Poor solubility is a significant and growing issue for the small molecule pharma pipeline. Active pharmaceutical ingredient (API) processing can address API solubility early in drug development. Salt and cocrystal formation are two options that present different benefits and challenges. MilliporeSigma’s Dr. Thomas Briel, Strategic Marketing Manager – Excipients Liquid Application SAFC® Raw Materials, and Dr. David Lüdeker, Principal Scientist – Solid Formulation SAFC® Raw Materials, discuss how their company is developing solutions that can help drug developers select optimum salts and co-formers for their unique and complex APIs with Pharma’s Almanac Editor in Chief David Alvaro, Ph.D.
David Alvaro (DA): Can you put the growing challenge of API solubility into the context of API development today?
Thomas Briel (TB): Solubility is critical to all drug products, irrespective of the dosage form. For liquid solution formulations, the APIs must dissolve in the formulation solution when first produced and then remain in solution during storage over an extended period of time. This is even more important when it comes to parenteral products which need to be free of visible particles, or high-concentration formulations. For solid formulations, whether tablets, capsules, powders, or other solid oral dosage forms, the API must dissolve latest at the site of absorption so that it can be taken up by the body. Without dissolution, there is no absorption and thus no activity at the systemic site of action, and the API is excreted from the body without having any effect.
Roughly 40% of APIs in marketed drug formulations fall into BCS (Biopharmaceutics Classification System) class 2 or 4, meaning that they have low solubility or are insoluble. This percentage is expected to increase significantly in the near future, as 70–90% of candidates in the pipeline fall into these two BCS classes.1 In addition, many APIs with high bioactivity are not being commercialized because they are poorly water soluble and as a result exhibit bioavailabilities that are too low to support efficacious products.
Finding adequate solutions to address this challenge is therefore becoming increasingly important, because poor solubility presents a real risk that a product may ultimately fail. While solubility can be addressed by sophisticated pharmaceutical formulations, that means leaving optimization until later development stages. Using API processing solutions, however, allows for improvement of drug solubility early on during drug development. Several techniques are available from which to choose, allowing matching of solutions to the specific characteristics and needs of individual APIs.
DA: Why do new drug molecules increasingly face solubility issues? Is it a factor of complexity, and most of the simpler molecules have been exhausted?
David Lüdeker (DL): It is not only the increasing molecular complexity of API candidates that contributes to poor solubility. Many of these molecules also are more hydrophobic or lipophilic than compounds developed in the past. Today’s high-throughput screening and target-oriented drug discovery approaches often result in challenging, poorly water-soluble APIs.
In large part, that is due to the fact that APIs are first designed to bind to specific targets, such as certain active centers within proteins. Then, potential toxicity issues are evaluated. Solubility is generally considered later on.
DA: Can you talk about the full range of different approaches for addressing poor solubility, from molecular design through formulation?
TB: Solubility can of course be addressed through formulation development, and there are diverse excipients designed for solubility enhancement. MilliporeSigma within its SAFC® portfolio brand offers e.g., silica carriers from the Parteck® excipient range and complexation agents like cyclodextrins, excipients for the formation of amorphous solid dispersions (ASDs) via hot-melt extrusion or spray drying, and many more.
Formulation alone, however, might not be sufficient to increase solubility. In addition, it comes at a late stage of development and might turn out to be too late. With API processing, it is possible to address solubility as early as possible during API development, thus eliminating surprises. There are diverse processing options, e.g., polymorph screening and salt- and cocrystal- formation, as well as particle engineering. Each method provides a unique route for addressing solubility; nanomilling is about physical modification, salt formation is about ionizing the API, and polymorph screening is about identifying the most suitable crystalline form.
DA: Is it typical to explore all of these options simultaneously?
TB: It is important to recognize that there often is no single solution for addressing poor solubility. Often, it must be addressed using different techniques in parallel or in series. Polymorph screening is typically performed first, as it is necessary to meet regulatory requirements. Once the right polymorph has been identified, then salt or cocrystal formation might be warranted, or in some cases both could be possible, and screening in parallel could help identify the optimal choice. Nanomilling can be performed on salts, cocrystals, and other API forms, because it is a physical modification done to increase the surface area of the particles as a means for increasing dissolution. It is therefore possible for an API to go through three or four processing steps in a row to boost solubility. After that, formulation techniques can be employed to further enhance the solubility.
DA: Where does particle sizing fit in?
DL: One of the most common methods of particle engineering is nanomilling, and there are some companies that subject all of their APIs to this form of physical modification. It is a straightforward process that allows for enhancing your dissolution behavior. However, there are issues that can occur related to the stability of the formulation. Indeed, significant mechanical stress is transferred to the API during nanomilling, and more complex APIs often are more prone to chemical decomposition or undergoing phase transitions. Hence, after polymorph screening, typically salt or cocrystal formation is assessed. Simultaneously, enabling formulation approaches such as the formation of ASDs or co-amorphous formulations can be an option. After having explored these avenues, I would consider using particle engineering techniques, such as nanomilling.
DA: Can you explain the concept of salt formation briefly and the key benefits and considerations?
TB: Salt formation is appropriate for ionizable APIs, including those that are anionic, cationic, or zwitterionic. The process involves ionization of the API, typically via hydrogen ion transfer, using a basic or acidic counterion depending on nature of the API. This technique is well-established, simple, cost-effective, and accepted by regulatory agencies as it has been employed for several decades, with roughly 50%2 of all APIs on the market formulated as salts.
Why are so many APIs generated as salts? Salt formation can help improve desired properties of the API and consequently the final dosage form. While poor solubility is the most prominent reason for forming an API salt (along with improvement of the dissolution rate and enhancement of bioavailability), other parameters such as API purity, crystallinity, stability, toxicity, processability, particle size, flowability, and others, can be positively influenced by this process.
DL: There are, of course, some challenges that may arise with salt formation. First, the API must be ionizable. This processing approach will not work otherwise. Second, finding the optimum API–counterion combination for salt formation is critical and thus should happen as early as possible in the drug development process. Third, the common ion effect, disproportionation, and hygroscopicity can be issues.
DA: Is salt formation fundamentally the same process from API to API? Are there a limited number of counterions to choose from?
DL: I think every API is different. It’s not exactly a science — it is also very much an art. Typically, companies have their lists of counterions that work for them. However, salt formation and especially cocrystal formation is an API-specific task; there is no one-fits-all approach to it. The issue isn’t to find a salt or cocrystal that can be generated, but to find the right one having optimal performance.
DA: Is the increasing complexity of APIs making them less amenable to salt formation and leading to more exploration of cocrystal formation?
DL: There isn’t really a correlation. Finding a co-former that will generate a cocrystal of a more complex molecule is a bit more challenging owing to the increased degrees of freedom (higher number of rotatable bonds etc.). It is likely that the increasing use of the cocrystal approach reflects more awareness of the technique.
TB: The ease of finding salts of APIs is probably one reason why developers tend to explore this option first. They have experience and knowledge, not only in terms of knowing which counterions tend to work best but in getting salts approved by regulatory bodies. Few companies have deep knowledge of co-formers or experience developing cocrystals, which means that finding the optimum solution takes more effort. Access to computational tools that simplify the process is helping to some extent.
DL: I agree that developers will try salt formation first because it is a technique with which they are more familiar. They might look at cocrystal formation if the salt approach is not effective. I should add that there are some companies that perform cocrystal screening specifically for intellectual property reasons. They are looking to build IP around their API to protect it from generic competition. In fact, cocrystals are interesting compounds from an IP perspective.
DA: Can you discuss and contrast salt and cocrystal formation in terms of the unique benefits that cocrystals offer and the challenges that they present?
TB: As mentioned previously, to form a salt, the API must be ionizable. Not all are. On the other hand, almost every API can form a cocrystal with the right co-former. In addition, the common ion effect and disproportionation are not issues for cocrystals. They also tend to be less prone to hygroscopicity.
For cocrystals, a crystal structure is built by the API and a co-former based on non-covalent bonding interactions, such as van der Waals forces, hydrogen bonding, and others. Like salts, cocrystals provide diverse benefits such as improved drug solubility, dissolution rate, bioavailability, physical and chemical stability, and processability. They may also enable improvement of the API purity, crystallinity, particle size, and flowability.
However, because the co-former remains in the final drug’s crystal structure, finding the right co-former in sufficient quality is key. Indeed, the challenge is to identify the right co-former and develop a controllable crystallization process that generates a good cocrystal. Therefore, experience is needed not only for co-former selection but also in process development. Another challenge is that cocrystals are relatively new, with fewer than 10 drugs marketed with the API generated in this form.3
DL: On the other hand, at least in the United States, the FDA allows cocrystals to be submitted for review via the abbreviated new drug application (ANDA) process, which is not possible with salts.
DA: What can you tell me about matching an API with a co-former? Do you use computational prediction in combination with screening?
DL: Co-former selection is an API-specific task that is mainly dependent on the purity, structural features, and physical properties (e.g., solubility, thermal stability) of the API. Screening may be done in-house by pharma companies or outsourced to CROs (contract research organizations), but regardless should be done as comprehensively as possible to ensure full IP protection.
In many cases, sophisticated experimental approaches (e.g., solvent evaporation, solvent-assisted grinding, sublimation, slurrying, crystallization from melt) are supported by computational methods. The latter are particularly important if there is limited supply of the API, as computation screening is a promising approach for saving time and resources.
Especially in the last two years, several papers have been published on the application of machine learning tools for optimal cocrystal identification. Our own internal evaluations have revealed that most do not work especially well. They are not applicable, generally owing to data inconsistency issues, as they use data from different literature sources that were generated using different methods. That makes the value of these models rather limited.
DA: Is the actual process of forming cocrystals fundamentally similar from API to API?
DL: There are in general two major cocrystal-formation routes. In the first approach, the cocrystal can be generated in the solid state, which is a good screening approach to find a hit. The mechanical energy input leads to cocrystal formation if the co-former is appropriate.
The second route starts with the API and the co-former in solution. It requires that both compounds have comparable solubility in that solvent mixture to ensure that they don’t crystallize as single components during crystallization.
DA: Can you talk a little bit about the regulatory guidances covering salts and cocrystals?
DL: There have been two statements, one each from the European Medicines Agency (EMA)4 and the FDA.5 The main difference is that EMA does not strictly separate salt formation and cocrystal formation, while the FDA distinguishes the two.
DA: What main messages do you want readers to take away from this discussion?
TB: Poor solubility is a significant and growing issue for the small molecule pharma pipeline. API processing can address API solubility early in drug development. Salt formation is well established and has many upsides, but is only applicable for ionizable APIs. It also bears some stability risks, like disproportionation, hygroscopicity, and common ion effects. Cocrystals are still rare but strongly emerging. They represent a possible means for increasing solubility and bioavailability for nearly every API, while avoiding the key issues that may be observed with salts.
DL: There are some considerations that must be weighed when looking at cocrystals, however, selection of the best co-former for a given API remains challenging, regulatory differences between EMA and the FDA exist, and the IP landscape is evolving. If overcome, however, cocrystals can offer excellent opportunities to enhance API properties. Both cocrystal and salt formation should definitely be part of every challenging small molecule development program.
References
- Ting, J.M. “Advances in Polymer Design for Enhancing Oral Drug Solubility and Delivery.” Bioconj. Chem. 29: 939–952 (2018).
- Paulekuhn, G. Steffen, Jennifer B. Dressman, and Christoph Saal. “Trends in Active Pharmaceutical Ingredient Salt Selection based on Analysis of the Orange Book Database.” J. Med. Chem. 50: 6665–6672 (2007).
- Bandaru, Ravi Kumar, et al. “Recent Advances in Pharmaceutical Cocrystals: From Bench to Market.” Front. Pharmacol. 12:780582 (2021).
- Reflection paper on the use of cocrystals of active substances in medicinal products. European Medicines Agency. 2015.
- Regulatory Classification of Pharmaceutical Co-Crystals – Guidance for Industry. U.S. Food and Drug Administration. Feb. 2018