Dr. Jason Smithers - Congenital Diaphragmatic Hernia Outcomes

Others, um, so, uh, we all know and love Jason. Um, he is, uh, give us a privilege of hear him talk about, uh, our view of our congenital diaphragmatic outcomes. Um, he is someone who, uh, did a surgery residency in Portland at OHSU. Um, he also did his research time in, uh, critical care time here, started his training here during that portion, then did his fellowship here and has been on staff ever since. Um. Been uh extremely productive member of our oesophageal, uh, and airway, uh, the EA team, um, and, uh, also works in the AFAC and, um, uh, the diaphragm program. So, um, so it's a great pleasure that I introduce, uh, see Jason Smithers. I All right, this sounds OK? All right, this will be a bit of a whirlwind. I, there's some extra slides at the end, some of which I'll fly through, but uh people could have them for reference later if they wanted. Uh, no conflicts of interest or financial disclosures. Um, just talking about you diaphragmatic hernia, brief, uh, Uh, briefly just to see that there's a classification and a spectrum of severity in, in patients with uh type A and B defects, which are smaller, have a much different type of course generally than patients with larger defects. And, uh, much different treatment strategies that you need to employ. So today, I'm gonna take a brief detour at the beginning just to talk about thoracoscopic repair, um, which remains controversial. Then, uh, we'll go into, uh, reviewing our outcomes, uh, here at Children's. Uh, we'll talk about prenatal risk stratification. Which is, shouldn't be too controversial, but there's still some thoughts of does it really matter what the, what the lung volumes are. Then we'll focus on our ECMO cohort, which is, is the problematic group and uh talk a little bit about the, the remaining nemesis that we have of pulmonary hypertension and, and hypoplasia. So briefly, I, I think it is well established that thoracoscopic repairs for diaphragmatic attorney have, have higher recurrence rates than open repairs in general. Uh, a recent consensus, uh, statement article, uh, published, uh, out of a Canadian journal. Uh, they, they say they don't, they don't have any randomized trials or, or true high-level data. But nonetheless, they would strongly recommend, uh, not doing minimally invasive approach for CDH due to, due to, to the uh recurrence rate. So, um, and we've certainly had our recurrences here as well. And so a few years ago, we set out to just look at were there reasons we could find for that. And uh one thing we noticed, and, and other people have noted this too, if you look here, certainly, neonatal repairs are more problematic than, than kids repaired uh later in life. Uh, and the cutoff is around one month of age. So for, for children of that age or older, the, the recurrence rate here has been 0. And uh so that remains a group that you, you wouldn't necessarily ban thoracoscopic approach for. We, we noticed interestingly that, the, the, the right side of repairs, we didn't have recurrences, although this is a small number, so it's not statistically significant. Um, the, the thing that I, that struck me was that patients with a sac type diaphragmatic hernia. When we repaired those arthroscopically, they actually had a very low recurrence rate. And the method we employed from the thoracoscopic vantage point was uh keeping the hernia sac intact and pushing it below the diaphragm, and then repairing the, uh, the defect with essentially that sac intact, but it's pushed to the abdominal side. And so, based on that finding, and other people I've talked about similar strategies of, of buttressing repairs uh with an underlay uh patch or, or mesh of some type, uh, although it hasn't really been published before, uh, we, we did some technical modifications of what we were doing. Uh, and this is an example of this. So, uh, looking, uh, from the chest down on the diaphragm, this is a type A defect. The contents have been reduced already. And we're essentially doing a primary repair. It's just buttress with a sandwich. Uh, in this case, I use bovine pericardium underneath and Marlex on top. And then these are horizontal mattress sutures, uh, uh, to suture that buttress in place. So there's both an underlay, uh, and an overlay. The overlay Marlex is to allow a weave for, for better scarring. And on the recurrences we had had, we had noted that you'd often see pieces of omentum uh sticking between the sutures uh when you went back. Uh, and so, uh, similar to ventral hernias, which have a high recurrence rate, uh, an underlay component of a, of a buttress, uh, was the thought here. And then, uh, the same idea for a patch, um, this is a, a CDH that was more of a type D defect or excuse me, type B. So required a small patch. And we're using a dual layer patch. Uh, in this case, uh, using Gore-Tex underneath and Marlex on top. And with each, um, each suture placement, the rim of the diaphragm is sandwiched between the two. So, uh, you also, uh, in both cases, it's almost like the sutures are plated with this buttress material on, on both sides. And uh this is just uh showing this recurrence rate in another way that uh older children uh have a low recurrence rate. So this certainly remains a viable option in that group. Uh, again, the, the sac type had a much lower recurrence rate. And when we did these buttress repairs, we have not had recurrences, uh, as of yet. So, uh, ultimately, uh, I think we'd say that thoracoscopy is a, a valid option for neonatal CH repair in select stable patients. Um, we've only chosen to do this in, in kids who didn't need ECMO and, and who are clinically more straightforward to stabilize at birth. Other centers do talk about using thoracoscopy more widely. Um, but I, I personally haven't been pushing for that. This paper, we did not talk about any kids on ECMO cause we haven't, uh, used that approach. So we can't specifically speak to that. But I think, um, uh, I think these technical modifications have, have so far eliminated the recurrence problem. Does that mean they're better than the open repairs in those patients? Uh, I don't think you can necessarily say that either. But it, uh, I think you can get equivalent results. So, uh, these next slides will be a little bit busy, and I'll, I'll go, go through them. Um, first of all, this is just showing uh a range of our annual statistics. At the bottom, this actually reflects statistics from 95 until now. Um, 95 is, is, uh, from that year forward is what, what uh I was able to find good data for. And, and you see that uh in that period, we've had 400/400 diaphragmatic hernias. And about almost 40% of them have required ECMO and 60, uh 62% have not. Uh, a little over half of them are inborn, and it's about the same rate of, of prenatal diagnosis. Let's see how do we make this go away. Um, this, uh, this section is, you, you can't see the label, it, it's left versus right. But ultimately, we have a little over 80% of our diaphragms are, are left sided. These two columns, this first column is uh patients with either A or B defects, and this column is uh C or D defects. So, it's a pretty similar rate to our ECMO rate, uh, about 40% of the larger defects. This, uh, column represents patients with, uh, severe cardiac defects that require repair, uh, or chromosomal anomalies. And overall, low rates, um, 5% for chromosomal anomaly. Friends, Friends syndrome is the most common. And then if you look at our overall survival, um, uh, in these columns, I've, I've listed, for example, here, this year, we had 88% survival with uh 100% adjusted survival. And that's basically adjusted for, uh, children who had, um, severe chromosomal anomalies and basically care was redirected without, uh, uh, attempting to treat them. Um, so you, you can see that overall, we have, uh, almost 100% survival for CH repairs who don't require ECMO. In other words, we, uh, we don't allow children to die without at least trying ECMO. So that in some ways, you could say that's artificially high. Uh, uh, and then this is our overall, um, adjusted survival for patients with ECMO, uh, which over the years has been pretty steady at 50, around 58%. The national average is right around 50%. Um, so we're, we're on the higher side of that. I highlight a couple of years. This was uh uh a year where 22 out of 6 children made it. So that was a low year on the ECMOs survival. This most recent year so far has been our highest ECMOs survival, um, Uh, uh, by far, uh, we had a total of 7 kids. Uh, interestingly, only one of these had, um, uh, a PPLV below 15. We had one patient with 10%. Uh, all the other ones were 16% or higher. Um, this has led to so far a pretty high survival for, for 2018 of 94%, uh, and an overall survival in our, in our patients of, uh, 84%. So talking about prenatal risk uh stratification, the typical, uh, prognostic factors people talk about, certainly chromosomal anomalies or syndromes. And some of these you can diagnose prenatally. Um, but, uh, it certainly happens that we, we figure this out postnatally as well. I mentioned that Friends syndrome has been the most common one that, that results in redirection of care. Um, But, but you can, uh, but there are certainly other ones come up. Significant cardiac anomalies, uh, especially hypoplastic left heart syndrome or any, or any variety of uh AV canal defect that ends up acting like a single ventricle. Um, it is a bad prognostic sign. Right-sided CDH is classically thought to be a worse prognosis. Uh, with these kind of general numbers reported in the literature of, of a survival of 50% versus 75%, a higher ECMO rate, and a higher patch rate. Our survival here for right-side is actually not that different than left. Uh, it's pretty close. Uh, we do have, um, not quite double the, the rate of ECMO use for right-sided. Uh, and the patch rate is a little bit higher. So our numbers, uh, over that twenty-year span are a little bit different than generally reported. Liver herniation overall is considered the most reliable prenatal predictor. Um, I, I think essentially because it's the easiest to, for anyone to figure out. Um, going into more lung volume measurements, um, lung head ratio has been, been used, uh, uh, for a long time. Uh, it, it, you improve the accuracy by doing observed to expected, and the expected LHR is calculated, uh, with sort of complicated formulas based on gestational age. Uh, with these kind of numbers, um, reported for, uh, how severe you think they're gonna be. And these are generally the, the numbers that the fetal trial, uh, centers are using to, uh, include or exclude patients for fetal intervention. Um, uh, a more thorough and sort of accurate way to measure fetal lung volume is by MRI. Um, here, uh, we use a% predicted lung volume. Uh, probably more centers use observed to expected total lung volume. Uh, both, both, uh, calculations would, would have the same method of, uh, deriving the observed lung volume. Uh, especially looking at the MRI and there's 3D programs to sort of calculate the volume. The expected total lung volume, there's 6 or 7 different formulas for calculating that based on gestational age, uh, based on natural logs. I, I'm not smart to figure out how they derive those formulas. But I would say that there's quite a range of expected total lung volume you'll get depending on which formula you use. Uh, the good thing about MRI is you are accounting for all the observable lung tissue rather than a single shot, uh, view. Uh, so this is, um, Uh, looking at a, an LHR measurement on ultrasound. And you basically take a shot of the chest with a four-chamber heart view. And whatever lung tissue you see in that view, uh, kinda gives you your, your, your lung measurement. As you can see, uh, this is the, the same patient on an MRI. Uh, the, the blackout part here is the heart, and this is sort of the lung you can see in that, in that view. Uh, but a few clicks, uh, further through the series, and you see a, a, a, a much larger portion of right lung. And so, uh, I, I think that's why MRI is so much more accurate in, in lung volume measurements. In our patients, uh, when you see the reported LHRs, there's always, uh, quite a range that is often reported. And to know exactly where they fall in that range, uh, can be difficult, but the MRI, uh, version, uh, should improve that. And this is uh uh first paper in 2007. We started having, um, doing MRI measurements in around 2005. And this was, um, one of the first papers talking about that. Uh, this is showing around 15% to be a critical cutoff level where you start, start to see a, a big, um, drop in the survival. He's the more recent paper that Hester helped with, um, uh, uh, with Doctor Buck Miller, and showing the value of, uh, not only having one MRI measurement, but having, uh, probably at least two throughout gestation. One is typically in, uh, around the first visit, which is in the ballpark of 20 weeks, and then a subsequent, uh, shot around 32 to 34. Um, And what you, what you saw is, um, this is showing patients who, uh, start low. Uh, and, and remain low. So both measurements were below 15%, both at 20 weeks and, and 32 weeks. And that group has a, a low survival and a very high ECMO rate. For patients who start higher than 15 and drop below that, they're, they still have a very high ECMO utilization, but they, uh, the survival is much better. So that late gestational MRI is preparing you for what to expect in the delivery room and, and those, that first day. Uh, and then finally, if you stayed below 15 on both, uh, shots, then your survival was even better. And this is a just a statistical curve, uh, showing the relationship between uh uh PPLV and, and survival. And it's a very significant uh probability curve. Um, so I think this remains probably our best, uh, overall prenatal, uh, uh, risk stratifier. Others like, uh, percent of liver, uh, above, uh, the diaphragm is one that's used, um, and, and other measurements on, on fetal echo. But probably the bottom line is, is the, the best lung volume measurement you can get. is gonna be the most accurate. So now, if we look at our uh PPLB results for patients that we've had under 10%, uh, they've had a 100% rate of ECMO. And we only had one survivor in that group. And that patient was in the hospital well over a year, and, and with severe developmental delay. Um, Uh, we, we haven't seen her probably in 5 or 6 years. Um, she was here when I was a fellow. In this group of uh 10 to 12%, uh, still 100% ECMO use, uh, survival, um, is up to 40%. None of these patients, uh, went without ECMO. In this group, 13 to 15%, still a very high ECMO use, but not 100%. Um, the survival, uh, Uh, for the ECMO group was 58 versus 20 for a total survival of 67% in that group. Um, so over since 2005, for, we've had three kids, uh, in the 15 or under range not need ECMMO. Um, When you get above 15%, uh, uh, 16 to 20%, there's still a pretty high ECMO utilization, 61%. The survival is actually a little bit lower for whatever reason. I think these are overall small patient numbers. Um, the non-ECMO rate is much higher. And, and giving you a, a total survival, that's a little bit higher. Uh, once you get above 20%, the ECMO rate, uh, again is falling. Uh, even these children who need ECMO, when you're above 20%, the survival is, is, uh, much higher, uh, um, a higher non-ECMO rate, and then you're pushing 95% survival, which goes up, um, above 25%. When you look at how things are in the delivery room, uh, in these kids with a PPLV less than 10, uh, this is where we've had a pretty high CPR rate in the delivery room. Uh, even, and these are all inborn, uh, kids where you're, you're sort of expecting them. We did have one of these patients survive. Um, it was that, that one I mentioned earlier. They all had very, uh, the average time to cannulation from birth was 1 hour. Um, and, uh, a range of initial pHs. Some of these not too bad, but an average of 6.9. In the 10 to 12%, the, the CPR rate in the delivery room was 32%. Uh, 2 of those, uh, children actually, uh, lived. The time decannulation average is a little higher. But still, uh, 14, almost half the kids, uh, or a little more had a cannulation time, uh, in under an hour. Uh, and then the numbers, uh, improve a little bit after that. Over 13% or higher, the CPR rate drops quite a bit. Um, And the average time decation is getting higher. And so this leads to sort of our typical stratification when we're talking to families in the, in the AFCC that uh when you have a, a PPLV greater than 25%, that's a 97% survival. We, we still try to attend those deliveries. Uh, but it's a very low ECMO rate. Um, Uh, in these, these cases, uh, again, that was the 95%, uh, group. And the ECMO utilization, uh, was still 32%. So we have Just a team aware. Uh, and, and going on from there. Uh, for kids in this range that are under 15, uh, we'll often have ECMO ready and waiting, uh, for them. And perhaps the delivery strategy for these is still up for uh discussion. Uh, we, we basically aren't really having survivals in single digit PPLVs. Um, and whether that means that's just lethal, and we'll never improve on that, or that's an opportunity to figure it out. Um, all right. So, now focus on just the, the CDH kids who, who require ECMO. And, uh, I think, uh, a topic of debate, uh, here, uh came up at, at least the last 5 years, maybe a little longer. And, uh, what should be the timing of repair. Uh, uh, here in Boston, we've, we've been a proponent for early repair on ECMO for, for eons, probably one of the early adopters of that. But, but we had a run where we were sort of debate, debating that. When you look at other major centers, um, certainly Philadelphia has, has classically been, uh, a group that recommends late repair, usually after, after eCO decannulation. Uh, uh, Children's of LA, uh, um, recommends that as well in their articles. Um, but there's still a high mortality for kids that can't come off ECMO. Uh, Michigan, they actually go back and forth a little bit. Texas Children's, uh, uh, or all Children's in Florida, uh, are some of the, and, uh, and ourselves probably are, uh, generally go for early repair. The CDH registry papers have argued both sides of this. And, um, a paper that won an award at the AAP that Doctor Wilson was the, the primary author on. Um, looked at, um, you know, a big cohort. Uh, they were looking at 630 patients. And this is based on uh centers that basically routinely repaired on ECMO or not, like what percentage they did that. They weren't, they weren't grouping the patients in the timing of repair per se. So in the group that did the majority of repairs on ECMO, their average time to repair after cannulation was 6 days, which most people wouldn't consider early. Um, But nonetheless, that paper seemed to show that all the different timing, all the different groups had similar uh survival of around 52%. Uh, it was noted that the, the centers that did 90+% repairs on ECMO had, that group had a higher CPR rate, still, it's still achieving that same survival. So maybe that was a clue that repairing on ECMO uh is helpful, but it's not certainly very strong. And that, uh, I don't think that paper is gonna end up being published, but I'm not sure about that. Um, I've included some, some other articles at the end of the talk that people can look at if they, if they want to. But a recent one in 2015 was a pretty decent review article of all the, all the Timing papers. Um, the, the one thing they did say is that using an adjunct to decrease the bleeding risk is probably helpful. Either Amicar, which is what we've classically used here, or TXA. Other than that, they would say the literature is not great on the topic. Um, but either early repair while on ECMMO or delayed repair after decannulation appear to be superior than delayed repair while on ECMO. And certainly, there, I don't, it's hard to find centers that, that sort of support that they, they prefer to do delayed repair on ECMO. I think most centers in that category re uh prefer to repair after the ECMO run. But if you can't come off, they're stuck doing a, a delayed repair. And that's probably setting up a pretty, pretty high mortality group. The most recent uh CDH registry type paper has gone back to the side of slightly preferring repair on ECMO. It's a little bit of a complicated paper, but nonetheless, they grouped centers, um, Uh, high low here meant they were a high volume center doing more than 10, uh, CH repairs a year. And low meant they had a low rate of non-repair. Uh, so you, so you might just say that this high low center would be an aggressive center. They aggressively try to repair, uh, kids on ECMO. And and they had, uh, an overall improved, uh, survival compared to certainly low volume centers and centers that had a high rate of non-repair. Um, so, so they said, for every 100 patients, you might get almost 3 extra survivors beyond uh expected. Um, an issue with the registry in general, though, is, is that centers have such different philosophies and strategies that it's a little bit difficult to compare. Um, so the paper that we're trying to finish up, and we presented at the AAP, um, a little over a year ago. Uh, was looking at this. And, um, Basically, as we've said, the optimal uh timing is, is not clear. Surgical bleeding, uh, is really the main, uh, problem that centers, uh, debate. Uh, and, and, uh, in centers that don't like to do stage repairs on ECMO, it's the concern for, for bleeding risks that is, is primarily an issue. Uh, so in this, uh, case, we, we looked at our, our, uh, results with, uh, CH repair on ECMO, uh, either early versus late. Uh, and this is defined as 48 hours. We had 133 patients, uh, in this time. It, it, the last couple of patients were in the early part of this year. Um, again, the definition of early versus delayed based on 48 hours from the time of, uh, emocannulation. And bleeding is defined as a need for re-exploration of either the chest or the, or the abdomen. Um, interestingly, the CH registry doesn't currently report bleeding. So that's not something you can sort out, uh, of the registry, which would have higher, uh, patient numbers. So in this group, uh, uh, early repairs had, uh, 90 patients and, and delayed was 43. Uh, some of the prenatal risk factors are all quite similar, including the birth weight, uh, the rate of inborn patients, uh, prenatal diagnosis, left versus right, liver up, uh, PPLV. Um, we don't have numbers on all the patients. Uh, defects of either C or D. Um, the delay group, uh, did have a higher rate of CPR prior to cannulation. Uh, so that was a significant difference. Um, We're sort of going back and forth about uh calling patients unstable at say 24 hours after cannulation. And, and what does that mean? Uh, I think many, certainly, Wilson, uh, would argue, well, you know it when you see it. And I, and I think in many ways, that's true. Um, when you look through the notes, uh, for these patients, there's certainly a, a 30% group that were defined by the, the people involved at the bedside that this patient's not doing well. And at one time, we had the strategy, they're not doing well, therefore, we're gonna go forward with repair to see if that'll make them better. And at other times, we've said they're not doing well, we think we should wait. Uh, none of this, uh, there is a difference here of almost 14%, but that wasn't statistical. Um, another interesting thing that we noted is how high the BUN was for patients in the delayed repair group, uh, with a median of 72 versus 14. And that was a, a big difference. The survival, uh, was, was, was similar. Uh, no statistical difference there. Uh, the bleeding, uh, was drastically different in the delayed group. And, uh, as was the ECMO duration by about 5.5 days, uh, uh, for the, for the mean. And so this, this graph shows uh the, the uh timing of repair and sort of what happened with the, the rate of bleeding. Uh, and you can see it, it marched up, um, Uh, uh, over time. Now, uh, initially at the, when we did the presentation, We were, we're also looking at, uh, like breaking down sub subgroups. And, and we took a, uh, the subgroup of kids who were uh cannulated in the 1st 48 hours after birth, which is the majority of patients. But we saw that if you, uh, if babies will last, uh, greater than 24 or, or certainly 48 hours before requiring emocannulation, that's a much, uh the, the baseline, uh, better group as far as survival than kids who are cannulated before that. Uh, and the difference was in the ballpark of, uh, 80% survival versus 50%. And in our initial analysis, it showed a survival advantage to do early repair in this group. But, um, But when we looked closer, uh, uh, I went back after this, uh, after this presentation to get the exact time of birth, the exact time of cannulation, and the exact time of repair, uh, to the hour rather than the, the date. And uh so on a, uh, which means that this, this survival advantage is not something that I, that we see now when you have the, the hours more granular. We still do see that for kids who require early emocannulation from the time of birth, their baseline survival is lower than kids candidated after that, um, 53 versus 75. And that's a significant difference. So, uh, as far as how early kids are candidated, that sort of sets up a baseline difference in how, how they're gonna do. There's a survival difference in this worst group with early versus delayed repair. But it's not, it's almost 20%, but it's not statistical uh for this sort of uh patient group size. Um, interestingly, it, it flip-flops, uh, for, for the higher survival group, uh, where delayed repair had a little higher survival, but that, that again, was not significant. Um So now if we look uh through this data set for logistic regression for the different variables, when we look at, at the deaths, so there are 55 in this group. Uh, a type D defect was a significant predictor. Uh, low birth weight was a significant predictor. Having a cardiac anomaly was, as was renal failure and eCO duration. Uh, the timing of repair wasn't significant. Uh, having CPR wasn't, or a stroke, uh, or this somewhat nebulous, uh, clinical instability after, after cannulation. Um When we looked at, at the risk factors for bleeding, uh, certainly, as you saw, timing was of repair was a big risk factor. Interestingly though, if you include, uh, the, the degree of uremia or, or the BUN level, that overtakes the timing of repair. Um, uh, so that, that remains a strong predictor. Uh, and that the median, again, uh, BUN for bleeders was 50, was 78 versus 16 for non-bleeders. Um, So, so this is, is something that we can modify. If, if we're stuck doing a delayed repair, I, I think a lot of kids, we push the BUN high with diuretics, trying to get them, their volume status improved. So I think to a degree, these, doing repairs with the high BUN is something that we can potentially avoid. Uh, perhaps not always. Uh, when you, um, do the multivariable analysis for the ECMO duration, The things that uh That were important were having renal failure. Uh, bleeding, increased it by an average of 10 days. Defect size, uh, by an average of 2 to 3 days per stage, and timing of repair by an average of 5 days. Um, The, the only uh significant predictor for having a stroke was uh renal failure. A CPR interestingly was not uh significant for that or cardiac defects. So overall, when we took a detailed look at our repairs on ECMO, we definitely see a much lower bleeding rate, uh, and a, and a shorter ECMO duration when you do the repair within the 1st 48 hours of, of cannulation. Um, high BUN at the time of operation is an independent risk factor, uh, which we need to keep in mind. Patients that require early emocanulation during the 1st 24 hours, uh, represent a higher baseline mortality group. Um, and possibly would benefit from early repair. I think we would need, uh, a larger study or to see if we can look at the registry for that question. Um, Uh, and again, this is, uh, the limitations of a single center study where we've strongly preferred repairs on ECMO. Um, this is our numbers for CH repairs, uh, for patients who had ECMO, but the repair was not, uh, while they were on it. That was either they had a CH repair and required ECMO afterwards, uh, which we've had 13 cases of, uh, over that 20-year span. It's, uh, a 77% survival. So that's higher than the baseline of 58. And an average eCO run of almost 2 weeks. The strategy that many centers prefer to repair after they decanulate, we've only done 3 times. So the, uh, 22 of them lived, and the duration was 9 days. But it's such a small number in our cohort that it's not something that we can really comment on. Um And, and I think it, it, sins tend to choose one philosophy or the other. So to get a single group where you can really control all the factors, know all the data, uh, is not happening that, and perhaps the registry will make some ground, uh, on that over time. Um So, some, um, Uh, some data that I have from the registry for this 10-year period. It didn't include the exact birth times or, or cannulation time. So I, I couldn't work on that part. But nonetheless, just that this is a, a good comparison for uh to compare to us. The baseline survival for uh kids who required ECMA was 51. Uh, so again, we're averaging 58 above that. So it's above that. Um, kids who had repair before ECMMO, uh, in their, in the registry was survival was 57%. Ours was, uh, 75%. Uh, uh, a pretty high number of kids repaired after the ECMO run with a high survival, 78%. Average emo duration of 10 days. Um, The issue is, it's just that a large number of kids who had no repair. Uh, and those kids aren't survivors. Uh, if you look at early versus delayed repair while on ECMO from the registry, Um, there's no significant difference if you make the early and delayed cutoff at 2 days, as, as we did in our paper, um, the 53 versus 47. Interestingly enough, if you make the cutoff 10 days, that, that's statistically significant. Um, But it's sort of an odd timing to choose though. So it seems a little, uh, I'm not sure really what to make of that. When you look at an area under the curve calculation for timing of repair, it shows up as significant. But nonetheless, uh, our analysis was that timing of repair is a predictor, uh, but it, it's certainly not a strong one. The strongest predictors in the registry data were the both the defect size and the ECMO duration. Uh, the ECMO duration may simply be, uh, looking at what's the baseline risk of the patient. Um, The, the ones that are, are worse off have longer ECMO runs. So even to say ECO long eCO duration leads to higher mortality, it, uh, it's sort of redundant almost. Um, so overall, when we look at timing of repair, I think we would say it's not, uh, we can't say that it's a strong predictor of survival, survival. But there are definitely differences relating to bleeding and eCO duration in our, in our cohort. Um, I think it is safe to say that if you can successfully wean from ECMO without repair, that is a, uh, a good prognostic sign for that patient. That's probably a better baseline patient group. Um, and, and perhaps some registry papers can, can work to prove that. Um, but ultimately, optimal timing may still depend on each patient's specific risk, uh, and, and how they present. So now, uh, still, um, focused on our ECMO cohort, uh, in, in some ways, to state the obvious obvious, but just to look at, uh, uh, a series of patients who died. Uh, actually, this is, uh, includes some kids who, who died without, uh, um, ECMO. But in, in the patients that we attempted to treat, you would look at pulmonary hypoplasia and pulmonary hypertension as implicated in, in over 90% of the deaths. Um, uh, cardiac, uh, significant cardiac defects, uh, were present in 23%, chromosome anomalies in 9%. Um, uh, strokes, uh, that, uh, led to redirection of care, almost 20%. Uh, sepsis was fairly common, as was renal failure. So where does that leave us now? I think things to try to fix, uh, severe pulmonary hypertension. This, this may be the easier uh uh thing to fix that. Perhaps we're making some headway on. We'll talk about that a little bit. Uh, the most recent advances that we're doing our management of the, the ductus arteriosis and prostacyclins. Uh, when we think about pulmonary hypertension treatment, um, sort of breaking down into different aspects of it, and, and what the goals are, they're supporting the right ventricle function with the Milranone or other inotropes. There's offloading pressure from the right ventricle, especially if it's super systemic pressure. Uh, ductal patency can achieve that. Um, perhaps surgical pot shunt would be another option. Uh, that hasn't been exercised to my knowledge, uh, anywhere. Uh, ECMO, uh, certainly, uh, does that. Pulmonary vasodilators like nitric, sildenafil, uh, prostacyclins, um, can be used. And people talk about for a more chronic phase of vascular remodeling with some of those same agents, uh, but adding, uh, to it bosentan. Uh, and this is a typical algorithm you might see for uh patients who come in, they're, they're on the, the scan for ventilator management. Uh, you get an echo, and you could classify them as subsystemic, ecosystemic, or super-systemic. Uh, in addition, you would look at how stable they are, uh, as far as their ventricular function, how their event settings are, and, and so forth, which is to say that you'll have plenty of kids who are super systemic in the first echo at less than 24 hours, who are on very low event settings, and sort of overall stable. Uh, but nonetheless, these are the options you may consider for kids who are unstable. Then, then a lot of those will end up on ECMO. But you can think of some of these agents and perhaps even Remodulin uh to avoid ECMO. And then in a chronic phase, uh, commonly after an ECMO run, um, we have, uh, similar, similar options. We'll consider a cardiac cath for these patients. Some of the remodeling drugs like the Remodulin or um bosentan. Um, and perhaps managing the ducts, uh, in some method if they're super systemic. So, um, so this brings us to, uh, a talk that we gave at the ECMO meeting last year on, on the role for prostaglandins, um, to, to maintain the ductal patency. Uh, and, and we basically looked at our experience with, with using glandins here. Um, we, uh, did it over this 10-year period, uh, at the time. And the, the literature on using prostaglandins goes back at uh over a decade, for sure. Um, a couple of papers out of Japan and Australia, uh, talking about it. Uh, a more recent one, and actually quite a good paper just, uh, recent this year is out of CHP. Uh, they had 57 patients, uh, with CDH that they use glandins for. Um, so this, this reference is definitely worth, uh, worth reading. In that group, I think 40% or so of the kids were on ECMO and getting glandins. Um, So for the patients that I evaluated for that talk, uh, we had 30, uh, 30 patients. The overall survival was 60%. ECMO utilization was 70%. The typical split of left versus right, um, much more commonly used on kids with the liver up, um, in a preponderance of C and D defects. Uh, the mean PPV was 16. And we basically had different groupings of kids who got glandins, um, a group that had ductal dependent cardiac lesions. Um, that's sort of a, a no-brainer that you might try that at first. Uh, a group where we attempted to reopen a duct that had closed, uh, either after an ECMO run in order to avoid ECMO or while on ECMO. Uh, in a group where we decided to maintain the duct, uh, open from the beginning, uh, in the absence of, uh, a cardiac lesion. And we did that for 10 kids on ECMO and for 8 kids in order to avoid it. And basically, what we saw in our, when we had attempts to reopen a duct that already closed, the success rate of, of achieving that was almost 40%. Um, And you could say that the RV function improved in, in two out of those 3 patients. Um, and, and, uh, but it didn't necessarily lead to a high survival. Um, 40% of these kids improved even though the duct stayed closed as far as the RV function, but still with a low survival. When we use gland is to uh maintain a duct, we were uh almost always successful when you start glandins while the duct was still open. And uh, Uh, had a 60% survival in the kids who are on ECMO, which is right around our average, which is, you know, was 58. And for the group that you avoided ECMO, um, Uh, you avoided it in 7 out of 8 cases, uh, but the one who went on to ECO didn't make it. So just briefly on the, the kids with ductal-dependent heart lesions, um, one was a co-arch with a hypoplastic arch. That patient did end up, uh, dying, uh, after the, uh, cardiac surgery. A patient with a transposition. Uh, of the great arteries, uh, had a, had a stroke, uh, about a month into their hospital course after PA banding, uh, and was redirected. Uh, this patient with both a single ventricle and anomalous veins, um, Uh, actually ended up doing well. Their, their CDH was a mild one, a type A, repaired on day of life one. They later had a uh, a BT shunt, then a Glenn and a Fontan and, and have actually done well long term. And then this patient with uh this sort of going back and forth between the AV canal and a tetrology diagnosis, um, was the worst CDH at type C, uh, but underwent a cardiac repair at 2 months, uh, and did well. Um, this is, uh, just a little more details on the kids that we attempted to reopen a duct after it closed. Um, These, uh, and these were the kids who were, uh, on, uh, on ECMO. Um, this, uh, this one, we started immediately after going on ECMO, and it varied in time, all the way up to a month later. Uh, interestingly, even at a month, uh, of age, uh, we're able to get the duct back open. Um Uh, that pigeon still didn't survive though. Uh, and then this is, uh, just some details of the group that we, uh, maintain the duct open from the beginning. Um, and I'll, I'll skip through the, the details. Uh, this is, uh, one of the we've had a couple of patients with ductal stents. Um, this was one from a few years back, uh, who, who ended up doing well. Uh, and again, the group to, uh, uh, avoid ECMO. So overall, when we look at our experience with, with using glandins, um, It's hard to say anything super conclusive. But I, I think we, we can say when we've done it, we've, we've generally used it in kids who are super systemic. And if they become subsystemic RV pressure, especially if it's a lot, then, then we generally have not wanted to keep the duct open. i.e. we don't want a lot of left to right shunting, especially uh on ECMO for concern of pulmonary hemorrhage. So, uh, for those kids that we've maintained the duck well on ECMO, we've done around uh, uh, echoes twice a week in general. Um, and, uh, without any clear evidence of causing problems in that cohort. Um, could, could doing this contribute to the need for a duct ligation later? Uh, that, we had one case in this series, um, who had a brief run of glandins for only 2 or 3 days. Uh, a month later, the duct was still open, and ended up getting a, a, a surgical ligation. Uh, that was, uh, uh, a more mild CDH who, who did not require ECMO. Uh, we had one case of, uh, of a patient who, who had necrotizing gallbladder infection while still in the neonatal period. Uh, I don't know if we can implicate PGE directly, but we did note that. And, and one patient who had a distinct, uh, ileus when we stopped glandins, um, after, after they had been tolerating full feeds for quite some time. Um, Uh, so overall, we're using this only on kids who are super systemic. And, and I don't think we're demonstrating clear complications from doing it. Um, so we can say that if you want to maintain ductal patency, you, you can have a high success rate with glandins. Um, if you want to reopen a duct that's closed, the success rate is going to be less than 50% in, in our cohort, but may still be possible even at advanced age. So, uh, potentially worth trying. Um, I think it, uh, it may play a role to improve the RV function, uh, in the kids who are super systemic, uh, possibly to either help avoid ECMO, uh, or to have that option when you, when you come off ECMO. Um, uh, we actually, on a recent case, we're able to reopen a duct in the cath lab. We've had cases in the past where I say they're a month of age and you're trying to reopen the duct in the cath lab and you can't. Uh, but a recent case, we're able to accomplish that. Um, It, it, it may have a role if, if, if you're a proponent of trying to use VV uh for CDHs. Some, some centers, uh, uh, do that. But I, I would say that we haven't been systematic in how we've applied it. Um, and so it's hard to make definitive statements about the role of glandins other than we have a reasonably large experience without showing complications. Some cautions. Um, does the duct actually promote a higher RV pressure or make it seem like they're super systemic, um, just by the fact that they have the duct open? Um, possibly. Uh, we had a couple of cases that were super systemic until you let the duct close, and then they were subsystemic almost immediately after the duct closed. Um, Uh, and, and we've done ductclusion test in the cath lab to try to sort that out. Um, And I think it, it, it also is safe to say that having an open duct alone is not enough to, to address uh pulmonary hypertension. Um, and it doesn't, uh, guarantee, uh, that you're gonna prevent RV failure. But we've had, uh, several kids who still developed RV failure with an open duct. Either we were keeping it open on purpose, or, or they just happen to maintain a, a large duct even without trying. And then we've had 4 cases to suggest that if the duct is too large, that they may actually be counterproductive in, in promoting uh uh fetal-type circulation. Uh, and, uh, that was the reason why we actually turned off glandins and downsized with the ductal stent on, on one patient. That patient was, is still doing well. They were on Remodulin for the better part of a year, um, with that. I think it's still an option to think about a surgical pot shunt. Um, it is now a standard treatment for primary pulmonary hypertension. Uh, I think it's more ideal if you can predict who would benefit from an open duct and do it by another means than, than a cardiac surgery. But nonetheless, that may be something to think about. Where Um I'll fly through the last few minutes. So, the other thing to fix severe pulmonary hypoplasia. Uh, probably the worst kids are even more susceptible to ventilator associated lung damage. And it's not really clear if we're making progress. Uh, this is certainly an effort from the, the fetal intervention trials, uh, which may be showing promise. I think it's unclear yet. Um, And uh with some centers showing such a high survival with postnatal management, that that's gonna be counter to the fetal intervention trials. So we'll have to see where those things uh fall out. Um, but I think we've, we've had numerous cases of severe CDHs where everything goes pretty well. You have a pretty good pulmonary hypertension regimen, including that, uh, with or without an open duct. They have good heart function, but the lungs, uh, still crap out. Um, This is briefly showing a VQ scans. The only point I wanted to show is that on the ipsilateral lung, uh, these are type C and D defects. You have a VQ mismatch that is present and either persists the same or gets worse over time. So the, the, the two lungs, to me, I, I don't think we should consider them equivalent. Uh, we see this as well in kids with cardiac cath. This is the left CDH, about a month of age. But you can tell it's not doing well cause they're still on ECMO and you're doing a cath. But nonetheless, they have hypoplastic, uh, vasculature on the involved side. And that would correlate with the VQ mismatch. So this lung, although it fills the chest, has, uh, is, has more dead space. So a question to me to think about for the pulmonary hypoplasia problem is should we think about uh uh growing the lungs differently from one side to the other? Um, I just flip through to finish. Uh, there's a slide showing that the lung growth schedule with perflubron. My only point saying perflubron never really worked out to improve survival. And it may be because it preferentially grows the ipsilateral lung, uh, which is prone to have more dead space. Um, these are some papers about that. So I think things that, that I, that I personally think about are, can you do lung isolation to promote growth of, of the better contralateral lung? That could be possible. Maybe you would, could use proflubron with that. Maybe you could even do the fetal tracheocclusions of the right main stem rather than the trachea. Um, But I think suffice it to say that the that the lungs aren't the same. And the, the worst lung is the one that more easily grows in the neonatal period. Uh, I think we could, and, uh, 01 quick plug for, for Doctor Peter, who I think can help solve the hypoplasia problem. And that's is, can we start using growth factors for that purpose. Uh, so animal work that he's working on with VEGF, enhancing uh lung growth. Uh, interestingly, heparin impairs that VEGF, uh, uh, improvement in lung growth, which we have all the severe kids on heparin. Uh, this, this paper also showed that, um, using bivalve rudin as the anticoagulation does not have that interaction. So that that could be something to think about. Um, And so I think growth factor manipulation, and basically looking at things like this is a baby with a severely small lung. Can we get that organ to grow in the way that we want? Uh, we know, for example, with long gap esophageal resia, if you do it just right, you can get the esophagus to grow well. We've also shown and made a career out of showing 1000 ways to mess it up. And so I think the same will be true for, for lethal pulmonary hypoplasia. If you do it just right, there's probably a way to get uh a very high success rate. But there's probably gonna be a lot of uh uh traps along the way. Thank you. Jason, I seem to be without a microphone I think. Thank you for putting together a great massive collection of data about CD trying to do to improve your. I, I had a couple of questions and I'm sure they'll give you some. So when, when you looked at the bleeding effect relative to the timing of Meeting on um Timing of repair, you clearly showed that. Elevated UN and creatinine is associated with. Noted in the, um, Actor analysis, analysis. That into consideration and the timing was not significant. Well, it, it to me, it seemed like a, a quirk of statistics. Like in this multivariable analysis, and with a small overall number of bleeders, you can't put a lot of variables in at the same time. But if we had this factor of a high BUN creatinine ratio, then the timing was more significant. If you use BUN only, then that becomes more significant. So, um, I think ultimately, uh, we, we need to be worried about both. And, uh, And I think the good news is that the BUN is potentially modifiable. Um, some of these kids who are getting long on the tooth on an ECMO run, you seem to think that getting their fluid balance, uh, in a, a bright spot where they're not blowing up like a balloon, so you could do the repair would be helpful. Maybe you have to bite the bullet and do repairs when they're blown up like a balloon, and just see what happens with that. But nonetheless, when you do it on a super high BUN and you're anticoagulated, that certainly was a very strong risk factor for bleeding. Yeah. And the second issue was around the prostate glandins and Open, so the, so the quandary is that if you increase the shunt to the, to the right to the left side. Take the pressure off the RV. Can you increase the. Saturation with the, circulation. Yeah, I think in our series. My sense has been that there's an optimal size. Um, some of these kids on Glennon's had ducks that were 56 millimeters in diameter, which is basically the same as the aorta. Uh, and certainly on one particular autopsy, you, you couldn't differentiate the duct from the aorta based on how thick they were or the size. But when we've had the duct morph 2 to 3 millimeters, which is what we've generally done with the stents, that seems to allow pressure offloading without full-blown fetal circulation. But, but there's, there's sort of a balance to try to be gained there, I think. Um, and again, the chop paper from this year is a very good article on prostaglandins. Uh, it's certainly being used more and more widely. So there's some role for it, I think, defining that role, and it's only part of one of many things to do for pulmonary hypertension. But uh I think it will continue to have a role, but I think we need to clarify the size and who exactly needs it. Bottom line, still ends up being uh. Increasing the pulmonary. Vascular supply which is computers. Yeah. Additional questions for Doctor Smith. So, Jason, uh, thank you for an excellent talk, and, uh, uh, uh, you certainly brought a ton of data to us. So you accurately point out, eCO is a stabilization strategy. It's not a treatment strategy. Yeah. And there are two things we have to do to control it. I didn't think it's gonna grow the one. One thing you did not mention was nutrition. Per catabolic individual. I should remember that Patrick helped show that. I was like sitting beside him in the research lab, but I forgot. Yeah. And it was actually uh Steve Chu who was probably born, he might, he might have even been born when Steve did his initial studies, but at any rate, you lose about 5% of your lean body mass per day on that. That we did 8 days of that, you're dead. So, basically, we uh ignore that very important thing. And yesterday, I was making rounds in the ICU and one of the dietitians asked me, do you really want to start again on the first day of ACO? Of course you do. And that is one thing that when you look at the delayed repair and the early repair, it's very important to Figure out what the nutritional supplementation is. That's one thing we can do by giving protein in 2 to 3 g. OK, which is what Patrick showed, that really will cause one to grow the lungs and very seldom, uh, Yeah, I think we're, you're in like a, a bind in, in the worst patients, unfortunately, where you're trying to fluid restrict, and yet they're on a bunch of pressors, which is giving them 8 cc's an hour of fluid. But I think you, uh, you're totally right. It has to be, uh, a big focus, because ultimately, it's a problem of getting the lungs to grow. That probably is what fixes the pulmonary hypertension, too. Uh, so. Yeah, and I would put that another concept could be early ultrafiltration and giving them Quantity of nutrition. Yeah No, that's a great point. Additional questions or? It was more of a comedy. Um, I've been here for like 2 Never seen CDH. Yeah she was bigger. No, it'll, we'll, we'll make it go somewhere. Does this kind of data acquisition is new. I we we we've all pull out our perspective to try to support. Post I think that's improving as we speak, uh, with Doctor Buck Miller, we're, uh, working on coalescing uh several different data sets that have a, a truly grand master CDH data set that is all-inclusive of all the different data features. And once that is sort of set and robust, which is taking some work, but is underway, then I, I think we can uh use that to keep answering these questions. Jason thanks very much for your presentation.