Thrombogenicity of total parenteral nutrition solutions: I. Effect on induction of monocyte/macrophage procoagulant activity

Although thrombosis is a frequent complication of total parenteral nutrition (TPN), its pathogenesis has received little scientific attention. We have studied, in vitro, the effects of the component solutions of TPN on the induction and modulation of human monocyte procoagulant activity, an initiator of coagulation. Human peripheral blood mononuclear cells were cultured with (a) 200 μl of dextrose solution (10%, 15%, 20%, 25%, and 50%), (b) 200 μl of amino acid solution (full, one-half, and onequarter strength), and (c) 200 μl of isosmolar 10% lipid emulsion (LE). Cocultures of LE and 20% dextrose, LE and full-strength amino acid solution, and LE and bacterial lipopolysaccharide were also studied. Cells cultured with lipopolysaccharide or medium alone constituted positive and negative controls, respectively. In addition, cocultures of LE and 20% dextrose, LE and full-strength amino acid solution, and LE and lipopolysaccharide were also studied. Cells were incubated for intervals of 12–72 h, washed, frozen, and assayed for monocyte procoagulant activity (MPCA). Milliunits of MPCA were derived from a standard thromboplastin curve. In addition, spontaneous MPCA levels were measured in healthy volunteers (n = 4) and “home” total parenteral nutrition patients (n = 4) before and after a 2-h infusion of 500 ml of LE. Our results show that, in vitro, hypertonic dextrose and fullstrength amino acid solutions induce significant levels of MPCA. Induction of MPCA by dextrose was lymphocyte-independent. Although a significant increase in MPCA by full-strength amino acid solution was seen in cultures of isolated monocytes, a lymphocyte requirement was demonstrated for full MPCA. In contrast, LE significantly inhibited the induction of MPCA by 20% dextrose and fullstrength amino acid solution. This inhibitory activity was at the monocyte level. Subfractionation of the LE into triglyceride and phospholipid phases showed the inhibitory capacity to reside in the former. In vivo, patients on home total parenteral nutrition expressed higher spontaneous MPCA levels than normal controls. Ten percent lipid emulsion infusion abolished MPCA expression in both groups. These corroborative in vitro and in vivo data suggest a mechanism for the thrombogenicity of total parenteral nutrition solutions and that the inhibitory properties of LE may be of practical advantage in preventing thrombosis.