PENETRATING DUODENAL ULCER AS A CAUSE OF INTRAHEPATIC ABSCESS CASE REPORT

Jeffrey C. Allard, MD

Ewa Kuligowska, MD

Boston University Dept. Radiology

One of the most alarming complications of peptic ulcer disease is perforation. 80% of all peptic ulcers occur in the duodenal bulb and when perforations occur, they are usually free into the peritoneal cavity. In some chronic cases duodenal ulcers will become adherent to adjacent organs and penetrate into these organs or ligaments and mesenteries related to them, thereby becoming "walled off" perforations. Most peptic penetrations involve the pancreas or biliary system, but penetration of the liver parenchyma has rarely been reported. We describe a case of duodenal ulceration with fistulous tract to an intrahepatic abscess as a complication.

CASE REPORT

A 57 year old alcoholic was admitted with sudden onset of epigastric pain and near syncope. He had several days of fever, chills, and anorexia prior to admission. Examination was remarkable for postural vital signs, mild right upper quadrant tenderness, a right inguinal hernia, guaic positive stools, and leukocytosis. Working diagnoses were incarcerated hernia and possible myocardial infarction.

On the second hospital day a barium enema was reported as normal. The patient was noted to have intermittent fevers to 103 F. Subsequently, an upper gastrointestinal study was reported as showing an ulcer of the superior portion of the bulb. Liver function tests worsened and three of twenty blood cultures taken returned with variable gram positive bacteria.

On the thirteenth day computed tomography was reported as normal. The following day, gentamycin and clindamycin were started for probable intra-abdominal sepsis, possible strangulated hernia. Ultrasound of the next day showed a localized echogenic area within the liver with reverberation suspicious for a gas containing abscess. Plain films and repeat UGI floroscopy showed a large air/ fluid level just superior to the previously documented ulcer. Percutaneous drainage was performed and contrast injected showed fistulous communication between the abscess and ulcer. Cultures grew enterococcus and after 5 days the patient was doing well clinically, and therefore the drainage catheter was removed...

DISCUSSION

Posterior to the stomach is the lesser sac and pancreas. From the pylorus, the duodenum passes superiorly, posteriorly, and to the right, until it turns sharply and becomes retroperitoneal, adjacent to the head of the pancreas. The duodenal bulb is just inferior to the quadrate lobe of the liver, to the left of the gall bladder, and anterior to the common bile duct. The free margin of the duodenohepatic ligament contains the portal vein, common bile duct, and hepatic artery, defining the anterior margin of the epiploic foramen, which connects Morrison's pouch with the lesser sac.

Understanding these anatomic relations explains the potential consequences of perforation. Anterior wall bulb ulcers usually perforate free into the peritoneal cavity. Posterior gastric ulcers can perforate into the lesser sac or penetrate the pancreas, and in fact 50% of peptic penetrations are related to the pancreas. Perforations can be limited by the various peritoneal reflections or omentums. Walled off collections or abscesses are most commonly seen in Morrison's pouch, the lesser sac, or in a subphrenic location. Evaluation of these peptic penetrations or walled off abscesses is best accomplished with CT, because of its more reliable demonstration of the pancreatic head, enlargement of which is often the only sign of penetration. Perihepatic and subphrenic collections are also apparent with ultrasound.

Cholecystoduodenal fistulas are usually due to gallstone perforation, but peptic ulceration accounts for 10- 20%. Although less common, choledochoduodenal fistulas are due to peptic perforation in 90% of cases. Plain films show air in the biliary system when this occurs, and barium studies will readily show the site of connection. When these fistulas result from peptic disease, a large percentage will close off with medical therapy, whereas those due to gallstones usually do not close.

Less common sites of peptic penetration include the kidney, colon, and liver. We have found only four cases in the literature of reported liver penetration by peptic ulceration. In none of these cases was the penetration complicated by intrahepatic abscess. In fact, septic complication of perforated ulceration is reported in only about 1% of cases, although prophylactic antibiotics are used in these "clean-contaminated" situations.

Intrahepatic abscess is usually pyogenic or amebic, and results from gastrointestinal or generalized sepsis or biliary disease. In about 10% of cases direct extension related to tumor or trauma is causal. Ultrasound, CT, and scintigraphy have all ben useful for the detection of these abscesses. 20 -30% of intrahepatic abscesses have gas collections, related to gas producing organisms. The interventional radiologist has had good results with percutaneous drainage, thus avoiding the need for surgery. Rupture of intrahepatic abscesses most commonly occurs with amebic abscesses. Most common sites of rupture include intrathoracic, peritoneal, or pericardial; with increasing mortality as a result. Of 2074 amebic ruptures, only one was into the duodenum and this case also involved the kidney at autopsy. The AFIP has collected 8 cases of intrahepatic abscesses rupturing into the gastrointestinal tract: 5 to stomach, 2 to duodenum, and 1 to colon.

That our case represents primary peptic disease is supported by the fact that initial UGI series demonstrated the superior bulb ulceration at a time when the intrahepatic gas collection was very small, suggesting that the abscess at this time would have been too small to rupture. That a tumor of the bulb (rare to begin with) was responsible for the penetration would be unlikely in the face of interval healing of the ulcer demonstrated by UGI. Our case points out the necessity for close scrutiny of the RUQ for abnormal gaseous collections. As in our case, CT has failed to identify intra-abdominal abscesses in 5 of 15 cases when there was fistulous connection to the bowel. Authors of this study suggest that correlation with planar film is necessary to identify that an air/fluid or air/contrast level is extraluminal. Ultrasound is demonstrated to be a superior imaging modality for pathology in the RUQ. We believe that percutaneous drainage, antibiotic, and antiulcer therapy can achieve good outcome