Seven Trends Shaping the Future of Pharmaceutical Formulation Development

In recent years, the discovery and development pace in the pharmaceutical industry has accelerated, including development timelines shortening significantly and new treatments have been launched that exceed the number of approvals of any previous year (1,2). This shows, on the one hand, the ability of the industry to respond fast to immediate needs, with RNA-based vaccines and RNA-based therapies being recent examples, which were triggered by the pandemic situation. On the other hand, new treatment approaches and new active substances often come with new challenges, which results in a clear need for new formulation approaches and technologies rising to meet this demand.

There is a general trend observed for advanced manufacturing technologies across the pharmaceutical sector—with a high focus on connected and more efficient processes, such as continuous manufacturing, new technologies suitable for personalized and on-demand medicine (such as 3D printing), and an unceasing effort to provide solutions for challenging compounds in the pipeline (such as solubility enhancement approaches.) Digitalization, challenging compounds, and a fast pace lead to a trend for modeling, predictive approaches, and digital collaboration in the pharmaceutical industry (3). In addition, the increasing number of biomolecules in the pipeline (2) comes with additional and specific challenges, such as protein purification and stability.

This article highlights several of these fast-evolving trends and developments that have the potential to change the pharmaceutical landscape and offer exciting opportunities for future treatments.

Continuous manufacturing and moving from powder to tablet

Tablets are important and popular dosage forms. Besides advantages in handling, shipping, and storage, good patient compliance and easy self-administration make them attractive (4). Currently, most tablets are manufactured in a complex, multistep approach that is batch manufacturing. Each step—such as blending, tableting, and coating—is performed as a stand-alone process with subsequent quality inspection and potential storage or shipping steps (5). This lack of flexibility and agility may pose a potential risk to the overall healthcare system. Therefore, a paradigm shift can be seen from traditional batch manufacturing to continuous manufacturing (CM) of the drug product, which is supported by authorities such as FDA (6). The process refers to starting materials (i.e., APIs and excipients) that are continuously fed to the solid dosage (e.g., tablets) manufacturing line, whereas the finished product is continuously taken out of the system with online and real-time monitoring of quality. This process can be described as “powder to tablet.”

FDA authorized the switch from batch production to continuous manufacturing of Johnson & Johnson’s (J&J’s)/Janssen’s drug Prezista in 2016, which is seen as an important landmark decision. With this switch, J&J/Janssen reduced manufacturing and testing cycle time by 80%, cut waste by one third, and reduced the overall process from seven to two rooms (7).

In addition to these benefits, supply chains can be secured, product quality improved, and a leaner and faster process from development to commercial introduced. Challenges arise due to new expertise that must be established, new requirements for equipment, and the process overall (8).

Supplementary to these challenges, the process itself requires raw materials to exhibit certain qualities that may have been insignificant in batch processing. The ICH Q13 draft guideline on continuous manufacturing defines two critical quality attributes (CQAs) concerning the raw material (9). For blend and content uniformity, the bulk density of the primary excipient, as well as the particle size distribution of the drug substance, were identified (9).

Looking at the continuous manufacturing equipment, the simplest set-up includes loss-in-weight feeders, dedicated to excipients and APIs, a continuous blender to mix the raw materials, a possible second blender for the lubricant, and the rotary tablet press. This can be expanded to additional dry or wet granulation units as well as a coater. Looking at feeding, disturbances in the process can easily occur if the material—especially the excipient—accumulates, flows poorly, or shows triboelectric tendencies. During blending, low-dose API formulations can lead to challenges in blending homogeneity (8). Therefore, the application of the right excipients will play a major role in ensuring a smooth and robust process. Understanding the critical material attributes (CMA), targeting CQAs, understanding critical process parameters, and implementing a control strategy to ensure a robust process and deliver an environment for continual improvement are part of the quality by design (QbD) approach that can be directly implemented when deciding for a continuous manufacturing setup (10).

This QbD approach can also be used to maintain the desired quality. Continuous manufacturing enables the integration of quality assurance concepts, such as real-time release testing. In this process, the product quality is monitored via process analytical technology, which eliminates lab-based testing procedures—thereby ensuring the intended quality throughout the process through in-line testing (11).

The continuous manufacturing approach paves the way for straightforward tablet manufacturing and displays numerous advantages but also challenges. It will be interesting to see to which extent regulatory authorities will further support this future trend.

Modeling and predictive approaches for solid formulation

Within the scope of industrial digitalization, statistical learning and physics-based modeling approaches leverage the pharmaceutical development process by reducing experimental effort and time. Formulation and pre-formulation problems are especially challenging when it comes to computational pharmaceutics due to their multi-scale nature. While data-driven modeling can be applied to all scales, nanoscopic and microscopic scales require quantum mechanical considerations or molecular dynamics, respectively, whereas classical mathematical modeling is typically used for mesoscopic and macroscopic orders of magnitude (12). Basically, quantum mechanical calculations allow for the accurate description of the electronic structure of molecular entities. Subsequently, different implementations of density functional theory (DFT) allow for the deviation of various properties for liquids, emulsions, and crystals (13) as well as amorphous materials (14). For solid formulations comprising crystalline and amorphous states, different simulation approaches have been established in recent years focusing on drug-excipient interactions—e.g., drug-excipient compatibility (15), co-former and counterion selection for cocrystal (16), and salt and solvate prediction (17). Especially, the combination of a physics-based DFT approach combined with statistical thermodynamics like the conductor-like screening model for realistic solvents (COSMO-RS) has become a powerful tool for the evaluation of thermodynamic state functions and deduced properties—such as the enthalpy of excess, solubility, and patrician coefficients—which can be utilized in various fields of drug product development (18–20). Most recently, these developments have been extended by data-driven approaches involving statistical models and machine learning. For instance, in case of coamorphous formulations, a model has been generated to predict the amorphous nature of specific API-excipient combinations. The model combines features of COSMO-RS and a partial least squares discriminant analysis (PLS-DA) focusing on API-amino acid combinations as input data (21). In the context of solid dispersions, an online platform has been launched (PharmSD) aiming to optimize the in-vitro performance of API-polymer solid dispersions based on physical stability and dissolution data of (amorphous) solid dispersions mined from literature (22,23).

The challenges for data-driven approaches are the quality (e.g., consistency) and quantity of the data used to train a reliable model as well as the access to “negative” data. What is more, proper external validation of the developed models needs to be ensured using suitable figures of merit (24).

For the description of thermodynamic state diagrams of solid dispersions, the perturbed chain statistical associated fluid theory has become a useful tool to evaluate physical stability, especially when it comes to the identification of polymer-API mixing gaps and amorphous-amorphous phase separation in dependence on relative humidity (25-30).

In addition to the reduction of experimental screening effort, computational predictions may support the rational development of industrial unit operations; for instance, blending, milling, granulation, and compression techniques tailored to the powder flowability and tableting properties (31,32).

Additive manufacturing’s impact on production

Advanced manufacturing concepts are on the rise and will eventually disrupt traditional manufacturing concepts. Being strongly data-driven, digitalization will play an important role in the pharmaceutical industry. This evolution will be amplified by rapidly evolving digital technologies and the integration of machine learning concepts into production processes.

3D printing brings the opportunities for formulation developers to a new level. Its application can range from enabling personalized medicine for individual patient needs to the supply of clinical material during drug product development (33).

The design and the performance of a final dosage form can be adapted on the fly. As a highly digitalized process, the technology allows for self-optimization as well as automated formulation development. The prediction of key process parameters and formulation characteristics during drug development will be key to enhancing the efficiency of the process (34). The opportunity for decentralized manufacturing allows a fast reaction to individual regional demands and ensures high flexibility.

Currently, the technology landscape is broad, ranging from powder-based systems to melt-based approaches (see Figure 1). Whereas powder-based systems are aiming for larger-scale production within an industrial framework, dedicated approaches for local production of personalized medications are also being explored. To ensure a successful implementation, a holistic approach needs to be considered. Fundamental aspects of engineering, formulation knowledge, as well as QbD approaches need to be well understood and implemented during the development of novel technology concepts. As the high individualization of medications is not suitable for classical quality control or batch-release mechanism, there is a high demand for elaborated in-process controls and extended process analytical technology to confirm the quality of the final dosage form. Regulatory bodies need to be included in the early development stages to assure that appropriate quality control measures are established and requirements relevant to the approval process are already integrated early in the development phase. Dedicated 3D printing technologies are also creating an emerging demand for specialized and functional excipients to leverage the full potential of the individual technology concept, providing interesting new opportunities for excipient providers and pharmaceutical manufacturers alike.

Overcoming solubility challenges of poorly-soluble APIs

For oral drug delivery systems, solubility is one of the most important parameters when formulating a drug. The solute—in this instance, the API—needs to fully dissolve in the solvent, which is the fluid in the gastrointestinal system to reach the desired concentration in the systemic circulation. Otherwise, therapeutic plasma concentrations cannot be reached (35). This leads to potential risks regarding pharmacokinetics (PK). From 2016-2020, 228 drugs were approved by FDA with 74% being small molecule new molecular entities. Compared to older drugs, these molecules show higher molecular weights and are overall more hydrophobic. These factors can lead to low solubility, increased efflux, and elevated metabolism (36). Looking at the market, approximately 40% of marketed drugs are poorly soluble in water (37); but in perspective, this value is estimated for pipeline drugs to reach between 70% (38) and 90% (see Figure 2) (37, 39).

Today, the biopharmaceutical classification system is widely used to classify drug substances based on their aqueous solubility and intestinal permeability (40). Four classes are defined and can be seen in Figure 3. Drug substances categorized in class II and IV pose a major challenge in the formulation of a respective oral solid dosage form. Without any modification, these drugs would need to be administered in high doses to reach the therapeutic plasma concentrations resulting (e.g., in large tablets or the potential PK risks).

Strategies to overcome the poor solubility of drugs include physical and chemical modifications. Examples of chemical alterations are changes in the structure of the drug to reach respective salts or generate prodrugs. Physical modifications range from simple particle size reductions to the generation of solid dispersions (35).

Amorphous solid dispersions (ASDs) are enabling formulations to improve solubility. The stable crystalline lattice of the API needs to be disrupted to enhance solubility. The formation of the amorphous state can lead to increased apparent solubility and faster dissolution, but this higher energy form is less stable than the respective crystalline state of the drug (41). Therefore, this amorphous state needs to be stabilized, which is performed by an auxiliary material (e.g., a polymer matrix). The polymer plays a key role in protecting the drug from re-crystallization by reducing molecular mobility (42). The behavior of increased apparent solubility, leading to a supersaturated state by the amorphous form, is described as a “spring” process; whereas, subsequently, this state needs to be maintained, which is aided by the carrier material and referred to as the “parachute” part (43).

The crystalline lattice of the drug can be broken by converting it to a liquid state. Melting through heat or dissolving in a suitable solvent are the methods of choice. In the next step, the material is either rapidly cooled or dried, making hot melt extrusion and spray drying the methods of choice and accounting for most registered products (44).

Recently, the formation of ternary systems in ASDs, containing two synergistically acting polymers, has gained attention (45). However, not only ASDs play a great role in enhancing solubility, the application of surfactants and solubilizers is another method to aid this process. Additionally, the inclusion of complex formation using cyclodextrins (35) or drug encapsulation into mesoporous silica (37) are promising strategies.

Due to the expected rise in poorly soluble compounds, it will be interesting to see which approach proves to be most effective and what new concepts will appear.

Improving protein purification

The production of antibodies and fragment crystallizable (Fc)-fusion proteins involves several downstream processing unit operations. The widely used purification template—with protein A chromatography, virus inactivation at low pH, and subsequent ion exchange chromatography steps—is mostly able to remove impurities like aggregates, host-cell proteins, and viruses, which could affect the safety and efficacy of the product.

The low pH condition during elution in protein A chromatography, as well as during virus inactivation, may induce aggregation. Therefore, there is a need in the biopharmaceutical industry to define methods to reduce the risk of aggregation during low pH steps in downstream processing. Preventing protein aggregation during these unit operations, instead of removing the multimeric forms during subsequent polishing steps, would be a more efficient strategy.

Several strategies can be applied to minimize aggregation formation during downstream processing, including the selection of solution conditions and stationary phases (for chromatography) that prevent aggregation and stabilize proteins under the conditions required for purification (46). Excipients can be used as well—not only to minimize aggregate formation during the low pH steps in downstream processing but also to improve the purification performance in protein A chromatography (47).

Results from a recent paper by Stange et al. demonstrated the addition of polyol or sugar excipients, such as sucrose, trehalose, mannitol, and sorbitol, can prevent the generation of aggregate during the low pH elution or virus inactivation (47). Moreover, the addition of polyethylene glycol 4000 (PEG4000) during protein A chromatography with pH gradient elution led to sharper elution peaks, reduced pool volume, and enabled a greater reduction of host cell proteins (HCPs), which results in higher pool concentration and can lower processing costs in the following unit operation (e.g., a smaller tank needed for virus inactivation step) and shorten process times for steps such as sample loading in cation exchange chromatography. Furthermore, positive effects in both protein A chromatography and virus inactivation steps, such as higher HCP clearance, lower elution pool volume, and stabilization of the antibody can be achieved by the addition of excipient combinations of PEG4000 and sucrose without performing additional buffer exchange such as diafiltration between the two steps. It has been demonstrated that these excipients have no negative impact on virus reduction and the binding capacity of cation exchange chromatography. Therefore, there is no need for the removal of excipients during the purification process.

Overall, the addition of excipients can improve not only the purification performance in protein A chromatography but also the protein stability in virus inactivation without harming subsequent chromatographic steps. As such, the excipients or combination of those could also be applied for continuous downstream processing.

Highly concentrated protein formulations for subcutaneous application

Biologics are integral therapeutic drugs utilized in the therapy of distinct disorders (e.g., autoimmune diseases and cancer) (48). Monoclonal antibodies (mAbs) are an important class of biologics. As of 2021, 100 mAbs have been approved by FDA (49). Despite increasing interest and the variety of therapeutic applications, the dosage remains a major drawback for mAbs as a therapy. Usually, high doses of mAbs (multimilligrams/kg body weight) are administered either intravenously or subcutaneously. Even though most mAb applications are intravenous, subcutaneous administration assures better patient convenience and adherence due to fewer hospital visits, the possibility of self-administration by patients, and low healthcare costs. However, subcutaneous administration has one major limitation—namely, the low application volume (1.5-2 mL) (50,51). To prevent tissue back pressure and injection pain, highly concentrated protein formulations (HCPFs) are preferred in subcutaneous applications (52). HCPFs have a broad concentration range varying from > 50-200 mg/mL (53). At concentrations above 100 mg/mL, some proteins (e.g., mAbs) tend to undergo intermolecular protein-protein interactions (PPIs), possibly leading to high solution viscosity, higher susceptibility to aggregation, and lower physical stability (51,54). Concentrations above 200 mg/mL are more favorable for cluster forming, which result in the crowding of protein molecules. In this case, the distance between protein molecules is decreased, leading to direct influences on physical protein stability (i.e., unfolding, aggregation susceptibility, and viscosity enhancement).

Several methods to reduce physical instability and high viscosity of HCPF have been applied, varying from genetic manipulation of protein sequence to the addition of viscosity-reducing excipients. The latter is of great interest since the excipients can be added to the base formulation of HCPF without interfering with its biochemical properties. Therefore, excipients either directly interact with proteins or with solvents in protein formulations, which then affect protein stability. The excipient molecules can belong to a variety of chemical classes: amino acids, sugars, surfactants, salts, detergents, polymers, etc. (51). Among others, basic amino acids asserted themselves as suitable viscosity-reducing excipients. For instance, Qian and coworkers investigated the impact of distinct sodium salts and basic amino acids on the viscosity reduction of two IgG1 mAbs and found that, depending on the nature of the antibody, both excipient classes have concentration-dependent viscosity-reducing effects (55). Additionally, Inoeu et al.highlight specific amino acid side chain-driven interactions with aromatic functional groups in mAbs, resulting in the viscosity reduction of HCPF (56).

Despite these advances, one problem remains. Intermolecular PPIs and intramolecular stabilizing PPIs share the same molecular origin; thus, resulting in a potential destabilizing effect for some excipients, which directly interacts with proteins. In fact, concentration-dependent destabilizing effects for excipients have been shown in literature (57,58).

To overcome this issue, combinations of different excipients to turn the net PPI repulsive can be used. The first examples of combinations of salts and organic acids with amino acids have already been demonstrated (59–61). For instance, Kumar et al.showed that two excipients (basic and acidic molecules) in equimolar quantity (0.25 M), respectively, decreased gelation of gelatine, and casein at high concentrations and distinct pH values (60). However, the viscosity-reducing effect was exclusively attributed to the acidic component. The basic amino acid as a sole excipient did not prevent gelation (gelatine) or only for 24 h (casein), whereas the acidic excipient prevented gelation for a long period or even completely inhibited it for the proteins.

Even though a synergistic effect could not be ruled out, a more detailed presentation of the impact of combinations of distinct excipients on viscosity reduction of HCPF can be seen in a white paper by MilliporeSigma, the life science business of Merck KGaA, Darmstadt, Germany (61). It demonstrates that combinations of amino acid-based excipients with organic acids/esters such benzenesulfonic acid (BSacid) and thiamine phosphoric acid ester chloride dihydrate (TMP)—as well as basic vitamin B6 salt pyridoxine hydrochloride (Pyr-HCl)—resulted in a synergistic viscosity reducing effect on two model antibodies at high concentrations, infliximab and evolocumab, respectively. Additionally, the structural integrity was assessed, and it was shown that combining destabilizing viscosity-reducing excipients with amino acids yields an improved monomer content in a forced degradation study. However, it should be noted that some excipients are temperature-sensitive, and studies are conducted at high temperatures, while stable at 2-8°C (62,63). Thus, forced degradation studies may not be appropriate to assess the impact on protein stability since a reacting molecule may introduce artifacts. In such cases, accelerated storage studies would give more reliable insights.

To summarize, the field of mAb therapeutics is constantly growing, and the subcutaneous application route is a favorable and convenient method for patients (64,65). Providing patients with subcutaneous mAb formulations that can be self-injected or received in the familiar environment of a family physician can be achieved via HCPF formulations with the right choice of excipients. Constant efforts to reduce the viscosity of HCPF, while retaining structural and functional integrity, is a cumbersome yet desirable task—as replacing intravenous administrations improves patient convenience, which is, in turn, related to patient compliance and ultimately potential treatment success.

In this section, these efforts with a special focus on distinct excipient combinations are highlighted, leading to the conclusion that the future of highly-concentrated mAb therapeutics is bright and will ensure a more patient- and healthcare-friendly approach to battle a variety of diseases.

mRNA technology: status and future potential

Messenger RNA (mRNA) holds promise against unmet clinical needs. Within cells, mRNA is transcribed from DNA and then translated into proteins. Exogenous mRNA can encode for antigens (vaccination) or for a protein of interest (gene editing and protein therapy). Exogenous mRNA resembles natural occurring mRNA, it is single-stranded and includes the protein-coding open reading frame marked by start and stop codons, which is flanked by untranslated regions with a 5′ cap and a 3′ poly(A) tail. In recent years, substantial efforts and advancements have been made to improve the stability, translation, and safety of mRNA. Some examples of this are nucleoside-modified, sequence-optimized, or circular mRNA (66).

mRNAs are large, negatively-charged, hydrophilic molecules, which are highly susceptible to nuclease degradation, and delivery vehicles are required to stabilize and protect mRNA from endonucleases, improve cellular uptake by specific tissues, and promote endosomal escape of mRNA for further protein translation. A variety of carriers have been explored to improve mRNA translation and safety, such as lipid nanoparticles (LNPs), liposomes, polyplexes, protamine conjugates, nanoemulsions, and other nanoparticles (67).

The most promising and clinically-advanced platform for mRNA delivery is the LNPs, which are the delivery vehicles of the two marketed COVID-19 mRNA vaccines: SpikeVax by Moderna and Comirnaty by Pfizer/BioNTech. LNPs are composed of four main components: 1) the ionizable cationic lipid, which interacts with mRNA and fosters the endosome escape; helper lipids, such as 2) phospholipid and 3) cholesterol, which provide membrane integrity; and 4) a PEG-lipid, which provides steric hindrance (see Figure 4A).

Historically, the most common method used to load lipid-based formulations with RNA payload was the hydration of a lipid “thin film” with the aqueous RNA solution. This approach came with several pitfalls, primarily, the generation of a heterogeneous particle population that required homogenization by extrusion, low encapsulation of RNA, and a challenging scaling up. In the early 2000s, a scalable and extrusion-free method was developed for the production of RNA-loaded lipid formulations, based on spontaneous vesicle formation by ethanol dilution (68). This technique entails the rapid mixing of an ethanolic solution of lipids with an aqueous solution of RNA at acidic pH, using either a microfluidic or a T-tube mixer (69,70). The resulting LNP dispersion is then further purified to remove organic solvents (e.g., ethanol) and is introduced to a normal saline buffer, either by dialysis or tangential flow filtration approaches to allow injection to patients (see Figure 4B).

An important aspect related to the clinical translation of LNPs are their storage (71). By adding cryoprotectants (such as sucrose, trehalose, or mannitol) to the LNP formulation suspension, the LNP can be preferably stored at freezing temperatures (below -70oC) to ensure the integrity of the RNA payload. However, this requirement compromises the accessibility of the universal distribution of LNP-based vaccines and therapeutics, and it is a topic that became increasingly relevant due to the recent COVID-19 pandemic. On that note, both authorized COVID-19 mRNA vaccines are stored in the presence of sucrose in a frozen state. To circumvent this issue, several studies have suggested recently that lyophilization is a promising approach to increase the stability of RNA-loaded LNPs (72) and allows storage at room temperature. Still, thorough process development is required, taking into consideration the stresses generated during the process, which can lead to loss of activity and aggregation of the particles. To that end, Pfizer has recently initiated a Phase III study to compare the lyophilized Comirnaty formulation to its equivalent frozen liquid formulation in terms of safety and tolerability (NCT04816669).

Upon systemic administration, LNPs predominantly accumulate in the liver. LNPs are typically sub-100 nm in size, carry a neutral charge, and exhibit a high affinity towards the binding of apolipoprotein E (ApoE), which further mediates their uptake by hepatocytes using ApoE-low-density lipoprotein receptor pathway. Therefore, initial clinical studies for RNA-LNP as therapeutics primarily focused on treating liver diseases. However, in recent years, various strategies have been developed that enable RNA delivery to extrahepatic tissue(73). This opens the potential of RNA-based therapies beyond vaccines and liver-specific diseases and for the treatment of a variety of unmet diseases.

Additional approaches to enhance treatment options, by mediating a tissue- and cell-specific delivery in vivo, include attaching targeting ligands (such as antibodies, small molecules, or peptides) to the surface of lipid nanoparticles. Next to this, a significant scientific effort is being invested in the rational improvement of the lipidic component design and the addition of charged components to the LNP in order to modulate the nanoparticle composition and its physicochemical properties—and, therefore, tune the interaction between nanoparticles and biomolecules (e.g., proteins) (74). This will eventually determine the in-vivo biodistribution and fate of the nanoparticle (75) and further enable treatment for a broad range of emerging diseases—thus, hopefully improving health care in the near future.

Future outlook

Despite being often perceived as slow and cautious, the pharmaceutical industry has picked up considerable speed. Successful approaches in the presented development areas include the use of novel lipids in marketed COVID-19 vaccines and the application of predictive tools to improve development timelines and effectiveness.

However, it is still a barrier to innovation that novel approaches and novel excipients face considerable regulatory hurdles and are associated with an increased risk for the drug manufacturer. It is encouraging that FDA started a dedicated program to review novel excipients before they are used and submitted as part of a drug formulation (76). This is especially relevant for applications currently lacking suitable options within the existing excipient offering.

Another example of encouraging initiatives of FDA as a regulatory agency is the advanced manufacturing initiative that supports the transition from classical batch manufacturing to more advanced manufacturing approaches (e.g., continuous manufacturing, portable manufacturing platforms, and the use of artificial intelligence in drug manufacturing) (77). This initiative and other ones like it could lessen the risk of manufacturers using novel technologies/excipients and enable excipient providers to bring more innovative technologies into the market. The contribution of excipient providers and their support of drug manufacturers in identifying suitable excipients and solutions targeting specific formulation challenges and new technologies is essential. Additionally, their strong support in providing the information required for approval helps to reduce effort and risk on their customers’ end.

The coming years will prove a turning point in the pharmaceutical sector with many game-changing technologies on the rise. Clearer guidelines by regulatory bodies and close collaboration of excipient suppliers and drug product manufacturers are key to accelerating progress, allowing the development of novel solutions, easing the effort to bring these innovative drug products into the market, and unlocking the innovation potential in the field.

References

1. T. Yang, S. Shah, and C. Chang, “The Future of Biopharma:, Reimagining Traditional Business Models in 2040,” www2.deloitte.com, (March 23, 2020).

2. IQVIA Institute for Human Data Science, Global Trends in R&D 2022: Overview Through 2021(Feb. 10, 2022).

3. Frost & Sullivan, Global Pharmaceuticals Outlook, 2022 (March 8, 2022).

4. S. Dai et al., Int. J. Pharm. DOI:10.1016/j.ijpharm.2019.118742 (Oct. 21, 2019).

5. M.S. Arshad et al., Adv. Drug Deliv. Rev.,DOI:10.1016/j.addr.2021.113840 (November 2021).

6. S.L. Lee et al., J. Pharm. Innov.,DOI:10.1007/s12247-015-9215-8 (September 2015).

7. J. Wechsler, Pharmaceutical Technology40 (2) 14-16 (Feb. 2, 2016).

8. J. Wahlich, Pharmaceutics 13 (8) 1311-1341 (Aug. 22, 2021).

9. EMA, Draft ICH Guideline Q13 on Continuous Manufacturing of Drug Substances and Drug Products – Step2b(July 29, 2021).

10.L.X., Yu et al., AAPS J. 16 (4) 771-83 (May 23, 2014).

11. D. Markl et al., Int. J. Pharm., DOI:10.1016/j.ijpharm.2020.119353 (April 20, 2020).

12. W. Wang et al., J. Control. Release,DOI:10.1016/j.jconrel.2021.08.030 (Aug. 19, 2021).

13. J. Hoja et al., Sci. Adv., DOI: 10.1126/sciadv.aau3338 (Jan. 11, 2019).

14. B. Schammé et al., Mol. Pharm.15 (3) 1112–1125 (2018).

15. N. Wang et al., Int. J. Pharm., DOI:10.1016/j.ijpharm.2021.120962 (July 31, 2021).

16. C. Loschen and A. Klamt, J. Pharm. Pharmacol. 67 (6) 803–811 (April 7, 2015).

17. C. Loschen and A. Klamt, Pharm. Res.33 (11) 2794–2804 (July 28, 2016).

18. C. Loschen and A. Klamt, Ind. Eng. Chem. Res.53 (28) 11478–11487 (2014).

19. C. Loschen and A. Klamt, “New Developments in Prediction of Solid-State Solubility and Cocrystallization Using COSMO-RS Theory” in Computational Pharmaceutical Solid State Chemistry, Y. A. Abramov, Ed., pp. 211-233., (John Wiley & Sons, Inc., Hoboken, NJ, 2016).

20. A. Pozarska et al., Eur. J. Pharm. Sci. 49 (4) 505–511 (April 2013).

21. L. I. Chambers et al., Eur. J. Pharm. Sci.,DOI:10.1016/j.ejps.2020.105636 (Feb. 1, 2021).

22. J. Dong, H. Gao, and D. Ouyang, Int. J. Pharm.,DOI:10.1016/j.ijpharm.2021.120705 (July 15, 2021).

23. R. Han et al., J. Control. Release,DOI:10.1016/j.jconrel.2019.08.030 (October 2019).

24. M. A. Lones, arXivLabs, DOI:10.48550/arXiv.2108.02497 (accessed September 2022).

25. K. Lehmkemper et al., Mol. Pharm. 14 (1) 157–171 (Jan. 3, 2017).

26. C. Luebbert and G. Sadowski, Int. J. Pharm. 532 (1) 635–646 (Oct. 30, 2017).

27. K. Lehmkemper et al., Mol. Pharm.14 (12) 4374–4386 (2017).

28. C. Luebbert, F. Huxoll, and G. Sadowski, Molecules22 (2) 296 (Feb. 15, 2017).

29. C. Luebbert, M. Wessner and G. Sadowski, Mol. Pharm. 15 (2) 669-678 (2018).

30. K. Lehmkemper et al., Eur. J. Pharm. Biopharm., DOI: 10.1021/acs.molpharmaceut.6b00763 (January 2017).

31. P. Pandey and R. Bharadwaj, Predictive Modeling of Pharmaceutical Unit Operations(Woodhead Publishing, 2017).

32. Y. A. Abramov, G. Sun, and Q. Zeng, J. Chem. Inf. Model. 62 (5) 1160–1171 (2022).

33. I. Seoane-Viaño et al., Adv. Drug Deliv. Rev.,DOI:10.1016/j.addr.2021.05.003 (July 2021).

34. M. Elbadawi et al., Trends Pharmacol. Sci. 42 (9) 745–757 (July 5, 2021).

35. K. T. Savjani, A. K. Gajjar and J. K. Savjani, International Scholarly Research Notices, DOI: 10.5402/2012/195727(2012).

36. A. Avdeef and M. Kansy, J. Solution Chem., DOI:10.1007/s10953-022-01141-7 (September 2022).

37. T. Loftsson and M. E. Brewster, J. Pharm. Pharmacol. 62 (11) 1607–1621 (November 2010).

38. A. Budiman and D. L. Aulifa, Pharmaceuticals 15 (6) 730 (2022).

39. J. M. Ting et al., Bioconjug. Chem. 29 (4) 939–952 (April 18, 2018).

40. V. P. Shah and G. L. Amidon, AAPS J.16 (5) 894–898 (Sept. 16, 2014).

41. D. H. Kim et al., Pharmaceutics13 (8) 1318 (Aug. 23, 2021).

42. Y. He and C. Ho, J. Pharm. Sci. 104 (10) 3237–3258 (October 2015).

43. H. R. Guzmán et al., J. Pharm. Sci.,DOI: 10.1002/jps.20906 (Oct. 1, 2007).

44. S. V. Bhujbal et al., Acta Pharm. Sin. B. 11 (8) 2505–2536 (August 2021).

45. F. Pöstges et al., Int. J. Pharm.: X, DOI: 10.1016/j.ijpx.2022.100115 (December 2022).

46. A. A. Shukla et al., BioProcess Int.,3 36-44 (January 2005).

47. C. Stange et al., J. Chromatogr. B,DOI: 10.1016/j.jchromb.2021.122848 (Aug. 1, 2021).

48. R. Lu et al., J. Biomed. Sci. 27 (1) 1–30 (Jan. 2, 2020).

49. A. Mullard, Nat. Rev. Drug Discov. 20 (7) 491–495 (July 2021).

50. C. Narasimhan, H. Mach, and M. Shameen, Ther. Deliv., 3 (7) 889–900 (July 2012).

51. G. R. Strickley and J. W. Lambert. J. Pharm. Sci. 110 (7) 2590–2608 (July 2021).

52. M. Holstein et al., Biotechnology and Bioengineering117 (11) 3591–3606 (November 2020).

53. P. Garidel et al., Eur. J. Pharm. Biopharm. 119 353–360 (October 2017).

54. E. Binabaji, J. Ma, and A. L. Zydney, Pharm. Res. 32 (9) 3102–3109 (September 2015).

55. S. Wang et al. Mol. Pharm. 12 (12) 4478–4487 (Dec. 7, 2015).

56. N. Inoue et al., Mol. Pharm. 11 (6) 1889–1896 (June 2, 2014).

57. S. V. Thakkar et al., J. Pharm. Sci. 101 (9) 3062–3077 (September 2012).

58. M. J. Robinson et al., Mol. Pharmaceutics, DOI:10.1021/acs.molpharmaceut.8b01024 (Feb. 4, 2019).

59. MilliporeSigma Patent GmbH, “Viscosity reduction of highly concentrated protein formulations,” US Patent 11207412B2, 2019.

60. A. Kumar and A. M. Klibanov, Appl. Biochem. Biotechnol. 182 (4) 1491–1496 (August 2017).

61. S. Braun et al., MilliporeSigma, "The Viscosity Reduction Platform: Viscosity-Reducing Excipients for Protein Formulation," Whitepaper, September 2021.

62. A. Schnellbaecher et al., Biotechnol. Bioeng. 116 (6) 1537–1555 (June 2019).

63. A. L. Voelker, L. S. Taylor, and L. J. Mauer, BMC Chem. 15 (1) 47 (August 2021).

64. J. Jin et al., Patient Prefer Adherence, DOI:10.2147/PPA.S87271(July 2015).

65. M. Viola et al., J. Control. Release, DOI:10.1016/j.jconrel.2018.08.001 (Sept. 28, 2018).

66. R. Verbeke et al., Nano Today, DOI:10.1016/j.nantod.2019.100766 (October 2019).

67. Y. Yan et al., J. Control. Release, DOI:10.1016/j.jconrel.2022.01.008 (February 2022).

68. E. Samaridou, J. Heyes, and P. Lutwyche, Adv. Drug Deliv. Rev., DOI:10.1016/j.addr.2020.06.002 (June 8, 2020).

69. L. B. Jeffs et al., Pharm. Res. 22 (3) 362–372 (March 2005).

70. M. J. W. Evers et al., Small Methods, DOI:10.1002/smtd.201700375 (2018).

71. A. J. Barbier et al., Nat. Biotechnol. 40 (6) 840–854 (May 9, 2022).

72. H. Muramatsu et al., Mol. Ther. 30 (5) 1941–1951 (May 4, 2022).

73. D. Loughrey and J. E. Dahlman, Acc. Chem. Res., DOI:10.1021/acs.accounts.1c00601 (Jan. 4, 2022).

74. X. Hou et al., Nature Reviews Materials6 (12) 1078-1094 (Aug. 10, 2021).

75. J. Simon et al., Cells 10 (1) 132 (Jan. 12, 2021).

76. FDA, “Novel Excipient Review Pilot Program,”www.fda.gov, accessed August 2022.

77. M. Kopcha and P. Cavazzoni, “FDA’s Advanced Manufacturing Initiatives Helping to Provide Quality Human Drugs for Patients,” www.fda.gov, accessed August 2022.

About the authors

Can Araman is the principal scientist, biomolecule formulation; Moritz Beck-Broichsitter is the associate director, formulation service; Nelli Erwin is the head of protein stability; Supriyadi Hafiz is the senior scientist formulation; Adela Kasselkus is the technical communication manager; Thomas Kipping is the head of drug carriers; David Luedeker is the principal scientist, solid formulation; Aditi Mehta is an associate director, mRNA process and delivery; Lena Mueller is the head of functional excipients; Sara S. Nogueira is a scientist, RNA drug delivery; Tobias Rosenkranz is the head of biomolecule formulation; Eleni Samaridou is theprincipal scientist formulation service; and Johanna Simon is the principal scientist formulation service, all for MilliporeSigma, the life science business of Merck KGaA, Darmstadt Germany.