Developing Multi-Component Solid Formulation Strategies for PROTAC Dissolution Enhancement

Abstract

PROTACs are an emerging class of beyond-rule-of-5 molecular drugs currently under clinical investigation for the treatment of malignant diseases and are capable of degrading previously “undruggable” protein targets. They are poorly crystallizable due to their structure, consisting of two ligands joined chemically by a flexible linker, yet the inherent insolubility of their amorphous phases hinders their development into sufficiently bioavailable medicines. Formulation approaches to improve the dissolution properties of PROTACs are required as a result, but research in this area is made even more challenging by the scarcity of available samples. In this work, amorphous solid dispersion (ASD) formulations of four cereblon-recruiting PROTACs ‘AZ1–4’ using hydroxypropyl methylcellulose acetate succinate (HPMCAS) as a polymer excipient are described. ASDs of AZ1 show up to a 2-fold increase in drug supersaturation compared to the pure amorphous API, observed up to a drug loading of 20% w/w. Preparing the ASDs by slurry conversion offers greater solubility enhancement over those prepared by solvent evaporation and maintains the dissolution advantage up to a higher drug load. Positive deviations from theoretical T_g values coupled with a lack of spectral evidence of drug-polymer hydrogen-bond interactions suggest that the ASDs may differ from ideal mixtures via predominantly dispersive drug-polymer interactions. ASDs that provide a dissolution enhancement were stored at elevated temperature and humidity for one month and showed no sign of plasticization or loss of physical stability. Coamorphous formulations using low-molecular-weight excipients, by contrast, showed no dissolution advantage despite evidence of drug–coformer hydrogen-bonding interactions. This work demonstrates that ASDs may be an effective strategy for improving PROTAC bioavailability and producing commercializable solid forms for oral administration despite the lack of well-behaved solid phases of PROTACs. It also highlights the need for a deeper understanding of how to develop successful formulation approaches for bRo5 compounds.

Introduction

Proteolysis targeting chimeras (PROTACs) are an emerging class of molecular drugs in a chemical space beyond Lipinski’s rule of 5 (bRo5), (1) designed to degrade target disease-causing proteins previously considered “undruggable” by conventional small molecule drugs. Unlike traditional small molecule inhibitors, PROTACs are capable of degrading a protein by harnessing the ubiquitin-proteosome system, rather than only blocking its enzymatic function by binding to the enzymatic site. (2) This affords PROTACs the benefits of lower dosing potential, since they act via an event-driven mechanism and do not have to remain bound to the protein of interest to have a therapeutic effect. However, their large size and structural complexity present challenges regarding their cell permeability, solubility, and pharmacokinetic properties. Very poor aqueous solubility impedes their development into drug products with sufficient oral bioavailability, (3) and with numerous ionizable atoms, the conventional processes of pharmaceutical salt and cocrystal screening are made much more complicated. (4)

Multicomponent solids, which can offer advantageous chemical and physical characteristics compared to a pure active pharmaceutical ingredient (API) without modifying its chemical structure or pharmacology, (5−8) usually consist of two molecular components bonded by noncovalent interactions such as hydrogen-bonding, π–π stacking, van der Waals forces or halogen-bonding, and as such they are often designed using an understanding of the possible supramolecular motifs that can form between the two components. (5) The resulting solids can be either crystalline or amorphous. The latter possess no long-range order but often show improved solubility and/or dissolution rate compared to their cocrystalline counterpart through their higher free energy. (8−10)

Compared to an inherently unstable pure amorphous drug, coamorphous formulations are often less likely to recrystallize into the crystalline drug form and lose their dissolution advantage. (11) The glass transition temperature (Tg) of the resulting amorphous solid is usually between that of the two compounds, which would suggest greater molecular mobility and lower kinetic stability than the single component with the higher Tg; however, the coamorphous phase is usually further physically stabilized by intermolecular interactions between components. This is in part because the formation of strong heterodimer interactions can disrupt short-range molecular ordering of one component, where the formation of homodimers often precedes recrystallization and the loss of the dissolution advantage provided by the coamorphous phase. (12) Strong intermolecular interactions between components in a coamorphous solid can usually be identified by the Tg value deviating from the theoretical Tg calculated via the Gordon–Taylor equation, (13−16) which predicts the Tg of a homogeneous coamorphous solid assuming no specific interactions between components (ideal mixing) and ideal free volume additivity of the two components. (12) The greater physical stability that they impart allows the more soluble and faster dissolving amorphous solid to persist for a longer duration in contact with solution, maintaining supersaturation of the API for a longer duration before recrystallization. This type of dissolution kinetics is often referred to as a “spring and parachute.” In some systems, such as coamorphous naproxen/cimetidine, heterodimer formation involving hydrogen bonds has also been shown to cause synchronous dissolution of the individual components, as well as enhancing the dissolution rate. (17) Hence, producing a coamorphous formulation involving strong noncovalent bonds between a poorly soluble drug that has a low dissolution rate and a more soluble coformer may facilitate dissolution improvement of the low-solubility API, even compared to its own pure amorphous form.

Coamorphous formulations using low-molecular-weight excipients can include non-pharmaceutically active, generally recognized as safe (GRAS) compounds. Alternatively, a second drug molecule can be used to provide a dual-action pharmaceutical formulation, such as coamorphous indomethacin/paracetamol, (18) naproxen/cimetidine, (17) and indomethacin/naproxen, (19) which all show greater physical stability and enhanced dissolution rates as well as evidence of intermolecular interactions between components. Other drug–drug coamorphous solids, such as simvastatin/glipizide (20) and ritonavir/indomethacin, (21) show dissolution and stability improvements without any evidence of specific intermolecular interactions. However, in both of these cases, while the dissolution rate of one drug component was increased compared to its pure amorphous form, the second component did not see such an increase due to the lack of synchronous dissolution. Coamorphous formulations containing polymer excipients instead of small molecules are usually referred to as amorphous solid dispersions (ASDs), in which the stabilization of the amorphous API depends on its solubility in the carrier polymer. (22) Below the miscibility limit, the drug is molecularly dispersed and stabilized against recrystallization by physical separation of the drug molecules between polymer chains. (22) The polymer matrix usually raises the Tg of the ASD compared to the pure amorphous drug, inhibiting crystallization through a reduction of molecular motion to impart greater physical stability, (23) and it has been shown that differences in the types of intermolecular interactions between a given polymer and drug affect both the ASD dissolution performance and the maximum drug-loading (DL) capacity. (9,23,24) Rapid dissolution of drug from an ASD produces a supersaturated solution with an enhanced free drug concentration that can exceed even the amorphous API solubility in some systems; in these cases, the drug separates from the bulk aqueous phase via liquid–liquid phase separation to produce colloidal nanoparticles, which act as a reservoir to maintain the elevated drug concentration. (9,24,25) This behavior is generally at lower drug loadings where drug and polymer are released congruently at the rate of polymer dissolution─the limit of this congruent release is referred to as the “LoC” of the ASD formulation. The method used to prepare ASDs with acidic polymers has been shown to have a strong impact on the degree of proton transfer between drug and polymer, particularly for combinations of drug and polymer that differ in pKa by several log units, where the greater charge separation between drug and polymer causes greater aqueous solubility. (26) ASDs of the malaria drug lumefantrine and poly(acrylic acid) (PAA) produced by slurry conversion showed 70% protonation of the drug by the polymer in the resulting solid, whereas the same ASDs prepared by melt quenching showed only 20% protonation and had 6-fold lower apparent solubility in simulated gastric fluid. (26) On the downside, ASDs are also quite often hygroscopic and, with water ingress, can cause a reduction in Tg (plasticization) that can lead to phase separation of the coamorphous solid and drug recrystallization. (27) Furthermore, the DL of many ASDs is unable to exceed 25% w/w due to the limited miscibility of the drug with polymers, meaning that large volumes of polymer are required to produce the final dosage forms. Co-amorphous formulations with low-molecular-weight excipients are usually produced at a 1:1 molar ratio, by comparison. (28) Designing an ASD formulation with optimized polymer selection and drug loading also relies solely on experimental information and cannot currently benefit from in silico screening protocols such as COSMOtherm, (29) which aid in the development of low-molecular-weight multicomponent solids. (30)

While there are no reports in the literature of PROTAC coamorphous formulations using low-molecular-weight coformers, there are several recent examples of successfully applying ASD formulations to enhance the solubility of PROTAC compounds. Pöstges et al. first produced ASD formulations of an initially amorphous androgen-receptor PROTAC “ARCC-4” at 10% and 20% DL with hydroxypropyl methyl cellulose acetate succinate (HPMCAS) and Eudragit polymers by vacuum compression molding, with nonsink dissolution studies showing a pronounced supersaturation enhancement without drug precipitation. (31) Hofmann et al. also demonstrated significant supersaturation enhancements for spray-dried ASDs of an initially amorphous cereblon (CRBN) PROTAC with Soluplus and Eudragit polymers at 10% DL, compared to the pure amorphous API. (32) Mareczek et al. later demonstrated dissolution enhancements for spray-dried ASDs of both an initially semicrystalline PROTAC “ARV-110” and an initially amorphous PROTAC “SelDeg51” with poly(vinyl alcohol) (PVA) at 30% DL, indicating that the ASD formulations using the crystalline API were physically stable for at least 4 weeks. (33) Most recently, Zhang et al. studied ASDs of a CRBN PROTAC with HPMCAS, Eudragit and Soluplus prepared at 5, 10, and 20% DL by solvent evaporation, confirming the presence of drug-polymer hydrogen-bonding interactions using FTIR and showing that HPMCAS ASDs could be produced as high as 40% DL, although ASDs at higher drug loading showed poor dissolution performance. (34) They also showed, however, that the limited dissolution enhancement of the high DL ASDs could be improved greatly by adding sodium dodecyl sulfate as a surfactant to produce a ternary ASD system. (34) However, these studies did not show how the dissolution enhancement of a given PROTAC ASD varied with the ASD preparation method used and, with the exception of work by Mareczek et al., whether the selected preparation method could be applied successfully to more than one PROTAC compound. Furthermore, the previous studies have not investigated the slurry conversion method for producing ASDs, which was shown by Neusaenger et al. to produce lumefantrine–PAA ASDs with the highest degree of drug protonation and therefore the greatest aqueous dissolution enhancement compared to other methods. (26) In this work, we compare the dissolution behavior of coamorphous formulations of a CRBN PROTAC “AZ1” using either small molecule or polymeric excipients and investigate the nature of noncovalent interactions between drug and excipient components. We also investigate the effect of polymer selection and DL and ASD preparation methods on the dissolution performance of AZ1 ASDs and establish that the results are general across four structurally distinct CRBN PROTACs.

Download the full article as PDF here Developing Multi-Component Solid Formulation Strategies for PROTAC Dissolution Enhancement

or read more here

Materials

PROTAC compounds “AZ1–4” were supplied by AstraZeneca and were prepared according to previously published methodology. AZ1 amorphous form A was prepared by neat milling of AZ1 powder as synthesized using a Retsch MM200 Mixer Mill for 20 min at 20 Hz in a stainless steel milling jar. Hydroxypropyl methyl cellulose acetate succinate (HPMCAS-LG–AQOAT) was supplied by Shin-Etsu Chemicals (Tokyo, Japan). Polyvinylpyrrolidone vinyl acetate (PVPVA–Kollidon VA64) was supplied by the BASF Corporation (Ludwigshafen, Germany). All other chemicals and solvents were available from commercial sources and used without further purification. Infrared spectra were recorded between 4000 and 550 cm^–1 using a PerkinElmer 100 FT-IR spectrometer with a μATR attachment. Unless otherwise specified, powder X-ray diffraction (XRPD) patterns were collected at room temperature using a Bruker AXS D8 Advance GX003410 diffractometer with a Lynxeye Soller PSD detector, using Cu Kα radiation at a wavelength of 1.5406 Å and collecting from 2° ≤ 2θ ≤ 40°.

Martin A. Screen, Sean Askin, James F. McCabe, Esther Jacobs, Akosua Anane-Adjei, Clare S. Mahon, Mark R. Wilson, and Jonathan W. Steed, Developing Multi-Component Solid Formulation Strategies for PROTAC Dissolution Enhancement, Molecular Pharmaceutics Article ASAP, DOI: 10.1021/acs.molpharmaceut.5c01107