Lymphangiography is the most popular procedure for CA treatment as it is a minimally invasive diagnostic tool. In the past, lymphangiography was performed by cutting down the dorsum of the foot to expose the lymphatic vessels and puncturing them with a needle; however, this method is very time-consuming, technically difficult, and invasive, rendering this procedure obsolete.

Intranodal lymphangiography is an innovative novel lymphatic intervention − initially described in 2011, it provides a preferred alternative for pedal lymphangiography, as it is less technically challenging with a shorter procedure time [31]. Under ultrasound guidance, a fine needle is placed in the inguinal lymph node at the junction between the hilum and cortex. Lipiodol is then injected slowly to allow the contrast agent to move up the pelvic and retroperitoneal lymphatic chain to the CC and reach the TD.

Lymphangiography has been shown to have a therapeutic effect on chylous leak, probably due to its embolic or inflammatory effect on the lymphatic vessels [9].

The therapeutic effect of lymphangiography is volume-dependent, and its efficacy is higher in patients with low-output chylous leaks. A previous report has shown that its efficacy was 70% on low-output patients and 35% on high-output (>500 mL/day) patients, including 43 patients with postoperative chylothorax, CA, lymphocele, or lymphatic fistulas [32]. The efficacy of therapeutic lymphangiography for only postoperative CA remains unclear.

The most common finding of lymphangiography is extravasation due to lymphatic leaks. However, a recent report has shown a detection rate of 55% for leakage sites in patients with CA, and this rate was lower than that of chylothorax [2]. Lack of leak visualization may be due to a small intestinal, or hepato-enteral lymphatic leak [30].

Dolan et al. [33] reported complications after lipiodol lymphangiography in 166 of 522 patients (31.8%), including fever (18.6%), nausea/vomiting (4.4%), pain (3.3%), and respiratory signs/symptoms (1.3%). Potentially severe complications include a spectrum of pulmonary complications as well as cerebral and visceral oil embolization injuries [34]. Serious respiratory complications, including pulmonary infarction (0.25%), lipid pneumonia (0.04%), pulmonary edema (0.03%), and hemoptysis (0.03%), have been reported [35]. Extremely rare respiratory complications, such as acute respiratory distress syndrome which developed into pulmonary fibrosis requiring domiciliary oxygen therapy, have been reported; additionally, an allergic inflammatory reaction to lipiodol has been postulated, with the author of the report suggesting that early steroid administration may prevent lung fibrosis [36].

Overdose injection of more than 20 mL increases the risk of embolic complications. A review in the 1960s, long before intranodal lymphangiography was developed, investigated 522 lymphangiograms stratified by volume of administered lipiodol showed that injections of less than 18 mL had a 13% risk, between 18 and 20 mL had a 24% risk, and more than 20 mL had a 48% risk of complication [34]. It is reported that most lymphangiography procedures can be performed with 15 mL of lipiodol or less, although prescribing information stated 20 mL as the maximum lipiodol dose [34]. A systematic review by Sommer et al. [37] stated a maximum dose of 15 ml for lipiodol in intranodal lymphangiography based on data from several small case series. In the dose-proportional pedal lymphangiography complication rate study, the 15 ml for intranodal lymphangiography figure seems reasonable, as the risk of complications increased with doses above 18 ml [34, 37].

No scientific evidence or consensus has been reached regarding children; however, several papers state 0.2 ml/kg as the maximum dose, and prescribing information states 0.25 ml/kg [31, 38].

Moreover, the risk-benefit ratio of lipiodol lymphangiography should be considered, and other modalities such as MRL should be chosen as alternative diagnostic tools for patients with contraindications to lipiodol lymphangiography, such as right-to-left shunt and known pulmonary disease [39].

TD injury after abdominal surgery is unlikely. Performing TDE for CA without a detailed preoperative evaluation may worsen the outflow of CA by obstructing the TD. Therefore, when percutaneous transabdominal TDE (PTTDE) is performed in the case of CA, a detailed preprocedural evaluation such as lymphoscintigraphy or CL is necessary, and the leak point must be embolized without fail when embolization is performed. In TDE for chylothorax, a catheter tip can be advanced near the leak point and isolation can be performed, whereas TDE for CA requires blocking downstream of the TD with an embolic material such as a coil and then embolizing the leak point located upstream from the catheter tip with n-butyl-2-cyanoacrylate (NBCA) in a retrograde fashion. Yokokawa et al. [40] reported a CA case after laparoscopy-assisted distal gastrectomy and D2 lymph node dissection, which was successfully treated using PTTDE with coil embolization of the upper site of the TD and a 20% NBCA: Lipiodol injection. However, to date, no other clinical research has been conducted on the use of PTTDE for CA treatment.

Although PTTDE is a very low-risk procedure, there can be rare complications. Schild et al. [41] analyzed 35 patients by CT after PTTDE and found one perihepatic hemorrhage and one periaortic hematoma without any additional treatment. Solitary cases of bile peritonitis and pancreatitis have also been noted. They mentioned that preprocedural ultrasound scanning may avert puncturing the gallbladder [41]. Should a patient show symptoms after embolization, other authors recommend that CT should be performed without hesitation [34].

Sclerotherapy or embolization can also be adapted for CA treatment when a patient with an lymphopseudoaneurysm (LPA) (Fig. 4), which is defined as a small cavity of lymph collection that is directly extravasated from the lymphatic vessels or lymph nodes. It is contained in the surrounding tissue before draining into larger spaces such as the peritoneal cavity, or into lymphoceles [40, 42]. There are some reports of CA treatment with LPA embolization using NBCA with or without coils [9, 42, 43]. Kwon et al. [42] retrospectively investigated six patients who had refractory CA after retroperitoneal surgery, and all patients were successfully treated using 33% to 50% NBCA injected from a placed catheter without any complications. Nadolski et al. [9] reported that a chylous LPA was punctured with a 22 G needle under fluoroscopy guidance, after which a microwire was advanced into the cavity and a needle was exchanged for a microcatheter. Embolization was then performed with coils and NBCA from the inserted microcatheter [9]. Another case report was treated with CT-guided catheter insertion and embolization [43]. However, no reports could be found using another agent for abdominal chylous leaks.

After lipiodol lymphangiography is performed, if lymphatic extravasation was identified along the course of the pelvic sidewall or retroperitoneal lymphatic vessel, the nearest upstream lymph node should be punctured using a small-sized needle (22-25 G). It is not necessarily a single lymph node that is involved, and if possible, the surrounding lymph nodes should also be considered for embolization. This technique is useful for treating leaks from small lymphatic vessels which cannot be catheterized. Hur et al. [22] reported direct puncture of the lipiodol-stained lymph node could be easily performed under fluoroscopy and/or CT. The ratio of glue to lipiodol should be adjusted according to the extent of the target embolization taking care of premature polymerization [22, 44].

Clinical failures may result from a complex network of collaterals or multiple leakage sites that cannot be visualized by intranodal lymphangiography. Therefore, repeated embolization should be considered [45]. Nadolski et al. [9] reported six CA cases were treated successfully using proximal lymph node embolization.

The retrograde transvenous (RTV) approach is a primary or secondary method for TD approach intervention, which was introduced in 2008. Kim et al. [46] reported that the retrograde TD cannulation rate was as high as 70%. In this study, intranodal lymphangiography was performed to evaluate the morphology of the terminal TD before starting RTV, and it was concluded that the engagement of a shaped diagnostic catheter into the jugulovenous junction and the presence of a dominant channel in the terminal TD were prognostic factors for a successful RTV [46].

RTV can be a bailout technique not only for PTTDE, but also for the treatment of CA. A few case reports of treatment with retrograde TDE for CA have been published [47, 48]. When antegrade lymphangiographies such as inguinal intranodal lymphangiography or lymphatic scintigraphy could not confirm the leak point of CA due to disrupted pelvic or retroperitoneal lymphatic ducts caused by lymphatic dissection, or when the leak point was out of line from the lymphatic chain due to intranodal injection, retrograde transvenous lymphangiography and embolization may be used to approach the leak point (Fig. 5). Although some authors reported without balloon catheter embolization [47], theoretically, balloon-occluded embolization may be reasonable, both in terms of delivering contrast material or embolic liquid to the leak point and preventing pulmonary embolisms due to reflux [48]. Catheter adhesion is also less likely because, unlike blood, the polymerization time of NBCA in lymph fluid is longer, especially in chyle [49]. NBCA polymerizes upon exposure to anions, zwitterions, and free radicals. Although the electrolyte levels in plasma and lymph are similar, the concentrations of fat (chylomicrons, triglycerides (TGs), and fat-soluble vitamins) are quite different. Previous studies have shown that a high TG concentration extends the polymerization time of NBCA [50].

Liver lymphatic flow is divided into deep and superficial systems. The deep lymphatic tract is further divided into the hepatic vein tract and portal tract. Lymphatic flow toward the hepatic hilum in the portal tract can be the cause of peritoneal lymphorrhea and is located in the periportal area as well as the intrahepatic bile duct.

Cope demonstrated the liver lymphatic duct in 47% of patients for attempting transhepatic cholangiography using a 27 G micropuncture needle [51]. This finding has been applied to intrahepatic lymphangiography and embolization. Guez et al. [52] demonstrated hepatic lymphangiography and subsequently embolized with Onyx for hepatic lymphorrhea using 21 G and 27 G coaxial needle techniques.

Hepatic lymphatics communicate with the intestinal lymphatics through multiple channels; liver lymphangiography can sometimes elucidate CA, but not always [53]. There are few case reports that treated liver lymphatic embolization for CA. Nguyen et al. [54] reported a case of treated CA after pancreaticoduodenectomy using Aetoxisclerol foam. At our institution, we treated one case (not published) with hepatic lymphatic embolization using NBCA (Fig. 6).

Peritoneovenous shunt (PVS) is a permanent placement catheter system under the skin that connects the peritoneal cavity to the superior vena cava. This system has two catheter limbs with an interposed single or double-valved unidirectional pump. A valve should be placed upon the rib to facilitate pressing the pump button.

This system is a permanent placement device and indirectly affects chyle leaks. In addition, there have been a few reports of disseminated intravascular coagulation (DIC), as well as problems such as CA recurrence due to thrombotic occlusion; device replacement is needed in this case. Therefore, this system is considered an important last resort, this device should be placed after considering the patient's condition, providing sufficient information regarding the need for regular follow-up, and explaining the burden following its use. Yarmohammadi et al. [55] investigated 20 patients who had been placed PVS for postoperative CA. The clinical success rate was 90%, leading to shunt removal in 85% of patients between 46 and 481 days. The total complication rate was 40%, with the most common complication being occlusion (30%). Subclinical DIC was observed in one patient without additional treatment [55]. In this study, patients were taught to press the button 20 times in the morning and before going to bed. However, the number and intervals of pressing the button seem low and should be increased to prevent occlusion. There was no mention of anticoagulant administration, and although there is no consensus, direct oral anticoagulant should be considered to prevent catheter occlusion.