As surgical results of palliation of patients with functionally univentricular hearts continue to improve, continued high interstage mortality has stimulated the development at many centres of clinical programmes and research focussed on improving interstage outcomes. The high mortality experienced by centres that care for infants with hypoplastic left heart syndrome has been reduced or nearly eliminated in some programmes that have adopted interstage home monitoring, whereas others continue to report interstage mortality as high as 22%.Reference Ghanayem, Hoffman and Mussatto1–Reference Ohye, Sleeper and Mahony6 Causes of interstage mortality are multifactorial and unlikely to be improved by implementing a therapeutic strategy that targets a single aetiology.Reference Hehir, Dominguez and Ballweg4, Reference Bartram, Grunenfelder and Van7 The goal of interstage monitoring is therefore to provide heightened surveillance of physiologic instability that may precede important haemodynamic deterioration, allowing for early intervention and prevention of potentially fatal interstage events. Studies of interstage outcomes have highlighted the prevalence of failure to thrive during infancy, and the impact of programmes of feeding and nutrition on interstage growth and outcomes after the superior cavopulmonary anastomosis.Reference Anderson, Beekman and Border8–Reference Davis, Davis and Cotman11 We review
• the existing evidence regarding somatic growth in this population,
• published programmes regarding feeding and nutrition, and
• existing programmes of interstage home monitoring that have led to improved results.
In addition, we will discuss data presented at HeartWeek 2011, from Cardiology 2011, the 15th Annual Update on Pediatric and Congenital Cardiovascular Disease and the 11th Annual International Symposium on Congenital Heart Disease.
Interstage surveillance
Structured interstage home monitoring was first reported by Ghanayem et alReference Ghanayem, Hoffman and Mussatto1 nearly a decade ago and was found to be associated with a marked reduction in interstage mortality, which is defined as mortality after discharge from the hospitalisation for the Norwood (Stage 1) operation, and before the cavopulmonary anastomosis. These results have been reproduced by others (Table 1), including the experience from Kiel, Germany, demonstrating a similar reduction from an interstage mortality of 15– 0% following adoption of home monitoring in 2005.Reference Furck, Uebing and Hansen2 Programmes of interstage home monitoring have since been developed by many centres, and have taken many forms.Reference Schidlow, Anderson and Klitzner12 Recently, the National Pediatric Cardiology Quality Improvement Collaborative of the Joint Council on Congenital Heart Disease has made the reduction of interstage mortality its primary mission, and has involved many centres across the United States of America in a collaborative approach to reducing interstage mortality.
Studies reporting interstage mortality.
BTS = Blalock–Taussig shunt; RVPA = right ventricle to pulmonary artery conduit
At the Children's Hospital of Wisconsin, the programme of interstage home monitoring continues to function as a multi-disciplinary, adjunctive clinical service. Since its inception, the programme has undergone frequent cycles of review and modification, with the goal of ongoing quality improvement. In addition to the support provided by their paediatricians and primary cardiologists, families have weekly phone calls and return for visits to the clinic every 2–4 weeks with the team of the programme of home monitoring. Before discharge, parents are trained to obtain and record daily saturations of oxygen, weight, and volume of enteral intake. Once discharged, parents are instructed to notify the team of the programme of home monitoring for the following criteria:
• enteral intake less than 100 millilitres per kilogram per day,
• weight loss greater than 30 grams,
• failure to gain at least 20 grams over a 3-day period, and
• oxygen saturation below 75% or above 90%.
Failure to meet the criteria triggers evaluation by the multi-disciplinary team. In our recent experience, 89 out of 145 (55%) patients experienced violations of the protocol resulting in evaluation by the team of the programme of home monitoring, with 46 out of 89 isolated to a problem of feeding or growth resulting in nutritional intervention alone. These interventions included
• increasing the total volume, caloric density, or modifying the type of formula,
• admission to the hospital for acute loss of weight or dehydration, and
• interval placement of either gastrostomy tube or nasogastric tube.
Growth of infants with hypoplastic left heart syndrome
Failure to thrive is a common clinical concern in congenital cardiac disease, and is particularly problematic in infants with functionally univentricular hearts undergoing palliation. Failure to thrive is a well-described sequelae of significant congenital cardiac lesions and may indicate the presence of decompensated cardiac failure.Reference Davis, Davis and Cotman11, Reference Golbus, Wojcik, Charpie and Hirsch13, Reference Menon and Poskitt14 Following the Norwood (Stage 1) operation, the physiologic state of the shunt-dependent infant with parallel circulation is marked by cardiac failure that is often tenuous and challenging to manage. This physiology, often combined with any of the following potential co-morbidities, may contribute to impaired growth of the infant:
• chronic hypoxia,
• persistent tachypnea,
• venous congestion,
• overload of fluid,
• gastro-oesophageal reflux,
• inability to feed or intolerance of feeding,
• genetic syndromes, and/or
• additional non-cardiac anomalies.
Malnutrition is a significant risk factor for surgical outcomes, and specifically has been associated with worse outcomes at the time of cavopulmonary anastomosis and in the long term.Reference Anderson, Beekman and Eghtesady15, Reference Anderson, Kalkwarf, Kehl, Eghtesady and Marino16 Most reports of growth in infancy in patients with hypoplastic left heart syndrome have documented a high incidence of failure to thrive. With few exceptions, there have been retrospective and descriptive studies describing the state of the problem. In contrast, in a study of 145 consecutive patients prospectively enrolled in the programme of interstage home monitoring at Children's Hospital of Wisconsin, patients achieved an interstage velocity of growth of 26.8 grams per day, which parallels the normal growth of an infant, and experience “catch-up growth” during the early interstage period, with improvement of the weight-for-age z-score from the time of neonatal discharge from the hospital to Stage 2 palliation (Hehir, unpublished data presented at Cardiology 2011).
Early growth: the hospitalisation for the Norwood (Stage 1) operation
Although patients with hypoplastic left heart syndrome typically have near-normal weight at birth, they may experience significant failure to thrive during the hospitalisation for the Norwood (Stage 1) operation. Pre-operative enteral nutrition is often withheld because of a variety of concerns:
• the potential for gastrointestinal complications resulting from the dependence of systemic output on the patency of the arterial duct,
• the usage of prostaglandin,
• pre-existing dysfunction of organs and cardiac failure, and
• the presence of umbilical catheters.
It is likely that patients undergo delayed introduction of enteral nutrition post-operatively as well, because of ongoing concerns for gastrointestinal complications, with little evidence to support the delayed advancement of feeding as part of a routine post-operative protocol. In addition, patients typically require inotropic support, mechanical ventilation, and diuretics, and may suffer from any of the following post-operative complications, all of which may conspire to increase caloric requirements and/or further delay reaching goals of nutritional intake:
• infection,
• overload of fluid including pleural effusions and ascites,
• intolerance of feeding, and
• dysfunction of a vocal cord.
It is also possible that guidelines for post-operative nutrition in this population may under-estimate actual needs, and empiric caloric goals should be re-assessed. Li et alReference Li, Zhang and Herridge17 found a surprisingly high energy deficit in the early post-operative period following the Norwood (Stage 1) operation. Neonates in this study were found to be hypermetabolic and did not reach a positive energy balance until post-operative day 3 despite the use of standard regimens of parenteral nutrition.
Multiple studies have demonstrated an acute decline in weight-for-age z-scores in the early post-operative phase. In a study of 61 infants undergoing Norwood operation, Medoff-Cooper et alReference Medoff-Cooper, Irving and Marino18 found that the mean change in weight-for-age z-score during the neonatal hospitalisation, from date of surgery to discharge, was −1.5 plus or minus 0.8. In this study, patients discharged on all oral feeds grew better, and 60% of the variance in weight-for-age z-score change could be explained by
• a lower weight-for-age z-score at birth,
• the presence of moderate or greater atrioventricular valvar regurgitation,
• longer post-operative time of ventilation, and
• placement of an additional central venous line.
Srinivasan et alReference Srinivasan, Jaquiss and Morrow19 demonstrated that weight-for-age z-scores steadily decline following the Norwood (Stage 1) operation until the time of Stage 2 palliation, and longer length of stay in the hospital following the Norwood (Stage 1) operation results in lower weight-for-age z-scores at the time of discharge. Similar trends of declining weight-for-age z-scores during the hospitalisation for the Norwood (Stage 1) operation have been reproduced and have illuminated a number of potentially modifiable risk factors for failure to thrive, leading to interventions ranging from initiation of pre-operative feeding to implementation of interstage nutritional protocols.Reference Anderson, Beekman and Border8–Reference Vogt, Manlhiot, Van, Russell, Mital and McCrindle10
Growth during the early interstage period
With few exceptions, existing studies document ongoing failure to thrive and deterioration of weight-for-age z-scores during the early interstage period (Table 2). Interestingly, both Anderson et alReference Anderson, Beekman and Border8 and Kelleher et alReference Kelleher, Laussen, Teixeira-Pinto and Duggan9 found higher arterial saturation to have an inverse relationship with gaining weight during the early interstage period. This finding likely reflects the relationship of gaining weight and the severity of pulmonary overcirculation in this population; a higher saturation may represent an increase in the flow of pulmonary blood, resulting in greater volume and workload on the functionally univentricular heart, leading to impaired growth.
Studies reporting growth of infants with hypoplastic left heart syndrome.
BDG = bidirectional Glenn; TGA = transposition of the great arteries; WAZ = weight-for-age z-score
Growth at the time of Fontan completion and beyond
Long-term failure to thrive is well recognised in patients with functionally univentricular hearts and may have significant implications for clinical outcomes. Anderson et alReference Anderson, Kalkwarf, Kehl, Eghtesady and Marino16 found that patients with a weight-for-age z-score of less than −2 at the time of Fontan completion had a higher rate of post-operative infection, a significant determinant of length of stay. The early deterioration in weight-for-age z-scores following the Norwood (Stage 1) operation is typically reversed following cavopulmonary anastomosis, with “catch-up growth” demonstrated through the time of Fontan completion.Reference Vogt, Manlhiot, Van, Russell, Mital and McCrindle10, Reference Srinivasan, Jaquiss and Morrow19–Reference Hasan, Bendaly, Alexy, Ebenroth, Hurwitz and Batra22 This finding has been attributed to the alleviation of the volume load on the functionally univentricular heart by transitioning to series circulation. Some studies suggest that an earlier age of surgical unloading promotes “catch-up growth”.Reference Vogt, Manlhiot, Van, Russell, Mital and McCrindle10, Reference Srinivasan, Jaquiss and Morrow19–Reference Stenbog, Hjortdal, Ravn, Skjaerbaek, Sorensen and Hansen21 The importance of the volume load on the functionally univentricular heart is further demonstrated in the finding by Vogt et alReference Vogt, Manlhiot, Van, Russell, Mital and McCrindle10 that patients with two or more venous collateral vessels following cavopulmonary anastomosis experienced a decline in weight-for-age z-score, compared with normal patterns of growth found in those patients without collaterals, or who had their collaterals occluded.
In addition to the aforementioned risk factors, the Fontan circulation may impart specific challenges to normal growth, including loss of protein due to ascites and protein-losing enteropathy. In a comparison of serial growth measurements of 80 patients after the Fontan operation to 66 patients after the Mustard operation, both groups experienced “catch-up growth” following their respective surgeries.Reference Hasan, Bendaly, Alexy, Ebenroth, Hurwitz and Batra22 However, the patients after the Fontan operation continued to exhibit low height, whereas patients after the Mustard operation normalised their height over time. Although long-term failure of linear growth is common, it is poorly understood. Chin et alReference Chin, Stephens, Goldmuntz and Leonard23 have investigated the relationship between failure of linear growth, low cardiac index, and protein-losing enteropathy in a study that found alkaline phosphatase, a serum biomarker, to be useful in following and managing these patients. This study sheds light on the complex interaction of growth and cardiopulmonary physiology in patients with functionally univentricular hearts, and highlights the demand to develop more sensitive methods of evaluating failure such as biomarkers, so that early intervention can be performed before the onset of failure to thrive.
Challenges related to feeding in patients with hypoplastic left heart syndrome
Difficulties related to feeding are common in infants with congenital cardiac disease, and can be due to a variety of congenital and acquired anatomical and physiological barriers including:
• non-cardiac anomalies,
• genetic syndromes,
• cardiac failure,
• tachypnoea,
• gastro-oesophageal reflux,
• dysphagia,
• oral aversion, and
• dysfunction of a vocal cord.
The incidence of dysfunction of a vocal cord following the Norwood (Stage 1) operation has been found to be as high as 45% with routine screening, and may contribute to difficulties with feeding in some patients.Reference Skinner, Halstead, Rubinstein, Atz, Andrews and Bradley24 Therapies for improving oral feeding in the presence of dysfunction of a vocal cord may include speech therapy and injection of the vocal cord.
The Sano modification of the Norwood (Stage 1) operation has been proposed as a solution to the diastolic run-off seen in patients receiving a systemic-to-pulmonary arterial shunt. This diastolic run-off is thought to impair splanchnic perfusion and hence potentially lead to gastrointestinal complications. This theoretical benefit has not been observed in several clinical studies.Reference Harrison, Davis and Reid25, Reference Johnson, Ansong and Li26 In an adjunctive study to the Single Ventricle Reconstruction trial sponsored by the Pediatric Heart Network of the National Institutes of Health of the United States of America, Johnson et alReference Johnson, Ansong and Li26 found that although resistive indices in the splanchnic vessels are improved through use of the Sano modification, no observable difference in gastrointestinal outcomes is seen between the types of shunt.
Protocols for feeding patients with hypoplastic left heart syndrome
Owing to the high incidence of difficulties related to feeding and gastrointestinal complications in the infant with hypoplastic left heart syndrome, as well as tremendous variability in clinical practice, interest exists in developing post-operative protocols for feeding in an attempt to standardise management and improve outcomes. Braudis et alReference Braudis, Curley and Beaupre27 demonstrated that implementation of a protocol for post-operative feeding in 36 patients with hypoplastic left heart syndrome resulted in a shorter length of stay, shorter time to achieve goals of caloric intake, and less requirement for parental nutrition when compared with historical controls. Del Castillo et alReference del Castillo, McCulley and Khemani28 documented a reduction in the incidence of necrotising enterocolitis following initiation of a programme for feeding; however, patients experienced delayed initiation of enteral feeds, longer duration to attain full caloric goals, and longer length of stay in the hospital following initiation of the protocol. In a study of infants excluding patients with functionally univentricular hearts, Pillo-Blocka et alReference Pillo-Blocka, Adatia, Sharieff, McCrindle and Zlotkin29 demonstrated that rapid post-operative advancement to formula with high caloric density was safe and resulted in better growth in the hospital as well as shorter length of stay. At the Children's Hospital of Wisconsin, post-operative patients are evaluated daily by a multi-disciplinary team including the inpatient intensivist and cardiologist, cardiology nurse practitioners, and registered dieticians. Initiation and advancement of enteral feeds are guided by patient-specific indicators of readiness such as non-invasive assessment of cardiac output and degree of cardiopulmonary support. Initial goals of 100 cubic centimetres per kilogram per day and 120 kilocalories per kilogram per day are targeted, and then adjusted based on clinical factors and the goal of gaining 20–30 grams of weight per day. At the time of discharge, patients consume 110–120 kilocalories per kilogram per day via fortified breastmilk or infantile formula with a caloric density of 24–27 kilocalories per ounce. Patients are required to demonstrate good gain of weight on the above regimen before discharge from the hospital. This strategy has resulted in interstage growth that is similar to that of the normal growth of infants.Reference Ghanayem, Cava, Jaquiss and Tweddell30, Reference Ghanayem, Tweddell, Hoffman, Mussatto and Jaquiss31
Pre-operative feeding in patients with hypoplastic left heart syndrome
The decision to feed pre-operative patients with a functionally univentricular heart remains controversial. Many practitioners withhold enteral feeds from those patients with ductal-dependent flow of systemic blood because of concerns for insufficient splanchnic perfusion and the risk of necrotising enterocolitis.Reference Pearl, Nelson, Schwartz and Manning32 However, some benefits to pre-operative feeding may exist, including:
• initiation of normal patterns of feeding,
• promotion of salutary hormonal effects,
• development of beneficial native flora in the alimentary tract, and
• promoting maternal bonding.
If and when the decision to feed is made, patients must be monitored closely for evidence of low cardiac output and for the development of intolerance to feeding. Some reports include the use of nasogastric feeds while advancing to full enteral feeding pre-operatively, whereas others advocate only oral feeding without an emphasis to reach full goals of feeding.Reference Stieh, Fischer and Scheewe33 In a study of a heterogeneous group of 67 neonates undergoing surgery within the first year of life, 62 had feeding initiated pre-operatively, with 47 achieving 100 millilitres per kilogram per day and 22 reaching full feeds.Reference Natarajan, Reddy and Aggarwal34 Ductal-dependent lesions were diagnosed in 52 of these 67 patients. This study reported a low incidence of gastrointestinal-related morbidities, with necrotising enterocolitis diagnosed in two infants; however, only 60.7% of patients were discharged home without gavage feeds despite early pre-operative initiation of enteral feeding. This study made the following conclusion:
The majority of infants with congenital heart disease achieve moderate enteral intake prior to surgery, even while on prostaglandins. Despite this and the early initiation of postoperative enteral feeds, many infants need gavage feeds at discharge. Evidence-based feeding strategies for this high-risk population are critical to improving outcomes.
Conclusions
Patients with hypoplastic left heart syndrome continue to experience significant interstage mortality, which may be improved or eliminated by implementation of a structured programme of interstage home monitoring. Ensuring normal growth and nutrition has become a cornerstone of interstage management. Using an aggressive nutritional plan in the interstage period can result in normal patterns of growth, which is associated with improved outcomes.
