Introduction
Unrecognized injury to the main thoracic duct or its tributaries during surgery is
the usual cause of traumatic chylothorax. In cases unresponsive to conservative management,
surgical revision with ligation of the thoracic duct is performed. Based upon experimental
studies, an interventional technique to occlude the thoracic duct has been described
[4]
[5]. In the following we want to report a modified technique for percutaneous, transabdominal
occlusion of the thoracic duct.
Patient and Technique
A 68-year-old patient had undergone esophageal resection for a distal carcinoma. A
small amount of fluid was drained by the right chest tube postoperatively. When oral
food intake was started on the 9th postoperative day, the volume of drained fluid,
which turned out to be chylous, increased drastically to about 2600 ml per day; this
persisted despite placing the patient on total parenteral nutrition. Over the next
days the patient's general condition deteriorated secondary to continued high output
chylous leakage, such that he was considered a poor candidate for surgical revision.
Eventually, after 24 days of unsuccessful conservative management, we decided to attempt
interventional occlusion of the chylous leak.
A lymphogram was performed via cut down on the dorsum of the left foot; a prior attempt
to catheterize a lymph vessel on the right foot had failed. After intubation of a
lymph vessel (Lymphangiography Set, Cook, Mönchengladbach), about 10 mL of Lipiodol
Ultra Fluid (Guerbet Sulzbach) were carefully injected over 45 minutes with the patient
on the intervention table.
Passage of the contrast medium was intermittently observed with fluoroscopy. About
an hour after start of the injection, segments of larger abdominal lymphatic ducts
were opacified. Under local anesthesia a larger lymph vessel at L 1 level was transabdominally
punctured with a 40 cm long fine needle (0.7 mm diameter; Cook, Mönchengladbach) which
was armed with a F 4 vessel dilator (Fig. [1]). To minimize trauma, the dilator was kept on the needle segment outside of the
patient until - after several needle passes - a lymph vessel was entered with the
needle tip as judged by fluoroscopy. A 0.012 guide wire (Tracker Therapeutics/Boston
Scientific, Ratingen) was carefully advanced until the non-floppy part was securely
positioned in the lymph vessel. At this time, the distal guide wire had entered the
right pleural space (Fig. [2]). The dilator was advanced over the needle and guide wire into the lymphatic duct,
the needle removed and a Tracker catheter system exchanged. The dilator served as
an access sheath and to stabilize and stiffen the system to avoid buckling.
With the catheter in the thoracic duct, direct lymphography was performed via careful
injection of water soluble contrast medium (Iopromide, Ultravist 300, Schering, Berlin)
through the indwelling catheter. This revealed the leakage site into the right pleural
space at the TH 8 level. However, some of the oily contrast medium from the transpedal
lymphography was not only seen in the right, but as well in the left pleural cavity,
indicating bilateral lymphatic leakage.
The catheter was pulled back to the TH 11 level well below the leakage point, where
a total of seven “Tracker” coils (3 mm diameter) were placed (Fig. [3]). Careful contrast medium injection after twenty minutes revealed only incomplete
occlusion. To further embolize the duct we injected about 0.3 ml of a 1 : 3 mixture
of bucrylate (Histoacryl, Braun Melsungen) and Lipiodol Ultra Fluid (Guerbet, Sulzbach),
and immediately removed the catheter system and the dilator to avoid inadvertent fixation.
The patient was returned to the ward and was kept with nothing per os over night.
The immediate postinterventional period (12 hours after the procedure) was uneventful,
except for some mild abdominal discomfort that did not require any specific therapy,
and that subsided by the next morning.
The amount of chylous drained by the chest tube decreased in the immediate post interventional
period from 2300 mL to 170 mL per day. A CT study on the day after the intervention
showed neither ascites nor fluid around the access site.
Oral nutrition was resumed on the second day, when 180 mL/24 hrs of chylous fluid
were still drained by the chest tube. Over the next days, the drainage volume continued
to decrease, such that on day six after the intervention, the chest tube was removed.
After recovery from an intercurrent pneumonia, the patient was dismissed on day seventeen
after intervention. Follow-up studies three and ten months after the procedure revealed
neither recurrence of pleural fluids, nor accumulation of ascites.
Discussion
Injuries of the thoracic duct with chylothorax are a rare though well known complication
of esophageal surgery, and are reported in 0.3 to 4 % of the patients [1]. Conservative management consists of chest tube drainage, nil per mouth and total
parenteral nutrition.
Surgical revision with ligation of the thoracic duct is indicated if conservative
treatment fails [2]. Patients unresponsive to conservative treatment are usually those with high output
leaks (2000 - 2500 ml/day), and/or with no signs of reduction of the daily amount
of drainage within one week of conservative treatment. Prolonged loss of major amounts
of chylous fluid places the patient at a significantly increased risk of infection
owing to an impaired immune competence, and is associated with a substantial deterioration
of the patient's general condition. Therefore, patients with persisting loss of chylous
fluids are considered poor candidates for surgical interventions. This is even more
important as the surgical approach to the thoracic duct may be difficult, and has
a reported mortality of up to 25 % [1]
[3]
[6].
Cope et al. [4] demonstrated in animal studies that it is feasible to percutaneously catheterize
the thoracic duct. Based on their experiments, they described a percutaneous approach
to the abdomino-thoracic lymph system, which we modified. In our patient, a monolateral
lymphangiogram via pedal cut down sufficiently opacified larger abdominal lymph ducts.
Thus bilateral lyphangiography seems not to be mandatory. Because of the lymph flow,
and the oily contrast medium used for lymphography, opacification of the lymph ducts
was segmental and non-continuous, necessitating intermittent fluoroscopy for puncture
guidance. In their report Cope et al. used an 8 French metal cannula for the puncture
of the lymph duct. In our patient a 4 French dilator provided access with sufficient
stability and avoided guide wire or catheter buckling.
Occlusion of the thoracic duct with multiple coils did not result in immediate and
complete occlusion in our patient, which probably reflects the reduced amount of components
responsible for coagulation in chyle when compared to blood. This prompted us to also
seal the thoracic duct with a small amount of bucrylate. As we had to withdraw the
catheter system immediately after glue injection, we could not prove that glue injection
indeed helped to occlude the leakage site. As the observed reduction in drained chylous
volume might have also been attributable to intraabdominal lymph leakage through the
puncture site, we performed a CT study the day after the intervention. This examination,
however, did not reveal any intraabdominal fluid collections, in particular not around
the lymphatic puncture site at the L 1 level. We assume that the glue injection helped
seal the lymphatic duct at the puncture site.
The volume of chylous drained further decreased over the next days so the chest tube
was removed seven days after the intervention. The pleural effusion on the left decreased
as well; only a small residual amount of pleural fluid was seen here after two weeks
of follow up. Ten months after the intervention the patient is in good general condition
and free of pleural and abdominal effusions.
Based on our experience, the report by Cope et al. (1999) and a report published during
the review of our manuscript [7] we feel that percutaneous interventional occlusion of the thoracic duct should be
considered as an alternative to surgery when (traumatic) chylothorax requires treatment.
The procedure can be performed through a 4 French access; this is less traumatic than
Cope's technique and may improve procedure safety.
However, percutaneous thoracic duct occlusion may be technically impossible in some
patients when lymph duct anatomy precludes successful puncture, or chylous extravasation
comes from a small thoracic duct collateral, or from a ligated but reconstituted duct.
In these cases surgery may still be required; when performed thoracoscopically, as
has recently been reported [8] this may also be less traumatic for the patient than conventional operative treatment.
Fig. 1Transabdominal fine needle puncture of a lymph vessel, which is segmentally opacified
by lymphographic contrast medium.
Fig. 2The thoracic duct is outlined by contrast medium injection via the dilator, which
was advanced into the punctured lymph vessel and ends at level TH 11. The guide wire
enters the right pleural cavity through a defect in the thoracic duct at TH 7/8 level.
After lymphography droplets of oily contrast medium are seen in both pleural spaces.
Fig. 3The thoracic duct was embolized with seven platinum coils (3 mm diameter). Situation
before injection of tissue adhesive. Note: cysterna chyli outlined by contrast medium
reflux at
the lower part of the image.
