Author + information
- Received February 7, 1996
- Revision received June 17, 1996
- Accepted November 5, 1996
- Published online March 1, 1997.
- Joseph M Forbess, MDA,
- Nancy Cook, DScA,
- Alain Serraf, MDA,
- Redmond P Burke, MDA,
- John E Mayer Jr., MD, FACCA and
- Richard A Jonas, MD, FACCA,*
- ↵*Dr. Richard A. Jonas, Department of Cardiovascular Surgery, Children’s Hospital, 300 Longwood Avenue, Boston, Massachusetts 02115.
Objectives. The aim of this study was to perform a retrospective analysis of an institutional experience with a consecutive series of patients with post–stage I palliation for hypoplastic left heart syndrome (HLHS).
Background. In a recent review of 212 consecutive patients who underwent stage I operations for HLHS at our institution between 1983 and 1993, we identified risk factors related to stage I mortality. We sought to examine the outcome for these patients at subsequent palliative procedures.
Methods. All patients who underwent stage I reconstruction between January 1983 and June 1993 and also underwent subsequent palliation at our institution were included. Seventy patients underwent palliative procedures and two underwent heart transplantation. Patient-specific factors and features of the stage II operation were analyzed for impact on stage II mortality and actuarial survival.
Results. The only independent risk factor for stage II mortality was the performance of a nonfenestrated Fontan operation (p < 0.001). There were nine in-hospital deaths (69%) in the 13 patients undergoing the nonfenestrated Fontan procedure at stage II. Fifty patients underwent intermediate superior vena cava to pulmonary artery anastomosis at stage II, with 4 (8%) early deaths. Pulmonary artery augmentation was performed in 19 patients (38%) at stage II, without increased operative risk. Hypoplastic left heart syndrome anatomic subtype did not influence stage II mortality. The modified fenestrated Fontan procedure has been performed as a third stage in 32 patients whose median age was 28.7 months, with one early death at a median follow-up of 24.5 months.
Conclusions. A second-stage bidirectional cavopulmonary anastomosis for HLHS reduces second-stage mortality and improves intermediate survival. The modified fenestrated Fontan operation may then be performed as a final palliative stage with low operative risk.
(J Am Coll Cardiol 1997;29:665–70)
Hypoplastic left heart syndrome (HLHS) is a common congenital heart defect that is responsible for more neonatal deaths in the United States than any other anomaly (). Since the first successful reconstructive palliative procedure was described by Norwood in 1983 (), several centers have reported sizeable clinical series using this approach to HLHS ([3–8]). The original surgical strategy proposed by Norwood involved a neonatal stage I reconstruction, wherein the main pulmonary artery is connected to the aorta with pulmonary blood flow established through a systemic to pulmonary shunt, followed by a second-stage Fontan operation (). As the high risk nature of the Fontan operation as a second stage was recognized, the bidirectional cavopulmonary anastomosis was introduced in 1988 as a second and intermediate stage before completion of the Fontan operation (). During the third stage, we currently undertake a fenestrated Fontan operation ().
We have previously reported our 10-year institutional results for stage I palliative reconstruction in a consecutive series of 212 patients who underwent first-stage palliation between January 1983 and June 1993 (). This report summarizes the results of subsequent palliative reconstructions for this same consecutive series. Specifically, we sought to identify any patient- or procedure-specific risk factors related to second-stage palliation for HLHS. In addition, we sought to examine the outcome for those patients who have undergone third-stage modified fenestrated Fontan operations.
All patients who underwent stage I reconstruction at our institution between January 1983 and June 1993 and also underwent subsequent palliative procedures at our institution were included in this retrospective analysis. There were 114 stage I operative survivors. Of these, two patients crossed over to heart transplantation and 70 underwent further palliative reconstructions. Patient data were compiled by a review of clinical records, including operative reports and preoperative imaging studies. The diagnosis of HLHS was based on angiographic or two-dimensional echocardiographic evidence of a diminutive ascending aorta, aortic atresia or stenosis and a hypoplastic left ventricle. Operative survival was defined as survival >30 days and the ability to leave the hospital. Late follow-up was obtained from direct patient contact or from the patient’s primary cardiologist for 71 of 72 patients during a 1-month period ending February 28, 1995. This comprised a total of 279 patient years, with a mean follow-up of 3.4 years and a follow-up range of 0.5 to 8.0 years from the date of second-stage palliation.
1.2 Statistical analysis.
Potential risk factors considered for operative (early) mortality included patient characteristics and surgical procedures. Rates were compared in contingency table analyses using the chi-square or Fisher exact test. Dichotomous variables denoted the presence or absence of each condition. For continuous variables, tests for trend in proportions across quartile categories were performed. Age at stage II procedure was classified into four categories with cutoff points at 8 months, 1 year and 2 years of age. Right ventricular function and tricuspid regurgitation were classified as none/normal, mild, moderate or severe. Categoric variables such as anatomic subtype and stage II procedure were tested both with an overall chi-square test and as dichotomous variables, comparing each condition to the rest. Multivariate analysis of stage II operative mortality with these risk factors was performed using stepwise logistic regression. For late mortality, survival time among those who initially survived the procedure was estimated using Kaplan-Meier curves. Survival estimates according to type of stage II procedure were compared using the log-rank test. All analyses were performed using a standard, commercially available software package (SAS Institute, Inc.).
2.1 Hypoplastic left heart syndrome anatomic subtypes.
We have previously reported the strong influence of anatomic subtype on both stage I survival and pre–stage II mortality in our institutional experience ([5, 6]). The significantly better stage I survival of those patients with mitral stenosis and aortic stenosis is reflected by the fact that although this group comprised only 24% of the original cohort, 44% of all stage II palliative procedures were in patients with mitral stenosis and aortic stenosis (Fig. 1).
2.2 Pre-stage II interim procedures.
Post–stage I patients underwent a total of 38 interim operative procedures before stage II (Table 1). The most frequently performed interim procedure was a tube gastrostomy, which was performed for failure to thrive in 12 patients. Ten patients underwent balloon dilation of either the neoaorta or descending thoracic aorta, and five patients required surgical aortoplasty to treat neoaortic obstruction or persistent coarctation. Four patients underwent surgical pulmonary artery augmentation, three of whom also had their Blalock-Taussig shunts enlarged during this procedure. Three patients required pericardial windows for persistent hemodynamically significant pericardial effusions.
2.3 Stage II procedures.
As shown in Table 2, a variety of second-stage procedures were used during the study period. A total of 18 patients underwent some form of modified Fontan procedure. Six patients had a right atrial to pulmonary artery anastomosis performed with a pulmonary venous to tricuspid valve baffle. Seven patients underwent the modified Fontan procedure using the lateral atrial tunnel technique ([11, 12]). Five patients underwent the lateral atrial tunnel Fontan procedure with a fenestration of the prosthetic inferior vena cava to superior vena cava baffle. Beginning in 1988, 50 patients underwent an intermediate superior vena cava to pulmonary artery anastomosis in preparation for completion of the Fontan procedure. Forty-seven patients underwent the bidirectional Glenn anastomosis. In seven of these with bilateral superior vena cavae, bilateral anastomoses were performed. Three patients underwent the hemi-Fontan procedure, wherein a double superior vena cava to right pulmonary artery anastomosis was performed and the internal orifice of the superior vena cava was closed with a prosthetic patch. Two patients had classic Glenn shunts performed with the placement of left modified Blalock-Taussig shunts, and two patients underwent orthotopic heart transplantation.
The distribution of the stage II procedure (Fig. 2) shows that, since the earliest application of the bidirectional cavopulmonary anastomosis used to treat HLHS at our institution in 1988, this has become our procedure of choice as a second-stage palliative procedure. The second-stage bidirectional cavopulmonary connection is typically performed at a much earlier age than was the modified Fontan operation (Fig. 3). The bidirectional cavopulmonary anastomosis was performed at a median age of 8.25 months during the study period.
Twenty-eight concomitant corrective procedures were performed at the time of second-stage palliation (Table 3). Nineteen patients required pulmonary artery augmentation. Four patients underwent patch aortoplasty. Two patients had their pulmonary artery to aorta synthetic tube conduits removed and replaced by a direct pulmonary artery to ascending aortic anastomosis, augmented by a gusset of pulmonary homograft. Three patients underwent atrial septectomy for a restrictive atrial septal defect detected at pre–stage II catheterization.
2.4 Stage II mortality.
Table 4summarizes the results of all stage II procedures performed during the study period. The early mortality rate for the completed, nonfenestrated modified Fontan procedure was 69%, with an overall mortality rate of 77% during the study period. In the 50 patients who underwent some form of bidirectional cavopulmonary shunt, there were four operative deaths (8%). Three of these 46 bidirectional cavopulmonary shunt operative survivors died during follow-up. Two patients died suddenly out of the hospital, and one patient died during the stage III Fontan procedure.
2.5 Stage II risk factor analysis.
Patient-specific anatomic and physiologic data, as well as procedural features of the stage II operation, were evaluated as potential risk factors for stage II operative mortality in both univariate and multivariate analyses (Table 5). Performance of a nonfenestrated Fontan procedure as a stage II procedure, as well as patient age and weight greater than median, were significant risk factors for stage II mortality in univariate analysis. Also, a systemic arterial saturation ≥81% at the time of pre–stage II catheterization approached significance as a risk factor in univariate analysis (p = 0.051). In multivariate analysis using stepwise logistic regression, the sole independent risk factor for stage II mortality was the performance of a nonfenestrated Fontan procedure, although the power for multivariate analyses was low in these data.
2.6 Post–stage II actuarial survival.
Actuarial survival curves for the three predominant stage II procedures are shown in Fig. 4. Actuarial survival was significantly improved for patients undergoing a bidirectional cavopulmonary anastomosis as a second-stage procedure (p < 0.001). The actuarial survival rate at 3 years for all patients undergoing a bidirectional cavopulmonary anastomosis as a second-stage procedure was 86%. The median time to performance of a stage III fenestrated Fontan operation after stage II procedures was 1.26 years.
2.7 Stage III fenestrated Fontan operation.
During the study period, a total of 52 patients, including the two patients who had classic Glenn procedures, underwent pre–stage II Fontan procedures. To date, 32 subsequent fenestrated Fontan procedures have been performed, resulting in one death (3%) from a presumed ventricular arrhythmia on the day of hospital discharge. Twenty-five of the 31 stage III Fontan survivors are presently in New York Heart Association functional class I or II. Six patients have undergone subsequent transcatheter closure of the baffle fenestration, and spontaneous closure has been documented by echocardiography in seven patients. Nine patients have required subsequent cardiac intervention. In addition to the six patients undergoing device closure of their baffle fenestration, two patients have undergone implantation of permanent pacemakers for sinus node dysfunction, and one patient has undergone balloon dilation of both pulmonary arteries.
The systemic arterial oxygen saturations of patients undergoing three-stage palliation are shown in Fig. 5. Mean prehospital discharge saturation after stage I palliation was 75%. After the bidirectional cavopulmonary connection, mean arterial oxygen saturation increased to 86%. After a stage III fenestrated Fontan procedure, the mean systemic arterial oxygen saturation at latest follow-up was 90% for those patients with a patent fenestration and 96% for those patients with transcatheter device closure or echocardiographic evidence of spontaneous fenestration closure.
3.1 Anatomic subtype does not influence outcome for stages II and III.
Although the first successful completion of the Fontan operation for HLHS was performed over a decade ago (), the reconstructive approach to this anomaly continues to challenge clinicians who treat these patients. We have previously described our results with stage I palliation of HLHS (). In that consecutive institutional experience, patients with atresia of either the mitral or aortic valve were noted to have higher operative mortality and higher pre–stage II mortality than patients with mitral stenosis and aortic stenosis. The present study has demonstrated that HLHS anatomic subtype does not influence outcome at second- or third-stage operations.
3.2 Staged approach to fenestrated Fontan procedure improves HLHS survival.
The addition of the bidirectional cavopulmonary anastomosis as a second stage, before completion of the fenestrated Fontan operation, has been associated with improved intermediate survival for stage I operative survivors. The high early mortality after completion of the stage II Fontan procedure in this group of patients was attributable to a progressively low cardiac output during the early postoperative period ([13–15]).
We have performed the fenestrated Fontan operation as a third-stage procedure in this group with low operative risk, stable intermediate survival and good intermediate cardiac functional outcome in most patients. These data are in agreement with those of Jacobs and Norwood (), who have reported improved results with staging toward the Fontan operation and improved Fontan results using baffle fenestration.
The theoretic considerations, which these data support, for this operative sequence have been discussed previously ([4, 9, 17–21]). The major beneficial contribution of the bidirectional cavopulmonary anastomosis to stage I anatomy is the removal of the excess ventricular volume loading inherent in arterial shunt-dependent single ventricle physiology. The majority of stage I survivors in this series received a 3.5-mm modified Blalock-Taussig shunt. At the time of pre–stage II evaluation, these growing patients were usually demonstrating increased cyanosis from inadequate shunt flow. For this reason, we also noted an increase in mean arterial saturation after the stage II bidirectional cavopulmonary anastomosis (Fig. 5).
3.3 Fontan circulation versus cardiac allotransplantation for HLHS.
Some investigators have expressed concern about the long-term outcome for patients with a Fontan circulation versus those with a biventricular circulation ([22–24]). This has led some to advocate cardiac allotransplantation for neonates with HLHS ([25–27]). This approach is limited by inadequate numbers of appropriate organ donors (), however, and the unknown effects of lifelong immunosuppression. In our HLHS patients with Fontan anatomy, we observed stable ventricular performance and cardiac functional status over the time frame analyzed. Cardiac interventions unrelated to fenestration closure have been the exception rather than the rule.
3.4 Factors potentially influencing composition of the study group.
We have previously reported increased stage I mortality and pre–stage II attrition in these patients with HLHS with either mitral or aortic atresia compared with patients with mitral stenosis and aortic stenosis (). We have speculated that this difference may be the result of technical variations in stage I procedure types used for these anatomic variants. We have presently standardized our stage I operation to what we have previously labeled a type 1 procedure, wherein the ascending aorta is anastomosed in a side to side fashion to the divided main pulmonary artery and the neoaortic arch is augmented using a homograft gusset (). Additional analysis of a current cohort of patients would be required to determine if anatomic subtype continues to influence pre–stage II attrition for stage I survivors, since this standardization.
Clearly the pre–stage II attrition for stage I survivors of all anatomic subtypes has decreased. During the second half of this study period, 85% of all stage I hospital survivors have undergone subsequent second-stage operation (). We are encouraged by these results and continue to offer staged reconstruction to all patients with HLHS who are adequately resuscitated, have satisfactory ventricular function and do not have additional severe congenital anomalies. Only continued follow-up of these patients with Fontan anatomy will reveal the long-term outcome for this treatment strategy for HLHS.
☆ This study was presented in part at the 44th Annual Scientific Session of the American College of Cardiology, New Orleans, Louisiana, March 1995.
- Received February 7, 1996.
- Revision received June 17, 1996.
- Accepted November 5, 1996.
- The American College of Cardiology
- Murdison KA,
- Baffa JM,
- Farrell PE,
- et al.
- Forbess JM,
- Cook N,
- Roth SJ,
- Serraf A,
- Mayer JE,
- Jonas RA
- Meliones JN,
- Snider AR,
- Bove EL,
- Rosenthal A,
- Rosen DA
- Bridges ND,
- Jonas RA,
- Mayer JE,
- Flanagan MF,
- Keane JF,
- Castaneda AR
- Bridges ND,
- Lock JE,
- Castaneda AR
- Seliem MA,
- Muster AJ,
- Paul MH,
- Benson DW
- Gewillig M,
- Daenen W,
- Aubert A,
- Van der Hauwert L
- Jonas RA
- Chang AC,
- Hanley FL,
- Wernovsky G,
- et al.
- Driscoll DJ,
- Offord KP,
- Feldt RH,
- Schaff HV,
- Puga FJ,
- Danielson GK
- Fontan F,
- Kirklin JW,
- Fernandez G,
- et al.