HPB P44 Incidence of pancreatitis induced splanchnic venous thrombosis in a tertiary care hospital

Abstract

Abstract Background Splanchnic venous thrombosis (SPVT) is a well-recognized vascular complication of acute pancreatitis (AP). Exact incidence is not clear, however in recent years, the incidence of SPVT is on the rise. This is partially attributed by the availability of computed tomography. Most commonly involved veins are splenic vein (SV), superior mesenteric vein (SMV) and portal vein (PV) in the same order of frequency. Systemic anticoagulation is the presumed treatment for SPVT but evidence of benefit is still not clear. Further complications such as gastrointestinal bleeding, intestinal ischemia, anaemia, thrombocytopenia, infection and oesophageal or gastric varices has been documented which may worsen patient's outcomes. Currently, no data on incidence has been documented in Scotland and there are no national guidelines on the anticoagulation management for SPVT in acute pancreatitis. This study aims to determine the incidence of SPVT complicating AP, the possible aetiological risk factor, the correlation of severity of AP with the development of SPVT, the current practice in management via anticoagulation and outcome of these patients, all in a single health board. Methods A cross-sectional study was conducted, including patients admitted to Aberdeen Royal Infirmary (ARI) and Dr. Gray's Hospital (DGH) with the primary diagnosis of AP from 1st January 2017 to 31st March 2018. Data was obtained from the health intelligence as per NHS Grampian protocol. Data collection was conducted retrospectively by assessing patient's electronic records via TrakCare on NHS Grampian computers, then inputted into REDCap software. Variables instruments heading patient background, admission/diagnosis background, pancreatitis severity, microbiology, organ support, feeding and nutrition, imaging/complications, SPVT, inpatient intervention and follow-up was set up. CT severity index (CTSI) was used to stratify pancreatitis severity. These scores were made based on interpretation of CT scans formal reports. Outcomes parameters were measured by length of hospital stay (LOS) in days and death during admission. Lastly, data was exported from REDCap and statistical analyses were conducted in IBM SPSS Statistics Version 27. Mean, standard deviation and range were calculated for continuous variables while frequencies and percentages for categorical variables. Chi-square test was used for correlation between categorical variables. P value of < 0.05 was set as level of significance. Results Of 219 patients admitted with AP, 202 had either CT scan or/and ultrasound scan conducted, recording incidence of SPVT to be 16/202(7.92%). Patients without scans were excluded as presence of undiagnosed due to its asymptomatic nature. Among SPVT group, 8/16(50%) were gallstone-related, 4/16(25%) alcohol, 1/16(6.25%) post-instrumental(iatrogenic) and 3/16(18.75%) idiopathic. Location of thromboses categorised into SV only in 7(43.8%), SMV only 2(12.5%), PV only 1(6.3%), SV+SMV 4(25%) and SV+ SMV+PV 2(12.5%). 125 patients received CT scan, SPVT group displayed 6/125(4.8%) CTSI ranging 0–5 and 9/125(7.2%) for 6–10, while non-SPVT group 101/125(80.8%) and 9/125(7.2%) respectively, p=0.000. Mortality rate due to SPVT recorded 1/10(10%) and LOS ranged from 0–360 days, mean:14.91, SD:41.43, p=0.000. 4/16(25%) SPVT patients had therapeutic anticoagulation, 3/16(18.8%) not anti-coagulated and 9/16(56.4%) had no data. In relation to thromboses site, 1/7(14.29%) SV only thrombosis received anticoagulation, ½(50%) for SMV only, ¼(25%) for SV+SMV and ½(50%) for SMV+SV+PV. On discharge, 5(62.5%) received anticoagulation, 3 months for 4 patients and 6 months for 1, 2/5(40%) oral and 3(60%) injectable. Conclusions Incidence of pancreatitis induced SPVT was 7.29% during the study period. No aetiological risk factor was determined. Severity of pancreatitis measured by CTSI was directly correlated with increased incidence of SPVT. Besides that, a high rate of idiopathic pancreatitis (37.9%) was found which is not in line with NICE guidance. Hence, further investigations should be offered to establish the aetiology of pancreatitis. In NHS Grampian, we observed mainly supportive management in SV thrombosis while anticoagulation commenced for PV and SMV involvement. Extension of thrombosis also prompted initiation of treatment. For future studies, state of splanchnic circulation after anticoagulation, locations of collaterals, acute and late complications among SPVT can be better analysed.