Using Flash Nanoprecipitation To Produce Highly Potent and Stable

---

Subscriber access provided by TUFTS UNIV

Article

Using Flash Nanoprecipitation to Produce Highly Potent and Stable Cellax Nanoparticles from Amphiphilic Polymers derived from Carboxymethyl Cellulose, Polyethylene Glycol and Cabazitaxel Joseph Bteich, Simon A. McManus, Mark J Ernsting, Mohammed Z. Mohammed, Robert K Prud'homme, and Kenneth K. Sokoll Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/acs.molpharmaceut.7b00670 • Publication Date (Web): 25 Sep 2017 Downloaded from http://pubs.acs.org on October 5, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Molecular Pharmaceutics is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 42

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Using Flash Nanoprecipitation to Produce Highly Potent and Stable Cellax Nanoparticles from Amphiphilic Polymers derived from Carboxymethyl Cellulose, Polyethylene Glycol and Cabazitaxel AUTHORS Joseph Bteich1†, Simon A. McManus3†, Mark J. Ernsting1,2, Mohammed Z. Mohammed1, Robert K. Prud’homme3* and Kenneth K. Sokoll4* 1

Drug Delivery and Formulation, Drug Discovery Program, Ontario Institute for Cancer Research, MaRS Centre, West Tower, 661 University Avenue, suite 510, Toronto, Ontario, Canada, M5G 0A3

2

Faculty of Engineering and Architectural Science, Ryerson University, Toronto, Ontario, Canada, M5B 1Z2

3 Department of Chemical and Biological Engineering, Princeton University, Princeton, New Jersey 08854, United States 4 Fight Against Cancer Innovation Trust, MaRS Centre, West Tower, 661 University Avenue, suite 510, Toronto, Ontario, Canada, M5G 0A3 †

These authors contributed equally

Page 1 of 32 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT:

We report the use of Flash Nanoprecipitation (FNP) as an efficient and scalable means of producing Cellax nanoparticles. Cellax polymeric conjugates consisting of carboxymethyl cellulose functionalized with PEG and hydrophobic anticancer drugs, such as cabazitaxel (coined Cellax-CBZ), have been shown to have high potency against several oncology targets, including prostate cancer. FNP, a robust method used to create nanoparticles through rapid mixing, has been used to encapsulate several hydrophobic drugs with block co-polymer stabilizers, but has never been used to form nanoparticles from random copolymers, such as Cellax-CBZ. To assess the potential of using FNP to produce Cellax nanoparticles, parameters such as concentration, mixing rate, solvent ratios, and subsequent dilution were tested with a target nanoparticle size range of 60 nm. Under optimized solvent conditions, particles were formed that underwent a subsequent rearrangement to form nanoparticles of 60 nm in diameter, independent of Cellax-CBZ polymer concentration. This intra-particle relaxation, without inter-particle association points to a delicate balance of hydrophobic/hydrophilic domains on the polymer backbone.

These particles were stable over time, and the random amphiphilicity did not

lead to inter-particle attractions, which would compromise the stability and corresponding narrow size distribution required for parenteral injection. The amphiphilic nature of these conjugates allows them to be processed into nanoparticles for sustained drug release and improved tumor selectivity. Preferred candidates were evaluated for plasma stability and cytotoxicity against the PC3 prostate cancer cell line in vitro. These parameters are important when assessing nanoparticle safety and for estimating potential efficacy, respectively. The optimal formulations showed plasma stability profiles consistent with long circulating nanoparticles, and cytotoxicity comparable to free CBZ. This study demonstrates that FNP is a promising technology for development of Cellax nanoparticles.

KEYWORDS: Nanoparticle, Nanocarrier, Flash Nanoprecipitation, Carboxymethyl cellulose, Cabazitaxel, Polyethylene Glycol, Conjugated Polymers

Page 2 of 32 ACS Paragon Plus Environment

Page 2 of 42

Page 3 of 42

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

INTRODUCTION

CBZed-Nano™ is a novel polymer conjugate formulation consisting of a hydrophobic drug (Cabazitaxel) covalently conjugated to a cellulose polymer modified with polyethylene glycol (termed a Cellax™ polymer) that assembles into well-defined nanoparticles. Docetaxel (DTX) was the original hydrophobic molecule selected to develop with this technology and several proof of concept in-vivo efficacy studies have been published to date exhibiting the utility of the approach 1-4. Promisingly, biodistribution studies have shown that the use of Cellax nanoparticles results in increased concentration of the drug in tumor in several cell lines, presumably due to esterase activity hydrolyzing DTX from the Cellax backbone2. With Cellax NP treatment, DTX concentrations above the IC50 of DTX were maintained for more than 10 days. A generic scheme summarizing the chemistry and formulation steps is provided in Figure 1.

Page 3 of 32 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1. (A) Generic synthetic scheme; Cellax polymer is synthesized from an acetylated carboxymethyl cellulose polymer (CMC-Ac), a hydrophobic drug (e.g. Docetaxel or Cabazitaxel) and polyethylene glycol (PEG). (B) Cellax polymers can be engineered to self-assemble within a biologically relevant range of sizes, by various processes that induce nanoprecipitation.

Page 4 of 32 ACS Paragon Plus Environment

Page 4 of 42

Page 5 of 42

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Similar to DTX, cabazitaxel (CBZ) is a semisynthetic tubulin-binding taxane derivative, however unlike DTX, CBZ exhibits poor substrate affinity for P-glycoprotein (P-gp) an ATP-dependent efflux pump which is over-expressed in DTX-resistant tumors, including prostate cancer 5, 6. Clinical development of CBZ is limited by concerns regarding its severe dose-limiting cytotoxicity 7-9 and, thus, this study focused on evaluating improved process methods to formulate nanoparticles with the goal of enhancing the targeting of CBZ to prostate tumors, increasing efficacy and reducing side effects.

Translation of nanomaterials from the laboratory to GMP production at scale has been a significant challenge. Scott McNeil, the head of the National Cancer Institute's Nanotechnology Laboratory recently commented: "(a) big hurdle in developing nanomedicines is scaling up the synthesis of the particles to meet Good Manufacturing Practice standards required for the clinic. …developing a synthesis that yields particles with those precise properties on a consistent basis… is still a difficult process." (C&ENews. ACS.org, June 20, 2016, p. 19). The process we present for the processing of CBZed-Nano™ is Flash NanoPrecipitation (FNP), which addresses this barrier to translation. FNP has been shown to be a robust and scalable method for creation of nanoparticles with low polydispersity and is amenable to a variety of materials 10, 11. The process typically involves dissolving hydrophobic molecules of interest and amphiphilic block co-polymers in a water-miscible organic solvent and rapidly mixing (on the order of 1 ms) with an antisolvent, usually water. This rapid mixing leads to uniform nucleation and growth of precipitates of the hydrophobic molecules. Growth of the nanoparticles is halted by binding of the hydrophobic blocks of the amphiphilic copolymers to the particle surface and the hydrophilic block acting as a steric barrier to Page 5 of 32 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

prevent further aggregation. This process has been used to encapsulate therapeutic compounds, imaging agents, and combinations to produce theranostic nanoparticles 12. The amphiphilic nature of the Cellax polymer makes it an interesting candidate for FNP. The unimolecular composition of hydrophobic drug and PEG appended to the same backbone makes the assembly process different from those previously studied. In those studies the hydrophobic compound precipitated independently from the attachment of the amphiphilic stabilizing polymer. For Cellax-CBZ polymers the stabilizing PEG and nucleating hydrophobic drug are forced to assemble cooperatively and stoichiometrically. Possible outcomes of FNP using Cellax-CBZ include the rapid precipitation of the Cellax-CBZ, leading to trapping of both the PEG chains and the drug in the core, or sufficient internal mobility within the nanoparticles to allow PEG chains to rearrange to the particle surface. The role therefore of precipitation kinetics (antisolvent conditions), rate of nanoparticle growth (governed by total solute concentrations) and polymer composition on size and stability requires clarification.

A key objective of this work is screening the suitability of the nanoparticles manufactured by the FNP process for potential use in efficacy studies. An assessment of the potency of the nanoparticles is provided by in-vitro cytotoxicity studies determining the IC50 of free CBZ versus CBZed-Nano™ formulations in PC3 cells whereas plasma stability studies provide indications of safety including reduced toxicity and improved tolerability.

Page 6 of 32 ACS Paragon Plus Environment

Page 6 of 42

Page 7 of 42

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

EXPERIMENTAL SECTION

Cellax-CBZ Polymer Conjugate Synthesis. Cellax-CBZ polymer is synthesized from an acetylated carboxymethylcellulose polymer (CMC-Ac), drug (CBZ), and polyethylene glycol (PEG). A variant of the method used to prepare the prototype Cellax polymer comprising DTX and PEG 13 was synthesized using an optimized ratio of coupling agents (2:4:2 molar equivalence of EDC.HCl:NHS:DMAP). The synthesis reaction was executed stepwise with PEG (0.15mol equivalence) feed first, followed by the addition of CBZ (0.75 mol equivalents) at 5 hours after the addition of PEG, with stirring at RT. The reaction was quenched after 24 hours by precipitation in MTBE followed by repeated washing procedures. For maximum recoveries the process was optimized employing MTBE (2X) followed by a water (2X) to remove residual couplings agents and unreacted starting materials.

Polymer Composition and Dose Concentration by qNMR. Cellax-CBZ polymer and CBZed-Nano™ formulations used in IC50 cytotoxicity studies were analyzed by qNMR (Bruker, 500 MHz NMR) for composition1 using 2-methyl-5-nitrobenzoic acid (Sigma, 638307) as internal standard, and dimethyl sulfoxide-d6 as solvent providing the wt% of CBZ and PEG respectively. Nanoparticle drug concentrations were also characterized using qNMR, to provide the corresponding dose of CBZ after undergoing lyophilization, dissolution with internal standard solution and sonication. Residuals amounts of unconjugated CBZ and PEG are determined by LC/MS.

Page 7 of 32 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

CBZed-Nano™ Formation by Multiple Inlet Vortex Mixing. Cellax-CBZ polymers were precipitated by FNP into CBZed-Nano™ using a two-stream multi-inlet vortex mixer (MIVM) as described previously 14. Cellax-CBZ was dissolved in acetonitrile at 10-30 mg/g and loaded into a 25 mL glass syringe. Normal saline was loaded into a 100 mL glass syringe and used as the antisolvent. Both syringes were driven by separate Malvern syringe pumps to allow individual control of the solvent and antisolvent flow rates, which were varied as described for each experiment. After initiation of FNP, the first 10 mL of the resulting solution was discarded, followed by collection, and 0, 1, or 2 fold dilution in normal saline, depending on experiment. Particles were stored at 22 or 4oC depending on the experiment.

Dynamic Light Scattering Measurements. Particle sizes were determined by dynamic light scattering (DLS) using a Malvern Nano ZS instrument at room temperature. Samples were diluted with normal saline until water-clear and measured at a scattering angle of 173 degrees with a 632 nm helium neon laser. Reported sizes are the average of three measurements. The measurements shown are the mean of three measurements of 10 runs each. Reported Z-average diameters are the intensity-weighted diameters obtained from dynamic light scattering measurements and PDI is the polydispersity index obtained from the cumulants fitting program, and reported by the DLS instrument15.

TEM Experiments. Transmission electron microscopy (TEM) was carried out on unprocessed FNP nanoparticle samples on a ZIESS LEO Omega 912 energy filtered transmission electron microscope equipped with a 7.5 megapixel Hamamatsu Orca

Page 8 of 32 ACS Paragon Plus Environment

Page 8 of 42

Page 9 of 42

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

EMCCD camera. Carbon-coated copper grids were first glow discharged under vacuum. The CBZed-Nano™ solution was added to the grid for 1 minute and wicked off. The grid was then stained with 0.2% uranyl acetate solution for 20 sec, following by wicking. Samples were then imaged at 90 kV at 12.5K times direct magnification. TEM was also carried out on processed (i.e. dialyzed, sterilized, and concentrated) nanoparticle lots on a Hitachi H-7000 TEM using an acceleration voltage of 75 kV and a direct magnification of 70 000x. Images were captured on an Advanced Microscopy Technology XR60 CCD camera. Samples were prepared in similar fashion.

Nanoparticle Dose Preparations. CBZed-Nano™ lots were dialyzed three times against 2L of sterilized normal saline, for 2 hours, then overnight and finally 2 more hours, using Slide-A-Lyzer™ G2 10 kDa MWCO cartridges. Once dialyzed, samples were sterile filtered by first filtering via 0.45 µm Millex HP PES 33 mm filter, followed by a 0.22 µm Millex GV PVDF 33 mm sterile filter. Once sterile filtered, nanoparticle lots were concentrated using Vivaspin® 10 kDa MWCO PES centrifugal concentrator tubes. Concentrated samples were then transferred to sterile Eppendorf tubes and stored at 4°C. All doses were handled in a BSC using sterile/aseptic technique.

Cell Proliferation Assay. The ATPLite proliferation assay was performed according to manufacturer recommendations. 1000 cells per well were seeded into ViewPlate-96 Black from PerkinElmer® and cells were incubated overnight at 37° C with 5% CO2. After an overnight incubation, cells were treated with varying concentrations of CBZedNano™ formulations over 72 hours. On day 3, plates were left to equilibrate to room

Page 9 of 32 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

temperature for 15-30 minutes. Using a multichannel pipette 125µl of ATPLite reagent was added to each well and the top and bottoms of plates were sealed and then shaken on an orbital shaker for 3 minutes. The total luminescence was measured using an Envision plate reader equipped with a US-Luminescence detector. The results were then analyzed using prism 5.0 Graphpad software. A non-linear regression analysis was used to fit a Hill curve on the data points using sigmoid dose-response.

The PC3 (Prostate, ATCC® CRL-1435™) cancer cell line was grown in DMEM culture media supplemented with 0.06 g/l penicillin, 0.1 g/l streptomycin (Life Technologies 15140-122) and 10% fetal bovine serum (ThermoFisher Scientific, Canada). Only cultures with a plating efficiency of over 80% were used for the analysis.

Plasma Release Assay. The analysis of drug release from Cellax-CBZ nanoparticles in mouse plasma (Balb/c, Na-heparin) at 370C over 3 days was carried out by LC/MS (Agilent 1200 HPLC System /AB SCIEX QTRAP 5500) in conjunction with a Waters Xbridge C8, 3.5m, 2.1x100mm column or equivalent, and d6-CBZ (Alsachim) as an internal standard. Released CBZ and plasma were separated by mixing with a solution of MeCN+0.5% formic acid, followed by centrifugation. The resulting supernatants were analyzed and sample concentrations were compared to a calibration curve (range: 0 µg/mL to 100 µg/mL CBZ) and quantified. Each time point was sampled in triplicate.

Page 10 of 32 ACS Paragon Plus Environment

Page 10 of 42

Page 11 of 42

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

RESULTS

Polymer Characterization / Properties. The Cellax-CBZ polymer analogue produced was characterized by MALS (multi-light angle scattering, Wyatt Helios II), operated in batch mode and was found to have a molecular weight of 8.8 kDa. Cellax polymers possess variable hydrophobic and hydrophilic domains and do not effectively elute in size exclusion columns. Thus characterization via traditional GPC chromatography was not applicable. Assays for residual unconjugated CBZ and PEG, which may be present post purification were completed employing LC/MS and found to be below detection limits. Cellax-CBZ polymer composition was characterized by qNMR, a technique described in detail for a Cellax-DTX analog in earlier publications 1. For this study a Cellax-CBZ polymer with compositions of 29.5 wt% CBZ, 11.7 wt% PEG and 58.8 wt% CMC-Ac was selected.

Flash Nanoprecipitation of Cellax-CBZ Polymer into CBZed-Nano™. Particle sizes can be influenced by many factors. The parameter space chosen was based on previous experience with both FNP and formulation of CBZed-Nano™ by manual mixing. Table 1 outlines the parameters tested. For FNP, two mixing geometries can be used. The Confined Impinging Jet (CIJ) mixer geometry 10 requires a 1:1 ratio of the organic and aqueous antisolvent streams. This limits the range of supersaturations that can be achieved. In contrast, the Multi Inlet Vortex Mixer (MIVM) 14 enables variable ratios of the organic and aqueous streams, and is, therefore, more amenable for studies of solvent quality on nanoparticle formation.

Page 11 of 32 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 42

Table 1. Parameters Tested for CBZed-Nano™ Formation by FNP

Flow rate ratio (mL/min)

Dilution ratio (saline:product)

Total Flow Rate (mL/min)

2:1

3:1

4:1

0

1:1

2:1

14

Cellax-CBZ concentration (mg/g)

25

10

50

20

70

30

The first two parameters tested involved the composition of the solvent mixtures during and immediately after FNP. During FNP with two input streams, rapid solvent mixing results in a homogenous solvent mixture dictated by the relative flow rates of the two streams. Different solubilities of Cellax-CBZ polymer in different solvent mixtures leads to varying amounts of supersaturation and nanoparticle growth rates. FNP was carried out at flow rate ratios of 2:1, 3:1, and 4:1 saline to acetonitrile and the nanoparticle size was recorded over time by dynamic light scattering (DLS). 10 mg/g Cellax-CBZ in the acetonitrile stream was chosen based on previous work with Cellax-CBZ dropwise precipitation that showed low variability at this concentration and previous observations that FNP with other systems in this concentration range produces nanoparticles in the desired size range.

As seen in Figure 2, immediately after FNP, nanoparticles made with different flow rate ratios followed a trend of decreasing size with higher aqueous to organic flow rate ratios; 4:1 ratios producing 102 nm nanoparticles, 3:1 producing 150 nm nanoparticles, and 2:1

Page 12 of 32 ACS Paragon Plus Environment

Page 13 of 42

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

producing large particles that were too large to be measured by DLS. The increasing size with increasing aqueous phase ratio is consistent with smaller particles being produced at higher supersaturations. Greater aqueous phase concentration (antisolvent concentration) creates higher supersaturation of the hydrophobic components, and high supersaturation leads to higher nucleation rates, which lead to smaller nanoparticle sizes.

Interestingly, nanoparticles produced using 3:1 and 4:1 ratios and no post-formation dilution decreased in size over time (solid lines in Figure 2A and 2B), with 3:1 shrinking from a diameter of 150 nm to 100 nm, and 4:1 from 100 nm to 60 nm. In contrast, particles that were diluted after FNP either 1x or 2x in saline, did not change in size over time when made with 4:1 or 3:1 flow rate ratios at either dilution, or 2:1 with 2x dilution. The one diluted formulation that showed reduction in nanoparticle size was obtained using a 2:1 flow rate ratio and 1x dilution, which decreased in size from 180 to 60 nm. Looking for commonalities between the formulations that reduce in size over time, they all have similar final solvent compositions between 25 and 16.5% acetonitrile. In contrast, formulations with lower final acetonitrile concentrations, such as 2x diluted samples with 9% or less acetonitrile did not show reduction in nanoparticle size over time.

This decrease in size with time has not been observed previously in FNP processing. Previous studies have always used small hydrophobic core materials and amphiphilic diblock copolymers. With adequate hydrophobicity to control precipitation, particles always had a fixed size after assembly. For a few systems, Ostwald ripening occurred

Page 13 of 32 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

over time 16, where particle size increases as smaller particles dissolve due to the higher interfacial energy of highly curved surfaces, and larger particles grow. This process involves core material transferring from one particle to another – an inter-particle mass transfer process. The particle decrease in size observed for the CBZed-Nano™ formulations, by as much as 80%, is new. Its origin comes from the rapid precipitation process that captures the amphiphilic chains by a kinetically arrested, diffusion limited process. Rubenstein and Semenov 17, 18 have modeled polymers networks of associating polymers with “stickers”, which correspond to the hydrophobic drug in the CBZ. At high sticker interaction energy the gel rearrangement of stickers is prevented and the network is arrested in the state in which it was kinetically formed. This state is not the lowest free energy, but is kinetically trapped. At lesser sticker interaction energies the stickers can rearrange over time to relax towards the lowest free energy state, which means a smaller, denser nanoparticle. The value of the interaction energy depends on the hydrophobicity of the cabazitaxel group (“stickers”) and on the polarity of the solvent phase. It appears that acetonitrile concentrations between 16.5 and 25 vol% provide the correct polarity to allow the nanoparticles to rearrange to the lowest energy state, to produce nanoparticles on the order of 60 nm. Once relaxed to smaller size, the particles have reached their thermodynamic minimum energy and are size stable. For example, the formulation made with a 4:1 flow rate ratio and no further purification retains the same particle size and polydispersity index for over 3 months, as shown in Figure 5D and Table 2. As formulations made with 4:1 aqueous:organic flow rate ratio relaxed in size, did not require further dilution, and were stable at the desired particle size of 60 nm, this flow rate ratio was used in subsequent optimization of other parameters.

Page 14 of 32 ACS Paragon Plus Environment

Page 14 of 42

Page 15 of 42

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Figure 2. Flow rate ratios and dilutions. DLS measurement over time for CBZedNano™ produced at (A) ratios of saline:solvent stream of 4:1, (B) 3:1, and (C) 2:1 with no dilution, 1x dilution and 2x dilution. The formulation with 2:1 with no dilution produced large flocculates/aggregates that could not be measured by DLS. (D) DLS distribution of 10 mg/g Cellax-CBZ with 4:1 flow rate ratio and no dilution over time, showing the size to stabilize to 60 nm, at 22oC.

An unexpected dependence of nanoparticle size on mixing intensity was also found. We tested the effect of total flow rate on CBZed-Nano™ formation by FNP. The flow rate of the solvent and antisolvent streams dictates the speed of mixing, with higher mixing flow rates resulting in faster mixing. We tested mixing at total flow rates of 14, 25, 50, and 70 mL/min (corresponding to Reynolds numbers of 1075, 1920, 3841, and 5377, respectively) using 4:1 aqueous:organic flow rate ratio and 10 mg/g Cellax-CBZ in the Page 15 of 32 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

organic feed stream. For each formulation, the size was monitored over time by DLS. As shown in Figure 3a, the initial size of the nanoparticles is flow rate dependent, with higher Reynolds numbers resulting in smaller nanoparticles. Over time however, CBZed-Nano™ particles produced at all four Reynolds numbers all relax in size to ~60 nm, as shown in Figure 3b. Interestingly, the particles made under poorer mixing conditions still trap the same mass of polymer during the aggregation process. But all particles relax to the same final small, dense size because rearrangement occurs under these solvent conditions. This insensitivity of nanoparticle size to mixing rates is a highly beneficial feature of FNP processing under these solvent conditions, and polymer composition.

Figure 3. Effect of flow rate (A) Using 4:1 flow rate ratio, 10 mg/g Cellax-CBZ show that initial size is dependent on flow rate. (B) The final size of 60 nm is independent of flow rate between 14-70 mL/min total flow rate. 11.2 mL/min saline, 2.8 mL/min CHCN, Re = 1075, 20 mL/min saline, 5 mL/min CHCN, Re = 1920 40 mL/min saline, 10 mL/min CHCN, Re = 3841, 56 mL/min saline, 14 mL/min CHCN, Re= 5377

As producing nanoparticles at higher concentrations reduces the processing volumes necessary to produce dosage forms, we tested whether CBZed-Nano™ of 60 nm could be Page 16 of 32 ACS Paragon Plus Environment

Page 16 of 42

Page 17 of 42

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

produced using higher concentration of Cellax-CBZ in the feed stream. We doubled and tripled the concentration to 20 mg/g and 30 mg/g Cellax-CBZ and performed FNP under the parameters that were optimized above (4:1 flow rate ratio and 70 mL/min total flow rate). As shown in Figure 4, the nanoparticles produced at higher concentrations behaved similarly to those made with 10 mg/g Cellax-CBZ. Both initially formed larger particles, around ~100 nm, that decreased in size and stabilized at 60 nm. This behavior is similar to that seen with small molecule surfactants. Small molecule surfactants have a preferred packing parameter that dictates the sizes of the final micelle they form. Going to higher concentration produces more micelles, but the sizes are nearly constant above the critical micelle concentration (CMC). Once again, the CBZ formulation seems to

Figure 4. Final relaxed size is concentration independent between 10-30 mg/g. (A) Size distribution over time for 20 mg/g Cellax-CBZ. (B) Size distribution over time for 30 mg/g Cellax-CBZ. display a preferred “packing parameter” that drives its equilibrium surface-to-volume ratio and size. That this occurs for the chemically complex CBZ derivatized cellulose is

Page 17 of 32 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

unexpected, but it is highly desirable because it makes nanoparticle size insensitive to processing concentrations, when formulated via FNP. As formulations with higher concentrations are desirable to reduce the need for downstream concentrating during manufacturing, the feed concentration was kept at 30 mg/g for subsequent tests and optimization.

We tested the effect of temperature, both before and after particle relaxation. CBZedNano™ were made (4:1 flow rate ratio, 70 mL/min total flow rate, 30 mg/g Cellax-CBZ) with one portion being immediately transferred to 4oC while another was left at 22oC. As shown in Figure 5a, in contrast to the same formulation at 22oC, the fraction at 4oC initially underwent particle growth to 180 nm and remained stable at that size. To test the influence of temperature on CBZed-Nano™ formulations that have already relaxed to 60 nm, particles that had already stabilized at 60 nm were placed at 4oC. Under these conditions the particles remained at 60 nm for at least 6 months (Figure 5b and Table 2), demonstrating that once relaxed these particles can be stored at refrigerated temperatures for prolonged periods of time and remain size stable.

Page 18 of 32 ACS Paragon Plus Environment

Page 18 of 42

Page 19 of 42

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Figure 5. Effect of temperature on relaxation. (A) Relaxation only occurs at 22oC (not 4oC). (B) Once relaxed, particles are stable at 60 nm at either 4OC or 22oC. Particles shown were relaxed for 16 h at 22oC and transferred to 4oC. These particles are stable for at least 6 months.

To investigate the morphology of CBZed-Nano™ before and after relaxation, transmission electron microscopy was performed. As shown in Figure 6, particles imaged immediately after FNP were a mixture of large (>100 nm) and smaller particles.

Figure 6. TEM of Cellax-CBZ nanoparticles (A) Immediately after FNP. (B) 3 days after FNP. Page 19 of 32 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

In contrast, images taken 3 days after FNP shows all particles were of smaller size (