Abstract & Commentary

Endotipsitis — Difficult to Diagnose, Difficult to Treat

By Stan Deresinski, MD, FACP, Clinical Professor of Medicine, Stanford, Associate Chief of Infectious Diseases, Santa Clara Valley Medical Center, is Editor for Infectious Disease Alert.

Synopsis: Infection of transjugular intrahepatic portosystemic shunts presents difficult diagnostic and therapeutic challenges.

Sources: Mizrahi M, et al. Endotipsitis-persistent infection of transjugular intrahepatic portosystemic shunt: Pathogenesis, clinical features and management. Liver Int. 2010;30:175-183; Kochar N, et al. Tipsitis: Incidence and outcome at a single centre experience. Eur J Gastroenterol Hepatol. 2010;22:729-735.

I was asked earlier this week to see a patient with candidemia. The patient had severe cirrhosis and had previously had placement of a transjugular intrahepatic portosystemic shunt (TIPS) for control of complications of portal hypertension. While the source of the infection has not be determined, and he appears to be responding to therapy, the possibility of infection of his TIPS is currently under investigation. This patient, in fact, brought to mind one I saw several years ago with recurrent Enterococcus faecalis bacteremia who relapsed after each of three prolonged (up to 8 weeks) treatment courses with ampicillin plus gentamicinfor possible endocarditis. It was finally concluded that the patient had infection of his TIPS and he was placed on suppressive therapy with orally administered amoxicillin, which was successful in preventing recurrence of his infection for the rest of his life, which ended several years later and which resulted from hepatic failure. At the time, I was not aware that his infection had a clever but kitschy name — endotipsitis (or, alternatively, just tipsitis).

Mizrahi and colleagues in Jerusalem have reviewed the literature and identified 36 reported cases of tipsitis. While a definitive diagnosis can only be made by examination and culture of the device, this is not possible except at post-mortem examination or at the time of liver transplantation, since the device cannot be removed except together with the liver. A clinical diagnosis of definite infection can, however, be accepted in the presence of continuous bacteremia together with the visualization of vegetations or intraluminal thrombus in the absence of another source of the infection. TIPS infection can be considered probable with continuous bacteremia in the absence of another source, even if imaging of the device fails to demonstrate vegetation or thrombus. In addition to fever, most patients present with increasing jaundice.

In the series derived from the literature, the most frequently identified pathogens were members of the Enterobacteriaceae, Enterococci, Staphylococci, and Streptococci. Three Candida infections, including two due to C. glabrata and one due to C. albicans, were identified.

Mizrahi et al suggest that the initial evaluation should include color Doppler ultrasound examination to determine the patency of the device. Echocardiography should be performed to search for evidence of endocarditis. In some cases, endoscopy (upper and lower) and retrograde cholangiopancreatography should be performed to search for evidence of a TIPS-biliary tract fistula. Other studies should be performed as indicated.

Kocahr and colleagues in Edinburgh, in a single-center study, report that probable endotipsitis occurred in 8 of 785 (1%) of patients > 14 years of age in whom a TIPS had been placed. Symptoms of the infection began at a mean of 24.7 months after device placement. None had identifiable device vegetations, but the TIPS was occluded in four. All the infections were bacterial. Patients received antibiotic therapy for a median of 3 months (range, 10 days to 3 months). Three patients, two of whom had received antibiotics for only two weeks, had recurrence of their infection. The TIPS infection contributed to the deaths of four patients.


While apparently quite uncommon, infection of a TIPS can be difficult to manage, as are all foreign body infections. In this case, however, similar to the circumstance with infections of ventricular-assist devices, the device cannot be removed in the absence of organ transplantation. Performing organ transplantation in the face of an ongoing infection presents its own challenges, however. Thus, the optimal approach, once the diagnosis is made, is to administer the most effective bactericidal antibiotics available for a prolonged period of time and, in the cases in which relapses occur, to consider the use of suppressive antibiotic therapy for the rest of the patient's life.