There is increasing interest in the role of polymorphonuclear (PMN) leukocytes in the evolution of focal cerebral infarction. Surgical preparation of focal cerebral ischemia models may alter leukocyte reactivity and thereby make interpretation of leukocyte function following ischemia/reperfusion difficult. The effects of surgical preparation and of experimental ischemia/reperfusion on granulocyte function have been examined prospectively in a baboon model. Twenty-six adolescent male baboons underwent surgical preparation, of which 21 underwent middle cerebral artery occlusion/reperfusion. Four additional animals served as nonsurgical controls. Peripheral venous blood specimens were taken for performing assays of leukocyte function at defined intervals before and after both the surgical preparation (i.e., the overall procedure for implantation of the middle cerebral artery occlusion device) and occlusion/reperfusion. A stress-related elevation in total leukocyte number was attributed mainly to an increase in the number of circulating PMN leukocytes. Values rose from 13.9 +/- 4.9 x 10(3) to 27.8 +/- 5.8 x 10(3)/microliters, (+/- SD; n = 21) for total leukocyte number, with p < 0.001, and from 4.3 +/- 2.1 x 10(3) to 15.9 +/- 4.7 x 10(3)/microliters (n = 21) for PMN leukocytes, with p < 0.001. Surgical preparation had no effect (p > or = 0.4) on the ability of PMN leukocytes, isolated 24 h after the implantation procedure, to display polarization, O2.- production, or beta-glucuronidase release when stimulated with human C5a. A moderate decrease in the chemotactic response to C5a resolved within the 7-day postsurgery (preocclusion) period. Three-hour middle cerebral artery occlusion and 1-h reperfusion resulted in a significant reduction in C5a-induced polarization.(ABSTRACT TRUNCATED AT 250 WORDS)