Multicenter Clinical Research in Congenital Heart Disease: Leveraging Research Networks to Investigate Important Unanswered Questions

Abstract

Congenital heart disease (CHD) is the most common birth defect in the United States. Neonates with CHD are often cared for by neonatologists in addition to cardiologists. However, there is a paucity of rigorous evidence and limited clinical trials regarding the management of neonates with CHD. In this review, we will describe some of the challenges of research in this field. The Pediatric Heart Network serves as an example of how a research network can effectively overcome barriers to conduct and execute well-designed multicenter studies.

INTRODUCTION

Congenital heart disease (CHD) is the most common birth defect, affecting an estimated minimum of 40,000 infants each year in the United States, approximately 25% of whom (~1.4 per 1,000 live births) will require invasive treatment in the first year after birth. CHD is the most common cause of birth defect–related mortality. (1) Maternal risk factors for CHD include smoking, drinking, obesity, diabetes, folate deficiency, and various infections such as rubella and chlamydia. (1) Pulse oximetry screening for critical CHD was instituted as part of the newborn screening panel in 2011, and by 2018, it was operationalized in all 50 states and the District of Columbia. (2) Initial targets for screening were for the following 7 CHDs: hypoplastic left heart syndrome (HLHS), pulmonary atresia, tetralogy of Fallot, total anomalous pulmonary venous return, transposition of the great arteries, tricuspid atresia, and truncus arteriosus. This was then expanded to 12 key CHDs targeted for screening. Since then, other definitions for critical CHD have emerged; some experts define it as the need for intervention in the first year after birth, and others use this term only for ductal-dependent conditions. Between 2007 and 2013, in states that had pulse oximetry screening policies in place, screening was associated with a significant decrease in infant cardiac deaths compared to states without such policies. (3)

Given how common CHD is among infants, it is likely that neonatologists will assume care for children with CHD in the NICU. In fact, severe CHD has been noted more commonly in preterm infants (25–32 weeks’ gestation) compared with term infants in 1 study that used administrative data from the Kids’ Inpatient Databases (2003–2012). (4) Severe CHD in this study included conditions such as truncus arteriosus, tetralogy of Fallot, total anomalous pulmonary venous return, HLHS, coarctation, interrupted aortic arch, pulmonary atresia, tricuspid atresia, Ebstein anomaly, aortic stenosis, atrioventricular canal defects, transposition of the great arteries, and double outlet right ventricle. (4) The birth prevalence of CHD was 116 in 1,000 preterm births (25–32 weeks’ gestation), and severe CHD was seen more commonly in preterm infants (25–32 weeks’ gestation) compared with term infants (7.4/1,000 very/extremely preterm births versus 1.5/1,000 term births, P<.001). (4) Another study demonstrated that being born early term (37–38 weeks’ gestation) was associated with worse postoperative outcomes after neonatal cardiac surgery when compared to those born after 39 weeks. (5) Preterm birth is a known risk factor for both in-hospital and postoperative mortality in infants with CHD. (6)(7) We therefore need to expand the evidence base for the care of such fragile preterm infants with CHD through research efforts.

Clinical trials are essential in testing the safety and efficacy of new therapies to make progress in medicine. However, neonatal and CHD clinical trials often face significant challenges, many of which are similar. Given the rarity of some neonatal conditions, there is often a limited number of potential research participants at any 1 center. Unfortunately, this has led to a paucity of evidence on treatments and management for neonates. (8) The Best Pharmaceuticals for Children Act, enacted in 2002, has the goal of improving the safety and efficacy of medication use and dosage for children through rigorous clinical studies in the pediatric population. (9) However, drug trials are still particularly scarce in neonatology, leading to the widespread use of unlicensed or off-label medications. (10) These same dilemmas are faced by the CHD research community. (11)(12)(13)

Under federal regulations, all children, including neonates, are vulnerable and require extra protection in the conduct of research. (14) Neonates can neither consent nor assent to research participation; therefore, consent must be obtained from the parents or legal guardian. (10)(11) Due to the circumstances of neonatal research, parental consent is typically sought during stressful times when the neonate’s health and parents themselves may be vulnerable. (15) Parents often weigh the potential risks associated with participating in a trial with the potential direct benefit to their child, as well as the benefit of contributing to research knowledge. (15) Beyond the informed consent process, additional protections are provided through a review of protocols and progress by Institutional Review Boards and Data and Safety Monitoring Boards, also known as Data Monitoring Committees. These groups monitor participant safety, study progress, data quality, and overall trial performance.

Clinical trials in CHD face unique challenges. First, there is marked heterogeneity in types of CHD, which creates obstacles to adequately powering clinical trials at a single site. Multicenter collaboration is required to enroll enough participants to ensure robust sample sizes and power. Second, many patients with CHDs have (fortunately) low mortality rates; therefore, we rely on surrogate outcome measures to determine the efficacy of a treatment or intervention. Many outcome measures have not been validated in this unique population. (13)(16) Third, CHD trials often involve surgical and catheter-based interventions, which may be confounded by variations in institutional practice preferences and operator skills. (16) Concerns have also been raised that the expertise of cardiac centers and clinicians can significantly influence the treatment effect in trials involving interventions. (16) While blinded clinical trials are the gold standard, masking participants and study personnel to the intervention is often difficult or impossible. Additional steps are therefore required to ensure masked evaluation of trial endpoints by an independent third party, such as a core laboratory. Lastly, pharmacologic clinical trials in CHD have their own challenges, including the lack of availability of liquid formulations and the need to understand pharmacodynamics and pharmacokinetics over a potentially wide range of patients’ ages, weights, and physiologic states. (16)

RESEARCH NETWORKS

Many neonatologists are familiar with the Neonatal Research Network (NRN), which was established in 1986 by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD). The goal of the NRN is to conduct rigorous studies to improve the management and health outcomes of critically ill new-borns. (8)(17) However, neonatologists may be less familiar with the Pediatric Heart Network (PHN), which also conducts multicenter studies on neonates, as well as older children and adults with CHD. Due to a similar need for well-designed multicenter studies, the National Heart, Lung, and Blood Institute (NHLBI) established the PHN in 2001. The PHN is composed of 9 main clinical sites, a data coordinating center (DCC), auxiliary sites, and NHLBI. (12)

The PHN and NRN are both examples of networks conducting large multicenter studies that have influenced clinical practice. The PHN’s clinical trials on single-ventricle patients span from the neonatal period to adulthood. The PHN has also completed clinical trials in individuals with Marfan disease, Kawasaki disease, and other pediatric heart diseases. Several ongoing PHN clinical trials in the neonatal realm will be discussed here.

RESEARCH NETWORK ADVANTAGES AND CHALLENGES

Establishing a research network has a number of advantages. A research network allows for multicenter involvement and collaboration. With many rare neonatal conditions and CHD, each individual center often has a limited number of potential participants for any given study. Multicenter collaboration leads to an increased number of potential participants in studies, which is especially valuable with rare diseases. (8) Research networks can also leverage the expertise of investigators and research coordinators from multiple centers. (8)(12) Investigators and research coordinators from different centers can share ideas and lessons learned. In addition, there is a separate DCC that provides support for the many logistics involved in running a multicenter clinical trial. (8) Networks provide a fertile ground for training the next generation of investigators in a nurturing environment.

Networks provide a level of efficiency that does not exist for single trials in that new infrastructure does not need to be created for each successive study. This was particularly valuable during the COVID-19 pandemic, when there was a critical need to launch research studies very quickly to understand the disease process and develop diagnostic and therapeutic strategies. Many networks funded by the National Institutes of Health (NIH) were involved in this effort. To better understand how COVID-19 specifically affects children, the NIH launched an effort called the Collaboration to Assess Risk and Identify LoNG-term outcomes for Children with COVID (CARING for Children with COVID). This program, led by NHLBI and NICHD in collaboration with the National Institute of Allergy and Infectious Disease, used existing infrastructure from the PHN and the Pediatric Trials Network to develop new research protocols and studies. (18) In pediatric cardiology, the PHN was able to enroll 600 children in a study of the multisystem inflammatory syndrome in children associated with COVID-19 in just 2 months and to complete enrollment of more than 1,200 children in slightly more than a year.

Through having the research infrastructure in place, the PHN has been seen as favorable for industry partnerships and partnerships across institutes at NIH. Pharmaceutical companies have collaborated with the network to investigate drugs in CHD, with the goal of receiving pediatric labeling in the future. (12)(19) Collaborating with industry has allowed the PHN to expand its reach to involve centers outside the United States and enhance representation and generalizability of the results. Challenges of such collaborations include aligning academic, government, and industry timelines and infrastructures, and developing data-sharing processes that allow industry to meet their regulatory and business obligations yet also allow investigators to pursue secondary analyses beyond the main results of the trial.

Nurses have played an essential role in the PHN, as well as in other research networks. Nurses in the NICUs and cardiac intensive care units (CICUs) often have the most consistent day-to-day contact with families and form unique bonds with them. For a clinical trial to be successful, it is important to have buy-in from the entire NICU and/or CICU team, especially the bedside nursing staff. Families often seek guidance from their bedside nurse about participating in research. Recognizing the importance of bedside nursing staff, some PHN trials have provided trial-specific education to nursing staff. In addition, nurses are involved in study protocol development, providing invaluable feedback during the initial stages of setting up a clinical trial. The PHN also currently has a Nursing Research Committee and an ongoing nurse-led trial, Training in Exercise Activities and Motion for Growth (TEAM 4 Growth), which will be discussed in more detail below.

Research networks are not immune to experiencing challenges. First, networks have to establish integrated infrastructure across centers to be able to collaborate, including contracts, data sharing agreements, investigator and study coordinator support, data management systems, project management and statistical support, and policies governing publications and presentations. Overcoming regulatory hurdles, including Institutional Review Board approval and subcontracts, is another challenge, but one that the networks are familiar with and able to help centers navigate. (12) Since 2018, NIH has required that multicenter studies use a single Institutional Review Board to serve as the Institutional Review Board of record for domestic sites, with the goal of enhancing and streamlining the review process by eliminating duplicative review without diminishing protections of human subjects. The expectation is that all centers will rely on the single institutional review board for ethical review of the study. While this idea has merit, in practice, many centers have not demonstrated a willingness to fully rely on the single Institutional Review Board and instead are performing duplicative reviews of study protocols, which leads to more inefficiency. Variations across centers in practice style, policies, or guidelines may vary and can pose obstacles in developing study protocols. (8)(12) In addition, at different institutions, there may be differences in opinions of equipoise, that is, uncertainty on the relative value of approaches being compared in a trial. Each investigator may have their own opinion about which approach they prefer—an opinion likely to have been influenced by their own experience with similar patients or by their own interpretation of the literature. Yet, as long as there is uncertainty among experts in the field and in the clinical community (termed clinical equipoise), it remains ethical and important to randomize participants to either approach and to try and prove or disprove our own therapeutic biases and dogmas. (20)

The field of pediatric cardiology has seen significant improvement in outcomes of those living with CHD. Unfortunately, there remain significant racial, ethnic, and socioeconomic disparities. (21)(22) Racial and ethnic minorities are under-represented in clinical trials, including in cardiovascular clinical trials. (23) Previous studies have identified mistrust or lack of comfort with the clinical trial process, language barriers, time and resource constraints, lack of information about clinical trials, and stigma associated with clinical trials to all be barriers to trial recruitment. (23)(24) Through working across many different clinical sites with diverse research teams, the PHN strives to overcome these barriers to research participation. The next funding cycle of the PHN places a particular emphasis on diversity, equity, and inclusion, and is requesting that sites propose a plan to increase racial and ethnic diversity in PHN studies through community participation strategies. (25)(26)

CLINICAL TRIALS

Several ongoing clinical trials in neonates and infants with CHD are being conducted through the PHN. We will discuss these studies to highlight how leveraging multicenter collaboration through a research network can work to answer some of the challenging questions facing the fields of pediatric cardiology and neonatology (Table).

Table.

Three Ongoing Pediatric Heart Network Trials in Neonates and Infants with Congenital Heart Disease

Clinical Trial	Team 4 Growth	Use of Oxandrolone to Promote Growth in Infants with HLHS	COMPASS Trial
Population of Interest	Hypoplastic heart syndrome or other single right ventricle anomalies following Norwood procedure	Hypoplastic heart syndrome or other single right ventricle anomalies following Norwood procedure	Neonates with ductal-dependent pulmonary blood flow
Intervention	Passive range of motion exercise program	Oxandrolone	Ductal artery stent
Comparison	Standard of care	No oxandrolone	Blalock-Taussig-Thomas shunt
Outcomes	Growth indices	Safety/tolerability, optimal dosing, preliminary efficacy on growth indices	Serious adverse events (i.e., death, ECMO, unplanned interventions), days alive outside of hospital, length of stay, pulmonary artery growth and inpatient costs

COMPASS=Comparison of Methods of Pulmonary blood flow Augmentation in neonates: Shunt versus Stent, ECMO=extracorporeal membrane oxygenation, HLHS=hypoplastic left heart syndrome

Training in Exercise Activities and Motion for Growth (TEAM 4 Growth)

It is well known that infants with CHD have impaired growth. Infants with single-ventricle physiology are particularly at risk, which is thought to be secondary to inadequate nutrition, hypoxemia, complications from surgery, and higher energy expenditure. (27)(28) The neonate is at greater risk for malnutrition from the prolonged metabolic response to stress occurring with surgery. (28)(29) Given the many variables involved, it is not surprising that a focus on adequate nutrition alone does not ensure adequate weight gain after Norwood operation for single-ventricle heart disease (30)(31)

Previous NICU studies on passive range of motion therapy in healthy preterm infants have demonstrated increased bone mineralization and enhanced weight gain. (32)(33) However, no previous research had evaluated passive range of motion in infants with CHD.

The initial pilot study conducted by the PHN, TEAM 4 Growth, enrolled 20 infants following stage I single-ventricle palliation with the Norwood procedure. This study found a passive range of motion exercise program to be safe and feasible in this high-risk population. (30)

Currently, the PHN is conducting a phase III randomized trial of a passive range of motion exercise program for infants with HLHS or other single right ventricle anomalies following the Norwood procedure. (34) TEAM 4 Growth is the first nurse-led clinical trial in the PHN. An ancillary study will also be conducted evaluating families’ perceptions of participating in research. This will help inform the design for future research studies, and illustrates another flexibility of networks, which is that pilot studies can be performed to provide data to design the definitive trial. Previous PHN studies evaluating the experiences of parents and families of infants with HLHS demonstrated that quality of life, well-being, and family resources were significantly lower than adult norms. (35)(36) Family characteristics and individual traits, such as anxiety and coping skills, were the strongest predictors of outcomes. (35) The TEAM 4 Growth Perceptions Ancillary Study will build on this previous work to better understand family considerations in research participation. (35)

Use of Oxandrolone to Promote Growth in Infants with HLHS

The PHN recognized that there are likely several avenues to address the problem of growth failure in the single-ventricle population. In addition to evaluating passive range of motion therapy, the PHN recently started enrollment for a study on oxandrolone’s effect on growth. Oxandrolone is an anabolic steroid used in children to prevent short stature in Turner syndrome and Duchenne muscular dystrophy and to prevent malnutrition associated with severe burns. (37)(38)(39)(40) Comparisons have been made between the inflammatory cascade from cardiopulmonary bypass and the inflammatory response from severe burn injury. (40)

This is a phase I/II randomized trial of oxandrolone versus no oxandrolone to assess the safety/tolerability, optimal dosing, and preliminary efficacy of this drug in neonates with HLHS or other single right ventricle anomalies after undergoing the Norwood procedure. (41) The primary aim is to evaluate if oxandrolone is safe and tolerable in this population. The secondary aim is to investigate the efficacy of oxandrolone in improving growth and nutrition indices in these patients. (41)

This clinical trial was developed based on the preliminary results of a previous single-center study of oxandrolone. (40) This study of 13 participants post–Norwood procedure had found no evidence of serious adverse events, changes in safety laboratories, or evidence of virilization in those receiving oxandrolone therapy. A significant difference in change in weight-for-age z score was seen among the participants, with those receiving oxandrolone in medium-chain triglyceride oil preparation having the lowest decline during the study period. (40)

Similar to the TEAM 4 Growth clinical trial, an ancillary study evaluating families’ perceptions of participating in research will also be conducted for this clinical trial.

Comparison of Methods of Pulmonary Blood Flow Augmentation in Neonates: Shunt versus Stent

Neonates with ductal-dependent pulmonary blood flow (pulmonary blood flow supplied by the systemic circulation via the patent ductus arteriosus) were traditionally initially palliated with a systemic-to-pulmonary artery shunt, also known as the Blalock-Taussig-Thomas shunt. In more recent years, an alternative option for establishing a stable source of pulmonary blood flow has arisen. Ductal artery stenting is a cardiac catheterization procedure where a stent is placed in the patent ductus arteriosus to achieve the same result as a surgical shunt. However, data comparing the outcomes of these 2 interventions remain limited, and all studies up to this point have been retrospective. (42)(43) Given the lack of prospective, multicenter, randomized trials, uncertainty remains regarding which intervention has better outcomes, which was the impetus for designing Comparison of Methods of Pulmonary blood flow Augmentation in neonates: Shunt versus Stent (COMPASS).

The COMPASS trial is a multicenter, randomized interventional trial for neonates with ductal-dependent pulmonary blood flow. Participants will be randomized to receive either a ductal artery stent or a systemic-to-pulmonary artery shunt, and will be followed through the first year for serious adverse events such as death, extracorporeal membrane oxygenation, mechanical circulatory support, or unplanned interventions. The study will also compare days alive out of the hospital, growth of the pulmonary arteries, length of stay, and inpatient costs between treatment arms. (44) This trial is also unique in leveraging existing data from 2 established CHD registries, the Pediatric Cardiac Critical Care Consortium and the Congenital Cardiac Research Collaborative, for data collection. The COMPASS trial will harvest many of its data elements from existing clinical registries at each center rather than having study coordinators manually enter the same data onto study case report forms. This model of leveraging registries for clinical research will hopefully improve efficiency and reduce errors in data collection for research.

CONCLUSIONS

Multicenter clinical research in the neonatal and CHD populations continues to pose challenges. Research networks, such as the PHN and NRN, serve as examples of how collaboration can overcome many of these obstacles. In the future, the PHN and NRN may benefit from sharing strategies and best practices in clinical trials and for collaborative studies examining outcomes of mutual interest, such as neurodevelopment. However, for those involved in research, there will always be new questions and a desire to further improve outcomes.

As neonates with CHD now live into adulthood, there is still much to be understood about the neurodevelopmental and quality-of-life outcomes. (45) The research interests of neonatologists and cardiologists in these important areas align well and may foster fruitful collaborations. In addition, while there has been an overall improvement in mortality in the CHD population over time, there remain significant disparities in health outcomes by race, ethnicity, and socioeconomic status. (21)(22) Including diverse participants and perspectives in clinical research studies will continue to be important in addressing these health inequities. (45) Lastly, research efforts must continue to evolve to meet the research needs of the patients and families by harnessing the tools of data science and precision medicine to improve outcomes. (45)

EDUCATION GAPS.

Clinical trials are essential in testing the safety and efficacy of new therapies to improve outcomes in the neonatal population with congenital heart disease. However, researchers in this area often face significant challenges, which are important to recognize when planning research studies.

OBJECTIVES After completing this article, reader should be able to:

1. Identify challenges faced by researchers in neonatal and congenital heart disease research.

2. Describe some of the benefits and challenges of research networks.

3. Describe a few current clinical trials in the neonatal population with congenital heart disease.

American Board of Pediatrics Neonatal-Perinatal Content Specifications.

Identify the study design most likely to yield valid information about the benefits and/or harms of an intervention.

ABBREVIATIONS

CHD

congenital heart disease

CICU

cardiac intensive care unit

COMPASS

Comparison of Methods of Pulmonary blood flow Augmentation in neonates: Shunt versus Stent

DCC

data coordinating center

HLHS

hypoplastic left heart syndrome

NHLBI

National Heart, Lung, and Blood Institute

NICHD

Eunice Kennedy Shriver National Institute of Child Health and Human Development

NIH

National Institutes of Health

NRN

Neonatal Research Network

PHN

Pediatric Heart Network

TEAM 4 Growth

Training in Exercise Activities and Motion for Growth