Daniel Joseph Price, Ph.D., Strategic Marketing Manager for the SAFC^® portfolio of solid formulation excipients at MilliporeSigma, explains how the Developability Classification System is enabling the development of better excipients for poorly soluble compounds and simplifying the formulation process for these challenging APIs, in conversation with Pharma’s Almanac Editor in Chief David Alvaro, Ph.D. 

David Alvaro (DA): How have solubility issues with small molecule APIs evolved, what approaches are currently being pursued, and where do they face limitations?

Daniel Joseph Price (DJP): In order to have a therapeutic effect in the arena of oral drug delivery — predominantly tablets, but also capsules and suspensions — the drug substance must be soluble in the gastrointestinal (GI) tract so that it can permeate through the intestinal cell membranes and pass into the systemic circulation.

Back in the 1970s and 1980s, solubility was less of an issue, as most molecules with therapeutic efficacy also had reasonable physicochemical properties, including solubility. They could be put into tablets without any complex modification or formulation. Candidate drug molecules have become more challenging since then, with more complex structures and chemistries and, often, increased lipophilicity. As a result, they have become less and less soluble, to the point where modification — either to the molecule or formulation — is necessary to address the low solubility.

Several different approaches have been pursued. Medicinal chemists can modify the chemical structure of the API. They can change existing functional groups or add new ones to make the molecule more polar and thus more soluble. Another option, if the candidate molecule has an acid or base substituent, is to generate a salt form with higher solubility. A third possibility is to generate a prodrug, which typically involves converting a carbonyl group in the molecule into a very polar ester substituent to increase the polarity and improve the solubility. This approach is sometimes referred to as a “Trojan horse formulation” because the molecule is hydrolyzed in the acidic environment in the intestinal tract to generate the active drug molecule.

These chemical approaches are considered early in the development of the drug candidate. However, for many APIs, changing the structure of the molecule, particularly one that has been carefully designed to have strong affinity for the biologic site of action, can dramatically reduce the efficacy, which is unacceptable.

For these molecules, formulation scientists must identify a means of increasing the solubility without changing the molecular structure. Here, too, there are several options, such as forming amorphous solid dispersions; incorporating specific excipients, for example lipids; and/or reducing the particle size. 

DA: How was the Biopharmaceutical Classification System established, and how does it correlate the in vitro and in vivo performance of molecules?

DJP: Solubility is important for absorption from the gastrointestinal tract. If a drug is not in solution in the gastrointestinal fluid, then it won’t have the opportunity to pass through the intestinal membrane. Another important parameter is permeability, which has to do with how well the drug, if soluble, can actually pass through the intestinal membrane. Solubility and permeability are equally important with respect to absorption.

This creates challenges, because optimizing either solubility or permeability alone can have a negative impact on the other. For instance, a lipophilic, fatty, “greasy” molecule with no ionizable groups is ideally suited for passing through the intestinal membrane, but, owing to its lipophilic structure, will have low solubility in the aqueous gastrointestinal fluid. As a result, it can sometimes be very difficult to achieve that sweet spot of really good solubility and really good permeability.

The Biopharmaceutical Classification System (BCS) is a framework created in the 1990s by a U.S. academic group in cooperation with the U.S. Food and Drug Administration (FDA) that describes molecules on the basis of these two parameters. BCS Class 1 molecules have both good solubility and good permeability. BCS Class 2 compounds, which are probably the most common, have low solubility but good permeability. BCS Class 3 molecules have the opposite properties: good solubility and low permeability. The most difficult-to-formulate drug candidates, which have both low solubility and low permeability, fall into BCS Class 4.

The BCS was primarily designed as a tool to assist formulators in their regulatory submissions. It allows formulators and regulatory scientists to understand the risks associated with biowaivers, which apply to certain types of molecules being developed as generic drugs for certain indications under defined circumstances. When all of the conditions apply, it is not necessary for drug developers to conduct in vivo pharmacokinetic studies to establish bioequivalence. For BCS Class 1 molecules with good solubility and permeability, in vitro data is sufficient, and clinical trials are not necessary. This saves resources while also minimizing risks to patients. The biowaiver can be applied.

DA: So far, we have been considering oral dosage forms. Are solubility and permeability similarly critical for other dosage forms or routes of administration?

DJP: Solubility and permeability are important issues for topical, nasal, and respiratory delivery. Solubility is often addressed through solubilization in a spray or other formats. Permeability is particularly crucial for topical formulations, because the drug substance has to pass through several layers of skin. While there isn’t as much of a barrier in the nasal cavity, there are still a lot of tissues and many cells through which the API must pass. Permeability is also important for respiratory delivery due to the need for the drug substance to pass through various membranes.

The BCS Class of a molecule is one of the factors taken into consideration when determining the most appropriate route of administration for drug candidates.

DA: Can you explain how shortcomings in the BCS led to the creation of the Developability Classification System?

DJP: First, we have to think about what solubility actually means from a molecular perspective. Essentially, the solubility of a molecule is the concentration at which the rate of precipitation of the molecule is equal to the rate of dissolution and solvation. Simply put, the thermodynamic solubility is the maximum concentration at which the drug is stable indefinitely without precipitation. A molecule with low solubility will reach the thermodynamic equilibrium at a low concentration, while a molecule with high solubility will have a high concentration.

Because solubility is a thermodynamic descriptor, it does not take kinetics into account. Kinetics are very important, because they determine the dissolution rate — how long it takes to reach the thermodynamic equilibrium. A highly soluble drug substance that takes 17 hours to reach thermodynamic equilibrium when formulated in a tablet will never have a chance to reach thermodynamic equilibrium, because the API will no longer be in the system.

The BCS serves its purpose very well from a regulatory perspective. However, it doesn’t give formulators any idea of how best to formulate their molecules. To address this issue, Dressman and Butler from Goethe University in Frankfurt modified the BCS to make it more biologically relevant and more applicable to formulation development. The result was the Developability Classification System (DCS).

Specifically, rather than determine solubility in simple buffers, which is what the BCS does, the DCS determines solubility in fasted simulated intestinal fluid (FaSSIF), a freeze-dried, powdered mixture of phospholipids, bile salts, and buffer components that is reconstituted in water and mimics the fluids in the gastrointestinal tract. This makes the DCS more biorelevant. The cutoff point between soluble and not soluble was also increased, making the DCS more lenient with respect to what is considered highly soluble.

However, the most crucial change in the DCS was the addition of two new classes. BCS Class 2 was split into two different classes separated by the solubility-limited absorbable dose (SLAD) line, which represents a linear correlation between solubility and permeability. Molecules that fall above the SLAD line – DCS 2a molecules - have a slow dissolution rate. DCS  2a molecules are dissolution limited with apparently low solubility. Increasing the dissolution rate — getting these molecules into solution more quickly — can improve the absorption, because these molecules have high permeability in relation to the dose/solubility ratio. Class 2b molecules fall below the SLAD line and are solubility limited. Here, the dissolution rate doesn’t matter, because the solubility is limiting and the permeability is low in relation to the dose/solubility ratio, so no improvement in absorption can be achieved with an accelerated dissolution. For DCS 2b molecules, the root cause — low thermodynamic solubility — must be addressed.  

The DCS is arguably one of the biggest advances in oral solid dosage formulation science in the past 10 years. It allows formulation scientists to truly understand the root causes of their molecules’ challenges — and it is only when you understand the root cause of a molecule’s challenges that the problem can most effectively be solved.

DA: Are these properties always determined empirically, or can structure be used as a predictor of solubility or permeability?

DJP: Much work remains to be done to establish how computational tools might be used to predict permeability, solubility, and the root cause of solubility (i.e., solubility limited vs. dissolution limited). This is an area of considerable interest within our SAFC^® excipients portfolio at MilliporeSigma. We are developing a system, but it remains in the investigational stage. In the future, however, we believe we will have algorithms that can effectively predict the DCS Class for a molecule based on its structure.

DA: Can you tell me about the formulation strategies used for DCS Class 2a molecules?

DJP: These are molecules limited by their dissolution rate. The dissolution rate of any molecule is defined by the Noyes–Whitney equation, which describes the rate of dissolution as a function of the solubility, the diffusion coefficient, the surface area, and the diffusion boundary layer.

As I mentioned, the diffusion coefficient is related to how immersed the molecule is in the solution, which we judge on the basis of certain factors — whether it comes into contact with “solvent” molecules, whether it has a similar polarity, whether it mixes well with the dissolution media, and so on. The surface area refers to the surface area of the API. The diffusion (or Olmstead) boundary layer is essentially the bubble of fluids that surrounds each API molecule, the width of which influences the rate of dissolution. The wider or thicker the bubble of fluid is, the longer it will take for the drug substance molecule to come in contact with the bulk media and go into solution.

In a bioreactor, the Olmstead boundary layer can be minimized by vigorous stirring. That is not possible in the GI tract, so its value is assumed to be constant. Of course, the solubility can be improved, but for DCS 2a molecules this is not necessary,, and so it can be considered a constant, Consequently, the dissolution rate essentially becomes a function of the diffusion coefficient (or how well-integrated the molecule is with the aqueous medium) and the surface area. Molecules that have a high affinity to the dissolution media and have very small surface areas will exhibit high dissolution rates.

Understanding the Noyes–Whitney equation as it relates to a particular API can be used to design effective formulations for DCS 2a molecules. Since the dissolution rate is a function of the molecule’s surface area, one of the most common ways to enhance dissolution is to reduce the particle size — smaller particles have greater surface area. Techniques such as micronization and nanosizing are ways to reduce the particle size, increase the surface area, and ultimately accelerate dissolution.

Another approach is to improve the diffusion coefficient by more effectively increasing the affinity of the drug with the media. Hydrophobic molecules in the GI tract tend to form clumps or aggregates to minimize their contact with the aqueous solution in the gut. One solution is to use surfactant excipients, such as our Parteck^® PLX 188, which is an amphiphilic copolymer comprising a hydrophilic component in polyethylene oxide and a hydrophobic component in polypropylene oxide.

When Parteck^® PLX 188 is used for dissolution enhancement, the polypropylene oxide segments of the polymer bind to the hydrophobic drug molecule, and the polyethylene oxide segments bind to the GI fluids. As a result, the API is brought in contact with the GI fluids, allowing diffusion to occur more rapidly. Increasing the coefficient will then increase the rate of dissolution, according to the Noyes–Whitney equation.

DA: Are there other methods to alter the diffusion coefficient for Class 2a APIs beyond surfactants?

DJP: Absolutely. Incorporating disintegrants into a tablet can also accelerate contact between the drug molecule and the GI fluids. In our portfolio, Parteck^® CCS, which is based on croscarmellose sodium, is a polymeric superdisintegrant that essentially disintegrates the tablet immediately upon contact with the gastrointestinal fluids. The faster the tablet disintegrates, the quicker the molecule comes into contact with those fluids and the quicker it will go into solution.

Another option is to remove any unnecessary hydrophobic excipients in the formulation. The more hydrophobic ingredients in a tablet blend, the more opportunities there are for a hydrophobic drug molecule to form aggregates and minimize contact with the GI fluids. One classic example is magnesium stearate, the gold standard for lubrication of the tableting process, which essentially serves to prevent excessive wear on tablet punches. While it is an essential component, magnesium stearate is incredibly hydrophobic and carries the risk of reducing the dissolution rate, particularly for APIs already likely to have slow dissolution rates on their own. A good strategy is to replace magnesium stearate with a hydrophilic lubricant.

In our case, that would be Parteck^® PLX 188, which has dual functionality and acts as both a surfactant and lubricant. Using Parteck® PLX 188 enhances the dissolution rate via the surfactant mechanism and by reducing unwanted hydrophobic interactions. This functional excipient was designed specifically for oral solid dosage forms, has an optimized particle size distribution profile, and is compatible with direct compression. In addition, Parteck® PLX 188 is multi-compendial and exceeds all of the regulatory requirements for formaldehyde concentration. It is also an Emprove^® Essential qualified product.

Currently, we are working to demonstrate proof of concept for use of Parteck^® PLX 188 in continuous manufacturing, 3D printing, and many other novel manufacturing processes.

DA: Can you describe the formulation strategies for Class 2b drug substances?

DJP: Molecules that are limited by their thermodynamic solubilities are a challenge, because solubility is a fundamental property that cannot be easily changed; it requires something fundamental. One common approach is to pre-dissolve very lipophilic molecules in lipids to change their solid-state structures.

Most molecules are crystalline in structure, with molecules bonded or interacting with one another in a crystal lattice. This feature is a positive one, because crystalline solids are in a state of thermodynamic equilibrium. For most molecules, the fact that the equilibrium is stable is attractive, because it means that the drug substance will be stable over a long shelf life. For hydrophobic molecules, however, it is a negative attribute because it is very difficult to solubilize crystalline materials that have strong crystal lattices.

To dissolve such a material, the bonds that hold the molecules within the interconnected web that comprises the lattice must be broken, and each individual molecule removed from the web and inserted into the surrounding media. These two distinct steps — bond breaking and insertion into the surrounding solution — each have their own energy requirements. The first step requires sufficient energy to overcome the crystal lattice energy, while the second step requires enough energy to create a cavity in the solution suitable for insertion of the molecule. The latter depends on the polarities of the molecule and the solvent, with the solubility greatest when the polarities are similar.

Overall, then, the solubility of a crystalline material is a function of the strength of the crystal lattice and the affinity of the drug for the solvent. In a bioreactor, the solution can be heated to provide the energy needed for dissolution of crystalline materials. That is not possible in the GI tract. The temperature remains fairly constant at approximately 37 °C. Similarly, the polarity is an innate property of the molecule and cannot really be changed at this point in the drug development process.

The answer is instead to change the nature of the material in the solid state and formulate the drug substance in its amorphous form rather than using its crystalline form. The chemical structure is the same, but instead of having the individual molecules bonded strongly to one another, they only interact weakly. Removing a molecule from the “web” takes much less energy, and the solubility is thus increased.

The challenge with this approach is that amorphous forms are not at thermodynamic equilibrium; the molecules have excess energy and are at risk of recrystallizing to release it. Therefore, to use the amorphous form of a molecule to enhance its solubility in a drug formulation, it is necessary to stabilize the molecule in that higher energy form.

Within our portfolio, Parteck^® MXP and Parteck^® SLC have the primary function of stabilizing the amorphous forms of different molecules to achieve enhanced solubility and absorption.

DA: Can you give me an overview of how the Parteck^® MXP and SLC systems work? 

DJP: Parteck^® MXP is a synthetic polyvinyl alcohol (PVA) — specifically, PVA 4-88. The 4 refers to the percentage of viscosity in water and is indicative of the molecular weight of the polymer, which correlates linearly. The 88 is the percentage of acetate ester groups that were hydrolyzed to hydroxyl groups during polymerization. In this case, the higher the number, the more hydrophilic the molecule. So, this PVA is a relatively low-molecular-weight, hydrophilic polymer.

Parteck^® MXP is used as a carrier for amorphous solid dispersions formed via hot-melt extrusion (HME). In this process, a mixture of drug and polymer is melted and mixed together in the molten form to generate a homogeneous solution. The solution is then extruded through a die — a very small hole — and rapidly cooled on a conveyor belt. The resultant solid-state material comprises the API in its amorphous form suspended and stabilized within the polymeric matrix. When ingested, such an amorphous solid dispersion (ASD) provides accelerated solubility and increased absorption in the GI tract.

One of the key benefits of Parteck^® MXP is its high flexibility with respect to downstream formulation development. Generated ASDs can be formulated into capsules, directly molded into tablets, ground into powders, or broken into pellets. They can also be used in sustained-release products. Just as importantly, Parteck^® MXP has a very high thermostability. It can be extruded at very high temperatures without any chemical degradation.

Parteck^® SLC consists of mesoporous silicon dioxide. Silicon dioxide has been used for decades in pharmaceutical applications, primarily as a glidant, which means it has a long regulatory history. Specifically, Parteck^® SLC is multi-compendial — the United States Pharmacopeia and European Pharmacopeia — and a generally-recognized-as-safe (GRAS) material according to the U.S. FDA.

But Parteck^® SLC is not like a typical silicon dioxide glidant, because it is mesoporous with a tightly controlled pore size of approximately 6 nm in diameter. It is highly porous like a sponge and thus has an incredibly high surface area of 500 m²/g. To put that into context, a vial containing 10 grams of Parteck^® SLC would have approximately the same surface area as a standard premier league football (soccer) pitch — so, huge amounts of surface area.

The tightly controlled pore size and very high surface are the key attributes of Parteck^® SLC. When and combined with a concentrated solution of a crystalline API solubilized in an organic solvent such as ethanol, acetone, or methanol (to break down the crystal lattice), it will absorb a large amount of the drug molecules inside its porous network. Once the solvent is rapidly removed, the drug molecule is then trapped inside the 6-nm pores of the Parteck^® SLC, and this nano confinement results in unparalleled amorphous stability.

The powder that is generated in this process can be formulated as a tablet, capsule, or suspension. Once the formulated product enters the GI tract, the GI fluids are absorbed onto the surface of mesoporous silica — the Parteck^® SLC. This establishes a concentration gradient, with lots of drug molecules on the inside of the silica and none on the outside. As a natural consequence, drug molecules are released into the gastrointestinal environment, generating high concentrations of the amorphous form, which, with its increased absorption properties, has accelerated solubility and increased absorption.

DA: How do you determine when the Parteck^® MXP and SLC systems are the appropriate approach?

DJP: Parteck^® MXP is great for molecules that have a solubility limitation — so, DCS Class 2b APIs. Most molecules that have a low solubility, when extruded with Parteck^® MXP, result in stable formulations that exhibit enhanced absorption from the gastrointestinal tract.

For molecules with a poor stability in the amorphous form, Parteck^® SLC may be a good option. With Parteck^® SLC, the drug molecules are truly, unchangeably confined inside 6-nm pores. They cannot migrate or recrystallize. There is no other technology on the market that stabilizes the amorphous form better than Parteck^® SLC.

DA: Are the formulation strategies leveraged for DCS Class 2a and 2b molecules applicable to DCS Class 4 compounds?

DJP: Absolutely. BSC Class 4 and DSC Class 4 are essentially the same. In recent years, the strategy for these molecules has been to focus on increasing the solubility in order to increase the absorption to some extent. There has also been some work on permeation enhancers. Overcoming the challenges presented by DSC Class 4 APIs is a very active area in the literature.

We are currently exploring options within our SAFC^® portfolio that are early stage but which we believe will enhance permeability. I can’t say too much about these efforts yet, as they are still in the incubation period, other than that we are working on some options that we think will enhance permeability.

One example from the literature that I can share is the use of excipients such as Vitamin E TPGS, which has been shown to enhance permeability in situ. When formulated into tablets, it dissolves in the GI tract, and somehow — we’re still not sure how, and neither is the literature — that enhances permeability. Overwhelmingly, however, the bulk of efforts in this space focus on solubility.

DA: What are the implications of the DCS for practical oral dosage drug formulation?

DJP: Why is the DSC so important? We can see the reasons in our three excipients: Parteck^® PLX 188, Parteck^® MXP, and Parteck^® SLC. The first is a poloxamer that simply requires mixing into the tablet blend to achieve accelerated dissolution, because it is designed for APIs with poor solubility due to limited dissolution rates — DSC Class 2a molecules.

Using Parteck^® MXP and Parteck^® SLC involves more complex processes — extrusion, solvent dissolution, and mixing, followed by evaporation and processing into the final dosage form. But these excipients help formulators achieve enhanced solubility for APIs that are solubility-limited — the more challenging DCS Class 2b molecules.

With the greater understanding and clearer classification of molecules provided by the DSC, formulators can develop the most efficient formulation for each individual API. If a drug substance is determined to be a DCS Class 2a molecule, time and resources don’t need to be wasted on the development of formulations using complex formulation technologies, saving time and money. They can use Parteck^® PLX 188. For DSC Class 2b molecules, Parteck^® MXP is often the ideal solution, but in some cases Parteck^® SLC might be more appropriate.

Within our SAFC^® portfolio, we offer a toolbox of solutions to address the different scenarios that can arise. Once we understand the root cause of a formulation issue, we can help identify the right formulation technology for that specific molecule, and, in the process, eliminate the time, effort, and money associated with trial-and-error approaches.

Learn More