AJR 2005; 184:S70-S72
© American Roentgen Ray Society
Pylephlebitis After CT-Guided Percutaneous Liver Biopsy
Richa Tandon1,
Ashley Davidoff2,
Michael G. Worthington3 and
John J. Ross3
1 Department of Medicine, Caritas St. Elizabeth's Medical Center, Boston, MA
02135.
2 Department of Radiology, Caritas St. Elizabeth's Medical Center, Boston, MA
02135.
3 Division of Infectious Diseases, Caritas St. Elizabeth's Medical Center, 736
Cambridge St., Boston, MA 02135.
Received February 6, 2004;
accepted after revision April 19, 2004.
Address correspondence to J. J. Ross
(jrossmd{at}cchcs.org).
Introduction
In the preantibiotic era, pylephlebitis, or septic thrombophlebitis
of the portal venous system, was a familiar, usually lethal consequence of
intraabdominal sepsis. It is encountered today as a rare complication of
diverticulitis, appendicitis, malignancy, inflammatory bowel disease, and
vasculitis. Pylephlebitis is often not suspected clinically and may be a
serendipitous finding on abdominal CT. Mortality is still substantial, at up
to 32% [1,
2].
We describe a case of pylephlebitis after CT-guided percutaneous liver
biopsy. To our knowledge, pylephlebitis has not been previously reported as a
procedural complication. Factors influencing the risk for infectious
complications of percutaneous liver biopsy are discussed.
Case Report
A Whipple procedure was performed in a 46-year-old woman for primary
resection of a pancreatic vasoactive intestinal peptide-secreting tumor
(VIPoma). Two years later, numerous liver nodules were found on an abdominal
CT scan. Percutaneous biopsy was performed under CT guidance with the
skinny-needle technique and an 18-gauge needle, which revealed a metastatic
VIPoma (Figs. 1A and
1B). Coaxial technique for the
biopsy was used, with several passes of the skinny needle through the outer
needle to optimize tissue sampling. One day after the biopsy, the patient
developed a high fever with chills and rigors. On admission, she was febrile
to 39°C. Abdominal examination revealed normal bowel sounds with mild
right upper quadrant tenderness without hepatomegaly or ascites. The remainder
of the examination was unremarkable. The WBC was 3.5 x 109/L,
with 59% neutrophils and 6% band forms. Serum bilirubin was normal. Serum
aspartate aminotransferase was 42 U/L (normal range, 14-36), alanine
aminotransferase 87 U/L (9-52), and alkaline phosphatase 259 U/L (38-126).
Blood cultures grew Klebsiella oxytoca, sensitive to all antibiotics
tested except ampicillin. Repeat abdominal CT scan showed a low attenuation
perfusion defect in the right lobe of the liver and a hypodense area in the
right portal vein (Figs. 1C,
1D,
1E). This finding was
compatible with the diagnosis of right portal vein thrombosis and was not
present on the preprocedure CT scan.
View larger version (126K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —Hepatic metastasis from VIPoma in 46-year-old woman. CT scan
shows hepatic metastasis (thick arrow) close to jejunal aspect of
choledochojejunostomy (two thin arrows). Portal vein branch to
anterior segment of right lobe is patent (single thin arrow).
|
|
View larger version (128K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. —Hepatic metastasis from VIPoma in 46-year-old woman.
CT-guided needle biopsy, with needle adjacent to metastasis (thick
arrow), in close proximity to choledochojejunostomy site (two thin
arrows).
|
|
View larger version (129K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C. —Hepatic metastasis from VIPoma in 46-year-old woman.
Pylephlebitis after CT-guided percutaneous liver biopsy. CT scan 1 week after
needle biopsy (C) reveals portal vein thrombosis of anterior segmental
branch to previously patent right lobe (single thin arrow). CT scan 1
week after needle biopsy (D) with portal vein thrombosis, multicentric
perfusion defects, and extensive involvement of segment VII (thin
arrows). CT scan 6 months after biopsy (E) shows improved
perfusion but persistent thrombosis in one portal radicle (thin
arrow). No intrahepatic biliary dilatation is noted. Thick arrow in
C denotes hepatic metastasis.
|
|
View larger version (125K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D. —Hepatic metastasis from VIPoma in 46-year-old woman.
Pylephlebitis after CT-guided percutaneous liver biopsy. CT scan 1 week after
needle biopsy (C) reveals portal vein thrombosis of anterior segmental
branch to previously patent right lobe (single thin arrow). CT scan 1
week after needle biopsy (D) with portal vein thrombosis, multicentric
perfusion defects, and extensive involvement of segment VII (thin
arrows). CT scan 6 months after biopsy (E) shows improved
perfusion but persistent thrombosis in one portal radicle (thin
arrow). No intrahepatic biliary dilatation is noted. Thick arrow in
C denotes hepatic metastasis.
|
|
View larger version (129K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1E. —Hepatic metastasis from VIPoma in 46-year-old woman.
Pylephlebitis after CT-guided percutaneous liver biopsy. CT scan 1 week after
needle biopsy (C) reveals portal vein thrombosis of anterior segmental
branch to previously patent right lobe (single thin arrow). CT scan 1
week after needle biopsy (D) with portal vein thrombosis, multicentric
perfusion defects, and extensive involvement of segment VII (thin
arrows). CT scan 6 months after biopsy (E) shows improved
perfusion but persistent thrombosis in one portal radicle (thin
arrow). No intrahepatic biliary dilatation is noted. Thick arrow in
C denotes hepatic metastasis.
|
|
Gatifloxacin, cefepime, and heparin were administered by peripheral vein.
The patient's fever abated. She was discharged 8 days later on warfarin and
oral ciprofloxacin but was readmitted with recurrent fever the next day.
Persistent portal venous thrombosis was present, but the perfusion defect was
improved. Several repeat blood cultures were negative. The patient was treated
with IV ceftriaxone, and anticoagulation was continued with warfarin. She
responded to this therapy, was discharged after 6 days, and completed an
8-week course of ceftriaxone. Anticoagulation was continued for 3 months after
discharge. Serial CT scans after discharge showed a gradual improvement in
liver perfusion, possibly secondary to increased blood flow through the
hepatic artery. However, the portal vein thrombus did not resolve
completely.
The patient was still alive 1 year after completing antibiotic therapy. She
responded poorly to chemotherapy and recently underwent radiofrequency
ablation and wedge resection of liver metastases.
Discussion
Pylephlebitis is classically associated with intraabdominal infections such
as diverticulitis, appendicitis, and cholangitis. It may also be seen in
hypercoagulable states, when bacteria seed a preexisting portal vein
thrombosis [3]. Patients
present with fever, are usually bacteremic, and typically, but not invariably,
have abdominal pain. Surgical or percutaneous drainage is sometimes required
in addition to antibiotics [1,
2]. The benefits of
anticoagulation remain unproven
[3]. Refractory cases may
require placement of a percutaneous drainage catheter into the portal vein
[4].
Pylephlebitis has not been reported as a complication of liver biopsy. In
this patient's case, unique anatomic considerations may have played a role.
The passage of skinny needles through the small bowel does not normally lead
to complications, presumably because the bowel is a muscular, self-sealing
tube [5,
6]. With a coaxial technique,
the depth of each pass is varied. Given the close apposition of the liver
lesion and the choledochojejunostomy, it is conceivable and probable that the
bowel was transgressed. Normally, as bile does not favor microbial growth, the
bacterial burden in the upper small bowel is very low, from 10-1,000
organisms/mL [7]. However,
after a Whipple procedure, loops of bowel may have stasis and higher bacterial
colonization, increasing the potential for infection after needle passage.
Patients with bilioenteric anastomosis are also at an elevated risk for liver
abscess from portal bacteremia
[8]. Thus, the patient's
previous Whipple procedure may have pre-disposed her to pylephlebitis. It may
be relevant that Klebsiella oxytoca bacteremia is most commonly
reported in the setting of hepatobiliary disease
[9].
Pylephlebitis in this patient may have arisen from another mechanism.
Asymptomatic bacteremia occurring within 24 hr of liver biopsy is documented
in 2-13% of patients [10]. The
patient's pancreatic neoplasm may have caused a hypercoagulable state. A
preexisting thrombus in the portal system could have become infected from
postbiopsy bacteremia. However, no thrombus was identified on prebiopsy CT
images.
Although hypercoagulability may have pre-disposed this patient to
pylephlebitis, the intimate temporal relationship of the liver biopsy to the
development of pylephlebitis and the rarity of pylephlebitis strongly suggest
that the procedure was the proximate cause. Pylephlebitis is seldom seen, with
only 19 case reports in the English language literature from 1979 to 1993.
Most episodes of pylephlebitis are clearly related to another focus of
intraabdominal infection [1].
The lack of a primary source of infection in this case also argues for the
liver biopsy as the inciting factor.
Although the penetration of loops of bowel by skinny needles is generally
viewed as safe, infectious complications may increase when loops of bowel with
bacterial stasis are involved. Interventional radiologists should be aware of
the possibility of pylephlebitis as an unusual complication of liver
biopsy.
References
- Plemmons RM, Dooley DP, Longfield RN. Septic thrombophlebitis of
the portal vein (pylephlebitis): diagnosis and management in the modern era.
Clin Infect Dis1995; 21:1114
-1120[Medline]
- Singh P, Yadav N, Visvalingam V, Indaram A, Bank S. Pylephlebitis:
diagnosis and management. Am J Gastroenterol2001; 96:1312
-1313[Medline]
- Baril N, Wren S, Radin R, Ralls P, Stain S. The role of
anticoagulation in pylephlebitis. Am J Surg1996; 172:449
-453[Medline]
- Pelsang RE, Johlin F, Dhadha R, Bogdanowicz M, Schweiger GD.
Management of suppurative pylephlebitis by percutaneous drainage: placing a
drainage catheter into the portal vein. Am J
Gastroenterol 2001;96:3192
-3194[Medline]
- Cho KJ, Elta GH. Interventional radiology. In: Yamada T, ed.
Textbook of gastroenterology, 4th ed. Philadelphia,
PA: Lippincott William & Wilkins, 2003:3247
-3273
- Shankar S, vanSonnenberg E, Silverman SG, Tuncali K. Interventional
radiology procedures in the liver. Biopsy, drainage, ablation. Clin
Liver Dis 2002;6:91
-118[Medline]
- Finegold SM. Anaerobic bacteria: general concepts. In: Mandell GL,
Bennett JE, Dolin R, eds. Mandell, Douglas, and Bennett's
principles and practice of infectious diseases, 5th ed.
Philadelphia, PA: Churchill Livingstone, 2000:2519
-2537
- Kubo S, Kinoshita H, Hirohashi K, Tanaka H, Tsukamoto T, Kanazawa
A. Risk factors for and clinical findings of liver abscess after
biliary-intestinal anastomosis. Hepatogastroenterology1999; 46:116
-120[Medline]
- Kim BN, Ryu J, Kim YS, Woo JH. Retrospective analysis of clinical
and microbiological aspects of Klebsiella oxytoca bacteremia over a
10-year period. Eur J Clin Microbiol Infect Dis2002; 21:419
-426[Medline]
- Moreira Vicente VF, Hernandez Ranz FM, Ruiz Del Arbol L, Bouza EP.
Septicemia as a complication of liver biopsy. Am J
Gastroenterol 1981;76:145
-147[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Introduction
Case Report
