Morbidity and Mortality Risk Factors in Adults With Congenital Heart Disease Undergoing Cardiac Reoperations

Transcription

1 ORIGINAL ARTICLES: SURGERY: The Annals of Thoracic Surgery CME Program is located online at To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal. Morbidity and Mortality Risk Factors in Adults With Congenital Heart Disease Undergoing Cardiac Reoperations Alessandro Giamberti, MD, Massimo Chessa, MD, PhD, Raul Abella, MD, Gianfranco Butera, MD, Concetta Carlucci, MD, Halkawt Nuri, MD, Alessandro Frigiola, MD, and Marco Ranucci, MD Department of Cardiac Surgery and Grown Up Congenital Heart Unit and Department of Cardiothoracic-Vascular Anesthesia and Intensive Care Unit, Istituto di Ricovero e Cura a Carattere Scientifico, Policlinico San Donato, San Donato Milanese, Milan, Italy Background. Reoperations represent relatively frequent events in adults with congenital heart disease (ACHD). Cardiac operations in these patients present major difficulties in management and technique. Although reoperations in ACHD are becoming increasingly frequent, limited knowledge exists regarding perioperative risk factors. Methods. The study included 164 ACHD patients who underwent cardiac reoperations between January 2002 and December 2007 at our institution. Preoperative and intraoperative data were analyzed to identify morbidity and mortality risk factors. Results. Reoperations included pulmonary valve implantation or conduit replacement in 60, aortic valve/root procedures in 36, residual atrial or ventricular septal defect closure in 19, and Fontan operation/conversion in 19. Hospital mortality was 3.6%. The mean mechanical ventilation time was 26 hours. Mean intensive care unit stay was 3.1 days. Severe postoperative complications occurred in 24 (15.1%). Cardiopulmonary bypass time (p 0.001), Fontan operation/conversion (p 0.001), preoperative hematocrit (p 0.004), previous number of operations (p 0.001), and preoperative congestive heart failure (p 0.021) were associated with severe morbidity. No factor was associated with death. Conclusions. Reoperations in ACHD are mostly due to right ventricular outflow tract lesions and were associated with a low mortality rate if performed in a center with a considerable activity and a dedicated program. Severe morbidity is relatively frequent and is generally associated with the preoperative (high hematocrit due to cyanosis, congestive heart failure, and the number of previous operations) and operative (Fontan operation/ conversion and cardiopulmonary bypass duration) conditions of the patient. (Ann Thorac Surg 2009;88: ) 2009 by The Society of Thoracic Surgeons The number of adults with congenital heart disease (ACHD) will inevitably increase in the near future [1]. Recent data suggest that the number of ACHD, whether repaired or not, approaches the number of children with the disorder [2]. Cardiac operations in these patients present major difficulties in management and technique. This especially applies to cardiac reoperation, which is one of the most frequent interventions [3 7] and the most common cause of death in these patients [3]. Reoperations in ACHD may be divided according to their nature into a first corrective repair, reoperations Accepted for publication May 20, Address correspondence to Dr Giamberti, Department of Cardiac Surgery and Grown Up Congenital Heart Unit, IRCCS Policlinico San Donato, Via Morandi 30, San Donato Milanese, Milan, 20097, Italy; alegia@hotmail.com. after corrective repair, and further palliation after a first palliative operation. Many ACHD require reoperation after successful repair owing to residual defects or complications, conduit degeneration or outgrow, or the need for valve replacement or repair. Many others require reoperations for definitive repair after palliative operations, for the development of long-term sequelae, or for additional palliative procedures after the first palliative operation. Complicating surgical management and worsening the prognosis are acquired heart diseases, such as coronary artery disease, valvular disease, and multisystemic anomalies that may also occur with increasing frequency as the ACHD population ages. Only a few studies have been published and limited knowledge has been obtained in this field [3 7]. The present study reviews our experience with a large population of 164 consecutive ACHD patients who under by The Society of Thoracic Surgeons /09/$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg GIAMBERTI ET AL 2009;88: REOPERATIONS IN ADULT CHD PATIENTS 1285 went cardiac reoperations, with the aim of investigating the risk factors for postoperative morbidity and mortality. Material and Methods This study was approved by the Scientific Committee of our hospital. All patients gave informed consent to surgical procedures and to the anonymous treatment of their data for scientific purposes. Patients During the study period, 2086 patients with congenital heart defects were operated on in our institution. Of these, 480 (23%) were ACHD patients, and 164 (34% of the ACHD patients) had reoperations. ACHD was defined as patients with congenital heart defects aged older than 16 years. Operative Details The approach was through a repeat sternotomy in 151 patients (92%), and the remaining required a lateral thoracic approach. Femoral vessels were harvested in 25 patients (15%) and femoral cannulation was used in 11 (7%). Cardiopulmonary bypass (CPB) was used in 155 (95%), with 104 (63%) receiving aortic cross-clamping and 51 (31%) undergoing beating heart operations. Deep hypothermic circulatory arrest was not used. Aside from the main cardiac intervention, 88 patients (54%) received 115 associated cardiac surgical procedures during the reoperation, including interventions for arrhythmias, 31; right ventricular (RV) remodeling, 29; tricuspid valve repair, 16; pacemaker implantation, 15; residual VSD closure, 8; residual ASD closure, 8; and pericardial patch enlargement of pulmonary artery branches, 7. Table 1. Demographics, Preoperative Profile, and Operative Details of the Patient Population (N 164) Variables No. (%) or Mean SD Preoperative data Age, y Gender male 92 (56) Weight, kg Left ventricular ejection fraction Congestive heart failure 10 (6.1) Serum creatinine value, mg/dl Hematocrit, % Serum bilirubin value, mg/dl Antiplatelets drugs 10 (6.1) Warfarin 15 (9.1) Previous cardiac operations One 100 (61) Two 44 (27) Three 15 (9) Four or more 5 (3) Surgical data Pulmonary valve a 60 (36.6) Aortic valve b 36 (21.9) Fontan operation/conversion 19 (11.6) Mitral valve repair/replacement 8 (4.9) Residual ASD/VSD 19 (11.6) Others 22 (13.4) On-pump procedures 154 (94) CPB duration, min Aortic cross-clamping time, min a Implant/conduit replacement. b Replacement/repair/Bentall or Ross operation. ASD atrial septal defect; CPB cardiopulmonary bypass; VSD ventricular septal defect. Data Collection and Definitions All data were retrieved from our institutional database. Preoperative data included demographics, left ventricular ejection fraction, values for serum creatinine (mg/dl), hematocrit, and serum bilirubin (mg/dl); use of antiplatelet drugs or warfarin, the main diagnosis leading to the reoperation, and the number of previous cardiac operations. Operative data included CPB and aortic cross-clamping duration in minutes and type of operation performed. Operations were grouped into homogeneous categories and analyzed according to these groups. The following outcome data were collected: allogeneic blood products transfusion rate, mechanical ventilation time (hours), intensive care unit (ICU) stay (days), peak postoperative creatinine value (mg/dl), low cardiac output state (need for inotropic support for more than 48 hours), acute renal failure requiring renal replacement therapy, stroke, sternal wound infection/mediastinitis, surgical reexploration; arrhythmias (including supraventricular tachycardia, atrial fibrillation, ventricular tachycardia and complete atrioventricular block), and hospital (30-day) mortality. To avoid multiple comparisons, the outcome variables admitted to the analysis were restricted to allogeneic blood transfusions, length of ICU stay, low cardiac output syndrome (LCOS), severe morbidity (defined as need for reexploration, sternal wound infection/mediastinitis, permanent stroke, renal failure, or mechanical ventilation for more than 48), and hospital death. Statistics Data in the text and tables are expressed as the number and percentage or the mean and standard deviation. Univariate associations between risk factors and outcome variables were tested with linear or logistic regression analyses when appropriate. Factors significantly (p 0.05) associated with outcome variables were entered into a multivariable analysis (step-wise forward linear or logistic regression analysis) to determine factors independently associated with outcome variables. Multicollinearity of the models was considered and excluded by tolerance statistics. A tolerance value above 0.50 was considered acceptable. A value of p 0.05 was considered to be significant for all the statistical tests. All statistical tests were two-sided. Statistical calculations were performed using SPSS 13.0 software (SPSS Inc, Chicago, IL).

3 1286 GIAMBERTI ET AL Ann Thorac Surg REOPERATIONS IN ADULT CHD PATIENTS 2009;88: Table 2. Outcome Variables Outcome Variable Number (%) or Mean SD Table 4. Multivariable Analysis for Outcome Variables Linear or Logistic Multivariable Regression Outcome Variable Factors a SE p Value Blood products transfusions 89 (54.2) Packed red cells 76 (46.3) Fresh frozen plasma 57 (34.7) Platelets units 18 (11) Intensive care unit stay, d Mechanical ventilation time, h Low cardiac output state 18 (11) Arrhythmias 18 (11) Surgical reexploration 10 (6.1) Acute renal failure 8 (4.9) Stroke 0 (0) Mediastinitis 0 (0) Severe morbidity 24 (15.1) Hospital mortality 6 (3.6) Results The demographics, preoperative and operative variables of our patient population are summarized in Table 1. Pathologies of the RV outflow tract (RVOT) were the most frequent, including 37 pulmonary valve regurgitations (PVR), 17 pulmonary conduit obstructions, and 7 RVOT obstructions. Pathologies of the left ventricular outflow tract (LVOT) were present in 28 patients plus 7 patients with ascending aorta aneurysm. Less frequently, residual atrial septal defect (ASD) was present in 16 and failing Fontan in 13. In patients with RVOT disease, we performed 43 pulmonary valve implantations (PVI) and 17 conduit replacements. In patients with LVOT disease, we performed 15 aortic valve replacement/repairs, 12 Bentall operations, and 9 Ross operations. Other frequent reoperations were residual atrial septal defect (ASD) closure in 16, conversion to extracardiac Fontan in 12, and mitral valve repair/replacement in 8. According to the type of operation, 148 patients (90%) Table 3. Univariate Association Between Outcome Variables and Perioperative Factors a Factor Transfusions ICU Stay Low Cardiac Output Severe Morbidity Fontan operation/ conversion CPB time, min Age, mon Preoperative hematocrit Previous operations, No. Congestive heart failure a Analysis was by linear or logistic regression analysis. The p values for associations are reported, with significance set at p 0.5. CPB cardiopulmonary bypass; ICU intensive care unit. ICU stay, days Constant Fontan operation/ conversion CPB time, min Low cardiac output Constant CPB time, min Hematocrit, % Severe morbidity Constant Previous operations, No CPB time, min a Regression coefficient. CPB cardiopulmonary bypass; ICU intensive care unit; SE standard error. had a reoperation after a repair, 9 (6%) had a palliation after a palliation, and 7 (4%) had a repair after a palliation. A Fontan-type operation was considered a repair after a palliation because this procedure represents the definitive palliation for patients with univentricular heart. Conversion to an extracardiac Fontan operation was considered a reoperation after a repair. The operations were grouped as (1) pulmonary valve/ conduit operations, (2) aortic valve/root operation, (3) Fontan operation/conversion, (4) mitral valve repair/ replacement, (5) residual septal defects, and (6) others. The distribution of these groups is reported in Table 1. Outcome details are presented in Table 2. Six (3.6%) hospital deaths occurred. Two deaths occurred in patients who were very cyanotic. They were aged 52 and 26 years old and had univentricular hearts not suitable for Fontan operation because of high pulmonary artery pressure and compromised ventricular function. They died after a systemic-pulmonary artery shunts on postoperative days 1 and 4, respectively. Two patients with severe pulmonary regurgitation, aged 40 and 48 years, died after PVI, intervention for arrhythmias, and tricuspid valve repair: 1 during the operation of untreatable left lung bleeding and 1 on postoperative day 20 for severe RV dysfunction. This last patient had a preoperative severe PVR and a RV dilatation and dysfunction (RV/LV ratio of 2.1) with severe tricuspid valve regurgitation. One 40-year-old patient with obstructive atriopulmonary Fontan connection died of severe multiorgan failure on postoperative day 8 after conversion Fontan. The last patient, with subaortic obstruction 16 years after an extracardiac Fontan operation, underwent ventricular septal defect (VSD) enlargement. He had severe LCOS and ventricular arrhythmias in the postoperative period and died after 5 days of extracorporeal membrane oxygenation (ECMO) assistance. Only one intraoperative major injury (0.6%) occurred in the 151 repeat sternotomies. The patient was a 20-yearold man with severe aortic regurgitation after arterial switch operation. A laceration was created in the pulmo-

4 Ann Thorac Surg GIAMBERTI ET AL 2009;88: REOPERATIONS IN ADULT CHD PATIENTS 1287 Fig 1. Severe morbidity likelihood according to the cardiopulmonary bypass time and the number of previous cardiac operations. nary artery trunk by the oscillating saw. He was placed on femoral bypass, the sternotomy was completed, the pulmonary artery trunk laceration was sutured, and the rest of the planned Bentall operation was uneventful. During the immediate postoperative period, 10 patients (6%) underwent surgical reexploration: 8 for bleeding, 1 after a conversion Fontan to remove a thrombus on the right atrium, and 1 Fontan patient after VSD enlargement to start ECMO assistance for severe LCOS. A 56-year-old patient with tetralogy of Fallot, previously (43 years before) underwent surgical pulmonary valvotomy and a complete repair in association with RV remodelling, tricuspid valve repair, right-side Maze procedure, and left pulmonary artery path enlargement. Early severe biventricular dysfunction developed, and the patient left the operating room on ECMO support. The patient was weaned from ECMO 6 days later and discharged from the hospital on postoperative day 56. Upon univariate analysis (Table 3), need for allogeneic blood products, ICU stay, low cardiac output, and severe morbidity were outcome variables that demonstrated a significant association with preoperative and operative factors. The other outcome measurements, including hospital mortality, did not demonstrate any association with preoperative and intraoperative factors. The CPB time and a Fontan operation/conversion were significant risk factors for all of the outcome variables. A high preoperative hematocrit (cyanotic patients) was a risk factor for ICU stay and LCOS. The previous number of operations was a risk factor for LCOS and severe morbidity. Age at the time of the reoperation was associated with the need for transfusions, and finally, the presence of preoperative congestive heart failure was associated with severe morbidity. When tested using the multivariable analysis (Table 4), the CPB time remained an independent risk factor for increased ICU stay, LCOS, and severe morbidity. Fontan operation/conversion remained independently associated with an increased ICU stay, and the number of previous operations with severe morbidity. The preoperative hematocrit remained independently associated with LCOS only. Figure 1 shows the likelihood of severe morbidity based on CPB duration and number of previous cardiac operations according to the multivariable logistic regression analysis. Comment Although reoperations in ACHD have become very frequent today, our knowledge about this growing patient population is limited. This is a timely topic of increasing clinical interest, as shown by recent publications [6 9]. Some of these recent reports [8, 9] suggest that ACHD should be operated on by pediatric cardiac surgeons in children s hospitals. In our hospital, all the cardiac surgical procedures for ACHD have been performed by pediatric cardiac surgeons in association with adult cardiac surgeons when procedures such as coronary artery bypass grafting or aortic operations were needed. Independently of the hospital, we believe that complex operations for ACHD, including reoperations, should be performed in a few experienced centers where reoperations are frequently performed. As previously [3] and more recently [6, 7] suggested, it is important to concentrate resources, patients, funding, and professional experiences in specialized high surgical and cardiology care units. A repeat sternotomy was required in 93% of our patients. This approach represents a technical challenge for the surgeon and carries a significant risk of patient morbidity and death. The few articles that are available demonstrate a 5% to 10% risk of a significant injury upon repeat sternotomy [10, 11]. Our data and the most recent series [12] demonstrate that if approached meticulously

5 1288 GIAMBERTI ET AL Ann Thorac Surg REOPERATIONS IN ADULT CHD PATIENTS 2009;88: and in large centers, where reoperations are frequently performed, repeat sternotomy can represent a negligible risk of injury and of subsequent morbidity and death. Some comments must be offered comparing our data with the other series in term of incidence of reoperations and type of operations performed. Even if our data underline an incidence of reoperation comparable with the more recent articles [4 7] of 23% to 34%, this percentage is dramatically smaller than the 58% incidence of reoperation in the series published by Dore and coworkers [3] 10 years ago. This reduction can be explained by the improvements in cardiac operations resulting from the introduction of new surgical techniques in the last 20 years and by the increasing number of interventional procedures in the catheterization laboratory. More interesting is the analysis related to the type of operations performed. As pointed out by Srinathan and coworkers [5], ACHD are a changing patient population. The PVI for PVR, for example, is the reoperation that is most frequent today despite the recent introduction of the pulmonary valve transcatheter implantation. In our experience, this operation represents 25% of the activity, which is similar to the recent Srinathan experience [5]. Conversely, PVI represented 11% of the reoperations in the study of Berdat and coworkers [4] and only 3% in the oldest series presented by Dore and coworkers [3]. Moreover, we are now performing operations such as Fontan conversion (7% of our activity), which were not previously performed, and we are observing that operations such as recoarctation have practically disappeared, which is nowadays a prerogative of the interventional cardiologists. Few studies have been published on perioperative risk factors for these patients [3 7]. Two are descriptive of a series of ACHD who have undergone operations and reoperations [3, 5]. Two are multicenter studies presenting the results of operations for ACHD where the heterogeneous experience of several centers was assembled [6, 7]. One study [4] presents an analysis of early outcome of ACHD undergoing cardiac reoperation. Unfortunately, the number of patients in that study is relatively small and very heterogeneous. The study also covers a long period of time, during which treatment algorithms, surgical techniques, and perioperative management have considerably changed. In this context, our study presenting the data of 164 patients undergoing reoperations in a 5-year period in 1 center should be quite representative of the actual surgically related perioperative risk factors of these complex patients. Mortality was 3.6% in our series and compares favorably with the few data available from the literature [3 7]. No perioperative risk factors were found for death. All of our patients who died were very complex cases; however, it is likely that we could not find mortality risk factors due to the limited number of events (type II error). Two deaths occurred in patients who were 40 and 48 years old. They had severe pulmonary regurgitation and died after a complex operation of PVI, intervention for arrhythmias, and tricuspid valve repair. One died of untreatable left lung bleeding and 1 of severe RV dysfunction. Patients with chronic PVR and RV dysfunction might have benefited from earlier reoperations before irreversible myocardial deterioration was established, as proposed by other authors [11, 12]. Pathologies of the RVOT were the most frequent in our series, and the PVI was the most frequent operation. The question of when to perform a PVI, and in whom, is becoming increasingly pressing [13]. Clear guidelines to assist in this decision have proved difficult to identify. The presence of symptoms, as for other valvular lesions, is an indubitable indication for reoperation. At our institution, all the potential candidates for PVI undergo electrocardiography, echocardiography, magnetic resonance imaging (MRI) to assess ventricular function, ventricular volume and RVOT morphology, cardiopulmonary exercise testing, and electrophysiologic study. Indications for reoperation include severe PVR (regurgitation fraction 35% on MRI imaging); RV dilatation (RV end-diastolic volume/lv end-diastolic volume 1.4 with symptoms or 2.0 without symptoms); RV systolic pressure of 2/3 or more, systemic pressure with symptoms or 3/4 or more without symptoms, and impaired exercise capacity (peak oxygen consumption 65% of predicted). We considered very important in our decision making the progressive evolution of these data in the follow-up, the association with pulmonary stenosis/pvr, and the presence of peripheric pulmonary arteries stenosis, severe tricuspid regurgitation, and supraventricular/ventricular arrhythmias. The surgical PVI is generally performed in the patients without intracardiac communications on normothermic CPB without aortic cross-clamping. The percutaneous approach now offers a less invasive treatment that may potentially shift the indications toward earlier intervention. All the patients presented in our series were considered as not good candidates for percutaneous approach for RVOT morphology or associated cardiac anomalies, or both. Two other deaths were patients with failing Fontan: one died after subaortic obstruction relief and one of multiorgan failure after Fontan conversion. For these patients, the timing of reoperation is easier to identify. Arrhythmias such as atrial reentry tachycardia or atrial fibrillations or obstruction in the atriopulmonary Fontan connection are clear indications for extracardiac Fontan conversion plus arrhythmias interventions when present. These patients with failing Fontan, as well as the previous PVR group, may also benefit from earlier reoperation or from heart transplantation. Finally, 2 cyanotic patients with univentricular heart not suitable for Fontan operation died after a second palliation. Patients not suitable for corrective repair, frequently cyanotic patients, have a very high operative mortality in adulthood [3, 7]. The early mortality higher [3,7] is not only higher, but the length of hospital stay is also longer and the risk of postoperative complications is higher [3, 7]. Thanks to improvements in pediatric cardiology and cardiac surgery, the rate of these very complex adult patients presenting for reoperation can be expected to decrease as time goes by.

6 Ann Thorac Surg GIAMBERTI ET AL 2009;88: REOPERATIONS IN ADULT CHD PATIENTS 1289 A very high number of our patients (54%) received several (n 115) associated cardiac surgical procedures during the reoperation. The most frequent were performed for arrhythmias, RV remodelling, and tricuspid valve repair. RV dysfunction and dilatation, arrhythmias, and developing of secondary tricuspid valve regurgitation are complications related to long-standing/chronic PVR. Once again this suggests that we are probably operating on these patients too late [12 17]. Mechanical ventilation time, ICU stay, and postoperative hospital stay in our series compare favorably with the other published series [3 7]. Postoperative morbidity was relatively high, with severe complications occurring in 15.1% of the patients. LCOS, arrhythmias, and acute renal failure, were the most frequent complications. We could identify three major sources of morbidity: preoperative conditions, timing of reoperation, and CPB duration. High preoperative hematocrit (cyanotic patients), number of previous operations, and presence of preoperative congestive heart failure indicate the worst patients for reoperation. Repair after palliation (as it happens in Fontan operation/conversion) and age at reoperation suggest once again the not optimal timing to reoperate on these patients. Finally, the CPB time was the most important risk factor for all of the outcome variables that were analyzed. CPB duration is, of course, a general marker of the complexity of the procedure, and it is not surprising that it is an independent risk factor for severe morbidity. All these risk factors were considered in the past [3, 7] as independent predictors of early mortality for ACHD that underwent operation or reoperation. In conclusion, reoperations in ACHD are very frequent today and are mostly a result of RVOT lesions. Reoperations in these patients can be performed with low risk if approached meticulously and in larger centers where reoperations are frequently performed. Severe morbidity is relatively frequent and is generally associated with the patient s preoperative (high hematocrit due to cyanosis, congestive heart failure, and the number of previous operations) and operative (Fontan operation/conversion and cardiopulmonary bypass duration) conditions. References 1. Chessa M, Cullen S, Deanfield J, et al. The care of adult patient with congenital heart defects: a new challenge. Ital Heart J 2004;5: Webb GD. Care of adults with congenital heart disease. A challenge for the new millennium. J Thorac Cardiovasc Surg 2001;49: Dore A, Glancy DL, Stone S, Menashe VD, Somerville J. Cardiac surgery for grown-up congenital heart patients: survey of 307 consecutive operations from 1991 to Am J Cardiol 1997;80: Berdat PA, Immer F, Pfammatter JP, Carrel T. Reoperations in adults with congenital heart disease: analysis of early outcome. Int J Cardiol 2004;93: Srinathan SK, Bonser RS, Sethia B, Thorne SA, Brawn WJ, Barron DJ. Changing practice of cardiac surgery in adult patients with congenital heart disease. Heart 2004;91: Vida VL, Berggren H, Brawn WJ, et al. Risk of surgery for congenital heart disease in adult: a multicentered European study. Ann Thorac Surg 2007;83: Padalino MA, Speggiorin S, Rizzoli G, et al. Midterm results of surgical intervention for congenital heart disease in adults: an italian multi center study. J Thorac Cardiovasc Surg 2007;134: Mahle WT, Kirshbom PM, Kanter KR, Kogon BM. Cardiac surgery in adults performed at children s hospital: trends and outcomes. J Thorac Cardiovasc Surg 2008;136: Karamlou T, Diggs BS, Person T, Ungerleider RM, Welke KF. National practice patterns for management of adults congenital heart disease. Circulation 2008;118: Russell JL, LeBlanc JG, Sett SS, Potts JE. Risks of repeat sternotomy in pediatric cardiac operations. Ann Thorac Surg 1998;66: Elahi MM, Kirke R, Lee D, Dhannapuneni RR, Hickey MS. The complications of repeat median sternotomy in paediatrics: six-months follow-up of consecutive cases. Interact Cardiovasc Thorac Surg 2005;4: Morales DLS, Zafar F, Arrington KA, et al. Repeat sternotomy in congenital heart surgery: no longer a risk factor. Ann Thorac Surg 2008;86: Harrild DM, Berul CI, Cecchin F, et al. Pulmonary valve replacement in tetralogy of Fallot. Impact on survival and ventricular tachycardia. Circulation 2009;119: Therrien J, Siu SC, McLaughlin PR, Williams WG, Webb GD. Pulmonary valve replacement in adults late after repair of tetralogy of Fallot: are we operating too late? J Am Coll Cardiol 2000;36: Frigiola A, Tsang V, Bull C, et al. Biventricular response following pulmonary valve replacement for right ventricular outflow tract dysfunction. Is age predictor of outcome? Circulation 2008;118(14 suppl):s Frigiola A, Giamberti A, Chessa M, et al. Right ventricular restoration during pulmonary valve implantation in adults with congenital heart disease. Eur J Cardiothorac Surg 2006;29(suppl):S Giamberti A, Chessa M, Abella R, et al. Surgical treatment of arrhythmias in adults with congenital heart defects. Int J Cardiol 2008;129: INVITED COMMENTARY The work of Giamberti and colleagues [1] has several interesting features. Reoperations in adult congenital heart disease (ACHD) patients due to the right ventricular outflow tract, the left ventricular outflow track, residual septal defects, and in Fontan patients were associated with a low rate of mortality (3.6%). With just 6 deaths, it is not surprising that statistical analysis of the data did not identify any clinical features associated with in-hospital death, although the fact that 4 deaths occurred among the subset of 13 patients with a failing Fontan circulation supports the view that such patients require very careful preoperative assessment. Severe morbidity, as defined in their article, was frequent at 15% and associated with a number of patientspecific features (high hematocrit/cyanosis, congestive heart failure, and number of previous operations), operation details (Fontan operation/conversion), and the cardiopulmonary bypass time. We would suggest that such 2009 by The Society of Thoracic Surgeons /09/$36.00 Published by Elsevier Inc doi: /j.athoracsur