Intravenous injection of liposomes can cause significant pulmonary hypertension in pigs, a vasoconstrictive response that provides a sensitive model for the cardiopulmonary distress in humans caused by some liposomal drugs. The reaction was recently shown to be a manifestation of "complement activation-related pseudoallergy" (CARPA; Szebeni J, Fontana JL, Wassef NM, Mongan PD, Morse DS, Dobbins DE, Stahl GL, Bünger R, and Alving CR. Circulation 99: 2302-2309, 1999). In the present study we demonstrate that the composition, size, and administration method of liposomes have significant influence on pulmonary vasoactivity, which varied between instantaneously lethal (following bolus injection of 5 mg lipid) to nondetectable (despite infusion of a 2,000-fold higher dose). Experimental conditions augmenting the pulmonary hypertensive response included the presence of dimyristoyl phosphatidylglycerol, 71 mol% cholesterol, distearoyl phosphatidylcholine, and hemoglobin in liposomes, increased vesicle size and polydispersity, and bolus injection vs. slow infusion. The vasoactivity of large multilamellar liposomes was reproduced with human C3a, C5a, and xenoreactive immunoglobulins, and it correlated with the complement activating and natural antibody binding potential of vesicles. Unilamellar, monodisperse liposomes with 0.19 +/- 0.10 microm mean diameter had no significant vasoactivity. These data indicate that liposome-induced pulmonary hypertension in pigs is multifactorial, it is due to natural antibody-triggered classic pathway complement activation and it can be prevented by appropriate tailoring of the structure and administration method of vesicles.