- Email: [email protected]
To Nuss or Not to Nuss? Two Opposing Views
To Nuss or Not to Nuss? Two Opposing Views Francis Robicsek, MD, PhD,* and André Hebra, MD† Although the issue of the appropriate approach for the repair of pectus excavatum remained unsettled for decades, just when we thought that the consensus was clear, an entirely new method was introduced: the Nuss operation. This technique now challenges not only the previously established standards, but also the basic conceptual views of pectus surgery. In the following text, 2 opposing views on the subject are presented: the angle from which Francis Robicsek, a pioneer in conventional pectus excavatum surgery, views the issue, and the opinion of Andre Hebra, who has extensive experience with the Nuss operation. Semin Thorac Cardiovasc Surg 21:85-88 © 2009 Elsevier Inc. All rights reserved. KEYWORDS open repair, Nuss procedure
Francis Robicsek, MD, PhD Since Donald Nuss introduced first his operation entitled, “Minimally invasive technique for the correction of pectus excavatum” in 1997, at the 29th meeting of the American Pediatric Surgical Association, there has been an abundance of publications describing the technical details as well as the short- and midterm results of this procedure. Most, if not all, of these studies were neither prospective nor randomized and compared the Nuss operation to that described by Ravitch in 1949. Although the Ravitch operation indeed paved the way for modern pectus surgery, because of the need for extensive exposure and the less than optimal late results achieved, it has been largely abandoned. Most surgeons experienced in the repair of pectus deformities now use modifications that include not only bilateral resection of the costal cartilages, but also sternal remodeling and permanent sternal support. During the last decade, complying with the trends of modern surgery, these methods have been further altered such that they are performed through a small (8-10 cm) submammary skin incision. Although the jury is still out as far as late results are concerned, there are no data that even remotely suggest that the outcomes of the Nuss operation exceed those of “open” procedures performed by experienced hands. Also, the operating time of the Nuss operation, especially if endoscopically *Department of Cardiovascular and Thoracic Surgery, Carolinas Medical Center, Charlotte, North Carolina. †Division of Pediatric Surgery, Medical University of South Carolina, Charleston, South Carolina. Address reprint requests to Francis Robicsek, MD, PhD, 1001 Blythe Blvd., Ste. 300, Charlotte NC 28203. E-mail: FRobicsek@carolinashealthcare. org
1043-0679/09/$-see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1053/j.semtcvs.2009.03.007
assisted, is longer and costs more to perform. Double, if one considers the need for frequent follow-ups and the necessity of reoperation to remove the bar. Complications are more frequent and, if they occur, because of the sustained presence of the metal bar(s), they might be far more serious than those encountered after “open” repair. One may add to the above that, although the skin incisions for the Nuss operation may be considered “minimally invasive” compared to the extensive exposure of the classic Ravitch procedure, the two 3- to 4-cm incisions (plus the hole for the videoscope) needed to perform the Nuss procedure certainly “add up” to the single 7- to 8-cm submammary incision needed to perform an up-to-date limited-exposure “open” pectus repair. Also, how can anybody call an operation “minimally invasive” in which 2 ⫻ 12-cm metal rods are driven through both pleural cavities, passed by the width of a hair between the heart and the sternum and left there for extended periods, and then 2-4 years later the same procedure is performed in “reverse”? This particular aspect of the “left-in” precordial metal bars which, like the sword of Damocles threatens health and life, outweighs all the perceived advantages of the Nuss operation. That this is indeed the case is proven by the virtually unprecedented number of case reports describing various and often serious complications heretofore unknown: fracture, rotation, displacement of the bar, metal allergy, infection, and hemorrhage occurring months, even years, after surgery, life-threatening, even deadly, injuries to adjacent organs (primarily the heart), either intraoperatively or late after surgery, stubborn pericardial and pleural effusions, obstruction of the thoracic inlet and/or the caval veins, erosion of the sternum, mammary artery pseudoaneurysm—and so 85
86
Figure 1 The average skin incision used in recent pectus excavatum and carinatum repairs.
on. Unfortunately, most of these events occur not only as part of the “learning curve” but also in experienced hands! Considering that the need for correction of pectus excavatum, even half a century after its introduction, is still unsettled and in the lion’s share of cases indication is cosmetic rather than
F. Robicsek and A. Hebra physiological need, it is rather difficult to reconcile it with the words of Hippocrates: “Thou should do no harm.” Recent, large series report complications of 20%. This is unacceptably high. The argument that open repair of pectus excavatum may induce development of acquired thoracic restrictive dystrophy does not hold water. We strongly believe that restrictive thoracic dystrophy should never occur following a properly performed open pectus excavatum repair. Furthermore, the repair may be performed even at a very early age, but it has to be performed correctly. By contrast, it is a specific concern as how the Nuss bars may affect the costal cartilages of the growing child? These metal girdles (sometimes as many as 3) not only splint but may also restrict the growth of the anterior chest wall. An issue that is clearly related to the changes observed (fractures, angulation) in the elongated and unresected rib cartilages that are forcefully reshaped by the insertion of a bar is the prolonged postoperative pain after the Nuss procedure, present especially in older patients. Another important matter that should be pointed out is the length of postoperative evaluation of the outcome of surgical correction. The term “postoperative” after the Nuss procedure should be used only after the metal bars are removed. This could take up to 4 years! After all the above is said, one may wonder how the Nuss operation became so popular and is now considered by many the “gold standard” of pectus excavatum repair? The answer is complex and controversial. First, the proponents of the procedure successfully injected into the debate the magic words of contemporary medicine: “nonsurgical” and “minimally invasive”—nonetheless that the Nuss procedure requires 2 operations and 4-5 incisions and that it is more invasive than any other pectus operation save Juro Wada’s sternal turnover. The second reason for the popularity of the
Figure 2 Metal introducers used in the course of a “minimally invasive” Nuss operation. (Photo courtesy of Dr. Hans Pilegaard.)
Open vs. Nuss repair Nuss procedure may be the lack of necessity of partial chest wall resection. This allows specialists who are not formally trained in thoracic surgery to enter a new area uncharted by all but thoracic surgeons. As usually happens in the effort of turf protection, many thoracic surgeons readily followed, especially because the news that “pectus excavatum now may be corrected without surgery” induced patients and pediatricians alike who otherwise may have been reluctant to proceed with correction of the anomaly (Figs. 1 and 2).
There Are Some Overdue Tasks Regarding the Nuss Procedure We need well controlled, evidence-based trials that compare the major clinical and economical aspects of this innovative procedure, not the long-outmoded Ravitch operation, but up-to-date, much more effective and less invasive open operations, such as the 1-stage, mesh-supported sternoplasty. In the meantime, it is our obligation that relevant information already available on the virtues and shortcomings of the Nuss operation be revealed to the parties involved, in a way to allow them to render a truly informed consent of choice.
Andre Hebra, MD The history of medicine has taught us that surgery is an always evolving art and science. Many innovative and sometimes revolutionary ideas introduced by surgeons were met with great resistance before becoming acceptable surgical techniques, frequently replacing more traditional methods of treatment. Notable examples include the management of chronic cholecystitis with open cholecystectomy and the treatment of peptic ulcer disease with radical gastric and duodenal resections. One must be careful when evaluating new methods of treatment for any type of disease or condition. However, one must be mindful that change in surgical techniques usually brings benefits to patients and opens the door for extended applications in the practice of surgery. There is no doubt that the term “minimally invasive surgery” does not mean that the surgeon will not invade body cavities or that the procedure will be pain-free. For pectus excavatum, minimally invasive repair (MIRPE) implies that the correction of the deformity can be achieved without the creation of large incisions and without the removal of bone or cartilage. Because one of the major concerns of patients with pectus excavatum is related to the appearance and perception of their physical attributes, the creation of small surgical scars that are located in an area that makes them almost invisible to others is an important component of the treatment outcome. Surgeons have never proclaimed that the procedure is painfree. As a matter of fact, since the inception of MIRPE, great effort has been placed on the management of postoperative pain. Protocols have been developed that allow for the optimal use of modern pain management techniques that minimize discomfort and facilitate postoperative recovery. The acceptance and popularity of MIRPE developed quickly since its introduction in 1997. Proponents of this new approach have argued that the principal advantages of
87 the MIRPE is based on the fact that an anterior chest wall incision with creation of pectoralis muscle flaps, resection of several ribs and cartilages, and performing a sternal osteotomy are not needed. This leads to a much shorter operating time, minimal blood loss, and minimal anterior chest wall scar. Moreover, the stability and strength of the chest wall is not compromised as it is with the open repair. Part of the significant increase in the number of procedures being performed using the minimally invasive technique was patient driven. In a very short period since inception of this new technique, several series were reported in which MIRPE was the primary mode of operative technique. This was followed by reports of various complications, with an overall complication rate of almost 20% in some of the early but large series of patients. Analysis of the reported data clearly demonstrated that many surgeons were reporting outcomes during their learning curve. Additionally, the techniques for bar stabilization had not yet been refined as the popularity of MIRPE continued to increase. Today, bar displacement is considered a rare occurrence. Fortunately, most factors that may lead to complications and poor results were related to early inexperience; these factors have been corrected. Moreover, the introduction of thoracoscopy when performing MIRPE has significantly enhanced the surgeon’s ability to pass the bar precisely behind the sternum, avoiding the risk of cardiac or vascular injury. Reassuringly, only one reported case of cardiac perforation occurred before the routine use of thoracoscopy. It is true that the length of the surgical scarring that results from the open repair of pectus excavatum has decreased over time, but it still leaves behind a scar of at least 8-12 cm in the anterior chest, in a quite visible location. Such scar is known to increase in size as the child grows. This is difficult to compare to the scars caused by the MIRPE, because their location and size is so different. Most patients that undergo MIRPE will have 2 small (less than 2 cm scars) in the lateral chest at the level of the midaxillary line, typically hidden behind the arms. This has been one of the main incentives reported by patients and parents as to the preference of MIRPE over the open procedure. To date, limited comparative prospective studies between the 2 techniques are available. The few comparative reports in the literature have demonstrated that the MIRPE and open procedures, in experienced hands, have similar results and outcomes. At the present time, there is 1 on-going prospective multi-institution study underway that is analyzing comparative data (including open and MIRPE techniques) in more than 500 pectus patients. Unfortunately, despite the excellent study design that would allow a fair comparison of outcomes, patient recruitment in the open surgery arm of the study has been very limited. This is due primarily to the fact that many patients and families do not wish to be submitted to the open repair, despite objective informed consent. Preliminary reported results have indicated that patients and their parents reported significant positive postoperative changes. Improvements occurred in both physical and psychosocial functioning, including less social self-consciousness and a more favorable body image. Ninety-seven percent
F. Robicsek and A. Hebra
88 of patients thought that surgery significantly improved their body image. Among the 326 study subjects, only 15% underwent the open operation. Although the open operation has been performed successfully since 1949, very few reports are available that have analyzed long-term outcomes, particularly related to pectus recurrence. It is recognized today that numerous patients that underwent open corrective surgery before 10 years of age have experienced recurrent deformity. A few patients in this category (including young adults) have undergone corrective surgery using the minimally invasive technique. Dr. Nuss has reported successful repair of failed Ravitch procedures in 39 patients with a median age of 16 years. The lack of long-term outcomes studies of the Ravitch procedure or any of its modifications is surprising, considering that it has been performed extensively since the early 1950s. The few reports available are by very few surgeons that have been able to accumulate a large series of patients over the years. Even these surgeons have reported significant modifications to their open techniques over the years that have allowed for better outcomes and less complications. The Nuss technique has been incorporated into modern surgical practice in the last 12 years. Several modifications to the original operation have been adopted and improved outcomes continue to be reported. Even the patient and family awareness has been significantly increased since the introduction of the MIRPE. Another significant advantage of MIRPE over the open surgical procedure is that the dreaded complication of “thoracic
constriction” (Jeune syndrome) does not seem to occur with this new technique. Chest wall constriction has been described in a few patients following extensive open pectus excavatum operations. Apparently, the bone growth center can be affected, which results in restriction of chest wall growth with marked limitation of ventilatory function. Such patients are very symptomatic and are unable to compete in any running games. The forced vital capacity and forced expiratory volume at 1 second is typically decreased by more than 50% of predicted reference range levels. With the MIRPE, because no resection or incision is made on ribs or cartilages, such unusual complication does not appear to be a concern. Once the cartilage and bony structures are remodeled, normal or improved pulmonary function is established and the flexibility and malleability of the chest remains unaffected, particularly once the metal strut is removed. Similarly, another dreaded complication of open surgery is the lack of regrowth of bone and cartilage after resection, which can lead to a partial flail chest. This condition appears to be related to injury to the perichondrium and periosteum during resection. Providing reconstruction and protection to mediastinal structures in this setting is a complex problem.
Acknowledgment Portions of this article have been adapted from Robicsek F: When four plus four is less than eight: the Nuss operation. Eur J Cardiothorac Surg 35:559-560, 2009.
Francis Robicsek, MD, PhD Since Donald Nuss introduced first his operation entitled, “Minimally invasive technique for the correction of pectus excavatum” in 1997, at the 29th meeting of the American Pediatric Surgical Association, there has been an abundance of publications describing the technical details as well as the short- and midterm results of this procedure. Most, if not all, of these studies were neither prospective nor randomized and compared the Nuss operation to that described by Ravitch in 1949. Although the Ravitch operation indeed paved the way for modern pectus surgery, because of the need for extensive exposure and the less than optimal late results achieved, it has been largely abandoned. Most surgeons experienced in the repair of pectus deformities now use modifications that include not only bilateral resection of the costal cartilages, but also sternal remodeling and permanent sternal support. During the last decade, complying with the trends of modern surgery, these methods have been further altered such that they are performed through a small (8-10 cm) submammary skin incision. Although the jury is still out as far as late results are concerned, there are no data that even remotely suggest that the outcomes of the Nuss operation exceed those of “open” procedures performed by experienced hands. Also, the operating time of the Nuss operation, especially if endoscopically *Department of Cardiovascular and Thoracic Surgery, Carolinas Medical Center, Charlotte, North Carolina. †Division of Pediatric Surgery, Medical University of South Carolina, Charleston, South Carolina. Address reprint requests to Francis Robicsek, MD, PhD, 1001 Blythe Blvd., Ste. 300, Charlotte NC 28203. E-mail: FRobicsek@carolinashealthcare. org
1043-0679/09/$-see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1053/j.semtcvs.2009.03.007
assisted, is longer and costs more to perform. Double, if one considers the need for frequent follow-ups and the necessity of reoperation to remove the bar. Complications are more frequent and, if they occur, because of the sustained presence of the metal bar(s), they might be far more serious than those encountered after “open” repair. One may add to the above that, although the skin incisions for the Nuss operation may be considered “minimally invasive” compared to the extensive exposure of the classic Ravitch procedure, the two 3- to 4-cm incisions (plus the hole for the videoscope) needed to perform the Nuss procedure certainly “add up” to the single 7- to 8-cm submammary incision needed to perform an up-to-date limited-exposure “open” pectus repair. Also, how can anybody call an operation “minimally invasive” in which 2 ⫻ 12-cm metal rods are driven through both pleural cavities, passed by the width of a hair between the heart and the sternum and left there for extended periods, and then 2-4 years later the same procedure is performed in “reverse”? This particular aspect of the “left-in” precordial metal bars which, like the sword of Damocles threatens health and life, outweighs all the perceived advantages of the Nuss operation. That this is indeed the case is proven by the virtually unprecedented number of case reports describing various and often serious complications heretofore unknown: fracture, rotation, displacement of the bar, metal allergy, infection, and hemorrhage occurring months, even years, after surgery, life-threatening, even deadly, injuries to adjacent organs (primarily the heart), either intraoperatively or late after surgery, stubborn pericardial and pleural effusions, obstruction of the thoracic inlet and/or the caval veins, erosion of the sternum, mammary artery pseudoaneurysm—and so 85
86
Figure 1 The average skin incision used in recent pectus excavatum and carinatum repairs.
on. Unfortunately, most of these events occur not only as part of the “learning curve” but also in experienced hands! Considering that the need for correction of pectus excavatum, even half a century after its introduction, is still unsettled and in the lion’s share of cases indication is cosmetic rather than
F. Robicsek and A. Hebra physiological need, it is rather difficult to reconcile it with the words of Hippocrates: “Thou should do no harm.” Recent, large series report complications of 20%. This is unacceptably high. The argument that open repair of pectus excavatum may induce development of acquired thoracic restrictive dystrophy does not hold water. We strongly believe that restrictive thoracic dystrophy should never occur following a properly performed open pectus excavatum repair. Furthermore, the repair may be performed even at a very early age, but it has to be performed correctly. By contrast, it is a specific concern as how the Nuss bars may affect the costal cartilages of the growing child? These metal girdles (sometimes as many as 3) not only splint but may also restrict the growth of the anterior chest wall. An issue that is clearly related to the changes observed (fractures, angulation) in the elongated and unresected rib cartilages that are forcefully reshaped by the insertion of a bar is the prolonged postoperative pain after the Nuss procedure, present especially in older patients. Another important matter that should be pointed out is the length of postoperative evaluation of the outcome of surgical correction. The term “postoperative” after the Nuss procedure should be used only after the metal bars are removed. This could take up to 4 years! After all the above is said, one may wonder how the Nuss operation became so popular and is now considered by many the “gold standard” of pectus excavatum repair? The answer is complex and controversial. First, the proponents of the procedure successfully injected into the debate the magic words of contemporary medicine: “nonsurgical” and “minimally invasive”—nonetheless that the Nuss procedure requires 2 operations and 4-5 incisions and that it is more invasive than any other pectus operation save Juro Wada’s sternal turnover. The second reason for the popularity of the
Figure 2 Metal introducers used in the course of a “minimally invasive” Nuss operation. (Photo courtesy of Dr. Hans Pilegaard.)
Open vs. Nuss repair Nuss procedure may be the lack of necessity of partial chest wall resection. This allows specialists who are not formally trained in thoracic surgery to enter a new area uncharted by all but thoracic surgeons. As usually happens in the effort of turf protection, many thoracic surgeons readily followed, especially because the news that “pectus excavatum now may be corrected without surgery” induced patients and pediatricians alike who otherwise may have been reluctant to proceed with correction of the anomaly (Figs. 1 and 2).
There Are Some Overdue Tasks Regarding the Nuss Procedure We need well controlled, evidence-based trials that compare the major clinical and economical aspects of this innovative procedure, not the long-outmoded Ravitch operation, but up-to-date, much more effective and less invasive open operations, such as the 1-stage, mesh-supported sternoplasty. In the meantime, it is our obligation that relevant information already available on the virtues and shortcomings of the Nuss operation be revealed to the parties involved, in a way to allow them to render a truly informed consent of choice.
Andre Hebra, MD The history of medicine has taught us that surgery is an always evolving art and science. Many innovative and sometimes revolutionary ideas introduced by surgeons were met with great resistance before becoming acceptable surgical techniques, frequently replacing more traditional methods of treatment. Notable examples include the management of chronic cholecystitis with open cholecystectomy and the treatment of peptic ulcer disease with radical gastric and duodenal resections. One must be careful when evaluating new methods of treatment for any type of disease or condition. However, one must be mindful that change in surgical techniques usually brings benefits to patients and opens the door for extended applications in the practice of surgery. There is no doubt that the term “minimally invasive surgery” does not mean that the surgeon will not invade body cavities or that the procedure will be pain-free. For pectus excavatum, minimally invasive repair (MIRPE) implies that the correction of the deformity can be achieved without the creation of large incisions and without the removal of bone or cartilage. Because one of the major concerns of patients with pectus excavatum is related to the appearance and perception of their physical attributes, the creation of small surgical scars that are located in an area that makes them almost invisible to others is an important component of the treatment outcome. Surgeons have never proclaimed that the procedure is painfree. As a matter of fact, since the inception of MIRPE, great effort has been placed on the management of postoperative pain. Protocols have been developed that allow for the optimal use of modern pain management techniques that minimize discomfort and facilitate postoperative recovery. The acceptance and popularity of MIRPE developed quickly since its introduction in 1997. Proponents of this new approach have argued that the principal advantages of
87 the MIRPE is based on the fact that an anterior chest wall incision with creation of pectoralis muscle flaps, resection of several ribs and cartilages, and performing a sternal osteotomy are not needed. This leads to a much shorter operating time, minimal blood loss, and minimal anterior chest wall scar. Moreover, the stability and strength of the chest wall is not compromised as it is with the open repair. Part of the significant increase in the number of procedures being performed using the minimally invasive technique was patient driven. In a very short period since inception of this new technique, several series were reported in which MIRPE was the primary mode of operative technique. This was followed by reports of various complications, with an overall complication rate of almost 20% in some of the early but large series of patients. Analysis of the reported data clearly demonstrated that many surgeons were reporting outcomes during their learning curve. Additionally, the techniques for bar stabilization had not yet been refined as the popularity of MIRPE continued to increase. Today, bar displacement is considered a rare occurrence. Fortunately, most factors that may lead to complications and poor results were related to early inexperience; these factors have been corrected. Moreover, the introduction of thoracoscopy when performing MIRPE has significantly enhanced the surgeon’s ability to pass the bar precisely behind the sternum, avoiding the risk of cardiac or vascular injury. Reassuringly, only one reported case of cardiac perforation occurred before the routine use of thoracoscopy. It is true that the length of the surgical scarring that results from the open repair of pectus excavatum has decreased over time, but it still leaves behind a scar of at least 8-12 cm in the anterior chest, in a quite visible location. Such scar is known to increase in size as the child grows. This is difficult to compare to the scars caused by the MIRPE, because their location and size is so different. Most patients that undergo MIRPE will have 2 small (less than 2 cm scars) in the lateral chest at the level of the midaxillary line, typically hidden behind the arms. This has been one of the main incentives reported by patients and parents as to the preference of MIRPE over the open procedure. To date, limited comparative prospective studies between the 2 techniques are available. The few comparative reports in the literature have demonstrated that the MIRPE and open procedures, in experienced hands, have similar results and outcomes. At the present time, there is 1 on-going prospective multi-institution study underway that is analyzing comparative data (including open and MIRPE techniques) in more than 500 pectus patients. Unfortunately, despite the excellent study design that would allow a fair comparison of outcomes, patient recruitment in the open surgery arm of the study has been very limited. This is due primarily to the fact that many patients and families do not wish to be submitted to the open repair, despite objective informed consent. Preliminary reported results have indicated that patients and their parents reported significant positive postoperative changes. Improvements occurred in both physical and psychosocial functioning, including less social self-consciousness and a more favorable body image. Ninety-seven percent
F. Robicsek and A. Hebra
88 of patients thought that surgery significantly improved their body image. Among the 326 study subjects, only 15% underwent the open operation. Although the open operation has been performed successfully since 1949, very few reports are available that have analyzed long-term outcomes, particularly related to pectus recurrence. It is recognized today that numerous patients that underwent open corrective surgery before 10 years of age have experienced recurrent deformity. A few patients in this category (including young adults) have undergone corrective surgery using the minimally invasive technique. Dr. Nuss has reported successful repair of failed Ravitch procedures in 39 patients with a median age of 16 years. The lack of long-term outcomes studies of the Ravitch procedure or any of its modifications is surprising, considering that it has been performed extensively since the early 1950s. The few reports available are by very few surgeons that have been able to accumulate a large series of patients over the years. Even these surgeons have reported significant modifications to their open techniques over the years that have allowed for better outcomes and less complications. The Nuss technique has been incorporated into modern surgical practice in the last 12 years. Several modifications to the original operation have been adopted and improved outcomes continue to be reported. Even the patient and family awareness has been significantly increased since the introduction of the MIRPE. Another significant advantage of MIRPE over the open surgical procedure is that the dreaded complication of “thoracic
constriction” (Jeune syndrome) does not seem to occur with this new technique. Chest wall constriction has been described in a few patients following extensive open pectus excavatum operations. Apparently, the bone growth center can be affected, which results in restriction of chest wall growth with marked limitation of ventilatory function. Such patients are very symptomatic and are unable to compete in any running games. The forced vital capacity and forced expiratory volume at 1 second is typically decreased by more than 50% of predicted reference range levels. With the MIRPE, because no resection or incision is made on ribs or cartilages, such unusual complication does not appear to be a concern. Once the cartilage and bony structures are remodeled, normal or improved pulmonary function is established and the flexibility and malleability of the chest remains unaffected, particularly once the metal strut is removed. Similarly, another dreaded complication of open surgery is the lack of regrowth of bone and cartilage after resection, which can lead to a partial flail chest. This condition appears to be related to injury to the perichondrium and periosteum during resection. Providing reconstruction and protection to mediastinal structures in this setting is a complex problem.
Acknowledgment Portions of this article have been adapted from Robicsek F: When four plus four is less than eight: the Nuss operation. Eur J Cardiothorac Surg 35:559-560, 2009.