AODWE-010 Severe gastrointestinal dysmotility related intestinal failure: chronic intestinal pseudo-obstruction, enteric dysmotility or a ‘pragmatic’ approach? experience from a national referral centre

IntroductionGastrointestinal dysmotility (GID) is a recognised cause of Type 3 Intestinal Failure (IF). Diagnosis is challenging, with uncertain merits of subclassifying into chronic intestinal pseudo-obstruction (CIPO) and Enteric Dysmotility (ED)1. A ‘pragmatic’ algorithm for GID has been proposed to address these difficulties2. Retrospectively, we evaluated the performance of ED/CIPO diagnostic criteria and pragmatic GID criteria in predicting outcomes of patients with ‘dysmotility’ requiring home parenteral nutrition (HPN).MethodPatients with primary dysmotility referred for HPN to a national IF Unit (1999–2015) were identified. 2 investigators reviewed casenotes for clinical histories, motility tests, imaging, full-thickness biopsies (FTB), breath tests, HPN and survival outcomes. Cases were excluded if they did not meet pragmatic GID criteria (morbidity + either >1 region abnormal motility test or positive FTB + small bowel (SB) involvement). Patients satisfying GID criteria were then sub-categorised into; ED (objective SB dysmotility), CIPO (ED criteria + non-mechanical SB dilatation) and ‘non-CIPO/ED’ if neither criteria were met. Outcomes were compared between subgroups.Results44 patients met GID criteria (median age 40, 30 females, 18/44 with abnormal FTBs (myopathy n=12, neuropathy n=6) and median follow-up 5 years)), whereas 7 patients not meeting pragmatic criteria were excluded. HPN improved BMI (median pre 18.3 vs. post 21.2 kg/m2, p=0.01). HPN outcomes: survival at 1, 5 and 10 years was 95%, 77% and 47%, median catheter-related blood stream infection rate (CRBSI) 0.5/1000 days, and 18% achieved HPN independence. 25/44 (57%) met criteria for CIPO, 12/44 (27%) ED and 7/44 (16%) were ‘non-CIPO/ED’ patients that met GID criteria. CIPO was associated with bacterial overgrowth (13/25 vs. 3/19, p=0.03), and HPN dependency (1/25 weaned off HPN vs. 7/19, p=0.01) when compared to the ED and ‘non-CIPO/ED’ groups. Other factors including the yield of FTBs, opiate-dependence and HPN complications (CRBSI, IF associated liver disease and catheter related venous thromboses) did not significantly differ between the three groups.ConclusionOur data emphasise important prognostic implications of CIPO diagnosis in the IF/GID population as a predictor of HPN dependency. Combining pragmatic evidence-based criteria with imaging to exclude CIPO is likely to give the maximum diagnostic and prognostic yield across the spectrum of GID.References. Wingate, et al. Journ