Interventional procedure overview of percutaneous thoracic duct embolisation for persistent chyle leak
Closed for comments This consultation ended on at Request commenting lead permission
Summary of key evidence on percutaneous thoracic duct embolisation for persistent chyle leak
Study 1 Power R (2021)
Study type | Systematic review |
|---|---|
Country | Ireland |
Search details | Search period: up to December 2020; databases searched: Medline, Embase, and Web of Science. Hand searching of references from relevant systematic reviews was also done for additional studies. |
Study population and number | n=7 retrospective case series (with 455 patients who had chyle leaks from multiple aetiologies [including 180 after oesophageal resection]). |
Age and gender | Not reported |
Study selection criteria | Inclusion criteria: randomized trials or retrospective studies that assessed management of chyle leaks after oesophageal resection; studies with chyle leaks of multiple aetiologies, interventions and comparators such as conservation management, surgical TDL, TDD, and TDE; reporting outcomes such as technical success rate, clinical success rate, time to resolution of chyle leak and complications; studies in English, and in humans. Exclusion criteria: case reports, small case series (n < 10), conference abstracts, commentaries, editorials, duplicate studies, and those reporting only incidence, risk factors, or prognosis. |
Technique | Lymphangiography with thoracic duct embolisation or disruption. |
Follow up | Varied across studies. |
Conflict of interest/source of funding | The authors declared that they have no conflicts of interest. |
Analysis
Study design issues: The protocol was registered on PROSPERO and study was done according to preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines. Searches were comprehensive, 2 authors independently screened, extracted data and assessed the quality of studies (using Cochrane risk of bias tool or using the National Institute of Heart Lung and Blood (NHLBI) quality assessment tool for non-randomised case series). Studies were retrospective with significant risk of bias. Any disagreement was resolved by discussion and consensus. As studies were heterogenous (in patient groups, management regimens, treatment modalities and definitions used), a qualitative analysis was done.
Other issues: 18 studies that described the conservative management of chyle leaks, 17 by surgical ligation of the thoracic duct, 5 by pleurodesis, were not considered in this overview as they are out of the remit of this guidance.
There is some overlap of primary studies between study 1 and 2.
Key efficacy findings
Number of patients analysed: 455
Study | Patient aetiology | Intervention | Technical success % | Clinical success % | Hospital stay (mean) |
|---|---|---|---|---|---|
Boffa 2008 | Unknown indications/referrals | TDE (n=21) | 93 | 57 | 8 days |
TDD (n=4) | 50 | 19 days | |||
Itkin 2010 | Unknown indication/referrals | TDE (73) | 97 | 74.6 | NR |
TDD (n=18) | 72 | NR | |||
Nadolski 2018 | Failed TDL/referral | TDE (n=49) | 98 | 98 | NR |
TDD (N=1) | 100 | 100 | NR | ||
Pamarthi 2014 | Unknown indication/referrals | TDE/TDD (n=50) | 86 (n=43 post oesophagectomy) | 56 | NR |
Reisenauer 2018 | 1.1 litre daily output | Surgical TDL (n=48) | - | 85 | NR |
TDE (n=40)^ | 48 | 38 50 post oesophagectomy (n=22) | NR | ||
Yannes 2017* | Presence of post-operative chyle leak and failed conservative management | INL alone (n=7) | 100 | 71.4 | Median TTR 14 days |
INL+TDE for leak<500 L/day(n=21) | 90.5 88.6 (post esophagectomy [n=9]) | Median TTR 3 days | |||
INL+TDD (n=12) | 41.7 75 (n=4 post esophagectomy) | Median TTR 7 days |
*no difference in clinical success rates between the groups was reported (p=0.19).
^inability to cannulate the cisterna chyli in 48% patients.
Reoperations were not needed in patients who had TDE.
In 4 studies the median clinician success rate was 57% (ranged 38-98%); 4 studies reported TDE and TDD separately, with a median clinical success rate for TDE of 75% (range 57% to 98%) and a median clinical success rate for TDD of 72% (range 41.7% to 100%).
Key safety findings
Complications | % (n) |
|---|---|
Minor complications | Range 4-6% |
Further chyle leak (managed by endoscopic cholangiography and bile duct stenting) | 1 (Boffa 2008) |
Leg and pedal oedema (resulting in wound infections) | 2 (Itkin 2010) |
Asymptomatic pulmonary embolisation | 1 (Itkin 2010) |
Inconsequential coil misplacement | 1 (Itkin 2010) |
Mortality | 3% (1/40) (Reisenauer 2018) 2% (1/57) ( Yannes 2017) |
Study 2 Kim PH 2018
Study type | Systematic review and meta-analysis |
|---|---|
Country | Republic of Korea, China |
Search details | Databases searched: MEDLINE, EMBASE, and Cochrane databases were searched until March 2017. Hand searching of references from relevant systematic reviews was also done for additional studies including Google Scholar. Authors of studies contacted for further individual patient data. |
Study population and number | 9 retrospective case series (n=407 patients) chylothorax aetiology: iatrogenic (82.6%; 336/407), malignancy (5.4%; 22/407). High-output chylothorax in 76.9% (70/91) Previous unsuccessful surgery 17.6% (40/228) |
Age and gender | Mean 60 years; 43% (177/407) male |
Patient selection criteria | Inclusion criteria: studies regarding patients with chylothorax treated with lymphangiography, TDE, or TDD, in English language. Exclusion criteria: studies with less than 10 patients, case reports, review articles, letters, and conference abstracts, with no extractable data, or data included in subsequent articles or duplicate reports. |
Technique | Percutaneous lymphatic interventions for chylothorax (LAG, TDE, TDD) |
Follow up | Varied across studies. |
Conflict of interest/source of funding | None |
Analysis
Follow-up issues: overall 6 patients were lost to follow-up (1 after LAG and 5 after TDE).
Study design issues: the review was conducted according to the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines. Searches were comprehensive, 2 authors independently screened, extracted data using standardised form. All studies included were retrospective case series and quality of studies were fair to good according to the U.S. National Institutes of Health Quality Assessment of Case Series studies tool. Any disagreement between the authors was resolved by discussion and consensus. Meta-analysis was done using random effect model and heterogeneity across studies was also assessed. There is significant heterogeneity in sample size, patient groups, definition of outcomes.
Other issues: there is some overlap of primary studies between study 1 and 2.
Key efficacy findings
Number of patients analysed: 407
Technical and clinical outcomes
Technical success of LAG was defined as successful injection of contrast agent into the lymphatic system; technical success of TDE was defined as total occlusion of the target lymphatic duct.
Clinical success was defined as complete resolution of chylothorax without further surgical treatment.
% (n) | |
LAG | n=407 |
Pedal LAG | 88.7% (361/407) |
Intranodal LAG | 11.3% (46/407) |
Technical success | 95.1% (387/407) |
Technical failure^ | 4.9 (20/407) |
Clinical success (n=89) | 56.2% (50/89) |
TDE | 80.1% (310/387) |
Technical success | 62.9% (195/310) |
Technical failure^^ | 37.1 (115/310) |
Clinical success | 79.5% (151/190) |
TDD (in those with TDE technical/clinical failure) | 24.8 (77/310) |
Clinical success | 61 (47/77) |
^ data available for 16 patients show that they were due to unsuccessful cannulation of the pedal lymphatic vessels in 87.5% (14/16), over-sedation in 6.3% (1/16), and injector malfunction in 6.3% (1/16) of the patients.
^^data available for 68 patients show that the causes included unsuccessful cannulation of the cisterna chyli or thoracic duct in 86.8% (59/68) and unsuccessful negotiation of the guidewire into the thoracic duct in 13.2% (9/68).
Pooled technical and clinical success rates (on a per-protocol basis)
Outcome | Number of studies | % (95% CI) | P value, I2 |
|---|---|---|---|
Technical success rate of LAG | 6 studies | 94.2 (88.4–97.2) | p=0.059, I2=46.7% |
Clinical success rate of LAG | 6 studies | 56.6 (45.4–67.2) | p= 0.382; I2=5.4% |
Technical success rate of TDE | 6 studies | 63.1 (55.4–70.2) | p=0.157, I2=37.3% |
Clinical success rate of TDE | 6 studies | 79.4 (64.8–89.0) | p=0.008, I2=68.1% |
Clinical success rate of TDD | 5 studies | 60.8 (49.4–71.2) | p=0.830, I2=0% |
Overall clinical success rate of lymphatic interventions (LAG, TDE, TDD) | 6 studies | 60.1 (52.1–67.7) | p=0.025, I2=54.3% |
Meta-regression analysis showed that aetiology of chylothorax was identified as a significant source of heterogeneity for the pooled clinical success rate of TDE (p=0.012) and overall clinical success rate (p=0.002).
Key safety findings
Complications of LAG | % (n) |
|---|---|
Major- aspiration | 0.3 (1/407) |
Minor | 1.3 (8/407) |
Over-sedation | 5 |
Pedal incision site injury | 2 |
Oedema in leg | 1 |
Complications of TDE | |
Major (bile leak) | 1.0 (2/195) |
Minor (in 6 studies) | 4.1 (8/195) |
Non-target embolisation to lungs | 3 |
Non-target embolisation to the portal vein | 1 |
Guidewire fracture | 3 |
Perihepatic hematoma | 1 |
Complications of TDD | 0 |
Complications were categorized as major or minor according to the Society of Interventional Radiology clinical practice guidelines.
Pooled major complications
LAG (6 studies) | 1.9% (95% CI, 0.8%–4.3%), p= 0.940; I2 = 0% |
TDE (6 studies) | 2.4% (95% CI, 0.9%–6.6%), p= 0.236; I2= 26.4%) |
Study 3 Jun H 2022
Study type | Retrospective case series |
|---|---|
Country | Republic of Korea |
Study details | 2016-2019 |
Study population and number | N=45 patients with postoperative chylothorax |
Age and gender | Mean 62 years; 62% (28/45) male |
Patient selection criteria | Patients with postoperative chylothorax undergoing lipiodol LAG for intended TDC and TDE, with milky" chylous effusions > 110 mg/Dl, fat-free or low-fat diet or total parenteral nutrition, failed to stop the chylous leakage were included. |
Technique | First patients underwent intranodal lipiodol LAG under ultrasound guidance. If targetable central lymphatic vessels were identified, standard TDC (antegrade transabdominal approach) was attempted. The retrograde approach was applied as a bail-out method for TDC in few failed cases. TDE (embolisation), was performed after confirming leakages in the trans-TDC catheter lymphangiography. Micro-coils or Concrerto were deployed at the cervical position of the thoracic duct and then glue embolisation of the segment of the duct was done. Extensive embolisation of thoracic duct or super-selective embolisation of culprit lymphatic channels were done. |
Follow up | Average 434 days |
Conflict of interest/source of funding | No conflicts of interest. Study funded by the Ministry of Science, ICT and Future Planning of Korea. |
Analysis
Follow-up issues: 2 patients were lost to follow-up. Follow-up data was highly variable.
Study design issues: all procedures were conducted by 1 interventional radiologist. Radiological and clinical data were retrospectively reviewed by 3 radiologists. Electronic medical records and picture archiving and communication systems were accessed for review. Technical success and clinical success of TDE group (with all 3 steps of LAG, TDC and embolisation) and non-TDE group were compared.
Key efficacy findings
Number of patients analysed: 45
Technical and clinical outcomes
Per protocol % (n) | ITT % (n) | |
|---|---|---|
LAG | 100 (45/45) | |
Technical success rate of TDC (antegrade approach) | 78 (31/40) | 69 (31/45) |
Failure of TDC | 22 (9/40) | |
Technical success rate of TDC (antegrade + bail-out retrograde approach in 8) | 93 (37/40) | 82 (37/45) |
Failure | 7 (3/40) | |
Clinical success of TDE* (n=35) | 89 (31/35) | 80 (36/45) |
Non-TDE^ (n=10) | 50 (5/10) |
^ the reasons for non-TDE: lack of targetable lymphatics for TDC in LAG (n = 5), technical failure of TDC (n = 3), and lack of visible leakages in the transcatheter lymphangiography (n = 2).
* defined as resolving the lymphatic leakages within 2 weeks after final TDE, regardless of its technical success. TDE was in the form of extensive embolisation of thoracic duct (n = 27) or super selective (n = 8) embolisation of culprit lymphatic channels.
Key safety findings
n | |
|---|---|
Procedure related major complication | |
Bile peritonitis (caused by needle penetration of the distended gallbladder at average 434 days, triggered extreme abdominal pain after TDE, needed emergency percutaneous cholecystectomy). | 1 |
Death (due to persistent lymph leak in TDE failure patient) | 1 |
Minor complications | |
Fever (treated conservatively) | 4 |
Asymptomatic non-target glue embolisation of the pulmonary artery | 1 |
Study 4 Schild HH 2020
Study type | Retrospective case series |
|---|---|
Country | Germany |
Study details | 2014-18 |
Study population and number | N=35 patients with chylous effusions refractory to conservative therapy who had TDE. Aetiology of the chylothorax: traumatic/postoperative in 25 patients, idiopathic in 9, and leukaemia-related in 1. |
Age and gender | Mean 57 years; 63% (22/35) male |
Patient selection criteria | Patients with TDE and post-procedural CT data were included. |
Technique | Transabdominal TDE: first, conventional oily lymphangiography was performed. For initial duct obstruction, coils were placed into the thoracic duct to act as a scaffold, followed by injection of liquid embolic mixture (mix of tissue adhesive agent and Lipiodol) under fluoroscopic guidance. During embolisation, the microcatheter was slowly pulled back until the point of entry into the lymphatic system was sealed. A thin line of radiopaque embolisation material outlined the access route on postprocedural imaging. CT studies were obtained 30–60 minutes after TDE. |
Follow up | Mean follow-up, 678 days (range, 44–2,619 days) |
Conflict of interest/source of funding | One author is a paid consultant for Philips Healthcare. |
Analysis
Follow up issues: complete follow-up.
Study design issues: all procedures were performed by 2 interventional radiologists. Data were gathered from electronic medical records. Procedures with post intervention CT data were retrospectively analysed by 2 experienced radiologists for abdominal structures and organs transgressed by the access route, signs of complications, and distribution of embolic material. Findings were correlated with clinical course. Adverse events were graded according to CTCAE version 5.
Other issues: only the final access route and its related complications were analysed. Other previous unsuccessful punctures and long term complications of TDE were not assessed in this study.
Number of patients analysed: 35
Key safety findings
Study 5 Gurevich 2022
Study type | Retrospective case series |
|---|---|
Country | USA |
Study details | 2014-20 |
Study population and number | N=52 patients with nontraumatic chylous pleural effusions and/or chylopericardium. Aetiologies: idiopathic 58% (30/52), malignancy 40.3% (21/52), genetic 1. Previously failed conservative treatments: non-fat diets, total parenteral nutrition, and octreotide. Patients were symptomatic for an average of 283 days before referral. |
Age and sex | Mean 54 years (range 11-89 years); sex not reported |
Patient selection criteria | Patients with various abnormalities that led to chylothorax are identified using pre-intervention dynamic contrast-enhanced magnetic resonance lymphangiography (DCMRL). |
Technique | Using DCMRL 3 lymphatic patterns were identified (abnormal perfusion from thoracic duct needing TDE [n= 31], abnormal perfusion from retroperitoneal networks needing interstitial lymphatic embolisation [n=10], and presence of ascites with a normal/absent thoracic duct, similar fluid characteristics within ascites and chylothorax, extravasation of contrast in the abdomen and requires treatment for chylous ascites [n=11]). All embolisation procedures done under general anaesthesia or moderate sedation. Thoracic duct embolisation (TDE): intranodal lymphangiography was followed by transabdominal thoracic duct catheterization and thoracic duct embolisation. Embolisation was performed using mixture of endovascular coils and/or n-butyl cyanoacrylate n-BCA glue or autologous blood and oil-based contrast material (Lipiodol). Interstitial lymphatic embolisation delivers an embolisation agent (n-BCA glue or lipiodol directly into cystic or cavernous retroperitoneal lymphatic patterns. |
Follow up | Median 396 days (range 824) |
Conflict of interest/source of funding | None |
Analysis
Follow up issues: long term follow-up was available in 87% (27/31) patients in the TDE only group.
Study design issues: small single centre study, data was collected retrospectively from electronic medical records and/or health system medical records. In a few cases data was supplemented by calling patients.
Other issues: DCMRL was not performed for 3 patients in thoracic duct group because of contraindications to MR imaging. They underwent intranodal lymphangiography.
Number of patients analysed: 52
Key efficacy findings
TD only (n=31) | Retroperitoneal lymphatic networks (n=10) | TD and lymphatic networks (n=41) | Ascites (n=11) | |
|---|---|---|---|---|
Mean number of procedures | 1.3 | 2.4 | 1.5 | |
Embolisation technical success (ability to access and embolise the duct), % | 97(30/31) | 80 (8/10) | 93 (38/41) | (6/11) |
Clinical success (output reduced by 85% within 2 weeks or eliminated within 2 months), % | 97(30/31) | 80 (8/10) | 93 (38/41) | (4/6) |
Time to resolution (mean, range) days | 8 (1–30) | 15.4 (5-53) | 9.5 (1-53) | NR |
Follow-up (median, range) days | 275 (640) | 751 (573.5) | 396 (824) | NR |
Key safety findings
Complications were scored according to the National Institutes of Health common terminology grading system (CTCAE version 5.0)
Complications in TDE and lymphatic embolisation groups only | 14.6 (6/41) |
|---|---|
Hypotension (grade 1) | n=1 |
Hypoxemia (grade 1) | n=1 |
Fluid overload (grade 1) | n=1 |
Systemic inflammatory response syndrome (grade 2) | n=1 |
Atrial fibrillation with rapid ventricular response and acute pulmonary oedema (grade 2) | n=1 |
pulmonary embolism due to stress-induced cardiomyopathy 2 days after procedure (grade 5) | n=1 |
Complications in ascites group (n=11) | 0 |
Study 6 Laslett D2012
Study type | Retrospective case series |
|---|---|
Country | USA |
Study details | 1994-2010 |
Study population and number | N= 106 patients with technically successful TDE for symptomatic chylous effusions. |
Age and sex | Average age 58 years; 50% women |
Patient selection criteria | patients' multiple medical conditions who underwent TDE for symptomatic chylous effusion. |
Technique | Thoracic duct embolisation. |
Follow up | Mean 34 months (range 2-134 months) |
Conflict of interest/source of funding | None |
Analysis
Follow up issues: long term follow-up was available in 74% (78/106) patients.
Study design issues: data was collected retrospectively from electronic medical records and hospital databases. Patients who were alive and 3 family members were surveyed and interviewed to determine the post-operative status and rate of long-term complications.
Number of patients analysed: 106
Clinical success of TDE (chylous effusions resolved): 93% (99/106)
Key safety findings
% (n) | |
|---|---|
Overall long-term complication rate | 14.3 (26/46) |
Death unrelated to TDE | 41 (32/78) |
Chronic leg swelling in patients alive (probably related to the procedure) | 8 (4/46) |
Abdominal swelling (unrelated to the procedure) | 6 (3/46) |
Chronic diarrhoea (4 probably related to the procedure as onset was after the procedure and needed medication) | 12 (6/46) |
How are you taking part in this consultation?
You will not be able to change how you comment later.
You must be signed in to answer questions