Keywords
postop collection drainage - interventional radiology - computed tomography guided
drainage
Introduction
Development of intra-abdominal abscesses or collections is a common complication following
various intra-abdominal surgeries.[1] If these abscesses or collections are left untreated, it could result in a series
of complications such as intra-abdominal bleed, sepsis, and mortality ranging from
45 to 100%.[2] Over the past 20 years, image-guided percutaneous drainage has surpassed surgical
drainage and has become the standard therapy for intra-abdominal abscesses.[3] However, conventional percutaneous approaches for abscess drainage may not be suitable
for deeply located and inaccessible abscesses or collections because of many obstacles,
such as surrounding viscera, like liver, stomach, bowel, spleen, and kidney, and intervening
vascular structures along the catheter tract.[4] Few sporadic studies have described the safety and efficacy of transhepatic drainage
of inaccessible intra-abdominal abscesses under computed tomography (CT) or ultrasonographic
(USG) guidance.[5]
[6]
[7] Here, we report a short series of five patients who developed deep intra-abdominal
abscess after surgery and were successfully managed by transhepatic catheter drainage.
We also review the literature on transhepatic drainage.
Materials and Methods
This retrospective study was approved by the institute ethics committee and informed
consent was waived. Between April 2019 and December 2020, five patients (mean age:
45.8 ± 15 years; 3 men and 2 women) underwent CT-guided transhepatic drainage of postoperative
abdominal abscess in our department and were included in this case series.
Patient Data
The demographic and clinical data, surgical details, laboratory results, indication
for drainage, and postprocedure data were obtained from institutional medical records.
The imaging and procedure-related details were obtained from the departmental picture
archiving and communication system and patient records. All abscesses were confirmed
on the contrast-enhanced CT scan and the decision to drain was made after a discussion
between the gastrointestinal surgeons and the interventional radiologists. In all
the five patients, the collections were surrounded by liver, bowel, and vessels ([Figs. 1] and [2]) and hence a transhepatic route was chosen for drainage. Informed consent for the
procedure was obtained from each patient before the procedure.
Fig. 1 Schematic diagrams (A, B) showing a deep-seated collection (C) surrounded by bowel
loops and vessels with drainage performed through transhepatic route (B).
Fig. 2 Axial contrast-enhanced computed tomography (CT) image in a patient after Whipple's
procedure showing a well-defined collection (white asterisk) surrounded by liver and
duodenum on right side, kidneys and major vessels posteriorly, superior mesenteric
vessels and bowels loops anteriorly and to left.
Drainage Procedure
All procedures were performed under CT guidance on a dual energy CT scanner (Somatom
Definition, Siemens, Erlangen, Germany) under local anesthesia and with patients lying
either in supine or left decubitus position. The transhepatic trajectory of the drainage
was planned on the diagnostic CT scan. A peripheral portion of the liver parenchyma
was chosen as the entrance site to avoid any major hepatic vascular and biliary injury.
Seldinger technique of drainage was used in all patients. An 18-gauge 15-cm long diamond-tipped
two-part puncture needle was used to enter the abscess through the liver ([Fig. 3A]). Then the stylette of the puncture needle was removed and the fluid was aspirated
to ensure the correct placement of needle. Subsequently, the needle was removed over
an extrastiff guidewire (Cook Medical, Bloomington, United States) and the tract was
dilated. Finally, an 8F (n = 3) or 10F (n = 2) pigtail catheter was inserted over the extrastiff guidewire and the position
was confirmed on CT scan ([Fig. 3B]). The catheter was fixed to the skin using sutures and connected to a collecting
bag.
Fig. 3 (A) Axial computed tomography (CT) section with same patient as in [Fig. 2] in left decubitus position showing an18-gauge, 15-cm long diamond-tipped two-part
puncture needle (white arrow) reaching the abscess through the right lobe of liver
(segment 6), between duodenum and right kidney.(B) Axial CT scan at the end of the procedure shows the catheter with tip in the abscess
cavity.
Postprocedure Evaluation
All patients were monitored clinically after the procedure and the drain volume was
recorded daily. Any complications developing during this time was recorded. The catheter
was removed when the patient showed clinical improvement (no fever or leukocytosis)
and when there was no output in the catheter for more than 48 hours. In doubtful cases,
CT scan was repeated to assess response. The technical and clinical success of the
procedure and the duration of catheter drainage were also assessed.
Results
Demographic and clinical details of the patients are described in [Table 1]. The surgical procedures were Whipple's pancreatoduodenectomy in four patients (80%)
and gastrectomy in one patient (20%). All patients presented with fever and elevated
total leucocyte counts (mean: 16.5 ± 8.7 × 103/µL). Mean duration between the surgery and insertion of transhepatic drainage catheter
was 12.6 ± 5.9 days (range: 9–23 days). All catheter drainages were performed within
24 hours of diagnosis of the abscess by CT scan. The mean longest diameter of the
abscess was 8.3 ± 1.15 cm.
Table 1
Demographic clinical and laboratory details of the patients
|
Patient 1
|
Patient 2
|
Patient 3
|
Patient 4
|
Patient 5
|
|
Age, sex
|
43, male
|
42, female
|
60, male
|
60, female
|
24, male
|
|
Presenting complaints
|
Pain in RUQ and progressive jaundice × 2 mo
|
Pain in RUQ and progressive jaundice × 3 mo
|
painless progressive jaundice × 1 mo
|
Fever and painless progressive jaundice × 7 mo
|
Corrosive ingestion under alcohol influence
|
|
Primary diagnosis
|
Periampullary carcinoma
|
Periampullary carcinoma
|
Periampullary carcinoma
|
Mid-CBD cholangiocarcinoma
|
Acute corrosive injury with sealed gastric perforation
|
|
Surgery performed
|
Whipple's procedure
|
Whipple's procedure
|
Whipple's procedure
|
Whipple's procedure
|
Total gastrectomy with tube duodenostomy and feeding jejunostomy
|
|
Symptoms suggesting collection/abscess during post operative period
|
Pain abdomen
Fever
|
Pain abdomen
Fever
|
Pain abdomen
Fever
|
Pain abdomen
Fever
|
Pain abdomen
Fever
|
|
Time between surgery and symptoms (days)
|
20
|
10
|
7
|
8
|
7
|
|
WBC counts prior to drainage (n × 103/µL)
|
12.7
|
12.5
|
11.5
|
12
|
33.9
|
|
Duration between surgery and transhepatic drainage (d)
|
23
|
12
|
9
|
9
|
10
|
|
Access route for drainage/ Segment of liver
|
Right lobe of liver/ segment 6
|
Right lobe of liver/ segment 6
|
Right lobe of liver/ segment 6
|
Right lobe of liver/ segment 6
|
Left lobe of liver/ segment 3
|
|
Size of drainage catheter
|
8-Fr
|
8-Fr
|
10-Fr
|
10-Fr
|
8-Fr
|
|
Duration of the procedure (min)
|
31
|
21
|
22
|
37
|
35
|
|
Procedure or catheter-related complications
|
Nil
|
Nil
|
Nil
|
Nil
|
Nil
|
|
Duration of catheter drainage (days)
|
9
|
15
|
16
|
8
|
12
|
|
WBC counts at catheter removal
(n × 103/µL)/ fever status
|
10.2/ afebrile
|
8.8/ afebrile
|
4.96/ afebrile
|
2.46/ afebrile
|
10.7/ afebrile
|
Abbreviations: CBD, common bile duct; RUQ, right upper quadrant; WBC, white blood
cell.
All catheters were placed via the preplanned route with a technical success rate of
100%. Four out of five (80%) abscesses were drained through the right lobe of liver,
while one (20%) was through the left lobe. The mean total time for catheter drainage
procedure from the initial scan including patient positioning and preparation, till
the catheter fixation to the skin using sutures was 29.2 ± 7.4 minutes.
There were no procedure-related complications in any patient. Mean duration of catheter
drainage was 12 ± 3.5 days. All patients had successful drainage of the abscess with
resolution of fever and normalization of total leucocyte count. Thus, the clinical
success was 100%. Two patients had follow-up CT scans after a mean period of 12 days
after catheter drainage. The scans showed near-complete resolution of the abscess
([Fig. 4]). All patients were discharged after catheter removal in stable condition. The findings
are summarized in [Table 1].
Fig. 4 Axial computed tomography (CT) scan during follow-up of the same patient after catheter
removal showing near-complete resolution of the abscess (white arrow).
Discussion
Image-guided percutaneous drainage with catheter is the treatment of choice in the
management of patients with intra-abdominal abscesses.[3] In a majority of cases, the percutaneous access route is straightforward and the
abscesses are drained under either USG or CT guidance. However, uncommonly, the simple
percutaneous approach may not be possible because of the deep and inaccessible location
of the abscess and the intervening critical structures like bowel, liver, and vessels.[4] Various CT-guided approaches with modifications and improvisations have been described
in the literature for the drainage of such intra-abdominal and pelvic abscesses.[4] This entails the use of angled gantry technique, hydrodissection to displace the
critical structures, and traversing less critical structures like stomach, rectum,
vagina, and liver.[8]
The transhepatic approach for the drainage of abscesses was first described by Mueller
et al in 1985.[6] They successfully and safely performed this procedure in 12 patients, 8 of whom
had postoperative collections requiring drainage. Subsequently, the safety and efficacy
of this approach in the drainage of deep-seated abscesses or collections were confirmed
by Yamakado et al[7] in 12 patients, Ciftci et al[5] in 30 patients, and Zhao et al[9] in 32 patients. Image-guided catheter drainage of deep liver abscesses and obstructed
biliary ducts routinely involves a pathway via the liver parenchyma and is considered
safe.[10]
[11] However, compared with liver abscess or biliary drainage, transhepatic drainage
of deep-seated abscesses necessitates the passage of the catheter with breach of liver
capsule at two points. Few suggestions have been made for safe transhepatic drainage.[4] These include confirming a normal coagulation profile, using the shortest path through
the liver, and avoiding major blood vessels, dilated bile ducts, and other organs
like the gallbladder. The third and sixth segments of liver parenchyma are usually
preferred for transhepatic access, because of the absence of major vascular and biliary
branches.[5] Furthermore, it is important to ensure that the side-holes of the catheter are in
the abscess cavity and not in the liver parenchyma to avoid liver infection. However,
when there is uncorrectable coagulopathy, liver cirrhosis, and dilated bile ducts,
this procedure should be avoided.[6] CT is generally preferred over USG as the guiding modality since it allows better
visualization of the deeper structures in the abdomen and is not limited by bowel
gas or obesity. Further, in the postoperative period, inflammation also affects visualization
by USG. Hence, we and most of the cases in the studies reported used CT guidance for
transhepatic drainage. However, one study has reported the safe use of USG as guidance
for transhepatic drainage of abscesses.[5]
Mean time required for the procedure as reported in the published series ranged from
12 to 19 minutes which was less than our series. However, the studies did not define
what constituted procedure time. The technical success of transhepatic drainage of
abdominal abscesses ranges from 97 to 100% in the reported studies.[5]
[6]
[7]
[9] The clinical success rates, which are based on clinical improvement and resolution
of the abscess on imaging, ranged from 94 to 100%. The reasons for clinical failure
were the need for surgery, recurrence of the abscess after catheter removal, and death
due to sepsis.[5]
[7]
[9] Our short series had 100% technical and clinical success rates. The mean duration
of catheter drainage ranged from 18 to 33 days with patients having a fistula or leak
requiring placement for a longer time.[5]
[7]
[9] The common sizes of the catheters used were 8F and 10F. Our series also used 8F
and 10F catheters for a mean duration of 12 ± 3.5 days.
No major procedure or catheter-related complications occurred in any of our patients.
Various complications associated with transhepatic intra-abdominal abscess drainage
include vascular and biliary injury, bacteremia, liver abscess due to secondary bacterial
contamination through the catheter track, and catheter-related minor complications
such as dislodgment, accidental removal, obstruction, and kinking of drainage catheters.[5]
[7] Fistulous communications with small bowel and biliary tree have also been reported.[5] Catheter-related complications were significantly more frequent with 8F compared
to 10F or 12F catheters.[5] Although accidental catheter dislodgement is uncomplicated, a case of fatal liver
injury and bleeding has been reported.[12] Although there is a theoretical possibility of higher procedure-related pain due
to the puncture of liver capsule at two sites, this has not been addressed in the
published literature.
There have been reports in literature regarding endoscopic ultrasound (EUS)-guided
drainage of postoperative abdominal collections and abscesses.[13]
[14] The clinical success rates of this procedure were reported to be 70 to 80%.[13] Significant advantage of EUS-guided drainage over percutaneous or surgical drainage
is the lack of longstanding fistula.[13] There is also a decreased risk of injury to interposed blood vessels. However, due
to modified bowel anatomy in postoperative cases, many of these abscesses and collections
may not be accessible under EUS guidance. To our knowledge, there are no reports comparing
the outcomes of EUS-guided drainage and image-guided percutaneous drainage for the
treatment of postoperative intra-abdominal abscesses.
In conclusion, the transhepatic approach is safe and effective for the drainage of
inaccessible postoperative abdominal collections or abscesses where a standard percutaneous
approach is not possible.
