Conclusions: Gastric varices are a noteworthy issue in patients with portal hypertension, primarily due to the associated risk of bleeding, which occurred in 20% of our patients. Although not significantly linked to a fatal outcome, patients with IGV-1 or large varices and those without portal hypertensive gastropathy or NSBB appear to have an increased risk of bleeding. In these circumstances, primary

prophylaxis should be considered, and the use of NSBB seems to be a suitable option. Disclosures: The following people have nothing to disclose: Tiago Curdia Goncalves, Joana Magalhaes, Carla M. Marinho, Jose Cotter Purpose: Liberal use of blood products to correct coagulopathy prior

to invasive procedures in patients with cirrhosis is a common practice, but evidence increasingly supports learn more a restrictive transfusion strategy for blood products in a variety of clinical scenarios, prompting this systematic review. Methods: Using a pre-established search engine, two reviewers independently evaluated and retrieved papers in EMBASE, PUBMED RG-7388 and reference lists, with no date or language restriction until December 2013. Only reports with cirrhotic patients undergoing elective minor procedures including data on bleeding complications were included. Papers on major surgeries were excluded. Results 3,972 abstracts were reviewed, of which 12 met the inclusion criteria. Four studies (4 case series) evaluated upper endoscopy (EGD), colonoscopy, and Orotic acid transesophageal echocardiography (TEE). Data from these studies are shown in the TABLE. Four studies (3 case series and 1 randomized controlled trial (RCT)) evaluated liver biopsy and found overall low rates of bleeding often without correction of coagulopathy, though the study of mini-laparascopic biopsy found bleeding

requiring hemostasis in 60% of cirrhotics. Three studies (2 RCTs and one case series) examined dental procedures and found bleeding in <6% of patients in the two RCTs (both in groups receiving and not receiving blood products) and in 23% in the case series. Finally, a RCT using eltrombopag vs placebo in cir-rhotics with platelets <50K followed by a supplemental platelet transfusion if needed prior to invasive procedures found bleeding complications in 17% of the eltrombopag group vs 23% of the placebo group (95% CI -15 to 3). Conclusion: Most of the evidence for correction of coagulopathy in cirrhotic patients comes from case series. Overall the data suggest that bleeding complications are infrequent. Given the potential complications of the over-transfusion of blood products, further RCTs or high-quality observational studies are needed to determine the best transfusion strategy in cirrhotic patients undergoing invasive procedures. TABLE.