compounds showed proliferation >50% of control proliferation rate and >20% absolute difference between knockout and wildtype H69 cholangiocytes. Of these, 7 compounds showed higher proliferation in wildtype than PRKCSH knockout cells. These included 3 anti cancer drugs, 1 haemostatic drug, 1 antipsychotic drug, 1 antimalarial drug, and 1 antimicrobial agent. Octreotide showed no difference in proliferation between incubated cells and controls, nor between knockout and wildtype cells. Conclusion: We identified 7 FDA approved drugs that reduce proliferation rates in PRKCSH knockout cholangiocytes without large effect on proliferation rates in wildtype H69 cells. These drugs target different pathways than somatostatin analogues, and may become future pharmaceutical options for patients unresponsive to current treatment. SAT-051 Splenic iron deficiency is characteristic for HFE-associated hemochromatosis A. Viveiros 1 , B. Schaefer 1 , M. Tobiasch 1 , A. Finkenstedt 1 , C. Kremser 2 , M. Plaikner 2 , B. Henninger 2 , H. Tilg 1 , H. Zoller 1 . 1 Medical University Innsbruck, Internal Medicine, Innsbruck, Austria; 2 Medical University Innsbruck, Radiology, Innsbruck, Austria Email: andre.viveiros@i-med.ac.at Background and Aims: Non-invasive hepatic iron quantification using magnetic resonance imaging (MRI) is a staging method for patients with known iron overload and a diagnostic tool for the work- up of patients with suspected or unexplained iron overload. The aim of the present study was to assess the utility of spleen iron concentrations for the evaluation of hyperferritinemia. Method: Of all patients who presented with high serum ferritin between July 2001 and August 2015, 443 were HFE genotyped and underwent non-invasive liver and spleen iron quantification by T2* weighted MRI, where the degree of iron content is expressed as R2* (=1/T2*). Liver iron overload was defined as a R2* > 60 sec −1 and spleen iron overload as a R2* > 50 sec −1 . Results: Clinical, biochemical and radiological data are summarized in Table 1. Fifty-five patients were homozygous for the p.C282Y mutation in the HFE gene. When the latter were compared to patients with all other genotypes, median R2* was significantly higher in the liver (179.3 vs. 68.6, p <0.001) but also significantly lower in the spleen (38.3 vs. 47.1, p=0.003). Similar results were found when patients with compound heterozygosity for the p.C282Yand p.H63D mutations and homozygous for the p.H63D mutation were excluded from the analysis. Patient groups did not differ in terms of age, BMI and serum ferritin. Median transferrin saturation (%) was signifi- cantly higher in the p.C282Y homozygous group (77 vs. 35, p < 0.001). Table 1: Median values are represented and the 25th and 75th percentiles are given between parentheses. p.C282Y homozygous (n = 55) all other HFE genotypes (n = 388) p value Age, years 53 (36–62) 52 (42–63) 0.33 BMI, kg/m 2 25.3 (23.8–27.1) 26.3 (23.2–29.6) 0.14 Serum iron, μmol/L 34.1 (25.8–40.6) 21.5 (17.3–28.2) <0.001 Serum ferritin, μg/L 624 (282–1344) 618 (388–961) 0.70 Transferrin, mg/dL 204 (180–220) 247 (221–271) <0.001 Transferrin saturation, % 77 (51–83) 35 (28–46) <0.001 Alanine aminotransferase, U/l 29 (22–50) 38 (24–66) 0.049 R2* liver, sec −1 179.3 (95.9–342.1) 68.6 (53.0–91.5) <0.001 R2* spleen, sec −1 38.3 (28.8–52.3) 47.1 (34.7–63.8) 0.003 Conclusion: High ferritin is a common clinical finding that can indicate iron overload, inflammation and cell death. Our study shows that p.C282Y homozygous patients have significantly higher hepatic iron. In accordance with the low hepcidin state of hemochromatosis, reduced splenic iron concentrations in this condition indicate uncontrolled iron release and high transferrin saturation. Splenic iron can therefore be used to guide genetic testing beyond HFE genotyping, where patients with high liver but low spleen iron should preferably be tested for non-HFE mutations. In patients with high spleen iron, dietary and metabolic factors should be considered the main cause of high ferritin. SAT-052 Transcatheter recanalization with angioplasty and/or stenting: a novel, minimally-invasive treatment option for children with chronic portal vein thrombosis with cavernous transformation H. Justino 1 , K. Kukreja 2 , J.A. Hernandez 2 , Y. Pham 1 , S. Harpavat 1 , L. Karam 1 , P. Hertel 1 , B. Carter 1 , B. Shneider 1 , T. Miloh 1 . 1 Texas Children’s Hospital, Pediatrics, Houston, United States; 2 Texas Children’s Hospital, Radiology, Houston, United States Email: hjustino@bcm.edu Background and Aims: Chronic portal vein thrombosis (CPVT) with cavernous transformation of the portal vein (PV) is a leading cause of gastrointestinal (GI) bleeding in children. We soughtto evaluate the feasibility of performing transcatheter recanalization of CPVT with balloon angioplasty (BA) and/or stenting for CPVT. Method: Retrospective study of patients (pts) taken to catheteriza- tion (cath) for recanalization of CPVT from April 2014 to November 2017. Results: 13 pts (7 female) underwent 18 caths; median weight= 13.9 kg (range 7.5–88) and age = 4.1 years (1–15.3). Pre-procedural imaging included MRI and Doppler ultrasound in all, with CPVT and cavernous transformation diagnosed in all. 4 pts had history of prematurity and prior umbilical venous catheterization. CPVT manifestations included GI bleeding (10 pts), hyperammonemia (3), thrombocytopenia (11) and splenomegaly (11). Right heart cath & hepatic vein wedge angiography were performed, followed by percutaneous trans-splenic and/or direct puncture of intrahepatic portal veins. Additional findings at cath: mild pulmonary hyperten- sion (1 patient) and mild hepatopulmonary syndrome (1). Median cath time was 7.2 hours (range 2–11.5). Recanalization of CPVT was successful in 4 pts (single cath in 3 pts, 4 caths in 1 patient), unsuccessful in 7, and not attempted in 2. Successful procedures entailed BA of main and right PV with rheolytic thrombectomy in left PV in one case, BA of main PV and 4 branches of right PV with rheolytic thrombectomy in another case, BA of main, right and left PV in another case, and BA of main PV and 6 intrahepatic PV branches followed by re-occlusion requiring re-angioplasty of 4 intrahepatic PV branches and placement of multiple stents during 3 subsequent caths in fourth case. Diffuse intrahepatic PV occlusion with non-confluent left and right PVs was found in 2 of 4 successful cases. There were no major complications or mortality. Minor complication of bleeding from splenic puncture requiring transfusion in 3 caths (2 among unsuccessful cases). All 4 successful cases were treated with low molecular weight heparin, with additional dual antiplatelet therapy in 1. At follow-up from final cath (range 4–20 months), the 4 pts with successful CPVT recanalization have patent PVs on ultrasound, with no recurrence of GI bleed and improved platelet count and spleen size. Conclusion: Transcatheter recanalization of CPVT with BA/stenting is feasible in a subset of pts, including those with diffuse intrahepatic PV occlusion that would not otherwise be candidates for meso-Rex shunting. Although transcatheter recanalization is lengthy and technically challenging, it can be attempted with low risk. Further research is needed to determine the most suitable candidates, and whether this approach should be considered as a first-line option for children with CPVT. POSTER PRESENTATIONS S628 Journal of Hepatology 2018 vol. 68 | S605–S842