Daniel G. Remick and Steven L. Kunkel,     Department of Pathology, University of Michigan Medical School, Ann Arbor, Michigan 48109

Publisher Summary

This chapter discusses the pathophysiologic alterations that can be induced by tumor necrosis factor (TNF). The most sensitive parameter for detecting an effect of TNF is the alteration in peripheral blood constituents. In a study described in the chapter, rHuTNF is injected into mice, which results in the rapid induction of lymphopenia and neutrophilia. The kinetics of these peripheral blood alterations are extremely rapid, with statistically significant alterations occurring within 1 hr. The lymphopenia and neutrophilia are both absolute and relative, that is, there is both a decrease in the percentage of lymphocytes and a decrease in the total number of circulating lymphocytes. In a microscopic examination described in the chapter, the toxicity of TNF appeared to be dose-related. At 1 μg/mouse, there were occasional foci of necrosis of the mucosa in the small intestine. This necrosis was observed primarily at the tips of the villi. At the higher dose of 10 μg, there were more severe changes. TNF-treated animals had blunting of the villi with frank necrosis of the mucosa at the tips of the villi. In a study performed through electron microscopy, the small intestine was examined by electron microscopy to confirm the light microscopy changes and to provide further insight into the mechanism of damage.

I Introduction

Tumor necrosis factor-α (TNF) is a small peptide mediator secreted primarily by cells of monocyte lineage. This 17,000-Da cytokine exerts multiple effects both in vitro and in vivo. TNF was first described as an oncolytic agent directed against solid tumors (Carswell et al., 1975), but further work began to disclose its broad range of activity. TNF has been implicated in the pathogenesis of several diseases and inflammatory conditions, including rejection of transplanted solid organs (Maury and Teppo, 1987), congestive heart failure (Levine et al., 1990), arthritis (Saxne et al., 1988), parasitic infections (Scuderi et al., 1986), glomerulonephritis (Remick, 1991), and acquired immunodeficiency syndrome (Lahdevirta et al., 1988). It must be mentioned that this is by no means a complete list of diseases in which TNF has been implicated in the altered pathophysiology.

The strongest evidence for TNF participation in a disease state is found in the data describing TNF and septic shock. Data supporting the hypothesis that TNF mediates septic shock have been provided by multiple independent laboratories and consist of four parts. First, in experimental animal models of septic shock, TNF is produced and secreted into the circulation within 1 to 2 hr. The rapid production of TNF is observed in humans (Michie et al., 1988), rabbits (Beutler et al., 1985b; Mathison et al., 1988), and rodents (Waage, 1987; Remick et al., 1989). The second line of evidence for the role of TNF is the detection of TNF in the serum of patients in septic shock (Waage et al., 1987). In this now classic study, TNF was present in the serum of 10 of 11 patients who died, but was in the serum of only 6 of 68 survivors. All patients with greater than 100 pg/ml of TNF in their serum died. More recent work has confirmed this earlier report (Debets et al., 1989; Marks et al., 1990). The third piece of information is given by experiments wherein antibodies to TNF will prevent the lethality observed after injection of endotoxin (Beutler et al., 1985a) or live gram-negative bacteria (Tracey et al., 1987a). The study with live bacteria raised some concerns about potential endotoxin contamination in the antibody preparation, because the antibody needed to be given 2 hr prior to the bacteria in order to be efficacious. Work by Chong and Huston (1987) had demonstrated that endotoxin contamination in antibody preparations could confer nonspecific protection if the antibodies were given 2 hr prior to lipopolysaccharide (LPS). These doubts were alleviated by Hinshaw et al (1990), who started the antibody treatment after 30 min of infusion of bacteria, and were still able to confer protection. The last piece of evidence for the role of TNF became available when sufficient amounts of recombinant material could be provided to investigators. Workers in several labs have been able to inject this purified material into experimental animals and induce altered pathophysiology and organ injury. Tracy et al (1986) first reported that injection of recombinant human TNF (rHuTNF) would induce shock and tissue injury. Since that initial report, we and other groups have provided additional evidence of the effects of TNF injection into experimental animals. These data represent the focus of this review.

II TNF-Induced Peripheral Blood Alterations