International Journal of Pharmaceutics 682 (2025) 125952

Contents lists available at ScienceDirect

International Journal of Pharmaceutics

journal homepage: [Link]/locate/ijpharm

Development of inhalable spray dried nalbuphine hydrochloride powders

Waiting Tai a , Hong-Jaan Wang b , Dipesh Khanal a , Pancy Tsz Hei Kwong a, Patricia Tang a ,
Chih-Chin Shih c,* , Hak-Kim Chan a,*
a
Advanced Drug Delivery Group, Sydney Pharmacy School, Faculty of Medicine and Health, The University of Sydney, Sydney, NSW 2006, Australia
b
School of Pharmacy, National Defense Medical Center, Taipei, Taiwan, Republic of China
c
Department and Graduate Institute of Pharmacology, National Defense Medical Center, Taipei, Taiwan, Republic of China

A R T I C L E I N F O A B S T R A C T

Keywords: Nalbuphine hydrochloride (NAL) is a potent opioid analgesic that has been used clinically for decades. However,
Nalbuphine hydrochloride current delivery methods are either invasive or inefficient due to low bioavailability. Therefore, an alternative
Opioids route of administration is needed. Pulmonary delivery is a promising option as it can provide high bioavailability
Inhalation
and a rapid onset of action. Beyond systemic analgesia, NAL has also shown local therapeutic benefits in the
Pulmonary delivery
lungs. This study aimed to develop inhalable NAL powders using spray drying, with leucine and human serum
Dry powder formulation
Leucine albumin (HSA) as dispersion enhancers. The resulting powders were amorphous and contained <4 % residual
Human serum albumin solvent by mass. The spray dried particles were spherical, 3–4 µm large, and exhibited corrugated surface. Spray
dried NAL alone were inhalable, with a fine particle fraction <5 µm (FPF) at 25 %. The addition of HSA improved
the FPF to 33 % and 37 % for 10 % and 20 % w/w HSA, respectively. Similarly, the FPF increased to 39 % after
adding 10 % w/w leucine, reaching a maximum of 47 % with 20 % w/w leucine. The latter formulation also
demonstrated good chemical and physical stability after one month of storage at 25 ◦ C under both 30 % and 60 %
relative humidity. These formulations can be used as a platform to deliver NAL and other opioid analgesics by
inhalation.

1. Introduction inclusion of paediatric subjects, who naturally have more variable PK

profiles than adults (Kuznetsov and Tymko, 2024). In contrast, intra­
Nalbuphine hydrochloride (NAL) has been used clinically for de­ nasal administration demonstrated a bioavailability of 65 % (Tymko
cades as a potent analgesic. It is a semi-synthetic opioid that is equi­ et al., 2024) and has proven effective in treating pain associated with
potent to morphine (Liang et al., 2020). However, unlike morphine, conditions such as prehospital trauma (Pietsch et al., 2021) and post­
which is a full mu opioid receptor agonist, NAL acts as a mu opioid re­ operative recovery (Tymko et al., 2024), supporting its clinical feasi­
ceptor antagonist and a kappa opioid receptor agonist (Chen et al., bility. Based on this, a pharmaceutical company in Ukraine developed a
2024). These mechanistic differences confer NAL a more favourable NAL nasal spray, Apain® (Tymko et al., 2024; Tymko et al., 2023).
safety profile, with fewer adverse effects and a lower risk of addiction. Although its official dosing instructions are unavailable, human PK and
Due to its low abuse potential, NAL is the only opioid in the United States efficacy studies used a 10.5 mg dose (administered as three sprays) at
that is not classified as a controlled substance (Narver, 2015). intervals of every 5–6 h, consistent with the 5-h plasma half-life of NAL
NAL is primarily administered as a solution for intravenous, intra­ (Tymko et al., 2024; Tymko et al., 2023). However, frequent intranasal
muscular, and subcutaneous injection. Although oral NAL tablets are administration may damage the nasal mucosa. Moreover, concentrated
available, their use is limited due to low oral bioavailability (11–25 %) aqueous NAL solutions (i.e. 25 mg/mL in water) are chemically unstable
(Tymko et al., 2024; Wang et al., 2014). To improve its delivery, pre­ due to oxidative degradation and precipitation at low temperatures
vious studies have explored alternative route of administration, (Kuznetsov and Tymko, 2023). For instance, Apain solutions reportedly
including rectal and intranasal delivery. While rectal NAL was rapidly changed colour from light blue to pink after prolonged storage, although
absorbed and provided adequate analgesia, it exhibited high inter- they remained stable at room temperature for at least 28 days
individual pharmacokinetic (PK) variability, likely due to the (Kuznetsov and Tymko, 2024). Such instability raises concerns about

* Corresponding authors.
E-mail addresses: ccshih@[Link] (C.-C. Shih), [Link]@[Link] (H.-K. Chan).

[Link]
Received 15 May 2025; Received in revised form 4 July 2025; Accepted 7 July 2025
Available online 9 July 2025
0378-5173/© 2025 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license ([Link]
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

Table 1 and minor surgeries (Narver, 2015). Beyond systemic analgesia, inhaled
Solute mass ratios of the spray dried formulations. NAL may also offer local therapeutic benefits in the lungs. For example,
Formulations Nalbuphine Leucine Human serum albumin (HSA) NAL demonstrated anti-inflammatory effects on lung macrophages
(Zeng et al., 2020) and has been used in patients undergoing thoraco­
SD NAL 100 % 0% 0%
NAL10L 90 % 10 % 0% scopic lobectomy to reduce postoperative inflammation (Zhang et al.,
NAL10HSA 90 % 0% 10 % 2017). It may also relax pulmonary arteries (Sun et al., 2006) and sup­
NAL20L 80 % 20 % 0% press the growth of non-small cell lung cancer cells (while sparing
NAL20HSA 80 % 0% 20 % normal epithelial cells) via kappa-opioid receptor activation (Kuzumaki
et al., 2012). Recent clinical trials further suggest that NAL may reduce
cough in patients with idiopathic pulmonary fibrosis (Maher et al.,
Table 2 2023) and protect the lungs when co-administered with dexmedetomi­
The production yields, volumetric diameters, and spans of the spray dried dine by improving oxygenation and reducing pulmonary fluid content
formulations. (Wang et al., 2022). These emerging therapeutic potentials support the
Formulations Production yield D10 D50 D90 Span rationale for pulmonary delivery of NAL.
(%) (µm) (µm) (µm) This study aimed to develop inhalable NAL powders using spray
SD NAL 80.4 0.7 ± 2.8 ± 6.3 ± 2.0 ± drying, with and without the addition of L-leucine (hereafter referred to
0.1 0.1 0.3 0.1 as leucine) or HSA, which were used as dispersion enhancers and
NAL10L 84.5 0.6 ± 3.2 ± 7.7 ± 2.3 ± powder stabilisers. Leucine is an endogenous, generally regarded as safe
0.1 0.1 0.9 0.2
NAL10HSA 83.9 0.9 ± 3.5 ± 7.1 ± 1.8 ±
(GRAS) amino acid used as an excipient in inhalable dry powder for­
0.1 0.1 0.8 0.2 mulations due to its distinct physicochemical and aerodynamic advan­
NAL20L 84.6 0.7 ± 3.7 ± 6.8 ± 1.6 ± tages (Alhajj et al., 2021). Leucine preferentially migrates to the particle
0.1 0.2 0.3 0.1 surface during spray drying and forms a hydrophobic shell around
NAL20HSA 79.2 1.1 ± 3.9 ± 7.7 ± 1.7 ±
particles, reducing cohesive forces and improving dispersibility (Ke
0.1 0.1 0.2 0.1
et al., 2022). Leucine can also produce wrinkled or corrugated particles
to enhance lung deposition (Ke et al., 2022). On the other hand, human
efficacy loss, exposure to degradation products, and safety risks. serum albumin (HSA) is a non-immunogenic and highly biocompatible
To address these issues, a promising alternative is pulmonary de­ protein used in clinical settings as an excipient or stabiliser, demon­
livery of NAL in the form of inhalable powders. Pulmonary delivery has strating regulatory acceptance (Chow et al., 2019). For example, it has
been shown to achieve high bioavailability, rapid onset of action, less been used in approved injectable formulations, such as Abraxane® and
adverse effects, and improved convenience and tolerance across various Victoza® (Murphy et al., 2025). Although HSA is an endogenous and
populations (Tai and Kwok, 2022). These advantages have already been abundant protein present in the lungs (Bosquillon et al., 2001), limited
demonstrated with inhaled opioids such as fentanyl, which achieved human data are available on its inhalation (Todisco et al., 1990) and its
comparable bioavailability and time to reach maximum concentration potential adverse pulmonary effects remain unknown. Nonetheless, HSA
than intravenous injection (Macleod et al., 2012). A fast onset of action contributes to a more stable powder matrix and optimises particle
is particularly important for NAL, given its use in emergency medicine morphology, and has therefore been used in multiple inhalable powders

Fig. 1. Scanning electron microscopy images of (a) SD NAL, (b) NAL10L, (c) NAL10HSA, (d) NAL20L, and (e) NAL20HSA.

2
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

use.

2.3. Particle size distribution

The particle size distribution of the powders was measured at 4 bar

air pressure using a Mastersizer 3000 with an Aero S dry powder
accessory (Malvern Panalytical Ltd., Malvern, United Kingdom). The
refractive indices of NAL and leucine were 1.702 and 1.463, respec­
tively. The refractive index of HSA was not reported before, and hence
that of bovine serum albumin (1.602) was used instead. For formula­
tions, the refractive indices were calculated based on the mass ratio of
each component, as shown in Table 1. Measurements were conducted in
triplicate. The reported D10, D50, and D90 were the volumetric equiva­
lent diameters at 10 %, 50 %, and 90 % undersize, respectively. The span
((D90-D10)/D50) were also reported to describe the width of the size
distribution.

2.4. Particle morphology

The morphology of spray dried particles was investigated using a

scanning electron microscope (SEM; Zeiss Sigma VP HD; Zeiss, Ober­
kochen, Germany). Samples were spread on the carbon tabs (PELCO®;
Ted Pella, California, United States), and then coated with 15 nm of gold
using a CCU-010 sputter coater (Safematic, Zizers, Switzerland). Images
were acquired at 5 kV beam accelerating voltage and 20,000 x
magnification.

2.5. Droplet contact angle

Droplet contact angles were measured using a goniometer (Ramé-

Fig. 2. Droplet contact angles of (a) nalbuphine hydrochloride solution (NAL),
Hart instruments, Succasunna, NJ, USA). A 5 µL drop of NAL solution,
(b) nalbuphine hydrochloride + 10 % w/w leucine solution (NAL10L), (c)
nalbuphine hydrochloride + 10 % w/w human serum albumin solution
with and without excipients (leucine or HSA) at a concentration of 16
(NAL10HSA), (d) nalbuphine hydrochloride + 10 % w/w leucine solution mg/mL (the same as used for spray drying) was applied onto a glass
(NAL20L), and (e) nalbuphine hydrochloride + 20 % w/w human serum al­ slide. Image of the droplets was acquired immediately and analysed
bumin solution (NAL20HSA). using ImageJ (Schneider et al., 2012). Triplicate measurements were
performed for each sample.
(Chow et al., 2019; Li et al., 2025; Maa et al., 1997; Seville et al., 2007;
Tai et al., 2025). In this study, the physiochemical properties and aerosol 2.6. Optical Photothermal Infrared Spectroscopy (O-PTIR)
performance of the spray dried powders were examined. The one-month
storage stability of the most dispersible powder was also assessed. The NAL-excipient interactions and the particle surface enrichment
of excipients were determined using optical photothermal infrared
spectroscopy (O-PTIR). Powders were sprinkled onto calcium fluoride
2. Materials and methods
substrates (10 × 0.35 mm; Crystan, Dorset, United Kingdom) and ana­
lysed using a mIRage-LS microscope (Photothermal Spectroscopy Cor­
2.1. Materials
poration, California, USA) equipped with a 40 x all-reflective Cassegrain
objective and a silicon photodetector. A HYPERspectra Quantum
NAL was purchased from PhytoHealth Corporation (Taipei, Taiwan,
Republic of China). Leucine, HSA, sodium dihydrogen phosphate, and Cascade Laser (tunable wavelength range: 2990–2700 and 1890 to 790
cm− 1), collinear with a continuous-wave visible detection laser (532
high-performance liquid chromatography (HPLC) grade acetonitrile
were bought from Sigma-Aldrich (St. Louis, United States). A Milli-Q® nm), was used. The excitation laser was pulsed at 100 kHz with a pulse
width of 100 ns to collect the O-PTIR spectra at a spectral resolution of 2
Direct Water Purification System (Millipore, Massachusetts, United
States) was used to obtain ultrapure water (resistivity = 18.2 MΩ⋅cm). cm− 1. The IR and probe beam power was 20 % and 6 %, respectively. O-
PTIR spectra for more than 100 particles were averaged followed by
second derivative analysis using a five-point Savitzky–Golay smoothing
2.2. Spray drying function with a third-order polynomial.
Chemical maps were obtained at wavenumbers specific to NAL
Powders (Table 1) were prepared using a Büchi S300 spray dryer (1117 cm− 1), leucine (1465 cm− 1), and HSA (1542 cm− 1) with a step
coupled with a conventional two-fluid nozzle (BUCHI Labortechnik AG, size of 300 nm. Overlaid images of NAL to excipient (leucine or HSA)
Flawil, Switzerland). These formulations contained either 10 % w/w or were used to qualitatively evaluate particle surface enrichment in each
20 % w/w of leucine (NAL10L or NAL20L) or HSA (NAL10HSA or sample. All spectral acquisition, image capture, and analysis were per­
NAL20HSA) as excipients. NAL was spray dried individually as a control formed using PTIR studio 4.0 software (Photothermal Spectroscopy
(SD NAL). Feed solutions (16 mg/mL in ultrapure water) were atomised Corporation, California, USA).
at a feed rate of 4 mL/min, with an inlet temperature of 140 ◦ C, outlet
temperature of 70–72 ◦ C, atomising gas flow of 742 L/min, and aspi­ 2.7. X-ray powder diffraction (XRD)
ration rate at 23 m3/h. The powders were collected into a plastic
container inside an acrylic box maintained at relative humidity (RH) < Powder crystallinity was evaluated using a SmartLAB SE powder
15 % and stored in a desiccator at 25 ± 2 ◦ C and 30 % RH prior to further diffractometer (Rigaku Corporation, Tokyo, Japan). The powders were

3
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

Fig. 3. O-PTIR chemical maps of NAL10L, NAL10HSA, NAL20L, and NAL20HSA. The left panel shows the chemical maps of nalbuphine hydrochloride (NAL)
collected at 1117 cm− 1. The middle panel depicts the chemical maps of leucine (collected at 1465 cm− 1) or human serum albumin (HSA; collected at 1542 cm− 1). The
right panel presents the overlaid maps of the two corresponding wavenumbers.

first loaded on zero diffraction plates and then subjected to copper Kα powders. Approximately 5 mg powders were weighed in an open
radiation at room temperature. The current was 45 mA, while the alumina crucible and heated from 25-160 ◦ C at a constant heat rate of
voltage was 40 kV. The scattered intensity of the samples was recorded 10 ◦ C/min under 20 mL/min nitrogen flow. The mass change of the
from 5 to 50◦ at a scan speed of 5◦ per min and step size of 0.01◦ . powder was recorded.

2.8. Thermogravimetric analysis (TGA) 2.9. Differential scanning calorimetry (DSC)

A TGA/DSC 1 with the STARe System (Mettler Toledo, Zürich, The thermal events of the spray dried powders were determined
Switzerland) was used to determine the residual solvent content in the using a differential scanning calorimeter (DSC 1 model with the STARe

4
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

Fig. 4. X-ray powder diffraction patterns of raw materials and spray dried powders containing nalbuphine hydrochloride (NAL), leucine, and human serum albu­
min (HSA).

increments, with the equilibrium moisture content criterion of each step

at ≤0.002 % per min.

2.11. Density measurement

The bulk densities (ρb) of the formulations were measured by accu­

rately weighing 1–2 cm3 of powder in a 10 cm3 graduated cylinder. The
powder in the cylinder was then tapped approximately 1200 times using
a tapping apparatus (Surface Measurement Systems Ltd, London, United
Kingdom) and the final volume was recorded to determine the tapped
density (ρt). Triplicate measurements were performed for all samples.
The flowability of the powders was assessed by calculating Carr’s index
using Equation (1).
(ρt − ρb )
Fig. 5. Thermogravimetric analysis of raw materials and spray dried powders Carr’s index = × 100 (1)
containing nalbuphine hydrochloride (NAL), leucine, and human serum albu­
ρt
min (HSA).
2.12. Aerosol performance
System; Mettler Toledo, Zürich, Switzerland). Aluminium pans loaded
with about 5 mg samples were crimp-sealed with lids with pinhole. The The aerosol performance of the spray dried powders was examined
thermograms were obtained by heating the samples from 25 to 350 ◦ C at using the Next Generation Impactor (NGI; Copley, Nottingham, United
a constant rate of 10 ◦ C/min under 20 mL/min nitrogen flow. Kingdom) with the United States Pharmacopeia throat. Prior to disper­
sion, the collection cups inside the NGI were coated with a thin layer of
silicone (Slipicone, Victoria, Australia) to avoid particle bounce. Powder
2.10. Dynamic vapour sorption (DVS) (20 ± 2 mg) was filled into a size 3 hydroxypropyl methylcellulose
capsule (Capsugel, Sydney, Australia) and dispersed using a high resis­
The moisture sorption–desorption profiles of the powders were ob­ tance RS01 inhaler (Plastiape S.p.A., Osnago, Italy). This device was
tained using dynamic vapour sorption (DVS; DVS-Intrinsic, Surface selected because it is an approved dry powder inhaler for therapeutic use
Measurement Systems, London, United Kingdom). Approximately 10 mg (not diagnostic testing) and requires only moderate inhalation effort
powder was equilibrated in the chamber at 25 ◦ C/0% RH under from patients to achieve sufficient lung deposition at a low airflow rate
continuous nitrogen flow. It was then subjected to two adsorp­ (50–70 L/min), which also helps reduce oropharyngeal deposition
tion–desorption cycles from 0-90 % RH. RH was increased in 10 % (Yang et al., 2016). The inhaler was connected to the throat with a

5
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

The ratio between the MMAD and the aerodynamic diameter at 16 %

undersize was defined as the GSD.

2.13. Stability test

Stability test was conducted using the formulation with the best
aerosol performance (i.e. the highest FPFs). It was stored as a loose
powder in a plastic container and directly exposed at 25 ◦ C/30 % RH (as
a control) and 25 ◦ C/60 % RH for one month. After storage, the aerosol
performance of the powder was re-evaluated using the same method
described in Section 2.9.

2.14. Drug quantification

NAL was quantified with a HPLC system consisting of a Shimadzu

CBM-20A controller, a LC-20AT pump, a SIL-20A HT autosampler, a
SPD-20A UV/VIS detector, and a Waters Nova-Pak C18 column (4 µm;
150 × 3.90 mm). The mobile phase was 70:30 v/v 0.05 M sodium
dihydrogen phosphate buffer (pH 3.5):acetonitrile, constantly flowing at
1 mL/min. The UV detection wavelength was 230 nm. The standard
solutions were prepared in ultrapure water by serial dilution. The cali­
bration curves were linear over the concentration range of 3.9 – 1000
µg/mL (r2 > 0.99).

2.15. Statistical analysis

Data was analysed using one-way analysis of variance (ANOVA)

followed by Tukey’s post hoc test. Differences were considered statisti­
cally significant with a p < 0.05 and α < 0.05.

3. Results

The production yields of the spray dried powders were 80–85 %, as

Fig. 6. Differential scanning calorimetry thermograms of (a) raw materials and shown in Table 2. The D50 of the particles in all formulations were below
(b) spray dried powders containing nalbuphine hydrochloride (NAL), leucine,
4 µm, with span values about 2 (Table 2). SEM images showed that SD
and human serum albumin (HSA).
NAL particles were corrugated (Fig. 1a), while those containing leucine
or HSA appeared more wrinkled (Figs. 1b–e). The surface of NAL20L
appeared rougher than that of NAL10L (Figs. 1b and d), whereas the
Table 3
surfaces of NAL10HSA and NAL20HSA were similar (Figs. 1c and e).
Thermal events (in ◦ C) of raw materials and spray dried powders containing
Fig. 2 shows the droplet contact angle of the NAL and excipient-
nalbuphine hydrochloride (NAL), leucine, and human serum albumin (HSA).
containing solutions. The contact angles of NAL solution (50.8 ± 3.0◦ )
Formulations Crystallisation Sublimation Melting Decomposition
and NAL10L solution (45.4 ± 1.5◦ ) were comparable. Both were
Raw NAL / / 293.8 308.3 significantly larger than those of NAL10HSA (39.5 ± 2.4◦ ), NAL20L
Raw leucine / 319.3 / / (37.5 ± 1.2◦ ), and NAL20HSA (34.3 ± 1.5◦ ), which were comparable to
Raw HSA / / / /
each other.
SD NAL 211.8 / 290.5 308.5
NAL10L 208.4 257.7 280.9 305.0 Chemical maps (Fig. 3) collected from spray-dried NAL particles
NAL10HSA 211.8 / 281.0 304.9 containing either leucine or HSA at specific wavenumbers—1117 cm− 1
NAL20L 201.7 / 263.8 301.7 (NAL, C–O stretching), 1465 cm− 1 (leucine, CH2 bending), and 1542
NAL20HSA 218.3 / 274.6 301.6
cm− 1 (HSA, N–H bending)—confirmed the homogeneous distribution of
both NAL and the excipients. The overlaid chemical maps of NAL with
silicone adapter. The dispersion was conducted at 4 kPa pressure drop either leucine or HSA showed a pink coloration, indicating the overlap of
(equivalent to 57 L/min). The aerodynamic cut-off diameters of the NGI blue (excipient) and red (NAL) signals, which suggests surface enrich­
stages (from Stages 1 to 8) are >8.29, 8.29, 4.58, 2.89, 1.70, 0.97, 0.57, ment of the particles with the respective excipient. However, depending
and 0.35 µm. The duration of dispersion was 4.3 s, allowing 4 L of air to on the concentration of leucine or HSA, faint blue regions were still
pass through the inhaler. The drug deposited on the capsule, inhaler, visible, indicating areas with lower NAL presence. In the overlaid map of
adapter, throat, and the NGI stages was rinsed with 5 mL ultrapure water NAL20L, particles exhibited more red and pink clusters compared to
and quantified by HPLC. Triplicate dispersions were conducted for each NAL10L, suggesting greater surface coverage at higher leucine concen­
formulation. Emitted fraction (EF), fine particle fractions (FPFs) <5 µm trations. A similar trend was observed between NAL10HSA and
and <3 µm, mass median aerodynamic diameter (MMAD), and geo­ NAL20HSA. At equivalent mass concentrations, HSA appeared to pro­
metric standard deviation (GSD) were calculated. The EF was the per­ vide more extensive surface coverage than leucine.
centage of emitted dose (sum of mass deposited on adaptor, throat, and Raw NAL and leucine were crystalline, as evidenced by their distinct
the NGI stages) with respect to the recovered dose. The FPFs were peaks in the X-ray diffractograms in Fig. 4. In contrast, raw HSA showed
defined as the percentages of fine particle doses (<5 µm and <3 µm) a halo pattern, indicating its amorphous nature. Similarly, all spray
relative to the recovered dose. The aerodynamic diameter that split the dried formulations also showed halo patterns (Fig. 4), confirming their
aerodynamic particle size distribution in half by mass was the MMAD. amorphous state. However, a faint peak at 6◦ was observed in NAL20L.
TGA results are depicted in Fig. 5. Raw NAL contained 4.5 % residual

6
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

Fig. 7. O-PTIR spectra of (a) raw leucine, SD NAL, NAL10L, and NAL20L and (b) raw human serum albumin (HSA), SD NAL, NAL10HSA, and NAL20HSA.

solvent. Spray drying reduced this to 3.6 % in SD NAL. The addition of desorption, SD NAL retained 8 % w/w more mass than baseline (Fig. 8a),
HSA slightly decreased the residual solvent content to 3.0–3.4 %, while close to the theoretical water content (8.38 % w/w) required for com­
leucine reduced it further to 2.5–2.8 % (Fig. 5). plete transformation to the dihydrate form. The other recrystallised
Fig. 6a shows the DSC results of the raw materials. No thermal event powders retained only 6–7 % moisture (Figs. 8b–e), indicating partial
was observed for raw HSA. Raw leucine displayed an endothermic peak transformation.
beginning at 284.8 ◦ C (Fig. 6a), indicating decomposition (Pokorný Table 4 presents the bulk and tapped densities of the spray dried
et al., 2020). Raw NAL showed an endothermic melting peak at 293.8 ◦ C powders. The bulk and tapped densities of SD NAL was 0.2 ± 0.0 and 0.3
followed by an exothermic decomposition peak at 308.3 ◦ C (Fig. 6a) ± 0.0 g/cm3, respectively. The addition of 10 % w/w leucine to the
(Hanko and Vlahova, 2011). These thermal events were also observed in formulation increased the bulk density (0.3 ± 0.0 g/cm3) but did not
the spray dried powders, with melting between 263.8–290.5 ◦ C and affect the tapped density, which remained at 0.3 ± 0.0 g/cm3. For
decomposition between 301.7–308.5 ◦ C (Fig. 6b; Table 3). An NAL20L, both bulk and tapped densities increased to 0.3 ± 0.0 and 0.5
exothermic event between 208.4–218.3 ◦ C was present in all formula­ ± 0.0 g/cm3, respectively. In contrast, the addition of 10 % and 20 % w/
tions (Fig. 6b; Table 3), corresponding to crystallisation from the w HSA decreased the bulk density to 0.1 ± 0.0 g/cm3 and tapped density
amorphous state. NAL10L also exhibited an additional endothermic to 0.2 ± 0.0 g/cm3. The Carr’s index of SD NAL was 14.5 ± 4.3 %,
peak at 257.7 ◦ C, indicating leucine sublimation (Li et al., 2016), which indicating good flowability (Seville et al., 2007). NAL10L exhibited the
was not observed in NAL20L (Fig. 6b; Table 3). lowest Carr’s index at 4.1 ± 2.1 %, classifying it as a powder with
The O-PTIR spectra of the raw materials and formulations are shown excellent flowability (Seville et al., 2007). However, NAL10HSA,
in Fig. 7. SD NAL exhibited peaks at 1559 (C-N stretching), 1385, 1323 NAL20L, and NAL20HSA had Carr’s index values of 36.8 ± 2.6 %, 41.0
(–CH2 wagging), 1463 (CH2 bending), and 1167 cm− 1 (C-O stretching) ± 2.8 %, and 27.9 ± 4.6 %, respectively, suggesting poor flowability
(Khanna et al., 2025; Mestek, 2025). Similar peaks at 1385, 1323, and (Seville et al., 2007).
1167 cm− 1 were also observed in NAL10L, but the peak at 1559 cm− 1 The aerodynamic particle size distributions of the spray dried pow­
was shifted to 1564 cm− 1. Additional peak shifts were found in NAL20L, ders are illustrated in Fig. 9. SD NAL showed the highest drug retention
including from 1323 to 1305 cm− 1 and from 1167 to 1191 cm− 1. While in the capsule and inhaler, as well as throat deposition. These were
the peak at 1385 cm− 1 disappeared in the spectrum of NAL20L, new reduced by adding 10 % or 20 % (w/w) of either leucine or HSA, with
peaks emerged at 1407, 1591, and 1359 cm− 1. On the other hand, for greater improvements observed at 20 % (Fig. 9). Leucine was more
NAL10HSA and NAL20HSA, the peaks around 1664 cm− 1 (amide I band effective than HSA at the same mass concentration, resulting in less
of HSA) shifted toward 1641 cm− 1 with decreasing HSA content. New retention and more drug deposition in the later NGI stages (Fig. 9). The
peaks appeared at 1063 and 1031 cm− 1 in both NAL10HSA and MMADs of SD NAL (2.9 ± 0.2 µm) and NAL20L (3.1 ± 0.2 µm) were
NAL20HSA spectra. A decrease in peak intensity with increasing HSA comparable and significantly smaller than those of NAL10L (3.7 ± 0.2
content was observed at multiple wavenumbers, including 1509, 1375, µm), NAL10HSA (3.9 ± 0.1 µm), and NAL20HSA (4.0 ± 0.0 µm). The
and 1115 cm− 1. GSD values of SD NAL (2.1 ± 0.0), NAL10HSA (2.1 ± 0.0), and
Moisture sorption profiles showed that SD NAL adsorbed 14 % w/w NAL20HSA (2.0 ± 0.0) were comparable, and all were significantly
moisture at 80 % RH, with a sharp mass loss (Fig. 8a). Similar behaviour lower than those of NAL10L (2.7 ± 0.3) and NAL20L (4.5 ± 0.3), with
was observed in NAL10L (12 %), NAL20L (11 %), and NAL10HSA (16 %) the difference between the two leucine formulations also being
(Figs. 8b–d). NAL10L also showed a slight mass decrease at 60 % RH, significant.
suggesting leucine recrystallisation. In NAL20HSA, recrystallisation The EF of SD NAL was 78.4 ± 0.7 %, which was significantly lower
occurred at 90 % RH after adsorbing 18 % w/w moisture (Fig. 8e). Upon than those of the other four formulations (Fig. 10). While comparable

7
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

Fig. 8. Moisture sorption profiles of (a) SD NAL, (b) NAL10L, (c) NAL10HSA, (d) NAL20L, and (e) NAL20HSA.

± 3.5 %; Fig. 11a and Fig. 12a) and NAL20L (7683.6 ± 344.0 µg/47.5
Table 4
The bulk and tapped densities and Carr’s index of the spray dried powders. ± 4.1 %; Fig. 11a and Fig. 12a) had comparable, but higher FPDs and
FPFs. The FPF < 5 µm of NAL20L (47.5 ± 4.1 %) was significantly higher
Formulations Bulk density (g/ Tapped density (g/ Carr’s index
cm3) cm3) (%)
than that of NAL10L (39.2 ± 3.5 %). Furthermore, NAL10L exhibited
significantly higher FPD and FPF than NAL10HSA. Although NAL10L
SD NAL 0.2 0.0 0.3 ± 0.0 14.5 ± 4.3
±
had a significantly higher FPD than NAL20HSA, their FPFs were com­
NAL10L 0.3 ± 0.0 0.3 ± 0.0 4.1 ± 2.2
NAL10HSA 0.1 ± 0.0 0.2 ± 0.0 36.8 ± 2.6 parable. Both FPD and FPF of NAL20HSA were significantly lower than
NAL20L 0.3 ± 0.0 0.5 ± 0.0 41.0 ± 2.8 those of NAL20L. On the other hand, the FPDs and FPFs < 3 µm of SD
NAL20HSA 0.1 ± 0.0 0.2 ± 0.0 27.9 ± 4.6 NAL (3151.1 ± 146.9 µg/16.4 ± 0.3 %; Fig. 11b and Fig. 12b),
NAL10HSA (3146.1 ± 233.9 µg/16.5 ± 1.3 %; Fig. 11b and Fig. 12b),
and NAL20HSA (3365.4 ± 273.9 µg/18.9 ± 1.1 %; Fig. 11b and
EFs were observed in NAL20L (91.5 ± 2.1 %) and NAL20HSA (91.3 ±
Fig. 12b) were comparable. In contrast, the addition of 10 % and 20 %
0.5 %), they were significantly higher than those of NAL10L (87.1 ± 1.2
w/w leucine significantly increased the FPDs/FPFs < 3 µm to 4270.9 ±
%) and NAL10HSA (86.4 ± 1.9 %) (Fig. 10). For FPD and FPF <5 µm, the
454.1 µg/22.0 ± 2.7 % and 5128.4 ± 144.6 µg/31.7 ± 2.2 %, respec­
lowest ones were from SD NAL (4780.9 ± 412.8 µg/24.8 ± 1.4 %;
tively (Fig. 11b and Fig. 12b). Notably, the FPD < 3 µm of NAL10L was
Fig. 11a and Fig. 12a). Adding 10 % or 20 % w/w leucine or HSA
already significantly higher than those of the HSA-containing
significantly improved the FPDs and FPFs < 5 µm. Comparable FPDs and
formulations.
FPFs < 5 µm were observed from NAL10HSA (6258.7 ± 302.5 µg/32.8
Fig. 13 shows the aerodynamic particle size distribution of NAL20L
± 1.5 %; Fig. 11a and Fig. 12a) and NAL20HSA (6258.7 ± 302.5 µg/37.0
after one month of storage at 25 ◦ C/30 % RH and 25 ◦ C/60 % RH. The
± 1.8 %; Fig. 11a and Fig. 12a), while NAL10L (7636.6 ± 576.7 µg/39.2
drug recovery was close to 100 % and no additional peaks were observed

8
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

Fig. 9. Aerodynamic particle size distributions of spray dried powders containing nalbuphine hydrochloride (NAL), leucine, and human serum albumin (HSA). Data
presented as mean ± standard deviation (n = 3).

in the HPLC chromatograms, indicating the chemical stability. The

aerodynamic particle size distributions were comparable under both
conditions (Fig. 10), suggesting good physical stability. This was further
supported by the consistent FPFs < 5 µm/< 3 µm at 25 ◦ C/30 % RH (47.9
± 2.3 %/32.1 ± 1.3 %; Fig. 13) and 25 ◦ C/60 % RH (46.9 ± 0.8 %/34.9
± 0.7 %; Fig. 13).

4. Discussion

Inhaled opioids have previously been investigated in human PK

studies, demonstrating highly promising profiles. For example, the
bioavailability of inhaled fentanyl was about 100 % and its time to reach
peak concentration was comparable to that of intravenous injection
(Macleod et al., 2012). Both intravenous opioids and inhaled fentanyl
also provided similar analgesic efficacy (Thompson and Thompson,
2016). Another example is inhaled morphine, which exhibited a 59–100
% bioavailability and a comparable time to reach peak concentration as
intravenous infusion (Dershwitz et al., 2000; Ward et al., 1997). These
findings suggest that pulmonary delivery is an efficient and non-invasive
route for administering opioids. However, these studies employed either
a nebuliser (using a liquid formulation) (Thompson and Thompson,
2016), a soft mist inhaler (Ward et al., 1997), or a flow-independent
condensation aerosol inhaler (Staccato® with a drug-only formulation
as a thin film coated on a substrate for heating) (Macleod et al., 2012),
and none have investigated the use of dry powder inhalers. Therefore,
the current study is the first to develop inhalable opioid powders. While
multiple opioids are available, NAL was formulated in this study due to
its widespread clinical use and better safety profile compared to other
opioids.
In this study, NAL powders were produced via spray drying, which is
Fig. 10. Emitted fractions of spray dried powders containing nalbuphine hy­ a rapid, reproducible, and scalable one-step process (Marante et al.,
drochloride (NAL), leucine, and human serum albumin (HSA). Data presented 2020). The high feasibility of this method was demonstrated by the good
as mean ± standard deviation (n = 3; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; production yields at 80–85 %. The spray dried particles were physically
****p ≤ 0.0001). small (3–4 µm; Table 2) and exhibited a corrugated surface, as shown in

9
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

Fig. 11. (a) Fine particle dose < 5 µm and (b) fine particle dose < 3 µm of spray dried powders containing nalbuphine hydrochloride (NAL), leucine, and human
serum albumin (HSA). Data presented as mean ± standard deviation ((n = 3; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001).

Fig. 12. (a) Fine particle fraction < 5 µm and (b) fine particle fraction < 3 µm of spray dried powders containing nalbuphine hydrochloride (NAL), leucine, and
human serum albumin (HSA). Data presented as mean ± standard deviation ((n = 3; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001).

the SEM images in Fig. 1. Although SD NAL already displayed surface shrinks with water evaporation to form corrugated particles (Ke et al.,
corrugation, the addition of leucine and HSA further increased surface 2022). For HSA, it occupies the particle surface as its high molecular
roughness. Leucine is a hydrophobic amino acid with weak surfactant weight limits its diffusion during spray drying. When spray drying at a
properties (Chang and Chan, 2022). It is also more hydrophobic than high temperature (i.e. 140 ◦ C in the current study), rapid water evapo­
NAL, as indicated by its lower aqueous solubility (22 mg/mL at 25 ◦ C) ration induces formation of a crust on the particle which collapses and
compared to that of NAL (35.5 mg/mL) (Ordoubadi et al., 2023; Pfizer, causes surface wrinkling (Maa et al., 1997). The effects of HSA and
2025). These characteristics enable leucine to accumulate at the droplet leucine on the droplet and particle surfaces were confirmed by the
surface during spray drying, where it reaches supersaturation, pre­ droplet contact angle measurements and the O-PTIR chemical maps. The
cipitates, and forms a hydrophobic shell (Ke et al., 2022). This shell surface activity of leucine and HSA in the droplet was demonstrated by a

10
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

Fig. 13. Aerodynamic particle size distributions, fine particle fraction <5 µm, and fine particle fraction <3 µm of NAL20L after one month of storage at 25 ◦ C under
both 30 % and 60 % relative humidity. Data presented as mean ± standard deviation (n = 3).

significant decrease of contact angle compared to that of the NAL so­ reflected in the corresponding changings in melting points across for­
lution (Fig. 2), indicating their accumulation at the gas-liquid interface. mulations. Similar molecular interactions have also been reported in
The larger droplet contact angle observed for the NAL solution con­ other co-spray dried formulations, such as ciprofloxacin with tryptophan
taining leucine, compared to that containing HSA, suggested that (Chang et al., 2022) and budesonide with arginine (Lu et al., 2019). In
leucine had lower surface activity and was a weaker surfactant, poten­ this study, the intermolecular interactions between NAL and excipients
tially due to its limited amphiphilic nature. After spray drying, the were supported by peak shifts, peak disappearance, and the emergence
particle surfaces were enriched with leucine or HSA, as evidenced by of new peaks in the O-PTIR spectra of the formulations (Fig. 7). For
slight peak shifts and reduced intensities in multiple spectral peaks, as NAL10L and NAL20L, the spectral changes suggested the formation of
well as the dominance of pink to red colours in the ratio maps (Fig. 3). hydrogen bonds or ionic interactions between leucine and NAL. The
Traces of faint blue observed in the ratio maps indicated incomplete slight shift of the peak from 1559 to 1564 cm− 1 in NAL10L indicated a
surface coverage. Increasing the mass concentration of excipients from weak hydrogen bond involving the protonated tertiary amine of NAL
10 % to 20 % w/w enhanced the redness of the chemical maps and and the amino or carboxyl group of leucine. Increasing leucine content
reduced the presence of blue, suggesting that 20 % w/w excipient led to potentially stronger hydrogen bonds, as shown by the significant
resulted in larger surface coverage of the particles. shift of the peak from 1323 to 1305 cm− 1 and 1167 to 1191 cm− 1 in
All spray dried NAL powders were amorphous (Fig. 4) and contained NAL20L, corresponding to CH2 wagging and C-O stretching, respectively
<4 % residual solvent by mass (Fig. 5), indicating efficient drying during (Mestek, 2025). The C-O shift might originate from the formation of
production. Since the spray dried formulations were amorphous, their hydrogen bonds between the phenolic hydroxyl group of NAL and the
DSC thermograms in Fig. 6 differed from that of the crystalline raw NAL, amino group of leucine. The appearance of new peaks at 1591, 1407,
which only showed an endothermic melting peak at 293.8 ◦ C, followed and 1359 cm− 1, was assigned to the C=C aromatic shift, COO– sym­
by an exothermic decomposition peak at 308.3 ◦ C. These results are metric stretch, and C-N stretch, respectively (Mestek, 2025). These
consistent with previous observations (Hanko and Vlahova, 2011). In further supported the formation of ionic interactions. The disappearance
contrast, the amorphous spray dried powders first crystallised (as indi­ of the peak at 1385 cm− 1 suggested potential steric hindrance by
cated by exothermic events at 208.4–218.3 ◦ C; Table 3), then melted and leucine. Similarly, increasing HSA content in the formulations led to
decomposed (with endothermic melting events at 263.8–290.5 ◦ C, fol­ potentially stronger hydrogen bonding and conformational changes of
lowed by exothermic decomposition event at 301.7–308.5 ◦ C). The HSA. This was evidenced by the peak shift from 1641 to 1664 cm− 1,
addition of excipients influenced the crystallisation temperature and corresponding to the amide I band of HSA. The new peaks at 1063 and
subsequently the melting point of NAL. While SD NAL, NAL10L, and 1031 cm− 1 were attributed to C-O and C-N vibrations, suggesting po­
NAL10HSA crystallised at similar temperatures (208.4–211.8 ◦ C), tential interactions between NAL and HSA side chains (Mestek, 2025).
crystallisation occurred earlier for NAL20L (at 201.7 ◦ C) and later for The decreased intensity of peaks at 1509, 1375, and 1115 cm− 1 indi­
NAL20HSA (at 218.3 ◦ C). These shifts in crystallisation peaks suggest cated the binding of NAL within the HSA binding pockets, resulting in
potential molecular interactions between NAL and either leucine or the shielding of NAL functional groups.
HSA. Such interactions may have affected the packing of NAL crystals, as Although amorphous powders can enhance drug dissolution and

11
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

bioavailability (Bhujbal et al., 2021), they are also prone to recrystal­ (and therefore expected to disperse well), while the addition of excipi­
lisation, particularly in hygroscopic compounds. SD NAL is moderately ents increased the powder cohesiveness (and was thus believed to have
hygroscopic as it absorbed about 14 % w/w moisture at 25 ◦ C/80 % RH lower FPFs). However, the trends in bulk and tapped densities and Carr’s
before undergoing recrystallisation, as shown in Fig. 8 (Newman et al., index did not correspond to those of the FPFs. This discrepancy arises
2008). While the addition of leucine in NAL10L and NAL20L did not because Carr’s index assesses the flowability of powders with discrete
delay recrystallisation, it reduced moisture uptake to 10–12 % at 80 % particles (Seville et al., 2007). In contrast, spray dried powders typically
RH. In NAL10L, a small mass drop at 60 % RH indicated the recrystal­ form aggregates with varying strengths, which can lead to misleading
lisation of amorphous leucine. This recrystallised leucine acted as a Carr’s index value and inaccurate flowability classification. Therefore,
physical barrier to moisture penetration, as crystalline leucine is non- bulk and tapped densities and Carr’s index cannot reliably predict
hygroscopic and has a low propensity for water uptake (Chang and aerosol performance. This limitation was also reported in another study,
Chan, 2022). Similar protective effects and leucine recrystallisation at which showed a lack of correlation between FPF and Carr’s index in co-
60 % RH have been observed in other spray dried powders, such as spray dried salbutamol sulfate and lactose formulations containing
salbutamol sulfate (Li et al., 2017) and disodium cromoglycate (Li et al., various amino acids (i.e. arginine, leucine, and phenylalanine) (Seville
2016) with 5–20 % w/w leucine. In contrast, NAL20L only showed one et al., 2007). Nonetheless, increasing the proportion of leucine or HSA
recrystallisation event at 80 % RH because leucine was crystalline in the from 10 % to 20 % w/w in the formulations enhances surface enrich­
formulation, consistent with the faint peak observed at 6◦ in the XRD ment during spray drying (as confirmed by the O-PTIR chemical maps in
results in Fig. 4. On the other hand, NAL10HSA absorbed more moisture Fig. 3), thus reducing particle cohesion more effectively and improving
than SD NAL and both leucine-containing formulations at 80 % RH, aerosol performance. Besides promoting powder dispersibility, the
owing to the high hygroscopicity of HSA. Interestingly, HSA delayed the addition of leucine also contributed to the chemical and physical sta­
recrystallisation of NAL20HSA until 90 % RH. This may be due to the bility of the powders during storage (Chang et al., 2019). This was
high molecular weight of HSA, which limits molecular mobility and thus supported by the stability test results in Fig. 13, which showed no sig­
hinders crystallisation. Consequently, more water (acting as a plasti­ nificant drug degradation or statistical differences in FPFs <5 and <3 µm
ciser) was required to enable recrystallisation. Another possible expla­ after storage at 25 ◦ C/30 % RH and 25 ◦ C/60 % RH for one month. The
nation is the extent of particle surface coverage by the excipients. As good stability of NAL20L highlights the potential of dry powder for­
depicted in the O-PTIR chemical overlaid maps in Fig. 3, HSA appeared mulations to address concerns related to NAL degradation in solution,
to cover a larger portion of the particle surface than leucine at equiva­ such as in nasal formulations like Apain®. Further long-term stability
lent mass concentrations. Hence, HSA may absorb more moisture before studies are required to confirm the shelf-life of NAL20L under various
enough accumulates to trigger recrystallisation of NAL in the formula­ storage conditions.
tion. Future studies are warranted to investigate the actual mechanism The feasibility of clinical use of the formulations can be estimated
of how HSA delays moisture-induced recrystallisation. Nonetheless, the using in vitro FPD, although this may not represent the actual inhaled
ability of HSA on delaying drug recrystallisation has also been reported dose in vivo (Newman and Chan, 2020). While NAL has some potential
previously (Tai et al., 2025). local therapeutic effects in the lungs, its primary indication is for anal­
Although all five spray dried powders were inhalable, SD NAL had gesia, which requires the drug particles to reach the small airways or
the highest drug retention in the capsule and inhaler, as shown in Fig. 9. alveoli for systemic absorption. Hence, the FPD <3 µm of NAL20L (5.1
It also had the most drug deposition in the throat, potentially increasing mg per capsule) was used for dose estimation. The recommended dose of
the risk of local oropharyngeal adverse effects in patients. These findings NAL injection for a 70 kg adult is 10 mg, which can be repeated every 3
indicated that the aerosol performance of SD NAL required improve­ to 6 h when required (Pfizer, 2025). A similar intranasal dose (10.5 mg)
ment, warranting the need of excipients. Two commonly used excipients has also been used in human PK and efficacy studies (Tymko et al., 2024;
are leucine and HSA, which have previously been applied as dispersion Tymko et al., 2023). While this intranasal requires three puffs of spray
enhancers (Chow et al., 2019; Li et al., 2017; Li et al., 2016; Li et al., (as per the PK study), patients would only need to inhale two capsules to
2025). Their effects were demonstrated not only by reducing capsule NAL20L to receive an equivalent dose (10.2 mg). Inhaling two capsules
and inhaler retention and throat deposition but also by increasing drug every 3–6 h is practical, as similar dosing regimens are already used with
deposition on the latter stages of the impactor. These improvements commercial inhalers, such as Ventolin® metered dose inhaler (salbuta­
were observed with the addition of 10 % w/w excipient and became mol sulfate; one to two puffs every 4 to 6 h when needed)
more pronounced at 20 % w/w. Compared to SD NAL, the other four (GlaxoSmithKline Pharmaceuticals, 2025) and Bricanyl® Turbuhaler®
formulations had more NAL deposited on Stages 1–3, resulting in larger (terbutaline; one to three inhalations every 4 to 6 h when required) (NPS
MMADs. The only formulation with a comparable MMAD as SD NAL was Medicinewise, 2025). Therefore, inhalable NAL powders may offer a
NAL20L as it had more particles deposited at the latter stages of the NGI. feasible alternative to injection or intranasal administration, potentially
The extensive deposition of NAL20L, ranging from Stage 1 to the micro- improving pain management and patient adherence. Further studies are
orifice collector, also explained its significantly broader GSD. warranted to investigate the PK profiles of the inhalable spray dried NAL
Among all five formulations, NAL20L achieved the highest EF powders.
(Fig. 10), FPDs (Fig. 11), and FPFs (Fig. 12) at both cutoff diameters (<5
and <3 µm), suggesting that it could potentially deliver the most NAL to 5. Conclusions
the lungs and systemic circulation. This also implied that leucine was a
more effective dispersion enhancer than HSA. In fact, the FPF <5 µm of NAL powders, with and without leucine and HSA, were successfully
NAL10L (39 %) was comparable to those of both NAL10HSA (33 %) and produced via spray drying. The spray dried formulations exhibited good
NAL20HSA (37 %), indicating that 10 % w/w leucine was almost as production yields and were inhalable. The addition of leucine and HSA
effective as 20 % w/w HSA. For FPF <3 µm, both HSA-containing for­ further increased surface roughness, which enhanced powder aerosoli­
mulations did not show any significant improvement compared to SD sation. Compared to HSA, leucine was more effective at enhancing the
NAL, indicating that HSA had limited ability to enhance the fraction of aerosol performance. These formulations can be used as a platform to
fine particles <3 µm. This may be explained by the less corrugated deliver NAL and other opioid analgesics by inhalation.
particle surfaces of NAL10HSA and NAL20HSA. Reduced particle
corrugation increases contact surface area and interparticle cohesion, Declaration of Generative AI and AI-assisted technologies in the
which in turn worsens aerosol performance (Chew and Chan, 2001). writing process
Interestingly, the powder densities and Carr’s index in Table 4 suggested
that NAL10L was the least cohesive among all the spray dried powders During the preparation of this work the author(s) used ChatGPT in

12
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

order to improve the readability and language of the manuscript. After Hanko, J., Vlahova, P., 2011. Crystalline and amorphous forms of nalbuphine
hydrochloride, US8536191B2.
using this tool/service, the author(s) reviewed and edited the content as
Ke, W.R., Chang, R.Y.K., Chan, H.K., 2022. Engineering the right formulation for
needed and take(s) full responsibility for the content of the published enhanced drug delivery. Adv. Drug Deliv. Rev. 191, 114561. [Link]
article. 10.1016/[Link].2022.114561.
Khanna, K., Sharma, N., Karwasra, R., Kumar, A., Nishad, D.K., Janakiraman, A.K., Ram
Mani, R., Rajagopal, M., Tayyab, S., Goel, B., 2025. Exploring nalbuphine loaded
CRediT authorship contribution statement chitosan nanoparticles for effective pain management through intranasal
administration: a comparative study. J. Drug Target. 33 (1), 99–110. [Link]
org/10.1080/1061186X.2024.2397800.
Waiting Tai: Writing – original draft, Methodology, Investigation,
Kuznetsov, I.E., Tymko, V.G., 2023. Nalbuphine nasal spray: proven medication and new
Data curation, Conceptualization. Hong-Jaan Wang: Writing – review capabilities. J. Basic Clin. Pharma. 14 (4), 290–293.
& editing, Conceptualization. Dipesh Khanal: Writing – review & Kuznetsov, I.E., Tymko, V.G., 2024. Nalbuphine nasal spray: analgesic efficacy and
safety. J. Pharm. Pharm. Res. 8 (2), 011. 10.21767/ipipr.8.02.011.
editing, Investigation. Pancy Tsz Hei Kwong: Writing – review &
Kuzumaki, N., Suzuki, A., Narita, M., Hosoya, T., Nagasawa, A., Imai, S., Yamamizu, K.,
editing, Investigation. Patricia Tang: Writing – review & editing, Morita, H., Nagase, H., Okada, Y., Okano, H.J., Yamashita, J.K., Okano, H.,
Conceptualization. Chih-Chin Shih: Writing – review & editing, Su­ Suzuki, T., Narita, M., 2012. Effect of kappa-opioid receptor agonist on the growth of
pervision, Methodology, Investigation, Conceptualization. Hak-Kim non-small cell lung cancer (NSCLC) cells. Br. J. Cancer 106 (6), 1148–1152. https://
[Link]/10.1038/bjc.2011.574.
Chan: Writing – review & editing, Supervision, Methodology, Li, L., Leung, S.S.Y., Gengenbach, T., Yu, J., Gao, G.F., Tang, P., Zhou, Q.T., Chan, H.K.,
Conceptualization. 2017. Investigation of L-leucine in reducing the moisture-induced deterioration of
spray-dried salbutamol sulfate power for inhalation. Int. J. Pharm. 530 (1–2), 30–39.
[Link]
Declaration of competing interest Li, L., Sun, S., Parumasivam, T., Denman, J.A., Gengenbach, T., Tang, P., Mao, S.,
Chan, H.K., 2016. L-Leucine as an excipient against moisture on in vitro
aerosolization performances of highly hygroscopic spray-dried powders. Eur. J.
The authors declare that they have no known competing financial Pharm. Biopharm. 102, 132–141. [Link]
interests or personal relationships that could have appeared to influence Li, M., Cao, Y., Chan, H.K., 2025. Optimizing performance of inhalable bacteriophage
powders using human serum albumin (HSA). Int. J. Pharm. 125709. [Link]
the work reported in this paper.
10.1016/[Link].2025.125709.
Liang, R.J., Shih, Y.N., Chen, Y.L., Liu, W.Y., Yang, W.L., Lee, S.Y., Wang, H.J., 2020.
Acknowledgements A dual system platform for drug metabolism: Nalbuphine as a model compound. Eur.
J. Pharm. Sci. 141, 105093. [Link]
Lu, W., Rades, T., Rantanen, J., Yang, M., 2019. Inhalable co-amorphous budesonide-
The authors would like to thank Dr. Philip Chi Lip Kwok for his arginine dry powders prepared by spray drying. Int. J. Pharm. 565, 1–8. [Link]
valuable comments and insights to this study. org/10.1016/[Link].2019.04.036.
Maa, Y.F., Costantino, H.R., Nguyen, P.A., Hsu, C.C., 1997. The effect of operating and
formulation variables on the morphology of spray-dried protein particles. Pharm.
Data availability Dev. Technol. 2 (3), 213–223. [Link]
Macleod, D.B., Habib, A.S., Ikeda, K., Spyker, D.A., Cassella, J.V., Ho, K.Y., Gan, T.J.,
2012. Inhaled fentanyl aerosol in healthy volunteers: pharmacokinetics and
Data will be made available on request.
pharmacodynamics. Anesth. Analg. 115 (5), 1071–1077. [Link]
ANE.0b013e3182691898.
References Maher, T.M., Avram, C., Bortey, E., Hart, S.P., Hirani, N., Molyneux, P.L., Porter, J.C.,
Smith, J.A., Sciascia, T., 2023. Nalbuphine tablets for cough in patients with
idiopathic pulmonary fibrosis. NEJM Evid. 2 (8), EVIDoa2300083. [Link]
Alhajj, N., O’Reilly, N.J., Cathcart, H., 2021. Leucine as an excipient in spray dried
10.1056/EVIDoa2300083.
powder for inhalation. Drug Discov. Today 26 (10), 2384–2396. [Link]
Marante, T., Viegas, C., Duarte, I., Macedo, A.S., Fonte, P., 2020. An overview on spray-
10.1016/[Link].2021.04.009.
drying of protein-loaded polymeric nanoparticles for dry powder inhalation.
Bhujbal, S.V., Mitra, B., Jain, U., Gong, Y., Agrawal, A., Shyam Karki, S., Taylor, L.S.,
Pharmaceutics 12 (11), 1032. [Link]
Kumar, S., Zhou, Q.T., 2021. Pharmaceutical amorphous solid dispersion: a review
Mestek, O., 2025, Infrared Spectroscopy. [Link]
of manufacturing strategies. Acta Pharm. Sin. B 11 (8), 2505–2536. [Link]
6/0003~~[Link]?redirected (accessed 3rd July 2025).
10.1016/[Link].2021.05.014.
Murphy, G., Brayden, D.J., Cheung, D.L., Liew, A., Fitzgerald, M., Pandit, A., 2025.
Bosquillon, C., Lombry, C., Preat, V., Vanbever, R., 2001. Influence of formulation
Albumin-based delivery systems: recent advances, challenges, and opportunities.
excipients and physical characteristics of inhalation dry powders on their
J. Control. Release 380, 375–395. [Link]
aerosolization performance. J. Control. Release 70 (3), 329–339. [Link]
Narver, H.L., 2015. Nalbuphine, a non-controlled opioid analgesic, and its potential use
10.1016/s0168-3659(00)00362-x.
in research mice. Lab. Anim. (N.Y.) 44 (3), 106–110. [Link]
Chang, R.Y.K., Chan, H.K., 2022. Advancements in particle engineering for inhalation
laban.701.
delivery of small molecules and biotherapeutics. Pharm. Res. 39 (12), 3047–3061.
Newman, A.W., Reutzel-Edens, S.M., Zografi, G., 2008. Characterization of the
[Link]
“hygroscopic” properties of active pharmaceutical ingredients. J. Pharm. Sci. 97 (3),
Chang, R.Y.K., Li, M., Chow, M.Y.T., Ke, W.R., Tai, W., Chan, H.K., 2022. A dual action of
1047–1059. [Link]
D-amino acids on anti-biofilm activity and moisture-protection of inhalable
Newman, S.P., Chan, H.K., 2020. In vitro-in vivo correlations (IVIVCs) of deposition for
ciprofloxacin powders. Eur. J. Pharm. Biopharm. 173, 132–140. [Link]
drugs given by oral inhalation. Adv. Drug Deliv. Rev. 167, 135–147. [Link]
10.1016/[Link].2022.03.003.
10.1016/[Link].2020.06.023.
Chang, R.Y.K., Wallin, M., Kutter, E., Morales, S., Britton, W., Li, J., Chan, H.K., 2019.
Medicinewise, 2025, Bricanyl® Turbuhaler® - Consumer medicine information.
Storage stability of inhalable phage powders containing lactose at ambient
[Link]
conditions. Int. J. Pharm. 560, 11–18. [Link]
[Link] (accessed 7th May 2025).
ijpharm.2019.01.050.
Ordoubadi, M., Shepard, K.B., Wang, H., Wang, Z., Pluntze, A.M., Churchman, J.P.,
Chen, Y.L., Lai, Y.H., Huang, E.Y., Wang, H.J., Hung, H.Y., 2024. Nalbuphine-6-
Vehring, R., 2023. On the physical stability of leucine-containing spray-dried
glucuronide is a potent analgesic with superior safety profiles by altering binding
powders for respiratory drug delivery. Pharmaceutics 15 (2), 435. [Link]
affinity and selectivity for mu-/kappa-opioid receptors. Life Sci. 340, 122441.
10.3390/pharmaceutics15020435.
[Link]
Pfizer, 2025, Nalbuphine hydrochloride injection - Prescribing information. [Link]
Chew, N.Y., Chan, H.K., 2001. Use of solid corrugated particles to enhance powder
[Link]/[Link]?id=4505 (accessed 7th May 2025).
aerosol performance. Pharm. Res. 18 (11), 1570–1577. [Link]
Pietsch, U., Berger, Y., Schurter, D., Theiler, L., Wenzel, V., Meuli, L., Grünenfelder, A.,
1013082531394.
Albrecht, R., 2021. Nasal nalbuphine analgesia in prehospital trauma managed by
Chow, M.Y.T., Qiu, Y., Liao, Q., Kwok, P.C.L., Chow, S.F., Chan, H.K., Lam, J.K.W., 2019.
first-responder personnel on ski slopes in Switzerland: an observational cohort study.
High siRNA loading powder for inhalation prepared by co-spray drying with human
SJTREM 29 (36). [Link]
serum albumin. Int. J. Pharm. 572, 118818. [Link]
Pokorný, V., Červinka, C., Štejfa, V., Havlín, J., Růžička, K., Fulem, M., 2020. Heat
ijpharm.2019.118818.
capacities of L‑alanine, L‑valine, L‑isoleucine, and L‑leucine: experimental and
Dershwitz, M., Walsh, J.L., Morishige, R.J., Connors, P.M., Rubsamen, R.M., Shafer, S.L.,
computational study. J. Chem. Eng. Data 65 (4), 1833–1849. [Link]
Rosow, C.E., 2000. Pharmacokinetics and pharmacodynamics of inhaled versus
10.1021/[Link].9b01086.
intravenous morphine in healthy volunteers. Anesthesiology 93 (3), 619–628.
Schneider, C.A., Rasband, W.S., Eliceiri, K.W., 2012. NIH image to ImageJ: 25 years of
[Link]
image analysis. Nat. Methods 9 (7), 671–675. [Link]
GlaxoSmithKline Pharmaceuticals, 2025, Ventolin HFA (albuterol sulfate inhalation
Seville, P.C., Learoyd, T.P., Li, H.-Y., Williamson, I.J., Birchall, J.C., 2007. Amino acid-
aerosol), for oral inhalation use - Prescribing information. [Link]
modified spray-dried powders with enhanced aerosolisation properties for
ent/dam/global/hcpportal/en_US/Prescribing_Information/Ventolin_HFA/pdf/VEN
[Link] (accessed 7th May 2025).

13
W. Tai et al. International Journal of Pharmaceutics 682 (2025) 125952

pulmonary drug delivery. Powder Technol. 178 (1), 40–50. [Link] solution for injection in healthy volunteers. Br. J. Clin. Pharmacol. 89 (11),
10.1016/[Link].2007.03.046. 3311–3323. [Link]
Sun, X., Ma, S., Zang, Y.M., Lu, S.Y., Guo, H.T., Bi, H., Wang, Y.M., Ma, H., Ma, X.L., Wang, H.J., Hsiong, C.H., Ho, S.T., Lin, M.J., Shih, T.Y., Huang, P.W., Hu, O.Y.P., 2014.
Pei, J.M., 2006. Vasorelaxing effect of U50,488H in pulmonary artery and Commonly used excipients modulate UDP-Glucuronosyltransferase 2B7 activity to
underlying mechanism in rats. Life Sci. 78 (21), 2516–2522. [Link] improve nalbuphine oral bioavailability in humans. Pharm. Res. 31 (7), 1676–1688.
10.1016/[Link].2005.10.042. [Link]
Tai, W., Khanal, D., Arnold, J.C., Chan, H.K., Kwok, P.C.L., 2025. Solubilising and Wang, Y., Han, H.Y., Jia, Y.P., Zhao, L.Y., Li, Y., Zuo, B.J., Zhang, J., 2022. Protective
aerosolising cannabidiol using methyl beta-cyclodextrin and human serum albumin. effect of nalbuphine combined with dexmedetomidine on the lungs of children with
AAPS PharmSciTech 26 (5), 120. [Link] atelectasis and foreign body in the bronchus during the perioperative period. Front.
Tai, W., Kwok, P.C.L., 2022. Recent advances in drug delivery to the central nervous Physiol. 13, 799183. [Link]
system by inhalation. Expert. Opin. Drug Deliv. 19 (5), 539–558. [Link] Ward, M.E., Woodhouse, A., Mather, L.E., Farr, S.J., Okikawa, J.K., Lloyd, P., Schuster, J.
10.1080/17425247.2022.2074975. A., Rubsamen, R.M., 1997. Morphine pharmacokinetics after pulmonary
Thompson, J.P., Thompson, D.F., 2016. Nebulized fentanyl in acute pain: a systematic administration from a novel aerosol delivery system. Clin. Pharmacol. Ther. 62 (6),
review. Ann. Pharmacother. 50 (10), 882–891. [Link] 596–609. [Link]
1060028016659077. Yang, M.Y., Verschuer, J., Shi, Y., Song, Y., Katsifis, A., Eberl, S., Wong, K., Brannan, J.D.,
Todisco, T., Dottorini, M., Palumbo, R., Calvitti, M., Vernelli, C., Rossi, F., Iannacci, L., Cai, W., Finlay, W.H., Chan, H.K., 2016. The effect of device resistance and
Grandolini, S., Cosmi, E.V., 1990. Fate of human albumin microsphere and inhalation flow rate on the lung deposition of orally inhaled mannitol dry powder.
spherocyte radioaerosols in the human tracheobronchial tree. Lung 168 (Suppl), Int. J. Pharm. 513 (1–2), 294–301. [Link]
665–671. [Link] Zeng, S., Zhong, Y., Xiao, J., Ji, J., Xi, J., Wei, X., Liu, R., 2020. Kappa opioid receptor on
Tymko, V.G., Tsapko, G.V., Filipenko, V.A., Khvysiuk, O.M., Kovalova, K.V., pulmonary macrophages and immune function. Transl. Perioper. Pain Med. 7 (3),
Kuznetsov, I.E., 2024. A randomized, double-blind study to compare the efficacy and 225–233. [Link]
safety of nalbuphine nasal spray and injectable solution in patients after orthopaedic Zhang, Y., Jiang, Q., Li, T., 2017. Nalbuphine analgesic and anti-inflammatory effects on
interventions and traumatological procedures. Br. J. Clin. Pharmacol. 90 (7), patients undergoing thoracoscopic lobectomy during the perioperative period. Exp.
1728–1740. [Link] Ther. Med. 14 (4), 3117–3121. [Link]
Tymko, V.G., Tsapko, G.V., Kovalova, K.V., Mashchenko, S.V., Oliinykov, D.S.,
Kuznetsov, I.E., 2023. Comparative pharmacokinetics of nalbuphine nasal spray and

14