C activation may play a key role in HSRs to Doxil. However, low-level C activation does not necessarily entail clinical symptoms, highlighting the probable involvement of further, as yet unidentified, amplification factors.

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.

Pegylated liposomal doxorubicin (Doxil) and 99mTc-HYNIC PEG liposomes (HPL) were reported earlier to cause hypersensitivity reactions (HSRs) in a substantial percentage of patients treated i.v. with these formulations. Here we report that (1) Doxil, HPL, pegylated phosphatidylethanolamine (PEG-PE)-containing empty liposomes matched with Doxil and HPL in size and lipid composition, and phosphatidylglycerol (PG)-containing negatively charged vesicles were potent C activators in human serum in vitro, whereas small neutral liposomes caused no C activation. (2) Doxil and other size-matched PEG-PE and/or PG-containing liposomes also caused massive cardiopulmonary distress with anaphylactoid shock in pigs via C activation, whereas equivalent neutral liposomes caused no hemodynamic changes. (3) A clinical study showed more frequent and greater C activation in patients displaying HSR than in non-reactive patients. These data suggest that liposome-induced HSRs in susceptible individuals may be due to C activation, which, in turn, is due to the presence of negatively charged PEG-PE in these vesicles.

Cardiac anaphylaxis is a severe, life-threatening manifestation of acute hypersensitivity reactions to allergens and drugs. Earlier studies highlighted an amplifying effect of locally applied C5a on the process; however, the role of systemic complement (C) activation with C5a liberation in blood has not been explored to date. In the present study, we used the porcine liposome-induced cardiopulmonary distress model for 1) characterizing and quantifying peripheral C activation-related cardiac dysfunction; 2) exploring the role of C5a in cardiac abnormalities and therapeutic potential of C blockage by soluble C receptor type 1 (sCR1) and an anti-C5a antibody (GS1); and 3) elucidating the role of adenosine and adenosine receptors in paradoxical bradycardia, one of the symptoms observed in this model. Pigs were injected intravenously with different liposomes [Doxil and multilamellar vesicles (MLV)], zymosan, recombinant human (rhu) C5a, and adenosine, and the ensuing hemodynamic and cardiac changes (hypotension, tachy- or bradycardia, arrhythmias, ST-T changes, ventricular fibrillation, and arrest) were quantified by ranking on an arbitrary scale [cardiac abnormality score (CAS)]. There was significant correlation between CAS and C5a production by liposomes in vitro, and the liposome-induced cardiac abnormalities were partially or fully reproduced with zymosan, rhuC5a, adenosine, and the selective adenosine A1 receptor agonist cyclopentyl-adenosine. The use of C nonactivator liposomes or pretreatment of pigs with sCR1 or GS1 attenuated the abnormalities. The selective A1 blocker cyclopentyl-xanthine inhibited bradycardia without influencing hypotension, whereas the A(2) blocker 4-(2-[7-amino-2-(2-furyl)[1,2,4]triazolo[2,3-a][1,3,5]triazin-5-ylamino]ethyl)phenol (ZM-24135) had no such effect. These data suggest that 1) systemic C activation can underlie cardiac anaphylaxis, 2) C5a plays a causal role in the reaction, 3) adenosine action via A1 receptors may explain paradoxical bradycardia, and 4) inhibition of C5a formation or action or of A1-receptor function may alleviate the acute cardiotoxicity of liposomal drugs and other intravenous agents that activate C.

No abstract