Discussion| Volume 145, ISSUE 4, P1025-1027, April 2013


    Open ArchivePublished:January 31, 2013DOI:
        Dr V. Mohan Reddy (Stanford, Calif). This study primarily sought to evaluate the effect of a strategy of repair on late functional outcomes and exercise capacity, especially to see if the biventricular strategy was superior to the single-ventricle or 1.5-ventricle repair strategy. Despite the limitation that only 39% of the survivors participated in the study and only 63% to 77% completed all the tests, the study provides important inferences and is definitely a major contribution to the field.
        The major strength in these 3 groups—namely, the biventricular, 1.5-ventricle, and single-ventricle groups—is that they are diagnosis–concordant and that the initial strategy of surgical repair is based on sound principles. It also appears that forcing biventricular repair in patients with lower z-scores reflects our obsession with 2-ventricle repairs, and perhaps this data will compel us to rethink that strategy. Having said that, I have several questions. First, were there enough patients in each group who completed all the tests? If so, did you analyze them separately?
        Dr Karamlou. That's a very important question, as was the initial question, which was to determine if there is a tricuspid valve cutpoint that can guide all of our clinical decision making. Unfortunately, we were not able to determine a definitive cutpoint based on this study. I can tell you that if you look at the data and you segregate things, it looks like a tricuspid valve z-score of about -2 would give you a statistically significant difference between a univentricular repair on one hand and a biventricular repair on the other hand. But unfortunately, when you look at this as a continuum, we couldn't determine the threshold that you're seeking.
        Dr Reddy. Thank you. The second question is: Within the biventricular group is there a cutoff of initial tricuspid valve z-score that predicted the functional capacity and the exercise capacity?
        Dr Karamlou. That's an important question that we sought to answer and determine if there is a cutpoint that can guide all of our clinical decision making. Unfortunately, we were not able to determine a definitive cutpoint based on this study. I can tell you that if you look at the data and you segregate things, it looks like a tricuspid valve Z-score of about minus 2 would give you a statistically significant difference between a univentricular repair on one hand and a biventricular repair on the other hand. But unfortunately, when you look at this as a continuum, we couldn't determine the threshold that you're seeking.
        Dr Reddy. Within the biventricular group did you see if presence or absence of atrial septal defect influenced functional and exercise capacity?
        Dr Karamlou. Yes. Unfortunately, in this particular study we did not have echocardiographic information that was concomitant with all of the functional health status testing. That's something we hope to achieve in the future. So we didn't with absolute certainty know the status of the atrial septum. Now, if we knew it to be closed based on the yearly follow-up that the data center maintains on this cohort, we included that. But we cannot be certain that we know the status of the atrial septum contemporaneous with our evaluation.
        Dr Reddy. That may be an important factor to see if having an atrial septal defect would increase enough cardiac output vis-à-vis the cavopulmonary shunt in 1.5-ventricle repair patients.
        Dr Karamlou. Yes, absolutely, I agree.
        Dr Reddy. The next question to you is: How do you explain that the initial tricuspid valve size also correlated with peak oxygen uptake (VO2) even in patients with a single ventricle?
        Dr Karamlou. We were perplexed initially by that relationship. But having looked at this retrospectively as well among our other cohorts, we saw essentially the same finding in the patients with critical aortic stenosis and aortic valve atresia, whereby the status of the mitral valve was influential in outcome even within patients who were triaged to a single-ventricle repair. So exactly the converse on the other side was true in that cohort.
        Additionally, there are a couple of other things that one can speculate. One is that among other types of diagnoses there is ventricular interdependence and there is a ventricular interaction, so it could be that patients along the spectrum in the univentricular group with larger right ventricles—and therefore a large right ventricular mass—have some interaction with the systemic cardiac output, and that may influence their aerobic capacity. Second, perhaps the tricuspid valve z-score is surrogating other unfavorable or favorable anatomic characteristics like pulmonary arborization that unfortunately we could not look at.
        Dr Reddy. I encourage you to add a paragraph to that effect in your discussion.
        Dr Karamlou. Yes, that's a good point.
        Dr Reddy. The next question: Peak VO2 was higher in 1.5-ventricle repair compared with biventricular repair regardless of the z-score, which suggests that tricuspid valve is not the only factor but right ventricle end-diastolic pressures or tricuspid insufficiency may be important and a cavopulmonary shunt might offset these factors. So my question based on this is: Should we consider the 1.5-ventricle option for patients with pulmonary atresia with intact ventricular septum, for all patients pulmonary atresia with intact ventricular septum, regardless of the initial tricuspid valve size?
        Dr Karamlou. Yes, I think that question is anticipated based on these data, and I don't think what we're trying to convey is that for every anatomic subtype you should always do a 1.5-ventricle repair. Again, we have to be circumspect. This was a small group of patients—only 14 in the 1.5–ventricle-repair group—and although statistically significant, the range of tricuspid valve z-scores on either side of those lines is small. There is a physiologic correlate to our findings in that patients with a 1.5-ventricle strategy may be able to augment their cardiac output and therefore improve their aerobic capacity. But again, it's a small group of patients and warrants further study.
        Dr Reddy. Do you plan to investigate any neurodevelopment-related outcomes in these 1.5-ventricle group patients that might suggest that to increase the pressure in the superior vena cava may have some implications in developmental outcomes? That might offset recommending 1.5-ventricle repairs as opposed to just relying on peak VO2.
        Dr Karamlou. I only presented the peak VO2 data in the interest of time. There are obviously comprehensive exercise variables that we plan to look at in a separate manuscript.
        In terms of the neurodevelopment-related outcomes, although we could look at these data cross-sectionally, all would be confounded by not knowing what the patients started with initially. We know from data from Children's Hospital of Philadelphia that these patients start out with very different neurologic phenotypes that probably influence outcome. So although we could look at it late, I'm not sure we could make a strong conclusion.
        Dr Reddy. It might simply show if there is any effect of increased superior vena cava pressure.
        Dr Karamlou. That's true.
        Dr Reddy. My last question is: We all know that tricuspid valve size changes as patients grow, depending upon various factors—whether or not there is the presence of ASD, because that might influence the amount of blood going across the tricuspid valve; what the right ventricle end-diastolic pressure is; and what the right ventricular function is. So if the tricuspid valve size is definitely influenced by growth, did you make any attempt to compare the exercise capacity outcomes vis-à-vis the current tricuspid valve z-score rather than the initial tricuspid valve z-score?
        Dr Karamlou. Unfortunately we don't have the echocardiographic data contemporaneous with the exercise testing. So the thought of the current investigation was to correlate late functional health status with initial morphology. We don't really know what the tricuspid valve z-score in all patients is contemporaneous with the functional status or the exercise testing.
        Dr Reddy. I think it's very relevant because it's important to know what the current size of the tricuspid valve is. And how it has grown compared to the original size.
        Dr Karamlou. Yes, absolutely. Another point to mention is that patients on a Fontan pathway may actually have a different trajectory than biventricular repair patients. Their course isn't static, and so getting the echo and other ancillary data about how the patients are doing clinically at the time of their functional status evaluation would give us additional context in which to interpret these results.
        Dr John Foker (Minneapolis, Minn). It's important to have all the variables when you design a study. The main issues in these patients are right ventricular and tricuspid hypoplasia. So the important question and variable is the possibility of growth. It does not seem logical to decide to do a single or a biventricular repair if catch-up growth will erase the hypoplasia.
        My question was going to be about growth; however, you have already discussed that you did not analyze how well the biventricular repairs grew or the difference in tricuspid valve scores with time. My group’s previous research found that even a -4 tricuspid valve will grow up to normal size if you achieve total relief of right ventricular outflow obstruction, a restrictive atrial septal defect to encourage tricuspid valve flow yet maintain an adequate cardiac output, and ligation of all the important coronary artery fistulas or connections. My recommendation would be to have a growth arm in these studies in the future.
        My final point is that some children with biventricular repairs will start to miss their pulmonary valve at about age 15 years. They may have reduced exercise capacity until somebody screws in a pulmonary valve. With a pulmonary valve in place these patient should have improved exercise capacity.
        Dr Karamlou. Growth observation was actually in our initial grant application and for pragmatic purposes we chose to present the data we have. Certainly serial echocardiographic data are a critical component when studying this patient population.
        Dr Charles Fraser (Houston, Tex). Would you speculate a little bit more about the subjective findings in these patients? To me that's the most encouraging thing that you presented. Although we want to make decisions using objective scientific data, if patients perceive themselves as doing well that's a very good thing. And it's fascinating.
        I'll share a quick anecdote if I might. I just interviewed last week a 77-year-old woman who was operated on by Dr Blalock at age 5 years. She lived with a Blalock shunt for 50 years. She's had a wonderful life. She says “I love doctors.” Her quality of life, from her perspective, was outstanding; from ours, she looked very disabled. What do you think about that and what can we learn from that going forward?
        Dr Karamlou. It is very encouraging and important that there is a discrepancy between the subjective and the objective measures. Such a dichotomy has been seen in other studies; patients with congenital heart disease often believe they're doing better than control subjects. And it may be that there are multiple environmental factors. For example, it may be that people pay more attention to these patients as children. Psychosocial and emotional functioning is something we don't really understand and it's probably much more complex than putting somebody on a treadmill. Or, it may be that domains like bodily pain are calibrated differently. If you've experienced multiple operations, well, your idea of severe pain is probably different than somebody who has not undergone surgical procedures. It is fodder for additional studies.