Bringing doctors together

or

A free, private network for verified physicians

Leading voices

Dr. Nishant Shah

Preventive Cardiology · Duke Hospital

View

Dr. Amit Khera

Preventive Cardiology · UT Southwestern

View

Dr. Martha Gulati

Preventive Cardiology · Houston Methodist

View

Dr. Pradeep Natarajan

Preventive Cardiology · MGH

View

Dr. Christopher Cannon

Preventive Cardiology · BWH

View

Dr. Nicholas Marston

Preventive Cardiology · BWH

View

Dr. Dorairaj Prabhakaran

General Cardiology

View

Dr. Pugazhendhi Vijayaraman

EP · Geisinger

View

Dr. Robert Schaller

EP · HUP

View

Dr. Trey Woods

General Cardiology · Emory Healthcare

View
  • Discover your community

  • Stay current

  • A profile that represents you

  • Find your people

About RoonPrivacyTerms

    Overview

    Percutaneous Thoracic Duct Externalization and Autotransfusion in Patients With Lymphatic Failure Due to Elevated Central Venous Pressure.

    Sep 4, 2025DOI: 10.1161/circinterventions.125.015336Journal ArticleAHA

    Written by: Yoav DoriErin PintoLauren BirocMatthew J GillespieRyan CallahanJessica TangMichael L O'ByrneBrooke FordDanish VaiyaniAaron G DeWittDenise MerrillDigvijay ShindeEmmanuelle FavillaRachel ShustakFernando A EscobarGanesh KrishnamurthyAbhay SrinivasanJonathan J RomeChristopher L Smith

    Methods

    This is a retrospective review of treatment outcomes for patients who underwent TD externalization and autotransfusion from January 1, 2018, to August 1, 2022. This study was approved by an institutional review committee and was approved by our institutional review board. The authors declare that all supporting data are available within the article and its Supplemental Material. Patient demographic data, including age, cardiac diagnosis, comorbidities, prior cardiac surgical history, complications, laboratory data, and input/output data, were collected. Data from dynamic contrast-enhanced magnetic resonance (MR) lymphangiography (DCMRL), T2-weighted imaging, lymphatic embolization procedures performed, and clinical outcomes were also analyzed. The indication for externalization in all cases was severe, life-threatening, multicompartment lymphatic disease that was unresponsive to medical management.The primary end points of the study were successful TD externalization with autotransfusion and subsequent decannulation. The secondary end points included the weight-adjusted output of all liquid drainage, including chest tubes, peritoneal drains, watery diarrhea, and urine output. Additionally, all administered supplements including blood, fresh frozen plasma, 5% or 25% albumin, intravenous fluid, and platelets during the 3 days before externalization compared with the last 3 days of the autotransfusion if autotransfusion was done for less than a week or the average of 3 days after a week of the autotransfusion. Laboratory end points including creatinine, glomerular filtration rate, patient weight, absolute lymphocyte count, and calcium levels. Values on the day of the autotransfusion were compared with those recorded on the last day of the autotransfusion or after 1 week of the procedure.Externalization procedures were performed under general anesthesia. For transvenous retrograde, TD externalization access was obtained in the left basilic vein with a 4F sheath under ultrasound guidance. The TD outlet was crossed retrograde with a microcatheter and 0.018″ wire, and the microcatheter was advanced inferiorly to the cisterna chyli. The wire was then exchanged for an 0.018″ V-18 wire (Boston Scientific, Marlborough, MA), which was positioned in the cisterna chyli. A 4F, 45 cm Flexor sheath (Cook Medical, Bloomington, IN) was then advanced retrograde and positioned below the diaphragm at the level of the cisterna chyli.For the transcervical approach, the distal TD was filled with iodinated contrast or Lipiodol (Guerbet, Villepinte, France). Using fluoroscopy guidance, a 25-G spinal needle was advanced transcervically to puncture the distal TD. A 0.010″ Transend wire (Boston Scientific, Marlborough, MA) was then inserted through the needle and positioned in the TD. The needle was removed, and a Renegade microcatheter and 0.018″ V-18 wire were advanced over it and positioned in the cisterna chyli. The microcatheter was then removed, and a 4F or 5F Flexor sheath was advanced over the wire and positioned in the cisterna chyli.
    After externalization, patients were connected to an Arrow autotransfusion drainage chamber (Getinge, Wayne, NJ). The collected fluid was autotransfused via clinical intravenous tubing and an intravenous pump into a central line (Figure S1). The pump rate was adjusted manually each hour based on previous hour’s drainage into the Arrow chamber. The duration of autotransfusion was determined by clinical indication.
    All MR imaging was performed using a 1.5-Tesla MR scanner (Siemens Healthineers, Malvern, PA). DCMRL was conducted as previously described.15 Briefly, all access points—including bilateral inguinal lymph nodes (intranodal), liver (intrahepatic), and mesentery (intramesenteric)—were cannulated using 25 gauge needles under ultrasound guidance. Proper access was confirmed with iodinated contrast injection under fluoroscopy.
    Imaging began with a heavily T2-weighted noncontrast turbo spin-echo sequence. This was followed by the slow injection of gadobutrol (Gadavist, Bayer Healthcare) at 0.5 to 1.0 mL per minute at 1 access site. Time-resolved MR angiography was initiated 1 minute after injection, with imaging sequences performed at 20- to 30-second intervals over a total duration of 7 to 15 minutes. Finally, a navigated T1-weighted 3-dimensional spoiled gradient-echo sequence was obtained.
    CVP was measured during cardiac catheterization at the time of externalization using a fluid-filled end-hole catheter positioned in the right atrium or the Fontan conduit. For the single patient with Glenn circulation, CVP was recorded as the lower value between right atrial and superior vena cava pressures.
    Statistical analysis was performed using SPSS version 27 (IBM, Armonk, NY). Laboratory data—including serum creatinine, glomerular filtration rate, weight, supplemental fluid administration, liquid drainage volume, urine output, absolute lymphocyte count, and calcium levels—were compared using the nonparametric related samples Wilcoxon signed-rank test. Categorical data, such as pulmonary edema incidence, were analyzed using the McNemar Change test for nonparametric related samples.

    Results

    The evaluation included 8 patients (P1–8). Five patients (62.5%) had right heart failure secondary to congenital heart disease, whereas 1 of 8 (12.5%) had a lymphatic conduction disorder due to trisomy 21, 1 of 8 (12.5%) had a lymphatic conduction disorder due to both Noonan syndrome and single ventricle heart disease, and 1 of 8 (12.5%) presented with severe chronic lung disease due to prematurity (Table 1).
    Demographics and Medical History
    CVP was elevated in all patients, with a median of 15.5 mm Hg and a range of (range, 12–28 mm Hg). Lymphatic diagnosis included plastic bronchitis in 1 of 8 (12.5%), pleural effusions in 3 of 8 (37.5%), protein-losing enteropathy in 5 of 8 (62.5%), ascites in 6 of 8 (75%), and anasarca in 8 of 8 (100%). All patients presented with fluid overload and were all critically ill, having failed medical management, including maximal diuretic therapy. The median follow-up duration following TD externalization was 199 days (range, 47–1377 days).
    All patients demonstrated multicompartment lymphatic failure as evident by T2-weighted MRI and DCMRL. All the 5 of 8 (62.5%) patients with protein-losing enteropathy leaked contrast into the duodenum with intrahepatic DCMRL (Figure 1A). Additionally, the 4 of 8 (50%) patients with thoracic involvement all had pulmonary lymphatic perfusion syndrome on intranodal DCMRL (Figure 1B).
    Dynamic contrast lymphangiography (DCMRL) in patients with heart disease and lymphatic failure. A, A maximum intensity projection (MIP) coronal projection of intrahepatic (IH)-DCMRL in a patient with protein losing enteropathy (PLE) demonstrating a leak into the duodenal lumen (arrow). B, A MIP coronal projection of intranodal (IN)-DCMRL demonstrating bilateral mediastinal perfusion (arrow).
    Indications for the procedure are summarized in Table 1. All patients had multicompartment lymphatic failure. One patient of 8 (12.5%) underwent drainage due to multiorgan system failure and fluid overload, whereas 1 of 8 patients (12.5%) was drained due to multiorgan system failure complicated by respiratory failure. Four of 8 patients (50%) had severe respiratory failure and were unstable despite conventional ventilatory strategies, whereas 3 of 8 patients (37.5%) underwent drainage due to severe fluid overload not responding to conventional diuretic therapies. Drain outcomes are summarized in Table 2.
    Drainage and Autotransfusion Duration and Outcomes
    Seven patients (87.5%) were externalized with a 4F flexor sheath, whereas 1 of 8 (12.5%) was externalized with a 5F sheath. Five patients (62.5%) were externalized using the trans-cervical approach and 3 of 8 (37.5%) were externalized using the transvenous approach. The median age at externalization was 8.2 years (range, 0.6–19.3 years), and the median days of autotransfusion was 11.5 (range, 6–126 days).
    The strategy was used as a bridge to a lymphatic embolization procedure in 3 of 8 (37.5%) patients, to a TDD in 3 of 8 (37.5%) patients, and to lymphovenous anastomosis in 1 of 8 (12.5%) patient. Seven patients (87.5%) were successfully decannulated, and 1 patient (12.5%) died due to a bradycardic arrest and was not decannulated. Overall, there were 4 of 8 (50%) nonprocedural deaths due to various causes (Table 2).
    Four patients (50%) developed central venous catheter-associated bloodstream infections.
    The 2 patients with the longest externalization durations (P4 and P5) experienced multiple infections, with a median infection rate of 0.5 episodes per patient (range, 0–4). No detectable drain infections were observed. However, drain occlusion occurred multiple times in patients with prolonged externalization, requiring tPA (tissue-type plasminogen activator) administration. The median number of tPA treatments was 1 (range, 0–12). A total of 4 additional procedures were required for drain replacement, and 3 procedures were performed for drain repositioning.
    Patient clinical outcomes are summarized in Table 3. There was a significant reduction in serum creatinine levels, from a median of 0.63 mg/dL (range, 0.3–2.4 mg/dL) to 0.36 mg/dL (range, 0.16–0.8 mg/dL; P=0.017; Figure S2A and S2B). Additionally, there was a notable increase in GFR, urine output, and absolute lymphocyte count. There was also a significant reduction in patient weight, from a median of 26.8 kg (range, 9.2–143.7 kg) to 20.6 kg (range, 7–118.6 kg; P=0.012). The volume of drainage decreased significantly, from 76.1 mL/kg per day (range, 13.5–295.9 mL/kg per day) to 24.8 mL/kg per day (range, 0–105.5 mL/kg per day; P=0.012).
    Clinical Outcomes
    The amount of supplemental fluids required also decreased significantly, from 36.5 mL/kg per day (range, 1.5–508.2 mL/kg per day) to 7.4 mL/kg per day (range, 1.3–268 mL/kg per day; P=0.012; Figure S2B). Additionally, 6 out of 8 patients (75%) demonstrated improvement in pulmonary edema (P=0.031).
    In patient P2, renal function remained stable for a short period following TDD; however, due to persistently elevated ventricular end-diastolic pressure, renal dysfunction recurred within weeks of the procedure (Figure S1B).

    Discussion

    This report summarizes our experience with 8 patients who underwent TD drainage and autotransfusion (internal drainage using an external chamber) for lymphatic failure secondary to elevated CVP of various causes. All patients presented in life-threatening conditions after failing all available medical management strategies and were too unstable for further surgical or interventional therapies.
    Our findings demonstrate that draining the lymphatic system into an external low-pressure collecting chamber with subsequent fluid reinfusion into the venous system resulted in improved clinical symptoms, including reduced fluid drainage, weight loss, improved renal function, and resolution of pulmonary edema. These results suggest that, in critically ill patients experiencing fluid overload due to elevated CVP, TD drainage with autotransfusion may lead to clinical and fluid status improvement, allowing for more definitive interventions.
    There are multiple causes of elevated CVP, including heart failure and severe lung disease. Regardless of the underlying cause, elevated CVP contributes to tissue congestion and edema through a common mechanism of increased lymphatic production and impaired lymphatic drainage (Figure 2). When tissue fluid formation surpasses the TD’s drainage capacity, fluid accumulates in the tissue (edema) and leaks into the pleural and peritoneal spaces, resulting in effusions and ascites. Additionally, tissue congestion leads to end-organ dysfunction.
    Flow diagram of cause and consequences of lymphatic dysfunction in patients with elevated central venous pressure (CVP). BPD indicates bronchopulmonary dysplasia; and PLE, protein losing enteropathy.
    In the kidney, capsular constraints prevent expansion, leading to cardiorenal syndrome, which includes acute kidney injury, diuretic resistance, fluid retention, and chronic kidney disease.15 Our results demonstrate that renal function improved with TD drainage and autotransfusion, even in patients with acute kidney injury in the setting of chronic kidney disease, suggesting that renal dysfunction may be at least partially reversible. Furthermore, we observed that drainage of extralymphatic fluid, such as ascitic fluid, did not produce the same effect (Figure S2A; light green region). We hypothesize that, unlike TD drainage, the removal of extravascular and extralymphatic fluid (which has already extravasated from the vascular and interstitial compartments) does not alter TD or interstitial pressure and, consequently, does not relieve tissue congestion. This may also explain why TD drainage leads to a progressive decline in fluid output, whereas extralymphatic fluid drainage does not.
    Although TD cannulation with external drainage (without fluid return) can also decompress the lymphatic system, it may lead to significant fluid shifts, alterations in cardiac preload, and metabolic disturbances, making clinical management more challenging and necessitating extensive fluid replacement. In contrast, internal drainage of lymphatic fluid represents a more physiological alternative. In patients with single-ventricle heart disease, this approach has been shown to provide long-term resolution of lymphatic failure.9,10,13
    Here, we used an external chamber as a low-pressure compartment and an IV pump to reinfuse lymphatic fluid into the venous system. This strategy served as a bridge to additional procedures such as TDD, lymphovenous anastomosis, or lymphatic embolization. Due to the small and heterogeneous patient population, the duration of the procedure varied widely. Although our goal was to maintain autotransfusion for as long as possible after drain placement, factors such as drain occlusion, malfunction, infections, and limited chamber availability sometimes restricted the duration of treatment. Additionally, bloodstream infections contraindicated continued autotransfusion.
    Further studies are needed to better define the optimal duration of this approach for various indications, its durability, the appropriate volume of autotransfusion, and its long-term effects on patient outcomes.
    This work demonstrates that TD drainage with autotransfusion can be used to stabilize critically ill patients and alleviate symptoms of fluid overload in cases of lymphatic failure secondary to elevated CVP. Previously described lymphatic embolization techniques aim to occlude abnormal lymphatic pathways and prevent fluid leakage.16 Other interventions, such as surgical or percutaneous TD decompression, provide more durable relief by connecting the TD to a lower-pressure system.9,13,16 This strategy has been effective in patients with single-ventricle heart disease, leading to symptom resolution. However, it is associated with hypoxia and has not been shown to yield the desired outcomes in patients with elevated atrial pressures.
    For patients with lung disease or elevated left-sided heart pressures, current TD decompression strategies are not feasible. In these cases, we used prolonged TD drainage to stabilize the patient, allowing time for lung function to improve. For patients eligible for more permanent TD decompression procedures, this approach was used to stabilize them before undergoing additional interventions. Further research is required to determine whether short-term TD drainage with autotransfusion or prolonged drainage without autotransfusion can produce lasting improvements in lymphatic function. Additionally, studies should focus on defining the precise indications for this interventional approach.
    An alternative strategy for internal TDD in patients with Fontan circulation is the innominate vein turn-down procedure, as described by Hraška.12 This technique directly decompresses the lymphatic system by redirecting TD flow into the systemic venous system at a lower pressure and has shown promise in mitigating early lymphatic failure. Although our current study focuses on TD drainage with autotransfusion as a temporizing measure in critically ill patients, future studies should compare these approaches. In particular, performing innominate vein turn-down concomitant with the Fontan operation, as described by Kreutzer et al,10 may warrant consideration as a preventive strategy against early lymphatic complications.
    In 2 neonates (P4 and P5), TD drainage was used for several months. For P5, this strategy provided clinical stability and allowed the patient to grow, improving candidacy for lymphovenous anastomosis. For P4, the goal was to enable lung development and disease improvement. These cases demonstrate that internal TD drainage can be sustained for extended periods, resulting in clinical stabilization.
    Before TD drainage, both neonates experienced severe fluid losses requiring large-volume replacement, leading to metabolic instability and multiorgan system failure. Drainage with autotransfusion resulted in clinical stabilization, reduced fluid losses, and decreased need for replacement therapy (Figure S3A and S3B). These findings highlight potential resource utilization benefits and warrant further investigation.
    There were no major procedure-related complications. Minor complications included drain obstruction, which responded well to tPA administration or drain replacement. In long-term drainage patients, Enterococcus central line-associated bloodstream infections were diagnosed and successfully treated with antibiotics. No infections of the TD drain were observed.
    This is a small, retrospective review of medical records for a select heterogeneous group of patients with severe lymphatic disorder and elevated CVP, many of whom had multiple comorbidities. As a result, there is an inherent selection bias, and the observations may not be representative of all lymphatic disorder patients or generalizable to all types of lymphatic failure due to elevated CVP. Additionally, multiple factors influence outcomes, such as serum creatinine and fluid drainage. Further studies are needed to determine the specific effects of this procedure on these parameters.
    TD drainage with autotransfusion is a feasible intervention for patients with elevated CVP, resulting in improved symptoms of tissue congestion, reduced lymphatic failure, and enhanced organ function, including renal function. For critically ill patients, autotransfusion offers clinical and fluid status improvement, serving as a bridge to further treatment and interventions. Future research is needed to better define the role of this therapeutic approach and assess its long-term outcomes.

    ARTICLE INFORMATION

    None.
    None.
    Figures S1–S2
    EPDr. Edward Pinto

    Dr. Edward Pinto

    Cardiology · Preventive Cardiology
    Bridgeport Hospital
    edwardpinto