Improving liposome integrity and easing bottlenecks to production

The importance of liposomes as an effective drug delivery system is well accepted in the pharmaceutical industry, but their handling remains a challenge. The delicate nature of liposomes makes their separation/concentration/diafiltration particularly problematic. Even minor shear and heat can disintegrate these nanometer-sized phosphilipid entities. Principally, tangential flow filtration (TFF) is a suitable option for liposome handing (concentration and buffer exchange); however, optimization is painstaking because of the fragile nature of liposomes. We studied different methods of TFF to better understand which factors could be modified to improve liposome handling. Our investigation showed that lowering transmembrane pressure, delta pressure and process temperature, as well as performing concentration followed by direct diafiltration helps maintain liposome integrity and minimize cost.

Liposomes are minute spherical oil droplets separated from an aqueous medium by a phospholipid bilayer. The small size of liposomes enables them to pass through epidermis and act as a carrier for liquid substances, (Figure 1) which are enclosed within the liposome’s hollow internal cavity.

In today’s pharmaceutical industry, liposomes are considered to be one of the best drug delivery vehicles available because of their unique properties; a liposome can carry hydrophilic solutes inside its cavity, as well as hydrophobic chemicals in its surrounding membrane. Meanwhile, its lipid bilayer can fuse with other bilayers, such as the cell membrane, to effectively deliver an enclosed drug substance. Because of their unique nature, liposomes can deliver a considerable variety of therapeutic agents that cannot diffuse through the membrane normally, such as a solution of DNA, hydrophilic drugs, hydrophobic chemicals, drugs requiring a slow diffusion rate and cytotoxic drugs.

Liposomes are created by continued high‑shear sonication whilst a recently introduced method — extrusion — is specifically used to create human‑use liposomes.

Although liposomes may be the ideal drug delivery vehicles for many therapeutic agents, their delicate nature and integrity have created several bottlenecks in their production.

An introduction to TFF
Filtration is an effective technique that can be used to separate and study the properties of liposomes. Broadly, there are two types of filtration: normal flow filtration (NFF) and tangential flow filtration (TFF). TFF is a modified filtration technique that was developed to overcome the limitations of NFF.

In NFF, pressure acts on the perpendicular direction to the membrane filter along with product flow, resulting in irrecoverable (during process) plugging. In TFF, pressure acts both perpendicular and parallel (along with product flow) to the membrane filter; perpendicular pressure (TMP) facilitates 'filtration', whereas parallel pressure (Δ) protects the membrane from plugging (Figure 2).

First step: liposome diafiltration
We conducted a study to buffer exchange liposomes from one buffer (3.3% ammonium sulfate/9% sucrose solution) to another (9% sucrose solution) using a TFF cassette. The study was divided into two parts: in the first, direct diafiltration was evaluated; in the second, the liposomes were concentrated and then diafiltered to minimize the process time (because lower diafiltration volumes were required). For both parts, the integrity of liposomes was measured after the process, as was diafiltration efficiency (conductivity measurement), process time, membrane area requirement and recovery. The experimental set‑up is explained in Figure 3.

In the first part of the study, the liposome solution was directly diafiltered, after which three main parameters were observed and analysed: 1) recovery of liposome in retentate (size measurement/physical integrity checking); 2) gradual decrease of salt (ammonium sulfate) concentration permeate or diafiltration efficiency (conductivity measurement); 3) process flux.

A Millipore Pellicon 2 Mini System was installed (Figure 3) with a Pellicon mini 0.1 m² cassette (300 kD membrane) and the system was flushed with 8.2 L of water for injection (WFI). Cassette integrity was measured using the manual air diffusion method and was found to be normal. Normalized water permeability (NWP = flux of water per unit pressure) was measured to evaluate whether the cassette had been cleaned properly. Pre‑use NWP was 62.76 LMH/psi (litres/m²/hour psi), which was within the acceptable limit of 50 LMH/psi. After flushing the system/cassette, a diafiltration study was conducted on the liposome solution. The temperature was kept at ∼9 ºC to maintain liposome integrity.

We then performed step diafiltration. Briefly, 400 mL of 9% sucrose solution (no conductivity) was added to 400 mL of liposome in 3.3% ammonium sulfate/9% sucrose solution (conductivity ∼20 mS/cm). By maintaining a TMP of 0.2 bar and ΔP of 0.4 bar, ∼400 mL of permeate was passed though the membrane and the retentate volume was brought back to 400 mL. TMP and ΔP were kept intentionally low, anticipating the possible physical stress on the liposomes. The above step was repeated seven times until conductivity of ∼20 mS/cm fell to ∼0.1 mS/cm. Details schematically are presented in Figure 4.

In the previous step, 400 mL of 9% sucrose was added to 400 mL of liposome solution. It was brought back to 300 mL after removing 500 mL in permeate. Throughout the process, TMP was maintained at 0.2 bar. The retentate was subjected to 100 mL (50 mLx2) of 9% sucrose wash to make the final volume of 400 mL. The permeate was checked for conductivity during each step.

Process parameters observed in direct diafiltration of liposome are presented in Table 1.

Change of flux versus is presented in Figure 5.

The permeate was clear without any liposome passage and no significant change in liposome size was observed throughout the TFF process. There was no flux drop with the maintenance of constant TMP throughout the process; average flux was 32.6 LMH at 0.2 bar TMP. Initial conductivity was 20 mS/cm, which gradually came down to 0.09 mS/cm (as observed in the permeate of the eighth diafiltration volume). Mass balance and brief results are presented in Table 2.

After the process, 5 L of WFI was recirculated for 5 min and then drained off — recirculation was performed twice. The system was sanitized with 0.3N NaOH solution, which was recirculated for 10 min and then drained off. The cassette was again washed with 5 L of WFI. Post-use NWP (58.70 LMH/psi) was measured to check cleaning efficiency, which was acceptable.

Second step: intrinsic concentration followed by diafiltration
The main objective of the second part of the study was to reduce the number of diafiltration steps (by concentrating the solution first), thus reducing process time. To achieve this, the liposome solution was concentrated and then diafiltered. This was followed by observation and analysis of the three key parameters observed in the previous study.

A Millipore Pellicon Mini System, which was installed with a Pellicon mini 0.1 m² cassette (300 kD membrane), was used. Installation and equilibration conditions were as previously explained. Pre-use NWP was 58.70 LMH/psi, which was within the acceptable limit. After flushing the system/cassette, a concentration and diafiltration study was conducted on liposomes in 3.3% ammonium sulfate/9% sucrose solution. The temperature was kept at ∼10 °C to maintain liposome integrity.

Step diafiltration was then performed. Briefly, 400 mL of 9% sucrose solution was added with 400 mL of liposome in 3.3% ammonium sulfate/9% sucrose solution. By maintaining a TMP of 0.25 bar, ∼600 mL of permeate was passed through the membrane, and the retentate volume was brought back to 200 mL. Again, 600 mL of 9% sucrose solution was added to increase the volume to 800 mL. Details schematically are presented in Figure 6.

In the last step, 300 mL of 9% sucrose was added with 200 mL of liposome solution; this was reduced to 175 mL after removing 325 mL in the permeate. Throughout the process, TMP was maintained at 0.25 bar. The retentate was subjected to 225 mL of 9% sucrose wash to give a final volume of 400 mL. In each step, the permeate was checked for conductivity. Process parameters observed in direct diafiltration of liposome are presented in Table 3. Change of flux versus time is presented in Figure 7. Permeate was clear without any liposome passage.

The permeate was clear without any liposome passage, and no significant change in liposome size was observed throughout the TFF process. There was a flux drop with the maintenance of constant TMP throughout the process; average flux was 18.6 LMH at 0.25 bar/g TMP. Initial permeate conductivity was 21.9 mS/cm, which gradually came down to 0.1 mS/cm (as observed in the permeate of the fifth diafiltration volume). Mass balance and brief results are presented in Table 4 and Table 5.

What did we find out?
From the comparative evaluation (Figure 8), we concluded that, for better handling of liposomes during diafiltration by TFF, the following factors should be considered:

• Lower TMP and ΔP help to maintain liposome integrity.
• Less process temperature (preferably 8–10 °C) is favourable for maintaining the integrity of the liposomes.
• Direct diafiltration could be a better option than concentration and diafiltration, if exchange buffer/media is not expensive (flux decreases sharply with the inclusion of a concentration step). Moreover, liposome recovery is better in direct diafiltration rather than concentration and diafiltration.
• Concentration followed by diafiltration may be considered, if exotic buffer/media is used to minimize cost.

References
1. V.P. Torchilin, Adv. Drug Deliv Rev., 58(14), 1532–55 (2006).Protein Concentration and Diafiltration by Tangential Flow Filtration — An Overview, Millipore Technical Publications, Lit. No. TB032.
2. R.P. Lenk et al., Method for size separation of particles, US Patent 5948441 (1999).
3. H. Lautenschläger, “Liposomes,” in A.O. Barel, M. Paye and H.I. Maibach, Eds,