Author + information
- Received November 16, 2016
- Revision received February 1, 2017
- Accepted March 1, 2017
- Published online May 8, 2017.
- aChildren’s Hospital of Philadelphia, Division of Cardiology, Philadelphia, Pennsylvania
- bChildren’s Hospital of Philadelphia / Hospital of the University of Pennsylvania, Center for Lymphatic Imaging and Interventions, Philadelphia, Pennsylvania
- ↵∗Address for correspondence:
Dr. Maxim Itkin, Center for Lymphatic Imaging and Interventions, Hospital of University of Pennsylvania, 3400 Spruce Street, Philadelphia, Pennsylvania 19104.
Background Post-operative chylothorax in patients with congenital heart disease is a challenging problem with substantial morbidity and mortality. Currently, the etiology of chylothorax is poorly understood and treatment options are limited.
Objectives This study aimed to report lymphatic imaging findings, determine the mechanism of chylothorax after cardiac surgery, and analyze the outcomes of lymphatic embolization.
Methods We conducted a retrospective review of 25 patients with congenital heart disease and post-operative chylothorax who presented for lymphatic imaging and intervention between July 2012 and August 2016.
Results Based on dynamic contrast-enhanced magnetic resonance lymphangiography and intranodal lymphangiography, we identified 3 distinct etiologies of chylothorax: 2 patients (8%) with traumatic leak from a thoracic duct (TD) branch, 14 patients (56%) with pulmonary lymphatic perfusion syndrome (PLPS), and 9 patients (36%) with central lymphatic flow disorder (CLFD), the latter defined as abnormal central lymphatic flow, effusions in more than 1 compartment, and dermal backflow. Patients with traumatic leak and PLPS were combined into 1 group of 16 patients without CLFD, of whom 14 (88%) had an intact TD. Sixteen patients underwent lymphatic intervention, including complete TD embolization. All 16 patients had resolution of chylothorax, with a median of 7.5 days from intervention to chest tube removal and 15 days from intervention to discharge. The 9 patients with CLFD were considered a separate group, of whom 3 (33%) had an intact TD. Seven patients underwent lymphatic intervention but none survived.
Conclusions Most patients in this study had nontraumatic chylothorax and dynamic contrast-enhanced magnetic resonance lymphangiography was essential to determine etiology. Lymphatic embolization was successful in patients with traumatic leak and PLPS and, thus, should be considered first-line treatment. Interventions in patients with CLFD were not successful to resolve chylothorax and alternate approaches need to be developed.
Post-operative chylothorax in patients with congenital heart disease (CHD) is a challenging clinical problem with substantial morbidity and mortality (1). The incidence of chylothorax after cardiothoracic surgery has been reported between 2% and 5% (2). A recent analysis of the Pediatric Health Information System (PHIS) database found that the overall incidence of chylothorax in pediatric patients after congenital heart surgery or heart transplantation was 2.8%, with an increased incidence from 2.0% in 2004 to 3.7% in 2011 (3). In this PHIS cohort, the procedure codes associated with the highest incidence of chylothorax were cavopulmonary anastomoses (Glenn and Fontan surgeries), repair of transposition of the great arteries, and heart transplantation. Additionally, the development of chylothorax was associated with a significantly longer length of hospital stay (p < 0.0001), increased risk of in-hospital mortality (odds ratio: 2.13), and higher cost of hospitalization (p < 0.0001).
Multiple diagnostic algorithms for chylothorax have been developed and theories have been proposed to suggest the etiologies of traumatic and nontraumatic chylothorax (4). The cause of traumatic chylothorax is direct injury or surgical laceration of the central thoracic duct (TD) or 1 of its lymphatic tributaries. The causes of nontraumatic chylothorax have been reported to include lymphatic malformations, malignancy (e.g., lymphoma), infection (e.g., tuberculosis), extension from chylous ascites, and congenital syndromes (e.g., Down, Noonan, or Turner syndromes) (5).
Management of chylothorax can be challenging and includes both conservative and interventional treatments. The goal of the conservative approach is to reduce intestinal lymphatic flow through dietary modifications (such as a low-fat diet or total parenteral nutrition) and medications (e.g., octreotide or somatostatin) (6,7). If conservative management fails, then surgical procedures such as TD ligation, pleurodesis, and pleuroperitoneal shunts are considered (8,9). The PHIS analysis reported that TD ligation or pleurodesis was performed on patients a median of 18 days (interquartile range [IQR]: 7 to 28 days) after the cardiac procedure and patients were discharged from the hospital a median of 22 days (IQR: 10 to 47 days) after surgical treatment of chylothorax (3). More recently, percutaneous TD embolization has emerged as a minimally invasive alternative for the treatment of chylothorax (10,11).
One of the difficulties in determining the etiology of chylothorax has been the lack of methods to image the central lymphatic system. Dynamic contrast-enhanced magnetic resonance lymphangiography (DCMRL) is a new imaging technique that uses an intranodal injection of gadolinium-based contrast agents to visualize the anatomy and flow characteristics of the central lymphatic system, with good spatial and temporal resolution (12,13). DCMRL recently showed abnormal pulmonary lymphatic flow from the TD toward the lung parenchyma and/or lymphatic perfusion of the mediastinum in patients with single ventricle physiology and plastic bronchitis; this was termed pulmonary lymphatic perfusion syndrome (PLPS) (14). Percutaneous embolization of this abnormal lymphatic flow resulted in symptom resolution for the majority of patients.
For several years, our institutional approach to post-operative chylothorax has been to perform magnetic resonance lymphangiography on all patients before any lymphatic intervention or procedure in order to define the lymphatic anatomy and determine the mechanism of chylothorax. The objective of this study was to report the lymphatic imaging findings and outcomes of percutaneous lymphatic embolization for the treatment of chylothorax in patients with CHD.
This study is a retrospective analysis of patients with CHD and post-operative chylothorax who presented to our institution for lymphatic imaging and intervention between July 2012 and August 2016. Permission from our institutional review board was obtained before study initiation.
Data collection included patient demographics, cardiac diagnoses, surgical histories, prior therapies, weight-adjusted volume of chest tube drainage, imaging findings, results of lymphatic interventions, and the clinical course post-intervention. The diagnosis of chylothorax was established by the presence of a high percentage of lymphocytes (>70%) and/or a high concentration of triglycerides (in patients on a regular fat-containing diet) in the pleural fluid. In some cases, a high-fat food challenge test was performed and, if the concentration of triglycerides increased, then the diagnosis of chylothorax was confirmed.
The primary endpoints of the study were resolution of chylothorax and patient survival. The secondary endpoints were the weight-adjusted volume of chest tube drainage during the 7 days before intervention compared to the 7 days after intervention and the duration of chylothorax before intervention compared to the duration of chylothorax post-intervention (with resolution of chylothorax defined as removal of all chest tubes).
Lymphatic imaging and intervention
Imaging and interventions were performed in an XMR suite that combines a magnetic resonance image (MRI) scanner with a cardiac catheterization laboratory. All procedures were performed under general anesthesia. Patients initially underwent DCMRL using the technique previously described by Dori et al. (14). Briefly, the inguinal lymph nodes were first accessed under ultrasound guidance in the catheterization laboratory, using a 25-gauge spinal needle. A small amount of water-soluble iodinated contrast agent was injected to confirm the position of the needle inside the lymph nodes by fluoroscopy; then, the needle was secured in place.
Patients were then transported to the adjacent MRI suite equipped with a 1.5-T scanner. Heavy T2-weighted MRI lymphatic imaging using a respiratory-navigated and cardiac-gated 3-dimensional turbo spin echo sequence was completed. For DCMRL, a weight-based amount of gadobutrol was injected into each inguinal lymph node. One min after the injection, scanning was initiated using a syngo time-resolved angiography with interleaved stochastic trajectories sequence that acquired 1 set of images every 20 to 40 s over 10 to 15 min. This was followed by additional scans with a high-resolution navigator-gated 3-dimensional flash inversion-recovery sequence. At completion of the MRI, patients were transported back to the cardiac catheterization laboratory for lymphatic intervention.
Lymphatic interventions were performed as previously described by Dori et al. (14), Itkin et al. (15), and Nadolski et al. (16). The interventions included intranodal lymphangiography and transabdominal thoracic duct access, followed by TD embolization using a combination of ethiodized oil (Lipiodol), endovascular coils, and n-butyl cyanoacrylate glue. All patients were then transported to the cardiac intensive care unit for post-procedure monitoring.
Statistical analyses were performed using SPSS Statistics version 22.0 (IBM, Armonk, New York). The baseline characteristics of each group were reported as numeric values with percentages or median values with 25% to 75% IQRs. Continuous variables were compared between the groups using the Wilcoxon rank sum test and categorical variables were compared using the chi-square or Fisher exact tests. The weight-adjusted volume of chest tube drainage in the 7 days before intervention was compared to the weight-adjusted volume of chest tube drainage in the 7 days after intervention (within each group), using the Wilcoxon signed rank test.
A total of 25 patients with CHD and post-operative chylothorax who presented to our institution for lymphatic imaging and intervention were included in the study. Patient demographics, cardiac diagnosis, surgical history, and prior unsuccessful therapies are listed in Table 1. The details of lymphatic imaging, etiology of chylothorax, and type of intervention are described in Table 2. A total of 24 patients had T2-weighted MRI to assess for soft tissue edema, 23 patients had DCMRL to identify lymphatic flow patterns, and all 25 patients had intranodal lymphangiography. Lymphatic intervention was performed in 23 of 25 patients (92%). In 15 of these 23 cases (65%), the imaging studies and the lymphatic intervention were performed on the same day. The procedures were technically successful and the TD was able to be engaged in 16 of the 17 patients (94%) who had an intact TD.
Based on the imaging results, we identified 3 groups of patients with distinct etiologies of post-operative chylothorax: 2 patients (8%) with traumatic leak from a branch of the TD (Figure 1), 14 patients (56%) with PLPS (Figure 2), and 9 patients (36%) with central lymphatic flow disorder (CLFD) (Figure 3). PLPS has been described as a condition with abnormal pulmonary lymphatic flow from the TD toward the lung parenchyma through abnormal lymphatic networks in the chest (14). We now define CLFD as a condition with abnormal (reduced or absent) central lymphatic flow, effusions in more than 1 compartment, and the presence of dermal backflow through lymphatic collaterals in the abdominal wall. In our cohort of patients, this constellation of imaging findings that are diagnostic of CLFD was only observed in infants younger than 1 year.
For the statistical analyses, the 2 patients with traumatic leak and the 14 patients with PLPS were combined into 1 group of 16 patients “without CLFD.” The other 9 patients were considered a second group “with CLFD.” Baseline characteristics before intervention indicated that the 9 patients with CLFD were significantly younger in age (median: 0.3 years; IQR: 0.3 to 0.5 years) than the 16 patients without CLFD (median: 1.9 years; IQR: 0.5 to 4.3 years; p = 0.004) (Table 3). The patients with CLFD had a significantly higher incidence of concurrent ascites (9 [100%] vs. 5 [31%]; p = 0.001) and a higher incidence of concurrent pericardial effusion (7 [78%] vs. 4 [25%]; p = 0.016), compared to the patients without CLFD. Patients with CLFD also showed more transposition of the great arteries and more genetic syndromes (such as Noonan syndrome and trisomy 21), but these trends were not statistically significant.
In the 23 total patients who underwent lymphatic intervention, there was no difference in the median weight-adjusted volume of chest tube drainage between both groups during the 7 days before intervention (p = 0.154) (Table 4). However, there was a significant difference in the median weight-adjusted volume of chest tube drainage between both groups during the 7 days after intervention (p = 0.027). The average weight-adjusted volume of chest tube drainage per day for patients with and without CLFD on each of the 7 days before intervention and the 21 days after intervention is graphically represented in the Central Illustration.
Regarding the primary outcomes for the 23 patients who underwent lymphatic intervention, there was resolution of chylothorax in all 16 patients (100%) without CLFD compared to only 1 of the 7 patients (14%) with CLFD (p < 0.0001). All 16 patients without CLFD survived from intervention to hospital discharge whereas none of the 7 with CLFD survived from intervention to hospital discharge (p < 0.0001).
Of the 16 patients without CLFD, 14 had an intact central TD, 1 patient did not have an intact TD due to TD ligation, and 1 patient had congenital absence of the TD. All 16 patients without CLFD underwent lymphatic intervention, including selective lymphatic duct embolization and complete TD embolization (Table 2).
Regarding the outcomes for the 16 patients without CLFD who underwent lymphatic intervention, median chest tube drainage decreased significantly from 7 days before (21 ml/kg/day; IQR: 10 to 35 ml/kg/day) to 7 days after intervention (11 ml/kg/day; IQR: 5 to 20 ml/kg/day; p = 0.008) (Table 4). All 16 patients without CLFD had resolution of chylothorax after intervention, with a median of 7.5 days (IQR: 4.3 to 16 days) from intervention to chest tube removal. All patients without CLFD survived, with a median of 15 days (IQR: 6.0 to 36 days) from intervention to hospital discharge.
Of the 9 patients with CLFD, there were 3 patients with an intact central TD (but significantly reduced antegrade flow), 4 patients without an intact TD due to TD ligation, and 2 patients with congenital absence of the TD. Seven of the 9 patients with CLFD underwent lymphatic intervention, including Lipiodol-only embolization and complete TD embolization (Table 2).
Regarding the outcomes for these 7 patients with CLFD who underwent lymphatic intervention, there was no difference in chest tube drainage from 7 days before (median: 31 ml/kg/day; IQR: 27 to 61 ml/kg/day) to 7 days after intervention (median: 38 ml/kg/day; IQR: 16 to 111 ml/kg/day; p = 0.866) (Table 4). Only 1 of the 7 patients (14%) with CLFD had resolution of chylothorax after intervention (but subsequently died of an unrelated cause) and 0 of the 7 patients with CLFD survived from intervention to hospital discharge (median: 94 days; IQR: 33 to 187 days) from intervention to death.
The causes of death for patients with CLFD after lymphatic intervention included multisystem organ failure in 3 patients, and the following in 1 patient each: multisystem organ failure in the setting of sepsis; progressive CLFD (worsening effusions, ascites, anasarca) with withdrawal of care: unrelated stroke and intracranial hemorrhage (about 5 months after lymphatic intervention); and cardiac arrest due to ventricular fibrillation (about 3 months after lymphatic intervention).
Procedural risks and complications
In addition to being effective for specific patients, percutaneous lymphatic embolization is also a safe, minimally invasive procedure. The median procedure time for lymphatic interventions in our cohort was 220 min (IQR: 155 to 276 min). Regarding the risk of radiation exposure, the median total air kerma dose was 577 mGy (IQR: 210 to 1,205 mGy) and the median total dose-area product was 1,491 μGy/m2 (IQR: 524 to 2,844 μGy/m2). Only 2 cases exceeded the skin dose threshold of 2,000 mGy at a single site (from a single camera) at which skin injury has the potential to occur (17). However, both these patients had a skin dose <5,000 mGy and no radiation-induced skin injury was seen.
The minor symptoms that were most commonly experienced after lymphatic intervention included mild abdominal pain, transient abdominal distention, isolated fever without infection, and brief hypotension that did not require inotropic support. Post-procedural complications included systemic inflammatory response syndrome (SIRS) in 4 patients, transient hypotension requiring inotropic support in 3 of those 4 SIRS patients, and pulmonary edema in 1 patient. The observed SIRS cases were possibly due to an inflammatory reaction to the Lipiodol. No other major complications, including pancreatitis, sepsis, hemorrhage, or stroke related to the procedure, were observed in our cohort.
Chylothorax after CHD surgery can have high morbidity and to date has been poorly understood. In this study, we used MRI and conventional lymphangiography to determine the etiology of chylothorax in patients with CHD. The primary endpoints of chylothorax resolution and patient survival were statistically significant in patients with PLPS or traumatic leak (those without CLFD) compared to patients with CLFD. All 16 patients without CLFD had resolution of chylothorax after lymphatic intervention and survived to hospital discharge. However, only 1 of the 7 patients with CLFD who underwent intervention had resolution of chylothorax and none of the 7 survived to discharge.
This striking difference in patient outcomes highlighted the importance of developing a thorough understanding of the etiology of post-operative chylothorax. Lymphatic imaging techniques, such as DCMRL and intranodal lymphangiography, have proven critical in defining the anatomy and flow characteristics of the central lymphatic system before intervention in both patient populations. Both imaging modalities together can differentiate between causes of chylothorax, such as PLPS and CLFD, and, more importantly, predict the outcome of the disease process.
Lymph from the lower extremities, liver, and intestine normally enters the TD at the level of the cisterna chyli in the upper abdomen, and then is transported via the TD to the innominate vein. Phang et al. (18) performed a systematic review to describe the many variations in TD anatomy. This anatomic variability has significant clinical implications for treating chylothorax, especially in patients with CHD. It also explains why traditional surgical procedures, such as TD ligation, have inferior results (19,20).
In our cohort, 5 patients without CLFD and 5 patients with CLFD had a previous procedure for chylothorax. All 10 of these previous procedures were clinically ineffective, given that the patients presented to us with persistent chylothorax. Five of 10 procedures were also technically unsuccessful, because those 5 patients had an intact TD by DCMRL. Now that imaging modalities exist to visualize the central lymphatic system and guide lymphatic procedures, image-directed lymphatic interventions have largely replaced surgical TD ligation, pleurodesis, and pleuroperitoneal shunts.
Etiology of post-operative chylothorax
In this study, we found imaging to be fundamental in determining the etiology of post-operative chylothorax, especially since the majority of patients had nontraumatic chylothorax due to abnormal pulmonary lymphatic perfusion and complex central lymphatic anatomy. In our cohort, there were only 2 patients with traumatic leak, but even in these patients, the central TD was intact (Figure 1). Their leaks were from traumatic injury to an accessory lymphatic vessel connected to the lung or from a dilated branch of the TD connected to the pericardium, a deviation from our previous understanding of post-operative chylothorax in patients with CHD, where surgical trauma to the central TD was thought to be the primary problem. The TD is typically located posterior to the pericardium and therefore should not be injured during cardiac surgery, unless the aortic arch is involved, because the aorta is in the proximity of the distal TD.
Recently, PLPS was shown to be the underlying lymphatic abnormality in patients with single ventricle physiology and plastic bronchitis (14). It has been observed that a history of chylothorax is a risk factor for development of plastic bronchitis, but in the past, this association was poorly understood (21,22). In our cohort, we found similar imaging patterns of PLPS in all 14 patients with persistent chylothorax that was not due to either traumatic leak or CLFD (Figure 2), and 2 of these patients had a concurrent diagnosis of plastic bronchitis. Therefore, it appears both chylothorax and plastic bronchitis have the same underlying lymphatic abnormality of PLPS, which can present as either post-operative chylothorax or plastic bronchitis, depending on the clinical context.
We defined CLFD as a condition with abnormal central lymphatic flow, effusions in more than 1 compartment, and the presence of dermal backflow through lymphatic collaterals in the abdominal wall. The retrograde lymphatic flow through networks of lymphatic collaterals can extend to the genitalia and lower extremities (Figures 3 and 4). In our cohort of patients with CLFD, the types of central lymphatic flow abnormalities included congenital absence of the TD, anatomic TD outlet obstruction, and absence of central TD flow due to TD ligation. DCMRL is very sensitive to diagnose decreased antegrade central lymphatic flow, but the additional finding of retrograde flow to the genitalia and lower extremities can also be seen by intranodal lymphangiography. Interestingly, the diagnosis of CLFD was only observed in infants younger than 1 year and 4 patients had genetic syndromes, such as Noonan syndrome, which are known to be associated with chylothorax.
The dermal decompression through lymphatic collaterals seen by DCMRL appears to be the hallmark of CLFD. The redirection of lymphatic fluid through cutaneous tissues results in the rapid reconstitution of lymphatic flow in the chest and the development of chylous effusions and ascites, which are difficult to treat. For these reasons, interventions on the central lymphatic system in these patients often fail.
There were 2 patients with CLFD (with or without intervention) who had resolution of chylothorax and chest tube removal: Patient #23 and Patient #24. Patient #23 was found to have occlusion of the lymphovenous junction with no normal connection from the TD to the innominate vein, resulting in CLFD with significant dermal decompression (Figures 4A and 4B). Surgical lymphovenous anastomosis (LVA) was performed twice. Two weeks after the second LVA, repeat lymphangiography showed a patent lymphovenous junction, more antegrade central lymphatic flow, less dermal backflow, and less soft tissue edema (Figure 4C). About 3 months after the second LVA, the chylothoraces resolved and chest tubes were removed. Unfortunately, the patient died of an unrelated stroke and subsequent intracranial hemorrhage before discharge.
Patient #24 intentionally did not have any lymphatic intervention performed and the chylothorax was managed conservatively with diet modification and sildenafil. The patient had spontaneous resolution of chylothorax after a prolonged period of chylous drainage and was discharged home about 9 months after the initial cardiac surgery.
Treatment approaches for chylothorax based on etiology
The overall results of this study suggested that the etiology of post-operative chylothorax, as defined by pre-procedural imaging, predicts the likelihood of patient survival and clinical success following percutaneous lymphatic embolization. Therefore, different treatment approaches are required based on the type of underlying lymphatic abnormality. Patients with post-operative chylothorax due to traumatic leak or PLPS can be effectively treated with percutaneous lymphatic embolization, and thus lymphatic intervention should be considered first-line treatment for these patients (similar to the treatment for plastic bronchitis).
Conversely, lymphatic interventions to occlude lymphatic networks in patients with CLFD have not been successful to resolve chylothorax thus far. In patients with CHD who are born with effusions, performing surgery might theoretically disrupt their equilibrium, reveal a previously undiagnosed lymphatic flow disorder, and result in clinical instability. Any procedure or intervention could potentially be problematic and lead to significant worsening of their condition. Therefore, any infants with unexplained effusions, ascites, or edema should be considered for screening with lymphatic imaging before surgery.
In our cohort, patients often presented to our institution after previous procedures for chylothorax were unsuccessful at referring hospitals. They did not routinely have MRI performed before their initial procedure, so it is possible that these patients had findings of underlying CLFD before their cardiac surgery. We hypothesize that surgical procedures for chylothorax, such as TD ligation and pleurodesis, often make patients with CLFD significantly worse and should be avoided. Conservative management with diet modification and medications should be continued. Alternative approaches to this disorder, such as surgical LVA or TD externalization, must be explored.
This study was a retrospective analysis and the results might differ from a prospectively evaluated cohort. The study was subject to sampling bias because only patients referred for lymphatic imaging and intervention at our institution were included. There might have been different imaging findings in the patients with CHD and post-operative chylothorax who improved quickly with conservative management and did not require any lymphatic imaging or intervention. Therefore, this study’s generalizability is limited to patients with CHD and persistent chylothorax.
In addition, most patients without CLFD had their cardiac surgery performed at our institution, whereas the majority of patients with CLFD were referred from other hospitals, specifically for lymphatic intervention. The potential delay in transferring patients with CLFD to our institution might have contributed to the fact that patients with CLFD had a longer duration of time from the start of chylothorax to intervention than patients without CLFD. However, the 3 patients with CLFD who had cardiac surgery performed at our institution had a similar time to intervention and similar outcome compared to the rest of that group.
Most CHD patients with post-operative chylothorax in this study had nontraumatic chylothorax and lymphatic flow abnormalities, such as PLPS or CLFD. DCMRL and intranodal lymphangiography are essential to determine the etiology of chylothorax and guide the type of lymphatic intervention and should be performed in all patients with persistent chylothorax before intervention. Percutaneous lymphatic embolization was successful in patients with traumatic leak and PLPS, with timely resolution of chylothorax in all patients; therefore, this should be considered first-line treatment for these patients. Lymphatic interventions to occlude lymphatic networks in patients with CLFD were not successful to resolve chylothorax and most of these patients died. Surgical procedures, such as TD ligation, should be avoided in patients with CLFD and further research is needed to develop alternate approaches to treatment.
COMPETENCY IN MEDICAL KNOWLEDGE: Post-operative chylothorax in patients with congenital heart disease can result from traumatic leakage from a branch of the thoracic duct, PLPS with flow from the thoracic duct through abnormal lymphatics, or central lymphatic flow disorder with attenuated central lymphatic flow, multicompartmental effusions, and backflow through abdominal wall lymphatic collaterals. The most common cause in patients with congenital heart disease is PLPS.
COMPETENCY IN PATIENT CARE: In patients with persistent chylothorax after cardiac surgery, lymphatic imaging, such as DCMRL, might help identify etiology, guide percutaneous intervention, and predict outcomes. Percutaneous lymphatic embolization is generally safe and effective for patients with CHD and persistent post-operative chylothorax caused by traumatic leak or PLPS.
TRANSLATIONAL OUTLOOK: Further investigation is needed to determine the optimal timing of lymphatic imaging and intervention in patients with persistent chylothorax after surgery for CHD.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Itkin and Dori contributed equally to this work.
This study was presented as an oral abstract at the American Heart Association’s Scientific Sessions on November 14, 2016.
- Abbreviations and Acronyms
- congenital heart disease
- central lymphatic flow disorder
- interquartile range
- magnetic resonance imaging
- pulmonary lymphatic perfusion syndrome
- thoracic duct
- Received November 16, 2016.
- Revision received February 1, 2017.
- Accepted March 1, 2017.
- 2017 American College of Cardiology Foundation
- Mery C.M.,
- Moffett B.S.,
- Khan M.S.,
- et al.
- Tutor J.D.
- Fogg K.L.,
- DellaValle D.M.,
- Buckley J.R.,
- Graham E.M.,
- Zyblewski S.C.
- Yeh J.,
- Brown E.R.,
- Kellogg K.A.,
- et al.
- Krishnamurthy R.,
- Hernandez A.,
- Kavuk S.,
- Annam A.,
- Pimpalwar S.
- Dori Y.,
- Keller M.S.,
- Rome J.J.,
- et al.
- ↵Cutaneous Radiation Injury: A Fact Sheet for Clinicians. U.S. Department of Health & Human Services, Centers for Disease Control and Prevention (CDC). December 2015. Available at: https://emergency.cdc.gov/radiation/criphysicianfactsheet.asp. Accessed January 31, 2017.
- Phang K.,
- Bowman M.,
- Phillips A.,
- Windsor J.
- Bang J.H.,
- Kim S.H.,
- Park C.S.,
- Park J.J.,
- Yun T.J.