Products of cyclooxygenase-2 catalysis regulate postoperative bowel motility

Laparotomy involving manipulation of the small intestine causes injury, ini tiating an inflammatory cascade in the small bowel wall, which generates ei cosanoids and proinflammatory cytokines. We have shown that ketorolac and s alsalate, nonselective cyclooxygenase (COX) inhibitors, ameliorate postoper ative small bowel ileus in a rodent model. Others have shown that interleuk in-1 receptor antagonism improves postoperative gastric emptying. We examin ed whether inhibition of the proinflammatory cytokines, tumor necrosis fact or alpha (TNF alpha) and interleukin-1 (IL-1), or selective blockade of cyc looxygenase-2 (COX-2), the COX isoform induced during inflammation would ac celerate postoperative small bowel transit in our model. Duodenostomy tubes were inserted into male Sprague-Dawley rats. One week later, animals were randomized to receive TNF-binding protein (TNF-bp), IL-1 receptor antagonis t (IL-1ra), or saline (NS) prior to standardized laparotomy. Additional rat s were gavaged preoperatively with a selective COX-2 inhibitor (NS-398) or NS. Small intestinal transit was measured as the geometric center (GC) of d istribution of (CrO4)-Cr-51 at 30 min, 3 h, or 6 h (n = 5-9 rats/group) fol lowing laparotomy. Selective inhibition of COX-2 significantly increased po stoperative small bowel transit compared to controls (GC 2.9 +/- 0.3 vs 2.2 +/- 0.1 at 30 min, GC 2.9 +/- 0.3 vs 2.5 +/- 0.2 at 3 h, and GC 3.3 +/- 0. 3 vs 2.8 +/- 0.2 at 6 h, P < 0.05). In contrast, neither TNF-bp nor IL-1ra altered postoperative small intestinal transit in this model. Use of select ive COX-2 inhibitors may accelerate recovery of postoperative bowel dysmoti lity without the undesirable effects (e.g., gastrointestinal irritation and anti-platelet effect) of nonselective COX inhibitors. (C) 1999 Academic Pr ess.