A drug is considered poorly soluble when it fails to generate the concentration in the gastro-intestinal fluids required to produce a therapeutic response. In this context, solubility-limited is a more overarching theme and can represent a variety of limitations.
When a drug is administered to a patient the first potential limitation is the dissolution rate of the drug (Figure 1). The dissolution rate is the amount of drug that dissolves as a function of time. This means that even if the solubility of the drug in the gastro-intestinal fluids is high enough, the slow dissolution rate of a drug can prevent the drug from reaching the equilibrium solubility during the absorptive window. The next limitation is the actual equilibrium solubility of the drug, the concentration at which the solubilized drug is in equilibrium with its crystalline form present. The third limitation, although there is some debate, is the aqueous boundary layer (ABL) that is present before the enterocytes, the cells that form the intestinal lining. The ABL is defined as a water layer where diffusion dominates over convection. This means that the ABL can become limiting when the drug is lipophilic enough. Lipophilic compounds have no issue permeation across the cell membrane, this can cause a concentration gradient from the proximity of the cell membrane back towards the bulk solution where convection dominates.
Figure 1 . Steps and solubility limitations before absorption of the drug. From left to right: Dissolution from crystal lattice to generate a saturated solution, crossing the Aqueous boundary layer, absorption in enterocyte cell.
There are some common techniques available to formulators that can help to overcome the solubility limitations of a drug. Cyclodextrins, lipids, surfactants, nanocrystals, salt formation, using the amorphous form (ASDs) or a dissolved form (eutectic solvents, cosolvents) are techniques that can enhance the dissolution rate and/or the equilibrium solubility of a drug in the gastro-intestinal tract (Figure 2). Each of these techniques generates a micro-environment (a solubilising condition) for the drug which is in equilibrium with the freely dissolved drug.
Figure 2. Some examples of different micro-environments (solubilising conditions) that can be present in the small intestine. The micro-environments are in equilibrium with freely solubilised drug. However, the microenvironments themselves are usually not directly bioavailable.
Different kinds of solubility values used in formulation development
At SeraNovo, when we talk about solubility it usually means an equilibrium that is established with one of three commonly used forms. Firstly, the crystalline solubility, also known as thermodynamic solubility. It is the solubility found in any solvent where the drug is in equilibrium with its crystalline form present in the solvent. It is at this point the solution is said to be ‘saturated’, and the free energy of the drug-in-water system has reached its minimum.
Secondly, the amorphous solubility. This value is obtained when no crystalline form of the drug is present in the solvent and instead an amorphous phase has established an equilibrium with the solvated drug. The amorphous phase is highly disordered compared to a crystalline lattice and is therefore in a higher state of energy. The free energy difference between the amorphous and the crystalline form dictates the concentration difference found when both systems are in equilibrium with a solvated phase.
At the equilibrium, the free energy of mixing the drug and water has reached its minimum. This means that ‘taking out’ drug molecules from the solid form and ‘force’ them in the solubilized form is accompanied by a net energy gain of the system and thus will not occur spontaneously.
If this is forced, however, the system will reach a metastable point where any tiny fluctuation in diffusion processes that leads to a local drug-rich phase and a local water-rich phase causes a net decrease of free energy to the system. Therefore, this process does happen spontaneously, and the two liquid phases will separate out over time. This situation can be forced into being by adding a stock solution of the drug (in a water-miscible organic solvent) to water, with a target end concentration that exceeds the amorphous solubility. This liquid-liquid phase separation, or LLPS, drives the formation of amorphous drug particles until the concentration of freely solubilized drug has reached the amorphous solubility (Figure 3).
Figure 3. Simple representation of different solubilities. Sc is the crystalline solubility; it is in equilibrium with the crystalline form of the drug. Sa is the amorphous solubility; it is supersaturated with respect to the crystalline solubility and a meta-stable state from which crystallization occurs. Liquid-liquid phase separation is the maximum supersaturation before spinodal decomposition is favored, the formation of a drug rich amorphous phase.
The last type of solubility is the apparent solubility (Figure 4). At SeraNovo, we define the apparent solubility as the solubility value that is obtained after a certain sample manipulation was performed prior to concentration assessment (e.g. filtering or centrifugation). However, colloquially it is applied when one of the before-mentioned micro-environments is present (e.g. micelles, cyclodextrins). It is another way of saying that we cannot differentiate between what is molecularly dissolved and what is dissolved by a micro-environment.
This is important because only the molecularly dissolved, unionised drug species can be passively absorbed. A poor in vitro-in vivo correlation will be found if the apparent drug solubility in biorelevant media is mistaken for the molecularly dissolved drug fraction, as the true indicator of permeability is its thermodynamic activity.
Figure 4. The apparent solubility is tied to the method of determination and encompasses every form of solubility allowed by the method of determination. For example, the apparent solubility obtained by 1 mm filtration includes all drug-rich entities that can pass a 1 mm filtration.
Solubility vs thermodynamic activity as a proxy for in vivo exposure
The formulation performance industry standard is a USP-II apparatus dissolution test in which the formulation is exposed to fasted-state simulated gastric fluid (FaSSGF, GF) or fasted-state simulated intestinal fluid (FaSSIF, IF). Here at SeraNovo we use a modified form of the standard dissolution test. We use the same biorelevant media, but instead of two static experiments we perform a single dynamic experiment. We add the formulation to GF medium, and after a certain amount of time we simulate gastric emptying, and we shift to IF conditions in the same container. The drug concentration in the medium is measured after separating the solid phases from the solubilized phases. After several timepoints we can plot a normalized dissolution curve. If we do this for multiple formulations, we can rank order them based on their dissolution performance.
However, like mentioned before, if apparent solubility is used as the measuring stick for performance, then a low in vitro-in vivo correlation can be expected. If the sample manipulation (prior to concentration measurements) does not separate out the micro-environments that have a high affinity for the drug, then the level of supersaturation and thus performance is overestimated. In general, the highest thermodynamic activity is reached at the amorphous solubility, and any excipient added that enhances the apparent solubility will reduce its thermodynamic activity
At SeraNovo, we know that high apparent drug solubility does not equal high thermodynamic activity of the drug. When we develop a formulation for a poorly soluble drug, we consider the impact of the formulation excipients on the thermodynamic activity of the drug. It is a balance of adding enough excipient to facilitate consistent formation of a stable monodisperse drug-rich phase, whilst not adding so much excipient that the thermodynamic activity if affected much (Figure 5). The way we explore this is by quantifying the different solubility fractions that make up the apparent solubility. Centrifugation, ultracentrifugation, dynamic-light scattering, filtration and phase-solubility diagrams are some of the tools we employ to dissect the various micro-environments found in biorelevant media after full dissolution of one of our formulations.
Figure 5. Impact of solubilising conditions (apparent solubility) on maximum thermodynamic activity. The apparent solubility usually is a summation of the crystalline solubility under excipient conditions, the amorphous solubility under excipient conditions, and the LLPS portion of the total apparent solubility.
One can circumvent the dissection of the various fractions in the apparent solubility to gauge thermodynamic activity. It involves doing a side-by-side permeability experiment. In this experiment, a donor chamber is separated from an acceptor chamber by a lipid-impregnated membrane that mimics the cell membrane of the intestinal lining.
Typical experiments employ membranes with a low surface area compared to the volume of the compartments. Practical examples would use a drug concentration on the donor side of 2 mg/mL, but after several hours of running the experiment about 1% of the drug mass has transferred across the membrane (Figure 6). Real life examples have shown full absorption of the drug in less than 15 minutes. As the drug is absorbed, the supersaturating conditions quickly diminish and the excipient-to-drug ratio changes as well as other solubility parameters. The rate of crystallization and the thermodynamic activity of the drug in the in vitro experiment might only reflect what happens in vivo in the first several seconds. In practice this means that it is possible to encounter a situation in which formulations A and B will yield similar flux values on a side-by-side permeability experiment (Figure 6). However, when tested in vivo, formulation A will significantly outperform formulation B because the thermodynamic activity of the drug is influenced by the excipients present from the formulation, which become relevant once the drug is partially absorbed.
Figure 6. Left, schematic of a simple side-by-side perfusion experiment in which a donor compartment is separated by a lipid impregnated membrane and an acceptor compartment. Right, both A and B will give identical Thermodynamic activity (Flux) under quasi-steady state donor conditions, even though in practice, the flux in B will decline sooner as a function of the drug that is absorbed.
Formulation rank ordering using a dynamic thermodynamic activity assessment
This means that at a dose of 2 mg/mL, most of the drug is present as a phase-separated drug rich mixture (LLPS) since there are no real solubilizing conditions present (Figure 7). Then, as the dose is absorbed and the LLPS phase is depleted, any additional absorption that will drive the concentration below the amorphous solubility will be reflected as a decrease in measured flux. The flux will continue to decrease as the amount of drug that is left for absorption approaches 0, or more specifically the equilibrium between the sink condition in the acceptor compartment and the biorelevant condition in the donor compartment. But solubilizing conditions can have a profound impact on the flux as more of the total dose gets absorbed.
Figure 7. An example of how absorption influence the flux. As more of the drug is absorbed, there will be a point at which the remaining drug concentration will go below the amorphous solubility, this is the point where the flux and thermodynamic activity of the drug will start to decline.
Consider a cyclodextrin-based formulation. Cyclodextrins themselves are not bioavailable (limited by permeability). If we add this formulation to a side-by-side permeability experiment at a dose of 2 mg/mL we get the situation shown in Figure 7. in this example, both the crystalline and amorphous solubility are impacted by the presence of the cyclodextrin, and this increases the concentration were the thermodynamic activity (and thus Flux) is maximized. This means that as the drug gets absorbed, the point where the LLPS is depleted will be at higher apparent drug concentrations than that the co-solvent example. This will cause the flux to decrease when there is still a lot of drug left that has to be absorbed and could impact the total fraction absorbed in vivo if the flux gets too low.
Figure 8. An example of how absorption influence the flux. As more of the drug is absorbed, there will be a point at which the remaining drug concentration will go below the amorphous solubility, this is the point where the flux and thermodynamic activity of the drug will start to decline. Solubilising conditions tend to increase the concentration at which maximum thermodynamic activity is measured.
At SeraNovo we are constantly trying to improve the predictability our in vitro data with respect to in vivo exposure. This means that the resulting thermodynamic activity of the drug after formulation dissolution must be taken into consideration when a solubilising condition is present. A way to directly measure the thermodynamic activity of a drug under the influence of excipients is to perform a side-by-side perfusion experiment and measure the flux for each of the formulations. There are currently several limitations to using thermodynamic activity to predict in vivo exposure of formulations that generate supersaturation. The low mass transfer of the experimental setup cause quasi-steady state donor side concentration, and thus maintenance of maximum supersaturation, that will cause an overestimation of de-supersaturation due to drug crystallisation. A solution to this problem is to artificially simulate absorption. This can be done by diluting the donor side with intestinal medium that has all the excipients pre-dissolved at initial concentrations. The result is that only drug disappears on the donor side. This approach allows the measurement of flux as a function of donor concentration which can be used to appropriately rank order formulations that generate supersaturation under solubilising conditions.
Exploring solutions for poorly soluble molecules? Connect with our experts today to delve into advanced formulation strategies.