Endoscopy 2009; 41: E56-E57
DOI: 10.1055/s-2008-1077564
Unusual cases and technical notes

© Georg Thieme Verlag KG Stuttgart · New York

Recurrent bacteremia due to retained embolized glue following variceal obliteration

G.  Wright1 , W.  R.  Matull1 , L.  Zambreanu1 , S.  O’Neill1 , R.  Smith2 , J.  O’Beirne1 , M.  Y.  Morgan1 , 2 , 3
  • 1The Royal Free Sheila Sherlock Liver Unit, The Royal Free Hospital, Pond Street, London, UK
  • 2Department of Medical Microbiology, Royal Free Hospital, Pond Street, London, UK
  • 3The Centre for Hepatology, Royal Free Campus, Royal Free & University College Medical School, Rowland Hill Street, Hampstead, London, UK
Further Information

Publication History

Publication Date:
24 March 2009 (online)

Cyanoacrylate glue injection is the endoscopic treatment of choice for gastric variceal hemorrhage [1]. Though generally safe [2], glue is rarely associated with systemic embolization and significant bacteremia [3]. Prolonged bacteremia (≥ 4 weeks) has only been reported in two patients [4], in whom retained variceal casts acted as infected foci. Here we report a patient, 24 weeks after variceal injection, with recurrent life-threatening septicemia due to retained “embolized” glue.

A 38-year-old man presented with decompensated alcoholic liver disease complicated by spontaneous bacterial peritonitis (SBP). Intravenous tazocin (piperacillin/tazobactam) treatment was commenced after ascitic cultures isolated extended spectrum β-lactamase (ESBL)-producing Escherichia coli sensitive to tazocin and ertapenem. Between days 3 and 14, recurrent gastric variceal hemorrhage required repeated endoscopic glue injection to achieve hemostasis. On day 14, renal function suddenly deteriorated (despite terlipressin and albumin); CT imaging revealed glue in the gastric varix and inferior vena cava extending to the left renal vein ([Fig. 1]). Renal function improved with supportive measures. From week 3, recurrent swinging pyrexia proved unresponsive to empiric treatment with cefotaxime and later tazocin. Investigation, which included full viral and bacteriological screens, transthoracic echocardiogram, dental radiographs, gallium scan, and gastrointestinal endoscopies, failed to identify an infective source until, at weeks 7, 14, and 22, following antibiotic-free periods, ESBL-producing Escherichia coli sensitive to ertapenem was isolated from blood, and repeat CT at week 24 showed persistence of the glue observed at week 3 ([Fig. 2]). Only 6 weeks of intravenously administered ertapenem, following on from shorter courses, successfully treated the pyrexia, with negative culture and sustained improvement in hepatic function.

Fig. 1 a – d Abdominal/pelvic CT scan showing multifocal radiopaque densities in the gastric fundal varix and also inferior vena cava (IVC) extending to the left renal vein (LRV) consistent with retained systemically embolized glue, within days of repeated endoscopic injections of glue/lipiodol for gastric variceal hemorrhage.

Fig. 2 a, b Abdominal CT scan demonstrating the persistence of the multifocal radiopaque densities in the inferior vena cava (IVC) extending to the left renal vein (LRV) at 24 weeks after glue injection therapy.

The exclusion of other infective foci suggests that the embolized glue acted as a nidus for bacterial seeding on or within the adhesive casts during the original procedure. Breakdown of the casts over time could have resulted in the intermittent release of colonized microbes into the systemic circulation, which would have been only partially responsive to short courses of intravenous antibiotics. In conclusion, embolized glue may persist beyond 24 weeks and should be considered if recurrent infection develops which may necessitate protracted courses of antibiotics. This reiterates the need for caution with glue, and the benefit of endosonographic or fluoroscopic monitoring.

Endoscopy_UCTN_Code_CPL_1AH_2AC

References

M. Y. MorganMB ChB, FRCP 

The Centre for Hepatology
Royal Free Campus

Royal Free & University College Medical School
Rowland Hill Street
Hampstead
London NW3 2PF
UK

Fax: +44-207-4332870

Email: mymorgan@medsch.ucl.ac.uk

    >