|
|
||||||||
a Division of Cardiothoracic and Vascular Surgery, Department of Surgery, Brody School of Medicine, East Carolina University, Greenville, North Carolina
b Division of Cardiology, Department of Medicine, Brody School of Medicine, East Carolina University, Greenville, North Carolina
c Department of Biostatistics, Brody School of Medicine, East Carolina University, Greenville, North Carolina
Accepted for publication April 30, 2007.
* Address correspondence to Dr Rodriguez, Division of Cardiothoracic and Vascular Surgery, East Carolina University, 600 Moye Blvd, Teaching Annex #257, Greenville, NC 27858 (Email: rodrigueze{at}ecu.edu).
Presented at the Fifty-third Annual Meeting of the Southern Thoracic Surgical Association, Tucson, AZ, Nov 8–11, 2006.
| Abstract |
|---|
|
|
|---|
Methods: Data were collected contemporaneously on 289 patients operated on from May 2000 to September 2006. Every patient underwent preoperative transesophageal echocardiography. Follow-up consisted of serial echocardiograms, clinic visits, and phone conversations with patients and their physicians.
Results: A total of 66 patients (anterior leaflet, n = 14; and bileaflet, n = 52) were identified. Mean age was 52.6 ± 7.1 years, and 57 (86%) patients had New York Heart Association functional class II or III symptoms. Cardiopulmonary bypass and cross-clamp times were 171 ± 52 and 132 ± 39 minutes, respectively. The 30-day and late mortality rates were 3% (n = 2) for each time point. There were no device-related or perfusion-related complications or sternotomy conversions. Complications included 2 strokes (3%), 2 bleeding reexplorations (3%), and 10 pleural effusions requiring intervention (15%). The length of hospital stay for surviving patients was 5 ± 3 days, and time to extubation averaged 9.5 ± 13 hours. A total of 6 (9%) patients required valve reoperation. Mean follow-up was 795 ± 495 days, and echocardiographic mitral regurgitation (n = 60) was none or trace (n = 35, 58.3%), mild (n = 19, 31.6%), moderate (n = 2, 3.3%), and severe (n = 4, 6.7%).
Conclusions: Robotic mitral valve repair for anterior leaflet and bileaflet prolapse is feasible and safe. Outcomes and degree of late mitral regurgitation are similar to series using conventional techniques. Long-term follow-up is required to formally address the efficacy of robotic repair techniques.
| Introduction |
|---|
|
|
|---|
Robot-assisted MV repair (RMVP), using a less invasive approach with the da Vinci surgical system (Intuitive Surgical, Inc, Sunnyvale, CA), has become an established technique, providing three-dimensional visualization, ergonomic dexterity, and enhanced global precision, when working in small spaces such as the left atrium [6, 7]. In a recent prospective, multicenter; phase II Food and Drug Administration trial, robot assistance was proven to be effective for PL repairs with 1-month echocardiographic results similar to published series using conventional repair techniques [7]. As incremental expertise and good results have been attained, RMVP indications have expanded to include patients with severe AL and BL prolapse, as a result of either redundant or ruptured chords. This report analyzes both efficacy and short-term clinical results of 66 consecutive robotic AL and BL prolapse MV repairs.
| Material and Methods |
|---|
|
|
|---|
Echocardiographic Operative Planning
All operative repairs were based on detailed intraoperative two-dimensional transesophageal echocardiographic studies (TEE) done in concert with an anesthesiologist. As MV dysfunction is best analyzed during dynamic cardiac activity, optimal repair strategies were planned before cardiopulmonary bypass from echocardiographic measurements, which included mid-commissural views to determine AL lengths and individual P1, P2, and P3 heights above (lengths from) the annulus. Both annular and left ventricular outflow tract diameters were measured. The long-axis left ventricular outflow tract dimensions were used to estimate both intertrigonal and intercommissural distances by using the formula intertrigonal distance equals left ventricular outflow tract divided by 0.8 [8]. Moreover, four-chamber TEE views were obtained to measure the degree of leaflet prolapse above the annular plane as well as to quantify levels of leaflet prolapse. Lastly, transgastric views were helpful for defining segmental areas of prolapse as well as to isolate regurgitant leaks. Using these metrics, chordal lengths, required for AL restoration and competency, were determined. In addition, the requisite extent of PL resections or height reductions was planned in most cases from these measurements. These data, combined with both AL and ventricular outflow tract measurements, helped us both predict and avoid systolic anterior motion of the AL (SAM) after weaning from cardiopulmonary bypass.
Transesophageal echocardiographic anterior and PL lengths in many cases were confirmed by linear measurements made using the robot in the arrested heart. Intercommissural, intertrigonal, and transverse diameters of each Cosgrove-Edwards annuloplasty band sizer were measured, and a nomogram was constructed to compare these data with TEE-derived measurements. Using this methodology, the minimal acceptable band size was determined. In patients with severe Barlows disease (AL > 3.5 cm), bands were often oversized by one step (eg, 36 mm measured, 38 mm used) to assure the avoidance of SAM. As illustrated below, each BL repair was combined with PL height reduction.
Operative Technique
Details of the da Vinci surgical system set-up, instrument arm placement, and operative approach for robotic MV repairs at our institution have been published previously [9]. The most important technologic addition subsequent to most of these reports has been the incorporation of a robot-driven, left atrial retractor that functions by means of Endowrist (Intuitive Surgical, Inc) technology [10]. The left atrial retractor is activated by a fourth robotic arm, which is placed through a 0.8-cm entry port, positioned 1 to 2 cm lateral to the right internal thoracic vein and one intercostal space cephalad to the 3- to 4-cm working or camera incision. The left atrial retractor not only enabled ideal exposure for troublesome trigonal sutures and papillary muscle neochord implantations but also could be relaxed during valve saline testing.
Our basic robotic repair technique for patients with severe BL prolapse (Barlows disease) is shown in Figures 1 and 2.
Technical details of our RMVP for BL, isolated AL, and isolated PL prolapse have been illustrated in other publications [9]. In every patient a Cosgrove-Edwards annuloplasty band was used for the reduction annuloplasty. Briefly, for BL repairs a large P2 resection is carried to the annulus with chords left intact along the anterior one fourth of the resected segment. These P2 chords, along with an attached valve tissue segment, are transferred to either ventricular or atrial side of A2. By changing the chordal angle of the posterior chord with reference to the papillary muscle origin, radial rotation to the new AL site, de facto, reduces the segmental prolapse. The degree of reduction also is governed by the attachment distance from the coapting edge of A2.
|
|
Data Collection and Follow-Up
Preoperative, intraoperative, and postoperative data were collected prospectively and recorded in a graphic spreadsheet. Follow-up included scheduled clinic visits along with extensive telephone communication with cardiologists and primary care physicians, as well as with the patients themselves. In addition, operative notes from reoperations were studied.
We recommended follow-up transthoracic echocardiograms at 3 to 6 months and 12 months, as well as yearly thereafter. Survival data were ascertained in all 66 patients, and postdischarge echocardiogram results were obtained in 60 patients (91%). Three patients were lost to echocardiographic follow-up, 1 patient underwent reoperation, and 2 patients died during the initial hospitalization. Every attempt was made to follow all patients.
Statistical Analysis
Data are expressed as mean ± standard deviation. The Students t test results are shown for the comparison of operative times between AL and BL groups. The Kaplan–Meier curves were calculated in R version 2.4.1 (The R Foundation for Statistical Computing, Sunnyvale, CA, ISBN 3-900051-07-0) using the survival package (version 2.30).
| Results |
|---|
|
|
|---|
|
|
There were no cases of device-related or perfusion complications. Postoperative complications are shown in Table 3. The average time to extubation (n = 63, surviving patients) was 9.5 ± 5 hours, and 60 patients (95%) were ventilator free within 24 hours of surgery although their mean hospitalization was 5 ± 3 days. Seven (11%) patients were readmitted within 30 days of surgery for either pulmonary complications (n = 3), gastrointestinal distress (n = 1), dysrhythmias (n = 1), pericardial effusion (n = 1), or valve dysfunction requiring reoperation (n = 1).
|
|
|
|
| Comment |
|---|
|
|
|---|
There is no doubt that AL and BL prolapse repairs are challenging independently of whether a sternotomy or robotic approach is used. However, it is clear from our operative times that these procedures take longer than when performed using conventional techniques. Nonetheless, our cardiopulmonary bypass and cross-clamp times for these complex repairs are similar to other robotic MV series that included a more heterogeneous group of MV disease [13]. Despite longer cross-clamp times as compared with sternotomy repairs, there was only 1 in-hospital death secondary to pump failure, and the overall results were comparable to published national data [14]. In addition, the rate of postoperative morbidities was low and similar to other series.
Reported MV repair series have demonstrated excellent long-term results with an 80% to 93% freedom from reoperation between 10 and 20 years [1, 2, 5, 15]. However, BL prolapse and especially AL prolapse have been associated with increased reoperation rates, as high as 10% to 15% at 5 years and up to 30% at 15 years [2, 5, 15–17]. There is only one previously reported series of minimally invasive MV repairs for AL and BL prolapse [12]. Lapenna and colleagues [12] reported outstanding results without any in-hospital deaths or major complications, including no reoperations at a 22 months follow-up. This same group reported similar outstanding results for their sternotomy group with no difference in outcomes between AL or PL prolapse [18]. These results have not been matched by anyone in the field.
Our rate of reoperation was 9% with a mean follow-up of approximately 2 years. This rate of reoperation may appear high; however, these groups of patients (AL and BL prolapse) have been reported to have a higher incidence of reoperation [2, 5, 15–17]. In addition, the median and mean reported times for reoperation in other series were 2.4 years [19] and 15.6 ± 2.5 months [20]. Therefore, it appears that failures appear to occur early after repair and then plateau with sporadic and rare late failures [19, 20]. We expect that we have seen the peak rate for reoperation for this series.
Of the 6 patients who required a reoperation, 3 of them were reoperated on at our institution and 3 at other centers. Three of the early failures underwent reoperations at 15, 16, and 58 days. These cases represented early technical failures. One case was secondary to SAM. This patient had mild to moderate SAM on intraoperative TEE, which improved after volume loading and discontinuation of inotropic agents. Despite these measures, the patient returned with persistent SAM requiring reoperation. In retrospect, we should have reduced the PL a little bit more potentially by performing a posterior sliding plasty or decided during the first operation to replace the valve. The other two technical failures were secondary to residual AL restriction and hemolysis after Gore-Tex neochords insertion for AL repair. None of these patients had demonstrated SAM or mitral regurgitation on immediate postoperative TEE. Others have described similar complications in the literature that required reoperation after initial MV repair [19–21].
Late failures could be caused by technical problems or progressive myxomatous disease. Reoperations were required on 3 patients at 305, 487, and 825 days for late failure. Findings at reoperation included partial dehiscence of the annuloplasty band in each case. The sutures or U-clips were attached to the bands, but the bands were partially detached from the annulus. It is difficult to pinpoint whether the cause was secondary to either technique or rather progressive disease with anterior annular dilatation increasing stress on the band and ultimately dehiscing the annuloplasty band. We are exploring the possibility of using complete annuloplasty rings to reduce late failures. Again, late dehiscence has not been uncommon during reoperation in other series [19–21]. Two of these patients had mild mitral regurgitation at the end of the procedure, and it is well recognized that this is a risk factor for later reoperation.
There is an obvious learning curve with robotic MV surgery. The lack of tactile feedback could potentially limit ones ability to assess depth of suturing as well as tension on the sutures, resulting in repair failure. However, visual feedback can accomplish the same goals by learning to observe tissue displacement and deformation. In fact, it is not clear to us that the lack of tactile feedback was responsible for any of the failures. On the other hand, robotic technology offers superior visualization and precise dexterity, which could theoretically improve the ability to perform complex MV repairs.
There are a few limitations with this study. Even though this is the largest series for these complex RMVP, it is still small compared with conventional MV repair reports. Another limitation of the study is that we do not have multiple echocardiographic evaluations before the study that determined severe mitral regurgitation and need for reoperation on the three late failures. Therefore, it is difficult to determine whether this is true disease progression or this was an unrecognized early failure likely to be technical in nature as are most cases of reoperations after MV repairs for degenerative disease [2, 19, 20]. We have implemented a thorough follow-up on all MV repairs at our institution with postdischarge echocardiograms at 3 months to 6 months, at 12 months, and yearly thereafter. In addition, these patients are evaluated in our valve clinic at those times. Finally, the follow-up is short because this technology has only been used for the last 6 years in the United States.
Robotic MV repairs for complex diseases are feasible and confer results similar to series using conventional techniques. Nonetheless, longer-term follow-up is needed to determine whether long-term results will be comparable to the 10- and 20-year series published by others. In the meantime, we believe that this technique is feasible, and as technology continues to improve, these procedures will become easier and more reproducible and better results will likely follow. Data on quality of life and return to work are necessary to assess one of the benefits that have been demonstrated for other robotic cardiac procedures [22].
| Discussion |
|---|
|
|
|---|
My major question to you is this. You have learned a lot. Every one of us causes a ruckus in the operating room. What are you going to do differently on the next 58 patients? Thank you for the chance to discuss this.
DR RODRIGUEZ: Doctor Kron, thank you very much for your kind comments, and you are exactly right. Looking at the mortality, we had 4 patients, 3 patients who were within 90 days and 1 patient who was at the time of reoperation done at an outside institution 400 days from the initial operation. And yes, the cross-clamp times are much longer than by conventional techniques, and out of those 3 early deaths, 1 patient died of stroke, the other 1 had multiple organ failure, and 1 patient developed postpump failure although at the end of the operation had normal ventricular function; however, 6 hours later developed biventricular failure. This patient had a prolonged cross-clamp time.
Things that we are looking into include addition of retrograde cardioplegia in addition to our standard use of antegrade cardioplegia. Perhaps this would result in better myocardial protection. We are working on some minimally invasive cannulas for retrograde perfusion.
As far as the failures, we have been using a combination of sutures and U-clips. We went from sutures to clips back to sutures, and we actually looked at all our failures to see whether it was a clip or suture-related problem. What we have been doing now is we have been trying to use more sutures, although when we looked at the whole 300-patient series, there were 12 failures, there was no way we could tell whether it was a clip or suture problem. However, we are moving to more sutures than clips. In addition, we have been reexploring whether we should use a complete ring again, and we are basically working with a minimally invasive complete ring, although we acknowledge that there are excellent series showing excellent results using a semi-complete ring. But those are the three things that we are looking at as potential changes for the future.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. G. Castillo, A. C. Anyanwu, A. El-Eshmawi, and D. H. Adams All anterior and bileaflet mitral valve prolapses are repairable in the modern era of reconstructive surgery Eur J Cardiothorac Surg, April 26, 2013; (2013) ezt196v1. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Ibrahim, C. Rao, and T. Athanasiou Artificial chordae for degenerative mitral valve disease: critical analysis of current techniques Interact CardioVasc Thorac Surg, December 1, 2012; 15(6): 1019 - 1032. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. Chan, M. Ruel, S. Chaudry, S. Lambert, and T. G. Mesana Clinical and echocardiographic outcomes after repair of mitral valve bileaflet prolapse due to myxomatous disease J. Thorac. Cardiovasc. Surg., April 1, 2012; 143(4_suppl): S8 - S11. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. M. Suri, H. M. Burkhart, R. C. Daly, J. A. Dearani, S. J. Park, T. M. Sundt III, Z. Li, M. Enriquez-Sarano, and H. V. Schaff Robotic mitral valve repair for all prolapse subsets using techniques identical to open valvuloplasty: Establishing the benchmark against which percutaneous interventions should be judged J. Thorac. Cardiovasc. Surg., November 1, 2011; 142(5): 970 - 979. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Speziale, G. Nasso, G. Esposito, M. Conte, E. Greco, K. Fattouch, F. Fiore, M. Del Giglio, R. Coppola, and L. Tavazzi Results of mitral valve repair for Barlow disease (bileaflet prolapse) via right minithoracotomy versus conventional median sternotomy: A randomized trial J. Thorac. Cardiovasc. Surg., July 1, 2011; 142(1): 77 - 83. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Modi, E. Rodriguez, and W. R. Chitwood Jr. Robot-assisted cardiac surgery Interact CardioVasc Thorac Surg, September 1, 2009; 9(3): 500 - 505. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Umakanthan, M. Leacche, M. R. Petracek, S. Kumar, N. V. Solenkova, C. A. Kaiser, J. P. Greelish, J. M. Balaguer, R. M. Ahmad, S. K. Ball, et al. Safety of Minimally Invasive Mitral Valve Surgery Without Aortic Cross-Clamp Ann. Thorac. Surg., May 1, 2008; 85(5): 1544 - 1550. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |