| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 All authors: Department of Radiology, Division of Abdominal Imaging and Intervention, Massachusetts General Hospital, 55 Fruit St., White 270, Boston, MA 02114.
Received June 9, 2004;
accepted after revision September 7, 2004.
Address correspondence to D. A. Gervais
(dgervais{at}partners.org).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. A retrospective evaluation of all gallbladder drainages performed over a 5-year period revealed 163 patients (range, 7–98 years) who underwent percutaneous cholecystostomy catheter placement. Medical records and imaging studies were reviewed to assess the events at catheter removal (e.g., inadvertent removal, controlled removal with cholangiography without tract imaging, or controlled removal with cholangiography with tract imaging) and the incidence of major and minor bile leaks.
RESULTS. The events at catheter removal were assessed in 66 patients. Group 1 was 45 patients whose catheters were removed after a minimum of approximately 3 weeks with a cholangiogram that established cystic and common duct patency and no imaging of the tract. Catheters were not removed until the patient recovered from acute illnesses that contributed to acalculous cholecystitis. Group 2 was 11 patients managed similarly to group 1 except that tract imaging was performed at catheter removal. Group 3 was 10 patients whose tubes came out inadvertently without cholangiogram or tract imaging. Two major (group 2 and group 3) and two minor (group 2) bile leaks occurred. No bile leaks occurred in group 1 (p = 0.006).
CONCLUSION. Major bile leaks occurred in 3% of patients, and minor leaks occurred with equal frequency. Tract imaging may not be necessary in patients with small-bore gallbladder catheters who have recovered from critical illness, show patent cystic and common ducts, and have had catheters for 3–6 weeks.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although rare, one potentially serious complication of percutaneous cholecystostomy is a bile leak into the peritoneum, resulting in bile peritonitis and, occasionally, sepsis [2, 11, 12]. Clinically significant bile leak is also a potential complication of catheter removal when the gallbladder puncture site can allow bile to leak out of the gallbladder and into the peritoneum. For this reason, D'Agostino et al. [13] and other investigators have emphasized the importance of the establishment of a mature or sealed tract before catheter removal [13]. The tract forms around the catheter and matures over time as the cellular and physiologic processes involved in wound healing seal it off from adjacent tissues, thus preventing fluid from leaking into surrounding tissues. D'Agostino et al. found that tracts matured by 20 days in all cases in their series. Thus, over this time, catheter removal is expected to become safer with respect to potential bile leaks.
We undertook this study to assess the incidence of bile leak in our experience with percutaneous cholecystostomy catheter removal and to report our experience with a catheter removal strategy based on a minimum duration of cholecystostomy catheter drainage and patient recovery from contributory illness. We also sought to use our experience to evaluate clinical and imaging guidelines to ensure safe catheter removal.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Medical Record Review and Definitions of Bile Leak
Duration of drainage before catheter removal and the events at removal
(controlled versus inadvertent removal, development of bile leak) were noted.
Clinical evidence of bile leak was defined as the acute onset of abdominal
pain with abdominal rigidity and peritoneal signs and/or hemodynamic
instability (hypotension) occurring within 4 hr of catheter removal. Imaging
evidence of bile leak was defined as intraperitoneal leak of contrast material
injected into the gallbladder or cholecystostomy tract, or biloma occurring
within 30 days of catheter removal. Bile leaks were further subdivided into
major and minor based on standards adopted by the Society of Interventional
Radiology [14]. A major bile
leak was defined as one that required surgery, another interventional
procedure, prolonged hospital stay, or ICU admission. A minor bile leak was
defined as a one that was self-limiting and did not require surgery,
interventional procedures, or ICU admission.
Percutaneous Cholecystostomy Placement and Management
Percutaneous cholecystostomy was performed in critically ill patients with
strong clinical and imaging evidence of cholecystitis or in patients with
cholecystitis otherwise too ill to undergo immediate surgery. In all cases,
percutaneous cholecystostomy catheter placement was performed jointly by an
interventional radiology fellow under supervision of a staff interventional
radiologist. Locking pigtail catheters (8.5-French Dawson-Mueller, Cook) were
used in all cases. The gallbladder was identified with a sonogram
(Fig. 1), and a puncture site
was chosen and prepared with a sterile field. Patients who could travel to the
radiology department were administered IV sedation consisting of fentanyl
citrate, 25–150 µg, and midazolam hydrochloride, 0.5–2 mg. Most
ICU patients were already moderately sedated based on their comorbid
conditions. If needed, the ICU nurses gave additional IV narcotics to these
patients for sedation. For local analgesia, lidocaine 1% (Xylocaine MPF,
Astra) was administered in the subcutaneous tissues and along the liver
capsule. A transhepatic approach, defined as the catheter tract traversing the
liver parenchyma, was attempted when possible. The catheter was placed into
the gallbladder using the free-hand trocar technique under real-time sonogram
imaging guidance (Fig. 2A,
2B). CT was used only if the
gallbladder could not be well seen via sonogram. Sonogram guidance was used in
160 patients and CT guidance in three patients. The catheters were managed
jointly by the Interventional Radiology Service and the ICU team. Each nursing
shift, catheter flushes with 0.9% saline were performed to maintain catheter
patency.
|
|
|
Cholecystostomy Catheter Removal
Our standard practice for catheter removal in controlled circumstances
required a minimum of approximately 3 weeks of drainage and recovery from the
illness that required ICU admission and/or predisposed to acalculous
cholecystitis. This minimum duration of drainage was to allow adequate time
for a well-formed mature tract to develop and was based on previous work by
D'Agostino et al. [13]. The
requirement that the patients were not ill ensured that the patients did not
remain at high risk of developing recurrent cholecystitis. Thus, all patients
were out of the ICU before catheter removal and were pending discharge at
catheter removal or were discharged with the catheter in place and returned
later for elective catheter removal.
An additional component of our standard catheter removal strategy was a cholecystogram/cholangiogram obtained by injection of the cholecystostomy catheter with contrast material under fluoroscopy [15] (Fig. 3). Frontal and oblique spot films were obtained to confirm patent cystic and common ducts, thereby excluding an obstruction that would predispose to recurrence of cholecystitis after catheter removal. In some patients, at the discretion of the staff interventional radiologist supervising the cholangiogram and catheter removal, imaging of the catheter tract was performed (Fig. 4). All tract imaging was performed by a single staff radiologist. Tract imaging was performed by catheter injection through a dilator after catheter removal over a wire or by injection of the gallbladder catheter withdrawn over a wire with injection via a Y-adaptor. The wire allowed access to the gallbladder for catheter replacement in the event that an immature tract was shown by a leak of contrast material into the peritoneum.
|
|
The plan for all patients was a cholangiogram with or without a tract injection. However, in a few cases, the tube came out inadvertently. Thus, the study population formed three groups: those in whom the catheter was removed with a minimum duration of drainage and exclusion of cystic and bile duct obstruction by cholangiogram (group 1), those in whom both a cholangiogram and tract imaging were performed before catheter removal after minimum drainage period (group 2), and those in whom the tube was removed inadvertently without the benefit of a cholangiogram (group 3).
Statistics
Student's t test was used to assess for possible differences among
group means for duration of catheter drainage. Differences in incidence of
bile leak among groups was assessed with Fisher's exact test. A p
value of 0.05 or less was considered significant.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Thus, a total of 66 patients were evaluated for bile leak after percutaneous cholecystostomy catheter removal. Of these patients, the initial cholangiogram confirmed cystic and common duct patency in all but five of the patients. These five patients underwent continued catheter drainage and a subsequent cholangiogram showed patent ducts before tube removal.
Forty-five of these 66 patients formed group 1, who had the tube removed after a catheter cholangiogram confirming cystic duct and common duct patency, with a mean of 42.6 days of drainage (range, 17–223 days). Eleven of the 66 patients formed group 2 in whom the catheters were removed with catheter cholangiogram, confirming ductal patency and tract imaging at a mean of 40.9 days after catheter placement (range, 21–66 days). Group 3 consisted of the remaining 10 patients in whom the catheter was removed inadvertently at a mean of 18.2 days (range, 1–71 days) after catheter placement. The number of patients who underwent catheter removal at each weekly interval is detailed for each group in Table 1.
|
Incidence of Bile Leaks
A total of four bile leaks, two major and two minor leaks, were identified
in the 66 patients, for an overall incidence of 6.0%, and a 3.0% rate each for
minor and major leaks. The leaks resulted from catheter removal at 14, 27, 29,
and 42 days after catheter placement. None of the bile leaks occurred in group
1. Three bile leaks (one major, two minor) occurred in group 2 at an incidence
of 27% (3/11). One of the bile leaks occurred in group 3, the group of 10
patients whose tube came out inadvertently, a 10% incidence. The lower
incidence of bile leak in group 1 was significant when compared with the
incidence in group 2 (p = 0.006). Moreover, there was no significant
difference in the duration of drainage between group 1 (cholangiogram only)
and group 2 (cholangiogram and imaging of the tract), (p = 0.89). As
would be expected, the catheters inadvertently removed had a significantly
shorter duration of drainage than the catheters removed under controlled
conditions (p = 0.04).
Events at Major Bile Leaks
In group 3, a 64-year-old man was combative because of an altered mental
status, and he removed his catheter traumatically 14 days after percutaneous
cholecystostomy. He immediately experienced transient hypotension that was
successfully relieved with fluid infusion. No further immediate intervention
was needed, but a sonogram 3 days after catheter removal showed a large biloma
that was successfully managed with percutaneous drainage. Thus, this patient
had both clinical and imaging evidence of bile leak.
The other major bile leak occurred in a 27-year-old HIV-positive man from group 2, who was undergoing treatment for lymphoma. Cholecystostomy catheter removal was performed electively at 29 days with a cholangiogram and tract imaging (group 2). The tract imaging showed no leak, and the catheter and wire were removed. The patient then experienced an immediate onset of right upper quadrant pain, abdominal rigidity, and peritoneal signs, followed by hemodynamic instability requiring urgent cholecystectomy. Surgical findings confirmed intraperitoneal bile. This patient had clinical evidence of bile leak.
Events at Minor Bile Leaks
The first of the minor leaks occurred in a 42-year-old woman in group two
with extensive hepatic metastases from colorectal cancer. Catheter removal was
elective at 27 days, and both cholangiogram and tract imaging were performed
(group 2). Although the tract imaging showed no leak of contrast material, she
experienced immediate abdominal pain, requiring narcotics and overnight
admission for pain management on the basis of clinical evidence of a bile
leak. The patient remained hemodynamically stable throughout her course and
was discharged the next morning.
The second minor bile leak occurred in a 39-year-old diabetic man with renal failure from group 2. On day 42 after catheter placement, the patient presented as an outpatient for tube removal. The tract imaging showed a small leak, but access was lost and could not be reestablished. He experienced 5 minutes of severe right upper quadrant pain that resolved completely without treatment. The patient did not experience hypotension. Thus, this patient had imaging and clinical evidence of a bile leak. He was discharged home without further incident.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our rate of bile leaks at catheter removal is similar to that of Picus et al. [11], who reported 58 consecutive patients who underwent cholecystolithotomy, but our rate of major bile leaks is lower. Based on their four bile leaks, Picus et al. recommended tractograms in all cases. However, all four of their patients required surgery or an additional drainage procedure, compared with only two of our patients. This difference may be explained by the fact that the tracts in our series differed from Picus et al. and that of other investigators. The tracts in the Picus et al. series were larger, dilated up to 18 French for stone removal, and underwent serial manipulation (tract dilatation and stone removal later) as the gallbladder access was used for stone removal in many cases. The smaller 8-French tracts used in our series for simple gallbladder drainage and the lack of subsequent manipulations may account in part for our lower rate of surgery and/or percutaneous drainage when bile leaks did occur in our series. In a separate report, Picus et al. [12] also reported two cases of persistently immature tracts. In one patient with renal failure and cardiovascular disease, a 10-French tract remained immature at 63 days. In a second patient, an 18-French tract remained immature at 40 days. In a commentary based on these two cases, Vogelzang [16] also recommended tract imaging before removal of gallbladder catheters.
Our study includes the largest cohort of patients reported to date addressing the issue of cholecystostomy catheter management at removal. D'Agostino et al. [13] studied cholecystostomy tract formation by imaging at various time points in 28 patients and reported that a 20-day period was sufficient for tract maturation to occur. Hatjidakis et al. [17] performed a prospective evaluation of the events at catheter removal in 33 patients. In three of these 33 patients, the catheter was inadvertently removed 3 to 5 days after drainage, resulting in bile leaks in two patients. Thirty of these 33 patients were evaluated with a cholangiogram and tract imaging, confirming that most tracts (28/30) were mature within 21 days and all were mature by 30 days. Hatjidakis et al. were able to compare the tract maturation in gallbladder drainages performed via the transhepatic approach with those performed with the transperitoneal approach and showed that the transperitoneal tracts required significantly longer to mature than the transhepatic tracts.
Our experience of no clinically evident bile leaks in the 45 patients in group 1 who underwent removal of small cholecystostomy catheters after a mean duration of drainage of 42.6 days, clinical recovery from contributory underlying illness, and catheter cholangiogram confirming cystic and bile duct patency provides strong empiric support for this management strategy. Based on our experience, imaging of the percutaneous cholecystostomy tract may not be necessary in patients who have recovered from underlying contributory illness with patent cystic and common ducts and with a minimum duration of drainage of 21 days. Moreover, if the minimum duration of drainage further extends to 6 weeks, we speculate that all but one (a minor leak) of the bile leaks in our series may have been avoided.
Management by clinical assessment is grounded in the basic science of wound healing. There are three phases to wound healing. The first phase is the inflammatory phase, in which white blood cells, platelets, and clotting factors are recruited to the wound site. This phase begins immediately. The second phase is the proliferative phase, which begins between 3 and 7 days in the immunocompetent. During this phase, fibroblasts proliferate around the fibrin-fibronectin matrix and begin laying down predominantly type 3 collagen and to a lesser degree, type 1 collagen. Collagen equilibrium is reached at 2 to 3 weeks, and this begins the remodeling phase. During this phase, type 3 collagen is exchanged for the stronger type 1 collagen. Thus, by 3 weeks, patients who have recovered from underlying illness should have a tract that does not allow a bile leak. Disease states and pharmacologic agents that affect the immune response such as chemotherapeutics, steroids, illness requiring ICU support, and immunocompromised status have a deleterious affect on wound healing. Patients whose predisposing illnesses persist after gallbladder drainage may require a longer period for their cholecystostomy tracts to mature. A persistent debilitated state may in part also explain the two patients with persistent immature tracts reported by Picus et al. [12]. It may also explain our most serious complication, the bile leak requiring immediate surgery, as this patient was receiving treatment for lymphoma. Although he was an outpatient, his tract likely formed more slowly based on his chemotherapy, and ongoing and/or recent chemotherapy can be added to the criteria that would necessitate leaving the tube in place. A tract injection was performed in this case but did not identify the bile leak.
Our rate of bile leaks in the patients who underwent tract injections compared with those who did not is interesting, but it is difficult to draw firm conclusions that explain this result. The higher incidence cannot be attributed to shorter duration of drainage because the duration of drainage did not differ between patients who had tube and tract injections and those who had only the tube injection. Although the additional manipulation necessitated by tract injections might conceivably predispose to bile leak by disrupting a tenuous tract, the small number of tract injection cases and the retrospective nature of our study make this assertion impossible to prove. A prospective randomized comparison of the two catheter removal strategies, tube injection only versus tube and tract injection, would be required to confirm that there is a real and significant difference in the incidence of bile leaks between the management strategies.
The literature strongly suggests tract imaging be performed in all cases before catheter removal [11, 12]. However, many clinical series include patients with much larger tracts and with multiple manipulations. Our experience, on the other hand, supports a clinical management strategy as described herein. Removal of a transhepatic 8-French cholecystostomy catheter after 6 weeks of drainage in patients who are well recovered from their antecedent illnesses, with documented patent cystic ducts and common bile ducts, is not likely to result in a clinically significant bile leak, and tract imaging is not necessary.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
