Hypoplastic left heart syndrome (HLHS) represents 1.4–3.8% of congenital heart diseases but is responsible for 23% of cardiac deaths during the first week of life.1–3 HLHS is characterized by variable degrees of underdevelopment of left heart structures.4–6 Traditional management of HLHS, either with staged surgical palliation (Norwood, Glenn, Fontan)7,8 or cardiac transplantation,9 remains a challenge in most centers, particularly in high-risk patients such as those with low birth weight (<3 kg) and diminutive aortas (<2 mm).10–12 For this subgroup of patients, a hybrid approach – combining transcatheter and surgical techniques – was devised in the last decade, aiming to combine the best characteristics of surgical and interventional cardiology techniques.13–15

Originally reported in 1993,16 hybrid stage 1 palliation consists of bilateral branch pulmonary artery banding and stenting of the ductus arteriosus, without cardiopulmonary bypass, during the neonatal period. Assurance of a nonrestrictive atrial septal defect is the next step, through either balloon atrial septostomy or stenting. Later in infancy, a more complex comprehensive stage 2 is performed at approximately six months of age, in the form of a Norwood-type reconstruction combined with a bidirectional cavopulmonary (Glenn) anastomosis. Although initially suggested for high-risk candidates for the classic Norwood stage 1 and 2 palliation,14,17 it has since been adopted as the preferred first option by several centers.14,18,19

In 2007, our center started a hybrid program for HLHS palliation in high-risk cases. Our aim is to report our initial results with this technique, focusing on its efficacy, safety and late outcome, with particular emphasis on morbidity and mortality, need for unplanned reintervention, and the current status of the patients.

We report our initial results with the hybrid approach for palliation of patients with HLHS. These results reflect our technical challenges and learning curve.

In recent years, several strategies have had a noticeable impact on the disease burden of HLHS, most notably the right ventricle to pulmonary artery Sano modification21 and the development of dedicated teams for close monitoring of HLHS patients between stages 1 and 2 of palliation.22 It has been shown, however, that there are subgroups of patients within the spectrum of HLHS which have poorer outcomes when treated by traditional surgical palliation.10,23 Also, the classical staged palliation approach appears to have reached a stagnation point,24 no further improvements being possible with the currently available technology, with recent cumulative early and interstage mortality of 5%–30% for standard-risk patients25–27 but as high as 30%–50% in high-risk patients.11,12 A recent Pediatric Heart Network-sponsored multicenter randomized trial compared the Norwood procedure with a modified BT shunt versus the Sano modification,26 with the primary endpoint of death or transplantation at one year and the secondary endpoints of hospital course, RV function by echo, pulmonary artery size by angiography, unintended cardiovascular interventions, and serious adverse events and complications. This study failed to show clear superiority between any of the approaches in the long term. Although the Sano modification group was found to have a statistically significant lower risk of mortality at the 12-month endpoint, this was no longer significant with a longer follow-up. The complication rate was also higher in the Sano group, although the percentage of infants with at least one complication was the same in both groups.26

Hybrid palliation has evolved as an alternative to Norwood palliation based on the assumption that a less invasive stage 1 palliation in the hybrid strategy would improve overall survival and neurological outcomes.28 The goals of hybrid stage 1 palliation include the following29: (1) to ensure unimpeded systemic cardiac output through a patent ductus arteriosus; (2) to obviate the need for and complications associated with long-term prostaglandin E1 (PGE1) infusion; (3) to improve balance of the pulmonary and systemic circulations by restricting pulmonary blood flow; and (4) to relieve any obstruction to pulmonary venous return at the atrial septum level. One of the most important advantages of hybrid palliation is to defer the most complex surgical intervention to later in infancy, reducing the risk of neurological complications. In our series the majority of patients were extubated in the first 48 hours and were discharged home after a median hospital stay of 24 days, significantly shorter than for the traditional Norwood procedure, and have a normal neurological assessment to date.

The hybrid approach is also more flexible than the conventional staged approach.30 Some authors have even modified the hybrid approach and propose a partial hybrid stage I palliation in the form of bilateral PA banding with continuous PGE1 administration.31 It may be used as a bailout or salvage approach,32 providing excellent palliation for patients for whom transplantation is considered the most appropriate long-term approach. In such circumstances the time pressure to find a suitable donor heart is relieved and the child is likely to be exposed to fewer blood products, reducing problems with later organ matching. Likewise, some patients with a borderline left ventricle may be palliated using the hybrid approach, allowing further growth or recovery of cardiac function so that a two-ventricle repair may ultimately be attempted.33,34 While it is possible to take down a Norwood stage 1, this is a much larger undertaking than converting a hybrid procedure.

Despite the constant evolution of interventional techniques, the hybrid approach still faces certain limitations. A particularly serious one is the occurrence of retrograde aortic arch obstruction, leading to suboptimal cerebral and coronary artery perfusion.18,15 In our series the two deaths after stage 1 palliation were directly attributable to obstructed systemic cardiac output through the ductus arteriosus. Retrograde aortic arch obstruction can occur in utero, postnatally before surgical treatment begins, or afterward in acute or more insidious fashion after ductus stenting.35 Its incidence ranges from 8% to 30%. Some centers have managed this problem by stenting the stenotic isthmus,36 thus rendering reconstructive arch surgery even more challenging, or by creating a prophylactic reverse Blalock-Taussig shunt,37 while others consider evidence of restricted flow into the retrograde transverse aorta from the ductus arteriosus an exclusion criterion for the hybrid procedure.15 Nevertheless, it is highly desirable to avoid a hybrid procedure in patients with obstruction to retrograde arch blood flow, which leads to poorer results and significantly increased mortality.

Furthermore, several authors argue that the hybrid approach fails to attain some of the essential prerequisites for an optimal Fontan-type circulation, classically described by Choussat and Fontan.38 For these authors specific physiological risk factors for a failing Fontan prevail, related to ventricular performance, atrioventricular and systemic valve function, and pulmonary and systemic circulation. Some of the potential problems reported include: (1) risk of residual aortic arch obstruction, due to either inappropriate stent deployment leaving distal ductal tissue uncovered or to intrastent occlusion15,39; (2) increased single ventricle afterload due to the non-compliant/non-expansible nature of the stent placed in the ductus arteriosus and to excessively tight PA bands that may also (3) fail to provide adequate pulmonary blood flow or to promote adequate PA growth or even40 (4) induce iatrogenic PA stenosis, necessitating PA reconstruction in the comprehensive stage 2 operation41; (5) presence of increasingly restrictive atrial communication, leading to higher pulmonary venous resistance; (6) decreased single-ventricle function (systolic, diastolic or both), due to all the above-mentioned factors and leading to (7) increased atrioventricular valve regurgitation. In our series, prior to the comprehensive stage 2, the echocardiographic assessment of right ventricular function was graded as normal in all but two patients, in whom it was graded as mildly depressed and who had mild to moderate tricuspid regurgitation. These patients had their aortic arch obstruction corrected percutaneously in the interstage cardiac catheterization procedure. In three patients PA reconstruction was deemed necessary. These patients are now waiting for completion of their bicavopulmonary circulation, but our series is still small and lacks a longer follow-up.

In our limited experience, the timing of the atrial interventional procedure is the most difficult decision. In general we tend to perform it around 10 days after the hybrid procedure but, either due to our limited experience or because of morphologically difficult atrial septums, several of our patients needed surgical septectomy.

To obviate these potential limitations, the hybrid technique has evolved during the last decade.15 Several strategies have been proposed, including: (1) deferring the atrial septostomy to a second procedure, immediately before discharge, allowing for a more durable septostomy by using a larger balloon in a larger left atrium; (2) use of cutting and static balloons for the thick atrial septum; (3) use of self-expanding ductal stents; and (4) performing an earlier comprehensive stage 2 procedure to avoid (a) prolonged branch PA banding, promoting enhanced PA growth and obviating the need for branch PA reconstruction; (b) prolonged single-ventricle strain against a non-compliant stent, preserving ventricular function and avoiding atrioventricular valve regurgitation; (c) pulmonary venous return obstruction, so that the patient is in a better condition for the final Fontan operation.

In high-volume centers with modern surgical, anesthetic and intensive care practices, surgical results for average-risk patients born with HLHS to completion of stage 2 palliation are still superior to those offered by the hybrid approach. This is mainly due to mortality during the comprehensive stage 2 surgery, a long and complex operation. For high-risk patients, the results appear to be equivalent, with 50% anticipated survival. Our results are in agreement with reported international series, with survival to completion of stage 2 palliation of 66%. However, there is a learning curve effect that would suggest that results of hybrid palliation will improve with more experience, even for average-risk patients.

In conclusion, there has been considerable progress with the hybrid approach. Though it is relatively early, current international results suggest there may be subgroups of patients that are better palliated with this approach than by surgery. In most institutions surgical results are currently superior for other subgroups of patients. However, as experience in interventional palliation increases, this position will need to be reassessed with multicenter prospective trials comparing both classic and hybrid approaches. Above all, the existence of dedicated teams for the diagnosis and management of these patients, preferably in high-volume referral centers, is of major importance in this condition.

The authors have no conflicts of interest to declare.