Pylephlebitis

INTRODUCTION — Pylephlebitis, or infective suppurative thrombosis of the portal vein, is a serious condition with significant morbidity and mortality, which can complicate intra-abdominal sepsis of any etiology. Although universally fatal in the pre-antibiotic era [1], the outcome of this infection has improved somewhat with modern diagnostic and therapeutic modalities. Curiously, however, reports of the diagnosis have increased in the last 15 years, possibly as a result of newer, more sensitive imaging techniques [2,3].

PATHOGENESIS — The portal vein is formed by the union of the superior mesenteric with the splenic veins. The portal system drains blood from the abdominal section of the gastrointestinal tract, with the exception of the lower part of the rectum.

Pylephlebitis begins with thrombophlebitis of small veins draining an area of infection. Extension of the thrombophlebitis into larger veins leads to septic thrombophlebitis of the portal vein, which can extend further to involve the mesenteric veins [4]. In a review of case reports, the superior mesenteric vein was involved in 42 percent [5]. Thrombus involved the splenic vein and intrahepatic branches of the portal vein in 12 and 39 percent of cases, respectively. Mesenteric vein involvement can lead to bowel ischemia, infarction and death.

An associated hypercoagulable state is found in some cases of pylephlebitis [6]. As an example, in one series of 44 cases, 18 were hypercoagulable due to clotting factor deficiencies, malignant conditions, or HIV/AIDS [7,8]. However, in another study that included 33 patients who underwent testing for a hypercoagulable condition in the setting of pylephlebitis, the rate of positive tests was low and results did not impact management [9].

EPIDEMIOLOGY — Pylephlebitis can complicate any intra-abdominal or pelvic infection that occurs in the region drained by the portal venous system, especially diverticulitis and appendicitis (figure 1) [2,7,10,11]. In a 1948 report, the most common inciting infection was appendicitis, accounting for all 21 cases [12]. However, the incidence of pylephlebitis following appendicitis significantly decreased with the introduction of antibiotics [1]. One of the most frequent foci is diverticulitis, which accounted for 6 of 18 cases in one series [2]. Although diverticulitis is a common cause of pylephlebitis, pylephlebitis remains a rare complication of diverticulitis [1]. In a 2023 systematic review of published cases between 1971 and 2022, diverticulitis (26.5 percent) and appendicitis (22 percent) were the two most common underlying causes of pylephlebitis [3].

Contiguous infection (eg, cholangitis or infected choledocholithiasis) can also lead to this complication [2,13]. In addition, pylephlebitis has been associated with inflammatory bowel disease [10,14-18], pancreatitis [5,19], hemorrhoidal banding [20], transrectal prostate biopsy [21], and sigmoid colonic polypectomy [22]; it has also occurred as a complication of the intragastric migration of a silicone gastric band [23] and following a computed tomography (CT)-guided liver biopsy [24]. A precipitating focus was identified in 13 of 18 cases (68 percent) in one report [2]. Pylephlebitis due to Bacteroides in a patient with Strongyloides infection has also been reported [25]. In Taiwan, primary liver abscess due to Klebsiella pneumoniae has emerged as a precipitating factor for pylephlebitis [26]. (See "Invasive liver abscess syndrome caused by Klebsiella pneumoniae".)

Pylephlebitis is more common in males and the age range of affected patients is wide (6 to 75 years), which may partially be influenced by the number of patients with appendicitis as an inciting etiology [3]. The mean age of patients with pylephlebitis is approximately 42 years, and in one review, 24 percent were aged <18 years [5].

MICROBIOLOGY — Bacteremia associated with pylephlebitis is frequently polymicrobial. In a 2023 systematic review of 220 cases, positive blood cultures were reported in 70 percent of cases; of these, 61 percent had a single pathogen and 39 percent had polymicrobial infection [3]. In another review of 19 cases, bacteremia occurred in 88 percent. The most common bloodstream isolates were Bacteroides fragilis, Escherichia coli, and streptococcus spp, but other organisms (eg, Aeromonas hydrophila, streptococci, Proteus mirabilis, K. pneumoniae, anaerobic streptococci, and Clostridium spp) were also isolated [2]. Fusobacterium, Parvimonas, and Actinomyces have also been reported [27-29]. In another series of 95 cases, bacteremia occurred in 44 percent, and the most common isolate was Streptococcus viridans [19]. Histoplasma pylephlebitis has also been described in a liver transplant recipient [30].

Yeasts, other streptococci, staphylococci, and Citrobacter spp have also been described, but enterococci are an uncommon cause of this entity [2,7].

CLINICAL MANIFESTATIONS — Abdominal pain, which is not necessarily severe, and fever are the most common presenting symptoms of pylephlebitis. In one report, all 44 patients had these two complaints [7]. By contrast, 100 percent had fever but only 14 (74 percent) complained of abdominal pain in another review of 19 cases [2], and in another study of 67 patients, 87 percent had abdominal pain and 70 percent had fever [9]. In a 2023 review of 220 cases, 76 percent had fever and 65 percent had abdominal pain [3]. Other symptoms have included rigors, nausea, vomiting, and headache [1,2,31].

Physical examination may elicit right upper quadrant or generalized abdominal tenderness, hepatomegaly, and jaundice. However, jaundice appears to be unusual in pylephlebitis unless there is associated cholangitis or liver abscess [2]. Splenomegaly has been reported in 10 to 25 percent of cases of pylephlebitis [5,32].

Patients may have minimal or no clinical features referable to the primary focus of infection, similar to patients with liver abscess [10] (see "Pyogenic liver abscess"). Although pylephlebitis usually presents as an acute illness, it may also present with a chronic or stuttering course over weeks [4,33].

Laboratory findings — Leukocytosis is a common finding [1,9,14,31,34], but both a normal leukocyte count [2,32,34] and neutropenia [10] have been described. Abnormal liver function tests, especially three- to fourfold increases in alkaline phosphatase [1,10,14,31] and 5- to 10-fold increases in gamma glutamyl transferase [11,14], are noted in most patients [1]. However, in keeping with the relative infrequency of clinical jaundice, two- to sixfold increases in serum bilirubin occur only in selected cases [1,10,14,31]. The inciting intra-abdominal sepsis or complicating hepatic abscesses may also contribute to hepatic dysfunction.

Complications — Pyogenic liver abscesses can complicate cases of pylephlebitis; in one series, they were reported in 37 percent of cases [1,5,35-37]. Bowel ischemia has also rarely been reported with one case of small bowel infarction necessitating bowel resection [5,6]. Portal hypertension may be a long term complication of pylephlebitis and has been described in a small number of cases [5,10]. Further sequelae of portal hypertension can include a dilated splenic vein and numerous venous collaterals in the hepatoduodenal ligament [32]. (See "Pyogenic liver abscess".)

DIAGNOSIS — The diagnosis of pylephlebitis requires the demonstration of portal vein thrombosis (pylethrombosis) usually accompanied by bacteremia in a febrile patient [34]. The diagnosis is frequently delayed because pylephlebitis is not considered since it is an uncommon condition, or because symptoms are nonspecific, clinical signs are typically lacking from the primary focus of infection, and there may be poor visualization of the portal vein. The diagnosis was most often made at laparotomy or postmortem in the past, but modern imaging techniques now facilitate an earlier diagnosis.

Blood cultures — Bacteremia has been documented in 23 to 88 percent of cases [2,7,9,19]. For this reason, blood cultures should be obtained in any febrile patient with abdominal pain, optimally before initiation of antimicrobial therapy.

Imaging studies — The demonstration of thrombus in the portal vein is an important component of the diagnosis of pylephlebitis but is not sufficient by itself. The diagnostic imaging modality of choice depends on the expertise at the institution where the studies are performed. Both computed tomography (CT) scanning and ultrasonography can demonstrate thrombus in the portal vein (image 1), with or without associated mesenteric vein thrombus [32]. In up to 18 percent of cases, gas can be visualized in the portal system [5]. We prefer the CT scan as the imaging study because of the additional ability to identify an underlying focus of infection elsewhere in the abdomen or pelvis.

Computed tomography — The CT scan is diagnostic when it demonstrates portal vein thrombosis in a patient with a clinical picture consistent with pylephlebitis (image 1) [1,32,38]. CT scanning has also been useful in defining a precipitating focus of infection.

Ultrasound — Ultrasonography can demonstrate echogenic material within the lumen of the portal vein and can also be used to evaluate progression or extension of clot and subsequent recanalization [10]. Color flow Doppler ultrasonography may improve diagnostic accuracy [1]. Intraoperative ultrasonography has also been used to confirm the diagnosis of pylephlebitis when this entity is suspected preoperatively [32].

Other studies — Magnetic resonance imaging (MRI) has been used to demonstrate portomesenteric vein thrombosis [9,19]. There are occasional but limited reports of fluorodeoxyglucose positron emission tomography (FDG-PET) being useful in the diagnosis of pylephlebitis in patients with fever of unknown origin [39].

TREATMENT — Antibiotics constitute the major treatment approach to pylephlebitis.

Antibiotics — The choice of empiric antibiotics in pylephlebitis depends on the probable source of infection and the likely organisms (see 'Microbiology' above). Infection is often polymicrobial, with both gram-negative aerobes and anaerobes, especially B. fragilis. Successful antibiotic regimens have included ampicillin [40], third generation cephalosporins [10], fluoroquinolones [10], metronidazole [10], gentamicin [2], clindamycin [2], and lincomycin (not available in the United States) [2].

Recommended regimens — Pylephlebitis is an uncommon infection, and thus, no randomized controlled studies have evaluated empiric antibiotic regimens for this disease. Treatment regimens are based on coverage of the most likely organisms to cause infection in this setting and small case series. We recommend that empiric treatment include broad-spectrum antibiotics until culture results are available. Patients should be treated at least initially with parenteral antibiotics.

Appropriate antibiotics include any of the following parenteral regimens:

●Combination therapy with metronidazole (500 mg every eight hours) PLUS one of the following, ceftriaxone (2 g daily) OR cefotaxime (2 g every six hours) OR ciprofloxacin (400 mg every 12 hours) OR levofloxacin (500 mg daily)

●Monotherapy with a beta-lactam/beta-lactamase inhibitor, such as, piperacillin-tazobactam (4.5 g every six hours) OR ampicillin-sulbactam (3 g every six hours)

●Monotherapy with a carbapenem, such as imipenem (500 mg every six hours) OR meropenem (1 g every eight hours) OR ertapenem (1 g daily)

The initial antibiotic choice should be modified if blood or surgical (from the underlying focus of infection) culture results and antimicrobial susceptibility testing are available. Susceptibility testing for anaerobes are not routinely performed but should be considered, especially in patients who are failing therapy.

Duration of therapy — The typical duration of antibiotic therapy is at least four to six weeks [2]. Parenteral antibiotics are given until there is significant clinical response, usually about two to three weeks. The remainder of the antibiotic course can be completed with oral agents, such as the combination of oral metronidazole and a fluoroquinolone.

Anticoagulation for select patients — There are no prospective randomized controlled studies and no consensus on the use of anticoagulation in pylephlebitis. We do not use anticoagulation for all patients with pylephlebitis. However, we suggest anticoagulation for patients who have progression of thrombosis on repeat imaging or continued fever or bacteremia despite antibiotic therapy. Thrombosis that extends beyond the portal vein into the mesenteric vein on initial evaluation and an underlying hypercoagulable state are additional potential indications for anticoagulation. Since Bacteroides spp appear to promote coagulation, possibly through production of enzymes that break down heparin, as well as bacterial surface components that promote fibrin clotting [4], we are more likely to anticoagulate patients when this organism is isolated.

When anticoagulation is given, the optimal approach is uncertain. We typically initiate anticoagulation with low-molecular-weight heparin (LMWH). If needed (eg, on discharge), LMWH can be switched to an oral anticoagulant There have also been reports of the use of direct oral anticoagulants (DOACs) including apixaban for initial anticoagulation [28]. Doses are similar to those used for other venous thromboembolic conditions and are discussed elsewhere. (See "Venous thromboembolism: Initiation of anticoagulation", section on 'Anticoagulant agents'.)

The duration of anticoagulation is also uncertain. It is reasonable to discontinue anticoagulation once the patient has improved clinically and imaging suggests that the thrombus is stable and not extending. Although this may occur within two to four weeks, patients often receive anticoagulation for at least three months [9]. Recanalization has been documented in some cases, but the length of time required for this to occur is unknown. One study suggested a mean time to thrombosis resolution of four and a half months [9].

The rationale for anticoagulation in acute pylephlebitis is to prevent sequelae of thrombus extension. Most of the evidence informing the impact of anticoagulation is from small observational studies and of low quality. In one report of 100 patients with pylephlebitis, among whom 35 received anticoagulation, anticoagulation was associated with a lower mortality rate (6 versus 22 percent) [5]. Similarly, in another retrospective study, none of the 12 patients who received anticoagulation died compared with 5 of the 32 who did not receive anticoagulation [7]. However, in one retrospective study of 67 patients, anticoagulation was associated with a higher rate of portal vein thrombus resolution (58 versus 21 percent) but was not independently associated with lower mortality after adjusting for severity of illness [9]. A case report of anticoagulation after extension of thrombus documented prompt resolution of the clot [10].

Thrombolytic therapy for thrombosis progression has been reported [10,41], although we do not routinely pursue this. There is minimal clinical experience with thrombolysis for pylephlebitis and no conclusive evidence of efficacy.

Percutaneous techniques — Catheter insertion into the portal vein with aspiration of thrombus and/or pus has been reported [42,43]. Intraportal infusion of antibiotics has also been described [42]. Further clinical experience is required before this technique can be routinely recommended.

Surgery — Surgery is not usually required for the management of pylephlebitis. However, surgical drainage of the precipitating focus may be necessary in some cases.

FOLLOW UP — We typically perform follow-up imaging (either computed tomography scanning or ultrasonography) of the portal vein approximately five to seven days following initiation of antibiotic therapy. Incomplete resolution of the thrombus is common and does not preclude clinical improvement [5]. However, repeat imaging can identify thrombus extension or other complications that may prompt changes in management. (See 'Anticoagulation for select patients' above.)

Screening endoscopy is not routinely performed because of pylephlebitis. However, it may be warranted if Fusobacterium bacteremia occurs in this setting, since this has been associated with occult gastrointestinal malignancy [27].

OUTCOME — Pylephlebitis was uniformly fatal in the 20 cases reported in 1948, secondary to appendicitis [12]. Even with newer available antibiotics, the condition continues to have an appreciable mortality rate, ranging from 11 to 32 percent [2,7,19].

SUMMARY AND RECOMMENDATIONS

●Definition – Pylephlebitis is infective suppurative thrombosis of the portal vein and is associated with significant morbidity and mortality; it can complicate intra-abdominal sepsis of any etiology. (See 'Introduction' above.)

●Pathogenesis – Pylephlebitis begins with thrombophlebitis of small veins draining an area of infection. Extension of the thrombophlebitis into larger veins leads to septic thrombophlebitis of the portal vein, which can extend further to involve the mesenteric veins. (See 'Pathogenesis' above.)

●Risk factors – The most common predisposing infections leading to pylephlebitis are diverticulitis and appendicitis. (See 'Epidemiology' above.)

●Microbiology – Bacteremia, which is frequently polymicrobial, occurred in 88 percent of cases in one series; the most common bloodstream isolates were Bacteroides fragilis and Escherichia coli. (See 'Microbiology' above.)

●Symptoms – Abdominal pain, which is not necessarily severe, and fever are the most common presenting symptoms of pylephlebitis. (See 'Clinical manifestations' above.)

●Diagnosis – The diagnosis of pylephlebitis requires the demonstration of portal vein thrombosis (pylethrombosis) usually accompanied by bacteremia in a febrile patient. We prefer the computed tomography (CT) scan as the imaging study because of the additional ability to identify an underlying focus of infection elsewhere in the abdomen or pelvis (image 1). (See 'Diagnosis' above.)

●Antibiotic therapy

•Empiric antibiotic selection – Treatment of pylephlebitis involves empiric antibiotic therapy; until culture results are available, we suggest empiric treatment with a parenteral regimen with broad gram-negative and anaerobic coverage (Grade 2C). Typical treatment regimens include one of the following (see 'Antibiotics' above):

-Combination therapy with metronidazole (500 mg every eight hours) PLUS one of the following, ceftriaxone (2 g daily) OR cefotaxime (2 g every six hours) OR ciprofloxacin (400 mg every 12 hours) OR levofloxacin (500 mg daily)

-Monotherapy with a beta-lactam/beta-lactamase inhibitor, such as, piperacillin-tazobactam (4.5 g every six hours) OR ampicillin-sulbactam (3 g every six hours)

-Monotherapy with a carbapenem, such as, imipenem (500 mg every six hours) OR meropenem (1 g every eight hours) OR ertapenem (1 g daily)

•Tailored antibiotic therapy – The initial antibiotic choice should be modified once culture results and antimicrobial susceptibility testing are available. (See 'Recommended regimens' above.)

•Duration of antibiotics – The typical duration of antibiotic therapy is at least four to six weeks. (See 'Duration of therapy' above.)

●Limited role of anticoagulation – We do not routinely use anticoagulation for pylephlebitis. However, for patients with progression of thrombosis on repeat imaging or persistent fever or bacteremia despite antibiotic therapy, we suggest anticoagulation (Grade 2C). Thrombosis extension into the mesenteric vein on initial evaluation, a hypercoagulable state, and isolation of Bacteroides spp are other potential indications for anticoagulation. We continue anticoagulation until the patient has clinically improved and imaging documents stability of the thrombus without extension. (See 'Anticoagulation for select patients' above.)