Sepsis Following Liver Biopsy in a Liver Transplant Recipient: Case Report and Review of Literature

Abstract

Percutaneous liver biopsy is a relatively safe procedure with low complication rates. Infections following liver biopsy are uncommon and can lead to a poor outcome. There are limited data on liver biopsy–related infections among liver transplant (LT) recipients. Also, there is a paucity of data regarding the use of prophylactic antibiotics in LT patients undergoing percutaneous liver biopsy. We report a case of systemic sepsis following percutaneous liver biopsy in a LT recipient with choledochojejunal anastomosis. This was followed by severe rejection and deterioration of liver function and recurrence of primary sclerosing cholangitis (PSC) to the extent that he has been listed for retransplantation. This case report emphasizes the potential risk of sepsis in LT recipients with bilioenteric anastomosis undergoing percutaneous liver biopsy. This increased risk may warrant periprocedural broad spectrum antibiotic prophylaxis, in this subgroup of patients.

Percutaneous liver biopsy is a relatively safe procedure, and the reported complication rates are less than 1%.¹ Infections following liver biopsy are uncommon and can lead to a poor outcome. The incidence of infectious complications following a percutaneous liver biopsy ranges from 0% to 3.15%.² There are limited data on liver biopsy–related infections among liver transplant (LT) recipients. The risk factors for biopsy-related infections in hepatic allograft recipients remain an area of contention. Also, there is a paucity of data regarding the use of prophylactic antibiotics in LT patients undergoing percutaneous liver biopsy. We report a case of percutaneous liver biopsy induced sepsis in a deceased donor liver transplant (DDLT) recipient with bilioenteric anastomosis.

Case report

A 22-year-old male, who underwent orthotopic DDLT with Roux-en-y choledochojejunostomy 9 months back, was found to have deranged liver function tests (LFTs) on routine follow-up investigations (Figure 1). He underwent transplantation for decompensated chronic liver disease due to primary sclerosing cholangitis (PSC) and autoimmune hepatitis (overlap syndrome) (Figure 2). He underwent a choledochojejunostomy in view of the primary diagnosis of PSC. The explant biopsy showed features of cirrhosis with extensive canalicular cholestasis. Bile ducts showed concentric periductal fibrosis. Sections examined from the resected end of common hepatic duct showed ulceration of lining epithelium with underlying inflammatory cells. Sections from the gall bladder showed features suggestive of chronic cholecystitis.

Figure 1.

Showing patient's LFT/TLC values, hospital course and interventions. (TLC: total leucocyte count median range [4000––11000 cells/microliter]; AST: aspartate aminotransferase U/L [5–40]; ALT: alanine aminotransferase U/L [5–45]; ALP: alkaline phosphatase U/L [80–240]; total serum bilirubin mg/dL [0.2–1.2]).

Figure 2.

Sections from liver explant show cirrhotic liver with nodules, separated by fibrous septae (arrows) [A x 40]. The portal tracts show interface hepatitis with regenerative bile ducts [B × 200], and hepatocytes show extensive canalicular cholestasis (arrows) [C ×100]. One subhepatic bile duct shows epitheliotropic dense lymphoid cell infiltrate (arrows) [D x 40].

One month post DDLT, he had an episode of acute cellular rejection that was managed successfully with methylprednisolone pulse therapy (500 mg IV daily for 3 days). Subsequently, he received oral steroids, tacrolimus, and mycophenolate mofetil (MMF) as part of the immunosuppression regimen with good compliance. Eight months following DDLT, he was detected to have a serial elevation of serum aspartate transaminase (AST) and alanine transaminase (ALT), despite adequate immunosuppression. At this presentation, he was on tablet tacrolimus 2 mg BD, MMF 750 mg BD and tablet prednisolone 5 mg OD. Serum tacrolimus level at the time was 7.3 ng/mL.

On ultrasound doppler evaluation, the vessels showed normal color flow pattern and spectral wave form. A contrast-enhanced MRI with MRCP showed no anastomotic stricture, normal intrahepatic biliary radicals and normal hepatic parenchyma. To rule out the possibility of rejection, ultrasound-guided percutaneous liver biopsy was done under aseptic precautions using a coaxial biopsy needle (18 gauge). Histopathological examination showed maintained liver architecture with mild mixed inflammatory cell infiltrate around the portal tracts. Immunohistochemical staining for C4d was negative. No viral inclusions were noted (Figure 3a).

Figure 3.

a: Index post-transplant liver biopsy shows, however, normal liver biopsy cores without significant inflammation in the portal tracts with normal portal vein branches and bile ducts [E and F × 40; G and H × 100]. b: Last post-transplant liver biopsy taken during recent flair shows expansion and dense inflammation of all portal tracts, lymphocytic bile duct infiltration (arrows) [I x 200], portal venulitis (arrows) [J × 200], central venulitis and perivenular hepatocyte necrosis (arrows) [K x 200]. The reticulin stain does not show significant fibrosis [L x 100]. CD8 stain shows lymphocytic venulitis and bile duct injury (arrows) [M ×200], suggesting recurrence of severe acute cellular rejection.

Twelve hours following the liver biopsy, the patient developed high-grade fever with chills. Clinical examination of the biopsy site showed no signs of inflammation. His blood investigations showed raised total leucocyte count (TLC) along with a raised procalcitonin (36 ng/mL; reference value: <0.15 ng/mL). Three subsequent blood cultures taken at the time of fever spikes grew multidrug-resistant Escherichia coli, sensitive only to colistin and tigecycline. Fungal cultures and cytomegalovirus testing were negative. Initially, third-generation cephalosporin was administered parenterally, followed by intravenous colistin in full therapeutic dose (2 MU TDS IV). An ultrasound of the abdomen, 3 days after the start of intravenous antibiotics, showed a small 2 × 2 cm collection in segment 6 of the liver (Figure 4a). Aspiration of the liver lesion yielded a small amount of biliohaemorrhagic fluid (Figure 4b) that was sterile on culture.

Figure 4.

a: Ultrasound image showing 2 x 2 cm an echoic lesion in segment VI of liver with irregular margins (arrow). b: Ultrasound-guided aspiration of the liver lesion yielded 10 mL of bilio-haemorrhagic fluid.

After 2 weeks of intravenous antibiotics, fever resolved and TLCs got normalized. However, his LFTs continued to worsen (Figure 1). Despite a normal-looking MRCP, to rule out the biliary obstruction, a percutaneous transhepatic biliary drainage (PTBD) was done, 3 weeks after the first liver biopsy. PTBD was done in an attempt to decrease the serum bilirubin values. Cholangiogram showed smooth strictures of bile ducts in the perihilar and perianastomotic regions, possibly suggesting the recurrence of sclerosing cholangitis (Figure 5). However, serum bilirubin levels continued to rise. In view of persistently deranged liver functions, a repeat percutaneous liver biopsy was done, 4 weeks after the first liver biopsy. It showed features of severe acute cellular rejection (Banff grade 8/9) with complete loss of bile ducts (Figure 3b). Patient received three doses of methylprednisolone pulse therapy.

Figure 5.

Cholangiogram through percutaneous transhepatic biliary drainage (PTBD) catheter showed smooth strictures of the perihilar ducts (arrows) with mild p.

As he continued to have deranged LFTs, plasma exchange was done. Each time 50 mL/kg body weight of plasma volume was exchanged with fresh frozen plasma (FFP) and human serum albumin as the main replacement fluids. His serum bilirubin showed no improvement despite seven cycles of plasma exchange (Figure 1). In view of the refractory rejection, possible recurrence of PSC and persistent liver dysfunction, the patient was listed for retransplantation. Investigations during the listing for retransplantation have been shown in Table 1.

Table 1.

Showing Blood Investigation (Viral and Autoimmune Markers) at the Time of Listing for Retransplantation.

a) Viral markers	Result (Laboratory reference value/titre)
HBsAg	Nonreactive
Anti-HCV antibodies	Nonreactive
HIV I and II	Nonreactive
CMV IgG	500 IU/mL (>1 IU/mL positive)
CMV IgM	1.76 IU/mL (>1 IU/mL positive)
CMV DNA PCR	Negative
b) Auto antibodies
ANA (Method: IF)	Negative (Titre: 1:100)
ANCA (Method: IF)	Negative (Titre: 1:100)
Anti ds DNA (Method: ELISA)	5 IU/mL (0–100 IU/mL)
ASMA (Method: IF)	Negative (Titre: 1:100)
Total IgG (Method: nephelometry)	283 mg/dL (650–1850 mg/dL)
Total IgA (Method: nephelometry)	86 mg/dL (70–360 mg/dL)

Abbreviations: CMV: cytomegalovirus; DNA: deoxyribonucleic acid; PCR: polymerase chain reaction; ANA: antinuclear antibodies; ANCA: antineutrophil cytoplasmic antibodies; ds DNA: double-stranded deoxyribonucleic acid; ASMA: antismooth muscle antibody; IF: immunofluorescence; ELISA: enzyme-linked immunosorbent assay; IgG: immunoglobulin G; IgA: immunoglobulin A.

Discussion

We report a case of systemic sepsis following percutaneous liver biopsy in an LT recipient with choledochojejunal anastomosis. This was followed by severe rejection and deterioration of liver function to the extent that he has been listed for retransplantation.

Complications occurring as a result of liver biopsies in LT recipients are reported to range from 0.6% to 2.7%.^3,4 The risk does not seem to vary with living versus cadaveric LTs. The overall complication rates of liver biopsy between transplant and nontransplant patients are shown to be similar.^5,6 Infectious complications associated with liver biopsy are an unusual occurrence occurring in approximately 1 in 10,000 liver biopsies.⁷ Transient bacteremia following percutaneous liver biopsy has been reported with an estimated incidence of 2%–13%.^8,9 Most cases are asymptomatic, without evidence of fever, chills, or hypotension. The major risk factor for symptomatic bacteremia, leading to sepsis, is preexisting biliary obstruction.

Despite an extensive literature search, we could find only a few published cases of sepsis following percutaneous liver biopsy in LT recipients over the last two decades (Table 2). In 1991, Bubak et al¹⁰ in their series of 950 percutaneous liver biopsies, in 150 hepatic allograft recipients, reported an infectious complication rate of 0.6% (6 out of 950 biopsies). They found that five of the six septic complications occurred in patients with choledochojejunostomies. Based on these findings, they suggested using prophylactic antibiotics to reduce septic complications in this subgroup of patients with choledochojejunostomies. In another study, authors reported Pseudomonas bacteremia occurring more frequently in LT recipients with a Roux-en-Y biliary anastomosis than in patients with a choledochocholedochostomy.¹¹ The plausible hypothesis is that choledochojejunostomy and Roux-en-Y biliary anastomosis allows colonization of the hepatic allograft with bowel flora thereby making it susceptible to post-biopsy infection. Another study by Larson et al¹² on infectious complications following percutaneous liver biopsy also showed that the frequency of infection was higher (9.8%) in patients with choledochojejunostomy compared with those with choledochocholedochostomy (1.4%). They concluded that antibiotic prophylaxis before percutaneous liver biopsy was required in LT recipients with biliary tract abnormalities. In patients with a choledochojejunostomy and a Roux-en-Y biliary anastomosis, infections are related to the enteric flora and are usually polymicrobial. These findings suggest that the route of infection is the gastrointestinal tract rather than the more likely skin route in other patients undergoing liver biopsy.¹³

Table 2.

Review of Literature Describing Sepsis Following Liver Biopsy in Patients with Hepatic Allograft. (∗Incidence in Patients with Choledochocholedochostomy, # Incidence in Patients with Choledochojejunostomy).

Reference	Intervention	Total number of cases (biopsies)	Number of cases with sepsis post biopsy	Biliary Reconstruction		Interpretation/recommendation
				CC∗	CJ#
Cristina et al, 201414	PCLB/TJLB	286(667)	8	7.4%	2.3%	No increased risk in patients with CJ. Routine antibiotic prophylaxis not recommended.
de Diego et al, 199715	PCLB	196(666)	2	0%	1.5%	Antibiotic prophylaxis in patients with CJ may decrease the frequency of infectious complications.
Ben-Ari et al, 19962	PCLB	165(434)	5	1.5%	2.1%	Liver transplant recepients with CJ, undergoing liver biopsy do not require prophylactic antibiotics.
Galati et al, 199413	PCLB	92 (310)	1	1(0.8%	0%	Cannot recommend routine antibiotic in liver transplant recepients with CJ undergoing liver biopsy.
Bubak et al, 199110	PCLB	136(950)	6	1%	12.5%	Antibiotic prophylaxis at the time of liver biopsy may be appropriate in patients with CJ.

Abbreviations: CC: choledochocholedochostomy; CJ: choledochojejunostomy; PCLB: percutaneous liver biopsy; TJLB: transjuglar liver biopsy.

On the other hand, Galati et al¹⁴ in their retrospective series, compared 192 liver biopsies in patients with choledochojejunostomy with 118 liver biopsies in hepatic allograft recipients with choledochocholedochostomy but could not demonstrate a significant difference in infectious complications. They attempted to explain the conflicting findings by a difference in the length of the small bowel segment used for the choledochojejunostomy. Ben-Ari et al² also noted no significant increase in infectious complications following liver biopsy in patients with choledochojejunostomy. They concluded that prophylactic antibiotics are not necessary prior to liver biopsy in transplanted patients. The current literature on the use of prophylactic antibiotics is inconclusive, and we feel the need to generate further evidence for the same.

Transjuglar liver biopsy is a viable alternative in patients with an established contraindication to percutaneous biopsy like ascites or coagulopathy. Although considered safer than a percutaneous biopsy, it is however associated with its own unique set of complications like supraventricular tachycardia and access-related complications like neck hematoma.¹⁵ Direct evidence comparing the infective complication rates between percutaneous and transjuglar liver biopsy in the post-transplant setting is lacking.

The pathogenic relationship between liver biopsy and gram-negative septicemia appears well illustrated in the presented case. The possible risk factors in our case included the use of immunosuppressant and a bilioenteric anastomosis. Following the episode of bacteremic sepsis, the patient had deterioration of liver functions and required intravenous antibiotics and stoppage of immunosuppressive medications. The finding of severe acute cellular rejection with complete loss of bile ducts in the second liver biopsy 4 weeks following the first biopsy (which showed no rejection), remains unexplained. Possible explanation could be the suspension of immunosuppressants during the episode of sepsis or as a result of the septic process per se. Finally, it also appears that the patient has developed an early recurrence of his original disease (PSC) as seen in PTBD cholangiogram.

Conclusion

This case report emphasizes the potential risk of sepsis in LT recipients with bilioenteric anastomosis undergoing percutaneous liver biopsy. This increased risk may be mitigated by periprocedural broad spectrum antibiotic prophylaxis although evidence for this is limited.

Credit author statement

Lokesh Agrawal: Writing-Original manuscript preparation, Involved in patient care.

Sachin Jain: Involved in patient care.

Kumble S. Madhusudhan: Writing- Reviewing and Editing , Involved in patient care (Imaging, Liver biopsy).

Prasenjit Das: Histopathology review.

Dr. Shalimar: Writing- Reviewing and Editing , Involved in patient care.

Nihar Ranjan Dash,MS, MCh : Involved in patient care and transplant surgery.

Peush Sahni: Involved in patient care and transplant surgery.

Sujoy Pal: Writing- Reviewing and Editing , Involved in patient care and transplant surgery.

Conflicts of interest

All authors have none to declare.

Sources of financial or nonfinancial support

None to declare.