Inhalable Nanocrystal-in-Microparticle of Linezolid in Pulmonary Tuberculosis

Abstract

Background

Tuberculosis (TB) remains a major global health challenge, necessitating the development of advanced pulmonary drug delivery systems to improve therapeutic efficacy. Linezolid (LZD), a second-line oxazolidinone antibiotic, is effective against multidrug-resistant Mycobacterium tuberculosis but is associated with systemic adverse effects when administered orally for prolonged periods.

Objective

This study aimed to engineer and optimise LZD nanocrystals (LZD-NCs) in microparticles for dry powder inhalation (DPI) using a Quality by Design (QbD)-based Central Composite Design (CCD) approach to enhance localised pulmonary delivery.

Methods

LZD-NCs were initially prepared by high-pressure homogenization (HPH) and subsequently converted into respirable nanocrystals in microparticles (LZD-NIM) using spray drying. The LZD-NCs formulation was optimised using CCD. The developed formulation was evaluated for physicochemical characteristics, including solid-state characterisation, in vitro drug-release study, kinetic study, and aerosolisation behaviour and stability.

Results

The optimised formulation exhibited an average particle size of 239.5 ± 16.30 nm, polydispersity index (PDI) of 0.285 ± 0.042, and surface charge of −34.5 ± 1.84 mV, indicating acceptable colloidal stability. Later, NIM effectively retained the physicochemical integrity of LZD-NCs. The developed NIM formulation exhibited improved drug release of 61.39% over 24 h, compared with pure LZD at 37.26%. An in vitro aerosolisation study demonstrated appropriateness for deep lung deposition of NIM with a mass median aerodynamic diameter (MMAD) of 2.9 ± 0.15 µm and high fine particle fractions (FPF) of 76.12 ± 1.3%.

Conclusion

The engineered LZD-NIM successfully yielded a highly stable, aerodynamically efficient DPI platform with sustained drug release kinetics. The engineered platform demonstrates its potential as an effective pulmonary drug-delivery system for TB treatment.

Graphical Abstract

Introduction

Tuberculosis (TB), caused by Mycobacterium tuberculosis (Mtb), remains one of the deadliest infectious diseases worldwide, affecting approximately one-third of the global population.^1,2 Although Mtb can disseminate to extrapulmonary sites, including lymph nodes and bones, the lungs are the primary target organ.² The emergence of multidrug-resistant (MDR-TB) and extensively drug-resistant (XDR-TB) strains has severely limited therapeutic options and complicated treatment outcomes, with approximately 50% of MDR-TB patients failing to achieve a successful cure. According to the World Health Organisation (WHO) 2025 Tuberculosis Report, approximately 10.7 million new TB cases were reported globally in 2024.^3,4 Although TB-related mortality decreased slightly to 1.23 million deaths in 2024, significant challenges persist regarding the availability of effective diagnostics and treatment options.⁴ The limitations of conventional systemic TB therapy stem from multiple factors, including lengthy treatment regimens with poor patient adherence due to adverse effects, particularly hepatotoxicity with traditional drugs used in the Directly Observed Treatment Short Course (DOTS).⁵ These challenges have spurred research into advanced drug-delivery strategies.

The pulmonary route of administration offers distinct advantages by enabling direct delivery of antitubercular agents to the primary site of infection. This targeted approach minimises systemic exposure and associated toxicities while improving therapeutic efficacy compared to oral administration. Furthermore, pulmonary delivery enables sustained, controlled drug release, reducing treatment duration and dosing frequency, thereby enhancing patient compliance.^6,7 To achieve effective lung delivery, formulation design must balance multiple parameters. Particles with aerodynamic diameters of 1–5 μm deposit in the deep alveolar region where Mtb resides in infected macrophages.^8,9 In contrast, isolated nanocrystals (typically <1 μm) are generally exhaled and do not accumulate in the lungs; therefore, microparticles with embedded nanoparticulate components may provide an optimal strategy by combining the lung retention benefits of micrometre-sized particles with the enhanced solubility, cellular uptake, and tissue penetration properties of nanostructures.^10,11 In this context, Jadhav et al developed inhalable clofazimine nanocrystals that overcome poor biopharmaceutical properties. Subsequent nanoclusters show enhanced aerodynamic deposition and demonstrated reduced mycobacterial load in TB-infected mice.¹²

Linezolid (LZD), the first approved oxazolidinone antibiotic, has demonstrated significant activity against Mtb in both preclinical and clinical studies and represents a valuable option for treating drug-resistant TB.¹³ Standard oral or intravenous LZD treatment for MDR-TB is severely limited by dose-dependent systemic adverse effects during the required prolonged treatment regimens.^14,15 In this context, Huang et al fabricated controlled-release microspheres loaded with rifapentine and LZD that sustained drug release for 14 days and prolonged bronchial mucosal retention for up to 20 days.¹⁶ Zarogoulidis et al optimised nebuliser designs to enhance aerosol delivery efficiency of LZD and other antibiotics,¹⁷ while Sharma et al developed inhalable polycaprolactone microspheres– demonstrating good aerosolisation efficiency (fine particle fraction of 70.90%) and sustained release over 24 hours.¹⁸

In this study, we developed LZD-nanocrystals in microparticles (NIM) for dry powder inhalation (DPI) as a therapeutic strategy for pulmonary TB. LZD-NCs were prepared using high-pressure homogenization (HPH) and optimised using a Quality by Design (QbD) approach with a Central Composite Design (CCD). These optimised NCs were subsequently converted into microparticles via spray drying. We hypothesised that this NIM formulation would efficiently target the lungs, provide sustained drug release, and potentially enhance bioavailability, making it a promising strategy for treating pulmonary TB.

Materials and Methods

Materials

LZD was kindly supplied by Lupin Ltd. Chhatrapati Sambhajinagar, India, and Poloxamer 188 (Kolliphor^® 188), Poloxamer 407 (Lutrol^® F 127), polyvinylpyrrolidone K-30 (PVP K-30), polyvinyl alcohol (PVA), polyvinyl caprolactam-polyvinyl acetate-polyethylene glycol graft copolymer (Soluplus^®; PVCL-PVAc-PEG), Polysorbate 20 (Tween^® 20), Polysorbate 80 (Tween^® 80), Sodium lauryl Sulfate (SLS), and Mannitol were purchased from Sigma-Aldrich, India, HPLC grade methanol, and water were procured form Merck, India. All other chemicals and reagents used in this study were of analytical grade.

Preparation of LZD Nanocrystals (LZD-NCs)

LZD-NCs were fabricated by a high-pressure homogeniser (HPH) (Panda plus, Gea Niro Soavi, Italy), using a previously reported method.¹² Briefly, 100 mg of LZD was weighed accurately and slowly added to the 0.2% w/v Poloxamer 188 solution, which served as the stabiliser, under constant stirring to obtain an initial drug dispersion (drug: stabiliser;10:2). The dispersion was pre-milled using a homogeniser (IKA, Germany) at 10000 rpm for 5 min to obtain a microsuspension. The microsuspension was initially processed by HPH at 150 bars for 14 cycles and subsequently processed with pressure ranging from 300 to 900 bar and cycles from 10 to 30. The resulting NCs were centrifuged for 30 min at 15,000 rpm (Cooling centrifuge C 8, Remi India) to remove the excess stabiliser and recover the NCs as a pellet. The LZD NCs were redispersed in MilliQ water for physicochemical characterisation, including particle size, zeta potential, and drug content analysis by UV spectroscopy. At the time of preliminary formulation screening, alternative stabilisers were evaluated, including poloxamer-407, PVA, Polysorbate 20 and 80, PVCL-PVAc-PEG, PVP K-30 and SLS. Among these, poloxamer-188 was chosen as the optimal stabiliser based on formulation properties such as particle size, PDI and surface charge.

Optimisation of LZD-NCs Using CCD

The fabrication of LZD-NCs was optimised by selecting critical process parameters, including homogenization pressure and the number of homogenization cycles, which significantly impacted the average particle size, PDI, and zeta potential (Surface charge). To maintain consistency, other parameters, such as drug concentration, stabiliser amount, and media volume, were kept constant. To optimise these critical parameters, a CCD was employed. The CCD approach was chosen to estimate the relationships between process parameters and the resulting responses. A two-factor, three-level CCD was used, consisting of 13 experimental batches. These included a two-level full factorial design (4 batches), two axial point batches for each factor (+1 and −1) (4 batches), and 5 centre point batches. Replicate centre points were included to determine experimental error and enhance the precision of the estimations. The axial points were set at ±1.41421 to fulfil the CCD’s rotatability criterion.¹⁹ The composition of the CCD batches is summarised in Table 1, and all experiments were conducted in triplicate (n = 3).

Table 1.

Independent and Dependent Variables: Factors and Levels in the Central Composite Design

Factors	Levels
Independent variables	Low (−1)	Medium (0)	High (+1)
X1: Pressure (bar)	−1 (300)	0 (600)	1 (900)
X2: Number of Cycles	−1 (10)	0 (20)	1 (30)
Dependent variables	Constrains
Y1: Particle Size (nm)	Minimum
Y2: Polydispersity Index	Minimum
Y3: Zeta Potential (mV)	In range (±10 to 35)

Design of Experiments (DoE) Analysis and Model Selection

The results of the experimental batches were statistically evaluated using Design Expert^® 13.10 software (Stat-Ease^®). The effects of the independent variables and their interactions were assessed, and a mathematical relationship between the independent variables (process parameters) and dependent responses (average particle size, PDI, and zeta potential) was established.^20,21 Polynomial equations were generated to model these relationships, enabling the identification of optimal process conditions.

Preparation of NIM

The DPI of LZD-NCs was developed using a spray-drying process, in which the LZD-NCs were atomised in the presence of mannitol. Briefly, a nanodispersion containing LZD-NCs and 10% (w/w) mannitol relative to the total amount of drug and excipients was prepared in MilliQ water. The obtained dispersion was continuously stirred using a magnetic stirrer, and spray-dried powder was prepared using a 0.5 mm nozzle atomiser of a laboratory spray dryer (Spray Mate JISL, Mumbai, India). The spray dryer working settings were set as inlet temperature: 110 ^°C, outlet temperature: 60 ^°C, atomisation pressure: 1.2 kg/cm2, feed flow rate: 4 mL/min, and the vacuum was maintained at 100–110 mmHg. The LZD-NCs were fed to the spray dryers using a pump.^21,22 Lastly, spray-dried powder was collected from the collecting vessel and the cyclone separator, stored in airtight glass vials, and kept in the vacuum desiccator until further use for investigation. Response variables used to evaluate the spray-drying process and the resulting NIM dry powder included product yield, particle size, surface morphology, and the aerodynamic characteristics of the DPI.

Physicochemical Characterisation

Particle Size, PDI, and Zeta Potential

The average particle size, PDI and zeta potential of LZD-NCs were determined via dynamic light scattering with Nano ZS 90 Malvern Zetasizer (Malvern, UK). The 50 µL of all samples were diluted with 2 mL of deionised water in disposable polystyrene cuvettes, and measurements were performed at 25 °C, with a medium viscosity of 0.8872 mPas and a refractive index of 1.330. The number of runs is set automatically by the instrument.²³ All the samples were prepared separately, and experimental values were recorded in triplicate, and the results were represented as mean ± SD (n=3).

Surface Morphology

The surface morphology of LZD-NCs was examined using field-emission scanning electron microscopy (FE-SEM) (JSM IT300LV, JEOL, Japan). The NCs of LZD were diluted to 1 mg/mL, placed on the silica wafer, and allowed to dry. A very thin film of a spray-dried LZD-NCs powder sample was uniformly spread on the carbon tape attached to the aluminium stub. Then, the sample was mounted on the sample holders and sputter-coated with a 90 s gold layer using a coater (JEC-3000FC, JEOL, Japan). The sample holder was placed inside the FE-SEM, and surface morphology was captured using a secondary electron detector and an accelerating voltage of 2 kV (LZD-NIM) and 15 kV (LZD-NCs).^24,25

Re-Dispersibility Index

The ability of the fabricated spray-dried powder to re-disperse into NCs was studied. Briefly, 2 mg of LZD spray-dried powder was dispersed in 2 mL of MilliQ water and stirred magnetically for 15 min at room temperature. The obtained dispersion was evaluated for average particle size and zeta potential, and the redispersibility index (RDI) was determined as described in the hitherto method.^24,25

Determination of the Product Yield

The product yield of LZD-NIM in dry powder form was determined by weighing the collected product. The percentage yield was then calculated using the following formula.²⁶

(1)

Determination of Drug Content

The drug content was determined using the following procedure. Briefly, 5 mg of LZD-NIM were accurately weighed, and the LZD-NCs were dissolved in a PBS (pH 6.8) buffer to dissolve the LZD. A 10 µL sample was diluted with 10 mL of PBS buffer, and the absorbance was measured at 250 nm (Shimadzu UV–VIS Spectrophotometer, UV-2600, Japan). The content of LZD in the DPI formulation was quantified using UV spectroscopy.²⁶ The amount of LZD drug was calculated using the following formula:

(2)

Solid State Characterisation

Fourier Transform Infrared Spectroscopy (FT-IR)

The FT-IR spectra of Pure LZD and LZD-NIM were recorded to examine the interaction between excipients and drug using a Spectrum Two Spectrometer (Perkin Elmer, USA) assembled with an ATR (Attenuated total reflectance, single reflection diamond) accessory. The very small quantity of each sample was located in the ATR cell for spectral analysis. The FT-IR spectra were recorded at 4 cm⁻¹ resolution under a dry air purge with 8 scans accumulated in ATR mode, and the spectrum obtained in the range of 4000 to 400 cm⁻¹.²⁷

X-Ray Diffraction Analysis (XRD)

To further investigate the physical characteristics, particularly crystallinity, of pure LZD and LZD-NIM formulation, XRD analysis was performed using the D8 advance Eco powder diffractometer (Bruker, Germany) with radiation source (Cu-Kα; λ =1.54056 Å) and scanned over the 2θ range (5 to 60°) at scanning rate of 1°/min; time step per 0.5s and instrument was set operated at 40 kV and 40 mA. The samples of pure LZD powder and LZD-NIM were kept on the sample holder rings using the above-mentioned experimental setup.²²

Differential Scanning Calorimetry Study (DSC)

Thermal characteristics of pure LZD and LZD-NIM were determined using DSC 7020 (Hitachi High-Tech Corporation, Japan). 5 mg of each sample was placed in the aluminium crucibles and crimped, and scanning was executed in the nitrogen gas flow (20 °C/min⁻¹) maintained, scanning was completed at temperature in the range of 30 °C to 300 °C. The obtained thermograms were analysed using TA7000 software (Hitachi High-Tech Corporation, Japan) to evaluate the endothermic/exothermic peaks of the powder samples. Each sample was analysed in a single determination, except pure LZD, for which a second determination was carried out on an independent sample of the drug substance.²⁸

Thermogravimetric Analysis (TGA)

The thermal properties of pure LZD and LZD-NIM were analysed using a TGA-55 thermogravimetric analyser (TA Instruments, UK). Approximately 5 mg of each sample was placed in a platinum pan and heated at 20 °C/min to 600 °C under a nitrogen (N₂) purge at 50 mL/min. The results were processed using TA Instruments Universal Analysis 2000 software (version 4.5A), and weight-loss curves were plotted as a function of temperature.²²

In-vitro Release Study and Release Kinetics

An in vitro drug release study for pure LZD and LZD-NIM was conducted using the dialysis bag method. The dialysis bag, made of cellulose membrane (Sigma, USA, molecular weight cutoff: 12,000 g/mol), was filled with 50 mg of each formulation. The drug release medium consisted of 900 mL of phosphate buffer (pH 6.8), maintained at 37.5 ± 0.5 °C with a stirring speed of 60 rpm. At predefined intervals (15 min, 30 min, 1 h, 2 h, 4 h, 6 h, 8 h, 12 h, 18 h, 24 h, 48 h, and 72 h), 1 mL aliquots were withdrawn and immediately replaced with an equivalent volume of fresh dissolution medium to maintain sink conditions. The collected samples were filtered through 0.2 μm syringe filters to remove residual content, and the supernatants were analysed for drug content using UV spectroscopy at a wavelength (λ max) of 250 nm. Release data were fitted to the zero-order, first-order, Higuchi, and Korsmeyer-Peppas models, and R² values were determined.²⁹

In-vitro Aerodynamic Performance

The in-vitro aerodynamic performance of LZD-NIM was assessed using a Next Generation Impactor (NGI, Copley Scientific, UK) connected to a Rotahaler device (Cipla, Mumbai). In brief, 50 mg of LZD-NIM was accurately weighed and filled into hard gelatin capsules of size 3. These capsules were then placed in the sample reservoir of the Rotahaler inhaler, which was securely attached to the induction port of the NGI apparatus. The NGI apparatus was operated at an airflow rate of 60 L/min for 10 s, powered by a high-capacity pump (HCP5, Copley, UK), and the airflow was measured using a flowmeter (DFM2, Copley Scientific, UK). After a single actuation, the Rotahaler apparatus pricked the capsule and released the LZD spray-dried powder with a continuous airflow rate of 60 L/min. The spray-dried powder was distributed across the various stages of the NGI based on its aerodynamic particle size, with effective cut-off diameters of 8.06, 4.46, 2.82, 1.66, 0.94, 0.55, and 0.34 μm, respectively. Individual stages of NGI correspond to distinct effective cut-off sizes, which correspond to distinct locations along the respiratory tract. The separate stages of the NGI are associated with defined effective cut-off diameters that correlate with specific anatomical regions of the respiratory tract.^12,30 The LZD-NIM was collected from each NGI stage and analysed using the gravimetric method; all experiments were conducted in triplicate (n=3). Aerodynamic parameters, including mass median aerodynamic diameter (MMAD), geometric standard deviation (GSD), and fine particle fraction (FPF), were assessed.

Short-Term Stability Study

The stability of the optimised LZD-NIM formulation was assessed over six months under storage conditions at 25 °C (room temperature, RT), following a previously reported procedure.³¹ The samples were stored in glass vials sealed with rubber stoppers and aluminium caps to maintain integrity. These formulations were maintained under the specified conditions to enable comparative stability assessment. Colloidal characteristics, including mean particle size, PDI, and zeta potential, were measured to identify any changes over time. These parameters were evaluated using the methods described in the particle size, PDI, and zeta potential section. All experiments were performed in triplicate (n = 3) to ensure reproducibility.

Statistical Analysis

Experimental results are reported as mean ± standard deviation (SD) from at least 3 independent trials. Statistical analysis was conducted using one-way ANOVA, with a significance threshold set at p < 0.05. The analysis was performed using OriginPro 10.1.5.132 (Learning Edition) software (OriginLab Corp).

Results

Preparation and Characterisation of LZD-NCs

HPH is a top-down approach used to reduce large, micron-sized particles into nanoscale systems such as NCs. This technique offers significant advantages over traditional methods, including uniform particle size within the nanoscale range and a narrow PDI. The mean particle size, PDI, and zeta potential values of the optimised LZD-NCs formulation (Batch 5) were found to be 239.5 ± 16.30 nm, 0.285 ± 0.042, and −34.5 ± 1.84 mV, respectively (Figure 1A and B). The surface morphology of LZD-NCs fabricated by the HPH technique, as observed by FE-SEM, revealed cubic and irregularly shaped particles (Figure 1C).

Figure 1.

Physicochemical characterization of LZD nanocrystals (LZD-NCs) prepared by HPH. (A) particle size distribution and (B) Zeta potential of LZD-NCs measured by dynamic light scattering (DLS). (C) Field-emission scanning electron micrograph of LZD-NCs showing cubic to irregularly shaped discrete particles. (D) Higher-magnification field-emission scanning electron micrograph of the same sample. All measurements were performed in triplicate and are reported as mean ± SD (n = 3).

The primary objective was to reduce the NCs’ mean particle size. To achieve this, a minimum of ~100 mL of a dispersion/macro-suspension containing pure LZD with a stabiliser was processed through the homogeniser. The dispersion passed through the outlet channel after being circulated via the inlet chamber assembly for a defined number of cycles at different homogenization pressures.

During the initial phase of fabrication, selecting a suitable surfactant or stabiliser was critical, particularly one considered safe for pulmonary applications. Screening of stabilisers for nanocrystals preparation was performed at a drug-to-stabiliser ratio of 10:2. The trials involved 30, 20, and 10 homogenization cycles at initial pressures of 150 bar, 300 bar, and 600 bar, respectively, followed by 900 bar.

Poloxamer 188 demonstrated superior performance, with uniform particle size, narrow PDI values, and acceptable zeta potential, thereby ensuring colloidal stability. Other stabilisers, including Poloxamer 407, Polysorbate 80, PVA, PVP K-30, SLS, and PVCL-PVAc-PEG, were also evaluated but exhibited less favourable results. Consequently, Poloxamer 188 was selected for further optimisation of nanocrystal formulation using a Design of Experiments (DoE) approach based on the CCD.

Furthermore, process parameters such as homogenization pressure and the number of cycles were identified as critical factors influencing the physicochemical characteristics of NCs, including mean particle size, PDI, and zeta potential. Controlling these parameters was essential for achieving the desired stability and efficacy of the formulation.^32,33

Impact of Homogenisation Pressure

The effect of the critical process parameter on the NCs based formulation, such as pressure and number of cycles on the particle size, PDI, and zeta potential, directly impacting the stability and quality of the final product shown in Figure 2A. By keeping the lower pressure (bar) with cycles 30 cycles and drug (50 mg) and stabilizer concentration (0.1%, ie poloxamer-188) by preparing the NCs prepared using HPH. In this condition, the particle size (197.7 nm) was achieved with PDI (0.418), which indicates poor uniformity and zeta potential (−19.2 mV), which denotes acceptable colloidal stability. While pressure increases to 300 bar with 20 cycles, the particle size slightly increases, ie 231.9 nm, while PDI (0.345) and zeta potential (−26.5 mV) reflect the improved stability.

Figure 2.

Effect of formulation and process parameters on the mean particle size, PDI, and zeta potential of LZD NCs produced by HPH. (A) Influence of homogenization pressure (150–1200 bar, with the cycle number reduced from 30 to 5 as pressure increased) at a Poloxamer 188 concentration of 0.1% w/v. (B) The same pressure series at a Poloxamer 188 concentration of 0.2% w/v. (C) Influence of the number of homogenization cycles (14, 10, and 6 cycles at 150, 300, and 600 bar, respectively). (D) Influence of stabiliser type, comparing PVCL-PVAc-PEG, Poloxamer 188, Poloxamer 407, Polysorbate 80, sodium lauryl sulfate, PVP K-30, and polyvinyl alcohol at a fixed drug-to-stabiliser ratio. Data are mean ± SD (n = 3).

Further, increasing the pressure to 600 bar with 15 cycles yields a particle size (222.7 nm), PDI remains excellent at 0.228 and excellent zeta potential (−28.4 mV), indicating the stable and uniform NCs. Additionally, at 900 bar with 10 cycles, the particle size increases to 247.0 nm, PDI remains excellent at 0.288, and zeta potential reaches its highest value, i.e −29.2 mV, ensuring robust stability. At the highest pressure (1200 bar) with 5 cycles, the particle size increases to 259.8 nm, PDI rises to 0.286, and zeta potential drops significantly to −18.6 mV, which reflects the acceptable colloidal stability, likely due to the excessive energy input leading to particle aggregation. These results highlight moderate to high pressure (600–900 bar) with sufficient cycles to optimise the nanosuspension, achieving appropriate particle size reduction, uniformity, and stability, aligning with the principles of sustainable formulation by minimising energy input while maintaining high formulation quality. The corresponding trends obtained at the higher stabiliser concentration (0.2% w/v Poloxamer 188) are presented in Figure 2B.

Impact of Homogenization Cycles

Another parameter, ie the number of cycles and pressure settings, critically influences particle size, PDI and zeta potential. The relationship between these variables and the number of cycles and pressure determines the particle size and stability at 14 cycles and 150 bar, resulting in particle size (233.7), PDI (0.30), and zeta potential (−35.1 mV), as shown in Figure 2C.

The lower pressure setting provides a lower energy input per cycle, requiring a larger number of particles to achieve particle size reduction and improved stability. However, a zeta potential of −35.1 mV indicates excellent electrostatic stability. It has a slightly higher PDI, reflecting minor heterogeneity in particle size, which may be due to prolonged mechanical stress associated with the increased cycles. The reduced number of cycles (10) at increased pressure (300 bar), compared to the 14-cycle condition, resulted in a smaller particle size of 228.9 nm and a PDI of 0.26. LZD-NCs exhibited a higher zeta potential (−32.3 mV), providing sufficient electrostatic repulsion for nanosuspension stability at minimal cycles and moderate pressure, thereby reducing particle size. The combination of moderate pressure and cycles achieves the optimal particle size distribution and stability, making it an optimal setting for these LZD-NCs. Decreased cycles (i.e 6) and increased pressure (600 bar) resulted in LZD-NCs with a slight increase in average particle size (237.7 nm) and a low PDI (0.250), indicating a uniform particle size distribution. Additionally, the zeta potential (−21.5 mV) indicates stable, acceptable colloidal stability of LZD-NCs. The correlation between pressure and the number of homogenization cycles indicates that high pressure can be achieved with fewer cycles, which can reduce processing time but may not provide adequate formulation stability. Our study shows, at 10 cycles with 300 bar emerges as optimal condition, offering the average particle size 228.8 nm with uniform size distribution (PDI ie 0.260) and excellent stability (zeta potential i.e −32.2 mV) at high cycles and low pressure gives best stability but increased PDI and processing times while at 6 cycles at 600 bar achieves uniform the particle size distribution suboptimal surface charge which resulting the acceptable colloidal stability of LZD-NCs. This highlights the need for a balanced approach to choosing process parameters to achieve the required NC characteristics.

Impact of Various Stabilisers

The choice of a suitable stabiliser was critical for the development of LZD-NCs. The surfactants play a critical role in the fabrication of NCs with desired physicochemical characteristics, such as average particle size, PDI, and surface charge (zeta potential), for the overall stability of the nanoformulation. Our study highlights the impact of various polymeric stabilisers on NCs performance, as illustrated in Figure 2D. The tested stabiliser, PVCL-PVAc-PEG, demonstrated the most effective stabilisation, producing a particle size of 148.8 with moderate PDI (0.400) and indicating reasonably uniform particle size. Similarly, Poloxamer 188 produced favourable results, with a particle size of 289.4 nm and a low PDI of 0.295, demonstrating its potential for achieving both size reduction and uniform particle size. In contrast, Poloxamer-407, chemically similar to Poloxamer-188, resulted in a larger particle size of 349.1 nm and a higher PDI of 0.514, indicating a broader particle size distribution. Polysorbate 80 and SLS exhibited distinct physicochemical characteristics in the nanoformulation. Use of Polysorbate 80 showed an average particle size (181.4 nm) with a higher PDI (0.660), indicating significant heterogeneity of particle size distribution. On the contrary, SLS achieved a lower PDI (0.190), indicating excellent uniformity, but it also resulted in a high particle size (586.2 nm). The use of PVP K-30 and PVA further highlights this, as they produced the largest particle sizes of 481.3 nm and 618.7 nm, respectively, with high PDIs (0.686 and 0.952), suggesting wide size variability and poor particle size distribution.

Impact of Stabiliser Concentration

The concentration of Poloxamer-188 and the applied pressure significantly influence the average particle size, PDI and zeta potential of fabricated NCs produced using the HPH, shown in Figure 2A and B. The low pressure and higher cycles (150 bar with 14 cycles) with 10 mg and 20 mg of Poloxamer-188 per 100 mL distilled water with constant LZD concentration, which resulted in the particle size (383.4 nm and 243.1 nm) with high PDI (0.624 and 0.429) and lower zeta potential (−16.7 mV and −37.3 mV), respectively. Due to inadequate pressure, the stabiliser concentration has minimal effect, as there is insufficient energy for particle size reduction and stabilisation. Moreover, at the moderate pressure and number of cycles, i.e 300 bar with 10 cycles, which resulted in the particle size (307.9 and 239.1 nm), improved the uniformity with PDI (0.435 and 0.285) and enhanced stability (zeta potential: −25.6 and −34.5 mV), demonstrating the beneficial effect of moderate pressure and higher stabiliser for appropriate properties of the formulation. Moreover, at high pressure with a minimal number of cycles (600 bar with 5 cycles), the particle size for 10 and 20 mg poloxamer-188 was 236.0 nm and 247.2 nm, and further improved PDI (0.271 and 0.277) and zeta potential (−33.7 and −24.3 mV), respectively. In these experimental trials, higher pressure consistently reduced particle size and improved PDI and zeta potential at both concentrations. However, the combination of moderate pressure (300 bar) and higher concentration (0.02%) achieved the best balance of small particle size (239.1 nm) and low-value PDI (0.285), and higher zeta potential (−34.5 mV), underscoring the importance of optimising the stabiliser concentration and processing conditions for stable NCs.

Optimisation via Central Composite Design

The CCD was employed to study the effect of the two interdependent factors, mainly homogenization (X₁) and number of cycles (X₂), on the response variables such as particle size (Y₁), PDI (Y₂) and surface charge (Y₃). CCD is a robust statistical tool for optimisation, which enables the analysis of linear and interaction effects among factors at multiple levels. In our study, three levels were altered by ±1.414, facilitating a comprehensive analysis of their effects on the physicochemical properties of LZD-NCs. The experimental design comprised 13 runs, revealing significant changes in particle size, PDI, and zeta potential (Table 2). The primary aim of this study was to investigate the effects of homogenization pressure and cycle number on the physicochemical properties of LZD-NCs to optimise their performance for drug delivery.

Table 2.

Independent Variables and Corresponding Outcomes for Dependent Variables in the Central-Composite Design

Std	Run	Factor 1	Factor 2	Response 1	Response 2	Response 3
		A: Pressure (bar), Actual (Coded)	B: Number of Cycles, Actual (Coded)	Particle Size (nm)	PDI	Zeta Potential(mV)
6	1	1.41421	0	289.2	0.363	−22.3
7	2	0	−1.41421	311.7	0.218	−24.8
10	3	0	0	239.2	0.331	−22.6
11	4	0	0	318.6	0.349	−25.4
8	5	0	1.41421	231.4	0.261	−30.1
2	6	1	−1	287.9	0.26	−26.3
13	7	0	0	258.3	0.24	−24.5
9	8	0	0	360.6	0.293	−23.32
12	9	0	0	233.7	0.362	−27.2
1	10	−1	−1	422.5	0.295	−23.7
4	11	1	1	219.1	0.524	−29.5
3	12	−1	1	337.7	0.224	−32.2
5	13	−1.41421	0	328.9	0.26	−28.9

Notes: Actual factor values were calculated from the coded levels defined in Table 1 (pressure: −1 = 300 bar, 0 = 600 bar, +1 = 900 bar; cycles: −1 = 10, 0 = 20, +1 = 30).

Impact on Particle Size (Y1)

The particle size of NCs was significantly affected by both X₁ and X₂, with a linear model requiring the best fit for the response. The linear model was deemed the best fit for Y₁ based on the highest R² value. Allowing for the analysis of the design of experiment data using Analysis of Variance (ANOVA), the subsequent polynomial equation (3) was produced for the Y₁ response:

(3)

The negative regression coefficients in the polynomial equation suggest that the factors have an opposing effect on particle size: as X₁ and X₂ increase, particle size decreases, which is important for nanoscale particles. The opined particle size range of 219.1 to 422.5 ± 59.49 nm exhibits effective particle size reduction within the desired nano range when both X₁ and X₂ were superior. ANOVA analysis revealed a significant p-value of 0.0339, checking that the linear model adequately captured the effects of these factors on particle size. The non-significant lack-of-fit (p > 0.05) supports the model’s consistency, indicating a good fit between experimental and predicted data. The negative coefficients for both factors propose that higher homogenization pressure and more cycles advance the generation of nanosized particles, which are crucial for increasing solubility and bioavailability in pharmaceutical applications. Still, the relatively low R² value of 0.3900 suggests that unaccounted-for factors may continue to influence particle size, requiring further model refinement to improve predictability. Figure 3A illustrates the 2D contour plot for Y₁, showing a steep drop in particle size as both X₁ and X₂ increase. Figure 3B shows the 3D surface plot of particle size, proving that higher homogenization pressure and a larger number of cycles lead to a smaller particle size.

Figure 3.

Design-of-experiments optimisation of LZD-NCs using a two-factor, three-level CCD. Two-dimensional contour plots and corresponding three-dimensional response surface plots show the effect of homogenization pressure (X₁) and number of homogenization cycles (X₂) on (A and B) mean particle size (Y₁), (C and D) PDI (Y₂), and (E and F) surface charge (Y₃). Axial points were set at ±1.41421 to satisfy the rotatability criterion.

Impact on Polydispersity Index (PDI)

The PDI, a critical parameter for assessing the uniformity of nanocrystal size distribution, increased with higher X₁ and more X₂. The impact of independent variables on the PDI of LZD-NCs was modelled using a two-factor interaction (2FI) based on the highest R² value. The polynomial equation (4) represents the influence of factors on the LZD-NCs PDI with coded values:

(4)

The positive coefficients for both X₁ and X₂ in Equation 4 denote that growing whichever factor marginally increases PDI, leading to a less uniform particle size distribution. The significant interaction term indicates that the combined effect of homogenization pressure and the number of cycles increases adaptability in particle size. This regimen focuses on a trade-off between obtaining smaller particles and maintaining uniformity, as excessive homogenization pressure or cycles may lead to aggregation or irregular particle sizes.

The stated PDI range of 0.218 to 0.524 ± 0.0825 falls within the adequate range (PDI < 0.7), pointing toward good uniformity. ANOVA exposed that the 2FI model was significant (p-value = 0.0107), and the adjusted R² of 0.5994 points to a moderate fit to the data. The non-significant lack-of-fit (p > 0.05) further provides for the model’s adequacy in describing the observed PDI. These results suggest that optimisation should settle homogenization pressure and cycle number to reduce PDI and maintain consistent performance in NIM formulations. Figure 3C illustrates the 2D contour plot for Y₂, indicating the minor rise in PDI as both X₁ and X₂ are higher. Figure 3D presents a 3D surface plot of PDI, showing that homogenization pressure and cycles contribute to a broader particle size distribution, thereby increasing PDI.

Impact on Zeta Potential (Surface Charge)

zeta potential, a main indicator of NCs stability, illustrated contrasting impacts altering on the process parameters. The linear model for zeta potential data was represented in terms of coded factors as given in the polynomial equation (5):

(5)

The positive coefficient for X₁ in Equation 5 indicates that increasing X₁ slightly increases zeta potential, improving the colloidal stability of LZD-NCs. In conflict, the negative coefficient on X₂ indicates that more cycles decrease the surface charge.

The observed zeta potential ranges from −22.3 to 3.14 mV, reflecting the balancing act required to achieve stable LZD-NCs, as an optimal zeta potential is necessary to prevent aggregation. ANOVA yielded a p-value of 0.0372, implying that the linear model accurately depicts the effect of the factors on zeta potential. However, the adjusted R² value of 0.3786 implies that additional factors not included in the model may affect zeta potential. Despite this, the non-significant lack-of-fit (p>0.05) supports the model’s suitability and specifies that it signified the experimental data well. Such results highlight the importance of varying the homogenization pressure and cycles to achieve the desired surface charge, ensuring the long-term stability of the LZD-NCs. Figure 3E shows the 2D contour plot for Y₃, showing the opposing effect of homogenisation pressure and number of cycles on surface charge. Figure 3F illustrates the 3D surface plot of zeta potential, emphasising the optimal balance required to achieve the desired zeta potential of LZD-NC stability.

Characterisation of LZD-NIM

Product Yield

The product yield of spray-dried NIM is approximately found to be 46.78%.

Surface Morphology

The FE-SEM study demonstrated that the LZD-NCs shown in Figure 1C and D, which were developed by HPH, were predominantly discrete, slightly irregular in shape, with relatively smooth surfaces.

Redispersibility Index

The RDI of LZD-NCs confirmed the efficient redispersion of microparticles into NCs when suspended in distilled water. The reconstituted NCs exhibited a mean particle size of 289.2 ± 18.25 nm and a PDI of 0.292 ± 0.07, and zeta potential of 23.12 ±4.86 mV. The RDI is a measure of the redispersibility of NIM formulation. The LZD-NIM shows an RDI of 1.026 ± 0.21, close to 1, indicating that the dry powder readily redisperses into NCs.

Drug Content

The LZD content in NIM was determined using UV spectroscopy as described in the method section and was found to be 96.48 ±0.30%.

Solid-State Characterisation

Fourier Transform Infrared Spectroscopy (FT-IR)

The FTIR spectrum of pure LZD proves the existence of its characteristic functional groups (Figure 4A). A peak at 3361.40 cm⁻¹ corresponds to N–H stretching of the amide group, while the peak at 2858.53 cm⁻¹ specifies –OCH₃ (methoxy group) stretching. The sharp peak at 1729.13 cm⁻¹ is endorsed to C=O (carbonyl) stretching vibrations, and 1119.09 cm⁻¹ corresponds to C–O stretching. Furthermore, a peak at 1586.93 cm⁻¹ corresponds to C=C stretching in the benzene ring of the LZD structure. These recorded peaks are consistent with stated data, confirming the structural integrity and purity of LZD. In the NIM, LZD retains its core functional groups, as proved by the continuation of characteristic peaks. Conversely, interactions with Poloxamer 188 and mannitol result in shifts and the appearance of new peaks, indicating molecular interactions and structural modifications. Poloxamer 188 contributes important peaks in the FTIR spectrum, like broad O-H stretching bands at 3766 cm⁻¹ and 3737 cm⁻¹, and a peak at 2861 cm⁻¹ related to aliphatic C-H stretching. These peaks are similarly observed in the LZD-NCs spectrum, indicating successful incorporation of Poloxamer 188 into the NIM formulation. Likewise, mannitol contributes peaks at 3410–3480 cm⁻¹ for O-H stretching and at 1080–1150 cm⁻¹ for C–O stretching, which are also evident in the LZD-NCs spectrum, further proving its incidence in the formulation. Shifts in specific LZD peaks in the NCs spectrum suggest interactions with excipients. Intended for example, the carbonyl (C=O) peak at 1729.13 cm⁻¹ shifts to 1516 cm⁻¹, possibly due to hydrogen bonding with Poloxamer 188 or mannitol. New peaks at 1448 cm⁻¹ and 1330 cm⁻¹ may reveal changes in molecular vibrations involved by encapsulation. In addition, in the fingerprint region, broadened peaks at 1085 cm⁻¹, 1022 cm⁻¹, and 931 cm⁻¹ suggest structural integration and potential bonding between LZD, Poloxamer 188, and mannitol. Lower-wavenumber peaks, such as 880 cm⁻¹ and 711 cm⁻¹, correspond to bending vibrations and support the stability of LZD within the NCs matrix.^26,34 These findings confirm that the NIM successfully integrates LZD, maintaining its structural integrity and ensuring molecular compatibility with the included excipients.

Figure 4.

Comparative solid-state characterisation of pure LZD, Poloxamer 188 (P-188), mannitol, and NIM. (A) Fourier-transform infrared spectra recorded from 4000 to 400 cm⁻¹, showing retention of the characteristic LZD bands together with the O–H and aliphatic C–H bands contributed by Poloxamer 188 and mannitol. (B) Differential scanning calorimetry thermograms recorded from 30 to 300 °C at 20 °C/min, showing the two melting endotherms of pure LZD at 152.5 and 184.6 °C, the melting endotherm of mannitol at 178.9 °C, and the single endotherm of the spray-dried formulation at 169.9 °C. (C) Thermogravimetric curves recorded from 20 to 600 °C at 20 °C/min under nitrogen, showing 0.10% residual mass for pure LZD versus 12.25% for the spray-dried formulation. (D) X-ray diffractograms recorded over a 2θ range of 5–60°, showing the sharp reflections of crystalline LZD and their retention, with reduced intensity, after spray drying.

Differential Scanning Calorimetry Study (DSC)

DSC analysis provides insight into the thermal behaviour and crystallinity of a drug. The DSC thermograms of pure LZD, mannitol and the NIM are shown in Figure 4B. Pure LZD exhibited two endothermic events: a smaller endotherm at 152.5 °C followed by a return towards the baseline, and a sharp principal melting endotherm at 184.6 °C. This melt-recrystallisation-melt sequence is characteristic of the polymorphism of LZD, in which a metastable form melts at the lower temperature and recrystallises to the thermodynamically stable form that subsequently melts at 184.6 °C; it does not indicate impurity of the drug substance, and the same profile was reproduced in an independent determination on a separate sample of LZD, which gave corresponding endotherms at 151.8 °C and 185.6 °C. Mannitol alone showed a single sharp melting endotherm with an onset at 162.0 °C and a peak at 178.9 °C, confirming its crystalline state. The spray-dried LZD-NIM exhibited a single sharp endotherm at 169.9 °C. No discrete melting event was observed at 52–55 °C, where free Poloxamer 188 melts; this event was absent when the nanosuspension was spray-dried without the preceding centrifugation step, confirming that centrifugation effectively removes the excess stabiliser from the formulation. The endotherm at 169.9 °C lies below the melting points of both pure LZD (184.6 °C) and mannitol (178.9 °C). It is therefore attributed to the combined melting of the two crystalline components, whose individual events are depressed and merged as a consequence of the reduced crystallite size produced by HPH and spray drying and of the intimate mixing of drug and carrier within the microparticles. Mannitol is therefore retained in the spray-dried formulation, as evidenced by the X-ray diffraction data presented below, where its characteristic reflections remain clear. The overall data of DSC suggest that LZD remains mainly crystalline after nanonization and spray drying process, with the observed melting point depression reflecting reduced crystalline size and closely related with mannitol rather than conversion to an amorphous form.

Thermogravimetric Analysis (TGA)

The thermal stability of pure LZD and LZD-NCs and LZD-NIM was analysed using TGA, as shown in Figure 4C. The results showed notable variations in thermal behaviour between pure LZD and the optimised NIM formulation, specifically in the final stages of heating. Pure LZD endured nearly complete degradation, with only 0.10% residue lasting by the end of the process. In contrast, the spray-dried LZD-NIM retained 12.25% of the formulation, suggesting enhanced thermal stability and incomplete degradation, likely due to the protective outcome of the nanocrystal matrix.

X-Ray Diffraction Analysis

The effects of nanosizing and particle-size reduction on the structural properties of the LZD were analysed using XRD, as shown in Figure 4D. The XRD patterns of pure LZD exhibit sharp and distinct peaks at 2θ values of 8.04, 10.48, 15.16, 17.72, 20.86, 22.56, 23.32, and 29.84, confirming its crystalline nature. Compared with the NIM formulation, the peaks are slightly shifted at 2θ, such as 8.14, 10.46, 14.26, 18.74, 21.16, 22.92, and 25.42. Few peaks present in pure LZD, namely 15.16 and 17.72, are either reduced in intensity or absent in the NIM formulation, while a new peak emerges at 183.74. 21.16 and 40.24. This observation suggests that the partial reduction in crystallinity may be due to nanonization during the HPH process. The crystalline peaks of mannitol, observed at 2θ values such as 11.46, 12.44, 15.62, 17.54, 19.76, and 22.08, remain clear in the NIM formulation. This reveals that mannitol retains its crystalline nature throughout the process. This reduced intensity and shift in LZD peaks, along with the mannitol peaks, suggest that spray drying mainly affects the LZD properties and reduces its crystallinity. This change in structural transformation enhances solubility and dissolution while maintaining the stability of the NIM formulation. No significant differences were observed between the pure LZD and the nanocrystal formulations, whether mannitol was included. The diffraction patterns of the NIM were largely similar to those of pure LZD and LZD-NCs containing mannitol, but with reduced peak intensity. NCs agglomerates, both with and without matrix formers, showed broadening and a further reduction in peak intensity compared to the starting materials. These results can be attributed to the smaller particle size of the NCs, dilution with stabilisers, and reduced strain during milling and processing. Our findings indicate that the crystalline nature of LZD remained unchanged in the spray-dried formulation, even after HPH and spray drying.

Surface Morphology of DPI

FE-SEM analysis of the LZD-NIM (Figure 5A and B) revealed predominantly discrete, nearly spherical particles with rough and wrinkled surfaces. Minimal agglomeration was observed, suggesting efficient spray drying and good particle separation. The wrinkled morphology is characteristic of spray-dried powders and is favourable for pulmonary drug delivery.

Figure 5.

FE-SEM of NIM. (A) Descriptive image illustrating mostly discrete, near-spherical particles with rough, wrinkled, and corrugated surfaces characteristic of NIM. (B) Morphology of the powder, indicating a comparatively uniform particle distribution with partial agglomeration.

In-vitro Release Study

Figure 6A depicts in-vitro drug release of pure LZD and LZD-NCs from NIM. The release study, conducted at pH 6.8, showed that the NIM exhibited slightly faster drug release than pure LZD. In particular, after 2 h, the cumulative pure LZD release was 7.93% and 18.44% for the LZD-NIM. The NIM formulation demonstrated a significantly improved release pattern, with 61.39% LZD released from the NIM at 24 h, compared to 37.26% for pure LZD. At 72 h, cumulative release reached 53.62% for pure LZD and 84.24% for the LZD-NIM, which highlights the markedly improved solubility and dissolution behaviour of LZD after HPH and spray drying.

Figure 6.

In vitro drug release and kinetic study for pure LZD and NIM (A) Cumulative drug release, (B) the Higuchi model (R² = 0.9918), (C) the first-order model (R² = 0.9597), (D) the zero-order model (R² = 0.9126), (E) the Hixson–Crowell model (R² = 0.9479), and (F) the Korsmeyer–Peppas model (R² = 0.5036). Data are mean ± SD (n = 3).

In Vitro Release Kinetics

The LZD release from the NIM was evaluated using different mathematical models to identify the best-fit model and elucidate the underlying release mechanism. Among these, the Higuchi model yielded the highest coefficient of determination (R² = 0.9918), indicating diffusion-controlled release, as shown in Figure 6B. The first-order model (R² = 0.9597) also provided a good fit, indicating that the release rate depends on the concentration of the drug remaining in the formulation, as shown in Figure 6C. The zero-order model (R² = 0.9126) indicates constant release over time, as shown in Figure 6D. The Hixson-Crowell mode (R² = 0.9479) indicates that changes in particle surface area and shape at the time of dissolution, which contribute to the release process, reflect microparticle erosion or restructuring, as shown in Figure 6E. Moreover, the Korsmeyer-Peppas model (R² = 0.5036) gives a poor fit, suggesting anomalous transport or a non-Fickian mechanism, which is not dominant in this system, illustrated in Figure 6F. Overall, this data suggests that LZD release from NIM is primarily controlled by diffusion, with minor contributions from erosion and concentration-dependent release.

In vitro Aerodynamic Performance

The aerosolisation behaviour of NIM was evaluated using NGI, including key parameters such as MMAD, FPF, and GSD. The results of this study demonstrated excellent in vitro aerosolisation properties, including an MMAD of 2.9 ± 0.15 µm, which ensures an optimal NIM size for deposition in the alveolar region. The FPF of 76.12 ± 1.3% mainly confirmed that the high proportion of particles was respirable, thereby enabling efficient delivery to the lower respiratory tract, where most Mtb infections occur. Moreover, the GSD of 1.6 ± 0.1 denotes a narrow, uniform particle size distribution, which is critical for consistent deposition of particles in the lungs. As shown in Figure 7, the NGI stage-wise deposition profile of NIM confirmed deep lung targeting with minimal retention in the upper respiratory tract.

Figure 7.

In vitro aerodynamic performance of NIM determined with a Next Generation Impactor (NGI) operated at 60 L/min, with the powder delivered from size 3 hard gelatin capsules using a Rotahaler device. The stage-wise deposition profile is shown for the induction port and impactor stages 1–7 (effective cut-off diameters 8.06, 4.46, 2.82, 1.66, 0.94, 0.55, and 0.34 µm). Data are mean ± SD (n = 3).

Short-Term Stability Study

Stability was assessed for the optimised LZD-NIM over a 6-month period at room temperature (25 ± 2°C). Important parameters, including particle size, PDI, and zeta potential, were assessed throughout the study. These parameters showed no significant changes during the stability testing, indicating that the formulation remained stable particle size, PDI, and zeta potential were selected as indicators of colloidal stability. Numerical values were recorded to monitor any alterations, with no observable changes in the colour of the NCs in the nanosuspension. Data for the mean particle size, PDI, and zeta potential are provided in Figure 8A and B. The NCs, stabilised with poloxamer 188, exhibited an average particle size of 276.37 ± 15.45 nm and a low PDI of 0.260 ± 0.099, indicating uniformity in particle size distribution and zeta potential −27.09 ± 1.748 mV.

Figure 8.

Short-term stability of NIM stored for six months at 25 ± 2 °C in sealed glass vials. (A) Mean particle size and PDI, and (B) zeta potential of the reconstituted formulation as a function of storage time. Data are mean ± SD (n = 3).

Discussion

In the present study, the development of the NIM system was designed to address two critical and interrelated challenges in pulmonary delivery of LZD: achieving aerodynamic behaviours suitable for efficient deposition in the deep lung while simultaneously enhancing the dissolution of this poorly water-soluble oxazolidine following deposition. By combining NCs within inhalable microparticles, the NIM platform was envisioned to provide appropriate particle size and aerosolization behaviour for alveolar delivery, along with the increased surface area and dissolution rate afforded by the nanocrystalline drug form. This approach addresses important challenges in delivering drugs via the pulmonary route, as NCs offer a high surface area and an enhanced dissolution rate; their very small aerodynamic size might lead to inefficient deep-lung deposition and augmented exhalation. In contrast, conventional microparticles generally possess more favourable aerodynamic properties for pulmonary deposition but may exhibit slower dissolution due to their larger particle size and lower specific surface area. The NIM approach, therefore, integrates NCs within inhalable microparticles to achieve a balance between aerodynamic deposition and rapid drug dissolution following deposition in the deep lung.^10,11 Embedding LZD-NCs in a mannitol matrix that rapidly disintegrates upon contact with lung lining fluid was intended to resolve the competing constraints of pulmonary deposition and rapid drug dissolution. The redispersibility index of 1.026 ± 0.21 supports effective recovery of the NCs.

following rehydration, which indicates that incorporation within the microparticle matrix did not substantially compromise the nanoscale characteristics of the LZD-NCs. During the development of LZD-NCs, the choice of stabiliser significantly influenced the physicochemical properties of the resultant nanosuspension. PVCL-PVAc-PEG stabilised nanosuspension exhibited the smallest particle size, ie 148.8 nm, whereas the SLS-stabilised formulation showed the narrowest particle size distribution (PDI 0.190). However, neither stabiliser provided a favourable balance between particle size and size distribution, as the low PDI obtained with SLS was accompanied by a substantially larger particle size of 568.2 nm. Poloxamer 188 was chosen because the resulting nanosuspension exhibited a more balanced profile, with an acceptable particle size (289.4 nm), a relatively low PDI (0.295), and a zeta potential indicative of satisfactory colloidal stability. This choice is also consistent with the established use of non-ionic poloxamers as stabilising excipients in pharmaceutical formulations. In contrast, the PVP K-30 and PVA stabilised formulation exhibited larger particle sizes (481.3 and 618.7 nm, respectively) and broader size distribution (PDI 0.686 and 0.952), suggesting that these stabilisers were comparatively less effective under the applied HPH conditions. The CCD provided useful insights into the influence and interactions among the process variables, rather than serving solely as a predictive optimisation tool. Analysis of variance revealed that the fitted models for mean particle size and zeta potential were statistically significant (p = 0.0399 and p = 0.0372 correspondingly), with a non-significant lack of fit revealing model capability. Even though, the coefficients of determination (R² =0.3900 and 0.3786) reflect the inherent complexity of the colloidal system, they demonstrate that the selected formulation parameters exert a measurable and statistically reliable influence on both critical quality attributes, thereby confirming their relevance for preliminary factor screening. The statistical significance of the model (p < 0.05) confirms that the selected formulation and process parameters show a measurable impact on average particle size and surface charge. While the modest R² value reflects the inherent complexity of the colloidal nanosuspension, the PDI model yields a substantially high adjusted R² (0.5994), revealing a significant 2FI interaction between the homogenisation pressure and the number of homogenisation cycles, demonstrating that the design effectively captures the main interdependence governing the size distribution. Collectively, these findings support the practical utility of the design space approach for identifying a robust operating region that optimally balances particle size reduction, uniform size distribution, and colloidal stability, contributing a more versatile approach than relying on a single predicted optimum. The dissolution performance of the NIM system represents a substantial improvement over pure LZD, with significantly improved release throughout the study period. The excellent fit to the Higuchi model (R² = 0.9918) provides compelling evidence that diffusion-controlled release is the predominant mechanism, affording a well-characterised and probable release profile. In particular, the sustained-release behaviour of the present formulation compares favourably with LZD-loaded polycaprolactone microspheres (Sharma et al¹⁷) and rifapentine-LZD PLGA microspheres (Huang et al¹⁵), while offering the distinct advantage of maintaining an inhalable dosage form suitable for pulmonary administration. The present NIM system, therefore, signifies a favourable approach for achieving sustained pulmonary delivery of LZD. Solid-state analysis showed that LZD remained mostly crystalline even after formulation and spray-drying. This was confirmed by XRD, in which the main crystal peaks were still present, though less intense. The melting point also shifted from 184.6 °C to 169.9 °C. These changes are attributed to the formation of nanocrystalline domains, a reduction in primary crystal size, and dilution of the drug within the mannitol carrier matrix. Retention of the crystalline state of LZD in NIM formulation confirms a significant advantage for physical stability, as the amorphous state is more prone to recrystallisation during storage, which may affect the dissolution and aerosolisation behaviour of the formulation. The developed NIM formulation has excellent colloidal stability, as evidenced by the absence of noticeable changes in mean particle size, PDI, and surface charge during storage for six months. Our findings collectively suggest the feasibility of the NIM system for normal storage and handling under non-refrigerated conditions. In particular, a preliminary stability study suggests that a cold chain storage facility may not be a prerequisite for maintaining formulation integrity that can confer extensive logistical merit. Additionally, this hypothesis remains to be conclusively validated via extended and long-term along with accelerated stability studies which are important for establishing the definitive storage guidelines and predicting the shelf life of NIM formulation.

Conclusion

An LZD-NCs-based inhalable dry powder formulation was fabricated using HPH followed by spray drying. CCD was used to optimise the formulation. The critical process parameters, specifically the homogenization pressure and the number of cycles, were identified as the main factors influencing formulation performance. The selected parameters significantly affected essential properties, including average particle size, PDI, and zeta potential, as well as yield, redispersibility, dissolution rate, and the preservation of the XRD pattern of LZD-NCs relative to pure LZD. The approach used to develop NIM is simple and results in improved LZD solubility. Our findings demonstrate the important role of HPH and spray drying in developing the novel NIM formulation with an enhanced dissolution rate, supporting effective pulmonary delivery. In vitro aerodynamic testing confirmed the formulation’s suitability for inhalation therapy and demonstrated its potential for TB treatment.

Funding Statement

This study was financially supported by the Anusandhan National Research Foundation (ANRF), New Delhi, in terms of Empowerment and Equity Opportunities for Excellence in Science (EEQ/2023/000060).

Data Sharing Statement

All data supporting the findings of this study are available from the corresponding author, Dr Eknath Kole, upon reasonable request.

Author Contributions

All authors made a significant contribution to the work reported, whether in the conception, study design, execution, acquisition of data, analysis, and interpretation, or in all these areas; participated in drafting, revising, or critically reviewing the manuscript; gave final approval of the version to be published; agreed on the journal to which the article has been submitted; and agreed to be accountable for all aspects of the work.

Disclosure

The authors declare that they have no known competing financial interests or personal relationships that could have influenced the work reported in this paper.