Cynthia A. Challener is a contributing editor to Pharmaceutical Technology.

Cynthia A. Challener

Autologous chimeric antigen receptor (CAR) T-cell therapies have been shown to be effective for the treatment of many types of cancers. Challenges with scalability and manufacturing associated with autologous cell therapies have led to rapid advancement of allogeneic off-the-shelf cell therapies. However, allogeneic cell therapy products have their own challenges including problems with excessive immune responses and concerns about graft-versus-host disease (GVHD).

T cells are not the only immune cells, however, and there is significant interest in leveraging natural killer (NK) cells in a similar manner. NK cells are lymphocytes that have the ability to target tumor cells, releasing chemokines and cytokines that activate the innate and adaptive immune systems. They can be isolated from a variety of sources and do not have to be patient-specific. Progress is being made in the development of scalable production processes. Clinical trial data from engineered allogeneic NK cells is already in the public domain, additional clinical data continues to emerge from early-generation products, and data read-outs will soon begin for next-generation engineered products.

NK cells are important lines of defense the body has against infectious agents and tumors. They target any cells lacking a major histocompatibility complex (MHC), a group of genes that code for proteins found on the surfaces of cells that are unique to each individual and part of the human leukocyte antigen (HLA) system, according to Heather Stefanski, vice president of medical services at the National Marrow Donor Program (NMDP)/Be The Match. “Each T cell has a unique T cell receptor that recognizes a cognate antigen presented in the MHC complex; the repertoire of human T cells includes millions of unique T cells. Tumor cells can evade the immune system as tumors often do not have antigens that T cells can recognize, and therefore, the T cells do not attack them,” Stefanski says.

NK cells have the intrinsic ability to mediate tumor killing through innate recognition pathways without prior sensitization, comments Ryan Larson, vice president and head of translational science at Umoja Biopharma. “These intrinsic anti-tumor properties can be augmented by engineering NK cells to express synthetic tumor targeting proteins, such as CARs or adapters compatible with tumor-specific monoclonal antibodies that can be administered in combination,” he notes.

Next-generation CAR cell therapies, contends William Rosellini, co-founder, president, and director of CytoImmune, are focused on NK cells because they are one of nature’s best killing machines. “The NK cell has unbelievable tumor-killing activity with significantly reduced safety issues like on-target/off-tumor effects, GVHD, and cytokine release syndrome (CRS),” he explains.

NK cells do not have endogenous T cell receptors, nor do they cause the secretion of interleukin 6 and other cytokines that can cause undesired immune responses, Stefanski notes. As a result, NK cells typically have shown reduced cytokine production and expansion, and thus reduced rates of some of the serious side effects of CAR T-cell therapies, such as CRS and immune effector cell-associated neurotoxicity syndrome (ICANS), according to Larson. In addition, NK cells do not express antigen-targeting receptors that are tuned specifically to their host, thus eliminating the graft-versus-host risk that exists with allogeneic T-cell therapies.

That latter attribute also makes the development of NK-cell therapies less complex than the development of allogeneic CAR T-cell therapies, Larson adds. “For any allogeneic T-cell product, a genetic manipulation must be made to eliminate T-cell receptor expression. This step is not required for NK cells, and thus simplifies their engineering and manufacturing processes, as well as removing a potential safety issue,” he observes.

The ability to produce allogenic products from healthy starting cells is another advantage. “Autologous, or patient-specific, CAR T-cell therapies are produced using cells collected from patients who are very sick. Those T cells are potentially not ideal for a cell therapy product as some may not be good at expressing CAR proteins. In addition, the types of T cells present in each person will vary, and those differences can have a direct influence on the ability to modify those cells and generate an effective therapy,” Stefanski explains.

Overall, concludes Chris Nowers, CEO of ONK Therapeutics, allogeneic NK-cell therapies have the potential to deliver treatments with compelling clinical profiles that are better tolerated and more easily administered, logistically simpler to manufacture, and produced at lower cost. “We believe that through the reduction of costs and a growing accumulation of clinical data demonstrating the positive longer-term benefits, we will be better able to demonstrate measurable value,” he states.

NK cells are present throughout the body and can be harvested from a number of sources. They can be separated from peripheral blood or isolated from umbilical cord blood or the placenta for the development and production of allogenic therapies.Another option involves differentiation of human pluripotent stem cells (iPSCs).NK cells may also be generated from CD34+ stem cells present in bone marrow.

According to Rosellini, NK cells isolated from umbilical cord blood samples have been shown to have a higher cytotoxic capacity and span as much as 30% of white blood cells recovered, while NK cells isolated from peripheral blood tend to be a more mature population exhibiting more refined cytotoxic capabilities with a lower 15% of white blood cells recovered.

The lymphoma-derived NK-92 cell line is one established cell line designed to be easy to maintain and expand, as well as easily genetically modified using both viral and non-viral methods. It does require irradiation prior to clinical use, however, which can impact activity.

All of these different sources have been used to produce CAR NK-cell therapies undergoing clinical development. ONK Therapeutics, for instance, uses umbilical cord blood provided by Anthony Nolan Cell & Gene Therapy Services for development of its engineered NK-cell therapies. “We believe that cord blood offers a widely available source of NK cells that has been clinically validated and provides cells with a potential attractive phenotypic makeup,” Nowers observes.

Larson’s perspective is that we don’t know yet whether the source will affect the application. “It is likely that some combination of source, CMC [chemistry, manufacturing, and controls] processes, engineering/synthetic biology, and tumor-targeting approach (e.g., CAR) will determine the best therapeutic applications,” he states.

As with CAR T-cell therapies, once the NK cells for development of the therapy have been isolated, they must be expanded and genetically modified to express the CAR protein. The methods for cell expansion and cell engineering are similar for NK cells and T-cells, with some specialization accruing to each cell type, according to Larson. In most cases, large-scale expansion using cytokines and feeder cells is required to produce sufficient NK cells for clinical applications, notes Nowers. This approach facilitates efficient engineering.

Rosellini adds that retroviral, lentiviral, or non-viral delivery of genetic material is used, with viral-vector gene delivery systems currently considered the most effective. ONK Therapeutics is an exception. The company uses a combination of clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) editing and transposition of transgenes employing a non-viral TcBuster transposon approach (Bio-Techne) to produce large numbers of edited CAR-NK cells, according to Nowers.

Larson believes that the method that gives the largest scale possible will be the best to commercialize, as long as it can accommodate some engineering to make sure that the cells will perform well, including in the tumor microenvironment (TME) for solid tumors. He does note, however, that there are schools of thought that think it best to use minimally expanded cells.

While the theoretical unlimited expansion capacity of iPSCs makes them an attractive option for the production of CAR NK-cell therapies, Larson also points out there have been challenges to translating this feature to downstream yield of relevant immune cell progenitor populations in an efficient and scalable format. “Solutions for scaling a directed differentiation manufacturing process will realize the full value of iPSC as a starting material from a CMC perspective-in terms of simplicity, reducing cost, and patient access,” he remarks.

In this vein, Umoja recently entered a collaboration with TreeFrog Therapeutics to develop CAR NK-cell therapies (1). Umoja’s iPSCs are engineered with a synthetic rapamycin-activated cytokine receptor (RACR) to drive differentiation to, and expansion of innate cytotoxic lymphoid cells, including but not limited to natural killer (NK) cells in the absence of exogenous cytokines and feeder cells. TreeFrog’s proprietary C-Stem technology relies on the high-throughput encapsulation (>1000 capsules/second) of iPSCs within biomimetic alginate shells that promote in vivo-like exponential growth and protect cells from external stress.

One of the less-desirable attributes of CAR NK-cell therapies when compared to CAR T-cell therapies is their lack of persistence in the body. NK cells have a much shorter half-life on the order of weeks. Administration of cytokines such as interleukin 2 to patients can prolong NK cell lifetimes, but Stefanski notes this approach often comes with harmful side effects.

The need for lymphodepletion is another important issue associated with allogeneic CAR NK-cell therapies. Prior to administration, patients receive a short course of chemotherapy to kill their existing T cells as a means for increasing persistence. “Lymphodepletion adds a veneer of immune system damage that provides short-term benefits in enabling engraftment of therapeutic cells but is detrimental to immune system function and may contribute to recurrences or other problems in the long term,” Larson says.

It is important to employ strategies to improve drug exposure through engineering strategies and be able to dose more than once at high dose levels. “Scalable manufacturing solves a lot of problems,” Larson states. Developing NK cells that secrete their own cytokines, such as interleukin 15, is also an approach to improving NK persistence, according to Nowers. He notes that this strategy can be synergistic with the knockout of the cytokine inducible SH2-containing (CISH) protein.

The development of stealthy NK cells that are edited to evade the immune system is another strategy being pursued by several companies, according to Nowers. ONK Therapeutics is also looking to increase the ability of its NK cells to target specific tumors by using cells that express high levels of favorable chemokine receptors such as CXCR4 and CXCR3 and E-selectin ligands. “These features ensure a higher likelihood that the NK cells will traffic to the site of the tumor within their natural lifetimes,” says Nowers.

Other challenges for CAR NK-cell therapies, according to Stefanski, are similar to those faced by CAR T-cell therapies: difficulty infiltrating the TME of solid tumors, the use of lentiviral vectors for delivery of genetic material, and the need to know the target antigen. In addition, although CRS is avoided by using NK cells, other off-targets are possible that must be avoided. There are also complexities associated with garnering regulatory approval for cell therapies, Rosellini adds.

As next-generation treatments, the development of CAR NK-cell therapies is not as far along as that of CAR T-cell therapies. Most programs are preclinical to early-phase clinical, including both Phase I and II trials. Cancer is the focus, with the greatest number of candidates targeting blood cancers including various leukemias, lymphomas, and multiple myeloma, but there are trials underway in solid tumors such as glioblastoma, non-small cell lung cancer, head and neck squamous cell carcinoma, and prostate cancer.

Rosellini points out that academic research groups are hard at work understanding how CAR-NK cells could be used on other diseases too. “Similar to cancer cells, we could train NK cells to target other malfunctioning cells in inflammatory diseases,” he notes. Stefanski suggests there is a potential for genetically modified NK cells to fight viruses that infect patients following stem-cell transplants and to treat autoimmune diseases driven by malfunctioning B cells. “Theoretically, CAR NK-cell therapies could be used to treat any disease that involves proliferation of abnormal cells,” she concludes.

The CAR NK-cell therapy field is still quite new. Additional clinical data from Affimed and Fate and readouts for some of the hypo-immune cells from Century are expected during the next 6–12 months, according to Nowers.

Stefanski anticipates seeing the number of Phase II trials continuing to grow. Within the next year, she believes at least one hematological malignancy candidate may advance to Phase III. Five years from now, she hopes there will be at least one CAR NK-cell therapy on the market.

CAR NK-cell therapies, Stefanski also suggests given their limited persistence, may have the potential to get patients into remission and in a position to undergo stem-cell transplants they wouldn’t be eligible for otherwise. The limited persistence of NK cells may also be an advantage from a safety point of view, according to Nowers, because it could lead to reduced risk of prolonged on-target, off-tumor toxicity.

Improvements in manufacturing technologies are also anticipated, whether using NK cells isolated from umbilical cord blood or differentiated from iPSCs. “Enhanced manufacturing procedures and the increasing use of automated closed platforms both have the potential to significantly reduce cost of goods,” Nowers notes.

More advanced cell modifications are also being explored. Approaches that include multiple receptors and different ligands on the NK cells have shown potential. “Researchers are beginning to figure out what is best for a given application, because solutions that work well for one type of cancer aren’t necessarily optimal for others,” Stefanski says. As information is gained about the current targets and how existing CAR NK-cell therapy candidates behave clinically, Rosellini believes that much will also be learned about how to improve their safety and potency.

For instance, Larson observes that organizations including Umoja will introduce second-generation approaches that avoid lymphodepletion, include engineering strategies to enhance engraftment and pharmacokinetic profiles, and enable the production of larger numbers of doses with higher cell numbers per dose. “As these advances are layered on, I expect we will see better and better clinical results,” he states.

Similarly, companies that are exploring gene editing to benefit the cytotoxic potential, metabolic health, and durability of engineered cells in-vivo will demonstrate enhanced persistence and metabolism, improved tumor homing, reduced exhaustion and immune invasion, and maximization of antigen-independent killing, according to Nowers.

Longer follow up and more mature data from ongoing studies should give greater confidence in the durability of CAR-NK approaches as well as in their safety, says Nowers.“The breadth of ongoing research and an increased move to better tolerated, logistically simpler allogeneic NK cell-therapy platforms will allow evolution toward a more traditional treatment model for advanced cell therapies. Simultaneously, the use of big data and artificial intelligence to direct selection of the best patient candidates will lead to better efficacy, greater patient benefits, and increased understanding,” he contends.

The end result will potentially be significantly reduced list price costs combined with a growing clinical database demonstrating the positive longer-term benefits of CAR NK-cell therapies, according to Nowers. “Stronger demonstration of value will help support the willingness to pay of payors and subsequently drive broader patient access and adoption,” he reiterates.

1. Umoja Biopharma, “Umoja Biopharma and TreeFrog Therapeutics Announce Collaboration to Address Current Challenges Facing Ex Vivo Allogeneic Therapies in Immuno-Oncology,” Press Release, June 10, 2022.

Cynthia A. Challener, PhD is a contributing editor to 


Vol. 46, No. 8
August 2022
Pages: 24–26

When referring to this article, please cite it as C. Challener, “Progress in the Development of CAR NK-Cell Therapies,” 

Articles

Further advances in construct design and manufacturing scalability are still needed.

Progress in the Development of CAR NK-Cell Therapies

With many blockbuster biologic drugs soon to lose patent protection—Humira and Keytruda, among others—the number of biosimilars under development is increasing. In jurisdictions with robust regulatory standards, such as the United States and European Union, the approval pathways for biosimilars, while distinct from the pathway for originator biologics, are rigorous and scientifically appropriate, according to Jeff Yant, executive director of biosimilars operations at Amgen.

These approval pathways require demonstration of biosimilarity between the biosimilar candidate and the approved innovator drug in terms of efficacy, safety, and purity. To meet this requirement, biosimilar developers must evaluate the physiochemical and structural properties of the biosimilar candidates and reference products and conduct clinical studies to compare the safety and efficacy profiles of the products.

Biosimilars must have the same route of administration, the same dosage, and the same side effects profile as the innovator drug, notes Todd Stawicki, global marketing manager for biopharma at SCIEX. Therefore, to develop a biosimilar, it is necessary to fully characterize the reference product so the biosimilar can be built to be highly similar in all of these respects, including the different variants and impurities that are present, adds Susan Darling, senior director of product and market management for CE and Biopharma at SCIEX.

Biosimilarity is first demonstrated with analytical data supporting structural and functional similarity, then through clinical studies assessing immunogenicity and pharmacokinetics (PK), and in some cases pharmacodynamics (PD). Additional comparative clinical studies, says Ricardo Infante, global medical lead for Organon Biosimilars, are conducted between the reference product and the proposed biosimilar to ensure there are no clinically meaningful differences in efficacy and safety.

“The goal of a biosimilar development program is to demonstrate that the candidate biosimilar is highly similar to and has no clinically meaningful differences from its originator/reference product, not to independently establish de novo the safety and efficacy of the biosimilar,” adds Yant.

The array of data the manufacturer of a proposed biosimilar product must generate for the proposed product and the FDA-approved reference product in order to demonstrate biosimilarity include, according to Yant, analytical studies demonstrating that the biological product is highly similar to the reference product, notwithstanding minor differences in clinically inactive components; animal studies, including an assessment of toxicity; and a clinical study or studies sufficient to demonstrate safety, purity, and potency of the proposed biosimilar product in one or more of the indications for which the reference product is licensed.

Yant also emphasizes that no single study or endpoint should be considered “pivotal” in a biosimilar marketing application. Rather the totality of data and information submitted should support the demonstration of biosimilarity (1).

Developing a biosimilar is more nuanced than developing a small-molecule generic product. With small molecules, the API is a single molecule with a highly defined structure and chemical composition. There may be different synthetic routes to producing the compound that could potentially lead to different impurities, but the actual molecule itself will be identical in the branded and generic versions. There are no possible variants to consider.

Like all biologics, Yant notes that biosimilars are produced through an intricate, multistep process using living cells. The molecule is in fact determined by the process, Stawicki stresses. Every step in the process and every change in process conditions affects the nature of the biologic product. “There is a significant change in mentality needed when developing biosimilars vs. generics,” he says, “with the focus shifting to defining and controlling and characterizing the process to make the highly similar product.” There is also a need, Stawicki notes, for biosimilar developers to have a good understanding of how potential changes to the molecule can change side effects.

Complicating this issue is the fact that for most biologics, the majority of details about the production processes for these reference products are not publicly available. “The cell line and manufacturing process of the reference product are proprietary and belong only to the reference product
manufacturer,” says Yant.

As a result, biosimilar developers basically need to start from scratch and use their own processes. “Characterizing the reference molecule, figuring out which cell line to use, and essentially reengineering the manufacturing process is tremendously expensive. For this reason, some of the companies with the most demonstrated success in biosimilars are also developers of novel drugs—they are able to leverage the same expertise, processes, and toolset for both,” Darling observes.

While the same analytics are used in biosimilar development that are used for general protein-based new-product development, it is even more important to link structure to function, Darling remarks. “It is unlikely that a biosimilar production process will yield the same variants in the same percentages as the process developed for the innovator drug, and it is therefore necessary to be able to demonstrate that the differences in variants and variant percentages don’t impact the function of the molecule,” she explains.

Analytics, adds Infante, are the foundation of biosimilar development. “Analytical testing occurs during initial development to characterize the structure and function because analytical data [are] more sensitive than clinical data in detecting differences between a biosimilar and its reference product should they exist,” he says.

Procurement and analysis of multiple batches of the reference product help to establish a baseline that informs subsequent structural and functional similarity tests. “To ensure high confidence in structural and functional similarity between a proposed biosimilar product and its reference product, comprehensive analytical comparison covering a broad disciplinary of analytical techniques should be considered,” Yant states. He adds that analytical similarity assessment should be performed using scientifically sound, fit-for-purpose, state-of-the-art techniques that are sensitive and capable of detecting small differences.

Peptide mapping is particularly important for understanding sequence variations and differences in post-translational modifications, but standard techniques such as capillary electrophoresis, mass spectrometry, liquid chromatography, and binding assays are all crucial, Darling observes.

All of the analytics used to characterize the innovator drug also guide process development both upstream and downstream, according to Stawicki. “There are multiple steps in upstream process development and even more in downstream development, and every single one requires multiple types of analyses. The better the analytical capability—the more sensitive and more informative—the better the process, and optimum processes are needed for every step that can have an impact on quality, efficacy, and yield. It is not possible to design and execute such complicated biosimilar processes without fully characterizing every step,” he insists.

Yant adds that new analytical technologies may help address uncertainties about biosimilarity between the proposed biosimilar and the originator product, but any potential limitations also must be understood.

Advances in analytical technologies can actually create some challenges. Older biologics that have been on the market for some time may have been developed using technologies that are no longer accepted by the regulatory authorities. One example highlighted by Darling is sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE). State-of-the-art techniques must be used today to meet regulatory expectations, which requires the development and validation of new methods. Newer modalities such as bispecifics and fusion proteins can also present challenges to biosimilar development with respect to analytical method development.

“The difficulties associated with biosimilar analysis really are dependent on the molecule that is involved. Ion exchange chromatography might work for a monoclonal antibody but not for a bispecific. It isn’t so much that the molecule being developed is a biosimilar, but more that each specific type of molecule requires the right specialized hands-on expertise,” Darling comments.

Established platform technologies invaluable

In most cases, biosimilar developers seek to use standard methods that can be identified using information found in publicly available documents filed on the reference compound or in the general literature. “Although much of the process knowledge about reference compounds is kept proprietary, information about analytical methods is often available, and wherever possible biosimilar developers prefer to leverage the same methods used for the innovator molecule because getting the same results with these methods speaks strongly to regulators,” Darling observes. “If different methods are used, then they must be justified and more work is needed to provide assurance of similarity,” she adds. Ideally, those standard methods are based on platform technologies and standard workflows.

The use of statistical methods to aid in assessment of analytical similarity results has been a topic of debate among regulatory agencies and continues to be discussed by FDA and the European Medicines Agency (EMA). The lack of final regulatory guidance has not, however, prevented successful approval and marketing of biosimilar products in both the US and EU, according to Yant. “Finding analysis methodologies that can be broadly accepted by regulatory agencies for comparative assessment of analytical similarity will streamline biosimilar regulatory submissions and approval processes globally,” he contends.

Evolving design strategies to inhibit biosimilar development

Even with more specific regulatory guidance, the development of biosimilars may get more difficult going forward. “When new biologic drugs are developed today, innovators are thinking from the outset about reducing the ability of competitors to develop biosimilars. Innovator molecules are designed with very tightly defined characterization and structural features—for instance very specific glycans—that make it difficult for anyone to exactly reproduce them,” Stawicki explains. “The strategy today,” he adds, “is to design innovator molecules that have a smooth continuation from innovator molecule to very-difficult-to-copy biosimilar so the original developer can maintain a strong market position throughout the life of the drug.”

Regardless of the design strategy and the complexity of biosimilars, however, a focus on quality must be maintained. “Biologics development, whether for innovator or biosimilar products, is a complex process that requires a quality-based systematic approach. All variables and parameters critical to product quality are identified and rigorously monitored to ensure safe and effective products for patients,” Infante avers.

To create and preserve physician confidence, patient safety, and the integrity of the healthcare system, biosimilars must meet and maintain robust scientific standards before and after approval, Yant agrees. He notes, though, that there has been an increased debate around biosimilar development, with some stakeholders calling for lower regulatory standards around comparative clinical studies and demonstrating interchangeability during the development process. “Amgen firmly believes that rigorous, scientifically appropriate regulatory standards are necessary for the approval, manufacture, and uninterrupted availability of safe and effective biological products, including biosimilars,” he says.

“Analytical data alone,” Yant stresses, “are insufficient to determine biosimilarity. There will always be subtle structural and functional differences between a proposed biosimilar and an originator product. Analytical methodologies alone cannot definitively evaluate the potential impact of such differences on clinical outcomes, including pharmacokinetics, efficacy, and immunogenicity.”

Yant concludes that as biosimilar development continues to evolve, all biological products, including biosimilars, must continue to be held to the same manufacturing quality standards—following the same strict standards of quality, safety, and efficacy—expected for reference products.

FDA, Guidance for Industry. Clinical Pharmacology Data to Support a Demonstration of Biosimilarity to a Reference Product 

Cynthia A. Challener, PhD, is contributing editor for 

Biosimilar data analysis must demonstrate a correlation between structure and function.

Developing Methods for Analyzing Biosimilars

More than 25% of drugs on the market worldwide and candidates in the pharmaceutical pipeline were considered to be formulated with highly potent APIs (HPAPIs) in 2020 (1). The vast majority of these products were cancer therapeutics—approximately 60% of approved oncology treatments are based on HPAPIs. At the time, more than 115 contract manufacturing organizations had the capability to produce highly potent drugs, and many had recently invested or were in the process of implementing capacity expansions to meet growing demand for HPAPIs and products formulated with them.

In addition to specialized facilities and equipment and highly trained personnel, organizations developing and manufacturing highly potent drugs must have the requisite expertise to overcome numerous formulating challenges ranging from lack of sufficient toxicity data to managing poor solubility and bioavailability to ensuring uniform dosing of low concentrations of HPAPIs.

HPAPIs, typically defined as having an eight-hour time-weighted occupational exposure level (OEL) of10 micrograms or less, by their nature present risks to anyone handling or using them. Prior to any formulation efforts, therefore, it is essential to understand those risks so that appropriate engineering controls, appropriate personal protective equipment, and robust operator training programs are in place prior to establishing handling and containment policies, according to David Lyon, senior fellow, global R&D small molecules at Lonza. That can be difficult, says Lisa Caralli, senior director of science and technology for pharmaceutics with Catalent, because suitable toxicology data are often lacking in early development.

“These data are not only important to evaluate risks to patients, but also to
understanding how the compounds should be handled by formulators and manufacturing personnel,” Caralli observes. Without these data, she adds that by default HPAPIs are often classified as having a more conservative occupational exposure band (OEB) or OEL, which in turn may limit the outsourcing partners that can provide development and manufacturing support as well as the types of technologies that can be used in the formulation process. “Overall, this lack of crucial knowledge will certainly increase both the timelines and costs,” Caralli comments.

Experienced development and manufacturing partners, adds Caralli, will ask the correct questions regarding the existing studies and may propose a phased, risk-based approach to formulation development, as data needed for a more concrete assessment are gathered. She emphasizes that “any formulation developed must facilitate drug exposure at the intended site of action, in order to generate high quality and accurate toxicology data, which can then be used to refine the classification of the drug.”

When insufficient toxicology and safety information is available, formulators at Evonik focus on the mechanism of action (MoA) of the HPAPI (whether it is a new chemical entity [NCE] or new biological entity [NBE]), according to Kenny McCleary, senior international ESH advisor at Evonik Health Care. Information on the toxicity and safety properties of existing compounds or an existing class of compounds with the same or similar MoAs are considered along with any pieces of information specific to the new HPAPI to determine the OEL using conservative estimates.

This OEL, says McCleary, is then used to establish cleaning strategies and unit operation containment control strategies. As formulation development proceeds and as more safety and toxicology information become available, Evonik reassesses the OEL and acceptable daily exposure classifications.

“Such assessments are essential in multi-product facilities because of the challenges in preventing cross-contamination of one highly potent product into a subsequent product, as well as protecting workers while they are conducting formulation process activities,” McCleary notes. “In the end,” he stresses, “we have two safety priorities when handling highly potent compounds: the safety of the personnel handling these compounds and the safety of the patients who will be administered the drug substance.”

Overcoming poor solubility and bioavailability

HPAPIs can be found across the continuum of chemical space and, as such, can have virtually any combination of chemical and physical properties, according to Lyon. Like many candidates in the pharmaceutical pipeline, HPAPIs often suffer from some level of poor solubility and/or permeability, and thus less-than-desired bioavailability. Currently there are more than 1000 highly potent small molecules in development, notes Lyon, and approximately 70–90% of these compounds are poorly soluble and thus have low bioavailability.

Highly potent compounds can also be associated with a narrower therapeutic window. “When this is the case, it is extremely important to understand how the dosage form will perform in vivo across a wide range of patient types with well-defined co-morbidities,” Caralli says. HPAPIs with low solubilities, she says, should be formulated to ensure a high fraction remains soluble in the intestine where drug absorption occurs.

Some approaches to increasing solubility are better than others, however. If a high dose is used to overcome low bioavailability versus an enabling formulation that improves overall bioavailability, for instance, there is an increased risk of drug overexposure, Caralli observes. Toxic exposure levels can be reached if the drug exhibits factors such as a food effect (i.e., the drug is more soluble in the presence of food) or pH effect (where solubility is impacted by small differences in pH within the gastrointestinal tract). APIs that are CYP450 substrates can also be at risk for variable exposure as the result of genetic variability or drug-to-drug interactions that can lead to the inhibition or induction of CYP450 enzymes at different levels.

A number of different enabling technologies can be used to enhance the absorption of HPAPIs, even if the formulated doses are relatively low due to their high potency. Lyon points to amorphous solid dispersions (ASDs) as being essential for delivering poorly-soluble and highly potent compounds, and particularly spray drying, because it can somewhat mitigate the apparent potency of powdered HPAPIs through dilution in an appropriate excipient system.

HPAPIs, Lyon explains, are most potent in powder form. Therefore, the handling of the powder form must be carefully controlled. In spray drying, once the powder is introduced into a spray solution tank and dissolved in the desired spray solvent, the potency and concentration of the HPAPI are considerably reduced. Similarly, when the solution is dried, even though the API has been returned to powder form, it has been diluted by introducing a polymer or other excipient.

Liquid HPAPIs, meanwhile, are amenable to handling in a contained system and then formulated in a liquid or self-emulsifying lipid system, according to Lyon. “This approach is quite beneficial with respect to HPAPI handling because the API is dissolved throughout the processing equipment train and exposure risk can be effectively minimized,” he comments.

Overall, Caralli observes that using tools such as physiologically-based pharmacokinetic modeling is essential to assess which physicochemical parameters are associated with the highest risk. “The information that is obtained from these studies can be directly applied to the development of effective formulation strategies to ensure both safety and efficacy,” she contends.

The development of ASD formulations takes time, but drug manufacturers today face tremendous pressure to get into first-in-man studies as quickly as possible. One solution for the formulation of powdered HPAPIs during early development that is finding increasing use today is microdosing.

Microdosing, explains Lyon, is the precision-weighing and dispensing of powders directly into capsules or bottles using various specialized equipment platforms or techniques. “API-in-capsule or -bottle studies start with API characterization, from which an understanding of the API morphology, solubility, and other key attributes define the drug-product design and decision-flow diagram.”

Depending on the material’s physical attributes, API doses as low as 100 µg can be encapsulated with minimal variance in weight. That is important for HPAPIs, because these compounds provide therapeutic (and potentially toxic) effects already at very low doses. For instance, potent treatments delivered to the lung are often therapeutic in the 100s of microgram range, if not less, according to Lyon.

Microdosing of an undiluted HPAPI into a capsule using robust engineering controls can be an attractive alternative to other formulation approaches. “Using a drug-in-capsule approach for high potent drugs can be a good way to eliminate issues with blend uniformity that are associated with formulated capsules and tablets,” Caralli says. There are some limitations, however. The HPAPI must be sufficiently soluble and not require bioavailability enhancing strategies. In addition, it must have suitable flow characteristics to allow for accurate capsule filling, she observes. In vitro studies should also be performed to confirm the drug-release profile and explore the potential for undesirable drug-capsule interactions.

Microdosing, Caralli indicates, can accelerate entry to the clinic, because the HPAPI can be well contained, both during production and in the final capsule dosage form. It is not a good long-term dose-design strategy, however, as it is not suitable for scaling up for larger clinical trials or commercial applications.

Liquid-filled hard capsules for oral dosing of HPAPIs

For oral highly potent formulations, the use of liquid-filled capsules, either hard-shell or softgel, is an excellent approach. As is the case with ASDs, the solubilization of HPAPIs prior to filling reduces the exposure risks to formulators and manufacturing operators, Caralli notes. It also has the benefit of eliminating dose-content-uniformity issues that are associated with dry blends, and the process is directly scalable from early phase to commercial-scale batch sizes, which lowers overall program risks.

Lipid-based formulations, meanwhile, have the potential to offer even more advantages, as they can be used to enhance bioavailability through improved solubility, permeability, and for some APIs, may reduce metabolic loss through lymphatic transport, according to Caralli. The specific encapsulation approach, she says, is dictated by the fill solution properties and drug compatibility with the shell components.

While formulation of HPAPIs is most often associated with oral administration, they can be optimal for formulation as parenteral products. “The development of HPAPIs should always be performed with the compound’s target product profile in mind,” Lyon observes. For a compound developed to target lung disease, a locally acting, inhaled formulation would likely be best. Where possible, oral presentation is often more acceptable based on patient preference.

“Parenteral delivery, however, while more complicated due to the need for sterility and a delivery device (e.g., a pre-filled sterile solution in a syringe), can aid in patient compliance and assure that 100% of the dose is administered, particularly when stomach pH, presence of food in the stomach, or age may complicate oral administration,” comments Lyon. He also notes that parenteral administration can be useful in a clinical or hospital setting where the patient may not be able to take oral medications.

There are other benefits to parenteral delivery for HPAPIs with certain physicochemical properties. For instance, when working with hazardous APIs such as some chemotherapeutic agents, where localized tissue damage can occur, the injectable route of administration is clearly the preferred option, Caralli says. Drugs with a narrow therapeutic window that exhibit high first-pass loss or metabolic variability should be considered for injection as well.

Beyond these factors, the need for injectable delivery should be assessed as with non-highly potent drugs, according to Caralli. “Bioavailability challenges, patient dosing (e.g., hospitalized patients), and the onset of action should all be evaluated when considering the injectable route of administration,” she comments.

When sustained delivery is desired, formulation of HPAPIs as long-acting injectables (LAIs), such as bioresorbable microparticles and implants, is also highly attractive, according to Tom Tice, senior director of global technical marketing for Evonik Health Care. “In fact,” he states, “the most important criterion for LAIs is their potency, which translates to API mass.”

That is because API mass as well as excipient mass and injection vehicle mass contribute to the total injection volume, which is limited for parenteral formulations: from 1 to 2 mL for intramuscular/subcutaneous and 100 µL for intravitreal injections. “Many LAIs on the market deliver small-molecule and peptide HPAPIs for anywhere from one to four weeks and as long as six months following a single administration. Bottom line: HPAPIs are desirable candidates for long-acting injectables because of their potency, low dose, and small mass,” Tice concludes.

It is important to remember, according to Tice, that the concentration of API in the formulation, the physical form of the API, and the other excipients and carriers present do not change the inherent pharmacological properties of the compound. As such, all APIs must be managed based on their inherent toxicological properties, because whether they are potent or highly potent, the dosage and formulation of API material can impact the risk assessments, controls, and cleaning practices needed. When formulating HPAPIs in particular, Caralli emphasizes that the safety of both patients and drug development scientists are of equally high priority. “Performing the correct non-clinical studies at the right stage of development is critical to ensure the appropriate containment strategies are applied,” she says. Key to those studies, Caralli adds, is using drug formulation and delivery strategies that ensure appropriate exposure and result in high-quality data. “Because of the dangers inherent to HPAPIs, it is important to let the science drive the development strategy,” she asserts.

Given that HPAPIs will have a continuing strong presence in pharmaceutical pipelines, that potency is increasing, and that many HPAPIs have low solubility and/or low bioavailability, Lyon believes that there will be an ongoing need to further develop and refine current engineering controls, procedures, and use of personal protective equipment to ensure HPAPIs are handled in a robust manner while being formulated using increasingly effective enabling technologies.

1. V. Rees, “Research Finds 25 Percent of Drugs Contain Highly Potent Compounds,”

Cynthia A. Challener, PhD, is contributing editor to 

When referring to this article, please cite it as C. Challener, “Challenges to Formulation Development for Highly Potent APIs,” 

Lack of toxicity data and poor physicochemical properties must be overcome.

Challenges to Formulation Development for Highly Potent APIs

Coprocessed excipients are blends of different excipients subjected to physical processing that leads to modifications of their physical structures without causing any chemical changes. They possess unique functionality that cannot be achieved through simple blending and are increasingly important for addressing formulation issues posed by the highly complex APIs under development. Adoption of coprocessed excipients has been slow, however, because even when prepared using compendial excipients, they are considered “novel” by regulators. The lack of a separate excipient approval process means that drug developers must take on elevated risk when using such “novel” excipients in the formulation of drug candidates.

Reducing complexity and increasing functionality

Oral solid dosage forms are traditionally formulated with combinations of individual excipients with differing functionalities (filler, binder, disintegrant, lubricant, flow-aid, solubilizer, etc.), says Ashish Joshi, pharma technical and business manager at BASF Pharma Solutions. The processes used to formulate these excipients into an oral dosage form are complicated as well as time-, energy-, and labor-intensive (e.g., granulation, spray-drying, roller compaction, etc.). In addition, the use of many individual excipients requires bracketing or optimization studies and has the risk of non-uniform distribution, which can adversely affect final product performance.

The ideal process for tablet formation is direct compression, as it eliminates the need for granulation or other prior processing steps, according to Joao Marcos Assis, global technical marketing manager, BASF Pharma Solutions. “Direct compression formulations, however, typically require multiple excipients to achieve good flowability, high bulk density, excellent compressibility, and fast disintegration (for immediate release formulations) because individual materials do not possess these multiple attributes,” he observes.

Coprocessed excipients can be used to overcome these formulation challenges through their enhancement of the resulting material’s performance and processability, Assis says.“By reducing manufacturing complexity and thus drug development time, coprocessed excipients can significantly expedite time-to-market,” he continues. Furthermore, API and coprocessed excipients normally fulfill the minimum functional requirements for tableting, consolidating quality-by-design efforts by simplifying the formulation. Analytical testing and material handling and warehousing needs and documentation requirements are also reduced.

The benefits of using coprocessed excipients are increased when multiple excipients can be coprocessed together. “Initially, most coprocessed materials were based on just two excipients, but increasingly coprocessed products being introduced to the market comprise multiple excipients,” notes Yeli Zhang, technical service manager at IFF Pharma Solutions.

The ability to combine many individual excipients and functionalities into a single high-functionality excipient enhances the ease of formulation and final performance as well as enables process simplification and cost-savings, Joshi adds. “Consequently, the major trend in developing coprocessed excipients is to reduce the number of individual excipients used in a formulation, with a focus on improving the flow-properties, compressibility, disintegration, and lubrication of oral solid dosage forms using a single coprocessed excipient,” he comments.

This trend is particularly true with coprocessed excipients intended for us in orally disintegrating tablets (ODTs). “Many companies are launching coprocessed excipients targeting this segment that combine a number of excipients because ODT formulations require a high number of excipients and are typically formed using direct compression,” adds Vinay Muley, research and development leader at IFF Pharma Solutions. Joshi agrees that single coprocessed excipients can provide not only good flow,
compressibility, and fast-disintegration, but also excellent taste and a pleasant, non-gritty mouthfeel.

Coprocessed excipients are also increasingly being used to convert more complicated, labor-intensive formulation processes such as wet granulation into quicker direct-compression processes without affecting the final product quality or performance, according to Joshi. He also notes that coprocessed excipients are being used to improve the efficiency of continuous manufacturing operations given that only a limited number of feeder ports are typically available and it is crucial to precisely control the flow-rate of excipients as they are added.

Developers of coprocessed excipients can theoretically use any material in their products, but most commonly, excipients that already have pharmacopeial monographs are selected because these excipients have a history of being used in pharmaceutical product formulations, according to Zhang. The key when selecting different ingredients to use in a particular coprocessed excipient is to choose those materials that will provide the functionality required to address customer needs, Zhang observes. Coprocessed excipients can improve drug processability and/or drug product performance. “Mainly excipient developers are looking to solve particular customer problems or provide particular benefits, and excipients are chosen that can maximize the needed functionality,” she says.

Ideally, adds Muley, the materials used to create coprocessed excipients will also have synergistic interactions that lead to enhancements in functionality, so that the coprocessed excipient will not only simplify formulation, but provide improved performance. Joshi adds that in addition to having complimentary and/or synergistic functionalities, the individual excipients chosen should be amenable to use in commonly available commercial processes (agglomeration, spray-drying, etc.) and behave reproducibly under the processing conditions. Ingredients used in coprocessed excipients also need to be nonreactive and inert so they do not undergo any chemical changes during production, according to Muley.

Typically, one or more fillers are coprocessed with binders, glidants, or disintegrants depending on the intended purpose of the product, according to Diogo Gomes Lopes, product development scientific team leader, Softgel and Oral Technologies at Catalent. “Coprocessed excipients that are intended to overcome any manufacturing limitations of the API, such as poor compressibility or poor flowability, are often composed of fillers, which are coprocessed with binders or glidants,” he explains. These types of coprocessed excipients also have the potential to increase drug loading in the drug product, allowing for production of a smaller dosage form that is more acceptable to the patient, Gomes says.

Some coprocessed excipients specifically aim to improve the performance of the drug product. Here again, Gomes points to coprocessed excipients for the manufacture of ODTs. “These coprocessed excipients allow for immediate disintegration of the tablet after administration while providing sufficient mechanical strength for packaging and transport,” he notes.

The materials chosen for coprocessing can include both small molecules, such as sugars and esters, and polymers such as polyvinyl acetate copolymers, microcrystalline cellulose (MCC), and starch. Coprocessing such varied materials together can lead to improved functionality, including greater processing properties as well as improved content uniformity, palatability, stability, and enhanced sustained-release performance.

Processing methods provide intimate mixing

Several methods are used to manufacture coprocessed excipients. The one feature they all have in common, according to Zhang, is that they involve physical processing of two or more excipients such that they interact with each other on a micro level, providing intimate mixing.

The most common methods include spray drying, flash drying, co-extrusion, granulation, high-shear granulation, fluid bed granulation, and roller compaction. “All of these processes are well known and are standard operations in pharmaceutical manufacturing,” Assis underscores. The key advantage is that under the mild conditions—moderate temperature and pH values, for example—chemical reactions between the ingredients can generally be excluded. It is important, he emphasizes, to determine the potential for incompatibilities between the different excipients used that could lead to chemical changes and avoid any combinations that might present reactivity concerns.

One exciting new method that could play a role in the production of coprocessed excipients, according to Muley, is 3D printing. Fused deposition modeling, he notes, is a good example of an additive manufacturing method very applicable to the production of coprocessed excipients. “The use of such methods is at the very preliminary stage, with exploration of the technology just beginning,” he says. Such approaches fit well with the push by FDA to adopt more efficient production technologies.

Although the development of coprocessed excipients does not involve combinatorial chemistry to impart many different chemical and molecular changes, there still are a large number of excipients to choose from. “Increasing functionality by modifying physical interactions only can present challenges, and pre-formulation studies using a design-of-experiment (DoE) strategy can be very helpful at the initial development stages,” Muley observes.

Computational methods can also be used to better support formulators during the development of drug products using coprocessed excipients, according to Silke Gebert, project engineer, BASF Pharma Solutions. For example, she points to ZoomLab, BASF’s Virtual Formulation Assistant for predicting and optimizing formulations using advanced algorithms and thereby expediting drug development. “Marketed coprocessed materials, new products, and prototypes can be characterized using the ZoomLab-logic to generate a chart that identifies each material’s favorable and less favorable properties, such as particle size distribution, compressibility index, and flowability,” she explains.

The need to accelerate drug development and also increase the efficiency and reduce the cost of drug manufacturing is driving excipient suppliers to seek more comprehensive coprocessed product solutions. “The most significant development in the coprocessed field relates to the growing focus on all-in-one coprocessed excipients,” Gebert contends.

All-in-one excipients, Gebert explains, are strategically designed multifunctional materials containing all the functionalities required to produce tablets: diluent/filler, disintegrant, binder, and lubricant at a minimum. “These materials are processed to achieve optimal bulk density, flowability, blend processability, drug uniformity, excellent compressibility, fast disintegration, and outstanding product performance,” she says. The drug manufacturing process then only requires blending the API and the all-in-one coprocessed excipient, then compressing the blend.

For these all-in-one products, Assis notes that common fillers include lactose, mannitol, and MCC; binders are often PVA-polyethylene glycol copolymers, PVP, or pre-gelatinized starch; disintegrants can be crospovidone, croscarmellose, or starch glycolate; and lubricant choices include sodium stearyl fumarate and magnesium stearate.

“Each ingredient in these compositions needs to be expertly selected with the end-use in mind,” Assis stresses. Furthermore, the excipient manufacturing process needs to be designed specifically to reduce chemical and physical instabilities.

Continuous tableting processes will benefit the most from the use of all-in-one coprocessed excipients, says Assis. “In order to reduce variability and dosage accuracy issues during the continuous manufacturing process, it is crucial to have a highly stable and controlled feed rate. With an all-in-one coprocessed excipient, rather than five feeders for the API, binder, filler, disintegrant, lubricant, only two are required for the API and the all-in-one coprocessed material, affording a more straightforward process,” he explains.

In addition, because coprocessed excipients reduce blend variability, they can contribute to reduced sampling and analytical errors for continuous processes monitored using inline/online/atline sensors, according to Assis. “Process analytical technology is increasingly being used during formulation development as well as in manufacturing. Coprocessed excipients have been shown to be less sensitive to shear, and due to their coprocessed multicomponent nature, the blend has lower intrinsic variability. In effect, this leads to more accurate and robust PAT [process analytical technology] models that do not require extensive model maintenance and updates,” he observes.

Regulatory hurdles present biggest challenges

Beyond demonstrating that individual ingredients in coprocessed excipients have not undergone any chemical changes, there are a few other challenges to developing new coprocessed solutions. One, according to Zhang, is that coprocessed excipients have fixed ratios of the various ingredients and as a result they may not be as broadly applicable as simple blends of excipients. “The current solution for this challenge is to design the coprocessed excipient in a way so that it meets the majority of needs,” she says. They are also designed as fit-for-use.

Manufacturing coprocessed excipients that are truly highly functional is not easy either, Muley adds. “A deep understanding of the problem or issue that will be addressed by the coprocessed excipient and proper research and development efforts are both required to ensure that new coprocessed excipients are highly functional and produced using a robust manufacturing process,” he comments.

It is important, for instance, to understand and control the variability of the excipient’s material attributes, according to Dejan Lamešić, head of formulation and process development, Softgel and Oral Technologies with Catalent. “Doing so further supports pharmaceutical users in determining the potential impact of an excipient on a final drug product,” he says.

The higher cost of coprocessed excipients can be an issue as well, Lamešić notes. “Coprocessed excipients must demonstrate sufficient manufacturing process efficiency for customers,” he says. The uniqueness of coprocessed excipients, meanwhile, can be a concern for drug formulators from a security-of-supply perspective. “The excipients are usually only available from one supplier, which can increase supply risk and impact an innovator’s supply chain strategy,” Lamešić explains.

The biggest challenge to the adoption of coprocessed excipients, however, is a regulatory one. “Despite the significant positive feedback we have received about coprocessed excipients, there is a reluctance on the part of drug formulators to use these ‘novel’ excipients because of the higher risk they represent with respect to the regulatory approval process,” says Zhang.

The fact that many coprocessed excipients do not have an official monograph available is one of the major obstacles to their wider adoption in drug products, agrees Lamešić. “Although there have been developments in recent years, including the ongoing development of a general monograph on coprocessed excipients and guidance from the International Pharmaceutical Excipients Council on these products, there are still regulatory hurdles that need to be resolved,” he observes.

One of the fears surrounding the use of coprocessed excipients for formulators has been that regulatory agencies may request additional data to support the safety and efficacy of these materials, adds Joshi. He does note, however, that lately pharma companies have realized they cannot develop the drugs of tomorrow using traditional excipients that were developed decades ago.

In addition, Joshi points out that excipient manufacturers have put into place robust manufacturing and quality control mechanisms to ensure reproducible manufacture of high quality, coprocessed excipients. There is also significant regulatory support provided with Type IV drug master file (DMF) filings and well-documented analytical procedures. He also notes that FDA’s new pilot program to evaluate and approve new excipients before they are used in a drug filing is a significant positive step. “All these factors have significantly increased the acceptance and use of coprocessed excipients by the pharma industry, leading to their increasing use in FDA-approved formulations,” he concludes.

Continued growth with improved robustness

Despite these regulatory hurdles, the drive to reduce the time and cost for new drug development will continue to make coprocessed excipients attractive and lead to their growing use, according to Muley. The ability of these materials to help drug makers produce more robust oral solid dosage forms by overcoming manufacturing process challenges and reducing variabilities will also keep interest levels high, adds Assis. Growth will increase at an even greater rate if a regulatory pathway for coprocessed excipient approval is established, Zhang notes.

Cynthia A. Challener is contributing editor to 


Vol. 46, No. 6
June 2022
Pages: 24–26, 49

When referring to this article, please cite it as C. A. Challener, “Development of Coprocessed Excipients,” 

The right processes used with the right excipient combinations address evolving formulation needs.

Development of Coprocessed Excipients

Inhaled drug formulation involves consideration of multiple factors, including the properties of the API, the needs of the target patient population, the delivery requirements, and the choice of device. Predicting the performance of a specific API is, consequently, quite challenging. While some computational approaches can help, physical experiments provide the greatest insights.

Increasing applications for inhaled delivery

Historically, the inhaled route has been preferred for the local administration of drugs to treat pulmonary diseases. “Inhaled delivery has clear advantages because delivering the API directly to the intended target allows for more efficient delivery of less overall dose with maximized efficacy and minimized dose-related side effects when compared to simpler systemic dosing,” explains Mark Parry, technical director with Intertek Melbourn.

Inhaled delivery also provides an alternative route for systemic dosing with different pharmacokinetics when compared to traditional routes such as solid oral dose and is of increasing interest where rapid delivery and onset of action is the objective, Parry adds. APIs that have stability issues upon contact with gastrointestinal fluids or decreased bioavailability because of first-pass metabolism can also be considered for inhaled delivery, notes William Wei Lim Chin, manager of global scientific affairs at Catalent.

Today, this route of administration can also be used to deliver systemic drugs for indications such as diabetes, Parkinson’s disease, and for pain management, according to Chin. “The large surface area in the lungs and the thin-walled pulmonary vasculature allow transport of drugs into the bloodstream,” he explains.

Even large molecules such as peptides, antibodies, and various types of engineered proteins can now be delivered via the lungs, Chin observes. These molecules present considerable challenges for delivery via the traditional oral route. “As we see increasing importance of new modalities in the biopharmaceutical space coming through the drug development pipelines, inhalation provides an alternative to parenteral delivery of these active substances that may provide advantages both in terms of the pharmacokinetic behavior and the patient experience,” Parry states.

Indeed, one of the primary factors in determining if formulation for inhaled delivery is the right approach—and the specific type of inhalation device—for any given API is the needs of the target patient population. “If the device does not meet the needs of the patient population, the utility of the product will be severely limited,” observes Jennifer Wylie, director of analytical research and development at Merck.

For instance, while there is evidence that patients prefer inhaled products to injectable ones (1), inhaled devices are usually more complicated to use than oral or injectable dosage forms, Chin notes. “Strategies that include patient education or training can help to address these limitations, as can careful design or selection of the delivery device,” he says.

In another example. Wylie points to the fact that dry-powder inhalers (DPIs) cannot be used by children younger than five to six years of age. “If the indication being addressed has a significant pediatric population, it may therefore be necessary to select an alternate device that is more appropriate, such as a nebulizer,” she comments.

Parry adds that a nebulizer formulation for a soluble API could be the simplest route to develop and, if only used in a hospital setting for short-term treatment, may be well tolerated by the patient and make the most sense commercially. When looking at the management of chronic diseases such as asthma and chronic obstructive pulmonary disease (COPD), however, he comments that the benefits of multidose DPI and pressurized metered-dose inhaler (pMDI) products become more important. Furthermore, engaging in more complex development programs can make commercial sense when considering large patient populations.

In addition to considering the needs of the patient population, Chin points out that whether or not inhalation delivery is the correct choice depends on the desired site of action (local versus systemic), whether there is a need for rapid absorption, and the drug class, all of which must be balanced against the development costs associated with these relatively complex products.

Specific attributes of an API that can affect its suitability for inhalation delivery include its physicochemical and interfacial properties, the ease and robustness of its manufacturing process, its inherent stability and potential routes of degradation, the level to which it interacts with its biological target and how long it is retained in the lung, its local and systemic bioavailability, and its toxicity, according to Charles Evans, senior vice president of pharmaceutical development for MedPharm.

Other factors that are typically considered when selecting inhalation delivery as outlined by Marc Brown, co-founder of MedPharm and chair of the company’s scientific committee, include the intellectual property status; the funds available for formulation development; the main objective of the project (i.e., proof of concept or full commercial development); the delivery device; and the potential processing/production needs of the API, such as micronization, spray drying, freeze drying, coacervation, etc.

Because there are so many diverse classes of APIs considered for inhalation delivery, defining an ideal set of API properties is difficult. “With various platforms and production technologies available, almost any API candidate can be formulated for inhalation,” Chin observes. He does note, however, that certain overarching molecular properties will enable more successful delivery to the lung, such as the nature of the solid state, the level of permeability, and the specific surface area.

In fact, the ideal properties of an API do depend on the specific delivery device, according to Wylie. “Usually, a solution formulation is desired for a nebulizer, and therefore a stable molecule with adequate solubility is desired.For a dry powder inhaler, the ability to reduce its particle size to the appropriate size range is important,” she adds.

One of the first properties to consider for an API when deciding on a type of delivery, Parry agrees, is how soluble the material is, including related salt forms. Equally important is how practical particle engineering such as micronization or spray drying might be, which can be impacted by properties such as the melting point and any characteristic polymorphism. “These considerations will guide the feasibility and likely complexity of the different delivery options,” he says.

Brown adds that APIs with higher aqueous solubilities can be more applicable to nebulizer systems, including both portable and hospital units. “Nebulizer formulations can quite often be a faster route to clinical studies for proof of concept as they leverage the simplest approach,” he adds.

APIs that are soluble in more polar solvents such as alcohols, meanwhile, are often more suitable for MDIs, Brown says. Those with attractive interfacial properties, particularly post-micronization, can be appropriate for DPIs and suspension formulations. He also emphasizes that crystalline materials with good chemical and physical stability are preferred. A thorough understanding of the particle size and its aerodynamics is required for small-molecule APIs delivered via carrier-based or carrier-free DPIs, Chin adds. When an excipient carrier is involved, he stresses that particle-to-particle interactions, particularly the API morphology/shape and surface characteristics, are critically important.

The requirements for small and large molecules intended for inhalation formulations do differ somewhat. For small molecules, Chin notes that properties centered on the solid state of the drug (crystalline for stability purposes versus amorphous for solubility improvement), permeability, and particle size are critical drivers in influencing the drug absorption process in the lung. He adds that it has been reported that small-molecule inhalable drugs generally meet Lipinski’s Rule of Five, with a tendency to have slightly higher molecular weights, high polar surface areas, and lower lipophilicities (2).

For large molecules such as peptides, proteins, and antibodies, Chin observes that particle surface properties are probably more critical criteria for the disposition and retention of these molecules in the lung. Because of the structure of these large molecules, they are more amenable to formulation using a spray-drying process for dry-powder formulations.

“Most important,” Chin contends, “regardless of the type of API and delivery platform, is the ability to repeatedly and consistently deliver the drug molecules as an aerosol.”

Inhaled drug delivery is inherently complex requiring identification of compounds with an appropriate balance of lung retention, toxicity, and bioavailability according to Evans. Several studies have correlated an API’s physicochemical properties with the absorption of inhaled drugs, but Chin notes the current consensus is that there is a lack of reliability of in vitro methods to establish absorption and permeability at the lung level.

“For dry powder inhalers, the particle size of the API must be of a respirable size (usually considered below 5 microns).For any solution formulation (for example, a nebulized solution), the API must have adequate solubility in the proposed formulation. Other than that, it is difficult to predict the performance without any experimental data,” Wylie says.

The pulmonary biopharmaceutics classification system (pBCS), introduced in 2010 and inspired by the oral biopharmaceutics classification system (BCS), attempted to establish a framework by considering the physiology of the lung, drug physicochemical properties, and drug absorption (3). The pBCS, Chin explains, suggested that molecular size, lipophilicity, solubility, acid dissociation constant (pKa), protein binding, polar surface area, and number of rotatable bonds predict an API’s permeability across the lung epithelium.

“From the developability perspective, this was a turning point that generated considerable interest from academia, industry, and regulatory stakeholders,” Chin says, pointing to a workshop on the topic (4). Unfortunately, he observes that 12 years on, the predictive utility of the pBCS framework remains uncertain as there are still many factors other than physicochemical properties that affect drug disposition in the lung, including the variability in patient breathing patterns and compliance to treatment regimens.

On a positive note, MedPharm has found that some cell- and tissue-based models developed in-house at the company can be used to provide a means of comparing performance between APIs and devices, according to Evans.

Parry also points to significant research into better models for understanding the way APIs are deposited in the lungs and then dissolve, which could help improve in-vitro/in-vivo correlation. “While much of this work has been driven by a need to better support the development of generics, we now have an increasingly broad toolbox to understand and study the delivery of poorly soluble APIs that can aid the particle engineering and formulation development work in optimizing their performance,” he remarks.

Computational approaches for aiding inhalation formulation development are being investigated, but developing effective models is not easy. “Development of inhalation products relies heavily on the packaging and delivery device, not only the formulation, so understanding the behavior using computational methods can be very complex,” Brown observes.

Mathematical/computational models for predicting airflow and respiratory aerosol deposition have been reported in the literature, but Brown believes, like many such models, they have significant limitations due to the gross assumptions they require. He does expect the value of these models to increase over time, however, especially for large pharmaceutical companies with libraries comprising thousands of APIs and drug formulations available for screening.

One example is Cohesive Adhesive Balance (CAB), a screening tool for investigating the surface interfacial properties of secondary processed micronized APIs and their influence on blending dynamics, stability, and in vitro performance. “CAB directly quantifies the relative magnitudes of the cohesive and adhesive forces that govern formulation structure and ultimately product functionality,” Brown explains. It can be used throughout API process development, early-phase DPI formulation development, and scale-up.

Chin points to thermodynamic modeling approaches for predicting process parameters that affect the properties of spray-dried particles and computational fluid dynamics (CFD) simulations for understanding the complexity of device and formulation interactions, which he says are being used to improve the performance of DPIs. CFD has more recently been used to study air flow within devices and patients, while discrete element methods are used to simulate dispersion and aerosolization of particles, according to Brown.

Inhaled delivery development is primarily focused on getting efficient dosing of drug to the right area of the lung. “Particles larger than 5 microns may not travel very far into the lung, particles in the 3-5-micron range will tend to deposit in the larger airways, while sub-3-micron particles will reach the deeper lung; the desired target will depend on the API and the condition being treated,” Parry says.

Ultimately, therefore, developing effective inhalation formulations requires extensive physical evaluation and experimentation, Wylie agrees. “The API and its physicochemical characteristics can directly impact the performance of an inhaled product.It is very important to gain understanding of these properties throughout development,” she states.

In general, optimizing for the right particle size may have trade-offs in overall dose amounts delivered to the lung, so optimization for the overall product performance can be complex, observes Parry. “When dealing with solid APIs (DPI or suspension-based products), therefore, understanding the physical properties and behaviors will greatly inform the development objectives,” he says. Having early empirical data from well-designed feasibility and scoping studies is essential.

Typical experiments include solubility and stability studies, as well as extensive analysis to determine particle size, shape, and density; crystallinity, surface morphology, and polymorph content; and the molecule fingerprint of the API, according to Evans. Using these data, the best delivery routes are determined and preliminary formulations and devices are developed simultaneously.

“The specific approach for any given API and project depends on the data that [are] available,” Evans adds. For example, if particle-size data are available, then it is possible to assess whether the distribution will be suitable for an inhalation formulation or if micronization is required. However, he also notes that the potential for successful micronization is more difficult to predict without an understanding of the interfacial properties, which relate to the likelihood of agglomeration at smaller particle sizes. Solubility and stability data, meanwhile, can indicate whether a solution nebulizer system or MDI would be a more viable delivery platform.

Even though gathering experimental data is challenging, it is essential to the development of inhaled delivery formulations. Fortunately, evidence-based models do show promise for improving the development process, including proprietary in-house solutions as well as solutions accessible by the wider stakeholder community.

2. T.J. Ritchie, C.N. Luscombe, and S.J.F., 

4. Hastedt, et al. “Scope and Relevance of a Pulmonary Biopharmaceutical Classification System,” AAPS/FDA/USP Workshop March 16-17th, 2015 in Baltimore, MD. AAPS Open 2, 1 (2016).


Volume 46, Number 5
May 2022
Pages: 22–25

When referring to this article, please cite it as C. Challener, “Inhalation Formulation Development: Predicting API Behavior,” 

Inhalation formulations are complex, and empirical data are essential for realizing optimal solutions.

Inhalation Formulation Development: Predicting API Behavior

Drug products must be produced in a manner that ensures both their safety and efficacy. That requires consideration of the quality of all ingredients in a final formulation, including all excipients, whether they are functional or inert.

“Excipients are critical to the safety and efficacy of nearly every drug product, if not all drug products, regardless of route of administration,” says Joseph Zeleznik, technical director, North America at IMCD Pharma. They can impact drug substance solubility and permeability, stability, release rate, and absorption. Variability in excipients therefore needs to be part of the formulation risk assessment.

Evaluating how excipients affect performance is essential to ensuring drug product performance and achieving therapeutic success. Leveraging excipient supplier expertise facilitates pharmaceutical development processes. Good two-way communication contributes to the risk assessment and experimental design to obtain the enhanced understanding of formulation components and processing required quality-by-design (QbD) approaches.

Although not easy to do, it is essential because excipients can affect drug-product safety and efficacy.