Whether plasma exchange would improve survival in antibiotic-treated canines with septic shock was investigated. Escherichia coli O86H8 (1.4 x 10(10)) was surgically implanted as an intraperitoneal clot in 18 two-year-old (10-12 kg) purpose-bred beagles. Beginning 4 hours after surgery, all animals received cefoxitin and gentamicin for 5 days. Three treatment groups were defined: 1) a no apheresis, or control group, (n = 6); 2) a sham apheresis group, whose whole blood plasma was removed, separated, and then transfused (n = 6); and 3) a plasma exchange group from whom blood and plasma were removed and separated, to whom the blood was returned, and in whom infected plasma was replaced with compatible fresh-frozen canine plasma (n = 6). For the sham apheresis and plasma exchange groups, a commercial blood cell processor was used to separate 1.5 blood volumes of plasma at 5 and 24 hours after surgery. Serial radionuclide left ventricular ejection fractions and femoral and pulmonary arterial catheter hemodynamics were measured simultaneously in awake animals. All six animals in the plasma exchange group died. In both the sham and control groups, only one of six animals survived. Survival times were ordered (median in hours) (control [372 h] > sham apheresis [48 h] > plasma exchange [24 h] [p<0.038]). Decreases in mean cardiac index and mean arterial pressure (from before apheresis to after) at 5 to 7 hours after surgery were ordered (plasma exchange > sham apheresis > control; p<0.03). Thus, plasma exchange in this controlled trial of septic shock was associated with decreased survival and worsened hemodynamics. When extrapolated to humans, these results suggest that plasma exchange would be unlikely to provide therapeutic benefit and might produce harmful effects in patients with septic shock.
