Severe gastrointestinal bleeding after hematopoietic cell transplantation, 1987-1997: incidence, causes, and outcome.

OBJECTIVE: Severe GI bleeding after hematopoietic cell transplantation is commonly due to lesions that are unusual in nontransplant patients. The frequency of GI bleeding appears to have decreased over the last decade, but the reasons have not been readily apparent. We sought to determine the incidence of severe bleeding during two time periods, to describe the causes and outcomes of bleeding, and to analyze the reasons behind an apparent decline in severe bleeding over the decade covered. METHODS: During 1986-1987 and 1996-1997, we followed all patients with and without severe bleeding at our institution, a marrow transplant center. RESULTS: Over this decade, the incidence of severe bleeding declined from 50/467 (10.7%) to 15/635 (2.4%) (p < 0.0001). Overall mortality from intestinal bleeding declined from 3.6% to 0.9% (p = 0.002), but mortality in those with bleeding remained high (34% vs 40%). The onset (day 42 vs 47) and platelet counts (35,994 vs 37,600/microl) were similar, but the sites and causes of bleeding were different. During 1986-1987, 27/50 patients bled from multiple GI sites, viral and fungal ulcers, or graft-versus-host disease (GVHD). Over the decade, bleeding from GVHD had decreased 80% (p < 0.0001), and bleeding from viral (p < 0.0001) and fungal (p = 0.023) ulcers almost disappeared. CONCLUSIONS: The incidence of severe GI bleeding has declined significantly over the last decade because of prevention of viral and fungal infections and severe acute GVHD. However, severe bleeding after transplant remains a highly morbid event, particularly among patients with GVHD.

Histamine stimulates ion transport by dog pancreatic duct epithelial cells through H1 receptors.

Histamine affects pancreatic secretion, but its direct action on ion transport by pancreatic duct epithelial cells (PDEC) has not been defined. We now characterize the secretory effects of histamine on cultured, well-differentiated, and nontransformed dog PDEC. Histamine stimulated, in a concentration-dependent manner (1-100 microM), a cellular 125I- efflux that was inhibited by 500 microM 5-nitro-2-(3-phenylpropylamino)benzoic acid, 2.5 mM diphenylamine-2-carboxylate, and 500 microM DIDS and thus mediated through Ca2+-activated Cl- channels. Histamine-stimulated 125I- efflux was 1) inhibited by 100 microM diphenhydramine, an H1 receptor antagonist, 2) resistant to 1 mM cimetidine, an H2 receptor antagonist, 3) not reproduced by 1 mM dimaprit, an H2 agonist, and 4) inhibited by 50 microM 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid-AM, a Ca2+ chelator, suggesting that it was mediated through H1 receptors acting via increased cytosolic Ca2+. Histamine also stimulated a 86Rb+ efflux that was sensitive to 100 nM charybdotoxin and thus mediated through Ca2+-activated K+ channels. When PDEC monolayers were studied in Ussing chambers, a short-circuit current of 21.7 +/- 3.1 microA/cm2 was stimulated by 100 microM histamine. This effect was inhibited by diphenhydramine but not cimetidine, was not reproduced with dimaprit, and was observed only after serosal addition of histamine, suggesting that it was mediated by basolateral H1 receptors on PDEC. In conclusion, histamine, acting through basolateral H1 receptors, activates both Ca2+-activated Cl- and K+ channels; in this manner, it may regulate PDEC secretion in normal or inflamed pancreas.