Strategies to Formulate Lipid-based Drug Delivery Systems

Orally administered water-insoluble drugs have become increasingly important in therapy, and lipid-based drug delivery systems have become an essential tool in the development of formulations for these compounds. Lipid-based formulations have been marketed for a variety of drug classes such as HIV protease inhibitors (e.g., ritonavir, lopinavir, sequinavir, tipranavir, amprenavir), immunosuppressants (cyclosporine, sirolimus), and calcium regulators (e.g., calcitriol, paricalcitol). It is estimated that 40% of new drug candidates are water insoluble, and thus may require delivery in a system such as a lipid-based formulation.

A lipid-based drug delivery system typically is composed of lipids and surfactants, and may also contain a hydrophilic co-solvent. Many of them are characterized as Self-Emulsifying Drug Delivery Systems (SEDDS) such that they form an emulsion upon gentle agitation in water. Emulsions are considered metastable systems, with droplet sizes of 100-1000nm. Other formulations, Self-Microemulsifying Drug Delivery Systems (SMEDDS), form microemulsions that are thermodynamically stable systems. Microemulsions have droplet diameters <100 nm, and are visually transparent or translucent. The drug is generally present in the dosage form dissolved in the formulation, and should remain solubilized after dispersion of the dosage form in the GI tract. Absorption by the intestinal mucosal cells is facilitated by the rapid release of drug from the high surface area of the small emulsion or microemulsion droplets.

An ideal oral lipid-based dosage form must meet a number of demands:

1. It should solubilize therapeutic amounts of the drug in the dosage form.
2. It should maintain adequate drug solubility over the entire shelf-life of the drug product (generally 2 years) under all anticipated storage conditions.
3. It should provide adequate chemical and physical stability for the drug and formulation components.
4. It must be composed of approved excipients in safe amounts.
5. Once ingested, it should facilitate dispersion of the dosage form in the intestinal milieu and maintain drug solubilization in the dispersed form.
6. It should adapt to the digestive processes of the GI tract such that digestion either enhances or maintains drug solubilization.
7. It should present the drug to the intestinal mucosal cells such that absorption into the cells and into the systemic circulation is optimized.

These requirements can render formulation and evaluation of lipidbased systems quite challenging. Despite their importance, design of lipid-based formulations to meet these challenges is still largely an empirical exercise. This review will outline some guidelines and strategies in developing these formulations, addressing the challenges of each of these requirements and highlighting examples of successful lipid-based formulations.

Solubilization

Table 1    -    Excipients used in oral lipid-based formulations

Zoom In

Table 1 summarizes the excipients available to the formulator to solubilize a drug candidate in a lipid-based formulation. While the lipids (fatty acid derivatives) are the core ingredient of the formulation, one or more surfactants, as well as perhaps a hydrophilic co-solvent, may be required to aid solubilization and to improve dispersion properties. Surfactants are categorized by their Hydrophilic-Lipophilic Balance (HLB) number, with a low value (≤10) corresponding to greater lipophilicity and a higher value (≥10) corresponding to higher hydrophilicity. As a guideline as a starting point for formulation design, most of the lipids used in these oral formulations have a known “required HLB” value (generally available from the vendors), which corresponds to the optimal HLB for the surfactant blend necessary to emulsify the oil in water. For example, the required HLB of medium chain triglyerides (glyceryl tricaprylate/caprate) is ~11, while that of long chain triglycerides (vegetable oil) is ~6. The surfactants polysorbate 80 and sorbitan monooleate have HLB values of 15 and 4.3, respectively. Thus a blend of 63/37 blend of polysorbate 80/sorbitan monoleate will emulsify the former, and a 16/84 blend will emulsify the latter. Co-solvents used in lipid-based formulations include propylene glycol, ethanol, PEG400, glycerol, and diethylene glycol monoethyl ether.

Possible combinations for lipid-based formulations thus approach an endless number, but a consideration of the structure and physical-chemical properties of the drug candidate may allow one to narrow the search. For example, if the drug is an amine, it may be soluble in oleic acid by formation of an ion pair, as exhibited by marketed formulations of ritonavir and ritonavir/lopinavir. Hydrophobic non-ionizable drugs (generally characterized by a Log P_{octanol/water}≥3) may be solubilized by long chain or medium chain triglycerides and/or by combination of a lipid with a low HLB surfactant such as phosphatidylcholine/medium chain triglyceride or oleoyl macrogolglycerides. Less hydrophobic drugs (viz., Log P_{octanol/water} ≤3) may be solubilized by monoglycerides or propylene glycol monoesters, or by combinations of these lipids with high HLB surfactants or hydrophilic co-solvents. Often a combination of low HLB and high HLB surfactants give superior solubilization, which may also optimize the dispersion properties described later. Screening studies can be carried by mixing the required drug amount in the formulation and examining visually or microscopically for the presence of drug crystals, followed by more accurate determination of solubility values of drug in the most promising formulations by HPLC. High throughput screening systems have been employed to increase efficiency. For example, a robotic liquid dispenser was used to prepare a series of nilvadipine SMEDDS formulations by combinations of oil, surfactant, and ethanol, and identify the optimal low HLB/high HLB blend. Statistical and experimental design studies, such as a simplex lattice mixture or a central composite design, have also been used to optimize and develop SMEDDS formulations of celecoxib and bufalin, respectively. Nevertheless, as pointed out in the review by Rane and Anderson, prediction of drug solubility in lipid vehicles is difficult due to the dominant role played by interfacial effects and the possible presence of complex microstructures. While it is difficult to predict optimal formulations based solely on physicochemical properties of the candidate, some progress has been made. Thi et al. examined 10 different compounds with varying properties in SMEDDS formulations; optimal drug logP_o/w for solubility in the SMEDDS formulation was found to be between 2 and 4.

Frequently, toxicology studies will have been carried out with waterinsoluble drug candidates in lipid vehicles. Constraints may be different for the two situations; safety requirements of excipients will be more stringent for clinical formulations, but doses will likely be lower. Nevertheless, toxicology vehicles for a given drug may provide a starting point for development of a First-in-Human and subsequent formulations. 

Dispersion

Zoom In

Figure 1 - Ternary phase diagram of an oil-surfactant water system, based on a C12E10-oleic acid-water system.

Formulations that exhibit sufficient solubility of the drug candidate should be examined for emulsification and dispersion properties in aqueous vehicles. A preliminary screen can be carried out by microscopic observation of the formulation when mixed with water. Vigorous mixing, accompanied by diffusion and stranding mechanisms, occurring at the water/formulation interface is indicative of an efficient emulsification. Absence of drug precipitate after complete mixing of the formulation with aqueous medium is another requirement. Particle size measurement of emulsion droplets by laser light scattering or other techniques is useful to select promising formulations. Construction of ternary phase diagrams is a method frequently used to determine the types of structures resulting from emulsification and to characterize behavior of a formulation along a dilution path. An example is shown in Figure 1; the line from A to B represents dilution of a formulation consisting initially of 35% surfactant/65% oil, passing through regions of a water-in-oil microemulsion and a lamellar liquid crystal until reaching a stable bicontinuous oil-in-water microemulsion after dilution. It is often unnecessary to construct the entire phase diagram, but an understanding of the structures arising on a dilution path of a given formulation is important to assure formation of stable dispersed structures upon dilution. Appropriate combinations of low HLB and high HLB surfactants frequently lead to smaller emulsion droplet size than single surfactants. These more complex combinations can be examined by pseudo-ternary phase diagrams, wherein a given surfactant blend and/or oil blend would serve as the oil or surfactant apex of the diagram; co-solvents can be examined similarly as part of the oil or surfactant blend.

Dispersion properties can be examined and optimized as part of an experimental design study. For example, genistein SEDDS formulations were optimized by a three-factor, three-level Box-Behnken design. Droplet size after dilution, turbidity, and dissolution percentage of drug after 5 minutes and 30 minutes were the variables examined to optimize a formulation consisting of glyceryl monolinoleate, medium chain triglyceride, polyoxyl 35 castor oil, caprylocaproyl macrogolglyceride, and diethylene glycol monoethyl ether. Similarly, a D-optimal mixture design was used to select compositions of SMEDDS formulations of albendazole, with dispersion performance, droplet size, dissolution efficiency, and time for 85% drug release as the factors examined. The optimal formulation was polyoxyl 35 castor oil, polysorbate 80, acidified PEG-400, and propylene glycol monocaprylate. The same type of design was used to optimize a SEDDS formulation of lacidipine, with droplet size, optical clarity, drug release, and emulsification efficiency as the factors examined. The optimized formulation was composed of oleoyl macrogolglycerides, glycerol monocaprylate/caprate, polysorbate 80, and polyoxyl 35 castor oil.

Ideally, drug solubility and dispersion should be examined together. While presence of a co-solvent may enhance solubility of the drug in the dosage form, high amounts of hydrophilic co-solvent may lead to drug precipitation after dispersion due to diffusion of the co-solvent away from the emulsion droplet into the bulk aqueous phase. Mohsin et al. observed rapid precipitation of fenofibrate after dilution of formulations composed of high amounts of hydrophilic co-solvents (e.g., propylene glycol) and surfactants (e.g., polysorbates). If sufficient amounts of lipid (medium chain triglyceride) were present, formation of supersaturated systems sometimes resulted which nevertheless retarded precipitation for adequate time. Experimental design studies examining both drug solubility and dispersion properties can aid in identifying formulations with optimal behavior with respect to both factors. For example, a simplex lattice experimental design approach was used to optimize SEDDS formulations of curcumin based on ethyl oleate, PEG400, and polyoxyl 35 castor oil.

Digestion

The actions of intestinal lipases can have a profound effect on the behavior of lipid-based formulations in the GI tract, and must be considered in their design. It has long been recognized that non-dispersible but digestible lipids such as triglycerides can be metabolized by lipases to mono-/di-glycerides and fatty acids which will emulsify any remaining oil. Thus, the presence of high amounts of surfactants may be unnecessary to assure creation of the requisite small particle sizes and large surface areas for drug release. In 2000, Pouton proposed a classification system for lipid-based formulations based on the formulation components and the dependence on digestion to facilitate dispersion; this is shown in Table 2. Type I formulations, being composed simply of drug in triglyceride or mixed glycerides, require digestion in order to be dispersed. Dispersibility of the other classes of formulations is less dependent on digestion since low HLB (Type II) or high HLB (Type III) surfactants are included. A fourth category Type IV, was added in 2006, consisting of surfactant/cosolvent mixtures without lipid. However, the lower oil content of Type IIIA and especially of Type IIIB and Type IV formulations increases the risk of drug precipitation upon mixing with the intestinal milieu, due to diffusion of the co-solvents and high HLB surfactant away from the emulsion droplet. Furthermore, surfactants in the formulation may also be digested by intestinal lipases and hence lose their solubilizing power, leading to drug precipitation. Thus, in vitro dissolution testing in biorelevant media (i.e., simulated intestinal fluids) such as those developed by Dressman should be used to evaluate candidate formulations. Studies by Cuine et al. with a series of danazol lipid-based SMEDDS formulations showed that higher surfactant content led to smaller droplet size after in vitro dispersion, but under digestion conditions high surfactant content led to a greater occurrence of drug precipitation and lower bioavailability in dogs. More comprehensive evaluations can be carried out by in vitro lipolysis models as described by Mullertz. For example, in studies by Dahan, formulations of long chain triglycerides (LCT), medium chain triglycerides (MCT), short chain triglycerides (SCT), and aqueous suspensions of griseofulvin were examined in the in vitro lipolysis model and in rat bioavailability studies. Both the in vitro model and in vivo studies showed a rank ordering of drug solubilization of MCT > LCT > SCT > aqueous suspension, with a correlation r² >0.98.

Table 2    -    Pouton’s Classification of Lipid-Based Delivery Systems

Zoom In

Absorption

Zoom In

Figure 2 - Processes of dispersion, digestion, and absorption occurring for a lipid-based formulation.

Efficient absorption of the drug by the intestinal mucosal cells is of course the ultimate goal of any oral lipid-based formulation. Figure 2 shows the processes that occur in the intestinal milieu for a lipid-based drug formulation. First the components are dispersed to form lipid droplets (for Type I formulations) or emulsion droplets (for Types II-III), followed by lipolysis and solubilization of the digestion products by bile acids, forming colloidal mixed micelles. It is believed that drug then partitions from the emulsion oil droplets and bile salt mixed micelles to be absorbed by the mucosal cells of the intestinal wall. Even if only a very small proportion of drug is actually in solution in the intermicellar space available for absorption, the colloidal species quickly replenish the soluble portion to be absorbed. Testing in animals of candidate formulations is generally used pre-clinically to evaluate bioavailability of drugs in lipid-based systems. While rats and dogs are the species most widely employed, the pig has more recently been suggested as being more relevant and comparable to humans in eating behavior and in GI physiology. Alternatively, absorption can be monitored by isolated intestinal membranes. For example, absorption of ibuprofen from a medium chain triglyceride/diglyceryl monooleate/polyoxyl 40 hydrogenated castor oil microemulsion was evaluated in rat isolated intestinal membrane in an Ussing chamber. deally, absorption should be optimized along with solubilization and dispersion under biorelevant conditions. This is similar to the approach of Liu et al., who used a central composite design to optimize oil/surfactant/co-surfactant ratio in a SMEDDS formulation of oridonin, examining the factors solubility, droplet size after dispersion, and in situ intestine absorption rate.

Ultimately, the success of any lipid-based formulation will be determined in humans. Clinical results of experimental and marketed lipid-based oral formulations of poorly soluble drugs has been reviewed by Fatouros et al. One of the first compounds delivered by a SEDDS formulation was cyclosporine. The Sandimmune formulation, introduced in 1981, contains corn oil, ethanol, glycerol, and linoleoyl macrogolglycerides; it forms a crude emulsion upon dispersion in water. Neoral is a SMEDDS cyclosporine formulation introduced in 1994; it contains corn oil mono-di glycerides, polyoxyl 40 hydrogenated castor oil, ethanol, propylene glycol, and α-tocopherol. In clinical studies with transplant patients, Neoral gave a higher AUC (3028 vs. 2432 μg • h/L), a higher C_max (892 vs. 528 μg/L), and a shorter t_max (1.2 vs. 2.6 h) relative to Sandimmune. Several oral HIV protease inhibitors have also been formulated in lipid systems. Ritonavir was originally marketed in 1996 as a hard gelatin capsule with a semi-solid, lipid-based formulation composed of caprylic/capric triglycerides, polyoxyl 35 castor oil, citric acid, ethanol, polyglycolized glycerides, polysorbate 80, and propylene glycol. In 1998, after a less-soluble polymorph of the drug appeared, the product was re-formulated as a soft gelatin capsule with a liquid formulation containing ethanol, oleic acid, and polyoxyl 35 castor oil. A second generation product, ritonavir/lopinavir has been marketed in a oleic acid/polyoxyl 35 castor oil/propylene glycol. Another protease inhibitor, tipranavir, is marketed in a mono-/di-glycerides of caprylic/capric acids/polyoxyl 35 castor oil/propylene glycol/ethanol formulation. Another interesting example is the HIV protease inhibitor sequanavir: the lipid-based formulation (Fortovase), which contains medium chain mono-diglycerides, povidone, and α-tocopherol, gave a 3-fold higher oral bioavailability compared to a conventional capsule formulation of the drug’s mesylate salt (Invirase), even though the salt has a 1000-fold higher aqueous solubility than the free base used in Fortovase (2.2.mg/mL vs. 0.017 mg/mL).

Stability

Maintaining adequate chemical and physical stability of lipid-based drug formulations delivery systems can also present challenges. As already reviewed, unsaturated lipid components can undergo lipid peroxidation. This can be minimized by use of saturated medium chain (C6-C12) triglycerides and by use of appropriate anti-oxidants. Phenol-based anti-oxidants such as Vitamin E (α-tocopherol), butylated hydroxytoluene (BHT), butylated hydroxyanisole (BHA), and propyl gallate can act synergistically with oxygen scavengers such as ascorbic acid and its lipid-soluble counterpart, ascorbyl palmitate. With regard to physical stability, liquid or semi-solid fill formulations in hard or soft gelatin capsules must be carefully designed in order to ensure compatibility of the fill with the capsule shell.

Conclusions

While it is apparent that lipid-based formulations will continue to be an important tool to formulate poorly soluble drugs, design of these formulations can be a challenge. In their excellent review, Porter et al. recently outlined seven guidelines for design of lipid-based formulations, as summarized below:

1. It is critical to maintain drug solubility in the formulation, after dispersion, and after digestion.
2. Properties of the colloidal species formed after processing in the GI milieu are probably more important than properties of the formulation itself in enhancing absorption.
3. Higher proportions of lipid (>60%) and lower proportions of surfactant (<30%) and cosolvent (<10%) generally lead to more robust drug solubilization after dilution.
4. Medium chain triglycerides may afford greater drug solubility and stability in the formulation, but long chain triglycerides facilitate more efficient formation of bile saltlipid colloidal species and thus may afford higher bioavailability.
5. Type IIIB SMEDDS formulations give lower droplet sizes after dispersion. However, they are more dependent on the surfactant properties employed, and non-digestible surfactants generally give higher bioavailability.
6. Dispersion of Type IV formulations (surfactant/cosolvent) are likely more efficient if two surfactants are used rather than a single one.
7. Type IV formulations may give higher drug solubility, but must be designed carefully to assure that drug does not precipitate after dispersion.

These guidelines are important ones to keep in mind when designing oral lipid-based formulations for poorly soluble drugs. As further experience is gained with design and use of these formulations and the database of successful formulations grows, it is to be hoped that design of these formulations will become less of an empirical exercise and more rational in its approach. As this happens, the utility of lipid-based formulations can only grow.  Source: NDDS Tech Review, APDR, Drug Delivery system.

by Akshaya Srikanth, Pharm.D Intern