Background Therapeutic options for management of diabetic gastroparesis are limited. Failure to maintain upregulation of heme oxygenase (HO1) leads to loss of interstitial cells of Cajal and delayed gastric emptying in non-obese diabetic (NOD) mice. Our hypothesis was that hemin upregulation of HO1 would restore normal gastric emptying in humans with gastroparesis. Aims To compare effects of hemin and placebo infusions on HO1 activity and protein, gastric emptying, autonomic function, and gastrointestinal symptoms in diabetic gastroparesis. Methods In a single-center, double-blind, placebo-controlled, randomized clinical trial, we compared intravenous hemin, prepared in albumin, or albumin alone (placebo) in 20 patients, aged 41±5 (SEM) years with diabetic gastroparesis. After infusions on days 1, 3, and 7, weekly infusions were administered for 7 additional weeks. Assessments included blood tests for HO1 protein and enzyme activity levels, gastric emptying with 13C-spirulina breath test, autonomic functions (baseline and end), and gastrointestinal symptoms every 2 weeks. Results Nine of 11 patients randomized to hemin completed all study procedures. Compared to placebo, hemin increased HO1 protein on days 3 (P=.0002) and 7 (P=.008) and HO1 activity on day 3 (P=.0003) but not after. Gastric emptying, autonomic functions, and symptoms did not differ significantly in the hemin group relative to placebo. Conclusions Hemin failed to sustain increased HO1 levels beyond a week and did not improve gastric emptying or symptoms in diabetic gastroparesis. Further studies are necessary to ascertain whether more frequent hemin infusions or other drugs would have a more sustained effect on HO1 and improve gastric emptying.

Angiopoietin‐2 (Ang‐2) is known to influence blood and lymphatic vessel remodeling, but how Ang‐2 gene knockout influences experimental inflammatory bowel disease (IBD) is not known. We used Ang‐2 gene knockout and control mice in the 3% dextran sulfate model of IBD to investigate how the lack of Ang‐2 affects vascular remodeling and indices of disease activity (over 7d). Ang‐2 +/+ and ‐/‐ mice showed equivalent levels of disease activity (stool blood, form and wasting) and weight loss was similar in both groups. There was a signficant increase in both blood (MECA‐32+) and lymphatic (VEGFR‐3+) vessel densities in colitic (DSS‐treated) groups (p<0.05). These increases were significantly reduced in Ang‐2 KO mice, and were also accompanied by decreased concentrations of gut neutrophils (myeloperoxidase activity). Colitic inflamed Ang‐2 ‐/‐ gut tissue show evidence of dysplastic MECA32+/VEGFR3+ networks which were not observed in non‐colitic Ang‐2 KO's. Histopathological scores in DSS colitis were partially, but not completely, resolved by Ang‐2 gene deficiency, and suggest that Ang‐2 plays complex roles in the initiation of IBD possibly by controlling tissue remodeling and leukocyte infiltration of tissues.