Case Histories

A 67-year-old man was treated at another hospital for colostomy dysfunction and a peristomal hernia. In 2002, he was treated for a rectal cancer with radiotherapy, proctocolectomy and ileoanal anastamosis with protecting ileostomy. In 2003, he had ileostomy closure. In 2004, he had a loop sigmoid colostomy for recurrent disease with distal obstruction. In January 2006 he had colostomy relocation and repair of peristomal hernia. Shortly after discharge, he was admitted to our hospital with abdominal pain. On exam, he was afebrile and did not appear to be toxic. His abdomen was not tender. Lab values showed a leukocytosis with WBC elevated to 23,000/mm ³ . Other lab abnormalities included total bilirubin of 2.9 mg/dl. Alkaline phosphatase was 276. A CT scan was done which showed pneumopylephlebitis and air traced all the way into the inferior mesenteric vein. (Figures 1A, 1B, 1C) Since air in the portal venous system is most commonly seen in patients with ischemic bowel, the patient was explored. The surgeon felt that operative findings were basically normal considering the patients recent operation and nothing further was done. The patient improved with antibiotic therapy.

Figure 1

Figure 1A: Air in the liver was thought to be in branches of the portal vein

Figure 2

Figure 1B: Air in portal vein more centrally

Figure 3

Figure 1C: Air seen in Inferior Mesenteric Vein

A 56-year-old man was seen for a ten-day history of chills and fever. He did not have abdominal pain. His temperature was 39.8 degrees C. His abdominal exam was normal and stool Hemoccult was negative. His WBC was 15,700/mm ³ . Blood cultures were drawn which grew gram-negative rods. Serum bilirubin was 4.8 mg/dl. Other liver function tests were normal. Abdominal and pelvic CT scan showed air in the liver, in the portal vein and in the inferior mesenteric vein. (Figure 2A and 2B) In addition, portions of the inferior mesenteric vein were filled with clot or debris (Figure 2C). Diverticula were seen but there was no radiographic evidence of diverticulitis or abscess. The next day, portal venous ultrasound showed either clot or debris in both the right and left portal veins.

The patient was explored and a sigmoid colectomy and end colostomy with Hartmann procedure was done. When the inferior mesenteric vein was cut across, it appeared to contain fecal material. After his operation, the patient was anticoagulated because of concern that there was clot in the portal venous system. He also responded to antibiotics and was discharged from the hospital in ten days.

Figure 4

Figure 2A: Small gas shadow in liver

Figure 5

Figure 2B: Gas in inferior mesenteric vein

Figure 6

Figure 2C: Debris filling inferior mesenteric vein

Discussion

The presence of gas within the inferior mesenteric vein is usually found in patients with diverticular disease of the colon and indicates either a direct fistula with intestinal gas entering the vein through the fistula, the presence of gas producing organisms within the vein, or both of these conditions simultaneously. Jensen (₁) described two patients with pneumopylephlebitis and described inflammatory findings associated with the inferior mesenteric vein in two patients who were felt to have intramesocolic diverticular perforation. They suggested that a water-soluble contrast study could help identify and localize the site of communication between colon and portal venous system. They also emphasized that physical findings of active diverticulitis do not need to be present in these patients. Our first patient emphasizes this point further by showing that even intra-operative findings of septic phlebitis may be subtle unless a careful examination of the colon or the tissues surrounding the inferior mesenteric vein is done.

Table 1 summarizes cases of inferior mesenteric vein septic phlebitis that we have identified during the past 35 years. Thirteen patients are reviewed in addition to our two patients. Eleven of fourteen patients were men with ages ranging from 35 to 72. Abdominal pain was reported in only ten of the fourteen patients; in those ten patients, more than half reported only vague or mild pain. Nausea, vomiting and change of bowel habits were not reported. Blood in the stool was described in only one patient.

In the thirteen patients where abdominal exam is described, four had no tenderness, five had mild tenderness and only four had significant tenderness. Eleven patients were febrile. Leukocytosis was described in five of eight patients. Mild to moderate elevation of serum total bilirubin was present in half of the patients.

An abdominal CT scan was done in nine of the fifteen patients. Most patients had evidence of inflammatory process tracking along the course of the inferior mesenteric vein with streaking or stranding of the adjacent soft tissues. Six of the eight patients had gas within the inferior mesenteric vein and the eighth had barium from a recent barium enema within the IMV. Seven of nine patients also had radiographic findings of diverticulosis or diverticulitis with inflammation next to the sigmoid colon and mesentery. Thrombus or debris was seen in the inferior mesenteric vein in five patients. (₈,₉,₁₁,₁₂) In one patient, the only reported CT abnormality was gas seen near the head of the pancreas. (₁),

Six of the fifteen patients did not have a CT scan. Abdominal films showed portal venous or hepatic gas in three patients. (₁,₂,₆) One of those patients also had gas in the inferior mesenteric vein. (₂) Three patients had contrast enemas, two with barium and one with gastrograffin. (₃,₄,₅) All three studies had contrast material in the inferior mesenteric vein.

In addition to antibiotic therapy, fourteen of the fifteen patients had surgical treatment. In most of the patients, some degree of thickening of the sigmoid colon and its mesentery was found. The inflammatory process usually tracked along the course of the inferior mesenteric vein. Eight of the patients had thrombus within the inferior mesenteric vein either seen on CT scan or grossly seen in the operating room. Ten patients had sigmoid resection with end colostomy and Hartmann pouch creation. One patient also had drainage of a mesocolic abscess.

Two patients had a “negative” exploratory laparotomy and no resection done. One of these two patients recovered with antibiotics alone.(McClenathan) The other required a second operation with sigmoid resection and colostomy three weeks later to control sepsis.(₂) One patient had a loop colostomy of the transverse colon which failed to control sepsis, so a second operation to resect sigmoid colon and convert to an end colostomy was done.(₃) One operation was not described. The final patient was treated with antibiotics alone and eventually recovered without an operation.(₁₁) In fact, all but one of the patients reported in this review recovered from their disease and surgery.

The issue of anticoagulation therapy for treating patients with septic thrombophlebitis of the inferior mesenteric vein is not resolved by this review. Filling defects seen in the inferior mesenteric vein in these patients have usually been attributed to thrombus. Yet in one of our patients, the opened gross specimen revealed fecal material in the mesenteric vein that appeared to be identical to material found in the colonic lumen. In this series, only two of the patients were given anticoagulant therapy, and they recovered as did the other patients who did not get anticoagulation. (₁₁, McClenathan)

Conclusion

Septic phlebitis of the inferior mesenteric vein is an unusual condition which is usually a complication of diverticular disease of the colon. History, physical examination and even operative findings may deceptively underestimate the presence and severity of this septic process. Abdominal films and CT scanning are helpful identifying the presence of this condition. Most authors feel that septic phlebitis of the IMV warrants a surgical approach with resection of the abnormal colon and colostomy. Bowel preparation, resection and primary anastamosis was not done in any of these patients reviewed. Anticoagulation may not be necessary for several reasons. First, portal venous filling defects could represent fecal material in some patients. Second, patients in this series who were not treated with anticoagulants appear to have done as well as the few patients where anticoagulation was used.

Figure 7

Table 1