The immune microenvironment of human fracture/soft-tissue hematomas and its relationship to systemic immunity

Abstract

The immune environment of human soft-tissue injury is unstudied. We studied fracture soft-tissue hematomas (FxSTH) in 56 patients with high-energy bony fractures. FxSTH serum and mononuclear cells (MNC) as well as fracture patient plasma and blood MNC were studied. Twenty healthy controls donated plasma and MNC. Soluble tumor necrosis factor (TNF)-alpha, interleukin (IL-1 beta, IL-2, 6, 8, 10, 12, and interferon-gamma were studied by enzyme linked immunosorbent assay. Cells were studied by flow cytometry after cell-membrane stains for CD-14, TNF-alpha (mTNF), and human leukocyte antigen-DR, or intracellular stains for TNF (icTNF) and IL-10. Thirty-six patients with Injury Severity Score < 15 were analyzed further to evaluate the effects of isolated fracture on systemic immunity. Cytokines were rarely detectable in control plasma. TNF-alpha, IL-1 beta, IL-2, and interferon-gamma were rarely found in FxSTH serum or fracture patient plasma. All FxSTH sera were rich in IL-6, peaking before 48 hours (12,538 +/- 4,153 vs. 3,494 +/- 909 pg/mL, p = 0.02, U test). In Injury Severity Score < 15, IL-6 was not detectable in most early fracture patient plasma, but rose after 48 hours (p = 0.028). FxSTH serum IL-8 peaked after 48 hours (440 +/- 289 vs. 4,542 +/- 1,219 pg/mL, p = 0.006) and circulating IL-8 appeared after 72 hours. IL-6 and IL-8 showed gradients from FxSTH serum to paired PtS (p < 0.05, Wilcoxon). IL-10 was abundant (884 +/- 229 pg/mL) in FxSTH serum < 24 hours old. FxSTH serum IL-12 peaked late (3,323 +/- 799 pg/mL, day 4-7) then fell (p < 0.001, analysis of variance). Only IL-12 was higher in fracture patient plasma (1,279 +/- 602 pg/mL) than FxSTH serum (591 +/- 327 pg/mL) during the first 48 hours (p = 0.032, U test). On flow cytometry, control monocytes expressed 201 +/- 31 mTNF sites/cell, but icTNF was absent. mTNF was up-regulated after injury more in FxSTH monocytes (3,202 +/- 870 sites/cell) than peripheral blood monocytes (584 +/- 186 sites/cell) (p < 0.05 vs. peripheral blood monocytes by Wilcoxon, p < 0.001 vs. control monocytes by U test). Intracellular IL-10 was abundant in all MNC, but varied widely after injury. Fracture and peripheral blood monocytes expressed far less human leukocyte antigen-DR than control monocytes. Fractures create an inflammatory local environment. Proximal mediators are cell-associated and relatively confined to the wound, but soluble IL-6, IL-8, and IL-10 are abundant and probably exported. Systemic MNC have complex responses to local injuries. These may reflect the combined impact of multiple soluble cytokines initially generated within the wound. FxSTH appear to be a potentially important source of immunomodulatory cytokines in trauma.