Outcomes of a Multisite, Prospective, Randomized Clinical Trial Comparing Two Different Types of Shunts Used in the Norwood Procedure for Infants with Single (right) Ventricles
N Engl J Med 2010; 362(21):1980-92,
Reviewed by Suanne M.Daves, MD and Brian S. Donahue, MD, PhD
Vanderbilt Monroe Caroll Children’s Hospital
The article describes outcomes of a multisite, prospective, randomized clinical trial comparing two different types of shunts used in the Norwood procedure for infants with single (right) ventricles. This is the largest clinical trial and the first North American, multi-center, randomized trial comparing surgical therapies in congenital heart disease patients.1 The study, still ongoing, examines the outcomes of infants undergoing the modified Blalock-Taussig shunt (MBTS) versus the right ventricle (RV) to pulmonary artery (PA) shunt. It is conducted by the Pediatric Heart Network (PHN), a collaboration of clinical sites and a data-coordinating center established in 2001 by the National Heart, Lung, and Blood Institute, one of 27 institutes at the National Institutes of Health (NIH). The PHN was established to centrally coordinate multisite studies in congenital and acquired pediatric heart disease, given that the numbers of patients with a particular cardiac malformation at a single institution are often too small to produce meaningful outcomes research and that institutions may not have the resources to coordinate collaborative investigation.The study design, published in 2008, was intended to test the hypothesis that the RV-PA shunt is associated with better outcomes than the MBTS in terms of 12-month transplant free survival.2
Secondary outcomes included morbidity after the Norwood procedure (open sternum, duration of ventilation, ICU duration, etc) and stage II hospitalizations; the incidence of unintended cardiovascular interventional procedures on the shunt, pulmonary arteries, or neoaorta by 12 months; RV function and volume (by echocardiographic assessment), and degree of tricuspid valve regurgitation at three periods. The periods were 1) at discharge after the Norwood procedure, 2) before state II, and 3) at the age of 14 months. Prespecified subgroups included presence of aortic atresia, obstructed pulmonary venous return, birth weight, degree of tricuspid regurgitation before the Norwood procedure, deep hypothermic circulatory arrest versus regional cerebral perfusion, surgeon’s annual experience in performing Norwood procedures, and the annual volume of Norwood procedures at each center.
The investigators analyzed data from 15 participating North American centers, where 554 infants were randomized shortly after birth to undergo Stage 1 palliation with either a MBTS or an RV-PA shunt. All infants were followed for 14 months. Three analyses were performed; an initial analysis with intention-to-treat, an analysis according to the actual shunt used, and an analysis excluding two infants found at surgery to have anatomy that precluded the randomly assigned shunt. All analysis yielded similar results. Transplant-free survival at 12 months was 74% with the RV-PA shunt and 64% with the MBTS (P=0.01). However, in the set of patients in whom follow-up extended beyond 12 months, the difference in transplant-free survival was no longer significant between groups, primarily because of significantly greater early mortality in the MBTS group. Most deaths occurred from 30 days following the Norwood procedure until the time of the Stage II procedure.
With regard to secondary outcomes, the RV-PA shunt did result in more unplanned interventions and complications over the 12-month period. Subgroup analysis did not demonstrate significant interaction with shunt type although a preliminary analysis suggested that case volume at a specific practice site may lessen the advantage of the RV-PA shunt. Additionally, data suggested that infants with MBTS may have had more cardiac arrests during the Norwood hospitalization.
The strength of this study lies in the fact that it was done at all. This was a significant undertaking and in the words of Carolyn Bondy, MD, of the National Institute of Child Health and Development, it was ‘”heroic”.3 The study was powered to detect differences in the combined primary endpoint but may not have been adequately powered to detect differences in secondary endpoints or among the predetermined subgroups and the authors acknowledge this. The authors acknowledge expected variation in surgical technique (between institutions, surgeons, and over time), perfusion strategies, and perioperative care within each shunt group as potential limitations, but no information about the degree of standardization or clinical protocols is shared with the reader. Also, details of case distribution among the 15 sites is lacking, which may limit applicability of the findings to other centers. The care of the patient with HLHS occurs in complex, non-linear systems but further subgroup analysis may be enlightening. Investigators are continuing to follow this cohort and it remains to be determined whether there is a difference in shunt types in the long run. (study consent includes permission to follow patients up to 5 years-of-age).2
- US Department of Health and Human Services NIH News. Newer heart surgery for infants offers first-year survival benefit over traditional procedure. In: NHLBI; May 26, 2010.
- Ohye RG, Gaynor JW, Ghanayem NS, et al. Design and rationale of a randomized trial comparing the Blalock-Taussig and right ventricle-pulmonary artery shunts in the Norwood procedure. J Thorac Cardiovasc Surg. 136(4):968-75,2008
- Medpage Today. Shunt type matters for Norwood procedure survival. C. Phend, University of Pennsylvania School of Medicine; 2010.
