| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Michael E. DeBakey Department of Surgery, Division of Congenital Heart Surgery, Baylor College of Medicine, Houston, Texas
b Division of Congenital Heart Surgery, Texas Children's Hospital, Houston, Texas
Accepted for publication April 7, 2008.
* Address correspondence to Dr Morales, Division of Congenital Heart Surgery, Texas Children's Hospital, 6621 Fannin St, MC-WT 19345H, Houston, TX 77030 (Email: dlmorale{at}texaschildrenshospital.org).
Presented at the Forty-fourth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 28–30, 2008.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Methods: All RS between October 2002 and August 2006 were analyzed (602 RS in 558 patients). Median age was 3.6 years (range, 0.1 to 45.1); weight, 14.2 kg (2.0 to 112.2). Operations performed at RS were Glenn 22% (131), Fontan 21% (129), aortic valve repair/replacement 12% (72), right ventricle-pulmonary artery conduit 11% (67), Rastelli 7% (39), heart transplant 5% (31), and other 22% (133). Forty-seven percent of patients (280) had single-ventricle physiology. Incidence of second sternotomy was 67% (406), third 28% (166), fourth 4% (24), fifth 0.8% (5), and sixth 0.2% (1). A major injury upon RS was defined as one causing hemodynamic instability requiring vasopressor support or emergent transfusion; femoral cannulation or emergent cardiopulmonary bypass; and any morbidity. A minor injury is any other injury during RS.
Results: The incidence of a major injury was not different between RS (0.3%; 2 of 602) and first-time sternotomy (0%; 0 of 1,274; p > 0.1). Incidence of a minor injury was 0.66% (4 of 602). No injury resulted in hemodynamic instability, neurologic injury, or death. Two patients (0.3%) required a nonemergent blood transfusion secondary to injury. (Nonemergent was defined as adminstration rate of less than 0.2 cc/kg/min and less than 10 cc/kg in total.) Femoral cannulation was performed in 4 of 602 RS cases (< 0.6%). Sternal wound infection was 0.5% (3 of 602); reoperation for postoperative bleeding was 1% (8 of 602). Median intensive care unit stay was 3 days (1 to 174); median hospital stay was 7 days (1 to 202). Hospital survival was 98%.
Conclusions: Repeat sternotomy can represent a negligible risk of injury and of subsequent morbidity or mortality. Therefore, the choice of management strategies for patients should not be affected by the need for RS.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
|
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Operative Technique
Repeat sternotomy is approached in a stepwise fashion. No preoperative studies are performed specifically for RS. The xyphoid is split using electrocautery, and the lateral planes under the xyphoid are developed. Small Volkman rake retractors can now be placed under the xyphoid, and the posterior plate of the sternum is dissected off the underlying tissue. These dissection planes are developed laterally to allow the entire chest wall more mobility to be retracted upwards and, it is hoped, to enter the pleural space. The sternal wires are cut anteriorly but left in place. An oscillating saw is used to cut the anterior table and medulla of the sternum, using the wires as a guide posteriorly. Once completed, the wires are removed, and only a thin layer of the posterior table of the sternum should remain. Again the sternum is retracted upward and is widely dissected off the underlying tissue laterally, allowing for even more chest wall mobility. Therefore, the chest wall can be lifted quite far up, which allows for excellent visualization for meticulously dissecting structures (ie, aorta, right ventricle–pulmonary artery conduits) off the sternum. Once completed, the posterior table can simply be incised with the electrocautery or straight Mayo scissors. Entering the pleural space allows this process to proceed more quickly, as one can sweep from the open lateral pleural space medially to free the retrosternal space. Once the sternum is opened, there is still tissue attached to the peristernal chest wall, which is dissected off bilaterally and completely before placing the sternal retractor. The majority of RS operations were performed by the attending surgeon.
Definitions
A major injury upon RS was defined as one causing hemodynamic instability requiring inotropes, vasoconstrictors, or emergent blood transfusion; femoral cannulation or emergent cardiopulmonary bypass; or any morbidity. A minor injury is any other injury during RS, for example, an injury to the right atrium requiring only a simple suture repair and no other therapies. Hemodynamic instability was defined as a change in systolic or diastolic pressure by 10 mm Hg, change in heart rate by 10 beats per minute, or a decrease in the cerebral near-infrared spectroscopy by 5% as per the anesthesia record that records vital signs every 5 minutes. An emergent transfusion was defined as administration of blood or blood products at a rate of 0.2 cc · kg–1 · min–1 or more, or a transfusion of more than 10 cc/kg in total.
Statistical Analysis
Fisher's exact test and the 
2 tests were used to analyze binary variables. Means, medians, standard deviations, and ranges were used to describe continuous variables. Frequencies and percentages were used for categorical data. Determination of preoperative or operative risk factors for the primary endpoints (injury upon RS and subsequent morbidity and mortality) was not possible owing to the low incidence (< 1%) of the primary endpoints. Statistical significance was a p value of less than 0.05. All analyses were conducted with SPSS 15.0 (SPSS, Chicago, Illinois).
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
|
The incidence of a major injury in the RS group was 0.3% (2 of 602) as compared with 0% (0 of 1,274) in first-time sternotomies. This difference is not statistically significant (p > 0.1). One major injury was in a 5-year-old girl (16 kg) with a history of truncus arteriosus, dextrocardia, hypoplastic branch pulmonary arteries, and DiGeorge syndrome, undergoing her third sternotomy for conduit exchange. Her prior two surgeries were done at an outside hospital. The surgical plan was to resuscitate her pulmonary arteries, exchange her conduit, and close her atrial septal defect. A small laceration was created in the previously placed RV-PA conduit with the oscillating saw. Bleeding was controlled with direct pressure, and the patient remained hemodynamically stable without hypotension or need for vasoconstrictors or inotropes. The patient was placed on femoral bypass. The sternotomy was completed, the patient was cannulated centrally, and the rest of the surgery was uneventful. A blood transfusion had begun before the injury and was continued after the injury at the same rate of 0.13 cc · kg–1 · min–1 until cardiopulmonary bypass was instituted. Total transfusion was 8 cc/kg. Her immediate postoperative course was unremarkable; she was successfully extubated on postoperative day 1, transferred out of the intensive care unit on postoperative day 2 and discharged home on postoperative day 4 with no morbidity.
The next major injury occurred in a 16-year-old woman (50 kg) born with an atrioventricular canal defect who had undergone multiple operations at an outside institution. She subsequently developed end-stage congenital heart disease and underwent orthotopic cardiac transplantation at 8 years old. On this occasion, she presented with severe acute cellular rejection and required biventricular mechanical support. During this fifth median sternotomy, while dissecting free the retrosternal tissue, it was noted that the patient had severe biatrial enlargement secondarily to the severely depressed biventricular function. While dissecting the right atrium off the posterior table, a small laceration was made in the right atrium. There was no significant hemorrhage; however, it became apparent to the surgeon that he would not be able to separate the right atrium from the sternum safely. Therefore, he decided to place the patient on femoral bypass. The sternotomy was completed, and the small laceration was repaired without incident. The patient was not transfused nor did she suffer any hemodynamic instability from this injury. Once the sternum was opened, the patient was cannulated centrally. She required mediastinal exploration on postoperative day 2 secondary to bleeding and was extubated 10 days after this. She showed no improvement in ventricular function while mechanically assisted, and she was therefore relisted for transplantation. After 30 days, she underwent retransplantation and recovered well. She was extubated on postoperative day 1, discharged from the intensive care unit on postoperative day 10, and discharged from the hospital on postoperative day 28.
The incidence of minor injuries was 0.66% (4 of 602). Three of these were to the right atrium and 1 was to the innominate vein. All injuries occurred during dissection of the retrosternal space and were not caused by the oscillating saw. All of the minor injuries were secured by simple stitches, and none resulted in hemodynamic instability, need for femoral cannulation, emergent blood transfusion, or any identifiable morbidity.
Only 2 cases (0.3%) required a blood transfusion secondary to an injury (1 minor and 1 major injury). They both were nonemergent. Femoral cannulation was performed in 4 cases (0.6%); 2 were as a result of injury, and the other 2 were planned preoperatively. All 4 patients were eventually converted to mediastinal cannulation. There were no peripheral vascular or any extremity complications in these patients. Median time on cardiopulmonary bypass was 126 minutes (14 to 438), and median cross-clamp time was 43 minutes (2 to 286). Postoperative bleeding requiring mediastinal exploration occurred in 1.3% of RS patients (8 of 602). Sternal wound infection or dehiscence occurred in 0.5% of patients (3 of 602). The median intensive care unit stay and hospital length of stay postoperatively was 3 days (1 to 174) and 7 days (1 to 202), respectively.
Hospital survival was 98% (590 of 602). None of the 12 deaths was related to RS. No patient with a minor or major injury died.
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
In many congenital heart centers, RS is such a common occurrence that familiarity has lead to the presumption that RS, although perhaps challenging at times, does not really impact the outcome of a procedure. However, there are only a few manuscripts assessing resternotomy and even fewer assessing resternotomy in congenital heart surgery: three manuscripts over the past 2 decades [6–8]. These manuscripts report that the incidence of major injuries upon RS is 5% to 10%, but that these injuries did not increase mortality when compared with primary sternotomy. The current series establishes an even stronger claim that the risk of a major injury upon RS can be so low (0.3%) that it is not statistically different than for primary sternotomy (0%). Also, this series demonstrates that the morbidity and mortality from RS, even for patients who sustain an injury, is negligible. Even though 2 patients met this study's requirements for a major injury upon RS (both because they were cannulated peripherally), neither had hemodynamic instability or any morbidity as a result of the injury. Postoperative morbidities often associated with RS such as sternal wound infection and postoperative bleeding had a very low incidence in this series (0.5% and 1.3%, respectively) and would compare favorably to an incidence of these morbidities in any primary cardiac surgery series. Also, there were no blood transfusions given as a bolus or emergently. Two of the 6 patients injured (1 minor and 1 major) received a blood transfusion. One of these two blood transfusions was started even before the injury. Also, no patient with a RS injury died.
Even though no radiographic imaging is performed for a potentially hazardous resternotomy in our practice, if imaging is obtained for other reasons (namely, to define large psuedoaneurysms of an RV-PA conduit), the information (namely, the aneurysm eroding into the back of the sternum) is used in planning the operative approach. This explains how one of the femoral cannulations in the series was planned, as preoperative imaging demonstrated erosion of an aneurysm into the sternum. In general, however, we believe that our stepwise approach affords the surgeon a view of the vascular structures adhered to the sternum while freeing the retrosternal tissue and before entering the posterior table of the sternum. Therefore, imaging is not thought to be necessary. In 1 patient in this series, the surgeon observed during dissection of the retrosternal tissue that the RV-PA conduit had eroded into the sternum. The surgeon believed that he could not safely separate the conduit and the sternum and, therefore, electively went on femoral bypass before proceeding further.
Limitations
The current series is a retrospective study that does not have a comparison group. However, the main objective of the manuscript was to establish the risk of injury upon RS and the subsequent morbidity and mortality. The manuscript did establish this risk and compare it to patients undergoing primary sternotomy. Because there was no subsequent morbidity or mortality secondary to any injury, no comparison was made to primary sternotomy. The manuscript also does not compare the risk of major postoperative morbidities such as mediastinal bleeding, sternal wound infection, or need for blood transfusion after RS to primary sternotomy. The incidence of blood transfusion in a RS series is actually not necessarily representative of the patients' course during reoperation but rather their physiologic status. In this series, where almost half of all RS were in single-ventricle patients and the institution believes strongly in maximizing brain perfusion as measured by cerebral near-infrared spectroscopy, blood transfusions are sometimes given to maximize perfusion and not because of RS. Hemodynamic compromise from injury was measured through the anesthesia record that recorded vital signs every 5 minutes. The authors did not have access to continuous data of the vital signs retrospectively. Therefore, it is possible that a patient after an injury met the study criteria for hemodynamic compromise but recovered within a 5-minute timeframe and thus was not captured by those reviewing the anesthesia record. If that did occur in any of the 6 patients injured at RS, the study can at least state that the injury and subsequent recovery occurred in less than 5 minutes and did not require the institution of vasoconstrictors, inotropes, or an emergent blood transfusion.
In conclusion, the present series does not intend to claim that RS and primary sternotomy are the same. It is recognized that they represent different surgical tasks. A reoperation not only entails a RS but also the subsequent freeing of the heart and great vessels from adhesions, which can be onerous. However, the current series does demonstrate that if approached meticulously, RS can represent a negligible risk of injury and of subsequent morbidity or mortality. Therefore, the practice of not prioritizing the risk of RS when making clinical decisions is sound. The choice of management strategies for a patient should not be affected by the need for RS.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
DR MORALES: Yes. It is not our usual practice to use Gore-Tex or to close the pericardium. About a quarter of these patients were operated on at outside hospitals, and some of those patients did have their pericardium closed or had Gore-Tex membranes placed.
DR METRAS: But in your center, you don't close the pericardium?
DR MORALES: We do not.
DR JOSEPH J. AMATO (Chicago, IL): While I am not operating at the present time, I have extreme interest in resternotomies. I noticed that in your abstract you had mentioned CoSeal, and I thought you were bringing that up as a nonadhesive material There is also another product that I believe is called REPEL–CV which is also being touted as a biodegradable membrane that would "prevent injury." In 1996, I was part of a multi-institutional study and we published "Expanded PTFE membrane to prevent injury during resternotomy for congenital heart disease" (Ann Thorac Surg 1996;2:1778–82). We conducted a study of 1,085 patients and reported only 1 injury and no deaths. Both CoSeal and the REPEL-CV report an 88% effectiveness of preventing sternal adhesions and injury to the heart when opening the sternum. But I feel very strongly that the polytetrafluoroethylene (PTFE) membrane still represents the key to very little injury (1%) and no mortality if you use it properly underneath the sternum. My recommendation has been to both of these companies to place a thin strip of PTFE directly beneath the sternum in combination with a larger sheet of the bioabsorbable membrane combination around the rest of the heart.
DR MORALES: In about 10 of these patients, CoSeal was probably used. We are doing a randomized blinded study right now to assess if CoSeal decreases adhesions at reoperation.
DR SAM WEINSTEIN (New York, NY): David, excellent presentation. Two brief questions. One, can you just clarify, are you talking about the dissection from the sternotomy through cannulation as the time frame for recording the injuries? And the second question is, what data, if any, would you find preoperatively that might change your reopererative technique, such as perihperal cannulation before the sternotomy?
DR MORALES: We do not specifically get any radiographic imaging because of a potentially hazardous resternotomy. However, I think, as you saw, two of the femoral cannulations were planned. In these cases, imaging was obtained for other reasons (ie, define large psuedoaneurysms of an RV-PA conduit), but the information (ie, the aneurysm eroding into the back of the sternum) was used in planning the operative approach. And I'm sorry, what was your first question?
DR WEINSTEIN: I think you said it, and I may have missed it. The time frame for the injury, is it all the way through to cannulation, including the pericardial space dissection? Or are you counting just injuries performed at the sternotomy?
DR MORALES: Our intention was until cannulation. There were no major injuries during this time. However, all of the data were gathered retrospectively, so small injuries to right atriums and other structures that did not affect the flow of the procedure may not have been dictated.
DR CARL L. BACKER (Chicago, IL): The questions I want to ask focus on blood usage. The first question is, do you have information about how many blood products were used during the actual sternotomy reentry compared with patients who had, for instance, a primary sternotomy?
DR MORALES: There were two nonemergent blood transfusions secondary to an injury during resternotomy as seen on this slide. By reporting the rate and total volume transfused on this slide, I was trying to demonstrate that neither were rapid transfusions or boluses given for hemodynamic instability. One of these blood transfusions actually started even before the injury, but I included it since there was an injury and a blood transfusion at the same time. I don't have the data comparing it to first-time sternotomies. But unless you had an injury, getting a blood transfusion during resternotomy really has little to do with the resternotomy but rather the patient's physiology, who is getting the resternotomy. Therefore, we didn't collect these data.
DR BACKER: Nice timing on your discussion slide. The other part of the question relates to the use of aprotinin. Was aprotinin used in the whole series, or used selectively? The corollary to that is, have you noticed any difference in bleeding complications now that aprotinin is off the market?
DR MORALES: Aprotinin was used for most of the neonates and multiple reoperations who are included in the series. In respect to the second question, we have not noticed a difference since we have stopped using it, or at least I have not.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Giamberti, M. Chessa, R. Abella, G. Butera, C. Carlucci, H. Nuri, A. Frigiola, and M. Ranucci Morbidity and mortality risk factors in adults with congenital heart disease undergoing cardiac reoperations. Ann. Thorac. Surg., October 1, 2009; 88(4): 1284 - 1289. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Ferrari, C. Sulzer, E. Rizzo, and L. K. von Segesser Transcatheter stent-valve implantation in a stenotic pulmonary conduit via a sub-xyphoidian access Eur. J. Cardiothorac. Surg., September 1, 2009; 36(3): 595 - 597. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. M. Kirshbom, R. J. Myung, J. M. Simsic, Z. B. Kramer, T. Leong, B. E. Kogon, and K. R. Kanter One thousand repeat sternotomies for congenital cardiac surgery: risk factors for reentry injury. Ann. Thorac. Surg., July 1, 2009; 88(1): 158 - 161. [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 |
