Harvard Chan School of Public Health where I obtained my master's degree. Well, I've had a very enriching and diverse 3 years. The purpose of this talk, um, I'd like to focus on a couple of projects that interest me the most. The first project is contrast enhanced ultrasound for the evaluation of pediatric abdominal trauma, a project on which I worked with Doctors Mooney and Zaleikas. Abdominal solid organ injuries are common in blunt pediatric trauma. CT scan is currently the preferred imaging modality. However, up to 75% of scans are found to be normal. Children are more sensitive to the radiation required to perform these scans, resulting in a significant risk of cancer as high as 1 in 300 kids under the age of 5. Unfortunately, in children, FAST also has a sensitivity of only 55%. The lack of a suitable diagnostic tool leads us to pursue alternative. Um, imaging and pediatric trauma. So we performed a prospective clinical trial to compare the efficacy of contrast enhanced ultrasound to CT for kids with abdominal trauma. Con contrast enhanced ultrasound is analogous to grayscale ultrasound but includes the injection of an intravenous contrast composed of perfluorocarbon microbubbles. The contrast enhancement is used as a marker of organ perfusion. You can see this giant yellow arrow pointing out a um injury on the spleen here. Children aged 8 to 17 who were able to ascend hemodynamically stable and had liver, spleen, or pancreas injury confirmed on CT scan were enrolled. A grayscale ultrasound and contrast enhanced ultrasound were performed within 48 hours of injury, and you can see the contrast enhanced version here, um, lighting up compared to the gray scale and the injuries, um, much more clearly identified. Um, 2 bolus of contrast are given and the patients are closely monitored during and after for any adverse events which we did not have during our pilot study. Um, as you saw, the contrast bubbles enter the spleen through the splenic artery and are distributed to areas of perfusion. When compared to CT scan, all injuries were identified by contrast enhanced ultrasound in 16 of the 18 patients we enrolled. When comparing gray scale to contrast enhanced ultrasound, the sensitivity increased from 42.5% to 85.7%. The specificity increased from 96.4% to 98.6%. The positive predictive value increased from 79.2% to 94.7%, and the negative predictive value from 85.3 to 95.8%. The relatively lower proportion of fat and smaller body habitus of children make ultrasound a well established first line imaging modality in many pediatric diseases, as we well know. In our study, we failed to include exclusion criteria for patients who would typically be considered poor candidates for abdominal ultrasound, such as our 1 100 kg 17-year-old patient. Our study radiologists felt that this, um, there were inadequate gray scale and contrast ultrasound views of this, of patient's abdominal organs. The FDA required contrast dose was felt to be too low during the study to provide adequate enhancement. When evaluating our data using only patients who would be considered um ideal candidates for ultrasound, as in excluding the 100 kg patient. Um, we found that the sensitivity then increased to 94.5%, the specificity up to 99.2%, the positive predictive value to 97.3%, and the negative predictive value to 98.5%. That's learning for our future study to only consider patients who would be better candidates for abdominal ultrasound. Given the success of our pilot study, we're currently conducting a multi-center trial to further evaluate the potential of contrast enhanced ultrasound in pediatric trauma. Patient enrollment has begun, and we hope that this project will add momentum to the role of contrast ultrasound over the use of CT and pediatric abdominal solid organ injury. The second project I'd like to discuss is long-term outcomes in pediatric traumatic brain injury. Traumatic brain injury has been receiving a lot of attention in the media, specifically through the NFL. However, less is known about outcomes in children and in those with injuries more severe than simple concussions. This analysis was conducted using the Military Health System data repository. Um, this is a national cohort containing greater than 3 million child dependents of active and retired military members. Using diagnostic codes, we identified all children who sustained traumatic brain injury. We matched them 1 to 1 on age, sex, and injury severity score to children who had sustained non-TBI orthopedic injuries during the same time period. The primary outcome was incident diagnosis of neuropsychiatric disease defined as mental or intellectual disorder, headache, seizure, depression or anxiety, or brain damage. The Kaplan-Myer estimate was used to compare the outcome between cohorts. In the matched cohorts of children who sustained traumatic brain injury or orthopedic trauma, both groups containing 1260, 55%, 41%, and 4% had mild, moderate, and severe injury severity scores respectively. 33% were female and 48% were under the age of 5. In the TBI group, 492 children developed more than one neuropsychiatric diagnosis, including 15% having headache, 5% depression or anxiety, 15% a mental disorder, 13% an intellectual disorder, 4% a seizure, and 4% brain damage. In the orthopedic injury. Group, 197 or 16% of the children develop greater than one neuropsychiatric diagnosis, with 2% having headache, 3% being diagnosed with depression or anxiety, 4% being diagnosed with a mental disorder, 8% an intellectual disability, less than 1% seizure, and less than 1% brain damage. Five year estimates of freedom from new neuropsychiatric diagnosis were 59.6% in the traumatic brain injury group and 80.3% in the orthopedic cohort with a P value of less than 0.01. We determined that children who suffered TBI are at significantly increased risk of developing new post-traumatic neuropsychiatric disorders and may benefit from ongoing outpatient follow-up to facilitate early detection and intervention. Additional areas of focus during my research years included the following pneumato sales and pediatric trauma, pediatric renal injury outcomes, clinical pathway for non-op management of high grade pediatric pancreatic injuries, ICU admissions in children with high grade solid organ injuries, nutrition delivery and pediatric ECMO, surgical management of anterior cutaneous nerve entrapment syndrome syndrome, incorrect. Um, hernia recurrence following inguinal hernia repair, testicular atrophy following inguinal hernia repair, and opiate use after after appendectomy in children. I'd love to speak with uh any of you about these topics when more time allows. Um, given that this is an educational lecture, I thought I would end, um, this discussion with a pop quiz, um, of the age-old question. Who wore it better. Um, the good news is that this is multiple choice, but the bad news is that it's a trick question, so our winner, um, I cannot thank you enough for what you have given me over the last 3 years. Thank you to Doctor Weldon and Doctor Shamberger for making this possible. Um, to Doctor Mooney for your guidance, encouragement, and patience, uh, to the fellows, past and present, for always being such a great example for us and always taking the time for us, um, to all the ICU teams and to every surgical attending here, um, thank you for sharing your time and your knowledge with me. It's truly been an honor. You. Speak up the compared to the last time where we had all of the comments about the speakers at the end since Doctor Mooney has to get out to uh Waltham in the OR. First, are there questions that anybody would like to address to Lindsay? Certainly, you had a remarkable body of work that you completed by the end of your time with us. Not David, I'll let you. Oh, Doctor Fishman. With the with the contrast enhanced ultrasound, you're using CT scan as your gold standard, so that's 100% negative and positive predictive value. Once you excluded the obese children, you got pretty close to that, and I'm imagining a couple of percent that you missed both positively and negatively, were minor injuries. So is the conclusion that, and there's probably nothing, there's no big sequelae is if you missed it, is the conclusion that we might be able to avoid radiation and, and diagnosing and following these injuries. Yeah, that's, that's the hope and the intent and I think that this multi-center trial um will not only increase awareness of this imaging modality but will also help us in um potentially eliminating the use of CT scan for trauma because as you know, especially now that the management is changing and these kids are requiring, you know, less intervention, less ICU, you know, requirements and even Potentially, some are going home at some institutions with low-grade injuries, you know, giving them all this radiation just doesn't seem to fit their actual need. Um, Lindsay, can you see a blush on the contrast ultrasound? I mean, sometimes we use that on the CT scans as some sort of, you know, predictive of some sort of information. Yeah, you, there's really potential that you can. Um, it is seen sometimes we haven't, you know, specifically looked at that. Definitely not in our pilot study, but we have noticed that we could see that since it is organ perfusion. Um, so I think that's one of the potentials and this study that we're doing now is going to look at a lot more things including, you know, potentially looking at blushes, free fluids, um, comparing organs, and, and things like that. Doctor Murray. Thank you, Bob, and thank you for allowing me, uh, me this protocol violation. Lindsay, it's been great to work with you over the past couple of years. And uh you've done a wonderful job. Uh, you finished off this contrast enhanced ultrasound project which, uh, lesser fellows had stumbled over. And, uh, I look forward to, uh, following your career and torturing you when you return to Florida to finish off those last two papers. Um, and, uh, there's a, um, OK. And there's a uh a book and a certificate for you. Oh, this is heavy. I'm not used to holding textbooks. Thank you. Is that yours? That Yes. partner. OK, thank you. Good morning. Thank you for the opportunity to present some of my work from my time in the Vascular Anomalies Center. If you've ever been to one of our vascular anomalies conferences, you'll know how important nomenclature and terminology is in the field of vascular anomalies. Everything through ISFA, International Study for the Society of Vascular Society for the Study of Vascular Anomalies, has been broken down into tumors and malformations, and there are multiple different named, uh, diagnoses which we try and use within each group. One of my favorite quotes from our infamous textbook, Ambassador Anomalies has been attributed to Confucius. Beginning of wisdom is to call things by their right name. However, many times in vascular anomalies, the challenge I, I think, has been as much in identifying these things as a coherent diagnostic grouping so that the right name can be applied. Our center has been instrumental in the grouping and naming of many of these things, including cloves, fava, And the multiple different subtypes of hepatic hemangiomas. One such diagnostic grouping was identified long before our vascular anomaly Center was in existence. In 1860, Gascoon was one of the first to describe nevi scattered over various parts of the body and limbs, as well as studying the surface of the intestines, with the nevus visible beneath his perineal coat projecting considerably into the cavity of the bowel. However, it was William Bean in 1958 who actually detailed this association of multiple venous malformations in the skin with bleeding from similar anomalies in the GI tract. And it was he who called it by its current name, Blubber blue rubber blood nevus syndrome. The GI bleeding blue rubber blood nevus syndrome is the main complaint of this syndrome for many patients. Chronic anemia caused by this bleeding often leads to lifelong iron replacement and repeated repeated blood transfusions. In addition, these blebs obviously can present an inception risk. This entity has been treated by several methods over the years. However, there are currently two dominant therapies which have been written about by members of our own vascular anomalies Center. In 2005, Doctor Fishman described his initial experience with an aggressive surgical approach for the eradication of these blebs in 10 patients. The surgical therapy involves the identification of blebs with the help of intraoperative endoscopy and removal as many, as many identified blebs as possible. After this therapy, only one patient developed a recurrent blood transfusion requirement while 2 others were found to have recurrent or residual disease. After the surgery was pioneered, there was a medicine that came out, which has revolutionized the treatment of vascular anomalies. In a study which was done by Salu Mal but driven by Dr. Denise Adams, one of the members of our group, serolimus was used in 4 children with blue blood nevus syndrome who are symptomatic despite prior therapy. After serolimus, all had improvement in their lesion size, and the 3 patients with anemia requiring chronic transfusions became transfusion independent. The short-term successes and failures seen in these patients made us question the longer, the long-term effects of operative eradication, given its long-standing utilization of this institution. Therefore, since we had now a much larger group of patients who have undergone this operation, we looked at them using both retrospective review and structured interviews. Of the 62 patients who were actually seen here at Children's for the blue rubber blood nevus syndrome, 28 patients underwent this extensive procedure. Bleeding and anemia started at a young age for these patients at a median of 4 years, often years before a diagnosis of Bluebird blood nega syndrome. 93% of patients required blood transfusions with these patients needing a median of 18 units. Many patients attempted medical therapy or needed intestinal operation prior to the eradicative operation for bleeding or obstruction. Average age of first operation was 12 years, 8 years after the average onset of bleeding. 32% of patients had known lesions left at the end of the operation due to a large number of lesions in their bowel precluding complete resection within a reasonable anesthetic and fluid resuscitative period. Our average follow-up for these patients has been 9.1 years, although some of these patients are now decades out. Only 11% of patients who had complete eradications needed an additional resection or serolimus therapy. Very similar to the 1 of 10 patients or 10% from the original paper. Interestingly, 44 44% of the patients who had incomplete eradications have since been placed on erolimus therapy, mimicking the group of patients from the erroimus study mentioned earlier. The change in pre and postoperative hemoglobins for all these patients has been significant um postoperatively, and the overall upward trend for these patients is shown in the smoothing line. The proportion of patients who required a blood transfusion preoperatively compared to postoperatively, uh, postoperatively compared to preoperatively was only 22% compared to the 93% preoperatively. Overall, these results support the previous study showing overall decreased bleeding after the eradicated operation and a low need for operative or medical interventions in patients in whom all bloods are able to be resected. We did find that serolimus is not a perfect solution for every patient, as one patient who had new bles eventually needed a re-operation when he continued to bleed significantly on serolimus therapy. However, the relatively higher proportion of patients with a large number of bloods who required medical or further surgical therapy after their initial operation seems to support our current practice of trying serolimus therapy in patients with a very large number of lesions or serolimus therapy first in patients with a large number of lesions. While Bluebird blood nevus syndrome has been fairly well described in the literature, there's overall poor characterization of gastrointestinal vascular anomalies in general. The majority of studies in the literature describing these are single-center case reports, and going back to the original quote from this lecture, things were often not called by their right name. The same lesion will be variably called an AVM, hemangioma, cavernous hemangioma, or varices depending on the author. I discovered this firsthand when I went to explore a question brought up during one of our weekly conferences. It was during one of these sessions that we discussed the patient with colonic varices, and of course it came up that we've seen a bunch of these before and should look into them. Colonic varices, whether idiopathic or not, are a rare entity, only occurring in 0.07% of patients at autopsy. Typically, they're seen in patients with portal hypertension, and these are limited to only a section of the colon or the rectum. A liter a literature search revealed a multiple a multitude of case reports, mostly describing either one case or small familial series of these idiopathic clonic varices. These are some pictures from some of these representative case reviews. After the literature review, we set out to find this population within our own vascular anomalies database. While we initially only used a few search terms such as clonic varices, veers sail, etc. the search ended up being expanded as we quickly noted that even in our own center, these entities were often called by different names. We set about characterizing the presentation and treatment strategies used in patients with diffuse reticular venous malformations of the colon or these idiopathic colonic varices. This was done by retrospective chart review and review of radiologic images and pathology of the 20 patients we identified as having this, this vascular anomaly who've been seen at our center. These vascular anomalies are typically found during an evaluation for GI bleeding. Patients were fairly evenly split between presenting with melina and with hematochezia. All patients had anemia, and most had been transfused during their lifetime, although the overall number of these transfusions for these patients was much lower than the prior group at 2 units per patient. Most of this population did not have any any other significant findings, although 10% did have a history of another vascular anomaly. And we did have one family who had uh 3 affected siblings as well as a mother. Sorry, um, bleeding occurred a few times a year and acute self-limited episodes in this in this patient population. Endoscopy almost invariably showed panclonic involvement with terminal ile involvement in the majority. The rectum tended to be relatively spared. Angiographic features for most patients included tortuosity and early truncation of the colonic small arterial vessels, which is shown better in the in the picture on the left, as well as early opacification of colonic veins as seen in the picture on the right. Other mesenteric arteries and veins were normal on imaging and showed antigraded flow. Histology of these lesions shows large thin-walled tortuous venous channels in the colonic submucosa and also muscularis propria. They're not normally significant gross pathologic findings uh on any specimens that we've seen. Endoscopic the sclerotherapy or argon plasma coagulation has primarily been done at this institution um for treatment or for elective treatment of these vessels. It's been done in 60% of patients, although 33% of the patients who are treated did require repeated treatments. Only one patient underwent a hemicolectomy here, but 3 patients ultimately underwent bowel resections at outside institutions. So, to conclude, uh, that project, we saw over bleeding and diffuse reticular vascular malformations in the intestine, which generally happen in acute self-limited episodes. This bleeding is generally enough to cause anemia, but most patients don't have ongoing transfusion requirements, and it's usually seen in typical pancolonic involvement with frequent TI involvement and relative rectal sparing. Best treatment is still currently unknown. Varricelic colon, along with the term cavernous hemangioma, have also been used to describe a diffuse infiltrative intestinal lesion, most often involving the distal colon and rectum. Though they sometimes share the same name in the world literature, this is a completely different entity from the previously discussed diffuse reticular venous malformations. The location of these involves rather than spares the rectum before extending proximately for a variable distance as seen in several of these pictures. You can see the rectum being involved, and then there's usually a termination somewhere in the sigmoid or descending colon. Additionally, many of these patients have engorged tissue in the anal canal which can prolapse. These are just some pictures of operative findings showing the uh transmural involvement. And the involvement of the rectum. We similarly looked at these lesions for characterization and compared them to diffuse reticular venous malformations. These patients almost always presented with a hematochezia and rarely presented with melana. Bleeding was more chronic in this patient population, with almost 25% of patients bleeding daily or weekly. GI bleeding also uh started in much earlier in age, uh, with most, with patients sometimes recurring, uh, reporting bleeding as as early as infancy, although, although this is often attributed to fissures or milk protein allergies. While less patients required blood transfusions, the number of units given for those who needed it was significantly higher. In addition, at least half of these patients have visible external anomalies such as lower extremity, extensive venous malformations, or have uh Klippel-Toy syndrome. Treatments for this cohort have included anorectal sclerotherapy and pull-throughs given the distal burden of disease. So in conclusion, these, uh, these segmental transformmural venous malformations of the colon are very different than the diffuse reticular venous malformations. Over bleeding in these patients tends to occur chronically, um, and involves the rectum and sigmoid rather than the, um, rather than the pan colon with sparing of the rectum. Finally, I wanted to talk about one final project. Just a little near and dear to my heart. While retrospective data collection has been valuable in studying vascular anomalies given the rarity of most of these lesions. Charts and patient recall, and while charts and patient recall do a fair job of providing adequate medical information on disease presentation, workup, treatment, and some clinical outcomes, it's difficult to really glean patient reported outcome information on constructs such as pain, fatigue, and mental health, measures that truly affect quality of life. Though a large amount of research has been done looking at the genetics, clinical presentation and treatment of vascular anomalies, little has been done to examine these patient reported outcomes. The effect of disease and treatment on patients' daily lives is poorly understood by most clinicians. Hundreds of standardized measures have been developed to capture patient-reported outcomes in response to this phenomenon. The systematic use of information from these patient-reported outcome measures or prompts leads to better communication and decision making between doctors, and we can also use these measures for quality improvement and to monitor patient outcomes and progress in response to our treatment measures. A popular patient reported outcome is quality of life. Looking at quality of life in vascular anomalies, we have seen a very few studies. One study assessed the effects of serolimus treatment for patients with complicated vascular anomalies, and this showed significant improvement in QOL scores using the PESQL. Another study looked at the quality of life in patients with lower extremity vascular anomalies using the SF-36. Compared to the general population, these patients reported impaired vitality and higher levels of pain, while other health concepts demonstrated no difference. A similar study was done in patients with Coolani syndrome with the SF 36, and these patients showed significantly lower scores in almost all health concepts. One increasingly popular method for measuring patient-reported outcomes is the administration of these uh NIH health measures, particularly PROMISE measures, which measures self-reported global physical, mental, and social health measures. Specific measures have been developed for um many different constructs in each of these domains. These measures can be administered either as computer adaptive tests or in fixed short forms, and many measures have been created for children in addition to adults, while parent proxy versions also exist for those patients who are not able to participate due to age, disease burden, cognitive or communication deficits. Patients with chronic health conditions have been shown to have poor scores on these measures across all health status domains relative to people without a chronic disease, even after adjustment for socio-demographic factors. These scores have been standardized so that the same scale is used across the different outcomes measures and across age groups, allowing for comparisons. Scores are also relatively easy to interpret in regards to mild, moderate, or severe symptoms for each construct, as the standard deviations are 10 are used for most of these different designations. When looking for a patient group and who to start using these measures, we thought of a group of patients with FAVA or fibro adipose vascular anomalies. FAVA is a relatively newly described diagnostic category, as you may remember from one of my earlier slides describing an intramuscular vascular lesion which leads to painful limbs with associated dysfunction. Multiple modalities have been used to treat these lesions, but most commonly we're using either surgical excision, medical therapy with erolimus, or cryotherapy performed by interventional radiologists. One reason that we chose this diagnosis was, was that these patients already had a patient reported outcome being measured. Through a different, though a different scale was used, patients undergoing cryoablation had their pain pre and post-procedure measured using a brief pain inventory. Though the follow-up period has been relatively short, they noted baseline high pain scores and were able to show improvement in most patients after cryoablation, although this, this clinical response somewhat abated over time. Therefore, with some past data, we thought this would be a good diagnostic group in which to initially pursue some of these outcome measures. While pain was something that we obviously wanted to measure, we wanted to go beyond that and measure functional limitations caused by these diseases. In addition, we've been concerned for some time about how these diseases affect the mental health of our patients. While there's no good data to support it, we suspected a higher instance of depression, fatigue, and anxiety in our patients based on clinical acumen and just the strain of a chronic disease. These are just some representative pictures of what these short forms look like for pain interference, which is one of the constructs that we've been testing, um, both for the pediatric and for the adult population, as well as it's, uh, as it's, uh, in its red cap form, which is how we tend, tend to distribute the surveys. We designed this project so that patients are recruited during clinic visits or hospital visits, and these surveys are administered on recruitment more often in the initial time period at 1 month, 6 months, 1 year, and 18 months, and then yearly after that. If any procedure, if any patient undergoes an intervention, whether it's a pharmaco, whether it's starting a medication, uh, cryoablation, or having an operation, this timeline starts again, so we're able to measure relatively frequently these constructs right after the, um, the intervention is, is, uh, performed. Our aims are to describe these quality of life outcomes at presentation, to prospectively follow them, and to correlate any changes in quality of life measures with clinical changes, as well as to determine the effective treatment on these measures. We're very early in our days. We finally recruited, I think, the 1st 8 patients so far in our studies. And so these are the, these are the main baseline scores for initial patient population, which I hope will continue over time. If you remember, the mean score of the uh the kind of average baseline score for most people should be 50 and uh there's a standard deviation of 10 for these patients which indicates moderate um symptoms. So, looking at these, you can already tell the pain interference or pain interfering with life, at least in our initial patients does seem to be in the moderate category and anxiety is just Just about there at 59.7. Depression is, is, is still not noted, but mobility has been uh severely limited in these patients as most of them have lower extremity disease. These are very early outcomes, but we're excited to be able to follow those over time. I'm hoping that this can make a difference, um, not only in assessing treatment strategies, but hoping to identify potential mental health outcomes that will allow us to, um, To demonstrate that we need more and more resources for these patients beyond just medical and surgical therapy, but also social, um, social work and mental health services for this chronic disease population. Um, I'd like to acknowledge everybody in the Vascular Anomalies Center. Uh, they've taken me in over the last few years. They've helped me out. These are just a few of the people that have been helping with these projects, but there's so many more people that are on this slide. Thank you, Doctor Fisherman, for taking me in over the last couple of years. Thank you to the Weitzman's for helping fund me. Thank you to everybody in my center and everybody who I've worked with. I really appreciate my time. Question for them. Make the presentations at the end. OK. And Thanks for having me. I really appreciate everyone's uh time and, um, uh, just, uh, welcoming me into the uh surgical Critical Care fellowship. I've learned, uh, more than I ever thought I would and I will take, uh, uh, so much of that way, uh, with me, uh, over the years. It's been a pleasure. So I've worked with Doctor Romani on um uh a, a pretty big uh project um looking at the predictors of um mitral valve prosthesis and durability and after the valve replacement in children, we have the largest Patient series in the world for congenital heart disease, uh, for mitral valve replacement, so we wanted to look at it a little bit more closely. Um, some of the big, uh, background points is that a mitral valve replacement, of course, is pretty rare, um, uh, in infants and, and it's most often performed after a failed repair, and, and 30-day mortality has been shown to be anywhere between 10 and 36%, um, depending on what paper you look at, and altered left ventricular geometry based on um Uh, different, um, congenital heart defects, as well as, um, a marker of prosthesis-patient mismatch, uh, placed patients at greater mortality risk. Uh, and in children, um, prosthesis-patient mismatch is, um, usually defined as, uh, an, an oversized valve trying to fit, um, too large of a valve and too small of a child. Um, and, um, prosthesis survival, um, is negatively impacted by patient age and smaller valve size. And so given the poor outcomes seen in patients with a prosthesis patient mismatch in the youngest subset of patients and the lack of mechanical valves available for really anything under valves less than 16 millimeters in diameter. Uh, Boston Children's began performing mitral valve replacements, um, with melody valves, um, in 2010, uh, due to their non-fixed diameter, meaning they can be, um, enlarged in the cath lab, um, as the, the patient has somatic growth. Um. And our aim which is to compare, uh, specifically these melody valve outcomes, um, with the more conventional, uh, valves that have been used in the past. And so, uh, as you, you all are fairly well aware, um, a melody valve is a, it's a bovine jugular vein, um, valve sewn to, uh, a platinum stent. Uh, it comes in one size, it's crimped down to 6 millimeters in diameter, uh, and can be, uh, Expanded, um, as many times as you'd like up to 22 millimeters in diameter and over a period of years, it's been shown. And it was FDA approved in 2010 for percutaneous uh pulmonic valve replacement, usually for uh kids um after a tetrology of flow repair early on that had um incompetent pulmonic valves years down the line. Um, so, um, Our study was a single institution retrospective uh review, uh, looking back, um, over 24 years from, uh, between 1992 and 2016, um, of all the mitral valve replacements that happened here, um, in children, our primary outcomes defining what we, you know, call durability, our freedom from structural valve deterioration. It's Mostly based on um post-operative echocardiographic findings, um, include moderate or greater prosthetic stenosis, regurgitation, and, and para valve leak could also include, um, Thrombosis, valve stent fracture, embolization, leaflet perforation, or entrapment. Some of the important covariants that we included in our study include obviously which type of valve the child, children received, their age, diagnoses of LV hypoplasia, um. The two, main or three, main, uh, ones that we included were diagnosis of Shon syndrome, which is a constellation of left-sided obstructive, uh, lesions, and right dominant AV canal, and any variant of hypoplastic left heart syndrome. And What we did was a Cox regression, uh, repeated events analysis to kind of help us determine the association between time to structural valve deterioration, uh, or a re-replacement, uh, which is our other primary outcome, um, by valve. Um, so, fairly busy slide. The main take home, uh, is that we had, uh, 154 patients who received 213 valves. Um, what's uh Really important, uh, I think to take away from the study as, uh, was that melody valve patients were, um, across the board much younger, and their, the valves sizes that were placed in them were much smaller, and their markers of prosthesis prosthesis patient mismatch were increased. And because of their congenital heart defects and their size and the, the reason why they needed to get uh valves placed in the first place at such a young age, and they had much more prevalent altered left heart geometry. Um. We had out of the 213 valve replacements, 93 events of structural valve deterioration and 61 re-replacements, and what we did, because these patients' populations are clearly divergent, um, in many ways, what we tried to do is restrict our analysis to patients less than 3.2 years of age because that was the oldest child receiving uh a melody valve. Uh, and so this is a survival curve, um, based on for structural valve deterioration, uh, stratified by valve type, uh, and you can see mechanical, um, valves are the black line, and you can, of course, because they're the standard of care, uh, in children, a very, uh, robust, um, follow-up period, uh, with many patients, they have a very clean, uh, survival curve, um, and their freedom from structural valve deterioration. Um, was 66%. Uh, it was 75% for melody valves, which is encouraging, but clearly, uh, we only have, um, about 4 years of follow-up, uh, at this time, or I should say at the time of this analysis, and, and it's encouraging, but we need more time and more patience to prove that this red line, uh, keeps going the way we want it to, and, um, porcine and pericardial valves have performed less well. Restricting the analysis to younger patients less than 3 years of age, you can see, um, melody valves again, um, perform well, uh, compared to the other more conventional valve types. Uh, median times of structural valve deterioration are 4 years for um. Mechanical valves in less than a year for pine and pericardial valves and Melody valves, the structural valve deterioration does not reach 50%. Um, another busy slide, you, you may not be able to see it well, the, the top left hand, um, graph here and plots valve size on the Y axis to patient age at surgery on the x axis. Um, what you can clearly see is that there's a strong positive correlation between the two of them. In the red are the melody valves, and in the black circles are mechanical valves. The youngest and smallest valves, um, Our groups here at the bottom, which we expect. Um, and, um, mechanical valves were in patients that were larger and, um, the bigger valves replaced. Um, right next to it, we have, um, this marker of patient prosthesis mismatch. Um, which is valve size to patient weight ratio. There's a very strong negative correlation between that patient prosthesis-patient mismatch and age at surgery. The older you are, the less of a risk you have of having any sort of prosthesis patient mismatch because you're larger. You can, again, you can see the melody valves are grouped here all the way to the left on the steep end of the curve, meaning they're the youngest patients with the the highest amount of mismatch. Underneath that curve, you can again see um prosthesis patient mismatched graft on the Y axis, and on the X axis is valve size. The larger the valve that is placed is strongly negatively correlated with prosthesis-patient mismatch, meaning that, uh, the, the reason you can put the large valve in is because the patient is larger. Interestingly, these three graphs here, um, show that there's no correlation between preoperative, um, echocardiographic measurement of mitral valve annulus size and patient age at surgery, valve size that was placed, or markers of prosthesis patient mismatch was a little bit surprising to us. Again, to drive the point home, here you see um freedom from structural valve deterioration in mechanical valves stratified by age, valve size placed, and valve size to patient weight ratio. The oldest patients? With the largest valves and the small, smallest amount of prosthesis patient mismatch all had the best outcomes. You can see the same thing for porcine valves. Interestingly, we can't find that necessarily in melody valves, but importantly, something I didn't mention is that structural valve deterioration, um, the most common, um, Finding under that, uh, sort of umbrella is, um, moderate or greater prosthetic stenosis based on echo. That's not in the definition for melody valves because they, that can be addressed in the cath lab, so we did not include that as a definition for melody valves. OK. Just so that this is official and there's p values on here, smaller valve size, larger valve size patient weight ratio. Younger patient age at surgery. And left ventricular hypoplasia are all risk factors for structural valve deterioration, and the hazard of structural valve deterioration goes down, uh, as, uh, the valve size is increased. And Meaning mostly the patient is larger, not that you're fitting in. Larger bowel than a small patient. Uh, the hazard increases for, um, every Half of a unit increase in the valve size patient weight ratio and the hazard decreases for each year the patient gets older after adjusting for all these things, time to postoperative structural valve deterioration still significantly associated with which type of valve you put in. And more specifically, melody and mechanical valves are at a significantly lower covariant adjusted risk of structural valve deterioration than porcine and pericardial valves. And our hope is that with more follow-up and uh greater numbers that um melody valves can also outperform um the uh standard of care which is mechanical valves. Um, again, all the same, uh, sort of, uh, graphs and findings for, uh, our re-replacement analysis. You can see that the tried and true mechanical valves survive a lot longer than the three other, um, valve types. Again, this is restricted now to patients less than 3.2 years of age. You see mechanical valves still are outperforming all the others, but melody valves start to pull away from porcine and pericardial valves in this analysis. Median time to replacement was over 7 years for mechanical, over 4 years for melody valves, 1.5 for porcine valves, and less than 1 year for pericardial valves. Again, freedom from re-replacement. stratified by age. Valve size place and valve size patient weight ratio. Patients who are older, have larger valves placed and have lower markers of prosthesis patient mismatch, uh, have much greater freedom from re-replacement, not surprisingly. The same can be seen for porcine valves, and again, the, for our melody valve uh comparison here, there's, there's not a lot of separation between, uh, um. The two different groups age and. Over under 9 months, um, valve size over under 14 millimeters. And Valve size patient weight ratio over under 2. I think a couple of reasons why is that there are only 30 patients as opposed to over 100 for mechanical. And, as well as all of our patients are fairly tightly grouped in terms of numbers, um, on their age, uh, the valve size that was placed initially in their, um, prosthesis, uh, patient mismatch numbers. Very similar to our structural valve uh deterioration analysis, smaller valves. More patient prosthesis mis mismatch, younger patients and their diagnoses are all uh universally uh associated with uh Risk for replacement. And as uh the valve size increases, your, uh, Risk your hazard of re-replacement decreases, and as the patient prosthesis mismatch, uh, increases, uh, so does the risk of re-replacement, uh, and the inverse is true for age. The older you get, the lower chance you have of needing a re-replacement. And after adjusting for all these things, time to re-replacement is still significantly associated with valve type. Valve type does matter, and The take home from the re-replacement analysis is that the melody valves are uh significantly, um, they have a significantly lower risk of re-replacement than do pericardial valves, and there is a very strong, uh, but non-significant association. Um, with, uh, mechanical valves being better in the rear replacement analysis, but not, uh, significantly so the melody valves, and that, um, we could porcine valves do worse, but again, not quite significantly. I think with longer patient follow-up and, and better numbers, we'll, uh, have uh, a better handle on things, but, uh, to, to drive it home, valve type matters. For both the structural valve deterioration and need for re-replacement, and, after adjusting for all the things we know, uh, that put patients at risk. Um, melody and mechanical valves do better than other bioprosthetic valves across the board, um, and are more durable, um. What I'm so impressed with and with the melody and Outcomes is that The intercortile ranges of patient age and prosthetic valve size, um. Do not overlap at all, uh, when comparing melody to mechanical valves, they perform almost as good in, uh, patients much more at risk for bad outcomes, if that makes sense. And, um, patients receiving melody valves, uh, again, were significantly younger. Uh, requiring placement of much smaller prosthesis, uh, resulting in, uh, much more prosthesis-patient mismatch, um. We are currently updating since 2016, uh, uh, we're updating our database to reanalyze all these things to have even better numbers, uh, and, uh, about 50 more children have received, uh, mitral valve replacements in the last two years. Um, and they have received about 75 valves amongst those 50 children, and there have been 4 more melody valves placed for A total patient series of 200, a total valve series of 300, and melody valves now at 34. There's no larger, um, patient experience in the world. Um We're also gonna obviously be taking a look at melody valves. You don't have to uh anticoagulate long term, and, um, we have, um, but have not analyzed the data yet in terms of, uh, bleeding and thrombosis, uh, events by valve type, uh, and what, um, anticoagulation or antiplatelet platelet regimen the patient is on. Again, thank you so much for your time. I really appreciate it. So what happens to the melody valves that they end up needing to be replaced? How did they deteriorate? How did they deteriorate? Um, not, it's not a unique to melody valves. I think the most common type of structural valve deterioration for all types of valves is having a, um, uh, a gradient across the valve, uh, in diastole, um, so there's a moderate or greater prosthetic stenosis found on echo, um, for all of them, not just, uh, melody, again, um, because, uh, You can expand the melody in the cath lab, and we don't think that's quite as important. And so usually what happens with these melody valves, or again in any valve, but patients that are very small with altered geometry of their left hearts. And have much higher risk of leaflet entrapment, uh, thrombosis, and Even, um, more patient uh centric outcomes, uh, in terms of complete heart block, uh, left circumflex obstruction. There's a, there's a whole lot of reasons why they're at risk to do more poorly and why we're excited that they are performing, uh, quite well for the patient population. Any additional questions for Doctor Upchurch? It seems like you know you really shed some, some light and some insight on the durability over time um of these various types of mitral valves pericardial, the melody, mechanical, forcing, etc. What I'm curious about is, despite the adjustments for the other covariants and using like a COX model to look at like independent multivariable predictors of freedom from SVD and valve re-replacement, it doesn't seem like you're considering the recurrent re-replacements within the same patient. Um, using either a modulated renewal theory, a Nelson-Ahland estimator, instead of like the traditional Kaplan-Meyer curves that you showed us, which only look at first events within the same patient. So my question and my sort of recommendation would be, To shed a little bit more light on the durability over time of the different types of valves, to incorporate into your modeling here, a cumulative hazard that takes into account the serial recurrent time to events, multiple repeated, truly repeated events within the same patient. I think that would be appropriate to do. Uh, I, I, I appreciate that and, and, and we definitely consider that and are sort of reworking a lot of that and, and a lot of the, the more patients that we accumulate. Uh, none of this has quite been published yet, so. Thanks a lot, Patrick. So that, we'll start with Doctor Fishman. I'm gonna go without a microphone so I don't spread my virus. Doctor. Even with Nicole, I can speak very loudly. I, I wanna, um, just thank Eileen and, and I wanna clarify something she said. Um, you know, I think this to me is her PI, um, and, and that's a fallacy. Uh, the fact of the matter is she had a strong motivation to come to Boston, and I had, um, a strong group of people who could, um, benefit from her prior expertise and outcomes analysis and Uh, and, and take advantage of her passion and I can afford to pay her, um, but I made it very clear to her that given the current nature of my life that I would not be able to be her day to day mentor, and I'm sure in private she would confirm that that I have not, um, but there, but she has really been, um, mentored by a family, uh, in the National Center and some beyond. She has worked very closely with people she listed, um, Denise Adams, Belinda Tckney, Victor Fox, Sam Spencer, uh, Roger Shek, lot, lots and lots of people. Um, but in fact she's been part of the family up in back and has actually helped with some of the transition and the leadership of that, uh, and, uh, there's people out there who are mourning, uh, her leaving quite a bit, but we are, um, extremely grateful for all the work that she has done, just a small piece of what, um, that she's shown, she's going to finish all these managers plus more. I'm going to finish editing the ones that I'm involved in, um, and, um, and we're, uh, thrilled that she's gonna be uh training in Montreal and join the family of pediatric surgeons. we got that. You. Well, I think we'll let you touch the microphone. So Patrick called 3 months into his internship and asked to speak to me regarding the fellowship that had been advertised. So I called him back. It was said, can you call me after 9:30 at night? I said, sure, busy day. So I called Patrick up and it led to about an hour-long conversation and it pretty much convinced everything that we know about him. He's an independent thinker, and I think he's been able to trailblaze his own path. I'm speaking with Dr. Amani over his research time. I'm not going to mask my ignorance. I don't understand what he does necessarily. I certainly can't. Um, go into the data specifically looking at, uh, fine tuned. Projects and things, but I can safely say that I think the work you've done with Rahm will serve as a gold standard, and he has emphasized that over and over and over. There's nothing else out there like it. You've been able to essentially construct, populate, interrogate, analyze, and create work that will benefit children for decades to come, and I think that's truly a groundbreaking project for which you have pioneered. And put together on your own, obviously under his guidance, but you've spent many, many hours populating that, and I think you've done a fantastic job. Uh, I will also say that I think uh we're gonna lose a valuable surgical colleague, but anesthesia will absolutely get a powerful performer who will continue on great work. So, Patrick, congratulations and thank you very much for all your hard work. So I think we should have a final round of applause for all of our 3 presenters this morning. Yeah. uh Yeah You have. I see. Great job. uh, it's compared to all the other guys better in those other specific subset of patients. Is all right.
Click "Show Transcript" to view the full transcription (50579 characters)
Comments