Steroid dependency or excess occurred in 14.9% of British IBD patients (in 7.1% potentially avoidable). We demonstrated positive effects of service configurations (IBD multi-disciplinary team, dedicated IBD clinics). Routine recording of steroid dependency or excess is feasible and should be considered a quality metric.

Summary Background Patients with IBD are at risk of excess corticosteroids. Aims To assess steroid excess in a large IBD cohort and test associations with quality improvement and prescribing. Methods Steroid exposure was recorded for outpatients attending 19 centres and associated factors analysed. Measures taken to avoid excess were assessed. Results Of 2385 patients, 28% received steroids in the preceding 12 months. 14.8% had steroid excess or dependency. Steroid use was significantly lower at ‘intervention centres’ which participated in a quality improvement programme (exposure: 23.8% vs 31.0%, P < .001; excess 11.5% vs 17.1%, P < .001). At intervention centres, steroid use fell from 2015 to 2017 (steroid exposure 30.0%‐23.8%, P = .003; steroid excess 13.8%‐11.5%, P = .17). Steroid excess was judged avoidable in 50.7%. Factors independently associated with reduced steroid excess in Crohn's disease included maintenance with anti‐TNF agents (OR 0.61 [95% CI 0.24‐0.95]), treatment in a centre with a multi‐disciplinary team (OR 0.54 [95% CI 0.20‐0.86]) and treatment at an intervention centre (OR 0.72 [95% CI 0.46‐0.97]). Treatment with 5‐ASA in CD was associated with higher rates of steroid excess (OR 1.72 [95% CI 1.24‐2.09]). In ulcerative colitis (UC), thiopurine monotherapy was associated with steroid excess (OR 1.97 [95% CI 1.19‐3.01]) and treatment at an intervention centre with less steroid excess (OR 0.72 [95% CI 0.45‐0.95]). Conclusions This study validates steroid assessment as a meaningful quality measure and provides a benchmark for this performance indicator in a large cohort. A programme of quality improvement was associated with lower steroid use.

Introduction The incidence of esophageal and junctional adenocarcinoma has increased sixfold in the past 30 years. Despite aggressive chemotherapy and attempts at curative surgery, the 5-year survival rate remains at 20%. Unlike other epithelial cancers, targeted therapy is limited. We have identified TRIM44 as a putatative oncogene that is amplified in 8% of EA and 6% of breast cancers. Methods The aims of this project were to elucidate the molecular phenotype underlying the oncogenic role of TRIM44 as well as suggest ways to therapeutically target TRIM44 dysregulation. Results Using three microarray datasets representing EA (n¼37, n¼64) and breast (n¼997) we performed gene set enrichment analysis to identify signalling pathways dysregulated with TRIM44 high expression (EA and breast) or amplification (breast only). High expression of TRIM44 was associated with over-enrichment of targets of the mTOR pathway consistent across all three data sets. This association was validated using expression microarrays in a cell line (HSC39) with amplification of TRIM44 treated with siRNA. Using phosphorylation of p70S6K as a readout of mTOR activity we validated the link between TRIM44 and the MTOR pathway; knockdown of TRIM44 using siRNA in HSC39 and a cell line with high expression of TRIM44 (JIMT-1) resulted in a decrease in pathway activity. The connectivity map (http://www.broadinstitute.org/cmap/), a collection of expression array data derived from lines treated with bioactive small molecules was queried using signatures generated from the microarray data representing genes positively or negatively associated with TRIM44 (breast, EA, HSC39 +SiRNA datasets). The top consistent hits were sirolimus, an mTOR inhibitor and analogue of rapamycin and LY-294002, a PI3K inhibitor. These hits represent small molecules predicted to reverse the effects of high TRIM44 expression. Treatment with inhibitors in HSC39 and JIMT-1 demonstrated that these lines were highly sensitive (IC50<30 nM) to rapamycin, but less sensitive to PI3K inhibition (IC50>400 nM), consistent with a link at the level of mTOR. Conclusion We have demonstrated the ability to identify a previously unknown association between TRIM44 and the mTOR pathway using expression and copy number data. This phenotype was validated in cell line experiments and highlighted a potential therapeutic strategy using analogues of rapamycin, a small molecule inhibitor of the mTOR pathway, which are currently in clinical use as immunosuppressants and in clinical trials for other cancer types.

No abstract