performed since 2008 with the records of 51 consecutive TAVR procedures performed since 2012. Data included demographics, hemodynamics, The Society of Thoracic Surgeons (STS) risk score, extent of coronary disease, and ventricular function. Follow-up was 100% in both groups. Hospital financial information for both cohorts was obtained. Mean risk score for the TAVR group was 10.1% versus 17.6% for AVB group (p < 0.001). Results. Kaplan-Meier hospital rates of 3- and 6-month survival and of 1-year survival were 88%, 86%, 81%, and 61% and 89%, 83%, 83%, and 70% for the TAVR and AVB groups, respectively (p [ 0.781). Two patients who had undergone TAVR had a procedure-related stroke. The one

stroke in an AVB recipient was late and not procedure related. At discharge, mild and moderate perivalvular and central aortic insufficiency were present in 31% and 16% of TAVR recipients, respectively; no AVB valve leaked. Transvalvular gradients were reduced to less than 10 mm Hg in both groups. The average hospital length of stay for the AVB-treated patients was 13 days, and it was 9 days for the TAVR-treated patients. Median hospital charges were $253,000 for TAVR and $158,000 for AVB. Mean payment to the hospital was $65,000 (TAVR) versus $64,000 (AVB), and the mean positive contribution margin (profit) to the hospital was $14,000 for TAVR versus $29,000 for AVB. Conclusions. TAVR and AVB relieve aortic stenosis and have similar and acceptable procedural mortality rates. AVB-treated patients had 1.75 times the STS risk score when compared with the TAVR cohort. Hospital charges for TAVR were nearly twofold those of AVB. Hospital reimbursement was similar, but AVB had two to four times the profit margin of TAVR. Longer follow-up for the TAVR cohort will determine whether survival is comparable to that after AVB at 3 and 5 years.


offered these patients two treatment options. The first is TAVR, performed using the transapical or transfemoral approach. The second is the off-pump aortic valve bypass (AVB) operation with an apicoaortic conduit, a technique we have used since 1978. The AVB is a technically demanding operation and, like all other operations, has a learning curve. However we have followed patients up to 16 years postoperatively after this intervention.

evere and symptomatic aortic stenosis is a progressive disease and requires aggressive interventional therapy in addition to medical therapy to optimize affected patient outcomes [1, 2]. Before the advent of the transcatheter aortic valve replacement (TAVR) procedure, interventional or surgical therapy was often denied these high-risk patients [3, 4]. The definition of a high-risk patient population included those patients who had previous coronary artery bypass surgery, patients with a calcified aortic root or ascending aorta, and patients with chronic lung disease, renal failure, advanced age, or a history of diabetes or stroke [5]. Since 2012, we have Accepted for publication May 15, 2015. Presented at the Sixty-first Annual Meeting of The Southern Thoracic Surgical Association, Tucson, AZ, Nov 5-8, 2014. Address correspondence to Dr Brown, Section of Cardiothoracic Surgery, Indiana University School of Medicine, 545 Barnhill Dr, EH 215, Indianapolis, IN 46202; e-mail: [email protected]

Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier

(Ann Thorac Surg 2015;-:-–-) Ó 2015 by The Society of Thoracic Surgeons

Material and Methods We have been offering the AVB option to high-risk patients with aortic stenosis for 37 years at our institution (Indiana University Health and Indiana University School of Medicine, Indianapolis, IN). We began offering the TAVR procedure to similarly high-risk patients in 2012. In this study, we reviewed and compared our AVB clinical and financial data since 2008 (n ¼ 53) with our initial TAVR clinical and financial data (n ¼ 51). 0003-4975/$36.00



Demographics The average age at operation for the 53 AVB-treated patients was 73.7  13.0 years, and for the TAVR-treated patients it was 80.7  11.4 years (p ¼ 0.04). There were 28 (53%) and 24 (47%) male patients who underwent AVB and TAVR operations, respectively. All patients had either severe or symptomatic aortic stenosis, with a mean aortic valve area of 0.76  0.28 cm2 in the AVB group and 0.75  0.21 cm2 in the TAVR group. The average peak gradient across the aortic valve was 65.4  22.8 and 63.1  22.6 for the AVB and TAVR groups, respectively. The average left ventricular ejection fraction was 48.1%  15.4% in the AVB group and 49.2%  12.0% in the TAVR group. The Society of Thoracic Surgeons (STS) risk score was calculated for each patient in the study, and the frequency of each of the variables corresponding to this risk score is shown in Table 1. The average STS risk score was also used to calculate expected early mortality for each cohort. This study was approved by the Indiana University Institutional Review Board committee as an exempt study.

Table 1. Patients’ Demographic and Preoperative Risk Factorsa

Age (years) Male Nonwhite Weight (kg) Height (cm) Peripheral vascular disease Cerebrovascular disease Cerebrovascular accident Diabetes mellitus Preoperative creatinine Hypertension Endocarditis Chronic lung disease Immunosuppression Previous CABG Previous valve procedure Previous PCI Arrhythmia Previous myocardial infarction Congestive heart failure Coronary artery disease Ejection fraction (%) Mitral stenosis Aortic insufficiency Mitral regurgitation Tricuspid regurgitation STS risk score (%) a



p Score

73.7  13.0 28 (53%) 3 (6%) 82.8  22.1 165.1  10.9 5 (9%) 11 (21%) 5 (9%) 23 (43%) 1.31  1.03 31 (58%) 1 (2%) 5 (9%) 0 (0%) 24 (45%) 7 (13%) 10 (19%) 13 (25%) 3 (6%)

80.7  11.4 24 (47%) 3 (6%) 81  23.4 164.6  9.7 13 (25%) 11 (22%) 5 (10%) 16 (31%) 1.47  0.81 34 (67%) 0 (0%) 8 (16%) 0 (0%) 11 (22%) 2 (4%) 13 (25%) 16 (31%) 1 (2%)

0.004 0.556 0.961 0.703 0.787 0.03 0.919 0.949 0.205 0.41 0.389 0.324 0.335 1 0.011 0.092 0.416 0.437 0.327

43 (81%) 38 (72%) 48.1  15.4 3 (6%) 6 (11%) 13 (25%) 8 (15%) 17.5  10.9

51 (100%) 30 (59%) 49.2  12.0 4 (8%) 0 (0%) 15 (29%) 12 (24%) 10.1  6.1

0.045 0.168 0.697 0.657 0.013 0.575 0.275 0.0011

Differences in STS risk score system are shown in the right column.

AVB ¼ aortic valve bypass; CABG ¼ coronary artery bypass graft; PCI ¼ percutaneous coronary intervention; STS ¼ The Society of Thoracic Surgeons; TAVR ¼ transcatheter aortic valve replacement.

Ann Thorac Surg 2015;-:-–-

c2 tests were used to determine significance for categorical data. Student’s t test was used to determine significance for continuous data. The observed to expected mortality ratio, the standardized mortality ratio (SMR), was calculated and compared with the null by using a Z test. Survival curves were based on the Kaplan-Meier formula. The alpha value for this study was set to 0.05. All data were analyzed using SPSS 21 (IBM, Armonk, NY). Surgical Technique AORTIC VALVE BYPASS. Preoperatively, each patient has a chest computed tomography scan to rule out a heavily calcified descending thoracic aorta because this finding would be a contraindication to AVB. As long as there was a 5- to 6-cm portion of the descending thoracic aorta between the level of the left inferior pulmonary vein and the top of the left hemidiaphragm that was devoid of confluent calcification, for the distal anastomosis, we approved the patient for the AVB procedure. The operation was performed through a left anterior thoracotomy as described by our group previously [6, 7]. The AVB was done off cardiopulmonary bypass in all but 4 this series (92%). In one patient, bypass was required to insert the apicoaortic conduit because it was unclear whether the left ventricle was symmetrically cored with the cork bore. Two patients were electively placed on bypass before cardiac manipulation because of hemodynamic instability, severe pulmonary hypertension, and low cardiac output secondary to their severe aortic stenosis. The fourth patient was placed on bypass for a concomitant coronary artery bypass graft procedure. TRANSCATHETER AORTIC VALVE REPLACEMENT. The Edwards Sapien heart valve system (Edwards Life Sciences, Inc, Irvine, CA) was used at our institution when performing the TAVR. The procedure was performed in the hybrid operating room with the patient under general anesthesia, with biplane fluoroscopy and transesophageal echocardiography for assistance in positioning the prosthesis. Four sites for vascular access were used: transapical (n ¼ 27), transfemoral (n ¼ 20), transiliac (n ¼ 3), and transaortic (n ¼ 1). Leon and colleagues described, in great detail, the TAVR technique used in this series [8].

Results Morbidity and Mortality Five (9.4%) early deaths and 17 (35%) late deaths occurred in the AVB group over the last 6 years. There were six (11.8%) early deaths and six (13.3%) late deaths in the TAVR group over the past 2 years. The SMR for the AVB was 0.54 (p ¼ 0.12) and for the TAVR it was 1.16 (p ¼ 0.48). Of the 5 patients who had early AVB deaths, 2 died of respiratory insufficiency resulting from postoperative pneumonia and sepsis. One patient died of anemia because the patient and family refused postoperative blood products for religious reasons. One patient died of multisystem organ failure including bowel ischemia and required terminal ileum resection several days after AVB

Ann Thorac Surg 2015;-:-–-

placement. The fifth patient died of massive gastrointestinal bleeding. Of the six early TAVR deaths, three were the result of acute renal failure. Of the remaining 3 patients, 1 died of respiratory failure, 1 died of a cerebrovascular accident, and 1 died of sudden asystole. Three of the 6 patients who had late TAVR deaths died of unknown causes: 1 patient died of acute renal failure, 1 died of sudden asystole, and 1 died of a suspected myocardial infarction. Eleven of the 17 late AVB deaths over the next 6 years were of unknown causes. One each of the other six late AVB deaths resulted from CHF, hemorrhage, sepsis, endocarditis, respiratory arrest, and a myocardial infarction. The late death in the patient with hemorrhage occurred during drainage and debridement of a thoracotomy wound abscess. The actuarial survival at 6, 12, and 24 months was 83%, 70%, and 62%, respectively, in the AVB cohort. The actuarial survival at 6, 12, and 24 months was 81%, 61%, and 61%, respectively, in the TAVR cohort (p ¼ 0.78) (Fig 1). One AVB-treated patient had a stroke; this was late and secondary to a septic embolus. The patient was suffering from previously replaced aortic valve prosthesis endocarditis. Additionally, 3 AVB-treated patients experienced cardiac arrhythmias requiring pacemaker or implantable cardioverter-defibrillator placement, two early and one late. All these patients had been treated preoperatively for coronary artery disease. Two TAVR-treated patients had strokes, one early (fatal) and one late (nonfatal). Additionally, 1 TAVRtreated patient experienced complete heart block requiring pacemaker placement. Another TAVR-treated patient was being worked up for pacemaker placement at last follow-up.

Postoperative Hemodynamics and Hospital Length of Stay Intraoperative measurement of gradients for all AVB and TAVR cohorts showed complete relief of aortic stenosis. Postoperative echocardiograms of TAVR-treated patients

Fig 1. The Kaplan-Meier survival curves for the transcatheter aortic valve replacement (TAVR) at Indiana University (IU) versus aortic valve bypass (AVB) IU versus PARTNER (Placement of Aortic Transcatheter Valves) cohort B groups [17, 18].



showed a mild perivalvular leak in 13 (25%) patients and a moderate perivalvular leak in 3 (6%) patients. Seven (14%) patients had mild central valvular leaks, and 1 (2%) patient had a moderate central valvular leak. No patient had more than a moderate prosthetic valve leak. The average hospital stay for the AVB-treated patients was 13 days (median, 11 days; range, 5 to 39 days). The average length of stay for the TAVR-treated patients was 9 days (median, 7 days; range, 3 to 21 days; p ¼ 0.004).

Hospital and Physician Economics TAVR-related hospital charges were higher at $253,000 (range, $178,000 to $434,000) than AVB-related charges, at $159,000 (range, $76,000 to $415,000) (p < 0.001) (Table 2). On average, payment to the hospital was slightly more for TAVR at $65,397 versus AVB at $63,751 (p ¼ 0.849). The average direct cost for the hospital was significantly higher for TAVR procedures than for AVB procedures because of the higher cost of the TAVR prosthesis ($51,478 vs $35,397; p < 0.001). Finally, of the 51 TAVR cases, 17 cases (33%) had a negative hospital contribution margin versus only four of the 53 AVB cases (7.5%) (p ¼ 0.001). The average contribution margin for the AVB cases was $29,000 versus $14,000 for the TAVR cases (p ¼ 0.055). Table 3 shows the hospital payer mix and the contribution margin breakdown for TAVR and AVB for each payer group.

Comment Symptomatic severe aortic valve stenosis is a serious and progressive disease requiring intervention to relieve the symptoms of obstruction [1, 2]. Survival without intervention is poor [1]. Kapadia reported that only 6 of 179 patients in the medical treatment arm of the PARTNER B (Placement of Aortic Transcatheter Valves cohort B) trial survived 5 years [9]. Conventional aortic valve replacement (AVR) can relieve the patient’s obstruction, and it has been shown to have better survival rates than standard medical therapy [1, 10]. Conventional AVR carries significant risk of hospital morbidity and mortality in patients with comorbidities such as preoperative graftdependent coronary disease, chronic obstructive pulmonary disease, left ventricular dysfunction, and renal insufficiency with a serum creatinine greater than 2.5 [10]. To preclude these potential complications and to minimize recovery time, our institution has been offering AVB to high-risk patients for more than 37 years. We have performed AVB procedures in more than 100 elderly patients to date. The primary indications for AVB have been as follows: (1) aortic stenosis and coronary artery disease in patients who have already undergone coronary bypass surgery and are graft dependent; (2) a heavily calcified (porcelain) aortic root, ascending aorta, and aortic arch; (3) renal insufficiency; and (4) a small aortic root less than 20 mm in diameter. More recently, in 2012, our institution adopted TAVR for many of these high-risk patients. TAVR has been widely adopted in Europe and the United States for the last 5 to 10 years. This report represents our initial experience with TAVR.



Ann Thorac Surg 2015;-:-–-

Table 2. Summary of the Comparative Hospital Length of Stay, Hospital Charges, Hospital Collections, and Direct and Fixed Hospital Costs for Transcatheter Aortic Valve Replacement and Aortic Valve Bypass Procedures Indicators:



TAVR Ranges

AVB Ranges

ALOS (days) Average hospital charges ($) Average hospital payment ($) Average variable hospital direct cost ($) Average fixed hospital cost ($) Average contribution margin ($)

9 253,000 65,000 47,000 4,000 14,000

13 158,000 64,000 28,000 6,500 29,000

3–21 178,000–434,000 35,000–240,000 46,000–79,000 2,000–13,000 44,000–þ170,000

5–39 76,000–415,000 26,000–268,000 16,000–67,000 3,000–17,000 28,000–þ192,000

ALOS ¼ average length of stay;

AVB ¼ aortic valve bypass;

TAVR ¼ transcatheter aortic valve replacement.

The AVB cohort had a slightly lower early mortality rate compared with our TAVR group (five early deaths vs six early deaths). The STS risk score for the AVB group, however was significantly higher than that of the TAVR group (17.5 vs 10.1, p < 0.001) even though the AVB group was 7 years younger. The AVB group, however, had more coronary disease and earlier coronary bypass procedures. We tried to account for these differences by calculating SMRs for each cohort. Although neither cohort’s SMR rejected the null of 1 (p > 0.05), the SMR of the TAVR group was empirically higher than that of the AVB group, a finding that suggests the clinical superiority of AVB with respect to early mortality. The cause of late mortality in both patient groups in this study is less clear than the causes of early mortality. Few if any patients with late deaths had autopsy material to examine. The reason for the larger number of late deaths reported in the AVB group was that we had 6-year follow-up on the AVB group and only 2-year follow-up on the TAVR group. Most families of these patients who had late deaths reported that the deaths resulted from “old age.” Few if any of the late deaths in the AVB and TAVR groups were thought to be valve related. Previous papers have noted that the incidence of renal failure in TAVR-treated patients is similar to that in patients treated with conventional SAVR, approximately 5% [11]. Renal failure requiring dialysis was also observed at our center after TAVR, and half of the early deaths and one late death resulted from renal failure (4 of 51; 8%). Two of these four deaths were in patients with no earlier history of renal insufficiency. Neurologic events are not

Table 3. Payer Mix and the Contribution Margin for Transcatheter Aortic Valve Replacement and Aortic Valve Bypass Contribution Margin per Payer Payers Managed care Medicaid Medicare Other government

TAVR (%) AVB (%) TAVR (%) 8.00 2.00 86.00 4.00

AVB ¼ aortic valve bypass; replacement.

15.80 5.30 73.70 5.30

127,000 65,000 3,000 (1,500)

AVB (%) 112,000 34,000 12,000 16,000

TAVR ¼ transcatheter aortic valve

uncommon in earlier reports of TAVR [12]. One potential reason for this finding is that the native valve manipulation required for implantation of the TAVR valve disturbs the native valve calcification [13]. Because the AVB operation does not touch the ascending aorta or the native valve, the risk of potential procedure-related cerebral emboli is greatly reduced. Fortunately, only one procedural stroke occurred in our TAVR series. Recently, Kapadia reported on the TAVR PARTNER B trial. The 3- and 5-year survival rates were 46% and 28%, respectively. These results are sobering, but when compared with standard medical therapy with survival of 6%, they are not surprising. The 5-year survival rate for our AVB group, with a much higher STS score, was significantly superior at 46% [9, 14]. Hemodynamically, all patients who underwent the TAVR and the AVB had complete relief of aortic stenosis. In the AVB group, the aortic valve area was the sum of the native valve area plus the area of the valve inside the conduit. As noted in our previous AVB reports, 70% to 80% of the cardiac output leaves the left ventricle through the apicoaortic conduit, and 20% to 30% continues out the native aortic valve [15]. Even though the forward flow through the native valve and ascending and arch is decreased threefold, there is enough anterograde flow to provide adequate perfusion of the coronary and cerebral circulation. There was minimal retrograde flow up the descending aorta from the conduit to the left subclavian artery [15]. In our experience, no hemodynamic complications have been associated with AVB. Whereas TAVR relieved severe aortic stenosis, some of our TAVR-treated patients experienced central and periprosthetic aortic regurgitation. These leaks were small in 20 patients (40%) and moderate in 4 patients (8%). Although TAVR central and perivalvular leaks to date seem to have been of little significance in our TAVR cohort, they have been associated with negative outcomes in larger TAVR series [16]. The AVB and the TAVR procedures are both technically demanding and have learning curves. AVB-treated patients have occasionally required cardiopulmonary bypass (4 of 53; 8%) in our series, and our surgical times have been reduced to 3 hours for most cases. In particular, 1 patient required bypass because of a retained fragment of apical endocardium at the site of the left ventricular apex. To mitigate the anxiety and risk of bleeding at the time of left ventricular stent placement, an automated coring device has been developed. This device

Ann Thorac Surg 2015;-:-–-

is used in Europe, Canada, and Australia, but has not been approved for use in the United States. Cardiopulmonary bypass was necessary in 9 TAVRtreated patients (18%) and in only 4 patients (8%) in the AVB group. Determining the proper arterial access site through which to deploy the TAVR valve in these elderly patients with vascular disease and friable myocardial tissue can be challenging. Published cost analyses have shown that the transfemoral approach was the least expensive option [17]. We found that with respect to hospital stay, minimal difference existed between techniques. The average hospital stay was 7.6  3.4 and 7.0 -3.2 days (p ¼ 0.655) for the transfemoral and transapical approaches, respectively. There were nonsignificant differences between these two access points in survival at 6 and 12 months, but the sample sizes are too small to make definitive conclusions.

Financial Analysis Although the AVB recipients, on average, stayed longer in the hospital than did the TAVR recipients (4 days, p ¼ 0.004), the hospital collections for TAVR-treated and AVB-treated patients were similar ($65,000 vs $64,000, p ¼ 0.849). Seventeen TAVR-treated patients (33%) provided a negative contribution margin to the hospital because of the greater cost of the prosthesis, whereas only 4 (8%) of the AVB-treated patients provided a negative contribution margin (p ¼ 0.001). At our center, AVB cases on average provided twice the positive contribution margin to the hospital than did TAVR cases (Table 3, p ¼ 0.055). The negative hospital contribution margin produced by a significant number of TAVR recipients could be softened if more TAVR valves were implanted, if the payer mix included more patients with managed care insurance, or if reimbursement for TAVR increased.

Study Limitations This was a nonrandomized, noncontrolled retrospective review. The limitations were the unequal follow-up between the two groups. AVB-treated patients had a much longer follow-up than did TAVR-treated patients. Additionally, the two groups had disparate STS risk scores. The AVB-treated patients were a much sicker group, with 1.75 times the STS risk score (10% vs 17.5%).

Conclusions Patients undergoing AVB and TAVR procedures had similar survival at 12 and 24 months at our center. The AVB 5-year survival is superior to published TAVR 5-year results [14, 18]. Hospital and intensive care unit stays were longer for patients undergoing AVB at our center, but our AVB-treated patients had significantly higher STS risk scores. Perivalvular and central leaks, heart block, procedure-related stroke, and renal failure were rarely or not seen in our AVB series. These complications were more prevalent in our TAVR series and in published series [11, 12]. Cardiopulmonary bypass was necessary for twice as many of our TAVR-treated patients as for AVBtreated patients.



The hospital charges were nearly double for TAVRtreated patients at our center. Hospital reimbursement was similar for both groups at approximately $65,000. Hospital contribution margin was two to four times more for AVB than for TAVR procedures. Most of our TAVR cases were done by alternative access, and this may have contributed to the longer hospital stay and financial disparity. However, hemodynamic and survival outcomes for TAVR were comparable to those of AVB. The relatively high costs and relatively low reimbursement for TAVR cases bring into question the sustainability of TAVR programs for some hospital systems unless market competition substantially reduces the costs of the TAVR prosthesis. This study highlights the utility of AVB in the treatment of very high-risk patients with aortic stenosis and its lower complication rate. Despite the lower cost and postoperative complications of AVB, AVB was seldom offered as the initial strategy for patients at our center after our TAVR program started unless a patient was a poor candidate for TAVR. When the aortic annular diameter was too small or too large for TAVR, when the native coronary artery position was too low, when preoperative renal function was poor, or when the referring surgeon or cardiologist believed that TAVR posed too high a risk, then and usually only then would the patient be referred for AVB. We believe that a prospective randomized controlled trial is necessary to compare TAVR and AVB on a larger scale. We believe that AVB is an excellent operation to perform at high- or low-volume institutions and for all patients who wish to avoid the complications of perivalvular leak, heart block, renal failure, and stroke that are associated with TAVR. This project was supported by the Indiana University Health Indiana University School of Medicine Strategic Research Initiative.

References 1. Varadarajan P, Kapoor N, Bansal RC, Pai RG. Clinical profile and natural history of 453 nonsurgically managed patients with severe aortic stenosis. Ann Thorac Surg 2006;82:2111–5. 2. Kelly TA, Rothbart RM, Cooper CM, et al. Comparison of outcome of asymptomatic to symptomatic patients older than 20 years of age with valvular aortic stenosis. Am J Cardiol 1988;61:123–30. 3. Bouma BJ, van Den Brink RB, van Der Meulen JH, et al. To operate or not on elderly patients with aortic stenosis: the decision and its consequences. Heart 1999;82:143–8. 4. Iung B, Cachier A, Baron G, et al. Decision-making in elderly patients with severe aortic stenosis: why are so many denied surgery? Eur Heart J 2005;26:2714–20. 5. Thourani VH, Myung R, Kilgo P, et al. Long-term outcomes after isolated aortic valve replacement in octogenarians: a modern perspective. Ann Thorac Surg 2008;86:1458–64; discussion 1464–5. 6. Gammie JS, Brown JW, Brown JM, et al. Aortic valve bypass for the high-risk patient with aortic stenosis. Ann Thorac Surg 2006;81:1605–10.



7. Brown JW. Off-pump aortic valve bypass using a valved apical-aortic conduit. Oper Tech Thorac Cardiovasc Surg 2007;12:85–94. 8. Leon MB, Smith CR, Mack M, et al. Transcatheter aortic-valve implantation for aortic stenosis in patients who cannot undergo surgery. N Engl J Med 2010;363: 1597–607. 9. Wood S. TAVR Deaths at Five Years in PARTNER: Sobering and Remarkable. 2014, Medscape: Heartwire. Available at: Accessed November 5, 2014. 10. Thourani VH, Ailawadi G, Szeto WY, et al. Outcomes of surgical aortic valve replacement in high-risk patients: a multiinstitutional study. Ann Thorac Surg 2011;91:49–55; discussion 55–6. 11. Smith CR, Leon MB, Mack MJ, et al. Transcatheter versus surgical aortic-valve replacement in high-risk patients. N Engl J Med 2011;364:2187–98. 12. Miller DC, Blackstone EH, Mack MJ, et al. Transcatheter (TAVR) versus surgical (AVR) aortic valve replacement: occurrence, hazard, risk factors, and consequences of neurologic events in the PARTNER trial. J Thorac Cardiovasc Surg 2012;143:832–843 e13.

Ann Thorac Surg 2015;-:-–-

13. Tay EL, Gurvitch R, Wijesinghe N, et al. A high-risk period for cerebrovascular events exists after transcatheter aortic valve implantation. JACC Cardiovasc Interv 2011;4:1290–7. 14. Kapadia SR, Tuzcu EM, Makkar RR, et al. Long-term outcomes of inoperable patients with aortic stenosis randomly assigned to transcatheter aortic valve replacement or standard therapy. Circulation 2014;130:1483–92. 15. Stauffer CE, Jeudy J, Ghoreishi M, et al. Magnetic resonance investigation of blood flow after aortic valve bypass (apicoaortic conduit). Ann Thorac Surg 2011;92: 1332–7; discussion 1337–8. 16. Lerakis S, Hayek SS, Douglas PS. Paravalvular aortic leak after transcatheter aortic valve replacement: current knowledge. Circulation 2013;127:397–407. 17. Reynolds MR, Magnuson EA, Wang K, et al. Cost-effectiveness of transcatheter aortic valve replacement compared with standard care among inoperable patients with severe aortic stenosis: results from the Placement of Aortic Transcatheter Valves (PARTNER) trial (cohort B). Circulation 2012;125: 1102–9. 18. Makkar RR, Fontana GP, Jilaihawi H, et al. Transcatheter aortic-valve replacement for inoperable severe aortic stenosis. N Engl J Med 2012;366:1696–704.

DISCUSSION DR FAISAL BAKAEEN (Houston, TX): I would like to thank the Association for inviting me to discuss this very interesting and provocative paper. Dr Brown, you and your colleagues should be commended on your important work over the years in refining aortic valve bypass and for your excellent outcomes. In this study of high-risk patients with severe aortic stenosis, both TAVR and aortic valve bypass were effective in treating aortic stenosis, and hospital mortality and short-term outcomes were similar. You have nicely outlined the main advantages of aortic valve bypass, which include avoidance of TAVR-associated complications such as paravalvular leaks, heart block, coronary occlusions, and root rupture. In addition, aortic valve bypass does not require a hybrid room and is less resource intensive; therefore, not surprisingly, it is less expensive than TAVR, at least for now. It is important to note that there were significant differences between the TAVR and the aortic valve bypass cohort. The TAVR patients were older but had lower STS scores and, surprisingly, a low prevalence of congestive heart failure, only 16%! Interestingly, the transapical approach was performed more frequently than the transfemoral approach, and this goes against the direction nationally and internationally where the transfemoral approach is usually the more prevalent approach. I have two questions. My first question has to do with patient selection for TAVR versus aortic valve bypass. How do you factor in patient frailty, and could you please comment on the role of the heart team in the decision-making process? DR BROWN: First of all, all of these patients are evaluated by a group of cardiologists and surgeons, and the patients who are being sent to me now for aortic valve bypass are usually the TAVR rejects for one of the indications listed earlier. Patients had a previous renal transplant and have a creatinine of 2 or have a porcelain ascending aorta, although that is not a contraindication to TAVR, or have bypass grafts coming off the root or a number of other factors. I am seeing more referrals for aortic valve bypass since we started the TAVR program rather than a decrease. So it seems to be complementary to TAVR in many situations.

DR BAKAEEN: As you know, since the FDA approved the first TAVR device in November 2011, there has been a rapid dissemination of TAVR in the United States, and thousands of TAVR cases have been performed here and worldwide. In contrast, aortic valve bypass has been around for decades but is rarely performed and is limited to a few select centers such as yours. The low adoption rate for aortic valve bypass is probably related to the technical difficulty of the procedure, its steep learning curve, and its invasiveness. Although you have successfully performed aortic valve bypass off-pump, most surgeons feel more comfortable using cardiopulmonary bypass. Don’t you think that aortic valve bypass is a harder sell to the patient and the referring cardiologist when a percutaneous option that typically does not require cardiopulmonary bypass is available and feasible? In view of that, where do you see the future of aortic valve bypass both at your institution and beyond? Thank you and congratulations. DR BROWN: Well, thank you, Dr Bakaeen, for those provocative questions. I think that aortic valve bypass, if it could be done with instruments that don’t relate to bleeding, I think most everybody is afraid to cut a hole in a beating left ventricle. I just have been doing it for 40 years and have gotten used to it and it seems to be very straightforward and easy to do. There are devices that are coming on the market that will make the aortic anastomosis really quite simple that I think every cardiac surgeon would feel very comfortable in using. However, it is not available in the US today. So I think the aortic valve bypass can compete favorably. It may initially be used in those patients who have contraindications to TAVR or patients who don’t want to risk a neurologic event or renal failure. So I think it is going to have its place both tomorrow and 10 years from now in this very high-risk patient population, and I think we have been able to demonstrate that the outcomes, given the sicker patient population, seem to be just as good as for TAVR. It is more invasive, without question. DR VINOD THOURANI (Atlanta, GA): First of all, I want to thank John, because I do aortic valve bypasses, too, as you know. We have published and presented at the Southern regarding that

Ann Thorac Surg 2015;-:-–-

procedure. It was John’s paper that I actually read in order to do aortic valve bypass before TAVR came in. So, John, we owe you a lot of gratitude for really allowing us to reproduce your operation. So thank you for that. A couple of comments that I just want to make on this presentation is specifically that some of the comparisons you make are with the PARTNER B study and that is a little difficult for me, because you are comparing operable patients in the aortic valve bypass operation, and the TAVR PARTNER B population is an inoperable patient population. So we should expect to see a higher mortality risk and worse outcomes in those inoperative TAVR patients. So I think that we have to be a little careful comparing inoperable patients with a 73-year-old operative patient who underwent an aortic valve bypass.



But the other thing I wanted to raise is your cost comparison. We now looked at over 200 patients and their financial data for transcatheter valve and surgical valve patients, and for transcatheter valve patients right now for transfemoral performed in a minimalist technique (not for transapical though); our contribution margin is roughly $25,000 to $28,000 at Emory. This is beneficial for the hospital. Congratulations again on a great presentation. DR BROWN: Thank you for your comments. The thing that I want to do when I go home is to have our people at Indianapolis come down and figure out how you profit that much from your TAVR program, because it seems to be a difficult thing for our hospital system.