A novel liposomal bupivacaine formulation using an ammonium sulfate gradient

INTRODUCTION : Liposomal local anesthetics prolong analgesic duration in animals and humans. However, previously described liposomal local anesthetic formulations have relatively low drug/lipid ratios, with 0.26 being the greatest ratio reported to date (1). A low drug/lipid ratio necessitates administration of a large lipid load to achieve a prolonged analgesic effect. This may be problematic for subcutaneous administration of liposomes, where persistence of a large lipid mass would be conspicuous. We describe here a novel type of liposome using a remote loading technique to designed to achieve a greater drug/lipid ratio. METHODS: Lipid films were prepared from hydrogenated phosphocholine and cholesterol (CHOL) at 33% or 50% molar ratio. The films were hydrated with 250 mM (NH4)2SO4 to form multilamellar vesicles (MLVs). Some MLVs were homogenized to prepare small unilamellar vesicles by high pressure homogenization at 10,000-15,000psi, and the SUVs were used to prepare giant multivesicular vesicles (GMVs) by freeze-thawing 10 times. A trans-membrane (NH4)2SO4 gradient was established by removing unecapsulated (NH4)2SO4 from MLVs and GUVs, and bupivacaine was remotely loaded bupivacaine along the gradient. Liposomes were characterized for size using photon correlation spectroscopy, bupivacaine concentration using HPLC, and lipid content using Stewart's assay. The molar drug/lipid ratio was then calculated. Analgesic duration was quantified by response to cutaneous electrical stimulation in a mouse model. Current was delivered to the skin overlying the abdomen. Failure to vocalize in response to stimulation with threshold current was taken as analgesia. Mice were injected subcutaneously with 150 ul of 2% liposomal bupivacaine, 0.5% plain bupivacaine (as preliminary studies demonstrated that greater concentrations were often lethal), empty MLVs and GMVs (not loaded with drug) or normal saline (n= 8 mice/group). Analgesia was assessed at regular intervals, and continued until two successive tests resulted in vocalization (i.e., absence of analgesia). RESULTS: For 33% CHOL, MLV size was 5213±1101 and GMV size was 9542±2035. For 50% CHOL, MLV size was 5002±1283 and GMV size was 4318±1799. For 33% CHOL, MLV and GMV drug/lipid ratio was 0.3 and 1.1, respectively. For 50% CHOL, MLV and GMV drug lipid ratio was 0.5 and 2.3, respectively. Compared to plain bupivacaine, analgesia was prolonged for all liposomal formulations, with the 33% CHOL GMV producing the greatest analgesic duration. CONCLUSIONS: The remote loading of bupivacaine along an ammonium sulfate gradient into GMV liposomes resulted in much greater drug/lipid ratios than have previously been reported (1). Furthermore, the effective prolongation of analgesia by the liposomal formulations demonstrated that drug release was sufficient to block nerve conduction. The dual characteristics of a high drug/lipid ratio and effective analgesic prolongation are critical for designing a liposomal local anesthetic for eventual clinical use.