ORIGINAL ARTICLE
Aortic Valve Replacement via Right Minithoracotomy Versus Median Sternotomy A Propensity Score Analysis Donald D. Glower, MD, Bhargavi S. Desai, BS, G. Chad Hughes, MD, Carmelo A. Milano, MD, and Jeffrey G. Gaca, MD
Objective: The aim of this study was to define the relative role of a right minithoracotomy (RT) versus standard median sternotomy (ST) for open aortic valve replacement (AVR). Methods: A retrospective analysis was performed of all 1348 patients undergoing isolated, open AVR at a single institution during a 14-year period. Because relatively few patients were technically suitable for redo AVR with the RT approach (n = 20), all redo patients (n = 209) were excluded, leaving 1139 patients available for analysis. Patients converting from RT to ST approach (n = 15) were analyzed separately. Results: Relative to ST (n = 672), the RT patients (n = 452) were older with more stenosis but with more recent operation year, lower rate of congestive heart failure, higher ejection fraction, lower rate of endocarditis, and lower rate of renal disease than the ST AVR patients (all P G 0.0001). Right minithoracotomy AVR was associated with longer cardiopulmonary bypass times [157 (25) vs 131 (38), P = 0.0004] and clamp times [103 (20) vs 85 (27), P G 0.0001] but less transfusion (1.4 vs 3.4 U, P = 0.0003), less chest tube output (405 vs 950 mL, P G 0.0001), fewer reoperations for bleeding (0.4% vs 4%, P G 0.0001), shorter length of stay (6 vs 8 days, P = 0.03), and lower rate of atrial fibrillation (15% vs 20%, P = 0.03). Stroke, operative mortality, and survival were not significantly different between the groups. Conclusions: Given the biases of retrospective propensity-adjusted analysis, these data suggest that RT AVR is a safe alternative to ST AVR in selected patients, with advantages of avoiding sternotomy with associated bleeding, transfusion, and delayed wound healing, at the expense of longer pump and clamp times. Key Words: Aortic valve replacement, Minimally invasive, Propensity score analysis, Right thoracotomy. (Innovations 2014;9:75Y81) Accepted for publication December 30, 2013. From the Department of Surgery, Duke University Medical Center, Durham, NC USA. Presented at the Annual Scientific Meeting of the International Society for Minimally Invasive Cardiothoracic Surgery, June 12Y15, 2013, Prague, Czech Republic. Disclosure: The authors declare no conflicts of interest. Address correspondence and reprint requests to Donald D. Glower, MD, Department of Surgery, Duke University Medical Center, Box 3851, Durham, NC 27710 USA. E-mail: [email protected]. Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery ISSN: 1556-9845/14/0902-0075
Innovations & Volume 9, Number 2, March/April 2014
A
fter the first description of aortic valve replacement (AVR) via right thoracotomy (RT) by Rao and Kumar1 in 1993, the first large series of AVR via thoracotomy was reported by Cosgrove and Sabik2 using a right parasternal approach. Galloway and colleagues3 reported the first large series of a transverse right minithoracotomy (RT) for AVR in 1997. To date, most AVRs are performed via standard median sternotomy (ST). The most predominant minimally invasive approach to the aortic valve is the partial sternotomy,4,5 which still has disadvantages of dividing the sternum, despite a smaller incision. After the work of Galloway and colleagues,3 several other reports have demonstrated excellent results from the RT approach in series of 71 to 436 patients.6Y11 However, the few reports that directly compare RT versus ST AVR either combine RT and partial sternotomy approaches12,13 or are underpowered by 100 patients or fewer in the RT group.14Y20 The availability of transaortic insertion of percutaneous valves21 and the availability of sutureless aortic valve prostheses22,23 have revived interest in the RT for AVR.
METHODS With institutional review board approval, the records of all 1348 patients undergoing isolated AVR with stented prostheses during a 14-year period were retrospectively reviewed. Because only 20 patients underwent reoperative AVR via RT approach during that time, all 209 patients undergoing reoperative AVR (defined as any prior cardiac operation) were excluded. This left for analysis a total of 672 ST patients, 452 RT patients, and 15 patients who underwent initial RTapproach that was converted to ST. Patients were selected for the RT approach on the basis of either patient preference or surgeon preference. Patients with an ascending aorta diameter of greater than 4.2 cm, patients with previous sternotomy, or patients with a chest computed tomographic scan suggesting poor access to the ascending aorta via RT underwent ST. Standard median sternotomy AVR was performed as a full sternotomy using cannulation of the ascending aorta and the right atrium and anterograde and retrograde cardioplegia in most cases. The RT patients were positioned supine with a singlelumen endotracheal tube. An 8-cm right anterior minithoracotomy was made over the right third rib and carried into the second interspace in most patients, with detachment of the right second and third cartilages flush with the sternum. The right
75
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Innovations & Volume 9, Number 2, March/April 2014
Glower et al
TABLE 1. Demographics Demographics Age Operation year Male Weight, kg Smoking Diabetes Renal disease Hypertension Lung disease Cerebrovascular disease Peripheral vascular disease Coronary disease Ejection fraction Aortic stenosis Aortic regurgitation 91+ NYHA class Atrial fibrillation Urgent Aortic etiology Rheumatic Bicuspid Degenerative Infection Calcification
RT = 452
ST = 672
CONV = 15
P
67 (14) (20Y90)* 2005/2008/2012* 248 (55%) 84 (0.20) 130 (29%) 89 (20%) 37 (8%)* 376 (83%)* 53 (12%)* 29 (6%) 18 (4%) 46 (10%) 55 (9)* 398 (88%)* 141 (31%)* 2/2/3 60 (13%) 45 (10%)*
63 (14) (22Y92) 1999/2003/2008 388 (58%) 87 (24) 203 (30%) 157 (23%) 117 (17%) 508 (76%) 50 (7%) 55 (8%) 45 (7%) 51 (8%) 52 (12) 489 (73%) 294 (44%) 2/3/4 84 (13%) 222 (33%)
65 + 16 (41Y87) 2007/2008/2009* 5 (33%) 89 + 25 4 (27%) 3 (20%) 3 (20%) 15 (100%) 4 (27%)* 0 (0%) 2 (13%) 1 (7%) 53 + 10 14 (93%) 6 (40%) 2/2/3 0 (0%) 2 (13%)
9 (2%) 222 (49%)* 33 (7%)* 8 (2%) 180 (40%)
17 (3%) 205 (31%) 114 (17%) 60 (9%) 259 (39%)
1 (7%) 6 (40%) 1 (7%) 0 (0%) 7 (47%)
G0.0001 G0.0001 NS NS NS NS G0.0001 0.001 0.004 NS NS NS G0.0001 G0.0001 0.0001 G0.0001 NS G0.0001 G0.0001 NS G0.0001 0.02 G0.0001 NS
*P G 0.05 versus ST. CONV indicates conversion from RT to ST; NS, not significant; NYHA, New York Heart Association; RT, right minithoracotomy; ST, standard median sternotomy.
internal mammary artery and vein were divided. Cardiopulmonary bypass was initiated via the right femoral vein using a 25F or 28F percutaneous catheter. Cannulation of the ascending aorta was generally performed through the thoracotomy using a 15F to 19F wire wound arterial cannula. A 12F left ventricular vent was introduced into the left ventricle via the right superior pulmonary vein; a retrograde cardioplegia catheter was placed into the coronary sinus via the right atrium. The field was flooded with carbon dioxide at 2 L/min. The ascending aorta was clamped using a flexible clamp, and the heart was arrested using both anterograde cardioplegia via the ascending aorta and retrograde cardioplegia through the coronary sinus catheter. The ascending aorta was opened 1 to 2 cm above the right coronary ostium. The aortic valve was excised and replaced in standard fashion using standard stented prostheses. The heart was de-aired, the aortic clamp was removed, and the patient was weaned from cardiopulmonary bypass generally using temporary atrial epicardial pacing wires. In most cases, temporary right ventricular epicardial pacing wires could be placed if needed. Once the hemostasis was obtained, the second and third ribs were repaired using figure-of-eight No. 4 stainless steel wires, the ribs were approximated using No. 1 polyglyconate, and the thoracotomy was repaired in multiple layers. A single 19F Silastic drain was placed in the pericardium exited through the right pleural space. Two large 32F or 36F drains were placed in the pleural space for 12 hours. Selected patients underwent placement of a right paravertebral block catheter before anesthetic induction. The Silastic drain was removed in 4 days.
76
Variable definitions were in accord with those of The Society of Thoracic Surgeons guidelines. Continuous variables were compared by Student t test. Discrete variables were compared via W2 test. Patient survival was computed via the Kaplan-Meier technique, and group survivals were compared using the Cox-Mantel test. Cox proportional hazards model was used to determine independent predictors of survival using forward and backward analysis, including those variables found to be significant univariable predictors of survival at the P = 0.1 level. Propensity score analysis was performed using a logistic regression score for propensity to RT versus ST approach. For propensity score analysis of survival, transfusion, chest tube output, and new atrial fibrillation, the patients were broken into quintiles of similar propensity for RT approach. Data are presented as mean (SD) or as 25th/median/ 75th percentiles.
RESULTS A total of 452 patients underwent RT approach, and 672 underwent ST AVR. With experience, one surgeon ultimately completed 446 (96%) of 463 of selected, isolated, first-time AVR using RT approach. The RT and ST patients differed significantly, although these differences were of relatively small magnitude (Table 1). Older patients with more recent operative year, lung disease, hypertension, bicuspid disease, and aortic stenosis versus aortic regurgitation tended to get RT approach. The patients tended to get ST approach if they were urgent or had endocarditis, renal disease, or lower ejection fraction.
Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Innovations & Volume 9, Number 2, March/April 2014
Thoracotomy Versus Sternotomy AVR
TABLE 2. Operative Characteristics Characteristic Femoral cannulation Axillary cannulation Clamp time Pump time Valve size Mechanical valve
RT = 452
ST = 672
CONV = 15
P
15 (3%) 5 (1%) 103 (20)* 157 (25)* 23 (2) 81 (18%)*
9 (1%)* 2 (0.3%) 85 (27) 131 (38) 23 (3) 269 (41%)
1 (7%) 0 (0%) 90 + 18 148 + 29 22 + 2 2 (13%)
0.04 NS G0.0001 G0.0001 NS G0.0001
*P G 0.05 versus ST. CONV indicates conversion from RT to ST; NS, not significant; RT, right minithoracotomy; ST, standard median sternotomy.
The RT patients were slightly more likely to have femoral artery cannulation, although this approach involved only 3% of RT cases (Table 2). Aortic clamp times and cardiopulmonary bypass times were significantly longer in the RT versus ST patients. The RT patients were significantly more likely to receive a bioprosthesis. One surgeon accounted for 450 (99%) of 452 RT cases, whereas 629 (94%) of 672 ST cases were performed by nine surgeons. The RT approach was attempted but converted to ST in 15 patients. Conversion from RT to ST approach generally occurred because of unfavorable mediastinal anatomy, such as short aorta, aortic disease, or leftward displaced aorta. Length of stay (LOS) was significantly shorter in the RT versus ST patients (Table 3). Right thoracotomy was most dramatically associated with less chest tube output, less transfusion, and fewer reoperations for bleeding. Right thoracotomy also had benefits in lower rates of respiratory failure; new atrial fibrillation; and renal injury, defined as a rise in serum creatinine to 1 mg/dL higher than preoperative levels. No mediastinitis occurred in the RT group, but N was too small for significance. Two early RT patients developed late chest wall hernias that did not need treatment, but this complication has not recurred after using No. 1 polyglyconate pericostal
sutures. Stroke, new pacemaker, and operative mortality were not different between RT versus ST (Table 3). Follow-up was longer for the RT versus ST patients (1/3/5 vs 2/5/9 years, P G 0.0001). Late aortic valve reoperation was not different between the RT versus ST groups (Fig. 1), with most late reoperations being related to deterioration of biological prostheses. However, unadjusted patient survival was significantly higher in the RT group (Fig. 2A), with 10-year survival being 76% (5%) versus 57% (3%) for the RT versus ST groups (P G 0.0001). Univariable predictors of worse long-term survival included older age, earlier operation year, urgent operation, renal disease, lung disease, endocarditis, and New York Heart Association class III to IV, whereas RT approach and bicuspid disease improved survival (Table 4). By multivariable analysis using Cox proportional hazards model, independent predictors of worse long-term survival were older age, renal disease, and New York Heart Association class III to IV, whereas RT approach and bicuspid disease were independent predictors of improved long-term survival (Table 5). Although the number of patients converting from RT to ST was small (n = 15), outcomes were a little different from sternotomy, except that chest tube output was significantly higher than RT or ST in the conversion group, with a trend toward more transfusion than RT or ST. Conversion did not affect pump or clamp times or other outcomes.
Propensity Score Analysis Propensity score analysis revealed that the independent predictors of RT versus ST approach were later operative year, lung disease, higher ejection fraction, lower congestive heart failure class, nonurgent status, biological prosthesis, and bicuspid disease. For the 254 total deaths in this series, the patients were grouped into quintiles of equal propensity for RT versus ST approach. By propensity score analysis, independent predictors of less transfusion were RT approach (P = 0.0004), later
TABLE 3. In-Hospital Morbidity Propensity Adjusted In-Hospital Morbidity
RT = 452
ST = 672
CONV = 15
P
OR ($)
P
Length of stay, d Stroke Low cardiac output Wound infection Mediastinitis Pneumonia Respiratory failure Reoperation for bleeding Renal injury New atrial fibrillation New pacemaker Operative mortality Blood transfusion, U Chest tube output, mL
6 (4)* 7 (1.5%) 3 (0.7%) 0 (0%) 0 (0%) 7 (1.5%) 1 (0.2%)* 2 (0.4%) 5 (1.1%)* 66 (15%)* 14 (3%) 7 (1.5%) 0/0/2* 149/257/442*
8 (8) 8 (1.2%) 3 (0.4%) 5 (0.7%) 1 (0.1%) 8 (1.2%) 17 (2.5%) 27 (4%) 21 (3.1%) 136 (20%) 28 (4%) 22 (3.3%) 0/2/4 370/600/1010
7+3 1 (7%) 0 (0%) 1 (7%) 0 (0%) 1 (7%) 0 (0%) 0 (0%) 0 (0%) 4 (27%) 0 (0%) 0 (0%) 1/3/8* 523/760/952
G0.0001 NS NS NS NS NS 0.001 G0.0001 0.03 0.04 NS NS G0.0001 G0.0001
j1.0 ($)
0.03
0.18 0.64
j1.0 ($) j531 ($)
NS 0.03 NS 0.03
0.0004 0.0003
$ indicates propensity-adjusted mean difference for RT versus ST; CONV, conversion from RT to ST; NS, not significant; OR, odds ratio of relative risk for RT versus ST; RT, right minithoracotomy; ST, standard median sternotomy.
Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery
77
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Innovations & Volume 9, Number 2, March/April 2014
Glower et al
FIGURE 1. Freedom from late redo aortic valve replacement after right minithoracotomy (RT) versus standard median sternotomy (ST) aortic valve replacement.
operation year, younger age, no renal disease, and no endocarditis (all P G 0.01). The RT approach reduced blood transfusion by 1.0 (0.3) U. Propensity score analysis showed that independent predictors of less chest tube output were RT approach, recent operation year, and no renal disease (all P G 0.0003). The RT approach reduced chest tube output by 531 (65) mL. By propensity score analysis, the independent predictors of
reoperation for bleeding were ST approach (P = 0.03) and earlier year of operation (P = 0.01). The RT approach reduced the likelihood of reoperation for bleeding by an estimated factor of 5.5. By propensity score analysis, the only independent predictors of postoperative new atrial fibrillation were ST approach (P = 0.03) and older ager (P G 0.0001). The RTapproach reduced the likelihood of new postoperative atrial fibrillation by a factor
FIGURE 2. Unadjusted (panel A) and adjusted survival (panel B) after right minithoracotomy (RT) versus standard median sternotomy (ST) aortic valve replacement.
78
Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Innovations & Volume 9, Number 2, March/April 2014
Thoracotomy Versus Sternotomy AVR
TABLE 4. Univariable Predictors of Long-term Mortality RT Operation year Age Urgent Renal Lung SBE Bicuspid CHF34
P
Odds Ratio
95% CI
G0.0001 0.0004 G0.0001 G0.0001 G0.0001 0.03 0.048 G0.0001 G0.0001
0.50 0.94 1.040 1.81 3.40 1.51 1.60 0.38 1.96
0.35Y0.70 0.91Y0.97 1.03Y1.05 1.39Y2.35 2.56Y4.53 1.04Y2.18 1.004Y2.56 0.28Y0.52 1.48Y2.60
CHF34 indicates New York Heart Association class III to IV; CI, confidence interval; RT, right minithoracotomy; SBE, endocarditis.
of 1.6. Propensity score analysis showed that the independent predictors of renal injury were preoperative renal disease (P G 0.0001) and endocarditis (P = 0.03) but not RT or ST approach. Independent predictors of longer LOS were ST approach (P = 0.03), older age (P G 0.0001), and endocarditis (P G 0.0001). The RT approach shortened LOS by 1.00 + 0.45 days. By propensity score analysis, RT approach did not affect respiratory failure (P = 0.3). By propensity score analysis, the only independent predictors of worse long-term survival were older age (P G 0.0001), renal disease (P G 0.0001), and valve etiology other than bicuspid disease (P = 0.008). The RT approach was not a predictor of survival (P = 0.13) (Fig. 2B).
DISCUSSION Although prior reports of RT AVR have described favorable results, comparison with ST AVR has been difficult. Most studies reporting significant numbers of RT AVR have reported low mortality and low rates of bleeding, infection, or transfusion with early return to normal activity.6Y10 However, of six institutions reporting more than 100 cases of RT AVR,6,7,9,10,13,14 only two13,14 had control groups of ST AVR. Of these two controlled studies, one included partial sternotomy patients,13 and the other was underpowered with 117 RT AVR patients.14 Taken as a whole, prior comparisons of RT versus ST AVR have variously found longer operative, clamp, and pump times with RT AVR but have inconsistently reported shorter LOS, shorter intensive care stay, shorter ventilation time, and less transfusion.12Y20 Arom et al24 compared RT AVR with partial sternotomy, finding that RT AVR had earlier return to work and lower rate of atrial fibrillation but longer operating times.
More data regarding minimally invasive versus ST approaches are available in the mitral valve literature. Minimally invasive mitral valve literature has found mixed results with increased mortality, increased stroke, and increased exploration for bleeding in large national databases25 but earlier mobilization, less transfusion, lower rate of infection, and no significantly increased incidence of stroke in high-volume institutional series.26 In the mitral literature, the difference between the national series and institutional series probably results from higher complication rates in low-volume centers that dominate the national series. The current study is one of few comparing the RT approach with ST for AVR, with some attempt to control for baseline differences in patient characteristics.12,13,18 Of these three comparative studies, two reports12,13 included partial sternotomy patients, and the other study18 included only 58 RT AVR patients. These data do suggest that, in high-volume institutions, the RT versus ST AVR can provide advantages of less transfusion, less bleeding, lower rate of atrial fibrillation, and shorter LOS. Given that femoral cannulation was seldom used in the current series, the differences in stroke did not seem to be significant between RT versus ST AVR. There has been some concern that the less invasive approaches are more susceptible to stroke because of air embolism or retrograde perfusion.25 The current study is limited by not being a prospective, randomized trial. Despite the efforts to control for selection bias by multivariable analysis or by propensity score analysis, selection biases in referring patients for ST versus RT AVR cannot be excluded. In addition, because most thoracotomies were performed by a single surgeon who performed few sternotomies, one cannot exclude other technical differences in patient or operative management beyond the thoracotomy itself to account for differences in outcome in this series. In addition, retrospective data analysis cannot exclude the effects of other variables such as improving results over time, which could make outcome after RT AVR seem to be falsely superior to those from ST AVR. Finally, this study did not examine patient satisfaction, time to return to normal activity, or quality of life. Although a prospective randomized trial might be helpful to better define advantages and disadvantages of the RT approach, a prospective randomized trial of this type will inherently be unblinded and will always reflect some biases in selecting a small subset of patients for randomization. From a technical viewpoint, the RT approach has advantages relative to ST. By avoiding sternotomy entirely, the potential issues of sternal wound infection and sternal nonunion are eliminated. Deep infection or nonunion of the RT incision is theoretically possible but was not seen in this series
TABLE 5. Multivariable Predictors of Long-term Mortality Propensity Adjusted RT Age Renal Bicuspid CHF34
P
Odds Ratio
95% CI
P
Odds Ratio
95% CI
0.008 G0.0001 G0.0001 0.003 0.007
0.62 1.035 2.72 0.61 1.50
0.43Y0.88 1.024Y1.047 2.03Y3.63 0.44Y0.84 1.11Y2.00
0.13 G0.0001 G0.0001 0.008 NS
0.73 1.04 2.76 0.64
0.49Y1.09 1.03Y1.05 2.06Y3.69 0.45Y0.89
CHF34 indicates New York Heart Association class III to IV; CI, confidence interval; NS, not significant.
Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery
79
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Innovations & Volume 9, Number 2, March/April 2014
Glower et al
FIGURE 3. Yearly case volume for right minithoracotomy (white bars) versus standard median sternotomy (black bars) aortic valve replacement.
of 452 patients. Although the patient numbers were too small for statistical significance, clinically significant wound infections tended to be less common in the RT patients, suggesting that the RT approach may have a role in patients at high risk for ST wound complications, such as obese patients with diabetes, patients on immunosuppression, malnourished patients, and patients with prior chest radiation therapy. From a technical viewpoint, the RT approach has a number of disadvantages. First, the operative field is small, with compromised access and exposure to the aortic valve and surrounding structures because of the small, deep operative field, potentially leading to compromised technical outcomes and longer learning curve.27,28 Increased operative times for RT AVR may increase intraoperative costs relative to ST or partial sternotomy.4 As a result of the limited access to most of the heart, most concurrent procedures such as coronary bypass grafting, replacement of the ascending aorta or the aortic root, or mitral or tricuspid surgery are largely precluded by this RT incision. Moving the incision to the third interspace has allowed combined mitral and aortic valve procedures29,30 and, occasionally, grafting of the right coronary artery. Although reoperative AVR was done in this series in highly selected patients,15 we have found that a prior right coronary graft or dense mediastinal adhesions make the RTapproach to AVR more difficult than we are willing to tolerate in most patients. With experience, division and repair of the second and third ribs have made this incision reproducible in more than 90% of first-time, isolated AVRs. Finally, these data suggest that conversion from the RT incision to ST may increase postoperative bleeding and operative time.31 In a patient who will not accept transfusion, one might select patients carefully to avoid likelihood of conversion to ST approach. In summary, AVR using the RT approach can, with experience, be safe and applicable to most primary isolated
80
AVRs. Relative to ST, the RT approach is associated with shorter LOS, less transfusion, lower rate of exploration for bleeding, less chest tube output, and lower rate of postoperative atrial fibrillation, all at the expense of longer clamp and pump times. The RT approach to AVR could be considered in institutions with sufficient case volume and may be attractive for future procedures such as transaortic placement of a transcatheter AVR or placement of a sutureless AVR device (Fig. 3).21Y23
REFERENCES 1. Rao PN, Kumar AS. Aortic valve replacement through right thoracotomy. Tex Heart Inst J. 1993;20:307Y308. 2. Cosgrove DM III, Sabik JF. Minimally invasive approach for aortic valve operations. Ann Thorac Surg. 1996;62:596Y597. 3. Galloway AC, Ribakove GH, Miller JS, et al. Minimally invasive portaccess valvular surgery: initial clinical experience. Circulation. 1997; 96:I-508. 4. Estrera AL, Reardon MJ. Current approaches to minimally invasive aortic valve surgery. Curr Opin Cardiol. 2000;15:91Y95. 5. Tabata M, Aranki SF, Fox JA, Couper GS, Cohn LH, Shekar PS. Minimally invasive aortic valve replacement in left ventricular dysfunction. Asian Cardiovasc Thorac Ann. 2007;15:225Y228. 6. Kort S, Applebaum RM, Grossi EA, et al. Minimally invasive aortic valve replacement: echocardiographic and clinical results. Am Heart J. 2001; 142:476Y481. 7. Plass A, Scheffel H, Alkadhi H, et al. Aortic valve replacement through a minimally invasive approach: preoperative planning, surgical technique, and outcome. Ann Thorac Surg. 2009;88:1851Y1856. 8. Iribarne A, Karpenko A, Russo MJ, et al. Eight-year experience with minimally invasive cardiothoracic surgery. World J Surg. 2010;34:611Y615. 9. Glower DD, Lee T, Desai B. Aortic valve replacement through right minithoracotomy in 306 consecutive patients. Innovations. 2010;5:326Y330. 10. Cunningham MJ, Berberian CE, Starnes VA. Is transthoracic minimally invasive aortic valve replacement too time-consuming for the busy cardiac surgeon? Innovations. 2011;6:10Y14.
Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Innovations & Volume 9, Number 2, March/April 2014
11. Grossi EA, Schwartz CF, Yu PJ, et al. High-risk aortic valve replacement: are the outcomes as bad as predicted? Ann Thorac Surg. 2008;85:102Y106. 12. Sharony R, Grossi EA, Saunders PC, et al. Minimally invasive aortic valve surgery in the elderly: a case-control study. Circulation. 2003;108(suppl 1): II43YII47. 13. Sharony R, Grossi EA, Saunders PC, et al. Propensity score analysis of a six-year experience with minimally invasive isolated aortic valve replacement. J Heart Valve Dis. 2004;13:887Y893. 14. Hisamoto K, Sakwa MP, Shannon FL. Which surgical approach is better for minimally invasive aortic valve replacement?: a comparison of outcomes between right anterior minithoracotomy and ministernotomy. Innovations. 2013;8:157. 15. Pineda AM, Santana O, Reyna J, Sarria A, Lamas GA, Lamelas J. Outcomes of reoperative aortic valve replacement via right mini-thoracotomy versus median sternotomy. J Heart Valve Dis. 2013;22:50Y55. 16. Ruttmann E, Gilhofer TS, Ulmer H, et al. Propensity score-matched analysis of aortic valve replacement by mini-thoracotomy. J Heart Valve Dis. 2010;19:606Y614. 17. Brinkman WT, Hoffman W, Dewey TM, et al. Aortic valve replacement surgery: comparison of outcomes in matched sternotomy and PORT ACCESS groups. Ann Thorac Surg. 2010;90:131Y135. 18. Wheatley GH III, Prince SL, Herbert MA, Ryan WH. Port-access aortic valve surgery: a technique in evolution. Heart Surg Forum. 2004; 7:E628YE631. 19. Wei L, Wang C, Shen J, et al. Minimally invasive aortic valve replacement: comparative study of three different approaches [abstract]. Innovations. 2013;8:136. 20. Sansone F, Punta G, Parisi F. Right minithoracotomy versus full sternotomy for the aortic valve replacement: preliminary results. Heart Lung Circ. 2012;21:169Y173. 21. Lardizabal JA, O’Neill BP, Desai HV, et al. The transaortic approach for transcatheter aortic valve replacement: initial clinical experience in the United States. J Am Coll Cardiol. 2013;61:2341Y2345.
Thoracotomy Versus Sternotomy AVR
22. Laufer G, Wiedemann D, Vadehra A, Rosenhek R, Binder T, Kocher A. Mini-thoracotomy for sutureless-rapid-deployment aortic valve replacement: initial single center experience [abstract]. Innovations. 2013;8:105. 23. Miceli A, Murzi M, Gilmanov D, et al. Early experience with Perceval S sutureless bioprostheses in patients undergoing minimally invasive aortic valve replacement via right anterior minithoracotomy [abstract]. Innovations. 2013;8:104. 24. Arom KV, Emery RW, Kshettry VR, Janey PA. Comparison between portaccess and less invasive valve surgery. Ann Thorac Surg. 1999;68: 1525Y1528. 25. Gammie JS, Zhao Y, Peterson ED, O’Brien SM, Rankin JS, Griffith BP. J. Maxwell Chamberlain Memorial Paper for adult cardiac surgery. Lessinvasive mitral valve operations: trends and outcomes from The Society of Thoracic Surgeons adult cardiac surgery database. Ann Thorac Surg. 2010;90:1401Y1410, 1410.e1. 26. Falk V, Cheng DC, Martin J, et al. Minimally invasive versus open mitral valve surgery: a consensus statement of the International Society of Minimally Invasive Coronary Surgery (ISMICS) 2010. Innovations. 2011;6:66Y76. 27. Bridgewater B, Steyn RS, Ray S, Hooper T. Minimally invasive aortic valve replacement through a transverse sternotomy: a word of caution. Heart. 1998;79:605Y607. 28. Glower DD, Siegel LC, Galloway AC, et al. Predictors of operative time in multicenter port-access valve registry: institutional differences in learning. Heart Surg Forum. 2001;4:40Y46. 29. Colvin SB, Grossi EA, Ribakove G, Galloway AC. Minimally invasive aortic and mitral valve operation. Oper Tech Thorac Cardiovasc Surg. 2000;5:212Y220. 30. Kypson AP, Glower DD. Port-access approach for combined aortic and mitral valve surgery. Ann Thorac Surg. 2002;73:1657Y1658. 31. Foghsgaard S, Schmidt TA, Kjaergard HK. Minimally invasive aortic valve replacement: late conversion to full sternotomy doubles operative time. Tex Heart Inst J. 2009;36:293Y297.
CLINICAL PERSPECTIVE This retrospective case series examined over 1300 patients undergoing isolated, open aortic valve replacement at a single institution over a 14-year period. The authors performed a propensity score analysis on 672 median sternotomy patients and 452 right mini-thoracotomy patients. The minimally invasive approach was associated with longer cardiopulmonary bypass and cross-clamp time but less reoperation for bleeding, shorter length of stay and less atrial fibrillation. There was no difference in stroke, operative mortality and survival between the groups. These data suggest that right mini-thoracotomy is a safe alternative in selected patients and may avoid morbidity at the expense of longer pump and cross-clamp times. This is another series of excellent reports from this superb group from Duke University Medical Center under Dr. Glower’s leadership. It provides important information to readers on a large series of patients. While it is subject to all the biases involved in retrospective propensity-adjusted analyses, it is an important contribution to the literature and another piece of evidence testifying to the advantages of a minimally invasive approach for aortic valve replacement.
Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery
81
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Aortic Valve Replacement via Right Minithoracotomy Versus Median Sternotomy A Propensity Score Analysis Donald D. Glower, MD, Bhargavi S. Desai, BS, G. Chad Hughes, MD, Carmelo A. Milano, MD, and Jeffrey G. Gaca, MD
Objective: The aim of this study was to define the relative role of a right minithoracotomy (RT) versus standard median sternotomy (ST) for open aortic valve replacement (AVR). Methods: A retrospective analysis was performed of all 1348 patients undergoing isolated, open AVR at a single institution during a 14-year period. Because relatively few patients were technically suitable for redo AVR with the RT approach (n = 20), all redo patients (n = 209) were excluded, leaving 1139 patients available for analysis. Patients converting from RT to ST approach (n = 15) were analyzed separately. Results: Relative to ST (n = 672), the RT patients (n = 452) were older with more stenosis but with more recent operation year, lower rate of congestive heart failure, higher ejection fraction, lower rate of endocarditis, and lower rate of renal disease than the ST AVR patients (all P G 0.0001). Right minithoracotomy AVR was associated with longer cardiopulmonary bypass times [157 (25) vs 131 (38), P = 0.0004] and clamp times [103 (20) vs 85 (27), P G 0.0001] but less transfusion (1.4 vs 3.4 U, P = 0.0003), less chest tube output (405 vs 950 mL, P G 0.0001), fewer reoperations for bleeding (0.4% vs 4%, P G 0.0001), shorter length of stay (6 vs 8 days, P = 0.03), and lower rate of atrial fibrillation (15% vs 20%, P = 0.03). Stroke, operative mortality, and survival were not significantly different between the groups. Conclusions: Given the biases of retrospective propensity-adjusted analysis, these data suggest that RT AVR is a safe alternative to ST AVR in selected patients, with advantages of avoiding sternotomy with associated bleeding, transfusion, and delayed wound healing, at the expense of longer pump and clamp times. Key Words: Aortic valve replacement, Minimally invasive, Propensity score analysis, Right thoracotomy. (Innovations 2014;9:75Y81) Accepted for publication December 30, 2013. From the Department of Surgery, Duke University Medical Center, Durham, NC USA. Presented at the Annual Scientific Meeting of the International Society for Minimally Invasive Cardiothoracic Surgery, June 12Y15, 2013, Prague, Czech Republic. Disclosure: The authors declare no conflicts of interest. Address correspondence and reprint requests to Donald D. Glower, MD, Department of Surgery, Duke University Medical Center, Box 3851, Durham, NC 27710 USA. E-mail: [email protected]. Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery ISSN: 1556-9845/14/0902-0075
Innovations & Volume 9, Number 2, March/April 2014
A
fter the first description of aortic valve replacement (AVR) via right thoracotomy (RT) by Rao and Kumar1 in 1993, the first large series of AVR via thoracotomy was reported by Cosgrove and Sabik2 using a right parasternal approach. Galloway and colleagues3 reported the first large series of a transverse right minithoracotomy (RT) for AVR in 1997. To date, most AVRs are performed via standard median sternotomy (ST). The most predominant minimally invasive approach to the aortic valve is the partial sternotomy,4,5 which still has disadvantages of dividing the sternum, despite a smaller incision. After the work of Galloway and colleagues,3 several other reports have demonstrated excellent results from the RT approach in series of 71 to 436 patients.6Y11 However, the few reports that directly compare RT versus ST AVR either combine RT and partial sternotomy approaches12,13 or are underpowered by 100 patients or fewer in the RT group.14Y20 The availability of transaortic insertion of percutaneous valves21 and the availability of sutureless aortic valve prostheses22,23 have revived interest in the RT for AVR.
METHODS With institutional review board approval, the records of all 1348 patients undergoing isolated AVR with stented prostheses during a 14-year period were retrospectively reviewed. Because only 20 patients underwent reoperative AVR via RT approach during that time, all 209 patients undergoing reoperative AVR (defined as any prior cardiac operation) were excluded. This left for analysis a total of 672 ST patients, 452 RT patients, and 15 patients who underwent initial RTapproach that was converted to ST. Patients were selected for the RT approach on the basis of either patient preference or surgeon preference. Patients with an ascending aorta diameter of greater than 4.2 cm, patients with previous sternotomy, or patients with a chest computed tomographic scan suggesting poor access to the ascending aorta via RT underwent ST. Standard median sternotomy AVR was performed as a full sternotomy using cannulation of the ascending aorta and the right atrium and anterograde and retrograde cardioplegia in most cases. The RT patients were positioned supine with a singlelumen endotracheal tube. An 8-cm right anterior minithoracotomy was made over the right third rib and carried into the second interspace in most patients, with detachment of the right second and third cartilages flush with the sternum. The right
75
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Innovations & Volume 9, Number 2, March/April 2014
Glower et al
TABLE 1. Demographics Demographics Age Operation year Male Weight, kg Smoking Diabetes Renal disease Hypertension Lung disease Cerebrovascular disease Peripheral vascular disease Coronary disease Ejection fraction Aortic stenosis Aortic regurgitation 91+ NYHA class Atrial fibrillation Urgent Aortic etiology Rheumatic Bicuspid Degenerative Infection Calcification
RT = 452
ST = 672
CONV = 15
P
67 (14) (20Y90)* 2005/2008/2012* 248 (55%) 84 (0.20) 130 (29%) 89 (20%) 37 (8%)* 376 (83%)* 53 (12%)* 29 (6%) 18 (4%) 46 (10%) 55 (9)* 398 (88%)* 141 (31%)* 2/2/3 60 (13%) 45 (10%)*
63 (14) (22Y92) 1999/2003/2008 388 (58%) 87 (24) 203 (30%) 157 (23%) 117 (17%) 508 (76%) 50 (7%) 55 (8%) 45 (7%) 51 (8%) 52 (12) 489 (73%) 294 (44%) 2/3/4 84 (13%) 222 (33%)
65 + 16 (41Y87) 2007/2008/2009* 5 (33%) 89 + 25 4 (27%) 3 (20%) 3 (20%) 15 (100%) 4 (27%)* 0 (0%) 2 (13%) 1 (7%) 53 + 10 14 (93%) 6 (40%) 2/2/3 0 (0%) 2 (13%)
9 (2%) 222 (49%)* 33 (7%)* 8 (2%) 180 (40%)
17 (3%) 205 (31%) 114 (17%) 60 (9%) 259 (39%)
1 (7%) 6 (40%) 1 (7%) 0 (0%) 7 (47%)
G0.0001 G0.0001 NS NS NS NS G0.0001 0.001 0.004 NS NS NS G0.0001 G0.0001 0.0001 G0.0001 NS G0.0001 G0.0001 NS G0.0001 0.02 G0.0001 NS
*P G 0.05 versus ST. CONV indicates conversion from RT to ST; NS, not significant; NYHA, New York Heart Association; RT, right minithoracotomy; ST, standard median sternotomy.
internal mammary artery and vein were divided. Cardiopulmonary bypass was initiated via the right femoral vein using a 25F or 28F percutaneous catheter. Cannulation of the ascending aorta was generally performed through the thoracotomy using a 15F to 19F wire wound arterial cannula. A 12F left ventricular vent was introduced into the left ventricle via the right superior pulmonary vein; a retrograde cardioplegia catheter was placed into the coronary sinus via the right atrium. The field was flooded with carbon dioxide at 2 L/min. The ascending aorta was clamped using a flexible clamp, and the heart was arrested using both anterograde cardioplegia via the ascending aorta and retrograde cardioplegia through the coronary sinus catheter. The ascending aorta was opened 1 to 2 cm above the right coronary ostium. The aortic valve was excised and replaced in standard fashion using standard stented prostheses. The heart was de-aired, the aortic clamp was removed, and the patient was weaned from cardiopulmonary bypass generally using temporary atrial epicardial pacing wires. In most cases, temporary right ventricular epicardial pacing wires could be placed if needed. Once the hemostasis was obtained, the second and third ribs were repaired using figure-of-eight No. 4 stainless steel wires, the ribs were approximated using No. 1 polyglyconate, and the thoracotomy was repaired in multiple layers. A single 19F Silastic drain was placed in the pericardium exited through the right pleural space. Two large 32F or 36F drains were placed in the pleural space for 12 hours. Selected patients underwent placement of a right paravertebral block catheter before anesthetic induction. The Silastic drain was removed in 4 days.
76
Variable definitions were in accord with those of The Society of Thoracic Surgeons guidelines. Continuous variables were compared by Student t test. Discrete variables were compared via W2 test. Patient survival was computed via the Kaplan-Meier technique, and group survivals were compared using the Cox-Mantel test. Cox proportional hazards model was used to determine independent predictors of survival using forward and backward analysis, including those variables found to be significant univariable predictors of survival at the P = 0.1 level. Propensity score analysis was performed using a logistic regression score for propensity to RT versus ST approach. For propensity score analysis of survival, transfusion, chest tube output, and new atrial fibrillation, the patients were broken into quintiles of similar propensity for RT approach. Data are presented as mean (SD) or as 25th/median/ 75th percentiles.
RESULTS A total of 452 patients underwent RT approach, and 672 underwent ST AVR. With experience, one surgeon ultimately completed 446 (96%) of 463 of selected, isolated, first-time AVR using RT approach. The RT and ST patients differed significantly, although these differences were of relatively small magnitude (Table 1). Older patients with more recent operative year, lung disease, hypertension, bicuspid disease, and aortic stenosis versus aortic regurgitation tended to get RT approach. The patients tended to get ST approach if they were urgent or had endocarditis, renal disease, or lower ejection fraction.
Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Innovations & Volume 9, Number 2, March/April 2014
Thoracotomy Versus Sternotomy AVR
TABLE 2. Operative Characteristics Characteristic Femoral cannulation Axillary cannulation Clamp time Pump time Valve size Mechanical valve
RT = 452
ST = 672
CONV = 15
P
15 (3%) 5 (1%) 103 (20)* 157 (25)* 23 (2) 81 (18%)*
9 (1%)* 2 (0.3%) 85 (27) 131 (38) 23 (3) 269 (41%)
1 (7%) 0 (0%) 90 + 18 148 + 29 22 + 2 2 (13%)
0.04 NS G0.0001 G0.0001 NS G0.0001
*P G 0.05 versus ST. CONV indicates conversion from RT to ST; NS, not significant; RT, right minithoracotomy; ST, standard median sternotomy.
The RT patients were slightly more likely to have femoral artery cannulation, although this approach involved only 3% of RT cases (Table 2). Aortic clamp times and cardiopulmonary bypass times were significantly longer in the RT versus ST patients. The RT patients were significantly more likely to receive a bioprosthesis. One surgeon accounted for 450 (99%) of 452 RT cases, whereas 629 (94%) of 672 ST cases were performed by nine surgeons. The RT approach was attempted but converted to ST in 15 patients. Conversion from RT to ST approach generally occurred because of unfavorable mediastinal anatomy, such as short aorta, aortic disease, or leftward displaced aorta. Length of stay (LOS) was significantly shorter in the RT versus ST patients (Table 3). Right thoracotomy was most dramatically associated with less chest tube output, less transfusion, and fewer reoperations for bleeding. Right thoracotomy also had benefits in lower rates of respiratory failure; new atrial fibrillation; and renal injury, defined as a rise in serum creatinine to 1 mg/dL higher than preoperative levels. No mediastinitis occurred in the RT group, but N was too small for significance. Two early RT patients developed late chest wall hernias that did not need treatment, but this complication has not recurred after using No. 1 polyglyconate pericostal
sutures. Stroke, new pacemaker, and operative mortality were not different between RT versus ST (Table 3). Follow-up was longer for the RT versus ST patients (1/3/5 vs 2/5/9 years, P G 0.0001). Late aortic valve reoperation was not different between the RT versus ST groups (Fig. 1), with most late reoperations being related to deterioration of biological prostheses. However, unadjusted patient survival was significantly higher in the RT group (Fig. 2A), with 10-year survival being 76% (5%) versus 57% (3%) for the RT versus ST groups (P G 0.0001). Univariable predictors of worse long-term survival included older age, earlier operation year, urgent operation, renal disease, lung disease, endocarditis, and New York Heart Association class III to IV, whereas RT approach and bicuspid disease improved survival (Table 4). By multivariable analysis using Cox proportional hazards model, independent predictors of worse long-term survival were older age, renal disease, and New York Heart Association class III to IV, whereas RT approach and bicuspid disease were independent predictors of improved long-term survival (Table 5). Although the number of patients converting from RT to ST was small (n = 15), outcomes were a little different from sternotomy, except that chest tube output was significantly higher than RT or ST in the conversion group, with a trend toward more transfusion than RT or ST. Conversion did not affect pump or clamp times or other outcomes.
Propensity Score Analysis Propensity score analysis revealed that the independent predictors of RT versus ST approach were later operative year, lung disease, higher ejection fraction, lower congestive heart failure class, nonurgent status, biological prosthesis, and bicuspid disease. For the 254 total deaths in this series, the patients were grouped into quintiles of equal propensity for RT versus ST approach. By propensity score analysis, independent predictors of less transfusion were RT approach (P = 0.0004), later
TABLE 3. In-Hospital Morbidity Propensity Adjusted In-Hospital Morbidity
RT = 452
ST = 672
CONV = 15
P
OR ($)
P
Length of stay, d Stroke Low cardiac output Wound infection Mediastinitis Pneumonia Respiratory failure Reoperation for bleeding Renal injury New atrial fibrillation New pacemaker Operative mortality Blood transfusion, U Chest tube output, mL
6 (4)* 7 (1.5%) 3 (0.7%) 0 (0%) 0 (0%) 7 (1.5%) 1 (0.2%)* 2 (0.4%) 5 (1.1%)* 66 (15%)* 14 (3%) 7 (1.5%) 0/0/2* 149/257/442*
8 (8) 8 (1.2%) 3 (0.4%) 5 (0.7%) 1 (0.1%) 8 (1.2%) 17 (2.5%) 27 (4%) 21 (3.1%) 136 (20%) 28 (4%) 22 (3.3%) 0/2/4 370/600/1010
7+3 1 (7%) 0 (0%) 1 (7%) 0 (0%) 1 (7%) 0 (0%) 0 (0%) 0 (0%) 4 (27%) 0 (0%) 0 (0%) 1/3/8* 523/760/952
G0.0001 NS NS NS NS NS 0.001 G0.0001 0.03 0.04 NS NS G0.0001 G0.0001
j1.0 ($)
0.03
0.18 0.64
j1.0 ($) j531 ($)
NS 0.03 NS 0.03
0.0004 0.0003
$ indicates propensity-adjusted mean difference for RT versus ST; CONV, conversion from RT to ST; NS, not significant; OR, odds ratio of relative risk for RT versus ST; RT, right minithoracotomy; ST, standard median sternotomy.
Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery
77
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Innovations & Volume 9, Number 2, March/April 2014
Glower et al
FIGURE 1. Freedom from late redo aortic valve replacement after right minithoracotomy (RT) versus standard median sternotomy (ST) aortic valve replacement.
operation year, younger age, no renal disease, and no endocarditis (all P G 0.01). The RT approach reduced blood transfusion by 1.0 (0.3) U. Propensity score analysis showed that independent predictors of less chest tube output were RT approach, recent operation year, and no renal disease (all P G 0.0003). The RT approach reduced chest tube output by 531 (65) mL. By propensity score analysis, the independent predictors of
reoperation for bleeding were ST approach (P = 0.03) and earlier year of operation (P = 0.01). The RT approach reduced the likelihood of reoperation for bleeding by an estimated factor of 5.5. By propensity score analysis, the only independent predictors of postoperative new atrial fibrillation were ST approach (P = 0.03) and older ager (P G 0.0001). The RTapproach reduced the likelihood of new postoperative atrial fibrillation by a factor
FIGURE 2. Unadjusted (panel A) and adjusted survival (panel B) after right minithoracotomy (RT) versus standard median sternotomy (ST) aortic valve replacement.
78
Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Innovations & Volume 9, Number 2, March/April 2014
Thoracotomy Versus Sternotomy AVR
TABLE 4. Univariable Predictors of Long-term Mortality RT Operation year Age Urgent Renal Lung SBE Bicuspid CHF34
P
Odds Ratio
95% CI
G0.0001 0.0004 G0.0001 G0.0001 G0.0001 0.03 0.048 G0.0001 G0.0001
0.50 0.94 1.040 1.81 3.40 1.51 1.60 0.38 1.96
0.35Y0.70 0.91Y0.97 1.03Y1.05 1.39Y2.35 2.56Y4.53 1.04Y2.18 1.004Y2.56 0.28Y0.52 1.48Y2.60
CHF34 indicates New York Heart Association class III to IV; CI, confidence interval; RT, right minithoracotomy; SBE, endocarditis.
of 1.6. Propensity score analysis showed that the independent predictors of renal injury were preoperative renal disease (P G 0.0001) and endocarditis (P = 0.03) but not RT or ST approach. Independent predictors of longer LOS were ST approach (P = 0.03), older age (P G 0.0001), and endocarditis (P G 0.0001). The RT approach shortened LOS by 1.00 + 0.45 days. By propensity score analysis, RT approach did not affect respiratory failure (P = 0.3). By propensity score analysis, the only independent predictors of worse long-term survival were older age (P G 0.0001), renal disease (P G 0.0001), and valve etiology other than bicuspid disease (P = 0.008). The RT approach was not a predictor of survival (P = 0.13) (Fig. 2B).
DISCUSSION Although prior reports of RT AVR have described favorable results, comparison with ST AVR has been difficult. Most studies reporting significant numbers of RT AVR have reported low mortality and low rates of bleeding, infection, or transfusion with early return to normal activity.6Y10 However, of six institutions reporting more than 100 cases of RT AVR,6,7,9,10,13,14 only two13,14 had control groups of ST AVR. Of these two controlled studies, one included partial sternotomy patients,13 and the other was underpowered with 117 RT AVR patients.14 Taken as a whole, prior comparisons of RT versus ST AVR have variously found longer operative, clamp, and pump times with RT AVR but have inconsistently reported shorter LOS, shorter intensive care stay, shorter ventilation time, and less transfusion.12Y20 Arom et al24 compared RT AVR with partial sternotomy, finding that RT AVR had earlier return to work and lower rate of atrial fibrillation but longer operating times.
More data regarding minimally invasive versus ST approaches are available in the mitral valve literature. Minimally invasive mitral valve literature has found mixed results with increased mortality, increased stroke, and increased exploration for bleeding in large national databases25 but earlier mobilization, less transfusion, lower rate of infection, and no significantly increased incidence of stroke in high-volume institutional series.26 In the mitral literature, the difference between the national series and institutional series probably results from higher complication rates in low-volume centers that dominate the national series. The current study is one of few comparing the RT approach with ST for AVR, with some attempt to control for baseline differences in patient characteristics.12,13,18 Of these three comparative studies, two reports12,13 included partial sternotomy patients, and the other study18 included only 58 RT AVR patients. These data do suggest that, in high-volume institutions, the RT versus ST AVR can provide advantages of less transfusion, less bleeding, lower rate of atrial fibrillation, and shorter LOS. Given that femoral cannulation was seldom used in the current series, the differences in stroke did not seem to be significant between RT versus ST AVR. There has been some concern that the less invasive approaches are more susceptible to stroke because of air embolism or retrograde perfusion.25 The current study is limited by not being a prospective, randomized trial. Despite the efforts to control for selection bias by multivariable analysis or by propensity score analysis, selection biases in referring patients for ST versus RT AVR cannot be excluded. In addition, because most thoracotomies were performed by a single surgeon who performed few sternotomies, one cannot exclude other technical differences in patient or operative management beyond the thoracotomy itself to account for differences in outcome in this series. In addition, retrospective data analysis cannot exclude the effects of other variables such as improving results over time, which could make outcome after RT AVR seem to be falsely superior to those from ST AVR. Finally, this study did not examine patient satisfaction, time to return to normal activity, or quality of life. Although a prospective randomized trial might be helpful to better define advantages and disadvantages of the RT approach, a prospective randomized trial of this type will inherently be unblinded and will always reflect some biases in selecting a small subset of patients for randomization. From a technical viewpoint, the RT approach has advantages relative to ST. By avoiding sternotomy entirely, the potential issues of sternal wound infection and sternal nonunion are eliminated. Deep infection or nonunion of the RT incision is theoretically possible but was not seen in this series
TABLE 5. Multivariable Predictors of Long-term Mortality Propensity Adjusted RT Age Renal Bicuspid CHF34
P
Odds Ratio
95% CI
P
Odds Ratio
95% CI
0.008 G0.0001 G0.0001 0.003 0.007
0.62 1.035 2.72 0.61 1.50
0.43Y0.88 1.024Y1.047 2.03Y3.63 0.44Y0.84 1.11Y2.00
0.13 G0.0001 G0.0001 0.008 NS
0.73 1.04 2.76 0.64
0.49Y1.09 1.03Y1.05 2.06Y3.69 0.45Y0.89
CHF34 indicates New York Heart Association class III to IV; CI, confidence interval; NS, not significant.
Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery
79
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Innovations & Volume 9, Number 2, March/April 2014
Glower et al
FIGURE 3. Yearly case volume for right minithoracotomy (white bars) versus standard median sternotomy (black bars) aortic valve replacement.
of 452 patients. Although the patient numbers were too small for statistical significance, clinically significant wound infections tended to be less common in the RT patients, suggesting that the RT approach may have a role in patients at high risk for ST wound complications, such as obese patients with diabetes, patients on immunosuppression, malnourished patients, and patients with prior chest radiation therapy. From a technical viewpoint, the RT approach has a number of disadvantages. First, the operative field is small, with compromised access and exposure to the aortic valve and surrounding structures because of the small, deep operative field, potentially leading to compromised technical outcomes and longer learning curve.27,28 Increased operative times for RT AVR may increase intraoperative costs relative to ST or partial sternotomy.4 As a result of the limited access to most of the heart, most concurrent procedures such as coronary bypass grafting, replacement of the ascending aorta or the aortic root, or mitral or tricuspid surgery are largely precluded by this RT incision. Moving the incision to the third interspace has allowed combined mitral and aortic valve procedures29,30 and, occasionally, grafting of the right coronary artery. Although reoperative AVR was done in this series in highly selected patients,15 we have found that a prior right coronary graft or dense mediastinal adhesions make the RTapproach to AVR more difficult than we are willing to tolerate in most patients. With experience, division and repair of the second and third ribs have made this incision reproducible in more than 90% of first-time, isolated AVRs. Finally, these data suggest that conversion from the RT incision to ST may increase postoperative bleeding and operative time.31 In a patient who will not accept transfusion, one might select patients carefully to avoid likelihood of conversion to ST approach. In summary, AVR using the RT approach can, with experience, be safe and applicable to most primary isolated
80
AVRs. Relative to ST, the RT approach is associated with shorter LOS, less transfusion, lower rate of exploration for bleeding, less chest tube output, and lower rate of postoperative atrial fibrillation, all at the expense of longer clamp and pump times. The RT approach to AVR could be considered in institutions with sufficient case volume and may be attractive for future procedures such as transaortic placement of a transcatheter AVR or placement of a sutureless AVR device (Fig. 3).21Y23
REFERENCES 1. Rao PN, Kumar AS. Aortic valve replacement through right thoracotomy. Tex Heart Inst J. 1993;20:307Y308. 2. Cosgrove DM III, Sabik JF. Minimally invasive approach for aortic valve operations. Ann Thorac Surg. 1996;62:596Y597. 3. Galloway AC, Ribakove GH, Miller JS, et al. Minimally invasive portaccess valvular surgery: initial clinical experience. Circulation. 1997; 96:I-508. 4. Estrera AL, Reardon MJ. Current approaches to minimally invasive aortic valve surgery. Curr Opin Cardiol. 2000;15:91Y95. 5. Tabata M, Aranki SF, Fox JA, Couper GS, Cohn LH, Shekar PS. Minimally invasive aortic valve replacement in left ventricular dysfunction. Asian Cardiovasc Thorac Ann. 2007;15:225Y228. 6. Kort S, Applebaum RM, Grossi EA, et al. Minimally invasive aortic valve replacement: echocardiographic and clinical results. Am Heart J. 2001; 142:476Y481. 7. Plass A, Scheffel H, Alkadhi H, et al. Aortic valve replacement through a minimally invasive approach: preoperative planning, surgical technique, and outcome. Ann Thorac Surg. 2009;88:1851Y1856. 8. Iribarne A, Karpenko A, Russo MJ, et al. Eight-year experience with minimally invasive cardiothoracic surgery. World J Surg. 2010;34:611Y615. 9. Glower DD, Lee T, Desai B. Aortic valve replacement through right minithoracotomy in 306 consecutive patients. Innovations. 2010;5:326Y330. 10. Cunningham MJ, Berberian CE, Starnes VA. Is transthoracic minimally invasive aortic valve replacement too time-consuming for the busy cardiac surgeon? Innovations. 2011;6:10Y14.
Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
Innovations & Volume 9, Number 2, March/April 2014
11. Grossi EA, Schwartz CF, Yu PJ, et al. High-risk aortic valve replacement: are the outcomes as bad as predicted? Ann Thorac Surg. 2008;85:102Y106. 12. Sharony R, Grossi EA, Saunders PC, et al. Minimally invasive aortic valve surgery in the elderly: a case-control study. Circulation. 2003;108(suppl 1): II43YII47. 13. Sharony R, Grossi EA, Saunders PC, et al. Propensity score analysis of a six-year experience with minimally invasive isolated aortic valve replacement. J Heart Valve Dis. 2004;13:887Y893. 14. Hisamoto K, Sakwa MP, Shannon FL. Which surgical approach is better for minimally invasive aortic valve replacement?: a comparison of outcomes between right anterior minithoracotomy and ministernotomy. Innovations. 2013;8:157. 15. Pineda AM, Santana O, Reyna J, Sarria A, Lamas GA, Lamelas J. Outcomes of reoperative aortic valve replacement via right mini-thoracotomy versus median sternotomy. J Heart Valve Dis. 2013;22:50Y55. 16. Ruttmann E, Gilhofer TS, Ulmer H, et al. Propensity score-matched analysis of aortic valve replacement by mini-thoracotomy. J Heart Valve Dis. 2010;19:606Y614. 17. Brinkman WT, Hoffman W, Dewey TM, et al. Aortic valve replacement surgery: comparison of outcomes in matched sternotomy and PORT ACCESS groups. Ann Thorac Surg. 2010;90:131Y135. 18. Wheatley GH III, Prince SL, Herbert MA, Ryan WH. Port-access aortic valve surgery: a technique in evolution. Heart Surg Forum. 2004; 7:E628YE631. 19. Wei L, Wang C, Shen J, et al. Minimally invasive aortic valve replacement: comparative study of three different approaches [abstract]. Innovations. 2013;8:136. 20. Sansone F, Punta G, Parisi F. Right minithoracotomy versus full sternotomy for the aortic valve replacement: preliminary results. Heart Lung Circ. 2012;21:169Y173. 21. Lardizabal JA, O’Neill BP, Desai HV, et al. The transaortic approach for transcatheter aortic valve replacement: initial clinical experience in the United States. J Am Coll Cardiol. 2013;61:2341Y2345.
Thoracotomy Versus Sternotomy AVR
22. Laufer G, Wiedemann D, Vadehra A, Rosenhek R, Binder T, Kocher A. Mini-thoracotomy for sutureless-rapid-deployment aortic valve replacement: initial single center experience [abstract]. Innovations. 2013;8:105. 23. Miceli A, Murzi M, Gilmanov D, et al. Early experience with Perceval S sutureless bioprostheses in patients undergoing minimally invasive aortic valve replacement via right anterior minithoracotomy [abstract]. Innovations. 2013;8:104. 24. Arom KV, Emery RW, Kshettry VR, Janey PA. Comparison between portaccess and less invasive valve surgery. Ann Thorac Surg. 1999;68: 1525Y1528. 25. Gammie JS, Zhao Y, Peterson ED, O’Brien SM, Rankin JS, Griffith BP. J. Maxwell Chamberlain Memorial Paper for adult cardiac surgery. Lessinvasive mitral valve operations: trends and outcomes from The Society of Thoracic Surgeons adult cardiac surgery database. Ann Thorac Surg. 2010;90:1401Y1410, 1410.e1. 26. Falk V, Cheng DC, Martin J, et al. Minimally invasive versus open mitral valve surgery: a consensus statement of the International Society of Minimally Invasive Coronary Surgery (ISMICS) 2010. Innovations. 2011;6:66Y76. 27. Bridgewater B, Steyn RS, Ray S, Hooper T. Minimally invasive aortic valve replacement through a transverse sternotomy: a word of caution. Heart. 1998;79:605Y607. 28. Glower DD, Siegel LC, Galloway AC, et al. Predictors of operative time in multicenter port-access valve registry: institutional differences in learning. Heart Surg Forum. 2001;4:40Y46. 29. Colvin SB, Grossi EA, Ribakove G, Galloway AC. Minimally invasive aortic and mitral valve operation. Oper Tech Thorac Cardiovasc Surg. 2000;5:212Y220. 30. Kypson AP, Glower DD. Port-access approach for combined aortic and mitral valve surgery. Ann Thorac Surg. 2002;73:1657Y1658. 31. Foghsgaard S, Schmidt TA, Kjaergard HK. Minimally invasive aortic valve replacement: late conversion to full sternotomy doubles operative time. Tex Heart Inst J. 2009;36:293Y297.
CLINICAL PERSPECTIVE This retrospective case series examined over 1300 patients undergoing isolated, open aortic valve replacement at a single institution over a 14-year period. The authors performed a propensity score analysis on 672 median sternotomy patients and 452 right mini-thoracotomy patients. The minimally invasive approach was associated with longer cardiopulmonary bypass and cross-clamp time but less reoperation for bleeding, shorter length of stay and less atrial fibrillation. There was no difference in stroke, operative mortality and survival between the groups. These data suggest that right mini-thoracotomy is a safe alternative in selected patients and may avoid morbidity at the expense of longer pump and cross-clamp times. This is another series of excellent reports from this superb group from Duke University Medical Center under Dr. Glower’s leadership. It provides important information to readers on a large series of patients. While it is subject to all the biases involved in retrospective propensity-adjusted analyses, it is an important contribution to the literature and another piece of evidence testifying to the advantages of a minimally invasive approach for aortic valve replacement.
Copyright * 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery
81
Copyright © 2014 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
