2.2. Dialysis Membrane (DM) Method

The DM method is a widely used and versatile method for testing in vitro drug release of particulate formulations including microspheres [42], nanoparticles [43], and liposomes [44,45]. This method utilizes an appropriate dialysis membrane with a specific molecular-weight cut-off (MWCO) to physically separate the released drug molecules from microparticles by allowing the drug to pass easily through the membrane into the release medium. Drug release is usually assessed with samples taken from the external solution outside the dialysis membrane over time. Compared to the SS method, this method eliminates the need to separate the released compounds from microparticles, making sampling relatively easy and eliminating unwanted microparticle loss during sample preparation and handling [17]. However, slow equilibration with the outer medium limits an accurate measurement of initial drug levels [46]. In addition, the disadvantages of the DM method include the difficulty of achieving adequate agitation to prevent microparticle aggregation within the dialysis bag, the inability to use the drugs that bind to the polymer or the dialysis membrane, and the violation of sink conditions within the dialysis bag [15,16,17].

The DM method has been performed in a clipped bag of the dialysis tubing or in various types of dialyzers [42,47]. Dialysis tubing is an economical tool that is clear, flexible and durable, but this tubing has concerns about handling, closing and sample recovery [16]. Dialyzers are designed for a specific sample volume and are convenient and easy to use [48]. The most common used dialyzers are Float-A-Lyzer (Spectrum Laboratories, Rancho Dominquez, CA, USA), Slide-A-Lyzer (Thermo Scientific, Rockford, IL, USA), Pur-A-lyzer (Sigma-Aldrich, St. Louis, USA), D-Tube (Merck-Millipore, Billerica, MA, USA), and GeBA-flex dialysis tube (Gene Bio-Application Ltd., Kfar Hanagide, Israel).

Based on experimental setting, the DM method is classified into regular dialysis, reverse dialysis, and side-by-side dialysis method (Figure 2). In the regular dialysis, the microparticles enter the inside of a sealed dialysis tubing, and drugs released from the microparticles diffuse from the inner medium to the to the outer medium through the dialysis membrane. The diffusion of the drug through the dialysis membrane into the outer medium can be affected by stirring the contents of the container, thus minimizing the effect of the unagitated water layer [16]. Commonly used modes of agitation include a shaker [49,50], magnetic stirrer [51,52], and the United States Pharmacopeia (USP) paddle apparatus under agitation [53]. Unlike regular dialysis, reverse dialysis is a method where the microparticles are placed outside the dialysis tubing and sampling is performed inside the dialysis tubing containing only the medium [54,55]. Sampling is performed by opening the dialysis tubing and removing a certain amount of medium, or by removing the entire dialysis tubing and replacing it with a new one. The main advantage of the reverse dialysis method is that it can avoid the violation of sink condition that occurs in regular dialysis method. In regular dialysis, when the volume inside the dialysis tubing is low and the membrane surface area is small, the rapid drug diffusion of the drug into the bulk release medium of the outside container is not sufficient, resulting in a sink condition violation [15]. In the reverse dialysis method, the drug released from microparticles placed in the external release medium can easily diffuse into the dialysis tubing [56]. The third method is side-by-side dialysis, in which the donor and acceptor cells have the same volume capacity and are separated by a dialysis membrane. The drug released from the microparticles is evaluated by placing the microparticles on the donor cells and performing sampling on the receptor cells [54].

Dialysis membrane methods for in vitro drug release test of particulate formulations: regular dialysis (A), reverse dialysis (B), and side-by-side dialysis (C).

In addition to the experimental setup and agitation conditions, the MWCO of the dialysis membrane and the ratio between the internal and external release medium volumes are the main parameters for the successful DM method [15,46]. In particular, the selection of an appropriate MWCO is important for the dialysis membrane. The basic premise of the DM method is that the drug released from the particles diffuses through a semipermeable membrane with appropriately sized pores. Dialysis membranes with sufficiently high MWCO are used for in vitro studies to ensure they are not a limiting factor for drug diffusion [24]. Nonetheless, the selection of MWCO is somewhat subjective because the criteria are not clear. For example, MWCO of 3.5–5 kDa for loperamide [57], 8–14 kDa for risperidone [51], cyclic somatostatin [58] and cefquinome [59], and 100 kDa for beta-sheet peptide [60] have been used. The volume contained in the dialysis bag is much smaller than the external medium. To facilitate drug diffusion, the volume of the inner medium is kept 5–10 times less than the volume of the outer medium, providing the driving force to deliver the drug to the outside and maintaining sink conditions [16]. For example, inner medium volumes reported in the literature range from 1 to 10 mL, while external medium volumes are typically much larger, around 40 to 90 mL [61].

Qu et al. prepared cefquinome-loaded PLGA microspheres for lung targeting and evaluated in vitro cefquinome release by regular dialysis method using a dialysis bag (MWCO 8–14 kDa) [59]. The sealed dialysis bag containing the microspheres was immersed in PBS while stirring in a shaking water bath set at 100 rpm. For sampling, the medium was withdrawn to a volume of 2 mL and replaced with an equal volume of fresh release medium. The microspheres showed initial burst for 1 h and constant release for 36 h.

Chaurasia et al. prepared parenteral risperidone-loaded microspheres with different drug-to-polymer ratios (1:1.5, 1:1.75, and 1:2) and evaluated in vitro release profile by regular dialysis method using a dialysis bag (MWCO 12–14 kDa) [51]. Microspheres were immersed in 50 mL of PBS with continuous magnetic stirring at 100 rpm. For sampling, samples were taken from the dialysis bag at each time point and replaced with fresh medium to maintain sink condition. Microspheres with low porosity showed higher burst release followed by a longer lag phase and reached release plateau after 14 days, whereas microspheres with a highly porous structure showed a lower burst release followed by a shorter lag phase and reached release plateau phase within 14 days.

Zhang et al. prepared paclitaxel-loaded PLGA microspheres by the double-emulsion solvent evaporation method and evaluated in vitro release profile by regular dialysis method using a dialysis bag (MWCO 14 kDa) [62]. For the release study, the microspheres were suspended in 5 mL sodium salicylate/PBS medium in a dialysis membrane, and the dialysis bag was immersed separately in a screw cap tube containing 50 mL of the sodium salicylate/PBS medium. The tubes were placed horizontally in an orbital shaker. For sampling, 2 mL of solution in the tube was collected and the tube was supplemented with 2 mL fresh medium. When comparing smooth microspheres with internal sporadic porosity and rough microspheres with highly porous internal structure, the smooth microspheres exhibited roughly a slow linear release pattern, whereas the rough microspheres showed a faster S-curve release pattern.

Chen et al. investigated the release profile of various size fractions (5, 32, 70 and 130 μm) of gefitinib-loaded microspheres, which were prepared using the oil-in-water solvent evaporation method and then fractionated by wet sieving [42]. Prior to in vitro release study, this study embedded microspheres in methacrylated dextran hydrogels to prevent aggregation of microspheres during the incubation conditions without limiting the release of gefitinib from formulations. For in vitro release study, gefitinib microspheres-loaded methacrylated dextran hydrogels were casted in Slide-A-Lyzer MINI Dialysis Devices (MWCO 2 kDa) and incubation was performed in PBS containing 1% Tween 80 under constant shaking. The size-fractionated microspheres showed significant differences in drug release between small microspheres and larger microspheres. Microspheres smaller than 50 μm showed rapid diffusion-based release that reached completion within one week. However, the larger microspheres showed a pattern of sigmoid release that lasted for three months, where diffusion (early stage) and erosion (late stage) dominated drug release.

Zhang et al. studied drug release behavior from fenretinide-loaded PLGA microspheres by incorporating nonionic surfactants (Brij 35, Brij 98, Tween 20, and Pluronic F127) [63]. The in vitro release test was performed on a mesh bag (nylon material) with 1 μm pore size instead of a dialysis bag due to the interaction of the drug and the dialysis bag. The release medium was PBS containing 0.1% Tween 20. The samples were continuously agitated at a constant speed and the release medium was replaced periodically as needed to maintain the sink condition. At predetermined time points, each mesh bag was taken out, lyophilized, and analyzed for drug remaining in the microspheres by HPLC. Microspheres prepared with Brij 98 exhibited reduced initial burst and sustained release over 28 days. The release profile was dependent on the concentration of Brij 98 with a very significant increase in the release rate, especially when the surfactant level was increased from 10% to 20% w/w.