To define critical parameters concerning interferon (IFN) effects upon natural killer (NK) cells in vivo, we gave cancer patients serial weekly intramuscular injections of purified lymphoblastoid IFN in six doses ranging from 10(5) to 3 X 10(7) U. Dose sequences were determined by randomly allocating patients to one of six levels in a latin square ordering scheme. NK cell stimulation, a threefold peak increase above preinjection levels of cytolysis (P = 0.022), occurred in peripheral mononuclear cells (PMC) sampled 24 h postinjection, of 3 X 10(6) U, but was not detectable at any dose in PMC sampled 7 d postinjection. No blunting occurred in NK cell responsiveness to repeated injection of IFN dosages a second time at or several weeks after study completion. At IFN doses of 3 X 10(6), 10(7), and 3 X 10(7) U, a negative correlation existed between the amount of IFN injected and the average extent of NK cell activation (r = -0.423, P less than 0.05). This contrasted with the progressively increasing response of NK cells to in vitro incubation with increasing concentration of up to 3,000 U/ml of IFN. Overnight culturing of PMC sampled before IFN injections resulted in a mean 1.9-fold increase in cytolytic activity (P = 0.0005) and a mean 53% decrease in variance (P = 0.024) between serial preinjection NK cell activity determinations. Cell separation procedures may, therefore, have resulted in NK cell inactivation, from which overnight culturing permitted recovery. We found that maximal NK cell activation at a low IFN dose, decreasing NK cell responsiveness at higher doses, and the need to culture PMC to efficiently detect NK cell boosting may account for disparities in reported effects of IFN on NK cell function.
B S Edwards, J A Merritt, R C Fuhlbrigge, E C Borden