Anita J. Moon-Grady, MD, FAAP, director of the UCSF Fetal Cardiovascular Program, examines the reality of whether long-term outcomes are better for babies who had in utero interventions for certain heart defects, as well as the short-term complications for mothers. She draws on trial data and UCSF’s own registry, started more than 10 years ago, to illuminate complex treatment decisions, including ethical aspects.
Refer to Fetal Treatment Center
Welcome. Um This is Mary Donofrio. I'm directed the fetal heart program. Here are Children National and it is my extreme pleasure to introduce um perhaps one of the smartest people I know for sure, and definitely most dedicated. Um Anita Moon Grady, uh is a professor of pediatrics uh and director of the fetal cardiovascular program at UCSF. And she's going to tell us about fetal cardiac interventions. All right, well, thank you, mary. Thanks for inviting me to to speak. I am uh a little bit more comfortable in front of a big audience than on a webinar, believe it or not. So I'll do my best. But this is a little bit new form for me. So I was asked to talk about fetal cardiac intervention uh in particular how it may change outcomes and congenital heart disease. And as I thought about that a little bit more, um I uh wanted to also bring up some questions about change versus improvement. So, uh we will talk about it in that context. Now this is sort of the classic debate that we have in fetal cardiology. Um it's always entertaining to put people on a stage and have the debate the pros and cons of fetal intervention. Um and the usual pro is it prevents hypoplastic left heart syndrome. The con it doesn't work And I think it's time to really get get over this argument. It's over 20 years old and we have amassed a lot of data and so that's what I want to present today. Um 1st, then this concept of improving versus changing outcomes and who we are trying to improve outcomes for. And then I will present some fetal cardiac intervention, current applications and published data and uh all along the way perform a critical appraisal of this, of this data and where we need to go next. So first this defining of of change and improvement. Um and before we can even define improvement, we need to define who we are trying to improve things for. Is it for the patient? And if so who is the patient? Are we trying to improve things for the pregnant woman, for the woman herself or for for the pregnancy or for the baby? And we get into issues ethical issues of autonomy, autonomy versus selflessness. That as a prenatal care providers we deal with every day. But this certainly is a serious ethical issue for the fetus. Maybe we can improve things for the fetus or for the child. And keeping in mind that defining improvement for the child takes into account the context of the health care delivery system into which they are being born, the family into which they are being born and their resources. And so that also brings up this issue of who is the patient? Is it the family um and uh his improvement or change for them? Something that we need to discuss as well. So, a very complex ethical issue for for all of us that, as I said, fetal centres deal with this every day. Again. I want to back up though. Why is this even an issue? We have perfectly good treatment for uh for congenital heart disease these days? Right. We have developed single integral palley ations including the single integral affiliation to Fontane, which I have outlined on the following slides. Just for people who don't look at this every day. The Norwood operation involves reconstruction of a neo aorta from the pulmonary artery and whatever is present of the aorta with a right ventricle, the pulmonary artery conduit supplying pulmonary blood flow. Then a second stage. As the child grows at about 3-6 months, the right ventricle pulmonary artery conduit is taken down and uh anastomosis between the Superior Vena Cava and the pulmonary arteries took to provide pulmonary blood flow is performed. Then when they are toddlers 2 to 5 years of age, they'll have uh total cable, pulmonary connection or Fontane completion, in which they still have that bidirectional glenn. But now the inferior vena cava is routed via at our institution and extra cardiac gortex conduit to the pulmonary arteries, completing the Fontane circulation in some places. This is done by a lateral tunnel. So this is a lot of surgery um leaves you with your aorta and the timeline for the single ventricle patients. Uh really can be thought of as extending from the prenatal period onwards to the norwood that they have. And we'll we'll be discharged from the hospital at about a month of age back in the hospital 3 to 6 months for the bidirectional blend fontana, 2 to 5 years, Teenage years, they may start to have an increasing medical problems and as adults, uh the vast majority who are still alive do have some issues. And beyond that, outcomes are a little bit less well defined. So this timeline that starts in the prenatal period doesn't know thoroughly have a a defined uh sort of outcome that we can fix. Um So while there is neonatal therapy for hypoplastic left heart syndrome, it's not really as cut and dried as as that, we don't have a fix for this. In fact, in the single ventricle reconstruction trial, it was noted. This was a randomized controlled trial of of one type of shunt versus another and that's not important for this. What's important is that this is a relatively recent era randomized controlled trial in some of the best surgical centers with the best follow up. Uh And as we all know, patients in uh trials do tend to do better than uh so called historical controls. So the best possible scenario and still at um four years of life, there was only a 60 to 70% survival in uh in this cohort of hypoplastic left heart syndrome patients. So uh so not all good news. There are variants of hypoplastic left heart syndrome with restrictive or intact atrial septum left atrial hypertension in utero. That can make them even more high risk and make their survival much less than the standard H. L. H. S. Patients that I just presented with shown on the right hand side of the screen, muscular wrist, pulmonary vasculature and pulmonary hypertension. This entity of restrictive inter atrial septum complicates the immediate post natal management, makes them very unstable upon separation from the placental circulation. And even if they survived their initial resuscitation, that pulmonary vascular disease may preclude single ventricle palley ation, leaving the infant who survived the immediate post natal period subsequently inoperable. So very uh serious um a co morbidity with hypoplastic left heart syndrome. So, as I've outlined, the single ventricle palley ation isn't exactly transformative. What about pediatric or neonatal heart transplantation? Well, this too has good short term survival, but longer term survival in a cohort of pediatric heart transplant patients. This is from a few years ago. Um but it was only 40 years uh sorry, 40% of 20 years issues with pediatric Ortho topic. Heart transplantation include uh rejection, which is a lifelong issue infection as well. Allah graft, vascular apathy, malignancy and post transplant lymphatic diseases, renal disease. And uh an almost universal need for re transplantation. And so just as an example of one of these complications, the freedom from first episode of rejection is shown in the graph. At 10 years, more than half of patients have had at least one rejection episode. And so to sort of overview or sum up the current available post natal palley ations for hypoplastic left heart syndrome in particular, we have surgical options. Um Short term outcomes are good, but longer term outcomes are less clear. And just to mention a few others in um that are not directly related to surgical survival but but related to the longer term survival. This is a uh great uh hypoplastic left heart syndrome patient. We've had their fontana, extra cardiac conduit. Is there just a little bit of regurgitation? Good function. But here's another patient who's just status post glenn with very poor right ventricular function. So ventricular function is um is something that's sort of an intrinsic problem with the Fontane circulation. And these patients, some of it may stem from prenatal uh a dysfunction that's not very well understood. But as the years go on, ventricular dysfunction becomes more and more of an issue. They also have arrhythmias, protein losing inter apathy, hypoxia, mia, endocarditis, exercise limitations, elevations and PVR thrombin, symbolic complications. The paddock dysfunction and complications from long term cyanosis. So this is um in the patients who survive everything I showed you before, we're just survival curves, but survival with significant morbidity and some there's data on exercise physiology declining with age that I don't want to I'm not going to go into uh there are also uh I don't need to tell this audience issues with neuro development that's probably multi factorial and heart failure and crumble embolism, Ventricular failure at 10 years plus, post-fontanez almost uh universal and uh throw metabolic complications increase at over 15 years. Post Fontane, which sounds like a long time after Fontane. But remember these kids are 2 to 5 years old when they're having these, So you're talking about young adults, looking at going to college, starting to have profound issues with ventricular failure and thrombosis embolism. They also have at this point, problems with arrhythmias and another complication. That is almost unique. Not entirely unique but almost unique to Fontane. Populate Fontane population is excuse me, plastic bronc bronchitis, which is illustrated here these casts of pro tenacious material that deposit in the airways and um this is something that can be absolutely devastating protein losing and are up at the last second to last. Long term complication. Of montane um palley ation Historically, even though it only affects between one and 10% of Fontane patients Has a very dismal prognosis with 50% 5 year survival and a 75. Um I think that's a mistake. 25% transplant free survival and another. The final complication from our phantom population that we're just beginning to recognize as they are surviving longer is fontana associated liver disease, which may necessitate both heart and liver transplant In their 3rd and 4th decade of life. And so um so it is an issue that if we could potentially intervene in fetal life uh to improve outcomes for these patients to change their outcome, at least from hypoplastic left heart syndrome to something that allows them to have a bi ventricular circulation, that maybe they won't have all of these complications. So let's look at the data. Um There are several single center publications out there for foetal aortic stenosis. I'll talk about aortic stenosis and a little bit about atrial septal intervention. I'm not going to talk about pulmonary valve disease. But as far as intervention for aortic stenosis, uh they're single centered data. There are multi center um data from europe and some uh some issues that I'll go over with each one of these studies. But first, just to introduce the concept, if you haven't seen this fetal aortic valvular plastic is now done via per cutaneous access through the maternal abdomen uterus and then a per ventricular approach to the fetal left ventricle. Here's one of our patients, there's the needle going through the uterus and then puncturing through the left ventricle. And then a balloon tipped catheter is threaded through the needle and inflated across the aortic valve. This is uh sort of off the shelf uh balloon catheter for uh pc. And I guess I'm supposed to mention this is not exactly the labelled use of this product, but it is labeled for intravascular use. And so the use of it is allowed. Uh Here's a different patient of ours from UCSF who had a successful intervention. Dilated poorly functioning left ventricle as a fetus and after nine months in a rosco no operation has a normal size left ventricle at least. Although there's some diastolic dysfunction there. So I mentioned that there are single center data. The two biggest centers are Boston and Linz and they've both published extensively on this. The boston group published uh follow up to their initial report. Um this follow up, first authored by Lindsay Freud, who's at Columbia now talked about the outcomes of the 1st 100 patients. of those 1st 100, there were 38 who were by ventricular At the time of publication, 31 who are bi ventricular from birth and 11 procedure related losses. They had almost no cardiac deaths in there by ventricular final circulation and no cardiac deaths if the patients were by ventricular from birth. So that's the graph on the on the right. Uh, lance published their 1st 24 procedures several years ago now, where two thirds of their patients were by ventricular. And in addition, they showed that left heart obstruction with high drops responded favorably. They had a similar fetal loss rate about 10%, and also reported a significant number of patients over half had some sort of uh complication that did not result in death. So treatable braided cardio, pericardial effusion. Uh The newer boston data has also been published, Kevin Freedman just recently published uh separating their experience out to the earlier versus later era with much better success rates than actually getting the balloon inflated across the valve and a much higher. Almost 60% of patients were live born with bi ventricular circulation, probably because of more refined car patient selection. And that was that was shown here graphically. They did a recursive partitioning analysis uh cart analysis to show that the patients with higher left ventricular pressure and um and larger aortas at the time of feel intervention were more likely to have a successful intervention. So it's all looking pretty good so far from that single center data, at least from boston and and Linz. But to look at the sort of counter argument, the association of european paediatric cardiologist, fetal working group um in uh in 2000 sorry in the mid two thousands put together a multi center study. This was a two part study. So so present. First the Natural history uh study where they had 100 and seven fetuses who did not have fetal cardiac intervention. But they went back and retrospectively applied the so called boston criteria to identify ideal candidates. And then they looked at their survival and their final post natal circulation. They found that of the 40 who met the criteria for LV length, retrograde flow across the atrial septum and arch, Um 40 who met criteria, 33% of those were actually able to have a bi ventricular repair without fetal cardiac intervention. And in addition, 42% of the patients who were worse than Boston threshold score presentation were also by ventricular. So a significant number of patients who would if they had undergone fetal cardiac intervention would have been by ventricular, but the Europeans argued they would have been by ventricular anyway. Now, as part of the same study, they then also had patients who had fetal cardiac intervention. There were 59 of these, 40% of those were by ventricular um from birth. And uh the graphically the follow up is shown in uh in the chart on the on the right with survival versus follow up from birth and the fetal valvular plast e patients look like they're surviving better than the natural history cohort. But if you separate out the fetal valvular plastic who had a, you know, ventricular circulation. So a failed, essentially failed procedure versus a successful procedure. It's the, you know, ventricular patients who actually are driving that difference. Um And so, given they also had a 10% loss rate which isn't shown on this curve and a significant prematurity rate, It's questionable whether this procedures actually providing patients with better improved outcomes, although it did seem to confer a survival advantage in the universe ventricular patients, which is sort of curious, so mixed data than from the european group. A lot of other smaller papers have been published. Argentina had most of their procedures successful, but most of their losses were post natally because they are unable to at least at the time of this study, were unable to provide single ventricle palley ation and didn't have great neonatal surgery. So they all died post natally. Um and uh and groups from spain and um and there are groups in South America have also published smaller series. So what about longer term outcomes than the uh boston group again, has published a high rate of re intervention in those 1st 100 patients. The oldest one, I think was 10 years old when they published that with a high prevalence of pulmonary vessel dilator use and neurodevelopmental outcomes have not been shown to be better. So, putting all of this data together, it does seem, even the european group agreed that you're able to change the natural history of disease with a fetal valvular plasticky. But is it actually doing them any long term good? These are the uh not to to read this, but from that paper from boston, looking at long term outcomes, a long laundry list of additional surgical procedures that the patients have undergone. The vast majority of the so called successful patients have had additional surgical procedures, some of them multiple surgical procedures and to to go into why that is. And maybe put a little bit of context on that. I would say that most of these patients are still born with aortic stenosis, even if they aren't born with hyper plastic left heart syndrome. And the treatment for aortic stenosis in infancy is also balloon aortic valvular plasticky. Um, but many, many patients end up having this procedure, which is a ross procedure. It's moving the pulmonary valve which is normal into the aortic valve position. So giving the patient a neo aortic valve and then giving them a right ventricle pulmonary artery conduit. This is a pregnant, pretty big neonatal operation as well. And neonatal rosco no operation. And even balloon valvular plasticky does leave patients with some residual morbidity. And the mortality early mortality is not that high for neonatal a aortic stenosis, but the long term mortality is is significant. This paper from Dolph McElhinney reported 14% early deaths in a neonatal aortic stenosis cohort. And also found that even though trans catheter aortic valvular plastic is effective in congenital A. S there are steady long term hazards for surgical aortic valve, re intervention and replacement. Um And so uh the way that I look at this is that a successful fetal valvular plasticky sort of sets you up for having a patient with aortic stenosis as a neo Nate. And their outcome may be different from those with hypoplastic left heart syndrome, but it's still not clear that it's all that much better. There's a there's a survival Issues and then there are long term issues with pulmonary hypertension in four multiple re operations. So if the two then the fetal intervention versus natural histories, uh each have their their own advantages and disadvantages. One might ask, well, why not do randomized trial to see which ones actually better? Um And I usually do countered this argument with the cost and logistics of doing a large randomized controlled trial for this very rare disease and fetal life. The incidence of this is similar to milo, Meninga seal. And as as some of you may know, the mom's trial, which looked at fetal surgery versus post natal surgery for spina bifida took over Over eight years and something like $6 or $8 million dollars of NIH money um in order to show a difference for fetal surgery. So while it was great for fetal surgery, it was, it was a huge labor of love and I don't see the NIH uh funding another study um like that in the in the fetal realm. So uh so it cost logistics. There also is this issue that there's a variability and post natal approach and care as I showed with the Argentinean group who had successful value plastics but their neo Nate's still died because the availability of care. And then finally we've been at this now at least in Boston since 2001. And there really is a lack of equal poise. There are people who really believe that fetal intervention works and so randomizing patients would be quite an issue. So why then a registry? Well for all the reasons above. Um A lot of other uh areas we are moving towards registries. Uh They're very common in the congenital cardiac world And the there at least less costly and more feasible even though they give you different um different information. So we established in 2010, the international fetal cardiac intervention registry. It's housed here at UCSF. We currently have over 20 centres entering data and the slightest. Sorry the slightest. From last year we now have 648 records. Um the majority of those patients underwent intervention. We asked that all patients who are offered intervention be put into the registry or at least the ones that are being evaluated for intervention. And then we're recording procedural details and pregnancy outcomes. Our first publication was in 2005. At that time, only two centers were still requiring research consent speaking to the lack of equipoise. eight centres considered this to be a clinical care. Um most of the procedures by 2010 were being performed per cutaneous Lee with almost no one doing uh lap academies and no history history Adami's anymore. Most of the procedures are being done with without general anesthesia. There had been an evolution from general anesthesia to either local or spinal epidural. And so this just shows you some things that we can learn from the registry about evolution of the procedure towards better and better safety and less morbidity. At least for the mother. With per cutaneous access and avoiding general anesthesia, fetal cardiac intervention for atrial septal restriction. I haven't mentioned very much, but this is a similar approach, a per cutaneous approach to deploy a stent across the atrial septum. It does have a different goal um of decompressing the left atrium and allowing for um for a more stable neonatal and potentially for giving time for reversal of those pulmonary vascular changes that I showed you, boston, brazil and Toronto have all published small the small series of data. This is um a video of patient fetal patient undergoing stent placement. I'll just let that play for a second. So the stent is actually in the lower right lodged in the atrial septum and there is a colour flow demonstrating um pulmonary venus return to the right atrium and the pre and post pulmonary vein Doppler. These are both done in the operating room, one immediately before and one immediately after the stent deployment. So does seem to uh to be feasible. We've also looked at these patients through our registry approach And trade Johnson when he was in Michigan, published this in circulation last year. At that time, we had 46 fetal cardiac intervention patients, including 19 stents. And the rest of them were were just balloon dilation of the atrial septum. We should have disappointingly no difference in overall mortality, But did show a high procedural success. No difference between Septa plastic in stent. But the stents were more likely to stay open until delivery fetal complications were frequent but not often fatal. And if there was a procedural success in the fight to survive the procedure, there was improved neonatal stability and a trend towards better 30 days uh, survival. What about maternal outcomes? I did start this this talk saying that it's not just all about the fetus and baby. Um, we know from experience uh, with all types of maternal fetal procedures That complications are concerned, but not all that frequent, especially in uh in particular Tania's types of procedures. So this was a paper from 2006 from a single institutions actually from UCSF. of about 200 patients. And procedures where most of the maternal complications were in the history khatemi group, there were no deaths. But a significant number of women had pulmonary oedema and even needed. Icu care, coriander, Iranian, ex separation, uh preterm labor, preterm delivery uh were also common um even in the park cutaneous procedures. But a lot of these were um per cutaneous uh sort of history. Skah pick type of procedures, not needle based procedures. So um so there are complications associated with maternal fetal procedures. Um in the mother's, including all of the above and fetal loss, christoph. Well, most uh looked at data From another institution of all per cutaneous procedures, 53 of them with, again, no maternal deaths, no major complications. But post operative pain bleeding, necessitating transfusion and pregnancy loss are about 10%. And so Combining this experience has published experience and and others centers I think would agree that our general counseling for per cutaneous fetal procedure is a 10% pregnancy loss rates can be expected and um an additional concerns for mothers but no record of maternal deaths yet. Well, we looked at um at again at this atrial septal intervention group a little bit more closely with maternal complications in mind and found that our need for impact or exit procedures and for C sections was actually less in the fetal cardiac intervention group. So that was an interesting finding that potentially we can be um uh improving and changing the outcome for the mother while we are also trying to improve the outcome for the fetus. Um Also from that graph, I just want to point out that neonatal resuscitation and preoperative ventilation were also lower in the atrial septal intervention group. And if you think about a patient, a fetus with intact or highly restricted atrial septum and I tell you that they only have a 50-60% chance of needing to be intubated at birth. This is actually a real I think proof of concept that the fetal intervention can uh improve stability. We just weren't really able to show that because of the numbers in uh in this study was a difference but not statistically significant. So where are we now with with this uh concept then of the fetal interventions? Her cardiac disease? Uh overall. Um yeah. Are they a good idea? Are they not a good idea? Do we improve outcomes for the fetus or for the mother? There are uh schools of thought about fetal therapy. I did mention ethics earlier in the talk and I wanted to bring up this article from 1982 where Mike Harrison and some of his colleagues in the fetal therapy world, published in the New England Journal of Medicine. Some general guidelines of maternal fetal interventions. We recently rewrote these guidelines in the context of the current era, pointing out that when undertaking maternal fetal therapy, we have to first consider maternal safety and autonomy um and then have the necessary expertise and systems based for provision of these interventions, including care for the mother after intervention. If there is a complication, transparency and reporting was another basic tenant and criteria for ensuring and advancement and innovation so that we don't uh sort of squash innovate innovation by always insisting on randomized controlled trials but also um uh ensuring that this is done in in a safe scientific and transparent way. We also need to to uh continually consider training of new generations and also recommended that um that detailed recommendations for center infrastructure need to be uh need to be agreed upon uh very soon um before really maternal fetal intervention uh takes hold and is starting to be done in centers that are not providing appropriate infrastructure. So to come back then to this idea of changing or improving outcomes for fetuses with congenital heart disease, left heart obstruction and defining improvement. Now using the data again, the pregnant woman has issues of safety and autonomy. We do need guidelines but we do have some safety data and as we're moving towards less and less invasive, more per cutaneous procedures. Without general anesthesia things are improving. But there's still work to be done as far as the fetus is concerned. Are we improving things for the fetus? We cannot get probably better than 90% survival because of that 10% peri procedural risk associated with the fetal procedures. There's little in the way of analysis of volume outcomes relationship beyond just fetal death, fetal complications. And so one of the next registry efforts is actually to look at those 600 plus interventions and see where where the learning curve is. And if there should potentially be centralization of procedures to minimize those complications that I reported for the child. Uh We have really no data about this context issue and the ability of the health care system to deliver post natal care for the single ventricle affiliation versus ross. Um And so more work needs to be done there as well. And standardizing post natal care defining outcomes before we can assess for improvement, I think is key. We have to decide what it is we're trying to improve. Or are we simply talking about different options? Is the RosCO no operation and a neo Nate with pulmonary hypertension and uncertain long term survival and outcome. Is that really better than a norwood glenn Fontane pathway? We just don't know yet. So, so we have to define and maybe that's that definition of improved is going to be different for each family. So I would say that changing outcomes we definitely can change them, but improving outcomes is not quite clear. So, to summarize that pro con argument for aortic stenosis, the pros now are that it does give you the potential for avoiding Fontane, it's easy and experienced hands. There is very little maternal risk and to make confer a survival advantage, even if it's not successful, certainly is innovative and fun. On the other hand, successful patients still have aortic stenosis, there's a 10% fetal loss rate, there is some maternal risk and there are unknown effects on neuro development. We're assuming that it's equivalent, but we don't know that yet hasn't been systematically studied as far as the intact a restrictive atrial septum the pros. Well, these are really, really poor candidates for norwood Retaliation and long term survival is in some series less than 10%. And so there seems to be little downside to an attempt at a fetal intervention. It is feasible in experienced hands, but not easy. It may allow for a vaginal delivery. That was the one thing that we showed a decrease in the number of C sections. Um It may confirm survival advantage in the longer term. It's unclear. And again, it's certainly innovative and fun the downside. The patients still have hypoplastic left heart syndrome. So you're trying to take a bad hip hop. Last turn them into a good hippo class. There is a fetal loss rate. Stent, improvisation can't be dealt with in these fetuses and there's probably a little bit more maternal risk, especially for bleeding. With longer operative times. There's no difference been shown in short term survival and again, not systematically studied. And so overall what I want to leave you with is that maternal fetal intervention for fetal cardiac disease uh, deserves attention to the emerging standards for care delivery. Certainly a lot of places worldwide are doing these procedures. Continuing investigation, though, is necessary of both. We still don't completely understand the natural history, the defects nor the outcomes the so called unnatural history after intervention. Both short and long term transparent reporting of outcomes for both patients. The mother and the fetus are remaining necessary and the good news is collaborative efforts and iterative approaches to improvement and candidate selection, technique and definition of success are underway and are possible in today's medical world. So with that, I will um, take any questions. I just want to thank Dick Epcot's, who's in Leiden and has been my european counterpart for the international registry, uh, Wayne Gretzky and Helena Gardner for their lively debates on this subject. And, of course, a brew golf and Katie Archbold in administrative analyst who has been key in keeping the registry going, Thank you.