Pylephlebitis in patient with covered perforated rectosigmoid cancer

Summary

Acute portal vein thrombosis and septic thrombophlebitis of the portal vein represent serious, although rare cases in the non-cirrhotic population. The authors present a case report, in which nonspecific clinical and CT scan findings led to the difficult diagnosis of pylephlebitis due to perforated rectosigmoid cancer.

Key words:
portal vein thrombosis – pylephlebitis – colorectal cancer – cirrhosis

INTRODUCTION

Extrahepatic portal vein obstruction (EHPVO) in non-cirrhotic patients is seen in 5–10 % of patients and as such is the second most common cause of portal hypertension [1]. Its incidence is about 1% [2] with an annual mortality of approximately 13%. Given the current treatment possibilities, the cause of death is usually the primary disease, not complications of portal hypertension. Despite improvements in diagnostic methods, the cause of EHPVO is only uncovered in a portion of patients [3,4]. Pylephlebitis is defined as ascending septic infection of the portal venous system.

The following work presents a case report of our patient with an initially undetermined cause of pylephlebitis.

CASE REPORT

A 70-year-old normostenic woman was admitted to the Department of Surgery at the 2. Medical Faculty of Charles University and Central Military Hospital in Prague. At the time, she had been treated by corticosteroids for 20 years for rheumatoid arthritis. She used proton pump inhibitors preventively. Otherwise the patient was without any serious comorbidities. At the emergency room, the patient presented with a four-day history of progressive epigastric pain, subfebrile temperatures, vomiting and diarrhea. During the initial examination, the patient was normotensive, with tachycardia at a rate of 111 pulses/min., afebrile and anicteric. Physical examination revealed dominating abdominal findings. The abdomen was rigid upon palpation with peritoneal irritation in the right hypochondrium. Peristalsis could be heard in all quadrants, with no palpable mass, digital rectal examination did not find any pathology. Laboratory results showed the following pathology: leukocytosis 14x10⁹/l, Hb 103 g/l, significant thrombocytopenia 46x10⁹/l, creatinine 164 µmol/L, total bilirubin 57.6 µmol/l, ALT 0.76 µkat/l, AST 1.01 µkat/l, GMT 1.24 µkat/l, ALP 2.55 µkat/l, CRP 240.5 mg/l. Ultrasound examination described suspicion of perforation of an inflamed gallbladder with free fluid along the abdominal wall and pneumobilia. Due to diagnostic uncertainty, a CT scan was performed (Fig. 1−3), which described air in the portal venous system and in the venous system of the sigmoid colon, hilar occlusion of the portal venous system, infiltration of the hepatic hilum and suspicion of prior covered perforation of the gallbladder. Due to progression of the clinical findings and signs of peritoneal irritation, the patient was indicated for urgent surgical revision. Because of the suspicion of perforated cholecystitis, a right subcostal incision was selected. An ulcerophlegmonous gallbladder was discovered. The small intestine and accessible part of the bowel were without pathological findings. An uncomplicated choleystectomy was performed. Preoperatively, antibiotic therapy with piperacilin/tazobactam was administered in a therapeutic dose. Upon revision of the preoperative CT scan, the radiologist described prior complicated diverticulitis in the area of the sigmoid colon (Fig. 4). In the early postoperative course, the patient showed an overall worsening with signs of septic shock. Follow-up CT on the 2. postoperative day revealed gas in the portal venous system bilaterally with abnormal hepatic perfusion. The thrombus extended to the lienal vein (LV) and superior mesenteric vein (SMV). Due to suspicion of ischemic colitis, the patient underwent surgical revision. Peroperatively, the small bowel loops were dilated without a transitional zone as seen in paralytic ileus, the liver parenchyme was without visible ischemic zones. The entire colon was vital. A tumor was discovered in the rectosigmoid region along with several inflamed diverticuli and probable covered perforation (Fig. 5). We performed a left-sided hemicolectomy with terminal transversostomy. The subsequent hospitalization course was complicated due to the overall condition of the patient. There was secondary wound healing with high volume leakage of ascites due to hypoalbuminemia. We continued with realimentation. The patient received therapeutic doses of low molecular weight heparin (LMWH), which was increased to 0.8ml s.c. twice daily due to low anti Xa levels. After four weeks, follow-up Doppler ultrasonography revealed flow in the lienal vein. A week later, CT findings showed a forming abscess in the 6th liver segment, the portal vein (PV) was not patent. We newly administered amoxicillin with clavulanate for a period of three weeks. The follow-up laboratory results were as follows: leukocytes 16x10⁹/l, thrombocytes 451x10⁹/l, total bilirubin 108 µmol/l, ALT 0.35 µkat/l, AST 0.36 µkat/l, GMT 3.99 µkat/l, CRP 144 mg/l. The patient was without fever, passage through the stoma was fully renewed. The patient was then transferred to another facility for subsequent care. From this facility, the patient was released home after four weeks. Histological findings from the resected colon described adenocarcinoma GII pT3pN1(2/5)M0. Due to the patient’s poor performance status (ECOG 2-3, KI 40−50%) adjuvant chemotherapy was not considered at the time. The patient remains dispensarized at our workplace. Application of LMWH was discontinued, oral anticoagulation was not initiated. A follow-up CT is planned in three months.

Discussion

Etiological factors of EHPVO in patients without liver cirrhosis may be local as well as systemic, acquired or hereditary. Thrombophilic conditions are at the forefront. Myeloproliferative diseases are discovered in 20–48% of patients with EHPVO, predominantly with the newly described JAK 2-V617F mutation [2,4]. Other risk factors include malignancies, especially of the hepato-pancreato-biliary region, intraabdominal infections, vasculitis such as Behҫet’s disease, Budd-Chiari syndrome, inflammatory bowel diseases, sepsis, postoperative conditions (after liver transplantation, splenectomy, etc.), umbilical sepsis in children. Age is an independent risk factor.

EHPVO may be classified anatomically based on the extent of the thrombus into four categories: I – thrombus is isolated to the portal vein posterior to the confluence of the lienal vein and SMV, II– thrombus extends to the SMV and its patent branches, III – diffuse thrombosis of the splanchnic venous system with high-capacity collaterals, IV – extensive thrombosis with thread-like collaterals [4]. Anatomical classification has primarily a prognostic significance. The second classification is based on timing and is divided into acute and chronic EHPVO. The timing criterium defines the limit as 60 days from the creation of the thrombus to mesenterial symptomatology. In simplified terms, chronic EHPVO begins to manifest once complications of portal hypertension arise or remains clinically mute. The only abnormal findings may be pancytopenia. Acute thrombosis is often accompanied by non-specific symptoms, such as abdominal pain, fever, nausea, and diarrhea. Splenomegaly is often present, ascites is not common. Laboratory results are often unindicative. Liver functions may be altered. Thrombocytopenia has been described. If the thrombus extends to the SMV, EHPVO may be accompanied by bowel ischemia. Unlike in cirrhotics, however, the incidence is up to 5% [2,3,5]. Liver failure rarely occurs due to the so-called double rescue phenomenon due to the double blood supply of the liver with early collateral formation and cavernous reconstruction of the portal vein. Five-year survival of patients without cirrhosis and carcinoma with EHPVO is approximately 90%. The mortality is significantly higher, up to 20−50%, in the case of patients with a thrombus extending to the SMV and with infarction of the bowel [2]. Ascending septic infection of the portal venous system is referred to as pylephlebitis. The most common cause is intraabdominal infection of organs draining into the portal venous system. Primarily diverticulitis in 68% [3,5], less frequently appendicitis [3,5,6,7,8,9]. As for the infectious agent, a polymicrobial flora is described with Bacteroides fragilis.at the forefront. It is assumed, that by creating anticardiolipin antibodies, it is directly involved in causing EHPVO [10,11]. Hemoculture is positive in 80−88% of cases [3,5,7,8,12]. The endovascular character of the disease predisposes to the formation of septic metastases, including liver abscesses in approx. 53% [6,10,13,14]. The symptoms are similar to those seen in acute EHPVO, possibly with signs of sepsis. The disease course may mimic gastroenteritis [3,15]. The mortality remains high at around 32% [3,5], usually due to fulminant sepsis even before beginning antibiotic treatment.

Since 1979, when inflammation of a thrombus of the portal vein was first diagnosed using duplex ultrasonography, diagnostic imaging methods have advanced significantly. It seems that there has been an increase in the incidence of EHPVO; in reality, however, only the number of diagnosed cases has increased. Nonetheless, EHPVO remains an uncommon complication in patients without liver cirrhosis. Due to the shortage of randomized studies, views on treatment are inconsistent. The aim of EHVPO therapy has two directions. One is to prevent expansion of the thrombus to the mesenteric venous system and possibly succeed in its dissolution. Second is to treat complications stemming from portal hypertension. Treatment of complications of portal hypertension may be extrapolated from knowledge of treating cirrhotic patients. The course is milder in patients without cirrhosis. There is no unanimous opinion regarding anticoagulation therapy. Individual approaches are selected on a strictly individual basis. Spontaneous thrombus dissolution is possible and is mainly described in acute pancreatitis. The frequency is unknown however. In terms of anticoagulation therapy, chronic and acute EHPVO are viewed as individual entities. In order to predict the effect and duration of treatment, the cause of EHPVO must be known. If the local cause is unknown, we search for possible systemic prothrombophilic conditions. In acute EHPVO, it is presumed that anticoagulation may lead to thrombus dissolution in up to 80 % [4] and decidedly prevents its expansion or recurrence. The treatment of choice is LMWH. Their role is to prevent thrombus growth and prevent formation of new thrombi. The patient’s own fibrinolytic system is usually competent in dissolving the thrombus on its own. The effectiveness of therapy can be monitored by measuring the serum level of factor anti-Xa. The duration of anticoagulation therapy depends on the etiology of the thrombosis. It ranges from 3 to 6 months. It may last even longer if the thrombus has expanded to the SMV or if there is a systemic prothrombophilic condition present. In chronic EHPVO, the question of anticoagulation therapy administration is even more difficult. It is considered to decrease the risk of thrombus recurrence, without increasing the risk of hemorrhagic complications [2,4,16,17]. In cases where the thrombus is isolated to the PV and without a known clotting deficit, some authors recommend not administering anticoagulation drugs at all [17], and to carefully consider their administration in cases of hypersplenism.

Intervention treatment methods are rarely mentioned for treating acute thrombosis of the PV or SMV, usually only within case reports or on small sets of patients. Local thrombolytic therapy by percutaneous puncture of the PV, or intravascularly via the jugular vein, femoral vein or transhepatically, or even catheterization of the superior mesenteric artery are described. The risk of hemorrhage is high, in more than 50% of cases [18]. Another option is mechanical aspiration thrombectomy. The results are good if performed within 30 days from thrombus formation. Balloon dilation or stent implantation may be useful in the prevention of thrombus recurrence [19]. However such therapy is only possible in specialized centers. More extensive studies are lacking and conservative treatment is preferred.

The role of bypass operations in EHPVO is minor. Indications for surgical thrombectomy are sparse.

In terms of differential diagnosis, acute portal vein thrombosis or pylephlebitis is rarely considered. Often the correct diagnosis is made only after the result of CT examination. This was also true in the case of our patient, who upon admission complained of a number of nonspecific symptoms. During the first surgical revision, cholecystectomy of an ulcero-phlegmonic gallbladder was performed. It was not until subsequent revision of the CT findings and worsening condition of the patient, and due to suspicion of ischemic colitis, a second surgical revision was performed and a covered perforated rectosigmoid tumor with diverticulitis was discovered to be the source of the pylephlebitis. In the past, when gas was described in the portal venous system, it was considered to be due to mesenterial ischemia with extensive bowel necrosis with an expected fatal course. It was an indication for immediate surgical revision. Gas in the portal venous network, however, may have a number of causes and may not always necessitate surgical revision [5]. In our case, the cause was a perforated colorectal tumor in a terrain of diverticulitis. We therefore selected surgical resection. In a number of cases, antibiotic therapy supplemented by anticoagulation therapy is adequate [3,5]. Our patient received a therapeutic dose of low molecular weight heparin for two months at our workplace and subsequently for one month at a subsequent care facility. No additional administration of LMWH was indicated. The last CT examination did not show recanalization of the portal vein, flow in the SMV and lienal vein was detected. Newly created collaterals were observed (Fig. 6). The recommended anticoagulation therapy in acute EHPVO can identically be applied to the treatment of pylephlebitis. Anticoagulation and antibiotic therapy significantly decrease mortality [2,3,4,7]. Broad-spectrum antibiotics which are effective against anaerobes and Streptococcus species are selected, with subsequent correction based on cultivation results. Broad-spectrum antibiotics were administered to our patient prior to the first surgery. Twelve days after surgical treatment of the infection focus and improvement of the patient, antibiotic therapy was terminated and initiated again in the third postoperative week due to the formation of a liver abscess. We continued with conservative therapy for three weeks, without the need for surgical intervention or drainage. The literature states that antibiotic therapy should be extended to approximately 4 weeks due to the high risk of infection dissemination. In case of liver abscess formation, antibiotic therapy should continue for six weeks [3]. If the situation requires, the focus of infection is treated surgically. The literature reveals that it is not appropriate to surgically treat an infected blood vessel, and procedures such as surgical thrombectomies are accompanied by a high mortality [4,6].

CONCLUSION

Acute extrahepatic portal vein thrombosis and pylephlebitis remain uncommon complications of intraabdominal infections, especially in patients without liver cirrhosis. The varying clinical findings make diagnosis very difficult. Imaging methods play an integral role in the diagnosis- especially dynamic contrast CT and Doppler ultrasound. Prognosis is dependent on the primary disease, comorbidities, and possible bowel ischemia, which increases lethality to 50 %. Fortunately this complication is not common. In a number of cases, it is not necessary to indicate surgical treatment. The treatment of choice is early and long term administration of broad-spectrum antibiotic therapy and anticoagulation therapy. The duration of anticoagulation therapy is given by the etiology of the portal vein thrombosis. To date, more invasive methods leading to dissolution of the thrombus (thrombolysis, transhepatic angioplasty) are only described in individual cases. Larger studies are still lacking and there is no standard for therapy of EHPVO. Therapy of chronic EHPVO is the same as therapy of portal hypertension in patients with cirrhosis. However, complications are less frequent and patients have a significantly better prognosis. Bypass operations play a minor role in the therapy of EHPVO.

MUDr. Katerina Menclova, MD.

Surgery department, 2. Faculty of Medicine,

Charles University and Central Military Hospital, Prague

U Vojenské Nemocnice 1200

160 00 Prague 6

e-mail: katerina.menclova@uvn.cz