Manufacturing of liposomal nanomedicines (e.g. Doxil®/Caelyx®) is a challenging and slow production process based on multiple-vessel and batch processing techniques. As a result, the translation of these nanomedicines from bench to bedside has been limited. Microfluidic-based manufacturing offers the opportunity to address this issue, and de-risk the wider adoption of nanomedicines for drug delivery. Here we demonstrate the application of microfluidics suitable for continuous manufacturing of PEGylated liposomes encapsulating ammonium sulfate (250 mM). Doxorubicin was subsequently active loaded into these pre-formed liposomes. The process parameters and the effect of the formulation parameters were investigated and optimised. Critical process parameters and material considerations demonstrated to influence the liposomal product attributes included solvent selection and lipid concentration, flow rate ratio and temperature and duration used for drug loading. However, the total flow rate did not affect the liposome product characteristics, allowing high production speeds to be adopted. The final liposomal product comprised of 80 to 100 nm vesicles (PDI < 0.2) encapsulating ≥ 90% doxorubicin, with matching release profiles to the innovator product and is stable for at least 6 months. This demonstrates the ability to produce active-loaded PEGylated liposomes using microfluidics with comparable critical quality attributes to the originator Doxil®/Caelyx®.
Figure 1. Effect of the microfluidic process parameters on the physicochemical characteristics of HSPC:Chol:DSPE-PEG2000 and DSPC:Chol:DSPE-PEG2000 liposome formulations prepared in ethanol at 3:1:1 w/w. (A,D) The effect of flow rate ratio (FRR), (B,E) total flow rate (TFR), and (C,F) initial lipid concentration for HSPC:Chol:DSPE-PEG2000 (A,B,C) and DSPC:Chol:DSPE-PEG2000 (D,E,F) was measured. Unless varied, a FRR 1.5:1 and TFR 12 mL/min was used. Results represent the mean of at least 3 independent batches and the standard deviation is plotted as error bars.
Figure 2. Impact of the (A) solvent selection and (B) processing temperature on the physicochemical characteristics (particle size – bars; polydispersity – open circles; zeta potential - values) of HSPC:Chol:DSPE-PEG2000 (final concentration 4 mg/mL) prepared at FRR 1.5 and 3:1 and TFR 12 mL/min. Results represent the mean of at least 3 independent batches and the standard deviation is plotted as error bars. Significant differences are shown as *p <0.05.
Figure 3. Remote loading of doxorubicin (DOX) into liposomes. (A) Optimisation of the tangential flow filtration cycle for doxorubicin removal using free doxorubicin at 0.25, 0.5 and 2 mg/mL. (B) Effect of loading time and temperature using HSPC:Chol:DSPE-PEG2000 (10 mg/mL initial lipid concentration in ethanol, FRR 1.5:1 and TFR 12 mL/min) in ethanol with PBS (pH 7.4, 10 mM) or histidine-sucrose buffer (10 mM, 10%, pH 6.5) as external buffer. (C) Particle size (bars), polydispersity (open circles; PDI) and zeta potential (table) of HSPC:Chol:DSPE-PEG2000 liposomes (10 mg/mL initial lipid concentration in ethanol) at FRR 1.5:1 and TFR 12 mL/min using PBS or His-Suc as external buffer, before and after doxorubicin loading (from 10 to 30 min at 60°C). (D) Particle size and encapsulation efficiency (EE%) of the HSPC:Chol:DSPE-PEG2000 liposomes formulated with ethanol and [NH4]2SO4 250 mM at 1:1 to 3:1 FRR, final lipid concentration of 4 mg/mL, PBS external buffer and loaded during 10 min at 60°C. Results represent the mean of at least 3 independent batches and the standard deviation is plotted as error bars. Significant differences are shown as *p <0.05.
Figure 4. Effect of the solvent adopted using liposomal manufacturing on doxorubin active loading. Empty and doxorubicin-loaded HSPC:Chol:DSPE-PEG2000 (A,B,C) and DSPC:Chol:DSPE-PEG2000 (D,E,F) liposomes manufactured using microfluidics at FRR 1.5:1 and TFR 12 mL/min at an initial lipid concentration of 10 mg/mL. (A,D) Particle size (bars) and polydispersity (PDI; open circles), (B,E) zeta potential values and (C,F) encapsulation efficiency (EE%). Results represent the mean of at least 3 independent batches and the standard deviation is plotted as error bars. Significant differences, where not obvious, are shown as *p <0.05.
Figure 5. Manufacture and down-stream process of HSPC:Chol:DSPE-PEG2000 liposomes active-loaded with doxorubicin. (A) Physicochemical characterisation of HSPC:Chol:DSPE-PEG2000 FRR 1.5:1 TFR 12 mL/min at different processing stages: manufacture, purification, concentration, drug loading and filtration. (B) Drug recovery and (C) liposome and particle recovery after syringe filtration (filter 0.22 µm) of the doxorubicin-loaded HSPC:Chol:DSPE-PEG2000 formulation TFR 12 mL/min and FRR 1.5:1 (16 mg/mL final lipid concentration). A stability test of the doxorubicin-loaded liposomes at 4°C was carried out for 6 months and the physicochemical characteristics of the particles measured at selected time points: (D) particle size and polydispersity, (E) intensity plots of the particles and (F) zeta potential. (G) Doxorubicin loading as percentage of the initial amount loaded on day 0 of the experiment and day 180. Results represent the mean of at least 3 independent batches and the standard deviation is plotted as error bars.
Figure 6. Release profiles of doxorubicin-loaded liposomes. Different liposome formulations were incubated in (A) PBS media and (B) PBS + 1% human serum at 37°C. (C) Physicochemical characteristics of the liposomal formulations tested before and after the release study. Results represent the mean of at least 3 independent batches and the standard deviation is plotted as error bars.
Figure 7. Release studies using the USP-4 apparatus test assessing: (A) the lipid composition (HSPC:Chol:DSPE-PEG2000 vs SoyPC:Chol:DSPE-PEG2000 vs HSPC:LowChol:DSPE-PEG2000) (B) the influence of the FRR (1:1 vs 1.5:1) (C) and the incubation temperature (40°C vs 60°C) on the production of DOX-loaded PEGylated liposomes. (D) Physicochemical characteristics of the liposomal formulations tested before the release study. Results represent the mean of at least 3 independent batches and the standard deviation is plotted as error bars.
Figure 8. Physicochemical characteristics of the HSPC:Chol:DSPE-PEG2000 (3:1:1 w/w) formulation manufactured using Precision NanoSystems microfluidic platform (Bench scale and pre-clinical scale: 12, 60 and 90 mL/min) at FRR 1.5:1. (A) Particle size (bars) and polydispersity (open circles), (B) zeta potential (ZP), (C) encapsulation efficiency (EE%) and (D) CryoTEM images of the empty formulation prepared using the bench scale and scale-up production technology (scale bar 200 nm). Significant differences are shown as *p <0.05.
Figure 9. Drug release profiles of doxorubicin-loaded HSPC:Chol:DSPE-PEG2000 (1.5:1 FRR) using the bench scale (12 mL/min) and scale-up platform (12, 60 and 90 mL/min). Drug release was assessed using dialysis and ammonium sulfate pH 6.6 as release inducer at 45°C. Results represent the mean of 3 batches and the standard deviation is plotted as error bars.
Figure 10. Schematic representation of the continuous manufacturing pilot plan. Empty PEGylated liposomes with a pH gradient were successfully manufactured in a continuous manner using the scale-up microfluidic system attached to a three-filter tangential flow filtration system. Table shows the vesicle size, polydispersity, zeta potential and residual solvent content at each stage of the process. Further studies will focus on developing a continuous process for the active-loading of drugs.
Figure 11. Just-in-time personalised medicine. (A) Particle size, polydispersity and zeta potential of 50 mL batch of HSPC:Chol:DSPE-PEG2000 liposomes using the scale-up microfluidic system (1.5:1 FRR, 12 mL/min).. Particle size, polydispersity, zeta potential (tables) and EE% (graphs) of the aliquots loaded with doxorubicin (DOX), vincristine (VIN) and acridine orange (AO) respectively.