Formulation and Scale-up of Delamanid Nanoparticles via Emulsification for Oral Tuberculosis Treatment

Delamanid (DLM) is a hydrophobic small molecule therapeutic used to treat drug-resistant tuberculosis (DR-TB). Due to its hydrophobicity and resulting poor aqueous solubility, formulation strategies such as amorphous solid dispersions (ASDs) have been investigated to enhance its aqueous dissolution kinetics and thereby improve oral bioavailability. However, ASD formulations are susceptible to temperature- and humidity-induced phase separation and recrystallization under harsh storage conditions typically encountered in areas with high tuberculosis incidence. Nanoencapsulation represents an alternative formulation strategy to increase aqueous dissolution kinetics while remaining stable at elevated temperature and humidity. The stabilizer layer coating the nanoparticle drug core limits the formation of large drug domains by diffusion during storage, representing an advantage over ASDs. Initial attempts to form DLM-loaded nanoparticles via precipitation-driven self-assembly were unsuccessful, as the trifluoromethyl and nitro functional groups present on DLM were thought to interfere with surface stabilizer attachment. Therefore, in this work, we investigated the nanoencapsulation of DLM via emulsification, avoiding the formation of a solid drug core and instead keeping DLM dissolved in a dichloromethane dispersed phase during nanoparticle formation. Initial emulsion formulation screening by probe-tip ultrasonication revealed that a 1:1 mass ratio of lecithin and HPMC stabilizers formed 250 nm size-stable emulsion droplets with 40% DLM loading. Scale-up studies were performed to produce nearly identical droplet size distribution at larger scale using high-pressure homogenization, a continuous and industrially scalable technique. The resulting emulsions were spray-dried to form a dried powder, and in vitro dissolution studies showed dramatically enhanced dissolution kinetics compared to both as-received crystalline DLM and micronized crystalline DLM, owing to the increased specific surface area and partially amorphous character of the DLM-loaded nanoparticles. Solid-state NMR and dissolution studies showed good physical stability of the emulsion powders during accelerated stability testing (50 °C/75% RH, open vial).

For samples analyzed by TGA, crystalline delamanid was added to organic solvent at approximately 200 mg mL -1 , in excess of its expected solubility.After gentle vortexing, each vial was allowed to sit at room temperature overnight for approximately 12 hours, after which 100 μL of supernatant was removed for thermogravimetric analysis (TGA).Samples were analyzed using a Q50 TGA (TA Instruments, New Castle, DE).Under a nitrogen atmosphere, each sample was heated at a rate of 5 °C min -1 from room temperature to 120 °C, held isothermally for 20 min, and then cooled at 5 °C min -1 to room temperature.The residual solids mass was measured at the end of the isothermal step, and the solubility of delamanid was calculated as the ratio of the residual solids mass to the mass of the sample at t=0.Analysis was performed using the TA Instruments Universal Analysis software package.
Solubility in DCM was measured by adding a known mass of delamanid to a vial and gradually adding DCM volumetrically, stirring between additions.When all delamanid was visually solubilized, the final volume of DCM added was recorded and the solubility reported as a weight percent.For samples emulsified by high-pressure homogenization, the percentage of DCM lost was lower at higher batch size.The fact that DCM loss did not increase with increased batch size suggests that the mass of DCM loss may have a fixed component.Therefore, the effects of DCM loss may decrease at larger scales.Samples emulsified by probe-tip ultrasonication experienced comparatively minimal loss of DCM (<20%) over the 3 minute sonication duration.
For samples emulsified by probe-tip ultrasonication, 7 discrete samples of 5 mL each were prepared; each was sonicated for 0, 30, 60, 120, or 180 seconds in an ice water bath using the same conditions under which the delamanid emulsions were prepared.For high-pressure homogenization, sample volumes of 20 mL and 100 mL were tested.For each set of experiments, 5 discrete samples were prepared and homogenized for 0, 1, 2, 3 or 4 sequential passes using the same conditions as the delamanid emulsions.The Avestin EmulsiFlex C5 homogenizer was immersed in a room temperature water bath for the duration of all experiments.
Following emulsification, samples were volumetrically diluted 1:1 with methanol to form a single liquid phase.Reverse-phase HPLC (Phenomenex Kinetex C18, 100 Å, 150 x 4.6 mm, 5 μm particles) with an Agilent 1100 series HPLC was used to quantify the amount of DCM in each sample, Csample.DCM was eluted (RT = 2.8 min) with an isocratic mobile phase of 50:50 acetonitrile/water with 0.05% trifluoroacetic acid at a flow rate of 1 mL min -1 and column temperature of 35 °C, with UV detection at 220 nm.A linear standard curve was constructed for DCM concentrations between 0.5 and 10% v/v DCM in methanol (SI Figure 4).The concentration of DCM in the sample that did not undergo emulsification was termed C0.The DCM Loss percentage was calculated as 100 × Csample / C0.

SI Figure 4.
Exemplary HPLC standard curve for dichloromethane (DCM).Samples of known concentrations were prepared in methanol.The UV absorbance was highly linear with DCM concentration in the range of 1 -40% v/v, and a linear regression yielded  DCM = 0.0245(Peak Area), where  DCM is the concentration of DCM contained in the sample (as % v/v), and Peak Area is the integrated area of the UV absorbance peak at 220 nm.All samples were diluted into this concentration range for HPLC analysis.3 for batch sizes of 20 mL (A) and 100 mL (B).Sample 1 refers to the sample homogenized for one complete pass, Sample 2 for two passes, etc.All samples displayed similar temperature profiles, starting from room temperature of approx.22 °C and plateauing in the range of 30 -32 °C.Tabulated values are reported in SI Table 1. 1. Tabulated temperature values (°C) for samples described in SI Figure 5.    or sodium dodecyl sulfate (SDS) surfactant, measured by HPLC.The Tween 80 surfactant is nonionic, while the SDS is anionic.A saturated solution of 100 mg mL -1 DLM in THF was spiked into the aqueous buffer (100 μL THF into 10 mL buffer) to create a supersaturated solution.Samples were incubated at 37 °C overnight, after which they were centrifuged (21,000 g, 10 min) to pellet undissolved DLM.Samples were diluted with solvent as necessary and the concentration of dissolved DLM was quantified by HPLC.

SI Table
HEPES buffer with 3% added Tween 80 was selected for use as in vitro dissolution media for two reasons.First, the solubility of DLM in HEPES with 3% Tween 80 is sufficient to enable higher DLM dosing in the in vitro studies (approx.10 -15 μg/mL), enabling more accurate detection and quantification of DLM by HPLC (LLOQ = 0.7 μg/mL).Second, although SDS seems to provide enhanced solubility of DLM relative to Tween 80, the tabulated HPLC retention time data suggests that DLM and SDS may form an electrostatic complex.To avoid this electrostatic interaction, the nonionic Tween 80 was used, as it serves only as a micellar sink into which the hydrophobic DLM partitions.3 SI Figure 8. Exemplary HPLC standard curve for delamanid (DLM).Samples of known concentrations were prepared in tetrahydrofuran.The UV absorbance was highly linear with DLM concentration in the range of 1 -50 μg/mL, and a linear regression yielded  DLM = 0.1298(Peak Area), where  DLM is the concentration of DLM contained in the sample (in μg/mL), and Peak Area is the integrated area of the UV absorbance peak at 330 nm.The lower limit of quantification (LLOQ) was 0.7 μg/mL.All samples were diluted into this concentration range for HPLC analysis., measured by DLS.The particle size distributions across all timepoints are similar between formulations prepared by homogenization and ultrasonication, illustrating that the formulations exhibit good size stability over time and that both processing techniques can be used to produce nearly identical particles at different scales.As expected, due to the higher energy input of homogenization, particles produced by homogenization exhibited slightly smaller average size than those produced by ultrasonication.ASD t = 0 ASD t = 4 wk SI Figure 18.A) Particle size distributions over time (t = 0, 3, 24 h) for higher-loaded emulsion formulation containing 55% DLM by mass: 0.67 wt% DLM, 0.5 wt% lecithin, and 0.037 wt% HPMC (a 13.5:1 mass ratio of lecithin to HPMC).The dispersed phase was 15% v/v dichloromethane.

B)
In vitro dissolution profile for the higher-loaded emulsion formulation spray-dried, with 0.41:1 mass ratio of HPMC to nanoparticles added as a bulking agent.The resulting powder contained 43% DLM.The 43% DLM-loaded sample displayed rapid dissolution kinetics similar to those displayed by the 20% DLM-loaded sample.Dissolution kinetics for the as-received crystalline material and micronized crystalline DLM (prepared by spray drying from solution) as included for comparison.Tensiometry measurements were performed using a pendant drop of dichloromethane and an external phase of dichloromethane-saturated water.The stabilizer was dissolved in the water phase.A linear regression was performed for each measurement using an infinite-time asymptotic solution to the Ward and Tordai model for modeling adsorption of surfactant to a non-deforming surface: B1.Tabulated values of equilibrium interfacial tension,  ∞ , for a pendant dichloromethane drop in an external phase of dichloromethane-saturated water in the presence of stabilizer.
Images of experimental setups for A. probe-tip ultrasonication (VibraCell™ VC-50) and B. high-pressure homogenization (EmulsiFlex C5).During all experiments, the homogenizer was immersed in a room temperature water bath (Tbath ∼ 20 °C) such that the homogenizing valve assembly was entirely submerged.SI Figure 3. Dichloromethane (DCM) concentration (A) and corresponding loss percentage (B) as a function of homogenizer passes or ultrasonication time for blank DCM / water samples.Samples were prepared containing only 15% DCM and 85% water by volume, as the presence of drug and stabilizer interfered with quantification of residual DCM by HPLC.The amount of DCM lost during emulsification is important to quantify because significant loss of solvent in the system can lead to drug precipitation during emulsification.

5 .
Sample temperature as a function of pass number for high-pressure homogenized samples from SI Figure

6 .
Sample temperature as a function of homogenization pass Solubility of delamanid (DLM) in 150 mM pH 7 HEPES buffer with added Tween 80 Z-average hydrodynamic diameter of emulsion formulations A) F4, B) F5, C) F6, and D) F7 prepared by probe-tip ultrasonication using HPMC, HPMCAS-HF, HPMCAS-LF, or HPMCP as stabilizer (open bars) or a 1:1 mass ratio of lecithin to each of the cellulosic polymers (filled bars).The * indicates that the formulation was not prepared due to solubility limitations of the cellulosic stabilizer in the aqueous phase.The + indicates that the formulation gelled upon emulsification.Z-average hydrodynamic diameter of PCL core / HPMC-lecithin stabilized F1 formulation as a function of homogenizer pass number (for formulation prepared by high-pressure homogenization) or sonication time (for formulation prepared by probe-tip ultrasonication); B) Particle size distributions over time of PCL core / HPMC-lecithin formulation prepared by high pressure homogenization (4 passes), measured by DLS; C) Particle size distributions over time of PCL core / HPMC-lecithin formulation prepared by probe-tip ultrasonication (180 s soncation time)

12 .
SEM micrographs of A) delamanid crystals, B) F2 DLM/HPMC-lecithin spray-dried emulsion formulation at t=0, C) F2 DLM/HPMC-lecithin spray-dried emulsion after 4 weeks of accelerated stability testing (50 °C / 75% RH), D) DLM/HPMCP spray-dried solid dispersion at t=0, E) DLM/HPMCP spray-dried solid dispersion after 4 weeks of accelerated stability testing (50 °C / 75% RH).Micrographs were obtained using an FEI Verios 460 SEM at 10 KV acceleration voltage.All formulations exhibit a wrinkled sphere morphology due to the formation of a dried skin at the droplet surface during drying.No evidence of DLM crystallites was observed visually by SEM, confirming the need for the other characterization techniques applied in this work (XRD, DSC, ssNMR).Recrystallization studies of pure delamanid performed by DSC showing recrystallization from the melt state at cooling rates of A) 5 K min -1 , B) 10 K min -1 , C) 15 K min -1 , and D) 20 K min -1 .Such recrystallization behavior highlights the propensity of delamanid to recrystallize and would place it in Class I as defined by Baird et al. (J.Pharm.Sci.2010, 99 (9. 13C solid state NMR spectra of the DLM/HPMCP formulation before and after 4 weeks of open-vial storage at 50 °C / 75% RH, zoomed in on selected regions highlighting the DLM-only peaks.The estimated (lower limit) percent crystallinity of DLM in the different formulations is annotated on the left (calculated with respect to the mass of DLM in each formulation).The 4-week stability tested samples (open vial, 50 °C / 75% RH) are plotted in red and overlayed on the respective initial formulations.The dashed rectangles highlight the specific regions where the DLM signals are nominally well resolved and not overlapped by signals of the excipients.DLM HPMCP 4 wk SI Figure 14. 13 C solid state NMR spectra of the fresh (blue) and 4-week open vial, 50 °C / 75% RH (red) DLM HPMCP ASD formulation zoomed in on selected regions highlighting the DLM only peaks.The change in peak linewidths of selected DLM peaks at 143.63 ppm and 92.32 ppm are highlighted next to the peaks.SI Figure15.A) DLM crystals as received from the manufacturer, B) DLM spray-dried from a dichloromethane solution without addition of additional excipients.Differential scanning calorimetry (DSC) thermogram of DLM crystals as received from the manufacturer with melting enthalpy (on first heating) calculated by peak integration, B) DSC thermogram of DLM spray-dried from a dichloromethane solution (without addition of additional excipients crystals) with melting enthalpy (on first heating) calculated by peak integration.The ratio of the enthalpies indicates that the spray-dried material is 97% crystalline.Heating and cooling rates were 5 °C min -1 .SI Figure17.In vitro dissolution profile for an amorphous solid dispersion (ASD) formulation containing 20% DLM and 80% HPMCP (HP-50), prepared by spray drying from a 1:1 volumetric mixture of methanol and dichloromethane.Initially at t = 0 the ASD displayed fairly rapid and complete dissolution kinetics; however, after 4 weeks of storage at 50 °C / 75% RH the in vitro dissolution performance was significantly reduced.

Sample 2 Sample 3 Sample 4 Sample 1 Sample 2 Sample 3 Sample 4
number for emulsion formulations F1 and F2 with a 1:1 mass ratio of HPMC to lecithin used as stabilizer.Samples were prepared at a batch size of 100 mL.Samples were initially at room temperature and increased to 32 -33 °C after one pass, where they remained approximately plateaued for all successive homogenization passes.The Avestin EmulsiFlex C5 homogenizer was immersed in a room temperature water bath for the duration of all experiments.Tabulated values are reported in SI Table2.
SI Table2.Tabulated temperature values (°C) for samples described in SI Figure6.

.
Tabulated solubility and HPLC retention time of delamanid (DLM) in 150 mM pH 7 HEPES buffer with added Tween 80 or sodium dodecyl sulfate (SDS) surfactant.

Table A3 .
Average emulsion droplet size over time for emulsions stabilized by HPMCAS LG or a mixture of HPMCAS LG and lecithin (1:1 mass ratio) *Gelled during sonicationSI

Table A4 .
Average emulsion droplet size over time for emulsions stabilized by HPMCP (HP-50) or a mixture of HPMCP and lecithin (1:1 mass ratio)